The Role of Resources and Mortality in Life-History Evolution: A Model of Optimal Reproductive Effort and Offspring Size in the Trinidadian Guppy

Goal To improve understanding of life-history patterns in guppies by developing a model that predicts resource allocation decisions (interbrood interval, reproductive effort, and offspring size). Background The work of David Reznick and colleagues on life-history evolution in guppies (Poecilia reticulata) is viewed as strong support for age-specific demographic models for the evolution of reproductive effort (cf. Stearns 1992, p. 165). This work compares the life history of guppies from high-and low-predation sites. High-predation environments are those in which guppies co-occur with larger species of fish, such as the pike cichlid (Crenicichla alta), that frequently feed on guppies. Low-predation environments are found in the same drainages, but upstream of rapids or waterfalls that exclude the larger species of predators. At these sites, guppies co-occur with just the killifish (Rivulus hartii) that is limited to feeding on juvenile guppies (Liley and Seghers 1975). This difference in age-specific mortality risk leads to the prediction that guppies from high-predation locales should mature at an earlier age and have a higher reproductive effort than their counterparts from low-predation environment (following demographic models aligns such as Gadgil and Bossert (1970) and Law (1979)). These predictions have been well supported via comparative studies of the life-history phenotypes of wild-caught guppies (Reznick 1989; Reznick and Endler 1982; Reznick et al. 1996b) as well as through laboratory comparisons of the genetic basis of these life-history patterns (Reznick 1982; Reznick and Bryga 1996). Furthermore, replicated introduction experiments demonstrate that a change in predation environment is associated with the rapid evolution of life-history patterns in the predicted direction (Reznick et al. 1990; Reznick and Bryga 1987; Reznick et al. 1997). Nevertheless, our understanding of guppy life-history evolution is not complete. In their analysis of age-specific mortality as the mechanism of selection on guppy life histories, Reznick et al. (1996a) found an overall difference in mortality rates between Crenicichla and Rivulus locales, but no difference in size-specific mortality. Classic demographic theory predicts that in density-independent populations or in populations where density-dependent regulation affects all ages equally, a uniform change in mortality rate will not select for any change in reproductive effort or age at maturity (Gadgil and Bossert 1970; Michod 1979). If size-specific mortality is equated to age-specific mortality, then a specific form of density-dependence must be operating in order for predator-mediated selection to favor the observed changes in life history (Charlesworth 1980). Moreover, although predation is the dominant factor associated with …