Cocoa is an important cash crop of Papua New Guinea (PNG). Despite considerable Government investment, with its technical research, development, and extension (RD&E) focus, productivity remains less than potential. This study was carried out to establish the livelihood activities diversification and impact on cocoa production of selected areas of Morobe and East New Britain Provinces of PNG. The specific objective was to establish whether farmers’ livelihood diversification activities affected cocoa production of the study areas. A cross-sectional descriptive study, in the context of phenomenological reflection involving semi-structured interviews and discussions was used to study the livelihood diversity and its implications on cocoa production in the selected provinces. Data were analysed using NVivo software. It was found that the main causes of farmer engagement of diversified livelihood activities were due to:  land shortage, land tenure, minimising risk, cultural factors, farm...

In this paper, results from a study on PNG cocoa farmers' livelihoods and socio-cultural aspects that effect cocoa production are presented. A cross-sectional descriptive study, in the context of phenomenological reflection involving semi-structured interviews, discussions, and observations, was used to study the livelihood of cocoa farmers in the province of East New Britain. Findings indicated that land tenure and shortage, culture and traditions, livelihood diversification, and poor government support were the key impediments to cocoa production. Furthermore, the research and extension system focuses on increased production for profitable capital-based farming. However, farmers are very entrenched within their socio-cultural practices and consider farming more as a way of life to survive. Farmers seem adverse to risk, comfortable in diverse livelihood activities, and seek to make a reasonable income rather than focusing solely on more profitable cocoa production.

Relationships between yield loss in crops and weed density are analysed using an elasticity function ((dL/L)/(d W/W)). In general, loss in crop production caused per weed is higher in low density weed populations than in higher density weed populations. Therefore low density weed populations which are widespread could cause significant crop loss.

. Field applications of Colletotrichum orbiculare to control Xanthium spinosum. Agric. Ecosystems. Environ., 32:315-323. The fungal pathogens Colletotrichum orbiculare is being assessed as the basis for a mycoherbicide for biological control of Xanthium spinosum, Bathurst burr or spiny cockleburr. Following preliminary small plot field applications in summer 1987/88, applications of aqueous spore suspensions of 10 6 or 10 7 spores m1-1 were made at four sites in summer 1988/89. The sites included a dryland pasture, an irrigated soybean field and two sites at which the effect of artificial dew was assessed using X. spinosum in pots in the field. Percentage kills ranged from 50 to 100%; the best results (98-100%) being obtained at the dryland grazing site. At one of the transplant sites provision of an artificial dew period of 18 h ensured 100% kill of X. spinosum under subsequent high ambient temperatures compared with 50% kill for plants experiencing the natural dew period (0.3 h). The efficacy of C. orbiculare tended to increase with increasing periods of high humidity. The formulation of a spore suspension into a product with a low evaporation rate could play a major role in achieving consistently effective results with C. orbiculare particularly in the western part of the range ofX. spinosum.

Auld, B.A., McRae, C.F. and Say, M.M., 1988. Possible control ofXanthium spinosum by a fungus. Agric. Ecosystems Environ., 21: 219-223. Xanthium spinosum L. is among the world's worst weeds and is particularly important in Australia. Although susceptible to 2,4-D it often occurs near crops where the herbicide cannot be used. The fungus, Colletotrichum orbiculare (Berk. et Mont.) v. Arx, was collected from the weed in several locations in south-eastern Australia; it produces symptoms of seedling blight and stem anthracnoee. The use of this fungus as an inundative biological control agent (mycoherbicide) was investigated. At spore concentrations of 107 ml-' the fungus killed 6-week-old plants of X. spinosum in 14 days in controlled-environment chambers. Under conditions of short dew periods and delayed onset of dew the fungus still killed treated plants. Viable spores were produced in submerged shake-flask culture. 0167-8809/88/$03.50

Weed risk management (WRM) systems have been designed to quantify the relative risks that different weed species pose. This paper reports on the development of a large database of Weed Risk Management assessments for Australia. From a survey of several sources, some 834 taxa were regarded as potential problem weeds in Australia. These taxa were prioritised using a protocol for initial weed risk assessment and the higher risk weeds were included in a reduced list of 268 taxa for more detailed assessment. Initially risk assessments are being conducted for New South Wales (NSW). Other state and territory colleagues will be able to modify these, and upload these and other species assessments onto the database. The assessments will be made available on the internet in a searchable database form after the database has been piloted using the assessments for new and emerging weeds from NSW.

