Estimation of demographic parameters is central to research questions in wildlife management, conservation, and evolutionary ecology. I review the 7 major classes of mark–recapture models that investigators can use to estimate apparent survival and other parameters from live-encounter data. Return rates are the product of 4 probabilities: true survival (S), site fidelity (F), site propensity (δ), and true detection (p*). Cormack-Jolly-Seber (CJS) models improve upon return rates by separating apparent survival (φ = S × F) from the probability of encounter (p = δ × p*). The main drawback to mark–recapture models based on live-encounter data is that the complement of apparent survival (1 − φ) includes losses to mortality and to permanent emigration, and these 2 ecological processes are difficult to disentangle. Advanced mark–recapture models require additional sampling effort but estimate apparent survival with greater precision and less bias, and they also offer estimates of other useful demographic parameters. Time-since-marking or transient models control for individuals not encountered after the occasion they are first marked, a common feature of wildlife populations. Temporal symmetry models combine forward- and reverse-time modeling to estimate recruitment (f) and the finite rate of population change (λ). Multi-strata models include dynamic categorical information and offer state-specific estimates of apparent survival and encounter rates, as well as probabilities of changing states (Ψ). Robust design models subdivide sampling occasions into shorter periods, and they partition encounter rates (p) into estimates of temporary emigration (γ = 1 − δ) and true detection (p*). Joint models combine live encounters with other sources of information, including dead-recovery data, and decompose apparent survival into estimates of true survival (S) and site fidelity (F). Cormack-Jolly-Seber and multi-strata models have a large literature, but many of the advanced models have not yet received widespread use. In the future, wildlife ecologists should design field studies that take advantage of the best possible statistical procedures now that a range of alternative models and software tools are available.

Program MARK provides >100 models for the estimation of population parameters from mark–encounter data. The multistate model of Brownie et al. (1993) and Hestbeck et al. (1991) allows animals to move between states with a probability of transition. The simplest multistate model is an extension of the Cormack–Jolly–Seber (CJS) live recapture model. Parameters estimated are state-specific survival rates and encounter probabilities and transition probabilities between states. The multistate model provides a valuable framework to evaluate important ecological questions. For example, estimation of state-specific survival and transition probabilities between the biological states of breeders and nonbreeders allows estimation of the cost of reproduction. Transitions between physical states, such as spatial areas, provide estimates needed for meta-population models. The basic multistate model uses only live recaptures, but 3 extensions are included in MARK. A multistate model with live and dead encounters is available, although the dead encounters are not state specific. Robust-design multistate models are also included in MARK, with both open and closed robust designs. These models assume that animals move between states only between primary sessions of the robust design. For the closed robust design, we can specify 12 different data types for the modeling of encounter probabilities during the primary session, including 6 versions of the closed model likelihood incorporating population size (N) directly in the likelihood, and 6 versions of the Huggins model in which N is estimated as a derived parameter outside the likelihood. One assumption that is generally necessary to estimate state-specific survival rates in the multistate model is that transitions take place immediately before encounter occasions. Otherwise, survival rates over the interval between encounter occasions are a mix of survival rates over multiple states. Advantages of using MARK to estimate the parameters of the various multistate models include flexibility of model specification to include group, time, and individual covariates, estimation of variance components, model averaging of parameter estimates, and Bayesian parameter estimation using Markov chain Monte Carlo procedures on the logit scale.

Survival estimation is an important aspect of population ecology and conservation biology, and radiotelemetry is a major tool for assessing factors influencing survival time in free-ranging birds and mammals. Despite the advantage of telemetry in providing extensive and continuous survival information, telemetry-based survival estimates can be biased or imprecise when methods are misused. Simple cumulative survival estimators like the Heisey and Fuller and Kaplan-Meier methods have underlying assumptions and sampling requirements that commonly remain unverified by researchers. Telemetry studies often limit survival analysis to simple univariate tests that do not consider the range of factors potentially influencing mortality risk in free-ranging animals. Continuous-time modeling approaches like Cox Proportional Hazards or Anderson–Gill methods, or their discrete-time analogues, are superior because they are robust to a range of study design limitations and can handle multiple categorical or continuous covariates including those that vary with time or subject age. Parametric models may be difficult to fit in telemetry studies because the appropriate hazard function in wildlife populations usually is not known. The main assumptions in survival study design are that 1) subjects represent the population of interest, 2) mortality risk is independent between subjects, and 3) subjects are lost to follow-up (i.e., censored) randomly. These assumptions are prone to violation in telemetry research, and their assessment and possible remediation should be prioritized. Telemetry studies often are characterized by small sample size or short duration; both attributes lead to low numbers of mortalities and thus lack of precision in the survival estimate. I conclude that telemetry-based survival estimation will benefit from increased emphasis on modeling approaches designed to elucidate survival determinants in complex systems, combined with more rigorous attention to basic assumptions and study design limitations.

