Physics and Chemistry of the Earth 34 (2009) 786–789

Contents lists available at ScienceDirect

Physics and Chemistry of the Earth

journal homepage: www.elsevier.com/locate/pce

The effect of irrigation on synchronization of coffee (Coffea arabica L.) flowering

and berry ripening at Chipinge, Zimbabwe
M.T. Masarirambi a,*, V. Chingwara b, V.D. Shongwe a
a
Department of Horticulture, University of Swaziland, P.O. Luyengo Luyengo, Swaziland
b
Horticultural Research Centre, P.O. Marondera, Zimbabwe

a r t i c l e i n f o a b s t r a c t

Article history: Coffee (Coffea arabica L.) is a short day plant that flowers and ripens irregularly in many subtropical and
Received 6 February 2009 tropical production areas. This results in a prolonged, laborious and tiresome hand harvesting period
Received in revised form 29 June 2009 which can potentially compromise coffee quality. Synchronising or scheduling the ripening of coffee ber-
Accepted 30 June 2009
ries can help in reducing harvesting costs through reduction of the number of cycles and harvest trips.
Available online 14 August 2009
This study was carried out to assess the quantity and frequency of irrigation, on stimulation of early, uni-
form flowering and berry ripening of coffee in order to reduce harvesting trips. High soil moisture deple-
Keywords:
tion (25 cb or 30 cb) followed with increased irrigation levels resulted in increased number of flowers and
Coffee
Irrigation levels
subsequent number of berries per bunch at given assessment dates. Moisture depletion levels of 25 cb or
Synchronization 30 cb followed by application of 20 l or 25 l of water per tree advanced coffee flowering and subsequent
Flowering berry ripening in this experiment when compared to other moisture depletion levels followed by appli-
Berry ripening cation of 15 l of water per tree.
Ó 2009 Published by Elsevier Ltd.

1. Introduction 1.2. The effect of irrigation on coffee flowering and berry ripening

1.1. Coffee growing in Zimbabwe Unlike most other coffee producing areas, the high rainfall areas
(areas receiving over 1000 mm/year of rainfall) of Zimbabwe do
Coffee growing in many tropical and subtropical production not have a distinct dry season and wet season. Consequently, coffee
areas is characterised by poorly synchronized floral development ripens during most months of the year. The year to year variation
and a subsequently prolonged hand harvesting period (Clearwater in coffee yields and the number of harvests necessary to realise
et al., 1999). Coffee floral initiation in Zimbabwe occurs between the potential yield, are in response to factors such as rainfall pat-
May and July after cessation of extension growth and is associated terns, size of the previous crop, crop vigour and health, pruning,
with short days (Clowes, 1977). Flowering occurs from September fertilisation and irrigation practices. There is lack of knowledge
to November with lighter flowerings in December. In Zimbabwe, of how these factors can be integrated to reduce the yield variation
irrigated coffee exhibits three to six flowerings per season (Clowes, in order to control the coffee flowering and ripening period in or-
1977). Hand harvesting, in Zimbabwe is associated with coffee der to better manage labour. Crisosto et al. (1992) working at Ku-
samples, which includes various amounts of over and under ma- nia, Oahu, Hawaii, a hot dry area with cool moist winters, found
ture fruit as well as mature ripe and red cherry ripe berries. This that coffee flowered best when a dry period was followed by rain.
practice results in poor crop quality and subsequent reduction in They developed a rating scale for coffee flower bud development.
farm income. Coffee pickers mix berries of different maturity be- Only flower buds in the open white cluster stage (>4 mm in length)
cause selective coffee picking is tiresome and laborious, consider- responded to drought followed by rain or irrigation and moved to
ing that pickers are paid according to the quantity of their the next stage of development and then flowered in eight to ten
harvest. As ripening commences the coffee changes from green days (Crisosto et al., 1992). Contrary to Crisosto et al. (1992),
to red then to dark-red–purple and eventually black colour when Schuch et al. (1990) in Kona, Hawaii found that coffee that flow-
over mature. Coffee is harvested when red in colour, which is a ered at the open white cluster stage did not need drought and rain
stage where appropriate maturity has been attained. The time from to flower. Clowes (1977) recommended withholding of irrigation
flowering to berry ripening is about 9 months. 6–8 weeks before main rains came. Prediction of the main rains
is always a problem in developing countries because of inadequate
infrastructure and equipment. Consequently, there is need for
* Corresponding author. Tel./fax: +268 527 4441. precise irrigation timing in order to get uniform flowering and
E-mail address: mike@agric.uniswa.sz (M.T. Masarirambi). ripening.

