IAEA-TECDOC-888 Nuclear techniques to assess irrigation schedules for field crops Results of a co-ordinated research programme organized by the Soil Fertility, Irrigation and Crop production Section, Joint FAO/IAEA Division of Nuclear Techniques in Food and Agriculture INTERNATIONAL ATOMIC ENERGY AGENCY June 1996 The IAEA does not normally maintain stocks of reports in this series. However, microfiche copies of these reports can be obtained from INIS Clearinghouse International Atomic Energy Agency Wagramerstrasse 5 P.O. Box 100 A-1400 Vienna, Austria Orders should be accompanied by prepayment of Austrian Schillings 100,in the form of a cheque or in the form of IAEA microfiche service coupons which may be ordered separately from the INIS Clearinghouse. The originating Section of this publication in the IAEA was: Soil Fertility, Irrigation and Crop Production Section Joint FAO/IAEA Division International Atomic Energy Agency Wagramerstrasse 5 P.O. Box 100 A-1400 Vienna, Austria NUCLEAR TECHNIQUES TO ASSESS IRRIGATION SCHEDULES FOR FIELD CROPS IAEA, VIENNA, 1996 IAEA-TECDOC-888 ISSN 1011-4289 © IAEA, 1996 Printed by the IAEA in Austria June 1996 FOREWORD The increasing global demand for food and other agricultural products calls for urgent measures to increase water use efficiency which is, with plant nutrient availability, one of the two main limiting factors in crop production. Although only 20% of all cultivated land in the world is under irrigation, it provides 35-40% of all crop production. Because of higher yield under irrigated agriculture than rainfed dry-farming systems, investments for irrigation are usually a top priority. However, it has become a matter of serious concern in recent years that, despite their high costs, the performance of many irrigation projects has fallen short of expectations, as a result of inadequate water management both at farm and system level. Crop production increase has been well below the project targets. This TECDOC summarizes the results of a Co-ordinated Research Programme on The Use of Nuclear and Related Techniques in Assessment of Irrigation Schedules of Field Crops to Increase Effective Use of Water in Imgation Projects. The programme was carried out between 1990 and 1995 through the technical co-ordination of the Soil Fertility, Irrigation and Crop Production Section of the Joint FAO/IAEA Division of Nuclear Techniques in Food and Agriculture of the International Atomic Energy Agency. Fourteen Member States of the IAEA and FAO carried out a series of field experiments aimed at improving irrigation water use efficiency through a type of irrigation scheduling known as deficit irrigation. It had been previously shown that water stress developed at certain growth stages of field crops may not cause significant yield reduction. However very few studies had examined water stress response of field crops at different growth stages within the context of modifying irrigation scheduling to save water and thereby increase efficiency which would help to maintain soil productivity in arid-zone irrigation projects. Therefore the programme goals included the folio wings: to identify different growth stages of field crops when they are relatively less sensitive to water stress; to compare irrigation scheduling based on pre-planned timing of irrigation deficits with traditionally used irrigation schemes in relation to water use efficiency, and change of water and salt balance in plant root zone; to measure at least weekly field water status, using neutron moisture gauges, to estimate crop water use under each treatment. The co-ordinated research programme was organized by the IAEA according to the advice of a high level Review Committee of the Joint FAO/IAEA Division in 1988. Three research co-ordination meetings were held in 1992, 1993 and 1995. Venue of the first and the last meetings were Vienna, Austria. The second meeting was held in Fundulea, Romania. During the first meeting, research methods and field experimental layout were discussed and elaborated with the main scientific investigators. During the later meetings, experimental results were reviewed, additional field experiments, if needed, were suggested. The final report submitted by each participant in the programme was then edited with a view of publishing the present TECDOC. The FAO/IAEA Agriculture and Biotechnology Laboratory in Seibersdorf, Austria, assisted the programme. The Soil Fertility, Irrigation and Crop Production Section of the Joint FAO/IAEA Division wishes to express its appreciation to the co-ordinated research programme scientific investigators. Without their enthusiasm and commitment it would not have reached a successful completion. This publication was prepared by P. Moutonnet and C. Hera from the Soil Fertility, Irrigation and Crop Production Section of the Joint FAO/IAEA Division. It should also be acknowledged that each final report was thoroughly reviewed and edited by D.R. Nielsen, formerly with University of California Davis, USA, and at present retired. EDITORIAL NOTE In preparing this publication for press, staff of the IAEA have made up the pages from the original manuscripts as submitted by the authors. The views expressed do not necessarily reflect those of the governments of the nominating Member States or of the nominating organizations. Throughout the text names of Member States are retained as they were when the text was compiled. The use of particular designations of countries or territories does not imply any judgement by the publisher, the IAEA, as to the legal status of such countries or territories, of their authorities and institutions or of the delimitation of their boundaries. The mention of names of specific companies or products (whether or not indicated as registered) does not imply any intention to infringe proprietary rights, nor should it be construed as an endorsement or recommendation on the part of the IAEA. The authors are responsible for having obtained the necessary permission for the IAEA to reproduce, translate or use material from sources already protected by copyrights. CONTENTS SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Water stress effect on different growing stages for cotton and its influence on yield reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Prieto, C. Angueira Nuclear techniques to evaluate the water use of field crops irrigated in different stages of their cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . P.L. Libardi, S.O. Moraes, A.M. Saad, Q. de Jong van Lier, O. Vieira, R.L. Tuon The response of winter wheat to water stress and nitrogen fertilizer use efficiency ... Wang Fujun, Qi Mengwen, Wang Huaguo, Zhou Changjiu Water deficit imposed by partial irrigation at different plant growth stages of common bean (Phaseolus vulgaris L.) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . M. Calvache, K. Reichardt Field estimation of water and nitrate balance for an irrigated crop . . . . . . . . . . . . . . G. Vachaud, L. Kengni, B. Normand, J.L. Thony Crop yield response to deficit irrigation imposed at different plant growth stages ... T. Kovacs, G. Kovacs, J. Szito Sugarcane yield response to deficit irrigation at two growth stages . . . . . . . . . . . . . C.B.G. Pêne, G.K. Edi Yield response of groundnut grown under rainfed and irrigated condition . . . . . . . . . A. Ahmad Contribution to the improvement of sugarbeet deficit-irrigation . . . . . . . . . . . . . . . . M. Bazza, M. Tayaa Contribution to the improvement of irrigation management practices through water-deficit irrigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . M. Bazza Field response of potato subjected to water stress at different growth stages . . . . . . . M. Mohsin Iqbal, S. Mahmood Shah, W. Mohammad, H. Nawaz Some studies on pre-planned controlled soil moisture irrigation scheduling of field crops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 13 33 51 63 73 89 115 131 139 151 175 187 RA. Waheed, M.H. Naqvi, G.R. Tahir, S.H.M. Naqvi Water and nitrogen use efficiency under limited water supply for maize to increase land productivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 /. Craciun, M. Craciun Water balance and nitrate leaching in an irrigated maize crop in S W Spain . . . . . . . . F. Moreno, J.A. Cayuela, J.E. Femândez, E. Femàndez-Boy, J.M. Murillo, F. Cabrera Optimum irrigation schedules for cotton under deficit irrigation conditions . . . . . . . . M.S. Anaç, M. AU Ul, I.H. Tüzel, D. Anaç, B. Okur, H. Hakerlerler Yield response of cotton, maize, soybean, sugarbeet, sunflower and wheat to deficit irrigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Kirda, R. Kanber, K. Tülücü, H. Güngör Soil spatial variability considerations in salt emission and drainage reduction . . . . . . . J.W. Hopmans, S.O. Eching, W.W. Wallender 211 225 243 261 List of Participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271 SUMMARY The greatest potential for increasing food and other agricultural products is the more efficient use of naturally occurring precipitation in conjunction with improved soil fertility management. Until recently, regardless of the amounts and distribution of rainfall, irrigation practices were used almost exclusively to supplement the amount of soil water stored in the root zone to such an extent that the available soil water never allowed the crop to suffer from water stress throughout the growing season. As a result, even today farmers still tend to overirrigate to ensure a bountiful amount of water stored in the soil profile, and therefore many irrigation projects do not provide their full measure of anticipated success owing to the development of shallow water tables, salinization, over- or under-irrigation associated with inadequate or poorly designed irrigation scheduling, inefficient fertilizer use, etc. Through the use of deficit irrigation scheduling, crops are purposely irrigated less during plant growth stages that are relatively insensitive to water stress as regards the quality and quantity of the harvestable yield. With growth stages defined for each cultivar, field research is conducted for particular cultivars under local conditions of climate and soil fertility to establish optimal periods for allowing water stress conditions without significant yield reductions. Because the impact of a particular deficit irrigation and fertilization program for any one year usually has significant consequences for succeeding cropping seasons, field research is carried out to better understand the physical and biological processes accountable for crop response. Toward that better understanding, different crops were subjected to different local irrigated and environmental conditions in the countries represented by the scientific cooperators. Quantitative analyses of growing conditions and plant responses were aided by the use of neutron meters to monitor soil water retention and redistribution within and below the rooting zone of the crop in relation to field water balance strategies. 15N technology provided a unique tool to separately understand the behavior of fertilizer N and soil N for different combinations of deficit irrigation and fertilizer practices. Deficit irrigation schedules and other on-farm management practices for increasing water and fertilizer utilization efficiencies for particular crops are summarized below according to country. La Argentina, it was concluded that cotton, owing to its indeterminant growth habit, should always be grown under deficit irrigation practices. Irrigating cotton with unconstrained water applications causes excessive vegetative growth accompanied by a loss of fruits from the lower fruiting zones. Hence, deficit irrigation should be a required practice for reaching and sustaining maximum cotton yields. For both highest yields and water use efficiencies, available soil moisture should be maintained at a reasonable level (50%) during the vegetative and flowering growth stages, and no irrigation applied whenever the available soil water reaches 90-95%. The vegetative and bud formation growth period was the most sensitive period to water stress. If stress occurs during this period, the development of the cotton plant significantly reduces the number of positions for bud development. Water stress during the flowering period reduces yields primarily through fruit shedding. Gentle stress during the yield formation stage increases seed and lint yield by reducing vegetative growth without affecting production and retention of fruit. Under limited water availability, a single irrigation during the bud formation period would be a more successful irrigation strategy than the classical practice of irrigating at flowering. In Brazil, irrigating with 50% of the amount of water necessary to achieve maximum actual evapotranspiration and maximum production gave much higher bean and com grain yields than those of the national average. For both crops, field water use efficiency and crop water use efficiency were similar for each of the deficit irrigation treatments imposed, and were highest for the continuous stress treatment. Although a water stress during either the vegetative stage or flowering and yield formation stages of the bean crop significantly reduced the yield, while a stress during the ripening stage had no impact on yield compared with that obtained for a crop fully irrigated and without stress. Com grain yields obtained for each of the deficit irrigation treatments or for the fully irrigated treatment did not differ significantly, an outcome probably caused by the fact that rain occurred during the second and third growth development stages. In China, deficit irrigation treatments revealed that both grain yield and nitrogen use efficiency of winter wheat were sensitive to water stress during different growth development stages. The sensitivity of grain yields to water stress was greatest when the irrigation deficit was imposed during the booting to flowering stage. From this maximum, the sensitivity decreased for other stages of development in the following order: winter afterward to booting, flowering to milking, seeding to winter forward and milking to ripening. Values of field water use efficiency obtained when no water stress occurs during the entire growing season decreased by 5 - 20% whenever stress occurred during one or more growth stages. The extent to which nitrogen fertilizer is used by winter wheat depends upon the timing and rate of the fertilizer application as well as the impact of the water stress during particular growth stages. In the absence of water stress, fertilizer N recovery in the wheat is greatest during short periods soon after fertilizer application. During a water stress condition, the uptake of fertilizer was prolonged over a greater period. Smaller fertilizer use efficiencies were attained for high fertilizer rates owing to the leaching loss of fertilizer from the fine sandy soil. In Ecuador, the yield of deficit-irrigated bean crop was decreased severely if water stress occurred during the flowering stage of plant growth. Stressing the crop only during the flowering stage reduced bean yields to the same extent as when the crop was stressed throughout the entire growing season. Biological nitrogen fixation increased significantly when the bean crop was stressed during the flowering or yield formation stage. Compared with the traditional farmer practice, the only treatments having a greater crop water use efficiency were those receiving a full irrigation program or being water stressed only during the ripening stage. In France, a management strategy using intensive irrigation and f ertilization was sought to achieve a sustainable, high level of maize production and to simultaneously maintain and improve the quality of the environment including water quality in the agricultural region. Drainage and nitrogen losses from the soil were small or negligible during crop growth. However, during the early stage of the crop or during the inter crop period between October and April, drainage losses were about 90% of water inputs from naturally occurring precipitation. Nitrate leaching during these periods strongly depended upon the amount of nitrogen remaining in the soil profile at the end of harvest. A strategy of reducing the amounts of fertilizer applied drastically reduced annual nitrate losses with only a slight reduction of maize yield. In Hungary, using a special low pressure drip irrigation method, maize, soybean and potato were grown under water stress conditions planned during specific plant growth stages. Because of the high price of irrigation water and irrigation equipment, traditional irrigation practices of the Hungarian farmers compared favorably with the deficit irrigation treatments. In Côte dlvoire different degrees of deficit irrigation treatments were established on sugarcane during two different stages of plant growth development - tillering and boom (stem elongation) stages. Growth and yields of sugarcane declined when water stress was imposed during the stem elongation stage. A deficit irrigation practice to improve crop water use efficiency consists of omitting irrigation at tillering as soon as the crop is successfully established. Additionally, during stem elongation, a moderate water deficit of about 25% of a full irrigation regime appears to increase crop water use efficiency. In Malaysia, deficit irrigation treatments revealed that groundnut is especially sensitive to water stress during the flowering stage. At this stage, an inadequate water supply with a 40% deficit in evapotranspiration causes a 50% reduction of pod yield. Hence, to sustain groundnut production, an irrigation must be programmed so that sufficient water is particularly available during the flowering stage. Alternatively, with appropriate meteorological records, sowing time could be adjusted so that the flowering stage does not coincide with a normally dry period. In Morocco, improved management schemes based upon deficit irrigation concepts for sugarbeet and wheat production were developed. For sugarbeet, when water stress was applied early in the growing season, its effect could be almost entirely removed with adequate watering during the rest of the growing season. On the other hand, good watering early in the season, followed by a stress, gave a poor yield. The most crucial periods for adequate watering were those which correspond to late foliar development and root growth which coincided with the highest water requirement period. Stress throughout the growing season resulted in the highest yield reduction per unit of water deficit. An economic analysis taking into account the returns from the harvest and the expenses associated with water showed that the highest profit corresponds to the treatments irrigated at 40 cbar of soil water potential or 60 mm.m'1 of soil depletion. The performance of these treatments in terms of profit was higher or at least similar to that of the treatments which received larger amounts of water despite the low price of water. In the case of wheat, high water deficit occurred during the early stages and irrigation during these stages was the most beneficial for the crop. However, one water application during the tillering stage allowed the yield to be lower only to that of the treatment with three irrigations. Irrigation during the stage of grain filling caused the kernel weight to be as high as under three irrigations. The highest water-use efficiency value was obtained with one irrigation during the tillering stage. In Pakistan, yield responses of potato, wheat and cotton to deficit irrigation under different fertilization regimes indicated alternatives for higher water and fertilizer use efficiencies. The potato results obtained showed that stress imposed at the ripening stage caused least reduction in yield whereas that imposed at early development gave the greatest reduction. The traditional irrigation practice led to wasteful water application. The efficiency of water use was increased by applying deficit irrigation at appropriate growth stages with no adverse effect on yield. By using 1JN-labeled fertilizer it was shown that planting and earthing-up were equally important growth stages for applying N-fertilizer for its efficient utilization. Wheat was found to be most sensitive to moisture deficit during tillering and least sensitive during flowering. Additionally, wheat genotypes differed regarding their sensitivity to the timing of a water stress. For example, one genotype which was stressed only during booting produced a yield comparable to that of a crop fully irrigated throughout the growing season, while other genotypes could be stressed only at flowering or only at grain filling and still produce yields comparable to a fully irrigated crop. Substantial water savings were also achieved by deficit irrigation of two cotton genotypes. These genotypes differed in response to the timing of a water stress period. The order of contribution of irrigation to seed yield for one genotype was Vegetative Stage>Generative Stage>Maturity Stage, while that for the other genotype was Generative Stage>Vegetative Stage>Maturity Stage. In Romania, drought is the main environmental factor limiting crop productivity. The amount of rain and its distribution during the reproductive stage was found to be the main meteorological factor influencing yields of maize. High soil water content causes a reduction of maize grain yield in wet years owing to excessive water under irrigation conditions. Excessive water normally limits root development as a result of insufficient oxygen for respiration and lack of nitrate formation. The yield response of different hybrids to water deficit is of major importance in production management. For hybrids which are somewhat tolerant to water stress, overall grain production can be increased by extending the area under irrigation without fully meeting crop water requirements, provided water deficits do not exceed critical values. In Spain, an experiment was conducted in an intensively irrigated agricultural soil to determine the water flow and nitrate leaching below the root zone under an irrigated maize crop during and after the growing season (bare soil and rainy period). Two nitrogen fertilization rates were used - a high rate traditionally used by fanners in the region and another providing three times less nitrogen. Crop evapotranspiration was similar with both N-fertilization rates. Reducing N fertilization strongly decreased N03-N content in the soil profile at harvest, and consequently subsequent leaching was effectively reduced without decreasing yield. Strategies are being developed to effectively capture the mineralized N in low organic matter content soils together with that in irrigation waters in order to sustain maize production with minimal amounts of applied fertilizer. In Turisey, research on deficit irrigation showed that irrigation can be scheduled to increase the use efficiency of limited water supplies in the production of cotton, maize, soybean, sugarbeet, sunflower and wheat. Studies on deficit irrigation of cotton showed that yields depended not only upon the stage of plant development but upon the irrigation method and optimal fertilization management. A new cotton variety tested for its tolerance to stress revealed that it was most sensitive to deficit irrigation during its vegetative growth stage and least sensitive during boll formation. For limited water supply, maize irrigated only twice (once at early silking, and another during ripening) produced equal or better yields than when it was irrigated 3 times with a total application of 25 - 34% more water. Soybeans differ markedly between cultivars relative to their seasonal consumption of water. Good germination is assured only if the available soil water content is within a range of 50 - 80%. High yields are constrained by poor plant height and leaf extension caused by an inadequate supply of soil moisture. Because soybeans are most sensitive to water stress during flowering and pod filling stages, high yields can only be ensured if the crop is irrigated during these stages. By omitting an irrigation during the ripening stage of sugarbeets, a 20% savings of water is achieved without a decrease in yields. The tolerance of sunflower to water stress depends upon the growth stage during which soil water is available or in short supply. The growth stage most responsive to irrigation is flowering. Omitting an irrigation during either yield formation or vegetative stage caused smaller yield reductions than not irrigating at flowering. Guidelines for irrigating wheat were based upon climatic conditions prevailing in different regions of Turkey and the impact of a potential water stress on crop yield for different growth stages. Wheat responds well to irrigations at booting, heading and milking growth stages with number of irrigations being determined by the relative levels of rainfall. 10 In the United States of America, a practical approach for estimating deep percolation from a field with a subsurface drainage system in the presence of a regional ground water flow using an optimization scheme for salt load was developed. Crop ET is not required for the calculation of deep percolation. By estimating crop ET by an independent method, the amount of water used by the crop that comes from a shallow water table can be ascertained. In conclusion, the practice of deficit irrigation scheduling, especially in arid and semiarid regions where water shortage is the most important problem of agricultural production, can produce high crop yields with high crop water use efificiencies provided proper choices of irrigation timing and amounts are made. It also has the advantage of increasing fertilizer use efficiency as well as decreasing the potential for leaching losses of fertilizers and plant nutrients mineralized throughout the year. For the same amount of water savings, it is now recognized that for each kind of crop and local soil and climate conditions there are different irrigation management alternatives for creating water stress during a particular plant growth stage or for partitioning water stress to some degree throughout the season. It is further recognized that the economic returns and incentives for a farmer to practice deficit irrigation may greatly differ from those viewing a regional or national perspective. For the farmer considering the actual price of water, the irrigation system and other on-farm costs of production, maximum annual profit may be obtained by not practicing a deficit irrigation schedule but by using more water and never stressing the crop. On the other hand, on a regional basis, the water saved by each farmer using deficit irrigation can be used collectively to expand the total area cropped resulting in a net increase of production and profit for the country. It is anticipated that research utilizing the radiation and isotope techniques found successful in meeting the objectives of this project will be continued to examine the economics of deficit irrigation from both viewpoints - that of the individual farmer and that of a region. Next page(s) left blank 11 WATER STRESS EFFECT ON DIFFERENT GROWING STAGES FOR COTTON AND ITS INFLUENCE ON YIELD REDUCTION D. PRIETO, C. ANGUEIRA INTA-EEA Santiago del Estero, Santiago del Estero, Argentina Abstract An experiment was conducted during 5 growing seasons (1990 -1994) at the INTA-EEASE (Institute Nacional de Tecnolologia Agropecuaria-Estacion Experimental Agropecuaria Santiago del Estero) in Argentine with the objectives of i) determining cotton response to water stress on different growing stages, ii) define practical recommendations for water management and i i) to check the values of the crop coefficient Kc, the crop response factor ky and the water use efficiency WUE under local conditions. In a randomized block trial we provided eight treatments: (Tl) formation, (T3) (T6) without water stress, (T2) with water stress on vegetative period and early bud with water stress during flowering stage, (T4) with water stress during late ripening, (T7) peak flowering and (T8) with water stress in ripening, (T5) non irrigated, a traditional practice in the area with one irrigation during a traditional practice in the area with one irrigation during the early ripening period. Precipitation during the growing seasons ranged from 249 to 594 mm covering the normal variation in the region, while the range of measured actual evapotranspiration ETa was from to 649 to 914 mm. Results proved that the vegetative-bud formation period is the more sensitive growing stage (ky = 0.75) while the ripening period is the-lowest (ky = 0.20). Stress during flowering was intermediate (ky =0.48) and the response coefficient for the entire growing season was found to be the highest (ky = 1,02). The highest ETa lead to a yield reduction owing to an excessive vegetative growth and unfavorable conditions for fruit ripening in the bottom part of the plants. Owing to these conditions, the £7>yield curve was not a straight line and was found to be Y = - 8602 + 43,571£Ta - 0,0341(£ra)2. The recommended water management for a highest yield and WUE would be to maintain soil moisture content at a good level (50% 95%. in our case) during vegetative and flowering periods and to stop irrigation at 90 - If one irrigation is practiced (traditional method) it has to be done at bud formation, ky (for the entire period) and the WUE were greater than those reported by FAO 33 while Kc were similar. 1. INTRODUCTION Cotton (Gossypium hirsutum) is by far the most important irrigated crop in the Santiago del Estero, Pcie of Argentine. Cotton is also irrigated on the neighboring Pcies. of Cordoba, Catamarca and in small extend in La Rioja, Salta and Tucumân. Surface irrigation water is used in Sgo. del Estero, Tucumân, Salta and Cordoba for supplementing natural precipitation during the growing season. While in Catamarca and La Rioja, crop growth depends directly from irrigation but water is scarce and has to be pumped from the groundwater with high cost Both supplementary and total irrigation will benefit from deficit irrigation, but 13 cotton response to water applications has to be well known for a proper water managament under deficit irrigation strategies if economical profit wants to be maximized. Cotton response to water deficit is complex owing its indeterminant growing habit, its highly efficient deep rooting system, its low critical leaf water potential, its capacity of osmotic regulation of leaf turgor potential and a possibly different response according to previous moisture level ("conditioning") [1], The Er-production function provides a useful means of defining an efficient water managament. Direct use of evapotranspiration or other available indicator of plant water status described from it as the independent variable appear to have the greatest rigor and potential for transferability. However, applied water AW has a high practical usefulness because it is the controlled variable and the required variable for cost consideration [2]. A proper waterproduction function can be reached if it is based on data that utilize proper irrigation scheduling to give the least possible yield reduction. Sammis cited by Grimes and El-Zik [3] reported a linear relationship between cotton ET and lint yield. However, based on their own studies the former authors suggested a slight curvature to the function. The AW-lint function is always curvilinear as it progressively departs from the ET function as AW increases (larger deep drainage and water remaining in the soil profile at the end of the season) [3,9,10]. Water management on cotton haS been well described by Grimes and El-Zik [3], Doorenbos and Kassan [4] also summarized most of the literature on cotton response to water into their linear water response model. The permanent effects of a water stress in the vegetative period has been enphasized by Grimes and Yamada [5] who recommended a threshold value of the Mid Day Leaf Water Potential (MDLWP) of - 1.6 MPa for programming irrigation during this crop stage. Grimes et al. [6] pointed out the importance of the opportunity of the first irrigation. They found that both an early and late first irrigation depresed final yield. Thomas et al [1] however, found that plants that suffer a gentle water stress in the vegetative period presented a higher tolerance to water deficit in later growing stages ("conditioning"). They found that stomata closure started at MDLWP of - 1.8 MPa in well-watered plots and it was delayed until MDLWP of - 2.8 MPa in plants that suffer water stress in previous stages. The response to water stress during cotton flowering period is conditioned by its indeterminant growing habit and climate conditions late in the season. Grimes et al. [7] found that water stress has more serious effect on the mid and late flowering period, because the plant has not time to compensate the shedding of young fruits provoked by the restricted water conditions. However, it is during this stage that the crop shows its high capacity for an efficient water use, so Grimes and Yamada [5] suggested a MDLWP threshold value of - 1.9 MPa for irrigation scheduling during this period. The same value was found by Guinn and Mauney [8] to affect fruit retention. Meron et al. cited by Grimes and El-Zik [3] found that under small and frequent water applications (drip irrigation) the MDLWP should not decrease from -1.6 to -1.7 MPa during this stage. Literature on cotton water response in the late season (mostly fiber growth and development) is more scarce. From their study Grimes and Yamada [5] concluded 14 that fiber length and micronaire were not affected until a MDLWP of - 2.8 MPa, but were seriously depressed below this value. The objectives of the present studied were i) to determine cotton response to water stress on different growing stages in order to define practical recommendations for water management based on deficit irrigation and ii) to check under local conditions values of cultural coefficient KC, yield response factor ky, field irrigation efficiency £/-and crop water use efficiency Ec. 2. MATERIALS AND METHODS Soils of the treatments plots were classified as Aridic Haplustoll, moderately deep with no drainage problems. They have a mollic epipedon, 25 cm depth; and a cambic horizon 70 cm depth. At this depth, starts an calcium carbonate accumulation. Texture is silty loam throughout the profile. Soils are in a plain position with slope less than 1%. Field capacity and wilting point are 26 and 7% by volume, respectively. Unsaturated hydraulic conductivity was determined insitu by the instantaneous profile method on covered 2 x 2 m plots. Cotton varieties studied were Guazuncho-H INTA (from 1990-1991 to 1994-1995 growing season) and Quebracho-INTA (1990-1991 to 1992-1993). A randomized block design with four replications was used. The plot size was 5 rows, l m apart and 12 m long. Ten central m of the two central rows were harvested. The plants samples were collected in the 4th row, where also the neutron probe access tubes were located. Plots were seeded on 27 Nov 1990,12 Nov 1991, 17 Oct 1992, 18 Nov 1993 and 1 Nov 1994. Each plot was thinned to a constant number of 10 plants-nr1 by row, which is equivalent to 100,000 pl-ha4. Plots were cultivated to control weeds. Check irrigation was practiced. Plots were completely closed by earth edges and 3 m apart in order to avoid water run-off or seepage from one plot to another. Irrigation water was measured with a 3" Parshall flume. Soil moisture content 0 (cm3-cnr3) at 5 soil depths: 15,30, 80,125 and 175 cm were determined weekly with the neutron gauge Troxler 3333 during the first 4 years and hi 13 soil depths (15, 30,45, 60, 75, 80, 90,105, 120, 125,1 40, 160 and 175 cm) in 1994-1995. A unique calibration curve [0= 0.8 -f 0.096(- S.9+54L5CR - 43.9(CR)2 + 53.2(C7?)3] was found to be applied for the very homogeneous soil. Irrigation depth (mm) was calculated for each treatment to recover soil moisture to field capacity. All of the treatment plots received a pre-seeding irrigation which stored about 240 mm in a 1.5 m soil depth. Owing to its indeterminant growing habit, a definition of the cotton growing period is difficult. In this study 4 stages were separated (Table I) for defining water treatment oriented towards subjecting the crop to different watering regimes (Table ÏÏ). A threshold value of 60% of the available water was also used on treatments T2 to T6 in the periods without stress periods. Meteorological data [precipitation (mm), Class A pan evaporation (mm), sunshine, minimum and maximum temperatures (°C), wind velocity (km-lr1)] were collected in a traditional weather station at the Experiment Station near the trial site. Plant 15 TABLE I. MAIN GROWING STAGES FOR COTTON IN SANTIAGO DEL ESTERO Stage 1 Vegetative 0-65 Stage 2 Flowering + bud formation d StageS Early ripening Stage 4 Late ripening 95-110 d 110-140 d 65-95 d TABLE H. DESCRIPTION OF THE IRRIGATION TREATMENTS Treatment Tl T5 T7 T8 T2 T3 T4 T6 Growth stage 1 2 3 4 1 0 0 0 1 0 0 0 0 1 1 0 1 1 1 1 1 1 0 0 1 0 0 1 1 1 1 0 0 1 1 0 Description Controls Moisture content threshold 60% of AW at 0-100 cm soil depth Traditional practice 0 Not irrigated (only pre-seeding irrigation) Traditional practice 1 Traditional practice 2 One stress period Stress during stage 1 Stress during stage 2 Stress during stage 3 Stress during stage 4 height, number of nodes, green and dry matter of leaves, stem and reproductive parts were determined in eight plants selected at random from each treatment. Sampling was done during a fortnight from emergence to harvest Fruiting positions of sympodial branches were monitored from the beginning of stage 2 on eight plants from each treatment. Boll weight, lint %, seed index and lint quality were determined in 25 bolls of each treatment. Actual crop water use ETa was determined by the water balance method for time periods no shorter than 7 d. Deep drainage was measured directely in the 1994-1995 season by placing mercury tesniometers at the bottom of the monitored rooting zone. In the previous seasons it was estimated from the soil moisture readings and the pF curve. Only periods without rain and irrigation were considered in the calculations to avoid assumptions about efficiency of precipitation and irrigation because not logical values were found in periods with irrigation and rains, although irrigation depth was adjusted based on soil 16 moisture determination before and after water application and precipitation could be consider 100% effective. Crop coefficient KC [the relationship between reference evapotranspiration ET0 and the maximum crop evapotranspiration ETm determined from the water consumption in the no stressed treatment (Tl)] was analyzed for different time periods in order to determine the crop coefficient for each growing stage. Pan evaporation (Class A) and Kpan - 0.7 were used for calculating ET0 owing to its practical applicability in farmer conditions. 3. RESULTS AND DISCUSSION 3.1. Characteristics of the growing season From the irrigation point of view, the study was done under supplementary conditions because cotton grows during the rainy season of the area (October-March). For that reason achieving treatments based on water stress in different growing periods was highly dependent upon the rainfall distribution during the particular season. Table ÏÏI presents the hydroclimatic conditions of the studied season, while Table IV summarizes the statistical results. TABLE ffl. SEASON RAINFALL AND REFERENCE EVAPOTRANSPIRATION Climatic factor (mm) Rainfall ET0 1994-1995 1993-1994 Season 1992-1993 414 690 249 675 516 692 1991-1991 1990-1991 339 647 594 709 ET0 = Pan evaporation x 0.7 (Kpan). TABLE IV. SIGNIFICATIVE DIFFERENCES BETWEEN TREATMENTS Treatment Tl T2 T3 T4 T5 T6 T7 T8 1994-1995 a a a a a a a a 1993-1994 a ab ab a b a ab ab Season 1992-1993 1991-1991 a a a a a a a a a a a a a a a a 1990-1991 a be ab a c a be c 17 Only in two of the five seasons were yield differences between treatments statistically significant. Reasons for these findings will be discussed in Section 2.2. Judging from the rainfall distribution during each vegetative stage presented in Fig. 2, it is apprent that cotton yield response is closely related with the water regime during the first two growing stages. Fig. 2 shows the distribution of ET0 during the main growing stages. The vegetativebud formation period (0 - 65 d) has the higher values of ET0 owing to its duration and the fact that it happens during the months having high atmospheric demand (December - January). It is also noticable from the figure that the ET0 has a low variation during the flowering and early yield formation stages and its high variation between years in the first growing stage. 1994-1995 ^300 £, 1992-1993 n 1991-1992 es a 1990-1991 e •M M • • M i T MM1 MM MM •^ • M l 0-65 MMM 3 65-95 KXXXXXXXXXl V///////A IXXXXXXXXVI => — M^ • MM MM* • MM MM MM MM k\\\\V| » Hi O I • i• - M^ MMH MMM MMM M^ M^ y/////////. O 200 H ^ PRECIPIT^ 1993-1994 ra ^^ M^ PMM MMM MMM • M M ^•M MBH H 3 ^ ^ 95 -110 r É MMM MM MMM MM MMM MM MMM • M B • ^M l MMM MMM1 RnH ! 110 -140 DAYS FROM SOWING Fig. 1. Distribution of the rainfall during the growing seasons. ,——, r~t 1994-1995 <u It, cj -y l 100 « fcs3 n kXXXXXXV SP .5 200 y ^ E3 jjj ~ -^ ~— jj ^s ^ N / ro ^ X . o /\ 1993-1994 1992-1993 a s 1991-1992 1990-1991 a = - MM jjjH "• •J •• MM N MMM • J • ^ 0-65 V//////A ^300 _ o 1 - 65-95 ''/.. %'^ 95-110 1 >;1 110-140 DAYS FROM SOWING Fig. 2. Distribution ofET0 in the growing season. 18 ^ _ ^k 3.2. Crop response to water stress 3.2.1. Water stress during the vegetative period T2 (0111) The main results for T2 are shown in Table V where it can be seen that in those years with significant (5%) yield differences, the relative evapotranspiration deficit was above 20%. With the unique exception of the 1991 -1992 season, the yield response factor ky was close to 0.75. This value is substantially higher than 0.2 for the vegetative stage and 0.5 for the flowering stage reported by Doorenbos and Kassan [4]. TABLE V. BULK RESULTS FOR T2 DURING THE 5 GROWING SEASONS Season 1994-1995 1993-1994 1992-1993 1991-1992 1990-1991 ETa Relative Yield Relative yield Yield response WUE (mm) ET deficit (kg-ha-1) decrease factor £y (kg-nr3) 587 503 680 582 648 0.11 0.24 -0.06 0.16 0.29 4718 3891 5400 5500 3719 0.07 0.17 -0.05 -0.01 0.23 0.69 0.73 0.80 -0.06 0.80 0.80 0.77 0.79 0.95 0.57 Greater yield reductions occurred in those years (1990-1991 and 1993-1994) when the mean soil moisture content of the 2-m soil profile (Fig. 3) decreased below 18%. Owing to the pre-seeding irrigation required in the region, a water deficit was reached during the latter part of the period after the first bud. This early stage of cotton is characterized by a fast growth of its root system which provides the possibility of using water from greater soil depths. Hence, the water extraction pattern of the root zone is a sensitive indicator of the differences between each of the 5 seasons. From Fig. 3 which presents root water extraction patterns it is clear that in those seasons of higher atmospheric demand (1990-1991,1993-1994 and 1994-1995) the crop used water from deeper soil depths. Nevertheless, in spite of this favorable response for an efficient use of water, stress conditions were reached which was evident through most of the plant indicators such as plant height (Fig. 4). The limited water supply during the vegetative-bud formation period affected mainly the number of bolls on each plant. In spite of the indeterminant growth habit of the cotton plant, the water stress during the growing season of 1990-1991 and 1993-1994 caused an early cut-off of crop growth reducing the development of fruit on the hightest fruiting positions of the plant (See Fig. 5). Boll shedding, a physiologic process highly sensitive to water stress, was not evident With the other yield component (weight of each boll) not being affected under limited water availability during this phenologic period, the cotton plant reduces its production to a limited number of fruits per plant and concentrates its assimilate to each boll. 19 3.2.2. Water stress during the flowering period T3 (1011) There was greater variation in the relative evapotranpiration deficit between seasons when water stress occurred during the flowering period than in the case of T2 (Table VI). Although the high variation was embraced in the ky factor, its mean value of 0.48 revealed a lower sensitivity of the crop to water stress during this growing stage of peak flowering. The low values of the relative yield decrease and their uniformity are remarkable. l 1990-91 0.25 i 1991-92 0.25 0.15 gu Vegetative & Boll formation 0.05 Vegetative & Boll formation 0.( 60 120 60 l H fc W 0.25 H £ 1992-93 1993-94 0.25 Tl O 0.15 U 0.15 Vegetative & Boll formation • Vegetative • & Boll formation W H < 0.05 120 l 0 60 0.05 120 ____I_____i 60 120 1994-95 0.25 Tl 0.15 ^—— Vegetative ——j & Boll formation 0.05 l_____I 0 60 120 DAYS FROM SOWING Fig. 3 Soil water profile (2-m rooting depth) ofTreatmentsTl and T2 for différent growing seasons. 20 (cm3-cm-3) SOIL WATER CONTENT 0 0.1 0 0.2 1 70 63 0.3 u 56 42 1 V^v^// \S 40 80 - 1990-91 T2 80 1Y/ \V NX \ - 1991-92 T2 t 0 0.1 s ° 1 & 40 — H 80 Ck W 120 - 1992-93 T2 \\ \ i vO Y-i \ -" 160 inn 0.2 0.3 u ' 66 59 46 \l W A O 1 1 200 0.1 0 0 1 120 - 1994-95 - \V/ Al ~ M iM i\ 1 0.2 ! 03 42 ' 69 \\ T2 \ » V\ \ v 160 200 V\ M T2 yf"""~ 40 80 03 42 ' <\\\\ 160 \ 0.2 160 47 67 53 80 - 1993-94 120 ! 0.1 40 / 1 1 0 03 69 48 62 1 44 120 200 §200 0.2 1 \K 160 à Î6° 0.1 40 120 ^ 0 - i // wafer profiles of the main rooting zone for different sampling dates during the vegetative-bud formation stage. The soil moisture profiles of the 2-m rooting zone during the flowering period did not fully explain the relative evapotranspiration deficit However, after looking at the soil moisture extraction pattern in Fig. 7, it could be concluded that the crop used water deep in the soil profile more efficiently in those seasons with higher water availability in the previous stage. The evapotranspiration deficit during the flowering stage reduced the development of fruit position and increased fruit shedding. The latter was by far the most important effect inasmuch as reduction of the number of bolls per plant affected mainly the high zones that normally had a limited contribution to the yield [1 1] owing to the climatic conditions of the region during the late season. The percentage of fruit shedding from the most important zones 21 1990-91 120 80 120 — Vegetative & Boll formation T2 — Vegetative 80 & Boll formation 40 40 1991-92 l ° 0 0 120 60 60 120 —T~ 1992-93 _ K120 2 3 80 1993-94 - 120 — Vegetative & Boll formation — Vegetative & Boll formation 80 T2 H 40 40 Z <: J o 0 120 60 120 1994-95 - 120 — Vegetative & Boll formation 80 T2 40 0 60 0 0 60 120 DAYS FROM SOWING Fig. 5. Evolution of plant height during the different growing seasons. TABLE VI. BULK RESULTS FOR T3 DURING THE 5 GROWING SEASONS ETa Season 1994-1995 1993-1994 1992-1993 1991-1992 1990-1991 22 Relative Yield Relative yield Yield response WUE (mm) ET deficit (kg-ha-i) decrease (kg-m-3) factor ky 424 449 587 613 832 0.35 0.32 0.09 0.16 0.09 4033 4466 5925 5175 4508 0.21 0.05 -0.15 0.05 0.07 0.59 0.16 -1.76 0.46 0.78 0.95 0.99 1.01 0.84 0.54 (cirrW3) SOIL WATER CONTENT C3 p O C3 Ç3 t— Ui tO V/> C3 fi *• »*• **** O 1 1 1 o 1 i 1 1 II 1 à D p o 3 i- tO • i » ! i i i VO .& *< ^ v> — H iä Oco § £/-s o a. <3 «^ § i C& U . . 3 . / ^ 5 E M ^ p v/ -JT VO 1 /i LA 1 1 ^/ tT > -u» 1/ P, 1 1 i 1 1— « i p Î-» C3 ^ < < O to i i p 1 O tO C3 C3 l/> LA 1 1 _ VO o CD ^ >-*< S«. t-* W /Ö -j ' _ vb O ? ^ h™"^ KA ! o different growing - i—> i. » to o -r vo l 1 3 ^3' OQ Er ^9 P / i ~ i w^ l - • - • - - - - - - • - •- - • " • *• ————————^— — — 4i §gg C/l b S ^ 01 .....„......,.............,..,........,......> o« }> ; < to - to o l <- S ^N JO OJ 0 1 VO /H a C> ~ [^ l S- ö fi 1 -^ ~ o i VO l c/> Oo H i—> vo l s g /V — ST ^ 1 ! 8 l 0 to 1 1 ^^4 &• (J\ U Sr, O s=> h^j ^[> •£ a sS. t> ^ v*<* vS H c" X^^> °* //^H vx 1 1 o M i i p 1—A y l - / \ ^^ ^^ _ i) r " ^ s EL A H o 1 "} 1 </ ' 1 / - v-o w-• • //l - 0 a' k*^ i \\ 3 NUMBER OF BOLLS 1 A I <3 ö o p l E o - (1 to 4) was increased by the water stress during this period as can be seen from Fig. 8 for the 1991-1992 season. 3.2.3. Water stress from the early ripening period T4 (1100) With the bulk results from this treatment presented in Table VII, it can be seen that the relative evapotranspiration deficits reached in this treatment were smaller than those in the previous treatments. The main reason for this behavior is that the water requirement of the crop decreases during this stage and the deep crop rooting system has the capacity for extracting soil water, as shown in Fig. 9 for the period of 104 to 111 d from sowing in the 1994-1995 season. (cnrW3) SOIL WATER CONTENT 0 o.i 91l81 0.2 03 40 40 80 80 120 160 0.3 90 1991-92 T3 160 200 _L 0.1 0.2 83 77l 120 1990-91 T3 200 0.1 0 0 7 7 7 0 0.2 03 0 0 40 40 H 80 80 0 0.1 0.2 0.3 71 O, W 120 160 120 1992-93 T3 1993-94 160 L § 200 0 200 0.1 '69 T3 0.2 0.3 977 8 I g e 8 3 40 80 - 120 160 1994-95 T3 200 Fig. 8. Soil water extraction patterns of Treatment T3for the different growing seasons. A characteristic output from this treatment is the negative values of both the relative yield decrease and the ky in 4 of the 5 seasons. They reveal a £Ta-yield curve relationship that has 24 useful meaning for the irrigation water management of cotton inasmuch as they strongly suggest that it is not advisable to irrigate for ETm of this crop. TABLE Vn. BULK RESULTS FOR T4 DURING THE 5 GROWING SEASONS ETa Relative Yield Relative yield Yield response WUE (mm) ET deficit (kg-ha4) decrease factor^ (kg-nr3) Season 564 605 626 753 829 1994-1995 1993-1994 1992-1993 1991-1992 1990-1991 0.14 0.08 0.02 -0.09 0.09 -0.04 -0.07 -0.06 0.00 -0.14 5276 5039 5475 5463 5540 -0.27 -0.89 -2.63 0.02 -1.53 0.93 0.83 0.87 0.73 0.54 1UU ^^ _ 1991-1992 0T1 ÖT3 / o ' Z »•H g 50 — / 0 £C t« H P tf fc _ r — 0 7 / / { Jl [0 n\ 1 2 • ni 3 ; ^ / f 4 f~ ' / / ^ ^ f 5 V\ /~ / ^ / ' / j i_ 6 t ; ' - ; // * / / / / s 7 8 : : / / / : / / t / ; : • ' t f \ '• t t ; — — : : : , : 9 FRUCTIFICATION ZONE Fig. P. Fruit shedding within each fructification zone for Treatments Tl and T3 during the 1991-1992 growing season. The slightly better performance of this treatment is caused by a balanced ration between vegetative (branches) and reproductive growth (boh1 and lint). The observed yield reduction at the highest ETa (Tl) is caused by the indeterminant growth habit of cotton and the agroecological conditions of Santiago del Estero. Under continuously available high moisture conditions (Tl), the required cut-off of crop growth did not occur and the plants produced branches and flowers continuosly at their top. This late growth has a negative effect in the final yield because it produces fruit lost from the lowest part of the plants owing to extreme dark and 25 wet conditions. At the same time, that loss is not compensated by the production of new fruits at the top because temperature and light conditions of the local autumn are not enough for normal growth. 3.2.4. Traditional treatments T5 (0000), T7 (0010) and T8 (0001). The control treatments based on traditional practices of farmers can be divided into two groups, that without irrigation during the growing season [T5 (0000) only with pre-seeding irrigation] and those with water application in critical periods [T7 (0010) and T8 (0001)]. Table VHI summarizes results from these treatments. Each of their evapotranspiration deficits was greater than those in the previous cases. The treatment based on one irrrigation at flowering (T7) was the best (3 in 5 cases reached a yield as high as the best watered treatment). T5 had the worse performance while T8 was intermediate. TABLE Vffl. BULK RESULTS FROM TRADITIONAL TREATMENTS ETa Relative Yield Relative yield Yield response WUE 1 Season T5 1994-1995 1993-1994 1992-1993 1991-1992 1990-1991 400 430 599 495 570 0.39 0.35 0.07 0.28 0.38 3596 2685 ' 5275 5475 2880 0.29 0.43 -0.02 0.00 0.41 0.75 1.23 -0.36 -0.02 1.08 0.90 0.62 0.88 1.11 0.51 T7 1994-1995 1993-1994 1992-1993 1991-1992 1990-1991 515 417 631 476 736 0.22 0.37 0.02 0.31 0.20 4263 3236 5575 5781 4805 0.16 0.31 -0.08 -0.06 0.01 0.75 0.85 -4.97 -0.20 0.04 0.83 0.78 0.88 1.22 0.65 T8 1994-1995 422 1993-1994 418 1992-1993 622 1991-1992 447 1990-1991 630 0.36 0.37 0.03 0.35 0.31 4148 3229 5600 5681 3145 0.18 0.31 -0.09 -0.04 0.35 0.52 0.86 -2.86 -0.12 1.13 0.98 0.77 0.90 1.27 0.50 26 (mm) Er deficit (kg-ha* ) decrease factor fey (kg-nr3) Treatment The three treatments achieved yields similar to that of the fully watered control (Tl) during the 1992-1993 growing season, which was an unusual season with all the treatments reaching high yields (see Tables V to VUE). From Fig. 1 it can be seen that during this specific season rainfall was mainly concentrated during the first two periods while the last two periods (yield formation) were rather dry, especially the last one. This rainfall distribution allowed a good crop development during the first two periods. It also provided high moisture storage in the soil profile in the mid and late season from which the crop could complete its growth cycle without substantial evapotranspiration restrictions (less than 10%). As can be seen in Fig. 10, the performance of T7 depends on the water supplied during the vegetatitive period. When conditions during that period are good and the crop can get a normal development of its fruit position and root system, an excellent response can be expected for (cnrW3) SOIL WATER CONTENT 0 0 0.1 0.2 0.3 ^~^. I 50 ^ ffl H O 150 GO 1994-1995 Tl 200 Fig. 10. Evolution of the soil water profile for a typical period during the early ripening stage. one irrigation during the flowering period. See Fig. 11. It reduces the fruit shedding that would occur if a well developed crop would suffer a water stress during this period. g a 0.25 0.25 Oe ^ 0 ? T3 0.15 J O \/ 0.15 ——^Floweringj- ——• JFlowering]-' 1990-91 0.05 0 60 120 0.05 1993-94 i____i 0 60 120 DAYS FROM SOWING Fig. 11. Soil water profiles (2-m rooting depth) during the 1990-1991 and 1993-1994 growing seasons. 27 3.3 ETa-Yield Relationship and Water Use efficiency As was mentioned in the introduction, a knowledge of the £Ta-yield relationship is useful for defining irrigation management under supplementary and scarce sources of water. From data collected during the five growing seasons, that relationship was ascertained for the Guazuncho-INTA cotton variety. The curve departed from the classical straight line reported by several authors but agrees with that reported by Grimes and El-Zik [7]. The yield reduction under high water availabiliy caused by an excessive ratio between vegetative and reproductive growth accounts for the shape of our curve. The relationship between water use efficiency and ETa was also investigated (Fig. 12). Although the expected linear relationship was manifested, the resulting coefficient was unusually low compared with most values reported in the literature. This low value is attributed to the effects of water shortage during the different growing stages. ^ 8000 ^ 's A 6000 gj 4000 § 2000 T = -8602 +43.57 ETa -0.03406 fe r2 = 0.60 H H O U 0 0 300 600 900 1200 ETa (mm) Fig. 12. ETa - yield relationships for INTA-Guazuncho cotton variety. 3.4 Cotton crop coefficient Kc and yield response factor ky In order to achieve the second main objective of our study the Kc for Guazuncho was determined based on the ETa of the fully watered treatment and the ET0 using pan evaporation and Kpan = 0.7 (Fig. 13). A study using the Penman-Montheith equation is currently in process and will be reported elsewhere in the near future. Pan evaporation was used in this study owing to its wide potential applicability under farmer conditions. Values of Kc and their patterns during the growing season (Fig. 14) agree fairly well with those reported by Doorenbos and Pruitt [12] and Doorenbos and Kassan [4]. In relation to the response factor ky, Fig. 14 to 16 summarize the results for the main growing periods. Although we divided the growing period in a different way, our values appear to be somewhat higher than those reported by Doorenbos and Kassan [4]. Response factors for the yield formation period are not presented because most of our data from this period was negative. 28 1 ^.• < M u z W y 0.9 a 'S 0.6 — p^ « 0.3 — 1 , •• N-4 • 1 • • • .* .. * • * Hr .:. • - • — ta C^5 — H 1 0 0 i i i 300 600 900 1200 £rß (mm) Fig. 13. EIa - water use efficiency relationship for INTA-Guazuncho cotton variety. 1.6 + Guaz • A V O • • -I Queb Guaz Queb Guaz Guaz. Guaz x Guaz ° Guaz 0 Guaz ' H 1.2 Z HH U HH & 0.8 S U Initial 0 0 | Development |Mid-season) Late season 40 80 120 160 DAYS FROM SOWING Fig. 14. Kc values during the growing season for INTA-Guazuncho cotton variety. These negative values and the fact that some of the data from other periods suggest a relationship between ky and the evapotranspiration deficit make it clear that more work has to be done on this subject The ky values summarized the crop response analyzed in section 2.2. The vegetative-bud formation period appears as the more sensitive growth period while the lowest effect of water stress is found during yield formation. The flowering period is intermediate. The mean response factor for the total growing period was greater than those for individual growing periods suggesting additive effects during the entire season. 29 l - ETa.ETm'1 1.0 0.8 0.6 0.2 0.4 VEGETATIVE & BUD FORMATION 0 0 0.2 %ö 0? 0.4 *C^>* 0.6 « r-l 0.8 MEAN k = Q.15 1.0 Fig. 15. ky factors for the vegetative and bud-formation stage of INTA-Guazumcho cotton variety. 1.0 0.8 0.6 0.4 0.2 0 0.2 0.4 0.6 • T-l FLOWERING — 0.8 1.0 Fig. 16. ky factors for the flowering stage of INTA-Guazumcho cotton variety. 4. CONCLUSIONS Results of this study indicate that cotton response to water is a complex process heavily influenced by its indeterminant growth habit, its efficient root system and its physiologic capacity of adjusting leaf turgor. However, the results make it perfectly clear that deficit irrigation on cotton is not a recommended practice for a better use of irrigation water. Rather, it is a "required" practice to reach maximum yields, inasmuch as under unrestrained water application, there is a yield reduction owing to excessive vegetative growth and loss of fruits from the lower fruiting zones. 30 The vegetative and bud formation period (0 to 65 d from sowing) was the most sensitive period to water stress. Evapotranspiration deficit over 20% during this period lead to permanent effects that could not be overcome with ample water in later periods. Under this stress intensity mainly during bud formation, the yield is affected by an early plant cut-off that reduces significantly the development of fruit position. 1.0 0.8 0.6 TOTAL GROWING SEASON Fig. 17. ky factors for the total growing season oflNTA-Guazumcho cotton variety. Water stress during the flowering period (65 to 95 d from sowing) has less effect on yield because the main plant flowering zone (1 to 4) is not seriously affected if water application during the first period was not limited. Yield reduction is mainly produced by an increment of fruit shedding. Development of new fruiting positions is also reduced but this process affects higher fruiting zones which made small contributions to the final yield under normal conditions owing to low levels of temperature and light intensity. Gentle stress during yield formation had positive effects on seed and lint yield. It reduces vegetative growth and does not affect production and retention of the main fruiting zones. From an irrigation point of view, reducing water application during this period allows the full capacity of the cotton roots to absorb water stored at rather great soil depths if soil conditions are not limiting. The additive effects of water stress occurring during different periods are evident from the response factor ky for the entire growing period which increases to values greater than 1 and to values greater than those of the individual periods. A successful irrigation strategy based on one irrigation at flowering or early yield formation was dependent upon the soil water status during the bud formation period. According to our results, under limited water availability, one irrigation at bud formation period would be advisable rather than the classical practice of irrigating at flowering. Values and the seasonal pattern of the crop coefficient Kc were similar to those reported by Doorenbos and Pruitt. However, the ky values were substantially higher than 31 those proposed by Doorenbos and Kassan. Negative values were found for the yield formation period which are the result of yield reduction under extreme water supply conditions. REFERENCES [I] [2] [3] [4] [5] [6] [7] [9] [10] [II] [12] 32 THOMAS, J.C., BROWN, K.W., JORDAN, J.R., Stomatal response to leaf water potential as affected by preconditioning water stress in the field. Agron. J. 68 (1976) 706-708. CUENCA, R.H., Irrigation system design: An engineering approach. Prentice Hall, Englewood Cliffs, Ney Jersey 07632 USA (1989). GRIMES, D.W., EL-ZIK, "Cotton" Section VE. Irrigation of Selected Crops, In Stewart, B.A., Nielsen, D.R. (ed), Irrigation of Agricultural Crops, Agronomy Monograph No. 30, ASA, CSSA and SSSA, Madison, Wisconsin, USA (1990). DOORENBOS, J., KASSAN, A.H., Yield response to water, FAO Irrigation and Drainage Paper No 33. Rome, Italy (1979). GRIMES, D.W., YAMAD A, Y.H., Relation of cotton growth and yield to minimum leaf water potential. Crop Sei. 22 (1982)134-139. GRIMES, D.W., DICKENS, W.L., Cotton response to irrigation. California Agriculture 31 No. 5 Univ. Calif., USA (1977). GRIMES, D.W., EL-ZIK, Water management for Cotton. Coop. Ext. Univ. Calif. Bulletin 1904, USA (1982). TAYLOR, H.M., JORDAN,W.R., SINCLAIR, T.R., Limitations to Efficient Water Use in Crop Production. ASA, CSSA and SSSA, Madison, Wisconsin, USA (1983). VINK, J., Crop Water Use, Irrigation and Production. Lecture notes, M.Sc. Course on Soil Science and Water Management, Univ. Agr. Wageningen, Wageningen, Holland (1982). PETERLIN, O., Crecimiento y Desarrollo del cultivo de algodon en Santiago del Estero. Personal comunication, JJNTA-EEA Sgo. del Estero (1983). DOORENBOS, J. PRUITT, W.O., Crop water requirements. FAO Irrigation and Drainage Paper No 24, Rome, Italy (1977). NUCLEAR TECHNIQUES TO EVALUATE THE WATER USE OF FIELD CROPS IRRIGATED IN DIFFERENT STAGES OF THEIR CYCLES P.L. LIBARDI, S.O. MORAES University of Sâo Paulo, Center for Nuclear Energy in Agriculture, Piracicaba A.M SAAD Institute de Pesquisas Technolôgicas, Sâo Paulo Q DE JONG VAN LIER, O. VIEIRA, R.L TUON University of Sâo Paulo, Center for Nuclear Energy in Agriculture, Piracicaba Brazil Abstract Experiments were developed in a field site of the county of Guaira, SP, Brazil, located 495 m above sea level at a latitude of 20°27'37"S and a longitude of 48°19'30"W. According to the American classification, the soil is a Typic Hapludox having a deep, homogeneous profile. In order to investigate the relationship between crop yield and the effective crop water use as a function of water stress, estimation of actual evapotranspiration ETa is a necessity. The soil water balance method was used to estimate ETa. Measurements of the soil water balance components were carried out in all experimental plots of bean and com crops. The bean crop was established in 2.0 by 5.4 m plots in a randomized complete block design containing 6 treatments (0000, 0111, 1011, 1101, 1110 and 1111) with four replications. For plants of treatment 1111, available soil water was not limited in any of the four selected development stages. The plants of treatment 0000 were irrigated on the same days with 50% of the irrigation depth applied to treatment 1111. In treatments 0111,1011,1101, and 1110, water stress was allowed during only one development stage. The corn crop was also established in a randomized complete block design with four replications using 7.2 by 10.0 m plots. The number of treatments, however, was reduced to three (111, 010 and 000). As in the case of the bean experiment, treatment 0 corresponded to 50% of the irrigation depth applied to treatment 1. Therefore, in both crops (bean and corn) two irrigation regimes were established: (i) normal irrigation in order that the actual evapotranspiration should equal the maximum evapotranspiration (ETa = ETm) and (a) a deficit irrigation, i.e. ETm > ETa. Water balance considerations were made to a depth of 0.45 m depth for the bean crop, and to the depth of 0.90 m for the com crop. In order to measure the water balance, each experimental plot was instrumented with tensiometers and a neutron probe access tube. According to the results, the following conclusions were reached: (1) The neutron probe proved to be sensitive to measure the irrigation water in the soil profile. (2) Irrigating with 50% of the necessary amount of water to obtain maximum actual evapotranspiration and maximum yield, gave bean and com grain yields much higher than those of the national average. (3) In both bean and corn crops, field water use efficiency and crop water use efficiency values were similar for each water treatment with their highest values occurring in the continuous stress treatment. (4) Water stress during the second stage (flowering period) and water stress during the third stage (yield formation) had the highest effect on bean grain yields. 