Assessing the post-release effects of capture, handling and placement of satellite telemetry devices on narwhal (Monodon monoceros) movement behaviour

Abstract

Animal-borne telemetry devices have become a popular and valuable means for studying the cryptic lives of cetaceans. Evaluating the effect of capture, handling and tagging procedures remains largely unassessed across species. Here, we examine the effect of capture, handling and tagging activities on an iconic Arctic cetacean, the narwhal (Monodon monoceros), which has previously been shown to exhibit an extreme response to extended capture and handling. Using accelerometry-derived metrics of behaviour, including activity level, energy expenditure and swimming activity, we quantify the post-release responses and time to recovery of 19 individuals following capture and tagging activities considering the intrinsic covariates of sex and individual size and the extrinsic covariates of handling time and presence of a 'bolt-on' satellite telemetry device. From accelerometer-derived behaviour, most narwhals appeared to return to mean baseline behaviour (recovery) within 24 hours after release, which was supported by longer-term measures of diving data. None of the covariates measured, however, had an effect on the time individuals took to recover following release. Using generalized additive models to describe changes in behaviour over time, we found handling time to be a significant predictor of activity levels, energy expenditure and swimming behaviour following release. Individuals held for the longest period (>40 min) were found to display the largest effect in behaviour immediately following release with respect to swimming behaviour and activity levels. We also found some support for relationships between activity levels, energy expenditure and swimming activity and two other covariates: sex and the attachment of a bolt-on configuration satellite tags. Our results indicate that narwhals recover relatively quickly following capture, handling and tagging procedures, but we suggest that researchers should minimize handling time and further investigation is needed on how to mitigate potential effects of bolt-on satellite tags in these sensitive species.

Keywords: Arctic; accelerometry; behaviour; cetacean; handling response.

Figure 1

Images showing net deployment anchored perpendicular to shore (A) in Tremblay Sound (72°21.389 N, −81°05.855 W), an example of attachment of the bolt-on configuration (3 pins) for a TDR-10 satellite tag, and accompanying (B) Maritime Biologgers (MBL) deployment package with timed release mechanism attached to pins or (C) Acousonde unit attached via both a suction cup mounting and little Leonardo releases held across the spider wires holding the TDR-10 satellite tag to the dorsal ridge and (D) a Daily Diary (TDR-DD) tag attached with a suction cup under a bolt-on satellite tag (2 pins).

Figure 2

Example of behaviour across the full recording period (~ 22 hours) for a 3.2 m male narwhal 17–20, including the period immediately after release. Behavioural metrics include (a) activity levels (norm of jerk), (b) energy expenditure (sVeDBA) and (c) swimming activity (TBFreq), before the tag fell off just after 12:00 on the second day as indicated by the green arrows.

Figure 3

Plots of the interaction between time after release and sex as well as the additive effects relative to handling time, t_cap, for deviance associated with (A) activity levels and (B) energy expenditure. Changes in swimming behaviour (C) were best explained by the interaction of time after release and handling time. Modelled behavioural responses relative to the population mean (deviance = 0) is highlighted by the horizontal dashed line. Data for individuals that were held for the longest time (t_cap > 40 min) were only modelled to 50 hours as a result of limited accelerometry data. The spline calculated for male narwhal (B) was zeroed out as a result of the shrinkage penalty implemented to prevent overly complex relationships.

Figure 4

Mean population or across-individual hourly deviance with 95% confidence intervals (solid black and dashed lines, respectively) for accelerometer-derived behavioural metrics (A) activity levels, (B) energy expenditure, (C) swimming activity and (D, E) dive behaviour for the first 72 hours following release from capture/tagging events. Deviance was calculated by pooling hourly mean values for accelerometry data (36+ hours post-release, limited by measurement durations) and dive data (7–14 day hourly mean). Coloured lines represent loess-smoothed trend lines to further highlight trends in behaviour.

Figure 5

Summary of mean recovery time points (hours) after release as a function of handling time (t_cap: short, <30 min; medium, 30–40 min; long, >40 min). Thick black bars in the middle of the boxes indicate the median value (bounded by the interquartile range) for time to recovery for each of the handling durations.