Research institutes in Africa have had limited success in producing new technologies appropriate to the needs of small farmers because of an inadequate understanding of small farmers' goals and resource limitations and over-reliance on the transfer of technologies from other regions. Farming systems research (FSR) is an effective way of improving the focus of scientists on the problems farmers face. Vertically segmented thematic research programmes make institutionalization of FSR difficult within multi-commodity research centres, particularly the International Agricultural Research Centres. The Farming Systems Research Group has proved effective at overcoming such difficulties at the International Institute of Tropical Agriculture. Its organization, operation and effects on the performance of the research programmes are described.

Despite numerous altempts, the development of generalixed models capable of accurate predictions of the times from sowing to flowering (f) of crop plants in field environments has remained elusive. Models which seek to correlate ;ʃ with various states of environmental factors such as photoperiod, P, and temperature, T, using formal statistical procedures arc often complex. Here, we describe a family of photothermal responses (involving unambiguous parameters and limits) which quantify the linear, non-interacting effects of P and T not on ʃ but on 1/ʃ, i.e. on the rate of progress towards flowering. Based on these relations we suggest that the modelling of crop phenology will be simplified, more reliable and more biologically plausible.

Differences in the age at which okra pods were harvested (4, 7, 10 or 13 days after pod set) had no effect on vegetative growth or pod yield, but pods harvested more than seven days after pod set were of poorer quality. The reduction in pod quality arose mainly from an increase in crude fibre and a reduction in the moisture, crude protein and ash content of older pods.

A multidisciplinary team has attempted to find a sustainable alternative to slash and burn agriculture in deciduous forest in the Bolivian lowlands for those who do not have access to tractor power. Farm size ranges from 30 to 50 hectares and labour and capital are major constraints to development. Fertilizers are not used because of their high price. The alternative proposed is that the farm should contain an area of not more than one hectare in alley cropping producing the family's requirements in farinaceous foods and vegetables, and an area, limited by labour constraints, of perennial crops for cash income. The balance of the farm would be in a silvipastoral system producing various tree products, beef and cheese for sale and sheep for home consumption.

A selection programme conducted on young clonal tea plants grown in large sleeved bags and in a field trial established criteria for selecting high yielding field clones on the basis of their growth in early years. Bush area, number of branches and shoots per bush, and foliar phosphorus content at the second crop peak were correlated with yield but the correlation between stem circumference at ground level and yield was inconsistent. It is therefore concluded that all these bush and growth characters, apart from stem circumference, may be reliably used to predict and select high yielding tea clones.

The effects of variety and fertilizer on wheat and barley production, and of inoculation with rhizobium on Vicia crop production under rain-fed conditions were observed in farmers' fields in Balochistan from 1985 to 1988. Weeding and application of phosphate fertilizer had little effect on wheat yields, but in the ‘good‘ rainfall year, 1986/87, application of nitrogen fertilizer resulted in increased economic yields. Barley varieties from Syria generally produced better grain yields and, with the exception of Arabi abiad, poorer straw yields than a local variety. Genotype–environment analyses indicated that Arabi abiad could be expected to produce more grain, similar amounts of straw and larger gross benefits than the local variety in all except the most severe environments, when crop failure was inevitable. V. villosa ssp. dasycarpa showed negligible amounts of cold damage, and produced greater herbage and straw yields than the other legume crops. Inoculation with Rhizobium leguminosarum produced large yield increases in 1986/87.

The economic returns from crop production were poor and variable, but the results suggest that the productivity of the crop–livestock system could be increased by greater emphasis on barley, and by the introduction of Arabi abiad barley and V. villosa ssp. dasycarpa.

D. J. Rees et al.: Cultivos regadospor la lluvia de las tierras altas de Balochistan.

The potential yield of wheat and the physiological basis of yield limitation in the lowland irrigated tropics was investigated in three cultivars planted at five sowing dates. Maximum grain yields were achieved by cultivars that flowered in January when the weather was coolest. The period from sowing to terminal spikelet initiation was similar for all varieties and all sowing treatments. The period from terminal spikelet initiation to ear emergence was increased when seed was sown in late November or early December rather than in October or early November. Manipulation of the sowing date in relation to the choice of cultivar had a considerable effect on grain yield.

H. M. Ishag and O. A. A. Ageeb: Rendimiento del trigo en las zonas tropicales.

Two early maturing semi-tall (120–130 cm) rice varieties, Kalinga-3 (85 days duration) and Banaprabha (95 days duration), were compared in pure and mixed crop stands under intermediate deepwater conditions (15–50 cm) with a modern late maturing (170 days duration) semi-dwarf (100 cm) variety, CR 1016. Parallel line seeding of CR 1016 and Kalinga-3 or Banaprabha in alternate rows (1:1 ratio) 20 cm apart yielded 19% more grain than a pure crop of CR 1016 alone. Mixed crop planting in a 1:2 or 2:1 ratio or with reduced inter-rovv spacing gave no additional yield advantage. Mixed row cropping prevented lodging in early varieties, whereas monocrops of the same varieties were partially lodged and showed some premature seed germination. The better and more stable yields obtained from such mixed cropping could be beneficial to resource poor farmers growing rice in conditions of excess water.