Agriculture around the world is affected by systemic risks, including production, market, financial, organisational, personal, and climatic challenges. To adapt to these risks and improve productivity, farmers rely on both formal and informal social networks. Through relationships and collaborations with peer farmers and agricultural experts, specialists, and organisations, farmers benefit not only from strong ties of trust and mutual support, but also from access to advanced scientific knowledge, infrastructural and financial resources and support through weaker ties with external actors. Along with this, the social network of agricultural organizations plays a vital role in facilitating farmers’ access to knowledge and resources and to support them in facing risks and issues. Thus, analysing the structure and characteristics of these networks and identifying key actors involved in policymaking and developing strategies that strengthen farmers’ resilience and adaptation is crucial. This study examines the social network of farmers and agricultural organisations involved in pistachio production in Central Iran. Pistachio is among the most significant agricultural products in this region, playing a significant role in economy, society, and culture of this area. However, producers face diverse risks, particularly water shortage and market-related challenges. By identifying the systemic risks affecting pistachio production and applying both qualitative and quantitative analysis of social network of local farmers and organisations, this study aims to provide insights into how different networks influence farmers’ adaptation to those risks and issues and their productivity.

Prior research indicates that in developing countries where agriculture is economically significant, links between farmers are important for transmitting information and generating social capital and support. This is especially the case in rural Ethiopia, which is also one of the countries most at-risk from climate change. To study how linkages between farmers impact adaptation behavior and climate resilience, an egocentric network survey using snowball sampling was launched in four Ethiopian kebeles (local administrative divisions) to study informational and support networks. The survey is part of a larger follow-up survey to a baseline household questionnaire conducted in additional kebeles. Results from the baseline survey and a geospatial model indicate that household responsiveness to livelihood shocks is highly localized, perhaps indicating a role for village networks in supporting resilience. In the network survey, we find important differences between kebeles in terms of access to extension services and output markets. We also find that, while informational and support networks are topologically similar within kebeles, cross-kebele differences are notable. In the next steps of the project, I simulate full networks from our partial network data, and then will use an agent-based model of diffusion to study how adaptation strategies propagate differently through these various networks. This work contributes to a growing literature on the importance of networks to learning and capacity-building in sub-Saharan Africa, as well as the relevance of such networks to climate change adaptation. Our results will also help to inform interventions targeting at-risk households and communities.

This paper explores the knowledge network and flow within the seed treatment sector, focusing on the development of alternative biological solutions for seed protection in response to the French Regulation’s 2018 EGALIM law (Article 83), which prohibits the use of unapproved plant protection substances. The purpose of this paper is to examine how diverse stakeholders, including seed producers, research laboratories, and biocontrol companies, collaborate to develop innovative biological seed treatments, addressing the complexity of this innovation that spans seed, seed technology, and biological sectors.

The study models the networks connecting stakeholders within Vegepolys Valley, a French innovation cluster specializing in plant breeding and agricultural solutions. Tools like RStudio and UCInet are used to quantify knowledge flows, with metrics such as centrality and density assessing the influence of individual actors and the cohesiveness of the network. The analysis also identifies key knowledge brokers and gatekeepers who control and facilitate information flow, shaping innovation pathways. The study further examines how knowledge moves between clusters within France, Europe, and globally.

The contributions of this paper are twofold. First, it provides a unique focus on biological seed treatment as a complex, multi-sector innovation involving collaboration across the seed, seed technology, and biological industries. Second, it applies quantitative methods to analyze knowledge flow within a heterogeneous cluster, emphasizing the economic and organizational dimensions of seed treatment innovation. This offers new insights into how regional clusters foster sustainable agricultural solutions in response to regulatory changes.

Periodic monitoring of pests and diseases in crops is essential for early detection, enabling preventative agrotechnical interventions while minimizing treatments in disease-free areas (Jeger et al., 2018). However, monitoring is often hindered by resource limitations, leading to delayed detection and unaddressed outbreaks, resulting in excessive pesticide use, escalating costs, environmental damage, and risks to worker and consumer health (Savary et al., 2019). This study employs social network analysis (SNA) to model the spatiotemporal dynamics of biotic stressors in two distinct agroecosystems (Garrett et al., 2018).

We investigated two systems: [1] the two-spotted spider mite (TSSM, Tetranychus urticae), affecting screenhouse-grown sweet peppers (Capsicum annuum) in southern Israel (Weintraub & Palevsky, 2008); and [2] white mold (WM, Athelia rolfsii), damaging open-field peanuts (Arachis hypogaea) in northern Israel (Dafny-Yelin, 2022). These systems represent contrasting environments and pathogen types, allowing for comparison across agricultural contexts.

During 2015–2017, TSSM abundance was monitored weekly in four pepper screenhouses by examining every 20th plant in every fifth row . In 2024, seven WM-infected peanut fields were monitored weekly or bi-weekly using a 20 × 1 m grid system.

To construct network models, each monitored field divided into grid. Each grid cell represented by a node. Directed edges (e_ij) between nodes indicated pathogen detection in grid cell i at time t followed by detection in grid cell j at time t+1 (Parry et al., 2014). Edge weights were calculated as the inverse distance between cells, reflecting proximity's influence on transmission probability (Sanatkar et al., 2015).

Statistical analysis using exponential random graph models (ERGMs) revealed probabilities of edge formation between grid cells (Robins et al., 2007). The models incorporated spatial covariates. Other covariates such as environmental parameters, and crop-specific factors can be added (Silk et al., 2017).

Results showed both common patterns and system-specific differences. In both networks, infection spreading to western grid cells had lower probability (p < 0.01). However, in TSSM networks, edge probability increased with distance, suggesting long-range dispersal capabilities possibly aided by human movement or equipment (Skirvin & Fenlon, 2003); conversely, in WM networks, edge probability decreased with distance, aligning with soil-based transmission through mycelial growth and localized sclerotia germination (Xu et al., 2012).

Network centrality measures identified 'hotspot' locations disproportionately influencing overall infection dynamics (Pautasso et al., 2010).

This research demonstrates the potential of applying SNA methodologies to agricultural crop protection (Shaw & Pautasso, 2014), offering a framework for optimizing sampling strategies by identifying high-risk locations and transmission pathways, enabling more efficient resource allocation in pest and disease management (Cunniffe et al., 2015).

Agri-food systems in Latin America face critical challenges arising from demographic exploitation, climate change, and migration. To address these, it is essential to understand the social structure embedded in these complex systems. However, there is a knowledge gap in how power dynamics, collaboration, and reciprocity within these networks affect the efficiency of these systems. This study addresses this gap by combining the Social Network Analysis method and "usable past" approach to assess the impact of CGIAR’s “AgriLAC Resiliente” Initiative implemented in Honduras, Guatemala, Colombia, Mexico, and Peru, on the transformation and configuration of collaborative networks in the agricultural sector. This approach allows understanding structural properties of the network by analyzing the relationships between actors before the intervention, which we used as a baseline for evaluating changes in the structure, connectivity, and collaboration dynamics among actors. The results show a significant expansion of the network, with increased participation of actors and connections, which has facilitated a more efficient flow of information, reduced distances between actors, and increased transitivity, suggesting better collaboration, knowledge generation, and innovation. Challenges include low network density and decreased reciprocity, highlighting the need to promote more equitable exchanges and deepen actor integration to strengthen long-term resilience. We provide empirical evidence on how network-based interventions facilitate exchange, collaboration, and innovation in agri-food systems to respond to major challenges, along with recommendations for improving the dynamics in the social structure of agri-food systems. These results can be used to inform public policies oriented at promoting more equitable and sustainable collaboration.