Geology Reports⌕ Search

SEARCH · Geology Reports

Results for “The Journal of Wildlife Management”

Search indexed USGS publications on groundwater, aquifers, geologic maps, mineral resources and earthquakes. Explore source records by subject and place.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 667 records · Page 37Linked to original sources
USGS2015

Roosting habitat use and selection by northern spotted owls during natal dispersal

We studied habitat selection by northern spotted owls ( Strix occidentalis caurina ) during natal dispersal in Washington State, USA, at both the roost site and landscape scales. We used logistic regression to obtain parameters for an exponential resource selection function based on vegetation attributes in roost and random plots in 76 forest stands that were used for roosting. We used a similar analysis to evaluate selection of landscape habitat attributes based on 301 radio-telemetry relocations and random points within our study area. We found no evidence of within-stand selection for any of the variables examined, but 78% of roosts were in stands with at least some large (>50 cm dbh) trees. At the landscape scale, owls selected for stands with high canopy cover (>70%). Dispersing owls selected vegetation types that were more similar to habitat selected by adult owls than habitat that would result from following guidelines previously proposed to maintain dispersal habitat. Our analysis indicates that juvenile owls select stands for roosting that have greater canopy cover than is recommended in current agency guidelines.

Washington↗
USGS2011

Monitoring bald eagles using lists of nests: Response to Watts and Duerr

The post-delisting monitoring plan for bald eagles ( Haliaeetus leucocephalus ) roposed use of a dual-frame sample design, in which sampling of known nest sites in combination with additional area-based sampling is used to estimate total number of nesting bald eagle pairs. Watts and Duerr (2010) used data from repeated observations of bald eagle nests in Virginia, USA to estimate a nest turnover rate and used this rate to simulate decline in number of occupied nests in list nests over time. Results of Watts and Duerr suggest that, given the rates of loss of nests from the list of known nest sites in Virginia, the list information will be of little value to sampling unless lists are constantly updated. Those authors criticize the plan for not placing sufficient emphasis on updating and maintaining lists of bald eagle nests. Watts and Duerr's metric of turnover rate does not distinguish detectability or temporary nonuse of nests from permanent loss of nests and likely overestimates turnover rate. We describe a multi-state capture–recapture model that allows appropriate estimation of rates of loss of nests, and we use the model to estimate rates of loss from a sample of nests from Maine, USA. The post-delisting monitoring plan addresses the need to maintain and update the lists of nests, and we show that dual frame sampling is an effective approach for sampling nesting bald eagle populations.

Virginia↗
USGS2015

Insufficient sampling to identify species affected by turbine collisions

We compared the number of avian species detected and the sampling effort during fatality monitoring at 50 North American wind facilities. Facilities with short intervals between sampling events and high effort detected more species, but many facilities appeared undersampled. Species accumulation curves for 2 wind facilities studied for more than 1 year had yet to reach an asymptote. The monitoring effort that is typically invested is likely inadequate to identify all of the species killed by wind turbines. This may understate impacts for rare species of conservation concern that collide infrequently with turbines but suffer disproportionate consequences from those fatalities. Published 2015. This article is a U.S. Government work and is in the public domain in the USA.

Journal of Wildlife Management↗
USGS2015

Pairing call-response surveys and distance sampling for a mammalian carnivore

Density estimates accounting for differential animal detectability are difficult to acquire for wide-ranging and elusive species such as mammalian carnivores. Pairing distance sampling with call-response surveys may provide an efficient means of tracking changes in populations of coyotes ( Canis latrans ), a species of particular interest in the eastern United States. Blind field trials in rural New York State indicated 119-m linear error for triangulated coyote calls, and a 1.8-km distance threshold for call detectability, which was sufficient to estimate a detection function with precision using distance sampling. We conducted statewide road-based surveys with sampling locations spaced ≥6 km apart from June to August 2010. Each detected call (be it a single or group) counted as a single object, representing 1 territorial pair, because of uncertainty in the number of vocalizing animals. From 524 survey points and 75 detections, we estimated the probability of detecting a calling coyote to be 0.17 ± 0.02 SE, yielding a detection-corrected index of 0.75 pairs/10 km 2 (95% CI: 0.52–1.1, 18.5% CV) for a minimum of 8,133 pairs across rural New York State. Importantly, we consider this an index rather than true estimate of abundance given the unknown probability of coyote availability for detection during our surveys. Even so, pairing distance sampling with call-response surveys provided a novel, efficient, and noninvasive means of monitoring populations of wide-ranging and elusive, albeit reliably vocal, mammalian carnivores. Our approach offers an effective new means of tracking species like coyotes, one that is readily extendable to other species and geographic extents, provided key assumptions of distance sampling are met.

