Over a 12-year period (1992–2003), we examined the impact of prescribed burning and hardwood removal on a population of nine-banded armadillos (Dasypus novemcinctus) located at Tall Timbers Research Station just north of Tallahassee, Florida. Although these armadillos are often found in close proximity to humans, there currently are no data on how they are affected by human impacts on the environment. Responses to annual burns between 1992–1997 indicated that in some years armadillos, particularly adults, avoided areas that had been burned, but effects were inconsistent and relatively weak. In contrast, hardwood removal during 1998–2000 coincided with a significant decline in population numbers that continued through 2003. However, interpretation of hardwood removal effects was complicated by the occurrence of a severe drought during the same time period. Comparisons between animals in logged and unlogged parts of the study area during the period of hardwood removal revealed few differences, suggesting drought was an important influence. However, because our population continued to decline after the drought ended, it seems likely that hardwood removal generated more persistent effects that were temporarily masked by the drought. We observed armadillos frequently in logged areas, probably because few other habitat choices were available. Armadillos weighed less during and after hardwood removal than prior to it. Although adult reproductive behavior appeared largely unaffected by logging, numbers of juveniles captured and recruited declined significantly with the onset of hardwood removal. There was no evidence that the disturbance from logging caused increases in distances moved by animals that remained in the study area. Our results may have broader implications for predicting how armadillo populations in Latin America will be affected by similar land management practices.

We investigated the effects of northern pocket gophers (Thomomys talpoides) on plant species diversity and biomass of herbaceous vegetation in aspen (Populus tremuloides) meadows through a series of 4 treatment plots. The treatments were control (no treatment), baited (pocket gopher removal), fenced (ungulate exclusion), and fenced-and-baited combined. When we compared baited plots to control plots or fenced-and-baited plots to fenced plots, we found that baiting reduced pocket gopher densities but had no effect on plant species diversity or herbaceous plant biomass, forb biomass, or grass biomass. However, there was a negative correlation between the number of pocket gopher mounds present in fenced plots and herbaceous plant biomass and grass biomass. At our study site, pocket gopher densities were not high enough to impact herbaceous vegetation.

We tested the efficacy of a snow-tracking-based model for predicting wolf (Canis lupus) distribution and environmental relationships, using n independent radiotelemetry data dataset. We documented tracks in snow on highway rights-of-way and adjacent transects in the central Rocky Mountains of Alberta, Canada between November and March, 1997–2000. Radiotelemetry data (ground and aerial) were collected in the same region for 2 wolf packs between 1991–1993. We assessed the relationship between wolf track data and topographic, vegetative, and prey metrics, using a Geographic Information System (GIS), logistic regression, and Akaike's Information Criterion (AIC). We transformed our optimal regression model into a probability surface in GIS and verified that surface using radiotelemetry data and a Receiver Operating Characteristic (ROC) curve. The optimal model showed that wolf presences were positively related to wetness (mature, possibly more complex forest), and elk (Cervus elaphus), and deer (Odocoileus sp.) track density and negatively associated with terrain ruggedness and open canopy. The ROC curve indicated that the track-based model was robust (AUC=0.78). We concluded that track data provide a reliable, cost-effective approach for determining distribution and predicting wolf–environmental relationships in mountainous regions.

Density estimation of wolves (Canis lupus) requires a count of individuals and an estimate of the area those individuals inhabit. With radiomarked wolves, the count is straightforward but estimation of the area is more difficult and often given inadequate attention. The population area, based on the mosaic of pack territories, is influenced by sampling intensity similar to the estimation of individual home ranges. If sampling intensity is low, population area will be underestimated and wolf density will be inflated. Using data from studies in Denali National Park and Preserve, Alaska, we investigated these relationships using Monte Carlo simulation to evaluate effects of radiolocation effort and number of marked packs on density estimation. As the number of adjoining pack home ranges increased, fewer relocations were necessary to define a given percentage of population area. We present recommendations for monitoring wolves via radiotelemetry.

Gray wolves (Canis lupus) likely will recolonize the northern Lower Peninsula of Michigan (NLP). As such, land managers would benefit from information on the amount, distribution, and quality of potential wolf habitat in this region. We estimated that 2,198–4,231 km² of favorable wolf habitat exist in the NLP, supporting an estimated population of 40–105 wolves. Favorable habitat was fragmented by road networks and was predominantly located in the northeastern part of the state on private land. We discuss the management of wolves in the NLP as a case study of wolf recolonization in a landscape that has a relatively high road density and agricultural lands that likely will be sources of conflict with wolves. We provide a hierarchical model for consideration in proactively managing landscapes that already or likely will contain several carnivore species concomitant with human land use. We suggest that this case study and our hierarchical model offer insight into how proactive land management should occur for wolves and other carnivores in the northern Great Lakes Region and other human-altered landscapes.

