Little information is available on the winter ecology of the small, geographically isolated, genetically-unique population of great gray owls (Strix nebulosa) in the central Sierra Nevada, California. This population is comprised of facultative, elevational winter migrants and access to winter habitat is an important component of their ecology. Winter observations and remotely sensed habitat variables were used to inform a predictive model of the environmental requirements and geographic distribution of this owl population. Using the modeled distribution map we assessed the distribution of 20% probability of occurrence classes relative to owl habitat associations, ownership, current development, and projected future development patterns. Our findings indicate that high probability class (81–100%) areas and the broader joint medium/medium-high/high probability class (41–100%) areas are uncommon on the landscape (0.2% and 5.0% of study area, respectively). High probability areas were characterized by Sierran Yellow Pine forest surrounding relatively small, flat areas of grassland, wet meadow, and riparian habitats, within the mid-elevation range. Approximately 32% of the high probability areas and 48% of the medium/medium-high/ high probability areas occur on private lands. Of the areas on private lands, 32% of the high probability and 42% of the medium/medium-high/high probability areas occur on currently developed lands. Projected future development on private lands indicated that an additional 12% of the high and 18% of medium/ medium-high/high suitability areas are slated for development by the year 2040. Future conservation planning efforts for the great gray owl in the Sierra Nevada will need to address management issues on both public and private lands. For future planning of development projects around great gray owl wintering habitat, the results from our study supplement current knowledge of breeding distributions to provide land and wildlife managers guidance on conservation priorities.

During the late 20th Century, due to decreases in both contamination and persecution, bald eagle (Haliaeetus leucocephalus) populations increased dramatically. Currently, mechanisms regulating eagle populations are not well understood. To examine potential regulating processes in the Pacific Northwest, where eagles are no longer primarily regulated by contaminants or direct persecution, we examined bald eagle reproductive success, breeding populations, winter populations, mortality, and salmon stream use. Wintering and breeding eagle populations in south-coastal British Columbia (BC) quadrupled between the early 1980s and the late 1990s, and have since stabilized. Density-dependent declines in reproduction occurred during 1986–2009, but not through changes in site quality. Mid-winter survival was crucial as most mortality occurred then, and models showed that density-dependent reductions in population growth rates were partially due to reduced survival. Wintering eagles in British Columbia fed heavily on chum salmon (Oncorhynchus keta) runs, and then switched to birds in late winter, when mortality was highest. Eagles tended to arrive after the peak in salmon availability at streams in BC as part of a migration associated with salmon streams from Alaska to northern Washington. Eagles were most abundant in southern BC during cold Alaskan winters and in years of high chum salmon availability. We suggest that eagle populations in the Pacific Northwest are currently partially limited by density on the breeding grounds and partially by adult mortality in late winter, likely due to reduced late winter salmon stocks forcing eagles to exploit more marginal prey supplies. Larger eagle populations have affected some local prey populations.

Predicting species presence requires knowledge of detection of individuals, scale of model variables, model selection uncertainty, and spatial autocorrelation. Our objective was to incorporate recent modeling advances to predict potential habitat occupancy of northern bobwhite (Colinus virginianus). From 15 May—15 August 2008 and 2009, we conducted repeat-visit surveys at 360 sites within Delaware to sample presence of bobwhite. We randomly selected half the data to model scale-dependent relationships of bobwhite presence with metrics of landscape- and site-scale habitat composition and configuration. The final averaged habitat-occupancy model fit the remainder testing dataset with an area under the receiver operating characteristic curve value of 0.62. At the site scale, bobwhite presence was negatively related to interspersion and juxtaposition of early successional habitat (ESH; grassland and shrubland), ESH to forest edge density, and agriculture to forest edge density, though relative effect sizes were weak to moderate after accounting for model selection uncertainty. At the landscape scale, bobwhite presence was negatively related to patch cohesion of human development within 2.5 km and positively related to patch cohesion of ESH within 2.0 km, with both variables exerting strong effects. Bobwhite presence was also weakly and positively related to percentage of shrubland habitat within 1.0 km of the sampling point. We applied our habitat occupancy model to map the predicted presence of breeding bobwhite within the Delmarva Peninsula, USA. The modeling results and distribution map will provide guidance to State and Federal private land management programs in the Mid-Atlantic to identify where habitat management efforts will be most effective. Our methodology can also serve as a basis for future habitat modeling of bobwhite and other grassland—shrubland species across their range.

