Introduction: What conservation problem or question does your study address?River dolphins are expected to be most vulnerable during the low water season when the flow decreases, fishing activity intensifies and fewer habitat patches are available for foraging. Robust information on factors affecting their distribution during this period could provide insights for prioritizing conservation efforts. We tested the effect of different kind of natural hydrological variables on the distribution of the Ganges river dolphin in Karnali River.
Methods: What were the main research methods you used? A total of 65 segments of each 1 km long were sampled in the Karnali river of Nepal and India to record the presence or absence of the Ganges river dolphin over one month December, 2013. In total four observers recorded sighting of dolphins from the forward, rear and left/right sides of the mechanized boat that follows upstream to downstream with constant speed of 5-7 km/h. The variables like depth and width of river channel were also measured in the middle of each segment. We maintained a fifteen minute search effort at favorable dolphin microhabitats like river confluences where more dolphins tend to congregate to reduce the sighting bias. We used logistic regression models to test the effect of hydrological covariates on the occurrence of the Ganges river dolphin. The method of Akaike information criterion (AIC) was used to select the best model from a set of logistic regressions that linked Depth, Width and Distance to upstream.
Results: What are your most important results? The best model included Distance to upstream and Width of river as covariates (AIC = 51.69). We found a positive relationship between Distance to upstream and the probability of presence of the dolphin (b = 0.16± 0.06; p = 0.006) i.e. probability of dolphin’s presence was found to be increased with the increase in distance from the upstream segments of the river. A negative relationship between Width and the probability of presence of the dolphin (c = - 0.01 ± 0.00; p = 0.023).
Discussion: What are your important discussion points and what is the relevance of your results to conservation (if any)?Our results show that the probability of dolphin presence increases in downstream segments, which matches with lower water velocity and more fish biomass in downstream sites. A negative relationship between river width and the probability of dolphin presence might be due to imperfect detection of dolphins. This demonstrates the need for accounting imperfect detection by occupancy models to effectively monitor the river dolphin populations.