Increasingly detailed insights in animal behaviours using continuous on-board processing of accelerometer data

Abstract

Background: Studies of animal behaviour, ecology and physiology are continuously benefitting from progressing biologging techniques, including the collection of accelerometer data to infer animal behaviours and energy expenditure. In one of the most recent technological advances in this space, on-board processing of raw accelerometer data into animal behaviours proves highly energy-, weight- and cost-efficient allowing for continuous behavioural data collection in addition to regular positional data in a wide range of animal tracking studies.

Methods: We implemented this latest development in collecting continuous behaviour records from 6 Pacific Black Ducks Anas superciliosa to evaluate some of this novel technique's potential advantages over tracking studies lacking behavioural data or recording accelerometer data intermittently only. We (i) compared the discrepancy of time-activity budgets between continuous records and behaviours sampled with different intervals, (ii) compared total daily distance flown using hourly GPS fixes with and without additional behavioural data and (iii) explored how behaviour records can provide additional insights for animal home range studies.

Results: Using a total of 690 days of behaviour records across six individual ducks distinguishing eight different behaviours, we illustrated the improvement that is obtained in time-activity budget accuracy if continuous rather than interval-sampled accelerometer data is used. Notably, for rare behaviours such as flying and running, error ratios > 1 were common when sampling intervals exceeded 10 min. Using 72 days of hourly GPS fixes in combination with continuous behaviour records over the same period in one individual duck, we showed behaviour-based daily distance estimation is significantly higher (up to 540%) than the distance calculated from hourly sampled GPS fixes. Also, with the same 72 days of data for one individual duck, we showed how this individual used specific sites within its entire home range to satisfy specific needs (e.g. roosting and foraging).

Conclusion: We showed that by using trackers allowing for continuous recording of animal behaviour, substantial improvements in the estimation of time-activity budgets and daily traveling distances can be made. With integrating behaviour into home-range estimation we also highlight that this novel tracking technique may not only improve estimations but also open new avenues in animal behaviour research, importantly improving our knowledge of an animal's state while it is roaming the landscape.

Keywords: Daily distance; Home range; Machine learning; Time-activity budget; Wildlife tracking.

Fig. 1

Continuous ethograms of six Pacific black ducks with date on the y-axis and time of day on the x-axis. Grey curves depict local sunrise and sunset times. For this representation, only the behaviour during the first 2 s of each half minute is used, the remaining 28 s being discarded

Fig. 2

Time-activity budget estimation differences between down sampled behaviour records and continuous behaviour records (using a total of 690 days of continuous behaviour records across 6 ducks). (A) Density plots of differences for 9 sampling intervals varying from 10s to 60 min of 8 behaviour types ranked by their overall percentages. The three vertical lines in each density plot represent 25%, 50%, and 75% percentiles. (B) Daily total time in the rarest behaviour recorded: flying. For each duck day the total time spent flying was calculated from continuous behaviour records (marked with black dots) and data down sampled with 10 min interval (grey line without dots). The blue shades indicate the difference between the two calculations. Vertical grey lines indicate the separations between ducks (e.g. data between the line marked with D5099_2 and the line marked with D5210 is the data of D5210). (C) As B but for the most frequently recorded resting behaviour

Fig. 3

Using the same data as Fig. 1, error ratio (i.e. uncertainty of time-activity budget estimation through permutations with 9 sampling intervals: 10 s, 30 s, 1 min, 3 min, 5 min, 10 min, 20 min, 30 and 60 min) as a function of sampling interval and daily mean behaviour proportion (p) for the eight different behaviours. Colored dots indicate values out of permutations and colored diamonds indicate values calculated using Eq. 2

Fig. 4

Measures of dependency between behavioural observations (cf. Eq. 3) as a function of sampling interval for all eight behaviours that were distinguished across the 690 days in the six ducks. Dependency was calculated using Eq. 3

Fig. 5

Continuous behaviour records and GPS fixes of duck 5210 from 21/11/2020 to 31/01/2021 (i.e. 68 days, excluding 4 days with failed fixes) for daily distance estimation. Scatterplot showing that the daily distances travelled based on continuous behavioural records are significantly larger than distances calculated from hourly GPS fixes (mean GPS-based distance = 5183 m, mean behaviour-based distance = 7665 m). The diagonal represents a 1:1 ratio. Insets [1] and [2] are schematic scenarios for two GPS fixes (i.e. with one hour difference in this study), where blue lines indicate GPS-based distances (i.e. Haversine distance between two fixes) and red dotted lines indicate possible true flight routes if behaviour-based distance is larger than GPS-based distance

Fig. 6

Home range estimation of duck 5210 based on energy expenditure approximation, behaviours and geographical and temporal information. (a). Map depicting duck 5210’s whereabouts between 21/11/2020 and 31/01/2021 (i.e. 72 days). A total of 18, 30 × 30 m cells on this map could be identified in which duck 5210 spent more than 24 h individually. Based on geographical proximity, these 18 cells could be allocated to four key sites (cells 12 and 13 belong to one site). The four insets depict enlargements of these 18 cells. Within each inset, the base map is on the left and on the right the colour of the number in each cell identifying the site and the background colour the energy percentage (see text). (b). Hierarchical clustering of the 18 cells based on (c) the behaviour (i.e. proportional distribution of behaviour) of duck 5210 in each cell. (d). Total time (over the 72 days) in each cell as a function of time of day. Colors of histograms represent hierarchical clustering based on hour percentage. (e). Total time in each cell as a function of date