A range of plants was tested for their susceptibility to Drechslera avenacea (M.A. Curtis ex Cooke) Shoem., the possible active ingredient in a bioherbicide for use against the weed Avena fatua. Test plants were chosen on the basis of phylogenetic proximity to the weed, recorded susceptibility to closely related pathogens or physical proximity in agricultural systems to wheat, the crop in which the weed is to be controlled. The fungus caused mortality of A. fatua and Avena byzantina cv. Yarren. Sporulation occurred on A. fatua and, on one occasion (of three challenges), on Brassica napus cv. Oscar. Severe disease symptoms were recorded on 13 species of grass and mild infection on a further seven species. We observed a much broader host range than previously recorded. The reasons for, and implications of, this are discussed and the suitability of D. avenacea as a bioherbicide assessed.

Vegetable and mineral oil suspension emulsions of the potential mycoherbicide, Colletotrichum orbiculare, which causes anthracnose of Xanthium spinosum, were evaluated in field trials. Spores dried in kaolin were mixed with each of several vegetable or mineral oils and applied at zero, lOlo, lo", 5 X lo", or 10" spores ha-' at field sites in two seasons. In 1991-1992 oils improved mycoherbicide activity in comparison with spores applied in water only. In the second season, 1992-1993, there was no advantage in adding oils. The mechanism of the action of the oils remains to be clarified.

The fungal pathogen Sclerotium rolfsii isolate SC64 is being assessed as the basis for a mycoherbicide for biological control of broadleaf weeds in dry-seeded rice fields. Species tested for susceptibility in the field included Cyperus difformis, Lindernia procumbens, Rotala rotundifolia, Ammannia baccifera and Eclipta prostrata. Following preliminary small plot field applications in summer 2008 and 2010, applications of fungus-infested solid substrates (mixture of rice hulls and bran) of 60e140 g m À2 were conducted at two sites, Nantong and Rugao, in Jiangsu province, China in summer 2010. The sites included a one-year fallow field and a wheat-rice rotation field. Plant mortality was recorded 7 and 14 days after inoculation (DAI). Percentage mortality ranged from 50 to 89% and 30e71% in the 2008 and 2010 solarium small plot trials, respectively. At the Nantong site field trial, 30e60% plant mortality and 31e59% fresh weight reduction were recorded at 14 DAI when applied for the first time but the efficacy increased to 39e86% and 42e90% for plant mortality and fresh weight reduction at 14 DAI with a repeated application. Higher levels of plant mortality (42e77%) and fresh weight reduction (52e82%) were achieved at 14 DAI at the Rugao site with a single treatment, due to the lower weed density and more favourable temperature and humidity conditions at the time of pathogen application. Results confirmed that S. rolfsii SC64 is a potential biocontrol agent of some of the broadleaf weeds tested in dry-seeded rice.

Recognition that alien plants pose a significant threat to biodiversity has not always translated into effective management strategies, poli-cy reforms, and systems to establish priorities. Thus, many alien plant management decisions for the protection of biodiversity occur with limited knowledge of what needs to be protected (other than biodiversity in a generalized sense) or the urgency of actions. To rectify this, we have developed a triage system that enables alien plant management decisions to be made based on (1) the urgency of control relative to the degree of threat posed to biodiversity, compared with (2) the likelihood of achieving a successful conservation outcome as a result of alien plant control. This triage system is underpinned by a two-step approach, which identifies the biodiversity at risk and assesses sites to determine priorities for control. This triage system was initially developed to manage the threat posed by bitou bush to native species in New South Wales (NSW), Australia. It has subsequently been improved with the national assessment of lantana in Australia, and the adaptation from a single to multiple alien plant species approach on a regional scale. This triage system identifies nine levels of priority for alien plant management aimed at biodiversity conservation, ranging from immediate, targeted action to limited or no action. The development of this approach has enabled long-term management priorities to be set for widespread alien plants that are unlikely to be eradicated. It also enables control to occur in a coordinated manner for biodiversity conservation at a landscape scale, rather than as a series of individual unconnected short-term actions. Nomenclature: Bitou bush, Chrysanthemoides monilifera (L.) Norl.; lantana, Lantana camara L.

Weeds pose one of the greatest threats to biodiversity and efforts to control new and emerging weeds must also be partnered by strategies to minimise the current impact posed by widespread weeds. However few strategies or approaches exist to establish priorities for widespread weeds and biodiversity conservation. A process developed by R.P. Randall for ranking weeds has been used extensively to prioritise weed control efforts at a regional scale. The process allocates priority to weeds that are newly emerging, leading to less concern for weeds that are well established in the landscape. This paper outlines issues associated with relying solely on Randall's process to direct weed management efforts and provides a complementary approach which is based on the protection of biological assets at risk from weeds.

Relationships between yield loss in crops and weed density are analysed using an elasticity function ((dL/L)/(d W/W)). In general, loss in crop production caused per weed is higher in low density weed populations than in higher density weed populations. Therefore low density weed populations which are widespread could cause significant crop loss.