New York↗
USGS2015

Estimating the phenology of elk brucellosis transmission with hierarchical models of cause-specific and baseline hazards

Understanding the seasonal timing of disease transmission can lead to more effective control strategies, but the seasonality of transmission is often unknown for pathogens transmitted directly. We inserted vaginal implant transmitters (VITs) in 575 elk ( Cervus elaphus canadensis ) from 2006 to 2014 to assess when reproductive failures (i.e., abortions or still births) occur, which is the primary transmission route of Brucella abortus , the causative agent of brucellosis in the Greater Yellowstone Ecosystem. Using a survival analysis framework, we developed a Bayesian hierarchical model that simultaneously estimated the total baseline hazard of a reproductive event as well as its 2 mutually exclusive parts (abortions or live births). Approximately, 16% (95% CI = 0.10, 0.23) of the pregnant seropositive elk had reproductive failures, whereas 2% (95% CI = 0.01, 0.04) of the seronegative elk had probable abortions. Reproductive failures could have occurred as early as 13 February and as late as 10 July, peaking from March through May. Model results suggest that less than 5% of likely abortions occurred after 6 June each year and abortions were approximately 5 times more likely in March, April, or May compared to February or June. In western Wyoming, supplemental feeding of elk begins in December and ends during the peak of elk abortions and brucellosis transmission (i.e., Mar and Apr). Years with more snow may enhance elk-to-elk transmission on supplemental feeding areas because elk are artificially aggregated for the majority of the transmission season. Elk-to-cattle transmission will depend on the transmission period relative to the end of the supplemental feeding season, elk seroprevalence, population size, and the amount of commingling. Our statistical approach allowed us to estimate the probability density function of different event types over time, which may be applicable to other cause-specific survival analyses. It is often challenging to assess the cause of death, or in this case whether the reproductive event was an abortion or live birth. Accounting for uncertainty in the event type is an important future addition to our methodological approach.

Wyoming↗
USGS2015

Truth, models, model sets, AIC, and multimodel inference: a Bayesian perspective

Statistical inference begins with viewing data as realizations of stochastic processes. Mathematical models provide partial descriptions of these processes; inference is the process of using the data to obtain a more complete description of the stochastic processes. Wildlife and ecological scientists have become increasingly concerned with the conditional nature of model-based inference: what if the model is wrong? Over the last 2 decades, Akaike's Information Criterion (AIC) has been widely and increasingly used in wildlife statistics for 2 related purposes, first for model choice and second to quantify model uncertainty. We argue that for the second of these purposes, the Bayesian paradigm provides the natural framework for describing uncertainty associated with model choice and provides the most easily communicated basis for model weighting. Moreover, Bayesian arguments provide the sole justification for interpreting model weights (including AIC weights) as coherent (mathematically self consistent) model probabilities. This interpretation requires treating the model as an exact description of the data-generating mechanism. We discuss the implications of this assumption, and conclude that more emphasis is needed on model checking to provide confidence in the quality of inference.

Journal of Wildlife Management↗
USGS2015

Feather isotope analysis reveals differential patterns of habitat and resource use in populations of white-winged doves

The white-winged dove ( Zenaida asiatica ) serves an important ecological role as a diurnal pollinator of the saguaro cactus in the Sonoran desert and an economic role as a highly sought after game bird in North America. White-winged doves are intimately linked to anthropogenic changes on the landscape and because of this, have experienced dramatic population fluctuations over the last 75 years in response, both positively and negatively, to anthropogenic changes on the landscape. To understand the factors driving population growth and decline of migratory species like the white-winged dove, it is imperative we study resource use on both their breeding and wintering grounds. To understand how populations are distributed on the wintering grounds, we tested an alternative to band recovery approaches by using stable isotope analysis. Before we could use isotope analysis to link breeding and wintering locations for this species, we first needed to determine if hydrogen (δ 2 H) and carbon (δ 13 C) stable isotopes in feather tissue (δ 2 H f and δ 13 C f , respectively) could differentiate among populations of white-winged doves across their breeding range in Texas, New Mexico, and Arizona. δ 2 H f and δ 13 C f not only differentiated between populations of white-winged doves that breed in the United States, but δ 2 H f also provided further differentiation in white-winged doves that breed in native Sonoran Desert and agricultural habitats in the western portion of their range. Ecological processes associated with desert resources and anthropogenic influences, specifically saguaro cacti and irrigated crops, largely determined δ 2 H f in some white-winged doves in Arizona whereas δ 2 H of precipitation (δ 2 H p ) largely determined δ 2 H f of doves in New Mexico and Texas. This study highlights the usefulness of stable isotope analysis to differentiate populations of animals across the landscape and the insight isotopes can provide into habitat and resource use. Published 2015. This article is a U.S. Government work and is in the public domain in the USA.

