Reproductive adaptations to reduce locomotor costs in viviparous fish (Poeciliidae)

Viviparity, a live-bearing mode of reproduction, has evolved over 100 times independently in vertebrate animals. Despite its frequent evolution, viviparity has a number of hypothesised disadvantages compared to the ancestral mode of reproduction, oviparity (egg-laying). One of these disadvantages is a reduction in locomotor performance during pregnancy, the period of internal development of the embryos. Adaptations to a live-bearing reproductive mode could have evolved to reduce these locomotor costs. In this thesis, I aim to find whether matrotrophy, post-fertilization nutrient provisioning (e.g. through a placental structure), and superfetation, the presence of multiple broods of different developmental stages, reduce the locomotor performance decline during pregnancy in the Poeciliidae, live-bearing fishes. In <strong>Chapter 2</strong>, we review the literature on the effects of pregnancy on morphology, performance and fitness. The biomechanics of each mode of locomotion (walking, swimming or flying) are distinct, and are affected differently by the added mass and volume of pregnancy. Furthermore, we list the possible adaptations that have evolved to reduce the locomotor costs of pregnancy, and divide them into three different categories: adaptations that reduce the locomotor costs of live-bearing, adaptations with which the locomotor costs of live-bearing are avoided, and adaptations to the life history of the animal. Lastly, we discuss hiatuses in the literature and experimental procedures to quantify the hypothesised benefit of adaptations. In <strong>Chapter 3</strong>, we compare the morphological changes during pregnancy in two closely-related species of live-bearing fish: <em>Poeciliopsis turneri</em> and <em>Poeciliopsis gracilis</em>. These species mainly differ in their mode of nutrient provisioning: <em>P. gracilis</em> is lecithotrophic and <em>P. turneri</em> is an extensive matrotroph. We tracked the morphological changes in 3D using a non-invasive method that creates three-dimensional body models. We find that <em>P. turneri</em> is more slender during the early stages of pregnancy, but increase in size more rapidly. This is in line with the locomotor costs hypothesis, which predicts that matrotrophic fish are more slender during the early stages of pregnancy, but that the difference between the body shapes of lecithotrophic and matrotrophic fish diminishes as pregnancy progresses. Our results indicate that matrotrophy could indeed provide a morphological advantage during pregnancy. Fast-start performance, a manoeuvre fish deploy to escape predatory strikes, is important for individual survival. In <strong>Chapter 4</strong>, we use state-of-the-art biomechanical methods to, for the first time, quantify this manoeuvre in three-dimensional space in adult fish (<em>Heterandria formosa</em>). We show that fish can orient their escapes in up- and downwards direction, and that this is correlated with a change in pitch angle of the body. Changes in roll angle of the body were not correlated with orientation of the fish. We furthermore demonstrate that stage 1 of the fast start, often described as a preparatory stage, can already contribute to propulsion. The results from <strong>Chapter 4</strong> indicate that three-dimensional measurements of fast-start manoeuvres provide novel insights that were often overlooked. Measuring fast starts in three-dimensional space is relevant in determining the adverse effects of pregnancy on locomotor performance. We did this by comparing three species of live-bearing fish: <em>P. turneri</em>, <em>H. formosa</em> and <em>Phalloptychus januarius</em>. In <strong>Chapter 5</strong>, we show that pregnancy-induced changes in abdominal width are correlated with a reduction in performance in the horizontal plane (maximal horizontal speed, change in yaw angle), but less so in the vertical plane (maximal vertical speed, change in pitch angle). Furthermore, we demonstrate that an increase in abdominal width is correlated with a decrease in abdominal curvature and, for some species, in a decrease in maximal curvature rate in the abdomen. Lastly, we show that the pregnancy-induced morphological changes depend on the level of superfetation: species with a high level of superfetation experience higher frequency, but smaller amplitude changes in the shape of the abdomen. Whether superfetation actually results in a more slender body shape, as predicted by the locomotor costs hypothesis, depends on the level of reproductive investment. In this thesis, I show that pregnancy induces changes in morphology which comes with a cost in fast-start performance. Both matrotrophy and superfetation affect how body shape changes due to pregnancy, but whether the latter provides beneficial changes depends on the level of reproductive investment. Furthermore, I reveal that fast starts can have a substantial three-dimensional component which is relevant both to biomechanicists that aim to understand the physical and physiological mechanisms underlying this manoeuvre and to evolutionary biologists that strive to answer performance-related questions.