Closed Population Models Research Articles

Decline of the Levantine Frog, Pelophylax bedriagae Camerano, 1882, in the western Aegean Region of Turkey changes in population size and implications for conservation

We estimated the annual population size, survival rates and capture probabilities for two populations of the Levantine Frog, Pelophylax bedriagae, using a Capture-Mark-Recapture (CMR) non-invasive digital photographic identification technique in Karagöl Lake and Soğan-harımı Pond in the western Aegean Region of Turkey, and used POLLOCK’s methodology to assess population parameters. Time specific capture probability, time specific survival rate and no temporary emigration were found to be the best appropriate models for our data. Population sizes were estimated as 245 adults in 2006, 301 in 2007, 67 in 2008 and 54 in 2009. Annual capture probabilities were estimated on average as 0.161, and annual survival rates across years were on average 0.301. Soğanharımı Pond was destroyed for recreational purposes after the first year of our study. To determine the size of the population, which was depleted during the construction of the recreational facilities, we used a closed population model, and concluded that in total 54 adults disappeared. In addition, we determined the possible negative factors that affect the population size and survival rates of the Levantine Frog including habitat destruction and the introduction of Siluris glanis and Astacus leptodactylus into the freshwater body.

The Shark Assemblage at French Frigate Shoals Atoll, Hawai‘i: Species Composition, Abundance and Habitat Use

Empirical data on the abundance and habitat preferences of coral reef top predators are needed to evaluate their ecological impacts and guide management decisions. We used longline surveys to quantify the shark assemblage at French Frigate Shoals (FFS) atoll from May to August 2009. Fishing effort consisted of 189 longline sets totaling 6,862 hook hours of soak time. A total of 221 sharks from 7 species were captured, among which Galapagos (Carcharhinus galapagensis, 36.2%), gray reef (Carcharhinus amblyrhynchos, 25.8%) and tiger (Galeocerdo cuvier, 20.4%) sharks were numerically dominant. A lack of blacktip reef sharks (Carcharhinus melanopterus) distinguished the FFS shark assemblage from those at many other atolls in the Indo-Pacific. Compared to prior underwater visual survey estimates, longline methods more accurately represented species abundance and composition for the majority of shark species. Sharks were significantly less abundant in the shallow lagoon than adjacent habitats. Recaptures of Galapagos sharks provided the first empirical estimate of population size for any Galapagos shark population. The overall recapture rate was 5.4%. Multiple closed population models were evaluated, with Chao Mh ranking best in model performance and yielding a population estimate of 668 sharks with 95% confidence intervals ranging from 289–1720. Low shark abundance in the shallow lagoon habitats suggests removal of a small number of sharks from the immediate vicinity of lagoonal islets may reduce short-term predation on endangered monk seal (Monachus schauinslandi) pups, but considerable fishing effort would be required to catch even a small number of sharks. Additional data on long-term movements and habitat use of sharks at FFS are required to better assess the likely ecological impacts of shark culling.

Augmenting mark–recapture with beach counts to estimate the abundance of little penguins on Penguin Island, Western Australia

Context Penguin Island supports the largest colony of little penguins in Western Australia. It is subjected to a suite of anthropogenic threats because of its proximity to an increasing urban population. For effective management of the colony, it is necessary to not only have knowledge of the size of the colony, but also the population trend of the colony. Aims To demonstrate a new cost-effective method of estimating the island-wide population of penguins on Penguin Island. Methods We estimated the island-wide population by combining mark–recapture sampling over 2 years on part of the island and beach counts of penguins arriving at night around the entire island. We estimated the abundance using closed population models, allowing for sex and time effects in capture probabilities. We had four capture occasions in 2008 only, and so considered heterogeneity of capture probabilities (Mh), using the Chao heterogeneity moment estimator. The proportion of all penguins counted that arrived at the four mark–recapture sites was then used to inflate the population estimate for the whole island. Key results In all, 62% of all penguins counted used the four mark–recapture sites. In 2007, there were an estimated 2369 ± 198 penguins, and 1543 ± 82 in 2008. When capture heterogeneity was allowed for in 2008, this estimate increased to 2069 ± 172. Conclusions Fewer eggs were laid and all measures of breeding performance were lower in 2008 than in 2007. Hence, the lower population estimate is most likely to represent fewer birds attempting to breed. However, further work on population estimates is required to determine whether capture heterogeneity occurs in both good and poor breeding years. Capture rates were affected by the presence of a full moon and high tides. Implications The estimate of the population can be used as part of the basis of a long-term monitoring program needed for effective management of the penguin colony. However, such studies must be coincident with the monitoring of a suite of reproductive and foraging parameters if short-term impacts of threats are to be recognised and well managed.

