Steroid hormones and immune function: experimental studies in wild and captive dark-eyed juncos (Junco hyemalis)

Monogamous and polygynous male songbirds generally differ in their breeding season profiles of circulating testosterone. Testosterone level spikes early in the breeding season of monogamists and then declines, but it remains high in polygynists. Male dark-eyed juncos (Junco hyemalis) are socially monogamous and exhibit the usual pattern, but experimental maintenance of high testosterone throughout the breeding season alters normal behavior and physiology and affects various components of annual reproductive success but not overall annual success. Because stabilizing selection predicts that alteration of naturally existing phenotypes should reduce lifetime reproductive success, we asked whether prolonged testosterone exposure might impair immune function and perhaps thereby reduce life span. We assessed immune function in captive and wild male juncos that we treated with either testosterone-filled or empty Silastic implants. Results indicate that prolonged elevation of testosterone suppresses antibody production in captive males and cell-mediated immunity in wild males. Together these results suggest that testosterone-treated males may be more susceptible to disease or parasitic infection. As earlier studies have shown, levels of corticosterone as well as testosterone are higher in testosterone-treated males, so it is unclear whether the immune suppression we observed is due to testosterone's direct effects on immunity or testosterone's influence on glucocorticoid production. We discuss results in the context of recent hypotheses regarding life-history theory and potential endocrine-immune interactions.     

Winter fattening in the dark-eyed junco: plasticity and possible interaction with migration trade-offs

Although fat often supplies the major source of metabolic fuel during winter fasts of birds, this critical life-history trait is little studied by ecologists. In the dark-eyed junco Junco hyemalis, we have in a series of studies investigated the extent of plasticity in the winter fat reserve. Earlier (Rogers et al. 1993), we reported (1) a highly variable pattern of geographic variation in the winter fat reserve of junco populations in eastern North America, (2) disappearance of statistically significant interpopulation variation after experimental displacement to a common latitude, and (3) post-displacement temporal variation in the fat reserve. In analyses reported here, recent temperature, recent snowfall (a measure of short-term predictability of resources), season (perhaps reflecting continued exposure to unpredictable resources) and daylength explained spatial variation in the fat store. Recent temperature explained temporal variation in the fat reserves of groups of displaced juncos. These results suggest that platticity in a life-history trait has evolved in an uncertain winter environment. Through environment-dependent fattening, the costs of fat can be avoided during warm periods and at locations where fat confers little benefit, whereas benefits of fat can be quickly gained if weather conditions become harsh and snowfall might restrict food. Three types of winter fatteners probably exist among birds: responders (fatten in response to the proximate environment), predictors (fatten in anticipation of long-term environmental conditions), and responder-predictors (combination of both types of regulation). Because dark-eyed juncos select different winter latitudes as they age, we hypothesize that the nonbreeding component of the life-history of juncos includes the co-adapted plastic traits of winter fattening and post-breeding migration. Life-history theory can apparently explain important traits related to fitness in the nonbreeding period.     

Comparative morphology of dark-eyed juncos Junco hyemalis breeding at two elevations: a common aviary experiment

Morphologies of bird species often vary along elevation gradients, yet causes of the variation have not been examined experimentally. We investigated variation in morphological traits of the dark‐eyed junco Junco hyemalis, breeding at 1,000 m a.s.l. (low‐elevation; i.e. low) and 2,000 m asl (high‐elevation; i.e. high) in the Rocky Mountains, Canada. Eight morphological traits were measured in free‐living birds. We found two consistent differences in populations between elevations: at high‐elevation sites, females had longer wings and males had longer tails than birds from low‐ elevation sites. Other age‐ and gender‐ specific results were observed in free‐living birds between elevations: tarsi were shorter in high‐elevation second year (SY) females and after second year (ASY) males, beak lengths were slightly longer in low‐elevation SY females, and high‐elevation ASY females tended to have lower fat than low‐elevation ASY females. Morphological differences may result from genetic differences between elevations, or phenotypic flexibility resulting from exposure to the different environmental conditions. To identify which mechanism caused the difference in morphometrics, hand‐reared birds from low‐ and high‐elevation habitats were raised in identical conditions with unlimited access to high quality food until they had replaced all feathers. The traits measured in the lab (wing and rectrix length, weight and fat score) tended to increase in magnitude compared to field values. Juncos from high‐ and low‐elevations had similar responses to the aviary environment, with one exception: males from high‐elevation sites had greater weight gain relative to free‐living juncos than males from low‐elevation sites. Thus, morphological traits in dark‐eyed juncos were phenotypically flexible, capable of growing larger in the laboratory environment. However, there were also persistent genetic or perinatal/maternal differences underlying population responses that prevented traits from converging under aviary conditions. As a result, trait size differences between high‐ and low‐elevation populations were maintained or exacerbated in the common aviary environment.     