Although numerous authors are investigating indirect effects of wolf recovery, the most fundamental ecological impact of the Greater Yellowstone Area wolf reintroduction, the effects of wolf predation on ungulate populations, remains unclear. We report on a 5-year comparative study of wolf (Canis lupus)–elk (Cervus elaphus) dynamics on an elk herd in the headwaters of the Madison River within Yellowstone National Park and the lower Madison elk herd that winters 40 km downriver outside the Park. A resident pack became established on the Madison headwaters area in 1997 and grew to multiple packs totaling 30–40 animals by 2002. During winter 1999 emigrates from Yellowstone established a pack on the lower Madison area. However, poor recruitment and low adult survival limited wolf population growth, with the area supporting a single pack, never exceeding 5 animals. Wolf kill rates on the lower Madison area were approximately double that documented for the Madison headwaters area. Moderate kill rates in the Madison headwaters, combined with high wolf densities and modest elk densities, resulted in an estimated 20% of the elk population being killed during winter and projections for a declining elk population. In contrast, high kill rates on the lower Madison area, combined with low wolf densities and high elk densities, resulted in winter predation estimates not exceeding 4% of the elk population. We suspect this level of mortality will be of little biological significance with respect to elk population trajectory. These results suggest that the effects of wolf predation on elk populations differ substantially over relatively small spatial scales, depending on a complex suite of interacting factors. Thus, we caution against generalizing the effects of wolf restoration on elk dynamics from any single study and encourage collaborations to develop comparative predator–prey studies that improve our understanding of wolf–ungulate interactions and enhance conservation.

Managing wolf (Canis lupus) depredation on livestock is expensive and controversial; therefore, managers seek to improve and develop new methods to mitigate conflicts. Determining which factors put ranches at higher risk to wolf depredation may provide ideas for ways to reduce livestock and wolf losses. We sampled cattle pastures in Montana and Idaho that experienced confirmed wolf depredations (n = 34) from 1994–2002 and compared landscape and selected animal husbandry factors with cattle pastures on nearby ranches where depredations did not occur (n=62). Pastures where depredations occurred were more likely to have elk (Cervus elaphus) present, were larger in size, had more cattle, and grazed cattle farther from residences than pastures without depredations. Using classification tree analysis, we found that a higher percentage of vegetation cover also was associated with depredated pastures in combination with the variables above. We found no relationship between depredations and carcass disposal methods, calving locations, calving times, breed of cattle, or the distance cattle were grazed from the forest edge. Most pastures where depredations occurred during the wolf denning season (April 15–June 15) were located closer to wolf dens than nearby cattle pastures without depredations. Physical vulnerability, especially of calves, also may increase risk of depredation.

Migratory mule deer (Odocoileus hemionus) and pronghorn (Antilocapra americana) populations rely on seasonal ranges to meet their annual nutritional and energetic requirements. Because seasonal ranges often occur great distances apart and across a mix of vegetation types and land ownership, maintaining migration corridors to and from these ranges can be difficult, especially if managers do not have detailed information on mule deer and pronghorn seasonal movements. We captured, radiomarked, and monitored mule deer (n=171) and pronghorn (n=34) in western Wyoming to document seasonal distribution patterns and migration routes. Mule deer and pronghorn migrated 20–158 km and 116–258 km, respectively, between seasonal ranges. These distances represented the longest recorded migrations for either species. We identified a number of bottlenecks along the migration routes of mule deer and pronghorn, but the most critical appeared to be the 1.6-km-wide Trapper's Point bottleneck, which was used by both mule deer and pronghorn during their spring and autumn migrations. Housing developments and roadways apparently have reduced the effective width of this bottleneck to <0.8 km. We estimate 2,500–3,500 mule deer and 1,500–2,000 pronghorn move through the bottleneck twice a year during spring and autumn migrations. Identification and protection of migration corridors and bottlenecks will be necessary to maintain mule deer and pronghorn populations throughout their range.

White-tailed deer (Odocoileus virginianus) populations in southern Québec have recently reached record densities after almost collapsing at the turn of the twentieth century. High densities of deer can dramatically alter forest vegetation and tend to increase use of tree plantations, orchards, and agricultural fields, which can induce severe financial loss to producers. We quantified grazing damage in young legume (i.e., clover [Trifolium sp.] and alfalfa (Medicago sativum]) fields that were bordered by woodlots near white-tailed deer wintering areas. We placed exclusion cages in 8 fields during two consecutive dormant seasons (October–May). We collected vegetation just prior to harvesting in exclusion and control plots to estimate dry biomass produced and legume, nitrogen (N), and fiber content of the hay. White-tailed deer grazing in autumn and spring caused an estimated loss ranging from 12–14% in subsequent annual yields in legume fields situated near woodlots. However, losses varied considerably among farms. Legume and chemical content (N and fibers) of the hay were not affected by grazing. Increasing deer harvest likely represents the best method to reduce damage to hay fields in southern Québec.

