Understanding biodiversity distribution is a key ecological goal, with numerous hypotheses linking gradients to environmental factors like energy, water, productivity, and heterogeneity. Most research focuses on wild species, while knowledge about agrobiodiversity distribution, particularly livestock, remains limited. Native livestock breeds, locally adapted intraspecific groups, are valuable for conserving animal genetic resources, ensuring food security, and providing agroecosystem services. Their distribution is shaped by natural and artificial selection, geographic isolation, inbreeding, genetic drift, ecological and historical processes, and human geography. Despite their recognized value, particularly in the context of climate change, knowledge gaps exist regarding the mechanisms of breed adaptation and the ecological perspectives of breed-environment interactions. Wild and domesticated diversity differ significantly in evolutionary pathways. While climate and biogeography strongly influence wild diversity, their impact on domesticated diversity is less understood, affected by human migrations and interactions with wild populations. Human pressures limit both wild and domesticated diversity, particularly in human-modified regions like the Mediterranean basin. This study investigates the association between livestock agrobiodiversity and environmental factors in mainland Spain, hypothesizing that breed richness responds to environmental factors (climate) according to species' physiological needs, similar to wild vertebrates, and that the current distribution relates to rural abandonment and agricultural intensification.

A substantial body of literature explores biodiversity gradients in wild species, linking species richness to environmental factors such as ambient energy, water availability, vegetation productivity, and environmental heterogeneity. However, research on the distribution of agrobiodiversity, particularly livestock, is significantly less developed. Existing studies highlight the role of human factors, including the diversity of production systems and agricultural area, in shaping livestock diversity. Conversely, agricultural intensification, breed replacement, crossbreeding, and economic factors contribute to livestock diversity erosion. While in situ conservation is often preferred, most research focuses on breed genetics and animal production, neglecting ecological perspectives that could enhance our understanding of breed-environment interactions. Studies on wild diversity demonstrate the significant influence of climate and biogeography, but their influence on the distribution of domesticated diversity is less clear. The distribution of domesticated animals is often associated with human migrations, alongside processes of local adaptation and intermixing with wild populations. Environmental heterogeneity is proposed as a driver of both wild and domesticated species richness, influencing niche options, species coexistence, and the probability of speciation. Climatic conditions also play a role, shaping distributional ranges and influencing breed adaptation through physiological mechanisms. Human factors such as the diversity of production systems are positively correlated with the number of reported breeds; however, these same factors also drive diversity erosion through intensification and abandonment of traditional breeds.

This study uses data on 133 extant and extinct local livestock breeds (bovine, ovine, caprine, asinine, equine, and porcine) in mainland Spain. Past distributions (before agricultural intensification) were determined by reviewing literature on breed origins. Current distributions (after intensification) were mapped using municipality-level data from the National Programme for the Conservation, Improvement and Promotion of Spanish Livestock Breeds (2017-2019), supplemented by data from breed associations and conservation programs. Richness indices (number of breeds per cell) were calculated at three spatial scales (10x10 km, 20x20 km, 50x50 km UTM grid cells) for both past and present distributions. Environmental data included annual mean temperature, annual precipitation, precipitation seasonality, and vegetation productivity seasonality (from WorldClim and MOD13Q1 satellite data). Geographically Weighted Regression (GWR) models were used to assess the relationship between breed richness and environmental factors, varying bandwidths (2.5%, 5%, 10%, 20% of data) to account for spatial heterogeneity. Model accuracy was evaluated using global quasi-R². Ordinal Logistic Regression (OLR) models were used to analyze the relationship between current breed richness (at the municipality level) and land-use changes (classified into seven categories based on Corine Land Cover data from 1990, 2000, and 2012) from less to more intensified agriculture, based on the dominant transition. All analyses were conducted using R software.

Environmental factors significantly influence the distribution of local livestock breed diversity, both historically and currently. For past distributions (before agricultural intensification), environmental factors explain up to 73% of the variation in species diversity (global quasi-R² values ranging from 0.39 to 0.73, varying by species and spatial scale). This percentage decreases for contemporary distributions (up to 43% of variation, global quasi-R² values ranging from 0.20 to 0.43), reflecting a weakening of the associations between breed distribution and environmental factors. The distribution of livestock diversity shifted from hotspots in southern and northeastern Spain (Guadalquivir basin and Pyrenees) to southwestern and western regions and Atlantic coastal areas. GWR models reveal contrasting effects of environmental factors on past and present distributions. Past richness showed mostly negative or non-significant associations with temperature and precipitation, except for a positive association in southern Spain. Contemporary distributions show mainly positive associations with temperature and precipitation across most of Spain. Precipitation seasonality had positive effects on past richness but negative effects in northern and southern Spain in contemporary distributions. Vegetation productivity seasonality had positive effects in both past and present distributions. Species-specific analyses revealed similar patterns for ruminants (bovine, ovine, caprine), with generally negative associations with environmental factors in past distributions and positive associations in contemporary distributions, except for precipitation seasonality. Equine and porcine breeds show different patterns. Analyses using coarser spatial scales showed a tendency towards positive associations between environmental factors and contemporary breed richness. Finally, higher livestock breed richness is found in areas undergoing agricultural abandonment, with municipalities experiencing afforestation being up to three times more likely to have high diversity compared to those undergoing intensification.

This study demonstrates that despite human influence on livestock breed differentiation and distribution, environmental factors significantly shape their spatial patterns. The observed weakening of the associations between environmental factors and breed distribution in contemporary settings reflects the increasing impact of human intervention, particularly the intensification of agricultural practices. The shift from past distributions negatively associated with temperature and precipitation in northern Spain to positive associations in contemporary distributions suggests that livestock diversity is adapting to more productive environments. This is consistent with the observation that areas with high breed richness are undergoing agricultural abandonment. The contrasting patterns between past and present distributions highlight the decoupling of livestock diversity from historical environmental constraints, largely due to human-driven changes in land use and agricultural practices. While environmental factors, consistent with the water-energy hypothesis, are important predictors of breed distribution, the underlying mechanisms differ from those of wild species. The study's findings underscore the importance of considering both environmental and human factors in conservation strategies for livestock agrobiodiversity.

This study confirms the significant role of environmental factors in shaping the distribution of livestock diversity, both historically and presently. Although the strength of these associations has weakened with recent land-use changes, it’s crucial for conservation to integrate ecological and sociocultural factors alongside genetic and production-oriented approaches. Future research should focus on higher-resolution data to refine our understanding of the temporal dynamics of breed distributions and on understanding species-specific responses to environmental change within the context of evolving human-driven land use shifts. Improved data regarding the timing of breed origins and more detailed environmental data would further strengthen future analyses.

The study's limitations include the resolution of breed distribution data, which limits precise temporal associations and prevents linking past distributions with specific past environmental conditions. The breed concept itself has a human dimension, influenced by farm environments and anthropogenic selection; this can lead to changes in breed recognition and management over time. However, this was mitigated by analyzing time dynamics (comparing distributions before and after intensification) and conducting sensitivity analyses across different spatial scales and breed groupings. The assumption of relatively stable environmental factors at the study’s scale across time periods, while plausible, could still introduce uncertainty. Finally, the dataset used may not capture the entire spectrum of livestock diversity, particularly smaller, less studied breeds.