Interspecific dominance and the foraging behaviour of juncos

We observed mixed groups of dark-eyed juncos (Junco hyemalis) and grey-headed juncos (Junco caniceps) at baited sites in northern Arizona during the non-breeding season. In interspecific and inter-racial conflicts, J. caniceps dorsalis was dominant to J. caniceps caniceps and to two races of dark-eyed juncos. Junco caniceps dorsalis also fed significantly faster than any of the other juncos. For both species, feeding rates were approximately the same in large and small mixed-species groups, though in larger groups, individual grey-headed juncos won conflicts at a higher rate and individual dark-eyed juncos lost conflicts at a higher rate. Also, dark-eyed juncos fed at a significantly lower rate in groups comprised mostly of grey-headed juncos than in groups of similar size but composed mostly of conspecifics. Residency times and recapture probabilities were similar for the two species, suggesting little difference in over-winter survival.     

Importance of Montane Riparian Forest and Influence of Wildfire on Nest-Site Selection of Ground-Nesting Birds

Abstract: We studied breeding populations of 2 coexisting ground-nesting birds, the red-faced warbler (Cardellina rubrifrons) and yellow-eyed junco (Junco phaeonotus), in high-elevation (>2,000 m) forested drainages of the Santa Catalina Mountains, Arizona, USA. From 2004 to 2005, we 1) estimated density and nesting success of breeding populations of red-faced warblers and yellow-eyed juncos, 2) identified nest-site characteristics for each species (i.e., used sites vs. random plots), 3) compared nest-site characteristics between the 2 species, and 4) examined effects of a recent (2003) wildfire on distribution of nests of both species. In addition, we estimated the areal extent of montane riparian forest (the preferred breeding habitat of both species) within high-elevation forests of the Santa Catalina Mountains. We found that red-faced warblers and yellow-eyed juncos were the 2 most common ground-nesting birds within our study area with an average density of 2.4 and 1.4 singing males/ha, respectively, along drainage bottoms. Compared to random plots, most red-faced warbler and yellow-eyed junco nests were located close (≤30 m) to drainage bottoms within a strip of montane riparian forest characterized by abundant brush, small woody debris, ferns, and forbs (both species), high number and diversity of saplings and small trees (red-faced warblers), and abundant shrubs and downed logs and less canopy cover (yellow-eyed juncos). Although both species nested in close proximity within montane riparian forest, nest-site characteristics differed between the 2 species, especially at finer spatial scales. For example, most yellow-eyed juncos nested adjacent to grass (principally Muhlenbergia spp.), whereas red-faced warblers nested adjacent to a variety of plant species, including grass, bigtooth maple (Acer grandidentatum), white fir (Abies concolor), and Douglas-fir (Pseudotsuga menziesii). Both red-faced warblers and yellow-eyed juncos avoided nesting in areas burned during a recent wildfire. In addition, nesting success was low for red-faced warblers (13%) and yellow-eyed juncos (19%) following the wildfire, suggesting an indirect negative effect of fire on breeding populations in the short-term. Montane riparian forest appears to provide important breeding habitat for red-faced warblers and yellow-eyed juncos. However, little research or conservation planning has been directed toward montane riparian forest in the region, even though this forest type is limited in its areal extent (<11% of high-elevation forest in the Santa Catalina Mountains) and increasingly threatened by disturbance. Results from our study can be used to facilitate the management and conservation of breeding populations of red-faced warblers and yellow-eyed juncos in high-elevation forests of the southwestern United States.     