Population estimation and monitoring of an endangered lagomorph

We conducted the most intensive estimate of the endangered Lower Keys marsh rabbit (Sylvilagus palustris hefneri) metapopulation to date using pellet surveys and capture–recapture methodology. We livetrapped 83 rabbits, evaluated 5 closed population models, and selected the model that best represented the data. We considered the variation in behavioral response model the best model and correlated (r2 = 0.913) its patch population estimates to patch pellet densities. From the prediction equation, we generated a range-wide metapopulation estimate of 317 rabbits, a western clade population of 257 rabbits, an eastern clade population of 25 rabbits, and translocated marsh rabbit populations of 35 and zero on Little Pine and Water keys, respectively. A subset of patches whose marsh rabbit subpopulations were last estimated in 1993 exhibited a 46% decline in abundance over 15 yr. Due to the low estimate of the eastern clade population, special effort should be initiated to avoid loss of genetic diversity. The prediction equation suffers from limited data at high pellet densities, patches with ≥5 pellets/m2. Future studies should investigate if the slope of the regression is indeed near 1 by sampling patches across the range of pellet densities, especially those with ≥5 pellets/m2. The equation provides managers a quick, efficient, and noninvasive method to estimate marsh rabbit abundance from pellet counts but the confidence of predicted rabbit densities from high pellet density patches is low. © 2011 The Wildlife Society

Parameter-expanded data augmentation for Bayesian analysis of capture–recapture models

Data augmentation (DA) is a flexible tool for analyzing closed and open population models of capture–recapture data, especially models which include sources of hetereogeneity among individuals. The essential concept underlying DA, as we use the term, is based on adding “observations” to create a dataset composed of a known number of individuals. This new (augmented) dataset, which includes the unknown number of individuals N in the population, is then analyzed using a new model that includes a reformulation of the parameter N in the conventional model of the observed (unaugmented) data. In the context of capture–recapture models, we add a set of “all zero” encounter histories which are not, in practice, observable. The model of the augmented dataset is a zero-inflated version of either a binomial or a multinomial base model. Thus, our use of DA provides a general approach for analyzing both closed and open population models of all types. In doing so, this approach provides a unified framework for the analysis of a huge range of models that are treated as unrelated “black boxes” and named procedures in the classical literature. As a practical matter, analysis of the augmented dataset by MCMC is greatly simplified compared to other methods that require specialized algorithms. For example, complex capture–recapture models of an augmented dataset can be fitted with popular MCMC software packages (WinBUGS or JAGS) by providing a concise statement of the model’s assumptions that usually involves only a few lines of pseudocode. In this paper, we review the basic technical concepts of data augmentation, and we provide examples of analyses of closed-population models (M0, Mh, distance sampling, and spatial capture–recapture models) and open-population models (Jolly–Seber) with individual effects.

Comparison of Electrofishing and Snorkeling Mark–Recapture Estimation of Detection Probability and Abundance of Juvenile Steelhead in a Medium‐Sized River

AbstractWe compared nighttime electrofishing‐ and snorkeling‐based mark–recapture methods for estimating the detection probability and abundance of juvenile steelheadOncorhynchus mykissin the Cheakamus River, British Columbia. The reliability of abundance estimates largely depends on the precision and accuracy of detection probability (the fraction of marked individuals detected) as well as a few key assumptions of closed population models that we evaluated in this study. There was minimal bias (−2.5%) in diver estimates of the fork lengths of juvenile steelhead, and the relationship between measured and estimated fork lengths was very precise (r2= 95%). With a hierarchical Bayesian model, estimates of the detection probability for smaller juveniles (40–60 mm) ranged from 0.4 to 0.6 with electrofishing and were near zero with snorkeling. In contrast, snorkeling‐based detection probability was 0.6 and independent of size for larger juvenile steelhead (>60 mm) and much greater than that with electrofishing. These results provide strong evidence that there is considerable individual heterogeneity in detection probability driven by fish size for both methods. Owing to these differences, the abundance of age‐0 steelhead based on snorkeling was underestimated by 50%, but that of larger, age‐1 fish was unbiased and more precise (10‐fold) than that based on electrofishing. The use of electrofishing during marking resulted in a substantive reduction in snorkeling‐based detection probability during recapture, but the converse was not true. Thus, there is strong evidence of behavioral heterogeneity in detection probability induced by electrofishing, but only when snorkeling is used to detect fish during recapture. The differences in detection probabilities among size‐classes and sampling methods were probably driven by differences in concealment behavior, spatial distribution, and fright responses to sampling. Our results indicate that snorkeling is the better way to estimate abundance for larger juvenile steelhead, whereas electrofishing is preferred for smaller fish.

