Spatiotemporal drivers of harbour porpoise occurrence and foraging activity
GAMs were run to compare the temporal trends and environmental predictors of harbour porpoise occurrence between C and F-PODs. The best models for detection positive minutes per hour (DPM/h) from both PODs retained all explanatory variables (all variables significant in both the C-POD and the F-POD models). Effect sizes of variables retained within the models were remarkably consistent between the C-POD and F-POD models, with the exception of noise, where this showed a considerably higher effect within the C-POD model (Table 1).
Similar temporal trends were highlighted by the C-POD and F-POD models in terms of harbour porpoise DPM/h per month and throughout the diel cycle (Figure 3). Both tidal range and time to high tide were significant, reflecting similar trends from both devices with an increase in detections during the ebb tide and during tidal ranges associated with spring tides (Figure 3), however, both variables had relatively small effect sizes within the model (Table 1). The C-POD model highlighted a much greater effect of noise on harbour porpoise detections, with a clear decrease in porpoise detections at higher noise levels, occurring at a much lower noise threshold for the C-POD than the F-POD (Figure 3). Harbour porpoise detections were found to decrease with increasing water temperature for both devices (Figure 3).
In contrast to the similar temporal patterns shown via the occurrence, strong differences occurred between models of foraging buzzes between C-PODS and F-PODS using buzz clicks per hour (BPH) as the response variable.
The C-POD foraging model did not retain any temporal variables despite them remaining the most influential covariates within the F-POD foraging model, likely due to a much-reduced sample size of identified feeding buzzes by the C-POD. Neither time difference to high tide or tidal range were found to influence harbour porpoise foraging behaviour in either C-POD or F-POD models. The F-POD model suggested a decrease in foraging buzzes between July and September, and an increase in foraging buzzes detected during the day (Figure 4). Similar to the detection model, foraging activity decreased with increasing water temperature, with highest buzz detections around 10° C. In contrast to the C-POD model, the F-POD foraging model found no significant effect of noise on foraging activity (Table 2).