Marine protected areas (MPAs) are an increasingly common conservation and management tool worldwide. Typically, managers expect that after fishing ceases inside an MPA, previously fished populations will steadily increase in abundance as they return to unfished levels. Hence adaptive management typically involves examining the ratio of fish density after:before MPA implementation or inside:outside MPAs. However, the expectation of a steady, positive increase in population density inside MPAs is complicated by two factors: A) the expected increase depends on the level of fishing pre-MPA, which is usually unknown, and B) high variability in larval recruitment to populations in MPAs, both over time (pulses and droughts) and over space (hotspots and coldspots) makes post-MPA transient population trajectories highly variable. Often, management decisions (e.g., keep or remove an MPA) must be made on time scales that are faster (e.g., 5 years) than the transient dynamics of the populations responding to protection in the MPAs. We use data on a kelp forest fish (kelp bass, Paralabrax clathratus) from southern California MPAs to show how a combination of dynamic and statistical models can address this problem.
Results/Conclusions
First, analysis of a dynamic age-structured model reveals that there are simple, analytical predictions for the expected magnitude and time scale of population increase inside an MPA, and the variance around that increase. We use those analytical results as the basis for statistical regression models that we fit to time series of observations inside and outside of five MPAs. We accounted for the variance introduced by year-to-year fluctuations in larval recruitment using a lagged index of recruitment from standardized larval collectors deployed at each site. Including that index term in regression models revealed that – with few exceptions – all short term ‘trends’ in abundance we explained primarily by the timing and magnitude of recruit pulses. However, the deterministic effects of protection inside MPAs was still evident in the population size structure: fish were getting larger, and the population was approaching the unfished size distribution. These approaches can help guide short-term management decisions in an uncertain and highly variable environment.