Ecoregional research has the potential to help address some of the huge challenges facing agriculture in developing countries by developing technologies that work under different agro-ecological conditions, and the processes by which these technologies can be adapted to work in other areas with similar conditions. The CGIAR system has been developing ecoregional research as a new paradigm for over a decade. In this paper we evaluate one of the most ambitious of these initiatives called the Benchmark Area Approach (BAA) pioneered by the International Institute of Tropical Agriculture. We evaluate the BAA against nine good practice criteria for ecoregional research and finding that the approach is delivering, or has the potential to deliver, on all nine. Many of the lessons learnt from this evaluation will be relevant to current and future attempts to undertake co-ordinated multi-locational research for development.

In this paper we attempt to remove the dichotomy created by distinguishing between participatory and non-participatory breeding programmes by using the degree of client orientation as the basis for an analysis. Although all breeding programmes are implicitly client-oriented, we examine how participatory approaches explicitly increase the extent of client orientation. We briefly review the history of participatory plant breeding (PPB) and analyse the participatory techniques used at different stages of the breeding programme. In common with several other authors, we find that farmer involvement in selecting in the segregating generations may not be an essential component of PPB. However, in some circumstances such collaboration is required and is the subject of a second paper in this series. The purpose of all the techniques used in PPB programmes is to better meet the needs of clients. Thus, breeding programmes can be differentiated by their extent of client-orientation removing the dichotomy involved with the term participatory. We discuss four techniques in the suite of techniques that have been employed by PPB: identifying the target market or clients; using germplasm that can best meet the needs of target clients; matching the environments of the target clients; and product testing in the target market with target clients. Most attention is paid to the last of these four that is often referred to as participatory varietal selection (PVS) and how it is done varies with circumstances. Rice varieties from a client-oriented breeding programme in Nepal were tested in mother and baby trials in Bangladesh. The rapid acceptance of these varieties by farmers illustrates the power of the participatory trials system and the success of a highly client-oriented breeding approach.

The translocation of C and N in a maize-Striga hermonthica association was investigated at three rates of nitrogen application in a glasshouse experiment. The objectives were to measure the transfer of C and N from maize to S. hermonthica and to determine whether the amount of N in the growing medium affected the proportions of C and N transferred. Young plants of maize were labelled in a 13CO2 atmosphere and leaf tips were immersed in (15NH4)2SO4 solution. The Striga×N interaction was not significant for any of the responses measured. Total dry matter for infected maize was significantly smaller than for uninfected maize from 43 to 99 days after planting, but N application increased total dry matter at all sampling times. Infected maize plants partitioned 39–45% of their total dry matter to the roots compared with 28–31% for uninfected maize. Dry matter of S. hermonthica was not affected by the rate of N applied. S. hermonthica derived 100% of its carbon from maize before emergence, decreasing to 22–59% thereafter; the corresponding values for nitrogen were up to 59% pre-emergence and up to 100% after emergence. The relative proportions of nitrogen depleted from the host (up to 10%) were greater than those of carbon (maximum 1.2%) at all times of sampling after emergence of the parasite. The results show that the parasite was more dependent on the host for nitrogen than for carbon.

Pearl millet topcross hybrids (inbred male-sterile seed parent × open-pollinated variety restorer) based on locally adapted varieties and publicly available seed parents provide an ideal entry point into the commercial hybrid seed business, which can stimulate commercial investment by prospective seed producers. To demonstrate this potential, fifteen topcross hybrids made with the widely adapted variety ICMV 221 were evaluated in Eritrea, Sudan and Kenya for overall field performance, and in India for mechanisms of expression of heterosis and for terminal drought tolerance. Across all evaluation environments, the mean yield heterosis was 8%, with a range of −1% to +19%; six hybrids had a statistically significant, positive across-environment yield heterosis (ranging from +11% to +19%). Significant grain yield heterosis in rainy season environments was a consequence of heterosis in both biomass and harvest index, but not necessarily in any specific yield component. Positive grain yield heterosis under terminal drought stress, a common occurrence in millet-growing environments, was related to positive heterosis for grain size. These results are discussed in terms of their support for topcross hybrids as an entry point for prospective millet seed producers, and a scheme presented for the rapid creation, evaluation and marketing of locally adapted topcross hybrids.

Reports of cassava root rot disease from different African countries have increased in recent times. Field studies were conducted from July 1998 to October 1999 to determine a reproducible disease assessment method that would allow the comparison of results from different locations and an evaluation of the relationship between disease severity and root yield. Single point disease assessments at 6, 9, 12 and 15 months after planting (MAP) were compared to multiple points assessment based on the area under a disease progress curve (AUDPC). Single point assessments at 12 and 15 MAP, and the AUDPC identified continuous variation (p≤0.01) among the genotypes. However, a consistent result across trials was obtained only with the assessment based on AUDPC. Root dry yield (DYLD) at 15 MAP showed a strong negative correlation with AUDPC (r=−0.74). Regression analysis also confirmed the negative relationship between yield and root rot severity. The five genotypes compared were separated into resistant (91/02324, 30572 and 92/0427) and susceptible (92/0057 and TME-1) groups. It was concluded that root rot disease may cause significant yield loss; however, the magnitude of the yield loss will depend on the susceptibility of the cassava genotype.

Maize and soybean are commonly intercropped in the drier zones of the western mid-hills in Nepal, but farmers report that productivity of soybean has been declining in recent years. Two researcher managed on-farm field experiments were conducted in the mid-hills environment of Nepal during 2001 and 2002, and one glasshouse experiment at the University of Wales, Bangor during 2003, to determine whether varying densities of maize and soybean influenced productivity of the system and to what extent soybean exhibited adaptation to shade. In neither season was maize yield affected by the presence of soybean, but grain yield of soybean was reduced in mixture by means of 59 and 53% during 2001 and 2002 respectively. Biomass and grain yield of maize were greatest at 53×103 plants ha−1 and least at the lowest density, whilst conversely biomass and grain yield of soybean increased. With increasing maize density, rates of accumulation of dry matter and leaf area index also increased, the latter resulting in decreasing transmission of light to the intercropped soybean. Soybean exhibited no photosynthetic adaptation to shade, but the specific leaf area was greater in artificially shaded and intercropped plants. Land equivalent ratios of all intercrops were greater than unity (1.30 to 1.45), indicating higher efficiency of intercropping compared to sole crops. Given the low plasticity in response of the maize canopy to variations in density, it is suggested that soybean could be better grown under maize by increasing between-row spacing of maize from 0.75 to 1.0 m to improve light transmission to the understorey, resulting in higher overall productivity of the intercropping system, and also that soybean germplasm be screened for adaptation to shade.