Premium
Abundance Estimation With a Transient Model Under the Robust Design
Author(s) -
CLAVEL JOANNE,
ROBERT ALEXANDRE,
DEVICTOR VINCENT,
JULLIARD ROMAIN
Publication year - 2008
Publication title -
the journal of wildlife management
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.94
H-Index - 111
eISSN - 1937-2817
pISSN - 0022-541X
DOI - 10.2193/2006-328
Subject(s) - acrocephalus , mark and recapture , population , estimation , abundance (ecology) , population model , statistics , population size , transient (computer programming) , warbler , econometrics , vital rates , computer science , ecology , population growth , mathematics , biology , demography , engineering , systems engineering , sociology , habitat , operating system
A common situation in capture‐mark‐recapture (CMR) studies on birds and other organisms is to capture individuals not belonging to the studied population only present during the short time of the capture session. Presence of such transient individuals affects demographic parameter estimation from CMR data. Methods exist to reduce biases on survival estimates in the presence of transients and have been shown to be particularly efficient within the Robust Design framework (several secondary capture sessions within a short time interval during which the studied population can be assumed closed). We present a new model to estimate population size accounting for transients. We first used simulated data to show that the method reduces positive biases due to transients. In a second step, we applied the method to a real CMR dataset on a reed warbler ( Acrocephalus scirpaceus ) population. Population size estimates are reduced by up to 50% when correcting for the presence of transients. Many field studies on managed animal populations use capture‐recapture methodology to obtain crucial parameters of the focal population demography. The resulting data sets are used either to estimate population size ignoring the presence of transients, or to estimate vital rates, accounting for transients but overlooking abundance estimation. Our method conciliates these 2 approaches.