Although metal legbands have been an important scientific tool, their use for estimation of harvest and survival relies on samples of dead birds harvested by hunters using shotguns. We hypothesized that the force of steel pellets discharged from a shotgun, within the range of conditions normally experienced by goose hunters, was sufficient to reduce probability of band retention. We conducted 8 experimental trials to estimate retention per round fired at aluminum bands normally applied to arctic-nesting geese in relation to effects of 1) target range (20 m vs. 40 m), 2) steel pellet size (4.57 mm [BB] vs. 3.81 mm [number 2]), 3) cartridge size (76.2 mm [3 in.] vs. 69.9 mm [2.75 in.]), and 4) number of rounds fired (up to 25). There was nearly complete band retention (0.999/round) at 40 m regardless of shot size or shell size used. Retention per round fired at 20 m declined to between 0.984 and 0.987 for number 2 shot and between 0.968 and 0.974 for BB shot. Our conclusions apply to unworn bands, so we recommend further simulations to assess how retention may change with age of bands as they erode or corrode on free-ranging geese. Bias in estimates associated with loss of older bands from shotgun discharge could be adjusted if bias is estimated as done in this article.

Waterfowl and other migratory birds commonly store nutrients at traditional staging areas during spring for later use during migration and reproduction. We investigated nutrient-storage dynamics in the midcontinent population of greater white-fronted geese (Anser albifrons; hereafter white-fronted geese) at spring staging sites in the Rainwater Basin of Nebraska during February–April and in southern Saskatchewan during April–May, 1998 and 1999. In Nebraska, lipid content of white-fronted geese did not increase, and protein content changed little over time for most age and sex categories. In Saskatchewan, lipids increased 11.4 g/day (SE = 1.7) and protein content increased 1.6 g/day (SE = 0.6) in the sample of adult geese collected over a 3-week period. A study conducted during 1979–1980 in the Rainwater Basin reported that white-fronted geese gained 8.8–17.7 g of lipids per day during spring, differing greatly from our results 2 decades later. In addition, lipid levels were less in the 1990s compared to spring 1980 for adult geese nearing departure from staging sites in Saskatchewan. This shift in where geese acquired nutrient stores from Nebraska to more northern staging sites coincided with a decrease in availability of waste corn in Nebraska, their primary food source while staging at that stopover site, and an increase in cultivation of high-energy pulse crops in Saskatchewan. White-fronted geese exhibited flexibility in nutrient dynamics during spring migration, likely in response to landscape-level variation in food availability caused by changes in agricultural trends and practices. Maintaining a wide distribution of wetlands in the Great Plains may allow spring-staging waterfowl to disperse across the region and facilitate access to high-energy foods over a larger cropland base.

Sommers et al. (2010) reported that recolonizing predators increased bovine calf mortality rates in the Upper Green River Cattle Allotment in western Wyoming. However, Sommers et al. (2010) failed to consider multiple competing hypotheses explaining calf loss rates, increasing the likelihood that their results are actually spurious. I reanalyzed their data using a multiple competing hypotheses framework that considered effects of livestock density, summer precipitation, bias in reporting rates, and whether mortality by different predator species was compensatory. I found support for a confounded web of factors influencing calf losses. Calf losses increased with livestock density (which increased during the study), but also during drier summers and with increasing rancher reporting rates. Although both wolves (Canis lupus) and grizzly bears (Ursus arctos) did increase calf losses, the presence of just grizzly bears alone did not significantly increase calf losses. Unconditional estimates of the effects of wolves and grizzly bears on calf losses were only 2.0% (95% CI 0.53–3.81), compared to 3.6% reported by Sommers et al. (2010). Most importantly, however, I report bias in favor of livestock producers in the authors' assumptions that cast further doubt on the rigor of their results. In conclusion, I recommend managers not consider the spurious predator compensation factors reported by Sommers et al. (2010) to be reliable.

