Sea lamprey (Petromyzon marinus) habitat and population models in Michigan river networks: Understanding geomorphic context and boundaries.

Abstract

Rivers are hierarchical, heterogeneous products of climate and the landscapes through which they flow. This nature motivates two interesting and practical questions in river ecology: how are populations of riverine organisms and their habitat influenced by their geomorphic context, and how best should spatial boundaries for conservation and management be defined? I address both questions with respect to the sea lamprey (Petromyzon marinus), an invasive species in this region with a history of negative impacts on Great Lakes fisheries. High management costs (> $14 million/year) provide a strong practical motivation for developing a deeper understanding of sea lamprey ecology. In a first analysis, I built a series of regression models to predict larval sea lamprey habitat across the lower peninsula of Michigan. I used the geomorphological concept of process domains as a justification for using regression trees to both stratify the dataset and fit a regression model to each stratum. Trees identified useful partitions of the lower peninsula of Michigan and the importance of geomorphic influences on habitat distribution. In a second analysis, I developed an individual-based model (IBM) to explicitly test the hypothesis that river network structure influences the distribution of larval sea lamprey. In the model, patterns of larval aggregation varied strongly with network structure across a broad range of parameterizations. My results suggest that larval distribution across a watershed results from a system of emergent feedbacks shaped by the network structure of each watershed and the past migratory and spawning behavior of adults. In a final analysis, I expanded the IBM model to include heterogeneous distributions of larval habitat, and explored how habitat availability, network structure, and fish migratory behavior interact to shape larval distributions. Using this model I demonstrate how and when larval distribution could be influenced by both network structure and the distribution of larval habitat. My IBM results suggest that the watershed is an important functional boundary for sea lamprey. I argue for a view of each watershed as a unique place, where larval distribution depends on geomorphic structures that alter feedbacks between larval habitation and adult migratory choices.