Arizona, California, New Mexico, Texas↗
USGS2015

Selection of vegetation types and density of bison in an arid ecosystem

Understanding species habitat selection and factors that drive selection are key components for conservation. We report the first resource selection functions (RSFs) for bison inhabiting an arid ecosystem and use them with density estimates of bison to estimate the number of bison that could be supported if the bison range were expanded to federal lands in the San Luis Valley of Colorado. We derived RSFs for vegetation types using locations of plains bison collected weekly over 3 years from 2005 to 2007. Bison selected for wet or mesic grassland habitats in all seasons. Wetland selection by bison was predicted to be 18 times greater than that of rabbitbrush vegetation, the reference category, and selection of meadows was predicted to be 11 times greater than that of the rabbitbrush type. Willow-dominated plant communities were strongly avoided. Cottonwood communities were also avoided, with the exception of some moderate levels of selection in fall. The willow and cottonwood communities have an understory with low biomass of herbaceous species and low productivity in this arid system. Based on the RSFs we predicted that in the San Luis Valley of Colorado up to 2,379 bison could be supported in similar habitats under Fish and Wildlife Service (FWS) jurisdiction, and up to 759 bison could be supported on adjacent National Park Service (NPS) land. This modeling framework provides a conservation tool for the restoration of bison to their historical habitats, and has utility for application to other terrestrial species where assumptions are met.

Colorado↗
USGS2015

Connectivity among subpopulations of Louisiana black bears as estimated by a step selection function

Habitat fragmentation is a fundamental cause of population decline and increased risk of extinction for many wildlife species; animals with large home ranges and small population sizes are particularly sensitive. The Louisiana black bear ( Ursus americanus luteolus ) exists only in small, isolated subpopulations as a result of land clearing for agriculture, but the relative potential for inter-subpopulation movement by Louisiana black bears has not been quantified, nor have characteristics of effective travel routes between habitat fragments been identified. We placed and monitored global positioning system (GPS) radio collars on 8 female and 23 male bears located in 4 subpopulations in Louisiana, which included a reintroduced subpopulation located between 2 of the remnant subpopulations. We compared characteristics of sequential radiolocations of bears (i.e., steps) with steps that were possible but not chosen by the bears to develop step selection function models based on conditional logistic regression. The probability of a step being selected by a bear increased as the distance to natural land cover and agriculture at the end of the step decreased and as distance from roads at the end of a step increased. To characterize connectivity among subpopulations, we used the step selection models to create 4,000 hypothetical correlated random walks for each subpopulation representing potential dispersal events to estimate the proportion that intersected adjacent subpopulations (hereafter referred to as successful dispersals). Based on the models, movement paths for males intersected all adjacent subpopulations but paths for females intersected only the most proximate subpopulations. Cross-validation and genetic and independent observation data supported our findings. Our models also revealed that successful dispersals were facilitated by a reintroduced population located between 2 distant subpopulations. Successful dispersals for males were dependent on natural land cover in private ownership. The addition of hypothetical 1,000-m- or 3,000-m-wide corridors between the 4 study areas had minimal effects on connectivity among subpopulations. For females, our model suggested that habitat between subpopulations would probably have to be permanently occupied for demographic rescue to occur. Thus, the establishment of stepping-stone populations, such as the reintroduced population that we studied, may be a more effective conservation measure than long corridors without a population presence in between.