Ecological theory predicts that individual survival should vary between sex and age categories due to differences in allocation of nutritional resources for growth and reproductive activities. During periods of environmental stress, such relationships may be exacerbated, and affect sex and age classes differently. We evaluated support for hypotheses about the relative roles of sex, age, and winter and summer climate on the probability of mountain goat (Oreamnos americanus) survival in coastal Alaska. Specifically, we used known-fates analyses (Program MARK) to model the effects of age, sex, and climatic variation on survival using data collected from 279 radio-marked mountain goats (118 M, 161 F) in 9 separate study areas during 1977–2008. Models including age, sex, winter snowfall, and average daily summer temperature (during Jul–Aug) best explained variation in survival probability of mountain goats. Specifically, our findings revealed that old animals (9 yr) have lower survival than younger animals. In addition, males tended to have lower survival than females, though differences only existed among prime-aged adult (5–8 yr) and old (9 yr) age classes. Winter climate exerted the strongest effects on mountain goat survival; summer climate, however, was significant and principally influenced survival during the following winter via indirect effects. Furthermore, old animals were more sensitive to the effects of winter conditions than young or prime-aged animals. These findings detail how climate interacts with sex and age characteristics to affect mountain goat survival. Critically, we provide baseline survival rate statistics across various age, sex, and climate scenarios. These data will assist conservation and management of mountain goats by enabling detailed, model-based demographic forecasting of human and/or climate-based population impacts.

We used an individual-based population model to perform a viability analysis to simulate population growth (λ) of 167 elk (Cervus elaphus manitobensis; 71 male and 96 female) released in the Cumberland Mountains, Tennessee, to estimate sustainability (i.e., λ > 1.0) and identify the most appropriate options for managing elk restoration. We transported elk from Elk Island National Park, Alberta, Canada, and from Land Between the Lakes, Kentucky, and reintroduced them beginning in December 2000 and ending in February 2003. We estimated annual survival rates for 156 radio-collared elk from December 2000 until November 2004. We used data from a nearby elk herd in Great Smoky Mountains National Park to simulate pessimistic and optimistic recruitment and performed population viability analyses to evaluate sustainability over a 25-year period. Annual survival averaged 0.799 (Total SE = 0.023). The primary identifiable sources of mortality were poaching, disease from meningeal worm (Parelaphostrongylus tenuis), and accidents (environmental causes and unintentional harvest). Population growth given pessimistic recruitment rates averaged 0.895 over 25 years (0.955 in year 1 to 0.880 in year 25); population growth was not sustainable in 100% of the runs. With the most optimistic estimates of recruitment, mean λ increased to 0.967 (1.038 in year 1 to 0.956 in year 25) with 99.6% of the runs failing to be sustainable. We suggest that further translocation efforts to increase herd size will be ineffective unless survival rates are increased in the Cumberland Mountains.

Sierra Nevada bighorn sheep (Ovis canadensis sierrae) experienced a severe population decline after European settlement from which they have never recovered; this subspecies was listed as endangered under the United States Endangered Species Act (ESA) in 1999. Recovery of a listed species is accomplished via federally mandated recovery plans with specific population goals. Our main objective was to evaluate the potential impact of disease on the probability of meeting specific population size and persistence goals, as outlined in the Sierra Nevada bighorn sheep recovery plan. We also sought to heuristically evaluate the efficacy of management strategies aimed at reducing disease risk to or impact on modeled bighorn populations. To do this, we constructed a stochastic population projection model incorporating disease dynamics for 3 populations (Langley, Mono, Wheeler) based on data collected from 1980 to 2007. We modeled the dynamics of female bighorns in 4 age classes (lamb, yearling, adult, senescent) under 2 disease scenarios: 5% lower survival across the latter 3 age classes and persistent 65% lower lamb survival (i.e., mild) or 65% reduced survival across all age classes followed by persistent 65% lower lamb survival (i.e., severe). We simulated management strategies designed to mitigate disease risk: reducing the probability of a disease outbreak (to represent a strategy like domestic sheep grazing management) and reducing mortality rate (to represent a strategy that improved survival in the face of introduced disease). Results from our projection model indicated that management strategies need to be population specific. The population with the highest growth rate (; Langley; = 1.13) was more robust to the effects of disease. By contrast, the population with the lowest growth rate (Mono; = 1.00) would require management intervention beyond disease management alone, and the population with a moderate growth rate (Wheeler; = 1.07) would require management sufficient to prevent severe disease outbreaks. Because severe outbreaks increased adult mortality, disease can directly reduce the probability of meeting recovery plan goals. Although mild disease outbreaks had minimal direct effects on the populations, they reduced recruitment and the number of individuals available for translocation to other populations, which can indirectly reduce the probability of meeting overall, range-wide minimum population size goals. Based on simulation results, we recommend reducing the probability of outbreak by continuing efforts to manage high-risk (i.e., spatially close) allotments through restricted grazing regimes and stray management to ensure recovery for Wheeler and Mono. Managing bighorn and domestic sheep for geographic separation until Sierra Nevada bighorn sheep achieve recovery objectives would enhance the likelihood of population recovery.

