Natural Selection of Parental Ability to Vary Resource Partitioning among Offspring?

Abstract

Hassell and May (1985) suggest that our understanding of population dynamics would be advanced by integrating behavioural and population studies. An increasing number of published reports reflect the benefits of this approach. However, it may also be profitable for ecologists to review with hindsight some of their older population studies. In particular, one would point to those analyses which have relied on interpretation of population dynamics in terms of macro-ecological factors, like the effect of temperature on breeding patterns, rather than examining other aspects of the species' biology. In this note I want to illustrate a possible example, based on speculative hypothesis rather than a new model, re-interpreting seasonal patterns of egg production in terms of adaptive sexual strategies rather than tactical responses to climate, though these are not mutually exclusive influences. Marine and freshwater amphipods (Crustacea, Gammaridae) do not produce eggs of a constant size. Within species there is significant seasonal variation in the size of eggs, which are usually larger in winter. However, within samples at any season females produce eggs of a similar mean size. Clutch sizes vary significantly between populations, with season and with female size but total clutch volume for a given female size varies to a much smaller extent (Hynes 1955, Kinne 1959, Kolding and Fenchel 1981). The conclusions are that there is, unsurprisingly, an evolutionarily stable egg size but that the adaptive optimum varies with site and season. Why should optimal egg size vary? Skadsheim (1984) modified and expanded a model of Kolding and Fenchel (1981). The model assumes that relative reproductive effort (clutch volume) is fixed for a given female at any given season but that resource partitioning is subject to opposing selective forces to maximise total fecundity (i.e. clutch size) and juvenile survival (which is proportional to egg size). Survival increases with egg size and larger eggs give rise to larger offspring with better reserves and faster growth rates. Net selection maximises effective fecundity: the product of total fecundity and survival probability. Survival rates of any given size class vary seasonally as a consequence of climate and food supply. The implications of this model for optimal egg size are fully explained by Skadsheim (1984: Fig. 6). He concludes that, tactically, mothers respond to harsher weather by producing smaller clutches of larger eggs. A consequence of this reproductive trait which has not so far been considered is that there is variation in individual resource allocation between progeny produced at different times of the year. Winter offspring are effectively more expensive than summer offspring. If there were no other differences between the offspring this observation would be trivial. However, it is now evident that the mechanism of sex determination in at least some amphipod species will also produce a seasonal variation in the production of sons and daughters (Adams et al. 1987). For example, sex in Gammarus duebeni Lilljeborg is determined by environmental conditions (Bulnheim 1978). Male Gammarus carry their mates in precopula (as is widespread among Malacostraca), but are restricted in their freedom of mate choice by other males (Ward 1986) and by loading constraints (Adams and Greenwood 1987, Adams et al. 1989). Larger males have a wide choice of mates but small males may be unable to carry any females. Fitness also increases with size for females at any given season but small females can nonetheless breed successfully. Relative changes in size related fitness are shown in Fig. 1. Animals which will become large should differentiate as males and animals which will be small should be females. Field and laboratory data confirm that, under the influence of photoperiod, most immature animals early in the breeding season differentiate as males but as females later in the season. The critical day length for the shift from male to female varies with latitude. The adaptive nature of this environmental response is that, in a seasonal habitat, those animals produced earlier which grow for longer and will be larger adults in the next breeding season will also be male (Naylor et al. 1988a, b). Photoperiod is a highly predictable index of season, available growth time and future fitness expectations and environmental sex determination (ESD) is the product of a highly flexible system of sex ratio control. Although the relative genetic and environmental components vary between populations of G. duebeni the system is present at all sites sampled (Naylor 1986, H.-P. Bulnheim, pers. comm., P. J. Watt, pers. comm.). Some other species of Gammarus have also been shown to have ESD (Bulnheim 1987, Adams et al. 1987) and the genetic potential for environmental sensitivity could be widespread amongst amphipods. Since apparently non-ESD populations may nonetheless carry the trait