Research Papers

Rain does not fall on one roof alone. Farmers' management strategies to cope with precipitation variability among rainfed marginal systems with A. Mantino, G. Ragaglini, C. Fadda, M.E. Pè and A. Nuvolari

Under submission

In a changing climate, the awareness of precipitation variability is key to food security of smallholder farmers in rainfed marginal systems. In agricultural areas dependent on rainfall phenomena, households able to interpret correctly short-term precipitation deviations and extremes do retain an advantage in terms of resilience. Despite the communitarian nature of these rural societies, we know little about how collective processes influence rainfall risk perception and households’ awareness. In our study we join data from agronomic surveys, daily rainfall estimates and socioeconomic surveys to explore links between the growth of households’ rainfall awareness and social learning dynamics. We build a repeated cross-section dataset, interviewing 280 smallholder households in the Ethiopian highlands in the spring of 2013 and 2019, while computing the mean, the standard deviation and the maximum of three rainfall parameters during the crop growing seasons. By analysing the growth of the household’s precipitation risk perception and the growth rate of the three rainfall parameters, we identify a measure of farmer’s awareness to short-term rainfall variability, which shows high heterogeneity among neighbouring households. Regressions reveal the role of collective learning, which withstand controls for farmers’ age, gender and income. Instrumenting the idir, an informal institution present in Ethiopian societies, we investigate further the collective formation of rainfall awareness. Our findings suggest that collective mechanisms of learning should be considered in mitigation and risk reduction programs. Our work adds to the growing body of literature screening socioeconomic determinants of risk perception in natural hazards environment; it contributes to the debate on reducing households’ climate vulnerability, advising the consideration of social learning mechanisms in rural policy development.


How the diverse farmers-scientists programs contribute to the resilience of smallholder farmers in marginal agroecosystems: the case of participatory varietal selection initiatives in the Trifinio area with K. de Sousa, M. Dell'Acqua, J. Sellare, K. Paredes, J. Robalino and J. van Etten

Working paper available upon request


Farmers' local knowledge plays a key role in ensuring the sustainability of marginal areas, but its tacit nature often prevents the inclusion in scientific debates. For centuries, traditional and agroecological information have been processed locally by farmers, in the effort to cope at best with the environmental ecosystem which surrounds them. The significance and scale of the potential of local knowledge in community-based development may not be immediately obvious until one reflects on the diversity and complexity of local knowledge systems and their importance in agricultural production and the development agenda. As of today, participatory farmers-scientists approach is a unique strategy to include local ecological knowledge in sustainable studies. Innovative approaches, such as crowdsource-based initiatives, have been recently developed to overcome known deficiencies of state-of-the-art participatory techniques; however, the outcomes in terms of households' resilience are yet to be fully explored. This paper addresses this gap through a study linking a traditional form of participatory approach, the participatory varietal selection process (or PVS), and a crowdsource-based approach, the so called Crowdsourcing Crop Improvement (or CCI). Investigating these two different mechanisms of farmers’ engagement, we focus on their effects on inputs’ diversification (specifically linked to SDG1 and SDG13) and on food security (specifically linked to SDG2). Stemming from a diversified participatory experiment in the Trifinio area (at the crossroad between Guatemala, Honduras and El Salvador), this paper aims at understanding if different forms of participatory engagement influence smallholders’ performances. Randomly assigning 2331 smallholder farmers among 148 communities to two different participatory treatments (PVS or CCI) and one control group (no participatory approach iterated), we are able to quantify the effect of the different degrees of farmers’ engagement on households’ diversification and food security. Regressions reveal that farmers involved in the PVS initiative as well as the CCI program have a higher degree of on-farm varietal diversification with respect to control farmers. Effects on household food security are less straightforward: both participatory initiatives increase the odds of experiencing low levels of household food insecurity after three years from the start of the engagement program, but the effect is stronger and more significant for traditional techniques. On the other hand, the crowdsource-based program ensures a lower self-induced attrition rate, thus ensuring a lower drop-out rate in terms of farmers’ participation. Results call for further research on the comparison among different farmers’ empowering initiatives.


agriLOVE: an agent-based evolutionary model of the agricultural sector with explicit environmental boundaries with M. Coronese, F. Lamperti and A. Roventini

Working paper available upon request

This work present a novel agent-based model of land use and agriculture. The model focuses on the interactions between technological change and food production, in a spatially explicit economy populated by boundedly rational farmers which interacts with the environment and among themselves, and compete to fulfill an increasing demand for food. It represents a flexible tool to investigate the impact of innovation diffusion, patterns of imitation, behavioral and geographical factors on food production and land use. We firstly show the model to be able to replicate key stylized facts of the agricultural sector, and then extensively explore it across several scenarios featuring different institutional and behavioral settings. Finally, we employ the model in two distinct applications on soil degradation and climate impacts, showing a poor ability to cope with inter-temporal trade-offs arising from soil over-exploitation, hysteresis and non-trivial spatial propagation effects after weather-related shocks.


Is green the new black? An ABM approach on transition toward sustainable agriculture with M. Coronese, F. Lamperti and A. Roventini

Working paper available upon request

Since the last decades economic and population growth increasingly pressured the Earth system. Land management is essential to ensure sustainability of future food security, but it needs to cope with rising global environmental risks and constraints. In this work we extend the agriLOVE agent-based model of an agricultural sector to investigate interactions between different agricultural regimes in an economy exposed to explicit environmental boundaries. In particular, we study the ability of the system to favor a transition to a sustainable regime when prolonged cultivation with conventional techniques yields a slowdown in productivity dynamics, due to soil degradation. We investigate transition dynamics under several behavioral, environmental and policy scenarios. Our results point to strong path-dependece and show how the agricultural sector has a very limited capacity to ease transition towards a sustainable regime if not supported by appropriate policies. Intuitively, the transition is negatively affected by the velocity and timing of the phenomenon of soil-erosion. The presence of incomplete information has an ambiguous role: being endowed with more information paradoxically lower transition likelihood, although it speeds up the velocity of the transition itself. We further show the existence of trade-offs between distinct spatial configuration, particularly between different degrees of sustainable farms clusterization. Finally, we demonstrate how subsidies to sustainable farming are relevant to avoid lock-ins of conventional farming, but they are only marginally effective in fostering transition. Taxes on conventional farming can help the diffusion of sustainable practices, but conditional on being severe and targeted at R&D activities.


How informal share of agricultural sector is shaping agrobiodiversity among smallholder farmers in marginal areas: the case of idir in Ethiopia

Work in progress