The round goby (Neogobius melanostomus) invaded Oneida Lake in 2014, however, the extent of their spread is currently unknown. Elsewhere, gobies have both positive and negative effects on native fish populations, acting as a source of food for larger species and competing with smaller fish. Round gobies are known to prey on fish larvae and eggs, which can decrease recruitment of both small and large native species. However, the current use of bottom trawling to survey fish populations in Oneida Lake may not accurately estimate the goby population as gobies are known to reside within complex substrata and in shallow water depths where trawls cannot be used to survey. We used video imagery to survey gobies in the lake across habitats and compared video-based densities with trawl data. To do this, a GoPro camera was attached to an underwater frame and used to record two 10-minute videos at sites surrounding Shackelton Point from June 22, 2017 to July 12, 2017. For each of the 79 videos recorded, 7 screenshots were taken, and a count of round gobies and their lengths was recorded in ImageJ.
Results/Conclusions
Shallow (0-4m), rocky areas had the highest density (6.8 gobies/m2, SE 7.5) and biomass (24.76 g/m2, SE 11.6). Shallow, sandy areas had a density of 2.7 gobies/m2, SE 1.8, and biomass of 10.2 g/m2, SE 1.3. In deeper sandy areas, density was 1.2 gobies/m2, SE 0.8, and biomass of 6.7 g/m2, SE 1.8. No gobies were present in mucky substrate. Trawl data also showed few gobies in mucky areas. Accounting for habitat proportions within the lake, we calculated a lakewide density of 0.98 gobies/m2 and biomass of 3.71 g/m2. These estimates are much higher than those using bottom trawls (0.002 gobies/m2). Bottom trawling sampled deep waters and mucky substrata, missing the habitats where majority of round gobies reside. Given these high goby densities in shallow waters, we expect profound impacts on the fish community of this important lake. The use of low cost video images proved to be a useful method for estimating goby abundance, but the method could be improved by decreasing the distance of the camera above the substrate, and adjusting the timing of the survey based on season to accommodate breeding times.