Estimating cause-specific mortality is often of central importance for understanding the dynamics of wildlife populations. Despite such importance, methodology for estimating and analyzing cause-specific mortality has received little attention in wildlife ecology during the past 20 years. The issue of analyzing cause-specific, mutually exclusive events in time is not unique to wildlife. In fact, this general problem has received substantial attention in human biomedical applications within the context of biostatistical survival analysis. Here, we consider cause-specific mortality from a modern biostatistical perspective. This requires carefully defining what we mean by cause-specific mortality and then providing an appropriate hazard-based representation as a competing risks problem. This leads to the general solution of cause-specific mortality as the cumulative incidence function (CIF). We describe the appropriate generalization of the fully nonparametric staggered-entry Kaplan–Meier survival estimator to cause-specific mortality via the nonparametric CIF estimator (NPCIFE), which in many situations offers an attractive alternative to the Heisey–Fuller estimator. An advantage of the NPCIFE is that it lends itself readily to risk factors analysis with standard software for Cox proportional hazards model. The competing risks–based approach also clarifies issues regarding another intuitive but erroneous “cause-specific mortality” estimator based on the Kaplan–Meier survival estimator and commonly seen in the life sciences literature.

We conducted a 13-year survival (i.e., time survived since birth) and cause-specific mortality study, divided into 2 phases (Phase I = years 1–6; Phase II = years 7–13), of 302 female white-tailed deer (Odocoileus virginianus) ≥0.6 years old at capture. The study spanned a period of extreme variability in winter severity (maximum winter severity indexes [WSI] of 45–195) and hunting pressure. Most studies of survival and cause-specific mortality of northern deer have assumed constant survival rates for adults of each sex (≥1.0 yr old pooled) and examined fawns (0.6 ≤ x ≤ 1.0 yr old) separately. We observed U-shaped hazard (i.e., instantaneous risk of death) curves for both phases of the study, indicating that risk of death is highest for younger and older individuals. The estimated hazard for Phase II was generally lower and relatively constant for adults 2–10 years old compared to Phase I, where the instantaneous risk of death began to increase at age 6 years. This difference likely reflected differences in winter severities, associated changes in magnitude of wolf (Canis lupus) predation, and changes in hunting pressure between the 2 phases. The age distribution of our study cohort was relatively stable over the study period. Subsequently, when we included 76 neonates (i.e., ≤0.6 yr old) in the study cohort, the descending arm of the all-causes hazard began its descent at a hazard rate of 2.3 (vs. 1.0 without neonates), clearly demonstrating that the greatest risk of mortality occurs in the first year of life. We compared cumulative survival estimates for these data using the generalized Kaplan–Meier (GKM) and the iterative Nelson estimator (INE), and we illustrate the potential for bias when applying the GKM to left-truncated data. Median age of survival for females was 0.83 years old (90% CI = 0.79–1.45 yr old) using the INE and 0.43 years old (90% CI = 0.17–0.78 yr old) using the GKM. Lastly, we used a simulation approach to examine the potential for bias resulting from pooling adults. These simulations suggest that models using the constructed discrete time variable give nearly unbiased survival estimates and provide support for researchers and managers applying age-specific hazards derived during study periods to determine the reliability of adult age-pooled survival estimates. As indicated by our data, it is important to consider environmental variation and its interactions with natural mortality forces (e.g., predation) and age distribution of the population when setting harvest goals.

Our objective was to determine rates of rabies infection in bats submitted to the Indiana State Department of Health (ISDH) between species, over time, and in normally behaving bats. Those are important questions for public health reasons and also for protection of bats. Bats were tested by ISDH by using the immunofluorescent method. We tested 8,262 bats for rabies at the ISDH from 1966 to 2003, of which 445 (5.4%) tested positive. The 2 most common species, the big brown bat (Eptesicus fuscus; n = 5,584; 173 [3.1%] rabid), and the red bat (Lasiurus borealis; n = 1,512; 169 [11.2%] rabid), accounted for 85.9% of submissions. We found the highest rates of rabies in the hoary bat (L. cinereus; 54 of 178 [30.3%] rabid) and eastern pipistrelle (Pipistrellus subflavus; 41 of 314 [13.1%] rabid). Rates of rabies in other species ranged from 3.7% in the silver-haired bat (Lasionycteris noctivagans) to zero in Indiana myotis (Myotis sodalis) and evening bats (Nycticeius humeralis). The above data were mostly for incapacitated bats that people found sick or dead rather than normally behaving bats. None of 259 normally behaving big brown bats examined from areas where rabid bats had occurred were rabid. Levels of rabies in the big brown bat remained relatively stable throughout the study period with peak activity during late summer and fall.

On islands in the lower Florida Keys, endangered Key deer (Odocoileus virginianus clavium) aggregate in urban areas where deer densities are high, potentially increasing browsing impacts in these areas. To quantify impacts, we compared plant densities and diversity in hardwood hammocks located adjacent to (urban) and distant from (exurban) developed areas on islands with high (>17 deer/km²) and low (<1 deer/km²) Key deer densities. Although we detected no differences in total plant densities or diversity on islands with high densities of deer, we observed lower mean (±SE) densities of deer-preferred plant species within reach of deer in hammock stands in urban areas (84.9 ± 35.9/ha) than exurban areas (694.5 ± 428.4/ha). No such patterns were evident on islands with low deer densities. We also experimentally tested deer browsing intensity on islands of high deer density using potted plants. We found that mean (±SE) browsing pressures (measured as proportion of leaves remaining on potted plants) were higher on preferred plant species in hammocks near urban areas (0.451 ± 0.226) than exurban areas (0.591 ± 0.230). Stimuli (e.g., human handouts) that attract deer to urban locales should be controlled and further development restricted to deter heavy browsing pressures by Key deer near urban areas and to prevent degradation of state-imperiled hardwood hammock plant communities.