Louisiana↗
USGS2022

Choosing an optimal duck season: Integrating hunter values and duck abundance

State wildlife agencies have long struggled to identify optimal hunting season dates for migratory game bird species that meet the diverse and often competing interests of stakeholders. Many approaches have been used to ensure the regulated community participates in the decision-making process, including public hearings, hunter season-date preference surveys, and hunter task forces or committees. Although hearings, surveys, and task force approaches include portions of the regulated community (typically the most avid) they may not necessarily reflect the opinions and values of all stakeholders. Additionally, these approaches rely heavily on limited anecdotal observations that may be unduly influenced by hunter avidity (e.g., days spent afield), hunter density, species preferences, and property access. To address the challenges caused by engaging only portions of the stakeholder community, we used a structured decision-making framework that included a 2017 duck hunter survey to elucidate values of a representative sample of the regulated community in each waterfowl hunting zone in the state of New York, USA. Rather than asking duck hunters about their specific duck hunting season date preferences, we asked them to rank 6 objectives describing what they value in their hunting experience (e.g., maximizing the opportunity to see mallards [ Anas platyrhynchos ] and black ducks [ Anas rubripes ], maximizing the number of weekend days). Four of the 6 objectives described duck species availability (i.e., abundance or immigration) and the remaining 2 described considerations that affect an individual hunter's opportunity or limitation to going duck hunting (i.e., holidays or conflicting hunting seasons). We used spatiotemporal abundance models derived from eBird citizen science data to estimate abundance and immigration rates of ducks in each waterfowl zone. We evaluated up to 9 unique season date alternatives developed by duck hunter task forces to determine which season date alternative best satisfied the competing objectives of duck hunters in each zone. The approach we developed allowed for selection of optimal duck hunting season dates and successfully involved avid duck hunters in the regulation development stages, while ensuring that the values of a representative sample of all stakeholders were directly considered through a clear and transparent decision-making process.

New York↗
USGS2004

Dynamics of intertidal foraging by coastal brown bears in Southwestern Alaska

Shoreline areas provide early season foraging opportunities for coastal bears in Alaska. We investigated use by brown bears (Ursus arctos) of soft-shelled (Mya arenaria) and Pacific razor (Siliqua patula) clams at Katmai National Park, Alaska, USA, to identify the potential importance of these clams to bears. We used direct observations of bear foraging behavior in the summers of 1998, 1999, and 2001 to model the nutritional importance of clamming behavior. We also used previously described models to estimate the relative importance of clamming and vegetative foraging in meeting the maintenance requirements of bears. At the harvest rate that we observed (0.69 ?? 0.46 clams/min), bears achieved higher rates of digestible energy intake than those foraging on vegetation. Although clams are available for only a few hours per day, bears could significantly reduce their total daily foraging time by utilizing clams. Smaller single bears and females with dependent young were the most represented groups of bears using intertidal areas. Large male bears, faced with higher energy requirements, likely are unable to efficiently exploit these intertidal resources. Depending on the relationship between clam size and tissue mass, the relative quality of clams differed by species. Bears foraging on Pacific razor clams required the fewest hours to meet maintenance, followed by bears consuming soft-shelled clams. Our findings highlight the significance of intertidal habitats for coastal bears, especially females.

Journal of Wildlife Management↗
USGS2004

Isolation of Snake River islands and mammalian predation of waterfowl nests

In 1990–1992, we studied predation of waterfowl nests by mammalian predators on 30 islands in a 64-km reach of the Snake River in southwestern Idaho, USA, to identify river flows necessary to protect and enhance migratory bird use of Deer Flat National Wildlife Refuge. We monitored 235–314 Canada goose ( Branta canadensis ) nests each year and 122 duck nests, primarily mallard ( Anas platyrhynchos ), in 1991. We calculated island isolation as the flow energy (width × average water velocity [m 2 /sec]) a predator encountered in crossing to an island. Density of goose nests increased as island isolation increased ( P < 0.001). In contrast, visits of terrestrial predators to islands, mammalian predation of goose nests, and variance in predation rates decreased as island isolation increased ( P < 0.01). Nests were most frequently depredated by raccoons ( Procyon lotor ), coyotes ( Canis latrans ), badgers ( Taxidea taxus ), and mink ( Mustela vison ). Multiple regression models explained 67% of the variance in nest density and 48% of variance in predation rates ( P < 0.01). Variables of secondary importance relative to island isolation were island size and percentage of farmland on the adjacent mainland. Fewer geese nested ( P < 0.01) and predation rates were greater ( P = 0.06) on large islands, and more geese nested on islands adjacent to farmland ( P = 0.04). We also analyzed rates of nest predation relative to river flows for 21 islands from 1953 to 1992. Mammalian predation of goose nests decreased as island isolation increased with greater river flows ( P < 0.01). To identify levels of isolation needed to protect use of islands by nesting waterfowl, we estimated rates of change in nest densities and predation rates with island isolation for a range of regression quantiles. We then calculated levels of island isolation at river flows ranging from 110 to 340 m 3 /sec. Below flows of 250 m 3 /sec, nest predation rates increased rapidly. Flows ≥340 m 3 /sec adequately isolated Snake River islands for nesting waterfowl. Predator visitation and nest predation rates were at a low level, and variability in the rates was at a minimum for these flows.