Addressing challenges in non invasive capture-recapture based estimates of small populations: a pilot study on the Apennine brown bear

It is often difficult to determine optimal sampling design for non-invasive genetic sampling, especially when dealing with rare or elusive species depleted of genetic diversity. To address this problem, we ran a hair-snag pilot study on the remnant Apennine brown bear population. We used occupancy models to estimate the performance of an improved field protocol, a meta-analysis approach to indirectly model capture probability, and simulations to evaluate the effect of genotyping errors on the accuracy of capture-recapture population estimates. In spring 2007 we collected 70 bear hair samples in 15 5 × 5 km cells, using 5 10-day trapping sessions. Bear detectability was higher in 2007 than in a previous attempt on the same population in 2004, reflecting improved field protocols and sampling design. However, individual capture probability was 0.136 (95% CI = 0.120–0.152), still below the minimum requirements of capture-mark-recapture closed population models. We genotyped hair samples (n = 63) at 9 microsatellite loci, obtaining 94% Polymerase Chain Reaction success, and 13 bear genotypes. Estimated PIDsib was 0.00594, and per-genotype error rate was 0.13, corresponding to a 99% probability of correct individual identification. Simulation studies showed that the effect of non-corrected or filtered genetic errors on the accuracy of population estimates was negligible only when individual capture probability was >0.2. Our results underline how the interaction among field protocols, sampling strategies and genotyping errors may affect the accuracy of DNA-based estimates of small and genetically depleted populations, and warned us about the feasibility of a survey using only traditional hair-snag sampling. In this and similar cases, indications from pilot studies can provide cost-effective means to evaluate the efficiency of designed sampling and modelling procedures.

How has the remnant population of the threatened frogLeiopelma pakeka(Anura: Leiopelmatidae) fared on Maud Island, New Zealand, over the past 25 years?

Despite widespread global reports of declining amphibian populations, supporting long-term census data are few, limiting opportunities to study changes in numbers and survival over time. However, in New Zealand, for the past 25 years (1983–2008), we studied Leiopelma pakeka, a threatened, terrestrial frog that inhabits rocky boulder banks under forest on Maud Island. Using night sampling at least annually on two 12 × 12 m plots, we had 5390 captures of 1000+ individuals, 327 on one plot (grid 1), 751 on the other (grid 2). The mean (±SE) number of frogs found per night was 11.3 (±0.6) on grid 1 and 25.6 (±1.4) on grid 2. We used capture-recapture models to estimate population size, proportion of animals remaining beneath the surface and survival rate. The mean (±SE) population estimate was 131 (±14.7) frogs on grid 1 and 367 (±38.7) on grid 2. Over 25 years the estimated population increased on grid 1 and fluctuated on grid 2. Some frogs were captured on most sampling visits, others less often, evidently failing to emerge from cover each visit. Using a combination of open and closed population models, we estimated the mean (±SE) proportion remaining underground was 0.63 (±0.12) on grid 1 and 0.53 (±0.07) on grid 2. Our research represents one of the longest-run population studies of any frog, and we recorded significant longevity, two males reaching 35+ and 37+ years, a female 34+ years. No significant differences occurred between mean annual survival rates of apparent females and males, or between the two sites. The number of toes clipped for individual identification had little influence on the return rate, once the effect of time of first capture was removed.

A Comparative Study of Bayesian Model Selection Criteria for Capture-Recapture Models for Closed Populations

Capture-Recapture models estimate unknown population sizes. Eight standard closed population models exist, allowing for time, behavioral, and heterogeneity effects. Bayesian versions of these models are presented and use of Akaike's Information Criterion (AIC) and the Deviance Information Criterion (DIC) are explored as model selection tools, through simulation and real dataset analysis.

Using Mixtures to Model Heterogeneity in Ecological Capture‐Recapture Studies

Modelling heterogeneity of capture is an important problem in estimating animal abundance from capture-recapture data, with underestimation of abundance occurring if different animals have intrinsically high or low capture probabilities. Mixture models are useful in many cases to model the heterogeneity. We summarise mixture model results for closed populations, using a skink data set for illustration. New mixture models for heterogeneous open populations are discussed, and a closed population model is shown to have new and potentially effective applications in community analysis.

