Variation in top-down and bottom-up control of marine bivalves at differing spatial scales

At large spatial scales (>1 km), the forces structuring marine communities are diverse and include oceanographic, geological, and human processes, as well as availability of regional species pools. Few studies in marine and estuarine systems have examined the joint effects of predation (top-down) and food availability (bottom-up) in controlling populations at these scales. We compared the relative influence of top-down and bottom-up factors to population abundance of the infaunal clam Macoma balthica in two ecosystems differing in spatial extent by an order of magnitude: the York River, 50 km long, and the Rhode River, 5 km long. In both rivers, two habitat types could be differentiated by either high or low density of Macoma. In the York, Macoma abundance was tenfold greater upriver than downriver, predator abundance was greater in the high clam density habitat (i.e. upriver), and predation was only evident where clams were dense (i.e. upriver, where food availability for predatory crabs was elevated). In this large-scale system, predators were apparently unable to travel between upriver and downriver habitats where food availability differed. Moreover, sedimentary carbon (i.e. food for deposit-feeding clams) was significantly greater in the high clam-density habitat; this is consistent with the hypothesis that bottom-up factors principally dictate clam density, and moreover, that clam density drives predator density. In the Rhode River, a smaller-scale system, predators could easily move among habitats differing in clam density. Macoma abundance and predator abundance were both threefold greater in mud than sand. In contrast, predation intensity was higher in sand than mud, consistent with a hypothesis of top-down control. As in the large-scale system, sedimentary carbon was significantly greater in the high clamdensity habitat (i.e. mud). In this smaller scale system, both top-down and bottom-up factors contributed to habitat-specific patterns in clam population abundance. Hence, in this marine benthic ecosystem, bottom-up control was important at both large and small spatial scales, whereas top-down control was only important at the smaller spatial scale.