The developmental and scientific impacts of research carried out intermittently over a thirty year period on the tea crop (Camellia sinensis) in southern Tanzania are reported. It is an attempt to demonstrate to questioning donors that research can contribute, in many ways, to the alleviation of poverty in developing countries. During Phase I (1966–1972) the work was funded in large part by the UK Department for International Development (DFID), as it is now known, with the support of the tea industry in East Africa. During Phase II (1985–1991) a regional study in eastern Africa was largely funded by DFID with local field experimentation in Tanzania paid for by tea growers. In Phase III (1992–1997) the work was funded almost entirely by a consortium of commercial producers.

Throughout this period, despite political and economic uncertainties in Tanzania, the close contact with industry meant that the results of the research were exploited to commercial advantage whenever circumstances allowed. These interventions enabled productivity on a unit area basis to rise over time and companies to remain profitable. More recently major investments in new projects have taken place in southern Tanzania, again supported by the results of the research. All of this has led to the creation of new jobs and to continuity of employment in rural areas that have few other sources of paid work. The financial return on the research investment, although very difficult to quantify with precision, has been immense.

The science has not been neglected either. Over the same period more than 30 refereed papers have been published in the international scientific literature. These cover such topics as the responses of tea to climate and weather, including drought and irrigation. Scientists have also been trained in modern research techniques. In addition there have been ancillary benefits including the introduction of technical and management training programmes for staff in the tea industry. Frustration has been caused by the lack of continuity in funding due to changes in policies, and in the priority research topics, even within one donor organization, which have often put the work at risk. Ingenuity, patience and commitment have all been pre-requisites to ensure that the research continued.

The partnership with industry has been beneficial to all the stakeholders. It has led directly to the creation of the new, commercially managed Tea Research Institute of Tanzania serving both small- and large-scale producers throughout the country. Researchers, with the support of the tea industry and in partnership with the Tanzanian Government and DFID, can continue to play a significant role in the fight against poverty. Research for development is a continuing process: it can be successful.

Four surface sterilization agents and two hot-water treatments were evaluated with a view to eliminating bacterial contamination in sugarcane (Saccharum hybrids) buds used for in vitro initiation. None of the treatments achieved total elimination of contaminants. Disinfection in either a 2.7% solution of sodium hypochlorite (pH 6) or 1% mercuric chloride were equally good, resulting in 46.8% and 56.3% clean cultures respectively. However, sterilization with 2.7% sodium hypochlorite (pH 6) was most phytotoxic and 100% tissue necrosis and bud death were recorded. The Australian cold-soak hot-water treatment (consisting of a 48-h soak in running water followed by hot-water treatment at 50 °C for 3 h on the following day) reduced contamination from 70% to 25%. With the dual hot-water treatment (setts immersed in water at 52 °C for 20 min and then at 50 °C for 3 h on the following day) a decrease to 45% was observed. Bacterial contaminants were identified by fatty acid profiling and were found to belong mainly either to the Enterobacteriaceae family or the genus Bacillus.

Groundnut (Arachis hypogaea) cannot be grown economically on unamended acid soils because of poor pod-filling resulting from an inadequate supply of calcium. Gypsum improves pod-filling but in Mauritius it is now too expensive.

Field trials were conducted at two locations in 1994–95 and at four locations in 1995–96 to find out if cement could be used instead of gypsum. In three trials where the soil pH ranged from 5.4 to 6.3 and the soil calcium level from 2.4 to 6.9 cmol kg−1, there were few empty pods and no response to gypsum or cement was observed. At the remaining three sites, the soil pH was 5.5, 4.9 and 4.3 and the soil calcium levels were 1.8, 0.39 and 0.92 cmol kg−1, respectively. In the untreated plots at the three sites 17.4%, 27.5% and 50.6% of the pods were empty. Cement at rates of 525 and 700 kg ha−1 and gypsum at 1000 kg ha−1 applied to the foliage at flowering reduced the number of empty pods to less than 10% and significantly increased seed yield, shelling percentage and commercial-grade kernels. However, cement was not as effective when placed in the furrows at planting. At current prices, the use of cement was cost effective whereas that of gypsum was not. Cement may therefore be proposed as a replacement for gypsum for the improvement of pod-filling in groundnut.