CHARACTERIZING SOURCE–SINK DYNAMICS WITH GENETIC PARENTAGE ASSIGNMENTS

Source-sink dynamics have been suggested to characterize the population structure of many species, but the prevalence of source-sink systems in nature is uncertain because of inherent challenges in estimating migration rates among populations. Migration rates are often difficult to estimate directly with demographic methods, and indirect genetic methods are subject to a variety of assumptions that are difficult to meet or to apply to evolutionary timescales. Furthermore, such methods cannot be rigorously applied to high-gene-flow species. Here, we employ genetic parentage assignments in conjunction with demographic simulations to infer the level of immigration into a putative sink population. We use individual-based demographic models to estimate expected distributions of parent-offspring dyads under competing sink and closed-population models. By comparing the actual number of parent-offspring dyads (identified from multilocus genetic profiles) in a random sample of individuals taken from a population to expectations under these two contrasting demographic models, it is possible to estimate the rate of immigration and test hypotheses related to the role of immigration on population processes on an ecological timescale. The difference in the expected number of parent-offspring dyads between the two population models was greatest when immigration into the sink population was high, indicating that unlike traditional population genetic inference models, the highest degree of statistical power is achieved for the approach presented here when migration rates are high. We used the proposed genetic parentage approach to demonstrate that a threatened population of Marbled Murrelets (Braclhyrarmphus marmotus) appears to be supplemented by a low level of immigration (approximately 2-6% annually) from other populations.

Annual flooding, survival and recruitment in a rodent population from the Niger River plain in Mali

AbstractMultimammate rats of the genusMastomysare among the most widespread pest species in Africa. Previous studies ofMastomyspopulation dynamics have generally reported variation in abundance but few have investigated the demographic parameters underlying this variation, and in particular recruitment. Capture-mark-recapture data were collected forMastomys erythroleucusseveral times a year from 2000 to 2004 at a site annually flooded by the Niger River in Mali. Closed-population models were used to estimate population abundance. Both seniority (a parameter inversely linked to recruitment) and survival probabilities were estimated by capture-mark-recapture models. The impacts of water level, population abundance and cumulative rainfall were assessed for each demographic parameter. Survival probabilities (local survival) were negatively correlated with water level, suggesting that rodents emigrated out of the study zone during flooding. As for seniority probabilities, 86% of temporal variation was explained by a model with season, abundance, water level and the interaction between abundance and water level. This suggests that density-dependence in recruitment was mediated by intraspecific competition for food or refuge from floodwaters, or by predation. The flood of the Niger River greatly impactsMastomys erythroleucuspopulation dynamics, affecting both survival and seniority probabilities.

Estimating Abundance of Reef-Dwelling Sharks: A Case Study of the Epaulette Shark, Hemiscyllium ocellatum (Elasmobranchii: Hemiscyllidae)1

Benthic reef sharks play an important role in reef ecosystems, but little is known about their abundance or population dynamics. Abundance of the epaulette shark, Hemiscyllium ocellatum (Bonnaterre), on Heron Island Reef, Great Barrier Reef, Australia, was examined via a mark-recapture study. A total of 496 sharks was tagged between July 1994 and August 1997 in a 0.25-km2 area of reef flat, with 80 tagged sharks recaptured for a total of 102 recapture events. Captured individuals ranged in size from juveniles to adults (285–750 mm total length). Recaptured sharks were collected after 1–725 days at liberty and at distances of 0–329 m from their original capture point. The overall recapture rate was 20.6% with an estimated 17.5% tag loss. Population size was estimated using both closed and open population models. Closed population models produced various abundance estimates, with the Chao M(th) ranked best in model performance with an estimate of 2,224 sharks and 95% confidence intervals ranging from 1,730 to 2,916. Open population models produced lower estimates, with the Jolly D model producing an estimate of 559 individuals within the study site and confidence intervals ranging from 26 to 1,092. All models produced density estimations of 0.3 to 1.2 sharks per 100 m2. Based on thorough examination of model assumptions and results, open population models appear to provide the best population estimate within the study area.

