A video-based method for measuring small-scale animal movement ALISON E. POOLE * †, RICHARD A. STILLMAN * & KEN J. NORRIS † *Centre for Ecology and Hydrology (CEH), Dorset yCentre for Agri-Environment Research (CAER), University of Reading (Received 7 December 2005; initial acceptance 18 January 2006; final acceptance 16 February 2006; published online 22 September 2006; MS. number: SC-1289) The ability to measure movement distance or speed is important in understanding the interaction between animals and their habitat and competitors. Long move- ments (e.g. migration, dispersal or territoriality) can be measured by using radiotracking (see e.g. Walls et al. 2005; White et al. 2005) or markerecapture methods (see e.g. Peach et al. 2001; Butler et al. 2005; Hounsome et al. 2005). However, these methods are unsuitable for measur- ing small-scale animal movements in behavioural studies where measurements need to be accurate to centimetres rather than kilometres. Accurate measurement of small-scale movement is important in studies in which an animal’s speed and/or movement distance needs to be measured. Examples include, but are not limited to, measurement of searching speeds and/or distances to determine searching efficiency (Holling 1959) and measurement of distances moved dur- ing interactions with conspecifics. Searching could be for prey (e.g. Speakman & Bryant 1993; Ens et al. 1996; Eifler & Eifler 1999; Burrows et al. 2003), mates (e.g. Derivera 2005; Foellmer & Fairbairn 2005; Shine et al. 2005) or other resources such as nest sites (e.g. Liebezeit & George 2002; Boulton et al. 2003). Interactions with conspecifics could be within a nesting colony (e.g. Wolff 1994; Oswald et al. 2005) or kleptoparasitic attacks, the distances of which are key to understanding the strength of interfer- ence competition (Stillman et al. 2002). Previously, small-scale movements have been measured relative to reference markers placed at known locations within laboratory arenas (e.g. Carlson 1985) or natural habitats (e.g. Smith 1974a, b; Cresswell 1998; Cresswell et al. 2001). Using this technique to assess movement in natural habitats requires that wild animals move within the marked area. This is often difficult to arrange, espe- cially in species with unpredictable distributions or when only a small proportion of the available habitat can be marked. Animals may also avoid reference markers or the presence of markers may influence the behaviour of the animals (e.g. birds may perch on reference sticks rather than foraging on the ground between the sticks). What is ideally needed is a method that can measure small-scale animal movements without the use of refer- ence markers and without influencing behaviour. In previous studies of mammals and ground-foraging birds, researchers have measured movement distance without reference markers by counting paces and multi- plying the number of paces by pace length. Pace lengths have been measured either from footprints on the sub- strate (Pienkowski 1983) or from the distances covered be- tween recognizable features on the ground (Tye 1989). Both of these approaches have limitations because foot- prints form only on suitable substrates (e.g. mud) and not all habitats have recognizable features. These tech- niques also assume that pace length is constant, something that neither Pienkowski (1983) nor Tye (1989) tested. An alternative, video-based approach is to use bird length or another morphological feature as a calibration measure. Stillman et al. (2002) counted paces and mea- sured the onscreen bird length and distances run by oys- tercatchers, Haematopus ostralegus, and bar-tailed godwits, Limosa lapponica, running directly across the video screen, then calculated pace length in units of on- screen bird lengths. They used published bird lengths (Hayman et al. 1986) to convert pace lengths from units of bird lengths to metres, although, as in other studies, they did not test for variation in pace length. We tested the accuracy of this technique by using it to estimate the known movement distances of an artificial bird. We then applied the technique to several plover (Charadrii- dae) and thrush (Turdidae) species and studied the move- ment distances of birds that do not move directly across screen, without assuming that pace length is constant. Al- though we measured movement distance of birds in this study, the approach is equally applicable to movement Correspondence: A. Poole, CEH Dorset, Winfrith Technology Centre, Winfrith Newburgh, Dorchester, Dorset DT2 8ZD, U.K. (email: alpo@ceh.ac.uk). K. J. Norris is at the Centre for Agri-Environment Research (CAER), School of Agriculture, Policy & Development, University of Reading, Earley Gate, PO Box 237, Reading RG6 6AR, U.K. 1205 0003e 3472/06/$30.00/0 Ó 2006 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved. ANIMAL BEHAVIOUR, 2006, 72, 1205e1212 doi:10.1016/j.anbehav.2006.02.019