This study demonstrates a relationship between earliness of ripening and stalk elongation which could be useful in a sugarcane selection programme. Sixty-four sugarcane genotypes with contrasting ripening behaviour were grown in a quadruple-lattice design and ordinary logistic curves were shown to give the best fit between stalk height and time. Genotypes with early high sucrose contents gave logistic curves with earlier inflection points and steeper slopes at the point of inflection. The inflection point was shown to be a reliable indicator of time of ripening and it was suggested that stalk height per se could be considered as a selection criterion for earliness of ripening in sugarcane.

In two cultivar × density trials for oil palms (Elaeis guineensis) planted in Indonesia, single leaf area, number of green leaves per tree, leaf opening rate per year and rachis length of leaves were followed over fourteen years. The data were analysed to determine the time course of canopy leaf area and to predict the moment of canopy closure.

Growth functions were fitted to the observed data. Estimates of leaf area index (L) were based on single leaf area, number of green leaves, leaf opening and planting density. The time course of L was modelled on the basis of the fitted functions to the components. The moment of canopy closure was calculated from the planting density and the functions fitted for rachis length.

The modelled time course of L was considerably different from the function fitted to the single leaf area data. The expansion of L was not as rapid as expected from the area growth of single leaves and, after maximum L was reached, a steady decline was observed. The continuously declining number of green leaves was the main cause of these two observations.

The time course of L differed considerably between the two experiments. Not only were there large differences in the number of green leaves maintained per tree in the experiments but also the final area of single leaves differed between both experiments. The first factor was a result of the management of the experiments, whereas the second factor was most likely influenced by a difference in soil-related factors at the two locations.

Leaf areas and numbers of leaves per tree were different for each cultivar, as was L. This was also the parameter most sensitive to planting density. Individual leaf area and leaf number per tree were not affected by planting density, but rachis length was affected by the planting density treatment. The moment of canopy closure was similar in both experiments. Planting density was the main factor that determined the onset of canopy closure.

The effects of drought and plant density on fractional interception (f), radiation-use efficiency (RUE) and dry matter accumulation and partitioning to seeds in cowpea (Vigna unguiculata) need to be quantified in order to calibrate cowpea simulation models for West Africa. Accordingly, a short duration cultivar of cowpea, IT84S-2246-4, was grown at a range of plant densities in well-watered and drought conditions in 1990–91 at two contrasting locations in Nigeria, Ibadan (7°N) and Kano (12°N). Fractional interception of radiation (f) and total above-ground dry matter accumulation (TDM) were measured throughout the growing season and TDM and seed yield (SY) were measured at maturity. At Ibadan drought had no effect on phenology while at Kano drought increased the duration of the vegetative phase (by 5–6 days), but not of the reproductive phase. Drought reduced f and radiation-use efficiency (RUE), and hence TDM, and there were significant interactions between RUE and plant density at both Kano and Ibadan. Average SY in the well-watered treatments at Kano and Ibadan were 1600 and 700 kg ha⊃−1&/sup; respectively and drought reduced SY by 50% at both sites but did not interact with plant density. Across locations, plant densities and drought treatments SY was strongly correlated (r = 0.96) with TDM production between flowering and maturity. These data suggest that partitioning of dry matter during the reproductive phase is not affected by location, drought or plant density and the SY of cowpea can be predicted from dry matter accumulation.

Simple models of the growth and development of legume cover crops in different environments may aid the selection of species and genotypes most suitable for particular farming systems. The effects of temperature and photoperiod on the daily rate of seedling emergence (1/t50%) and the rate of leaf appearance (leaves d⊃−1&/sup;) were quantified for 12 diverse legume cover crop species. Six tropical or subtropical species were grown in 12 combinations of mean temperature (16.8, 21.8 and 26.8 °C) and photoperiod (11.5, 12.5, 13.5 and 14.5 h d⊃−1&/sup;). Another six genotypes of temperate species were grown in nine combinations of mean temperature (16.8, 21.8 and 26.8 °C) and photoperiod (12.5, 13.5 and 14.5 h d⊃−1&/sup;). Durations from sowing to 50% seedling emergence (t50%) were recorded and the number of leaves on the main stem counted during early vegetative growth. Photoperiod did not affect either 1/t50% or leaves d⊃−1&/sup;. However, the effects of temperature, photoperiod response group (short-day or long-day response), genotype or species and their interactions were all significant. An optimal temperature between 21.8 and 26.8 °C for leaves d⊃−1&/sup; was identified for Vicia sativa, Vicia dasycarpa and Lupinus mutabilis. Below the optimum temperature, 1/t50% and leaves d⊃−1&/sup; were positive linear functions of temperature for all legume species except Vicia dasycarpa. Estimates of base temperature (Tb) and thermal time (θ) for 1/t50% and leaves d−1 were obtained for each legume cover crop from these functions. Common values of Tb were found within each photoperiod response group: 9 °C and 1 °C for 1/t50%, and 10 °C and −2 °C for leaves d⊃−1&/sup; for short-day and long-day species respectively. Differences among species were represented by variant estimates of θ. The practical consequences of the combined effects of these values of Tb and θ for selecting legume cover crops for particular farming systems are discussed.