1474-7065/$ - see front matter Ó 2009 Published by Elsevier Ltd.

doi:10.1016/j.pce.2009.06.013
 M.T. Masarirambi et al. / Physics and Chemistry of the Earth 34 (2009) 786–789 787

Peasley (1990) reported that flowering could be controlled in accommodate all treatments. Coffee trees used in this experiment
areas with a reliable dry season during flowering (winter/spring) were not mulched. Hundred grams of compound J (14–6–20) (N–
and where there was irrigation, flowering could be controlled by P–K) and 120 g of ammonium nitrate were applied by hand three
deliberately water-stressing the trees and then heavily irrigating times a year in September, December and February per planting
them. Most researchers agree on the need for water stress before station. Hand-hoe weeding was done in August 2002, December
flowering. Drinnan (1990) reported that coffee required a plentiful 2002, February 2003 and August, 2003.
supply of water all year round except during the late stages of flo-
ral development (September–November). During this period, 2.2. Trial design
water stress can be used to manipulate flowering. Adequate rain-
fall (Went, 1957) or sometimes irrigation (Browning and Fisher Three blocks used in this experiment were marked, each block
1975) has been associated with the breaking of flower bud dor- had nine plots. Each plot had three planting stations bordered by
mancy in Coffea arabica. Matthew and Chokkana (1964) reported a single guard row to avoid possible chemical contamination due
that a pre-rains (or pre-irrigation) soil moisture stress induced to spray drift. The experiment was laid as a split plot with amount
heavier flowering in coffee, as reported by other researchers (Pirin- of irrigation water (three levels) as main plots and irrigation timing
ger and Borthwick, 1955; Alvim, 1960). Clowes and Wilson (1974), (three levels) as sub plots. Both main plots and sub plots were ran-
in their situation, reported that irrigating all year round resulted in domly assigned. A single block was selected at random for tensi-
continually late flowering as trees aged. Browning and Fisher ometer placement. Two tensiometers were firmly placed in
(1975) contradicted with other researchers when they found that auger-created holes for each treatment in the selected block on
a rapid drop in temperature could release buds from dormancy 16 September, 2002. The holes were 30 cm deep, a depth were
without changes in soil moisture status. Wriggley (1988) reported most of coffee feeder roots are located.
that the pattern of rainy and dry periods is important for growth,
budding and flowering in coffee. 2.3. Data collection and analysis
Although most researchers agree on the usefulness of moisture
stress before flowering and irrigation on stimulating heavy and Tensiometer readings were monitored daily at 08:00, 12:00 and
uniform flowering, there is a dearth of information on effects of 14:00 h by a group of experienced recorders. A hose and measuring
specific irrigation schedules on coffee flowering and ripening. This bucket were used for applying the irrigation water. Irrigation treat-
project sought to test whether irrigation scheduling could be suc- ments applied in this trial are shown in Table 1. The treatments
cessfully used to synchronize coffee flowering and berry ripening were repeated three times in the blocks.
using tensiometers to measure soil moisture.
2.4. Measurements
2. Material and methods
Flowering was recorded on the 23rd September, 24th October,
2.1. Description of study area and orchard management and 18th November, 2002. Number of flowers per replicate was ex-
pressed as a percentage of monthly totals. Ripe berries were hand
The study was done at Coffee Research Station which is 1132 m picked, counted and recorded before wet processing.
above sea level at 20° 1200 latitude and 32° 3700 longitude in Natural
Region 1, Zimbabwe, from August 2002 to September 2003. This 2.5. Statistical analysis
site was considered to be representative of a high rainfall area of
Zimbabwe because it receives on average more than 1000 mm/ Statistical analysis of data were done using MINITAB statistical
year of rainfall. However, rainfall is not always evenly distributed. package version 12.22 (Minitab Corporation, USA) licensed to the
Rainfall distribution for the period of August 2002–May 2003 at Coffee Research Station at Chipinge. The data for number of flowers
the Coffee Research Station, Chipinge is shown in Fig. 1. The soils was log transformed before analysis of variance because it was not
were a loamy sand with a pH of 5, which is ideal for coffee produc- normally distributed. Data was analysed by descriptive statistics
tion. In this experiment conducted at Chipinge, C. arabica L cultivar and analysis of variance (ANOVA). Mean separation, where signif-
SL 28, planted in October 1999 at a spacing of 3.0 m  2.4 m was icant differences existed was done by the least significance differ-
chosen because it was the only block which had young and healthy ence (LSD) test.
coffee plants. In addition, this was the only area, which could safely
3. Results and discussion
250
As the amount of water applied per tree increased there was a
200
general increase in the number of flowers produced per branch
(Table 2). The application of 25 l of water per tree irrespective of
Rainfall (mm)

150

Table 1
100 Irrigation treatments and irrigation timing of coffee (Coffea arabica L.).