Journal of Wildlife Management↗
USGS2004

Winter survival of lesser scaup in east-central Florida

The North American continental population of lesser scaup (Aythya affinis) has been declining since the mid-1980s. Seasonal survival estimates may provide insights about the ecological basis for this decline, but such data are not available. We estimated post-harvest winter survival of lesser scaup in east-central Florida, USA, where 62% of the Atlantic Flyway population winters. The Kaplan-Meier survival estimate from 11 January to 14 March 2002 was 0.95 ?? 0.04 (SE) for females and 0.90 ?? 0.09 for males. These estimates were not different (P = 0.64), and pooled survival was 0.93 ?? 0.04. Temporary emigration (movement out of and return to the study area) was exhibited by 25% of the birds during survey periods, but absences were short and were believed to have had little effect on precision of survival estimates. Our findings suggested that natural mortality at Merritt Island National Wildlife Refuge (MINWR) and surrounding estuarine areas was relatively low. Our results also indicate that habitat quality in this portion of east-central Florida was sufficient to meet overwintering requirements and likely contributed to the reported survival rates. Estimating survival during other stages of the annual cycle, as well as an overall winter estimate reflecting harvest mortality, is necessary to determine whether low survival rates are responsible for continental population declines.

Journal of Wildlife Management↗
USGS2005

A complete species census and evidence for regional declines in piping plovers

Complete population estimates for widely distributed species are rarely possible. However, for the third time in 10 years, an International Piping Plover ( Charadrius melodus ) Breeding and Winter Census was conducted throughout the species range in 2001. Nearly 1,400 participants from 32 U.S. states and Puerto Rico; 9 Canadian provinces; St. Pierre and Miquelon, France; Cuba; and the Bahamas visited 2,244 sites covering 11,836 km of shoreline habitat. During the winter census, 2,389 piping plovers were observed at 33.5% of potentially occupied sites ( n = 352). Of these, 56.8% had ≤ 10 birds present. The breeding census recorded 5,945 adults at 777 of 1,892 sites surveyed. More than 80% of sites with piping plovers present had ≤ 10 birds. Results indicated an 8.4% increase from 1991 but only a 0.2% increase since 1996. Regional trends suggest that since 1991, number of breeding birds increased on the Atlantic Coast by 78% (2,920 birds; 12.4% increase since 1996) and by 80% in the Great Lakes (72 birds; 50% increase since 1996). However, plovers declined 15% (2,953 birds; 10% decline since 1996) in Prairie Canada/U.S. northern Great Plains. Subregional trends since 1991 reflect a 32.4% decline in Prairie Canada (972 birds; 42.4% decline since 1996), a 2.5% decline in the U.S. northern Great Plains (1,981 birds; 24% increase since 1996), 5.5% decline in eastern Canada (481 birds; 14% increase since 1996), although a 66.2% increase on the U.S. Atlantic Coast (2,430 birds; 12% since 1996). While numbers were down in much of the U.S. northern Great Plains since 1996, an increase (460%, 1,048 birds; 67.7% increase since 1991) was detected on the Missouri River. Results from 3 complete species census efforts provide essential data for conservation planning and assessment and illustrate the utility of global censuses for species of concern.

Journal of Wildlife Management↗
USGS2005

Problems with sampling desert tortoises: A simulation analysis based on field data

The desert tortoise (Gopherus agassizii) was listed as a U.S. threatened species in 1990 based largely on population declines inferred from mark-recapture surveys of 2.59-km2 (1-mi2) plots. Since then, several census methods have been proposed and tested, but all methods still pose logistical or statistical difficulties. We conducted computer simulations using actual tortoise location data from 2 1-mi2 plot surveys in southern California, USA, to identify strengths and weaknesses of current sampling strategies. We considered tortoise population estimates based on these plots as "truth" and then tested various sampling methods based on sampling smaller plots or transect lines passing through the mile squares. Data were analyzed using Schnabel's mark-recapture estimate and program CAPTURE. Experimental subsampling with replacement of the 1-mi2 data using 1-km2 and 0.25-km2 plot boundaries produced data sets of smaller plot sizes, which we compared to estimates from the 1-mi 2 plots. We also tested distance sampling by saturating a 1-mi 2 site with computer simulated transect lines, once again evaluating bias in density estimates. Subsampling estimates from 1-km2 plots did not differ significantly from the estimates derived at 1-mi2. The 0.25-km2 subsamples significantly overestimated population sizes, chiefly because too few recaptures were made. Distance sampling simulations were biased 80% of the time and had high coefficient of variation to density ratios. Furthermore, a prospective power analysis suggested limited ability to detect population declines as high as 50%. We concluded that poor performance and bias of both sampling procedures was driven by insufficient sample size, suggesting that all efforts must be directed to increasing numbers found in order to produce reliable results. Our results suggest that present methods may not be capable of accurately estimating desert tortoise populations.

