A review of movement models in open population capture–recapture

Abstract

1. Understanding rates of survival and recruitment is critical to population management, and capture-recapture methods of estimation are widely used. Spatial models allow for a spatial detection process and can include the movement of activity centres between sampling times. Movement is often treated as a random walk with the step length governed by a probability kernel. However, the movement component of open population spatially explicit capture-recapture models (open SECR) has evolved haphazardly and comparison among studies is difficult. 2. We review published studies, document suitable probability kernels and address the issues of scale and buffer dependence in open SECR by a combination of simulation and case studies on ovenbirds Seiurus aurocapilla and tigers Panthera tigris.3. Flexible 2-parameter kernels, such as the bivariate t-distribution, fit better than the popular bivariate normal and resulted in higher estimates of survival. We reconcile different parameterizations of the bivariate t-distribution and identify a problem when the kernel is defined in terms of its margins.4. Movement models failed to separate mortality and emigration in simulated data when the data were a random mixture of long and short movements. Our estimates of ovenbird survival were buffer-dependent, and we interpret this as a sign that the data are inadequate for joint modelling of survival and movement.Estimates of tiger survival were more nearly asymptotic on buffer width. 5. We repeat the warning of earlier authors that movement models are effective for separating mortality and emigration only when the data span the range of movement. We appeal for more complete and consistent reporting of movement models and identify topics for future research.

Mechanistic home range capture–recapture models for the estimation of population density and landscape connectivity

“…Also, the generalisation of ADCR for dealing with temporally open populations driven by birth-death processes (Gardner et al, 2010;Efford & Schofield 2022) would be helpful for analysing long-term capture-recapture data. The development of software tools that enable users to benefit from the flexible nature of ADCR is an urgent task for expanding the application of the method.…”

“…The broad application of SCR in ecology can largely be attributed to its flexibility to incorporate heterogeneity in ecological processes underlying spatial capture recapture data, such as environment-dependent abundance and sex-biased detection (Efford, 2022). Future advances in ADCR should focus on similar functionality.…”

“…The log likelihood values of SCR with the least cost path were derived based on the R script by Sutherland et al (2015), with some modifications by Keita Fukasawa to deal with count detectors and the negative effect of landscape on the cost function. The basic SCR was estimated using the R package secr (Efford, 2022).…”

Dynamics of Humoral Immunity to Myxoma and Rabbit Hemorrhagic Disease Viruses in Wild European Rabbits Assessed by Longitudinal Semiquantitative Serology

“…The population density at each sampling session was estimated using Jolly-Seber-Arnason-Schwarz models, either spatially explicit for the free-ranging sites or nonspatial for the fenced sites ( 51 , 52 ). Both types of models were implemented using the package “openCR” ( 53 ) in R 3.6.1 ( 54 ).…”

Estimating spatially variable and density‐dependent survival using open‐population spatial capture–recapture models

“…When data are collected over several consecutive occasions, open‐population SCR (OPSCR) models can be used to estimate demographic rates and movement of individuals between occasions in addition to densities (Bischof, Milleret, et al, 2020 ). Modeling individual movement between occasions helps return unbiased estimates of survival (Gardner et al, 2018 ), but see Efford and Schofield ( 2022 ). These properties make OPSCR models well suited for drawing population‐level inferences about the drivers of demographic processes.…”