The results of the agronomy trials so far have provided a wealth of information, some of which presented in this podcaster. However, the agronomy team has also faced a number of challenges and results were not always as anticipated. The problems faced so far provided important lessons for the agronomy team leading to changes in the N2Africa research strategy. Below we share the the main lessons we have learned and the way we are adapting our approach.
Lessons learned
Institutional vs. biophysical challenges
While the trials have yielded interesting results, it is recognised that the lay-out and implementation of the trials in the previous seasons was often not ideal. In some trials, yields were very low and treatments effects were minimal. It is important here to distinguish, as far as possible, between poor yields caused by institutional problems and poor yields caused by biophysical challenges. Institutional problems (late contract signing, poorly-trained field staff, confusion about roles and expectations, etc.) sometimes led among others to a poor selection of sites, late planting, poor crop management, poor data recording and, in some cases, the complete abandonment of trials. Biophysical challenges (droughts, pests and diseases, poorly fertile soils) were often confounded with institutional problems. For instance, late planting may lead to a higher risk of drought during the crop cycle, or late arrangements with farmers led to the selection of highly infertile outer fields as trial sites (as these were the fields not yet planted). That said, we recognise that farmers have a wide range of planting dates due to labour constraints and other priorities. Further, much of the land available for crop expansion tends to be in a fairly degraded state in countries where population density is high. The emphasis needs to be on ensuring that technologies are tested across the full range of conditions encountered by farmers.
Data collection errors and omissions
In the first growing seasons, some errors were made during data collection. In a few instances, plant samples have gone missing or there were errors in labelling of soil samples. A common error was the use of coarse scales for weighing the total fresh pod weight. For instance, fresh pod weight was sometimes measured with an accuracy of 100 g while total weights from the harvested plot was only 500-1000 g. As this weight is the basis for calculating final dry grain yield, such errors seriously compromise the value of the yield data. In addition, precipitation data were not collected at most sites. Rainfall data currently available are estimated through satellites and may not always accurately reflect the situation on the ground.
Non-responsive soils
Trials were frequently located on very poorly fertile soils. While in some cases, N2Africa technologies led to spectacular yield increases (e.g. from 50 kg to 600 kg soyabean grain per ha as a result of using inoculants and P fertiliser in Zimbabwe) on such sites; in other cases there was barely any response to inputs. Therefore the soil may also be referred to as non-responsive soils. Further research in N2Africa will focus on identifying the biotic and abiotic factors other than N and P availability that limit growth on these non-responsive soils.
Covering the diversity of soils and climates
The type and lay-out of the agronomy trials in previous seasons (fully replicated randomised block designs with usually one trial representing a particular agro-ecology) obviously cannot cover the wide heterogeneity in soil fertility of fields within farms and between farms in a region. To draw conclusions on which technologies work where, and why they work, technologies need to be tested at a larger scale across the full range of environments. The demonstration and dissemination (D&D) trials, which are simple with a single replicate at each site, offer great opportunities to cover the heterogeneity of farmers’ conditions, given the large number of trials that are carried out. The Agronomy teams in the various countries have therefore started systematic data collection in a selection of D&D trials. This requires close collaboration between D&D and agronomy teams. The prime objective of the D&D trials remains the transfer of knowledge and skills to farmers and their communities, but we can learn much more from the D&D trials through measuring performance.
The way forward
- Where national partners experienced problems in execution of the trials we have increased the direct involvement of the local N2africa agronomist in the implementation of the trials. We recognise that the N2Africa project has an important role in capacity building. The need to implement trials well and the need to build capacity of national institutes, even things did not go well in the previous seasons, will be carefully balanced.
- Contract signing with partners and the planning of the agronomy trials has become timelier and will avoid delays during the growing season. This is likely to lead to a better management of the trials.
- More attention is given to timely site selection, so the trials can be located on a variety of soils with different fertility characteristics.
- Technicians have been further trained in the collection of yield data. There has been a major effort to ensure that all weighing scales used in the field are sufficiently accurate and that rainfall data is collected at each site with the help of rain gauges.
- Work on the identification of limiting factors in non-responsive soils has started or has been planned in the coming season (e.g. missing nutrient trials with pots in greenhouses have been initiated in Nigeria and Zimbabwe). Also more attention is devoted to the control of biotic stresses in trials. Some of this work is funded by the supplementary N2africa grant.
- The agronomy teams have started data collection activities in D&D trials. To facilitate this, a Field Book for Technology Evaluation has been developed. This includes the collection of general farm characteristics, planting and germination data, crop management data, yields and soil analyses. The aim is to monitor roughly 300 D&D trials per country.
The agronomy team is especially excited about the prospect of collecting data in large numbers of D&D trials. The team is also grateful that a supplementary grant to N2Africa from the Bill & Melinda Gates Foundation will allow the team to increase its research efforts through the appointment of an additional research assistant in each country and extra operational funding for work on non-responsive soils and soil analyses.
Linus Franke, Freddy Baijukya