Large-scale culvert replacement programs could benefit migratory fish populations by reconnecting reproductive and foraging habitats in fragmented watersheds. The objectives of this study were to: (1) identify stream and culvert characteristics contributing to fish passage barriers within an Appalachian watershed, U.S.A.; (2) quantify the total amount of Brook trout (Salvelinus fontinalis) reproductive habitat isolated above culverts; and (3) use an ecological currency to identify culvert replacement priorities and stream mitigation credit opportunities. We surveyed 120 state-owned culverts and used a fish passage assessment filter to determine the ''passability'' of each culvert. We then constructed a geographic information system stream network model to quantify the amount of trout reproductive habitat isolated by culverts. Ninetyseven percent of surveyed culverts were classified as obstacles or complete barriers to trout dispersal. Culvert impassability was higher in small streams with slopes exceeding 3-5%, suggesting a direct relationship between slope and impassability. Thirty-three percent of Brook trout reproductive habitat, representing over 200 km of stream, was isolated by culverts. This is a conservative estimate, because we did not survey privately or federally owned culverts. The top 20 prioritized culverts accounted for nearly half of the habitat loss. Our results indicate that standard culvert designs placed in streams with slopes exceeding 5% consistently produce trout dispersal barriers and should be avoided during new road construction. The process developed here provides an efficient method for identifying culvert replacement priorities and may be used to maximize watershed scale benefits of stream restoration.

Culvert Inventory and Effects on Fish Assemblages in a Central Appalachian Watershed Ira Poplar-Jeffers We surveyed 120 state-owned culverts within the upper Cheat River basin, West Virginia, and used a fish passage evaluation filter to quantify passage barriers. We quantified the length of stream and weighted potential brook trout recruitment area isolated by each culvert. We also sampled fish communities upstream and downstream of 24 culverts to quantify the impacts of culvert-related isolation on fish assemblages. We found 69 % of culverts to be completely impassable. We determined that 17 % of the stream length and 23 % of weighted potential recruitment area in the study area were isolated by impassable culverts. We found that impassable culverts are more likely to occur on small, steep, streams. We found that species richness and blacknose dace densities were negatively impacted upstream of culverts and culverts are more likely to impact species that are limited to larger streams, although culvertrelated impacts are more pronounced on small streams. study area) and 130 km of WPRA (23 % of study area) were isolated by impassable culverts. We found that Shavers Fork section of the upper Cheat basin had the highest extent and intensity of habitat isolation and we suggest that initial restoration efforts be concentrated in this watershed. The upper Cheat River basin is characterized by a high proportion of barrier culverts that cause discontinuities within the stream network. In addition to other factors, this problem has significant negative impacts on stream ecosystems by isolating small streams from larger mainstem rivers.