Rapid evolution of a sexually selected trait following population establishment in a novel habitat

Colonization of novel environments creates new selection pressures. Sexually selected traits are affected by the physical and social environment and should be especially susceptible to change, but this has rarely been studied. In southern California, dark-eyed juncos, (Junco hyemalis) naturally breed in mixed-coniferous temperate forests, typically from 1500 m to 3000 m in elevation. In the early 1980s, a small population became established in a coastal habitat, the University of California, San Diego campus, which has a mild, Mediterranean climate. I show that a sexually and socially selected signaling trait--the amount of white in the tail--has declined by approximately 22% as compared to mountain juncos. I address three main factors that could explain the difference between mountain and coastal juncos: phenotypic plasticity, genetic drift, and selection. Results indicate that the first two can be ruled out as the sole cause of the plumage change, which implies that selection contributed to the genetic differentiation from the mountain population. The estimated rate of evolution is about 0.2 haldanes, comparable with rates of change in systems where individuals have been artificially introduced into new environments (e.g., guppies and Drosophila). This is the first study to demonstrate evolution of a sexually selected trait after only several generations resulting from a natural invasion into a novel environment.     

Elevation impacts the balance between growth and oxidative stress in coal tits

The short favorable period of time available for the growth in seasonal environments could constrain the resources allocation between growth and other life-history traits, and the short-term fitness benefits of increased growth rate may prevail over other functions. Accelerated growth rates have been associated with long-term deleterious consequences (e.g., decreased lifespan), and recently oxidative stress (the imbalance between pro-oxidants generation and antioxidant defenses) has been suggested as a mediator of these effects. Here, we examined the impact of elevation on growth rate and self-maintenance parameters (resting metabolism, oxidative damage, and antioxidant defenses) of coal tit chicks (Periparus ater). We predicted that the shorter favorable season at the higher-elevation site could lead to a reallocation of resources towards growth at the expense of self-maintenance processes. We found that chicks at high elevation grew significantly faster in terms of body mass and body size. Chicks from the high-elevation site presented higher resting metabolism, higher oxidative damage level, but similar antioxidant defenses, compared to low-elevation chicks. Interestingly, the chicks exhibiting the better antioxidant defenses at 7 days were also those with the highest resting metabolic rate, and the chicks that grew at the faster rate within the high-elevation site were those with the highest levels of oxidative damage on DNA. Our study supports the idea that increasing elevation leads to a higher growth rate in coal tit chicks, possibly in response to a shorter favorable season. In accordance with life-history theory, a bigger investment in growth was done at the expense of body maintenance, at least in terms of oxidative stress.     

Females competing to reproduce: dominance matters but testosterone may not

The associations among aggression, testosterone (T), and reproductive success have been well studied, particularly in male birds. In many species, males challenged with simulated or real territorial intrusions increase T and levels of aggression, outcomes linked to higher dominance status and greater reproductive success. For females, the patterns are less clear. Females behave aggressively towards one another, and in some species, females respond to a social challenge with increases in T, but in other species they do not. Prior work on female dark-eyed juncos (Junco hyemalis) had shown that experimental elevation of T increases social status and intrasexual aggression. Here, we conducted two experiments designed to answer three questions: Are endogenous concentrations of T associated with dominance status in captive female juncos? Does dominance status influence readiness to breed in female juncos? And do captive females increase T in response to a challenge? In the first experiment, we introduced two females to a breeding aviary, allowed them to form a dominance relationship and then introduced a male. We found that dominant females were more likely to breed than subordinates, but that dominance status was not predicted by circulating T. In the second experiment, we allowed a resident male and female to establish ownership of a breeding aviary (territory) then introduced a second, intruder female. We found that resident females were aggressive towards and dominant over intruders, but T did not increase during aggressive interactions. We suggest that during the breeding season, intrasexual aggression between females may influence reproductive success, but not be dependent upon fluctuations in T. Selection may have favored independence of aggression from T because high concentrations of T could interfere with normal ovulation or produce detrimental maternal effects.     

Feather melanin and microstructure variation in dark‐eyed junco Junco hyemalis across an elevational gradient in the Selkirk Mountains