We describe landscape use of female brown bears (Ursus arctos) on the Kenai Peninsula, Alaska, USA. Radiocollars, fitted to 43 adult female brown bears, provided radio relocations, which we used to describe habitat use patterns by season and reproductive class at the landscape scale. Brown bears were associated with areas with low densities of human developments and roads, as well as riparian areas that were close to cover. Presence of streams and lakes that supported spawning salmon (Oncorhynchus spp.) positively influenced summertime distribution of bears. Female brown bears with cubs avoided concentrations of other bears. Resource managers may use this information to respond to brown bear conservation issues associated with increasing human populations and associated development in the establishment of road density standards, seasonal road closures, management of recreation sites, and vegetation management on the Kenai Peninsula.

Currently, 11 western states and 2 Canadian provinces use sport hunting as the primary mechanism for managing cougar (Puma concolor) populations. Yet the impacts of sustained harvest on cougar population dynamics and demographic structure are not well understood. We evaluated the effects of hunting on cougar populations by comparing the dynamics and demographic composition of 2 populations exposed to different levels of harvest. We monitored the cougar populations on Monroe Mountain in south-central Utah, USA, and in the Oquirrh Mountains of north-central Utah from 1996 to 2004. Over this interval the Monroe population was subjected to annual removals ranging from 17.6–51.5% (mean ± SE = 35.4 ± 4.3%) of the population, resulting in a >60% decline in cougar population density. Concurrently, the Oquirrh study area was closed to hunting and the population remained stationary. Mean age in the hunted population was lower than in the protected population (F = 9.0; df = 1, 60.3; P = 0.004), and in a pooled sample of all study animals, females were older than males (F = 13.8; df = 1, 60.3; P < 0.001). Females from the hunted population were significantly younger than those from the protected population (3.7 vs. 5.9 yr), whereas male ages did not differ between sites (3.1 vs. 3.4 yr), suggesting that male spatial requirements may put a lower limit on the area necessary to protect a subpopulation. Survival tracked trends in density on both sites. Levels of human-caused mortality were significantly different between sites (χ² = 7.5; P = 0.006). Fecundity rates were highly variable in the protected population but appeared to track density trends with a 1-year lag on the hunted site. Results indicate that harvest exceeding 40% of the population, sustained for ≥4 years, can have significant impacts on cougar population dynamics and demographic composition. Patterns of recruitment resembled a source–sink population structure due in part to spatially variable management strategies. Based on these observations, the temporal scale of population recovery will most likely be a function of local harvest levels, the productivity of potential source populations, and the degree of landscape connectivity among demes. Under these conditions the metapopulation perspective holds promise for broad-scale management of this species.

Scientists commonly use resource selection functions (RSFs) to identify areas important to large herbivores. Defining availability of resources is scale dependent and may limit inference on biological mechanisms of selection, particularly if variation in selection of resources is high among individuals within a population. We used logistic regression, the information-theoretic approach, and Global Positioning System (GPS) radiotelemetry data from 10 female woodland caribou (Rangifer tarandus caribou) and 5 wolf (Canis lupus) packs to model resource selection by individual caribou in the winter and late-winter seasons. We evaluated the influence of spatial scale on the relative importance of cost of movement and components of predation risk. We examined attributes of the risk of wolf predation within availability data at 2 spatial scales, and quantified variation in resource selection among individual caribou. Energetic cost of movement was the most important covariate for all caribou at a spatial scale defined by seasonal movement. Increasing distance to areas of high wolf risk was more important at the larger spatial scale of home range. Variation was high in the selection of resources among caribou, although commonalities among individuals enabled pooling data on use and availability into 2 selection strategies. Researchers and managers should conduct multiscale analyses with varied definitions of availability, quantify variation among individuals, and pool data into common selection strategies to identify mechanisms of selection and to map a population's selection for resources on the landscape.

Roads and traffic have a multitude of impacts on wildlife populations. Wildlife existing within the confines of fragmented reserves are particularly susceptible to fatalities on roads, especially those situated within urban and semirural matrices. The sustainability of many wildlife populations within reserve fragments are tenuous as roads further subdivide reserved areas and increase the frequency of animal–vehicle contact. Although many studies have assessed the quantity and diversity of fatalities from collisions, few studies have examined the long-term viability of wildlife populations living adjacent to roads. We chose to examine the effects of disturbances, including fatalities on roads, on a population of swamp wallabies (Wallabia bicolor) within the Royal National Park on the urban fringe of Sydney, Australia. Despite having an extensive range, researchers suspect that many local populations of this sole member of Wallabia are in decline. We used a combination of population modeling and sensitivity analysis to assess the impact of disturbances on the population. Under current conditions, the forecast of the population was to decline over the next 100 years with the possibility of becoming extinct. We found that female reproduction and breeding were most influential on the population model. Of the range of management options investigated, by far the most rewarding was the reduction of fatalities on roads, as only a 20% decrease in female fatalities on roads has the potential to reverse the current decline and represents the best option for maintaining long-term viability. We suggest that documentation and subsequent management of road impacts, within the context of other threats, is essential to the conservation of similar species in road-affected environments.

