Naturally and sexually selected traits in haplochromine cichlid fishes

Humankind seeks for explanations to describe the evolution of the astonishing biodiversity surrounding us. To understand organismal diversity, we first need to understand the evolutionary processes underlying it. However, we are still struggling with Darwin’s ‘mystery of mysteries’ (Darwin 1859), that is to understand how and why new species form (Coyne & Orr 2004). The establishment of reproductive isolation between divergent populations can evolve through barriers in post- (Snook et al. 2009) and pre-copulatory sexual selection (Darwin 1871). The two fundamental modes of Darwinian sexual selection are contests for mates (intrasexual selection) and mate choice by the opposite sex (intersexual selection) (Darwin 1871). Even though reproductive isolation could arise through sexual selection alone, it was hypothesized that it functions most effectively in conjunction with selection for species recognition or ecological selection (Ritchie 2007). Speciation through ecological selection drives adaptive diversification into a variety of ecological niches, which is described as ‘adaptive radiation’ in evolutionary groups that have exhibited exceptional extent of diversification (Schluter 2000). A textbook example of adaptive radiations and, therefore, an ideal system to study diversification are the perciform fishes of the family Cichlidae (e.g. Maan et al. 2006; Seehausen et al. 2008; Salzburger 2009). Their rapid speciation resulted in an estimated number of around 3’000 species (Snoeks 1994; Turner et al. 2001), turning cichlids into the most species-rich family of vertebrates (Salzburger & Meyer 2004; Salzburger 2009). Cichlids are distributed across South and Central America, Africa and parts of India. This distribution suggests a Gondwanian origin of the group (Salzburger 2009). Their centre of diversity, however, lies in the East African Great Lakes, which harbour extremely diverse and species-rich flocks of cichlid fishes and are therefore a prime model system in evolutionary biology (Meyer 1993; Turner et al. 2001; Seehausen 2006). In addition to the extrinsic environmental factors such as geologic and climatic events creating novel ecological niches (Fryer & Iles 1972; Sturmbauer 1998; Sturmbauer et al. 2001), several evolutionary key innovations have been hypothesized to have played a role in their rapid speciation and adaptations to a variety of ecological niches. Of particular importance are the special pharyngeal jaw apparatus (Fryer & Iles 1972; Liem 1973), the highly complex reproductive behaviour (Fryer & Iles 1972; Goodwin et al. 1998; Kornfield & Smith 2000) and the wealth of colour morphs. It was shown that colour and pigmentation patterns seem to play a central role in the explosively radiating cichlid fish lineages in the East African Great Lakes in general, and in haplochromine cichlids in particular (Seehausen et al. 1999; Kocher 2004; Turner 2007; Salzburger 2009). Haplochromines comprise about 80% of East African cichlid species including the entire species flocks of lakes Victoria and Malawi, the tribe Tropheini from Lake Tanganyika and many riverine species (e.g. Turner et al. 2001; Salzburger et al. 2005). Interestingly, all haplochromines are maternal mouthbrooders, with females incubating their offspring – until fully developed – in their buccal cavities (e.g. Fryer & Iles 1972; Salzburger et al. 2005). This special breeding behaviour evolved several times during cichlid evolution (Goodwin et al. 1998), but only the ‘modern haplochromines’ show a derived polygynous or polygynandrous maternal mouthbrooding system with males displaying the so-called egg-spots on their anal fins (Fryer & Iles 1972; Salzburger et al. 2005, 2007). These ovoid markings consist of a transparent outer ring encircling a brightly coloured yellow, orange or reddish centre (Wickler 1962; Fryer & Iles 1972). The conspicuous central area is formed by two chromatophore cell types, xanthophores and iridophores (Salzburger et al. 2007; Santos et al. 2014). Even though this trait is proposed to be a putative key innovation mediating the evolutionary success of haplochromines (Salzburger et al. 2005; Salzburger 2009), their function is not fully understood. Several hypotheses exist that seek to explain the function of egg-spots: Wickler (1962) associated the function of egg-spots with the special mouthbrooding behaviour, and suggested that egg-spots mimic real eggs and function as an attracting signal during courtship and as releasers for egg-uptake and, hence, to maximize fertilization. Support