Variation in feather melanism and microstructure can arise through sexual selection and ecological functional drivers. Melanin‐based plumage traits are associated with sexual dichromatism and the intensity of sexual selection in many avian species, but also have several ecological benefits such as protection against ultra‐violet (UV) radiation, camouflage, and feather strength. Additionally, feather microstructure influences thermoregulation. Plumage variation across species is well documented; however, the relative role of sexual selection and ecological drivers in intra‐specific and within‐population variation is less established. We investigated UV reflectance, melanism, and feather microstructure in a population of Oregon dark‐eyed juncos Junco hyemalis oreganus between high (1900–2200 m a.s.l.) and low (450–800 m a.s.l.) elevations in the Selkirk Mountains to evaluate potential sexual selection and ecological drivers of variation. We found no difference in UV reflectance or lightness (melanism) of head feathers between elevations, but individuals at high elevation had lighter (less melanism) and less brown (less pheomelanin) body contour feathers than at low elevations. High elevation individuals also had longer contour feathers with more pronounced plumulaceous regions. Sexual dichromatism did not vary between elevations, leading us to reject sexual selection in favour of ecological functional drivers of plumage variation in this system. To our knowledge, this is the first study to identify within‐population differences in feather melanism and microstructure between different elevations.     

The Effects of Experimentally Elevated Testosterone and Food Deprivation on Food Consumption and Prey Size Preferences in Male Dark-Eyed Juncos (Junco hyemalis, Emberizidae: Passeriformes)

Numerous studies have shown that the experimental elevation of circulating levels of testosterone reduces parental behaviour in male birds, particularly the provisioning of young. The mechanisms responsible for this change in behaviour are not fully understood. In this study, we examine the effects of elevated testosterone on food consumption and prey selection, both of which have potential consequences for nestling provisioning behaviour. We manipulated testosterone and performed two experiments on a captive, non-breeding population of male dark-eyed juncos ( Junco hyemalis ) on long day-lengths. In the first experiment, we subjected juncos to 3 h of food deprivation and compared food consumption and prey size selection by males with elevated testosterone (testosterone males) to that of control males. Testosterone males consumed more food than control males and showed a preference for larger prey. In a second experiment in which small prey were more abundant than large prey, food consumption and prey size preferences did not differ between testosterone and control males. We also manipulated the duration of food deprivation in the second experiment. Males of both treatments consumed more small prey under conditions of mild (1 h) or moderate (5 h) food deprivation and consumed more large prey under conditions of intermediate (3 h) food deprivation. We discuss our results and the effects that testosterone has on self-maintenance behaviour and male parental effort.     

Effects of captivity and testosterone on the volumes of four brain regions in the dark-eyed junco (Junco hyemalis)

This study investigates the effects of captivity and testosterone treatment on the volumes of brain regions involved in processing visual and spatial information in adult dark-eyed juncos (Junco hyemalis). We treated captive and free-living male juncos with either testosterone-filled or empty implants. Captive juncos had a smaller hippocampal formation (HF) (both in absolute volume and relative to telencephalon) than free-living birds, regardless of hormone treatment. Testosterone-treated males (both captive and free-living) had a smaller telencephalon and nucleus rotundus, but not a smaller HF or ectostriatum, than controls. We found that free-living testosterone-treated males had larger home ranges than free-living controls in agreement with earlier experiments, but we found no corresponding difference in HF volume. We discuss the implications of the effect of captivity on HF volume for past and future laboratory experiments.     

Non‐Redundant Social Information Use in Avian Flocks with Multisensory Stimuli

Animals generally live in multisensory worlds; however, our understanding of multisensory perception is rather limited, despite its relevance for explaining the mechanisms behind social interactions, such as collective detection while foraging in groups. We tested how multisensory stimuli affected the antipredator behavior of dark‐eyed juncos (Junco hyemalis) using alarm calls as an auditory signal and flushing behavior as a visual cue. We varied the degree of risk within the group by manipulating the number of group mates alarm calling and/or flushing using robotic birds. We assumed that alarm calling and flushing were redundant stimuli and predicted that they could generate one of three types of responses (enhancement, equivalence, or antagonism) depending on the mechanism of multisensory perception. We set up an artificial flock with three robotic juncos surrounding a live junco and controlled for multiple confounding factors (e.g., identity of the focal, body mass, food deprivation time). We found that the degree of alarm of live juncos increased when at least one robot flushed. However, the time it took the live individuals to react to the robots' behavior increased, rather than decreased, with at least one alarm call. This could be the result of an orienting response or sensory overload, as live juncos increased scanning behavior after being exposed solely to alarm calls. Contrary to some theoretical assumptions, alarm calling and flushing behavior elicited independent unimodal responses, suggesting that they are non‐redundant stimuli and that together they could reduce the occurrence of false alarms and facilitate flock cohesion.     

