Understanding the effect of multiple stressors using seagrass as a model system

Ecosystems around the world face an increasing frequency and intensity of human impacts, meaning that managing multiple co-occurring pressures is imperative. Unfortunately, our ability to predict ecosystem responses to multiple pressures is limited. Seagrasses are the foundation species of ecosystems that provide services such as carbon sequestration and improved water quality. Unfortunately, they are also heavily impacted by human activities. With an eye toward management, I use seagrass meadows as a model system to understand and predict the effects of multiple pressures across scales. First, I reconstructed area trends for 547 meadows. I found that one-fifth of the world's observed seagrass meadow area has been lost since the 1880s. However, losses were not consistent across time or space and only 10% of studies rigorously tested driver attributions, suggesting that local studies are important for informing relevant management actions. I then focused on one species, eelgrass (Zostera marina), to identify critical growth-related values that are pragmatic management targets and test whether these values change when pressures co-occur. I quantitatively reviewed studies of the effects of temperature and light on eelgrass performance. I found that two critical performance values, zero-growth rate and maximum growth rate, shifted across combinations of light and temperature values, suggesting that fixed thresholds are inappropriate targets. Finally, I used surveys of subtidal eelgrass meadows to examine the fine-scale (10s kms) variation in two metrics of eelgrass meadow health (shoot density and lesion prevalence) and relate this with variation in environmental conditions and human impacts. I found that environmental conditions were better predictors of eelgrass health than the human impacts considered (overwater structures and riparian modification), highlighting the importance of measuring conditions at the fine scale at which conservation and restoration efforts occur. As a whole, my research shows that although local context is important for seagrass management, there are generalisable patterns in how multiple stressors affect seagrass performance, which can be used to guide interventions.