Wildlife speed cameras: measuring animal travel speed and day range using camera traps

Abstract

Travel speed (average speed of travel while active) and day range (average speed over the daily activity cycle) are behavioural metrics that influence processes including energy use, foraging success, disease transmission and human-wildlife interactions, and which can therefore be applied to a range of questions in ecology and conservation. These metrics are usually derived from telemetry or direct observations. Here, we describe and validate an entirely new alternative approach, using camera traps recording passing animals to measure movement paths at very fine scale. Dividing the length of a passage by its duration gives a speed observation, and average travel speed is estimated by fitting size-biased probability distributions to a sample of speed observations. Day range is then estimated as the product of travel speed and activity level (proportion of time spent active), which can also be estimated from camera-trap data. We field tested the procedure with data from a survey of terrestrial mammals on Barro Colorado Island, Panama. Travel speeds and day ranges estimated for 12 species scaled positively with body mass, and were higher in faunivores than in herbivores, patterns that are consistent with those obtained using independent estimates derived from tracked individuals. Comparisons of our day range estimates with independent telemetry-based estimates for three species also showed very similar values in absolute terms. We conclude that these methods are accurate and ready to use for estimating travel speed and day range in wildlife. Key advantages of the methods are that they are non-invasive, and that measurements are made at very high resolution in time and space, yielding estimates that are comparable across species and studies. Combined with emerging techniques in computer vision, we anticipate that these methods will help to expand the range of species for which we can estimate movement rate in the wild.

The wild life of tick-borne pathogens

“…In essence, the sampling is then reduced to a line. A major advantage of this approach is that the capture rates obtain from this line only need to be corrected for EDD and day range (the distance that animals travel daily) to estimate density 186 , just as in line-transect estimation from indirect sign using the Formozov-MalyshevPereleshin formula 187 . Note that this is equivalent to an REM with a detection angle of zero 170 .…”

“…However, the speed with which animals move might affect encounter rate. Animals that move over a larger distance per day can do so by increasing the speed with which they move, or by increasing the time spent active 186 . There is however, to my knowledge, no information on the relationship between movement speed and the chance that an encountered tick will attach to the host.…”

Guidance on estimation of wild boar population abundance and density: methods, challenges, possibilities

“…The DR is the parameter most costly and time-consuming to be measured, but it can be estimated from photo trapping data (Rowcliffe et al 2016) rather than relying on fine resolution GPS or radiotracking data. However, comparative studies to finely describe the distances travelled among different regions, habitats and seasons are required to evaluate the potential practicability of REM.…”

Estimating day range from camera‐trap data: the animals’ behaviour as a key parameter

“…In addition, Rowcliffe et al . () recently developed an approach to estimate the DR based solely on the information derived from CT; in which, DR is estimated as the product of travel speed (i.e. average speed while active) and activity level (i.e.…”

“…For this purpose, based on the procedure described by Rowcliffe et al . () and considering wild boar as model species, we hypothesized that for those species with different behaviour (in terms of movement speed), it may be necessary to weight the speed and activity values to estimate a DR value. We assessed the new procedure by comparing DR results with those from other sources of data and procedures.…”