Biodiversity dynamics with complex genotype-to-phenotype architecture in multilayer networks

2 Abstract The genotype-to-phenotype architecture (GPA), defined by complex interactions such as pleiotropy, epistasis, and regulatory control, is a fundamental yet often overlooked driver of biodiversity dynamics. While empirical evidence suggests that traits mediating species interactions (biotic) and environmental responses (abiotic) are frequently correlated, most eco-evolutionary theories treat these traits as independent, leaving a gap in our understanding of how genomic architecture influences community-level outcomes. In this study, we contrast two distinct GPAs, modular (independent trait evolution) and correlated (integrated trait evolution), within a spatially explicit multilayer network framework. We evaluate their impact on biodiversity across varying regimes of selection, migration, and biotic and environmental filtering. Our results reveal a hierarchy of drivers: selection strength dictates the absolute magnitude of the architectural effect, while migration and context-dependent biotic and abiotic effects determine which architecture yields a diversity advantage. Correlated GPAs enhance species coexistence and diversity in low-migration landscapes characterized by strong selection and moderate, balanced biotic and abiotic pressures. In these contexts, trait integration serves as a buffer against selective noise. Conversely, modular GPAs support higher diversity under high migration and strong biotic interactions, where the decoupling of trait modules provides the adaptive flexibility necessary to navigate spatially conflicting selective pressures. Our findings demonstrate that genomic architecture acts as a critical filter for environmental perturbations. Integrating complex GPAs into multispecies models is essential for quantifying the co-evolutionary feedbacks among traits, population adaptation, and species persistence. Our framework provides a path for predicting how biodiversity emerges and persists across biological scales, from genomics to communities and food webs, under the accelerating pressures of global change. 1 Conclusions We integrate trait architecture to spatial biodiversity to show biodiversity patterns are not merely products of ecological interactions, but are fundamentally constrained by Genotype-to-Phenotype Architecture (GPA). By linking GPA to biodiversity we show the interplay between the complexity of an organism and community structure in determining diversity patterns. The hierarchy of Eco-Evolutionary Drivers: We establish a new conceptual hierarchy where selection strength acts as the fundamental governor of architectural impact, while the specific architecture predicting higher diversity (Correlational vs. Modular) is dictated by the interplay of migration scales and context-dependent biotic and abiotic dynamics. Selection-Migration contingency for coexistence: We provide a new hypothesis for species coexistence: Correlational selection serves as a stabilizing force under dispersal limitation, whereas Modular trait architecture provides the adaptive flexibility to maintain diversity in high-migration, spatially heterogeneous landscapes. Adaptive decoupling as a diversity engine: We propose that trait modularity functions as a “buffer” against extinction by decoupling phenotypic responses. This allows populations to navigate conflicting selective pressures, effectively facilitating evolutionary rescue in complex biotic environments. Methodological framework for empirical inference: To bridge the gap between theory and data, we provide a novel likelihood-based framework. This enables researchers to infer latent trait architectures from population genomic samplings, turning GPA from a theoretical construct into a measurable sampling variable in natural populations. We define a new roadmap for the next generation of eco-evolutionary modeling. By identifying the gaps between existing simulation engines, we provide a conceptual “blueprint” for a digital ecosystem that fully integrates complex genetic architecture with global bio-diversity dynamics.