33 1. INTRODUCTION The search for soil-water management systems that optimize water use of field crops should always be continued. The present Coordinated Research Program of the Joint Division FAO/IAEA has the objective to contribute to a better understanding of this subject by improving the use efficiency of water resources in irrigated agriculture. This project is a contribution to that program and sought to identify specified development stages of bean (Phaseolus vulgaris, L) and corn (Zea mays, L) during which the crops were less sensitive to water deficit. Experiments were carried out in a tropical soil of agricultural importance in a traditional irrigation field site of the State of Säo Paulo, Brazil. Neutron probe and tensiometers were used to determine the soil water balance in the different treatments. 2. MATERIALS AND METHODS Experiments were developed in a field site of the county of Guaira, SP, Brazil, located 495 m above sea level at a latitude of 20°27'37"S and a longitude of 48° 1930" W. The climate of the region is characterized by an annual average rainfall of 1,330 mm, an average air temperature of 24°C, an average relative humidity of 64% and an average wind speed of 2.6 m-S"1. According to the American classification, the soil is a Typic Hapludox having a deep, homogeneous profile with an average particle size distribution of 21% sand, 9% silt and 70% clay. Table I presents the soil profile characterization in terms of soil-water retention, soil bulk density and soil hydraulic conductivity. The soil-water retention curve for each soil depth was determined with undisturbed core samples (45 mm diameter by 45 mm height) using porous plate funnels (matric potential <j>m from 0 to -0.8 m) and porous plate pressure cells (matric potential from -3 to -150 m ). Soil bulk density values were also determined using undisturbed core samples of the same size. Soil hydraulic conductivity functions [1] were determined according to the equation in which K is the soil hydraulic conductivity, 8 the volumetric soil- water content , K0 the soil hydraulic conductivity for the field-saturated volumetric soil-water content 60 and. y an empirical parameter. Table I also presents the values of the parameters [2] in the equation e = er + (es - er )a+i a<t>\n rm (2) 2 as well as the fitting degree (r ) of the equation to the experimental data. The matric potential ^W is represented by <j> in the equation above. In order to investigate the relationship between crop yield and the effective crop water use as a function of water stress, estimation of actual evapotranspiration ETa is a necessity. The soil water balance method was used to estimate ETa. Measurements of water balance components were carried out in all experimental plots of bean and corn crops. In the water 34 TABLE I. PROPERTIES OF THE SOIL USED TO ASSESS THE SOIL-WATER BALANCE IN BOTH BEAN AND CORN CROPS. Soil-water retention curve Soil depth (m) Matric potential ^ (m) 0.0 -0.2 -0.4 -0.6 -0.8 -3.0 -6.0 -10.0 -150.0 0.10 0.20 0.30 0.45 0.60 0.75 0.90 1.05 1.20 0.59 0.51 0.45 0.42 0.39 0.32 0.30 0.28 0.26 0.58 0.50 0.43 0.40 0.38 0.32 0.31 0.30 0.29 0.60 0.49 0.41 0.38 0.35 0.31 0.28 0.28 0.24 0.60 0.48 0.40 0.37 0.35 0.30 0.29 0.28 0.25 0.61 0.54 0.43 0.38 0.35 0.29 0.27 0.26 0.23 0.60 0.52 0.45 0.41 0.38 0.31 0.29 0.29 0.27 0.61 0.54 0.47 0.43 0.40 0.32 0.30 0.30 0.28 0.62 0.54 0.45 0.41 0.38 0.31 0.29 0.28 0.26 0.63 0.55 0.44 0.39 0.36 0.29 0.27 0.27 0.25 4.20 0.444 1.798 0.278 0.609 1.000 4.23 0.465 1.867 0.262 0.618 0.999 4.51 0.465 1.869 0.246 0.631 0.998 Values of parameters for equation [2] Parameter a (m-1) m n 0r(m3-m-3) 3 3 ft (m -m- ) r2 4.40 0.396 1.660 0.278 0.576 0.999 5.87 0.371 1.594 0.259 0.581 0.998 7.30 0.391 1.644 0.238 0.612 0.996 8.59 0.369 1.585 0.244 0.596 0.997 4.40 0.461 1.855 0.233 0.610 0.991 4.51 0.435 1.770 0.264 0.600 1.000 Values of parameters for equation [1] Parameter (mm-h'1) 1.527 65.61 7 3 3 e0 (m -nr ) 0.576 Ko 2.899 69.13 0.581 4.876 8.018 11.540 14.498 18.187 23.030 27.849 61.75 56.30 52.19 51.87 49.61 45.71 43.25 0.612 0.596 0.610 0.600 0.609 0.618 0.631 Sou bulk density (kg-nr3) 1170 1150 1070 1050 1010 1040 1050 1010 960 35 balance équation, AS = P + I + C-ETa-R (3) where AS represents the change in soil water storage in the control volume of soil during a period At; P the rainfall and / the irrigation amount during At; C the capillary rise during At; D the drainage, that is, the integral of soil-water fluxes q at the lower soil boundary, also during At, ETa the actual evapotranspiration during At and R the run-off during At. The bean crop was established in 2.0 by 5.4 m plots (2.0 m parallel to the plant rows and 5.4 m perpendicular to the plant rows) in a randomized complete block design containing 6 treatments (0000, 0111, 1011, 1101, 1110 and 1111) with four replications. For the plants of treatment 1111, available soil water was not limited during any of the four selected development stages, in other words, the plants were irrigated according to the evapotranspiration losses. The plants of the treatment 0000 were irrigated on the same days with 50% of the amount of water applied to treatment 1111. Water stress was allowed during only one development stage for treatments 0111,1011,1101, and 1110. The four development stages selected were: stage 1, the vegetative period (12 to 53 d after sowing, DAS), stage 2, the flowering period (53 to 67 DAS), stage 3, the yield formation period (67 to 92 DAS) and stage 4, the ripening period (92 to 102 DAS). The corn crop was also established in a randomized complete block design with four replications using 7.2 by 10.0 m plots. The number of treatments, however, was reduced to three (111, 010 and 000). As in the case of the bean experiment, treatment 0 corresponded to 50% of the irrigation depth applied to treatment 1. Hence, treatment 111 is a fully watered treatment, 000 is a continuous stress treatment and 010 is a treatment in which water stress was allowed during two development stages. The three development stages selected were: stage 1, the vegetative period (17 to 72 DAS), stage 2, the flowering and yield formation period (72 to 104 DAS) and stage 3, the ripening period (104 to 138 DAS). Therefore, in both crops (bean and corn) two irrigation regimes were established: (i) normal irrigation in order that the actual evapotranspiration should equal the maximum evapotranspiration (ETa = ETm) and (ii) a deficit irrigation, i.e. ETm > ETa. According to soil, plant and climate conditions of the region, irrigation for the bean crop should be applied whenever the 0.1-m, 0.2-m and 0.3-m tensiometers (depending upon the crop stage) in treatment 1111 plots indicated a mean (j)m value of -6 m of water. For irrigation of the corn, the procedure was the same except that an additional tensiometer was used at the 0.4-m depth. Water balance considerations were made to a depth of 0.45 m depth for the bean crop, and to the depth of 0.90 m for the corn crop. In order to measure the water balance, each experimental plot was instrumented with tensiometers and a neutron probe access tube. In the bean experiment, mercury manometer tensiometers installed in each plot at depths of 0.1,0.2 and 0.3 m were read daily. Neutron probe readings in each plot at the depths of 0.15,0.30,0.45 and 0.60 m were taken immediately before and after each irrigation. In the corn experiment, tensiometers installed in each experimental plot at depths of 0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8, 36 0.9 and 1.0 m were read daily. Soil water storage within the soil planted to beans was measured with neutron probe readings. The calibration equation for the neutron probe (503 Hydroprobe of CPN Corporation) obtained from measurements within the soil profile at the same field site was 0 = -19.4 + 31.510? (4) 3 3 where 0is the soil water content (m -nr ) and CR the count ratio. For the corn crop, soil water storage was obtained from the tensiometer readings and the soil-water retention curves shown in Table I. Soil water fluxes q at the depth of 0.45 m in bean plots and at the depth of 0.90 m in the corn plots were estimated with Darcy-Buckingham's equation L (5) where K(6) are those functions previously determined at the site [1] having parameters for each soil depth given in Table I, z is the soil depth and ^ the total water potential head. The gradient at 0.45 m for the bean crop any time t was estimated by djt ..ft(0.3) -ft(0.6) dz 0.3 ^ } and that at 0.9 m for the corn crop any time t was estimated by 3&..&(0.8)-ft(LO) (7) IT ———02—— considering ^ = (<pm + fô) where ^ is the gravitational water potential head. The matric potential $m was calculated by the formula (t>m=-12.6h + hc+z (8) where h is the tensiometer reading (m of mercury), hc the distance (m) from the mercury level in the mercury manometer reservoir to the soil surface at the time of reading and z is the tensiometer porous cup installation depth (m). In both bean and corn experiments, hc = 0.3 m. The rainfall P was measured by rain gauges installed beside the experimental blocks. Run-off/? was considered zero. The irrigation equipment used to adequately irrigate the bean plots was a kind of tubular shower. The irrigation of the corn crop plots was made with microsprinklers fixed about 2.5 m above the soil surface in a special aluminum structure. In both irrigation systems, the water distribution uniformity tested to be more than adequate. A bean plant density of 240,000 to 300,000 plants-ha*1 was achieved by sowing seeds on May 7, 1993, at a distance of 0.54 m between rows, using 12 to 15 seeds per linear m. The beans were harvested from 5.4-m2 areas for grain production on August 17, 1993. Com seed was sown on July 29, 1994, at a spacing between rows of 0.80 m. The corn was harvested from 24m2 areas for grain production on December 15, 1994. 3. RESULTS AND DISCUSSION 3.1 Bean crop The variation of soil water matric potential with time after sowing until harvest was observed for all treatments. Fig. 1 is An example of these variations for the extreme treatments 37 (0000 and 1111) are shown in Fig.l for the 0.10-m soil depth. The response of the tensiometers to each irrigation was clearly observed. Also, as indicated by the tensiometers and expected during water deficit stages, the soil was always drier at the three depths. VEGETATIVE STAGE 0 FLOWER STAGE YIELD STAGE RIPENING STAGE A H -4 I [3 -8 • TREATMENT 1111 O TREATMENT 0000 -10 10 40 70 100 TIME AFTER SOWING (d) Fig. L Soil water matric potential during the growing season at the 0.1-m depth for treatments 111! and 0000. 40 VEGETATIVE STAGE 1 H YIELD STAGE ^ ' \ ? K \ f\ l RIPENING STAGE * ; \yWH/ v i H Z uO FLOWER STAGE 30 - l u . TREATMENT 1111 O TREATMENT 0000 20 10 40 70 TIME AFTER SOWING (d) 100 Fig. 2. Soil water content during the growing season at the 0.15-m depth for treatments 1111 and 0000. From neutron probe counts taken within the soil profile (depths of 0.15,0.30,0.45 and 0.60 m) for all treatments, volumetric soil-water contents at the 4 depths in each of the 24 38 experimental plots were frequently obtained during the growing season. Fig. 2., an example of the graphs of soil water content as a function of time, demonstrate that the neutron probe can be used effectively to monitor the soil water in each plot as a result of irrigation. The neutron probe clearly detected the differences of soil-water content among the different treatments. Bean grain yields for each treatment are shown in Table EL We note here that the Brazilian national mean production of irrigated beans is about 1500 kg-ha'1. Notice that treatment 0000 which received only one-half of the ideal irrigation amount of water produced a mean value of 2349 kg-ha4 - more than 1.5 times the national mean value. Hence, in the region of Guaira county, irrigation control by irrigating with 50% of the necessary amount of water to obtain the maximum production (treatment 1111) gives a yield much higher than the national mean. TABLE H. YIELD OF BEAN GRAIN (kg-ha4) FOR EACH EXPERIMENTAL PLOT AND MEAN AND STANDARD DEVIATION OF EACH TREATMENT Block Treatment 1 2 3 4 Mean ± std. dev. 1111 0111 1011 1101 1110 0000 3094.8 2963.4 3009.7 2522.8 3007.4 2536.1 2791.7 2671.7 2725.0 2699.3 2890.8 2233.3 2840.6 2566.1 2537.3 2217.4 2704.3 2491.5 2810.7 2733.8 2411.2 2569.9 2806.3 2135.4 2884.45 ± 141.67 2733.75 ± 168.02 2670.80 ± 260.13 2502.35 ± 204.10 2852.20 ± 128.53 2349.07 ± 195.21 mean Significancet level 5% 1% a a ab abc be c A A AB AB AB B 2665.44 ± 254.84 t Mean values followed by distinct letters are different at the indicated significance level. [DMS 5% = 328.62; DMS 1% = 414.35] Although the general mean was 2665.4 kg-ha-1 with a coefficient of variation of 9.56%, the relatively low coefficient of variation does not mean that there was no difference among treatments. By analyzing the Tukey test, data made after the confirmation that the F-test was significant, it can be concluded that for a 5% significance level, if a water deficit is provoked in the first, second or fourth development stage, the bean grain production does not differ from the situation without deficit (treatment 1111). But if the deficit is provoked in the third development stage (treatment 1101) or in all stages (treatment 0000) there is significant difference in the production with respect to the situation without deficit (treatment 1111). 39 Mean values (of the 4 replications) of the bean water balance components (considering the actual evapotranspiration as the unknown of the water balance equation) are shown in Table EH for all 6 treatments. As expected, from day 12 to day 102 after sowing, the actual evapotranspiration of treatment 1111 was 406.6 mm compared with 275.6 mm of treatment 0000. Extreme grain yields associated with extreme water treatments showed extreme plant water consumptions. Although the four intermediate treatments (0111,1011,1101 and 1110) resulted in intermediate values of ETa, they showed no correspondence with grain yield values. As can also be seen in Table IH, there was drainage only in the beginning of the first development stage owing to rainfall. In terms of actual evapotranspiration, the treatment stages that received water deficit, showed, indeed, lesser evapotranspiration than the others but the diferences were not so marked as in all stages of treatment 0000, except for stage 1. Table IV shows values of the field water use efficiency Ef, defined as the yield produced per unit of applied water (irrigation + rainfall), and values of the crop water use efficiency Ec, defined as the yield produced per unit of water taken up by bean plants. The values of Ef and Ec were similar for each treatment and among treatments 1111,0111,1011 and 1110. Treatment 0000 gave the highest values. In other words, applying irrigation with 50% of the ideal amount of water during the entire growing season improved both field water use efficiency and crop water use efficiency. Assume that during the time interval in which -0.8 > <j>m > -6 m (from field capacity to time to irrigate) the soil water storage in treatment 1111 is such that actual evapotranspiration was equal to maximum evapotranspiration [3] (ETa = ETm). Hence, for treatment 1111, the bean crop developed without water stress and as a result Ya = Ym (actual yield equals maximum yield). If this assumption is fulfilled, it is possible to establish the relationship between relative yield reduction (1 - YaY^1) and relative evapotranspiration deficits (1 - ETaET^). According this relationship, water stress during the fourth stage (treatment 1110) had limited effect on bean grain yield whereas that which occurred during the second stage (treatment 1011) had the greatest effect Between these two extreme cases, water stress in the third stage (treatment 1101) had a larger effect on the yield than water stress in all four stages (treatment 0000). Water stress in all four stages had a larger effect than water stress occurring during only the first stage (treatment 0111). 3.2 Corn Crop Unfortunately, there was considerable rainfall during development stages 2 and 3 which affected mainly stage 3 of treatment 010 and stages 2 and 3 of treatment 000. In spite of this trouble, the experiment was carried out completely. As in the bean experiment, the entire soil profile (0.1 to 1.0 m) was at field capacity at the beginning of the experiment. Hence, it was not until the middle of stage 2 (about 90 DAS) that differences between treatments were manifested. 40 TABLE m. BEAN SOIL WATER BALANCE COMPONENTS FOR ALL TREATMENTS: DAS = DAYS AFTER SOWING, D = DRAINAGE (OR CAPILLARY RISE), AS = SOIL WATER STORAGE VARIATION, / = IRRIGATION, P = RAINFALL AND ETa - ACTUAL EVAPOTRANSPIRATION DAS D (d) (mm) 12-19 19-34 34-40 40-48 48-53 53-58 58-63 63-67 67-73 73-77 77-82 82-86 86-92 92-95 95 - 102 -4.4 Total -19.3 -1.3 -2.2 0.0 0.0 -0.6 -0.9 -0.3 0.0 -0.1 -0.1 0.0 0.0 -0.2 -29.4 DAS D (d) (mm) 12-19 19-34 34-40 40-48 48-53 53-58 58-63 63-67 -11.7 -7.3 -1.5 0.0 -0.3 0.0 0.0 -0.5 TREATMENT 11 11 AS / P ETa (mm) (mm) (mm) (mm) -3.8 5.1 -1.5 -2.2 1.0 -1.7 -2.3 1.1 -3.8 -2.6 0.8 3.7 -9.9 5.7 3.4 -7.0 AS (mm) -2.9 1.6 -4.8 -5.5 4.7 -0.7 1.1 3.5 18.0 56.4 22.0 23.7 22.0 24.8 24.6 25.0 23.3 23.2 23.2 22.2 22.5 24.0 22.3 377.2 17.4 36.468.4 22.2 8.4 32.1 21.0 26.5 26.3 23.0 7.0 5L8 26.8 25.8 22.3 18.4 32.4 18.3 25.7 ETa (mm-cH) 2.5 4.6 3.7 4.0 3.8 5.3 5.3 5.8 4.5 6.5 4.5 4.6 5.4 6.1 3.7 406.6 mean = 4.5 TREATMENT 01 11 I P (mm) ETa (mm) (mm) 18.0 35.3 12.5 12.7 12.5 24.8 24.6 25.0 9.2 36.462.8 15.8 8.4 26.6 8.1 25.5 23.5 21.0 ETa (mm-d'1) 1.3 4.2 2.6 3.3 1.6 5.1 4.7 5.3 41 TABLE ffl. (continued) 67-73 73-77 77-82 82-86 86-92 92-95 95 - 102 Total -0.7 -0.4 -0.4 -1.9 -0.8 -0.5 -0.7 -26.2 -0.4 -2.5 1.2 4.7 -12.2 7.6 0.8 -3.8 23.3 23.2 23.2 22.2 22.5 24.0 22.3 7.0 326.1 23.0 25.3 21.6 15.6 33.9 15.9 27.8 3.8 6.3 4.3 3.9 5.7 5.3 4.0 51.8 355.6 mean = 4.0 TREATMENT 1011 (mm) AS (mm) I (mm) P ETa (mm) (mm) 12-19 19-34 34-40 40-48 48-53 53-58 58-63 63-67 67-73 73- 77 77-82 82-86 86-92 92-95 95 - 102 -1.4 -3.3 -1.5 -1.7 0.1 0.5 0.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.2 -3.8 7.5 -2.9 0.6 -8.0 -5.3 -6.6 -3.0 3.9 0.2 5.8 4.8 -6.9 7.7 1.7 8.0 56.4 22.0 23.7 22.0 13.6 13.8 13.5 23.3 23.2 23.2 22.2 22.5 24.0 22.3 20.4 36.4 82.0 23.4 8.4 29.8 30.1 19.4 20.8 6.5 19.4 23.0 17.4 17.4 29.4 16.3 27.4 7.0 Total 7.1 -4.4 DAS D (d) 42 343.7 51.8 ETa (mm-d'1) 2.9 5.5 3.9 3.7 6.0 3.9 4.2 4.1 3.2 5.7 3.5 4.4 4.9 5.4 3.9 392.8 mean = 4.4 TABLE m. (continued) (d) D (mm) AS (mm) 12-19 19-34 34-40 40-48 48-53 53-58 58-63 63-67 67-73 73-77 77-82 82-86 86-92 92-95 95 - 102 -8.1 -4.7 -3.0 -1.4 -1.2 -2.2 -0.5 -0.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 -2.2 DAS Total -21.4 DAS (d) (mm) 12-19 19-34 34-40 40-48 48-53 53-58 58-63 63-67 67-73 73-77 77-82 -8.8 -33.7 -5.4 -2.5 -2.9 -2.0 -1.3 -4.4 -0.6 -0.3 -0.1 D TREATMENT 1101 / P (mm) 6.8 -2.7 -2.1 0.2 1.8 -3.9 0.3 -7.6 -7.2 -2.8 -0.5 -8.9 4.3 8.0 18.0 56.4 22.0 23.7 22.0 24.8 24.6 25.0 13.2 12.0 12.0 11.7 12.1 24.0 22.3 -16.3 323.8 ETa (mm) (mm) ETa (mm-cH) 12.1 36.481.3 21.7 8.4 32.8 20.6 20.8 1.7 5.4 3.6 4.1 4.1 4.2 5.6 6.1 3.5 4.8 3.0 3.0 3.5 6.6 3.0 7.0 28.0 24.3 20.8 19.2 14.8 12.2 21.0 19.7 21.3 51.8 370.5 mean = 4.1 TREATMENT 11 10 AS 7 P ETa (mm) (mm) (mm) (mm) -2.7 6.6 -0.8 -1.8 1.1 0.9 -1.8 0.7 -2.4 -2.7 2.0 18.0 56.4 22.0 23.7 22.0 24.8 24.6 25.0 23.3 23.2 23.2 11.8 36.452.4 17.4 8.4 31.4 18.0 21.9 25.1 19.9 25.1 25.6 21.1 ETa (mm-cH) 1.7 3.5 2.9 3.9 3.6 4.4 5.0 5.0 4.2 6.4 4.2 43 TABLE ffl. (continued) 82-86 86-92 92-95 95 - 102 4.0 -12.7 1.1 -5.0 22.2 22.5 12.4 12.5 -63.9 -13.4 355.8 D (mm) TREATMENT 0000 AS I P (mm) (mm) (mm) 12 -19 19 -34 34 -40 40 -48 48 -53 53 -58 58 -63 63 -67 67 -73 73 -77 77 -82 82 -86 86 -92 92 -95 95 - 102 -7.4 -10.3 -0.7 0.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.7 0.7 -5.9 -4.8 -4.7 -4.0 -5.6 -1.3 -0.8 -3.3 1.9 1.7 3.6 1.7 -1.5 18.0 35.3 12.5 12.7 12.5 13.6 13.8 13.5 13.2 12.0 12.0 11.7 12.1 12.4 12.5 Total -18.1 -24.1 217.8 Total DAS (d) -2.0 -0.2 0.2 0.2 7.0 51.8 36.4 8.4 7.0 51.8 16.2 35.0 11.5 24.7 4.0 5.8 3.8 3.5 357.1 mean = 4.0 ETa (mm) ETa (mm-d'1) 12.3 60.6 17.7 26.2 17.2 17.6 19.4 14.8 14.0 15.3 10.1 10.0 8.5 10.7 21.0 1.8 4.0 3.0 3.3 3.4 3.5 3.9 3.7 2.3 3.8 2.0 2.5 1.4 3.6 3.0 275.6 mean = 3.1 The same procedures to calculate the soil-water balance of the bean crop were used to calculate the corn soil-water balance. Mean values of the corn water balance components are shown in Table V, VI and Vu for treatment 111,010 and 000, respectively. These tables show that, despite the rains at the end of the crop period, there was no drainage below a depth of 0.9 m during the entire growing season. Treatments that received more water during the growing season had larger evapotranspiration rates. As in the case of the bean experiment, the largest difference between evapotranspiration rates was manifested by treaments 111 and 000. In the 44 TABLE IV. FIELD WATER USE EFFICIENCY Ef AND CROP WATER USE EFFICIENCY EC FOR EACH WATER TREATMENT OF THE BEAN EXPERIMENT Treatment Ef (kg-m-3) (kg-m-3) 1111 0111 1011 1101 1110 0000 0.67 0.72 0.67 0.67 0.70 0.98 0.71 0.77 0.68 0.67 0.80 0.96 EC corn crop, the treatment stages that received water deficit (basically stages 1 and 2) also showed, as in the bean crop, lesser actual evapotranspiration. TABLE V. CORN SOIL WATER BALANCE COMPONENTS OF TREATMENT 111: DAS = DAYS AFTER SOWING, D = DRAINAGE (OR CAPILLARY RISE), AS = SOIL WATER STORAGE VARIATION, 7 = IRRIGATION, P = RAINFALL AND ETa = ACTUAL EVAPOTRANSPIRATION DAS (d) 17-25 25-32 32-36 36-40 40-45 45-51 51-55 55-59 59-63 63-68 68-72 72-76 76-81 81-86 D (mm) 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AS (mm) 7 (mm) P (mm) ETa (mm) ETa (mm-d-1) 15.6 1.6 -41.9 -18.0 21.3 -18.9 -13.0 -4.1 0.7 -3.9 -0.7 4.2 • 10.5 2.3 72.2 79.1 0.0 0.0 63.0 27.5 27.5 27.5 27.5 27.5 27.5 41.3 41.3 13.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 14.0 -56.6 -77.6 -41.9 -18.1 -41.7 -46.4 -40.5 -31.6 -26.8 -31.4 -28.2 -37.0 -30.8 -25.4 -7.1 -11.1 -10.5 -4.5 -8.3 -7.7 -10.1 -7.9 -6.7 -6.3 -7.1 -9.3 -6.2 -5.1 45 TABLE V. (continued) 86-93 93-98 98 - 104 104-110 110-115 115-119 119-124 124 - 132 132-138 Total 0.1 0.0 0.0 0.1 0.2 0.0 0.0 0.0 0.0 55.8 -52.3 -4.2 58.8 -17.2 -29.5 21.8 -3.0 -5.7 0.0 0.0 41.3 41.3 0.0 0.0 0.0 0.0 0.0 128.0 0.7 -19.8 558.0 361.0 0.0 3.0 69.0 55.0 0.0 41.0 36.0 15.0 -72.4 -52.3 -48.4 -51.6 -49.8 -29.5 -19.2 -39.0 -20.8 -10.3 -10.5 -8.1 -8.6 -10.0 -7.4 -3.8 -4.9 -3.5 -916.9 mean = -7.6 TABLE VI. CORN SOIL WATER BALANCE COMPONENTS OF TREATMENT 010: DAS = DAYS AFTER SOWING, D = DRAINAGE (OR CAPILLARY RISE), AS = SOIL WATER STORAGE VARIATION, 7 = IRRIGATION, P = RAINFALL AND ETa = ACTUAL EVAPOTRANSPIRATION DAS (d) 17-25 25-32 32-36 36-40 40-45 45-51 51-55 55-59 59-63 63-68 68-72 72-76 76-81 81-86 86-93 93-98 46 D (mm) 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.2 0.0 AS (mm) I (mm) 17.4 1.0 -43.6 -18.4 -6.7 -14.4 -5.3 -1.6 -0.8 -0.6 0.1 10.7 6.4 11.4 51.9 -54.1 72.3 79.2 0.0 0.0 34.5 13.8 13.8 13.8 13.8 13.8 13.8 41.3 41.3 13.8 0.0 0.0 P (mm) 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 o.p 14.0 128.0 0.0 (mm) ETa (mm-d"1) -55.0 -78.2 -43.6 -18.4 -41.2 -28.2 -19.1 -15.4 -14.6 -14.4 -13.7 -30.7 -34.9 -16.4 -76.2 -54.1 -6.9 -11.2 -10.9 -4.6 -8.2 -4.7 -4.8 -3.8 -3.6 -2.9 -3.4 -7.7 -7.0 -3.3 -10.9 -10.8 ETa TABLE VL (continued) 98 - 104 104-110 110-115 115-119 119-124 124 - 132 132 - 138 Total 0.0 0.1 0.1 0.0 0.0 0.0 0.0 -7.0 64.5 -17.2 -31.4 23.1 -2.6 -8.9 41.3 41.3 0.0 0.0 0.0 0.0 0.0 3.0 69.0 55.0 0.0 41.0 36.0 15.0 0.4 -26.1 447.7 361.0 -51.3 -45.9 -72.3 -31.4 -17.9 -38.6 -23.9 -8.6 -7.6 -14.5 -7.9 -3.6 -4.8 -4.0 -835.2 mean = -6.9 TABLE Vn. CORN SOIL WATER BALANCE COMPONENTS OF TREATMENT 000: DAS = DAYS AFTER SOWING, D = DRAINAGE (OR CAPILLARY RISE), AS = SOIL WATER STORAGE VARIATION, 7 = IRRIGATION, P = RAINFALL AND ETa = ACTUAL EVAPOTRANSPIRATTON DAS D (d) 17-25 25-32 32-36 36-40 40-45 45-51 51-55 55-59 59-63 63-68 68-72 72-76 76-81 81-86 86-93 93-98 98 - 104 104-110 (mm) AS (mm) 7 (mm) 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 -4.9 14.0 -45.2 -18.3 -8.3 -14.9 -5.6 -1.6 0.1 -1.5 0.5 1.6 -1.7 -0.2 74.2 -50.1 -17.1 60.1 71.6 78.3 0.0 0.0 34.1 13.6 13.6 13.6 13.6 13.6 13.6 20.5 20.5 0.0 0.0 0.0 20.5 20.5 (mm) ETa (mm) (mm-cH) 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 14.0 128.0 0.0 3.0 69.0 -76.6 -64.4 -45.3 -18.3 -42.4 -28.5 -19.3 -15.2 -13.5 -15.1 -13.1 -18.9 -22.1 -14.2 -53.8 -50.3 -40.6 -29.3 -9.6 -9.2 -11.3 -4.6 -8.5 -4.8 -4.8 -3.8 -3.4 -3.0 -3.3 -4.7 -4.4 -2.8 -7.7 -10.1 -6.8 -4.9 P ETa 47 TABLE Vn. (continued) 110-115 115-119 119-124 124 -132 132 -138 Total 0.1 0.0 0.0 0.0 0.0 3.2 -32.4 20.5 -7.7 -7.5 0.5 -42.9 0.0 0.0 0.0 0.0 0.0 347.5 55.0 0.0 41.0 36.0 15.0 361.0 -51.9 -32.4 -20.5 -43.7 -22.5 -10.4 -8.1 -4.1 -5.5 -3.7 -751.9 mean = -6.2 Corn grain yields for each treatment are shown in Table VHI. There were no significant yield differences between treatments. The overall mean was 10,141 kg-ha-1 with a coefficient of variation less than 2%. TABLE Vm. YIELD OF CORN GRAIN (kg-ha-1) FOR EACH EXPERIMENTAL PLOT AND MEAN AND STANDARD DEVIATION OF EACH TREATMENT Block Treatment 111 010 000 1 2 3 4 Mean ± std. dev. 10753 10265 9850 10256 10084 10117 19958 9948 10030 10263 ±219 10180 10529 9723 10128 ± 86 10033 ± 177 mean 10141 ± 161 From the data of Tables V to Vni values £/ (field water use efficiency) and Ec (crop water use efficiency) calculated for each treatment are presented in Table DC The Ef and Ec values were similar for each treatment and larger for the treatment that received less water (treatment 000). Plots of relative yield reduction versus relative evapotranspiration decrease [3] demonstrate the effect of water stress on yield decrease. In the present case, continuous stress and a single normal watering at flowering period + yield formation stage (treatments 000 and 010) had limited effect on the corn grain yield, probably because of rain during the second and the third development stages. 48 TABLE IX. FIELD WATER USE EFFICIENCY Ef AND CROP WATER USE EFFICIENCY Ee FOR EACH WATER TREATMENT OF THE CORN EXPERIMENT EC Treatment % (kg-m-3) (kg-m-3) 111 101 000 1.12 1.25 1.42 1.12 1.21 1.33 4. CONCLUSIONS According to the results, the following conclusions were reached: (1) The neutron probe proved to be sensitive to measure the irrigation water in the soil profile. (2) Irrigating with 50% of the necessary amount of water to obtain maximum actual evapotranspiration and maximum yield gave bean and corn grain yields much higher than those of the national average. (3) In both bean and corn crops, field water use efficiency and crop water use efficiency values were similar for each water treatment with their highest values occurring in the continuous stress treatment (4) Water stress during the second stage (flowering period) and water stress during the third stage (yield formation) had the highest effect on bean grain yields. REFERENCES [1] [2] [3] LffiARDI, P.L., REICHARDT, K., NIELSEN, D.R., BIGGAR, J.W., Simple field methods for estimating soil hydraulic conductivity, Soil Sei. Soc. Am. J., 44 (1980) 3-7. GENUTCHEN, M.T. van, A closed form equation for predicting the hydraulic conductivity of unsatureted soils. Soil. Sei. Soc. Am. J., 44 (1980) 892 - 897. DOORENBOS, J., KASSAM, A.H., Yield response to water. Irrigation and Drainage Paper 33, FAO, Rome (1979). Next page(s) left blank 49 THE RESPONSE OF WINTER WHEAT TO WATER STRESS AND NITROGEN FERTILIZER USE EFFICIENCY WANG FUJUN, QI MENGWEN, WANG HUAGUO, ZHOU CHANGJIU Laboratory for Application of Nuclear Techniques, Beijing Agricultural University, Beijing, China Abstract The response of winter wheat to water stress imposed at different crop growth stages by deficit irrigation and fertilizer use efficiency under several schemes of irrigation were evaluated on fine sandy soil and sandy loam soil. The results showed that according to grain yield response factor ky, the order of sensitive growth stages of winter wheat to water stress in decreasing sequence were booting to flowering (£,,=0.90), winter afterward to booting (£,,=0.69), flowering to milking (fcj,=0.44), seeding to winter forward (£,,=0.40) and milking to ripening (Jfej=0.25). Held water useefficiency of 16.7 kg-ftnnvha)'1 generally obtained when no water stress occurs during the growing season decreased by 5 to 20% whenever water stress occured in some growth stage. Although it was found that a high fertilizer application rate without split application would not significantly influence the yield on the fine sandy soil, scheduling of the irrigation affected the translocation of nitrogen in the plant When water stress occurred in later growth stages, the ratio of nitrogen use efficiency NUE in grain to straw decreased, and fertilizer was available for the crop only about one month after fertilizer application. The excessive fertilizer rate decreased the NUE by leaching nitrogen through the sandy soil. Total recovery of fertilizer at harvest was less than one-half that applied. 1. INTRODUCTION Improper irrigation may waste large amounts of water, leach soil nutrients and decrease crop productivity. Water for irrigation becomes more and more valuable owing to increasing costs of irrigation projects and a limited supply of good quality water. Therefore, especially for sand loam soils, we must learn how to prevent waste of irrigation water and degradation of land and improve irrigation methods for maximum crop prduction. A knowledge of crop, soil and weather is essential for these improvements. It is most important to know which growth stage of a crop is more sensitive to water stress and, according to its characteristics, establish a reasonable irrigation scheme [1,2,3, and 4]. An "Evapotranspiration-Yield" model to predict crop yield response to water stress [5 and 6] has established guidelines for irrigation scheduling suggested and recommended by FAO. This experiment was conducted to study the yield response of winter wheat to water stress imposed at different plant growth stages by deficit irrigation, to evaluate the sensitivity of winter wheat to water stress using the "ETa Yield" model and to assess nitrogen use efficiency NUE under deficit irrigation conditions. Hence, a reasonable irrigation and fertilization scheme was designed for the Hua Bei Plain - the largest wheat production plain in China. 51 2. MATERIALS AND METHODS 2.1. Experimental site Experiments were carried out in Da Ming experimental station of Beijing Agricultural University (E 114°30', N 36°46', altitude 45m). Although it is located in a semi-humid region having 589 mm average annual precipication during the last 20 years, only about 20% of the precipitation occurs during the growing season of winter wheat. The mean annual temperature is 13.3°C with annual evaporation from a free water surface being 2079 mm. 2.2. Soil characteristics The experimental plots were different for 1992-1993 and 1993-1994 with the textures of the two soils being fine sandy loam and sand loam, respectively. The soil bulk density and field capacity for the different layers of each soil are listed in Table I. The soil nutrient status of each soil is given in Table ÏÏ. TABLE I. BULK DENSITY AND FIELD CAPACITY OF THE SOILS FOR THE 1992-1993 AND 1993-1994 EXPERIMENTS Soil layer (cm) 1992-1993 Bulk density Field capacity (mm) (g-cnr3) 0-20 20-40 40-60 60-100 2.3. 1.28 1.44 1.45 1.50 39.0 49.5 56.6 131.9 1993-1994 Bulk density Field capacity (mm) (g-cnr3) 1.30 1.41 1.45 1.50 25.0 35.8 55.6 120.4 Experimental treatment 2.3.1. Plot experiment. A randomized complete block design was used for the two-year experiments consisting of 8 different irrigation treatments listed in Table m. Each treatment was replicated 4 times in 2 by 7.5 m plots. Each irrigation treatment of 1992-1993 was combined with two levels of Nfertilizer in addition to a base fertilization of 150 kg-ha-1 of ?2Os and 120 kg-ha-1 of N applied to all treatments. The two fertilizer levels were obtained by top dressing after winter just at beginning of re-greening with 90 kg-ha-1 N for low fertilizer level and 160 kg-ha'1 N for the high fertilizer level. Only irrigation treatments were made in the experiment of 1993-1994 with all plots receiving a base fertilization of 150 kg-ha-1 of P2Os, 100 K^O and 160 kg-ha-1 of N. 52 TABLE IL SOIL NUTRIENT CONTENT FOR 1992-1993 AND 1993-1994 EXPERIMENTS Soil analysis Organic matter (%) TotalN(%) Available N (mg-kg4)t Available P (mg-kg-1) Available K (mg-kg4) Soil laver during 1992- 1993 0-20 cm 20 - 40 cm 0.573 0.032 37.8 5.3 50.7 0.348 0.021 25.0 0.9 39.8 Soil laver durinsl993-1994 0 - 20 cm 20-40 cm 0.823 0.061 41.2 9.6 74.6 0.454 0.028 26.1 3.6 46.3 f Available N: alkali-hydrolysable N, Available P: Olsen P and Available K: exchangeable K A local variety of winter wheat (87-2) was seeded in October and harvested in June of, the following year. During the growing season, all procedures such as insect and weed control were the same as local practices. 2.3.2. 15 2.4. Data collection N subplot experiment. The 15N-labelled fertilizer experiment was conducted only in 1992-1993. Twenty four 65 x 60 cm subplots located in the middle of each plot were installed in 2 replicates of treatments A, D, E, F, G and TR, respectively. The date of application and amount of 15N-labelled fertilizer in each subplot were identical to those of the main plots receiving non-labelled fertilizer. The 15N abundance of the fertilizer was 2.28%. Soil samples for determining soil water content were taken by a soil auger at 10-cm intervals to a depth of 1 m once every 10 d, before and after irrigation or after a rain exceeding 15 mm. Meteorological data, such as temperature, humidity, precipitation, windspeed, sunshine and free water evaporation, were observed at a meteorological station located within 100 m from the experimental site. At harvest time, plant samples from 1.5 m2 and 0.5 m2 area in each plot were taken for determining yield and yield components, respectively. Plant samples for 15N analysis from each subplot were taken at the beginning of booting (11 April), flowering (4 May), milking (21 May) and ripening (5 June), respectively. At harvest, soil samples from the 0-20 cm and 20-40 cm soil layers were taken for 15N analysis. 53 TABLE III. AMOUNT OF IRRIGATION WATER (mm) FOR DIFFERENT TREATMENTS Year 1992 to 1993 1993 to 1994 Treatment^ A B C D E F G TR Seeding to Winter winter afterward forward to booting (60 d) (35 d) Booting to flowering (26 d) Flowering to milking (14 d) Milking to ripening (16 d) (1-1111) (0-0000) (0-1111) (1-0111) (1-1011) (1-1101) (1-1110) (0-1110) 58 12 12 58 58 58 58 0 60 20 60 15 60 60 60 100 60 20 60 60 20 60 60 100 40 0 40 40 40 0 40 100 50 0 50 50 50 50 0 0 A (1-1111) B (1-0000) 55 55 55 55 55 55 55 55 60 0 0 60 60 60 60 60 60 0 60 0 60 60 0 60 40 0 40 40 0 40 0 0 50 0 50 50 50 0 50 0 C (1-0111) D (1-1011) E (1-1101) F (1-1110) G (1-1001) TR (1-1100) t Treatments are defined as 1, no water stress in a given growth stage; 0, water stress in the growth stage provided by deficit irrigation; and TR, traditional irrigation methods. 2.5. Relèvent calculation The equation appropriate for the "Yield-£Ta model" is -1-1 i- 7=1 where Ya is the actual yield for a given treatment, Ym the maximum attainable yield without water deficit at any time during the growth season (corresponding to treatment A), ETa the actual ET during the jth day in the ith stage and ETp the maximum potential ET during the same time interval (corresponding to ETa under treatment A). Except for climate, actual evapotranspiration was affected by factors of plant and soil water content and was estimated using 54 where j^) is the crop coefficient having values of 0.76, 0.91, 1.23, 1.22 and 0.9 for each growth stage and ETp was potential ET calculated by the modified Penman formula [6]. The adjusted factor^) related to soil water content 6 [5] was taken â&fîs) - I when 6 > 6C, and when 6< 6C where 6C was the critical soil water content estimated to be 70% of field capacity. 3. RESULTS AND DISCUSSION 3.1. Actual evapotranspiration and yield Actual evapotranspiration for 1992-93 and 1993-94 is given in Table IV. Less evapotranspiration occurred during a deficit irrigation period. The slight soil water deficit during the flowering growth stage did not have a significant effect on ETa during the next growth stage when soil water content was restored. Grain and dry matter yields are shown in Table V. In the experiment of 1992-1993, yield differences between the two fertilizer levels as well as the cross effect between fertilizer and irrigation were not significantly different (LSD 5%), these results are discussed below in Section 3.4. The results from the irrigation treatments showed that when water stress occurred in a given growth stage, yields would decrease by 10 to 25%. A continuous water stress decreased yields by 20 to 30%. 3.2. Yield response factor Because the length of crop growth stages differ, the same percentage of yield decrease does not necessarily mean the same sensitivity of different plants to water stress. Hence, the sensitivity of wheat response to water stress was evaluated with the yield response factor ky being used in the "Yield-£Ta model". Values of ky for only one water stress period occurring within the entire growing season are given in Table VI. With similar results for both years, the most sensitive stage to water stress was that of booting to flowering with ky values ranging from 0.74 to 1. Another sensitive period is the stage from winter afterward to booting with ky values ranging from 0.64 to 0.72. The other growth stages were less sensitive to water stress. The " Yield-£Ta model" was tested by using the ky values in Table VI with values of ETa in Table IV to calculate the wheat yields listed in Table Vu. The calculated values of wheat yield agreed with the measured yield values for treatments TR, G and H inasmuch as the relative error was less than 10%. The error for treatment B which had water stress during the entire growing season ranged from 20 to 30% - a result probably caused by an accumulation of errors for all the different growth stage water stresses. 55 3.3. Water use efficiency Irrigation water use efficiency ^determined as yield per unit irrigation [kg-(mm-ha)'1] and field water efficiency Ec as yield per unit ETa are listed in Table VHL TABLE IV. ACTUAL EVAPOTRANSPIRATIONt (mm) OF DIFFERENT IRRIGATION TREATMENTS IN DIFFERENT GROWTH STAGES OF WINTER WHEAT Year 1992 to 1993 1993 to 1994 Treatment Seeding to Winter Booting Flowering Milking winter afterward to to to forward to booting flowering milking ripening Total (60 d) (35 d) (14 d) (26 d) (16 d) A B C D E F G TR (1-1111) (0-0000) (0-1111) (1-0111) (1-1011) (1-1101) (1-1110) (0-1110) 51.0 34.1 34.1 51.0 51.0 51.0 51.0 34.1 72.6 64.8 79.8 50.6 72.6 72.6 72.6 80.8 132.1 107.6 132.1 132.1 110.3 132.1 13t l 105.4 59.5 43.0 59.5 59.5 60.6 42.8 59.5 65.5 81.1 57.3 83.1 83.1 83.1 84.3 62.0 63.3 398.3 306.8 388.6 376.3 377.6 382.8 377.2 349.1 A B C D E F G TR (1-1111) (1-0000) (1-0111) (1-1011) (1-1101) (1-1110) (1-1001) (1-1100) 52.6 52.6 52.6 52.6 52.6 52.6 52.6 52.6 73.6 49.4 49.4 73.6 73.6 73.6 73.6 73.6 136.1 108.4 136.1 112.5 136.1 136.1 112.5 136.1 60.2 41.4 60.2 61.8 41.0 60.2 38.8 41.0 82.6 58.3 82.6 82.6 83.3 60.1 80.7 58.3 405.1 310.1 380.9 383.1 386.6 382.6 358.2 361.6 t ETa during the winter was omitted. 56 TABLE V. GRAIN AND DRY MATTER YIELDS Year Treatment A (1-1111) B (0-0000) 1992 C (0-1111) to D (1-0111) 1993 E (1-1011) F (1-1101) G (1-1110) TR (0-1 110) Year Treatment A(l-llll) B (1-0000) 1993 C(l-Olll) to D(l-lOll) 1994 ^E(l-llOl) F(l-1110) G(l-lOOl) H(l-llOO) Grain yield Low fertility High fertility (kg-ha-i) (%) (kg-ha-i) (%) 6200 a 4367 d 5370 b 4850c 5300 b 5600 b 5400 b 4810c 100 6313 a 70.4 4525 e 86.6 5500 b 78.2 5033c 85.5 5433 be 90.3 5263 be 87.1 5400 be 77.6 4850 d Grain vield (kg-ha- ^ 100 71.7 87.1 79.7 86.1 83.4 85.5 76.8 1 (%) 6384 a 4053 f 4878 e 5285c 5463c 5939 b 4969 de 5062 d 100 63.5 75.9 83.8 85.6 91.6 77.8 79.3 Drv matter vield Low fertility High fertility 1 (t-ha- ) (%) (t-ha-1) (%) 15.790 11.673 13.865 12.335 13.185 13.450 13.285 13.125 100 14.840 73.9 11.601 87.7 13.035 78.1 11.975 83.5 13.035 85.5 13.065 84.1 13.165 83.1 13.435 100 78.2 87.8 80.7 87.8 88.1 88.7 90.5 Drv matter vield (%) (t-ha-1) 100 66.4 77.8 83.4 89.8 94.6 75.2 82.9 16.825 11.186 13.105 14.054 15.140 15.936 12.679 13.969 TABLE VI. GRAIN AND DRY MATTER YIELD RESPONSE FACTORS *, Growth stage 1992-1993 High fertility Low fertility G.Y.t D.M. G.Y. D.M. Seeding to winter forward Winter afterward to booting Booting to flowering Flowering to milking Milking to ripening 0.403 0.719 0.855 0.345 0.488 0.368 0.722 0.999 0.528 0.625 0.389 0.675 0.845 0.590 0.569 0.367 0.637 0.737 0.426 0.446 1993-1994 G.Y. D.M., 0.717 0.993 0.452 0.256 0.677 1.144 0.319 0.199 G.Y.: grain yield, D.M.: dry matter. 57 TABLE Vn. THE TEST OF THE "YFBLD-ETa MODEL" Year Treatment Low fertility TR(O-lllO) 1992 Low fertility B (0-0000) High fertility TR(0-1 110) to 1993 High fertility B (0-0000) 1993 to 1994 B (O-OOOO) 0(0-1101) H (0-1 100) Calculated yield (kg-ha-i) Actual yield (kg-ha-1) Relative error 4570 3202 4754 3144 4810 4367 4850 4525 -4.9 -26.7 -2.0 -30.6 3091 4523 5083 4053 4969 5062 -23.70 -8.98 0.41 (%) Values of Ec and Ef between two fertilizer levels were not significantly different (LSD 5%). Except for treatments B and TR of 1992-1993, Ec values of the other treatments were also not significantly different (LSD 5%). Field water efficiency had its largest value of 15.7 kg-(mm-ha)"1 when there is no water stress during the entire growing season. However, if water stress occurred during one or two stages, the field water efficiency decreased 5 to 20%, and wheat yields were reduced to 12.9 kg-(mm-ha)-1 in treatment D in 1992-1993, and to 12.8 kg^mm-ha)-1 in treatment C of 1993-1994. The lower value of Ef of traditional irrigation TR of 1992-1993 showed that irrigation water applied each time exceeded that which could be held by the soil and was not used by wheat Because the amount of precipitation in the growing season of both years was no more than 50 mm, the higher value of JE^ in treatment B showed that during the summer, water stored deep in the profile supports some crop yield. 3.4. Fertilizer use efficiency The nitrogen fertilizer use efficiency values for different treatments in 1992-1993 determined by 15N-labelled fertilizer listed in Table IX showed that if no water stress occurred the fertilizer recovery would be the greatest during short periods after fertilizer application. One month after fertilizer applications, the fertilizer N recovery under the low fertilizer rate would no longer increase, and in the case of the higher fertilizer rate, it only increased 20-25%. Treatments having the same fertilizer rates manifested similar fertilizer use efficiency values. Fertilizer use efficiency at low fertilizer rates was slightly higher than that of high fertilizer levels, i.e. 40 and 30%, respectively. When water stress occurred in the stage after fertilizer application, maximum fertilizer recovery would be attained in a longer time, but final fertilizer use efficiency would not obviously decrease. Water stress in later growth stages would significantly decrease the ratio of fertilizer N recovery in grain to straw. See Table X. Because of the loss of nitrogen by leaching 58 TABLE Vm. WATER USE EFFICIENCY Grain Yield Year 1992 to 1993 Year Treatment Low High fertility fertility 1 [kg-(mm-ha)- ] (%) [kg-(mm-ha)-1] (%) A (1-1111) B (0-0000) C (0-1111) D (1-0111) E (1-1011) F (1-1101) G (1-1110) TR (0-1 110) 15.6 14.2 13.8 12.9 14.0 14.6 14.3 13.8 100 91.0 88.5 82.7 89.7 93.6 91.7 91.0 15.8 14.7 14.2 13.4 14.4 13.7 14.3 13.9 Grain Yield [kg-(mm-ha)"1] Treatment 15.76 13.07 12.81 13.80 14.13 15.51 13.87 14.00 A(l-llll) B (1-0000) 1993 C (1-01 11) to D (1-1011) 1994 E (1-1101) F (1-1 110) G (1-1001) H (1-1 100) 100 93.0 89.9 84.5 91.1 86.7 90.5 87.9 Dry matter yield Low High fertility fertility 1 [t-(mm-ha)- ] (%) 23.1 87.3 23.6 21.7 24.3 24.1 24.7 16.0 23.6 90.3 24.1 22.6 24.9 25.7 24.7 16.2 (%) Drv matter vield IXmm-ha)-1] 100 83.9 81.3 87.5 89.7 89.4 88.0 88.8 24.09 73.69 23.80 25.78 24.28 27.60 30.12 28.93 in a fine sandy soil, high fertilizer application rates result in a decrease of fertilizer use efficiency. At harvest, the fertilizer remaining in the 0-40 cm soil layer did not exceed 5% of the applied fertilizer. 4. CONCLUSION I One the basis of the 2-year experiment on fine sandy soil in the semi-humid region, the following conclusions were obtained. (1) Winter wheat manifests different levels of sensitivity to water stress during different stages of growth. According to the "Yield-£Ta model", the most sensitive period is during the flower form stage from booting to flowering. Another sensitive 59 TABLE IX. NITROGEN FERTILIZER USE EFFICIENCY Treatment April 11 May 4 May 21 Dry matter Dry matter Dry matter _______June 5________ Grain Straw Dry matter A (1-1111) 41.7 [100] D (1-0111) 37.7 [90] 42.4 [100] 40.7 [96] E (1-1011) F (1-1101) G (1-1110) TR(l-lllO) 38.4 [91] 42.2 [100] 42.2 [100] 43.7 [104] Low fertilizer!" 41.0 [100] 32.2 [100] 40.3 [98] 29.0 [90] 38.4 [94] 31.0 [96] 39.4 [96] 29.7 [92] 41.0 [100] 27.6 [86] 41.7 [102] 30.0 [93] 33.6 [100] 29.4 [87] 25.9 [77] 33.6 [100] 33.6 [100] 34.4 [103] High fertilizer''' 35.8 [100] 26.1 [100] 30.9 [86] 24.7 [95] 30.8 [86] 22.9 [88] 34.8 [97] 23.4 [90] 35.8 [100] 24.1 [93] 36.6 [102] 24.4 [94] A (1-1111) D (1-0111) E (1-1011) F (1-1101) G (1-1110) TR(0-1110) 41.7 [100] 41.7 [100] 41.7 [100] 42.8 [103] 25.3 [100] 17.9 [71] 25.3 [100] 25.3 [100] 25.3 [100] 28.8 [113] 10.2 [100] 42.4 [100] 10.7 [105] 39.7 [94] 9.1 [89] 12.1 [119] 12.4 [121] 13.3 [131] 7.7 [100] 7.4 [96] 8.0 [103] 8.5 [110] 8.8 [113] 9.4 [122] 40.1 [95] 41.8 [99] 40.1 [95] 43.4 [102] 33.8 [100] 32.2 [95] 30.9 [91] 32.0 [95] 32.9 [97] 33.9 [102] t First number is FUE (%). Second number in brackets is relative FUE compared with that of treatment A. TABLE X. RATIO OF NUE IN GRAIN TO STRAW Fertilizer treatment Low High Irrigation treatment A(l-llll) D(l-Olll) E(l-lOll) F(l-llOl) G(O-lllO) TR(0-1110) 3.15 3.37 2.70 3.34 3.40 2.86 2.45 2.75 2.22 2.75 2.25 2.59 period is the stage from winter afterward to booting. The other growth stages were less sensitive to water stress. (2) The " Yield-£ra model" is suitable for forecasting crop yield with a relative error between calculated yield and actual yield in the order of 10%. (3) For the purpose of evaluating water use efficiency, the use of field water efficiency is better than that of irrigation water efficiency. Field water efficiency has its largest value of 15.7 kg^mm-ha)*1 when there is no water stress during the entire growing season. This value decreased by 5 to 20% whenever 60 water stress occured in some growth stage. (4) If winter wheat is top dressed with fertilizer during time periods when there is no water stress, N uptake will increase. Water stress in later growth stages will significantly decrease the ratio of fertilizer N recovery in grain to that in straw. The N fertilizer use efficiency in sandy soil ranging from 30 to 40% decreases when the amount of nitrogen application increases during top dressing. REFERENCES [1] [2] [3] [4] [5] [6] DAY, A.D., INTALAP, S., Some effect of soil moisture stress the growth of wheat, Agron. J. 62 (1970) 27-29. RAS, N.H., SARMA, P.B.S., CHANDER, S., Irrigation scheduling under a limited water supply, Agric. Water Management 15 (1988) 165-167. REGINATO, R.J., HATFEELD, J.L., Winter wheat response to water and nitrogen in the north american plain, Agric. For. Meteorol. 44 (1988) 105-116. DOORENBOOS, J., KASSAM, A.S., Yield response to water. Irrigation and Drainage Paper. 33, FAO Rome Italy (1979) 35-40. LU JIEZHONG, Calculation and prediction of the water balance and the drought of the field, Acta Agriculturae Universitatis Pekinensis 8(2) (1982) 69-75 (in Chinese). TAO ZHUWEN, PEIBUXIANG, The calculation method of evapotranspiration and various value of water content in soil body, Acta Meteorological Sinica 37(4) (1979) 79-87 (in Chinese). Next page(s) left blank 61 WATER DEFICIT IMPOSED BY PARTIAL IRRIGATION AT DIFFERENT PLANT GROWTH STAGES OF COMMON BEAN (Phaseolus vulgaris L) M. CALVACHE Ecuadorian Atomic Energy Commission., Quito, Ecuador K. REICHARDT University of Sâo Paulo, Center for Nuclear Energy in Agriculture, Piracicaba Brazil Abstract The purpose of this study was to identify specific growth stages of common bean crop, at which the plant is less sensitive to water stress so that irrigation can be omitted without significant decrease in biological nitrogen fixation and final yield. The field experiment was conducted at "La Tola" University Experiment Station, Tumbaco, Pichincha, Ecuador, on a sandy loam soil (Typic durustoll). The climate was warm and dry (mean air temperature 16 °C and mean relative humidity 74%) during the cropping season and rainfall of 123 mm was recorded during the cropping period (July to October, 1992 and 1994). The treatments consisted of combinations of 7 irrigation regimes (II is all normal watering; 12 all stress; 13 traditional practice; 14 single stress at vegetation; 15 single stress at flowering; 16 single stress at yield formation and 17 single stress at ripening stages) and 2 levels of applied N (20 and 80 kg.ha-1). These 14 treatment combinations were arranged and analyzed in a split-plot design with 4 replications. The plot size was 33.6 m2 (8 rows each 7 m long) with a plant population maintained at 120,000 plants per ha. Differential irrigation was started after 3 uniform irrigations for germination and crop establishment. Soil moisture was monitored with a neutron probe down to 0.60 m depth, before and 24 h after each irrigation. The actual evapotranspiration ETa of the crop was estimated by the water-balance technique. Field water efficiency Ef.(kg.m-3) and crop water use efficiency EC (kg.m-3) were calculated by dividing actual grain yield (10% humidity) by irrigation and by ETa, respectively. Biological nitrogen fixation was calculated using N-15 methodology in the 20 kg N.ha-1 treatment. From the yield data, it can be concluded that treatments which had irrigation deficit had lower yield than those that had supplementary irrigation (1% probability). The flowering stage was the most sensitive to moisture stress. Nitrogen fertilization significantly increased the number of pods and grain yield. Biological nitrogen fixation was significantly affected by water stress at flowering and yield formation stages. The crop water use efficiency was the lowest at flowering period and the yield response factor ky was higher in treatments 12 (all stress) and 15 (stress at flowering). Comparing with traditional practice by farmers of the region, only treatments II and 17 had 13 and 10% higher crop water use efficiency. 1. INTRODUCTION Common bean (Phaseolus vulgaris L.) is an important crop in Latin America for its grain protein content. It has relatively shallow rooting depth, is a poor nodulator, requires frequent irrigation and large supplies of N fertilizer [1]. The increasing demands for limited water supplies and rising costs of nitrogenous fertilizers require their economic application without adversely affecting production. There is little information regarding the effect of drought on 63 nitrogen fixation of common bean, however studies in other grain legumes suggest that N2 fixation is sensitive to drought and even more so than NOs reduction [2, 3, 4]. An extended period of stress during the vegetative stage retards nodulation and depresses N2 fixation [3, 5, 6]. The loss of nodule activity is reversible if the nodules lose no more than 20% of the maximum fresh weight [7]. Water stress during the reproductive stage reduce viable rhizobia in roots and yield [2,8]. The objectives of this experiment were: (i) to identify specific growth stages of common bean crop, at which the plant is less sensitive to water stress so that irrigation can be omitted without significant decrease in final yield and (ii) to determine how the irrigation deficit will affect fertilizer use efficiency and biological nitrogen fixation using N-15 methodology. 2. MATERIALS AND METHODS Two field experiments were carried out at "La Tola" Agricultural Experimental Station in Tumbaco, Pichincha, Ecuador on a Typic Durustoll soil sandy loam texture (16% clay; 0.8% organic matter and pH 6.2). Seven water stress treatments were studied at different growth stages (El - Vegetative; E2 - Flowering; E3 - Bean formation and E4 - Ripening period) and two levels of N-fertilizer applied (Fl - 20 and F2 - 80 kg N-ha-1). Two levels of irrigation were imposed for water stress: 1) Normal watering when real evapotranspiration is equal to the expected maximum ETm and 0) Deficit irrigation (1/2 ETm). These 14 treatments are given in Table I. Physical and chemical characterization of the soil and neutron probe calibration were carried out before actual experimental work was initiated. Seeds of bean cv. Imbabello were planted on July 3,1992. Plots of 33.6 m2, 8 furrows of 7 m (2 seeds per 25 cm) were distributed in a split-plot design, maintaining a plant population of 120,000 ha'1. All the necessary data to calculate Penman reference evapotranspiration was collected from June to December, 1992 and 1994. Immediately after the crop establishment period irrigation was applied every week according to the experimental design. Soil water content was measured with a neutron probe from 20 to 70 cm with measurements every 10 cm. Soil moisture at the surface was determined by the gravimetric method. The actual evapotranspiration ETa of the crop was estimated by the water-balance technique. Field water efficiency £/0ig-nr3) and crop water use efficiency Ec (kg-nr3) were calculated by dividing actual grain yield (10% seed humidity) by irrigation and by ETa, respectively [9,10]. Biological nitrogen fixation was calculated using N-15 methodology in treatments Fl, using the isotope dilution method [11]. Bean plants were harvested according to their physiological stage at 120 dap, separated into pods and straw, dried at 65°C, weighed and ground. Sub samples were analyzed for total N and N-15 isotope ratio with a NOI6e analyzer [12]. 64 TABLE I. TREATMENT COMBINATIONS Growth stage No. Treatments 1 1 Fill F1I2 FID F1I4 F1I5 F1I6 F1I7 F2I1 F2I2 F2I3 F2I4 F2I5 F2I6 F2I7 1 0 0 1 1 1 1 0 2 3 4 5 6 7 8 9 10 11 12 13 14 3. Description 1 1 1 0 0 0 - - - 1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 0 0 - - - - 0 1 1 1 1 1 1 0 1 1 0 1 1 1 1 0 1 All normal watering All stress (deficit irrigation) Traditional practice Single stress at vegetation Single stress at flowering Single stress at yield formation Single stress at ripening All normal watering All stress (deficit irrigation) Traditional practice Single stress at vegetation Single stress at flowering Single stress at yield formation Single stress at ripening RESULTS AND DISCUSSION. The results of the physical analysis and the calibration curve of the neutron probe are shown in Table ÏÏ. Table lu illustrates the data of crop evapotranspiration according to the modified Penman method and irrigation requirements applied to irrigation treatments. From the yield data presented in Table IV it can be concluded that treatments which had irrigation deficit have lower yields than those that had supplementary irrigation (1% probability). The flowering stage is the most sensitive to water stress. This treatment has the same result as the one which has water stress during the whole growing cycle in two experiments (1992 and 1994). Comparing the results of fertilized treatments it can seen that there are significant differences (1% probability) between the two fertilization levels used in two experiments (1992 and 1994). The interaction between irrigation and fertilization was not statistically significant in first experiment (1992) and was statistically different in the second experiment (1994). In 1994 the highest yield was obtained in the treatment I4F2. The lowest yields that year in decreasing order came from treatments I5F2>I5F1>I2F2>I2F1. 65 TABLE n. PHYSICAL ANALYSIS AND MOISTURE CALIBRATION CURVE OF THE SOIL Depth (cm) Sand (%) Silt (%) Clay (%) 0-28 28-53 53-84 84-110 72 54 56 69 12 20 22 20 16 26 22 20 Bulk density Particle density Porosity (g-cnr3) (g-cnr3) (%) 2.62 2.51 2.55 2.62 1.45 1.36 1.37 1.43 44.7 45.8 46.2 44.0 A B 0.009 0.18 -0.05 0.21 . - TABLE ffl. CROP EVAPOTRANSPTJRATION AND IRRIGATION REQUIREMENTS Month Decade Stage Kc ETa Effective rain Irrigation Irrigation 1 (mm-d' ) (mm-decade' ) requirement requirement (mm-d'1) (mm-decade-1) 1 Jul Jul Jul Aug Aug Aug Sep Sep Sep Oct Oct Oct Nov Total 1 2 3 1 2 3 1 2 3 1 2 3 1 initial initial develop develop develop mid mid mid late late late late late 0.35 0.35 0.38 0.54 0.80 1.04 1.15 1.15 1.11 1.02 0.90 0.79 0.68 1.60 1.61 1.80 2.68 4.18 5.22 5.48 5.29 5.00 4.47 3.88 3.34 2.82 0.2 0.0 0.5 1.0 1.5 4.5 7.5 10.4 15.3 22.1 28.1 22.2 4.5 449.1 117.6 1.57 1.61 1.75 2.58 4.03 4.77 4.74 4.25 3.47 2.25 1.08 1.13 1.33 11.0 16.1 17.5 25.8 40.3 47.7 47.4 42.5 34.7 22.5 10.8 11.3 4.0 331.4 ETm during July 23 to October 20 was 374.5 mm. ETm during July 22 to November 8 was 400 mm. 66 TABLE IV. NUMBER OF PODS PER PLANT AND YIELD AT 14% HUMIDITY 1994 1992 Factors Number of pods Yield (kg-ha-1) Number of pods Yield (kg-ha-1) Irrigation 11 12 13 14 15 16 17 Fertilization ** ** ** ** 15.2abt 9.8d 13.7b 14.1b ll.lcd 12.1bc 17.2a ** 2723ab 1207d 2347abc 2552ab 1840c 2225bc 2825a ** 24ab 15d 19c 23ab 19c 21bc 24a * 2772b 2022d 2417c 2845a 2278cd 2704b 2558bc * Interaction I1F1 I1F2 I2F1 I2F2 I3F1 I3F2 I4F1 I4F2 I5F1 I5F2 I6F1 I6F2 I7F1 I7F2 12.7b 14.2a ns 13.6 16.8 10.3 9.4 13.7 13.8 12.8 15.4 10.8 11.4 11.9 13.9 15.7 18.6 2105b 2386a ns 2321 3125 1313 1101 2231 2464 2377 2727 1818 1861 2041 2408 2369 2715 20b 21a * 2470b 2648a * 23ab 24ab 15e 16de 18cd 20c 23b 24ab 19c 19c 19c 22b 235 25a 2628de 2916bc 1870h 2174g 2352fg 2483ef 273 lab 2951a 2227g 2329fg 2667cd 2740cd 2510def 2607de CV (%) 11.4 15.2 4 4 t Values followed by different letters are significantly different (P < 0.05), ns means not significant, * means significant to 5% and ** means significant to 1%. 67 Table V presents number and weight of nodules per plant in different irrigation and fertilization treatments. It can be seen that there are no significant differences between treatments in experiment 1 (1992) but, in experiment 2 (1994) there are significant differences between TABLE V. NODULE NUMBER AND NODULE DRY WEIGHT PER PLANT DURING FLOWERING STAGE UNDER WATER STRESS TREATMENTS Factors Irrigation 11 12 13 14 15 16 17 Fertilization Interaction I1F1 I1F2 I2F1 I2F2 I3F1 I3F2 I4F1 I4F2 I5F1 I5F2 I6F1 I6F2 I7F1 I7F2 CV (%) Number of nodules per plant ns 7.4 4.5 5.8 6.8 6.3 6.7 5.8 ns 6.0 6.4 ns 6.8 8.0 5.3 3.7 5.5 6.2 6.1 7.4 5.7 6.9 7.1 6.7 5.1 6.4 Weight of nodules per plant (mg) ns 105.0 33.7 55.6 72.5 51.0 173.1 81.2 ns 73.2 83.2 * 80.0a 130.0b 40.0a 27.5a 58.7a 52.5a 75.0a 70.0a 55.0a 47.0a 181.2b 115-Ob 22.5a 140.0b 19.4 t Values followed by different letters are significantly different (P < 0.05), ns means not significant and ** means significant to 1%. 