Cervids readily adapt to suitable human-altered landscapes and can cause several types of damage, including economic loss associated with landscape and agricultural plantings, human health and safety concerns, and adverse impacts on natural habitats. The need for effective, practical, and nonlethal tools to manage damage caused by elk (Cervus elaphus), mule deer (Odocoileus hemionus), and white-tailed deer (Odocoileus virginianus) has been heightened by the growing prevalence of locally overabundant populations and public demand for nonlethal wildlife management methods. Various frightening devices are available commercially, but most have not been subjectively evaluated. We used consumption measurements to evaluate the efficacy of a specific motion-activated light- and sound-emitting frightening device for urban mule deer and elk. The devices proved ineffective; deer and elk ignored them. As the demand for frightening devices to reduce deer and elk damage increases, it is important that research be conducted to evaluate the efficacy of new devices so that users know what level of efficacy to expect.

Recent work has examined ultimate factors limiting deer (Odocoileus spp.) and elk (Cervus elaphus) populations during winter, such as temperature and forage, but there has been inadequate examination of the influence of snow, especially in concert with foraging decisions. We examined deer and elk habitat selection on winter range in the temperate mountains of southeastern British Columbia. The life histories of radiocollared mule deer (O. hemionus) and elk included seasonal elevational migrations (1,000–1,400 m) and long-distance movements (up to 50–63 km, respectively). Late-winter, deep-snow habitat is limited in wet, mountainous environs and may explain the relatively low densities of ungulates present. Snow-track transects conducted during late winter suggested that deer (mule deer and white-tailed deer [O. virginianus] combined) avoided areas with >40 cm of snow and elk areas with >50 cm. Late-winter snow depth was positively related to elevation and negatively related to slope and solar radiation (hours/day), all of which can be obtained from existing databases and used to map relative late-winter snow depth. The snow-depth model can be used to map potential winter range regardless of current vegetation cover. During late winter deer selected older forests and stands with greater amounts of Douglas-fir (Pseudotsuga menziesii) in the overstory in forested sites. Older stands probably were selected because they had lower snow depths, while mature Douglas-fir trees offered more litterfall forage than other tree species. In contrast, elk selection among forested stands was weak, which suggested that forested stands were not preferred portions of their late-winter habitat. Track data during late winter confirmed that both deer and elk also used areas with low canopy cover, likely to acquire browse. We demonstrated that it is possible to map potential winter range using topographic variables as surrogates for relative snow depth, and we present a model applicable to portions of the temperate interior mountains. We suggest it is important to consider both forage production and snow interception in habitat management because winter energy budgets are a balance between nutrient intake and cost of locomotion.

Information on hunter-effort–harvest-size relationships relevant to managing and monitoring hunted populations with multiple hunt types has not been examined in detail. We estimated harvest size from hunter effort in 6 types of hunts for white-tailed deer (Odocoileus virginianus) and compared intercepts and slopes to assess whether harvest size differed with hunting effort. The 6 hunt types were (age–sex class of population hunted and weapons allowed) as follows: antlerless gun (shotgun, centerfire rifle), either-sex archery, either-sex gun, either-sex muzzleloader, buck-only gun, and buck-only muzzle-loader. The data set was a 12-year harvest record from a hunt program in western Tennessee. We measured hunter effort by hunter days (number of hunters and days they hunted). We detected differences in intercepts but not slopes among hunt types. Hunt types with largest harvest sizes were >4 times larger than hunt types with smallest harvest sizes, controlling for hunter effort. Because slopes of regressions were similar, hunter effort among hunt types can be adjusted so that a harvest size:hunter effort ratio can be used to track relative deer population change.

