Regressions Fit for Purpose: Models of Locust Phase State Must Not Conflate Morphology With Behaviour

Abstract Locusts are defined by their capacity to transform between two very distinct integrated phenotypes or ‘phases’ in response to changes in population density: a solitarious phase, which occurs when densities are low, and a gregarious phase, which arises as a consequence of crowding and can form very large and economically damaging swarms. The two phases differ fundamentally in their behaviour, physiology and morphology. A large body of work on the mechanistic basis of behavioural phase transitions has relied on multivariate logistic regression (LR) models to estimate the probability of behavioural gregariousness from multiple behavioural variables. Martín-Blázquez and Bakkali (2017, Entomologia Experimentalis et Applicata 163, 9–25) have recently proposed standardised LR models for estimating an overall ‘gregariousness level’ from a combination of behavioural and, unusually, morphometric variables. Here I develop a detailed argument to demonstrate that the premise of such an overall ‘gregariousness level’ is fundamentally flawed. Since locust phase transformations intrinsically entail a decoupling of behaviour and morphology, phase state cannot meaningfully be conflated onto a single axis. LR models that do so are therefore of very limited value for any analysis of phase transitions. I furthermore show why behavioural predictor variables should not be adjusted by measures of body size that themselves differ between phases. I discuss the models fitted by Martín-Blázquez and Bakkali (2017) to highlight potential pitfalls in statistical methodology that must be avoided when applying LR to the analysis of behavioural phase state. Finally, I reject the idea that ‘standardised models’ provide a valid shortcut to estimating phase state across different developmental stages, strains or species. The points addressed here are pertinent to any research on transitions between complex phenotypes and behavioural syndromes.