Lamprologus callipterus, a cichlid fish endemic to Lake Tanganyika, exhibits the most extreme sexual size dimorphism among animals, with males being bigger than females. Males are on average 13 times heavier than females in the population observed in this study and this ratio of may exceed 1:27 in particular cases. This size dimorphism corresponds with extreme levels of polygyny, with harem sizes reaching up to 30 females per male. The biggest males defend nests consisting of clumps of empty gastropod shells, which they collect from the vicinity of their nests, or steal from other nests. During the time of defending a territory, territorial males do not feed. Their body condition therefore decreases with increasing time of territory maintenance until they are outcompeted by heavier or larger males. Females ready to spawn enter these nests, lay their eggs inside a shell and perform broodcare therein for 10 to 14 days. During that time period, females usually stay permanently inside their shells and do not feed.

Usually, sexual selection is regarded to be responsible for sexual size dimorphisms in animals. Sexual selection can result either from mate choice (= intersexual selection) or from direct competition between individuals of the same sex (= intrasexual selection). In L. callipterus, additionally natural selection mechanisms may be responsible for the evolutionary origin and stability of the extreme sexual size dimorphism, because of the peculiar breeding substrate they use. In a study I conducted, I examined the influence of both, sexual and natural selection mechanisms on the evolution of large males and small females. Additionally, I used a life history model of sexual size dimorphism to test whether intrasexual selection mechanisms via male-male competition and natural selection mechanisms via fecundity advantages of large females could explain the extreme sexual size dimorphism in L. callipterus.

To test if intrasexual selection would influence male size I performed lab experiments with direct male-male competition for territories and for access to females. Larger males were more aggressive than their smaller companions and they could hold their territories for longer time periods. Therefore, larger males could spawn with more females and sired more offspring. In the field there also was an advantage of large male size due to superior competitive abilities. Large males were able to defend nests with more and larger snail shells and therefore, had a higher reproductive success (number of breeding females) than smaller males.

To test the influence of intersexual selection on male body size I performed female choice experiments in the lab, and measured the numbers of breeding females of males of different quality in the field. Neither in the lab nor in the field females were found to choose particular males, despite a great variation of available males in size and body condition. I concluded that intersexual selection seems to be only of minor importance for the evolution of sexual size dimorphism in L. callipterus, while intrasexual selection mechanisms clearly favour bigger males. I discuss that it is unlikely, however, that this mechanism is alone responsible for the extreme sexual size dimorphism in L. callipterus.

Natural selection mechanisms are usually limiting the evolution of extreme body sizes in one or the other of the two sexes. In contrast in L. callipterus, natural selection may operate in the same direction as sexual selection, i. e. increasing the size dimorphism of the sexes, but in opposite directions in males and females. The clue to understanding the extreme sexual size dimorphism in L. callipterus is probably the particular breeding substrate. Males construct nests of empty snail shells, that is they must be able to manipulate and carry these shells. Females enter the nests and breed in these shells, that is they must be small enough to fit into these shells.

I tested whether males must pass a minimum body size to be able to carry shells efficiently. Experiments in which male sizes and the sizes of shells were varied revealed that male minimum size is indeed limited by their ability to collect empty snail shells. Only males ³ 9.0 cm standard length were able to carry Neothauma shells at all, and larger males were able to carry the shells more efficiently than smaller males.

In lab-experiments I tested whether female size is limited to an upper threshold size by the physical properties of the breeding substrate. I found that females in smaller shells (data were corrected for female body size) sired less offspring than females that were breeding in larger shells. The largest shells were preferred by the largest females, whereas smaller females were spawning in shells that matched their own body size. For further insight in the evolutionary origin of the use of shells as a breeding substrate in L. callipterus, I examined whether snail shells were used also in other, non-reproductive contexts. If shells were also used as shelters, it is likely that the ancestors of L. callipterus were small shell breeding cichlids like today’s females, using shells primarily for hiding. If snail shells were used by L. callipterus only for reproduction it is probable that the ancestors of L. callipterus were larger than females are today and that natural selection worked in opposite directions on males and female sizes. In the lab and in the field I found that females were only breeding in snail shells and never on other structures. In experiments in which I tested the significance of snail shells for reproduction and for hiding from predators I found that females use shells primarily for breeding and only very rarely for shelter.

I further examined whether a general life history model of sexual size dimorphism in fish, developed by Parker (1992) may explain the extreme size dimorphism in L. callipterus. In this model, it is assumed that size is tuned by a trade-off of mortality against (i) intrasexual competition in males, and (ii) fecundity in females. If body sizes predicted from this model fit the body sizes of males and females observed in the field, natural selection might be responsible for female size (fecundity advantages of large females), and intrasexual selection might be responsible for male body size (competition advantages of large males). Under different theoretical approaches, predicted male switch size to maturity always exceeded female size, although not to the extent found in the field. The predicted female size derived from this model matched the body size found L. callipterus in the field quite well. However, the predicted male size was smaller than the average body size of territorial males found in the field. The difference between the predicted switch sizes to maturity in males and the observed size of territorial males were discussed. Sperm competition had an additional effect on optimal body size of males and therefore I concluded that strong male-male competition (intrasexual selection) may be one factor determining large male size and the size differences between the sexes, in addition to the limitations imposed on males by the physical properties of shells.

In a gain rate maximization model for females I assumed that females are selected to maximize their reproductive gain rates (number of offspring per unit time). In this model no mortality costs for large females were considered, but the predicted female body size was still within the range observed in the field. This indicates that the growth curve and the daily growth rates, which I used in the first model to calculate optimal switch times to maturity, were tuned by the disadvantage sustained by big females in finding a suitable breeding substrate. I attributed the small female size to the effects of natural selection acting to maximize number of offspring per unit time. Natural selection additionally restricts female size to a maximum limit, and is another factor determining size differences between the sexes.

I concluded that the divergent influence of natural selection mechanisms on male and female body size in combination with size dependent mechanisms of intrasexual competition among males, may explain the extreme sexual size dimorphism found in this species. The size dimorphism between the sexes is also stabilized by the properties of the peculiar breeding substrate. Snail shells impose different constraints on the body sizes of males and females. Females must be small enough to fit into shells and males must be large enough to be able to compete for and transport shells efficiently. I argue that natural selection should not be underrated as a potential mechanism for the origin and maintenance of sexual size dimorphisms in animals. □