In 2003 we conducted a study to determine the consequences of feral hog (Sus scrofa) invasions in several ecoregions of Texas. We examined the observations, experiences, and actions of landowners and managers concerning feral hogs on their property. We used purposive sampling of landowners and managers who fit 1 or more of 3 selection criteria. Landowners and managers were either sent a self-administered, mail-out questionnaire or given a copy of the questionnaire during pesticide applicator workshops. There were 775 survey participants. The effective response rate from those landowners and managers who received a mailed questionnaire was 62% (n=284). Nearly all (95%, n=491) of the pesticide applicator workshop participants turned in a completed questionnaire. Sampling error based on the farms (includes ranches) in Texas and in each region was ±3%, α =0.05. The majority (74%) of respondents were ranchers, and 18% were farmers. Most respondents felt that feral hogs came from the neighbor's property and were an agricultural pest. Rooting, wallowing, and crop damage were the major forms of damage caused by feral hogs. The average economic loss due to hog damage, over the lifetime ownership of the land by the respondent, was $7,515 (U.S). Hog control was an incidental process. The average cost for hog control over the lifetime ownership of the land by the respondent was $2,631 (U.S.). There was strong support for programs related to feral hog management and control, but only half of the survey participants responded to the question. The average quiz score of 11.5 indicated that respondents could correctly respond to <50% of the 26 questions. Region was found to have an effect (P≤0.05) on all questions tested except one. Management implications included the need for educational programs about feral hogs, how landowners can make better use of feral hogs on their property, ongoing education efforts about feral hogs, and the impact of this study on the public policy and decision-making process.

Overwater nesting structures for ducks have been used primarily to increase mallard (Anas platyrhynchos) production where success of ground-nesting hens was low (<15%), such as in the Prairie Pothole Region (PPR) in North America. However, managers have inquired about recommended numbers of nest structures for small wetlands (i.e., <2 ha) based on duck-production and benefit–cost evaluations, but such data were not available. Therefore, we conducted an experiment in the PPR in Manitoba, Canada, in 2001 and 2002 to test the effects of different treatment numbers of nest structures/wetland (1, 2, or 4) and wetland area (≤0.4 or 0.45–1.5 ha) on use of structures by nesting ducks, nest success, and number of ducklings produced, as well as to evaluate cost–benefits of the structures. Across duck species, mean use of structures by nesting hens increased 56% between years (F_1,107 = 14.29, P <0.001); however, by year 2 use did not differ among treatment numbers of structures and averaged 78% (t₁₅₇ =−0.15–1.08, P≥0.28). Apparent nest success averaged 99% across treatment numbers of structures in 2001, but it did not differ among treatments (F_2,72 = 0.93, P = 0.40) and decreased to an overall 48% in 2002, largely due to egg predation by corvids. Mean number of ducklings departing structures did not vary between years (F_1,54 = 2.38, P = 0.13), but treatment number of structures influenced mean number of ducklings (pooled over 2001 and 2002) exiting structures (F_2,54 = 4.54, P = 0.02). Mean numbers of ducklings departing 1 and 2 structures/wetland did not differ (t₅₄ = 0.29, P = 0.77; pooled x̄ = 4.61 ducklings), but each respective mean was nearly twice that from 4 structures/wetland (t₅₄ = 2.26–2.51, P≤<0.03). Neither wetland area nor any interactions influenced any analysis (F_{1–2,54–157} = 0–1.11, P≥0.33). We concluded that 2 structures/wetland was cost-effective ($2.16/fledged duck [U.S.]) for consideration among other management strategies to increase nest success of mallards in the PPR; however, managers should monitor duckling production from structures to determine whether our initial recommendation from a single study area warrants change.

During 1996 we examined whether providing predators with supplemental food would reduce predation on duck nests at 10 North Dakota sites. We randomly selected 5 sites to serve as treatment sites where supplemental food (chicken eggs) was provided at feeding stations during the duck-nesting season. The remaining 5 sites served as controls with no supplemental food. During Experiment 1 we distributed approximately 320 chicken eggs/site/week during May and early June, but daily survival rates (DSR) of nests at treatment sites (x̄ = 0.891, SE = 0.018) and control sites (x̄ = 0.939, SE = 0.001) did not differ (F_1,4 = 6.54, P = 0.06). During Experiment 2, we distributed 1,600 chicken eggs/site/week in late June and July, but DSR at treatment sites (x̄ = 0.941, SE = 0.016) and control sites (x̄ = 0.954, SE = 0.008) again were similar (F_1,4 = 0.72, P = 0.44). Our results indicate that supplemental feeding of predators is ineffective at reducing predation rates on upland duck nests.

