Check out recently accepted abstract for a talk to be given at the Southern Division of the American Fisheries Society annual meeting held at the Moody Gardens Hotel and Conference Center on Galveston Island, Texas from January 24-27, 2019.

Occupancy of four gar species in river-floodplain habitats: is there evidence for competitive exclusion and habitat mediated co-existence?

Authors: D. A. Schumann, M. E. Colvin, L. E. Miranda, D. T. Farrand

Anecdotal evidence suggests that co-existing gar species segregate among available habitats and may competitively exclude related species. However, the magnitude of these interactions among gar species are largely unknown. Interactions are potentially greatest between the closely related Longnose Gar (Lepisosteus osseus) and Shortnose Gar (Lepisosteus platostomus), but the ubiquitous distribution of Spotted Gar (Lepisosteus oculatus) are thought to be less influenced by other species. We used an existing dataset to evaluate co-occurrence patterns of four gar species in 62 floodplain habitats associated with the lower Mississippi River and its larger tributaries. Samples were collected at each floodplain or oxbow lake during the summer or fall from 2006-2012 to describe the fish assemblages using spatially-replicated boat electrofishing transects (i.e., 2–16 transects per lake). Spotted Gar were the most abundant and commonly encountered species (~78% of transects). Shortnose Gar (~17%) and Longnose Gar (~12%) were more rarely captured in the transects. No Alligator Gar (Atractosteus spatula) were captured in >600 electrofishing transects. Spotted Gar were found in all sampled floodplain habitats, whereas Shortnose Gar (52%) and Longnose Gar (39%) were captured in fewer systems. We estimated the occurrence probabilities of each species and the probability that each species is present conditional of the presence of other species while accounting for imperfect detection using occupancy modeling. Detection probability was 0.79 for Spotted Gar, 0.30 for Shortnose Gar, and 0.28 for Longnose Gar. Occupancy rate estimates for Shortnose Gar were 0.52 and for Longnose Gar were 0.39. The ratio of model estimated occurrence and co-occurrence of the gars identified weak species interactions with Spotted Gar. Strong interactions between Shortnose and Longnose gar suggest that these species typically co-occur in floodplain lakes. By describing habitat-wide environmental features, we identified the conditions under which these species co-occur which will inform future species distributional modeling efforts.