Applying integral projection models to marine systems: A case study with the eastern and pacific oyster

Background/Question/Methods

Local populations are often structured by a variety of traits (e.g. size, age, gender, life stage) that determine rates of growth, mortality, and fecundity. These differences in demographic rates often have important consequences for population viability and management. Recently, integral projection models (IPMs) have become a popular method used to describe terrestrial populations structured by traits that vary continuously. However, very few studies have applied this modeling approach to populations in marine environments. Here, we develop IPMs to describe the population dynamics of two marine ecosystem engineers: the pacific oyster, Crassostrea gigas, and the eastern oyster, Crassostrea virginica. We use these models to compare the long-term population growth rate (λ), the stable size distribution and reproductive values, and their elasticities between both oyster species at multiple spatial locations. 

Results/Conclusions

We show that numerous considerations arise when applying an IPM to marine populations. First, many marine organisms demonstrate non-negative growth. Thus, unlike past applications of IPMs, growth cannot be treated as a normally distributed random walk. Instead, we model growth as a gamma random walk and demonstrate that this formulation is a biologically realistic way to successfully account for this growth pattern. Second, a limitation in marine systems is the inability to fully monitor pelagic larval survival. We demonstrate that IPM output for individual species and locations is highly sensitive to this value, but that qualitative differences between species and locations is relatively unaffected. We find significant differences exist between C. gigas and C. virginica, particularly with regard to the sensitivity of λ to changes in size-specific demographic rates. Preliminary results indicate that for C. gigas λ is most sensitive to changes in the growth and survival of the smallest oysters, while for C. virginica λ is sensitive to the growth and survival of a much wider range of oyster sizes. Management actions such as stock enhancement may be most effective for C. gigas, whereas no-take marine protected areas may be most effective for C. virginica.