Estimating the dynamics of furbearer populations is challenging because their elusive behavior and low densities make observations difficult. Statistical population reconstruction is a flexible approach to demographic assessment for harvested populations, but the technique has not been applied to furbearers. We extended this approach to furbearers and analyzed 8 yr of age-at-harvest data for American marten (Martes americana) in the Upper Peninsula of Michigan. Marten abundance estimates showed a general downward trend from an estimate of = 1,733.3 ( = 861.3) animals in 2000 to =1,163.9 ( = 520.1) in 2007. The harvest probability of martens increased nearly 5-fold from 0.0542 ( = 0.0250) in 2000 to 0.2637 ( = 0.1154) in 2007, which corresponded to a 5-fold increase in trap-nights. Continued monitoring of martens in the Upper Peninsula, Michigan, and a reassessment of current harvest regulations are necessary given the estimated decreases. Moreover, we do not encourage the use of harvest indices as the sole technique to assess the status and trends of marten and fisher populations. Auxiliary studies in the Upper Peninsula, Michigan, will allow for continued use and improvement in the application of these models.

We compared the distribution and frequency of American marten (Martes americana) detections during historic surveys and a recent survey on the Sagehen Experimental Forest (SEF) in the Sierra Nevada Mountains, California. This area has been the location of 9 previous marten surveys during 1980– 1993, each involving a systematic detection/non-detection survey on the same grid. These data are a time series of information on the occupancy of martens that can be related to habitat change in the study area. Our objectives were to 1) resurvey martens in SEF using methodology similar to previous studies to assess current marten occupancy; 2) evaluate changes in marten occupancy during the period 1980–2008; and 3) examine associations between marten occurence and changes in habitat and landscape metrics. Current marten occupancy was estimated using surveys conducted in summer 2007, winter 2007–2008, and summer 2008. From 1978 to 2007 there was a decrease in predicted habitat patch size, core area, and total amount of marten habitat in the study area, as well as an increase in distance between patches. Marten detections in 2007–2008 were approximately 60% lower than in surveys in the 1980s. We detected no martens in the summers of 2007 and 2008, and 10 detections in winter 2007–2008 were limited to higher elevations in the southwestern portion of SEF. No martens were detected in the lower elevations where most of the recent forest management activity occurred. We suggest that the marten population at SEF has been negatively affected by the loss and fragmentation of habitat. We recommend that future management of forests in the Sagehen basin focus on restoring and connecting residual marten habitat to improve habitat quality for martens.

Models of resource selection are being used increasingly to predict or model the effects of management actions rather than simply quantifying habitat selection. Multilevel, or hierarchical, models are an increasingly popular method to analyze animal resource selection because they impose a relatively weak stochastic constraint to model heterogeneity in habitat use and also account for unequal sample sizes among individuals. However, few studies have used multilevel models to model coefficients as a function of predictors that may influence habitat use at different scales or quantify differences in resource selection among groups. We used an example with white-tailed deer (Odocoileus virginianus) to illustrate how to model resource use as a function of distance to road that varies among deer by road density at the home range scale. We found that deer avoidance of roads decreased as road density increased. Also, we used multilevel models with sika deer (Cervus nippon) and white-tailed deer to examine whether resource selection differed between species. We failed to detect differences in resource use between these two species and showed how information-theoretic and graphical measures can be used to assess how resource use may have differed. Multilevel models can improve our understanding of how resource selection varies among individuals and provides an objective, quantifiable approach to assess differences or changes in resource selection.

