COS 57-6 - A unified meta-ecosystem dynamics model: Integrating herbivore-plant subwebs with the intermittent upwelling hypothesis

Abstract: Determining the relative influence of biotic and abiotic processes in structuring communities at local to large spatial scales is best understood using a biogeographic comparative-experimental approach. In exploring these processes in rocky intertidal ecosystems, the Intermittent Upwelling Hypothesis (IUH) suggests that community dynamics vary unimodally along an upwelling-based productivity gradient. That is, ecological subsidy rates and species interactions strengths are hypothesized to increase from persistently downwelled sites to intermittently upwelled sites (where upwelling regularly alternates with downwelling) then decrease towards persistently upwelled sites. Evidence consistent with the IUH comes from the sessile invertebrate-predator subweb in certain rocky intertidal communities, but if herbivore-macrophyte subwebs follow this pattern is unknown.

Here we ask - Are herbivore-macrophyte and predator-sessile invertebrate subwebs both structured as predicted by the Intermittent Upwelling Hypothesis? We hypothesize that, unlike predator-sessile invertebrates subwebs, macrophyte/herbivore subwebs are driven by a new construct, the Grazing-Weakening Hypothesis (GWH). This states that macrophyte-herbivore interactions weaken monotonically with increasing nutrients, with strong (weak) herbivory and low (high) macrophyte productivity at low (high) nutrients.

We used a combined experiment to explore local-to-large spatial scale dynamics of both subwebs. Factorial experiments tested community development with and without herbivores and predators at ten sites differing in upwelling regime for >2 years (26-29 months). To capture the full range of persistent upwelling to persistent downwelling, the study was done in coastal ecosystems of the northern California Current and New Zealand Shelf Large Marine Ecosystems.

For the macrophyte-herbivore subweb, results were consistent with the GWH: herbivory was strongest at low nutrient, unproductive downwelled sites and declined with increasing nutrient inputs and macrophyte abundance. As expected, results for the sessile invertebrate-predator subweb were consistent with the IUH: predator effect size was unimodally related to upwelling. However, overall, herbivory explained more variation in community structure than did predation, especially in New Zealand, where grazers (limpets) were larger than in Oregon and predators (sea stars) were less abundant.

Combining herbivory-macrophyte and predator-sessile invertebrate results yields the unified Meta-Ecosystem Dynamics Model (MEcoDynaMo). With increasing upwelling strength, sessile invertebrates vary unimodally (as in the IUH), while macrophytes increase, then decrease due to negative interactions with the sessile invertebrate subweb, then increase as nutrients steadily increase thereby reducing herbivory intensity. If herbivores are absent, macrophyte abundance varies inversely to that in their presence, declining from high to intermediate levels with increasing upwelling. Thus, macrophyte-phytoplankton-sessile invertebrate interactions link the two subwebs.