Skip to main content

The consequences of placentation on the swimming performance of pregnant livebearing fish

Final Report Summary - SWIM (The consequences of placentation on the swimming performance of pregnant livebearing fish)

Livebearing is a reproductive strategy that confers both costs and benefits to females. One important cost is a reduction in locomotory performance of the mother during her pregnancy. This can detrimentally influence her feeding efficacy and her ability to avoid predation. Some livebearing organisms have evolved reproductive life-history adaptations that help minimize these locomotory costs. For example, in livebearing snakes the transition to an arboreal life style is accompanied by a repositioning of the paired ovaries along the length of the body, to reduce overlap; this reduces bodily distension and enhances climbing ability of the females during gestation.
I recently proposed that the placenta is a life-history adaptation that evolved in livebearing fishes to minimize adverse effects on locomotory performance during gestation (the Locomotor Performance hypothesis). The placenta is a complex, temporary organ that regulates embryo nourishment and waste product removal. It evolved many times throughout the animal kingdom, which suggests that it confers an adaptive advantage to organisms. I propose to test this hypothesis by comparing the consequences of pregnancy for locomotory performance between closely related livebearing species with and without a placenta.
I adopt a comparative multi-disciplinary approach that is structured around: (i) detailed knowledge of the species’ life-histories, (ii) an objective criterion for the level of placentation (the matrotrophy index), and (iii) a highly resolved phylogeny. The fish family Poeciliidae is used as a model system, because: (a) the placenta independently evolved several times within the family, (b) this family contains closely related living species with remarkable variation in the degree of placentation (ranging from species with no placenta to species with highly developed placentas), and (c) these fish have an excellent prior history as subjects in laboratory studies.
In this project new state-of-the-art methodologies were developed to study the 3D morphology and swimming performance using a cross-disciplinary approach in which we combine expertise from the fields of bio/fluid mechanics with evolutionary ecology. To date, the swimming performance of fish in an ecological and evolutionary context is usually studied in a simplified two-dimensional plane by placing fish in a very shallow swimming arena where swimming in an up- or downward direction is limited. The swimming performance in this forced 2D plane is then filmed from above using a single camera. Fish swimming studied in this way is likely to yield unrealistic estimates of performance, because fish typically live in a three-dimensional environment. Far more advanced methodologies are available in the high-tech field of biomechanics, however, true cross-disciplinary studies combining Evolutionary Ecology and Biomechanics are still largely lacking. In this project I took advantage of the state-of-the-art knowledge of the biomechanics of fish swimming available in the Experimental Zoology group at Wageningen University (collaboration with C.J. Voesenek, R.P.M Pieters, M. Fleuren, E.M. Quicazan Rubio, and J.L. van Leeuwen) to build a state-of-the-art swim arena where the fast startle response of females is filmed by means of three orthogonally positioned high-speed video cameras. These movies are analysed using novel in-house automated tracking software only available in our group. This ‘FishTracker’ software requires a ‘3D body model’, constructed with the Body model generator. To map the swimming kinematics and the body wave, the midlines of the female bodies are digitized using automatic tracking software or, for low contrast images, by manually digitizing the central axis of the fish silhouette using in house MatLab software. Three-dimensional analyses yield important information on the yaw, pitch and roll movements of fish that cannot be inferred from 2D views. To date, 3D swim analyses have not been performed within an ecological or evolutionary context, simply because they are too time consuming to allow the analyses of a sufficient number of individuals to obtain statistically reliable data. The FishTracker software, however, allows very rapid large-scale computerized analyses of multiply-synchronized fast-start videos, enabling the precise tracking of the three-dimensional movements of a sufficiently large number of fish. Kinematic parameters are automatically obtained from the interpolated midlines and mass distribution along the fish: instantaneous and mean swimming speed, and acceleration related to the centre of mass, tail-beat amplitude and tail-beat frequency. The swimming analyses made with the new experimental set up demonstrate that livebearing fish exploit their full 3D environment to make escape manoeuvres, confirming our expectations and the relevance of our major investment in developing this new technology.
In addition, we are in the process of designing a new state-of-the-art swim tunnel to study sustained swimming performance. In this setup we correct for the position the fish take in the flow tunnel by combining continuous video analyses (from both side and top views) and detailed information of the flow velocity profile in the flow tunnel as inferred from (a) Computational Fluid Dynamic computations and (b) experimental verification by means of Digital Particle Image Velocimetry. This enables us to make fine adjustments of the swimming speed based on the fish’s swimming position and the overall flow rate through the tunnel. This is crucial because cross-sectional velocity profiles are never uniform, amongst others due to ‘wall effects’. No standard solid blocking correction is applied, since the proposed fish species occupy <5% of the cross-sectional area of the swimming chamber. These techniques are common practice in the field of fluid mechanics, but again have never been applied to large-scale ecological or evolutionary studies.
The project has yielded one Nature paper1 and continuous to generate a large amount of high-quality data that will lead to a number of publications. To date several experiments are finished yielding a large data set with 3D swimming events that is currently being analyzed, leading to at least two publications (manuscripts in preparation). Other experiments are still in progress and we expect that these will lead to an additional 6 publications. Although parts of the work as outlined in Annex I is delayed due to the high technical challenges that had to solved, it continues past the project end date of 31 December of 2014.
The project has also generated a lot of spin-off. Apart from Dr Pollux (who continues on the project), we have now 2 PhD students working on the project. In addition, 6 MSc students and 6 BSc students have or are working on the project, and more will follow. The project also fostered the internal collaboration within the Experimental Zoology Group: a new spin-off project was started in collaboration with Dr Lankheet on visuo-motor control in livebearing fish.
Further information can be found on project’s web site: Contac details: PI: and Coordinator:

1Pollux BJA, Meredith RW, Springer MS & Reznick DN (2014) The evolution of the placenta drives a shift in sexual selection in livebearing fish. Nature 513, 233-236. doi:10.1038/nature13451.