Mechanistic simulations of kelp populations in a dynamic landscape of light, temperature, and winter storms

Abstract

Kelp forests are widely distributed across the coastal ocean, support high levels of biodiversity and primary productivity, and underpin a range of ecosystem services. Laminaria hyperborea is a forest-forming kelp species in the Northeast Atlantic that alters the local environment, providing biogenic structure for a diversity of associated organisms. Populations are strongly affected by light availability, temperature, and storm-related disturbance. We constructed a stage-based, two-season model of L. hyperborea populations along the coast of Great Britain and Ireland to predict biomass across a range of depths, drawing on extensive surveys and data from the literature. Population dynamics were driven by wave exposure, historic winter storm intensity, and simulated interannual variation in temperature and depth-attenuated light intensity, with density-dependent competition for light and space. High biomass was predicted in shallow depths across the domain on suitable substrate, with populations extending deeper in the north and west where light penetration was greater. Detritus production was heavily skewed across years, particularly at greater depths, with 10 % of years comprising more than 50 % of detritus on average below 10 m depth. Annual fluctuations in light and storm intensity produced opposing population oscillations with a ∼6-year period persisting for up to a decade but diminishing sharply with depth. Interannual variation in temperature had minimal impact. Biomass was most sensitive to survival and settlement rates, with negligible sensitivity to individual growth rates. This model highlights the need for an improved understanding of canopy and subcanopy mortality, particularly regarding increasingly frequent heatwaves. Estimations of kelp forest contributions to carbon sequestration should consider the high variability among years or risk underestimating the potential value of kelp forests. Process-based simulations of populations with realistic spatiotemporal environmental variability are a valuable approach to forecasting biotic responses to an increasingly extreme climate.