MEPS 547:233-246 (2016)  -  DOI: https://doi.org/10.3354/meps11648

ABSTRACT: We opportunistically assessed the relative abundance of leatherback sea turtles Dermochelys coriacea over 14 yr in a foraging area off the coast of Nova Scotia, Canada. We conducted haphazard, unmarked, non-linear transects (HUNTs) for leatherbacks from a vessel with an observation tower, recording the time and location of all sightings concurrent with opportunistic sampling of turtles in support of other research objectives. We calculated the sightings per unit effort (SPUE) and standardized it using both a generalized linear model (GLM) and a generalized additive model (GAM) accounting for variables that may have affected detectability. During the monitoring period, a total of 179 HUNTs occurred and 564 leatherbacks were sighted, with annual average nominal SPUE (leatherbacks h^-1 ± SE) fluctuating from a low of 0.02 ± 0.02 in 2005 to a high of 0.69 ± 0.15 in 2014. Both standardization approaches indicated no significant linear effect of year (GLM: p = 0.16, GAM: p = 0.17), suggesting that although the relative abundance of leatherbacks at this field site exhibits high inter-annual variability, it is likely stable. We present preliminary estimates of absolute abundance in years where a spatial component provided by vessel GPS tracks in a subset of HUNTs allowed for the use of spatial modelling techniques. These data suggest Canadian waters may have the highest density of foraging leatherbacks reported across the species’ range, providing further evidence of the importance of this region to leatherback conservation.

KEY WORDS: Abundance · Trends · Platform of opportunity · Sea turtle · CPUE standardization · Spatial models