RESEARCH

General Interests:

Quantitative ecology primarily focusing on terrestrial vertebrate population and community dynamics. I am also interested in evaluating current and developing new statistical methods for the estimation of demographic parameters in mark-resight studies where data collection methods are generally less expensive and less invasive than traditional mark-recapture methods.

Ph. D. Dissertation:

Generalized mixed effects models for estimating demographic parameters in mark-resight studies.

Committee Chair: Dr. Gary C. White, Fish, Wildlife, and Conservation Biology

Committee Members: Dr. Kenneth P. Burnham, Fish, Wildlife, and Conservation Biology

Dr. Jennifer A. Hoeting, Statistics

Dr. Michael F. Antolin, Biology

As a less invasive and less expensive alternative to traditional mark-recapture methods, my objective is to develop binomial logit- and Poisson-log normal mixed effects models for estimating animal abundance and related parameters in mark-resight studies. The models will be robust to a variety of sampling conditions often problematic in these studies, including sighting heterogeneity, sampling with replacement, unknown numbers of marked individuals, and imperfect marked individual detection. The ability to incorporate covariates in modeling sighting probabilities may also lead to improved precision of estimates. I ultimately intend for the models to combine the favorable qualities of previously available mark-resight and mark-recapture estimators into a generalized framework to estimate demographic parameters for a wide range of species and field conditions. They will initially be applied to black-tailed prairie dogs and burrowing owls on colonies in the Pawnee National Grassland as part of the Colorado State University Plague Project.

These models are now available in Program MARK. Documentation for running these models in MARK may be downloaded here.

Temporarily marking an anesthetized black-tailed prairie dog (left) and a banded burrowing owlet (right) on the Pawnee National Grassland, Colorado.

M. S. Thesis:

The mark-resight method: an application to bighorn sheep in Colorado and a simulation study evaluating estimators.

Committee Chair: Dr. Gary C. White, Fishery and Wildlife Biology

Committee Members: Dr. Kenneth R. Wilson, Fishery and Wildlife Biology

Dr. Bruce A. Wunder, Biology

Download my thesis here. This is a large file and may take some time.

Collaring and collecting samples from a ewe via helicopter net-gunning, March 2003 (photo by W. Rogers).

The “Bighorn Crew” (Mat, Kurt, Leslie, Justin, and Brett) posing tough before heading into the

backcountry, July 2003 (photo by B. Keller).

Brett and Dr. Gary C. White at the 3^rd International Wildlife Management Congress, December 2003.

Recent and Upcoming Publications

McClintock, B. T., G. C. White, and K. P. Burnham. 2006. A robust design mark-resight abundance estimator allowing heterogeneity in resighting probabilities. Journal of Agricultural, Biological, and Environmental Statistics 11: 231-248. Download Here.

McClintock, B. T., and G. C. White. 2007. Bighorn sheep abundance following a suspected pneumonia epidemic in Rocky Mountain National Park. Journal of Wildlife Management 71: 183-189. Download Here.

Magle, S. B., B. T. McClintock, D. W. Tripp, G. C. White, M. F. Antolin, and K. R. Crooks. 2007. Mark-resight methodology for estimating population densities for prairie dogs. Journal of Wildlife Management 71: 2067-2073. Download Here.

McClintock, B. T., G. C. White, K. P. Burnham, and M. A. Pryde. In press. A generalized mixed effects model of abundance for mark-resight data when sampling is without replacement. Environmental and Ecological Statistics.

McClintock, B. T., G. C. White, M. F. Antolin, and D. Tripp. In press. Estimating abundance using mark-resight when sampling is with replacement or the number of marked individuals is unknown. Biometrics.

McClintock, B. T., and G. C. White. In review. A less field-intensive robust design for estimating demographic parameters with mark-resight data.

Recent Presentations

EURING Technical Meeting January 14-20 2007, Dunedin, New Zealand

A generalized mixed effects model of abundance for mark-resight data when sampling is without replacement

Abstract:

In recent years, the mark-resight method for estimating abundance when the number of marked individuals is known has become increasingly popular. By using individually identifiable bands that may be resighted from a distance, these techniques can be applied to many bird species, and are particularly useful for relatively small, closed populations. However, due to the different assumptions and general rigidity of the available estimators, researchers must often commit to a particular model without rigorous quantitative justification for model selection based on the data. Here we introduce a nonlinear logit-normal mixed effects model addressing this need for a more generalized framework when sampling is with replacement. Similar to models available for mark-recapture studies, the estimator allows a wide variety of sampling conditions to be parameterized efficiently under a robust sampling design. Resighting rates may be modeled simply or with more complexity by including fixed temporal and random individual heterogeneity effects. Using information theory, the model(s) best supported by the data may be selected from the various candidate models proposed. Under this generalized framework, we hope the uncertainty associated with mark-resight model selection will be reduced substantially. We summarize the performance of our model in simulation experiments and compare it to other mark-resight abundance estimators when applied to mainland New Zealand robin (Petroica australis) data recently collected in Eglinton Valley, Fiordland.

