Factors Influencing Distribution and Survival of Migratory Fishes Following Multiple Low-Head Dam Removals on a North Carolina River

Migratory fish species are assumed to benefit from dam removals that restore connectivity and access to upstream habitat, but few studies have evaluated this assumption. Therefore, I assessed factors influencing distribution and survival of migratory fishes in the springs of 2007 through 2010 on the Little River, North Carolina, a tributary to the Neuse River with one partial and three complete dam removals. I tagged migratory fishes with passive integrated transponders (PIT) at a resistance board weir installed at a dam removal site (river kilometer (rkm) 56 in 2007, rkm 4 in 2008-2010) and followed migrations with upstream PIT antennas in 2008-2010. This gear proved very effective in low to moderate flows as thousands of fish were tagged and monitored, but less effective in high flows. Fish migrations were strongly influenced by river flow, with most movement occurring during freshets, high flow events following rain. Connectivity between reaches increased following dam removals, with use of restored habitat varying by species. For example, 24-31% of anadromous American shad Alosa sapidissima, 45-49% of resident gizzard shad Dorosoma cepedianum, and 4-11% of introduced flathead catfish Pylodictis olivaris passed the rkm 56 dam removal site. For these species, 17-28% did not pass the partially removed dam (rkm 8) while 20-39% remained downstream for more than a day before migrating upstream. This suggests the notched dam may impede or delay migrations, potentially limiting access to habitat while increasing energy expenditure and predation vulnerability. I further evaluated American shad, an ecologically and economically important species that has experienced prolonged population declines. During freshets, American shad weir captures increased, migrations were more extensive and faster, and diel activity shifted to daytime hours. Considerable weight loss occurred and displayed a positive relationship with thermal days (cumulative temperature during residence). Proportional weight loss was primarily less than 30% for males and 50% for females, indicating potential thresholds for survival to emigration. Spawning survival was low as estimates ranged from 0.07 to 0.17; no factors (e.g., sex, size, migrations) influencing survival were documented. Estimated American shad abundance increased from 2007 through 2009, but decreased in 2010. Flathead catfish were most common in May and in lower river reaches, increased annually from 2008 to 2010, and were documented consuming American shad. I estimated flathead catfish consumed 237 to 1,278 American shad (7-36% of estimated run size) in 2010. Lastly, I developed a linear spatial capture-recapture Cormack-Jolly-Seber model capable of analyzing the plethora of data collected at continuous monitoring stations (e.g., PIT antennas, telemetry receivers) and applicable to a variety of aquatic and terrestrial systems. I analyzed the model using a Bayesian framework and estimated survival and activity centers (i.e., mean location) and measures of uncertainty (i.e., credible intervals) in each time period. I demonstrated the model with 2010 American shad data. Estimated weekly survival (0.80) was consistent with a non-spatial Cormack-Jolly-Seber model. Activity centers and a movement parameter showed individuals migrated greater distances during higher flow conditions. This study provides strong support for efforts to restore currently inaccessible habitat through complete removal of derelict dams, substantially increases information on spawning American shad that could aid restoration efforts, and provides a model applicable for any linear array (e.g., rivers, shorelines, corridors), opening new opportunities to study demographic parameters, movement or migration patterns, and habitat use.