Large-scale, multispecies monitoring programs are widely used to assess changes in wildlife populations but they often assume constant detectability when documenting species occurrence. This assumption is rarely met in practice because animal populations vary across time and space. As a result, detectability of a species can be influenced by a number of physical, biological, or anthropogenic factors (e.g., weather, seasonality, topography, biological rhythms, sampling methods). To evaluate some of these influences, we estimated site occupancy rates using species-specific detection probabilities for meso- and large terrestrial mammal species on Cape Cod, Massachusetts, USA. We used model selection to assess the influence of different sampling methods and major environmental factors on our ability to detect individual species. Remote cameras detected the most species (9), followed by cubby boxes (7) and hair traps (4) over a 13-month period. Estimated site occupancy rates were similar among sampling methods for most species when detection probabilities exceeded 0.15, but we question estimates obtained from methods with detection probabilities between 0.05 and 0.15, and we consider methods with lower probabilities unacceptable for occupancy estimation and inference. Estimated detection probabilities can be used to accommodate variation in sampling methods, which allows for comparison of monitoring programs using different protocols. Vegetation and seasonality produced species-specific differences in detectability and occupancy, but differences were not consistent within or among species, which suggests that our results should be considered in the context of local habitat features and life history traits for the target species. We believe that site occupancy is a useful state variable and suggest that monitoring programs for mammals using occupancy data consider detectability prior to making inferences about species distributions or population change.

We derived 3 rangewide mark–recapture estimates for American black bears (Ursus americanus) in Minnesota, USA, over a 12-year period (1991–2002) to aid in management of this hunted population. We used tetracycline-laden baits to biomark 700–1,200 bears over an area of 85,000 km² during each of 3 summer marking sessions. Ingested tetracycline produced long-lasting marks in bones and permanent marks in teeth. We collected bone and tooth samples from bears harvested in both the year of marking and the following year and examined them for tetracycline marks. For each marking event, the cumulative recapture sample totaled 3,400–4,800, or 14–30%, of the estimated population. Estimates of population size, derived from a Lincoln–Petersen (L-P) estimator, pertained to the time of marking; loss of animals from the population after marking was thus inconsequential, so long as losses were unbiased with respect to marking. We detected tetracycline marks in a greater proportion of samples collected in the year of marking than the following year. We surmised that the harvest included a disproportionate number of marked bears compared to unmarked bears in the year of marking, probably because both marking and harvest involved the bears' attraction to bait. After removal of these most vulnerable marked bears, the proportion of marked bears declined the next year. We used population reconstruction and a data-driven model to help gauge the potential magnitude and direction of biases resulting from sex–age heterogeneity in mark–recapture estimates based on data from the first year's harvest (biased low), second year's harvest (biased high), or some combination of the two. Uncertainty regarding yearly estimates hampered assessment of population trend. We recommend biomarking with >1 year of recapture data for large-scale population estimation; however, we suggest modifications to our methodology to obtain larger and more representative samples of the population, thereby increasing accuracy and precision.

During waterfowl hunting a large number of birds are shot but not instantly killed. Some will die within a few days as a direct consequence of heavy injuries, whereas another proportion is only lightly injured and will survive for an extended period of time although their survival may still be affected. We predicted that embedded body shot (when not instantly lethal) will cause reduced body condition among common eiders (Somateria mollissima), as we assumed such birds to be physical disadvantaged (e.g., as to mobility and foraging) from their injuries. Among birds collected during 3 winters (2000–2002) by Inuit hunters and fishermen in Nuuk, southwest Greenland, we X-rayed and dissected 762 common eiders to extract information about embedded lead shot and body condition. After adjusting for structural body size, year and date of sampling, habitat, and sampling method, we found that embedded lead shot had a significant effect on juvenile body condition. Wounded juvenile birds carried on average 19% less fat than unwounded juveniles. In accordance with a priori predictions, we did not detect an impact of wounding on the body condition of immature and adult birds. For most of these older birds, the shooting incidence took place >1 year before they were collected, and the insignificant test results indicated the absence of a long-term effect on body condition once birds survived the initial effect of wounding. For juvenile birds, the wounding effect most likely added to other causes of mortality; however, additional knowledge about natural mortality is required to estimate the net consequence on population dynamic.

We aimed to clarify the factors that affect Japanese hare (Lepus brachyurus) densities and habitat availability in certain vegetation types. Using fecal pellet counts, we found that hare densities and habitat availability were higher in open–tree-canopy habitats with a dense ground cover of herbs and grasses, such as cleared areas in artificial forests. However, when we examined the relationships between fecal pellets as an indicator of hare density and various attributes of forest floor vegetation, no vegetation parameters had a significant correlation with fecal pellets among different plant communities. Thus, it is difficult to predict hare density based on forest floor vegetation, in general. However, in seven Japanese cedar (Cryptomeria japonica) plantation sites, high herb volume index and a low degree of canopy closure, calculated from fisheye-lens photographs, had a significant correlation with fecal pellet density (P = 0.029 and 0.006, respectively). Based on the relationships between fecal pellets and certain forest attributes (e.g., stand height and diameter at breast height of trees), we determined that fecal pellet density was higher in young stands in these plantations. This is likely because young stands have a larger volume of herbaceous species that provide food and shelter for hares, owing to the bright ground conditions resulting from the relatively open canopy. Appropriate plantation control such as cutting and planting trees is important for habitat conservation of L. brachyurus, because planted C. japonica forest occupies more than half of artificial forests in Japan.

