Fig. 3
From: Residency and space use estimation methods based on passive acoustic telemetry data
Examples of space use estimation with some of the methods described in this paper using the Dasyatis pastinaca data set. Land is grey and water is dark blue. a Minimum Convex Polygon at 100% (blue), 95% (green) and 50% (yellow) isopleths; b Characteristic Hull Polygons at the 100% (blue), 95% (green) and 50% (yellow) isopleths. Note the 100% CHP is equal to the 100% MCP. The estimations at lower percentage levels (95% and 50%) differ from their MCP counterparts because of how each method constructs polygons. Discontinuous area estimates and highly irregular shapes can occur depending on the characteristics of the real-life data used; c a-Local Convex Hull using a value of a = 600 m, with the 100% isopleth in blue, 95% isopleth in green and 50% in yellow. The inserted picture c2 shows another example using a value of a = 1000 m; d Kernel Utilisation Distribution estimation method using sample data at the 95% (solid white line) and 50% (dotted white line) isopleths using a fixed bandwidth of h=250. The inserted picture d2 shows another example using reference bandwidth href = 116 to highlight the effect of bandwidth selection; e Optimally weighted AKDEC (wAKDEC) range estimation with confidence intervals using an anisotropic OU movement model. The 50% core area (bold dashed line), 95% home range (bold solid line) and their respective confidence intervals (thinner dotted and solid lines) are shown; f Brownian Bridge Movement Model at the 95% (solid white line) and 50% (dotted white line) isopleths, using a location error of 250 m; g Biased Random Bridge kernel method at the 95% (solid white line) and 50% (dotted white line) isopleths, using a diffusion parameter of D = 4.44, Lmin of 50 m, a minimum smoothing parameter of h=250; h Dynamic Brownian Bridge Movement Model at the 95% (solid white line) and 50% (dotted white line) isopleths, using a location error of 250 m, margin size m = 11, and window size w = 31. Images a–d and g were created using the R-package adehabitatHR, e using the R-package ctmm following the R script provided by Silva et al., [178], f using the Animal Tracking Toolbox in the R-package V-Track, and h using the R-package move. All images were posteriorly edited for publication