California↗
USGS2005

Markov Chain Monte Carlo estimation of species distributions: A case study of the swift fox in western Kansas

Accurate maps of species distributions are essential tools for wildlife research and conservation. Unfortunately, biologists often are forced to rely on maps derived from observed occurrences recorded opportunistically during observation periods of variable length. Spurious inferences are likely to result because such maps are profoundly affected by the duration and intensity of observation and by methods used to delineate distributions, especially when detection is uncertain. We conducted a systematic survey of swift fox ( Vulpes velox ) distribution in western Kansas, USA, and used Markov chain Monte Carlo (MCMC) image restoration to rectify these problems. During 1997–1999, we searched 355 townships (ca. 93 km 2 ) 1–3 times each for an average cost of $7,315 per year and achieved a detection rate (probability of detecting swift foxes, if present, during a single search) of θ̂ = 0.69 (95% Bayesian confidence interval [BCI] = [0.60, 0.77]). Our analysis produced an estimate of the underlying distribution, rather than a map of observed occurrences, that reflected the uncertainty associated with estimates of model parameters. To evaluate our results, we analyzed simulated data with similar properties. Results of our simulations suggest negligible bias and god good precision when probabilities of detection on ≥1 survey occasions (cumulative probabilities of detection) exceed 0.65. Although the use of MCMC image restoration has been limited by theoretical and computational complexities, alternatives do not possess the same advantages. Image models accommodate uncertain detection, do not require spatially independent data or a census of map units, and can be used to estimate species distributions directly from observations without relying on habitat covariates or parameters that must be estimated subjectively. These features facilitate economical surveys of large regions, the detection of temporal trends in distribution, and assessments of landscape-level relations between species and habitats. Requirements for the use of MCMC image restoration include study areas that can be partitioned into regular grids of mapping units, spatially contagious species distributions, reliable methods for identifying target species, and cumulative probabilities of detection ≥0.65.

Kansas↗
USGS2005

Broad-scale predictors of canada lynx occurrence in eastern North America

The Canada lynx (Lynx canadensis) is listed as a threatened species throughout the southern extent of its geographic range in the United States. Most research on lynx has been conducted in the western United States and Canada; little is known about the ecology of lynx in eastern North America. To fill critical knowledge gaps about this species, we modeled and mapped lynx occurrence using habitat and weather data from 7 eastern states and 3 Canadian provinces. Annual snowfall, road density, bobcat (L. rufus) harvest, deciduous forest, and coniferous forest were compared at 1,150 lynx locations and 1,288 random locations. Nineteen a priori models were developed using the information-theoretic approach, and logistic regression models were ranked using Akaike's Information Criterion (AIC) and by our ability to correctly classify reserved data (Kappa). Annual snowfall and deciduous forest predicted lynx presence and absence for a reserved dataset (n = 278) with 94% accuracy. A map of the probability of lynx occurrence throughout the region revealed that 92% of the potential habitat (i.e., >50% probability of occurrence) was concentrated in a relatively contiguous complex encompassing northern Maine, New Brunswick, and the Gaspe?? peninsula of Quebec. Most of the remaining potential habitat (5%) was on northern Cape Breton Island in Nova Scotia. Potential habitat in New Hampshire, Vermont, and New York was small (1,252 km2), fragmented, and isolated (>200 km) from known lynx populations. When federally listed as threatened in the contiguous United States in 2000, inadequate regulations on federal lands were cited as the primary threat to Canada lynx. However, the majority of potential lynx habitat in the eastern United States is on private lands and continuous with potential habitat in Canada. Therefore, lynx conservation in eastern North America will need to develop partnerships across national, state, and provincial boundaries as well as with private landowners.

Journal of Wildlife Management↗