Biologists have suggested that prairie dog (Cynomys ludovicianus) towns provide important habitat for several prairie species including swift foxes (Vulpes velox). To examine this issue for swift foxes, we radiocollared and monitored 72 swift foxes in areas occupied by prairie dogs on the Rita Blanca National Grasslands and on fragmented private ranches in northwest Texas from January 2002 to April 2004. We overlaid swift fox locations on vegetation maps and found that swift foxes used prairie dogs towns proportionately less than expected (P < 0.001) in 5 of 6 comparisons based upon availability at both study sites. In one comparison swift foxes used prairie dog towns as expected. We also mapped swift fox den locations and found that den sites were located in prairie dog towns (2 of 106 den sites on national grasslands and 2 of 87 on private ranch for all years pooled) as expected based upon availability (P = 0.54 and P = 0.34). Previous research in these areas of Texas indicated that prairie dogs were a minor component of swift fox diet (<2%). Prairie dog towns do not appear to provide important habitat for swift foxes in northwest Texas or other similar sites because prairie dogs are not a significant source of food for swift foxes, because swift foxes generally avoid prairie dog towns during their normal activities, and because swift foxes only den in prairie dog towns in relation to their availability.

Polar bear (Ursus maritimus) numbers in M'Clintock Channel, Nunavut, Canada have decreased significantly since 1972. We used mark–recapture and recovery data collected from 348 marked polar bears from 1972 to 2000 to estimate demographic characteristics and harvest risks of the M'Clintock Channel polar bear population. Total (harvested) survival rates (±1 SE) from mark–recapture analysis were: 0.62 (±0.15) for cubs of the year, 0.90 (±0.04) for subadults (ages 1–4 yr), 0.90 (±0.04) for adult (age ≥5 yr) females, and 0.88 (± 0.04) for adult males. Mean litter size was 1.68 ± 0.15 cubs with a mean reproductive interval of 2.8 ± 0.2 years. By 6 years of age, on average 0.29 ± 0.47 females were producing litters; mean litter production rate for females aged >6 years was 0.93 ± 0.33. We estimated total abundance to average 284 ± 59.3 bears, of which 166.9 ± 35.4 individuals were female and 117.2 ± 26.4 were male. We incorporated our standing age and mark–recapture demographic parameters as input into a harvest risk analysis designed to account for demographic, environmental, and sampling uncertainty. Population growth rate was 0.946 ± 0.038 for the period 1993–1999. A harvest quota not exceeding 3 bears/year is required if the population is to increase in the short term. Slightly higher quota options are available if increased risk and recovery times are accepted by stakeholders.

Point counts are one of the most widely used and efficient approaches to survey land-bird populations. A new approach to point-count surveys involves the use of 2 observers, which allows the calculation of a detection probability for each bird species. Detection probabilities derived from 2 observers permit investigators to calculate a corrected abundance estimate that accounts for birds present but not detected. We evaluated 2 double-observer point count approaches: the dependent-observer approach and the independent-observer approach. The dependent-observer approach involves 2 observers recording data together on a single data sheet with one observer designated the primary observer and the other designated as the secondary observer. The primary observer verbally dictates the number of each species detected while the secondary observer records this information; the secondary observer also records birds that the primary observer did not detect. The independent-observer approach involves 2 observers recording data independently on separate data sheets without verbal communication between observers. This study compares the detection probabilities and associated levels of precision generated by both double-observer approaches to ascertain which technique generates data that are more accurate and more feasible to apply in the field. We conducted point counts at 137 point locations in northern West Virginia during the spring of 2000 and 2001 using both double-observer approaches. We generated detection probabilities and abundances from data collected using both double-observer approaches using program DOBSERV. The dependent-observer approach resulted in higher observer-specific and joint detection probabilities, as well as lower standard errors of detection probability across most cover types. Species-specific detection probabilities were higher across all cover types under the dependent-observer approach. Given the higher detection probabilities and associated precision combined with fewer logistical constraints, we suggest that the dependent-observer approach be used when investigators are interested in surveying birds using point counts.

