Populations are not stable. But despite decades of study, the drivers of population variability are yet to be fully understood. A new study in the Journal of Animal Ecology suggests that somatic growth variation can be as important as early life‐history survival in driving population fluctuations in some marine fish species. Quantitative ecologist Dr Christine Stawitz led the study as part of her dissertation research with Dr Timothy Essington at the University of Washington. She currently works as a population model developer at ECS Federal, LLC, in support of NOAA Fisheries.

Centuries ago, fishermen realized their fishery catches were not stable, but highly variable from year to year. This prompted much speculation as to why. For a long time, all variability was attributed to movement; in years of poor catch, the fish must be somewhere else. Then, in 1914, a scientist called Johan Hjort observed that the demographics of captured fish also varied from year to year. Some years, there would be a lot of small (young) fish in the catch, while other times, lots of older, bigger fish were caught. Hjort proposed this was due to differences in early life history survival between cohorts of fish born in different years. In good years, he assumed many fish would survive the critical larval stage and as a result, more fish would be caught. After bad years, when very few fish survived this critical life stage, populations would be smaller.

In the intervening century, we have learned much about population ecology, but Hjort’s hypothesis is still canon. We’ve learned that maternal investments into offspring matter, and therefore the mother’s condition may be very important to her offspring’s survival. We’ve learned that the strength of predation on a species can also introduce variability into their population beyond what is caused by early life history survival. We’ve learned that competition for food induces a density-dependent reduction in somatic growth rates, with consequences for populations. However, it is still often assumed environmental perturbations primarily affect fish populations only through early life history survival.

In this study, we used simulated populations along with empirical data to test if changing somatic growth rates could affect fish population biomass as much as changes in early life history survival. First, we used empirical data to quantify how somatic growth and early life history survival varied over time. Next, we used a stage-structured population model to generate time series of fish biomass that included only somatic growth variation, only early life history variation, and both types of variation. We examined the amount of variation in the generated biomass time series across these three scenarios, which gave us an idea of how much relative variation was caused by each process. To test the sensitivity of these results to different factors, we varied the amount of age truncation in the population, examined the effect of a size-dependent maturation curve, introduced correlation between length and weight of fish, and repeated the analysis for eight life history archetypes.

For three out of eight life history types, the contribution of somatic growth to biomass variability was equal to or larger than the effect of varying early life history survival. Increasing the level of age truncation made early life history variation more important, while introducing length-dependent maturation or covariation between weight and length made somatic growth variation more important. We thought that the life history of each species type would relate to which type of variability was more influential, but we didn’t see species with similar life histories responding similarly to variation types. Instead, characteristics of the empirical time series used to simulate growth or early life history variability determined which process had a bigger influence.

These results suggest it is important to consider environmentally-induced variation in growth, in addition to early life history survival, when analyzing and predicting population fluctuations. Particularly, if we do not consider the effect of variation in other life history processes, beyond early life history, we may underestimate how variable populations really are. Hjort was right: some population variability in fishes is certainly caused by varying survival through the larval stage. However, in this century, predicting population response to a changing environment is more critical than ever before. With the advantage of advanced statistical tools and computing power, we can include environmental effects on adult somatic growth, in addition to early life history survival, in population dynamics models.

More Info:

Stawitz, C.C. and Essington, T.E. (2018) Somatic growth contributes to population variation in marine fishes. Journal of Animal Ecology.