Habitat-based spatial patterns enable realistic reconstruction of individual distributions across complex landscapes

The spatial structure of individuals both reflects and drives key ecological processes. However, in large or densely populated landscapes, logistical constraints limit our ability to capture these processes at realistic population sizes and spatial scales. Spatially explicit individual-based models, for example, offer a powerful framework to explore how spatial processes influence ecological dynamics, particularly when direct observation is unfeasible. Yet, these models require detailed information on the initial spatial arrangement and traits of individuals—data rarely available for large, heterogeneous, or novel landscapes. Habitat heterogeneity further complicates this task, as it shapes both spatial patterns and individual traits in ways that vary across habitat types. Here, we quantified how individual spatial patterns and traits differ among habitat types, and used this information to reconstruct realistic, landscape-scale spatial distributions suitable for spatially explicit modeling. We mapped and recorded morphological and reproductive traits over 5,000 individuals of the Mediterranean dwarf palm (Chamaerops humilis) across 146 hectares spanning four distinct habitat types within Doñana Natural Space (SW Spain). Our results shown strong small-scale aggregation in all habitats, with significant variation in aggregation level, plant size, reproductive status, and sex ratio. Using replicated spatial point pattern analyses, we fitted habitat-specific Thomas cluster models that captured this variation and upscaled them to simulate spatial distributions across a heterogeneous 1,950-hectare landscape. Simulated individuals were assigned traits based on empirical, habitat-specific trait frequencies, generating >59,000 individuals whose spatial and demographic structure closely matched field observations. Our results underscore the strong influence of habitat on both spatial and individual traits distribution and demonstrate that incorporating this variation is essential for building ecologically realistic models of plant populations in heterogeneous landscapes.