Ecosystem engineering by eelgrass (Zostera marina) leads to population feedbacks in certain environmental contexts.

摘要

Ecosystem engineers are organisms that modify physical conditions in ways that affect other organisms or conspecifics. When engineers affect conspecifics, it can create a feedback loop between the engineer and environment, potentially causing nonlinear population dynamics. Unfortunately, our mechanistic understanding of how engineer-environment feedbacks create nonlinear dynamics is limited to theoretical and mathematical models. Thus, I empirically investigated the impacts of engineer-environment feedbacks in eelgrass (Zostera marina L.). Eelgrass is a convenient empirical system for studying ecosystem engineering, and it is of applied interest as an important conservation target in marine systems.;Rapid recovery of eelgrass following disturbance requires successful seedling establishment, since growth via branching covers only small distances. I found that seedlings occurred at sites in Washington state with higher sediment organic content, an indicator of calmer hydrodynamic environments. These results raised the question: Could eelgrass act as an engineer to mitigate harsh hydrodynamic conditions, thus facilitating seedling growth and potentially leading to nonlinear dynamics? I addressed this question by performing quantitative experiments in Willapa Bay, WA to determine how eelgrass impacts the recruitment of its own seedlings and recovery dynamics by engineering its hydrodynamic environment. In one of the first demonstrations of a gradient in engineering effects, relative water flow monotonically decreased as eelgrass shoot density increased, and at a consistent density, eelgrass decreased water flow at all locations across a gradient in hydrodynamic conditions. This engineering generated a feedback by modifying the performance of early life history stages of eelgrass. In conditions of higher hydrodynamic stress, an adult canopy facilitated seedling recruitment. Furthermore, eelgrass recruitment following experimental disturbance was density-dependent. Interestingly, in calmer hydrodynamic conditions and at higher eelgrass shoot densities, eelgrass had a negative effect, likely due to intraspecific competition. The context- and density-dependent feedbacks led to delayed recovery following experimental disturbance that removed eelgrass shoots and rhizomes, suggesting alternative stable states could emerge under slightly higher disturbance regimes. This mechanistic understanding of an ecosystem engineer clarifies that the abiotic environment underpins organism performance but may also be a function of organism abundance, thus dynamically coupling the abiotic and biotic components of an ecosystem.