The Wildlife Society 13th Annual Conference September 23-27 2006, Anchorage, Alaska, USA

A generalized mixed effects model of abundance for mark-resight data when sampling is with replacement

Abstract:

In recent years, the mark-resight method for estimating abundance when the number of marked individuals is known has become increasingly popular. However, due to the different assumptions and general rigidity of the available estimators, researchers must often commit to a particular model without rigorous quantitative justification for model selection based on the data. Here we introduce a nonlinear Poisson-log normal mixed effects model addressing this need for a more generalized framework when sampling is with replacement. Similar to models available for mark-recapture studies, the estimator allows a wide variety of sampling conditions to be parameterized efficiently under a robust sampling design. Resighting rates may be modeled simply or with more complexity by including fixed temporal and random individual heterogeneity effects. Using information theory, the model(s) best supported by the data may be selected from the various candidate models proposed. Under this generalized framework, we hope the uncertainty associated with mark-resight model selection will be reduced substantially. We summarize the performance of our model in simulation experiments and compare it to other mark-resight abundance estimators when applied to Alaskan brown bear data collected in the late-1980s.

The Wildlife Society 12th Annual Conference September 25-29 2005, Madison, Wisconsin, USA

Distribution and abundance of bighorn sheep following a suspected pneumonia epidemic in Rocky Mountain National Park, Colorado.

Abstract:

Anecdotal evidence of a pneumonia epizootic among bighorn sheep in Rocky Mountain National Park, Colorado, during the mid-1990s prompted Park officials to examine the current condition of the herds. Emphasizing the adult ewe population, we designed a mark-resight study to estimate abundance, recruitment, survival, fecundity, distributions, and movements. Fifty-nine adult ewes were captured, sampled, and radio-collared via helicopter net-gunning during winter 2002–2003. Mark-resight data for Bowden’s Estimator were collected in 2003 and 2004 from ground surveys conducted May–September. The 2003 adult ewe population was estimated at 165.1 (SE=9.3) sheep. Overall ratios to ewes produced estimates of 39.6 (SE=4.7) yearlings, 74.0 (SE=8.6) lambs, and 87.7 (SE=8.3) rams for a total 2003 population estimate of 366.3 (SE=30.9). In 2004, ewe abundance was estimated at 165.1 (SE=11.2). Overall yearling, lamb, and ram ratios produced abundance estimates of 33.0 (SE=4.7), 90.0 (SE=9.6), and 70.1 (SE=7.8), respectively, providing a total 2004 population estimate of 358.2 (SE=33.3). Minimum fecundity for 2003 and 2004 averaged 0.74 (SE=0.05) lambs per ewe, but the 1-year lamb recruitment rate per ewe from July 2003 to July 2004 was 0.15 (SE=0.07), and the 1-year lamb survival rate was 0.24 (SE=0.09). Previous abundance estimates suggest a Park-wide decline has occurred between the late-1980s and the suspected pneumonia epidemic of the mid-1990s. Sample analyses indicate the presence of numerous pathogens associated with pneumonia. Program MARK survival analyses suggest interactions of these pathogens may be related to mortality. Such interactions provide evidence of the “pneumonia-complex” associated with massive sheep die-offs, and suggest the recent decline was likely pneumonia-related. However, the decline was either not as severe as other documented bighorn die-offs, or the population has somewhat recovered in terms of abundance in recent years. Low yearling ratios indicate the herds may not be recruiting at levels needed for population growth and warrant continued monitoring.

The Wildlife Society 11th Annual Conference September 18-22 2004, Calgary, Canada

Variation in sighting probabilities and mark-resight closed population abundance estimators: a simulation study and its implications for model selection.

Abstract:

Model selection in mark-resight studies to estimate abundance of closed populations is heavily dependent on the nuisance parameter known as sighting probability (p) and its assumed levels of variation. Current methods of estimating this variation are plagued by Type II errors, and investigators are often left to educated guesswork when determining the model to use based on what (if any) levels of variation are present. Program NOREMARK includes two estimators that model this variation differently. The Joint Hypergeometric Estimator (JHE) is based on maximum likelihood (ML) theory and assumes little or no heterogeneity of individuals in their p. Bowden’s Estimator (BOWE) is not ML-based, but allows individual heterogeneity. Here we introduce the Beta-Binomial Estimator (BBE), a mark-resight model to estimate abundance of a closed population combining the favorable qualities of ML estimation and allowing individual heterogeneity. BBE holds the same assumptions as BOWE, but does not allow sampling with replacement. Simulations were conducted to evaluate the performance of the three estimators under four classes of variation in p: 1) no variation; 2) individual heterogeneity; 3) temporal variation; and 4) both individual and temporal variation. Performance was evaluated primarily on achieved coverage and confidence interval length. With sufficiently large data sets, all estimators were unbiased. JHE generally achieved the greatest precision in Class 1 and Class 3 scenarios, but performed poorly when individual heterogeneity was present. BBE performed marginally better than BOWE in Class 2, and the two performed equally well in Class 4 simulations. Because BBE performed as well or better than BOWE under simulated conditions of individual heterogeneity, we are currently investigating the possible advantages of selecting BBE when this variation is suspected. These include its ability to provide a ML estimate of individual heterogeneity and incorporating covariates in the modeling of p.

Return to Brett T. McClintock’s home page

Go to Brett T. McClintock’s links page