68 irrigation treatments. All treatments under stress at sampling date have less number of nodules, probably because at sampling time they were not in water stress. The percentage of nitrogen derived from the atmosphere and N fixed in the different irrigation treatments are presented in Table VI. Note that treatments II ( All normal watering) and 14 and 17 ( Stress at vegetation and ripening stages, respectively) have more N fixed than the other treatment. Treatment 12 (All stress) had the least N fixed in experiment 1 (1992) and 2 (1994). Table Vu presents the final data for variables: actual yield Ya, Irrigation /, field water use efficiency Ef, actual evapotranspiration Eta, crop water use efficiency Ec and yield response factor ky to compare two experiments. We can see that treatment II (all normal watering) and 17 ( stress at ripening stage) have larger values of Ec than the others owing to higher productivity. Treatment 12 (All stress) has a larger value of Ef, owing to the small quantity of applied irrigation water during the growing season. The ky is higher in treatments 12 and 15, and lower in treatments II and 17. Comparing with traditional practice by farmers of the region, only treatments 14 and 17 had 13 and 10% higher crop water use efficiencies. TABLE VI. PERCENTAGE OF NITROGEN DERIVED FROM THE ATMOSPHERE (Ndfa), FROM THE FERILIZER (Ndff) AND FIXED (kg-ha'1) IN DIFFERENT IRRIGATION TREATMENTS DURING 1992 AND 1994 (%) Wheat Ndff (%) Ndfa (%) *# ** *# ** 4.1bt 5.7a S.Oab 3.9b 9.2ab 8.1b 10.7a 55.5a 27.7b 53.3a 10.8a 9.3ab 10.3a 9.8ab 63.9a 55.8a 58.2a 56.9a 51.7ab 15.3e 50.3ab 62.6a 39.6b 43.1ab 59.3a 14.8 18.2 Bean Ndff Treatment Irrigation No stress All stress Traditional Stress at vegetation Stress at flowering Stress at yield formation Stress at ripening 4.1b 4.3b 4.2b CV(%) 12.9 8.5 Bean N fixed (kg-ha-1) t Values followed by different letters are significantly different (P < 0.05), ns means not significant, * means significant to 5% and ** means significant to 1%. 69 Table VII. ACTUAL YIELD AT 14% HUMIDITY Ya, IRRIGATION I, FIELD WATER EFFICIENCY £/, ACTUAL EVAPOTRANSPIRATION ETay CROP WATER USE EFHCIENCY Ec AND RELATIVE EVAPORATION DEFICIT ky. l-ETa-ETm-1 \-Ya-Ym~l Ec I Ef ETa Treatment Ya ~~~ (kg-ha-1) (mm-period-^Okg-nr3) (mm) (kg-nr3) ky I1F2 2188 207 1.06 289.5 0.76 0.227 0.006 0.024 I2F2 771 44 1.75 204.9 0.38 0.453 0.650 1.434 I3F2 1734 138 1.26 259.9 0.67 0.306 0.212 0.691 I4F2 1909 154 1.24 286.9 0.67 0.234 0.132 0.565 I5F2 1303 143 0.91 261.5 0.50 0.302 0.408 1.350 I6F2 1686 141 1.20 274.4 0.61 0.267 0.234 0.874 I7F2 2112 193 1.09 286.3 0.74 0.236 0.040 0.169 [1] [2] [3] [4] [5] [6] 70 REFERENCES CALVACHE, M., Biological nitrogen fixation in common bean and faba bean using N15 methodology and two control plants. Rumipamba, Quito. 6 (1991) 1-10. KDRDA C, DANSO S.K.A., ZAPATA, F., Temporal water stress effects on nodulation, nitrogen accumulation and growth of soybean. Plant and Soil 120 (1989) 49-55. SAITO, S.M., MONTANHEIRO, N.M., VICTORIA, R.C., REICHARDT, K., The effects of N fertilizer and soil moisture on the nodulation and growth of Phaseolus vulgaris. J. Agric. Sei. Camb. 103 (1984) 87-93. MILLER, A.A., GARDNER, W.R., Effect of soil and plant water potentials on the dry matter production of snap beans. Agron. J. 64 (1972) 559-565. SMITH, D.L., DIJAK, M, HUME, D.J., The effect of water deficit on N2 fixation by white bean and soybean. Canadian J. Plant Sei. 64 (1988) 957-967. ZABLOTOWICZ, R.M., FOCHT, D.D., CANNELL, G.H., Nodulation and nitrogen fixation of field growth cowpeas as influenced by well irrigated and drought conditions. Agron. 1.73(1981)9-12 [7] [8] [9] [10] [11] [12] SPRENT, J.I., The effects of water stress on nitrogen-fixing root nodules. New Phytol. 70 (1971) 9-17. ESPINOSA-VICTORIA, D., FERRERA-CERRATO, D., LARQUE-SAAVEDRA, A., LEPIZ-IDELFONSO, R., Competition and survival of Rhizobium phaseoli in water stress bean plants. Nitrogen fixation research progress (ed Evans, HJ.; Bottomley, P.J.; Newton, W.E.) Netherlands (1985) 405 p. FAO, Crop Water Requirement. FAO Irrigation and Drainage Paper No. 24 ( 1977). FAO, Yield Response to Water. FAO Irrigation and Drainage Paper No. 33 (1979). FRIED, M., MIDDELBOE, V., Measurement of the amount of nitrogen fixed by a legume crop. Plant and Soil 47 (1977) 713 - 715. FIEDLER, R., PROKSCH, G., The determination of N-15 by emission and mass spectrometry in biochemical analysis: A review, Anal, Chim. Acta, 78 (1975) 1-61. Next page(s) left blank 71 FIELD ESTIMATION OF WATER AND NITRATE BALANCE FOR AN IRRIGATED CROP G. VACHAUD, L. KENGNI, B. NORMAND, J.L. THONY Laboratoire d'Etude des Transferts en Hydrologie et Environnement, Grenoble, France Abstract The major objective of this pluriannual field experiment was to determine an optimal fertilizer application scheme in an intensive agricultural system, with the dual goal of maintaining the quality of the environment while maintaining a sustainable level of crop production. Maize was the crop studied inasmuch as it is a major crop produced in the irrigated area. The different terms of the water balance (consumption, drainage, soil storage) and of the nitrogen cycle (mineralization, plant uptake, leaching) have been obtained from intensive monitoring in the upper layer of the soil (0.8 m corresponding to the root zone of the crop) with the combined use of a neutron moisture meter, tensiometers and soil suction cups. To determine the specific fertilizer use, one subplot was fertilized with 13NH413NO3 (mean isotopic excess, 1.2788 atom % of 13N). With the combined use of soil solution extraction and soil water measurement, it was possible to obtain on a weekly basis the dynamics and the mass balance of water and fertilizer in the root zone layer during the entire growing season. The results show that, in terms of the water balance, irrigation is extremely efficient because drainage (and nitrogen) losses during crop growth are small or negligible. On the other hand, the situation is completely different during the early stage of the crop (April-June) or during the intercrop period (October to April) when the crop covers only a small portion of the soil and evaporation is very small. In winter particularly, drainage corresponds to about 90% of total inputs from precipitation, and nitrate leaching strongly depends on the amount of nitrogen remaining in the soil at harvest. It is clearly shown that reducing the amount of fertilizer from 260 kg N.ha"1 to 160 kg N.ha'1 leads to a drastic reduction of nitrate losses during this period with only a slight effect on yield. 1. INTRODUCTION Non-point source agricultural pollution of ground water has become a real threat to the environment over the last 30 years. Although subject to controversy, the relation between nitrate pollution and agricultural practices remains uncontested [1]. In Europe, studies have shown that only 50 to 70% of the fertilizer is generally used by crops with the rest being volatilized, denitrified or leached [2]. Leaching of nitrate is responsible for ground water pollution. There is a need to determine optimal fertilization rate in such a way that the amount of N remaining in the soil after harvest, and potentially leachable during winter, will be minimized without negative economic implications concerning the level of crop production. Our aim was to determine the proportion of the fertilizer N that leaves the system through the drainage water during and after the growing season for different strategies, and the effect of N application on crop production in order to optimize fertilization techniques with respect to sustainable agriculture and the protection of environmental resources. 73 The results presented in this paper are related to three continuous years of experimentation. They will serve to illustrate a methodology reported in detail [3] to obtain mass flux estimation in the unsaturated zone of the soil using repetitive, nondestructive measurement techniques. 2. MATERIALS AND METHODS The study was conducted on the Experimental Farm of La Côte Saint-André, located 40 km northwest of Grenoble, France. The site is a typical glacial terrace, with approximately 1 m thick soil resting on a 10-20 m thick layer of gravels and pebbles of high permeability. The upper layer (0-0.3 m), is a loamy sand, rich in organic matter (2 to 3%). Because the percentage and size of gravels and stones increase with depth, the root zone is practically no deeper than 0.8 m, and augering deeper than l m is virtually impossible. A very powerful water-table aquifer, undergoing large fluctuations with a depth varying between 6 and 15 m from the soil surface, is intensively used for irrigation and urban water supply. It has a high sensitivity to nitrate pollution, and Because the European limit of 50 mg NOa-L4 is reached or overpassed in a large number of wells within this aquifer, a strong conflict between human health protection and agriculture production exists. 2.1. The treatments In order to determine the amount of leaching (both water and nitrate) resulting from conventional farmers1 practices and to obtain information on the response of crop to fertilization, different levels of treatment were applied in order to develop a strategy satisfying environment protection and sustainable agriculture. Two treatments were instrumented for intensive measurements [4]: fertilization with ammonium-nitrate applied at sowing (first the conventional rate in the region: 260 kg N-ha-1 in 1991,160 kg N-ha4 in 1992 and 180 kg N-ha1 in 1993); and no fertilization (in 1991, 1992 only). In the fertilized treatment, tagged WN fertilizer was also applied on replicate microplots, one of those fully instrumented, as described later. Finally, in parrallel with measurements in the cropped plots, two subplots were also instrumented on bare soil, one fertilized and one without fertilization. The same irrigation scheduled was applied on every plot with a gun sprinkler 2.2. Measurements A series of measurements sites were installed on every treatment, each one with the following equipment (Fig.l) : - one neutron access tube to allow soil water monitoring every 0.1 m to a depth of 0.9 m - five mercury tensiometers installed vertically at 0.15,0.3,0.5,0.7 and 0.9 m, - six suction probes - two each (northern and southern) at 0.3,0.5 and 0.8 m in order to account for the variability of NOs concentration and the 15N enrichment in the soil solution, even at a short distance, 74 - one rainfall recorder, at the level of the canopy. 2.4m RAIN GAUGE • VTÂGGÈDÀRÈÂSI • 0.5m | D10.15m I 0.8m ° i °-3m O DI 0.5m I • 0.8m D10.7m I D | 0.9m • I 0.5m' SOWING LINES • SUCTION PROBE Q NEUTRON ACCESS TUBE Q TENSIOMETER Fig. 1. Experimental scheme: instrumentation monitoring sites. Altogether, eight measurements sites were available: 2 on bare soil (Tl fertilized, T8 unfertilized), 4 on fertilized maize, (T2, T3, T4, T6 one of those for 15N application), and 2 on unfertilized maize (T5, T7). Measurements of soil moisture were taken twice weekly from June to October (with the exception of daily measurements during intensive periods of irrigation) and every two weeks until mid-February. Tensiometer readings were recorded daily from June to harvest, and then every week until the onset of frost. Rainfall/irrigation and micrometeorological data were obtained every 30 min with the use of a data acquisition system connected by modem to the laboratory. Soil solution samples were extracted with ceramic porous cups once a week. The samples were immediately sterilized in-situ with the use of a 0.4 jim Millipore filter (MILLEX SLHV 025LS) and then deep-frozen. The samples were first analyzed by liquid chromatography to obtain the nitrate concentration. Samples taken from the 15N subplot were subsequently reduced by dehydration, and the isotope ratio determined by mass spectroscopy. 2.3. Determination of water balance The water balance was calculated from the mass conservation equation (1) where AS is the change in water storage in the root zone, R the rainfall and irrigation amount, D the drainage at the depth zr below the root zone and ACT the actual evapotranspiration with all 75 values being related to a period of time At. Inasmuch as runoff was practically nil on this particular field site, runoff was neglected. Rainfall or irrigation were measured with raing gauges, and the change in water storage was obtained by the difference of water storage in the root zone (from 0 to 0.8 m) as measured by the neutron probe between two consecutive dates. The remaining unknown terms in (1) are D and AET. For most of this work, the drainage component D was calculated from Darcy's law D = qAt = -K(e}grad(H}At (2) 1 where q is the mean volumetric flux density (mm-d" ) during At, K(9) is the hydraulic conductivity (mm-d-1) corresponding to the water content 6 at zr, and grad(H) is the hydraulic head gradient at this same depth. This method requires that K(8) be known. This relationship was determined through the use of the "zero flux plane" method described in detail [5], A typical result is given in Fig. 2. In order to increase the accuracy of the estimation of drainage, because of the high sensitivity of K with 6,K(0) was determined for every measurement site and computations were done on a daily basis from June to September, following a method detailed in [4], 102 I H i-* > i—< H = (18-107)014-85 1 l^io *o O U Q S* K 10 -1 0.2 03 SOIL WATER CONTENT 0.4 6 3 (cm^cnr ) Fig. 2. Relation between the hydraulic conductivity and the water content obtained in-situ on one site. 2.4. Nitrogen movement and balance By using different tests Kengni [3] showed that nitrogen in the soil was essentially in the NÛ3 form. Hence, the amount of NOs-N in a soil layer was calculated from ^=([N03-]-6^-^)-(4.42)-i (3) where [NOs-] is the nitrate content in the soil solution, 0^ the mean water content of the layer Az, N the nitrogen amount expressed in kg N-ha'1 and 4.42 the molar ratio between NOs" and N. It was assumed that the nitrate content of the soil solution extracted at 0.3 m was 76 representative for the 0 - 0.3 m layer, and that at 0.5 m for the 0.3 - 0.6 m layer, and that at 0.8 m for the 0.6 - 0.9 m layer. The summation of the 3 values obtained accordingly with the use of (3) gave the total amount of NOs-N in the 0 - 0.9 m soil profile, corresponding to the root zone. Finally, the amount of NÛ3-N leaching below the root zone was obtained from LN = D-C0.s (4) were D is the water drainage calculated at 0.8 m from (2) and CQ.S the nitrate content at that depth. Obviously, (4) does not consider the effect of dispersion. It has been shown [3] that the dispersive term of the full transport equation represents here no more than 6% of the value of the convective term, and can therefore be neglected. Finally, the partitioning between N derived from fertilizer and total N in (3) and (4) was based on the 15N-tagged fertilized site with the use of the % enrichment obtained for every soil layer NE = (NT-E)-Eo1 (5) where NE is the mineral N from the fertilizer, N? the total mineral N in the layer [eq.(3)], E the 15 N enrichment in the same layer and EQ the 15N enrichment of the fertilizer [6,7]. The amount of N derived from soil mineralization Ns was obtained through the difference NS = NT- NE. In a similar manner, the amount of fertilizer leached LE was calculated using (6) 3. RESULTS AND DISCUSSION From an agrometeorological point of view, the 3-year study was strongly influenced by not only important differences in total rainfall, but also in the distribution of rainfall: the spring was fairly dry and hot in 1991, very wet and cold the following year and an exceptionally large amount of rainfall in 1993. We present here some of the results in greater detail during the first year, and then merely comment on the differences with the other treatments during the two subsequent years. 3.1. Water balance, 1991 The total water input during the period April 1991 - February 1992 is reported in Table I. This period is divided into three parts: from sowing (April 22) to the 6-leaf stage (June 22), from then to harvest (October 4), and finally the intercrop season until February 12. Irrigation was applied with a sprinkler gun at approximately 10-d intervals from mid-June to September. The total amount of water input during the growing season was 563 mm which included 238 mm from 6 sprinkler irrigation applications. A characteristic of this soil owing to its sandy nature is its rapid response to a water input as illustrated in Fig. 3 for three soil depths. Very clearly, the tensiometer at 0.9 m responds to a water application on the same day. As discussed previously, it is because of this dynamic 77 TABLE I. WATER (mm) AND NITROGEN (kg N-ha'1) BALANCES FOR THE SITE WITH 15N FERTILIZATION, 1991 Rainfall & Change in soil Drainage Total actual Cumlative Cumlative Irrigation water storage at 0.8 m ET N leaching L^t N uptake A^t Phase (mm) AS (mm) D (mm) ,4uET(mm) Apr 22Junl9 141 21.6 44.2 Jun 19 Oct4 422 -18.4 25.3 Oct4Febl2 321 18.8 273.1 Total 884 22.0 342.6 75.3 (kgN-ha-1) 2.3 (0) 415.2 4.3 (0.2) 29.0 (kgN-ha-1) 288 (169) 154.5 (43.5) 519.5 161.1 (43.7) 288 (169) Nitrogen derived from the fertilizer is shown in parentheses. behavior that it becomes necessary to establish the water balance with the use of Darcy's law on a daily basis. Cumulative drainage, averaged for all cropped sites (with or without fertilization) is reported in Fig. 4. Indeed, it should be noticed that in terms of water balance, no difference could be found between the two treatments. Drainage occurred essentially either before emergence of the plant, or at the end of the growing season or after harvest when the physiological activity had stopped. A good indicator of the efficiency of irrigation is that drainage losses were negligible during the active growing season. During the intercropping period, water losses were important and amounted to approximately 90% of the rain. In terms of evapotranspiration, the cumulative values for each of the two treatments given in Fig. 4 manifest no significant differences. Results summarized in Table I for these terms are related to the 15Ntagged fertilized site (T6 in 1991). 3.2. Fertilizer recovery and plant response 3.2.1. Evolution ofthefertilizer-N. The evolution of total N available in the root zone with time together with the partition between the amount derived from the fertilizer resulting from the 15N application (filled 78 H £ 40 IRRIGATIONS -^ X RAINFALLS ^ |I2o •^ \ 1 - n 1,1 | , 1 JUN 1 JUL AUG | SEP Fig. 3. Daily tensiometer and hydraulic head gradient responses. symbols) and that derived from the soil (open symbols) obtained from measurements on site T6 is shown in Fig. 5. Details can be found in [3]. In order to interpret these results, two important facts must be considered: - the production of NOs-N by mineralization of soil organic matter amounted to approximately 150 kg N-ha-1 during this specific year (a result obtained on the non-fertilized sites), - the plant nitrogen uptake on the fertilized sites (resulting from plant analysis done on a two-week interval) represented a maximum of 12 kg N-ha^-d'1 during the first two weeks of July. 79 Re on betwe ~ — _ — 3 1 O' »o § c-, ft S o° k «JL, —, ' | - ' Ci 1. ü s* d S' 0 £*• ^ Ci *^< f \ ': 1 H 8 1 8u s* 3 IT Bö • • • —— ———. 1 1 1 1 1 \ — o ~ o D o D n ............. 1 3 o <**• S ; % __ — n •/ 'a^ • f+ 1 ! ^ r"«* 1 TOTAL N - s o - W Ul FROM SOIL ^ FERTILIZER £ CUMLATIVE WATER LOSS (mm) Ui 1 1 1 1 > 3- 5 in o OO W S NITROGEN CONTENT (kg-ha-1) PLANT UPTAKE OF N [kg.(ha-dH D -•— — -a — •1 d o o oo _0 .^HARVEST (— —— —— ——' O 81 1 Fig. 5 shows clearly that owing to interactions between nitrogen transformations and plant uptake, the total nitrogen available in the root zone increased sharply (a result of the combination between fertilizer input and mineralization of the soil organic matter), and was then followed by a rapid decrease, such that at the end of the period of maximum plant uptake (mid July) the soil was almost totally depleted. Then, between mid August and mid September, there is an increase of about 50 kg N-ha4 of nitrogen in the soil attributed to the mineralization of the previously immobilized N fertilizer, and perhaps to the decay of some plant roots. 3.2.2. Nitrogen and fertilizer leaching. The amount of leached fertilizer nitrogen was estimated by continuously monitoring the isotopic ratio in the soil solution concentration at the 0.8-m soil depth, and by using Darcy's law. During the crop cycle, leaching was almost nil, owing to the fact that drainage of water was negligible. At harvest, the mineral nitrate N residue was 156 kg N-ha4 with 32% of it coming from the fertilizer. Most of the mineral nitrate N was found in the 0 - 0.6 m layer. During the intercrop period with climatic conditions favorable for nitrate leaching, a total of 155 kg N-ha4 were leached. Table I illustrates the partition between that nitrate originating from the soil and that from the fertilizer. Note the increasing proportion of the fertilizer contributing to the leaching after the harvest (Fig. 6). Some 28% of the leaching during the entire intercrop period originated from the fertilizer. This represented 17% of the total input, which is proof that the traditional fertilization scheme is excessive. 3.2.3. Fertilization and crop production. The same approach described above in Section 3.2.2. was used on the unfertilized maize plot to characterize the influence of fertilization on water and nitrogen budget 100 I § H FROM SOIL 75 e- -o >—H Z s, 50 E 2S 0 FROM FERTILIZER SEP NOV _ JAN Fig. 6. Partition between nitrogen from fertilizer and nitrogen from the soil in the leached water during winter, 1991. 81 and on plant production. The most important results from the unfertilized maize plot were: (i) as already noted, there were no differences between treatments regarding water uptake, (ii) in terms of nitrogen balance, the total amount of nitrate N available in the root zone increased to a peak value of about 150 kg N-ha'1 during mid June, and subsequently steadily decreased to zero at harvest (This behavior is proof that this soil has an excellent mineralization capacity.), (iii) in terms of plant uptake, the total uptake from the unfertilized treatment was 176 (±4) kg N-ha'1 instead of 290( ±30) kg N-ha'1 for the fertilized treatment. However, not apparent effects could be detected on the plants (same height, same color) and the variations of grain yield were slight [10.6 (±l)-103kg-ha-1 versus 13 (tlJ-KPkg-ha-1. Consequently, it was clearly concluded that the conventional rate of fertilization was too great Not only for the experimental site, but also for several farms of the county, the suggestion was given to the farmers to reduce their fertilization applications to a value close to loOkgN-ha-1. 3.3. Intel-comparison with results in 1992, and 1993 During the two following years, the experimental conditions were completely different in terms of climate and fertilization. 3.3.1. Water balance For the period between April 1 to Dec. 31, the total rainfall was of 601 mm in 1991,810 mm in 1992 and 1096 mm in 1993. Values related to the growing season as well as total irrigation during the same period are given in Table EL TABLE IL TERMS (AVERAGE OF 6 SITES) OF WATER BALANCE (mm) DURING THE 1991, 1992 AND 1993 GROWING SEASON OF MAIZE Rainfall Irrigation Drainage AET PET 1992 1991 Apr 16-Oct4 Apr 27 - Oct 8 1993 Apr 15 - Oct 20 323 227±12 82±13 460±20 250 504 120±23 157±11 460±21 276 886 121+11 482±25 531±15 215 Note that in the spring of 1992, rainfall occurring during the growing season caused in a fairly large amount of drainage and a low mineralization rate in May - June. During the following year an unsual amount of rainfall (530 mm) occurred during September and October. As a result, the 82 total amount of cumulative drainage was substantially different from one year to another, as shown in Fig. 7. However, it should be noted that during the period of irrigation (July - August), there was systematically no drainage - proof that irrigation, as it was done, was very efficient and not responsible for nitrate leaching. It is also clear that rainfall occurring at the end of the crop cycle or after harvest, at a period where evapotranspiration is very small or nil, represents a real threat for groundwater pollution in case of any residual quantity of nitrogen in the root zone at that period. MAY JUL SEP NOV JAN Fig. 7. Intercomparison between values of cumulative drainage under the cultivated sites obtained in 1991,1992 and 1993 3.3.2. Fertilization In terms of fertilization, and with regards to the results obtained in 1991, two different scenarios were applied. In 1992, the partition between fertilized and unfertilized subplot was maintained, but with a different application rate of fertilizer: a total of 160 kg N-ha-1 with 50 kg N-ha-1 applied at sowing and 110 kg N-ha-1 at the 6-leaf stage. In 1993, considering the abrupt decrease of yield on the subplot unfertilized for 2 continuous years (Table ffi), and in view of results obtained in parallel on classical agronomic tests for crop response to fertilization, it was decided to apply a uniform application of 180 kg N-ha-1 to the entire area with 40 kg N-ha-1 applied at sowing and 140 kg N-ha-1 at the 6-leaf stage. During both years, one of the plots was again fertilized with 15N-tagged ammoniun nitrate. Two important results merit comment First, and in conjunction with the data already given on Fig. 5, the evolution of the total amount of nitrate N derived from fertilizer and detected in the root zone is given on Fig. 8. Second, values of cumulated losses of nitrate N by leaching for the fertilized and unfertilized sites are given Fig. 9. In 1992, no differences could be found between the fertilized and the unfertilized treatment - proof that fertilizer efficiency was very good (with the exception of some leaching of approximately 25 kg N-ha-1 occurring during May and June. In 1993, no differences could be found between the fertilized and unfertilized 83 Cl W 1 2 • 5? E» "5? ** era x cro £ S3* P • O^ R* £2. # H \S «5 f? &• era g CÖ *N> 2 >> -fi J^ P. CL a* CT " P v3 ^ -fi *-* p p E3 V^ § t S ty <-t 4 8 1 g w L- 13 i tO H- oo 0 It 0 1+ O O -<• o oo 4^ oo -n; to 0 G> Sr* -4 8 S ^ K«a ^ ,-. Q * ?* 5-. § R ^> gl - tO 1-4 g 0 ^g Ci 1 0 î~î ^ '-x O LU f-'s t—» H-' O 1' ^1 t£ ^ *o era «s» ^ p, H-i 11 È ils» 0 N> oo ^ ^ to O ë t— > À O V^J 1 1 ^ o °° . I ic tn ® , C , L-« ——— 50 ke N-ha"1 • " " ^^^ 5 ^( ~«< t"* S 1/lOknMha" 1 • ~ - ^^ L^ — *J\J AV^ IT Hu. . 2 • "•• " 1 1 1\J in Iv^ kpIN N-ha' 1 lia. "~ •• |R _ 1 * 1 1 0 d o s 2 S ^ s- 2 ^f; . *** ^S S S G ^sto > •—' LO * i| 8 ^ *1 s2 m" ^< v> ÉH. < j^ ffARVP'^T * -------- HAAV£5r ^B* • ^^ 1 g8 S S ^ ^ K, ^^5 LO 1 tn ° Jf è H- r § i (kçha'1) NITROGEN CONTENT ?* 1° | o . <> S^ 2 ^ to a\ S O 1. r** -«^tfl C N h-* ' ^ m—i ^ - vo 1 1 r — h-* V« ^5 I ^ 1 ^™ L.M r — HH i i i treatments. Leaching occurring during the fall was clearly the result of a second stage of mineralization during September and October, and not the result of excessive fertilization, as clearly shown by Fig. 8. JUU 1 1 1 1 1 1 1 1 1 1 1991 ^200 _ D E D— D u &100 O DD 02 02 • * *! * ° O 300 h-3 i 1 1 1 1 1 1 i 1 1 1992 O 200 — — S 100 _ _ O MULATIVE ? g _^-..-.— _ D 0 0 mniniFn"*" ni 1 1 l i i i i i l l l l 1 • S i i l l 1993 U 200 - — D 100 n i nlÜ-J-'-M1-i~ULJ—LiJL J—^ i MAY JUL SEP i i NOV ü n i i JAN Fig. 9. Cumulative losses of nitrates under the cultivated sites in 1991,1992 and 1993. 3.3.3. Effect on crop production Finally, in terms of plant production and plant uptake, two main conclusions can be drawn with the most important results given in Table lu. First of all, with the yield completely insensitive to the level of fertilization, the conventional rate of 260 kg N-ha'1 is too high with respect to plant uptake inasmuch as it does not take into account the production of nitrate N by mineralization of the organic matter of the soil (from 100 to 150 kg N-ha*1 depending upon the climatic condition). A reduction of 80 -100 kg N-ha*1 does not affect crop yield production but decreases drastically the amount of residual nitrate N in the root zone at harvest, thus decreases the risk of nitrate leaching. With this point clearly and quickly understood by the fanner 85 community, corresponding reductions of fertilization rates have been widely applied with success since 1992 The second conclusion is based upon the fact that the soil is very fertile. However, it has been shown that the soil cannot be cropped with success without an appropriate fertilization scheme. After 2 years without fertilization, the decrease of crop yield is drastic and beyond economic sustainable levels. Hence, it is remarkable that one year of fertilization is sufficient to restore the yield to the same level as for a plot continuously fertilized. 4. CONCLUSIONS If the hydraulic conductivity - soil water content relationship is known, this experiment has demonstrated that the method based on the use of tensiometers and the neutron probe is suitable for monitoring and assessing the water balance of an agricultural field. Clearly, under our experimental conditions, irrigation was not responsible for pollution by nitrate leaching during the growing season. On the other hand it is clear that a misestimation of N application rate may result in excessive nitrate N in the root zone at harvest which will be potentially leached during subsequent winter rains. The use of a suction cup for the determination of the concentration of the soil solution (either in terms of nitrate N or enrichment in15N) was entirely satisfactory. Indeed, sampling can be repeated during the growing season without altering water and (consequently) nitrate dynamics; the coupling with the determination of soil water flux gives a direct information on the mass flux of nitrate. This approach allows the simultaneous characterization of the amount of nitrogen taken up by the crop and that leached from the soil under field conditions. Finally, it was only through a continuous stream of communication between scientists and farmers in very close cooperation with a farmers' association that practical solutions to important environmental problems were achieved and that the multidisciplinary approach of this experimentation was successfull. ACKNOWLEGMENTS The authors wish to thank the farmers' association in collaboration with the Lycée Agricole, la Côte StAndré, and thé Service Central d'Analyse, CNRS Solaize for taking care of nitrates and 15N analysis. Funds for this work were provided mostly by the European Communitee through the STEP-DGXn program, and by "Programme Environnement", Centre National de la Recherche Scientifique (CNRS), Paris. 86 REFERENCES [1] [2] [3] [4] [5] [6] [7] SÉBILOTTE, M., L'excès des nitrates dans les eaux : un problème à résoudre, une lutte engagée. Perspectives Agricoles 115 (1987) 12-17. GUIRAUD, G., Contribution du marquage isotopique à l'évaluation des transferts d'azote entre les compartiments organiques et minéraux dans les systämes sol-plante. Thèse DocL d'Etat-Es Sciences Naturelles, Univ. P. et M. Curie (1984) pp 335. KENGNI, L., Mesure in-situ des pertes d'eau et d'azote sous culture de maâs irrigué : application à la plaine de la Bièvre (Isère). Thèse Doct Univ. Joseph Fourier Grenoble I, Grenoble (1993) pp 220. KENGNI, L., VACHAUD, G., THONY, J. L., LATY, R., VISCOGLIOSI, R., Field measurements of water and nitrogen losses under irrigated maize. J. Hydrol. (1993) Accepted for publication. VACHAUD, G., DANCETTE, C., SONKO, M., THONYJ.L., Méthodes de caractérisation hydrodynamique in-situ d'un sol non saturé. Application à deux types de sol du Sénégal en vue de la détermination des termes du bilan hydrique. Ann. Agronomiques 29 (1978)1-36. MARIOTTI, A., Apport de la géochimie isotopique à la connaissance du cycle de l'azote. Thèse de Doct, Univ. P. et M. Curie (1982) pp 476. OLSON, R.V., Fate of tagged nitrogen fertilizer applied to irrigated com. Soil Sei. Soc. Am. J. 44 (1980) 514-517. Next page(s) left blank 87 CROP YIELD RESPONSE TO DEFICIT IRRIGATION IMPOSED AT DIFFERENT PLANT GROWTH STAGES T. KOVACS, G. KOVACS, J. SZITO Research Institute for Irrigation, Szarvas, Hungary Abstract A series of field experiments were conducted between 1991-1994 using 7 irrigation treatments at two fertilizer levels. Nitrogen fertilizers used were labelled with 15N stable isotope to examine the effect of irrigation on the fertilizer N use efficiency by isotope technique. Irrigations were applied at four different growth stages of maize, soybean and potato (vegetative, flowering, yield formation and ripening) in 4 replicates. The study compared the impact of deficit irrigation (i.e. water stress imposed during one growth stage) with normal and traditional irrigation practices. Two irrigation regimes were established: (1) normal watering when available water AW was within the range of 60-90%, and (2) deficit irrigation when the AW was at 30- 60%. The reference evapotranspiration EJ0 was calculated according to the method of Penman-Monteith. The crop water requirement ETm and the actual evapotranspiration ETa were computed using CROPWAT, an FAO computer program for irrigation planning and management [8]. Every irrigation treatment was equipped with neutron access tubes in two replicates at a depth from 10 to 130 cm. Tensiometers were installed at depths of 30,50,60 and 80 cm in one replicate of treatments and were measured on a daily basis. Neutron probe measurements were taken weekly as well as before and after each irrigation. Fluctuations of die soil water table were also frequently monitored. A special low pressure drip irrigation method was used for water application. With the amount of rainfall and irrigation water supplied measured, the soil water content distribution profiles were known for the different treatments. The relationships between relative yield decrease and relative evapotranspiration as well as those between the crop yield and water use were determined. 1. INTRODUCTION The objectives of the study were: (i) to identify which of the four growth stages of field crops (i.e. vegetative, flowering, yield formation and ripening) are less sensitive to water stress, (ii) to improve traditional irrigation practices for the main irrigated crops of Hungary and (iii) to determine fertilizer interactions with irrigation using 15N-labelled nitrogen fertilizer. 2. MATERIALS AND METHODS The study compared the impact of deficit irrigation (i.e. water stress imposed during one growth stage) with normal and traditional irrigation practices. Two irrigation regimes were established: (1) normal watering when available water AW was within the range of 60-90%, and (2) deficit irrigation when the A W was at 30-60%. The reference evapotranspiration ET0 was 89 calculated according to the method of Penman-Monteith. The crop water requirement ETm and the actual evapotranspiration ETa were computed using CROPWAT, an FAO computer program for irrigation planning and management [8]. A special low pressure drip irrigation method was used for water application. Every irrigation treatment was equipped with neutron access tubes in two replicates at a depth from 10 to 130 cm. Tensiometers were installed in 1992 at depths of 30, 50, 60 and 80 cm in one replicate of treatments and were measured on a daily basis. Fluctuations of the soil water table were also frequently monitored. Neutron probe measurements were taken weekly as well as before and after each irrigation. During the 1991 season, a CPN neutron probe was used to monitor soil water content, and during the following years a VNP-1 probe was used. Neutron probe calibration curves were determined in the field. The calibration curve for the CPN probe was determined to be 6 = = -0.225 + 0.322 R where 6 is the soil water content (cm3-cnr3) and R the count ratio from the neutron sealer. In Fig.l, volumetric soil water contents derived from the calibration curves for the VNP-1 probe are plotted against those determined by oven drying. o GRAVIMETRIC METHOD • NEUTRON PROBE 25 50 75 SOIL DEPTH (cm) Fig. 1. Neutron probe calibration. 2.1 Location of 1991-1994 field experiments The experiment was conducted in southeastern Hungary (20.53° E, 46.86° N) at the Field Experimental Station of the Irrigation Research Institute, Szarvas. 2.2 Climate The climate of the area is continental with an average annual rainfall of 485 mm. Climatic data was collected using the measurements of the Agrometeorological Station located close to the experimental station. 90 2.3 Soil properties The soil is a clayey loam, medium in organic matter (2.1%), medium in available P content (12-61 ppm) and medium in available K content (110-262 ppm). The pH in the top 70 cm layer is neutral (7.1-7.6), while in the lower layers it is strongly basic (9.1-9.3). See Table I. The water table is found at a depth of about 3.5 m. The soil particle size distribution and volume relations are plotted in Fig. 2. Sand content varies with depth from 14.0 to 31.3%, silt is between 19.5 and 31.5%. Clay content is dominant in the soil profile and varies from 38.5 to 56.5%. The volume relations show that about 50% of the soil volume is occupied by solid particles,while among the pores filled with water, the unavailable volume is the highest TABLE I. CHEMICAL SOIL CHARACTERISTICS Soil depth (cm) pH 0-23 23-51 51-72 72-94 94-156 156-200 7.1 7.5 7.6 9.1 9.2 9.3 - Salt EC CaCOs content 1 (%) (mS-cnr ) (%) 0.10 0.15 0.55 0.50 0.50 0.45 3.00 4.50 13.57 12.88 12.88 12.25 0.0 2.3 0.1 1.6 11.9 10.0 Organic matter Total N (%) (%) 2.10 1.31 0.65 0.46 0.02 0.01 0.26 0.12 0.09 0.08 0.04 0.04 P K (mg-kg-1) (mg-kg-1) 61 32 13 12 22 18 262 237 218 214 166 110 Unsaturated hydraulic conductivity was measured experimentally using the instantaneous profile method whereby decreases in water stored within the profile after a heavy irrigation are monitored on a daily basis together with hydraulic gradients computered from tensiometer readings [3,4]. Values of soil water content 0 and K(Q) for the 30-40 cm and 50-60 cm soil layers as a function of time are depicted in Fig. 3. 2.4 Maize experiment, 1991 The crop studied was a father line of SC445 maize variety. The crop was seeded on 22 May,1991, at 0.2-m spacings in each row and 0.7-m spacing between rows. The yield plots were 4.2 x 6 m and the isotope plotswere 1.4 x 1 m. The isotopically labeled plots were harvested on 25 May, 1991, and yield plots on 28 September, 1991. Weeds were controlled by hand. Two fertilizer treatments were established: FI consisting of 80,40 and 40 kg-ha4 of N, P and K, respectively, and F2 consiting of 160, 80 and 80 kg-ha-1. The nitrogen fertilizer was ammonium sulphate labelled with 1% 15N atom excess to examine the effect of irrigation on fertilizer N use efficiency. 91 ( SILT VO.02-0.002 mm GRAVITATIONAL POROSITY 2.2% 25 50 VOLUME (%) Ffg. 2. Particle size distribution and volumetric relations of the soil The following five irrigation treatments including the rainfed were used: No, rainfed; TR, traditional; IE, one early as 0100; IL, one late as 0010; and 2X, 2 times that of 0110. All plots were arranged in a randomized complete block design with all treatments having 4 replications. 2.5 Maize experiment, 1992 In 1992 the same 1991 fertilizer levels were used as above. The maize variety was father line SC398. Seven irrigation treatments were established that included normal and stress waterings at the four growth stages of maize replicated 4 times. 2.6 Soybean experiment, 1993 Soybean variety Drina was tested using fertilizer levels FI (80, 100 and 100 kg-ha'1 of N, P and K, respectively), and F2 (120,150 andlSO kg-ha-1). Seven irrigation treatments were 92 0.36 I £ 0.34 0.32 DEPTH (cm) 30-40 — 50-60 MEASURED COMPUTED § ————— O 100 0 200 300 0.004 ; MEASURED COMPUTED 30-40 50-60 ^ 0.003 § DEPTH (cm) 0.002 - 0.001 - 100 O 200 TIME (h) Fig. 3. Change of mean soil water content and unsaturated hydraulic conductivity as a function oftime. performed in 4 replicates with 2 of the replicates containing 0.5-m2 isotope plots. The nitrogen fertilizer was ammonium sulphate labelled with 2% 15N atom excess. 2.7 Potato experiment, 1994 In 1994, potato variety Désirée was planted and treated with two N, P and K fertilizer levels: FI (160, 65 and 290 kg-ha-1) and F2 (200, 80 and 360 kg-ha'1) and 7 irrigation treatments. Ammonium nitrate (1% 15N atom excess) was applied in 0.5 m2 isotope subplots. 3. RESULTS AND DISCUSSION 3.1 Maize experiment results, 1991 Precipitation, which was 23% above the regional mean during the growing season of maize, totaled 258 mm. Its distribution was characterized by a wet and somewhat drier period 93 with a maximum precipitation of 131 mm taking place in July. Total water use for the growing season ETa is given in Table H TABLE II. TOTAL WATER USE ETa FOR MAIZE IN 1991 ACCORDING TO IRRIGATION TREATMENT Irrigation treatment ETa (mm) Rainfed Traditional 297 303 1 early (0100) 1 late (0010) Twice (0110) 382 396 396 ETm amounted to 396 mm. The initial soil water content was 223 mm-m-1. The range of applied water was within 35 to 100 mm. Fig. 4-5 show the water supply i.e. precipitation with irrigation and the soil moisture profiles under maize. According to these figures, no stress condition occurred inasmuch as the available water content AW of soil did not deplete down from 60 to 30% within the critical periods. Thus, the deficit irrigation regime could not be established. Normal watering regimes were applied when the A W content depleted from 80 to 50% at the yield formation and ripening stages of maize. Traditional irrigation, according to the seed agronomist of the institute, was applied before the flowering stage of maize. The greatest dry matter yield of maize was that for both fertilization levels irrigated traditionally (Table HI and Fig. 6). With fertilizer treatments FI, the 15N isotope analyses show that the %N derived from fertilizer is between 16.9 and 18.5, and 36 to 43% of the fertilizer N was utilized by the plants. With fertilizer treatments F2, the % Ndff ranges between 25.1 and 34.4, and the FUE varies between 26 and 37% (Table IV). In both cases, irrigation had the highest effect on the % Ndff and % FUE at the yield formation stage of maize (treatment IE). According to a statistical analysis of variance, there was no correlation upon the effect of fertilization between dry matter yield, Ndff and FUE, but it occurred upon the irrigation, moreover upon the fertilization by irrigation. 3.2 Maize experiment results 1992 Total water supply for maize during the 1992 growing season according to treatment is given in Table V. The range of applied irrigation water was within 120 to 230 mm. The calculated value of ETm was 618 mm. Total rainfall during the growing season was 160 mm. The initial water in the 94 60 GZ3 IRRIGATION -*- PRECIPITATION 120 0 0 25 A £50 W G 75 100 0 30 60 90 120 DAYS AFTER SOWING F/g. 4. Water supply and soil water content 6 (cm3-cm'3) measured by neutron probe in the 0110 cm soil profile. 95 i o o SOIL DEPTH (cm) a tri l O l ! ^ "a IRRIGATION (mm) tv> .U T l o o 0\ o r •^ s I I C/l CO I H os W o 0 § ^3 I TO ssTO K» .U 0\ O O O PRECIPITATION (mm) TABLE ffl. DRY MATTER YIELD OF MAIZE, 1991 Treatment Pert. Irrig. FI FI FI FI FI F2 F2 F2 F2 F2 NO TR IE IL 2X NO TR IE IL 2X Number Irrigation water of irrigations applied (mm) 0 1 1 1 2 0 1 1 1 2 Seed yield Straw yield Dry matter HI (kg-ha-1) (kg-ha-1) (kg-ha-1) % 11290+ 186 12574+ 441 12353+ 872 11160+ 619 12771 + 1046 12264+ 938 13784+1234 11746± 304 1237 1± 381 11973+ 651 20649+ 119 22353+ 785 22046 ±1224 20411+ 614 22155 + 1566 21784+1056 23483 ±1411 21174± 572 21842± 714 21533 ±1003 45 44 44 45 42 44 41 45 43 44 9360 ±229 9779 ±366 9693 ±355 9251 ±340 9384 ±523 9520±157 9699 ±336 9428 ±365 9470 ±338 9561 ±366 0 35 50 50 100 0 35 50 50 100 10.0 *£ 9.8 — 1 ~ 9.6 O 1 ^ 1991 • D *^"*"*~"""~ °^^-^c> V •— 1 f\ « AJ i&r £ 9.4 W r \^J ^*"^~ * 0.350 S3 3 « • 9 2 * ~ 0.0 260 FERTILIZER LEVEL 1 FERTILIZER LEVEL 2 --0- 1 1 320 380 440 ETa (mm) Fig. 6. Relationship between maize yield and-water use for two fertilizer levels during 1991. 97 TABLE IV. ISOTOPE PLOT RESULTS OF MAIZE, 1991 Dry matter yield Pert Irrig. (kg-ha-1) Treatment Fi Fi FI Fi Fi F2 F2 F2 F2 F2 NO TR IE IL 2X NO TR IE IL 2X 14679 15786 15000 14643 16196 14661 15143 14589 14268 14821 Total nitrogen <%) 1.17 1.17 1.12 1.17 1.19 1.12 1.18 1.20 1.21 1.21 Nitrogen yield (kg-ha-1) Ndff m 172 185 169 « 172 193 164 179 176 173 179 18.2 17.9 18.5 16.9 18.2 25.1 26.5 34.4 28.4 31.2 Fertilizer Fertilizer N N yield (kg-ha- ) utilization (%) 31.3 33.0 31.4 29.1 34.7 41.3 47.8 60.0 49.0 55.5 39 41 39 36 43 26 30 37 31 35 1 TABLE V. TOTAL WATER SUPPLY FOR MAIZE DURING 1992 Irrigation treat ETa (mm) 0111 483 1011 437 1101 468 486 1110 1111 TR 496 430 0000 395 soil profile was 123 mm-nr1. Yield response of maize in this year 1992 (TABLE VI) was different from that in 1991 (Table HI). As a result of dry and hot periods during the growing season (Fig. 7-9) a stress condition occurred especially during the flowering stage of maize. The traditional practice of irrigating with 150 mm of water resulted in the second lowest yield after the 0000 control. The FUE measured by isotope technique showed the same character (Table VII). Thus, the relationship between maize yield and water use was high at both fertilizer levels (Fig. 10). 3.3 Soybean experiment results 1993 Total water supply for soybean during the 1993 growing season according to treatment is given in Table Vu. The range of applied irrigation water was within 150 to 390 mm. The calculated value of ETm was 462 mm. Total rainfall during the growing season was 99 mm. The 98 initial water in the soil profile was 107 mm-nr1. Despite of the large number of irrigations and large amounts of applied irrigation water, soil water contents recorded with the neutron probe in the upper layers of soil were continuously low (See Fig. 11 and 12). Accordingly, the yield response of soybean owing to irrigation was abundantly apparent at both fertilizer treatmentss (Table DC. and Fig. 14). The lowest fertilizer N utilization occurred with the traditional irrigation treatment (Table X). The relationship between relative yield decrease and relative evapotranspiration show the same pattern (Fig. 13). 3.4 Potato experiment results, 1994 Total water supply for potato during the 1994 growing season according to treatment is given in Table XI. TABLE VI. WATER SUPPLY AND DRY MATTER YIELD OF MAIZE, 1992 Treatment Pert. Irrig. Fi 0111 Fi 1011 Fi 1101 Fi 1110 Fi 1111 Fi TR Fi 0000 F2 oui F2 1011 F2 1101 F2 1110 F2 un F2 TR F2 0000 Number Irrigation water of irrigations applied (mm) 6 6 6 6 6 4 6 6 6 6 6 6 4 6 220 170 200 220 230 150 120 220 170 200 220 230 150 120 Seed yield Straw yield Dry matter HI (kg-ha-1) (kg-ha-1) (kg-ha-1) % 5064±899abt 3709±849cd 4427±434bc 5320±547ab 5503±885a 3091±533d 1280±194e 4993±626ab 4105±856bc 5539±450a 5826±745a 5480±661a 34761437e 1518±305d 5580 ± 608b 6708 ± 514a 5419 + 564b 6245 ± 425ab 5502 ± 413b 4193 ± 724c 3948 + 614c 6077 ± 824a 6431 + 1415a 5944± 467a 6729 ± 752a 5819 + 961a 4021 ± 416b 3750 ± 862b 10644±1501a 10417±1308a 9846± 988a 11564± 972a 11005±1265a 7284±1255b 5228 ± 636c 11071±1292a 10536±2219a 11483± 855a 12556±1351a 11300±1506a 7497 ± 816b 5268±1158c 48 36 45 46 50 42 24 45 39 48 46 49 46 29 tMeans ± standard deviation followed by the same letter indicate means are not significantly different by Duncan's Multiple Range Test 99 TABLE VÏÏ. ISOTOPE PLOT RESULTS OF MAIZE, 1992 Treatment Feit. Irrig. Fi FI Fi FI FI Fi Fi oui 1011 1101 1110 un TR 0000 F2 oui F2 1011 F2 1101 F2 1110 F2 nu F2 TR F2 0000 Dry matter yield (kg-ha-1) Total nitrogen 5148ct 5278e 6278b 8225a 8462a 4415e 2252d 8525b 7588b 7182bc 10798a 11278a 5584cd 4B8d 1.17 1.22 1.18 1.22 1.16 1.15 1.44 1.57 1.35 1.43 1.34 1.43 1.57 1.60 (%) Nitrogen yield (kg-ha-1) Ndff 60cd 64bc 74b lOOa 98a 51d 32e 134a 102b 103b 144a 161a 88bc 66c Fertilizer N utilization (%) Fertilizer N yield (kg-ha-1) 32a 27a 24ab 17b 29a 25ab 30a 33a 24b 26ab 27ab 24b 27ab 30ab 19 17 18 17 28 13 10 44 25 27 39 39 24 20 24b 21bc 23b 21bc 35a 16bc 12c 27a 15b 17b 24a 24a 15b 12b (%) tMeans followed by the same letter are not significantly different by Duncan's Multiple Range Test The range of applied irrigation water was within 100 to 260 mm. The calculated value of ETm was 543 mm. Total rainfall during the growing season was 146 mm. The initial water in the soil profile was 108 mm-nr1. The amount of rainfall with irrigation water supplied and the soil water content distribution profiles for the different treatments are plotted in Fig. 15 and 16. Average potato yields are closely related to the amount of irrigation water applied. Fig. 17 gives the relationship between potato yield and water use. The traditional irrigation practice gave the lowest average yield. Note that yields were rather insensitive to water stress caused in the fourth stage of potato growth (Table XÏÏ). The relationship between relative yield decrease and relative evapotranspiration plotted in Fig. 18 shows that the yield decrease is perhaps somewhat greater for fertilizer treatment F2. With fertilizer treatments FI, the 15N isotope analyses show that the %N derived from fertilizer is between 33 and 56, and 6 to 12% of the fertilizer N was utilized by the plants. With fertilizer treatments F2, the % Ndff ranges between 35 and 62, and the FUE varies between 9 and 20% (Table XHI). 100 i -« SOIL DEPTH (cm) IRRIGATION (mm) o o -1 II 38 r i j i to 00 o I i o o PRECIPITATION (mm) i 60 EHD IRRIGATION —— PRECIPITATION Z O r 60 1992 TREATMENT 0000 40 § M 40 H < H O S20 0 20 i^L_l2 0 120 60 180 0 0 T'' / ' ' '' ^ « / /N K H cu §80 120 o 60 120 DAYS AFTER SOWING 180 Fig. 8. Water supply and soil water content 0 (cm3-cm:3) measured by neutron probe in the O130 cm soil profile. 102 i * l S % si SOIL DEPTH (cm) IRRIGATION (mm) o o o \ ^«S S5 ^ H <3\ I -i O i I K» O R i. s es S — l 00 _L <3\ O o o PRECIPITATION (mm) 1992 ss* H 2 N FERTILIZER LEVEL 1 —•— FERTILIZER LEVEL 2 --0 — _________I_________ 0 380 420 460 500 ETa (mm) Fig. 10. Relationship between maize yield and water use for two fertilizer levels during 1992. TABLE Vm. TOTAL WATER SUPPLY FOR SOYBEAN DURING 1993 Irrigation treat 0111 ETa (mm) 391 1011 1101 1110 1111 TR 427 429 454 454 0000 312 319 TABLE DC. WATER SUPPLY AND AVERAGE YIELD OF SOYBEAN, 1993 Irrigation treatment 0111 1011 1101 1110 1111 TR 0000 Number Irrigation water of irrigations applied (mm) 9 9 9 9 9 4 9 315 350 310 350 390 150 155 Soybean yield for fertilizer treatment Fi (kg-ha-1) Soybean yield for fertilizer treatment 1500±292cdet 1740±503bcd 1819±288bc 203 1± 84b 2672+220a 1106±264e 1357 ± 60de 1651±217c 6708 ± 76b 5419+ 92b 6245±602b 5502±114a 4193 ± 72d 3948 ± 70bc F2 (kg-ha-i) tMeans ± standard deviation followed by the same letter indicate means are not significantly different by Duncan's Multiple Range Test 104 TABLE X. ISOTOPE PLOT RESULTS OF SOYBEAN, 1993 Dry matter yield Pert. Irrig. (kg-ha-i) Treatment FI Fi Fi Fi Fi Fi Fi F2 F2 F2 F2 F2 F2 F2 oui 1011 1101 1110 1111 TR 0000 oui 1011 1101 1110 nu TR 0000 3400 3275 3716 3436 4077 2317 2257 4460 3762 3695 3737 4860 2721 3244 Total Nitrogen nitrogen yield (kg-ha-1) (%) 2.89 2.94 2.34 2.86 2.65 2.43 2.89 2.78 2.62 2.32 2.09 2.52 '2.37 2.12 98.3 96.3 87.0 98.3 108.0 56.3 65.3 124.0 98.6 85.8 78.1 122.5 64.5 68.8 Ndff (%) 33 30 33 30 29 27 28 32 31 37 28 27 28 30 Fertilizer . N yield (kg-ha-1) 32.4 28.9 28.7 29.5 31.4 15.2 18.3 39.7 30.6 31.8 21.9 33.1 18.1 20.7 Fertilizer N utilization (%) 41 36 36 37 39 19 23 33 26 27 18 28 15 17 105 60 EZD IRRIGATION -•- PRECIPITATION 1993 TREATMENT 45 90 135 DAYS AFTER SOWING 0000 _ 60 180 Fig.ll. Water supply and soil water content 6 (cm3-cnr3) measured by neutron probe in the 0130 cm soil profile. 106 60 I 1993 CZ33 IRRIGATION -*- PRECIPITATION TREATMENT 1110 60 s 40 S 040 H o hH HH 20 0 0 t v- 0 135 ISO 45 90 135 DAYS AFTER SOWING 180 45 90 20 3 Fig. 12. Water supply and soil water content 6 (cm3-cm'3) measured by neutron probe in the 0130 cm soil profile. 107 l - ETa-ETm-1 1.0 0.8 l 0.6 l 0.4 0.2 0 • COMPUTED D FERTILIZER LEVEL l o FERTILIZER LEVEL 2 Fig. 13. Relationship between relative yield decrease and relative evapotranspiration. O SH O y -^ W » > O FERTILIZER LEVEL l FERTILIZER LEVEL 2 0 300 340 380 420 460 ETa (mm) Fig. 14. Relationship between soybean yield and water use. 108 TABLE XL TOTAL WATER SUPPLY FOR POTATO DURING 1994 Irrigation treat 0111 384 ETa (mm) 1011 1101 424 401 1110 403 413 1111 TR 283 0000 429 TABLE Xn. WATER SUPPLY AND AVERAGE YIELD OF POTATO, 1994 Irrigation treatment 0000 0111 1011 1101 1110 TR mi Number Irrigation water of irrigations applied (mm) 6 6 6 6 6 4 6 Potato yield for fertilizer treatment Fi (kg-ha-1) Potato yield for fertilizer treatment 8665 ± 436bt 9225 ± 562b 10620± 766a 9020±1061b 10825 ± 158a 8900±1179b 11650± 625a 8910± lOOc 10615± 679b 10875 ± 355ab 9445±1450c 10930 ± 785ab 9170± 414c 11930+ 952a 165 250 225 230 240 100 260 F2 (kg-ha-1) tMeans ± standard deviation followed by the same letter indicate means are not significantly different by Duncan's Multiple Range Test TABLE Xm. ISOTOPE PLOT RESULTS OF POTATO, 1994 Treatment Pert. Irrig. FI TR Fi 0000 FI 1110 FI 1111 F2 TR F2 0000 F2 1110 F2 1111 Dry matter yield (kg-ha-1) 1518 1457 1589 2644 2314 2561 3115 3207 Total nitrogen (%) Nitrogen yield (kg-ha-1) Ndff 2.01 2.29 2.04 1.89 2.11 2.07 2.04 2.21 30.5 33.4 32.4 50.0 48.8 53.0 63.5 70.9 Fertilizer N utilization (%) Fertilizer N yield (kg-ha-1) 33 56 50 38 35 62 62 54 10.1 18.7 16.2 19.0 17.1 32.9 39.4 38.3 6 12 10 12 9 16 20 19 (%) 109 o TABLE XIV. MEASURED AND CALCULATED VALUES OF SOIL WATER CONTENT AND HYDRAULIC CONDUCTI VITYt e Average Computed Computed ------b(i) Soil depth e3 3 dB/dt ]n[z(d6/dt)] e K(6) K(6) 8- 0- B 3 3 1 3 3 (cm -cnr ) (cm-d-1) 30cm 40cm (cm -cm- ) (cm-d- ) (cm -cnr ) _ _ _ . _ _ _ 0.3635 0.3575 0.3605 0.3630 0.3520 0.3575 -0.00059 0.0030 -5.1369 217.93 0.0036 0.3649 0.0027 0.3615 0.3440 0.3528 -0.00029 -5.8569 0.0077 -8.55 0.0025 0.0023 0.3553 0.3520 0.3555 0.3538 0.0067 -0.00019 -6.2781 -72.21 0.0021 0.0024 0.3498 0.3505 0.3555 0.3530 -0.00014 0.0075 -104.85 -6.5770 0.3459 0.0018 0.0023 0.3485 0.3485 0.3485 -6.8088 -0.00011 0.0120 -84.85 0.0020 0.3429 0.0016 144 0.3430 0.3420 0.3425 -6.9977 0.0180 -0.000091 -67.06 0.0014 0.0016 0.3405 0.3425 0.3430 0.3428 0.0178 168 -0.000078 -7.1575 -77.01 0.0014 0.0016 0.3385 -0.000068 -7.2964 192 0.3365 0.3395 0.3380 0.0225 -66.92 0.0013 0.0013 0.3367 0.3290 0.3400 0.3345 216 -0.000060 -7.4186 0.0260 -62.61 0.0012 0.0012 0.3352 0.3165 0.3205 0.3185 -0.000045 -7.7174 288 0.0420 -45.88 0.0006 0.0010 0.3315 W=A *», d9/dt =:ABf-1, ]n[z(dO/dt)] --=A0(eo-0), A = 0.4126, B = -0.0386, r = 0.845, A, = -5.79, B0 = -57.19, r = -0.853, InKo = -5.7907, K0 = 0.003056 , b = -37.20 Time (h) 0 24 48 72 96 120 e Time <W Average depth 0 Soil 3 30cm 40cm (cm -cm-3) 0 24 48 72 96 120 144 168 192 216 288 0.3575 0.3560 0.3475 0.3485 0.3535 0.3335 0.3500 0.3375 0.3355 0.3435 0.3335 0.3615 0.3610 0.3590 0.3620 0.3575 0.3545 0.3555 0.3540 0.3480 0.3480 0.3580 0.3595 0.3585 0.3533 0.3553 0.3555 0.3440 0.3528 0.3458 0.3418 0.3458 0.3458 de/dt . -0.00026 -0.00013 -0.000086 -0.000064 -0.000051 -0.000042 -0.000036 -0.000032 -0.000028 -0.000021 In[z(d0/dt)] 00-0 (cm3-cnr3) b(i) . -5.9419 -6.6469 -7.0597 -7.3525 -7.5791 -7.7658 -7.9211 -8.0566 -8.1772 -8.4684 _ 0.0010 0.0063 0.0043 0.0040 0.0155 0.0067 0.0138 0.0178 0.0138 0.0138 . 535.52 -27.12 -137.01 -218.76 -71.08 -190.87 -104.99 -88.97 -123.62 -144.80 Computed 0 K(&) (cm-d-1) (cm3-cnr3) _ 0.0015 0.0011 0.0012 0.0012 0.0006 0.0010 0.0007 0.0006 0.0007 0.0007 _ 0.3593 0.3549 0.3524 0.3507 0.3493 0.3482 0.3472 0.3464 0.3457 0.3440 Computed K(0) (cm-d-1) _ 0.0015 0.0012 0.0010 0.00093 0.00086 0.00080 0.00076 0.00073 0.00070 0.00063 t A = 0.3799, B = -0.0175, r = 0.797, A0 = -6.48, B0 = -105.32, r = -0.799, \nK0 = -6.4774, K0 = 0.001538, b = -57.17 Uj J-4Î SOIL DEPTH (cm) IRRIGATION (mm) O SX I I S -t 1 I ** os ON O • ^.—'• • .• L^f^y" VO O g. • S1 1 PRECIPITATION (mm) i g p\ >-- îf. SOIL DEPTH (cm) IRRIGATION (mm) tj «t o\ o o o i l §*• *- I i i ii. •o o • s§ S s- 00 H* ON PRECIPITATION (mm) 12 H Q 10 O H < FERTILIZER LEVEL l FERTILIZER LEVEL 2 - I Pk 7 240 295 350 405 460 ETa (mm) Fig. 17. Relationship between potato yield and water use. - ETa-ETm'1 1.0 0.8 0.6 0.4 0.2 0 0 * COMPUTED — 0.2 D FERTILIZER LEVEL 1 o FERTILIZER LEVEL 2 0.4 0.6 0.8 y = 0.065 + 1.007* r=0.863 1.0 Fig. 18. Relationship between relative yield decrease and relative evapotranspiration. 4. CONCLUSIONS The main conclusions of the experiment were: (1) he neutron probe was found to be a suitable measuring device to monitor the soil water status for the experiments; (2) using tensiometers in only one replicate, the uncertainty of their readings precluded accurate monitoring of soil water status for each treatment; (3) measured field values of K(6) were small owing to the high clay content and sodic soil layers at and below the 60-cm soil depth; (4) except of the first season (1991) with an unusually large amount of precipitation, no deep drainage occurred below the roots owing to the methods of irrigation and the small magnitude of 113 K(6); (5) because of the high price of irrigation water and equipment for irrigation, traditional irrigation practices of the Hungarian farmers compared favorably with the deficit irrigation treatments. [1] [2] [3] [4] [5] [6] [7] [8] [9] 114 REFERENCES FAO 1979 Yield Response to Water. FAO Irrigation and Drainage Paper No. 33. Szalki, S., Szike Molnar, L., Almasi, Z., SOILWAT Computer Consultancy Model for Irrigation. Workshop on Soil Moisture Monitoring 8-12 Oct., Szarvas, Hungary (1990). LIBARDI, P.L., et al., Simple Field Methods for Estimating Soil Hydraulic Conductivity, Soil Sei. Soc. Am. J. 44 (1980). JONES, A. J., WAGENET, R. J., In Situ Estimation of Hydraulic Conductivity Using Simplified Methods, Water Resources Research 20 (1984) 1620-1626. JOHANSSON, W., Soil Physical Description of the Root Zone, Agrochemistry and Soil Science 3 (1989) 3-4. INTERNATIONAL ATOMIC ENERGY AGENCY, Use of Nuclear Techniques in Studies of Soil-Plant Relationships, Training Course Series No. 2., IAEA, Vienna (1990). KOVACS, T., KOVACS, G., Crop yield response of maize to deficit irrigation imposed at different plant growth stages (I.) Paper was presented at the "First FAO/IAEA Research Co-ordination Meeting on the Use of Nuclear and Related Techniques in Assessment of Irrigation Schedules of Field Crops to Increase Effective Use of Water in Irrigation Projects", 3-7 February, Vienna, Austria (1992). FAO CROPWAT. A computer program for irrigation planning and management. FAO Irrigation and Drainage Paper No. 46 (1992). KOVACS, T., KOVACS, G., Crop yield response of maize to deficit irrigation imposed at different plant growth stages (II.) Paper was presented at the "Second FAO/IAEA Research Co-ordination Meeting on the Use of Nuclear and Related Techniques in Assessment of Irrigation Schedules of Field Crops to Increase Effective Use of Water in Irrigation Projects", 24-28 August, Fundulea, Romania (1993). SUGARCANE YIELD RESPONSE TO DEFICIT IRRIGATION AT TWO GROWTH STAGES CBG PENE, GK EDI Institut des Savanes, (IDES S A), Bouaké, Côte d'Ivoire Abstract In order to increase crop water use efficiency, a field study in northern Ivory Coast on sugarcane (Saccharum officinarum L ) yield response to deficit irrigation during both tillering and stem elongation stages was earned out at Institut des Savanes (IDESSA) experimental station of Ferkessédougou The cane crop tested was Co 449, an early-matunng genotype of Indian origin This experiment was conducted for three consecutive years as virgin crop (from November, 1991 to December, 1992), first ratoon crop (from December, 1992 to January, 1994) and as second ratoon crop (from January, 1994 to January, 1995) The experimental design was a randomized complete block with 10 irrigation treatments m 4 replicates of plots 54 m2 Water was applied through an improved furrow irrigation system Crop water consumption was estimated using the water balance approach based on neutron probe and tensiometer measurements This field water balance method required the determination of soil hydraulic conductivity as a function of water content and the neutron calibration curve Data presented are related to the two ratoon crops for which field water balance measurements were investigated It has been shown in the study that sugarcane growth and yield decline owing to water deficit is significantly high during stem elongation as compared to that during tillering As a result, the sugarcane crop tested was much more sensitive to water stress at stem elongation than at tillering Therefore, deficit irrigation practice to increase crop water use efficiency might be recommended at tillering rather than stem elongation The water management strategy to be suggested here may consist of omitting irrigation during tillering (assuming that the crop is successfully established) for the benefit of stem elongation As far as stem elongation is concerned, a moderate water deficit of about 25% with respect to the full irrigation regime appears to increase crop water use efficiency 1. INTRODUCTION Except for crop establishment period of an early maturing sugarcane genotype (Saccharum officinarum L.) grown in Côte d'Ivoire, two important growth stages should be taken into account regarding irrigation water management, namely tillering and stem elongation [1,2,3]. The purpose of this study was to examine yield response curves to water stress imposed at both growth stages in order to determine a suitable practice for deficit irrigation. It was a goal to receive no significant yield decline owing to water deficit Water deficit irrigation may be a strategy to increase irrigated area and crop water use efficiency [4] in sugarcane and hence be a partial solution to the irrigation water supply problem being faced by the Ivorian sugar industry [5,6,7]. On the other hand, this strategy may sustain soil productivity under the Ivorian sugar industry irrigation schemes. Data reported are related to the two ratoon crops for which field water balance measurements were investigated. 