Species richness estimation and determinants of species detectability in butterfly monitoring programmes

Abstract 1. Species richness is the most widely used biodiversity index, but can be hard to measure. Many species remain undetected, hence raw species counts will often underestimate true species richness. In contrast, capture–recapture methods estimate true species richness and correct for imperfect and varying detectability.2. Detectability is a crucial quantity that provides the link between a species count and true species richness. For insects, it has hardly ever been estimated, although this is required for the interpretation of species counts.3. In the Swiss butterfly monitoring programme about 100 transect routes are surveyed seven times a year using a highly standardised protocol. In July 2003, control observers made two additional surveys on 38 transects. Data from these 38 quadrats were analysed to see whether currently available capture–recapture models can provide quadrat‐specific estimates of species richness, and to estimate species detectability in relation to transect, observer, survey, region, and abundance.4. Species richness over the entire season cannot be estimated using current capture–recapture methods. The species pool was open, preventing use of closed population models, and detectability varied by species, preventing use of current open population models. Assuming a closed species pool during two mid‐season (July) surveys, a Jackknife capture–recapture method was used that accounts for heterogeneity to estimate mean detectability and species richness.5. In every case, more species were present than were counted. Mean species detectability was 0.61 (SE 0.01) with significant differences between observers (range 0.37–0.83). Species‐specific detection at time t+ 1 was then modelled for those species seen at t for three mid‐season surveys. Detectability averaged 0.50 (range 0.17–0.81) for individual species and 0.65, 0.44, and 0.42 for surveys. Abundant species were detected more easily, although this relationship explained only 5% of variation in species detectability.6. These are important, although not entirely unexpected, results for species richness estimation of short‐lived animals. Raw counts of species may be misleading species richness indicators unless many surveys are conducted. Monitoring programmes should be calibrated, i.e. the assumption of constant detectability over dimensions of interest needs to be tested. The development of capture–recapture or similar models that can cope with both open populations and heterogeneous species detectability to estimate species richness should be a research priority.

TheRcapturePackage: Loglinear Models for Capture-Recapture inR

This article introduces Rcapture, an R package for capture-recapture experiments. The data for analysis consists of the frequencies of the observable capture histories over the t capture occasions of the experiment. A capture history is a vector of zeros and ones where one stands for a capture and zero for a miss. Rcapture can fit three types of models. With a closed population model, the goal of the analysis is to estimate the size N of the population which is assumed to be constant throughout the experiment. The estimator depends on the way in which the capture probabilities of the animals vary. Rcapture features several models for these capture probabilities that lead to different estimators for N. In an open population model, immigration and death occur between sampling periods. The estimation of survival rates is of primary interest. Rcapture can fit the basic Cormack-Jolly-Seber and Jolly-Seber model to such data. The third type of models fitted by Rcapture are robust design models. It features two levels of sampling; closed population models apply within primary periods and an open population model applies between periods. Most models in Rcapture have a loglinear form; they are fitted by carrying out a Poisson regression with the R function glm. Estimates of the demographic parameters of interest are derived from the loglinear parameter estimates; their variances are obtained by linearization. The novel feature of this package is the provision of several new options for modeling capture probabilities heterogeneity between animals in both closed population models and the primary periods of a robust design. It also implements many of the techniques developed by R. M. Cormack for open population models.

Population size and structure of whale sharks Rhincodon typus at Ningaloo Reef, Western Australia

We used photo-identification to produce estimates of population size and structure of whale sharks Rhincodon typus at Ningaloo Reef, Western Australia. We analysed photographs of whale sharks taken from 1992 to 2004. A combination of spot and stripe patterns behind the last gill slit and forward of the dorsal fin (lateral view), and distinctive scars and marks on the body and fins were useful for identifying individual sharks. These patterns appeared to be unique to individuals and distinctive markings could be recognized on some sharks for more than a decade. From 581 photographs, 159 individuals were identified. Of these, 74% were male, 16% were female and 10% were of indeterminate gender. Photographed sharks ranged in estimated size from 3 to 10 m total length (TL). The size distribution of sharks was bimodal with a large peak at 8 m and a smaller peak at 6 m TL. Sixty individuals were resighted during the study. Of these, 46 were resighted at different times during the same year (sometimes on multiple occasions) up to 4 mo after they were initially photographed, and 33 were resighted (4 on >2 occasions) in different years. The interval between inter-annual resightings was typically 1 to 3 yr; however, 2 sharks were resighted after a period of 12 yr. We estimated the super population of whale sharks that visit Ningaloo Reef to consist of approximately 300 to 500 individuals (95% confidence interval) based on closed population models, or 320 to 440 based on Jolly-Seber open-population models. Our study shows that photo-identification offers a practical, non-invasive and non-destructive means to obtain data on the population size and demography of whale sharks.

