Population dynamics and mean age at reproduction
In our metapopulation of house sparrows, annual variation in population size and the strength of density-dependent intra-specific competition appears to generate temporal variation in the mean age at reproduction. If individuals produce the same number of recruits, and in the absence of density regulation, it is expected that individuals that reproduce earlier in life and thus have shorter generation times will be selected for, because early reproduction will result in a higher individual growth rate. Our results suggest that population dynamics affects the mean age at reproduction because density-dependent competition constrains the reproductive output of younger individuals. This is therefore consistent with the idea that density-dependent competition constrains individual growth rates. Especially for males, older individuals seem to contribute more to population growth when intra-specific competition is stronger.
We found strong support for density regulation in the production of recruits. In years when population sizes were higher than average, individuals produced fewer recruits. The patterns of density regulation in recruitment could have been mediated by effects on the reproduction of parents or via differential juvenile survival. Thus, in years where population sizes were higher than average, individuals fledged fewer offspring and/or juvenile survival was lower. In our study, we did not find that adult survival was density dependent, but given that our measure of first-year recruitment combines investment in reproduction by the parents with the survival probability of their offspring, then juvenile survival may well represent the point in the life-histories of these house sparrows where survival is most closely density regulated.
When we analyzed how the mean age at reproduction was affected by population dynamics across years and populations, we found a trend suggesting that in years where population sizes where larger than average then the mean age at reproduction was older (Table 3). We also analyzed how the expected growth of the population, measured as the mean fitness of the population in a given year, affected the mean age of the reproducing parents. We found that when the mean fitness of the population was low then the individuals that managed to reproduce were older (Table 4, model 1A). In contrast, when populations were growing either all individuals, even the young ones, managed to reproduce or individuals that invest more in current reproduction at the expense of future reproduction or survival were able to contribute disproportionately to population growth, as predicted by density-dependent selection theory (MacArthur & Wilson 1967; Engenet al. 2013; Engen & Sæther 2016). Further analyses showed that this was not solely the result of the age structure of the population (Table 4, model 1B). These results are thus consistent with classic density-dependent selection theory predicting that in scenarios with stronger competition, the individuals being favored will be the ones that invest more in future reproduction (e.g. by investing in traits that increase survival and enable them to reproduce later). But is also consistent with the idea that ‘high-quality’ individuals are the ones that manage to grow old and can reproduce when density-dependent competition is strongest.