An information-geometric framework reveals size-dominant and context-dependent stomatal perturbations in cacao under pathogen challenge

Abstract

Comparing stomatal responses across heterogeneous settings is difficult because host, environment, and pathogen effects are often entangled. Here, we developed a control-anchored, information-geometric framework to quantify pathogen-associated perturbations using 7558 cacao (Theobroma cacao) stomata across two genotypes, multiple isolates, two photoperiods, and three time points. By anchoring each infected condition to its strictly matched trial control, we partitioned seven morphometric traits into size and shape modules. Responses were quantified using module-response norms, Gaussian 2-Wasserstein divergence, vector alignment to light/time reference axes derived from matched controls, and leave-one-trial-out robustness procedures. Across the state space, size-related responses exhibited a broader dynamic range than shape-related responses, indicating that pathogen-induced perturbations shift stomatal scale more than geometric balance. The module-response landscape revealed strong genotype- and photoperiod-specific organization, with SCA6 conditions showing larger size-axis displacements than corresponding Pound7 conditions. Alignment analysis demonstrated that pathogen-response vectors do not converge on a universal control-derived light/time direction; instead, their orientations are highly context-dependent. Replicate-aware analysis confirmed that these response patterns are reproducible, although stability varies across contexts. Overall, these results map a structured but context-dependent landscape of stomatal responses and provide a practical framework for prioritizing plant–microbe interactions for future physiological and molecular investigation.