Information on production of moist-soil seeds is necessary to determine resource availability in wetland habitats and evaluate management efforts. Traditional methods (e.g., core sampling and seed-head clipping) are time-consuming and labor-intensive. Methods to estimate seed production using seed-head characteristics tend to be complex and may have limited utility for some moist-soil plants and in some regions. We developed a simple method to evaluate percent cover and seed-head characteristics of 6 common moist-soil plant types in the Central Valley of California. We estimated percent cover (AREA) and seed-producing potential of each plant type (QUALITY) using an ordinal scale for 13 wetland units on private duck clubs. The product of the AREA and QUALITY scores was calculated for each plant type and then summed over all plant types to provide a single index of seed production (Seed Production Index, SPI) for each unit. To evaluate the reliability of this index, we regressed the value of SPI for each unit against estimates of seed production derived by core sampling. The SPI index was correlated with estimates of moist-soil seed biomass (kg/ha) obtained by core sampling (R²_adj=0.88, P< 0.0001). To further assess the utility of this method in a field situation, 2 observers estimated SPI independently for 183 wetland units during annual site visits for the California Comprehensive Wetland Habitat Program. Estimates of SPI required <15 minutes for most wetlands and were repeatable for the 2 observers (intraclass correlation coefficient =0.79, P<0.0001). We suggest that this technique will provide managers with a simple method to estimate seed production in moist-soil wetlands, track temporal changes in food abundance within wetlands and across landscapes, estimate wetland carrying capacity, and evaluate management actions with minimal resource investment.

Understanding how populations expand to recolonize former habitats is important to restoration efforts in wildlife management and conservation. Translocation of black bears (Ursus americanus) to Arkansas in the 1950s and 1960s has led to recolonization of former bear range in Oklahoma, with substantial increases in distribution and abundance of the species in Oklahoma over the last 15 years. We studied demographics of black bears in southeastern Oklahoma from May 2001 to November 2002 to provide insight into characteristics of recolonizing populations of large carnivores. We trapped 51 black bears (22 M, 29 F) 77 times and radiocollared 25 female bears. Sex ratios of adults and cubs were skewed toward females, and the age structure was younger than observed in other unharvested populations. Survival of adult females was estimated at 0.9 ±0.1, and fertility was estimated at 0.77 female young/female/year. Density on the study area was estimated at 0.21 bears/km² and the current finite growth rate (λ) of the study population was estimated to be 1.11/year. Demographic characteristics of the Oklahoma population of black bears were similar to those of other recolonizing populations of large carnivores.

Litter size, an important reproductive parameter used in the management and conservation of brown bears (Ursus arctos), is determined from reported observations by the public in some areas. We compared brown bear litter sizes based on reported public observations with those obtained by counting young from a helicopter or the ground by researchers. Mean litter sizes based on public observations were lower and showed more variance between seasons (spring and autumn) than mean litter sizes based on research methods. Public mean litter sizes showed significant variation among years, unless data from at least 6 years were analyzed. In south-central Sweden annual correction factors ranging from 1.120–1.260 must be used to correct the mean litter size based on public observations to agree with the mean litter size obtained by research, depending on how evenly public observations are spread throughout the year.

Indices of relative abundance allow managers and researchers to examine changes in population size over time or compare relative population sizes in different areas. In the Pisgah Bear Sanctuary, bait-station surveys were conducted in most years from 1983 to 2000 to follow trends over time in the black bear (Ursus americanus) population. Baited bear trapping also took place in the sanctuary during those years, and some trap lines coincided with bait-station lines. Because the same baits were used for both trapping and bait station lines, we hypothesized that visitation rates of bears to bait stations established in proximity to baited trap lines would differ from rates at bait stations that were not associated with baited trap lines. We modeled probability of bait stations being visited by bears on trapped and untrapped lines to estimate the effect baited trapping had on visitation rates. We found that population trends inferred from bait-station visits in areas that also were trapped with bait were biased high and that bias increased over time. Bears may have become habituated to the bait on trap lines and incorporated it as a regular food source. Bait-station indices should not be conducted near research sites that employ similar bait when both produce a tangible reward for the animals.

We used radiotelemetry and population modeling techniques to examine factors related to population establishment of black bears (Ursus americanus) reintroduced to Felsenthal National Wildlife Refuge (NWR), Arkansas. Our objectives were to determine whether settling (i.e., establishment of a home range at or near the release site), survival, recruitment, and population viability were related to age class of reintroduced bears, presence of cubs, time since release, or number of translocated animals. We removed 23 adult female black bears with 56 cubs from their winter dens at White River NWR and transported them 160 km to man-made den structures at Felsenthal NWR during spring 2000–2002. Total movement and average circuity of adult females decreased from 1 month, 6 months, and 1 year post-emergence (F_{2, 14} =19.7, P <0.001 and F_{2, 14} =5.76, P =0.015, respectively). Mean first-year post-release survival of adult female bears was 0.624 (SE=0.110, SE_interannual= 0.144), and the survival rate of their cubs was 0.750 (SE=0.088, SE_interannual=0.109). The homing rate (i.e., the proportion of bears that returned to White River NWR) was 13%. Annual survival for female bears that remained at the release site and survived >1-year post-release increased to 0.909 (SE=0.097, SE_interannual=0.067; Z=3.5, P<0.001). Based on stochastic population growth simulations, the average annual growth rate (λ) was 1.093 (SD=0.053) and the probability of extinction with no additional stockings ranged from 0.56–1.30%. The bear population at Felsenthal NWR is at or above the number after which extinction risk declines dramatically, although additional releases of bears could significantly decrease time to population reestablishment. Poaching accounted for at least 3 of the 8 adult mortalities that we documented; illegal kills could be a significant impediment to population re-establishment at Felsenthal NWR should poaching rates escalate.