Leaf senescence in the rice (Oryza sativa) cultivars FR13A and IR42 under submergence was assessed in terms of changes in total chlorophyll, soluble amino acids and protein concentrations and peroxidase activity in crude extracts. The objective was to determine whether delay in leaf senescence was related to the submergence tolerance of a rice cultivar. Submergence induced senescence and its extent was notably different in the two cultivars. Results indicated a reduction in chlorophyll and protein concentrations in leaves during submergence but an increase in amino acid concentration and peroxidase activity. These relative changes were more pronounced in submergence intolerant IR42 than in tolerant FR13A even at four days after complete submergence. When plants were desubmerged and returned to standard irrigated conditions after 10 days of complete submergence the rate of recovery was different for the two cultivars. FR13A showed a greater protein and chlorophyll restoring ability compared with the intolerant IR42. The differences observed between the responses of FR13A and IR42 to submergence were likely to be due to differences in proteolysis. Senescence of leaves due to submergence was similar to the senescence of non-submerged excised leaves. Results demonstrate that, in the two cultivars studied, leaf senescence is an important biochemical mechanism in plants under submergence and its slower development in tolerant cultivars is, in part, responsible for submergence tolerance.

Some agronomic, economic and ecological aspects of an agroforestry system combining Grevillea robusta trees, bananas (Musa spp.) and beans (Phaseolus vulgaris) were studied in the subhumid highlands of Burundi. Three densities of G. robusta, 208, 313 and 625  trees  ha−1 were interplanted in plots of bananas, beans and a banana–bean mixture. When G. robusta was interplanted with bananas, the tree had a positive effect on banana yield with maximum yield occurring at 300  trees  ha−1. In the presence of beans, banana yield was not influenced by the densities of the tree. Under low fertility regimes, the yields of beans interplanted with G. robusta were equal to or greater than those of the no-tree plots. Maximum bean yields were observed at between 283 and 295  trees  ha−1, representing a 25–135% yield increase above the no-tree control. When soil fertility was raised by the application of fertilizers, bean yields declined with an increasing density of G. robusta. Three years after planting, the growth of G. robusta was not affected by tree density. Mean height and mean basal diameter were 7.1   m and 13.2   cm respectively. Below ground competition between beans on the one hand, and bananas or bananas and G. robusta on the other, was more important than competition for light. Compared with the crops alone treatments, cumulative net present values of tree plots improved from being negative in the first year to being positive in the third year. Net present value was highest in two of three instances when the density of G. robusta was 625  trees  ha−1. A multistrata arrangement of bananas and beans with G. robusta is viable for the low input agriculture system of the highlands of Burundi.

Transient soil-moisture-deficit stress was imposed on groundnut (Arachis hypogaea) at three phenophases for different durations: long stress in the early vegetative phase (20 days after sowing); moderate stress in the early vegetative phase (20 days after sowing); stress at flowering (40 days after sowing); and stress at pod development (60 days after sowing). Stress was imposed for 30 or 25 days at the vegetative stage. Transient soil-moisture-deficit stress, at all phenophases, reduced the production of flowers. Soil-moisture-deficit stress for 25 days at the vegetative phase followed by two relief irrigations at an interval of 5 days, resulted in closely synchronized flowering. This factor contributed to a greater efficiency of conversion of flowers to pods and to higher pod yields. Total biomass accumulation was also higher in plants which experienced stress in the vegetative phase. Groundnut pod yields were increased by imposing a transient soil-moisture-deficit stress in the vegetative phase for 25 days, followed by two irrigations at an interval of 5 days. Thus stress in the vegetative phase was beneficial for groundnut growth and pod yields, but was highly detrimental when imposed at flowering and pod development.