Genetics of Brassica-campestris 1. Genetic constraints on evolution of life-history characters

-Energy allocation arguments suggest a possible tradeoff between timing and magnitude of reproduction: plants that postpone reproduction may accumulate greater resources and consequently produce more offspring. However, early reproduction may be favored when adult mortality is high. Tradeoffs among life-history characters may be a consequence of constraints imposed by genetic and environmental covariation among traits. In this paper we examine the genetic basis of the relationship between timing and magnitude of reproduction in an annual plant, Brassica campestris, by selecting to change flowering date and plant size in each of four directions (early and large, late and large, early and small, or late and small). There is a strong positive relationship between flowering date and flowering height. The response to selection was greatest along the axis of positive genetic covariation. Populations may evolve to become early flowering and small or late flowering and tall, but there is little response for the alternative combinations of characters. In this instance, the constraints imposed by quantitative genetics are in striking accord with predictions that might be made on physiological, energetic, or ecological grounds. Received October 3, 1989. Accepted July 28, 1990. Models of life-history evolution predict the optimal "strategies" or combinations of traits that should evolve in particular situations (Steams 1976, 1977; Boyce, 1984). Many models assume that alternative combinations of life-history characters result from differential allocation of resources (Lack, 1954; Cody, 1966; Williams, 1966) and that these trait combinations will maximize fitness, as natural selection operates on components of fitness (Falconer, 1981; Istock, 1983; Lande and Arnold, 1983; Mitchell-Olds and Rutledge, 1986). Patterns of allocation to growth and reproduction have received considerable attention in recent years. Energy allocation to early growth may affect adult size, and consequently survival and reproduction, because size is often positively correlated with survival and fecundity (Harper and White, 1974; Cook, 1980; Solbrig et al., 1980; Solbrig, 1981; Reinartz, 1984; Antlfinger et al., 1985). Age of first reproduction, or the timing of change from vegetative to reproductive growth, may also be correlated with fitness (Cohen, 197 1; Schaffer, 1977; King andRoughgarden, 1982, 1983; Roach, 1986; Mazer, 1987). The fitness effects of timing of reproduction may be a consequence of factors such as plant size (Law, 1979) or predator avoidance (Evans et al., 1989), which directly affect fitness. Fitness in plants with overlapping generations and continuous growth patterns is enhanced by early reproduction (Cole, 1954; Lewontin, 1965). In annual plants, which are limited to a single generation per year, the relationship between development time and fitness is more complicated. Early reproduction may be favored if adults suffer mortality due to predation or disturbance. Alternatively, deferral of reproduction until later in the life cycle may permit continuing procurement of resources, leading to greater total reproductive success (Cohen, 1966; Abrahamson and Gadgil, 1973; Gaines et al., 1974; Schaffer, 1974; Solbrig and Simpson, 1974; Wiley, 1974; Hastings and Caswell, 1979; Pitelka et al., 1980; Schaal and Leverich, 1981). Energy allocation arguments suggest a possible tradeoff between timing and magnitude of reproduction in annual plants (King and Roughgarden, 1983). The fitness consequences of this will depend on the ecological characteristics of the home environment. Tradeoffs among life-history characters or other components of fitness may occur at several levels, corresponding to different factors that influence character covariation. We can measure an individual's phenotype for several characters, obtaining the phenotypic correlation between traits, which may be further partitioned into covariation