Environmental integration: patterns of correlation between environmental factors, early life decisions, and their long-term consequences

Questions: How does the temporal correlation of key environmental variables (here temperature and precipitation) affect life-history decisions in early life, and what are the long-term consequences (accumulated damage, survival, and expected reproductive success)? Hypothesis: Strong environmental correlation (environmental integration) is an important signal for the development of life-history traits, such as growth rate, phenotype, and reproduction, leading to higher fitness; plastic and fixed developmental strategies have different fitness depending upon the level of environmental integration. Mathematical methods: A dynamic state-dependent model in which the state of the organism is characterized by mass, reproductive investment, and accumulated damage, all of which are affected by feeding activity and developmental costs mediated by the environment. Fitness is measured as expected lifetime reproduction. Key assumptions: We assume that at each time step the resources gained by an individual through foraging activity are determined by developmental phenotype, which itself is the result of a decision process, and are then allocated to somatic growth, repair of cellular damage (e.g. oxidative stress) or gonadal tissue. Results/Conclusions: (1) The differences in growth rate and reproductive investment between the plastic and fixed phenotypic strategies are greater at low levels (both positive and negative directions) of environmental integration. (2) Optimal resource allocation changes as a function of environmental gradient only for the plastic phenotypic strategy, and the difference in the onset of resource allocation between the plastic and fixed strategies is greater at low levels of environmental integration (i.e. the correlation between environmental factors and their fluctuation affects the reproductive timing decision though modification of resource allocation). (3) There is a marked difference in fitness when there is a low correlation and high fluctuation - conversely, the correlation has little effect when environmental fluctuation is low. (4) Even with costs, the investment in phenotype-environment matching has greater payoffs for individuals who are better able to track changes in their environment. Our results highlight that to understand the interactions between developmental decisions, we need to take into account not only the average environmental conditions but also their dynamics through time (variance and covariance).