Amount of water (l) Irrigation timing (cb)

50
15 20
15 25
0 20 30
20 20
r

ril
ry
er

y
r

ch

ay
r
st

be

be
be

ar

Ap
ua
ob
gu

ar

M
nu
m

20 25
em

br
ct

M
Au

ve

ce

Ja
O

Fe
pt

No

De
Se

25 30
Months 25 20
25 25
Fig. 1. Rainfall distribution for the period August 2002–May 2003 (Coffee Research 25 30
Station, Chipinge, Zimbabwe).
788 M.T. Masarirambi et al. / Physics and Chemistry of the Earth 34 (2009) 786–789

Table 2 Table 4
Effect of irrigation levels and irrigation timing on number of coffee (Coffea arabica L.) Effects of irrigation timing on coffee (Coffea arabica L.) berry ripening.
flowers during the first assessment date: 23 September 2003.
Irrigation timing (cb) Assessment date
Amount of water (l) Irrigation timing (cb) Number of flowers
3rd: 24 October 2002 4th: 18 November 2002
25 20 96.30 a* Number of ripe berries
25 25 96.20 a
30 8.68 a* 14.78 a
25 30 88.60 a
25 10.49 a 15.87 a
20 30 81.70 a
20 5.56 b 10.30 b
15 25 43.10 b
20 25 42.70 b Lsd 2.58 4.01
15 30 39.40 b Sed 1.18 1.84
20 20 37.20 b *
Means followed by different letters in the same column are significantly different
15 20 32.70 b
at P < 0.05.
Lsd 21.39
CV (%) 19.40
Sed 9.82
eral recommendation for irrigation is given as a range from 20 cb
*
Means followed by different letters within in the same column are significantly to 30 cb (Hegney and Hoffman, 2005). However, frequently irrigat-
different at P < 0.05.
ing coffee plants with less water resulted in reduced amount of
ripe berries while increased number of ripe berries was obtained
with application of more litres of irrigation water applied infre-
timing significantly (P < 0.05) produced the highest number of cof-
quently (Table 4). The results shown here (Table 4) are only those
fee flowers per tree for both the first and second assessment dates
for irrigation timing at 20, 25 and 30 cb for the third and fourth
(Tables 1 and 2). At the first assessment date 20 l of water applied
assessment dates because there were no differences in the other
at an irrigation timing of 30 cb per tree resulted in a similar num-
results. These findings concur in part with those of Shongwe et.
ber of flowers to trees supplied with 25 l irrespective of irrigation
al. (2008) who reported increased growth of tissue-cultured bana-
timing. Application of 15 l or 20 l of water per tree resulted in sig-
na with application of high levels of irrigation water. Water applied
nificantly (P < 0.05) lower number of flowers when compared to
to coffee trees is not static but moves down the soil profile while
25 l (Tables 2 and 3). When trees had more number of flowers in
some is absorbed by the roots in the feeder zone. There is a lot of
the early assessment dates as shown in this experiment it meant
controversy as to the most suitable condition for growing coffee
that the trees flowered earlier and thus subsequent earlier ripening
in relation to moisture availability. Evenly distributed rainfall is re-
of coffee berries.
quired to enable the tree to make good growth and maintain foli-
There were significant (P < 0.05) differences between the inter-
age in good condition (Foster and Wood, 1963). Rainfall or
active effect of amount of water and irrigation timing on the num-
irrigation may be more useful at a certain stage of plant develop-
ber of flowers per branch (Tables 2 and 3). Application of 25 l of
ment. For example, Dean (1939) observed in Hawaii, that the rain-
water per planting station interacted positively with any irrigation
fall received during the period of maximum tree growth was
timing (Table 2). Number of flowers increased with increase in the
significantly correlated with yield in the following season. Infre-
amount of irrigation water and degree of moisture stress. The least
quent irrigation timings at 25 cb and 30 cb stimulated early ripen-
number of flowers per branch was from the application of 15 l per
ing of coffee berries. This is in agreement with previous
planting station plus irrigation timing of 20 cb (Table 2) Applica-
observations at Chipinge (Coffee Research Station Annual Report,
tion of 20 l of water per planting station plus irrigation timing of
1996). However, Wrigley (1988) reported a ripening stimulatory
30 cb gave the second highest number of flowers per branch (Table
effect of rain showers in a dry environment.
3). However, this was not significantly (P > 0.05) different from the
impact of applying 25 l of water per station and timed at 25 cb (Ta-
ble 2). Coffee plants whose irrigation timing was at 30 cb and 25 cb 4. Conclusions and recommendations
gave significantly (P < 0.05) higher number of ripe berries than
those plants whose irrigation timing was 20 cb during the third From this study the following conclusions were made:
and forth assessment date (Table 4).
Irrigation timing did not influence flowering differently in all (1) Effecting soil moisture depletion of 25 cb or 30 cb can suc-
assessment dates. This could be explained by the fact that the gen- cessfully synchronize coffee flowering and berry ripening if
followed with relatively heavier irrigations of 20 l or 25 l.
Table 3 (2) Irrigation scheduling can be successfully used to stimulate
Effect if irrigation levels and irrigation timing on number of coffee (Coffea arabica L.) early uniform flowering and coffee berry ripening in the high
flowers during the second assessment date: 24 October 2002. rainfall areas of Zimbabwe like Chipinge.
Amount of water (l) Irrigation timing (cb) Number of flowers (3) It is recommended to investigate the nature of the relation-
25 20 49.40 a*
ship between irrigation scheduling and amount of irrigation
20 30 43.30 ab water applied per tree of coffee during different seasons. In
25 30 43.30 ab addition, plant responses to different moisture regimes,
25 25 42.10 abc whose responses are quantitative and qualitative need to
20 25 33.60 bcd
be understood more clearly in coffee and therefore the need
15 25 31.00 cde
15 30 26.40 de for further research.
20 20 24.80 de
15 20 19.70 e
References
Lsd 12.29
CV (%) 19.80 Alvim, P. de T., 1960. Moisture stress as a requirement for flowering of coffee.
Sed 5.64 Science 132, 354.
*
Browning, G., Fisher, H.M., 1975. Flower bud dormancy in Coffea arabica L. studies of
Means followed by different letters within the same column are significantly gibberellin in flower buds and xylem sap and abscissic acid in flower buds in
different at P < 0.05. relation to dormancy release. Journal of Horticultural Science 50, 207–218.
 M.T. Masarirambi et al. / Physics and Chemistry of the Earth 34 (2009) 786–789 789

