Condition-dependent maternal sex allocation in horses can be demonstrated using a biologically relevant, multivariable condition measurement

Abstract

Maternal condition is influenced by multiple variables that individuals experience at seasonal and local levels, and thus condition-dependent sex allocation is likely also multifactorial. Here, we test the Trivers-Willard Hypothesis (TWH) using a multivariable approach on a dataset of thoroughbred mare breeding records. There is no sex ratio variation when examined at univariable level, mirroring the usual approach to test condition-dependent sex allocation. Conversely, the multivariate model shows multiple variables interact to influence the likelihood of producing a male. Mare and management variables that represent better body condition is associated with an increase in likelihood of a male offspring. The magnitude and direction of sex ratio skew correspond closely with predicted mare energy balance, consistent with TWH predictions. Our findings therefore support the TWH and show that, while sex allocation is multifactorial, maternal energy status emerges as the dominant driver. The additive and agonistic interactions of different variables demonstrate that a multifactorial approach should be a key consideration in sex allocation studies.