The need for alternative predator capture techniques is increasing because of concerns about the efficiency, selectivity, and injury of currently available capture methods. There also is a need for comparative data evaluating new or seldom used methods. In an initial evaluation, we first surveyed wildlife managers for information on cage-trapping; using these data, we conducted a field study of 4 coyote (Canis latrans) capture systems for animal damage management. We tested the SoftCatch^®, Collarum^®, Wildlife Services–Turman, and Tomahawk^®, systems for capturing coyotes in Arizona and south Texas during 2001 and 2002. We determined capture efficiency and selectivity and performed whole-body necropsies to identify trap-related injuries. Surveys indicated that coyotes usually were captured in large (>1.6-m-length) cage-traps baited with meat or carcasses. In our field evaluation, we estimated a capture efficiency (percentage of coyote captures per capture opportunity) of 0% for the Tomahawk cage-trap, 87% for the Collarum, 88% for the WS–T throw arm, and 100% for the SoftCatch. Cage-traps were the least selective, capturing 34 noncoyote animals, and Collarums were the most selective, capturing no noncoyote animals. The WS–T and SoftCatch devices showed intermediate selectivity of 50% and 69%, respectively. All devices showed low injury scores relative to jawed devices in previous studies; 92%, 57%, and 92% of coyotes captured in the Collarum, WS–T, and SoftCatch showed no indicators of poor welfare, respectively.

Biologists need a variety of tools to determine the population and genetic status of the ocelot (Leopardus pardalis), an elusive Neotropical cat that favors dense habitats. We developed and tested a technique that entices ocelots to rub on scented hair snares and uses DNA analysis of the hair to determine species, gender, and individual identity. Twenty-seven (84%) of 32 captive ocelots rubbed against the scented pads. In field tests at Laguna Atascosa National Wildlife Refuge in south Texas, we detected a minimum of 6 ocelots, including at least 3 of 4 radiocollared animals. Using a 6-locus microsatellite analysis, we made individual identification for 10 of 20 samples. Scented hair snares can provide useful information on the population and genetic status of ocelots and identification of key areas and connecting linkages. We suggest that surveys for ocelots deploy 1 station per 25–50 ha and check them every 1–2 weeks.

Urine urea–creatinine ratio (U:C) often is used as an index of nutritional status of wild mammals, especially because urine can be collected from snow, providing a non-invasive index of nutrition. The rate of creatinine excretion has been assumed to be relatively stable; thus, creatinine concentration is used to normalize urea concentrations and control for dilution in snow. We present data from captive wolves (Canis lupus) to test how much variation in U:C is due to variation in urea vs. creatinine in relation to time since feeding. We found that U:C increases for ∼18 hours post-feeding but that this variation is primarily due to the 5-fold decrease in creatinine rather than the 2-fold increase in urea concentration. It is not clear what governs the variable creatinine concentrations in urine. Until a fuller understanding of this metabolite is achieved, ecological studies employing this nutritional index should report urea and creatinine values as well as the ratio. Physiological studies should explore the causes of variation in creatinine concentrations in urine.

Fallow-field borders along edges of crop fields have been promoted for increasing northern bobwhites (Colinus virginianus) on farms and are a component of recovery plans for this species. However, research on bobwhite population response to field-border practices is sparse. Previous research on 2 farms documented increased use of farm fields and greater reproduction by bobwhites on farms with field borders, but nesting success was low during May and June. Bobwhite population response to field-border practices may increase when they are combined with nest-predator reduction on farms. Effect of nest-predator reduction on bobwhite populations on farmed landscapes has not been investigated in the Southeast. Therefore, we tested the effects of field borders and mesomammal nest-predator reduction on bobwhite abundance on 12 farms in eastern North Carolina, 1997–1999. We applied treatments to farms as factorial combinations. Reduction of mesomammal nest predators, including raccoons (Procyon lotor), Virginia opossums (Didelphis virginiana), and foxes (Urocyon cinereoargenteus and Vulpes vulpes), occurred from February–May of each year. To assess bobwhite response to treatments, we measured summer abundance of males using variable-radius point counts and covey abundance on farms in September and October using morning covey-call surveys. Bobwhites were more abundant on farms with field borders during summer (P=0.08). On field-border farms we heard 1.8× the number of coveys heard on farms without field borders (P=0.004). Summer abundance of bobwhites did not differ as a result of predator reductions (P=0.37), and we heard slightly fewer coveys on predator-reduction farms (P =0.084) during autumn. However, we heard more coveys on farms with both field borders and predator reduction compared to all other farms (P=0.022). Field-border systems were a practical management technique to increase autumn abundance of bobwhites on individual farms in eastern North Carolina.