for Wickler’s hypothesis was only found with respect to the function in courtship since females of the species Astatotilapia elegans and Pseudotropheus (Maylandia) aurora preferred to lay batches with males with many egg-spots (Hert 1989, 1991), whereas females of Pseudotropheus (Maylandia) lombardoi preferably chose males with an artificially enlarged egg-spot over males with one natural or many egg-spots (Couldridge 2002). However, there was no influence of egg-spots on fertilization rate (Hert 1989). Further doubts about the egg mimicry hypothesis arose because egg-spots often do not resemble size, shape and colour of a species’ actual eggs (Jackson & van Lier Ribbink 1975; Goldschmidt 1991). This mismatch between real eggs and egg-spots may be due to a trade-off between attractiveness towards females and conspicuousness for predators (Goldschmidt 1991). An alternative explanation could be that egg-spots serve as species recognition signal (Axelrod & Burgess 1973). So far, the results from studies that aimed to evaluate the function and selection pressures on egg-spots are scarce, rather inconsistent and raise the necessity for new experimental work on their mode of action and their evolutionary origin. Part 1 of my thesis is therefore dedicated to the evolution and function of egg-spots. The first manuscript focuses on the evolutionary origin of anal fin egg-spots, more specifically, we tested the hypothesis whether a sensory bias has triggered the evolution of egg-spots in cichlid fishes (1.1). Mate choice trials were conducted to see if females of the basal haplochromine Pseudocrenilabrus multicolor (naturally showing no true egg-spot on its anal fin) prefer computer-animated photographs of males with an artificially added egg-spot. Additionally, colour preferences (outside a mating context) were tested in a phylogenetically representative set of East African cichlids. The next two chapters focus on the putative function of egg-spots in sexual selection in the two haplochromine species Astatotilapia burtoni (1.2 The function of anal fin egg-spots in the cichlid fish Astatotilapia burtoni) and Astatotilapia calliptera (1.3 Egg-spot pattern and body size asymmetries influence male aggression in haplochromine cichlid fishes), which both exhibit several egg-spots on their anal fin. In both species, mate choice trials were conducted to test if females prefer to lay eggs with males with many egg-spots over males with fewer or no egg-spots. Since carotenoid based colouration can be indicative for the health and strength of its bearer (e.g. Endler 1978, 1980; Hill 1992), egg-spots are also a prime example to examine if there is a function in intrasexual selection. Therefore, male aggression experiments were conducted in both species to test if egg-spots could play a role in the assessment of an opponent’s strength. Visual signals will most probably not only diverge due to sexual selection, but might be influenced by their environment and are therefore expected to evolve to a point where viability costs balance mating advantage (Darwin 1871; Zahavi 1975; Endler 1978; Andersson 1994). To examine how the egg-spot phenotype can be influenced by sexual and ecological selection, the next manuscript examines the variation of anal fin egg-spots along an environmental gradient in a haplochromine cichlid fish (2.1). This project constitutes the first of two studies of Part 2 describing adaptive divergence in lake-stream systems in A. burtoni. This species represents an ideal model organism to address questions about adaptive divergence in lake-stream systems in cichlids, since A. burtoni is one of only few cichlid species, which inhabits shallow zones of one of the East African Great Lakes as well as rivers and streams surrounding it (Fernald & Hirata 1977; Kullander & Roberts 2011). Populations of lacustrine and riverine habitats of four lake-stream systems were examined with regards to sex- and habitat-specific differences in egg-spot characteristics such as number, size and colouration. Finally, we tested for an association between the conspicuousness of male egg-spots and underwater light environment as well as the status of the immune system. However, not only visual signals - like egg-spots - can adapt to the respective environmental conditions, but lake-stream systems are also a unique system to study how populations experiencing different environmental conditions may diverge in general. So far, adaptive divergence in cichlids has mainly been investigated within lakes, e.g. along depth or habitat gradients (see e.g. Barluenga et al. 2006; Seehausen