Latitudinal populations of the Atlantic silverside, Menidia menidia, show substantial genetic variation in rates of energy acquistion and allocation. Reared in common environments, silversides from northern latitudes consume more food, grow faster and more efficiently, store more energy, and produce greater quantities of eggs than their southern conspecifics. The persistence of seemingly inferior southern genotypes in the face of ostensibly superior northern genotypes suggest that there are hidden evolutionary trade-offs associated with these elevated acquisition and allocation rates. We tested the hypothesis that rapid growth and high levels of food consumption trade-off against locomotory performance in M. menidia. We compared both aerobic (prolonged and endurance) and anaerobic (burst) swimming capacities between intrinsically fast-growing fish from the north (Nova Scotia, NS) and intrinsically slow-growing fish from the south (South Carolina, SC) and between growth-manipulated phenotypes within each population. We also compared swimming speeds and endurance between fasted and recently fed fish within populations. Maximum prolonged and burst swimming speeds of NS fish were significantly lower than those of SC fish, and swimming speeds of fast-growing phenotypes were lower than those of slow-growing phenotypes within populations. Fed fish had lower burst speeds and less endurance than fasted fish from the same population. Thus, high rates of growth and the consumption of large meals clearly diminish swimming performance, which likely increases vulnerability to predation and decreases survival and relative fitness. The submaximal growth rate of southern M. menidia appears to be adaptive, resulting from balancing selection on rates of somatic growth.

Abstract. The Atlantic silverside (Menidia menidia) exhibits countergradient latitudinal variation in somatic growth rate along the East Coast of North America. Larvae and juveniles from high-latitude populations display higher intrinsic rates of energy consumption and growth than genotypes from low-latitude populations. The existence of submaximal growth in some environments suggests that trade-offs must counter the oft-cited theoretical benefits of energy and growth maximization (e.g., ''bigger is better,'' ''faster is better'') in the immature life stages. We hypothesized that energy and growth maximization trades off against investment in defense from predators. We conducted laboratory selection experiments to compare vulnerability to predation of silversides from: (1) fast-growing northern (Nova Scotia, NS) versus slow-growing southern (South Carolina, SC) source populations; (2) phenotypically manipulated fast-growing versus moderately-growing NS fish; and (3) recently fed versus unfed NS and SC fish. Tests involved fish drawn from common-garden environments and were conducted by subjecting mixed-treatment schools of sizematched silversides to natural, common piscine predators. NS silversides suffered significantly higher predation mortality than SC silversides. Parallel results were found in phenotypic manipulation of growth: NS silversides reared on a fast-growth trajectory (ϳ1.0 mm/day) were significantly more vulnerable to predation than those growing at a moderate rate (ϳ0.5 mm/day). Food consumption also affected vulnerability to predators: Silversides with large meals in their stomachs suffered significantly higher predation mortality than unfed silversides. Differences in predation vulnerability were likely due to swimming performance, not attractiveness to predators. Our findings demonstrate that maximization of energy intake and growth rate engenders fitness costs in the form of increased vulnerability to predation.Key words. Countergradient variation, evolutionary physiology, genotype-environment covariance, growth rate, latitude, life-history evolution, locomotory performance, optimal foraging theory. At what rate should immature organisms grow to maximize their fitness? Ecological theory suggests that natural selection favors genotypes that maximize net energy acquisition and growth (Lotka 1922;Ware 1982;Stearns 1992). Because rates of survival and reproductive output generally increase with size in many organisms, individuals that grow faster would reap the benefits of large size earlier in life and experience higher survival and fitness (Sauer and Slade 1987;Sogard 1997). Despite evidence and theoretical arguments to the contrary (Priede 1977;Calow 1982;Sibly et al. 1985;Abrams et al. 1996;Arendt 1997;Conover and Schultz 1997;Gotthard 2000), this faster-is-better paradigm is pervasive: Many life-history studies employ growth rate as a direct scalar of relative fitness among juveniles (e.g., Hatfield and Schluter 1999), and habitat quality is routinely evaluated based on positive associations...

Understanding the evolution of growth rate requires knowledge of the physiology of growth. This study explored the physiological basis of countergradient variation (CnGV) in somatic growth across latitudinal populations of the Atlantic silverside, Menidia menidia. Energetics of northern (Nova Scotia, Canada) and southern (South Carolina, USA) genotypes were compared across resource levels, temperatures, and fish sizes to identify trade-offs to rapid growth. Offered unlimited resources, genotypes differed in both energy acquisition and allocation. Food consumption, growth, and efficiency of northern genotypes were consistently higher than in southern genotypes, across temperatures and body sizes. Feeding metabolism (specific dynamic action; SDA) was proportional to meal size, differing between genotypes to the extent that food consumption differed. Given limited resources, northern and southern genotypes displayed similar growth, efficiency, routine activity, and SDA across temperatures and fish sizes. Routine metabolism was equal at 17°C and 22°C, yet was significantly higher in northern fish at 28°C. Growth rates in M. menidia do not appear to trade off across environments or body sizes, i.e., at no temperature, ration, or size do southern fish outgrow northern conspecifics. Nor does submaximal growth result from increased costs of maintenance, tissue synthesis, or routine activity. Based on our findings, we propose that CnGV consumption and growth in M. menidia likely result from trade-offs with other energetic components, namely sustained and burst swimming.