115 2. MATERIALS AND METHODS 2.1. Location The study was carried out in northern Ivory Coast at IDESS A experimental station of Ferkessédougou (Latitude: 09°35'N, Longitude: 05012W, altitude: 323 m). 2.2. Climate Northern Côte d'Ivoire has a tropical and moderately humid climate with a dry season (from mid-October to March) and a rainy season (from April to mid-October). The peak of rainfall pattern takes place in August and September (heavy rainfall with about 420 mm for both the daily mean temperature ranges from 25 to 28°C. Climatic data needed to estimate crop water requirements were collected on a daily basis from the weather station which is part of the experimental station. The climate in Côte d'Ivoire, as part of the west African region is governed by the intertropical front (FIT) movement from north to south and vice-versa. The climate becomes either dry (due to the harmattan as a continental wind) or wet (because of the monsoon as a wind of ocean origin) depending on the location in the north or south of the FIT, respectively. 2.3. Soil The soil from the field experiment is known according to the French classification system as a reworked, moderately desaturated and ferrallitic. Some analyses were carried out over l m depth so as to determine soil particle size distribution and soil fertility. Soil bulk density was determined from cylinder samples of known volume. An internal drainage test [8, 9,10] was performed during the 1993-94 dry season in order to determine both the unsaturated hydraulic conductivity K(&) and soil water retention curve as a function of water content Öat three different depths (0.5,1 and 1.5 m). 2.3.7. Internal drainage method After sufficient water has infiltrated to fully wet the upper layer of a field soil, the soil surface is covered to eliminate evaporation and protect it from subsequent infiltration. Measurements of the soil water content 6(z, t) and soil water pressure head h(z, f) profiles are obtained from frequent observations of neutron probe count rates and tensiometers. At any time t, the amount of water stored in the soil profile to any given depth L obtained from the neutron probe measured profile of soil water content 6(z, f) is (1) Since the flux density q is always zero at z = 0, by assuming one-dimensional vertical flow, consideration of mass conservation simply gives M at 116 (2) According to Darcy's equation, is also defined as (3) * z=L where the total head H(z, f) = [h(z, t) + z] is obtained from tensiometer readings of soil water pressure head h(z, 0 and soil depth z. Combining equations (2) and (3), the hydraulic conductivity function can be deduced for the soil water content 0 that obtains at depth L and , time fusing As water drains from the soil profile, the value of 6 decreases and the K(d) function is obtained over a range of water contents. An empirical analytic expression of the hydraulic conductivity [11] is K(e} = K0t^[ß(e-e0)} (5) where ß is a constant and K0 and 60 are values of K and 0 during steady-state infiltration, respectively. The slope and the intercept of a semi-log plot of K(6) versus (0 - 60) yield direct measures of ß and K0, respectively. 2.3.2. Field calibration of the neutron probe A field calibration was carried out on randomly selected measuring sites of the experiment. Neutron count rates were measured at different soil depths known to vary in chemical composition and bulk density. Volumetric soil samples were taken from the same depths where neutron readings were made. Soil samples were taken close to the neutron probe access tube so that they represent a volume of soil which comes from the sphere of influence of the slow neutrons. Mean volumetric soil water content of the samples, measured gravimetrically was later regressed with neutron count ratios at that specific soil depth. For this purpose, neutron count rates were first measured in a standard medium, namely water, before field measurements were obtained. A wide range of soil water contents was obtained in order to have a reliable and accurate calibration curve. The linear regression between soil water content 0 and neutron count ratios R = N-N~l provided the calibration equation 8 = a + bR (6) where a and b are intercept and slope, respectively, and N and N0 are the neutron count rates in soil and water, respectively. For the experimental site, values of a and b were -30 and 1.21, respectively, with an excellent coefficient of determination (r2=0.92). 2.4. The sugarcane crop The commercial cane genotype used in this experiment is Co 449, an early maturing crop of Indian origin. The crop was planted on November 14,1991, and was harvested first on December 14,1992 (plant or virgin crop), secondly on January 11,1994, (first ratoon) and third on January 4,1995, (second ratoon). Each crop cycle takes about 1 year divided into 8 months 117 of vegetative growth (2 months tillering, 6 months stem elongation), 3 months for yield formation and 1 month for cane ripening. Tillering and stem elongation were both growth stages to be studied with respect to deficit irrigation practice. The stem elongation is well known as the boom stage or active vegetative growth stage in sugarcane. 2.5. Experimental design The experimental lay-out was a randomized complete block design with 10 irrigation treatments in 4 replicates. A single 54 m2 plot contained 6 cane rows of 6 m each separated by 1.5 m. No fertilizer treatment was investigated. Fertilizers were applied uniformly in the field according to common practices, namely K2O (210 kg-ha-1), P^Os (45 kg-ha-1) and N derived from urea (92 kg-ha-1). The irrigation treatments were defined as follows: (0,0) (0,1) (0.25,1) (0.5,1) (0.75,1) (1,0) ( 1,0.25) (1,0.5) (1,0.75) (1,1) 2.6. rainfed treatment which is the traditional method. However, this treatment was irrigated during one month for crop establishment For ratoon crops which are already established, no irrigation was applied, no irrigation at tillering, full irrigation at boom stage, 25% of normal watering at tillering, full irrigation at boom stage, 50% of normal watering at tillering, full irrigation at boom stage, 75% of normal watering at tillering, full irrigation at boom stage, full irrigation at tillering, no irrigation at boom stage, full irrigation at tillering, 25% of normal watering at boom stage, full irrigation at tillering, 50% of normal watering at boom stage, full irrigation at tillering, 75% of normal watering at boom stage, no water stress during both growth stages (normal watering regime from November until mid-July). Irrigation system Water was applied in cane interrows according to an improved furrow irrigation system which allowed water flow under a very low pressure (not by gravity) in 7.5-cm diameter pipes connected to a sprinkler irrigation network. Watering on cane interrows was allowed by siphons with the amount of irrigation water applied on each plot measured using a flow-meter and a small valve. 2.7. Irrigation scheduling Irrigation water allocation was made on pan evaporation basis. Except for both rainfed (0,0) and full irrigated (1,1) treatments, all treatments to be water stressed at one growth stage were fully irrigated during the other growth stage. Treatments (0,1), (0.25,1), (0.5,1) and (0.75, 1) which were water stressed at the tillering stage were fully watered at the boom stage. The reverse was true for treatments (1,0), (1,0.25), (1,0.5) and (1,0.75). 118 On the other hand, water was allocated on weekly basis with respect to the following 10day mean value crop coefficients: 0.5 (tillering: 2 months); 0.8 (1.5 months), 1.0 (stem elongation: 7 months) and 0.8 (flowering and yield formation: 1 month) [12]. 2.8. Cane stalk elongation measurements Some elongation measurements were performed twice a month on 15 primary cane stalks randomly chosen in each plot in order to examine the effect of soil water deficit on crop growth. The measurements of elongation made within one block for all treatments were based on plant height from the soil surface up to the top visible dewlap (third leaf). 2.9. Soil water balance measurements The water balance method based on neutron probe measurements was used to estimate crop water consumption, i.e. actual evapotranspiration ETa. The water balance equation written in terms of change in soil water storage AS is AS = S(t2)-S(t1) = I + P + C-ETa-D-R (7) where S(ti) denotes the soil water storage (m) in the control volume measured by means of a neutron probe at time t-t (d) for f 2 > h, I is the amount of irrigation water measured with a flowmeter (m), P the precipitation (m) recorded from the weather station located close to the field, C and D are capillary rise and drainage, respectively, at zz)=0-9 m estimated from Darcy's law and R the surface runoff (m) which is neglected here. Twenty plots of 10 treatments in 2 blocks were each equipped with a neutron access tube installed to a depth of 1.20 m. Each neutron measuring site was coupled with two tensiometers installed below the root zone at the vicinity of the depth where the drainage term was to be measured i.e. ZD = 0.9 m. The hydraulic gradients were obtained from tensiometer readings at depths 0.8 and 1.0 m. Beginning at the 0.15-cm depth, neutron readings were taken every 0.1 m within the soil profile. Depending on the rainfall pattern and irrigation schedule, neutron and tensiometer readings were made 2 or 3 times a week throughout the growing season. Water balance results were obtained by using the computer program "PROBE" [13]. 2.10. Irrigation efficiency estimate and yield response curves Two definitions of irrigation efficiency were used depending on either water applied to the field or water taken up by plants. Field water use efficiency Ef was based on the depth of irrigation water applied, and it reflected the characteristics of the adopted irrigation method. However, crop water use efficiency Ec in terms of yield produced per unit amount of water taken up showed the ability of plants to produce yields with a limited supply of water. Both ratios, namely Ef and Ec calculated with respect to the rainfed treatment [14] were £ / =(y / -F 0 ).^ 1 (8) and ~ (9) 119 where YÎ is the crop yield for any treatment other than that for the rainfed treatment, YO the crop yield for the rainfed treatment, £7^ the ETa for any treatment other than that for the rainfed treatment, ETa^ the ETa for the rainfed treatment and /,- the irrigation water applied on any treatment other than that for the rainfed treatment. For each growth stage studied, cane and sugar yields Y were regressed on crop water consumption ETa in order to obtain crop yield response functions ky. Relative yield reduction (1 - Ya • y^1) is plotted against relative evapotranspiration deficit (1 - ETa • ET^1) using (l-Ya-Y^) = ky(l-ETa-ETj) (10) where Ya is the yield under water deficit conditions, Ym the maximum yield obtained under full irrigation regime, ETa the ET under water deficit conditions and ETm the maximum ET related to the full irrigation treatment Values of Ay indicate sugarcane sensitivity to deficit irrigation. 3. RESULTS AND DISCUSSION 3.1. Climate As far as total rainfall is concerned, Fig. 1 shows an important decrease in 1993 (-320.4 mm) and a slight increase in 1994 (+90.7 mm) with respect to the 1973-94 average. Over the period of high values of crop water requirements (from April to June), total rainfall decreased by 83.3 and 159.1 mm respectively in 1993 and 1994 compared to the 1973-94 average. Also, the rainfall pattern is unimodal, August and September receiving the largest amounts. Lowest daily mean temperatures (23°C) were recorded in December and January, due to a dry and windy season, namely the harmattan. Otherwise, mean temperatures range from 25 to 28°C. AUU a RAINFALL (1993) o RAINFALL (1975-1994) ^300 ~~ 00 RAINFALL (1994) c — • ETp (1989-1994) o TEMPERATURE (1975-1994) S, M H 200 OH W -, 9 ^ R :i "-1 -< JÄ ,»—--S^. Cl V^*^ W 100 H ^ n a F M ~t \ ~ ftt \s "« A f \ \ •A s ~~~- S ?t v s \ ts7 s t t ', t ss // _. N ^ >':i s -f "•> ^ •^ s 3 1 x S x ts ' \ s s ^ ^ ''''. J A S _ 40 [D '»v N» ' s 60 w ^J y ^*~——_^_—• .f J- O ^ï ^^"^fï sf s — s ^> ?^ ' >s > ' ss > ' f *' O £ ü (3 S a 20 | W H m 5 M J S O N D MONTHS Fzg. 7. Climatic conditions during growing season. Note that air temperature values are plotted on a logritkmic scale. 120 J rJ3 fftl ^ M ^^^ 0f Q, 3.2. Soil characterization in the field experiment The sandy loam soil containing about 18% clay, 24% silt and 58% sand has a gravimetric water content at field capacity of 14.3% (ranging between a pF of 2.5 to 3.0) and a permanent wilting point of 5.7%. From the soil surface down to 1.2 m depth, the soil has an average bulk density of about 1.5 g-cnr3. Values of K0 and ß for the hydraulic conductivity function K(ff) for depths 0.5, 1 and 1.5 m found by regression with an r2 value of 0.86 were 1.48 and 184.8, respectively. This function was used for calculating the drainage term of the water balance equation at soil depth zD = 0.90m. Soil fertility analyses carried out during crop growing season showed that the soil was fairly acid and desaturated. 3.3. Water deficit effect on crop growth The depressive effect of water deficit on both cumulated cane stalk elongation and elongation rate was much more important at the boom stage than at tillering (Fig. 2 and 3), which shows how beneficial the practice of deficit irrigation is during tillering, as compared to the active growth period of sugarcane. Treatment (0,1) (no irrigation at tillering) and treatment (1,1) (full irrigation regime) had almost the same growth patterns. This was not the case for treatment (1,0) (no irrigation at the boom stage) which gave much smaller crop growth with respect to the full irrigated treatment 3.4. Analysis of variance There was no significant yield decline owing to water deficit practice at tillering (Tables I and ÏÏ). However, significant yield reduction owing to water stress at the boom stage was observed (Fig. 4). Hence, the sugarcane crop is less sensitive to water deficit imposed during tillering than during stem elongation. Also, no significant yield reduction was observed for treatment (1,0.75) (25% water deficit at stem elongation) compared to that of the full irrigated treatment (1,1). This comparison suggests the possibility of improving irrigation efficiency by practicing some mild water deficit during the stem elongation stage. Some significant differences of sugarcane juice quality were observed within treatments. For the first ratoon crop, irrigation treatment had a significant effect on fiber content. For the second ratoon crop, purity and extractable sugar percentages differed between irrigation treatments. In general, the second ratoon crop yields are lower than those of the first ratoon crop. The probable cause of these differences owing to crop age is the rainfall distribution after the irrigation season (i.e. from July to October during the Monsoon period), being more satisfactory hi 1994 than in 1993. 121 3.5. Water balance and irrigation efficiency Water balance results averaged from two replicates obtained over the entire growing season are given in Table ÏÏL Values of field irrigation efficiency and crop water use efficiency are shown in Table IV. Figure 5 shows that irrigation efficiency is lower in treatment (1,0) than in treatment (0,1). Averaging the two ratoon crops, treatment (1,0) yielded only 1.4 kg-nr3 of cane and 0.3 kg-nr3 of sugar while treatment (0,1) provided 11.9 kg-nr3 of cane and 1.5 kg-nr 3 of sugar. These yields are lower and higher, respectively, than those normally obtained (6.7 kg-nr3 of cane and 0.7 kg-nr3 of sugar) in field conditions of most tropical African countries [15] using optimal amounts of irrigation water required for sugarcane. Apparently, deficit • TREATMENT (1,1) TREATMENT (0,1) O TREATMENT (1,0) G TREATMENT (0,0) T3 • 5 2 - B ~ rs A O w o l 6 11 DATES OF MEASUREMENTS Fig. 2. Sugarcane growth measurements taken every 2 weeks from Jan 20 (datel) to Aug 18 (date 16) for different treatments for ratoon crop during 1993. 122 160 • + O D 1120 o TREATMENT (1,1) TREATMENT (0,1) TREATMENT (1,0) TREATMENT (0,0) S »80 40 - «3 0 5 9 DATES OF MEASUREMENTS Fig. 3. Sugarcane growth measurements taken every 2 weeks from Feb 21 (datel) to Jul 25 (date 12) for different treatments for raioon crop during 1994. irrigation practice at tillering is much more beneficial than at the boom stage of sugarcane. Hence, compared with deficit irrigation during the boom stage, irrigation can be omitted during tillering without any significant yield decline. The fact that the crop water use efficiency was almost the same for treatments (1,1) and (1,0.75), suggests that a 25 % water deficit irrigation with respect to treatment (1,1) imposed at the boom stage may be profitable in terms of cane or sugar yield per unit of water taken up by the crop. Yield response curves as a function of ETa obtained by using absolute yield and ETa data show that the slope of the straight line which representing crop water use efficiency is higher for the boom stage than for tillering. Hence, tillering is less sensitive to water deficit than the boom stage. As a result, the yield response factor to water [4] is smaller than unity for tillering and greater than unity for stem elongation (Fig. 6). 123 TABLE I. STATISTICAL ANALYSIS OF YIELD AND JUICE QUALITY FOR RATOON 1 Yields (t-ha-1) Juice quality (%) Treatment (1,1) (0,1) (0.25, 1) (0.5, 1) (0.75, 1) (1,0) (1,0.25) (1,0.5) (1,0.75) (0,0) Mean Treatment effect2 Block effect Standard deviation Coeff. Var. 1 Cane Sugar Sucrose 93.5a* 90.1a 87.0a 92.5a 90.6a 62.4c 75.8b 81.7ab 89.4a 58.4c 10.5a 10.4a 9.9ab 10.2ab 10.3ab 7.2de 8.5cd 9.0bc 10.2ab 6.7e 15.5 15.8 15.5 15.3 15.5 15.8 15.4 15.3 15.8 15.6 15.6 Purity . Fiber Extracted sugar 88.1 89.0 88.9 87.3 88.0 88.8 87.5 85.9 87.9 88.6 14.2 14.0 14.1 13.1 13.0 14.0 13.1 14.0 13.9 12.8 11.3 11.6 11.4 11.0 11.3 11.6 11.6 10.9 11.5 11.4 88.0 13.3 11.3 82.1 9.3 ** ** n.s. n.s. * n.s. n.s. n.s. n.s. n.s. ** n.s. 7.3 0.9 0.7 2.1 0.3 0.7 8.9 9.8 4.2 2.4 2.2 6.0 Mean values with the same letters in one column are not significantly different according to Duncan test 2 n.s. indicates not significant at 5% level, * indicates significant at 5% level and ** indicates significant at 1% level according to Fisher test 124 TABLE H. STATISTICAL ANALYSIS OF YIELDS AND JUICE QUALITY FOR RATOON 2 Yields (t-ha-1) Juice quality (%) Treatment Cane Sugar Sucrose Purity Fiber ll.Sabc 12.6a 11.5bc 11.3b 12.0abc ll.obc 12.2ab 12.5a 12.2ab 11.7e Extracted sugar. (1,1) (0,1) (0.25, 1) (0.5, 1) (0.75, 1) ( 1,0) (1,0.25) (1,0.5) (1,0.75) (0,0) 70.8a1 68.0ab 60. lab 62.9ab 55.3ab 48.6abc 56.4ab 54.9ab 66.3ab 32.0e 8.4a 8.5a 6.8ab 7.2ab 6.7ab 5.6bc 6.9ab 6.9ab 8.1a 3.7c 16.3 16.8 16.2 16.1 16.8 16.2 17.0 17.0 16.7 16.2 88.0abc 91.2a 85.5bc 84.4c 85.3bc 86.3bc 86.6b 89.1ab 88.3abc 86.3bc 14.2 14.0 14.1 13.1 13.0 14.0 13.1 14.0 13.9 12.8 Mean 57.5 6.9 16.5 87.1 13.6 11.9 ** ** n.s. * n.s ** n.s. n.s. n.s. n.s. n.s. n.s. 11.9 1.4 0.7 2.7 1.4 0.7 Coeff. Var. 20.8 20.5 4.3 3.1 10.6 6.0 Treatment effect2 Block effect Standard deviation 1 Mean values with the same letters in one column are not significantly different according to Duncan test 2 n.s. indicates not significant at 5% level, * indicates significant at 5% level and ** indicates significant at 1% level according to Fisher test 125 100 EZZ1 RATOON l i RATOON 2 /• ~ v ï 80 Ä 60 ä 40 l U 2 ° 0 /—x 12 Ï • d * "" 8 0 i (1,1) (0,1) (1,0) (0,0) TREATMENTS ig. 4. Water deficit effect on cane and sugar yields for the two ratoon crops. TABLE m. SOIL WATER BALANCE RESULTS Irrigation applied (mm) Treatment (1,1) (0,1) (0.25, 1) (0.5, 1) (0.75, 1) ( 1,0) (1,0.25) (1,0.5) (1,0.75) (0,0) Ratoon 1 576.9 432.1 468.2 504.1 540.6 144.8 252.6 360.8 468.8 0 ETa (mm) Ratoon 2 Ratoon 1 Ratoon 2 Ratoon 1 Ratoon 2 445.3 305.4 341.0 375.4 410.3 139.9 216.2 292.5 369.0 0 1236.7 1093.3 1129.3 1164.9 1201.1 888.6 991.9 1092.1 1178.8 743.8 1094.7 954.9 990.5 1024.9 1059.8 851.9 919.8 985.1 1046.2 712.0 123.3 121.9 122.0 122.3 122.7 39.2 43.7 51.7 73.0 39.2 542.2 542.1 542.1 542.1 542.1 479.5 487.8 498.9 514.4 479.5 Total rainfall for ratoon 1 was 783.1 mm ; for ratoon 2,1194.2 mm 126 Drainage (mm) 16 P771 RATOON l S5S RATOON 2 r Î8 =• 4 u -ZERO 0 J 0 -ZERO (1,1) (0,1) (1,0) (0,0) TREATMENTS Fig. 5. Crop water use efficiency Ec depending on cane and sugar yields for the two ratoon crops. TABLE IV. FIELD IRRIGATION EFFICIENCY Ef AND CROP WATER USE EFFICIENCY Ec FOR THE TWO RATOON CROPS Rl and R2 Ef Treatment (1,1) (0,1) (0.25, 1) (0.5, 1) (0.75, 1) (1,0) (1,0.25) (1,0.5) (1,0.75) 3 canef kg-nr ) Rl R2 6.1 7.3 6.1 6.8 5.9 2.8 6.9 6.4 6.6 8.7 11.8 8.2 8.2 5.7 0 11.3 7.8 9.3 Er 3 suear flce-nr ') Rl R2 0.6 0.8 0.7 0.7 0.7 0.3 0.7 0.6 0.7 1.0 1.6 0.9 0.9 0.7 0 1.5 1.1 1.2 3 cane fkg-m- ) R2 Rl 7.1 9.1 7.4 8.1 7.0 2.8 7.0 6.7 7.1 10.1 14.8 10.1 9.9 6.7 0 11.7 8.4 10.3 sugar Hee-nr3} Rl R2 0.8 1.0 0.8 0.8 0.8 0.3 0.7 0.7 0.8 1.2 2.0 1.1 1.1 0.9 0 1.5 1.2 1.3 EC are calculated with respect to the rainfed treatment (0,0). 127 l - ETa -ET m• i 0.8 1.0 0.6 0.4 0.2 0 0 0.2 SUGAR YIELD REDUCTION 0.4 0.6 • TILLERING STAGE O BOOM STAGE 0.8 1.0 CANE YIELD REDUCTION Fig. 6. Cane and sugar yield response factors kyfor two growth stages. 4. CONCLUSIONS It has been shown in this study that growth and yield decline of sugarcane crop owing to water deficit is significantly high during stem elongation as compared to tillering. As a result, the sugarcane crop was much more sensitive to water stress at stem elongation than at tillering. Therefore, deficit irrigation practices to improve crop water use efficiency may be recommended during tillering rather than stem elongation. Hence, the irrigation water management strategy suggested here consists of omitting irrigation at tillering as soon as the crop is successfully established, for the benefit of stem elongation. As far as stem elongation is concerned, a moderate water deficit of about 25 % with respect to the full irrigation regime appears to increase crop water use efficiency. REFERENCES [1] 128 IDESSA (INSTITUT DES SAVANES), Effet d'un stress hydrique à différents stades de développement de la canne à sucre, Rapport de campagne 1983-84, IDESSA, Bouaké (1985) 25-28. [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] MONTENY, B.A., ZELLER B., HAINNEUX G., Estimation des besoins en eau de la canne à sucre en zone soudano-sahélienne pour la conduite de l'irrigation, In: Crop water requirements (Proc. Int. Conf. Paris, 1984), INRA, Paris (1985) 79-102. QUIDEAU, P., PENE B.G., Variation de l'efficience de l'irrigation selon la saison et le stade de développement de la canne à sucre, In: 1ère rencontre internationale en langue franâaise sur la canne à sucre (Proc. Int. Conf. Montpellier, 1991), AFCAS, Montpellier (1991) 169-173. PAO (FOOD AND AGRICULTURE ORGANIZATION), Yield response to water, PAO Irrigation and Drainage paper No. 33 (revised 1979), FAO, Rome (1986) 193 pp. PENE, B.C., Sugarcane yield response to irrigation. Agron. Afr. 4 1 (1992) 35-45. PENE, B.C., Yield response of sugarcane (Saccharum officinarum L.) to water deficit at different growth stages as to improve irrigation efficiency, PhD Thesis (Water management), University of Abidjan, Côte d'Ivoire, Faculty of Sciences & Techniques (FAST), IDESSA, Bouaké (1994) 207 pp. PENE, B.C., EBOI P., MEL S., Yield of an early-season cane crop as influenced by deficit irrigation during the boom stage, Agron. Afr. 5 2 (1993) 85-98. HILLEL, D.A., KRENTOS V.K., STILIANOU Y., Procedure and test of an internal drainage method for measuring soil hydraulic characteristics in situ, Soil Sei. 114 (1972) 395-400. VACHAUD, G., VAUCLIN M., LATY R., CARHYD: A computer-aided system for characterization of the hydraulic conductivity of field soil, Soil Technol. 3 (1990) 1 SI144. VAUCLIN M, VACHAUD G., Caractérisation hydrodynamique des sols: analyse simplifiée des essais de drainage interne, Agron. 79 (1987) 647-655. LIBARDI, P.L., REICHARDT K, NIELSEN D.R., BIGGAR J.W., Simple field methods for estimating soil hydraulic conductivity, Soil Sei. Soc. Am. J. 44 (1980): 37. [12] [13] [14] [15] FAO (FOOD AND AGRICULTURE ORGANIZATION), Crop water requirements, FAO Irrigation and Drainage paper No. 24 (revised 1977), FAO, Rome (1992) 144 pp. CHOPART, J.L., SIBAND P., PROBE: Programme de büan en eau, Mém. & Trav. de TIRAT No. 17, CIRAD-IRAT, Paris (1988) 76 pp. BOS, M.G., Summary of ICDI definitions on irrigation efficiencies, In: Crop water requirements (Proc. InL Conf. Paris, 1984), INRA, Paris (1985) 899-910. LEMAIRE, Y., MAHEO R., DUGARD, G., Logiciel "Parcelle". Gestion du fichier de parcelles sur micro-informatique en exploitation de canne à sucre, In: 1ère rencontre internationale en langue franâaise sur la canne à sucre (Proc. Int. Conf. Montpellier, 1991), AFCAS, Montpellier (1991) 209-215. Next page(s) left blank 129 YIELD RESPONSE OF GROUNDNUT GROWN UNDER RAINFED AND IRRIGATED CONDITIONS A. AHMAD Malaysian Institute For Nuclear Technology Research (MINT), Bangi, Kajang Selangor, Malaysia Abstract A rainfed treatment and two irrigated treatments of groundnut were compared in terms of yield and water use. The rainfed treatment (Treatment A) was considered the control treatment. Treatment B consisted of irrigation at 7-day intervals taking into consideration rainfall. Treatment C was the other irrigation treatment consisting of irrigations made whenever readings of tensiometers at the 20-cm soil depth were equal to or less than -30 kPa. Results showed that there was no significant difference in yield between the two irrigation treatments. The average groundnut yields obtained from treatments A, B and C were 1.9, 3.1 and 3.2 t.ha-1, respectively. The crop in the rainfed plot was exposed to water stress during its flowering stage owing to limited rainfall in August. The total water use of groundnut for a 30-day period beginning 30 days after planting (DAP) was 64.5, 124.5 and 152 mm for rainfed, irrigated B and irrigated C, respectively. The low water use in the rainfed plot resulted in a low yield. As indicated by a continuously decreasing value of the hydraulic head value to < -70 kPa in the rainfed plot from 35 to 54 DAP, the soil at 20-cm depth was continuously dry. With the yield response factor ky for flowering stage being 0.74, the decrease in groundnut yield due to water deficit during flowering stage is relatively large. 1. INTRODUCTION With groundnut being a cash crop for many farmers, it is one of the important legume crops of Malaysia. However, the present-day total area of groundnut is small and estimated to be between 1200-1700 ha-y1. Hence, Malaysia has to import groundnut to cater for local consumption at a cost in excess of RM 37 million-y1 [1]. This import value is expected to increase if the groundnut industry in the country does not expand. A new high yield variety has been introduced to encourage groundnut production in the country. However, the full genetic potential of this new variety to obtain stable, high yields cannot be achieved if environmental conditions are not favorable for crop growth. Hence, groundnut yields are sometimes variable from season to season. Groundnut requires from 500 to 700 mm of water during its growing period [2], Although the total amount of rain is more than sufficient for rainfed agriculture, a 10- to 20-day drought often occurs during the cropping season owing to the duration and distribution of rainfall not being uniform [3]. Moreover, because Malaysian soils typically have a low capacity for retaining water which causes rapid drying of the soil surface during the dry season, an internal water deficit is induced in shallow rooting annual crops like groundnut which affect their growth. Therefore, rainfall patterns within each cropping season may account for variations in yield obtained from year to year. Water deficits affect groundnut growth depending on the stage of crop growth and the intensity of the drought stress causing large decrease in yield [4,5,6]. It is therefore thought that 131 research on water-management technology should include studies of soil-plant-water relations of groundnut. Furthermore, supplementary irrigation is now being practiced by farmers in the country. Hence information on how much water is used by the crop in relation to yield is very important The objective of this study was to determine the yield response and water use of groundnut to irrigation treatments compared with rainfed treatment 2. MATERIALS AND METHODS A field plot experiment was conducted on a Serdang colluvial soil, located in the Research area of Field 10B of the Agricultural University, Serdang. The area is almost flat and the soil is well drained at the surface to imperfectly drained below one meter. The particle size analysis of the soil shown in Table I [7] reveals that there is an increase in clay content with depth and the texture varies from sandy loam to clay loam. The experimental area was divided into three 18 x 30 m plots for rainfed and irrigated treatments. Seeds of a new groundnut variety MKT1 were sown at a spacing of 45 x 30 cm. The rates of fertilizer application (kg-ha4) were 34:60:60 of N:P2Os:K2O. The fertilizer was applied in a band between the crop rows. During a 2-week period following sowing, rainfall was sufficient for seedling establishment Two replicates of tensiometers and aluminum neutron meter access tubes were used in each plot Mercury tensiometers were placed at soil depths of 20,40,60, 80 and 100 cm. One 1m long aluminum access tube was installed at each location in all plots. A plot for calibrating the TABLE I. SOIL PARTICLE SIZE ANALYSIS OF THE AREA IN FIELD 10B Soil depth (cm) 10 20 30 40 50 60 70 80 90 100 132 % clay % silt % fine sand (<0.002 mm) (0.002-0.02 mm) (0.02-0.2 mm) 19.9 21.0 27.0 30.0 29.4 29.4 29.2 29.5 27.4 31.4 16.1 16.6 16.8 16.6 18.2 15.8 18.8 18.4 17.9 19.7 41.9 40.7 36.3 32.9 31.3 32.3 31.4 34.2 33.8 28.6 % coarse sand (0.2-2 mm) 22.1 21.7 19.9 20.5 21.1 22.5 20.6 17.9 20.9 20.3 Textural class sandy loam sandy clay loam sandy clay loam sandy clay loam sandy clay loam sandy clay loam sandy clay loam sandy clay loam sandy clay loam clay loam neutron meter was established next to the experimental site. Three treatments were applied : A was rainfed as a control, B was an irrigated treatment based on 7-day intervals and C was an irrigated treatment based on the 20 cm depth tensiometer reading -30 kPa or less. The amount of irrigation water applied for treatment B was the total atmospheric evapotranspiration demand for 7 days taking into consideration the cumulative rain during that period. For treatment C, the amount applied returned the soil water content to field capacity in the root zone. Irrigation water was applied using microsprinklers fitted with flow meters for both B and C treatments. The water content in the 100-cm soil profile of the irrigated plots was measured before and after each irrigation using a Troxler neutron probe model 4300. Prior to flowering stage, the hydraulic head of the soil water was measured with the tensiometers twice a week. After flowering, daily measurements of hydraulic head were taken for a period of at least 10-15 d. Class A pan evaporation, rainfall and wind speed were recorded daily. A water balance approach was used to evaluate crop evapotranspiration/crop water use for each treatment From daily values of weather data collected at a nearby weather station, the Pan Evaporation Method [8] was used to estimate ET0 (grass reference ET). Crop coefficient Kc values for the initial, mid-season and late season stages were obtained [8]. At maturity the crop was harvested with dry pod weight yields being measured from plants harvested from an area of 16 m2 from each plot 3. RESULTS AND DISCUSSION Figure 1 shows the yield response of groundnut to irrigation applied during flowering stage. The yields obtained with irrigation treatment B and C were 3.1 and 3.2 t-ha*1, respectively. A total application of 50.7 mm of water for treatment B with water applied during the flowering stage did not increase the yield over that for treatment C with an application of only 38.8 mm of water. However, the yields obtained from both irrigated treatments were nearly double the 1.92 t-ha-1 yield attained in the rainfed plot. These data indicate that groundnut variety MKT1 is responsive to irrigation. Irrigations (Treatments B and C) at flowering stage significantly affected the number of pods-nr2, number of seeds-nr2 and seed weight (Table H). As compared with the rainfed condition, irrigation treatment B and C increased numbers of pods-nr2 by 33% and 44%, respectively. Number of seeds-nr2 was highest with irrigation treatment C (500 seeds-nr2) but not significantly different from treatment B (475 seeds-nr2). However the number of seeds was lower (325 seeds-nr2) in the rainfed plot There were also increases in seed weight from 30.9 g in the rainfed plot to 42 g and 37 g in treatment B and C, respectively. Crop water use/crop evapotranspiration estimates and precipitation values for every 10 days during the growing period of groundnut is presented in Figure 2. The crop received a total amount of rain equal to 159 mm, 311 mm and 244 mm during the initial (July), mid-season (September) and late season (October) stages, respectively. In general, rainfall exceeded crop 133 Q -J 0 RAINFED IRRIGATED 3S.8 mm IRRIGATED 50.7mm Fig. L Yield response of groundnut to irrigation applied during flowering stage. TABLE ÏÏ. YIELD COMPONENTS OF GROUNDNUT AT FINAL HARVEST Number of pods-rrr2 Number of seeds-nr2 Weight of 100 seeds (g) Treatment 225 (±0.51)t 300 (±0.70) 325 (±0.73) RainfedA Irrigated B Irrigated C 325 (±0.80) 475 (±1.21) 500 (±1.09) 30.9 (±0.70) 42.0 (±1.05) 37.0 (±0.56) t Standard errors of the mean are shown in brackets. 200 T l • PRECIPITATION O EVAPOTRANSPIRATION g S JULY AUGUST SEPTEMBER OCTOBER FIG. 2. Precipitation and evapotranspiration profiles during the growing period. 134 evapotranspiration. However, the month of August was relatively dry with only 38 mm of rainfall. Hence, during this limited rainfall period which coincided with flowering stage (identified from the crop coefficient curve), evapotranspiration demand exceeded rainfall. Rainfed crops at this stage were exposed to water stress. Irrigation water was applied during this period to the irrigated plots, the total application being 50.7 mm and 38.8 mm for treatments B and C respectively. A plane of zero flux existed at 30 to 80 cm soil depth in all plots from 30 to 54 days after planting (DAP). Hydraulic head profiles for different values of DAP for the rainfed plot are shown in Figure 3. Crop water use/evapotranspiration was calculated using a water balance approach with the presence of the zero flux plane. Irrigation treatment C gave the highest evapotranspiration of 152 mm whereas that for treatment B was 124.5 mm for a period of 30 days after 30 DAP. The evapotranspiration was low (64.5 mm) in the rainfed plot Clearly the low water use of groundnut at flowering stage under rainfed condition resulted in low yield as shown in Figure 1. -80 HYDRAULIC HEAD (kPa) -60 -40 -20 100 FIG. 3. Hydraulic head profiles within the soil for different days after planting (DAP) in the rainfed plot Figure 4 shows hydraulic head versus time from 30 DAP to 60 DAP for all three treatments. Although irrigation water was applied at 22 (treatments B and C) and 25 DAP (treatment B), data on soil water status was not available during that period. There was 22 mm of rain from 26 to 28 DAP resulting in low values of hydraulic head (< -10 kPa) in all treatments at soil depths of 20 and 40 cm. The hydraulic head remained between -6 kPa to -16 kPa at the 20cm soil depth until 35 DAP in both irrigated B and C plots. Observations of hydraulic head at 20 cm depth as a function of time reveal that irrigation relieved the soil from being continuously dry. The hydraulic head at the 40-cm soil depth of both irrigated plots did not change markedly throughout the season. In contrast, the soil at 20 cm depth hi the rainfed plot becomes much 135 t——— IRRIGATIONS ————N RAIN H 0 0-25 w + t - U 2-50 -75 - -100 25 — — — —— — -¥• — A 20 cm B 20cm C20cm A 40 cm B 40 cm C40cm 35 45 DAYS AFTER PLANTING 55 Fig. 4. Hydraulic head at 20- and40-cm soil depths for different days after planting (DAP) for rainfed Treatment A, and irrigation Treatments B and C. drier before each rain as indicated by values of the hydraulic head eventually reaching -70 kPa. And at 40-cm soil depth in the rainfed plot the value of the hydraulic head gradually decreased from -9 to -74 kPa at 54 DAP. Hence, the rainfed crop with its roots densely located close to the soil surface was exposed to water stress. The water deficit developed in the plant to a point where crop yield was affected as shown by the low yield for the rainfed plot (Figure 1). Although high precipitation from 54 DAP until maturity (Oct.) relieved the water stress of the plants in the rainfed plot during that period, it was not effective in increasing the final yield when compared with those of the irrigated plots. The relationship between crop yield and crop water supply can be determined when crop water requirements (maximum evapotranspiration ETm) and crop water deficits (actual evapotranspiration £Ta) on the one hand, and maximum Ym and actual crop yield Ya on the other can be quantified [9]. Therefore, the response of yield to water supply is quantified through the yield response factor ky relating relative yield decrease (1 -7a-rm--0 to relative evapotranspiration deficit (1 -ETa-ETfn'1). The maximum yield Ym and evapotranspiration ETm levels in this experiment were most likely from the B and C irrigation treatments. The average maximum yield of 3.14 t-ha-1 and evapotranspiration of 138 mm were used from these treatments to compute the response of yield to water supply. The generalized effect of water deficit on yield for the flowering stage in the experiment is shown in Figure 5. The value of the yield response factor ky is 0.74, comparable with 0.8 reported for groundnut at the flowering stage [9]. The value indicates that the decrease in yield owing to water deficit during flowering stage is relatively large, i.e. a 40% deficit in relative evapotranspiration will cause a 50% decrease in yield. 136 l - ETa'ETm'1 Fig. 5. Relationship between relative yield decrease (1 deficit (1 -ETa-ETm1) for groundnut at flowering stage. and relative evapotranspiration Nageswara Rao et al. [4] has reported that pod yield decreased by 61% when 44% less water than the control (continuous irrigation) was applied from the start of flowering to the start of seed growth. 4. CONCLUSION The results of this investigation indicated that groundnut responds to irrigation applied at flowering stage during a dry period to produce a high yield. The total crop water requirement at flowering stage is between 120-130 mm. A water deficit during flowering stage reduced the pod yield of groundnut by 50% with a 40% deficit in evapotranspiration. Therefore, for groundnut grown under supplemental irrigation the water application must be programmed so that sufficient water is particularly available in the soil during the flowering stage. Alternatively, with appropriate meteorological records, sowing time could be adjusted so that the flowering stage did not coincide with a normally dry period. ACKNOWLEDGEMENTS The author wishes to thank Prof. Cevat Kirda for his help and guidance on data analysis and interpretation of the results. Thanks are also given to IAEA, Vienna for financially supporting this project through Research Contract MAL 6786/R1/RB and Ministry of Science and Technology, Malaysia through grant IRPA 4-06-05-014. 137 REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] 138 MARDI, Kacang Tanah Bara Varied MKT11. ISMAIL A.B., Laporan Khas Penanaman Kacang Tanah, Ed. Zaharah, H., MARDI (1991)11-13. NIEUWOLT, S., GHAZALLI, M. Z., GONIPATHAN, B., Agro-ecological regions in Peninsular Malaysia. MARDI, Serdang, Selangor (1982). NAGESWARA RAO, R. C., SINGH, S., SIVAKUMAR, M. V. K., SRIVASTAVA, K. L., WILLIAMS, J. H., Agron. J. 77 (1985) 782-786. SINGH, B., SANDHU, B. S., KHERA, K. L., AUJLA, T. S., Field Crops Research. 13 (1986) 355-366. STIRLING, C. M., ONG, C. K., BLACK, C. R., J., Experimental Bot. 40 (1989) 11451153. MAENE, L. M., MAESSCHALCK, G. G., LIM, K. EL, MOKHTARUDDIN, A. M., Sou Physics Project, Faculty of Agriculture. Annual Report Oct 1977 - Dec 1978, Universiti Pertanian Malaysia, 108-148. FAO, Crop Water Requirements. FAO Irrigation and Drainage Paper No. 24, Rome (1992). FAO, Yield Response to Water. FAO Irrigation and Drainage Paper No. 33, Rome (1986). CONTRIBUTION TO TEE IMPROVEMENT OF SUGARBEET DEMCIT-IRRIGAITON M. BAZZA, M. TAYAA Institut Agronomique et Vétérinaire Hassan II, Département de V Equipement et de V Hydraulique, Rabat-Instituts, Morocco Abstract A field study was conducted in the Doukkala region of Morocco to assess the feasibility of scheduling irrigation of sugar beet on the basis of soil water depletion and potential. It consisted of six treatments each with three repetitions with the following threshold values for applying water: 40, 60 and 80 mm-nr1 of soil water depletion in the root depth and 20,40 and 60 cbar of soil water suction. Water consumptive use was determined using the in situ water balance method using a neutron probe and estimated using Cropwat Model [1]. As the growing season was relatively dry, the amounts of water applied to the different treatments as well as the crop water consumptive use varied enormously between treatments. As a result both root and sugar yields and water use efficiency were also different The latter was highest under the stressed treatments and lowest when water was not limiting; moreover, as it varied in opposite direction with the yield, it could not allow for the best treatments to be identified. An economic analysis taking into account the returns from the harvest and the expenses associated with water showed that the highest profit corresponds to the treatments irrigated at 40 cbar of soil water potential or 60 mm-nv1 of soil water depletion. The performance of these treatments in terms of profit was higher or at least similar to that of the treatments which received larger amounts of water despite the low price of water. It was concluded that although the implementation of these and similar techniques can be achieved when water is available on demand, it still necessitates further research programs to identify scenarios of providing water when necessary in large irrigation projects. The collaboration of fanners, researchers, extension agents and water management agencies is also necessary to ensure success from such programs. 1. INTRODUCTION In most developing countries, agriculture is still the main source of income and employment, but it is now being hindered by several issues. Presently, water shortage is the most important problem facing agricultural production, especially in arid and semi-arid regions. Solutions are urgently needed to stop the continuous decline in water supply and quality and sustain production and to improve the efficient use of the available water resources. Water losses under irrigation are estimated to reach as much as 60% of the applied amounts. The low efficiency is the result of (1) inappropriate practices at the farm level, (2) poor management by agencies in charge of water distribution, (3) policies related directly or indirectly to water allocation, pricing and use, and (4) distribution networks. The efficiency could be drastically improved through simple actions consisting of applying research results proven to be sound and reliable. 139 In many irrigated projects worldwide, water is still being allocated on the basis of demand by farmers or a fixed schedule with the amounts delivered or applied following no rules except the experience of farmers which often leads to inadequate timing, excessive deficit prior to irrigation and important water losses during and following application. As a result, although the total applied amount during the growing season is near or even exceeds the requirements, only a small fraction benefits the crop. Hence, crop yield remains low compared to its potential. The present work was conducted in one of the major irrigated projects of Morocco and had the following objectives: -assessing the possibility of irrigation scheduling based on soil water content depletion -assessing the possibility of irrigation scheduling based on soil water potential -assessing the benefits from such practices compared to common practice -validating of the Penman-Monteith formula for estimating reference evapotranspiration ET0 and water requirements of sugar beet 2. MATERIALS AND METHODS The trial was conducted in a 0.5-ha field plot in the Doukkala irrigated Perimeter of Morocco. The initial fertilization consisted of the equivalent of 700 kg-ha*1 of a 8.75-10-30 fertilizer prior to the installation of the crop. Sowing of a genotype of the N type took place manually on the 8th of December, with a spacing of 25 and 50 cm along and between the rows, respectively. Two days later, a light irrigation enhanced germination. Crop emergence and thinning took place 3 weeks and 47 d after installation, respectively. After a growing season of 205 d, the crop was harvested on the 6th of July. For irrigation scheduling based on soil water potential, three treatments were defined. Irrigation was to be applied when the soil water potential reached 20 cbar for treatment PI, 40 cbar for P2 and 60 cbar for P3. Tensiometers were installed at the depths of 15 and 45 cm in each treatment. During the first three months of the growing season, irrigation was applied when the threshold value of soil water potential was reached at the shallow depth (15 cm) with the deeper tensiometer serving as an indicator of whether any excess water was applied. Later on in the growing season, when the roots had extended downward, irrigation was based on the same threshold values of water potential but at the greater depth of 45 cm. Similarly, three treatments were defined for irrigation scheduling based on soil water depletion. Irrigation was to be applied when the depleted depth of water per meter of root depth reaches 40 mm for treatment Dl, 60 mm for D2 and 80 mm for D3. Each treatment consisted of nine rows of sugarbeet out of which only the central five were considered (two rows on each side were left as borders). Inasmuch as the rows were 40 m long and 50 cm apart, each treatment corresponded to an area of 180 m2. To account for soil variability, three replicates were used for all treatments and the six treatments were randomly distributed within each replicate. 140 Except for the amounts of applied water which varied according to the experimental setup, all production techniques were the same for all treatments and were those recommended for maximum production in the region. Inasmuch as furrow irrigation is normally used in the region, it was also used in the present investigation. With the irrigation water first lifted into a large calibrated reservoir from which it flowed under gravity, it was possible to accurately measure the amount supplied to each treatment during each irrigation. Soil water content was monitored with a neutron moisture probe. A neutron access tube was installed in the central row of each treatment. Two gauge type tensiometers were also installed in the same row about 30 cm away from each access tube. Measurements were taken three to four times a week. Water potential was read directly from the tensiometer gauges, whereas measurements with the neutron probe were replicated twice every 10 cm between the soil surface and 80 cm deep. All measurements were performed between 7 and 8 am. The soil physical and chemical characteristics were determined in situ or by taking soil samples for analysis to the laboratory. The neutron probe was calibrated in situ. The maximum evapotranspiration ETm was estimated using the Penman-Monteith formula and the crop coefficient Kc determined in the region during previous studies [2]. The meteorological data required by this method was measured in a weather station located in the region. The actual crop water use ETa was determined using thé in situ water balance method within the root zone, neglecting the drainage and capillary rise components. Although the neutron probe had been calibrated in situ, owing to the lack of accuracy of the gauge type tensiometers, the drainage and capillary rise components could not be assessed. To overcome this limitation, the Cropwat model [1] was used to estimate the amount of water lost to drainage in each treatment. The latter was run using the soil and crop data and the dates and amounts of water applied to the different treatments. The model is based on the water balance method and Penman formula for estimating crop water requirements. The latter has been found to be valid for sugarbeet in the region [2]. The Cropwat model was run under the following conditions: - Crop stages: an initial phase of 47 d, a development stage of 60 d, a mid-season phase of 74 d and a late season phase of 24 d - Rooting depth: variable from 20 cm in December to 60 cm starting in May - Crop coefficients: 0.5 for the initial phase, 1.05 for the mid-season and 0.98 for the late season - Allowable depletion: 50 % - Total available soil moisture: 150 mm-nr1 - Initial soil moisture depletion: 100%. At complete maturity, the sugar beet root yield was determined by harvesting and weighing three 6.25-m2 sub-plots in each treatment These sub-plots were located respectively, 4 m from the upstream end, hi the middle and l m from the downstream end of the rows. A sample of 12 roots taken from each sub-plot served for determining the sugar content and the technological quality. 141 3. RESULTS AND DISCUSSION The crop density measured at harvest was 88,310 plants-ha*1. The soil physical and chemical characteristics are given in Table I. The soil is a sandy loam with a mean water holding capacity of 150 mm-nr1. The in-situ calibration of the neutron probe yielded the following equations for soil depths of (0 - 20 cm), (20 - 40 cm), (40 - 60 cm) and (60 - 80 cm), respectively 6 = -8.57 + 19.60AT,- • AC"1 r2 = 0.94 6 = -5.87 +18.54Ni • N~l r2 = 0.87 0 = -15.20 + 24.13AT; • A^1 r2 = 0.92 0 = -13.05 + 21.5UVrA71 r 2 =0.9l where the volumetric soil water content d is given in %, NI the mean count number from the neutron probe and Ns the standard count number. TABLE I. SOIL PHYSICAL AND CHEMICAL CHARACTERISTICS AS DETERMINED IN-SITU AND IN THE LABORATORYt ________Soil depth (cm)____________ Property 0-15 Fine sand (%) Coarse sand (%) Fine loam (%) Coarse loam (%) Clay (%) 12.4 41.4 38.4 6.3 1.5 12.1 41.3 38.0 6.1 2.5 9.3 30.2 5.5 2.4 52.7 9.4 30.4 3.9 4.6 51.5 Humidity at FC (%) Humidity at PWP(%) 27.3 12.3 27.4 12.4 27.6 12.2 28.1 12.5 7.76 1.65 54.4 225.5 2.2 0.8 7.54 1.60 17.8 189.5 2.1. 0.5 7.4 1.63 13.1 159.6 2.3 0.1 pH Bulk density (g-cnr3) Available P (mg-kg'1) Available K (mg-kg-1) ECeOrnnhos-cnr1) O. M. content (%) 7.6 1.58 74.4 285.4 2.4 0.8 15-35 t The soil samples analyzed were taken after the addition of fertilizers. 142 35-50 50-70 Total precipitation during the year amounted to 304 mm, out of which over 53% (164 mm) occurred during October and November and only 140 mm coincided with the growing season. During the period of high water requirements (March through June), total rainfall was only 22 mm. Temperature values were generally favorable for sugarbeet Mean temperature per decade ranged from 9.5 °C in late February to 25 °C in late June. The mean maximum temperature was 32 °C and that of the minimum varied between 1.5 °C during the second decade of January to 16.3 °C during the last decade of June. Mean relative humidity ranged from 66 to 91% while the minimum fluctuated between 20 and 60%. Wind speed was generally moderate with a mean of the maximum values of 2.4 m-s-1 late in the growing season. Reference evapotranspiration ET0, estimated by the Penman-Monteith method totaled 709 mm with the minimum in December (1.5 mm-d'1) and the maximum in May (5.2 mm-d'1). The latter coincided with the end of the mid-season and the beginning of the maturation stage. Applying the crop coefficient values of sugarbeet found locally over the previous four years to this reference evapotranspiration yields the maximum water requirements of the crop ETm. The latter amounted to 633 mm for the entire growing season. The values ofETm estimated by the Penman-Monteith method and the measured values of ETa in the treatment conducted with no water stress were very close throughout the measurement period (February 10 to early July). The correlation coefficient between these two variables was 0.97 which proves the validity of the Penman-Monteith method for estimating the water requirements of sugarbeet within the context of the region. During the entire growing season, the total water consumptive use estimated by the Cropwat model was 585,509,450,599,478 and 422 mm, respectively, for treatments PI, P2, P3, Dl, D2 and D3. The maximum corresponds to the treatments conducted under no water stress (irrigated when soil water depletion reached 40 mm-nr1 of root depth or when soil water potential reached 20 cbar) and is very close to total ETm (92% of ETm for PI and 95% for Dl). The water content under these two treatments remained stable near field capacity throughout the growing season, thus their water consumptive use ETa can be considered as equivalent to the crop ETm. These two treatments received the highest number of water applications (13 and 11, respectively). Their irrigation frequency turned out to be very close to that recommended initially by the project designer, i.e. once every 2 weeks during the normal period, and once every 10 d during peak water requirements. Treatments P2 and P3 which were irrigated at the threshold values of 40 and 60 cbar, respectively, received about 50% less water applications than treatment PI and were subjected to some water stress (20% for P2 and 29% for P3). During the second phase of the growing season, water application to treatment P3 was limited to 68 mm of precipitation while ETm amounted to 167.5 mm. As a result, this treatment was subjected to a water deficit equivalent to almost 60% of the water requirements of this period. Similarly, during the late stage of the growing season, ETa of treatment P2 was only about 60 mm compared to ETm which exceeded 143 110 mm (deficit of over 46%). The total water deficit of treatments D2 and D3 was 155 and 210 mm, respectively. For treatment D2, ETa during the late stage of the crop was 51 mm, whereas the corresponding ETm was 106 mm. Hence, the crop was subjected to a water stress of 52% of the requirements during this stage. Similarly, for treatment D3, the deficit was 55%. During this period, ETa of these treatments was by far larger than the applied amount (irrigation and precipitation) which means that the crop water use consisted essentially of water which was stored in the soil. Table n shows for all six treatments the depths of water applied through irrigation and precipitation, the crop water consumptive use ETa, the water deficit expressed as a percent of ETa to total water requirements ETm, and the efficiency of the applied amounts of water to satisfy the crop water use. It can be noted that the lowest efficiency values were associated with the high irrigation frequency (PI and Dl). The amounts of water lost under these treatments which had their water requirements almost fully satisfied varied between 20% for Dl and 25 % for PI. The actual loss is even higher than these figures because part of the water use by the crop was drafted from the water which was initially stored in the soil (the water content in the soil was much lower at the end of the growing season than during the crop installation.) TABLE II. TOTAL DEPTH OF WATER APPLIED, ITS OVERALL EFFICIENCY AND ESTIMATED WATER CONSUMPTIVE USE AND WATER DEFICIT OF THE DIFFERENT TREATMENTS Parameter PI Depth (mm) Irrigations £Ta(mm) Deficit (%) Efficiency (%) 775 13 585 7.6 75.5 P2 593 8 509 19.6 85.8 Treatments Dl P3 467 7 450 28.9 96.4 748 11 599 5.4 80.1 D2 D3 517 7 478 24.5 92.5 423 5 422 33.3 99.8 Root yield for the different treatments responded very positively to the amount of water applied through irrigation. Yields varied from a minimum of 60 t-ha'1 for treatment D3 irrigated at the threshold value of water depletion of 80 mm-m-1 to a maximum of 92 t-ha*1 for treatment PI irrigated whenever soil water potential reached 20 cbar. The other treatments resulted in a root yield of 73 for P3 (60 cbar), 74 for D2 (60 mm-m-1), 84 for P2 (40 cbar) and 88 t-ha-1 for Dl (40 mm-m-1). From the standpoint of absolute terms, the latter resulted in a yield 4 t-ha*1 lower than that of PI, although its water consumptive use was slightly higher by 14 mm. This result confirms previous findings regarding the fact that the sugar beet maximum yield seems to 144 be obtained under conditions of a slight water deficit. Although both treatments were subjected to very limited water deficit at the end of the growing season (prior to harvest), this water shortage was slightly higher under treatment PL Statistically speaking however, the yield of these two treatments are not significantly different and hence, their mean yield (90 t-ha-1) can be considered as the potential yield under the conditions of the experiment. These two treatments received the highest number of water applications (13 for PI and 11 for P2), but resulted also in the highest amounts of water lost to drainage and not used by the crop. Moreover, with two irrigations more than applied in Dl, treatment PI resulted in practically the same yield. Hence, with the extra water applications having no positive effect on yield, the lowest efficiency of the applied water was obtained under this treatment (75 %). Treatment D3 which was irrigated when water depletion in the root zone reached 80 mm-nr1 resulted in the highest efficiency of the applied water. Its water consumptive use was equivalent to the applied depth. The amount lost to drainage under the conditions of this treatment were fully compensated by the water uptake from the soil (depletion between the initial water status at the crop installation and that at harvest time) and capillary rise from soil layers below the root zone. The second best efficiency (96%) was obtained under the treatment which was irrigated at the threshold value of soil water potential of 60 cbar. The other treatments had intermediate efficiency values (92% for D2 and 86% for P2). Thus, considering all treatments, the applied water efficiency is inversely proportional to both the root yield and the amount of water applied. It should be noted that irrigation was withheld two weeks before harvest for all treatments, irrespective of when the last irrigation was applied. As a result, the treatments which were irrigated less frequently received no water application during periods as long as 40 d before harvest This practice must have been beneficial for the treatments which were conducted under no water stress, but harmful for those conducted under water deficit conditions. Previous investigations have found that under water stress conditions during the growing season, it is better to continue irrigating until one week or less prior to harvest in order to allow the crop to partially make up for the effect of such a stress. However, under very limited or no water stress during the growing season, it is better to stop water application 2 or 3 weeks prior to harvest so that some dehydration of roots occurs and sugar content increases [3]. Being a very important sugar beet parameter, the sugar content of the roots behaved differently than the yield. Its highest value was obtained under treatments D3 (19.4%), P2 (19.3%) and D2 (19.5%). These treatments were subjected to severe water stress prior to harvest in addition to that during the growing season. On the other extreme, treatments PI and Dl which received very limited stress at the end of the growing season resulted in the lowest sugar content with 16.5 and 16.6%, respectively. Treatment P3 resulted in an intermediate value of 18.4%. Thus, in general terms, the higher the amount of water applied, the lower is the sugar content. The period of no irrigation prior to harvest also has a great effect on sugar content Based on root yield, quality and sugar content, the extractable sugar yield was determined. The latter was highest under treatments P2 (13.67 t-ha4), PI (13.61 t-ha-1) and Dl 145 (13.27 t-ha-1), while the minimum was obtained under treatment D3 with 9.71 t-ha-1. Treatments D2 and P3 resulted in intermediate yields with 12.05 and 11. 13 t-ha'1, respectively. Thus, extractable sugar yield followed the root yield to some extent. However, the maximum in absolute terms corresponds to treatment P2 (irrigated at 40 cbar), although this yield is not significantly different from those of PI and Dl which resulted in the highest root yield. The relatively lower root yield of treatment P2 (84 tha'1) was compensated by its high sugar content The yield per unit of applied water or irrigation water use efficiency was also inversely proportional to the amount applied (Table lu). Irrigation water use efficiency is highest with treatments D3 and P3 which received the least amounts of water, while its lowest values correspond to treatments PI and Di. Treatments P2 and D2 had intermediate efficiency values with respect to both root and sugar yields. As such, this criterion (efficiency) does not allow the identification of the treatment which optimizes the yield inasmuch as it would be infinitely large under rainfed conditions although the yield would not be economically profitable. Such identification requires that an economic analysis is performed for comparison between treatments. TABLE IE. IRRIGATION WATER USE EFFICIENCY WITH RESPECT TO ROOT AND SUGAR YIELD Irrigation Number Treatment PI P2 P3 Dl D2 D3 13 8 7 11 7 5 Volume (m3-ha-1) 6350 4530 3270 6080 3770 2830 Yield Roots (t-ha-1) 92 84 73 88 74 60 Sugar (t-ha-1) 13.61 13.67 11.13 13.27 12.05 9.71 Yield resoonse Root mass Sugar mass (kg-m-3) (kg-m-3) 14.49 18.54 22.32 14.47 19.63 21.20 2! 