Umeå University's logo

Size variation among individuals born at the same time in a common environment (within cohorts) is a common phenomenon in natural populations. Still, the mechanisms behind the development of such variation and its consequences for population processes are far from clear. We experimentally investigated the development of early within-cohort size variation in larval perch (Perca fluviatilis). Specifically we tested the influence of initial variation, resulting from variation in egg strand size, and intraspecific density for the development of size variation. Variation in egg strand size translated into variation in initial larval size and time of hatching, which, in turn, had effects on growth and development. Perch from the smallest egg strands performed on average equally well independent of density, whereas larvae originating from larger egg strands performed less well under high densities. We related this difference in density dependence to size asymmetries in competitive abilities leading to higher growth rates of groups consisting of initially small individuals under high resource limitation. In contrast, within a single group of larvae, smaller individuals grew substantially slower under high densities whereas large individuals performed equally well independent of density. As a result, size variation among individuals within groups (i.e. originating from the same clutch) increased under high densities. This result may be explained by social interactions or differential timing of diet shifts and a depressed resource base for the initially smaller individuals. It is concluded that to fully appreciate the effects of density-dependent processes on individual size variation and size-dependent growth, consumer feedbacks on resources need to be considered.

Most organisms exhibit a substantial size variation among individuals due to individual differences in experienced biotic and abiotic environmental conditions and because individuals undergo growth and development during most of their life time. One important issue in this context is how size variation within cohorts may develop over time. Here we tested the hypothesis, in gape-limited animals such as fish, that size divergence among individuals within a cohort depends on the opportunity to undergo size-dependent diet shifts, by allowing initially larger individuals to make an early diet shift when the first resource becomes limiting. We used young-of-the-year perch (Perca fluviatilis) as our study organism. Competitive intensity and the opportunity to undergo a diet shift from zooplankton to macroinvertebrates affected both mean growth rates and the extent to which inter-individual variation in growth was manifested. As predicted, increased competition combined with the presence of both zooplankton and benthic macroinvertebrates increased the degree of size variation. However, size divergence was also observed among individuals when only the initial resource, zooplankton, was available. We argue that only non-exploitative interactions, such as dominance structures and social interactions could have caused this latter pattern, as exploitative competition is expected to lead to size convergence due to the superior competitive ability of smaller individuals. Our results suggest that diet shifts are not a prerequisite for size divergence in animal cohorts and that dominance and social interactions may have similar effects on size variation within cohorts. Finally, development of size variation is suggested to have strong implications for overall cohort performance.

Variation in growth rates among individuals leading to the formation of broad size distributions is commonly observed in animal cohorts. Here we use laboratory derived size-scaling relationships to identify mechanisms driving changes in size distribution patterns within cohorts during early ontogeny. We introduced young-of-the-year perch (Perca fluviatilis) cohorts with different variation in body size distributions in pond enclosures. We kept the exploitative competitive environment constant by adjusting the number of introduced fish such that metabolic requirements were constant between different treatments. Based on modelling results we theoretically derived relative growth rates of differently sized fish when only taken exploitative competitive interactions into account. In agreement with predictions we found that initial variation in body size was negatively correlated with subsequent changes in body size variation in the pond experiment. Corresponding results were obtained in a field study covering 13 studied young-of-the-year perch cohorts in a small lake. Besides having a lower maximum growth capacity, initially large fish also suffered more from resource limitation in our experiment. The results suggest that exploitation competition is a major factor behind growth patterns in young fish cohorts, generally leading to size convergence. To explain the commonly observed pattern of size divergence in animal cohorts, including fish, we suggest that differential timing of diet shifts or mechanisms not related to exploitative interactions must be taken into account. For diet shifts to lead to size divergence we suggest that individuals with an initial size advantage need access to an exclusive prey which has a high growth potential. This, in turn, allows initially larger individuals to surf on a wave of growing prey while individuals only capable to feed on a depressed initial resource experience low growth rates.

Th e degree to which growth in early life stages of animals is regulated via density-dependent feedbacks through preyresources is much debated. Here we have studied the infl uence of size- and density-dependent mechanisms as well as sizeselectivepredation pressure by cannibalistic perch Perca fl uviatilis on growth patterns of young-of-the-year (YOY) perchcovering several lakes and years. We found no infl uence of initial size or temperature on early body size development ofperch. In contrast, there was a negative relationship between reproductive output and the length of YOY perch at fi ve weeksof age. However, rather than an eff ect of density-dependent growth mediated via depressed resources the relationship wasdriven by positive size-selective cannibalism removing large individuals. Hence, given a positive correlation between thedensity of victims and predation pressure by cannibals, size-dependent interactions between cannibals and their victimsmay wrongly be interpreted as patterns of density-dependent growth in the victim cohort. Overall, our results support theview that density-dependent resource-limitation in early life stages is rare. Still, patterns of density-dependent growth mayemerge, but from variation in size-selective predation pressure rather than density as such. Th is illustrates the importanceof taking overall population demography and predatory interactions into account when studying growth patterns amongrecruiting individuals.

Cannibalistic interactions generally depend on the size relationship between cannibals and victims. In many populations, alarge enough size variation to allow for cannibalism may not only develop among age-cohorts but also within cohorts. Westudied the implications of variation in hatching period length and initial cohort size for the emergence of cannibalism andbimodal size distributions within animal cohorts using a physiologically structured population model. We found that thedevelopment of size bimodality was critically dependent on hatching period length, victim density and the presence of afeedback via shared resources. Cannibals only gained enough energy from cannibalism to accelerate in growth when victimdensity was high relative to cannibal density at the onset of cannibalism. Furthermore, we found that the opportunity forearly hatchers to initially feed on an unexploited resource increases the likelihood both for cannibalism to occur and sizebimodality to develop. Once cannibals accelerated in growth relative to victims size bimodality, reduced victim numbersand relaxed resource competition resulted. Th us, in addition to that cannibals profi ted from cannibalism through energyextraction, their potential victims also benefi ted as the resource recovered due to cannibal thinning. To ensure recruitmentsuccess, it can be critical that a few individuals can accelerate in growth and reach a size large enough to escape sizedependentpredation and winter starvation. Hence, within-cohort cannibalism may be a potentially important mechanismto explain recruitment variation especially for cannibalistic species in temperate climates with strong seasonality. However,the scope for size bimodality to develop as a result of cannibalism may be limited by low victim densities and size andfood-dependent growth rates.

In seasonal environments accumulated energy reserves are important to avoid starvation mortality during periods of low resource levels. Here we investigated patterns of energy accumulation and the importance of growth history for winter survival in young-of-the-year Eurasian perch (Perca fluviatilis). Under simulated winter conditions in aquaria’s we showed that high winter mortality most likely relate to the depletion of energy reserves in small perch. Correspondingly in a field study, using 4 lakes covering 3-6 lake years each, overwinter survival within cohorts was positively related to individual size. However, average size in autumn did not explain the variation in overwinter survival between cohorts. Instead we showed that seasonal growth history is an important factor. High growth rates late in season may increase cohort survival over winter irrespective of average size, related to a positive growth dependent increase in allocation to energy reserves when approaching winter. Mechanisms regulating within-season temporal dynamics of growth rates are therefore suggested to be important for overall cohort performance.