Coastal watersheds provide valuable ecosystem goods and services to society. Both anthropogenic and natural changes may alter the ability of natural systems to provide these goods and services. To maintain resilient ecosystems in the face of future change, we need to understand their structure and how they function. One indicator of ecosystem function is species diversity. Substantial evidence exists to suggest that more diverse systems can be more resilient to ecosystem alterations. Species diversity can be affected significantly by both human and natural ecosystem engineers. The beaver (Castor canadensis) is a renowned ecosystem engineer. Their dams can introduce discontinuities in the river network by creating barriers to flow and fish movement, and that may alter stream habitat by creating pools upstream and riffles or runs downstream. Depending on the natural flow regime, and fish mobility and life history, beaver dams have the potential to either create discontinuities that reduce natural fish-species diversity by fragmenting habitat or to increase diversity by establishing a mosaic of patches that include refuges from competition and predation. Habitat alteration could also have a negative or positive impact on fish diversity by changing species-specific habitat. The impact may not be equal across species. Specialized species (migrating or flow-dependent species) may be more affected by ecosystem alterations than generalist species. Furthermore, beaver dams are dynamic discontinuities that, at different discharge rates, may result in different effects on fish diversity. For this reason, at high or variable discharges, beaver dams may have little effect on fish communities through either fragmentation or habitat change. Beaver populations are increasing in many northeastern watersheds including the Ipswich River, MA, part of the Plum Island Long Term Ecological Research Site. To examine how beaver dams are affecting fish species diversity at a local scale, fish were sampled using minnow traps, hoop nets, and electrofishing immediately upstream (30 m) and downstream (30 m) of 14 beaver dams and 9 control locations within a 20 km tributary in the Ipswich River watershed in July and August, 2009. Habitat at and adjacent to dams was also quantified. Of these 14 dams, 10 dams were relatively impermeable, given the existing flow regime, and changed the habitat from pool upstream to run or riffle downstream. Four of these dams were permeable and did not change the habitat (e.g., run upstream, run below stream). To examine the temporally dynamic attributes of dams, parameters of a subset of 5 dams were measured at various rates of discharge. To quantify passive-dispersal potential, pseudo-fish (wooden dowels of three different sizes) were released 10 m upstream of 5 dams, and the success or failure of passage of the dowels from upstream to downstream were quantified at each dam at 5 min and 24 hrs. These local data on beaver dams, fish communities, and habitat are also being examined at a landscape scale. Together these multi-scale data on how beaver dams affect fish communities, alter instream fish habitat, and are affected by discharge, will provide insights into the structure and function of coastal ecosystems. In addition, this information can provide guidance to decision makers on environmental regulations of watersheds such as dam removal, water allocation, and responses to global climate change.