Selecting for invasive tendencies: the evolution of morphological, physiological, and movement behaviour traits associated with dispersal

Abstract

Dispersal, defined as any movement of an individual over various spatial scales that may contribute to gene flow, is an essential component of species ecology. It provides a mechanism that allows organisms to track optimal environmental conditions, regulate population density and interactions with conspecifics, and colonise new areas. Dispersal ability varies widely among individuals, and this variation has been strongly linked to a suite of physiological, morphological, and behavioural traits that together constitute a dispersal syndrome. The evolution of dispersal-related traits can occur not only by natural selection, but also by spatial sorting, where individuals that have traits facilitating their dispersal accumulate at range edges and are limited proximally to mating with other dispersive individuals. The relationships among traits constituting the dispersal syndrome, their dynamics with age, and evolution under spatial selection on dispersal have not been comprehensively explored. Furthermore, integrating evolution into the study of dispersal is imperative to understand the mechanisms that select or constrain the evolution of dispersal-related traits. The overall aim of this thesis was to investigate a suite of physiological, morphological, and movement behaviour traits associated with the dispersal syndrome using a model system: laboratory dispersal apparatuses and Tribolium castaneum (red flour beetle). The research presented in this thesis focused on the following traits: body size, locomotor apparatus size, metabolic rate, spontaneous activity, and movement behaviour in a maze (including speed, path length, displacement distance, and behavioural intermittence). The first specific aim was to determine how the onset of sexual maturity and age throughout early life affects dispersal-related traits (Chapter 2). I found that prior to sexual maturity, T. castaneum have low metabolic rate and moved significantly less than mature ones. The low energy expenditure was attributable to reduced energy demand and inactivity, which was hypothesised to be a protective mechanism while the cuticle is undergoing sclerotisation. The second specific aim was to determine the relationships among metabolic rate, body size, relative leg length and different movement behaviour traits (Chapter 3). A dominant axis of movement ability described variation in several movement traits and was positively related to relative leg length, but unrelated to body size or metabolic rate. A mechanistic relationship between stride length and movement ability is therefore likely. The data suggested that the dispersal syndrome may be more strongly tied to morphology rather than physiology. The third specific aim was to investigate the dispersal rate of T. castaneum through three-patch dispersal apparatuses to determine which design would be most effective for artificial selection on the basis of dispersal success (Chapter 4). The distance and slope of the tubing that connected between patches significantly affected dispersal rate, therefore the apparatus design that yielded the most consistent dispersal rate was chosen for subsequent experiments. The fourth specific aim was to investigate the evolutionary changes of a suite of traits under artificial selection for and against dispersal via spatial sorting (Chapter 5). Body size responded rapidly to selection; non-dispersers increased in size while dispersers decreased in size over seven generations. Once mass differences had been accounted for, limb length and movement behaviour did not clearly diverge between dispersers and non-dispersers, but metabolic rate was slightly higher in non-dispersers. Interestingly, despite imposing strict selection, the dispersal rate of the selected lines did not strongly diverge. The association between traits and movement behaviour in a maze was different to that of the association between traits and the dispersal apparatus, suggesting that selection for dispersal by climbing does not correlate well with movement along a flat surface. Variance in dispersal rate and movement was maintained even under intensive selection, indicating that individuals may maximise their fitness by producing offspring that exhibit a variety of dispersal behaviours. The fifth aim was to determine whether a trade-off between dispersal and reproduction occurs in T. castaneum by cross-breeding the selected lines (Chapter 6). Dispersal and reproduction are both energetically expensive behaviours that may compete for resources. However, the selected dispersers, non-dispersers, and crossed lines did not differ in reproductive output, which suggests that dispersal and reproduction do not necessarily trade off in this system. Overall, the findings presented in this thesis demonstrate that many phenotypic traits are important for movement behaviour and dispersal, particularly body size, and metabolic and locomotor efficiency, providing empirical support for spatial sorting contributing to phenotypic evolution. Taken together, these results highlight the importance of experimental approaches to gain a more comprehensive understanding of dispersal ecology and evolution.