Lesser scaup (Aythya affinis) populations have been declining since the late 1970s. One of the explanations to account for this decline, the spring-condition hypothesis (SCH), is based on the premise that scaup are limited by their ability to acquire or maintain nutrient reserves during migration to the breeding grounds, leading to an impairment of their reproductive potential. Available evidence suggests that endogenous reserves required for reproduction are obtained at a later stage of migration or after arrival at the breeding grounds, not wintering sites. However, only one study has addressed body-condition levels on a southern wintering site in the last decade, with results limited to the wintering grounds on the Mississippi Flyway. We documented foraging behavior, nutrient levels, and body mass of lesser scaup in east-central Florida, USA, where 62% of the Atlantic Flyway population overwinters, during the winters of 2002 and 2003. Diurnal foraging did not increase seasonally. Nocturnal foraging increased seasonally by 76% or 43 minutes per night in females and by 478% or 1.9 hours per night in males. Measures of body condition did not change seasonally during 2002 for either sex. Between early and later winter in 2003 corrected body mass (CBM) and lipid reserves of male scaup increased 77 g and 39 g, respectively. Our results suggest that lesser scaup maintain or may slightly improve their physiological condition in east-central Florida during winter. Lower body mass and differences in nutrient levels in east-central Florida, compared to a wintering site in Louisiana, likely stem from geographic variation and lower thermal requirements associated with the warmer Florida environment. Lesser scaup depart Florida with sufficient reserves to initiate spring migration, but they maximize nutrient reserves used during reproduction elsewhere during migration or on the breeding grounds. These results suggest that maintaining the ecological integrity of this wintering ground is critical in minimizing winter mortality and preventing it from becoming an ancillary factor in current declines. Future research should address understanding survival rates during spring migration and at critical staging areas to provide new insight into the ramifications of scaup leaving wintering habitats such as MINWR with lower body condition than at other wintering sites in other flyways.

Substantial efforts have been made to identify the most effective practices for the control and management of invasive vertebrate pest species, such as the feral pig (Sus scrofa). We investigated the demographics, abundance, and molecular ecology of a persecuted feral pig population that was subjected to control. We then applied methodologies to determine if we could retrospectively quantify any changes in the population structure or dynamics of these pigs. Feral pig demographic and abundance parameters indicated that in this population of feral pigs, there were very few detectable changes between the two aerial culling years. We observed this despite environmental conditions being optimal for control. Genetic results indicated that pigs culled in the latter 2004 cull were genetically identical to those pigs that inhabited the area a year earlier. The genetic population was geographically larger than the sample area. These findings indicate that the recovery in feral pig density witnessed in the controlled area was not a result of reinvasion from a separate, genetically distinct population, but rather, it was the result of reinvasion from feral pigs outside the study area but within the same genetic population. Importantly, we were unable to detect any recent genetic bottlenecks. This approach has considerable potential for auditing the effectiveness of control programs of pest species and assessing the feasibility of impacting upon or locally eradicating many other free-ranging pest species.

The lack of adequate documentation of wildlife translocations, particularly details regarding the source stock used, have potentially serious implications for wildlife management. Poor documentation of translocations may lead to unintentional mixing of distinct types, potentially causing problems for future management, design of harvest programs, and evolutionary stability. In Kansas, we employed molecular tools and assignment methods to uncover the cryptic distribution of wild turkey subspecies resulting from decades of poorly documented translocations. Pure forms of Eastern (Meleagris gallopavo silvestris) and Rio Grande turkeys (Meleagris gallopavo intermedia) remain in many portions of the state, and future translocation programs now have the option to keep these distinct by prohibiting translocations between regions containing different subspecies. In addition, we documented 3 zones of hybridization: 1) at the interface between Eastern and Rio Grande subspecies, 2) in southwest Kansas where immigrant Merriam's turkeys (Meleagris gallopavo merriami) are mixing with translocated Rio Grande turkeys, and 3) surrounding an undocumented translocation of Eastern turkeys within a region characterized primarily by Rio Grande turkeys. The DNA-based techniques employed in this study were extremely informative tools for characterizing the distribution of wild turkeys in Kansas and suggest that such tools could be applied in a multitude of similar situations in other wildlife species.

A decline in red knots (Calidris canutus rufa) has been attributed to horseshoe crab (Limulus polyphemus) egg shortages on the Delaware Bay, an important foraging area for migrating knots. We studied the movements and distribution of 65 radiotagged red knots on Delaware Bay from May to June 2004 and related movements to the distribution and abundance of horseshoe crab eggs and other prey and to other habitat characteristics. The number of horseshoe crab eggs was the most important factor determining the use of Delaware Bay beaches by red knots (logistic regression cumulative Akaike's Information Criterion adjusted for small sample size [AIC_c] w = 0.99). The knots shifted from emergent marsh and peat-beaches to sandy Delaware Bay beach when crab eggs became abundant, which also suggested the importance of crab eggs. While red knots used sandy beach zones more than expected, given their availability, 44% of red knot low tide locations were in bay and coastal emergent marsh. The abundance of Donax variabilis (AIC_c w = 0.95) and Mytilus edulis (AIC_c w = 0.94) spat, both food for red knots, had a relationship with red knot use of sandy beaches. Levels of disturbance and the abundance of laughing gulls (Larus atricilla) also were important factors in red knot sandy beach use, although secondary to prey resources (AIC_c w < 0.4). These results are consistent with the hypothesis that the abundance of horseshoe crab eggs on sandy beaches is driving movement and distribution of red knots and that there is little alternative food during the migratory stopover in Delaware Bay. Our findings that red knots disproportionately use Delaware Bay sites with abundant eggs and that there is a lack of surplus eggs at areas used and unused by red knots support the continuation of management for sustained yield of horseshoe crabs and other food resources at this stopover.

