NRM practices, ag. extension and measuring adoption

An oft-discussed issue when it comes to NRM practices (generated from CGIAR research) is what constitutes 'adoption'. While this is a result of the fact that practices are often a combination of recommended management techniques that farmers deploy differently – like in the case of conservation agriculture where a farmer may use zero tillage + residue retention in one season, but only practice residue retention in another – one can also start to see the deviation away from ‘blanket recommendations’ in agricultural extension over the decades, leaving more room for interpretation (and innovation) by farmers. Blanket recommendations for fertilizer application do have several issues in that they do not take diversity of farming environments into account, and hence are not always profitable for farmers. But this deviation from blanket recommendations brings up other issues.

Take the case of micro-dosing – during the December 2015 workshop on documenting adoption of NRM practices, Kizito Mazvimavi of ICRISAT spoke about how the recommendations made are typically in reference to top dressing, but the micro-dosing method can be used for basal application (during planting) as well. More importantly, micro-dosing is described as reduction in 'normal' fertilizer application (between 1/5th and 1/3rd of the quantity). This then leads to a question of what "normal" is, and whether by not defining what the recommended quantity is, there is a risk that farmers resort to something that is unsustainable (consider the scenario where plants take more nutrients out of the soil because not enough is provided by the farmer). What happens in areas where farmers have never used inorganic fertilizers before – how do they determine what optimal quantities are because this complicates how they will experiment and learn? Hence, for the CGIAR, the questions of efficacy and effectiveness of micro-dosing as a scientific practice remains. And for the purpose of impact assessments and SPIA's work, the question of how we determine adoption rates is as challenging as ever. Is a one-time cross-sectional survey with farmers reporting what their previous dosing rates were (recall method) sufficient in the absence of panel data? Should a subset of farmers be asked to maintain diaries across seasons (leaving aside issues of how the keeping of records – a farming journal in this case – in itself might be a learning intervention)? We would love to hear your thoughts!

For those unfamiliar with micro-dosing, here is a little video from ICRISAT