Generalized mixed effects models for estimating demographic parameters with mark-resight data
Date
2008
Journal Title
Journal ISSN
Volume Title
Abstract
Mark-resight methods constitute a slightly different type of data than found in traditional mark-recapture, but they are in the same spirit of accounting for imperfect detection towards reliably estimating demographic parameters. Compared to mark-recapture, mark-resight can often be a less expensive and less invasive alternative in long-term population monitoring programs. However, the mark-resight estimators developed to date do not provide a flexible framework allowing the efficient use of covariates in modeling the detection process, information-theoretic model selection and multimodel inference, and the joint estimation of abundance and related demographic parameters. Here I develop a series of mark-relight models for the sampling conditions most often encountered in these studies that address this need for a more generalized framework. In Chapter 1, I introduce the the logit-normal mixed effects model (LNE) for estimating abundance when sampling is without replacement and the number of marked individuals in the population is known exactly. I compare the model to other mark-resight abundance estimators when applied to mainland New Zealand robin (Petroica australis) data recently collected in Eglinton Valley. Fiordland National Park. I also summarize its relative performance in simulation experiments. It can often be difficult to achieve sampling without replacement or to know the exact number of harked individuals in a population. In Chapter 2, I address these limitations of LNE by introducing the (zero-truncated) Poisson-log normal mixed effects abundance model, (Z)PNE. I demonstrate the use and advantages of (Z)PNE using black-tailed prairie dog (Cynomys ludovicianus) data recently collected in Colorado. I also investigate the expected relative performance of the model in simulation experiments. In Chapter 3, I extend (Z)PNE to a full-likelihood robust design model analogous to that used in mark-recapture for the simultaneous estimation of abundance, apparent survival, and transition probabilities between observable and unobservable states. I illustrate the use of the model with additional New Zealand robin data collected in Fiordland National Park, New Zealand. I also report on a series of simulation experiments evaluating the performance of the model under a variety of sampling conditions.
Description
Rights Access
Subject
Bowden's estimator
Cynomys ludovicianus
demographic parameters
fiordland national park
individual heterogeneity
mark-resight data
New Zealand
population size
program mark
program noremark