Virtual fish for real science - Investigating the role of public information content for mate-choice copying in the sailfin molly (Poecilia latipinna) using live fish and computer-animated fish stimuli

The strategy of mate-choice copying has fascinated scientist for decades but still, little is known about its underlying mechanism. Investigating what aspects of public information are relevant for triggering copying behavior and how these might be intertwined is, therefore, a central objective in current research. During my PhD project, I investigated the use of public information for mate-choice copying in the livebearing sailfin molly Poecilia latipinna. Both males and females were shown to copy the mate choice of conspecifics which makes the sailfin molly the ideal model species. Aside from using a classic experimental approach with live fish stimuli to study copying, I also used a novel methodological approach, that is, computer animation. For this, my PhD project also involved the development of “FishSim Animation Toolchain” within the interdisciplinary “Virtual Fish Project” at University of Siegen. FishSim is a software framework for the creation, animation and presentation of virtual sailfin mollies during experiments. A positive validation of the new method showed that focal females were equally attracted to virtual males as to live males or those presented as videoplayback. In a series of copying experiments with live and virtual stimulus fish, I manipulated visual public information content provided by either a model female and/or her sexual interaction with a male during the observation period.
A model female’s body size is assumed to indicate her quality as a model with respect to her age and, consequently, her experience in mate choice. With larger models being older and, hence, more experienced in mate choice. In a comparative study using live and virtual fish stimuli, I found that the size of a model female did not affect copying behavior in observing female sailfin mollies. They copied the choice of a model female (live and virtual) for a prior non-preferred male irrespective of whether the model female was larger or smaller than the focal female. These results were in contrast to previous studies that demonstrated size-dependent copying behavior in the closely related guppy. I assume that possible rules underlying mate-choice copying may be species-specific and vary between different socio-environmental conditions.
In another study, I assumed that the gravid spot, which is a common feature of female poeciliids, might provide information about a model female’s quality with respect to her reproductive status. However, I could show that neither the presence nor absence of a visible gravid spot in virtual model females affected copying behavior in focal females. They copied the choice of a virtual model female in both conditions, indicating that the gravid spot does not provide a reliable source of information for observing females.
At least not in sailfin mollies.
To investigate whether the behavioral interaction of the model female and the male affects copying behavior in observing focal females, I manipulated courtship behavior of the interacting pair. I used FishSim to either present courtship behavior that was mutual, only female driven or only male driven. I could show that focal females copied the choice of a virtual model female in all three conditions. The strength of copying, however, was affected by differences in courtship behavior. Copying behavior was strongest when only the male was actively engaged in courtship and weakest when both male and female were actively engaged and also copulated. When no model female was present, a lack of male courtship behavior resulted in focal females losing interest in a prior preferred male. My results indicate that, albeit providing valuable information for the assessment of a male, a sexually active model female may nevertheless be perceived as a competitor which overall weakens copying behavior. This study, demonstrates a first attempt for a systematic manipulation of behavior to investigate its possible effect on mate-choice copying, which has not been done before in such a controlled way.
In a final experiment using live fish only, I manipulated the behavioral interaction of the model female and male by increasing the distance between them. I could show that focal females showed copying behavior when model and male were at close proximity (1 cm) to each other but also when they were at apart from each other (40 cm) and only able to communicate over distance. These results were in contrast to my prediction that increasing the distance between the pair would impede courtship or any sexual interaction, which is described to naturally occur at close distance. This study implies an even wider transfer of public information for mate choice than previously thought.
In summary, my findings support previous knowledge on the importance of the model female and further highlight the importance of including aspects about the behavioral interaction of model female and male in future research on mate-choice copying. Further, I was able to demonstrate a highly controlled and standardized approach for future studies on mate-choice copying using FishSim, which also has the ability to increase animal welfare in research.