The raccoon (Procyon lotor) is the definitive host of Baylisascaris procyonis, a large intestinal roundworm. The prevalence of infection among raccoon populations often is high, and in the midwestern United States B. procyonis is documented in 68–82% of raccoons. Because raccoon populations appear to be increasing in response to changes in human land use and because B. procyonis is considered an emerging zoonotic disease, it is important to determine reliable methods to monitor prevalence of infection among raccoons. We compared the relative sensitivity of 3 common methods used by wildlife biologists to determine prevalence in free-ranging raccoon populations. We determined prevalence of infection among midwestern raccoon populations from 456 raccoon fecal samples, 742 raccoon latrine samples, and 212 necropsies (gut analysis). We developed logistic regression models in order to predict the log likelihood of presence of B. procyonis in a given sample as a function of season, land use, and technique. Finally, we measured the sensitivity of fecal sampling by evaluating fecal samples taken from 72 necropsies for the presence of B. procyonis eggs. Necropsy analysis yielded the highest measure of prevalence (44%), followed by latrine samples (22.5%), and fecal samples (17.5%). Necropsy analysis explained the most variance in logistic models, suggesting that this is the most reliable method. However, this technique is labor-intensive and may not be the most efficient method for large-scale investigations of B. procyonis prevalence. Fecal analysis is a reliable method of determining prevalence among raccoons as we observed B. procyonis eggs in 66% of fecal samples taken from positive necropsies. Latrine sampling may be the most efficient method and perhaps the best estimator of zoonotic potential; however, wildlife managers should realize that this measure often is an underestimate of prevalence among raccoons and develop management plans accordingly.

To better understand the distribution and abundance of headwater-stream salamanders in managed conifer forests, we examined relationships between Cascade torrent salamanders (Rhyacotriton cascadae) and biotic and abiotic habitat attributes at landscape and within-stream levels in western Oregon, USA. In 2001 we found 145 torrent salamanders in 25% of 59 headwater (first-order) streams from randomly selected 2.58-km² sections of the study area. Landscape-level variation in torrent salamander distribution and relative abundance were related to the age of adjacent riparian forests and to the landform features parent geology and stream aspect. In 2002 we conducted a more detailed study of salamander occurrence and abundance within 49 10-m stream reaches, stratified by gradient, that were randomly selected from 15 streams known to contain salamanders. We recorded 475 salamanders from 33 (67%) of the stream reaches. Akaike's Information Criterion (AIC) model selection indicated that a streambed substrate model best explained salamander occupancy in stream reaches, but a model containing only the parameter distance to stream origin and another model containing abiotic landform features also received strong empirical support. In contrast, the distance to stream origin model was the best candidate model explaining reach-level salamander abundance. However, 2 additional models explaining abundance, including one that discriminated between the northern and southern portions of our study area and another that reflected stream reach habitat parameters, also received strong empirical support. Physical features of stream habitats appear to have an important influence on the distribution and abundance of torrent salamanders at multiple spatial scales, and these parameters should be addressed when designing management strategies (e.g., riparian buffers) to conserve these species.

To effectively monitor winter foraging activity by the cryptic, non-site-faithful Virginia opossum (Didelphis virginiana), we tested the use of small data-logging temperature sensors (iButton Thermachrons^®, Maxim/Dallas SemiConductors, Dallas, Tex.) attached to a standard radiocollar on 3 opossums over the winter of 2000–2001. Two replicate sensors were required to clearly show time periods with cooler temperatures, an indication that the animal was outside the den. These foraging bouts were consistent with the available radiotelemetry data. Daily duration of foraging showed a strong negative relationship with ambient temperature, quantitatively documenting for the first time a phenomenon previously known only anecdotally. The iButton Thermachron seems to be an effective, low-cost, and low-effort technology for monitoring foraging activities of any animal that rests and forages in different temperature environments.