Site Occupancy Models with Heterogeneous Detection Probabilities

Models for estimating the probability of occurrence of a species in the presence of imperfect detection are important in many ecological disciplines. In these "site occupancy" models, the possibility of heterogeneity in detection probabilities among sites must be considered because variation in abundance (and other factors) among sampled sites induces variation in detection probability (p). In this article, I develop occurrence probability models that allow for heterogeneous detection probabilities by considering several common classes of mixture distributions for p. For any mixing distribution, the likelihood has the general form of a zero-inflated binomial mixture for which inference based upon integrated likelihood is straightforward. A recent paper by Link demonstrates that in closed population models used for estimating population size, different classes of mixture distributions are indistinguishable from data, yet can produce very different inferences about population size. I demonstrate that this problem can also arise in models for estimating site occupancy in the presence of heterogeneous detection probabilities. The implications of this are discussed in the context of an application to avian survey data and the development of animal monitoring programs.

ESTIMATING SURVIVAL AND BREEDING PROBABILITY FOR POND-BREEDING AMPHIBIANS: A MODIFIED ROBUST DESIGN

Many studies of pond-breeding amphibians involve sampling individuals during migration to and from breeding habitats. Interpreting population processes and dynamics from these studies is difficult because (1) only a proportion of the population is observable each season, while an unknown proportion remains unobservable (e.g., non-breeding adults) and (2) not all observable animals are captured. Imperfect capture probability can be easily accommodated in capture–recapture models, but temporary transitions between observable and unobservable states, often referred to as temporary emigration, is known to cause problems in both open- and closed-population models. We develop a multistate mark–recapture (MSMR) model, using an open-robust design that permits one entry and one exit from the study area per season. Our method extends previous temporary emigration models (MSMR with an unobservable state) in two ways. First, we relax the assumption of demographic closure (no mortality) between consecutive (secondary)...

Adaptations of animals to commensal habitats: population dynamics of house mice Mus musculus domesticus on farms

Summary Commensal mammals live in habitats that appear to provide both benefits and costs compared with natural and semi‐natural (non‐commensal) habitats. These commensal habitats offer potentially rich food resources but are also characterized by instability in time and space. We expected to demonstrate that animals in these habitats have high reproductive rates to counter high mortality rates and show flexibility in spatial organization. House mice (Mus musculus domesticus) are unusual because they are able to persist entirely in both commensal and non‐commensal habitats, and so can provide a test of the distinctiveness of commensal populations. We studied populations of commensal house mice on two neighbouring farms in North Yorkshire, UK, for 2 years by capture–mark–recapture using the robust design. A total of 568 house mice were captured, with a total of 1053 recaptures. Population size varied from nine to 93 individuals, estimated with closed population models. Apparent survival was surprisingly low (0·54 per month) and was best modelled as constant across age, sex and time. In situ reproductive recruitment occurred throughout the study and was numerically more important than immigration. Immigration was important during only two intervals and was probably from untrappable areas within the study site. Breeding throughout the year allowed the population to persist despite low survival rates. The results suggested that births and deaths had more influence on the overall population dynamics than movement. The population was divided into subgroups to represent the territorial, demic structure known to be present in commensal house mouse populations. Dispersal between subgroups within the population was limited, representing only 6·6% of recaptures. The low rates of dispersal suggested that house mice responded to their environment as consisting of aggregated patches of suitable habitat. In comparison, non‐commensal house mice have lower mortality, seasonal breeding, and individuals move more frequently and further than commensal house mice. These differences illustrate the responses of house mice to the specific opportunities and demands of commensal habitats and demonstrate the importance of a flexible life‐history strategy for animals exploiting these habitats.

Loglinear models for the robust design in mark-recapture experiments.

The robust design is a method for implementing a mark-recapture experiment featuring a nested sampling structure. The first level consists of primary sampling sessions; the population experiences mortality and immigration between primary sessions so that open population models apply at this level. The second level of sampling has a short mark-recapture study within each primary session. Closed population models are used at this stage to estimate the animal abundance at each primary session. This article suggests a loglinear technique to fit the robust design. Loglinear models for the analysis of mark-recapture data from closed and open populations are first reviewed. These two types of models are then combined to analyze the data from a robust design. The proposed loglinear approach to the robust design allows incorporating parameters for a heterogeneity in the capture probabilities of the units within each primary session. Temporary emigration out of the study area can also be accounted for in the loglinear framework. The analysis is relatively simple; it relies on a large Poisson regression with the vector of frequencies of the capture histories as dependent variable. An example concerned with the estimation of abundance and survival of the red-back vole in an area of southeastern Québec is presented.