The base (Tb), optimum (To) and ceiling (Tce) temperature for 50% seed germination and seedling emergence, and the thermal time at sub-optimal temperatures (θ1), were examined in experiments with a range of cowpea genotypes from different habitats and latitudes. There was substantial genotypic variation in seed germination for Tb and To, but not for Tce. There was no evidence that genotypes from semi-arid habitats had higher values of To or Tce, but estimates of Tb for seed germination were related to latitude and were highest in genotypes originating from close to the equator. The significance of these responses as a component of developmental plasticity in relation to the adaptation of cowpeas is discussed.

Twenty-nine diverse genotypes of cowpea (Vigna unguiculata) were grown in approximately 30 photothermal environments in Nigeria and Niger, with mean temperatures ranging from 19° to 30°C, photoperiods from 10 to 16 h d−1 and saturation deficits from 0.5 to 3.1 kPa. Twelve of these genotypes were insensitive to photoperiod. The time of flowering of these photoperiodinsensitive genotypes showed a similar response to temperature. Time to flowering was also delayed by mean pre-flowering saturation deficits greater than 1.5 kPa. The implications of these responses to temperature and saturation deficit for phenological adaptation in West Africa are discussed.

Seventeen photoperiod-sensitive genotypes of cowpea (Vigna unguiculata) were grown in approximately 30 photothermal environments in Nigeria. Photoperiods ranged from 10 to 16 h d−1, mean temperatures from 19° to 30°C and times from sowing to flowering (f) from 32 to 140 d. Rate of progress towards flowering (1/f) was related to mean pre-flowering values of temperature and photoperiod using simple linear rate models comprising one, two or three planes (thermal, photothermal and insensitive). There were no significant differences (p > 0.25) among genotypes in response to temperature within the thermal plane and the common base temperature was estimated to be 7.6°C. Photoperiod-sensitivity varied by a factor of 15 among genotypes, and the critical and ceiling photoperiods varied from 12.2 to 13.4 and from 13.8 to more than 16 h d−1 at a mean temperature of 27°C, respectively. These simple models satisfactorily predicted f in an independent data set (R2 = 0.62) for plants grown in the main cowpea growing seasons at latitudes between 7° and 13°N. The utility of photothermal models and methods to screen for photothermal responses are discussed.

Ninety five cocoa clones were tested in Sabah, Malaysia at planting densities of 1096 and 3333 plants ha−1. Analysis of dry bean yields up to 60 months after planting showed that vigorous clones performed better at the lower planting density, some clones of intermediate vigour performed well at both planting densities and some weak clones performed well at high planting density. The conclusion that optimal planting densities are highly variable has significant implications for procedures in clone selection and suggests that the relation between the yield of a clone and its general combining ability will become clearer once planting density is taken into account.

A field experiment was conducted to assess the effect of competition between a leucaena hybrid and maize (Zea mays L.) when planted simultaneously in an alley cropping system. The leucaena hybrid (a cross between L. diversifolia and L. leucocephala) was planted at hedgerow spacings of 3 and 5.25 m, while maize was planted in rows 75 cm apart between the hedgerows. The spacing between the leucaena hedgerow and maize was varied by removal of 0, l or 2 rows of maize to give three spacing treatments of 37.5, 75 or 112.5 cm between leucaena and maize. A control plot of leucaena alone was also included in the treatments. The growth and yield of individual maize rows were virtually unaffected by the presence of leucaena, but maize had a significant influence on the growth and yield of leucaena. At full maize canopy development, photosynthetically active radiation reaching the leucaena was reduced in all treatments, resulting in a 75% yield reduction in leucaena at the closest spacing. Overall, maize grain yield reached 10.3 t ha−1 in the continuous maize plots (37.5 cm treatment). This was reduced by up to 40% after removal of two maize rows in the closest leucaena row spacing treatment. The implications of these results for the practical establishment of leucaena hedgerows with a maize crop are discussed.

A field experiment was conducted to study the effect of leaf stripping, de-tasselling and topping on the yield of maize and finger millet in a maize/finger millet intercropping system at Lumle in Nepal. De-tasselling of maize before pollen shedding reduced the number of barren plants, and increased test weight and grain yield but de-tasselling after pollination had no significant effect. Defoliation of maize by removal of leaves below and above the ear, or by topping 30 days after 50% silking, had no effect on either maize grain yield or on that of the relay planted finger millet. This contradicts the hypothesis that increased light penetration due to reduced shading improves the yield of millet under maize.

The response to nitrogen of four high yielding good quality clones of tea was investigated on fertile virgin soil. The yields of all the clones examined were maintained without additional nitrogen during the first two years after planting but Clone PC 81 responded to nitrogen in the second year after planting. This was the earliest recorded stage at which seedling or clonal tea had responded to nitrogen in Malawi. The other clones examined all responded to nitrogen in the third year after planting, and the yields of all the clones were substantially increased by the higher rates of nitrogen applied in the early years after planting, though the responses to nitrogen differed between seasons.

Length and weight of pigeonpea roots were measured weekly in different soil layers and compared with estimates obtained from a root simulation model using daily climatic data, soil physico-chemical properties and dry matter allocation to roots. Daily moisture content and temperature at different soil depths were well simulated using sub-routines from the CERES-Maize model. Daily allocation of dry matter to roots was calculated from logistic functions fitted to the growth data for shoots and roots. Although root length and weight tended to be underestimated by the model, regressions between measured and simulated root growth were highly significant so that the model could, with a few modifications, be used to predict root system development.

The best time to plant Calliandra calothyrsus in association with Pennisetum purpureum and the appropriate in-row spacing of Calliandra for fodder and fuelwood production when grown on contour bunds were investigated. Main plot treatments were Pennisetum planted in November 1988 and associated with Calliandra in February 1989, Calliandra planted in November 1988 and associated with Pennisetum in February 1989, and both species planted simultaneously in November 1988. Sub-plot treatments were three within-row spacings of Calliandra. Survival of neither species was affected by time of planting. Total treatment yields of the second and third years were not influenced by time of planting but lower yields of leaf and wood were obtained from the later-planted Calliandra. Total yields were not significantly influenced by the spacing, but leaf and wood yields of the Calliandra at the 25 cm spacing were greater than those at the 100 cm spacing although not significantly different from those at the 50 cm spacing. Neither the time of planting nor the spacing of Calliandra affected the yield of maize or beans grown in the interbund areas. Simultaneous planting of Calliandra and Pennisetum is recommended.

The most important technological needs and farmers' problems in agricultural entomology were identified by analysis of data collected from farmers in Tamil Nadu, India. This showed that damage due to leaf hoppers in rice was the most important research need and that due to brown plant hoppers in rice was the most important extension problem.

The rainfall erosivity (R) and soil erodibility (K) factors of the Universal Soil Loss Equation (USLE) were determined on two sites in the Colombian Cauca Department over a five year period when rainfall was mostly lower than average. The results showed that the high erosion potential of the soils can be attributed more to high rain erosivity than soil erodibility. The R factor explained between 59 and 81% of the variation in soil loss recorded on continuously clean-tilled fallow plots. The erodibility of Inceptisols in the study region is classified as low. Values for soil erodibility (K) ranged from 0.012 to 0.015 (measured in SI units) in the fifth year of permanent bare fallowing. K factors were higher in the rainy than in the dry season. Soils, previously under grass vegetation, were very resistant to erosion in the first two years of bare fallowing. In the third year erodibility increased sharply and continued to increase steadily until the sixth year. K factors predicted by the USLE nomograph underestimated the empirically-determined erodibility of these highly aggregated clay soils.