Our understanding of factors that limit mule deer (Odocoileus hemionus) populations may be improved by evaluating neonatal survival as a function of dam characteristics under free-ranging conditions, which generally requires that both neonates and dams are radiocollared. The most viable technique facilitating capture of neonates from radiocollared adult females is use of vaginal implant transmitters (VITs). To date, VITs have allowed research opportunities that were not previously possible; however, VITs are often expelled from adult females prepartum, which limits their effectiveness. We redesigned an existing VIT manufactured by Advanced Telemetry Systems (ATS; Isanti, MN) by lengthening and widening wings used to retain the VIT in an adult female. Our objective was to increase VIT retention rates and thereby increase the likelihood of locating birth sites and newborn fawns. We placed the newly designed VITs in 59 adult female mule deer and evaluated the probability of retention to parturition and the probability of detecting newborn fawns. We also developed an equation for determining VIT sample size necessary to achieve a specified sample size of neonates. The probability of a VIT being retained until parturition was 0.766 (SE = 0.0605) and the probability of a VIT being retained to within 3 days of parturition was 0.894 (SE = 0.0441). In a similar study using the original VIT wings (Bishop et al. 2007), the probability of a VIT being retained until parturition was 0.447 (SE = 0.0468) and the probability of retention to within 3 days of parturition was 0.623 (SE = 0.0456). Thus, our design modification increased VIT retention to parturition by 0.319 (SE = 0.0765) and VIT retention to within 3 days of parturition by 0.271 (SE = 0.0634). Considering dams that retained VITs to within 3 days of parturition, the probability of detecting at least 1 neonate was 0.952 (SE = 0.0334) and the probability of detecting both fawns from twin litters was 0.588 (SE = 0.0827). We expended approximately 12 person-hours per detected neonate. As a guide for researchers planning future studies, we found that VIT sample size should approximately equal the targeted neonate sample size. Our study expands opportunities for conducting research that links adult female attributes to productivity and offspring survival in mule deer.

We examined retention of butt-end aluminum leg bands on Rio Grande wild turkeys (Meleagris gallopavo intermedia) captured in Texas and Kansas, USA, 2000–2009. We examined 187 recaptured or harvested radiotagged wild turkeys to determine band retention and modeled band retention with Program MARK. We did not detect differences in band retention among age and gender classes or that band retention probability was time dependent. We estimated monthly probability of band retention was 0.990 (SE = 0.002). Band retention probability was 0.971 (SE = 0.006) at 3 months post-banding and 0.864 (SE = 0.028) at 15 months post-banding. Butt-end aluminum leg band retention was not 100% for wild turkeys marked during our work; however, our band retention rates were 3.7–5.7 times greater than described previously.

We performed genetic analyses of Microtus longicaudus populations within the Crook Point Unit of the Oregon Islands National Wildlife Refuge. A M. longicaudus population at Saddle Rock (located approx. 65 m off-shore from the Crook Point mainland) is suspected to be partially responsible for declines of a Leach's storm-petrel colony at this important nesting site. Using Amplified Fragment Length Polymorphism markers and mitochondrial DNA, we illustrate that Saddle Rock and Crook Point function as separate island and mainland populations despite their close proximity. In addition to genetic structure, we also observed reduced genetic diversity at Saddle Rock, suggesting that little individual movement occurs between populations. If local resource managers decide to perform an eradication at Saddle Rock, we conclude that immediate recolonization of the island by M. longicaudus would be unlikely. Because M. longicaudus is native to Oregon, we also consider the degree with which the differentiation of Saddle Rock signifies the presence of a unique entity that warrants conservation rather than eradication.