14 3.02 3.40 2.18 3.20 3.43 Table IV synthesizes the results of an economic analysis of all treatments, taking into account the returns which a farmer would have made by selling his crop and the expenses associated with water (price of irrigation water and labor). The difference between the two does not correspond to the net profit because it still includes the cost of all other inputs (seeds, labor, fertilizers, pest control products, etc.). However, these costs are the same for all treatments and hence, the difference allows a relative classification of the treatments with respect to profit optimization. In Table TV the cost of water is that practiced by the Governmental Agency in charge of water allocation and management in the irrigated project and corresponds to 0.19 146 TABLE IV. COMPARISON BETWEEN MONETARY RETURNS FROM SUGAR BEET ROOTS AND EXPENSES ASSOCIATED WITH IRRIGATION WATER Treatment PI P2 P3 Dl D2 D3 ____________Monetary resource parameter_____________ At B C D E F 313.7 377.9 357.7 316.0 381.6 380.9 1206.5 860.7 621.3 1155.2 716.3 537.7 455 280 245 385 245 175 28860.4 31743.6 26112.1 27808.0 28238.4 22854.0 1661.5 1140.7 866.3 1540.2 961.3 712.7 27198.9 30602.9 25245.8 26267.8 27277.1 22141.3 t A = price of one ton of sugar beet roots, B = cost of irrigation water, C = cost of irrigation labor, D = total monetary returns from sugar beets, E = cost associated with water (B + C) and F = returns minus costs associated with water (D - E). Moroccan Dirhams (MAD) per cubic meter [as of 1994, 1 US $ was equivalent to 9 MAD]. The labor cost for irrigation is 35 MAD-d'1 and it is assumed that it takes 1 d for irrigating 1 ha of sugar beets. The gross return is calculated on the basis of what is practiced by the factory and takes into account both root yield and sugar content: where p(t) is the price (MAD) paid by the factory for .1 1 of roots, t the sugar content (%), p(t0) the reference price of 313 MAD-f1 of sugarbeets with a sugar content t0 = 16.5 %. Based on the water price practiced in the region, the highest profit making treatment is P2 which received 453 mm of irrigation water (8 irrigations) in addition to precipitation. This treatment was irrigated on the basis of soil water potential (40 cbar). This result confirms those from several previous studies which have shown that sugar beet should be irrigated at the threshold value of soil water potential of 40 to 45 cbar if yield is not to be affected by water shortage [4]. The second best profit-making treatment was D2 which was irrigated whenever soil water depletion from the root zone reached 60 mm-nr1. It received a total of 377 mm of irrigation water partitioned into seven applications. Yet, its performance in terms of monetary returns is practically the same as that of treatment PI which received the highest number of water applications (13) totalling 635 mm. Thus, from the economic viewpoint, and despite the price of water which is still considerably low, it is more beneficial for the farmer not to satisfy the total 147 water requirements of the crop. Based on these results, the following yield-water consumptive use relationships were developed: 7roof=44.8 + 0.5747 r = 0.94 where FTOO, is yield of roots (t-ha-1) and I the depth of applied irrigation (mm) Yroot=1.0 + Q.153ETa r = 0.93 where ETa is the actual evapotranspiration (mm) ? sugar = 8. 1 1 + 0. 009247 r = 0. 85 where Ysugar is extractable sugar yield (t-ha4) Ysugar =2.67 + 0.0189£ra r = 0.85 Assuming the maximum ETa and the maximum yields are those corresponding to the mean of treatments PI and Dl, i.e. mean maximum yield for roots is 90 t-ha'1 and mean maximum evapotranspiration is 592 mm, the root yield response to water is: Y root • Y^L = 1-03 - 1.07(l - ETa -ETj) r = 0.94 And for sugar yield: r = 0.85 Very similar relationships have also been developed during previous growing seasons. They show that the sugar beet crop is very sensitive to water stress and that a high yield reduction results from water stress. Nevertheless, if the stress occurs during specified periods of the growing season, the yield reduction can be small [2, 3]. 4. CONCLUSIONS The study showed that important water savings can be achieved if sugar beet irrigation schedules are based on either of the two techniques tested. Root yield varied from a minimum of 60 t-ha'1 for the treatment irrigated at the threshold value of water depletion of 80 mm-nr1 to a maximum of 92 t-ha-1 for the treatment irrigated whenever soil water potential reached -20 cbar. However, the latter and that irrigated at 40 mnvnr1 of soil water depletion resulted also in the highest water loss. Water use efficiency was highest and water loss negligible when irrigation was applied at 80 mm-nr1 of soil water depletion in the root zone or at -60 cbar as a threshold value of soil water potential. However, their root and sugar yields were neither competitive nor economically beneficial given the price of water practiced. Intermediate efficiency values were obtained with the treatments irrigated at 60 mm-nr1 depletion and -40 cbar potential. Economic analysis taking into account the monetary returns from the harvest and the expenses associated with water showed that the highest profit corresponds to the treatments irrigated at -40 cbar of soil water potential or 60 mm-nr1 of soil water depletion. Moreover, the performance of these treatments in terms of profit was higher or at least similar to that of the treatments which received larger amounts of water. Thus, from the economic viewpoint, and despite the price of water which is still considerably low, it is more beneficial for the farmer not 148 to satisfy the total water requirements of the crop. When water is available on demand, the implementation of these and other techniques can be simple provided farmers are initiated on their use or pilot stations are at hand for advising farmers on when to irrigate and how much water to apply. Nevertheless, such implementation, although easy at the farm level, is difficult and may even be impossible when considering a large irrigated area with a common source of water. When water is being transported for long distances before it reaches farms and the distribution network is of a limited capacity, it would be impossible and impracticable to provide water when needed for different crops growing on different soils and at different stages of growth. However, scenarios for optimizing the network functioning can be found for providing water near the threshold values. To find these scenarios, comprehensive research and extension programs involving farmers, researchers, extension agents and water management agencies are needed. Moreover, policies related to water pricing, delivery and use have to be reviewed to fit the needs. REFERENCES [1] [2] [3] [4] [5] [6] SMITH, M., CROPWAT: A computer program for irrigation planning and management. FAO Irrigation and Drainage Paper No. 46. Rome, Italy (1992). BAZZA, M., Contribution to the improvement of irrigation management practices through water-deficit irrigation. Third Research Co-ordination Meeting on "The Use of Nuclear and Related Techniques in the Assessment of Irrigation Schedules of Field Crops to Increase Effective Use of Water in Irrigation Projects." Rabat, Morocco, April (1995) 24-28. BAZZA, M., Contribution to the improvement of irrigation management practices through water-deficit irrigation. Second Research Co-ordination Meeting on "The Use of Nuclear and Related Techniques in the Assessment of Irrigation Schedules of Field Crops to Increase Effective Use of Water in Irrigation Projects." Fundulea, Romania, August (1993) 24-28. TAYLOR, A.S., ASHCROFT, G.L., Physical Edaphology: The Physics of Irrigated and Nonirigated Soils. W H. Freeman and Co. San Francisco (1972). MARHABOU, T., Contribution au pilotage de l'irrigation par la tenslomètrie et la méthode du bilan hydrique: cas de la betterave à suere dans le Périmètre des Doukkala. Mémoire de Troisième Cycle, Option Génie Rural. Institut Agronomique et Vétérinaire Hassan u, Rabat, Morocco (1994). BAZZA, M., Contribution to the improvement of irrigation management practices through water-deficit irrigation. Second Research Co-ordination Meeting on "The Use of Nuclear and Related Techniques in the Assessment of Irrigation Schedules of Field Crops to Increase Effective Use of Water in Irrigation Projects." Vienna, Austria, February (1992) 3-7. Next page(s) left blank 149 CONTRIBUTION TO THE IMPROVEMENT OF IRRIGATION MANAGEMENT PRACTICES THROUGH WATER-DEFICIT IRRIGATION M. BAZZA Institut Agronomique et Vétérinaire Hassan II, Département de l'Equipement et de l'Hydraulique, Rabat-Instituts, Morocco Abstract The study aimed at identifying irrigation management practices which could result in water savings through water-deficit irrigation. Two field experiments, one on wheat and the other on sugar beet, were conducted and consisted of not supplying water during specific stages of the season so as to identify the periods during which water deficit would have a limited effect on crop production. The year was exceptionally dry with a climatic deficit throughout the growing season. In the case of wheat, high water deficit occurred during the early stages and irrigation during these stages was the most beneficial for the crop. However, one water application during the tillering stage allowed the yield to be lower only to that of the treatment with three irrigations. Irrigation during the stage of grain filling caused the kernel weight to be as high as under three irrigations. However, this advantage could not make up for the loss with respect to the other yield components which were higher under good watering conditions during the stage of tillering. The highest water-use efficiency value was obtained with one irrigation during the tillering stage, although it was not very different from that of the treatment with three irrigations. The lowest value corresponded to the treatment with one irrigation during grain filling and that under rainfed conditions. For sugar beet, when water stress was applied early in the growing season, its effect could be almost entirely removed with adequate watering during the rest of the growing season. On the other hand, good watering early in the season, followed by a stress, resulted in the second lowest yield. Water deficit during the maturity stage had also a limited effect on yield. The most crucial periods for adequate watering were those which correspond to late foliar development and root growth which coincided with the highest water requirements period. Resuming irrigation late in the season, after a long period of stress caused the formation of new leaves at the expense of the sugar already stored in the roots. Stress throughout the growing season resulted in the highest yield reduction per unit of water deficit Nevertheless, for the same amount of water savings through deficit irrigation, it was better to partition the stress throughout the season than during the critical stages of the crop. From an economics standpoint, maximum profit for farmers was obtained with the fulfillment of the entire crop requirements. However, at the national level, it would have been more important to practice deficit irrigation and increase the irrigated area. For both crops, high yields as well as high water-use efficiency values could have been obtained provided the right choice of the period of water application is made. Under the conditions of the growing season, appropriate irrigation management practices would have allowed for the area cropped by wheat to be doubled and that of sugar beet to be increased by at least 30% with no decrease in yield. With a limited yield reduction, the area cropped by sugar beet could have been also doubled, with a substantial increase in monetary returns. 151 1. INTRODUCTION The study was conducted on wheat and sugar beet and had the following objectives; (i) conduct field trials aiming at identifying crop stages during which the crops could withstand water stress with limited effect on yield and quality, (ii) draw conclusions on ways of managing irrigation so as to increase water-use efficiency and limit water loss, (iii) make inferences on water savings on a large scale (irrigation projects) through adequate irrigation management In addition, a variety of by-products were expected to come out of the study, such as; (i) identify appropriate means of estimating water requirements of the crops used, within the context of the study region, (ii) contribute to assessing the crop coefficients associated with the empirical and semi-empirical models identified, (iii) contribute to establishing the water response function of the two crops used, both for the entire growing season and for each stage of these crops, (iv) identify appropriate means of irrigation scheduling and management which could be used at the farm level and by the agency in charge of project management. 2. SUGAR BEET 2.1. Materials and methods The experiment was conducted on an area of 0.8 ha in the Doukkala region in the western part of Morocco. During the previous growing season, the field had been cropped to wheat and had remained fallow between the wheat harvest in June and the soil preparation for the present experiment in October. Sowing took place manually in mid-November, with a spacing of 50 and 20 cm, respectively between and along the crop rows, i.e., with a crop density of 100,000 plants-ha-1. Fertilization consisted of blending 100 UN, 150 UP and 300 UK. To enhance germination, an irrigation was applied right after sowing. Throughout the crop season, production techniques were those recommended for maximum production in the region. 2.2. Experimental design and layout The experiment was oriented towards subjecting the crop to different watering regimes. The crop growing season was divided into four major growth periods, namely the initial stage (PI), the period of vegetative development (P2), the period of yield formation or root growth (P3) and that of ripening (P4). Ten treatments were intended initially as indicated in Table I. The experimental design consisted of completely randomized blocks with four replicates. Within each block the ten irrigation regimes were randomly distributed. Inasmuch as furrow irrigation is the method used for applying water, each irrigation regime consisted of 9 furrows out of which only the central five were used for measurements. Each treatment was 50 m long and 4.5 m wide (area of 225 m2). Adjacent treatments were 0.5 m apart from each other. 152 TABLE L DESCRIPTION OF THE IRRIGATION TREATMENTS Treatment Controls 1111 0000 TR Qne period of stress 0111 1011 1101 1110 Two periods of stress 0011 1001 1100 Growth Stage/Period PI P2 P3 P4 1 0 - 1 1 1 0 0 0 - - - 1 1 1 1 1 0 1 1 1 1 0 1 1 1 0 0 1 1 1 0 0 0 1 1 1 0 0 0 Description All normal watering All stress Traditional practice Stress during PI Stress during P2 Stress during P3 Stress during P4 Stress during PI & P2 Stress during P2 & P3 Stress during P3 & P4 (1) Normal watering (i.e. ETa - ETm) (0) Water stress (i.e. ETa < £Tm) Because the irrigation water was first lifted into a large calibrated reservoir from which it flowed under gravity, it was possible to accurately measure the amount of water supplied to each plot During the irrigation periods ETa = ETm, plots were normally irrigated every 15 d, except during peak water requirement periods when they were irrigated every 10 d. Each irrigation regime was equipped with three access tubes for monitoring soil water content at different depths. Measurements of soil water were taken every 3-4 d from February 7 through harvest and served for determining crop water use. At harvest (6th of June), root yield was measured in all treatments within all four replicates. Moreover, the sugar content of the roots as well as their technological quality were measured. The soil physical and chemical characteristics were determined on samples taken at different locations of the field. Precipitation was measured with a rain gauge installed inside the experimental field, while other climatological parameters were monitored in a station located hi the study region. The water consumptive use of the crop was determined by means of the water balance method neglecting drainage. Consumptive use was also estimated using the two methods described by Doorenbos and Kassam [10]. For a better accuracy, all the necessary parameters of these methods were determined in situ. Moreover, the soil water reservoir in the root zone was monitored through the measurements of soil water content with the neutron probe, and the rooting depth was measured occasionally until it reached the maximum of 60 cm by the end of 153 March. The crop water requirements (ETm) Criddle, Radiation and Penman. 2.3. were estimated by means of three methods: Blaney- Results and discussion 2.5.7. Soil characteristics Table n shows the soil characteristics determined in the laboratory. With respect to these characteristics, there seems to be no limitation with regard to sugar beet production. TABLE u. SOIL CHARACTERISTICS Soil property Fine sand (%) Coarse sand (%) Fine loam (%) Coarse loam (%) Clay (%) Texture 9 fc(%) 0 pwp(%) pH Bulk density (g-cnr3) EC (mmho-cnr1) P(mg-kg-i)t K(mg-kg-i) Organic matter (%) 0-15 42.5 15.0 8.1 6.7 27.7 SCL 17.11 8.32 7.78 1.26 2.88 21.90 156.20 1.16 Soil deoth (cm) 15-35 35-50 41.4 15.8 7.9 1.5 33.4 SCL 17.56 8.28 7.86 1.28 2.16 12.20 132.70 1.00 34.3 11.259.20 10.38 34.87 CL 17.92 8.21 7.83 1.30 2.24 11.00 68.30 0.84 50-70 29.0 9.7 6.6 18.4 36.4 CL 18.15 8.17 7.85 1.29 2.52 6.00 62.50 0.76 î The soil analysis was performed after the addition of fertilizers. 2.5.2. Climatic characteristics of the growing season Total precipitation during the crop season was abnormally low (133 mm). This figure is below the threshold value of 140 mm below which no sugar beet production is to be expected in the region, independently of its distribution [1]. As a result, throughout the growing period, the climatic deficit was important and irrigation was necessary for crop production. 2.3.3. Irrigation of the different treatments Apart from minor deviations owing to low levels of precipitation, all treatments were conducted according to the initially intended framework. Thus, treatment one received thirteen 154 water applications totaling the equivalent of 6250 m3-ha'1 in addition to 133 mm of precipitation. The partitioning of these applications were such that the crop was never stressed. Table HI shows the depths of water applied to each treatment and the dates of application. TABLE HI. DEPTHS OF WATER APPLIED TO THE DIFFERENT TREATMENTS (mm) AND CORRESPONDING DATES OF APPLICATION Date 9 Dec 1992 23 Dec 1992 6 Jan 1993 20 Jan 1993 19 Feb 1993 22 Mar 1993 -2 Apr 1993 14 Apr 1993 25 Apr 1993 9 Mayl993 20Mayl993 29 Mavl993 Total Crop age (d) 23 37 51 65 95 126 137 149 160 174 185 194 Tl T2 T3 T4 T5 T6 T7 T8 T9 T10 35 30 35 30 45 60 65 70 65 60 60 70 — — — — — — — — — — — — 35 — — — 40 30 45 60 65 70 65 60 60 70 35 30 35 30 45 — — — — — 75 70 35 30 35 30 45 60 65 70 65 60 — _ — — — — — 35 30 40 — — — 80 — 65 60 — _ 35 30 — — — 70 65 70 65 60 60 70 70 65 70 65 60 60 70 — — — — — 75 70 35 30 35 30 45 — — — — — — _ 0 280 560 525 320 495 460 210 175 625 2.3.4. Reference evapotranspiration ET0 and crop water requirements Reference evapotranspiration as estimated by three different methods (Penman, Radiation and Blaney-Criddle) known to be well adapted for the conditions of the region amounted to 705, 687 and 715 mm, respectively. Maximum crop water requirements were obtained by multiplying the reference evapotranspiration values with the sugar beet crop coefficients previously determined in the region. ETm values were low in the beginning of the growing season, increased gradually to attain a maximum during April and early May and subsequently decreased slowly in late May. The validity of all three methods within the context of the region is confirmed when it is noted that the evolution of ETm through the entire season was practically the same for each estimation method. Total ETm of the sugar beet crop amounted to 610 mm, for a growing period of 205 days. Assuming the irrigation efficiency to be 80% which is generously optimistic, gross water requirements were 7625 m3-ha'1. If we consider that each precipitation benefited the crop, farmers should have applied about 6300 m3-ha'1 to satisfy the crop water requirements. 155 Inasmuch as treatment Tl was conducted under no water shortage, its water consumptive use was considered equal to ETm. The correlation of the latter on estimated ETm values throughout the growing season yielded the following relationships: ETa = 0.077 + 0.980 ETm Penman (r = 0.998) ETa = 0.027 + 1.025 ETm Radiation (r = 0.997) ETa = 0.092 + 0.963 ETm Blaney-Criddle (r = 0.996) Thus, within the climatic conditions of the region, all three methods of estimating the crop water requirements can be considered as valid for the sugar beet crop. Similar results have been found during the last three growing seasons. The ratios of estimated ETm to maximum ETa of treatment number one can also serve for assessing the crop coefficients. The latter, based on all three methods, are given in Tables IV and V. TABLE IV. MONTHLY CROP COEFFICIENTS Kc OF SUGAR BEET BASED ON THREE METHODS OF ESTIMATING CROP WATER REQUIREMENTS Month December January February March April May June KC Penman KC Radiation KC Blaney-Criddle 0.47 0.75 0.84 1.01 1.09 1.00 0.89 0.50 0.77 0.82 1.00 1.15 1.02 0.93 0.50 0.81 0.75 0.93 1.09 0.99 0.88 By comparison, these values are very close to those identified by Doorenbos and Kassam [10] and which are as follows: 0.45 for the initial stage, 0.8 for the crop development stage, 1.1 for the mid-season stage and 0.95 for the late season stage. 2.3.5. Water consumptive use Table VI gives crop water use under the different irrigation regimes. Crop period 1 (initial stage) was completely dry. However, because of the low water requirements during this stage as well as the fact that an irrigation was made after the crop was sowed, treatment 4 which was to be stressed during this period was subjected only to a minor stress at the end of the stage. The maturity stage was also completely dry. As a result, treatment 7 which received no water application during this stage used only 73 mm during this period. This amount was almost 36% below the maximum requirements. 156 TABLE V. SUGAR BEET CROP COEFFICIENTS Kc FOR THE MAIN GROWING PERIOD BASED ON THREE METHODS OF ESTIMATING WATER REQUIREMENTS Growing period KC Penman KC Radiation KC Blaney-Criddle 0.49 0.82 1.05 0.98 0.52 0.82 1.08 1.01 0.52 0.79 1.02 0.97 PI P2 P3 P4 TABLE VI. WATER CONSUMPTIVE USE (mm) AS MEASURED BY THE IN SITU WATER BALANCE METHOD AND ESTIMATED BY TWO DIFFERENT METHODSt Treatment Tl T2 T3 T4 T5 T6 T7 T8 T9 T10 Water balance Feb. 7 - June 6 Total 548 140 378 562 510 332 512 . 493 289 253 Daily mean 4.4 1.1 3.0 4.5 4.1 2.7 4.1 3.9 2.3 2.0 Method 1 Dec. 1 - June 6 Method 2 Dec. 1 - June 6 622 167 460 621 528 389 574 493 274 295 626 166 409 619 587 389 564 543 291 307 t ET a estimates from method 1 are based on intervals between water applications, while those from method 2 are based on monthly periods [10]. Because natural precipitation amounted to 67 mm, stage 2 was only partially stressed. Total water requirements during this stage were 115 mm while actual evapotranspiration of treatment 5 which was to be stressed during this period was over 51% below this requirement. Treatment 6 which received no irrigation during the third stage was the most severely stressed. Its consumptive use during this stage (116 mm) was over 66% below the requirements. For the entire growing season, water consumptive use of treatments 3,4,5,7 and 8 was over 74% of ETm. That of treatment 6 was over 62% of ETm, while treatments 2,9 and 10 used 157 less than 50% of the maximum requirements. In the case of treatment 2 which was conducted under rainfed conditions, total ETa is slightly higher than the total precipitation of the growing season, which means that the soil water reservoir has contributed to ETa. Maximum values of ETa varied from 2.23 mm-d'1 for treatment 2 and 5.7 mm-d'1 for treatments 1, 7 and 8. They occurred at different moments, depending on the irrigation schedule of each treatment During the period from February 7 through harvest where the in situ water balance method was applied, all three methods of ETa determination gave close results, except for a slight overestimation by the in situ water balance method due to drainage which was not accounted for by this method. 2.3.6. Yield and water use efficiency The mean crop density at complete maturity was 88,310 plants-ha4, with no significant difference between treatments. Table VII shows the root and sugar yield obtained under the different treatments. Had it not been for a root rot disease whose origin could not be absolutely identified (probably a boron deficiency), these yields could have been higher by at least 15%. The proportion of infested roots was the same in all treatments which excludes its relationship with watering conditions. Minimum yields of both roots (29.96 t-ha4) and sugar (6.56 t-ha4) were obtained under rainfed conditions (treatment 2). Treatments 1 and 4 which were not subjected to water stress resulted in the maximum yield of both roots (about 80 t-ha'1) and sugar (almost 15 t-ha-1). The TABLE VII. ROOT AND SUGAR YIELD AND WATER-USE EFFICIENCY OF IRRIGATION WATER APPLIED TO THE TREATMENTS Irrigation Numbert Volume Treatment Tl T2 T3 T4 T5 T6 T7 T8 T9 T10 12 0 5 10 9 7 10 7 4 5 6250 0 2800 5600 5250 3200 4950 4600 2100 1750 Yield Roots (t-ha-1) Sugar (t-ha-1) 81.38 29.96 64.05 79.40 75.09 61.87 71.80 69.40 45.44 37.72 14.72 6.56 12.26 14.98 14.20 11.53 14.03 12.32 8.02 8.14 The irrigation applied right after the crop sowing is not included. 158 Water-use efficiency Roots Sugar (kg-nr3) 13.02 22.87 14.18 14.30 19.33 14.50 15.08 21.63 21.55 (kg-nr3) 2.35 4.37 2.67 2.70 3.60 2.83 2.68 3.82 4.65 only difference between treatments 1 and 4 is that the latter received two water applications less than the former during the initial stage. In spite this difference, the water consumptive use was similar under the two treatments owing to the irrigation applied after sowing and to the low water requirements during this stage. Treatment 10 which was conducted under adequate watering conditions throughout the first half of the growing season, followed by a period of stress, resulted in the second lowest yield (37.72 t-ha-1 roots). The same tendency has been found previously and is attributed to the fact that adequate watering conditions early in the season leads to the development of an abundant leaf cover and a shallow root depth. When a severe stress follows, the crop rapidly depletes the soil water stored in the root zone and wilts before the completion of additional root development at greater soil depths. Treatment 9 received only four irrigations. Two were during the initial stage and the remaining two during the last stage (maturity). These last two water applications allowed the crop to partially recover inasmuch as its yield (45.4 t-ha*1) was higher than that of T10. This ability of sugar beet to partially recover the effect of early water stress has also been identified during the previous growing seasons. From both of these results, it can be concluded that under limited water, it is better to start subjecting the crop to stress early in the season. By doing so, the crop adapts to limited watering conditions with the stress not being severely concentrated in any one time period. This conclusion is confirmed also by treatment 3 which received the same number of irrigations as treatment 10 (with different timing), but whose yield was over 70% greater than that of T10. The water consumptive use of this treatment T3 was 345 mm lower than ETm, but this deficit was more or less evenly distributed during the growing season. Treatments 5 and 8 were irrigated essentially during the second half of the growing season. However, the former received two water applications before the latter which allowed the crop to partially avoid the stress during the period of intensive foliar development, thus realizing a highly competitive yield of 75 t-ha'1. Treatment 6 was subjected to a deficit of over 300 mm in the middle of the growing season. Resuming irrigation during the last stage allowed the crop to partially recover inasmuch as its yield is over 24 t-ha-1 higher than that of T10. A deficit occurring-only during the last stage is illustrated by treatment T7. Although the deficit was only 130 mm with respect to ETm, the yield reduction was almost 10 t-ha-1. This yield reduction would have been much greater had the crop been subjected to water stress during any of the previous stages. Similar results have been found during the previous growing seasons. It should be noted at this point that hi the study region, many farmers withhold irrigation during this last stage, as well as providing inadequate water throughout the growing season. This situation causes major yield reductions. The sugar content of the roots was also highly related to the watering regime. Maximum sugar contents obtained under treatments 2 and 10 (21.9% and 21.6%, respectively) are attributed to the effect of the water stress caused by these treatments, especially that late in the growing season (prior to harvest). They were followed by treatments 7 and 3 (19.6% and 19.2%, respectively) which were subjected to water stress during the last month of the season. 159 Treatments 1,4, 5 and 8 which continued to be irrigated until about one week prior to harvest had relatively lower values of sugar content (18.1,18.9,18.9 and 17.8%, respectively). The minimum sugar content was obtained under treatment 9 with 17.7%. Thus, in general, reducing the amount of water applied throughout the season and/or withholding irrigation prior to harvest increases the sugar content of the roots and vice-versa. 2.3.7. Yield - water relationship From the above discussion, it is completely reasonable that yield differences between treatments can be attributed essentially to their watering conditions. The following linear relationships between root or sugar yield (t-ha"1) and the depth of irrigation water applied to each treatment (mm) were obtained: Root yield = 29.95 + 0.087(Depth of irrigation) (r = 0.96 ) Sugar yield = 6.33 + 0.015(Depth of irrigation) (r = 0.95 ) Hence, about 90% of the yield variability between treatments can be attributed to the difference in the amounts of water applied through irrigation, independent of the timing of the applications. When considering the depth of water actually used by the crop during the period February 7 through harvest, the relationships remain linear: Root yield = 13.29 + 0.12 ETa (r = 0.97 ) Sugar yield = 3.47 + 0.02 ETa (r = 0.96 ) 2.3.8. Water production function Crop production functions were determined for the entire growing season as well as for various periods during which stress was applied. The highest yield response coefficient (0.86) was obtained when the stress took place during the entire season. This value puts sugar beet among the crops which require large amounts of water if the yield is not to be reduced by water stress. The second highest yield reduction took place when the water stress was applied during the last two stages of the growing season (root development and maturity) with a coefficient of 0.74. Intermediate but still important yield reduction resulted from the stress during the P2 and P3 stages with a coefficient of 0.64. Similar results for all of these stress periods and growth stages have been observed during the previous year. The yield response coefficient corresponding to the first two stages (PI and P2) this year was slightly higher than that of the previous growing season with a value of 0.42. During the previous year, the stress applied during these two stages had the smallest effect. It should be noted however that during the previous year owing to different precipitation, only limited stress could be applied during these stages. When the stress took place only during the last stage (maturity), it had only a limited effect on yield, with a coefficient of 0.38. Nevertheless, the smallest effect was observed when the stress was applied only during the second stage (P2). In conclusion, the same tendency as that of the previous year was found during the present growing season. When water stress is applied early in the growing season after the 160 plants have 6 to 8 leaves, high yields could easily be sustained provided adequate watering conditions take place during the rest of the growing season. However, the water deficit during the early two stages should not be too high (over 50% of the requirements) for a long period of time. In the same manner, limited effect on yield results from water deficit applied late in the season during the maturity stage. The most crucial periods for adequate watering are those which correspond to late foliar development and root growth. These periods coincide with the highest water requirements and the crop cannot withstand water deficits of more than 30% without important reductions of yield. Meeting the water requirements during the first two stages can be very harmful if water shortages occur during the remaining of the season. Good watering early in the season allows the crop to develop an important cover and a limited root system. When the stress occurs, the crop depletes the soil water fast and wilts irreversibly. Stress throughout the growing season results in the highest yield reduction per unit of water deficit. Nevertheless, for the same amount of water savings through deficit irrigation, it is better to partition the stress throughout the season rather creating stress during one of the critical stages of crop growth. 2.3.9. Economic considerations Economic calculations performed on the basis of the yield obtained and the actual price of water showed that maximum profit for the farmer was obtained with the fulfillment of the entire crop requirements. However, at the national level, it is more important to practice deficit irrigation and increase the irrigated area. With appropriate irrigation management, the amounts of water applied could be reduced by at least 30% without an important effect on yield. 3. WHEAT 3.1. Materials and methods 3.1.1. Experimental design Four durum wheat genotypes known to be highly productive in terms of yield were used in an experiment consisting of five treatments based on the stage(s) during which supplementary irrigation was applied. These genotypes studied last year include a soft genotype (Acsad 59) and four others called locally Isly, Marzak, Tassaout and Sarif. The following treatments were established: Treatment A: Naturally occurring rainfed conditions, Treatment B: One irrigation during the stage of grain filling, Treatment C: One irrigation during the stage of heading, Treatment D: One irrigation during the stage of tillering, Treatment E: One irrigation during each of the above three stages. 161 BLOCKS BLOCK 2 BLOCK 1 A 31 35 34 33 32 35 34 32 31 S3 33 31 34 32 35 B 34 33 31 32 S5 32 33 34 31 35 35 34 33 31 32 H Z W 31 35 32 33 34 D 35 32 34 31 33 —— *• 3m E 35 3 31 33 32 34 35 32 34 31 33 — 35 33 32 34 Gl — «l*« F— t i G3 34 31 32 35 a 32 33 31 35 34 Fig. L Experimental layout of the irrigation of five wheat genotypes. Symbols Gl through G5 represent Isly, Marzak, Tassaout, Acsad 59 and Sarif, respectively. The experimental layout consisted of a strip-plot design with three replicates. The first factor (irrigation regime) was attributed to the large plots, and the second factor which was randomized (genotype) to the smaller plots (12 x 4m). Treatments were 8 m apart and genotypes within each treatment were separated by 30 cm (Fig. 1). Sowing took place on the 17th of December and was performed with a mechanical sowing machine having six lines 30 cm apart from each other. Care was taken so as to have the same plant density for all genotypes which required the equivalent of 136,125,136,128 and 114 kg-ha-1 for Isly, Marzak, Tassaout, Sarif and Acsad 59, respectively. Areas between genotypes and treatments as well as those around the blocks were also cultivated by wheat in order to avoid border effects. Owing to a lack of precipitation before and after sowing, a light irrigation of 40 mm was applied in early January to all treatments to enhance germination. Each treatment was irrigated separately; the depth applied during each irrigation was 60 mm. The irrigation system consisted of sprinklers installed on a grid of 12 x 12 m that provided 164 a nominal flow rate of 6 mm-hr1. The dates of the different irrigation applications are given in Table VIEL 162 TABLE VÏÏL IRRIGATION DATES OF THE DIFFERENT TREATMENTS Irrigation date Treatment Tillering (D) Heading (C) Grain filling (B) Frequent (E) 21/02/93 03/04/93 28/04/93 21/02/93,04/04/93,29/04/93 3.1.2. Observations and measurements Climatic parameters were monitored throughout the growing season by means of an automatic weather station installed near the experimental field. Recorded parameters included solar radiation, relative humidity, wind speed, insolation, precipitation and temperature. Soil water content was measured with a neutron probe in all treatments of the first block at soil depths 20,40,60 and 80 cm. Measurements were taken every 2 to 3 d during the last 110 days of the growing season. A calibration curve for the neutron probe was determined during a period of two months using two sites located in the middle of the experimental field. Soil bulk density was also measured in situ at different depths. The crop water consumptive use was determined by means of the in situ water balance method during the last 115 days of the growing season. In addition, it was estimated using the method described by Doorenbos and Kassam [10] and soil water storage based on neutron probe measurements. To avoid the effect of crop destruction during measurements and monitoring of the various parameters on the final yield, each experimental plot was partitioned as indicated in Fig. 2. The left part (monitoring) served for all measurements performed on the crop, while the right part (harvest) was left undisturbed for estimating the final yield and its components. NEUTRON ACCESS TUBE / AND TENSIOMETERS BORDER 1m / ï !? MONITORING HARVEST 11 BORDER Fig. 2. Sampling and measurement design for each plot. 163 The plants were monitored for leaf area index, stem height and above ground dry matter. Leaf area was measured in all blocks and for all genotypes every 15 to 20 d. Samples were taken on two adjacent lines over a length of 30 cm and the area of all synthesizing leaves was measured with an electronic area-meter. The stem height was determined with the same frequency by measuring each time the length of ten stems taken randomly in each plot. The above ground dry matter in all treatments and for all five genotypes was determined every 20 d. The method consisted of taking all aerial parts of two adjacent lines over 50 cm and drying them at 80°C to obtain the dry weight At complete maturity, the final crop density and the number of tillers per plant were determined in all sub-plots used for estimating the yield. The yield components were determined by taking ten spikes randomly in each plot. The kernel weight was obtained by weighing 4 samples of 250 kernels each. Simultaneously, the grain and straw yields were determined by harvesting a 10 x 1.2 m area. 3.2. Results and discussion 3.2.1. Characteristics of the experimental site The trial was conducted in the same region as the previous one, except that the soil and the climatic characteristics of the growing season were slightly different Table DC gives the soil chemical and physical properties. The soil texture is sandy clay, organic matter content is low and the pH is lightly alkaline. The bulk density ranges from 1.44 to 1.54 g-cnr3 and the water holding capacity is low due to the high sand content 5.2.2. Characteristics of Ute growing season Total precipitation during the period from September to May amounted to 200 mm, which is less than two-thirds of the region's mean (Table X). Out of this total, only about 140 mm coincided with the growing season which is too low compared to the water requirement. Air temperature was characterized by a low amplitude with a maximum of 25°C and a minimum of 2°C. The mean temperature throughout the period from early January to late May was 19.1°C which is very favorable for wheat production. The climatic deficit, defined by the difference between precipitation and potential evapotranspiration, was negative throughout the growing season. With the magnitude of this deficit increasing through the growing season to a maximum of about 5.5 mm-d"1 during the month of May, crop production without irrigation is expected to be very low. Crop water requirements estimated by the Penman method varied from a minimum of 2.4 mm-d'1 to a maximum of 6.9 mm-d'1. 164 TABLE IX. SOIL PHYSICAL AND CHEMICAL CHARACTERISTICS Soil depth (cm) 0-20 20-40 40-60 60-80 CS 12.86 10.88 10.61 11.35 Sou depth (cm) 0-20 20-40 40-60 60-80 Constituents (%ï FS CL FL 38.08 7.28 7.01 42.41 5.635.28 40.63 7.60 4.51 39.67 7.533.68 Exchangeable cations Ca Mg Na K [meq-(100g)-i] 15.2 16.8 19.0 C 4.0 4.4 5.4 Bulk density (g-cm-3) pH 1.44 1.46 1.52 1.54 7.64 7.46 7.83 7.83 34.77 35.80 36.65 37.77 P2O5 O.M.E.G. 25°C F.C. PWP (mg-kg-i) (%) (mmho-cnr1) (%) (%) 1.56 0.27 1.84 0.23 1.82 0.20 1.71 0.27 15.5 7.6 6.3 6.3 0.80 2.33 0.78 2.39 0.80 2.40 0.73 2.35 24.47 24.58 24.90 22.38 11.45 12.07 13.22 13.14 TABLE X. PRECIPITATION DURING THE GROWING SEASON (1992-1993) AND MEAN PRECIPITATION DURING THE PERIOD 1976-1993 (mm) Month Sept Oct Nov Dec Jan Feb Mar Apr May 1992-1993 1976-1993 1.5 2.3 24.3 31.7 11.0 51.0 26.8 50.0 23.1 54.8 29.9 45.5 51.7 35.6 25.5 28.5 5.6 10.7 3.2.3. Yield analysis 3.2.3.1. Grain yield Table XI shows the grain yield obtained under the different irrigation regimes, for all five genotypes. Both factors as well as their interaction had a highly significant effect of this variable. Under the climatic conditions during the growing season, characterized by low precipitations, all irrigation treatments resulted in different yield values. The highest yield was obtained under three water applications, followed in decreasing order by those irrigated once during tillering, heading and grain filling. The lowest yield was obtained under rainfed conditions with 60% less water than that of the treatment irrigated three times. 165 TABLE XL GRAIN YIELD OF ALL FIVE GENOTYPES UNDER THE IRRIGATION REGIMES (100 kg-ha-1) Irrigation treatment Isly Marzak E D C B A 59.24 45.20 32.50 28.33 24.18 50.68 44.90 31.66 25.00 22.99 37.89 a 35.05 b Mean Genotvpe Tassaout Acsad59 Sarif Meant 51.97 43.20 35.17 26.15 18.17 54.23 46.34 33.33 26.15 22.02 47.15 43.60 30.45 21.11 17.18 52.65 a 44.65 b 32.62 c 25.35 d 20.91 e 34.93 b 36.41b 31.90c 35.24 t Figures with the same letter are not significantly different at the 5% level according the test of Neuman and Keuls. Conditions during this growing season were such that water deficit during all three stages (tillering, heading and grain filling) was higher compared to that of the previous two years. Within the treatments with one water application, irrigation during the stage of tillering resulted in a higher yield because of its superiority in terms of most yield components, especially the tiller density, the kernel density, the number of spikelets per spike and the number of kernels per spike. Irrigation late in the growing season (grain filling) caused the kernel weight to be as high as under three irrigations. However, this advantage could not make up for the loss with respect to the other weight components, especially kernel density. The latter is higher under good watering conditions during the stages of tillering and stem elongation. The different yield components are elaborated during different stages of the growing season, and watering conditions during each of these stages determines the performance of the genotypes with respect to these components. The performance of the five genotypes used were such that the grain yield was highest with Isly, followed by Acsad 59, Marzak and Tassaout, and finally Sarif. 3.2.3.2. Straw yield As it was in the case of grain yield, both irrigation treatment and genotype as well as their interaction had a significant effect on straw yield. In addition, all five irrigation treatments resulted in different straw yields with a decreasing order as follows: three irrigations, one irrigation during tillering, one irrigation during heading, one irrigation during grain filling and no irrigation (Table Xu). 166 TABLE XH. STRAW YIELD (100 kg-ha-i) UNDER THE DIFFERENT IRRIGATION REGIMES AND FOR ALL FIVE GENOTYPES Irrigation treatment Isly Marzak E D C B A 82.31 73.69 69.16 57.77 51.93 82.65 73.28 72.78 66.88 48.64 Mean 66.97 a 68.85 a Genotvpe Tassaout Acsad59 Sarif Meant 91.36 74.57 59.82 56.18 56.94 81.88 75.32 70.55 64.00 56.43 77.64 61.39 57.99 57.33 46.95 83.17 a 71.65 b 66.06 c 60.43 d 52.18 e 67.77 a 69.64 a 60.26 b 66.70 t Figures with the same letter are not significantly different at the 5% level according the test of Neuman and Keuls. Thus, the greatest straw yield was achieved when irrigation was applied during the vegetative stages. With respect to genotypes, only Sarif resulted in a relatively lower straw yield than the other four genotypes. The interaction was such that the maximum straw yield (9100 kg-ha*1) was obtained with the genotype Tassaout under three irrigations, and the minimum (4700 kg-ha*1) with Sarif under rainfed conditions. 3.2.3.3. Total above-ground biomass The total biomass production (Table Xffl) behaved in a similar fashion to both grain and straw yields taken separately. With respect to genotypes, only Sarif manifested a relatively lower value than those of the other four genotypes. This lower biomass .production for Sarif is explained on the basis of its smaller height as well as being relatively more sensitive to Xanihomonas-Translucens which attacked the crop during its vegetative stages. In general, early water applications improved biomass production. 3.2.3.4. Water consumptive use 3.2.3.4.1. Actual evapotranspiration ETa The actual water consumptive use of the crop was determined by means of the in situ water balance method during the last 115 days of the season. It was also estimated using the method described by Doorenbos and Kassam [10]. The measured values ofwater content under the different treatments were used to determine the available water in the root zone and this variable was introduced in the model for a better accuracy. The method was tested and found to 167 TABLE Xm. TOTAL ABOVE GROUND BIOMASS (100 kg-ha-1) AS AFFECTED BY GENOTYPE AND IRRIGATION TREATMENT Irrigation treatment My E D C B A Mean Marzak Genotype Tassaout Acsad59 Sarif Meant 141.55 118.88 101.66 86.10 76.11 133.33 118.18 104.44 91.88 71.63 143.33 117.77 94.99 82.33 75.11 136.11 121.66 103.88 90.15 78.45 124.77 104.99 88.44 78.44 64.13 135.82 a 116.305 98.68 c 85.78 d 73.09 e 104.86 a 103.89 a 102.71 a 106.05 a 92.15 b 101.93 t Figures with the same letter are not significantly different at the 5% level according the test of Neuman and Keuls. reflect the exact values of ETa not only for wheat but also for sugar beet. Table XIV shows the estimated ETa values of the different treatments during the last three months of the growing season. Considering the genotypes, the lowest water consumptive use value was obtained with Acsad 59 (197 mm) and the highest with Sarif (219 mm). The other three genotypes used exactly the same amount, i.e., 207 mm. As for the irrigation treatments, the water consumptive use reflected the grain and straw yields obtained. It decreased continuously from a maximum of 274 mm with three irrigations to about 150 mm under rainfed conditions. TABLE XIV. ACTUAL EVAPOTRANSPIRATION ETa OF THE DIFFERENT TREATMENTS (mm) DURING THE LAST THREE MONTHS OF THE SEASON Irrigation treatment My E D C B A Mean 269.5 226.3 198.8 184.4 159.8 207.8 168 Marzak Genotvpe Tassaout Acsad59 Sarif Mean 272.5 223.5 204.9 184.3 150.4 207.0 291.2 226.3 189.5 187.6 144.3 207.8 257.1 223.4 186.6 172.1 144.3 196.7 281.6 244.6 220.2 196.9 150.4 218.7 274.4 228.8 200.0 185.1 149.8 207.6 3.2.3.4.2. Yield-ETa relationship The best relationship obtained between grain and straw yields YLD and the depth of water used by the crop ETa is of the logarithmic form: where the yield is in 100 kg-ha'1 and ETa in mm. The regression coefficients a and b as well as the correlation coefficients r2 for the different genotypes are reported in Table XV. TABLE XV. REGRESSION COEFFICIENTS AND CORRELATION COEFFICIENTS FOR GRAIN AND STRAW YIELDS (100 kg-hr1) Straw vieldt Grain vieldt Genotype a b r2 a b Isly Marzak Tassaout Acsad 59 Sarif -335.94 -239.91 -230.04 -276.14 -249.05 70.26 51.74 49.91 59.40 52.36 0.96** 0.87* 0.94** 0.97** 0.86* -251.83 -224.37 -216.22 -159.00 -177.29 59.92 55.18 53.49 43.45 44.25 r2 0.96** 0.93* 0.84* 0.98** 0.87* t Probabilities of 0.05 and 0.01 are indicated by * and **, respectively. Thus for all genotypes, the yield variation between treatments can be explained essentially by variations in water consumptive use. These results also show that crop response to water is a good indicator for comparing genotypes. Similar results have been found during the previous growing season. 3.2.3.4.3. Water-use efficiency The yield per unit of water used by the crop is a important indicator of the relative performance of the different genotypes under different water availability conditions. Under the conditions of the experiment and the climate of the growing season, both factors, and especially the watering regime, had a significant effect on this variable (Table XVI). However, the variability between irrigation treatments was lower for water-use efficiency than for grain yield. With respect to treatments, the highest efficiency is obtained with one irrigation during the tillering stage, although it is not very different from that of the treatment with three irrigations. The lowest value corresponds to the treatment with one irrigation during grain filling and that under rainfed conditions. Thus, the water application during grain filling had absolutely no effect on water-use efficiency. The earlier the irrigation was, the higher is the water return in terms of both grain and straw yields. The values of the water-use efficiency are high because they are based only on the consumptive use during the last three months of the season. 169 TABLE XVI. WATER-USE EFFICIENCY FOR GRAIN PRODUCTION [kg-(ha-mm)-i] Irrigation treatment Genotvpe Tassaout Isly Marzak C B A 22 20 16 15 15 19 20 15 14 15 18 19 18 14 13 21 21 18 15 15 17 18 14 11 11 19.4 19.6 16.2 13.8 13.8 Mean 17.6 16.8 16.4 18.0 14.2 16.6 E D Acsad59 Sarif Mean Some differences with regard to water-use efficiency existed also between genotypes. Thus, Acsad 59 yielded the highest value, followed by Isly, then Marzak, Tassaout and finally Sarif. Similar results were obtained for water-use efficiency for straw production (Table XVII). TABLE XVH. WATER-USE EFFICIENCY FOR STRAW PRODUCTION [kg-(ha-mm)-i] Genotvpe Irrigation treatment Isly Marzak Tassaout Acsad59 Sarif E D C B A 53 20 51 47 48 53 20 51 48 48 52 19 50 44 52 54 21 56 52 54 43 18 40 40 43 51.0 19.6 49.6 46.2 49.0 Mean 50.2 49.8 49.4 53.8 42.0 49.0 4. Mean CONCLUSIONS The growing season was exceptionally dry with a climatic deficit throughout the season for both wheat and sugar beet. Except for minor changes, the experiment was a replicate of that conducted the previous year with many of the results relative to the critical stages of the crops being the same for both years. 170 In the case of wheat with high water deficit occurring during the early stages (tillering and stem elongation), irrigation during these stages was most beneficial for the crop. Withholding irrigation during these stages subjected the crop to a severe deficit which amounted to over 60% of the requirements for over two months and resulted in a low tiller and spike density. However, one water application during the tillering stage was enough to reduce this effect and to allow the crop to have a much higher density which is an important yield component. The final yield under this situation was lower only to that of the treatment with three irrigations. Within the treatments with only one water application, irrigating during the tillering stage produced a higher yield because of its superiority in terms of yield components. Irrigation late in the growing season (grain filling) caused the kernel weight to be as high as under three irrigations. However, this advantage could not make up for the loss with respect to the other yield components, especially density. The latter is higher under good watering conditions during the stages of tillering and stem elongation. The highest water-use efficiency value was obtained with one irrigation during the tillering stage, although it was not very different from that of the treatment with three irrigations. The lowest value corresponded to the treatment with one irrigation during grain filling and that under rainfed conditions. Thus, the water application during grain filling had absolutely no effect on water-use efficiency. The earlier the irrigation was, the higher was the water return in terms of both grain and straw yields. During the same season farmers applied an average of 3.4 irrigations totaling over 300 mm. With adequate management and the appropriate choice of the irrigation timing, two water applications with 60 mm each, in addition to a light irrigation following the crop sowing would have resulted in the same yield and practically doubled the irrigated area. For sugar beet, finding the same tendency as that of the previous year confirms most results. When water stress was applied early in the growing season, its effect could be almost entirely removed provided that adequate watering conditions took place during the rest of the growing season. On the other hand, good watering early in the season followed by a stress caused the development of a large leaf area and a shallow root system. As a result, the crop could not withstand water deficit during root growth and maturity due to high water requirements during these stages. The most crucial periods for adequate watering were those which correspond to late foliar development and root growth. These periods coincide with the highest water requirements and the crop cannot withstand water deficits of more than 30% without a decrease in yield. Resuming irrigation late in the season after a long period of stress caused the formation of new leaves at the expense of the sugar already stored in the roots. Water deficit late in the growing season (maturity stage) had a limited effect on yield. Stress throughout the growing season resulted in the highest yield reduction per unit of water deficit. Nevertheless, for the same amount of water savings through deficit irrigation, it was better to partition the stress throughout the season than during any critical stage of the crop. The treatment which was not subjected to water stress resulted in the highest yield in terms of both roots and sugar. The minimum was obtained under rainfed conditions. Economically speaking and considering the actual price of water, maximum profit for the farmer was obtained with the fulfillment of the 171 entire crop requirements. However, at the national level, it would have been more important to practice deficit irrigation and increase the irrigated area. The methods of Penman, Radiation and Blaney-Criddle were tested and found to be valid for estimating the water requirements of sugar beet in the study region. The same result has been found during the preceding three years. The crop coefficient of sugar beet were also determined and found to be close to those identified by Doorenbos and Pruitt [11]. For both crops, high yields as well as high water-use efficiency values can be obtained provided the right choice of the period of water application is made. Under the conditions of the growing season, appropriate irrigation management practices would have allowed for the area cropped to wheat to be doubled and that to sugar beet to be increased by at least 30% with no decrease in yield. With a limited yield reduction, the area cropped to sugar beet could have also been doubled with a substantial increase in returns. REFERENCES [1] [2] [3] [4] [5] [6] 172 BAZZA, M., Identification of wheat varieties suitable for arid and semi-arid conditions and of the characters for selection and breeding in regard to water-use efficiency. Report submitted to IAEA within the framework of the Research Coordination Programme on "The Use of Isotopes in Increasing and Stabilizing Plant Productivity in Low Phosphate and Semi-arid and Sub-humid Soils of the Tropics and Sub-tropics" (1993). BAZZA, M., Effets du déficit hydrique et de sa place dans le cycle sur le rendement et la qualité technologique de la betterave à sucre. Winter Congress of Institut International de Recherches Betteraviäres. Brussels. 10-12 February (1993). BAZZA, M., Contribution to the improvement of irrigation management through waterdeficit irrigation. Second FAO/IAEA Research Coordinated Meeting on "The Use of Nuclear and Related Techniques in Assessment of Irrigation Schedules of Field Crops to Increases the Efficient Use of Water in Irrigation Projects." Fundulea, Romania. 2428 August (1993). BAZZA, M., Besoins en eau et irrigation de la betterave à sucre dans le Périmètre des Doukkala. Rapport définitif. Report of a study submitted to the sugar factories "Sucrerie des Doukkala" and "Sucrerie des Zemamra". Morocco (1993). BAZZA, M., Besoins en eau et irrigation de la betterave à sucre dans le Périmètre des Doukkala. Rapport définitif. Report of a study submitted to the sugar factories "Sucrerie des Doukkala" and "Sucrerie des Zemamra". Morocco (1992). BAZZA, M., Identification of wheat varieties suitable for arid and semi-arid conditions and of the characters for selection and breeding in regard to water-use efficiency. Third Coordinated Research Meeting on "The Use of Isotopes in Increasing and Stabilizing Plant Productivity in Low Phosphate and Semi-arid and Sub-humid Soils of the Tropics and Sub-tropics". FAO/IAEA/SIDA. Tunis, Tunisia. 5-9 October (1992). [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] BAZZA, M., Taj, A.. 1990. Effets du régime hydrique et de la qualité de l'eau d'irrigation sur le rendement et la qualité technologique de la betterave à sucre dans le primètre du Gharb. Sucrerie Maghrébine. Numéro Spécial. Rencontre du Groupe méditerranéen de 1HRB (1990). BAZZA, M., Identification of wheat varieties suitable for arid and semi-arid conditions and of the characters for selection and breeding in regard to water-use efficiency. Report submitted to IAEA within the framework of the Research Coordination Programme on "The Use of Isotopes in Increasing and Stabilizing Plant Productivity in Low Phosphate and Semi-arid and Sub-humid Soils of the Tropics and Sub-tropics" (1994). CORLIER, L., Les besoins en eau de la betterave sucrière. Quatorze années d'expérimentation (1961-1974) dans le Tadla, le Gharb, les Doukkala et en Moulouya. MARA/DER. CE. Bureau des besoins en eau. No. 143/75. Mars (1975). DOORENBOS, J., KASSAM, A.H., Yield response to water, FAO Irrigation and Drainage Paper 33, FAO, Rome (1979) DOORENBOS, J., PRUITT, W.O., Guidelines for predicting crop water requirements. Irrigation and Drainage Paper 24, FAO, Rome (1977). EL GHALI, A., Performances de différentes variétés de blé à la résistance à la sécheresse sous irrigation de complément. Mémoire de Troisième Cycle. Option Génie Rural. IAV Hassan II. Rabat (1992). EL MASAOUDI, M., Comportement de variétés de blé vis-à-vis de l'irrigation de complément: Efficience d'utilisation de l'eau et critères de sélection. Mémoire de Troisième Cycle Agronomie. Option Génie Rural. IAV Hassan H. Rabat (1993). LAKBAIBI, A., Besoins en eau du blé dans la vallée du Draa: contribution à l'optimisation de la productivité de l'eau. Mémoire de Troisième Cycle. Option Génie Rural. IAV Hassan H. Rabat (1990). LAROUSSI, M., Efficacité de l'irrigation de complément de variétés de blé relativement tolérantes au stress hydrique. Mémoire de Troisième Cycle Agronomie. Option Génie Rural. IAV Hassan H (1991). SALAMA, N., Comportement de la betterave à sucre vis-à-vis de l'irrigation déficitaire à différents stades de son cycle de développement dans le Périmètre des Doukkala. Mémoire de Troisiäme Cycle Agronomie. Option Génie Rural. IAV Hassan IL Rabat (1993). Next page(s) left blank 173 FIELD RESPONSE OF POTATO SUBJECTED TO WATER STRESS AT DIFFERENT GROWTH STAGES M. MOHSIN IQBAL, S. MAHMOOD SHAH, W. MOHAMMAD, H. NAWAZ Nuclear Institute for Food and Agriculture, Peshawar, Pakistan Abstract Yield response of potato to planned water stress was studied in field experiments conducted during 1990-95. There were seven irrigation treatments comprising water stress and nonstress periods imposed at four growth stages. Each irrigation treatment was split into two fertilizer treatments where optimal and suboptimal fertilization levels were applied. The results obtained showed that the timing of water stress influenced the tuber yield differently. The stress imposed at ripening stage caused least reduction in yield whereas that imposed at early development followed by tuber formation stage the greatest reduction. A plot of relative yield reduction versus relative water use deficit revealed that the most sensitive period of growth to water stress, hence most responsive to irrigation, was early development compared to tuber formation and flowering. The traditional irrigation practice led to wasteful water applications. The efficiency of water use was increased by applying deficit irrigation at appropriate growth stages with no adverse effect on yield. The studies with 15N-labeIled fertilizer showed that planting and earthing-up were equally important growth stages for applying N-fertilizer for its efficient utilization. 