Autocorrelated data in telemetry studies: time to independence and the problem of behavioural effects

Abstract

Independence of locational fixes, to reduce the effects of autocorrelation, is often deemed a prerequisite for estimation of home range size and utilization when using data derived from telemetric studies. Three methods of estimating times to independence using movement estimates, along with a statistical method of estimating the level of autocorrelation of locational data, were examined for two species of mammal. Attempts to achieve statistically independent data by subsampling resulted in severe redundancy in the data and significant underestimation of range size and rates of movement. Even a sample interval of one fix per week did not guarantee independence and also resulted in underestimation of range size despite range asymptotes being reached. It would appear that the correct strategy for the best possible estimation of range size and use from telemetry would be the repeated use of as short a sampling interval as is possible over an extended period of time. Statistical methods to measure levels of autocorrelation in locational data may be useful for comparing rates of range use between different populations of the same species or between species, as long as the same sample interval is used.

Females go where the food is: does the socio-ecological model explain variation in social organisation of solitary foragers?

“…Locations were estimated in reference to the nearest grid point (<15 m) and subsequently transformed into x-and y-coordinates; prior to data collection, we tested observer errors in distance estimations to be <3 m. Home range analyses were based on location data from focal observations that were subsampled at 20-min intervals and data from sequential radio tracking. These data points were regarded as independent because individuals can cross their home range during this time interval (Rooney et al 1998). We calculated home ranges as 100% minimum convex polygons (MCP) using ArcView GIS 3.3 (Esri) Animal Movement Software (Hooge et al 1999).…”

Home Range Use of Free-Ranging Cheetah on Farm and Conservation Land in Botswana

“…The selected interval in GPS readings could result in inaccurate estimates of home range size (Rooney et al 1998), which could be misleading as to the total area the cheetah population would need in order to be self sustaining. The selected interval also influences the location and size of the core range, which may place an emphasis on the wrong size of areas and location of habitat actually required for conservation.…”

Rigorous home range estimation with movement data: a new autocorrelated kernel density estimator

“…Typically, correlations diminish as observations grow farther apart in time, but autocorrelations in movement data often persist over long time periods, e.g., months or years (McNay et al 1994, Rooney et al 1998, Boyce et al 2010, Fleming et al 2014a. The conceptual definition of home range given by Burt (1943) lacks an objective, mathematical description that can be statistically estimated from data.…”