Assimilating ecological theory with empiricism: Using constrained generalized additive models to enhance survival analyses

Integrating ecological theory with empirical methods is ubiquitous in ecology using hierarchical Bayesian models. However, there has been little development focused on integration of into models for survival analysis. Survival a fundamental process, linking individual fitness population dynamics, but incorporating life history strategies to inform estimation can be challenging because mortality processes occur at multiple scales. We develop an approach analysis, model constraints based species' strategy functional analytical tools. Specifically, we structurally separate intrinsic patterns that arise from age-specific (e.g. increasing during early stages due growth or maturation, versus senescence) extrinsic over different periods time seasonal temporal shifts). use shape constrained generalized additive (CGAMs) obtain hazard functions incorporate theoretical information classical survivorship curves the age component and capture factors component. compare performance our modelling standard tools do not explicitly species structure, metrics predictive power, accuracy, efficiency computation time. applied these two case studies reflect shapes underlying curves, examining age-period white-tailed deer Odocoileus virginianus Wisconsin, USA Columbian sharp-tailed grouse Tympanuchus phasianellus columbianus Colorado, USA. found included effects CGAMs outperformed did include explicit constraints. demonstrate data-driven easily extendable analysis by showing its utility rates probabilities, accounting heterogeneity across ages time, very species. show how models, statistical methods, enhances analyses.