1. INTRODUCTION Potato (Solanum tuberosum) is a leading food and vegetable crop in Pakistan. Its cultivation is intensive in Northern Punjab and North West Frontier Province. Three crops of potato can be grown in a year, a spring and an autumn crop in the plains and a summer crop in the hilly areas at high altitudes. The crop has attained great importance during the past decade. The area under the crop in Pakistan has increased from 25,000 ha in 1980-81 to 72,000 ha in 1990-91 and production from 0.28 to 0.75 million tons, respectively [1]. The yield, however, has remained stagnant at around 10 ton-ha*1. This low yield is caused, hi part, by improper and unscientific methods of irrigation. The crop is very sensitive to irrigation. The yield [2, 3, 4] keeping quality [5,6] and disease resistance [6,7] are greatly influenced by timing, amount and frequency of irrigation applied. The farmers, on the other hand, apply water to the crop without regard to whether the plant actually needs water at that stage. The objective of this project was to study the relationship between crop yield and crop water use as a function of water stress. Given the yield and crop water use under stressed and nonstressed conditions, the susceptibility of potato to water stress at various growth stages will be determined. The growth stages that can withstand water stress with no significant effect on yield and that are most sensitive to water stress will be identified. 175 The potato crop requires sufficient fertilization to produce high yields. The fertilizer utilization is likely to be influenced by stress irrigation. The effect of deficit irrigation on fertilizer use efficiency and yield will be studied at optimal and suboptimal rates of fertilizer application. 2. MATERIALS AND METHODS 2.1. Experimental Site Field experiments were conducted from 1990-95 at the Experimental Farm of Nuclear Institute for Food and Agriculture (NEFA), Peshawar (34° 4' N, 72° 25' W). The soils of the experimental site were silty clay to clay loam in texture, alkaline in reaction, moderately calcareous, deficient in nitrogen and organic matter and free from salinity. The field capacity and maximum water holding capacity were 25.0 and 52.0%, respectively. The bulk density of the surface soil was 1.7 g-cnr3. 2.2. Crop Three red-skinned varieties of potato were used: Cardinal (in 3 experiments), Ultimus (in 2 experiments) and Désirée (in 1 experiment). Cardinal is an old variety having medium size tubers, red flowers, low response to fertilizers and resistance to high and low temperature and moisture. Ultimus is a medium size variety with pink flowers, high responsiveness to fertilizers and susceptibility to high and low temperature and water. Désirée is a big-sized variety with white flowers, medium responsiveness to fertilizers and susceptibility to high and low temperature and water. The experiments were conducted during spring and autumn seasons. 2.3. Experimental Design There were seven irrigation and two fertilizer treatments replicated four times and laid out according to split plot design with irrigations forming the main plots and fertilizers the subplots, as detailed below: 2.3.1. Irrigation treatments The treatments comprised water stress and nonstress periods imposed at four growth stages. Tl - Full watering at four growth stages; establishment, flowering, tuber formation and ripening. T2 - Continued stress watering at all these growth stages. T3 - Traditional irrigation practice adopted by common farmers of the area. T4 - Same as Tl but one stress watering applied at establishment stage. T5 - Same as Tl but one stress watering applied at flowering stage. T6 - Same as Tl but one stress watering applied at tuber formation stage. T7 - Same as Tl but one stress watering applied at ripening stage. 176 In full watering, full requirements of the crop were met, that is, ETa (actual evapotranspiration) = ETm (maximum evapotranspiration). In other words, water was not a limiting factor in this treatment In stress watering, ETa < ETm. The criterion used for determining the amount of water to be applied was pan evaporation from class A standard evaporation pan. The amount of water applied in full watering equalled the amount of water lost through pan evaporation since the last irrigation which was weighted by a crop coefficient to account for the increasing crop foliage. In the stress treatment, one-half the amount of water for full watering was applied. The amounts of water, calculated as above, were applied to the furrows in between the potato ridges from a reservoir by means of a plastic pipe attached to the water pump. For applying water to T3 (Traditional treatment) some local farmers were invited to the experimental site for consultation on the time and amount of water to be applied. The farmers, as a general practice, apply water to potato at an interval of 10 to 15 days depending upon time of the year. 2.3.2. Fertilizer treatments Each irrigation treatment was split into two fertilization treatments wherein normal and low fertilizer levels were applied. Fl = 250 kg N + 150 kg P2O5 4- 250 kg K20 per ha F2 = 100 kg N + 50 kg P2O5 + 100 kg K2O per ha The fertilizer sources were urea or ammonium sulphate, single superphosphate and potassium sulphate. One-half of the N and all of the P and K were applied at planting, the remaining onehalf of N was applied at earthing-up 35 to 40 d after sowing. 2.3.5. Labelled fertilizer studies For the autumn 1992 experiment, 15N-labelled ammonium sulphate was used to study the effect of normal and stress watering on fertilizer use efficiency. The labelled fertilizer, having 5% 15N atom excess, was applied to two irrigation treatments only, i.e., all-normal (Tl) and continued stress (T2), at the lower nitrogen rate of 100 kg-ha'1 to one out of 6 rows of potato (4.0 x 0.8 m) in each plot The labelled fertilizer was applied in two splits: one-half at planting and one-half at earthing-up time (5-6 weeks after planting), one split in each treatment receiving labelled fertilizer alternatively. The 15N analyses of potato tubers and above ground parts were made in the Siebersdorf Laboratory of the IAEA in Vienna, Austria. 2.4. Instrumentation The experimental plots of 5 out of 7 treatments, in two replicates, of each experiment were thoroughly instrumented with access pipes for neutron moisture probe surrounded by tensiometers at three depths of 15, 30 and 60 cm. Soil moisture content and soil moisture tension were monitored before and after each irrigation, then daily during the first week, on alternate days during the second week and after every third day during third week onward until 177 the next irrigation. These measurements provided basis for determining actual evapotranspiration of the crop using the water balance method. 2.5. Determination of Actual (ETa) and Maximum (ETm) Evapotranspiration The ETa was determined by the water balance method relating the changes in water storage S (mm) in the root zone to differences in water input and output over a specified time period using the relationship S=I+P+C-D where S = [Sfe) - S(t\)] with time t^ (d) being larger than t\, I the amount of irrigation, P precipitation, C capillary rise and D drainage. For estimation of drainage or capillary rise, unsaturated hydraulic conductivity K(6), data are required. The K(Q) of the experimental site was determined by the Libardi method [8]. The ETm was determined using data on reference evapotranspiration [9]. 3. RESULTS AND DISCUSSIONS 3.1. Yield The results (Table I) showed that the highest tuber yield of 14.41 t-ha-1 was produced by continued full watering (Tl) and the lowest of 8.71 t-ha'1 continued stress watering (T2) treatments. The relative reduction in yield over Tl, also known as Crop Susceptibility factor [10], ranged between 4.3 and 39.6%. The stress imposed at ripening stage, T7, caused the least reduction in yield whereas the one imposed at establishment stage led to greatest reduction indicating that early development stages were more sensitive to water stress than the ripening stage. These findings are corroborated by work done elsewhere. Roth [3] reported that critical period for avoiding water stress was flowering to full maturity especially the three weeks after flowering. In India, water stress imposed at stolon initiation stage repressed yield by 30-65% whereas the effect was less pronounced at other stages [11]. In another study, the earliest drought treatment affected the number of stolons per stem depending upon the variety of potato. The later dry period did not affect number of stolons or tubers. The field data of these workers also showed significant linear relationship between number of tubers per stem and total rainfall received during first 40 days [12]. The yield in the present experiment was composed mainly of grade C (< 60 g) and grade B (60-100 g) tubers. The stress at early development stage produced the heighest proportion of grade B tubers whereas continued full watering produced the highest proportion of grade C tubers. The above ground dry matter yield of all the treatments was statistically similar except of T2 which was lower. Per unit of green dry biomass, Tl produced maximum of 18.7 t-ha*1 and T4 minimum of 14.1 t-ha*1. The relative reduction in tuber yield was plotted against relative evapotranspiration deficit (Fig. 1). Their ratio, known as Yield Response Factor ky, helps evaluate susceptibility of 178 different growth stages to water stress. It will be seen that the early development stage was most sensitive to water stress, hence most responsive to irrigation, followed by tuber formation stage. 11.0 0.8 0.6 0.4 0.2 0 0 0.2 - 0.4 0.6 0.8 1.0 Fig. 1. Relationship between relative tuber yield decrease and relative evapotranspiration deficit for potato during autumn 1994. As regards fertilization, the normal recommended level (Fl) produced a significantly higher yield than the lower level (F2). The latter resulted in a 20% reduction in yield over the farmer (Table I). In Salara (Chiniot, Pakistan), potato yield was increased by 14.1% when recommended dose of 250 N + 100 P +100 K kg-ha'1 was used [13]. The dry biomass, weights of A and B grade and number of potatoes were all significantly reduced at the lower fertilization level. The interaction among irrigation and fertilization treatments was also significant which showed that the normal fertilization level produced significantly higher yield compared to the lower level in all the irrigation treatments. 3.2. Water use efficiency While assessing water use efficiency by the irrigated crops, differentiation between water applied and water taken up by the plants is important Field water use efficiency £f relates to the former and crop water use efficiency Ec to the latter. The Ef, defined as yield produced per unit amount of water applied, reflects the characteristics of the irrigation method adopted and is calculated based on the depth of irrigation water applied. The Ec, defined as yield produced per unit amount of water taken up by the crop, shows the ability of the plants to produce good yield with limited supply of water and is calculated on the basis ofETa. The Ef and Ec for the present experiment, based on data in Table ÏÏ, are presented in Table DL It is evident that the highest Ef of 4.65 kg-nr3 was produced in Tl. The Ef of the other treatments except T3 were more or less similar. The Efof T3, where the maximum amount of 179 TABLE I. MEAN TUBER YIELD AND WEIGHT OF ABOVE GROUND DRY MATTER, GRADE A AND GRADE B TUBERS AS INFLUENCED BY IRRIGATION AND FERTILIZER TREATMENTS DURING AUTUMN 1994 Tuber yield Treatment (t-ha-1) Above ground Yield Grade A (>1002) reduction dry matter Weight Number 1 1 (l-Ya-Y - )? (t-ha- ) (t-ha-1) Irrigation treatment Tl 14.41a 0 T2 39.6 8.7 le 4.3 T3 13.79ab , 27.6 T4 10.44d 14.2 12.37bc T5 22.2 11.21cd T6 13.36ab 7.3 T7 Ffrtiliyatinr i treatment 13.36a 0 Fl F2 10.73b 19.7 GradeB(60-100e) (t-ha-1) 0.77a 0.48b 0.83a 0.74a 0.82a 0.74a 0.82a 2.00ab 0.47c 2.46a 1.30c 1.45bc 2.30a 2.20a 32.5ab 7.8c 41.7a 22.2b 25.2b 38.0a 37.7a 4.90b 3.74c 5.97a 5.67ab 5.29ab 6.02a 5.98a O.SOa 0.68b 2.33a 1.15b 40.0a 18.6b 6.32a 4.41b t Ym = maximum yield, Ya = actual yield water was applied, was the lowest of all the treatments indicating some wasteful water application by the traditional practice. The highest Ec of 10.96 kg-nr3 was produced in T2 reflecting the most efficient utilization of water by potato in this treatment. The stress at tuber formation stage led to the lowest EC of 7.5 kg-nr3. In Bangladesh, the average water use efficiency was 60.4,58.4 and 79.5 kg-ha'1 per mm irrigation water for 80, 60 and 40% depletion of available soil moisture [14]. The EC was generally higher in the treatments where a combination of stress and normal waterings was applied compared to when continuous full watering was applied. 3.3. Fertilizer use efficiency The split application of labelled ammonium sulphate at planting and earthing-up stages (Table TV) showed that proportion of nitrogen derived by tubers from the labelled fertilizer was higher (about 18.5%) when the labelled fertilizer was applied at earthing-up time for both Tl (all-normal) and T2 (all-stressed) treatments. The fertilizer nitrogen taken up by tubers was, however, similar (about 0.8 g per row) at both the stages in Tl but higher at earthing-up time in T2. The utilization of fertilizer was also similar at both the stages (5.0%) in Tl but was higher at earthing-up stage in T2. 180 TABLE H. DEPTHS OF IRRIGATION WATER APPLIED AND ETa AT DIFFERENT GROWTH STAGES OF DESIREE POTATO FOR VARIOUS IRRIGATION TREATMENTS, DURING AUTUMN 1994 Treatment General Irrigation Establish Flowering Tuber Ripeninis No.2 stage stage formulation staged No.l 1 2 3 4 5 6 7 8.30 8.30 8.30 8.30 8.30 8.30 8.30 7.50 7.50 7.50 7.50 7.50 7.50 7.50 Depths of water applied (cm) 7.30 4.17 3.71 2.08 1.41 1.56 5.21 5.21 4.20 7.30 4.17 3.71 2.08 1.41 3.71 7.30 4.17 7.30 1.56 7.30 3.71 4.17 Total (cm) m3-ha'1 30.98 20.85 37.72 25.76 28.22 28.83 30.98 3098 2085 3772 2576 2822 2883 3098 17.35 7.95 12.50 14.52 14.78 - 1735 795 1250 1452 1478 Actual evapotranspiration ETa (cm) 1 2 3 4 5 6 7 Ndtt Nd Nd Nd Nd - Nd Nd Nd Nd Nd - 7.23 3.14 2.91 7.36 6.96 - 4.92 2.86 5.23 2.86 5.52 - 5.20 1.95 4.36 4.30 2.30 - t The plots were not irrigated because of the onset of frost tt Not determined. TABLE m. FIELD WATER USE EFFICIENCY Ef AND CROP WATER USE EFFICIENCY EC OF DIFFERENT IRRIGATION TREATMENTS DURING AUTUMN 1994 Treatment Tl T2 T3 T4 T5 T6 T7 Tuber yield (t-ha-1) Depth of water applied (m3-ha-1) ETa 14.41 8.71 13.79 10.44 12.37 11.21 13.36 3098 2085 3772 2576 2822 2883 3098 1735 795 1250 EC (kg-m-3) 1452 1478 _ 4.65 4.18 3.65 4.05 4.38 3.89 4.31 (kg-m-3) 8.31 10.96 8.35 8.52 7.58 _ 181 TABLE IV. EFFECT OF SPLIT APPICATION OF 100 kg N-ha"1 AS AMMONIUM SULPHATE TO POTATO AT TWO STAGES ON NITROGEN DERIVED FROM FERTILIZER (Ndff), LABELLED FERTILIZER UPTAKE AND FERTILIZER UTILIZATION IN ALL-NORMAL Tl AND ALL-STRESS T2 IRRIGATION TREATMENTS Fertilizert splits SI S2 Ndff (%) Tl 50* 50 50 50* 14.2 18.7 50* 50 50 50* 17.8 17.7 50* 50 50 50* - N uptake (g per row) T2 Tl T2 FerLN uptake f g per row) Tl Potato Tubers 0.82 4.02 14.5 5.77 0.80 4.26 18.3 3.76 Above Ground Dry Matter 13.9 3.37 3.19 0.60 0.48 16.8 2.73 1.78 Total (Tubers and Above Ground Parts) 1.42 9.14 7.21 6.99 5.54 1.28 Fert.Utilization T2 Tl T2 0.58 0.69 5.13 5.00 3.63 4.31 0.44 0.30 3.75 3.00 2.75 1.88 1.02 0.99 8.88 8.00 6.38 6.19 t SI is planting time and S2 is earthing-up time. * Labelled with 15N. In India, splitting of N 50% at planting, 25% at emergence and 25% four weeks after emergence produced the highest tuber yield of 29.0 t-ha'1 [15]. The above ground dry matter behaved differently with respect to fertilizer N uptake and fertilizer utilization; these parameters were lower at S2, i.e., when labelled fertilizer was applied at earthing-up stage. The tubers stored more nitrogen than the above ground dry matter in both the treatments at harvest. The continued stress treatment reduced the fertilizer N uptake and fertilizer utilization by tubers and above ground biomass. The overall nitrogen utilization was less than 10% which is very low. The major reasons for this may be the late application, by about two weeks, of the fertilizer, and application of fertilizer at the base of the potato ridge where it was highly prone to leaching beyond the root zone through irrigation water. It can be concluded from the above results that earthing-up was as important if not better time than that of planting for N application to potato. 3.4. Potato Varieties Three varieties of potato (Cardinal, Ultimus and Désirée) were used for different field experiments. Their yields, averaged over one to three seasons, are presented in Table V where it 182 can be seen that their responses to irrigation treatments were generally similar. The maximum yield of these varieties, however, did not exceed 15 t-ha'1 in any experiment. This yield is considerably low compared to those of other potato growing countries and shows that these varieties were not suited to the extreme climatic conditions of this region [16]. TABLE V. AVERAGE YIELD (t-ha'1) OF THREE VARIETIES OF POTATO AS INFLUENCED BY IRRIGATION AND FERTILIZER TREATMENTS Treatment 4. Cardinal Tl T2 T3 T4 T5 T6 T7 7.80 4.35 6.04 5.04 6.61 5.44 7.09 Fl F2 6.44 5.36 Variety Ultimus Irrigation treatments 7.68 5.17 6.78 5.64 6.01 6.04 6.53 Fertilizer treatments 6.41 6.13 Désirée 14.41 8.71 13.79 10.44 12.37 11.21 13.36 13.36 10.73 CONCLUSIONS As expected, the highest tuber yield was produced by continuous full watering and the lowest by continued deficit watering. The relative yield reduction, or the crop susceptibility factor, varied between 0.04 and 0.54. A plot of relative yield reduction versus relative water use deficit revealed that early development was the most sensitive stage to water stress, hence most responsive to irrigation, whereas ripening was the least sensitive stage to water stress. The prevalent practice of common fanners of applying water at 10 to 12 d intervals without regard to actual needs of the plant is not efficient, it leads to lower yields with greater amounts of applied water. By applying less water at appropriate growth stages, the same or even better yields could be obtained. The suboptimal fertilization level caused 20% reduction in yield over the optimal level. Planting and earthing-up were found to be equally important times for fertilizer application to obtain higher yields and efficient fertilizer utilization. 183 ACKNOWLEDGEMENTS The work reported was conducted under an International Atomic Energy Agency Research Contract No. 6410/RB. We are thankful to the Agency for providing some essential pieces of equipment, the labelled ammonium sulphate fertilizer for field studies and for analyzing the plant samples for 15N assay. We are also thankful to Pakistan Atomic Energy Commission for allowing us to undertake work under this contract REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] GOVERNMENT OF PAKISTAN, Agricultural Statistics of Pakistan, 1991-1992. Ministry of Food, Agriculture and Cooperatives, Govt of Pakistan, Islamabad (1992). BARTOSZUK, W., Decrease in potato yield resulting from water deficit during the growing season. Potato Abst.15 1 (1987) Abs. No. 194. ROTH, R., The influence of variation in water supply at individual growth stages on yield of silage maize (Zea mays L.) and potato (Solarium tubersoum L.). Potato Abst. 15 1 (1990) Abs. No. 193. TREBEJO, I., MIDMORE, D.J., Effect of water stress on potato growth, yield and water use in a hot and a cool tropical climate. J. Agr. Sei 114 3 (1990) 321-334. VETTER, A., SCHMIDT, H., Influence of variation in soil moisture on the yield, quality and storability of seed and culinary potato. Potato Abst. 15 4 (1990) Abs. No. 956. CARR, M.K.V., Potato quality control with irrigation. Water and Irrigation Rev. 9 (1989) 28-29. MARUTANI, M., GUZ, F., Influence of supplemental irrigation on development of potatoes in the tropics. Hort. Sei. 24 6 (1989) 920-923. LffiARDI, P.L., REICHARDT, K., NIELSEN, D.R., BIGGAR, J.W., Simple field methods for estimating soil hydraulic conductivity. Soil Sei. Soc. Am. J. 44 (1980) 3-7. DOORENBOS, I, KASSAM, A.H., et al., Yield Response to Water. FAO Irrigation Drainage paper 33, FAO, Rome (1986). [10] [II] [12] [13] 184 KIRDA, C, ANAC, S., TULUCU, K., GUNGOR, H., Deficit irrigation in semi-arid zone irrigation projects: Case studies in Turkey. In Nuclear methods in soil-plant aspects of sustainable agriculture, IAEA-TECDOC-785 (1995) 183-194. MINHAS, J.S., BANSAL, K.C., Tuber yield in relation to water stress at stages of growth in potato {Solomon tuberoswn L.). J. Indian Potato Assoc. 18 1-2 (1991) 1-8. HAVERKORT, A.J., VAN DE WAART, M., BODLAENDER, K.B.A., The effect of early drought stress on number of tubers and stolons of potato in controlled and field conditions. Pototo Res. 33 (1990) 89-96. PARC. Potato - Fertilizer use. In PARC Annual Report 1992. Pakistan Agricultural Research Council, Islamabad (1993) pp 30. [14] ISLAM, T., SARKER, H., ASLAM, J., HARUN-UR-RASHIED, Water use and yiled relationships of irrigated potato. Agricultural Water Mgt, 18 (1990) 173-179. [15] SHARMA, S. P., SHEKHAR, J,. Effect of split application of nitrogen on yield of potato (Solanum tuberosum). Indian J. Agr. Sei. 18 (1989) 173-79. LEVERY, D., GENIZ3, A., GOLDMAN, A., Compatibility of potatoes to contrasting seasonal conditions, to high temperature and to water deficit: The association with time of maturation and yield potential. Potato Res. 33 (1990) 325-334. [16] Next page(s) left blank 185 SOME STUDIES ON PRE-PLANNED CONTROLLED SOIL MOISTURE IRRIGATION SCHEDULING OF FIELD CROPS R.A. WAHEED, M.H. NAQVI, G.R. TAfflR, S.H.M. NAQVI Nuclear Institute for Agriculture & Biology, Faisalabad, Pakistan Abstract The study aimed at improving the conventional irrigation management practices to enhance yield and water use efficiency for pre-planned irrigation scheduling of wheat and cotton crops. Five field experiments were conducted during 1990-94. A 3-year study of moisture deficit irrigations (MDI) to wheat V-85205 was continued with the same irrigation schedule(s) for two years. The third year irrigation schedule(s) were modified on the basis of the preceding year's results. The 3-year study indicated that the crop was most sensitive to moisture deficit at tillering stage and least sensitive at flowering stage. In the fourth experiment, genetic diversity of wheat to moisture deficit was investigated. Three pre-selected wheat genotypes; Sarsabz, LU-26S and Pasban-90 were imposed to moisture deficit at different stages. The genotypes showed different response to moisture deficit Comparable yields to respective conventional irrigation schedule (1111) were obtained by MDI schedules (1011), (1110) and (1101) for Sarsabaz, LU-26S and Pasban-90, respectively. The fifth experiment was conducted on the genetic diversity of cotton to moisture deficit Two pre-selected cotton genotypes NIAB-86 and FH-682 were imposed to moisture deficit at vegetative, generative or maturity stages. NIAB-86 yielded 7% more seed cotton at MDI (110) and FH-682 yielded 9% excess over conventional irrigation treatments saving 150 mm of irrigation water. Thus, saving of 75 and 150 mm of irrigation water were achieved by applying improved irrigation schedule without undergoing any significant yield loss. 1. INTRODUCTION In arid and semi-arid regions of the world the rainfall is far less than evapotranspiration of field crops leading to moisture deficit The moisture deficit during critical crop growth stages adversely affected wheat growth and production by inhibition of chlorophyll activity [1]. During heavy fruiting, mild water stress associated with long irrigation cycles triggers deterioration of the root system of cotton that is very slow to be reversed [2]. High soil matric potential survey inhibits root growth [3]. Therefore, irrigations are applied to field crops to increase yield and improve quality of protein [4]. Moisture deficit irrigations are often recommended to (i) minimize irrigation inputs without affecting the season total biomass yield, (ii) control the weed stand in crop [5] and (iii) increase water use efficiency. The present studies were conducted with the following objectives: - to improve traditionally adapted irrigation practices for efficient utilisation of water for crop production, - to pre-plan the irrigation scheduling for field crops by identifying specific crop growth stages sensitive to moisture deficit, - to study genetic diversity of crops to moisture deficit. 187 2. MATERIALS AND METHODS 2.1. Location of research station Experiments were conducted in a field site 200 m above sea level at the Nuclear Institute for Agriculture and Biology, Faisalabad, Pakistan located at latitude of 31°-26' N and longitude of73°-26'E. 2.2. Climate, soil and irrigation water The experimental station is situated in a semi-arid zone with 350 mm of mean annual rainfall and around 1650 mm-y1 class A pan evaporation. The monsoon rains during the months of July and August share about 60% of total annual rainfall. After the monsoon rains generally the field soils attain field capacity up to a limited depth depending upon the residence time of rain water and soil moisture characteristics. The mean maximum temperature varies from 19.5 to 43.1°C in January to June. The mean minimum temperature varies from 4.4 to 28.1°C in January to July. The soils are from Paleistocene to early Holocene age and is formed in loamy textured mixed Himalayan alluvial deposits. The soil profile is moderately developed as may be concluded from the Cambic to Argellic horizon. The soil description is shown in Table I. The top soil is slightly lacking structure. From 15 to 120 cm the profile is more or less homogeneous showing only some gradual increase in lime content. From 130 to 180 cm the Ci horizon consists of sandy loam layer which probably has different hydraulic properties. The €2 horizon from 180 to!95 cm has finer textures which may retard drainage. The soil bulk density is 1.2 g-cnr3 at 15 cm and 1.3 to 1.32 g-cnr3 at a depth of 110 cm. The soil organic matter is TABLE I. SOIL PROFILE DESCRIPTION OF EXPERIMENTAL SITE Depth (m) Horizon 0-0.15 0.15-0.95 0.95-1.20 1.20-1.80 1.80-1.95 1.95-2.10 2.10-2.40 2.40-2.55 2.55-3.10 >3.10 Ap Bw Bwk Ci C2 C3 188 Profile description Loam massive structure Loam week structure Loam with fine nodules (lime),week structure Very fine sandy loam, massive Silt loam; close to silry clay loam Very fine sandy loam Fine sandy loam Loamy fine sand Fine sand Medium sand less than 1%, total soil N is less than 0.05% and Soil EC isl to 1.5 dS-nr1, pH is 7.9 to 8.1 and SAR is less than 5. Two irrigation water sources were used - canal water (EC = 0.3 dS-nr1, pH = 7.9, SAR < 3) and during canal closure the subsoil pump water (EC = 1.1 dS-nr1, pH = 7.9 and SAR = 3) 2.3. Soil hydraulic characteristics The soil hydraulic characteristics were determined using combination of Neutron Hydroprobe No. CPN-503 and soil tensiometers. The neutron hydroprobe calibration curve for the 2-m soil profile was 0 = -9.8 + 40.93/ where 0 represents volumetric moisture content and/is the neutron count ratio. The goodness of the linear regression r2 between the calculated and observed moisture content values was 0.984. The soil hydraulic characteristics were calculated by a parameter optimization method [6,7]. The parameters determined by RETC and SFTT computer programs are given in Table u. 2.4. Water use efficiency The reference evapotranspiration ET0 was calculated according to Penman-Monteith. Soil moisture depletion was monitored from the neutron access tubes installed in the soil down TABLE H. VAN GENUCHTEN-MUALEUM EQUATION PARAMETERS (AVERAGE) DETERMINED BY SFIT AND RETC COMPUTER PROGRAMS FOR THE EXPERIMENTAL SITE AT NIAB, FAISALABAD, PAKISTAN Depth (cm) a n A (cnr1) Ks 6r 0S 1 3 3 3 (cm-hr ) (cm -cnr ) (cm -cm-3) r2 SFTT program 25 50 75 100 0.022 0.020 0.013 0.007 1.853 1.429 1.462 2.030 1.37 2.623 0.133 0.44 0.109 1.426 0.504 0.069 0.903 2.075 0.768 0.076 RETC program 0.344 0.354 0.360 0.367 0.993 0.999 0.998 0.999 25 50 75 100 0.018 0.008 0.011 0.009 1.763 1.533 1.716 1.949 5.774 0.0001 0.732 0.0001 0.340 0.349 0.358 0.367 0.855 0.915 0.966 0.952 6.761 0.851 0.404 0.815 0.135 0.000 0.158 0.187 189 to 1 ra for wheat and 1.65 m for cotton. Soil moisture tensiometers were installed in the root zone of the pre-selected irrigation treatment plots. Actual evapotranspiration of the crop ETa was determined using water balance method considering irrigation, rainfall, soil moisture depletion and drainage and runoff being nil in the plots. Actual grain yield Ya (kg-ha-1), maximum grain yield Ym (kg-ha'1), Irrigation /.(mm-period'1), maximum crop evapotranspiration ETm, field water use efficiency jE^[kg-(ha-m3)4] and crop water use efficiency EC [kg^ha-m3)'1] were each determined. The yield response factor ky was calculated using 2.5. Irrigations Conventional flood irrigations were applied with 75 to 100 mm of irrigation water when available water AW was within the range of 60-90% and deficit irrigations were applied when the AW was at 30-60%. The CSM treatments were maintained at maximum soil matric potential of -50 kPa and the amount of irrigation water was calculated on the basis of effective root zone depth. The irrigation schedule and moisture deficit of treatments is given in Table HI. 2.6. Fertilizer Fertilizers were applied to all treatments in split doses as given in Table IV. 2.7. Insecticide/pesticide application Polytrin-C, Novacron, Curacron and Thiodan were sprayed on the cotton crop when the pest population reached economical level according to recommendations from entomologists at NIAB. 2.8. Agronomy All experiments were conducted under normal agronomic practices. Sowing of all wheat experiments was conducted in mid-December each year and cotton experiment was started in the first week of June. Pre-sowing irrigation was applied to all experiments except wheat in 1991-92 when the soil had adequate moisture after a rice crop. The seed rate for wheat was 90 (kg-ha-1) and cotton 16 (kg-ha-1) of delinted cotton. The cotton plants were thinned to maintain an interrow distance of 75 cm and an inter-plant distance of 25 to 30 cm. The main plot size of all treatments was 10 x 7 m. Both crops were harvested at physical maturity. 2.9. Statistical analysis All experiments were conducted in a randomized complete block design with five replications except experiment No. 3 where three replications were maintained. Experiments No. 1 to 3 were conducted in split plot design with irrigation in the main plot and fertilizer levels in subplots. In experiment No. 4 the split plot design involved wheat genotypes in the main plots 190 and irrigations in the subplots. In experiment No. 5 the irrigations were placed in the main plot and cotton varieties in the subplots. Data were subjected to analysis of variance followed by Duncan's New Multiple Range Test [8]. All results were compared at P = 0.05. TABLE m. IRRIGATION SCHEDULE OF TREATMENTS CODEt Description Wheat 1111 0000 CSM 1100 0011 1001 1011 0111 1101 1110 CSM Ill 000 101 110 011 100 010 Conventional four flood irrigations at tillering, booting,flowering and grain filling. No irrigation up to 30% AW; rain fed. Maintaining maximum soil matric potential to -50 kPa Moisture deficit after booting. Moisture deficit up to booting. Moisture deficit at booting to flowering. Moisture deficit at booting only. Moisture deficit at tillering only. Moisture deficit at flowering only. Moisture deficit at grain filling only. Cotton Maintaining maximum soil matric potential to -50kPa. Conventional flood irrigations at vegetative, generative and maturity stages. No irrigation up to 30% AW; rain fed. Moisture deficit at generative stage. Moisture deficit at maturity stage. Moisture deficit at vegetative stage. Moisture deficit at generative to maturity stages. Moisture deficit at vegetative and maturity stages. 10 designates moisture deficit, and 1 designates irrigated. 3. RESULTS AND DISCUSSIONS 3.1. Yield and irrigations 3.1.1. Wheat experiment 1991-92 The experiment, with a pre-selected wheat genotype V-85205, was conducted on a rice field without a pre-sowing irrigation. The crop received 87 mm of rainfall well distributed over 191 TABLE IV. FERTILIZER APPLICATION TO WHEAT AND COTTON CROPS Fertlilizer input Medium dose NPK (kg-ha-1) Low dose NPK (kg-ha-1) Wheat Basal Dose With Irrigation 50:100:60 50:0:0 Cotton Basal Dose 1st Irrigation 3rd Irriggation 23:60:60 60:0:0 60: 0:0 25:50:30 25:0: 0 crop stages n to IV. The temperature cycle and humidity was normal. Maximum grain yield (Table V) was observed at Tl(1111), the conventional flood irrigation treatment. At the low fertilizer level maximum grain yield was the same for T7(1101), Tl(llll) and T2(1100) showing that at the lower fertilizer level even higher irrigation inputs could not be duly beneficial. Minimum grain yield was produced in T9(0000) - rain fed, as expected. Comparing two irrigations in Fig.l, T2(l 100) produced maximum and T3(0011) minimum grain yields. In the former case, 87 - 84% excess grain yield was produced applying the same quantity of irrigation water at the two fertilizer levels, respectively. This yield variation confirms those [9] who reported 65% loss in grain yield owing to moisture deficit at crop stages I and II. This shows that the same irrigation water flux if applied at the earlier crop stage lead to consumptive use of water, compared to later stages. Among three irrigation treatments the maximum grain yield was produced in T8(l 110) followed by T7(l 101) and minimum yields in T6(0111) at both fertilizer levels. Shifting moisture deficit from crop stage ICE to IV did not affect the contribution of grain yield significantly at the medium fertilizer levels. At the low fertilizer level T7(1101) clearly out yielded the other moisture deficit treatments and the grain production was even better than the 4-irrigation treatment Tl(llll). It showed that with one irrigation at any later stage could be saved without significant losses in grain yield. Similarly gram yield in T4 (1001) and T5(1011) was the same, again indicating that the irrigation at crop stage in did not contribute significantly to the wheat grain production. Minimum £/was observed at T3(0011) and T6(0111) - both missing an essential irrigation at crop stage I. £/ decreased with reduced fertilizer input within an irrigation but increased when exposed to moisture deficit for a prolonged period. It might be concluded that the moisture deficit at the later stages was not as deterimental towards grain yield as at earlier stages. 192 TABLE V. GRAIN YIELD Ya, IRRIGATION 7, EVAPOTRANSPIRATION ETa, WATER USE EFFICIENCIES Ef AND Ec AND YIELD RESPONSE FACTOR ky EC ky Treatment Irrigation Fertilizer Ya I Ef ETa 4 3 4 3 4 number treatment level (kg-ha ) (mm) [kg^na-m ) ] (mm) [kg-Oia-m ) ] Wheat 1991-92 Tl 1111 1 T2 1100 T3 0011 T4 1001 2 1 2 1 2 1 1011 2 1 T5 T6 0111 17 1101 T8 1110 T9 0000 Tl 1111 T2 1100 T3 0011 T4 1001 T5 1011 T6 0111 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 13.2 5131 387 10.4 387 4023 4942 237 20.8 3945 237 16.6 2633 237 11.1 2154 9.1 237 4501 237 19.0 3692 237 15.6 312 4377 14.0 3702 312 11.8 312 3533 11.8 2904 312 9.3 312 4738 15.2 4102 312 13.1 4773 312 15.3 3869 312 12.2 32.1 2603 81 1884 81 23.2 Wheat 1992-93 360 349 305 290 276 257 290 280 332 315 325 320 327 325 331 309 218 170 . 14.2 11.5 16.2 13.6 9.5 8.4 15.5 13.2 13.2 11.7 10.9 9.1 14.5 12.6 14.4 12.4 11.9 11.0 0.27 0.23 2.13 1.81 0.63 0.50 1.87 1.00 3.10 3.62 0.89 0.00 0.87 0.64 1.26 1.06 3911 3159 3397 3032 2633 1649 3404 2611 3496 2952 2688 394 379 263 253 264 235 284 258 316 303 314 9.9 8.3 12.9 12.0 10.0 7.0 12.0 10.1 11.1 9.7 8.6 _ — 0.54 0.24 1.12 1.31 0.64 0.62 0.80 0.50 1.75 341 341 191 191 191 191 191 191 266 266 266 11.5 9.3 17.8 15.9 13.8 8.6 17.8 13.7 13.1 9.7 10.1 _ - 193 TABLE V. Continued 2 1 2 1 2 1 2 8.6 2292 266 266 15.7 4170 12.3 3279 266 13.9 3667 266 266 12.1 3120 41 2087 51.0 36.8 1509 41 Wheat 1993-94 311 330 326 324 316 143 137 7.4 12.6 10.1 11.4 9.9 14.6 11.0 1.67 0.00 0.00 0.61 0.29 0.78 0.84 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 6321 414 15.3 14.2 414 5860 341 5662 16.6 5014 341 14.7 41 101.1 4132 41 75.7 3103 41.6 116 4831 4425 116 38.1 191 4990 26.1 23.9 191 4547 23.2 4412 191 21.7 191 4143 29.3 191 5605 23.9 4561 191 21.1 5619 266 18.0 4778 266 Wheat genotype Sarsabz 394 371 382 360 189 174 242 229 248 236 252 241 272 270 318 313 16.0 15.8 14.8 13.8 21.8 17.8 . 19.9 19.3 20.1 19.3 17.5 17.2 20.6 16.9 17.7 15.3 _ 3.33 4.67 0.67 0.89 0.60 0.63 0.57 0.61 0.83 0.83 0.35 0.81 0.58 1.06 T7 1101 T8 1110 T9 0000 Tl CSM T2 1111 T3 0000 T4 1000 T5 1001 T6 0101 T7 1100 T8 1101 Tl T2 T3 T4 T5 T6 CSM 1111 0000 1011 1101 1110 18.2 5561 306 15.0 341 5113 41 76.3 3130 19.8 5272 266 18.8 5011 266 19.2 266 5107 Wheat genotype LU26S 355 335 167 310 307 316 15.7 15.3 18.7 17.0 16.3 16.2 — 1.43 0.83 0.41 0.73 0.74 Tl T2 T3 CSM 1111 0000 5540 5106 3722 18.2 15.0 90.8 343 327 177 16.1 15.6 21.0 — 1.67 0.68 194 306 341 41 TABLE V. Continued T4 1011 T5 1101 T6 1110 4893 4565 5174 266 266 266 18.4 21.3 19.4 289 273 304 16.9 16.7 17.0 0.74 0.86 0.58 329 308 172 270 276 290 15.6 15.2 18.3 17.9 18.1 16.3 — 1.32 0.81 0.32 0.16 0.63 762 716 427 557 562 561 559 548 4.08 4.10 4.93 4.32 5.59 4.53 5.03 5.67 — 0.95 0.75 0.85 0.00 0.73 0.37 0.03 787 720 412 559 550 563 562 545 3.94 3.84 5.83 4.97 5.50 4.15 5.19 5.13 — 1.37 0.46 0.34 0.10 0.86 0.21 0.32 Wheat genotype Pasban-90 Tl T2 T3 T4 T5 T6 CSM 1111 0000 1011 1101 1110 5121 4690 3148 4830 4987 4740 306 341 41 266 266 266 16.7 15.2 76.8 18.1 18.7 17.8 Cotton genotype NIAB-86 Tl T2 T3 T4 T5 T6 17 T8 CSM 111 000 101 110 Oil 100 010 3112 2934 2106 2408 3144 2544 2814 3110 687 612 237 462 462 462 387 387 4.5 4.8 8.9 5.2 6.8 5.5 9.0 10.0 Cotton genotype FH-682 Tl T2 T3 T4 T5 T6 T7 T8 CSM 111 000 101 110 Oil 100 010 3102 2768 2404 2776 3024 2336 2916 2798 687 612 237 462 462 462 387 387 4.5 4.5 10.1 6.0 6.5 5.0 9.3 9.0 195 M[EDIUM FERTILIZER E - z o O U LOW FERTILIZER O " O Q CJU| m a tu Cd 8 a Z i Q O - < 33 ^ Rl 3 IRRIG. 2 IRRIG. 3 IRRIG. 2 IRRIG. Fig. 1. Wheat grain yield under moisture deficit irrigation during 1991-1992 for medium and low fertilizer regimes. Missing irrigations during tillering, booting, anthesis and grain filling stages are designated by A, B, C and D, repectively, with the least significant difference being designated as LSD. 3.1.2. Wheat experiment 1992-93 The experiment was conducted on Mung bean (Vigna radiata) fields applying a presowing irrigation for land preparation. During the year 1992-93 the crop season was relatively dry and hot with a high temperature cycle of 4.2°C (average) and no rain fall at DAS 41 to 60. The total rainfall of the season was 41 mm of which 35 mm was received in crop stage lu. The generative processes started 10 to 15 d earlier reducing the vegetative growth duration of crop. The overall grain yield (Table V) decreased compared to that of the previous year. This yield decrease was in accordance with the finding of [10] who reported an average grain yield loss of 428 kg-flia-C0)'1 with temperature during vegetative growth period in this region. Maximum grain yield was produced in T7(1101) at both fertilizer levels (Fig.2) with 6.6 to 4.0% increase over Tl(llll) - the conventional flood irrigation treatment saving at least 75 mm irrigation water. The lowest grain yield was produced in the rain fed treatment. Comparable grain yield at V 1 MEDIUM FERTILIZER LOW FERTILIZER 1-< 4 - M £ 3 .\ 1 -1 r- &3 2 N* A Z M K 1 § oQ i- a £ 3 < z Û U 3 IRRIG. 2 IRRIG. Q U u 2 - 8 § n 0 n TVENTIONAL I ^ a --, Q a < O 3 g a n * 3 IRRIG. 2 IRRIG. Fig. 2. Wheat grain yield under moisture deficit irrigation during 1992-1993 for medium and low fertilizer regimes. Missing irrigations during tillering, booting, anthesis and grain filling stages are designated by A, B, CandD, repectively, with the least significant difference being designated as LSD. 196 medium fertilizer levels were observed in T2(1100), T4(1001) and T5(1011) all involving irrigation at crop stage I. The later treatments did not differ significantly from each other showing no contribution of irrigation to grain yield at crop stage HI. Two stages of irrigation treatments showed grain yield increase of 29% and 84% in T2(1100) over T3(0011) at two fertilizer levels, respectively, applying the same quantity of the irrigation water. Similar results were reported by [11]. Thus, irrigating this wheat variety at early stages was more productive than irrigating at late stages as observed last year. The treatments with and without irrigation at crop stage I could be separated into different groups with 30 to 55% variation in grain yield. The Ef\vas maximum under rainfed conditions followed by T2(1100), T4(1001) and T7(l 101). The conventional flood irrigation treatment Tl(l 111) could be ranked as lowest efficiency group at both fertilizer levels. The £f decreased with reduction in fertilizer inputs within irrigations. 3.1.3. Wheat experiment 1993-94: The experiment was conducted on the field left fallow from monsoon rain up to wheat sowing in 1993-94. A pre-sowing irrigation was applied to facilitate land preparation. The crop observed a normal season with respect to temperature and rainfall (42 mm) over crop stages n to IV. The experiment was conducted with irrigation schedule modified on the basis of results obtained from the previous experiments of 1991-93. Adequate soil moisture in the soil profile at the crop stage I and uniformly distributed rainfall led to a good harvest. Maximum grain yield (Table V) produced at Tl(CSM) - no stress treatment i.e., 12 to 17% higher over conventional flood irrigation treatment T2(llll). Minimum grain yield was produced under rainfed conditions in T3(0000) with 35 and 47% loss at the two fertilizer levels, respectively. Comparable grain yield (Fig.3) produced from T2(llll), T7 (1100) and T3(1101) all were irrigated at early crop stages at both fertilizer levels, Efwas maximum in T3(0000) followed by o 3: a MEDIUM FERTILIZER LOW FERTILIZER ^ 1 - -, p ÜD 4 S Cd -, 2 1 o u Z . >_ 2 is s oi > —i o CJ Q U Q O tu CQ < a 0 3 IRR. 2 IRR. l IRR. 3 IRR. 2 IRR. l IRR. Fig. 3. Wheat grain yield under moisture deficit irrigation during 1993-1994 for medium and low fertilizer regimes. Missing irrigations during tittering, booting, anthesis and grain filling stages are designated by A, B, CandD, repectively, with the least significant difference being designated as LSD. 197 T4(1000). All deficit irrigation treatments observed higher Ef than those for Tl(CSM) and T2(lll) controls. The 3-year study on the same variety during three different seasons showed that early crop stages of wheat were more sensitive to drought. An irrigation at crop stage HI did not contribute significantly to the total variation in grain yield. 3.1.4. Genetic diversity of wheat to moisture deficit This experiment was conducted in 1993-94 crop season to verify the results obtained during 1991-94 experiments on a wheat genotype. Three pre-selected wheat genotypes (Sarsabz, Lu-26S and Pasban-90) were exposed to moisture deficit irrigations as per irrigation schedule given in Table lu. Under rainfed conditions, 33 to 44% yield loss occurred (Table V). Different pattern of stage sensitivity (Fig. 4) was observed in the three wheat genotypes. Sarsabaz in T4(1011) and T6(1110), LU-26S in T6(1110) and Pasban-90 in T5(1101) with the moisture deficit irrigation treatments produced wheat grain comparable to T2(l 111). Thus, at least 75 mm of irrigation water was saved without affecting the ultimate yield. Under rainfed conditions LU26S produced up to 12% higher grain yield than those of Sarsabz and Pasban-90. The field water use efficiency fy was maximum with rainfed irrigation treatment T3(0000) in all wheat varieties. Generally, the Ef was lower in T2(l 111) than all other irrigation treatments. LU-26S ï fi S 2— ^* Z £ On PASBAN-90 4 -- -~_ * § 2 m | | ü à r-] . V 8 ^ a 2 l IRRIG. p. - 2- 1 a 2 -- 1 o Ü CQ u ÇÛ - 4 - 3 - ENTION SARSABZ ^ O n l IRRIG. a (J 8 "^ ! CQ 0 § l IRRIG. Fig. 4. Grain yield of three wheat genotypes under moisture deficit irrigation during. Missing irrigations during tillering, booting, anthesis and grain filling stages are designated by A, B, C andD, repectively, with the least significant difference being designated as LSD. 3.1.5. Genetic diversity of cotton to moisture deficit A field experiment was conducted in field of wheat Pre-sowing irrigation was applied for land preparation. The crop observed a normal season with respect to meterology. A total of 164 mm of rain fall was received; 80 and 83 mm received at the generative and maturity stages, respectively which reduced the deficit period. Two pre-selected cotton genotypes, NIAB-86 and FH-682 were exposed to moisture deficit irrigations. Both varieties responded differently to moisture deficit. Maximum seed cotton yield (Table V) of both genotypes was observed in treatment Tl(CSM) and minimum in T3(000) under the rainfed conditions. However, the yield under rainfed conditions was higher than expected owing to favorable climatic conditions that 198 - prevailed during crop stages II and IÏÏ. The rainfall of 80 and 83 mm was probably used most efficiently under T3(000). Irrigation treatments T3(000) and T6(011) were the lowest yielding. T1(CSM) and T5(l 10) were not significantly different from each other indicating that irrigation at vegetative and generative stages was efficiently utilized. The seed cotton yields of NIAB-86 in treatments Tl(CSM), T2(III), T5(110) and T8(010) were not significantly different from each other (Fig. 5). The moisture deficit treatments with and without irrigation at crop stage n differed by 23% employing the same quantity of irrigation water. Similar results were reported by [2]. T5(110) and T8(010) produced comparable yields by employing 300 and 150 mm of irrigation water, respectively. These results showed that irrigation at crop stage n was most contributive to seed cotton yield of this variety. T5(110) yielded 23% and 12% more than T6(011) and T7(100), respectively. The over all order of contribution of irrigation to the seed cotton yield of NIAB-86 was Stage n > Stage I > Stage ffl. For variety FH-682, yields from T5(l 10) and T7(100) did not differ significantly from each other with each receiving an essential irrigation at the vegetative stage. The treatments, with and without irrigation at vegetative stage, differed by 29% employing the same quantity of irrigation water. Similarly T7(100) utilizing 150 mm less irrigation water produced 26% more seed cotton than did T6(011). In T6(011), the irrigation at crop stage HI rather lowered the seed cotton yield owing to the initiation of revegetation processes which probably limited the photo-synthate material transfer to cotton bolls. Thus, the order of contribution of irrigation to seed cottton yield for FH-682 variety was Crop stage I > Crop stage n > Crop stage ffl. 3.2. Water use efficiency EC and yield response factor ky 3.2.1. Wheat crop 1991-92 Actual water use efficiency EC of crop was maximum (Table V) in T5(1001) followed by T7(l 101) and T8(l 110). Lowest Ec was observed in T3(0011) and T6(0111) in which both missed an irrigation at crop stage I. The EC values of 17(1101) and T8(l 110) were comparable u V NIAB-86 FH-682 G S 4 *" ~ ta n - , o u S ag < oa O - 2 - 3 VENTIONAL I 8 la A "" ta VENTIONAL 1 5Ç «^^ P 8 U ua ça o - Z u5 n l IRRIG. A 2 IRRIG. 8 n l IRRIG. 2 IRRIG. Fig. 5. Seed cotton yield of two genotypes under moisture deficit irrigation during 1994. Missing irrigations during vegetative, generative and maturity stages are designated by A, B and C, repectively, with the least significant difference being designated as LSD. 199 at both fertilizer levels showing that moisture deficit at irrigation stage in and IV had similar effects. EC was maximum in T2(l 100) and minimum in T3(0011) showing that water was more efficiently utilized at earlier stages. The yield response factor ky was lowest in T2(1100) and Tl(l 111) while it was maximum in T6(0111) indicating maximum sensitivity to moisture deficit at crop stage I (tillering). At the low fertilizer level ky was minimum for T7(l 101) comparable to the conventional treatment Tl(llll) employing 25% less irrigation water. Within irrigation treatments ky increased with medium fertilizer inputs as compared to lower doses. These results confirmation those of [12] who reported that higher N fertilizer application enhanced evapotranspiration. 3.2.2. Wheat crop 1992-93 Highest water use efficiency EC was observed (Table V) in treatment T9(0000) followed by T2(1100), T7(1101), T4(1001), T8(1110) and T5(1011) all involving an irrigation at crop stage I. Lower EC values were observed in Tl(l 111), T3(0011) and T6(0111) owing to either over-irrigation or missing an irrigation at crop stage I. EC increased in the irrigation treatments with higher fertilizer inputs. The overall EC decreased owing to high evaporation as is obvious from ET0 being 345 mm for the year 1992-93 compared to that of 315 mm for 1991-92. T2(l 100) at the low fertilizer level gave a similar value as that for T4(1001) at medium fertilizer level. Thus, shifting the irrigation schedule from crop stage IV to n saved fertilizer inputs by 50% without affecting the EC. Maximum yield response factor ky was observed in T6(0111) and T3(0011) both missing an irrigation at crop stage I. The crop showed least sensitivity to moisture deficit at crop stage HI as evident in T7(l 101). 3.2.3. Wheat 1993-94 Maximum water use efficiency EC was observed (Table V) in treatment T3(0000) under rainfed conditions. Among the moisture deficit irrigation treatments, a maximum EC was observed in T7(1100) followed by T5(1001) both involving irrigation at crop stage I, and minimum in T6(0101) missing an irrigation in crop stage L At the lower fertilizer level T2(l 111) and T8(l 101) had the lowest EC values indicating that at low fertilizer input large amounts of irrigation did not maintain EC . The yield response factor ky was maximum in T2(l 111). At the medium fertilizer level, ky was minimum in T7(l 100) and maximum in T6(0101) supporting that the moisture deficit exposed at later stages did not affect the yield as compared to crop stage I. The 3-year study on the same variety showed that irrigation at crop stage I contributed most to the total variation in grain yield. On the other hand, the moisture deficit at crop stage HI did not affect the yield significantly. 3.2.4. Genetic diversity of wheat to moisture deficit Maximum water use efficiency EC was observed (Table V) in treatment T3(0000) but at the cost of 38 to 44% loss of grain yield. In all cases, EC of the conventional flood irrigation 200 treatments T2(llll) was lowest. Probably, the water was more affectively utilized in the vegetative growth as compared to that during grain filling as observed from the low values of harvest index (not reported here). For Sarsabz , the EC value of T4(1011) was maximum and lowest in T6(l 110). For LU-26S, the order was reversed with the EC value of T6(l 110) being higher than that of T4(1011). For Pasban-90, Ec was maximum in T5(1101). Thus, under moisture deficit conditions the three varieties .had different options for maximizing EC . The yield response factor ky was lowest for T4(1011) in Sarsabz, T6(1110) in LU-26S and T5(1101) in Pasban-90 showing least effect of moisture deficit at crop stages ÏÏ, IV and IÏÏ, respectively for the respective variety. 3.2.5. Genetic diversity of cotton to moisture deficit For cotton variety NIAB-86, a maximum EC value was observed (Table V) for treatment T8(010) followed by T5(110). Both treatments included an irrigation at the generative crop stage. The lowest EC was observed in Tl(CSM) and T2(lll) where the conventional flood irrigation system was used. For NIAB-86, among moisture deficit at two crop stages, treatment T5(110) was most efficient as compared to the other treatments. T8(010) had a higher EC than did T6(011) leading to 22% more seed cotton yield and saving 19 mm water. The lowest ky value was observed at T1(CSM) and T5(110) followed by T8(010) - all involving an essential irrigation at the crop stage ÏÏ. For variety FH-682, the rainfed treatment T3(000) scored the highest EC followed by T5(l 10) and T8(010). The lowest Ec was observed in T2(l 11) followed by T1(CSM). Among moisture deficit irrigations T6(011) was the lowest and T5(110) the maximum showing that moisture deficit at vegetative stage reduced the EC. After Tl(CSM) the lowest ky value was observed in T5(110) under moisture deficit at maturity stage. ky was maximum in T6(011) showing that moisture deficit at vegetative stage reduced the yield in this variety. The two varieties showed different behavior for ky in moisture deficit irrigations. The seed cotton yield of NIAB-86 and FH-682 was enhanced by irrigation at generative and vegetative stages, respectively. 4. CONCLUSIONS - The neutron hydroprobe proved to be a very useful tool for assessing root zone soil moisture in irrigation projects. - The moisture deficit irrigation approach helped hi the pre-planned irrigation scheduling of wheat and cotton crops with multiple options to utilize water more efficiently. - In wheat, irrigation at tillering was most sensitive to moisture deficit At other crop stages, the varieties responded differently to moisture deficit - In cotton, FH-682 and NIAB-86 showed maximum sensitivity to moisture deficit at vegetative and generative stages, respectively. 201 ACKNOWLEDGMENT The authors are highly grateful to the FAO/IAEA authorities for funding and providing valuable guidance for the research activity. Gratitudes are also given to the colleagues in the Soil Biology Division, Statistical Cell and Cotton group of Mutation Breeding Division and Entomology Division of NIAB for intimate and sincere contribution. REFERENCES [I] MUSICK, J.I., DUSCK, D.A., Planting date and water deficit effects on development and yield of irrigated winter wheat. Agron. J. 72 (1980) 45-52. [2] RADIN, J.W., MAUNEY, J.R., KERRIDGE, P.C., Water uptake by cotton roots during fruit filling in relation to irrigation frequency. Crop Sei. 29 (1989) 1000-1005. [3] SCHMIDHALTER, U., OERTLI, J.J., Germination and seedling growth of carrots under salinity and moisture stress. Plant and Soil 132 (1991) 243-251. [4] KN1EP, K.R., MASON, S.C., (1991) Lysine and Protein content of normal and opaque Maize grain as influenced by irrigation and nitrogen. Crop Sei. 31 (1991) 177-181. [5] NORRIS, R.F., AYRES, D., Cutting interval and irrigation timings in Alfalfa; yellow Fox-tail invasion and economic analysis. Agron. J. 83 (1991) 552-558. [6] VAN GENUCHTEN, M.T., A closed form equation for predicting the hydraulic conductivity of unsaturated soils, Soil Sei. Am. J. 44 (1980) 892-897. [7] MUALEM, Y., A new model for predicting the hydraulic conductivity of unsaturated porous media. Water Resour. Res. 12 (1976) 513-555. [8] STEELE, R. G. D., TORRIE, J. H., Principles and Procedures of Statistics. McGrawHill, New York (1980). [9] HASSAN, U.A., OGUNLELA, V.B., SINHA, T.D., Agronomic performance of wheat (Triticum aestivum 1.) as influenced by moisture stress at various growth stages and seeding rates. J. Agron. & Crop Sci.158 (1987) 172-180. [10] AGGARWAL, P.K., NAVEEN, K., Analyzing the limitations set by climatic factors, genotype and water and nitrogen availability on productivity of wheat ÏÏ. Climatically potential yields and management strategies. Field Crops Res. 58 (1994) 93-103. [II] STORRIER, R.R., The influence of water on wheat yield, plant Nitrogen uptake and soil mineral Nitrogen concentration. Aust J. ExpL Agric. Anim. Husb. 5 (1965) 310-316. [ 12] GAGRI, P.R., PRfflAR, S.S., ARORA, V.K., Interdependence of nitrogen and irrigation effects on growth and input-use-efficiency in wheat. Field crops Res. 31 (1993) 71-86. 202 WATER AND NITROGEN USE EFFICIENCY UNDER LIMITED WATER SUPPLY FOR MAIZE TO INCREASE LAND PRODUCTIVITY I. CRACIUN, M. CRACIUN Research Institute for Cereals and Industrial Crops, Fundulea, Romania Abstract Inasmuch as drought is the main environmental factor limiting productivity, the study of plant response to water deficit has been a major research topic. Increasing maize evapotranspiration ET does not always mean increasing efficiency because a large ET value does not always mean a large gram yield value. Mechanisms relating grain yield to ET are complex. The amount of rain and its distribution during the reproductive stage is the main meteorological factor influencing yield. In our study 1991, high soil water content caused a reduction of maize grain yield in the wet years owing to excessive water under irrigation conditions. As a result of insufficient oxygen for respiration and lack of nitrate formation, excessive water normally limits root development. The yield response of different hybrids to water deficit is of major importance in production management. The available water supply should be directed towards fully meeting the water requirements of hybrids having large values of their yield response factor ky. For hybrids having smaller values of kr overall production will increase by extending the area under irrigation without fully meeting water requirements, provided water deficits do not exceed critical values. 1. INTRODUCTION Environmental water management is a subject which a few years ago would not automatically have appeared in an article of this nature. The fact that it is being discussed highlights the changes in public perceptions and values over recent years. The word environment means different things to different people. Although areas under rainfed and irrigated agriculture have significantly expanded during the past decade, crop productivity and its sustainability have been constrained by many complex and interrelated factors. Deterioration of irrigation systems, and problems of water logging and salinization have decreased agricultural production. Hence, although the interrelationships between water and sustainable agricultural development are direct and vital, the latter is not possible without the former. Rising costs of irrigation pumping, low commodity prices, inadequate irrigation system capacities and limited water supplies are among the reasons that many irrigators deliberately apply less water than is required for maximum yield. The goal of deficit irrigation is to enhance economic returns by reducing water and/or energy use. Crops are exposed to varying levels of environmentally induced stresses during their growth cycles. Inasmuch as stress affects crop productivity, this relationship and methods of quantifying and monitoring crop stress have received intensive research attention [1,2,3,4,5]. 203 The purpose of this study was to report: (i) the effects of water deficit periods on growth and yield components of maize, (ii) to determine the seasonal evapotranspiration requirements of maize, (iii) to give information regarding the adaptation of maize for limited irrigation in a region of normally high evaporative demand climate, and (iv) to generate rapidly, at any time, in any location, accurate estimates of the functional relations between yield of irrigated maize and ET, and before the start of an irrigation season, to predict an optimal irrigation program. 2. MATERIALS AND METHODS The study was conducted in the field on a cambic chernozem during 1991-1994 at the Research Institute for Cereals and Industrial Crops, in Fundulea, Romania (44° 30' N and 24° 10' E). For each experimental year a line-source sprinkler design was established with one hybrid (F420) in four replications. The well-irrigated treatment designed to avoid plant water TABLE I. IRRIGATION TREATMENTS FOR DIFFERENT STRESS PERIODS (1991 1994) Irrigation Treatment 001 _____Growth stage________ 1(V) 2(G) 3(M) 0 0 v 0 1 1 0 1 g 1 v 1111 1 G l V 1 0 0 1 V M No intended stress, normal watering at all stages. m Stress at pollination and ripening stage. 0 g 1 0 G Stress at vegetation and pollination stage. M G V Description Stress at vegetation stage. 1 0 1101 M - Stress at ripening stage. m stress received water at the level of 50% of available water in the first 80 cm of the root zone. Measurement of sprinkled water collected in cans placed at the center and the ends of each plot were taken immediately after irrigation to calculate the amount of each irrigation. Soil water 204 content to a depth of 120 cm was measured in the center of the harvest area plot with a fieldcalibrated neutron probe. Actual evapotranspiration ETa of maize was estimated using a well known water balance method. Maize seeds were sown in rows spaced 70 cm apart in May. Plant populations averaged 7.5 plants-nr2. Weeds were controlled by application of Diizocab at a rate of 10 L-ha'1. Water deficit periods are shown in Table I. Grain yield was determined from four sub-samples hand harvested from each plot, oven dried to constant weight, and adjusted to 15.5% grain moisture content on a dry weight basis. 3. RESULTS The seasonal precipitations and temperatures (Fig.l and Fig.2) were different from year to year during 1991-1994. For example, from April to October the seasonal rainfall totaled 611 mm in 1991, 249 mm in 1992, 363 mm in 1993 and 461 mm in 1994. The average rainfall for this period during 1991-1994 was 421 mm compared with 378 mm for the 32-year average at Fundulea. Hence, depending upon the level of water supply, the plants were grown under varying soil water regimes that imposed different stress conditions during all vegetative periods. See Fig. 3. 120 / \ / / \1991-1994AVERAGE APRIL JUNE AUGUST OCTOBER Fig. L Summary of average precipitation data for 1991-1994 and 1962-1994 at Fundulea. 3.1. Evapotranspiration ET - yield relation The yearly results indicated that the seasonal £J ranged from 348 mm in 1992 (VGm110-treatment) to 556 mm in 1994 (VGM-111-treatment). In 1992, a dry year (only 249 mm rainfall during April to October) seasonal ET ranged from 348 mm to no more than 443 mm. Depending on the level of water supply under the varying soil water regimes, the variation in maize seasonal ET-was more than 20%. As an average (1991-1994), the variation in maize seasonal ET was only 12%, and the value ranged from 422 mm when the stress was created at 205 30 g 20 g 10 - 1991-1994 AVERAGE - 1959-1994 AVERAGE 0 APRIL JUNE AUGUST OCTOBER Fig. 2. Summary of average air temperature data for 1991-1994 and 1962-1994 at Fundulea. 