The need for and feasibility of diversification in agriculture

Diversification has become a critical strategy in farming systems for mitigating risks, while also providing socio-economic benefits and ecological resilience. However, the dominant conceptualization of diversification in land use science focuses primarily on crops or farm-level practices, overlooking its relevance across interdependent components namely science, policy, and practice, that are determinants of agricultural diversification. This narrow framing limits understanding of the benefits, challenges, and constraints of diversification, thereby restricting its potential to support sustainability. This thesis addresses this gap by examining the need for and feasibility of diversification across three domains: (i) knowledge production, (ii) farming systems, and (iii) decision evaluation processes. <br> A scientometric analysis of 161,909 peer-reviewed land-related research articles (2000-2021) reveals a lack of diversification in knowledge production, with women underrepresented (27%), and authorship dominated by White (62%) and Asian (30%), while Hispanic (6%), and Black (2%) women remain significantly marginalized. Intersectional inequalities are particularly stark, with Black and Hispanic women significantly underrepresented in lead authorship positions. These structural patterns of exclusion highlight the need for diversity of targeted efforts to diversify who produces land-related knowledge and to expand the epistemic basis of land-related science. <br> In terms of feasibility of scaling diversified farming systems based on niche modeling approach using maximum entropy (MaxEnt) with socio-economic variables and observed locations of profitable diversified farming systems, the analysis identified infrastructure as a key impediment. Globally, 47% of the land was suitable for profitable diversified farming systems, with higher suitability in the Global North and in areas close to cities in the Global South. These findings suggest targeted investments in infrastructure, especially in the Global South, could enhance the scalability and adoption of diversified farming systems. <br> In evaluating the decision to consolidate land among smallholder farmers through decision analysis approach, stakeholder diversification including different groups of smallholder farmers emerged as essential in capturing the full impacts of land consolidation. Engagements with stakeholders revealed that conventional land consolidation models inadequately address farmers' needs. Consequently, stakeholders co-developed alternative model that better aligned with their realities, offering potential for more responsive land consolidation policies and models. Probabilistic modeling of financial viability of land consolidation for smallholder farmers indicated that land consolidation was less preferable (27%) than maintaining small farms (72%), with maize and land prices as key determinants. This participatory modeling approach demonstrates the value of stakeholders under uncertainty and data scarcity. <br> Together, these studies demonstrate that diversification must be expanded beyond a narrow agronomic scope to include representation in knowledge production, context-sensitive decision-making, and enabling socio-economic conditions for diversified farming systems adoption. This broader conceptualization is not only necessary to reduce systemic risks associated with simplification but also feasible across key domains. The findings underscore diversification's central role in advancing agricultural sustainability and provide methodological and policy-relevant contributions for future interventions.