et al. 2008). The A. burtoni setting should therefore be established as the first lake-stream system in cichlids, which is described in the second study of Part 2 (2.2 Adaptive divergence between lake and stream populations of an East African cichlid fish). Here, we first established phylogeographic relationships and assessed the population structure as well as body shape differences in over 20 A. burtoni populations from the southern part of Lake Tanganyika. In a second step, we focused on four lake-stream systems in detail (the same systems as in chapter 2.1) and, in addition to the body shape and population-genetic surveys, we quantified other ecologically relevant traits (gill raker and lower pharyngeal jaw) as well as stomach contents. To test whether the shifts in the examined traits reflect ecologically based adaptive divergence (Berner et al. 2009; Harrod et al. 2010), we tested for an association between morphological variation and environmental factors, such as resource use and water velocity. Finally, a mating experiment was conducted to test for reproductive isolation among lake and stream populations. Adults and offspring from this common garden setting were further used to evaluate levels of phenotypic plasticity in the traits body shape and gill raker morphology. During the sampling trips for the study mentioned above, we observed a clear-cut barrier for the occurrence of A. burtoni in the streams surrounding Lake Tanganyika. At a certain elevation A. burtoni was absent and seemed to be replaced by another riverine cichlid, namely a species of the Pseudocrenilabrus philander complex. Interestingly, they both were found to co-occur in Lake Chila, a small lake 20 km south of Lake Tanganyika. The first side project of Part 3 concentrates on this P. philander complex with the manuscript about the phylogeographic and phenotypic assessment of a basal haplochromine cichlid fish from Lake Chila, Zambia (3.1). Here we report the discovery of a population of the normally riverine P. cf. philander in Lake Chila. We examined this lake population for increased morphological variation compared to riverine populations of P. cf. philander. With this dataset we wanted to test whether ecological opportunity in the form of a greater number and more diverse ecological niches promotes diversification in lakes compared to rivers (as seen in e.g. Stelkens & Seehausen 2009). The phenotypic variability of this Lake Chila population was evaluated in relation to other lacustrine and riverine populations by quantifying colouration and body shape. Additionally, phylogeographic history was investigated with attention to a case of hybridization of two distinct lineages. The second side project focuses on a special case of morphological variation, namely mouth asymmetry, by performing a field based assessment of attack strategy and feeding success in the scale- eating cichlid fish Perissodus microlepis (3.2 A fitness benefit for mouth dimorphism in a scale-eating cichlid fish). Perissodus microlepis is the most common and perhaps the most specialized lepidophagous cichlid in Lake Tanganyika (Takahashi et al. 2007) and exhibits a pronounced asymmetry with individuals that feature a mouth slightly bent to the right or to the left side in order to optimize feeding successes (Hori 1993). In this study the lateralisation dynamics in P. microlepis were reassessed in a semi-natural environment in order to confirm laboratory based findings about asymmetrical attack strategies and to test if dimorphic experimental populations of P. microlepis ultimately are more successful and show a higher feeding success than monomorphic experimental populations. All together, we aimed to disentangle causalities in the evolution of this system and to demonstrate the selective advantage of dimorphic mouth opening and attack strategy in scale-eaters. This is necessary to explain how such asymmetries have evolved and can be maintained in natural populations. In summary, my thesis consists of two main parts and a third part comprising two side projects. Part 1 investigates the trait egg-spots, which were mentioned to be a key innovation of haplochromines, the most species-rich tribe of cichlids. Three manuscripts deal with their mode of action as well as their evolutionary origin. Part 2 examines the divergence among lake and stream populations with respect to egg-spots and in a second project with respect to body shape and other ecologically relevant traits. Additionally, the phylogeographic relationships of A. burtoni populations from the southern part of Lake Tanganyika were established