400 H g 300 — FELD CAPACITY 8ce 200 WILTING POINT g 100 MAY JUN JUL ÄUG SEPT Fig. 3. Water deficit distribution for limited -water supply within the 0-80 cm soil profile cropped to maize (F-420) during 1991-1994 at Fundulea. pollination and ripening stage (Vgm-lOO-treatment) to 482 mm under the normal watering (VGM-111-treatment). See Table II. Water stress is a major factor accounting for high yield variability. During 1991-1994, the grain yields varied with climatic conditions of the year and the level of water stress. Grain yields ranged from 7,039 kg-ha'1 (hi 1992, a dry year) to 13,556 kg-ha4 (in 1991, a wet year with more than 600 mm ramfall from April to Oct). In 1992, a dry year, grain yield variation was 45% and values ranged from 7039 kg-ha'1 (in vgM-001-treatment) to 12,849 kg-ha'1 (in VGm110-treatment). In 1991, a wet year, the grain yield variation was no more than 7% and with 206 values ranging from 12641 kg-ha4 (in VGm-llO-treatment) to 13,556 kg-ha4 (in Vgm-100- treatment). In 1991, over irrigation reduced the yield. During the period of study (1991-1994), grain yields ranged on the average from 10,279 1 kg-ha" (in vgM-001-treatment) to 12,459 kg-ha4 (in vGm-011 -treatment) with yield variation for all treatments being only 17% which depended upon the climatic conditions from year to year and water stress level. See Table E. Although the relationship between ET and grain yield is generally linear, our results show that water stress at critical periods can make yields deviate from that strict linearity. The data shown that on an average (1991-1994), hybrid F420 had the largest ET value (482 mm) when it was well irrigated (VGM-111-treatment), but that treatment did not have the greatest yield. A greater grain yield of 12,459 kg-ha4 was achieved when stress was created at vegetative stage only (011) and still another large grain yield of 12,363 kg-ha4 was achieved when the stress occurred at the ripening stage(VGm-l 10-treatment). Again, see Table u. TABLE n. WATER DEFICIT EFFECTS ON ANNUAL AVERAGE TOTAL ET, MAIZE YIELD AND WATER USE EFFICIENCY (WUE) IN RELATION TO THE ANNUAL AVERAGE AMOUNT OF IRRIGATION WATER APPLIED ACCORDING TO TREATMENTS DURING 1991-1994. 001 vgM Irrigation water (mm) Totaler Maize yield (kg-ha4) WUE (kg-mm4) 3.2. 74 440 10,279 31 Oil vGM Treatment 1 11 100 VGM Vgm 124 445 12,459 36 198 482 12,291 27 1 10 VGm 74 422 123 468 12,169 42 12,363 43 Efficiency of irrigation water - WUE In an attempt to determine whether progress had been made in breeding maize for increased utilization of soil water, the relationship between crop hybrid variety and water stress was explored. Efficiency of irrigation water or water use efficiency WUE is defined as the ratio of the increase of yield (kg-ha4) owing to irrigation to the total amount of irrigation water supplied (mm-ha4). Depending on climatic conditions of the year and the level of water stress, values of WUE ranged from 11 kg-mrn4 in 1993 (when the maize was irrigated only at ripening .stage, i.e., 001-vgM-treatment) to 62 kg-mm4 in 1994 (at the same stage). In the period 19911994, on the average, values ranged from 27 kg-mm4 in the well irrigated soil (111-VGMtreatment) to 43 kg-mm4 when the stress occurred only at ripening stage (110-VGm-treatment). 207 The data show that it is possible to reduce the amount of irrigation by 70 mm (one irrigation) without any reduction in grain yield when the stress occurs only at the ripening stage, or to reduce the amount of irrigation by 120 mm with a reduction in grain yield of only 1%. Table El shows for the period 1991-1994 maize yield losses in relation to the amount of irrigation reduction for different stages of growth. Reducing the amount of irrigation by 46% when water stress occurred at vegetative and pollination stages reduced the grain yield of F420 by 18%. On the other hand, for the same reduction of irrigation when the water stress occurred at pollination and ripening stages reduced the grain yield by only 9%. For a smaller irrigation reduction of 29%, grain yield reduction was 8% when the stress was only at the vegetative stage, or only 3% when the stress was created at ripening stage. TABLE HI. MAIZE YIELD LOSSES IN RELATION TO THE AMOUNT OF IRRIGATION REDUCTION ACCORDING TO TREATMENTS DURING 1991-1994. __________________Treatment_____________ Irrigation water reduction (%) Maize yield reduction (%) 3.3. 001 vgM Oil vGM 111 VGM 100 Vgm 110 VGm 46 29 0 46 30 18 8 0 9 3 Maize yield response factor - ky The response of maize grain yield to water supply is quantified through the response factor ky which relates the relative yield decrease (l-Ya- Y^) to relative evapotranspiration deficit (1 - ETa • ET^1). For the total growth period the decrease in yield is proportionally greater with an increase in water deficit (ky > 1) for crops such as maize, for hybrids with a longer period of vegetation, or for hybrids with less drought resistance such as hybrid Progrès (ky = 2.0). See Table IV. Note that a higher drought resistance was manifested by Fundulea 340 (ky = 0.74) and Fundulea 322 (ky = 0.87). 4. CONCLUSIONS As drought is the main environmental factor limiting crop productivity, the study of plant response to water deficit has been a major research topic. An increase of maize ET does not 208 TABLE IV. YIELD RESPONSE FACTOR ky FOR DIFFERENT MAIZE HYBRIDS UNDER LIMITED WATER SUPPLY CONDITIONS DURING 1991-1994. _______________Maize hybrid________________ F 322 F 340 Progrès Robust Soim 1.23 1.79 2.00 0.74 0.87 always mean an increased efficiency because the highest ET value does not always coincide with highest grain yield. Among the mechanisms relating grain yield and ET, the amount of rain and its distribution during the reproductive stage is the main meteorological factor influencing yield. In our study 1991, high soil water content caused a reduction of maize grain yield in the wet years owing to excessive water under irrigation conditions. As a result of insufficient oxygen for respiration and lack of nitrate formation, excessive water normally limits root development. The yield response of different hybrids to water deficit is of major importance in production management The available water supply should be directed towards fully meeting the water requirements of hybrids having large values of ky. For hybrids having smaller values of ky, overall production will increase by extending the area under irrigation without fully meeting water requirements, provided water deficits do not exceed critical values. REFERENCES [1] [2] [3] [4] [5] WANJURA, D.F., HATFIELD, J.L., UPCHURCH, D.R., Crop water stress index relationships with crop productivity, Irrig.Sci. 11 (1990) 93-99. OSVALD, J., OSVALD, M., Consequences due to water stress for the development and yield of maize, sorghum, cabbage and tomato plants, Bioloski - VestnLk 39 (1991) 1-2, 129-135. NESMTTH, D.S., RITCfflE, J.T., Maize (Zea mays L.) response to a severe soil water deficit during grain filling, Field Crop Research 29 (1992) 23-35. CRACIUN, L, NAESCU, V., CRACIUN, M., The influence of irrigation on the main crops yield under Romanian Plain conditions, Lucrari stiintifice, Agricultural University lasi 36 (1993) 128-132. CRACIUN, L, CRACIUN, M., Irrigated maize response under limited water supply, Romanian Agricultural Research 31 (1994) 57-63. Next page(s) left blank 209 WATER BALANCE AND NITRATE LEACHING IN AN IRRIGATED MAIZE CROP IN SW SPAIN F. MORENO, J.A. CAYUELA, I.E. FERNÂNDEZ, E. FERNÂNDEZ-BOY, J.M. MURILLO, F. CABRERA Institute de Recursos Natural es y Agrobiologia de Sevilla, Sevilla, Spain. Abstract During three consecutive years (1991-1993) a field experiment was conducted in an intensively irrigated agricultural soil in SW Spain. The main objective of this study was to determine the water flow and nitrate leaching below the root zone, under an irrigated maize crop and after the growing season (bare soil and rainy period). The experiment was carried out on a furrow irrigated maize crop using one of the highest nitrogen fertilization rates traditionally used by farmers in the region [about 500 kg N-(ha-yr)'1] and another that represents one third of the former [170 kg N^ha-yr)'1] to provide data that can be used to propose modifications of nitrogen fertilization to maintain crop yield and to prevent the degradation of the environment. The terms of water balance (crop evapotranspiration, drainage and soil water storage), and the nitrate leaching were determined by intensively field monitoring the soil water content, soil water potential and extracting the soil solution by the combination of a neutron probe, tensiometers and ceramic suction cups. Nitrogen uptake by the plant and N03-N produced by mineralization were also determined. The results showed that, in terms of water balance, crop evapotranspiration was similar with both N- fertilization rates used. During the irrigation period drainage below the root zone was limited. Only in 1992 the occurrence of rainfalls during the early growing period, when the soil was wet from previous irrigation, caused a considerable drainage. Nitrate leaching during the entire experimental period amounted to 150 and 43 kg-ha*1 in the treatments with high and low N-fertilization, respectively. This leaching occurred mainly during the bare soil and rainy periods, except in 1992 when considerable nitrate leaching was observed during the crop season owing to high drainage. Nitrate leaching was not so high during the bare soil period as could be expected because of the drought during the experimental period. A reduction of Nfertilization strongly decreased nitrate leaching without decreasing yield. 1. INTRODUCTION Increases in the nitrate concentration in both groundwater and surface water are related to agricultural practices. The use of nitrogen fertilizers at rates higher than the rate of uptake by the plant increases the potential for increased nitrate leaching, as has been shown for nitrogenfertilized corn [1]. Addiscott et al. [2] showed that the increase of nitrate concentrations in both groundwater and surface water hi England, during the last 30 years, is related to the increased use of nitrogen fertilizer. In Mediterranean areas, farmers often use amounts of N-fertilizers that exceed the N requirements of crops [3,4], and thereby increase the amounts of potentially teachable nitrate in the soil. Under irrigated agriculture, drainage below the root zone is required to maintain salt 211 balance. The water flow below the root zone can produce nitrate losses. Water flow and nitrate leaching depend on the soil characteristics, amount of water applied by irrigation or natural precipitation, and the amount, timing and species of nitrogen applied. Sheperd [5] posed the question, "Are the effects of irrigation on nitrate leaching loss good or bad?", and goes on to point out "both opinions have recently been expressed". Some papers [3,6,7] have shown the relationship between the use of high N-fertilization and nitrate leaching in Spain. This leaching generally occurred during the rainy period when the soil was bare. The objective of this study was to determine the water flow and nitrate leaching below the root zone under an irrigated maize crop and after the growing season (bare soil and rainy period) in SW Spain during three consecutive years. Results presented in this paper correspond to the irrigation and fertilization practices normally used in this region, and the use of a reduced nitrogen fertilization rate. The study was conducted following a multidisciplinary approach to obtain data necessary for a better understanding of the problem, and to propose modifications of the rate of N-fertilization while maintaining crop yields and preventing the degradation of the environment 2. MATERIALS AND METHODS 2. 1. Experimental site The experiments were conducted at the experimental farm of the Institute de Recursos Naturales y Agrobiologia de Sevilla (ERNAS, CSIC) located at Coria del Rio close to Seville city in SW Spain (37° 17' N, 6° 3' W). The climate is typically Mediterranean with mild rainy winters and very hot, dry summers. The average annual rainfall (1971 -1992) is 550 mm and most falls between October and May. An experimental plot (0.1 ha) was used (Fig. 1). The soil is a sandy loam (Xerocrept), developed on limey sandstone of the Aljarafe Miocene, with a depth of more than 3 m. The -21m "T —I __ |A2 _ rA3_ T SUBPLOT A I r 1 .-I- ——i——r— i i __i__ iu_ 1 B T SUBPLOT B Fig. L Experimental layout at the field site. 212 spatial variability of some soil properties was studied after taking samples at 45 grid nodes of a 5 x 5 m cell mesh, at two depths, 0 - 0.5 m and 0.5 -1m. Mean textural values are, at 0 - 0.5 m and 0.5 -1 m, respectively: coarse sand 60.7±4.9% and 57.3±4.6%; fine sand 16.8±2.8% and 17.8±3.0%; sût 9.0±1.8% and 8.3±2.1%; clay 13.1±2.2% and 16.4±1.9%. Organic matter contents are 0.88±0.15% and 0.55±0.09% at depths of 0 - 0.5 m and 0.5 -1 m, respectively. 2.2. Crop management and treatments The experimental plot was divided into two subplots, A and B (Fig. 1), each of 450 m2, with the aim of establishing two nitrogen fertilization treatments. Both subplots were cropped with maize (cv. Prisma) during three consecutive years from 1991 to 1993. Planting was carried out on the 5th of April 1991,24th of March 1992, and 24th of March 1993. The rows were 0.8 m apart with a plant density of 75,000 plants ha4. Subplot A had 510 kg N-(ha-yr)-1, a rate widely used in the area. Subplot B at 170 kg N-(ha-yr)-1 received one-third of the normal rate. Fertilization was applied at three times: one deep fertilization of 1000 kg-ha*1 (15-15-15 complex fertilizer) some 10 days before planting, and two top dressings at about 45 and 75 days after planting. Each top dressing consisted of urea at 400 kg-ha-1 (46% N) in subplot A and one third of this amount in subplot B. Standard management practices typical for the Guadalquivir river valley, the main area for irrigated maize in the region, were used. The crop was irrigated by furrow in both subplots, but some sprinkler irrigations were applied between planting and the establishment of the furrows. Dates and quantity of irrigation are given in Fig. 2. Irrigation stopped at about the end of July, or the beginning of August, some 20 days before harvest The crop was kept healthy and free of weeds. With the land surrounding the experimental plot cropped every year with furrow or sprinkler irrigated crops (maize or cotton) advection was minimized. Rainfall during the experimental period is given hi Fig. 2. The soil of the plot was kept bare during the period between the harvest and the beginning of the next crop season. 2.3. Measurements Several measurement sites were installed in every subplot (three in subplot A and three in subplot B), each one equipped with the following equipment: - one access tube for the neutron probe to measure soil water content every 0.1 m to a depth of 2.3 m - five mercury tensiometers at depths of 0.3,0.5,0.7,0.9 and l.lm - three ceramic suction cups at 0.3,0.6 and 0.9 m to extract the soil solution - soil water content was monitored every five or seven days during the crop period. During the bare soil period these measurements were carried out every two weeks, and always after a rainfall. Tensiometer readings were recorded daily during the crop season, and one or two times per week during the bare soil period - rainfall and micrometeorological data were obtained from a meteorological station 200 m away from the plot situated hi the experimental farm, 213 some crop development parameters (crop height, leaf area index and root density), nitrogen uptake by the crop and yield were determined the soil solution was extracted with suction cups at least once per week when the soil water content allowed this extraction. The soil solution was analysed for nitrate content in the laboratory by ionic chromatography using a solution of 0.0013 M borate - 0.0013 M gluconate in acetonitrile (12% v/v) at pH 8.5 as eluent g 40 fa 20 II 0 z o 30 Ü S 0 0 300 60 900 DAY NUMBER Fig. 2. Rainfall and irrigation during the experimental period. Day 0 is 20 March, 1991. In 1992 one 2 x 2 m plot (C) was also established on bare soil without fertilization. This plot was irrigated with the same amount of water and on the same dates as subplots A and B. The object was to measure the water flow and the NOs-N concentration in the soil solution in order to calculate the NOs-N produced by the mineralization of the soil organic matter. 2.4. Determination of water balance and nitrate leaching The water balance was calculated from the mass conservation equation AS = R + I-D-AET (1) where S is the change in water storage (mm) in the soil profile exploited by the roots, R the rainfall (mm), I the depth of irrigation applied (mm), D the drainage (mm) at a depth fe) below the root zone, AFT the actual evapotranspiration (mm). Water runoff was neglected because it was practically nil on this field site. 214 The drainage component D was estimated using Darcy's law: D = qAt = -{K(8)gradH]At (2) 1 where q is the mean volumetric flux density (mm-d' ) during At, At the period of time (d), K(6) the hydraulic conductivity (mm-d4) corresponding to the water content 6 at a depth zr and grad H the hydraulic head gradient at the same depth. For the application of this method K(6) must be known. The K(&) relationship was determined by the internal drainage method [8] at a selected site of the plot, and by the application of the "zero flux plane" method [9] at every measurement site. A typical result is given in Fig. 3. 10 2 K = (7.49-10-6)exp(63.50) r2 = 0.84 ,10° O 10'2 - • INTERNAL DRAINAGE METHOD O ZERO-FLUX PLANE METHOD __________I____________ 0 0.1 SOIL WATER CONTENT 0.2 0 (cm Fig. 3. Relation between the hydraulic conductivity K and the soil water content 6 obtained in situ. The amount of NOs-N leached LN below the root zone at a depth of 0.9 m was obtained from the relation LN=DCo.9 (3) where D is the water drainage calculated at a depth of 0.9 m from (2) and Co.9 the NOs-N concentration in the soil solution at the same depth. The depth of 0.9 m was established for the calculation of drainage and nitrate leaching because results of the root length density obtained in this study showed that the root system was situated above this depth. The amount of NOs-N in a soil layer at a given time was calculated by multiplying the NÛ3-N concentration of the soil solution by the water stored in this layer. For this purpose we assumed that the soil solution extracted at a depth of 0.3 m was representative of the 0 - 0.4 m soil layer, that at 0.6 m of the 0.4 - 0,7 m soil layer, and that at 0.9 m of the 0.7 - 1 m soil layer. Summing the three values obtained in this way gave the total amount of NOs-N hi the soil profile at a given date. At harvest time it was not possible to extract the soil solution and soil samples at 0 - 0.3, 0.3 - 0.6 and 0.6 - 0.9 m were thus taken to determine the NOs-N content in the soil. 215 The years of the experimental period (1991 -1993) were characterized by total rainfalls lower than the annual average of the period 1971-1992 (Table I). It is noticeable that during the autumns of 1991,1992 and 1993, respectively, rainfall was 20,35 and 41% less than the average (Table I). Distribution of rainfall was also different from that normally observed in the region. TABLE I. COMPARISON OF RAINFALL DURING THE EXPERIMENTAL PERIOD WITH THE 20-YEAR AVERAGE IN THE REGION Period 1992 (Sep-Sep) 1990-1991 1991-1992 1992-1993 Rainfall (mm) Average rainfall Sep 1971-Sep 1992 Period Sep-Dec Rainfall (mm) (mm) 458 446 353 550 Average rainfall Sep-Dec, 1971(mm) 1991 1992 1993 201 160 147 247 3. 1. Water balance Fig. 4 shows the cumulative values of the actual evapotranspiration AET and drainage during the crop seasons of 1992 and 1993 for both subplots. Results of water balance for the crop season of 1991 are not presented in this way because measurements in the experimental plots started at the beginning of June when the crop was about 0.6 to 0.7 m high. The total water input (rainfall and irrigation) in both subplots during the crop seasons (March-August) of 1992 and 1993 were 731 and 666 mm, respectively. The A£Tin subplot A during the'crop season of 1992, amounted to 640 mm and the drainage was 142 mm. In contrast, in subplot B the AET was 578 mm, and the drainage below the root zone was 241 mm. During the crop season of 1993 the AET amounted to 646 and 637 mm in subplots A and B, respectively. Water losses by drainage during this crop season were 57 and 85 mm in subplots A and B, respectively. The fact that greater drainage was observed during the crop season in 1992 than in 1993 may be due to the rainfall distribution. In 1992 about 90 mm of rain fell during the early growth period, concentrated mainly in a few days, when the soil was wet from previous irrigations, and while water consumption by the crop was still low. After this rainfall the water uptake by the crop in subplot A was higher than in subplot B. This can be explained because the leaf area index LAI at this time was higher in subplot A than in subplot B. The higher crop evapotranspiration in subplot A than hi subplot B is in part responsible for a lower drainage in subplot A. In the 1993 crop season no heavy rain occurred early in the growth period, so the crop evapotranspiration was similar in both subplots. 216 1000 50 100 150 DAYS AFTER PLANTING Fig. 4. Cundative drainage and evapotranspiration during the crop seasons of 1992 and 1993 in subplots A and B. 3.2. Nitrate-N in the soil profile The maximum amount of NOs-N stored in the soil profile (0 -1 m) in plot C (bare soil not fertilized) during 1992 reached about 84 kg NOs-N-ha'1. This maximum was observed in April and also in the autumn after the first rainfalls. Results obtained in the laboratory from experiments of mineralization on soil samples, taken at a depth of 0 - 0.3 m showed that the amount of NOs-N produced in 8 weeks for this soil layer was 53 kg-ha*1. This clearly shows that NOs-N production from the mineralization of the soil organic matter is relatively important even though the organic matter content of the soil is low. In subplots A and B, the NOs-N contents in the profile (0 - 0.9 m) at harvest time in 1991,1992 and 1993 are shown in Table n. Most of this N03-N was found in the soil layers 0 0.3 m and 0.3 - 0.6 m. These amounts were very high in subplot A, but different from one year to other. In 1992 the NOs-N content in the profile was 265 kg-ha-1 (Table ÏÏ), lower than in 1991 217 and 1993 years in which drainage during the crop season was low. These results show that the nitrate potentially leachable during the rainy period is high when a N-fertilization rate such as that in subplot A is used. The NOs-N contents in the profile for subplot B were much lower than in subplot A in accordance with the lower N-fertilization. The highest NOa-N contents in the profile in 1993 for both subplots must be related to the lowest drainage observed during the crop season and to a lower crop performance than in 1991 and 1992. TABLE H. CONTENT OF NO3-N IN THE SOIL PROFILE (0 - 0.9m) AT HARVEST NO3-N (kg-ha-1) Year Subplot A Subplot B 199i 1992 1993 293.4±109.4 265.5±101.3 375.2± 89.7 61.2±36.9 49.2±11.1 125.4±25.1 3.3. Nitrate leaching An example of changes in the NOs-N concentration in the soil solution, extracted at three depths in both subplots is shown in Fig. 5. These results correspond to the period between the 10th of October 1991 (day number 209 from the beginning of the experiment) and the 25th of August 1992 (day number 524). Clearly, the NOs-N concentrations at all depths were higher in subplot A than in subplot B. With the data obtained by systematic monitoring of the concentration of NOa-N in the soil solution at a depth of 0.9 m, and the use of Darcy's law, the amount of NOs-N leached below the root zone was estimated. Fig. 6 shows the cumulative water drainage and the cumulative NOs-N losses for the entire experimental period. The total drainage observed in subplot B was higher than in subplot A. As mentioned in the section on water balance this greater drainage is related to a lower water uptake by the crop in subplot B than in subplot A, particularly during the crop season of 1992. Total drainage values represent 13.4 and 21.4% of the water applied by irrigation and rainfall in subplots A and B, respectively. NOs-N leaching generally occurred during fall and winter (rainy period) when the soil was bare. These nitrate losses were probably smaller than expected owing to a lower rainfall than the average for this period (Table HI). In contrast, considerable nitrate leaching was observed during the early growth period in the second crop season (1992), owing to rainfall (90 mm) when the soil was wet from the previous irrigation and the water consumption by the crop was still low, as has been earlier mentioned. Nitrate leaching was always higher in subplot A, where a high N-fertilization rate was applied, than in subplot B, even though the drainage was higher in subplot B. Total 218 SUBPLOT A 1000 L * l 1 30cm 1 500 - IL* if ""•' z •^ o z 0 1 300 -1 - 1 60cm l 1- 1 0 150 1 30cm " • ^ 100 - • 50 _ •«VjT»! «•• »W». • ! l 150 SUBPLOT B l i «• 1 1 • 1 V» 1 Vfcrf 1 60cm 1 100 - 200 - • t 100 * 0 1 300 -» • 200 - 100 ~ • n 1 • . * •* t « • » ! • * ' •! l1 1 rt/t 90cm 50 1 li - 0 150 i •* • *•. .'; i i i 90cm 100 • 1 »»^ F Si 50 n 1 1 * 1 r• • . * DAY NUMBER Fig. 5. NOs-N concentration in the soil solution at different depths in subplots A and B. Day 0 is 20 March, 1991. leached was 147.5 and 44.0 kg-ha*1 in subplots A and B, respectively. These amounts are not as high as could be expected under our conditions possibly because of the drought of the experimental period. This is particularly true in subplot A if we take into account the high NOsN contents in the soil profile at harvest time. Nitrate leaching was strongly reduced by decreasing the N-fertilization rate by one-third. 3.4. Crop response Table IV shows some parameters of the crop development and yield in both subplots. Plant heights measured when the crop was fully developed were not significantly different between subplots. In contrast, the leaf area index LA/, also measured when the crop was fully developed, was higher in subplot A than in subplot B, although mean values were not significantly different Bennett et al. [10] showed that in an optimally irrigated maize crop with a plant density similar to that used in our study the LAI in the treatment with low N-fertilization was lower than in the treatment with high N-fertilization. Crop yields were not significantly different between subplots in the three crop seasons. The only significant difference was between the 1000-kernel weight in 1991 (Table IV). The nitrogen exported by the crop (determined in the above ground part of the plant at harvest) was higher in subplot A than in subplot B in 1991 and 1992, but was practically the same in 1993 (Table IV). The nitrogen exported by the crop in subplot B (between 216 and 241 kg-ha4) was higher than the applied by fertilization [170 kg N-(ha-yr)- 1 ]. This indicates that the crop used NOs-N from the mineralization of the soil organic matter and from the irrigation water. 219 200 400 600 800 DAY NUMBER Fig. 6. Cunüative dranläge and NOs-N leached below the depth of 0.9m in subplots A and B during the experimental period. Day 0 is 20 March, 1991. In both subplots the yield decreased from 1991 to 1993, probably owing to the continuous cropping with maize and/or the unusual climatological conditions prior to flowering in 1992 and 1993. With the reduction of the N-fertilization rate we have not observed differences in the crop yield. In our conditions, the N-fertilization rate of about 500 kg N-(ha-yr)'1 is too high, but in contrast 170 kg N-(ha-yr)'1 is probably too low for a continuous maize crop in order to maintain an adequate fertility level in the soil. 4. CONCLUSIONS From results obtained in our experiment it seems that the use of the method based on measurements of soil water content by neutron probe and soil water potential by tensiometers to 220 determine the water balance is appropriate if we use the hydraulic conductivity-water content relationships determined in situ. From the drainage calculated at a depth below the root zone and the NOs-N concentration in the soil solution it was possible to calculate the NOs-N leaching below the root zone. The use of a high N-fertilization rate (500 kg-ha*1), traditionally used by farmers in our region, for an irrigated maize crop is excessive and causes a high NOs-N content in the profile at harvest time that is potentially leachable during the rainy period in autumn and winter when the soil is bare. During the experimental years rainfalls below the average and with an altered distribution did not produce leaching as could be expected. In contrast, during the crop season the occurrence of heavy rainfalls when the soil profile was wet from previous irrigations caused a considerable NOs-N leaching. With a N-fertilization rate that represents one-third of that TABLE HI. MEAN VALUES OF DRAINAGE AND N03-N LEACHING BELOW A DEPTH OF 0.9 m FOR DIFFERENT PERIODS Period Subplot A Drainage (mm) NOs-NOcg-ha'1) Subplot B Drainage (mm) 6 Jun 1991 25 Aug 1991 (crop season) 25.0± 6.8 3.2+ 1.0 47.8±13.5 0.3± 0.4 25 Aug 199124 Mar 1992 (bare soil) 79.7±14.0 49.7±10.6 127.1±21.1 1.3± 1.5 149.2±19.2 40.4±13.8 243.0+24.6 12.7±16.4 25 Aug 199224 Mar 1993 (bare soil) 36.6±11.7 25.3± 6.5 53.6± 9.6 12.9± 24 Mar 1993 25 Aug 1993 (crop season) 57.0±13.6 28.4± 8.7 85.0+.13.2 16.4+ 6.5 24 Mar 1992 25 Aug 1992 (crop season) 4.1 221 TABLE IV. MEAN VALUES OF PLANT HEIGHT.LEAF AREA INDEX LA/, 1000 KERNEL WEIGHT AND YIELD Treatment Year Plant heightt L4/f (m) 1000 kernel weight (g) Exported N (kg-ha-1) Yield (Mg-ha-1) 314b 334a 261 241 13.0a 13.2a 1991 Subplot A Subplot B 2.91att 2.94a 1992 Subplot A Subplot B 2.24a 2.30a 4. lia 3.47a 329a 320a 270 220 12.5a 12.5a 1993 Subplot A Subplot B l.SOa 1.83a 3.38a 2.78a 322a 328a 221 216 9.7a 9.6a t Plant height and LAI were measured when the crop was fully developed. tt Values followed by the same letter in the same column per year do not differ significantly at the level P < 0.05. normally used by farmers and with the same irrigation the NOs-N content in the soil profile at harvest was much lower and consequently leaching was strongly reduced. The reduction of Nfertilization did not affect the yield. A lower leaf area was observed in the crop with the low Nfertilization than in the crop with the high N-fertilization. Even though our soil has a low organic matter content the NOs-N produced by mineralization together with the NOs-N of the irrigation water was enough to recover partially the nitrogen needs of the crop with low fertilization, at least for three years. In our conditions 175 kg N-ha*1 of fertilization for a continuous maize crop is probably too low in order to maintain an adequate fertility level of the soil. ACKNOWLEDGEMENT Thanks are due to Mr O. Blazquez and Mr J. Rodriguez for help with field measurements. Research carried out in the framework of the contract STEP-CT90-0032 of the European Community. REFERENCES [ 1] [2] 222 ROTH, L.V., FOX, R.H., Soil nitrate accumulations following nitrogen-fertilized corn in Pennsylvania. J. Environ. Qua!., 19 (1990) 243-248. ADDISCOTT, T.M., WHITMORE, A.P., POWLSON, D.S., Farming, fertilizers and the nitrate problem. CB A International, Wallingford (UK) (1991) 170 pp. . [3] [4] [5] [6] [7] [8] [9] [10] CAYUELA, JA, FERNANDEZ, J.E., MORENO, R, MURILLO, J.M., CABRERA, F., Estimaciön de las pérdidas de nitrato en un suelo con cultivo de maiz y riego. Riegos y Drenajes XM, 75 (1994) 30-34. DANALATOS, N.G., Quantified analysis of selected land use systems in the Larissa region, Greece. Doctoral thesis, Agricultural University, Wageningen (The Netherlands). (1992) 370 pp. SHEPHERD, M.A., Effect of irrigation on nitrate leaching from sandy soils. Water and Irrigation Review (1992) 12 19-22. RAMOS, C., VARELA, M., Nitrate leaching in two irrigated fields in the region of Valencia (Spain). In: R. Calvet (éd.), Nitrate-Agriculture-Eau. Proc. International Symposium, Paris (1990) 335-345. ORDONREZ, R., GIRALDEZ, J.V., GONZALEZ, P., Nitrogen use on irrigated farms in the Guadalquivir Valley: Approach to a rational design after soil column leaching (1990). HILLEL, D., KRENTOS, V.D., STILIANOU, Y., Procedure and test of an internal drainage method for measuring soil hydraulic characteristics in-situ. Soil Sei. 114 (1972) 395-400. VACHAUD, G., DANCETTE, C., SONKO, M., THONY, J.L., Méthodes de caractérisation hydrodynamique in-situ d'un sol non saturé. Application à deux types de sol du Sénégal en vue de la détermination des termes du bilan hydrique. Ann. Agronomiques 29 (1978) 1-36. BENNETT, J.M., MUTO, L.S.M., RAO, P.S.C., JONES, J.W., Interactive effects of nitrogen and water stresses on biomass accumulation, nitrogen uptake, and seed yield of maize. Field Crop Res. 19 (1989) 297-311. Next page(s) left blank 223 OPTIMUM IRRIGATION SCHEDULES FOR COTTON UNDER DEFICIT IRRIGATION CONDITIONS M.S. ANAÇ, M. ALI UL, I.H. TÜZEL, D. ANAÇ, B. OKUR, H. HAKERLERLER Ege University, Agriculture Faculty, Irrigation and Dränage Department, Ismir, Turkey Abstract The study aimed at determining the following: water consumption, irrigation water requirements of new cotton variety N84, spécifie growth stages of cotton which are less sensitive to stress so that irrigation could be avoided without significant yield decrease and the interactions between deficit irrigation and nitrogen fertilizer use. The experiment was set up with six irrigation and three nitrogen fertilizer (0, 60 and 120 kg-ha"1) treatments. The irrigation treatments employed a single stress at vegetative, flowering and boll formation stages, in addition to full irrigation, continuous stress and the traditional practice. A stress condition was defined when the available soil water was depleted by 75-80%, whereas in normal irrigation the depletion was only 40% in the root zone (0.90 m). For the full irrigation treatment 8 irrigations were applied, whereas only 3 or 4 irrigations were applied for a continuous stress condition. The number of irrigations were 6 or 7 for the other stress treatments. Irrigation water applications varied from 424 to 751 mm. Seasonal ET ranged between 659 and 899 mm with the highest monthly ET occurring in August for all of the treatments. Daily ET varied from 2.2 to 12.1 mm-d"1. Seed cotton yields, ky values and yield-N indices indicated that the vegetative stage was most sensitive to water stress. The boll formation stage was least affected by water stress. Under limited water resource conditions, vegetative growth period of cotton should be given preference for irrigation, followed by flowering period. Omitting irrigation during the boll formation stage would result in a water savings of 4.3 to 9.1%. Yield changes with respect to N rates showed that high N doses are accompained by high yields. Nitrogen recoveries either from fertilizers or soil revealed high uptakes in full irrigation conditions. Average nitrogen use efficiencies during the three year study were 19% for stressed conditions and 29% for full irrigation. The distribution of N uptake in different plant parts showed that the seeds contained the greatest amount of N. 1. INTRODUCTION Inasmuch as rainfall is insufficient to meet plant water requirements for optimum crop growth and yield, irrigation is practiced in the majority of Turkey. Like in other arid and semi-arid regions of the world, investments for irrigation projects are of top priority. The total irrigated area is about 4 million ha, A new irrigation project, aimed at irrigating 1.8-106 ha of agricultural land in Southeast Anatolia is under construction. 225 Although the promising improvements have been achieved, major problems confront the future of irrigation development in Turkey. Chief among these are; rapidly decreasing surface and ground-water resources owing to drought conditions occurring particularly in the western regions, escalating cost of pumping and intensified competition for water from expanding urban areas and industry. The impact of progressive drought on the water resources of irrigation schemes in western Turkey has been of vital importance during the last 5 years. For example, in the Lower Gediz Irrigation Project which covers an area of about 105 ha, the reservoir storages decreased by 16 to 35% of those normally available since 1988. Water shortages limited irrigation deliveries to a great extent with water distribution performed for 3 or 4 weeks at the peak demand period of the major crops. Limited availability of irrigation water requires fundamental changes in irrigation management or urges the application of water saving methods. A generally applicable procedure is to assess the benefits of changing irrigation water management based on deficit (or partial) irrigation which is the practice of deliberately under-irrigating crops. In order to implement deficit irrigation successfully, specific growth stages of the major crops at which they can withstand water stress with no significant effect on plant growth and yield need to be ascertained. Thus, it will be possible to develop optimum schedules for implementing deficit irrigation programs. In the Aegean Region, cotton is one of the main crops and is grown on about 240,000 ha. The objectives of this study were to: (i) develop irrigation schedules for new cotton variety of N84 by assessing water consumption, irrigation water requirements and irrigation intervals, (ii) develop a reliable crop production function that relates water use to crop yield, (iii) understand the relation between high nitrogen fertilization and improved water stress tolerance of plants, (iv) examine nitrogen recoveries and use efficiencies under full irrigation and stress conditions and (v) identify specific growth stages of cotton during which it is less sensitive to water stress so that the irrigation can be omitted without significant yield decrease and thereby reduce the total irrigation water applied. 2. MATERIALS AND METHODS A local new cotton variety N84 was used in the experiment conducted in Bornova-Izmir (38° 24' N, 27° 10' E). The climate is of the Mediterranean type characterized by rainy and warm winters and dry and hot summers. The long term average annual precipitation is 525 mm. During the experiment the annual rainfall was 294, 554 and 485 mm in 1992, 1993 and 1994, respectively. Precipitation was concentrated mostly between October and May. Mean annual temperature and relative humidity are 17 °C and 61%, respectively. Annual class A pan evaporation is about 1,569 mm with a daily maximum of 9 mm in July. 226 The experiment was conducted on an alluvial clay loam soil. Some of its physical and chemical properties are given in Tables I and II. The water content at field capacity varied from 20.8 to 28.9%.The permanent wilting point varied from 11.7 to 13.7% on dry weight basis. Soil bulk density values ranged between 1.40 and 1.70 g-cnr3. The available water holding capacity for 0.90 m. soil profile was 181 mm. TABLE I. PHYSICAL PROPERTIES OF THE EXPERIMENTAL SOIL Texture Soil depth (cm) 0-30 30-60 60-90 90 - 120 clay loam clay loam clay loam clay loam F.C. P.W.P. (%) (%) 28.9 25.0 23.9 20.8 12.1 13.7 13.0 11.7 Bulk density (g.cm-3) 1.40 1.62 1.69 1.70 TABLE n. CHEMICAL PROPERTIES OF THE EXPERIMENTAL SOIL Soil depth (cm) pH 0-20 20-40 7.05 7.15 T.S.S. (%) CaCO3 (%) O.M 0.085 0.059 10.85 13.85 1.90 2.00 (%) C.E.C. [meq-(lOOg)-1] 32.08 28.25 The soil was neutral in reaction and low in organic matter. CaCOs and cation exchange capacities ranged between 10.85 and 13.85% and 28.25 and 32.08 meq-(lOOg)-1, respectively. Total soluble salts ranged between 0.059 and 0.085%. The available phosphorus and total nitrogen in soil were 0.25 ppm and 0.13%, respectively. The crop was seeded on May 13 in 1992 and May 18 in 1993 and 1994 with a row spacing of 0.70 m. and a within-row spacing of 0.25 m. The population density of the cotton crop was 57000 plants per ha. The experiment was set up in split plot design with six irrigation treatments as main plots and three nitrogen fertilizer rates ( 0,60 and 120 kg-ha*1) as subplots, with four replications.The harvested portion of each of the 4.2 by 6.0 m plots (25.2 m2) was 2.8 by 4.0 m (11.2m2). Inasmuch as vegetative growth continues during flowering and boll formation, and flowering continues during boll formation, the distinction of cotton growth stages is difficult. Therefore, the growth stages were divided into three phases following Doorenbos and Kassam [1]: 227 1. Vegetative stage from 25 days after emergence until 5% of flowering opening 2. Flowering stage from 5% until 70-80% flowering opening 3. Boll formation stage from 70-80% flower opening until opening first bolls TABLE HI. IRRIGATION TREATMENTS Treatment*" Vegetative Flowering B oil formation Description Controls 111 Tr. 000 1 — 0 1 — 0 1 — 0 Full irrigation Traditional practice Continuous stress 1 1 0 Stress at vegetative Flowering Boll formation One stress Oil 101 110 0 1 1 1 0 1 î (1) Normal watering, (0) Water stress In the full irrigation treatment the soil water content was allowed to be depleted to 60% of available water content of the plant root zone (0.90 m), whereas in the stress condition it was depleted to 20-25% during the specific growth stage. The soil water was at field capacity on all plots at sowing. Irrigation water was applied to furrows by perforated pipes. The traditional irrigation, generally adopted by the farmers in the region is to apply irrigation 4 or 5 times with a 15- to 25-d interval particularly during the flowering stage. Two nitrogen fertilizer rates were applied - a high rate of 120 kg N-ha'1 and a low rate of 60 kg N-ha'1. For a more reliable statistical analysis, control plots which received no nitrogen fertilizer were also included. In order to determine fertilizer N recovery and fertilizer use efficiency, 15N-labelled fertilizers were applied to microplots in the full irrigation (111) and the continuous stress (000) treatments, and in the plots receiving the high rate of fertilizer. All plots received 80 kg PaOs-ha-1 and 100 kg K2O-ha-1. Soil water content to a depth of 1.50 m was frequently monitored for each treatment with gravimetric sampling (0-0.20 m) and the neutron probe (0.20-1.50 m). The neutron probe was calibrated in situ. Within one replication, each of the irrigation treatments was equipped with two access tubes for determining soil water content in 0.15-m. increments. Around one of the access tubes, tensiometers were installed at depths 45, 60, 80 and 110 cm for determining soil water potential. Measurements of soil water content and water potential were taken frequently throughout the growing season in order to calculate crop actual evapotranspiration using the water balance equation 228 C-D±AS where ET is evapotranspiration (mm), / the irrigation (mm), P the precipitation (mm), C the capillary (upward) rise (mm), D the downward drainage (mm) and AS the change in soil water storage (mm). The amount of irrigation water to be applied for each irrigation was determined as the amount necessary to replenish the 0 to 0.90 m soil profile to field capacity. After cotton was harvested by hand in two pickings (October and November), seed cotton yields were determined. Water use efficiencies (WT/Ef)were calculated based on total depth of irrigation water and seasonal evapotranspiration (WUEc). In order to evaluate the sensitivity to water stress, yield response factor ky was calculated for each irrigation treatment. Yield response factor ky defined as the ratio of relative yield decrease to relative evapotranspiration deficit was calculated from Ym A ^\ ET - m) l L li where Ya is the actual yield (kg-ha-1), Ym the maximum yield (kg-ha-1), ETa the actual evapotranspiration (mm) and ETm the maximum evapotranspiration (mm). Yield-N index defined as the ratio of kg seed cotton production to kg of applied fertilizer N was calculated for all treatments. Total N and 15N contents of the cotton parts (leaf, stem, burs, seed and lint) were measured and the same procedure was repeated for the soil samples taken from the labelled microplots. The data was used to determine the total N yield (uptake), fertilizer N yield, soil N yield and nitrogen use efficiency. 3. RESULTS AND DISCUSSION 3.1. Water requirement and consumptive use Because the rainfall during the three growing seasons of experiment was extremely low, desirable levels of water stress were easily achieved. Rainfall during May through October totaled 7.6, 2.5 and 6.2 mm in 1992, 1993 and 1994, respectively. Each year, June, July and August were almost completely dry. Number of irrigations, seasonal evapotranspiration and depth of irrigation water applied to different treatments are given in Table IV. Eight irrigations were applied to plots receiving the full irrigation treatment. The continuous stress treatment was achieved with 3 or 4 irrigations throughout the growing season. Other treatments received 6 or 7 irrigations. The traditional irrigation treatment normally adopted by the farmers included 4 or 5 irrigations. Although drought conditions were very severe during the experimental period, irrigation schemes of Aegean Region were sufficient to apply 2 irrigations in July and August during the peak water consumption period of cotton. ET during the first year of the experiment in 1992 is not given because of problems in obtaining some water balance components. Seasonal ET increased with increased number and 229 TABLE IV. IRRIGATION TREATMENT EFFECTS ON NUMBER OF IRRIGATIONS, SEASONAL ET AND IRRIGATION WATER APPLIED Year 1992 1993 1994 Number of irrigations Seasonal ET (mm) Irrigation water (mm) Relative El Relative irrig, water (%) (%) 111 110 101 Oil 000 Tr. 8 7 6 6 4 4 — — — — — —— 576 552 515 521 471 462 — — — — — — 100.0 95.7 89.3 90.3 81.7 80.2 Ill 110 101 Oil 000 Tr. 8 7 6 6 4 647 588 5 834 763 701 734 659 702 544 574 476 553 100.0 91.5 84.1 88.0 79.0 84.2 100.0 90.9 84.1 88.7 73.6 85.5 Ill 110 101 Oil 000 Tr. 8 7 6 6 3 5 899 828 718 748 680 844 751 703 564 640 424 626 100.0 92.1 79.9 83.2 75.6 93.9 100.0 93.6 75.1 85.2 56.5 83.4 Treatment depth of irrigation water applied (Table IV). The highest seasonal ET occurred in the full irrigation treatment obviously owing to an adequate soil water supply during the entire growing season. As was expected the lowest ET occurred in the continuous stress treatment. On the other hand, ET values were nearly identical for treatments 101 and 011. Seasonal ET varied between 659 and 834 mm in 1993 and between 680 and 899 mm in 1994. Similar JET values have been reported by several researchers under the climatic conditions similar to those of this experiment [2,3]. Cumulative evapotranspiration versus time for 1993 and 1994 are shown in Figs. 1 and 2, respectively. The cumulative ET values reflect the soil water stress treatments created by different numbers and amounts of irrigations. The maximum ET rates were generally obtained in 230 mid-August during the peak blooming stage of crop growth. The period of peak ET rate was almost the same in both years. The seasonal ET in 1993 was lower than that in 1994. Some climatic factors may have caused this difference inasmuch as the temperature was higher in 1994. Except June (Fig. 3) the average monthly temperatures were continuously higher during growth period with the temperature variations in July, August, September and October being 0.6, 1.2,3.2, and 1.4 °C, respectively. 1 JUN I JUL r ÄUG I SEP T TREATMENT 111 110 Oil TR 101 000 900 0 \ OCT 600 300 1993 B 0 0 40 80 120 160 DAYS AFTER EMERGENCE Fig. L Cumulative ET in the different irrigation treatments (1993). l JUN l JUL l ÄUG SEP 900 OCT I TREATMENT 111 TR 110 Oil 101 000 600 1—4 H 1993 P U 0 0 40 80 120 160 DAYS AFTER EMERGENCE Fig. 2. Cumulative ET in the different irrigation treatments (1994). 231 Irrigation water amounts, determined on the basis of soil water storage depletion within the 0.90 m of plant root zone ranged between 462 to 576 mm in 1992,476 to 647 mm in 1993 and 424 and 751 mm. in 1994. In 1992,1993 and 1994,18,26. and 44% of irrigation water was saved in continuous stress condition treatments, respectively. In the treatments where water stress was created at different growth stages the irrigation water savings ranged between 4 to 25% (Table IV). Soil water balance calculations showed that downward drainage below the crop root zone (0.90 m) and upward flow from this zone were particularly larger in the stress created treatments. Immediately following an irrigation, downward flow initially occurred and several days after each irrigation, the flow reversed to the upward direction for almost all of the treatments. However, the upward flow amounts of the water balance component were larger than g g S - j May \ I1 - ,_ » MONTHLY AVERAGE TEMPERATURE ( °C) OT993 CÜ1994 Jun Jul j Aug -i '\ - Sep Oct Fi£. 3. Monthly temperatures (°C) during growth periods. the amounts of percolation in all treatments. The greatest percolation and upward flow amounts were obtained in the continuous stress treatment, whereas these amounts were least in the full irrigation treatment. Contributions of upward flow to the seasonal ET were 11% in 1993 and 15% in 1994 for the 000 treatment and 5% in 1993 and 3% in 1994 for the 111 treatment. Similar ratios of upward flow have been reported by van Bavel et al. [4] for sorghum. The researchers attributed some part of this contribution to direct root absorption from deeper zone rather than from capillary rise. 3.2. Seed cotton yield Seed cotton yields (Table V) showed no statistically significant interactions between irrigation and nitrogen treatments. On the average, the highest seed cotton yield was obtained in the full irrigation treatment and the lowest in the continuous stress treatment In comparison with the 111 treatment, the yield decrease ratios in 000 varied between 24 and 31%. Seed cotton yields in the treatments of 111 and 110 were close to each other with low (3 to 8%) yield 232 TABLE V. SEED COTTON YIELDS (kg-ha-1) UNDER DIFFERENT IRRIGATION AND NITROGEN TREATMENTS Year Treatment 1992 111 110 101 Oil 000 Tr. ON 2379 2326 2310 1906 1759 1674 Seed Cotton Yield 120 N 60 N 3055 2905 2545 2464 2331 2420 3228 3219 2884 2705 2410 2241 Sx 0.05 1993 111 110 101 Oil 000 Tr. 1994 S, 0.05 2888 2817 2580 2359 2166 2112 100.0 97.5 89.3 81.7 75.0 73.1 93.6 2985 2680 2677 2598 1951 2305 3553 3021 3096 3439 2401 2824 3647 3452 3263 3366 2503 2898 Sx 0.05 111 110 101 Oil 000 Tr. Average Relative yield (%) 3312 3051 3012 3134 2285 2676 100.0 92.1 90.9 94.6 69.0 80.8 177.2 3067 2803 2462 2629 2455 2742 3442 3135 3141 2830 2636 3419 3851 3765 3313 3111 2765 3468 3453 3234 2972 2856 2618 3210 100.0 93.7 86.1 82.7 75.8 92.9 157.6 233 decrease ratios. Yield decreases in Oil were higher than those in treratments 110 and 101 in 1992 and 1994. However, yield differences were not significant between these treatments in 1993 (Fig. 4 ). 3500 E31992 E£]1993 £31994 - ^ 1500 III 110 1 £ 101 .'• y , • V*" • "."•• * ,**y. Oil PI •'• •'*• v//////////* U 1 VI KXXXXXXXX^ O /.v V////////////// 2000 - kXXXXXXXXXXXXXt H H '"• "•* ^ 7?. '/////////////A G § m^ V7/////////777X 2500 ':• '. KXXXXXXXXXXXXXN 3000 _ 000 TR TREATMENTS Fig. 4. Seed cotton yields in the different irrigation treatments. ' The seed cotton yields were consistently higher in high nitrogen application for all irrigation treatments. These results reveal that cotton is most sensitive to water stress at its vegetative stage. The flowering stage was also sensitive. On the other hand, no significant differences were found between the full irrigation treatment and the treatment causing water stress at boll formation stage. In design and management of irrigation systems for deficit irrigation, the irrigation planners must rely upon water production functions that relate water consumption to crop yields. Based on average seed cotton yield and ET, the following linear functions were derived (Fig..5) for 1993 and 1994, respectively, F = -853 + 5.14£T r 2 =0.71 Y= 354 + 3.44£r r 2 =0.92 where Fis the seed cotton yield (kg-ha4) and ET the seasonal evapotranspiration (mm). Bilgel and Kanber [5], Bastug and Tekinel [6] and Sammis [7] reported linear production functions for cotton. Linear functions are most likely when ET deficits are distributed over several growth periods or when irrigation applications achieve programmed depletions of root zone water [8]. The water production function results indicate a strong relationship between seed cotton yield and ET. When such a relationship between yield and ET occurs, the sensitivities to water stress at different growth stages are close to each other. 234 3400 Ö J W 1993 -853 + 5.14 x ^ = 0.71* 1994 = 354 + 3.44 x r2 = 0.92** 3000 2600 c» 2200 600 700 800 900 ET (mm) Fig. 5. Seasonal ET - seed cotton yield relationships. 3.3. Effect of N rates Significant responses were observed between fertilizer N rates and seed cotton yields and increasing trends were apparent. Excluding the first year of study, second and third year results showed statistically similar tendencies (Table VI). TABLE VI. SEED COTTON YIELDS, FERTILIZER N RATES AND YEARS Seed cotton yield (kg-ha-1) Fertilizer N rate (kg-ha-1) 0 60 120 S, 0.05 1992 1993 1994 2059 2619 2781 2533 3014 3188 2692 3100 3378 57 54 68 Seed cotton yields ranged from 2059 to 2781, 2570 to 3188 and 2692 to 3378 kg-ha-1 for the years 1992,1993 and 1994, respectively. The yields for the second and third years were higher than those of the first year and may be attributed to the increased availability of soil N in the fertilizer added plots [9] of 1993 and 1994 study years. However, because the experimental field was fallow and unfertilized during the year preceding the 1992 experiment, N uptake by the first year crop might have been low. 235 An overall evaluation of the 3-year experiment showed that fertilizer N rates had quadratic effects on yield particularly in the 000, Oil and Tr irrigation treatments which are considered the "low yielding treatments". The quadratic relations indicate that 100 kg N-ha'1 is the optimum fertilization dose for such irrigation management. The results were also evaluated with respect to yield increases achieved by each unit of added fertilizer and were identified as "yield - N index". Each year and generally almost in every irrigation treatment, the index decreased implying that there is a decrease in crop production with high fertilizer N doses (Table VII). Table VII shows the yearly and average indices where the declines were much more typical. The 011,000 and Tr "low yielding treatments" generally had the low index which confirms the fertilizer N rate - yield relations in the present study. Similar results for wheat have been reported by Lathwal et al. [10]. TABLE VE. YIELD - N INDEX AND IRRIGATION TREATMENTS Yield - N index Irrigation treatment 111 110 101 Oil 000 Tr. 1992 60 N 120 N 1993 60 N 120 N 1994 60 N 120 N Average 60 N 120 N 11.3 9.7 3.9 9.3 9.5 12.4 9.5 5.6 6.9 14.0 7.5 8.9 6.3 5.5 11.3 3.4 3.0 11.2 9.0 6.9 7.4 8.9 6.7 10.8 7.1 7.4 4.8 6.7 5.4 4.7 5.5 5.2 4.8 6.4 4.6 4.9 6.5 8.0 7.1 4.0 2.6 6.0 6.4 6.8 5.6 5.7 4.2 5.2 Total N yields (uptake), fertilizer N yields, soil N yields and nitrogen use efficiencies of cotton were also examined under full irrigation and continuous stress conditions for high fertilizer N application rate of 120 kg N-ha-1. Before proceeding any further, if average total N concentration (%) of plant tissues and total dry matter production (kg-ha*1) are examined in relation to irrigation, apparent slight decreases in N concentration and increases in dry matter are seen which resulted in higher total N yields (uptakes) for full irrigation treatment (Table VIE). The average of three years data shows that 166 and 225 kg N-ha-1 were taken up by cotton in the 000 and 111 irrigation treatments, respectively. Similar effects of irrigation for cotton have also been reported by Doss and Scarsbrook [11]. Fertilizer N yield and soil N yield results revealed that recoveries were always greater in full irrigation conditions. And N derived from soil was always higher than N from fertilizers. 236 TABLE Vffl. NITROGEN CONCENTRATION (%) IN PLANT TISSUE AND TOTAL DRY MATTER PRODUCTION (kg-ha-1) 1992 000 Avarage total N concentration^) 111 1993 1994 000 111 000 111 1.87 1.80 1.83 1.60 1.81 1.75 8824 11377 9235 13572 9060 14514 Total dry matter product.(kg-ha-!) The situation is very distinct at the final year when 23 kg N-ha'1 was taken up from the fertilizer and 141 kg N-ha'1 from the soil in the 000 irrigation treatment and 40 and 214 kg N-ha4 in 111 treatment, respectively (Table IX). The apparent increase in the uptake of soil N with and without stress conditions could be attributed to the stimulation of microbial activity by the addition of N fertilizer, mineralization of organic matter and increased availability by the water [9,12]. As for N use efficiencies, on the average 19% of nitrogen in stress conditions and 29% in full irrigation were found efficient. These results were relatively low compared to previous findings [9] for wheat and corn [12], higher than reported by Atta and Van Cleemput [13] for sesame and sunflower and confirmed by Hearn and Constable [14] for cotton imposed to stress conditions. The conflicting data could be mainly due to crop variety, rooting system and other cultural managements. The distribution of N uptake in different plant parts i.e. leaves, stems, burs, seed and lint revealed that 45% of the total uptake in the 000 treatment and 39% in the 111 treatment were recovered by the seeds. The second highest recovery was achieved by the leaves. A parallel trend observed in the partitioning of labelled N in the various plant parts are in agreement with other investigations [13]. 3.4. Water use efficiency and yield response factor Water use efficiency calculations based upon irrigation water (WUEf) and upon seasonal evapotranspiration of crop (WUEC) are presented in Table X. Crop water use efficiency indicates the yield produced per unit of water consumption. WUEC values were the lowest in continuous stress conditions. For the treatments imposing water stress at different growth stages, the values of WUEC were nearly the same. Consequently, no clear distinction can be made between these stress treatments. In 1994, however, the value was the lowest in the 011 treatment 237 TABLE IX. TOTAL N YIELD, FERTILIZER N YIELD, SOIL N YIELD AND N USE EFFICIENCY OF COTTON Irrigation treatments Year 000 111 Total N yield (kg-ha-1) 1992 1993 1994 Average 165 169 164 166 203 217 254 225 Fertilizer N yield (kg-ha-i) 1992 1993 1994 Average 20 25 23 23 25 38 40 34 Soil N yield (kg-ha-1) 1992 1993 1994 Average 145 144 141 143 178 179 214 190 Nitrogen use efficiency (%) 1992 1993 1994 Average 17 21 19 19 21 32 33 29 The lowest yield response factor ky was obtained from treatment 110 - 0.63 in 1993 and 0.28 in 1994 (see Table X). The yield response factors to water deficit for the entire growing season were 1.49 and 1.15 in 1993 and 1994, respectively. Although scattered values of ky were obtained in stress created treatments, the highest yield response factor of 1.11. was found for the 011 treatment in 1994 (Fig. 6). Values of ky equal to 0.50 for flowering and 0.85 for the entire growing season have been reported by Doorenbos and Kassam [1]. They also explained the differences in ky values based upon variations in climate, level of EFand soil. Differences can be attributed to the variations in above factors and also varying response to water stress of the new cotton variety N84 used in the experiment 238 m 1.0 0.8 0.6 0.2 0.4 0 0 0.2 0.4 0.6 0.8 1993 1.0 l - ET.-ET» 1.0 0.8 0.6 0.4 1 0.2 0 0 0.2 0.4 0.6 0.8 1.0 Fig. 4. 