Sonar imaging surveys fill data gaps in forage fish populations in shallow estuarine tributaries

Forage fish are economically and ecologically important to many coastal marine ecosystems, but most forage fish species are not targeted in species-specific abundance surveys. In the Chesapeake Bay, Atlantic menhaden (Brevoortia tyrannus), an important forage species for many upper-trophic level consumers, are sampled with beach seines and midwater trawls. There is evidence that these traditional fish sampling gears are insufficient for observing menhaden abundance and distribution, which may lead to biased and inaccurate indices of abundance. We suggest this is due to omitting or under-sampling shallow estuarine creeks and nearshore areas. We tested this hypothesis in the Patuxent River, Maryland and its surrounding shallow-water creeks, which are representative of many estuarine tributaries with varied bathymetry. Sonar imaging surveys were conducted using Adaptive Resolution Imaging Sonars (ARIS 1800 and 3000) and a generalized additive model was fit to the resulting data to model forage fish density. Results indicated that menhaden and other forage fish were present at significantly higher densities in shallow-water creeks than in deeper-water areas. Further, many forage fish were observed in creek habitats frequently omitted by both seines and trawls, suggesting that these survey designs may result in biased forage fish indices of abundance. Therefore, we recommend current forage fish surveys in the Chesapeake Bay, especially those with the intent of developing juvenile menhaden indices of abundance, be supplemented with hydroacoustic surveys in shallow waters to improve accuracy and reduce bias.