Predation and life-history variation in Poecilia reliculata (Cyprinodontiformes: Poeciliidae)

Synopsis Reznick and Endler investigated natural variation in life-history traits of populations of Trinidad guppies exposed to one of three intensities of predation: (i) high predation directed primarily at adults, (ii) moderate predation directed primarily at juveniles, and (iii) low predation. They were able to document significant interpopulational differences in life-history traits associated with this differential predation on a trait-by-trait basis. However, the present extended multivariate analysis indicates that (1) life-history traits do not differ significantly between populations exposed to moderate versus low predation, although both differ greatly from high-predation populations; (2) life-history variation is strongly unifactorial; and (3) despite the importance of predation effects, approximately 17% of the variation in life-history variables cannot be accounted for by predation intensity. Residual variation has no obvious geographical patterns, but instead seems to reflect local environmental variability. Life-history differences between predation regimes are consistent with residual patterns of variation within regimes, suggesting that local variation provides the raw material for extrapolation in response to predation, but also that it influences the direction of correlated change in life-history traits.     

Effects of captivity and testosterone on the volumes of four brain regions in the dark‐eyed junco (Junco hyemalis)

This study investigates the effects of captivity and testosterone treatment on the volumes of brain regions involved in processing visual and spatial information in adult dark‐eyed juncos (Junco hyemalis). We treated captive and free‐living male juncos with either testosterone‐filled or empty implants. Captive juncos had a smaller hippocampal formation (HF) (both in absolute volume and relative to telencephalon) than free‐living birds, regardless of hormone treatment. Testosterone‐treated males (both captive and free‐living) had a smaller telencephalon and nucleus rotundus, but not a smaller HF or ectostriatum, than controls. We found that free‐living testosterone‐treated males had larger home ranges than free‐living controls in agreement with earlier experiments, but we found no corresponding difference in HF volume. We discuss the implications of the effect of captivity on HF volume for past and future laboratory experiments. © 2000 John Wiley & Sons, Inc. J Neurobiol 43: 244–253, 2000     

Ecological and phylogenetic determinants of life-history patterns among ten loricariid species

We analysed the influence of ecological factors, phylogenetic history and trade-offs between traits on the life-history variation among 10 loricariid species of the middle Paraná River. We measured eight life-history variables and classified the life-history strategies following the equilibrium–periodic–opportunistic (EPO) model. Principal-component analysis of life-history traits segregated species along a gradient from small opportunistic (low fecundity, low parental investment) to large equilibrium (low-medium fecundity, high parental investment) species. A clear periodic strategist was absent in the analysed assemblage. Variation partitioning by canonical phylogenetic ordination analysis showed both a component of variation uniquely explained by phylogenetic history (PH; 32.2%) and a component shared between PH and ecological factors (EF; 37%). The EPO model is a useful tool for predicting correlations among life-history traits and understanding potential demographic responses of species to environmental variation. Life-history patterns observed throughout Loricariidae suggests that this family has diversified across all three endpoint strategies of the EPO model. Our study indicates that evolutionary lineage affiliation at the level of subfamily can be a strong predictor of the life-history strategy used by each species.     

Social status does not affect resting metabolic rate in wintering dark-eyed junco (Junco hyemalis)

Studies of wintering birds have demonstrated a correlation between social rank and energy expenditures. It is assumed that dominance is energetically costly because of increased activity, possibly caused by elevated androgen levels. As winter acclimatization leads to an increase in metabolic rate, maintaining dominance status in a cold climate can be a substantial challenge. We measured resting metabolic rates in dominant and subordinate dark-eyed juncos (Junco hyemalis) living in small groups in a controlled winter environment. We found no significant effect of social rank when controlling for body size. It has been shown previously that high testosterone levels during the nonbreeding season can lead to higher body conductance, fat loss, and higher nocturnal body temperature. A hypothesis explaining our result is that for juncos it is preferable to maintain low androgen levels during winter and to maintain social rank using a mechanism other than higher agonistic activity.     