As part of a national experiment, the Fire and Fire Surrogate Project, we evaluated the effects of forest thinning on small mammal population densities and total small mammal biomass in ponderosa pine (Pinus ponderosa)–dominated forests at 2 study areas in northern Arizona and northern New Mexico, USA. We also evaluated the effects of wildfire on small mammal population densities and biomass after a wildfire burned a portion of one study area. Our statistical methods consisted of estimation of population densities in combined analyses across space and time, followed by a weighted regression analysis of treatment effects on densities. We hypothesized that habitat change postdisturbance would be the critical determinant of population responses to thinning and wildfire within 1 year of disturbances. Our results largely supported this hypothesis, as we documented predicted positive responses to thinning for deer mice (Peromyscus maniculatus), gray-collared chipmunks (Tamias cinereicollis), and least chipmunks (T. minimus). We also observed predicted positive responses to wildfire for deer mice, although our results did not support predicted negative responses to wildfire for least chipmunks. Total small mammal biomass generally increased following both thinning and wildfire. Our results suggest that fuel reduction treatments will have the largest positive impact on small mammal populations in areas where tree densities are especially high.

Despite the ongoing worldwide trade in reptiles, monitoring and management systems are only recently being developed for exploited lizards and snakes. We studied the tegu lizard (Tupinambis spp.) management and harvest monitoring program in Paraguay as a model for commercially exploited reptile management. Tegu lizards are among the most exploited reptiles in the world, with current quotas for Paraguay and Argentina of 300,000 and 1,000,000, respectively. We analyzed demographic data obtained from harvested skins of 2 species of tegu lizards at 2 trade levels (i.e., check stations and tanneries) over an 8-year period. There was significant annual variation in the number, size, and sex ratios of harvested tegus. Sex ratios were consistently male biased. The proportion of small skins in the harvest decreased over time, indicating compliance with the policy to ban commerce of small tegu skins. However, disparities among check stations and tanneries indicated that small skins might have been restretched by middlemen as they were bartered up the trade chain. Our results support the conclusion that tegu lizards are withstanding the harvest in Paraguay. There was no chronological trend of increased take or decreasing size that would indicate overharvest, and there were no indicators of local population decline. However, data on hunter effort are lacking, and it is important to implement a system for quantifying hunter effort and incorporate those data into harvest models. This is the first study to report long-term data on a harvested lizard. The approach we used here can serve as a model for monitoring and management of other commercially exploited terrestrial reptiles.

Although partial forest harvesting is practiced over large areas, managers know little about its impacts on sciurid rodents, particularly on northern (Glaucomys sabrinus) and southern flying squirrels (G. volans) in the northeastern United States and Canada. We examined habitat relationships of sciurid rodents (northern flying squirrels, southern flying squirrels, red squirrels [Tamiasciurus hudsonicus], and eastern chipmunks [Tamias striatus]) at 2 spatial scales in managed and unmanaged coniferous and hardwood forests of Algonquin Provincial Park, Ontario, Canada. We live-trapped rodents in 26 northern hardwood stands and in 16 white pine (Pinus strobus) stands from 2002 to 2004. Northern flying squirrel and red squirrel densities were significantly lower in recently harvested (3–10 yr since harvest) shelterwood stands than in unmanaged stands. In contrast, southern flying squirrel densities were higher in selection-harvested stands than in old-forest areas. The densities of northern flying squirrels and red squirrels had a strong relationship with the density of large spruce (Picea sp.) and hardwood trees and snags in conifer sites. Southern flying squirrel numbers had a positive association with the density of mast trees at the landscape level but not at the stand level in hardwood forests. Eastern chipmunk density had a positive correlation with the volume of old downed woody debris and the stems per hectare of declining trees. We recommend forest managers retain more large spruce and hardwood trees to mitigate the impacts of shelterwood harvesting on northern flying squirrels and red squirrels, and that they maintain high mast availability at the landscape level to ensure the persistence of southern flying squirrels.

Society faces a growing challenge in the management of zoonotic wildlife diseases. Unique attributes of zoonotic diseases and the shifting sociocultural contexts within which diseases are experienced create serious challenges for managers. We address 2 critical questions: how do uncertainty and severity associated with a zoonotic disease relate to public management attitudes and preferences, and do immigrant attitudes and preferences differ from those of long-term residents in rural areas of the Intermountain West? We addressed these questions using a personally administered questionnaire in Teton County, Idaho, USA. Based on 2 hypothetical zoonotic wildlife diseases, one less severe and more known (type A) and one more severe and less known (type B), we asked respondents to indicate their agreement with 13 statements regarding their perception of the disease and management preferences. We also asked respondents to indicate their support for different groups controlling management. Our compliance rate was 95% (n = 416, sampling error ± 4.8%). Respondents considered type B a greater risk to human and livestock health, and supported using lethal control methods, except hunting, to control it. Disease type, however, had less impact on public support for management options involving fencing and supplemental feeding. With only 2 exceptions, longer-term residents (LTR) supported lethal management options more than newer residents (NR). Further, NR hunted less than LTR, thereby restricting lethal management options. Respondents indicated some level of support for all management control options except giving authority to local civic leaders. Newer residents showed higher support for wildlife scientists and federal agencies making management decisions, whereas LTR preferred state livestock agencies. Demographic change in rural areas may lead to higher levels of support for federal and scientific control over zoonotic disease management but lower support for lethal management and ability to enact lethal management. Our results suggest 2 critical management needs: solicitation and consideration of public input for type A zoonotic disease management and promotion of hunting or developing a viable lethal management alternative.