6. Relationships between relative yield decrease and relative ET deficit. CONCLUSION This experiment has indicated that the sensitivity of the cotton plant to water stress at different growth stages is nearly identical. These results might be attributed to the high water holding capacity [15] of the soil and to the water moved in from soil outside the depleted root zone to alleviate stress [16]. On the other hand, Ayars et al., [17] and Taylor and Klepper [18] emphasize the adaptive mechanisms of the cotton plant to adjust root and shoot growth to alleviate the impact of water stress. Neverthless, the seed cotton yields, ky values of 1994 and yield - N indices revealed that the vegetative stage was more sensitive to stress conditions. On the other hand, the above parameters for the boll formation stage were only slightly affected. Therefore, under limited water resource conditions which exist in most of the irrigation projects 239 TABLE X. IRRIGATION TREATMENT EFFECTS ON WUEf> WUEC AND YIELD RESPONSE FACTORS Year 1993 1994 Treatment WUEf kg-(ha-mm)-1 111 110 101 Oil 000 Tr. 5.64 5.87 6.00 5.86 5.26 5.24 4.37 4.52 4.65 4.59 3.80 4.13 0.63 0.66 0.64 1.49 1.30 Ill 110 101 Oil 000 Tr. 5.13 5.36 5.87 4.86 6.52 5.54 4.28 4.55 4.61 4.16 4.07 4.11 — 0.28 0.69 1.11 1.15 1.63 WUEC kg-flia-mm)-1 ky — of Turkey, the vegetative growth stage of cotton should be given priority for irrigation, followed by the flowering stage. Under such conditions, irrigating cotton during only the vegetative and flowering stages would result hi only a 2.5 to 7.9 % yield reduction with a water savings of 4.3 to 9.1%. High seed cotton yields were achieved with high fertilizer N rates. Fertilizer N yields (uptake) and soil N yields revealed that recoveries were always higher in full irrigation conditions at high nitrogen rates. Apparent increases in soil N uptake with and without stress conditions were determined. The highest and next highest recoveries were achieved by the seed and by the leaves, respectively, in both full and continuous stress conditions. Nitrogen use efficiencies were always higher in the full irrigation treatments. It is concluded that irrigation has an important effect on the efficiency of fertilizers and that even high fertilizer N applications cannot improve the efficiency during water stressed conditions. REFERENCES [1] DOORENBOS, J., KASSAM, A.H., Yield response to water. FAO Irrigation and Drainage Papers No. 33, FAO, Rome (1979). [2] KAMBER, R. et al., Yields and comperative performance of different crop production 240 [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] functions of cotton as influenced by deficit irrigation, Doga, Turkish J. of Agric. and Forestry 14, Ankara, Turkey (1990). GRIMES, D.W. et al., Functions for cotton (Gossypium hirsutum L.) production from irrigation and nitrogen fertilization variables: I. Yield and evapotranspiration, Agron. J. 61 (1969). VAN BAVEL, C.H.M. et al., Hydraulic properties of a clay loam soil and the field measurement of water uptake by roots: II. The water balance of the root zone, Soil Sei. Soc. Amer. Proc. 32 (1968). BILGEL, L., KAMBER, R., The comparison of the irrigation season lengths on cotton under Harran Plain conditions in Urfa. Univ. of Çukurova, J. Sei. Eng. 2, Adana, Turkey (1987). BASTUG, R., TEKINEL, O., Water production functions of cotton under limited irrigation water conditions, Doga, Turkish J. of Agric. and Forestry 13, Ankara, Turkey (1989). SAMMIS, T.W., Yield of alfalfa and cotton as influenced by irrigation, Agron. J. 73 (1981). STEGMAN,E.C, et al., Irrigation Water Management - Adequate or Limited Water, Irrigation - Challenges of the 80's, Proc. Second Nat Irr. Symp. ASAE (1981). RAO, A.C.S. et al., Influence of added nitrogen interactions in estimating recovery efficiency of labelled nitrogen, Soil Sei. Soc. Amer. J. 55 (1991). LATHWAL, O.P. et al., Effect of N and irrigation levels on N use efficiency in wheat, Hayrana Agric. Univ. J. Res. 22, India (1992). DOSS, B.D., SCARSBROOK, C.E., Effect of irrigation on recovery of applied nitrogen by cotton, Agron. J. 61 (1969). REDDY, G.B., REDDY, K.R., Fate of Nitrogen -15 enriched ammonium nitrate applied to corn, Soil Sei. Soc. Amer. J. 57 (1993). ATTA, S.K.H., VAN CLEEMPUT, 0., Field study of the fate of labelled fertilizer ammonium-N applied to sesame and sunflower in a sandy soil, Plant and Soil 107 (1988). HEARN, A.B., CONSTABLE, G.A., Irrigation for crops in a subhumid environment VII. Evaluation of irrigation strategies for cotton, Irrig. Sei. 5 (1984). MUSICK, J.T., DUSEK, D.A., Irrigated corn yield response to water, Trans. ASAE 23 (1980). KOCK, de J. et al., The relative sensitivity to plant water stress during the reproductive phase of upland cotton (Gossypium hirsutum L.), Irrig. Sei. 11 (1990). AYARS, J.E. et al., Cotton response to nonuniform and varying depths of irrigation, Agric. Water Mgt 19 (1991). TAYLOR, H.M., KLEPPER, B., Water relations of cotton. I. Root growth and water use as related to top growth and soil water content, Agron. J. 66 (1974). Next page(s) left blank 241 YIELD RESPONSE OF COTTON, MAIZE, SOYBEAN, SUGARBEET, SUNFLOWER AND WHEAT TO DEFICIT IRRIGATION C. KIRDA, R. KANBER, K. TÜLÜCÜ University of Çukurova, Faculty of Agriculture Adana H. GÜNGÖR Rural Affairs Research Institute, Eskisehir, Turkey Abstract Results of several field experiments on deficit irrigation programs in Turkey are discussed. Deficit irrigation of sugarbeet with water stress imposed (Le. irrigation omitted) during ripening stage saved nearly 22% water, yet with no significant yield decrease. An experiment, conducted in Trakya Region the European part of Turkey, and aimed at studying water production functions of sunflower (i.e. yield versus water consumption) revealed that water stress imposed at either head forming or seed filling stages influences yield the least with 40% savings of irrigation water supply compared with traditional practices in the region. Water stress imposed at vegetative and flowering stages of maize hindered the yield most significantly. The results showed that deficit irrigation can be a feasible option under limited supply of irrigation if stress occurs during yield formation stage. A four year field experiment aiming at developing deficit irrigation strategies for soybean showed that soybean was most sensitive to water stress during flowering and pod filling stages, and irrigation during these stages would ensure high yields. Results of experiments on cotton showed that irrigations omitted during flowering and yield formation stage did not significantly hinder the yield. Similarly, wheat gives good yield response depending on wheather conditions if irrigated at booting, heading and milking stages. In areas where rainfall at planting is limited, supplementary irrigation during this period can ensure good establishment of a wheat crop. 1. INTRODUCTION Crop yields under irrigated agriculture are several fold higher than rainfed dry farming systems. Investments for irrigation are usually top priority in all countries of arid and semi-arid regions. However, it has become a matter of serious concern in recent years that despite their high costs the performance of many irrigation projects has fallen short of expectations as a result of inadequate water management both at farm and system level. Crop production has been well below the project targets. Among the causes of problem are poor standard of operation and maintenance of irrigation networks and low irrigation efficiencies resulting from traditional habits of farmers to apply excess water which contribute more than one-third of total water waste in the irrigation projects. Excess water application at farm level compounded with seepage water along irrigation networks additionally causes rising of ground water table, triggering further soil 243 salinity problem. Therefore measures to increase effective use of water at farm levels are critical in sustaining and increasing agricultural production in irrigated areas. An increase in world water crisis is a present day reality faced by all nations. Industrial use of water is competing heavily with agricultural use. Industrial use of water is favored to the detriment of agricultural production which is needed to sustain food and fiber production for increasing population. Therefore, high efficiency in water use is of high priority. Generally, irrigation and irrigation water requirement of crops were determined without any consideration of likely water deficient or limitation of available water supplies. Many irrigation schemes were designed for situations where water availability and deficiency may be a major constraint to plant growth and high yields. However, in arid and semi-arid regions, because of increasing allocations of water for municipal and industrial use, major changes came about in water use under irrigated agriculture. New inovations had to be tested and adapted to increase effective use of decreasing water allocations for agricultural use. Research effort has focused on developing new techniques to receive high returns from restricted supply of water. Recently published works suggest new innovations which can improve traditionally used irrigation practices and thereby increase effective use of water. For example, it has been reported that exposing field crops to water stress at specific growth stages may not cause significant yield reduction and therefore irrigation during these stages can be omitted and excess water left in the system can be diverted to other areas. Stegman [1] reported that corn yield, irrigated with trickle irrigation which maintained near zero water potential within the plant root zone, was not statistically different than the yield obtained with sprinkler irrigation which allowed 30-40% depletion of available water content between the irrigation intervals. Stöckle and James [2], using a growth simulation model, concluded that slight water stress for corn (i.e. the ratio of actual to potential transpiration > 0.89) could provide higher net benefit than full irrigation. Ziska and Hall [3] reported that cowpea had the ability to maintain seed yields when subjected to drought during the vegetative stage as long as subsequent irrigation intervals were not too great For example, an 8-day irrigation interval following a drought period during vegetative stage produced the highest yields and water-use-efficiency. Therefore, they could conclude that water use of cowpeas can be reduced while maintaining seed yields by deficit irrigation (i.e. planned water-deficit irrigation). Soybean was also subjected to extensive research regarding its response to deficit irrigation. Korte et al. [4] reported on genotypic response of soybean regarding effects of irrigation on reproductive ontogeny. They showed that minimum reduction occurs in pods to plant ratio, seeds to pod ratio and mean weight per seed relative to nonstressed control plants as long as water stress does not coincide with flowering and pod development stages. Results by Eck et al. [5] led them to conclude that soybean is also amenable to limited irrigation and is even more suited to limited irrigation than is corn. Specht et al. [6] found that soybean yield with delayed irrigation until the flowering stage or mid-pod elongation stage was not significantly different when compared with normally irrigated treatments where available soil water content 244 was maintained between 50 to 80% of the total plant available soil water content Kirda et al. [7] reported that earlier water stress followed by later watering was less detrimental to biological nitrogen fixation ability of soybean. Stegman et al. [8] indicated that short term water stress during the early flowering stage of soybean may result in flower and pod drop in the lower canopy, but this effect is frequently compensated by more pod set at the upper nodes if moisture is adequate at later stages of growth. Stress effects are most detrimental to yield if imposed in full pod and seed development stages, suggesting that plants can recover and minimum yield reduction occurs from water stress imposed at early growth stages if irrigation is resumed at later stages [8]. Deficit evapotranspiration is also among the techniques of increasing effective use of water. Deficit evapotranspiration can be used either through agronomic practices or through changing management schemes to decrease crop evapotranspiration ET. Crops are exposed to water stress either throughout the entire growth season or at certain growth stages. It is therefore possible to save irrigation water without significant yield decrease which implies that irrigated area can be increased without additional water supply available [35]. The main approach in deficit irrigation practice is to increase crop water use efficiency by eliminating those irrigations with the least impact on crop yield. In the areas where water supplies are limited and unit water costs are expensive, the best irrigation practice is not necessarily that which gives the highest yield. Additionally, in areas where energy supplies are limited and capital investments are short, deficit irrigation is used as a strategy to increase farmers' income [36]. Alternatively, deficit irrigation is also used to maximise or stabilise regional crop yields. Although deficit irrigation practice is subject to extensive discussions regarding its many features, it is indeed an innovative and noble technique used to maximize income and stabilize food production [37,38]. Among the benefits of deficit irrigation are low production costs, increased irrigation efficiency and a more acceptable cost-benefit ratio of irrigation water. Although crop yields are somewhat decreased under deficit irrigation, benefits from the saved water may be high. Deficit irrigation is promoted widely and used for some crops in Turkey. In this work, results of cotton, maize, soybean, sugabeet, sunflower and wheat are summarised. 2. CASE STUDIES IN TURKEY Irrigation projects are of top priority in all Middle-Eastern countries. Among those, Turkey is to put in operation a new irrigation project, targeted to irrigate 1.6 million ha of agricultural land in Southeastern Anatolia in the near future. Although presently there is no scarcity of water resources in Turkey, demand for water would steadily increase as a result of increasing population (projected to reach 100 million in 2010), rapid urbanization and rapid industrialization. Research teams in the national research institutes have been confronted with what could be done to increase effective and efficient use of water in agriculture. In addition to engineering measures to prevent waste of water at system level, new irrigation technologies like 245 trickle and sprinkler irrigation methods are being promoted to increase irrigation efficiency at the field level. Deficit irrigation is considered another option which could save additional irrigation water when it replaces traditional irrigation habits of growers. Several research institutes in Turkey launched extensive research programs to compare deficit irrigation practices with traditional irrigation schedules as commonly used by farmers. The foEowing sections review and summarize recent research findings on deficit irrigation practices of several field crops in Turkey. 2.1. Cotton Early irrigation experiments on cotton were conducted in 1940s in Çukurova Region [39]. Initial experiments were mainly of demonstrative propose with the main focus being to determine optimum irrigation requirement Deficit irrigation of cotton was first proposed by Tekinel and Kanber [40] who investigated crop water requirement and yield production functions of cotton. They found that a second degree polynomial relation could adequately describe yield response of cotton which showed that as much as 30% reduction in irrigation water application did not appreciably hinder cotton yield. Similar results obtained by Yalçuk and Özkara [41] in Aegean Region, Turkey, showed that a 40% reduction in irrigation water application would not significantly decrease the cotton yield. 11.0 0.8 0.6 0.4 0.2 0 0 0.2 0.4 0.6 COTTON 0.8 1.0 Fig. 1. Relative yield decrease of cotton as afonction of ET deficit [42] Ba§tu| [42] conducted open field irrigation experiments to study the effects of both seasonal and growth-stage specific deficient ET on cotton yield. He tested 3 growth stages of cotton: (1) vegetation, (2) flowering and yield formation and (3) ripening. The relation between 246 relative ET deficit and relative yield decrease (Fig. 1) followed the Steward equation [16]. The yield response factors were 0.99 for the entire season and 0.76 for flowering and yield formation stages of cotton. They indicate that the least yield reduction is obtained when deficit irrigation (water stress) is confined to flowering and yield formation stages than when a general ET deficit throughout the entire growth season. Doorenbos and Kassam [16] reported that seasonal and flowering stage water-deficit yield response factors of cotton grown in deep and medium textured soils changed markedly from 0.85 to 0.50. For cotton, Kanber et al. [43,44] obtained a rather low yield response factor (~ 0.40) for the seasonal ET deficit. It is well documented that yield response factors vary according to ET, wetting depth during irrigation, the particular irrigation program and crop yielding capacity [16,45,46]. Yavuz [47], compared yield response relations of cotton irrigated with different irrigation methods: furrow, drip and single-line sprinkler. Under each irrigation method, he had subtreatments where different levels of irrigation water applications were included. Different yield response factors ky obtained under different irrigation methods (Fig. 2) were attributed to differences in the wetting depth of the soil profile. Differences in wetting depth would normally cause differences in the volume of soil exploited by roots for plant nutrients. 11.0 0.8 0.6 0.4 0.2 0 0 0.2 i 0.4 0.6 g !x, * « TH 0.8 1.0 Fig. 2. Relationship between relative yield decrease and relative ET deficit of cotton under different irrigation systems [47]. Hence, it is not surprising that yield responses differ under different irrigation methods owing to differences in irrigation water application, £Tand crop yield as discussed by Garity et al. [58]. However, proper fertilization designed for different irrigation methods would have minimized the observed yield differences. 247 2.2. Maize In recent years, irrigated corn (Zea Mays) has expanded rapidly in coastal regions of Turkey and has become a widely grown field crop, particularly as a second crop after wheat or barley. Doorenbos and Kassam [16], Barret and Skogerboe [17] and many others discussed results of extensive research on irrigation programs, particularly on feasibility of deficit irrigation for corn under conditions of limited supply of water. Here, the results of a field experiment on deficit irrigation of maize will be summarized. Irrigation treatments were arranged considering three main growth stages: vegetative, silking and ripening stages. Additionally, aforementioned growth stages were further subdivided into sub-stages which therefore enabled to fine tune deficit irrigation programs to maximize yield and crop water use efficiency. Fig. 3, shows that significant savings in irrigation water could be made without any appreciable yield decrease. For example, with two irrigations alone at early silking and ripening stages, the same or even higher level of yield (4,510 kg-ha-1) could be obtained with irrigation water savings of 25 to 34% compared with 3 irrigations. 2.3. Soybean Soybean (Glycine max.), first introduced in 1972 after white fly (Bemisa tabaci) epidemics in cotton grown areas of Turkey, is planted in large areas. It is favored as a second crop after wheat, alternating with cotton. Its growing period is from early June to late September, a period of very dry weather with high evapotranspiration demand. Therefore, high yields can only be ensured with irrigation. 10 -r 8 V/A HIGH YIELD LOW YIELD J= Q J V(l, 3) V(2) 1X2,3)1X2,3) Y(3) Y(2,3) V(l,3) T(2,3) Y(2, 3) T(l) Y(l) W N 0 3 4 5 6 (347mm) (409mm) (490mm) (499mm) NUMBER OF IRRIGATIONS Fig. 3. Effects of deficit irrigation programs on maize yield. Numbers in parentheses show irrigation water application (mm). Main growth stages of maize (vegetation, tasseling and yield formation) are designated as V, T and Y, respectively. The numbers from 1 to 3 under the growth stages correspond to sub-stages from early to late. DRY 248 2 (297mm) Soybean is a crop with many cultivars of different irrigation characteristics [9]. Depending upon soil and climatic conditions, the seasonal water consumption of soybean cultivars have a wide range from 350 to 750 mm [10]. Although their roots can reach to 150 cm, they are usually confined within the top 60 cm of a soil profile [11]. Good germination can be ensured only if the available soil water content is within a range of 50 to 80% [12,13]. If not irrigated, the leaf water potential of soybean is always lower in early hours of the day becomes higher at midday when compared with irrigated conditions [14]. The major constraint to high yield is primarily insufficient soil water content [11] which directly hinders plant height and leaf extension [15]. Crop water requirements of soybean are indeed very complex and influenced significantly by plant ontogeny. Figures 4 and 5 show the soybean response to water-deficit irrigation. 2.4. Sugarbeet Among the earlier studies on sugarbeet (Beta vulgaris) yield response to water by Doneen [18], Erie and French [19], Okman [20], Ziba and Bügin [21] indicated that the level of soil water depletion preceding each irrigation for sugarbeet does not significantly influence yield. Salter and Goode [22], in their work to determine which growth stage of sugarbeet is most sensitive to water stress, found that stress at mid-vegetative stage would be more detrimental to final yield than stress occurring during late-yield formation and ripening stages. Oylukan [23] indicated that six was the most economical number of irrigations for sugarbeet in medium and heavy textured soils of Central Anatolia. Under similar conditions, Günbatili [24] recommended that irrigation of sugarbeet should start when available soil water content is depleted down to 65%. Winter [25] showed that deficit irrigation of sugarbeet could be a feasible option if irrigation water supply is limited. Fig. 4. Relationship between relative yield decrease and relative ET deficit for soybean. 249 EARL Y STRESS « 4 - LATE STRESS 3 - I o c« 0 6 - I-4 - 0 5 4 3 (8%) (18%) (30%) NUMBER OF IRRIGATIONS Fig. 5. Effect of early and late water stress on soybean yield and water use efficiency WUE. The experiment summarized here was conducted in Porsuk Plain of Central Anatolia of Turkey. The experimental site has alluvial soils of clay texture, developed at the delta cone of the Porsuk River. The experiment aimed to determine at which growth stage of sugarbeet could irrigation be omitted (i.e. deficit irrigation). Six irrigation periods were identified with a total of 64 irrigation treatments defined as different combinations of 6 periods and 6 irrigations. A local variety Turkseker I was sown on the date generally practiced in the region, within the first week of April. Treatments received 200 and 100 kg-ha*1 of N and P2Û5, respectively. Results convincingly showed that 6 irrigations of sugarbeet, as usually practiced by the farmers in the region do not bring extra yield benefit, and 5 irrigations (with one irrigation omitted during midripening stage) produced the highest yield (Fig. 6). 250 Depending on the period when irrigation is omitted, deficit irrigation practice may give significantly different yields, some of which may not necessarily be different than full irrigation. For example, in case of 4 irrigations alone, the same level of yield as was obtained with 5 irrigations can be attained if the omission of one irrigation is made during late vegetative stage (Fig. 6). Similarly, with 2 irrigations alone, yields similar to those for 3 or 4 irrigations can be obtained if the irrigations are applied during plant growth stages either tolerant to water stress or giving little yield response to irrigation. As confirmed with the results presented here, it appears that except during emergence and early growth periods, sugarbeet does not seem to be very sensitive to moderate water deficits, and better yield response is received if restricted water supply is made available to irrigation during early growth stages rather than late yield formation and ripening stages (Fig. 7). 2.5. Sunflower Sunflower (Heliantus annuus) oil is highly demanded not only for human consumption but also for chemical and cosmetic industries. After oil extraction, the remaining cake containing 30% protein, 19% carbohydrates, 8% oil and minerals is also on high demand as animal feed stuff. The sunflower heads, after harvesting seeds, are chopped, ground and also fed to animals. LOW YIELD J2/2 HIGH YIELD 4 3 NUMBER OF IRRIGATIONS Fig. 6. Sugarbeet yield as influenced by the number and timing of irrigations in relation to specific growth stages. Columns having the same letter are not significantly different at the 5% level. 251 XX/1 EARLY STRESS LATE STRESS 2 3 4 (40-45%) (25-35%) (10-20%) NUMBER OF IRRIGATIONS Fig. 7. Differential effects of water stress imposed at either early or later stages of sugar beet growth. (55-60%) Even the stems and seed shells are used as fuel in the villages. Sunflower is widely grown in the Trakya Region in the northwestern European part of Turkey. Although sunflower is known as a drought tolerant crop, substantial yield increases are achieved with irrigation. Although there are numerous research reports on yield response of sunflower to water, only a few will be cited here. Decau et al. [26] showed that irrigation of sunflower not only increases seed yield but oil content as well. Osman and Talha [27] in Egypt found that irrigation water quantity and frequency of irrigation both influence seed and oil yield of sunflower. Moreover, Karami [28] in Iran found similar results. Data by Browne [29] showed that the date of final irrigation of sunflower influences seed yield and all other yield attributes. Final irrigation made during early yield formation stage resulted a 19% increase of seed yield when compared with the final 252 irrigation being made during the vegetative stage. Results of Bhattaacharya and Sarkar [30] indicated that the higher the available soil water content maintained throughout the growing season of sunflower, the higher would be the plant growth and photosynthetic rates and leaf area index. Jana et al. [31] compared the effects of irrigation made during different growth stages of sunflower on yield, water consumption and water use efficiency WUE. Irrigations during vegetative and yield formation stages gave the highest WUE. Harman et al. [32] reported that sunflower irrigated throughout flowering stage gives better yield response when compared those having irrigations made either earlier or later growth stages. Rawson and Turner [33] found that highest sunflower yield could be ensured with frequent irrigations. However, under restricted supply of irrigation water, they found that a single irrigation made 3 weeks before pollination also gave a good yield. The experiment summarized here using sunflower variety Sunbred277cv. is courtesy of Karaata [34] who provided the data. A randomized complete-block field design was adopted to identify the most critical growth stages of sunflower to water stress on a non-calcareous brown soil. Three growth stages, (1) heading (late vegetative stage), (2) flowering and (3) yield formation stages were considered. Treatments of deficit irrigation were formed either omitting irrigation during a certain stage or cutting irrigation water by applying 40 or 60% less water than actually required. The treatment where irrigation applied throughout all 3 growth stages was used as a reference to decide irrigation times for the other treatments. Results showed that water stress developed during different growth stages influenced yield and all other yield-related attributes, leaf area index LAI, photosynthetic rate etc. in different manners. For example, irrigation confined to heading stage only caused more vegetative growth than an irrigation during other stages. On the other hand, irrigation during flowering promoted both vegetative and generative growth. Although irrigation during yield formation stage had no effect on vegetative growth, it increased yield. Tolerance of sunflower to water stress either for the entire season or during certain growth stages is illustrated using crop yield response relations in Fig. 8. The growth stage which is most responsive to irrigation was flowering stage. Compared with yield formation and late vegetative stages (Fig. 8), irrigation during flowering stage ensured the least yield reduction of sunflower. 2.6. Wheat Regarding irrigation of wheat, early studies in Turkey dealt with optimum number of irrigations which could be recommended in a given region. For example, in numerous studies completed in Central Anatolia, Eski§ehir, Konya, Isparta and Ankara, it was found that wheat requires 2-4 irrigations with the first irrigation being before or immediately after planting in October depending on the region. The other irrigations must be in the spring during either booting, flowering, heading or milking stages [48, 49,50]. Similar results in the same region were later obtained by Güngör and Ogretir [51], Aran and Kivanç [52], Uzunoglu [53] and Çetin [54]. 253 0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 l 1.0 0.8 0.6 0.4 L - ETa-ETm- 0.2 1.0 0 1 Fig. 8. Relationship between relative yield decrease and relative ET deficit for sunflower water stressed throughout the season or only irrigated during one or two growth periods. 254 Research on irrigation of wheat was also conducted in transitional areas from typical continental climatic zones of Central Anatolia, to mountainous high lands and to arid regions [55]; in the southeastern region of Turkey in Urfa [56]; in the high mountainous areas in Erzurum [57]. General guidelines for irrigating wheat in Turkey can be ascertained from the results summarized in Table I. If irrigated at booting, heading and milking growth stages, wheat gives good yield response. Nevertheless, climatic conditions should be considered. For example, in the transitional areas of high rainfall 2 irrigations are adequate, whereas in southeastern part of Turkey where rainfall is low, 3 or even 4 irrigations may be needed to achieve a good yield response. In the areas like Eastern and Central Anatolia, respectively, where spring and fall rains aie limited, irrigation of wheat at planting becomes important. Yield response curves from data summarized in Table 1 are plotted in Fig. 9. In Tokat, where annual rainfall is relatively high and uniformly distributed over the growing season of wheat, the lowest crop yield response factor was obtained. Hence, this low factor indicates that TABLE L EFFECT OF DIFFERENT PROGRAMS ON WHEAT YIELD Information source 1 Irrigationt Irrigation program 2 3 4 applied 5 (mm) Madanoglu, Ankara XXX Gimbatili, Tokat X Karaata, Urfa X X XX Sevim, Erzurum XX Yakanand Kanburoglu, Trakya XX X X X 228 ET Yield (mm) (kg-ha-1) 690 Yield difference with no irrigation (kg-ha-1) 4,160 +1,730 253 616 3,740 +390 440 728 5,070 +2,600 238 398 5,860 +2,080 356 786 5,390 +1,400 1. Preplanting, 2. Stem elongation, 3. Head development, 4. Flowering, 5. Milking stage 255 wheat yield response to irrigation is not significant. Whereas in Urfa where annual rainfall is low, the crop yield response factor is the highest, indicating that evapotranspiration deficit drastically hinders wheat yield and high yields can only be ensured with irrigation. Fig. 9. Wheat yield response to relative ET deficit. Crop water consumption of wheat depends upon on soil and climatic conditions, irrigation practice and wheat variety. As a result of differences in £Tand yields, different crop yield response factors were obtained for different regions of Turkey (Fig. 9). Yield response to deficit ET in the temperate regions of Turkey (for example, in Tokat and Erzurum) was less obvious than that in dry areas (Urfa and Ankara). 4. CONCLUSION Deficit irrigation practices can be a feasible option for improving irrigation schedules and thereby increasing efficient use of restricted water resources under irrigated agriculture. As shown in several case studies hi Turkey, exposing field crops to water stress at specific growth stages may not cause significant yield decrease and therefore, irrigation during these periods can be omitted and excess water left in the system can be diverted to other areas. Future research should examine water stress response of field crops at different growth stages within the context of modifying irrigation schedules to save water. REFERENCES [ 1] 256 STEGMAN, E.G., Corn grain yield as influenced by timing of evapotranspiration, Irrig. Sei. 3 (1982) 75-87. [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] STÖCKLE, C.O., Analysis of deficit irrigation strategies for corn using crop growth simulation models, Irrig. Sei. 10 (1989) 85-98. ZISKA, L.H., HALL, A.E., Seed yields and water use of cowpeas (Vigna unguiculata [L.] Walp.) subjected to planned-water-deficit irrigation, Irrig. Sei. 3 (1983) 237-245. KÖRTE, L.L., WILLIAMS, J.H., SPECHT, I.E., SORENSEN, R.C., Irrigation of soybean genotypes during reproductive ontogeny. I. Agronomic responses. Crop Sei. 23 (1983) 521-527. ECK, H.V., MATHERS, A.C., MUSICK, J.T., Plant water stress at various growth stages and growth and yield of soybean, Field Crops Research 17 (1987) 1-16. SPECHT, J.E., ELMORE, R.W., EISENHAUER, D.E., KLOCKE, N.W., Growth stage scheduling criteria for sprinkler-irrigated soybeans, Irrig. Sei. 10 (1989) 99-11L KIRDA, C., DANSO, S.K.A., ZAPATA, F., Temporal water stress effects on nodulation, nitrogen accumulation and growth of soybean, Plant and Soil 120 (1989) 49-55. STEGMAN, E.G., SCHATZ, B.G., GARDNER, J.C., Yield sensitivities of short season soybeans to irrigation management, Irrig. Sei. 11 (1990) 111-119. MASON, W.K., CONSTABLE, G.A., SMITH, R.C.G., Irrigation for crops in a subhumid environment, ÏÏ. Water requirement of soybeans, Irrig. Sei. 2 (1980) 13-12. TÜLÜCÜ, K., KIRDA, C., KUMOVA, Y., Water production function of soybean under limited water supply conditions, DOGA Turkish J. Agr. and Forestry 15 (1991) 814826. BOYER, J.S., JOHNSON, R.R., SAUPE, S.G., Afternoon water deficits and grain yields in old and new soybean cultivars. Agron. J. 72 (1988) 274-278. CONSTABLE, G., A., HEARN, A.B., Agronomic and physiological responses of soybean and sorghum crops to water deficits, I. Growth, developmant and yield, Aust. J. Plant Physiol. 5 (1978) 159-167. HERN, A.B., CONSTABLE, G.A., Irrigation for crops in subhumid environment, V. Stress day analysis for soybean and economic evaluation, Irrig. Sei. 2 (1980) 1-12. JUNG, P.K., SCOTT, H.D., Leaf water potential, stomatal resistance and temparature relations in field-grown soybeans. Agron. J. 72 (1980) 990. MOMEM, N.N.,CARLOS, R.E., SHAW, R.H., ARJMAND, O., Moisture-stress effects on the yield components of two sybean cultivars. Agron. J. 71 (1979) 86-90. DOORENBOS, J., KASSAM, A. H., Yield response to water, FAO Irrigation and Drainage Paper 33, FAO, Rome (1979). BARRETT, J.W.H., SKOGERBOE, G.V., Effect of irrigation regime on maize yields, J. Irrig, and Drainage Div. Proc. ASCE104, IR2 (1978). DONEEN, L.D., Some soil moisture conditions in relation to growth and maturation of sugarbeet, Plant Proc. 3, Am. Soc. Sugar Beet Tech. (1942) 54-62. ERIE, L.J., FRENCH, O.P., Water management of fall-planted sugarbeets in Salt River Valley of Arizona. Trans. Am. Soc. Agr. Engr. 11 (1968) 792-795. 257 [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] 258 OKMAN, C. Ankara §artlannda §ekerpancartnui su istihlakinin tayini uzerinde bir ara§tirma. Ankara Universités! Ziraat Fakiiltesi Yaytnlari. 780 (Doktora Cali§mast), Ankara, Turkey (1973). ZIBA, A., BILGIN, Y., Ankara'da §ekerpancari sulamasi ve bu s.artlar altinda sekerpancannin su istihlaki, Seker, Turkey 73 (1969) 37-47. SALTER, P.I., GOODE, J.E., Crop resposes to water at different stages of growth. In: Research Review No.2, Commonwealth Bureau of Horticulture and Plantation Crops, Commonwealth Act Bureaux, Farnharm Royal, England (1967). OYLUKAN, S., Cesitli mahsullerde ekonomik sulama saytstntn tesbiti denemesi sonuj raporu. Eski§ehir Böige TOPRAKSU Ara§tirma Enstitüsü Md. 71, Eskisehir, Turkey (1973). GÜNBATILI, F., Tokat'ta §ekerpancanmn su tüketimi. Tokat Böige TOPRAKSU ArasUrma Enstitüsü Md. Rapor No. 14, Tokat, Turkey (1978). WINTER, S.R., Suitability of sugarbeets for limited irrigation in a semi-arid climate. Agron. J. 72 (1980) 649-653. DECAU, J., ALL, A., The effect of climatic conditions and irrigation on the oil and protein production of three oil seed crops (rape, sunflower, soybean) in SW France. Comptes Rendus des Sciences de l'Acadamie d'Agriculture de France 59 (1973) 14641474. OSMAN, F., TALHA, M., The effect of irrigation regime on yield and sunflower seed oil content Nath. Res. Cent., Dokki, Cairo. Egyptian Jour, of Soil Sei. EGYPT 15 (1975)211-218. KARAMI, E., Effect of irrigation and plant population on yield and yield components of sunflower. Pahlavi Univ. Shiraz. Indian Jour, of Agricultural Sciences 47 (1977) 15-17. BROWNE, C.L., Effects of date of final irrigation on yield and yield components of sunflowers in semi-arid environment. Dep. of Agr., Leeton, N.S.W. Aust. J. Exp. Agriculture and Animal Husbundary 17 (1977) 482-488. BHATTACHAYRA, B., SARKAR, R., Relationship between growth parameters, assimilation rate and seed yield in sunflower under varying irrigation levels. Univ. Coll. Agr., Calcutta, Indian Agriculturist 22 (1978) 237-341. JANA, P.K., MISRA, B., KAR, P.K., Effect of irrigation different physiological stages of growth on yield attributes, yield, consumptive use, and water use efficiency of sunflower. Dep. of Agro., B.C. Krishi Viswa Vidyalaya, Kalyani, India. Indian Agriculturist. India 26 (1982) 39-42. HARMAN, W.L, UNGER, P.W., JONES, O,R., Sunflower yield response to furrow irrigation on fine textured soils in the Texas High Plains. Texas Agr. Exp. Station, Miscellaneous Pub. No. MP 1521 (1962) 35. [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] 43] [44] [45] [46] [47] RAWSON, H.M., TURNER, N.C., Irrigation timing and relationship between leaf area and yield in sunflower. Div. of Plant Industry, CSIRO, Canberra, Irrigation Sei. 4 (1983) 167-175. KARAATA, H., Ktrklareli ko§ullarinda aycicegi bitkisinin su-uretim fonksiyonlan, Tarim ve Koyi§leri Bakanligï, Köy Hizmederi Ara§tirma Enst., Rapor 24 (PhD Thesis), Kirklareli, Turkey. MERRIAM, J. L. A management control concept for determining the economical frequency of irrigation. ASAE Annual Meeting, Paper No.65-206, Georgia (1965) 1-10. ENGLISH, M.J., MUSICK, J.T., MURTY, V.V.N. Deficit irrigation. Management of Farm Irrigation Systems. Edit, by GJ. Hoffman, T.A. Howell, K.H. Solomon. An ASAE Monog., St. Joseph, MO (1990) 631-655. HILLEL, D. The relationship between field water balance and water use efficiency. Optimizing the Soil Physical Environment Toward Greater Crop Yields. Edit by D. Hfflel. Academic Press, New York (1972) 79-100. ENGLISH, M.J., Deficit irrigation: An analitical framework. J. ASCE (IR). 116(3) (1990) 399-412. ALAP, M. Pamuk raporu. Sulu Zir. Aict 1st. Yay. 15pp. Tarsus (1958). TEKlNEL, O., KANBER, R. Yield and water consumption of cotton as affected by deficit irrigation under Cukurova conditions. Soil-Water Res. Inst. Pub., Tarsus 98 (48) (1979) 39pp. YALCUK, H., ÖZKARA, H.M., The effects of omitted irrigation on cotton production in West Anatolia Region. Sou-Water Res. Ins. Pup. No. 107, Izmir (1984) 35pp. BASTUG, R., A study on determining the water production functions of cotton under Cukurova conditions. Univ. of Cukurova of Inst of Sei., Irr. and Drain. Dep. Ph. D. Thesis, Adana (1987) 120pp. KANBER, R., BASTUG, R., KÖKSAL, H., BAYTORUN, R, Yield and comperative performance of different crop production functions of cotton as influenced by deficit irrigation. Doga 14 (1990a) 442-455. KANBER, R., TEKÎNEL, O., ÖNDER, S., KÖKSAL, H., Comparison of irrigation length of cotton under Cukurova climatic conditions. Tr. J. of Agric. and Forest. 18 (1993b) 81-86. HANKS, R.J., Yield and water-use relations: An overview. Limitations to Efficient Water Use in Crop Production. Edi. by H.M. Taylor, WJ. Jordan, T.R. Sinclair. ASA, Madison, Wisconsin (1983) 393-410. VAUX Jr., H.J., PRUITT, W.O. Crop-water production functions. Advances in Irrigation, Edit by D. Hillel. Academic Press, New York 2 (1983) 61-93. YAVUZ, M.Y., Farklt sulama yöntemlerinin pamukta verim ve su kullammma etkileri. Univ. of Cukurova of Inst. of Sei., Irr. and Drain. Dep.,Adana, Ph. D. Thesis, Zir. Fak. Yay.19 (1993) 156-168. 259 [48] [49] [50] [51] [52] OYLUKAN, S., Nem azalma metoduna göre çesitli mahsullerin su istihlaklerinin tesbiti ara§tirmast. Böige TOPRAKSU, Eski§ehir, Arast Enst. Yay. 53 (1970) 12-13. OYLUKAN, S., Bugday ve §ekerpancan mahsullerinde sulama metodlartmn mahsul verim ve maliyeti üzerine tesirlerinin tesbit aras,tirmast Sonuç raporu (1967-1970). Böige TOPRAKSU, Eski§ehir, Arast Enst. Yay. 61 (1972) 19. MADANOGLU, K., Orta Anadolu kosullarinda bugday (Yektay 406) su tüketimi. Merkez TOPRAKSU, Ankara, Arast Enst Yay. 52 (1977) 67p. GÜNGÖR, H., ÖGRETlR, K. Eskis.eh.ir ko§ullarinda lizimetrelerde yeti§tirilen §ekerpancart, bug" day, mtsir ve patatesin su tüketimleri. Böige TOPRAKSU, Eskisehir, Arast Enst. Yay. 156(115) (1980) 32 p. ARAN, A., KTVANC, F. Konya ve Aksaray ovasi kosullartnda bu|day ve arpanin azotsu iliskileri. Köy Hizmet Ara§t. Enst. Yay., 131 (125) (1989) 51 p. [53] UZUNOGLU, S. Ankara yöresinde bugdaytn (Gerek 79) su tüketimi. Top. ve Güb. ArasL Enst. Yay., 183 (102) (1992) 23. [53] UZUNOGLU, S. Ankara yöresinde bu|daym (Gerek 79) su tüketimi. Top. ve Güb. ArasL Enst. Yay., 183 (102) (1992) 23. ÇEIÎN, 0. Harran ovasi ko§ullartnda farkli su ve azot uygulamalarimn bugday verimine etkisi ve su tüketimi. Köy Hizmet, Ara§t. Enst. Yay. 80 (54) (1993) 117 p. GÜNBATILI, F. Tokat-Kazova kosullaruida bugdayin su tüketimi. Böige TOPRAKSU Arast Enst. Yay. 45 (28) (1980) 26 p. KARAATA, H. (1987). Harran ovasinda bugdayin su tüketimi. Köy Hizm. Ara§t Enst. Yay. 42 (28) (1987) 34 p. SEVtM, Z. Erzurum ko§ullannda bu|dayin su tüketimi. Köy Hizm. Ara§t. Enst. Yay. 19 (1988) (16), 49 p. GARRTTY, D.P., WATTS, D.G., SULLIVAN, C.Y., GILLEY, J.R. Moisture deficit and grain sorghum performance: Effect of genotype and limited irrigation strategy. Agron. J. 74 (1982) 808-814 . [54] [55] [56] [57] [58] 260 SOIL SPATIAL VARIABILITY CONSIDERATIONS IN SALT EMISSION AND DRAINAGE REDUCTION J.W. HOPMANS, S.O. ECfflNG, W.W. WALLENDER Hydrologie Science, Department of Land, Air and Water Resources, University of California, Davis, USA Abstract A simple practical approach for estimating deep percolation from a field with a subsurface drainage system in the presence of regional ground water flow using an optimization scheme for the salt load has been presented. The effectiveness of the technique relies on the precise knowledge of the area of the field that is being drained, and on the assumption that the increase in drain-flow following an irrigation is owing to deep percolation. Crop ET is not required for the calculation of deep percolation. The amount of local regional ground water flow represents only the amount that is intercepted by the drain lines. However, a better understanding of ground water flow can be achieved by monitoring groundwater variations using piezometers. Such an experimental study is ongoing. Comparison of changes in storage with ET calculated with the yield function (85% ET from Westlands Water District) led to the conclusion that part of the crop ET was obtained from the shallow ground water. These results show the need for considering ground water table depth in irrigation management strategies. 1. INTRODUCTION It is generally accepted that current irrigation management practices should minimize subsurface drainage and deep percolation losses. In addition to increasing the water table with its concomitant water logging hazard and salt buildup in the root zone, large amounts of agricultural drainage waters cause increasing salt and trace element loads to surrounding surface waters. For example, in California the State Water Resources Control Board (SWRCB) recommended that total discharges in a study area of the San Joaquin River Basin would have to be reduced by about 40-70% to meet allowable selenium levels in the San Joaquin River downstream of the study area. It is furthermore expected that an increased supply of water will be required to meet California's growing demand and inasmuch as the amount of water used by agriculture is approximately 85% of the total water supply, agriculture is targeted to offer possibilities for water savings to be directed to other users. Certainly, the decrease in water quality and quantity available for agricultural use is not only an issue in California, but questions the sustainability of irrigated agriculture in general It is suggested that agricultural water conservation is the primary solution to overcome the State's increasing demand of water. In this regard, water conservation is viewed as a water management objective to minimize exploitation and degradation of the water resource, and specifically includes those practices that result in a decrease in the quantity of irrigation water lost by agricultural use. Because approximately 80% of California's irrigated agriculture is by 261 surface irrigation, half of which is by furrows, directions are needed to manage furrow irrigation systems in a more efficient manner. As growers tend to overirrigate, especially during the preplant and first irrigation, emphasis to reduce drainage waters should be placed on those first irrigations. The applied research, as reported here, emphasizes the importance of prior knowledge of soil properties and their field variability. This is in contrast to other studies that are directed at improving irrigation efficiency through increased water application uniformity. Although this is important, an equally important factor is soil variability and how it affects infiltration, water storage, deep percolation and plant growth and development The general objective of this study was to determine if soil spatial variability considerations in drainage and irrigation management reduce mass emission of salts through deep percolation or drainage. Although we made significant progress, we were not able to fully achieve this objective. Mostly, we were overwhelmed by the complexity of in situ soil-water systems at the field scale. As a result, even the estimation of a field water balance proved to be a challenging task. Therefore, a lot of effort was placed on field measurements, their field variability, and on methodology to describe field processes and their variability. The results described here are part of a comprehensive field study, with a focus on field soil spatial variability and its implication on irrigation water management. The study clearly has shown that the application of nuclear measurement techniques is beneficial for effective water use in irrigated agriculture. 2. MATERIALS AND METHODS The field investigated is a 32-ha field in Five Points, California, at the West Side of Fresno County (Diener Ranch). The soil is classified as Panoche loam [fine loamy, mixed (calcareous), thermic Typic Torriorthent]. The dominant soil textures are sandy loam, loam, and fine sandy loam. The soils are deep, well drained and are free of salt or are only slightly saltaffected. The field was furrow irrigated. Figure 1 shows the schematic of the field layout, with the irrigation ditch at 400 m dividing the field to shorten irrigation runs. Drain lines are 2.5 m deep and at 150 m intervals across the field. Part of the drainage system extends to 16 ha of an adjacent cotton field. Cotton was planted on April 1,1992. Most of the fields in the area are planted in cotton, onions or tomatoes. The field received 0.3 m of water during the pre-plant irrigation of January 6-10,1990. Water was applied to the field from an irrigation ditch at the head end of the field through PVC siphons. Soil water content was measured with a neutron probe at 0.15-m intervals from 0.15 to 0.9 m and at 0.3-m intervals from 0.9 to 2.1m. Soil water contents were measured six times thereafter for 125 d. The starting time for drainage (time = 0) corresponded with the time at which the soil profile water content was maximum. This maximum occurred 4 d after the beginning and 1 d after the end of irrigation. The 125-d measurement period extended from 24 262 MONITORING FURROW NUMBER 1 2 3 4 800 — 1 ol-ll i o 1 o ! o V ^~— LATERAL DRAIN x-s Z BZ ol-9 o ol-7 / o 60 ° 400 o o ^- NEUTRON PROBE ACCESS TUBE o o IRRIGATION DITCH -^ S 0 O^-J O O ol-3 o SITE ol-l o2-l03-1 o O j^ix/^ju^ ijk^ivir K U NORTH /i !j sS 1o U W g 1 o -1 o4Jl IRRIGATION DITCH -^ 1 1 i 0 100 200 300 400 DISTANCE (m) Fig. 1. Experimental Field Layout on Diener Ranch January to 30 May 1990. There was a water table in the field below the 2.1 m soil depth during this measurement period. The field was seeded hi cotton in late April. Evaporation from the soil surface during the cold months of February and March were neglected. Rainfall (37 mm) and estimated evapotranspiration (35 mm) in April and May were approximately equal. Evapotranspiration was calculated from crop coefficient and potential evapotranspiration data measured at a weather station 3 km from the field. Rainfall was also measured at the station. Hence, net changes in soil water content during the 125 d were attributed to drainage only. The field received also a pre-plant irrigation in the period January 7-10, 1992. Five additional post-plant irrigations were applied during the 1992 season. The drainage analysis to be described later is based on these five imgations. The irrigation water was a blend of well water and the Westlands Water District canal water. Irrigation on-flows were measured for irrigations 1 through 4 from flow meters at the pump and Westiands Water District canal meter. Electrical conductivities EC of irrigation waterfor imgations 2,3 and 4 were measured. Water was delivered to the furrows by siphons. All imgations were scheduled by the grower. Neutron probe access tubes were installed to a depth of 2.1 m at equal distances of 150 m along four monitoring cotton rows spaced 100 m apart. There were 6 access tubes in each monitoring row. The first and last tubes were placed about 50 m from the ends of the field. This spacing gave a grid pattern of 4 by 3 access tubes in each half of the field for a total of 24 access tubes. 263 Neutron probe measurements were taken just before, immediately after and one or two times between irrigations. Neutron probe readings were taken at 0.15-m intervals from 0.15 to 0.9 m and at 0.3-m intervals from 0.9 to 2.1 m. A single calibration curve was used to convert the neutron probe readings to water content within the 2.1 m profile. Water storage in the soil profile was calculated from the water content. The amount of infiltrated irrigation water was calculated by measuring the increase in water in the soil profile following irrigation, and adjusting for crop evapotranspiration during the irrigation period. Changes in storage within the 2.1-m soil profile were calculated for selected time intervals. Selection of the time intervals depended on availability of neutron probe data taken from the entire field on the same day. Field-average soil water storage values were estimated for each of the 24 measurement locations. A flow meter was installed in the drain sump on May 31, 1992. Dram flow rate and cumulative drainage were recorded from the meter three times a week. EC was measured on water samples taken at the same time of day beginning June 26. Fig. 2 shows the drain flow rate (m^d'1) and the electrical conductivity of drainage water (dS-nr1) as a function of time (d) with the beginning of irrigations shown by arrows. The increases in drainage rates are assumed to represent deep percolation from the field following an irrigation. Regional and field drainage 1200 8 800 g'? a iz —i 400 2 H U W EC O 0 0 30 60 90 TIME (d) Fig. 2. Drainage flow rate and salinity as a function of time since May 30. rates and amounts were estimated with an optimization scheme using the electrical conductivity and flow rate of the drain water as inputs. The salt load of the drain water (blend) is equal to the sum of salt load of the intercepted ground water, and that of field drainage water. Salt loads for selected periods were calculated by multiplying cumulative drainage for each period with the corresponding EC. Equation (1) relates drain water salt load to ground water and field drainage salt loads 264 Qr (1) 1 3 where EC is the electrical conductivity (dS-nr ), Q is cumulative drainage (m ) and subscripts b, I and r denote blend (drain water), local (field drainage) and regional (ground water), respectively. Salt concentration is given as EC without conversion to mass inasmuch as a linear relationship with total mass of dissolved salt was assumed. The exact partition between field and regional contribution to total flow is unknown inasmuch as drainage from the field does not cease at the beginning of the next irrigation. Therefore, line 1 in Fig. 2 was used as a first approximation of the regional ground water flow rate. Cumulative regional and field drainage were calculated based on the area under the curve. The EC of the drain water (blend) for a given time interval was assumed to be equal to the arithmetic mean of the EC over that interval. The blend salt load was calculated using this arithmetic mean. The calculated salt load, field drainage and regional drainage were substituted in Eq.(l) and EQ and ECr optimized by minimizing sums of squared deviations between measured blend water salt load and calculated salt load. Assuming that the flow at the end of the measurement period (day 97, Fig. 2) was owing to ground water only, the EC measured at this time was considered to be ECr. Following this assumption, ECr was fixed and only ECi was optimized for the selected time intervals. The optimization results with both ECi and ECr free were compared with results from optimizations with ECr fixed. The above analysis was repeated with base lines 2, 3 and 4 in Fig. 2. With flow at the end of the measurement period fixed, each line was calculated by changing the slope of line 1, The optimum base line was chosen based on the agreement between results of optimizations with both ECi and ECr free compared with results from optimizations with ECr fixed. A water budget approach was used to estimate the contribution of groundwater to crop ET. For a control volume which includes the water table (0-2.5 m), and for a time interval between irrigations it follows that: AST = ET+ DP (2) where ASx is change in water storage from the soil surface to the drain line at 2.5 m, ET is cumulative crop evapotranspiration for the time interval and DP is deep percolation beyond 2.5 m calculated using the drain-flow analysis. Water content measured with the neutron probe showed that the ground water table depth was near or below 2.1 m during the time intervals considered. Readings taken from piezometers installed 1 m away from each access tube confirmed the neutron probe readings. In order to calculate the contribution of groundwater to ET, ASr is split into its components ASj — ASQ_<2.i + • 4Sr2.i_2.5 (3a) with (3b) where ASo-2.i is change in storage within the neutron access tube zone, AS2.\-2.s is change in storage below the access tube and WTis flow from or to the 2.1-2.5 m soil layer. A negative 265 WT-value implies a capillary rise contribution to £T, whereas a positive value implies an increase in water storage in the 2.1-2.5 m soil layer. 3. RESULTS AND DISCUSSION Amount of applied irrigation water calculated from the flow meters at the pump and district turnout, and irrigation water salinities are presented in Table I. Also shown is the average amount of infiltrated water (storage) calculated from neutron probe readings. The large standard deviation is a result of soil spatial variability and the fact that some furrows were blocked off shortly after runoff began. TABLE I. IRRIGATION WATER AMOUNT AND SALINITY Date Irrigation Applied inflow (mm) 1 2 3 4 5 May 30-Jun 3 Jun 21-Jun 25 Jul 14-Jul 17 Jul 31-Aug4 Aug 16-Aug20 92 135 120 97 NA Measured water Neutron Probe (mm) Irrigation water salinity dS-m-1 88±34t 127 ±52 98 ±54 80±41 63 ±29 NA 1.7 1.2 1.2 NA NA = Not available t Standard deviation The drainage rate increases in Fig. 2 are assumed to represent contribution from the fields drained by the subsurface drainage system inasmuch as they occur just after the beginning of the irrigation period. The remainder of the drainage is attributed to regional ground water flow. Note that the second irrigation started after the drainage rate had already begun to increase. This increase is attributed to drainage from an adjacent tomato field which was not irrigated again for the remainder of the study. Table n shows the results of EC optimized with different slopes of the base line with ECr fixed at 6.93 (ECb at end of measurement period), and with both EQ and ECr free. ECfi ~ ECi and ECr « 6.93 (optimized regional EC is equal to the blend EC at the end of the measurement period) were used as criteria for an optimum slope. Line 3 fits the criteria and was chosen as the baseline partitioning total flow into regional (below baseline) and field (above baseline) flow. 266 TABLE H. ELECTRICAL CONDUCTIVITY VALUES ESTIMATED WITH DIFFERENT BASELINES Estimated EC (dS-nr1) Base line 1 2 3 4 ECf ECr ECi 2.32 2.91 3.64 3.93 5.73 6.00 5.73 5.08 3.75 3.49 6.87 7.22 t ECr = 6.93 fixed Using line 3, deep drainage from each of the five drainage events was calculated and presented in Table lu. Drainage from the field studied ranges from 667 to 3,316 m3 for the five considered time intervals. These quantities are based on the assumption that the regional ground water drainage rate decreases linearly with time. The amount of deep percolation from the adjacent cotton field is assumed to be proportional to the field area drained inasmuch as the soil, plant age and amount of water applied per unit area are the same as those of the field studied. Because the adjacent cotton field was irrigated at about the same time as the field studied, drainage contribution from it was considered to account for one third of the drainage represented by the peaks above the base line. Total drainage from the field studied for the five time intervals was 10,075 m3 (Table III), or 24% of total water that passed through the drain meter. This drainage corresponds to 32 mm of water loss from the soil profile above the drain line in the 32ha field during the five time intervals. Drainage from the adjacent cotton field (12% of drainflow) is not shown in Table ffl. The difference between total flow and drainage from the two fields is assumed to be from shallow regional ground water. The amount of regional water flow presented in Table I representing only the portion that is intercepted by the drainage system accounts for 64% of the total drain-flow volume during the study period. The function Y- -547.11 + 33.8 IET where 7 (kg-ha-1) is cotton lint yield and ET (cm) is crop evapotranspiration was used to calculate a total ET for the season. The mean lint yield was 1,400 kg-ha-1 with a standard deviation of 360 kg-ha'1. Seasonal ET was also calculated using the Wesuands Water District water conservation program. The total ET estimated with the yield function was 84.7% of the computer program ET. Therefore, 85% of program ET was used in all subsequent analyses and is presented in Table IV. The time intervals considered are the same as those for the drainage calculations. Also shown in Table IV are changes in storage estimated using the neutron probe data and water table contribution to ET calculated using Eq. (3b) and the estimated DP-values from the drainage analysis (Table IV). The use of Eq. (3b) to estimate ground water contribution to ET is justified because the water table was never above 2.1 267 TABLE m. TOTAL DRAIN FLOW, FIELD DRAINAGE, AND REGIONAL GROUND WATER FLOW Period Jun 5-Jun 21 Jim 26-Jul 14 Jul 18-Jul 31 Aug 4-Aug 16 Aug 20-Aug 28 Total Total flow (m3) Studied field drainage (mm) (m3) 14,258 13,100 7,452 4,533 2,328 3,128 3,316 1,943 1,021 667 10 10 6 3 2 9,566 8,126 4,538 3,001 1,328 41,671 10,075 31 26,559 Regional flow (m3) TABLE IV. CHANGES IN STORAGE, ADJUSTED PROGRAM ET, DEEP PERCOLATION AND WATER TABLE CONTRIBUTION TO ET Period Cumulative Change in storage ET from 0 to 2.1 m (mm) (mm) Deep percolation (mm) Water table contribution to ET (mm) Jun 5- Jun 21 Jun 26- Jul 14 Jul 18- Jul 31 Aug 4- Aug 16 Aug 20-Aug 28 76 115 106 66 32 94±20t 99 ±26 80 ±27 62 ±25 41 ±13 10 10 6 3 2 8 -26 -32 -7 7 Total 395 376 31 -65* Standard deviation Negative numbers only m. If the water table were at 2.1 m, the change in water content between 2.1 m and 2.5 m, would be zero, but there would still be upward or downward flux. At the beginning of the season, WT is 8 and at the end of the season (8/20-8/28) it is 7 mm. In most cases, however, WTis negative indicating that some of the ET is contributed by ground water. Hence, although some of the infiltrated water may have been lost from the 2.1-m profile by deep percolation, some of it is 268 later taken up by the crop. Calculations indicate that 65 mm of the £Tfor the periods considered comes from ground water. This water is equivalent to 16.6% of the total ET during the experimental period. 4. CONCLUSIONS A simple practical approach for estimating deep percolation from a field with a subsurface drainage system in the presence of regional ground water flow using an optimization scheme for the salt load has been presented. The effectiveness of the technique relies on the precise knowledge of the area of the field that is being drained, and on the assumption that the increase in drain-flow following an irrigation is owing to deep percolation. Crop ET is not required for the calculation of deep percolation. The amount of local regional ground water flow represents only the amount that is intercepted by the drain lines. However, a better understanding of ground water flow can be achieved by monitoring groundwater variations using piezometers. Comparison of changes in storage with ET calculated with the yield function (85% £Tfrom Westlands Water District) led to the conclusion that part of the crop ET was obtained from the shallow ground water. These results show the need for considerations of ground water table depth in irrigation management strategies. Next page(s) left blank 269 LIST OF PARTICIPANTS ARGENTINA D. PRIETO INTA-EEA Santiago del Estero C.C.N. 258-Jujuy 850, 4200 Santiago del Estero BRAZIL P.L. LIBARDI University of Säo Paulo, Escola Superior de Agricultura Luis de Queiroz and Center for Nuclear Energy in Agriculture, Piracicaba (SP) CHINA WANG FUJUN Laboratory for Application of Nuclear Techniques, Beijing Agricultural University, Beijing 100094 CÔTE DWOIRE C.B.G. PENE Institut des Savanes, B.P. 633, Bouaké 01 ECUADOR M. CALVACHE Ecuadorian Atomic Energy Commission, P.O. Box 2517, Quito FRANCE G. VACHAUD Laboratoire d'Etude des Transferts en Hydrologie et Environnement, B.P. 53, 38041 Grenoble cedex 9 HUNGARY T. KOVACS Research Institute for Irrigation, Szabadsag u. 2, 5540 Szarvas MALAYSIA A.AHMAD Malaysian Institute For Nuclear Technology Research (MINT)Bangi, 43000 Kajang Selangor MOROCCO M.BAZZA Institut Agronomique et Vétérinaire Hassan H, Département de l'Equipement et de l'Hydraulique, B.P. 6202, Rabat-Instituts 271 PAKISTAN M. M.IQBAL Nuclear Institute for Food and Agriculture, Tarnab, P.O. Box 446, Peshawar R. A. WAHEED Nuclear Institute for Agriculture & Biology (NIAB), P.O. Box 128, Faisalabad ROMANIA I. CRACIUN Research Institute for Cereals and Industrial Crops, Jud. Calarasi, 8264 Fundulea SPAIN F. MORENO Institute de Recursos Naturales y Agrobiologia de Sevüla (CSIC) P. O. Box 1052,41080 Sevilla TURKEY M.S.ANAÇ Ege University, Agriculture Faculty, Irrigation and Drainage Department, 35100 BornovaIZMIR C.KIRDA University of Cukurova, Faculty of Agriculture 01330 Adana USA J.W. HOPMANS Hydrologie Science, Department of Land, air and Water Resources, University of California, Davis, CA 95616 CONSULTANT TO THE PROGRAM D.R. NIELSEN Emeritus Professor, Hydrologie Science, University of California, 1004 Pine Lane, Davis, CA 95616 272 SCIENTIFIC SECRETARIES C. KIRDA (1991-1992) C. HERA (1992-1993) P. MOUTONNET (1993-1995) Joint FAO/IAEA Division, Soil Fertility, Irrigation and Crop Production Section, P.O. Box 100, A-1400 Vienna, Austria 273