Hydrologic Factors Associated with Detection and Occupancy of Riparian Areas By Asian Swamp Eel in the Chattahoochee River System, Georgia

Asian swamp eels (Monopterus albus) were recently discovered in backwater marsh areas of Bull Sluice Lake, an impoundment of the Chattahoochee River, Georgia.  This introduced population of M. albus has been notoriously difficult to sample and occupies a dynamic hydrologic environment influenced by releases for hydropower production.  We used occupancy modeling to estimate detection and occupancy probabilities under variable habitat and sampling conditions.  Abiotic and biotic model covariates were collected at sampling locations and incorporated in a suite of 4 detection and 27 non-collinear occupancy models.  Top models included hydrologic covariates for both detection and occupancy.  Mean detection probability was 0.174 ± 0.039 and increased with water depth and temperature (p < 0.05).  The two sub-global models of occupancy showed no evidence of lack-of-fit (p > 0.42), and permutations of these models showed probability was influenced by vegetation, depth, substrate, distance from invasion point, and mean and variance of water temperature.  Geostatistical interpolation methods were used to estimate covariates in areas not sampled, which allowed occupancy probabilities to be back-transformed from model averages to the entire riparian area of the study site. This study demonstrates the importance of considering hydrologic factors when determining detection and occupancy probability of introduced species.