Multiple stressors in estuaries can cause declines in native species and impairment of ecosystem goods and services. In contrast, one stressor -the introduction of non-native speciesactually leads to higher local richness. We examined the changes in ecosystem function associated with introductions into Willapa Bay, Washington, USA, a relatively undeveloped estuary with 45 documented exotic marine species. The replacement of native oysters by 2 new bivalve species has increased secondary production of harvested suspension feeders by 250% over peak historic values (3.3 × 10 5 vs. 0.9 × 10 5 kg dry wt yr -1), based on >150 yr of records of harvested biomass. Key aspects of aquaculture -particularly planted area -have remained constant over time, so we attribute much of the altered secondary production to higher growth rates of non-native species. The addition of 2 tracheophytes has increased primary production on the tideflats by > 50% (5.3 × 10 7 vs. 3.5 × 10 7 kg dry wt yr -1 ), which we calculated by scaling up local measurements of plant growth to the total area occupied by each species. These changes in production are also associated with altered detritus, water filtration, and biogenic habitat. Because other stressors are largely absent from Willapa Bay, the addition of particular exotic species has dramatically enhanced system production, while fundamentally reshaping the ecological character of the estuary. These strong ecological impacts of introduced species have rarely been measured at whole-ecosystem scales, and they occur in part because new species occupy habitats where similar native species were not present.

We applied Qualitative Network Models (QNMs) to evaluate the potential community effects of ocean acidification (OA) in a major shellfish-producing estuary (Willapa Bay, Washington). QNMs are well-suited to data-limited systems and only require information on the sign (+, −, 0) of the interactions between species. We examined qualitative predictions of community responses to 13 different OA scenarios that corresponded to 3 broad categories of hypothesized OA effects: (1) increased primary productivity, (2) reductions in bivalve populations, and (3) enhanced predation interactions between bivalves and their crab and gastropod predators. The cultivated bivalve Manila clam tended to respond negatively across scenarios, while primary producers (phytoplankton and eelgrasses) and Chinook salmon tended to respond positively. Tradeoffs between species were also assessed: Manila clam and Pacific oyster were predicted to decrease and increase, respectively, when direct OA effects were limited to eelgrasses and the reverse occurred when phytoplankton alone was stimulated by OA. We analyzed the QNMs to identify key linkages that influenced the sign outcome of community members and might therefore warrant future quantitative study. QNMs may be particularly relevant to researchers as a simple method for identifying conditions under which the sign response of species to OA, as inferred from single-species OA experiments, will likely hold in the wild. Given data limitations in most systems, QNMs are a practical alternative or complement to data-intensive quantitative approaches and may help accelerate our understanding of the community-wide effects of OA in marine systems.