Shellfish aquaculture is an expanding industry in coastal British Columbia, Canada, and occurs in important wintering areas for surf scoters (Melanitta perspicillata) and white-winged scoters (M. fusca). We quantified habitat use by scoters in relation to natural environmental attributes and habitat modifications associated with shellfish aquaculture. We found that, despite the extensive clam and oyster farming in our study area, densities of wintering surf scoters and white-winged scoters were related primarily to natural environmental attributes, particularly intertidal area, clam density, and sediment type; shellfish aquaculture variables were generally poor predictors of bird densities. We conclude that current levels and forms of shellfish aquaculture in our study site were not an important determinant of scoter distribution and abundance, suggesting that winter scoter populations and the shellfish aquaculture industry may be mutually sustainable. We caution that intensification or further industrialization of shellfish aquaculture in British Columbia could eventually lead to detrimental effects if some threshold level of habitat modification is exceeded.

We developed a habitat model to predict cerulean warbler (Dendroica cerulea) habitat availability in the Cumberland Mountains of eastern Tennessee. We used 7 remotely sensed vegetation and topographic landform explanatory variables and known locations of territorial male cerulean warblers mapped in 2003 as the response variable to develop a Mahalanobis distance statistic model of potential habitat. We evaluated the accuracy of the model based on field surveys for ceruleans during the 2004 breeding season. The model performed well with an 80% correct classification of cerulean presence based on the validation data, although prediction of absence was only 54% correct. We extrapolated from potential habitat to cerulean abundance based on density estimates from territory mapping on 8 20-ha plots in 2005. Over the 200,000-ha study area, we estimated there were 80,584 ha of potential habitat, capable of supporting about 36,500 breeding pairs. We applied the model to the 21,609-ha state-owned Royal Blue Wildlife Management Area to evaluate the potential effects of coal surface mining as one example of a potential conflict between land use and cerulean warbler conservation. Our models suggest coal surface mining could remove 2,954 ha of cerulean habitat on Royal Blue Wildlife Management Area and could displace 2,540 breeding pairs (23% of the Royal Blue population). A comprehensive conservation strategy is needed to address potential and realized habitat loss and degradation on the breeding grounds, during migration, and on the wintering grounds.

Spatial relationships among stocks of breeding waterfowl can be an important component of harvest management. Prediction and optimal harvest management under adaptive harvest management (AHM) requires information on the spatial relationships among breeding populations (fidelity and inter-year exchange), as well as rates of movements from breeding to harvest regions. We used band-recovery data to develop a model to estimate probabilities of movement for American black ducks (Anas rubripes) among 3 Canadian breeding strata and 6 harvest regions (3 in Canada, and 3 in the United States) over the period 1965–1998. Model selection criteria suggested that models containing area-, year-, and age-specific recovery rates with area- and sex-specific movement rates were the best for modeling movement. Movement by males to southern harvest areas was variable depending on the originating area. Males from the western breeding area predominantly moved to the Mississippi Flyway or southern Atlantic Flyway (Ψ_ij = 0.353, SE = 0.0187 and Ψ_ij = 0.473, SE = 0.037, respectively), whereas males that originated in the eastern and central breeding strata moved to the northern Atlantic flyway (Ψ_ij = 0.842, SE = 0.010 and Ψ_ij = 0.578, SE = 0.0222, respectively). We used combined recoveries and recaptures in Program MARK to estimate fidelity to the 3 Canadian breeding strata. Information criteria identified a model containing sex- and age-specific fidelity for black ducks. Estimates of fidelity were 0.9695 (SE = 0.0249) and 0.9554 (SE = 0.0434) for adult males and females, respectively. Estimates of fidelity for juveniles were slightly lower at 0.9210 (SE = 0.0931) and 0.8870 (SE = 0.0475) for males and females, respectively. These models have application to the development of spatially stratified black duck harvest management models for use in AHM.

Although the Florida panther (Puma concolor coryi) has a history of study spanning more than 2 decades, some researchers have criticized previous habitat analyses of the panther for using daytime telemetry locations to infer 24-hour habitat use, selective use of radiotagged animals, comparison of animal locations to an inappropriate set of available resources, use of land cover maps for time periods that poorly matched some animal locations, use of locations instead of the individual animal as the sampling unit, and failure to account for telemetry error. To address all but the first of these concerns, we analyzed over 57,000 radiolocations of 100 Florida panthers and 8 introduced Texas cougars (Puma concolor stanlyana) collected from 1981 to 2003 using a Euclidean distance-based analysis (EDA), a technique that measures linear distances between telemetry locations and habitat types to determine nonrandom habitat use. We matched panther locations to 1 of 3 land cover maps reflecting cover conditions spanning 23 years to determine second- and third-order habitat selection. Panthers selected a mosaic of habitats when establishing a home range, and selected forests and avoided open wetlands within their home range. We recommend that managers give forests highest consideration when conserving this endangered species and suggest the further evaluation of the habitat value of natural openings that occur in close proximity to forested panther range.