Clearwater, M.J., Meinzer, F.C., Osgood, R.V., 1999. Control of coffee flowering and Hegney, M., Hoffman, H., 2005. Using Tensiometers for Scheduling Irrigation of
ripening in Hawaii. Journal of Horticultural Science 30, 20–21. Potatoes. Government of Western Australia, Department of Agriculture and
Clowes, M.S.T., 1977. The Effects of ethrel on ripening Coffea arabica L. fruits at Food. <http://www.agric.wa.gov.au/PC_92500.html>.
different stages of maturity. Rhodesia Journal of Agricultural Research 20, 10– Matthew, P.K., Chokkana, H.G., 1964. Studies on the intake of water and nutrients
12. during development of flower buds to blossoms in coffee. Journal of
Clowes, M.S.T., Wilson, J.H., 1974. Physiological factors influencing irrigation Horticultural Science 25, 264–288.
management of coffee in Rhodesia. Rhodesia Agricultural Journal 71, 53. Peasley, D.L.,1990. In: Proceedings of the New South Wales Coffee Marketing
Coffee Research Station Annual Report, 1996. Effect of Irrigation and Nitrogen on Summit, NSW Agriculture, Murwillumbah, NSW, Australia.
Coffee Growth and Development. Chipinge, Zimbabwe, pp. 7–8. Piringer, A.A., Borthwick, H.A., 1955. Photoperiod responses of coffee. Turrialba 5,
Crisosto, C.H., Grantz, D.A., Meiner, F.C., 1992. Effect of water deficit on flower 72–77.
opening in coffee (Coffea arabica L.). Tree Physiology 10, 127–139. Schuch, U.K., Fuchigami, L.H., Nagao, M.A., 1990. Gibberellic acid causes earlier
Dean, L.A., 1939. Relationships between rainfall and coffee yields in Kona District, flowering and synchronizes fruit ripening in coffee. Plant Growth Regulation 9,
Hawaii. Journal of Agricultural Research 59, 217–222. 59–64.
Drinnan, J., 1990. Temperature and irrigation affect vegetative growth and Shongwe, V.D., Tumber, R., Masarirambi, M.T., Mutukumira, A.N., 2008. Soil water
flowering of coffee (Coffea arabica L.). Journal of Horticultural Science 70, 25– requirements of tissue-cultured Dwarf Cavendish banana (Musa spp. L.). Physics
34. and Chemistry of the Earth 33, 768–774.
Foster, L.A., Wood, R.A., 1963. Observations on the effects of shade and irrigation on Went, F.W., 1957. The experimental control of plant growth. Chronica Botany 17,
soil moisture utilisation under coffee in Nyasaland. Journal of Experimental 257.
Agriculture 31, 108–114. Wrigley, G., 1988. Coffee. Longman, Burnt Mill, Harlow, England.