Experimentally-elevated testosterone, female parental care, and reproductive success in a songbird, the Dark-eyed Junco (Junco hyemalis)

In male dark-eyed juncos (Junco hyemalis), experimentally elevated testosterone (T) decreases male parental care and offspring survival, but results in higher overall fitness through greater mating success. To help address the ensuing question of what prevents selection from favoring higher levels of T in male juncos, we manipulated T in female juncos. A previous study demonstrated no effect of experimentally elevated T on female incubation behavior, suggesting that female parental behavior might be insensitive to T. In this study we asked whether experimentally elevated T mediates other female parental behaviors and whether variation in T-mediated parental behavior might influence reproductive success. We videotaped free-living control- and T-females during nesting to quantify brooding behavior when young were 3 days old and provisioning behavior when young were 6 days old. Nest defense was measured by quantifying responses to a mounted predator placed near the nest. Reproductive success was assessed via fecundity, nestling quality, and nest survival. T-females spent less time than control females brooding but did not differ in provisioning rate. T-females performed fewer dives at the predator mount and, unlike controls, failed to increase defense as nesting progressed. T-females also had lower daily nest survival and lower nest success (odds of producing at least one fledgling). We conclude that some aspects of female parental behavior are sensitive to experimentally elevated T while others are not and consider the implications for the evolution of T-mediated characters in both sexes.     

Nitrogen deposition and herbivory affect biomass production and allocation in an annual plant

If environmental conditions vary, plasticity in life-history traits is predicted. A recent model indicates that males and females should differ in life-history traits, because sexes differ in optimal attributes depending on species ecology. In this study we test the impact of two biotic factors in combination (presence/absence of predators and low/high food level) on gender specific life-history traits in the damselfly Coenagrion puella (Odonata). Results show that predator presence and low food density decreased activity in both sexes. Additionally, individuals with less food grew more slowly, emerged later, remained smaller and had a higher mortality. At low food densities, however, and in contrast to former investigations, individuals from treatments with predator presence were the same size or larger than individuals without predators. Gender had a strong impact on larval activity and life-history traits and sexes differed in development. Females were less active and took longer to complete development, but emerged at a larger size, weight and fat content. This study highlights the importance of gender specific approaches in life-history research.     

Top-down effects of an invasive omnivore: detection in long-term monitoring of large-river reservoir chlorophyll-a

Elevation represents an important selection agent on self-maintenance traits and correlated life histories in birds, but no study has analysed whether life-history variation along this environmental cline is consistent among and within species. In a sympatric community of passerines, we analysed how the average adult survival of 25 open-habitat species varied with their elevational distribution and how adult survival varied with elevation at the intra-specific level. For such purpose, we estimated intra-specific variation in adult survival in two mountainous species, the Water pipit (Anthus spinoletta) and the Northern wheatear (Oenanthe oenanthe) in NW Spain, by means of capture–recapture analyses. At the inter-specific level, high-elevation species showed higher survival values than low elevation ones, likely because a greater allocation to self-maintenance permits species to persist in alpine environments. At the intra-specific level, the magnitude of survival variation was lower by far. Nevertheless, Water pipit survival slightly decreased at high elevations, while the proportion of transient birds increased. In contrast, no such relationships were found in the Northern wheatear. Intra-specific analyses suggest that living at high elevation may be costly, such as for the Water pipit in our case study. Therefore, it seems that a species can persist with viable populations in uplands, where extrinsic mortality is high, by increasing the investment in self-maintenance and prospecting behaviours.     

Disturbance regimes and life-history evolution

Disturbance regimes are ecologically important, but many of their evolutionary consequences are poorly understood. A model is developed here that combines the within- and among-season dynamics of disturbances with evolutionary life-history theory. "Disturbance regime" is defined in terms of disturbance timing, frequency, predictability, and severity. The model predicts the optimal body size and time at which organisms should abandon a disturbance-prone growth habitat by maturing and moving to a disturbance-free, nongrowth habitat. The effects of both coarse-grained (those affecting the entire population synchronously) and fine-grained disturbances (those occurring in a patch dynamics setting) are explored. Several predictions are congruent with previous theory. Infrequent or temporally unpredictable disturbances should have little effect on the evolution of life-history strategies, even though they may cause high mortality. Similar to seasonal time constraints on reproduction, disturbance regimes can synchronize metamorphosis within a population, resulting in a seasonal decline in body size at maturity. Other model predictions are novel. When disturbances cause high mortality, coarse-grained disturbances have a much stronger effect on life-history strategies than fine-grained disturbances, suggesting that population structure (relative to the scale of disturbance) plays a critical evolutionary role when disturbances are severe. When within-population variance in juvenile body size is high, two consecutive seasonal declines in body size at maturity can occur, the first associated with disturbance regime and the second associated with seasonal time constraints.