Evolution, 51(6), 1997, pp. 1946-19541946© 1997 The Society for the Study of Evolution. All rights reserved.OPTIMISTIC GROWTH: COMPETITION AND AN ONTOGENETIC NICHE-SHIFTSELECT FOR RAPID GROWTH IN PUMPKINSEED SUNFISH (LEPOMIS GIBBOSUS)JEFFREY D. ARENDT1 AND DAVID S. WILSON2Department of Biology, Binghamton University, Binghamton, New York 13902-60001E-mail: [email protected]: [email protected].—Intrinsic growth rate is emerging as an important life-history trait that can be modified by natural selection.One factor determining optimal intrinsic growth rates is the pattern of resource availability. Organisms that experiencechronically low resource levels tend to have slow intrinsic growth rates. However, this does not necessarily hold ifresource levels change as an organism grows. We present a theoretical model showing that rapid growth is favoredwhen resource levels for small size classes are low relative to resource levels for large size classes. We call such agrowth strategy "optimistic" because rapid growth is based on an expectation that resources will improve once aminimum size is reached. We provide empirical support for this hypothesis by examining the intrinsic growth ratesof pumpkinseed sunfish derived from three populations sympatric with bluegill sunfish (an important competitor withsmall size classes) to three populations allopatric with bluegill sunfish raised under common conditions. Rapid growthhas evolved in the sympatric fish to reach the size refuge from competition as quickly as possible.Key words.—Interspecific competition, intrinsic growth rate, Lepomis gibbosus, ontogenetic niche-shift, pumpkinseedsunfish.Received November 25, 1996. Accepted July 25, 1997.Life-history theory often postulates a trade-off betweengrowth and reproduction (Charnov 1991; Roff 1992: Stearns1992) but seldom treats intrinsic growth rate during the ju-venile stage as a parameter that evolves. Often, juveniles areassumed to grow as fast as possible (e.g., Sibly and Calow1985; Stearns and Koella 1986; Perrin and Rubin 1990;Hutchings 1993), in which case their growth rates are de-termined by properties of the environment such as temper-ature or resource abundance. In some life-history models(e.g., Kozlowski 1992) growth rate over an extended periodof time is described as a parameter that evolves, but the trade-off is actually between time allocated to growth versus re-production during this period, which again assumes that in-trinsic growth rate is maximized during the periods thatgrowth is occurring. In the few studies that have modeledgrowth rate as a parameter that evolves, growth rate–depen-dent mortality is the only trade-off considered other thanreproduction (Case 1978; Sibly et al. 1985).In contrast to the theoretical literature, empirical studiessuggest that growth rate is a parameter that evolves but israrely, if ever, maximized (Calow 1982). Differences in in-trinsic growth rate exist between closely related species(Ricklefs 1984; Werner 1986, 1994; Shine and Charnov1992) and local populations of the same species (Berven andGill 1983; Smoker 1986; Conover and Present 1990; Niewia-rowski and Roosenburg 1993; Gotthard et al. 1994). Intrinsicgrowth rate can be increased by artificial selection (Gjerde1986; Lilja and Marks 1991) and genetic engineering (Zhu1992). The list of possible trade-offs with growth rate, apartfrom reproduction, includes developmental rate (Ricklefs etal. 1994), developmental stability (Leamy and Atchley 1985),defense against pathogens (Smoker 1986; Kirpichnikov et al.1993), starvation resistance (Gotthard et al. 1994), and lon-gevity (Jonsson et al. 1992).One factor that is known to have a strong effect on theevolution of growth rates is the pattern of resource avail-ability. It is well known that plant species found under chron-ically poor resource conditions cannot grow fast even whennutrients are provided (Grime 1979; Chapin 1980). This"stress-tolerant" (Grime 1979) or "pessimistic" growthstrategy (Iwasa 1991) contrasts with plants that can respondto high nutrients with high growth rate (Iwasa's "optimistic"strategy). Although these patterns are well known in plants(Grime and Hunt 1975; Chapin 1980), they are only begin-ning to be described in animal species (Winemiller and Rose1992; Niewiarowski and Roosenburg 1992; Arendt 1997).Patterns of resource availability become more complicatedwhen organisms shift diet through ontogeny. For example,some species experience chronically poor resource conditionsas juveniles but not as adults (Werner 1988). It is likely thatnatural selection favors a high intrinsic juvenile growth ratein this situation, in contrast to the low intrinsic growth rateof species that experience poor conditions during their entirelife cycle. What we call "optimistic" growth is selected, notto exploit pulses of abundant resources in the juvenile stage,but to achieve the adult stage, in which resources are con-sistently abundant, as quickly as possible.In this paper we present a theoretical model that confirmsa pessimistic growth strategy for chronically poor resourcelevels and an optimistic growth strategy when resource levelsshift from low to high with a size-determined niche shift. Wetest the latter hypothesis by exploiting an interesting ecologicalrelationship between pumpkinseed sunfish (Lepomis gibbosus)and bluegill sunfish (L. macrochirus). As juveniles, both spe-cies are confined by predators to the littoral zone of lakes andcompete for similar resources. However, after reaching about70 mm standard length, predators are no longer an importantthreat and bluegills shift to a diet dominated by zooplanktonwhile pumpkinseed specialize on snails (Mittelbach 1984).Because zooplankton are a more abundant resource than snails,bluegill are numerically dominant and contribute many moreoffspring than pumpkinseed to the littoral zone (Mittelbachand Chesson 1987; Osenberg et al. 1988). The presence ofbluegill therefore dramatically alters the pattern of resourceOPTIMISTIC GROWTH STRATEGY IN SUNFISH 1947availability for the pumpkinseeds. By estimating age-specificgrowth rates from scales, Osenberg et al (1992) showed thatpumpkinseeds are most food limited during the juvenile stagein lakes inhabited by bluegill and most food limited during theadult stage in lakes in which bluegill are absent. We there-fore predict that juvenile pumpkinseeds that coexist with blue-gill will evolve a high