The southeastern United States population of the painted bunting (Passerina ciris) has decreased approximately 75% from 1966–1996 based on Breeding Bird Survey trends. Partners in Flight guidelines recommend painted bunting conservation as a high priority with a need for management by state and federal agencies. Basic information on home range and survival of breeding painted buntings will provide managers with required habitat types and estimates of land areas necessary to maintain minimum population sizes for this species. We radiotracked after-second-year male and after-hatching-year female buntings on Sapelo Island, Georgia, during the breeding seasons (late April–early August) of 1997 and 1998. We used the animal movement extension in ArcView to determine fixed-kernel home range in an unmanaged maritime shrub and managed 60–80-year-old pine (Pinus spp.)–oak (Quercus spp.) forest. Using the Kaplan-Meier method, we estimated an adult breeding season survival of 1.00 for males (n = 36) and 0.94 (SE = 0.18) for females (n = 27). Painted bunting home ranges were smaller in unmanaged maritime shrub (female: kernel x̄ = 3.5 ha [95% CI: 2.5–4.5]; male: kernel x̄ = 3.1 ha [95% CI: 2.3–3.9]) compared to those in managed pine–oak forests (female: kernel x̄ = 4.7 ha [95% CI: 2.8–6.6]; male: kernel x̄ = 7.0 ha [95% CI: 4.9–9.1]). Buntings nesting in the managed pine–oak forest flew long distances (≥300 m) to forage in salt marshes, freshwater wetlands, and moist forest clearings. In maritime shrub buntings occupied a compact area and rarely moved long distances. The painted bunting population of Sapelo Island requires conservation of maritime shrub as potential optimum nesting habitat and management of nesting habitat in open-canopy pine–oak sawtimber forests by periodic prescribed fire (every 4–6 years) and timber thinning within a landscape that contains salt marsh or freshwater wetland openings within 700 m of those forests.

White et al. (2005) estimated the area occupied by “active” colonies of black-tailed prairie dog (Cynomys ludovicianus) in Colorado as 255,398 ha (95% CI of±9.5%) based on data collected with aerial transect surveys during 2001–2002 by staff of the Colorado Division of Wildlife (CDOW). During 2004 we conducted on-the-ground examinations of a sample of the colony intercept data used by White et al. (2005) and found evidence of misclassifications that would yield significant overestimation bias. We found that 25.4% of the total length of the colony intercepts we examined was incorrectly classified as being a prairie dog colony (these segments had no prairie dog burrows of any age). We also found that 50.3% of the length of examined intercepts fell on currently inactive colonies or portions of colonies (vacant burrows but no living prairie dogs) and only 24.3% fell on active prairie dog colonies with signs of living prairie dogs at our examinations 2 years after the survey reported by White et al. (2005). Further, in Bent and Kiowa counties, where plague (Yertsina pestus) and poisoning were active, we examined 36.9 km of reported active prairie dog intercepts and subjectively classified only 1.6% as active at normal-appearing prairie dog densities. Our fieldwork demonstrated that the estimate by White et al. (2005) was based on data with substantial errors as well as overestimation biases that will be repeated if protocols are not modified for future surveys.

Miller et al. (2005) assert that the estimate of White et al. (2005) of 1.94% of eastern Colorado being occupied by black-tailed prairie dogs (Cynomys ludovicianus) was biased high. The subsequent surveys conducted by Miller et al. (2005) do not provide a valid bias correction for the estimate of White et al. (2005) because 1) their survey protocol could only result in estimating a negative bias, with no possibility of estimating a positive bias; 2) their nonrandom selection of intercepts did not provide valid inferences to the sampling frame used in White et al. (2005); 3) evidence suggests that they did not survey the same aerial tracks as surveyed by White et al. (2005); and 4) their surveys were conducted 2 years after the original surveys, thus not comparable given the temporal dynamics in prairie dog colonies.

Turner et al. (2004) developed a habitat selection model for a population of desert bighorn sheep (Ovis canadensis) in the Peninsular Ranges of southern California that is listed as a threatened and endangered population by the state of California and the federal government, respectively. We are concerned that the recent publication of an article by Turner et al. (2004) could be detrimental to the management and recovery of bighorn sheep in the Peninsular Ranges because it lends credibility to a flawed analysis of bighorn sheep habitat-use patterns. The model attempts to extrapolate conclusions from a limited subset of bighorn sheep data that is not representative of the study area and was not gathered in a manner conducive to the analysis methods used by the authors. The authors classified habitat pixels as “active” or “inactive” based on the presence–absence of bighorn sheep observations without considering monitoring intensity. Turner et al. (2004) also failed to consider the implications of basing their model almost entirely on a bighorn sheep subpopulation known to have atypical habitat selection patterns. This subpopulation in the northwestern Santa Rosa Mountains frequently used food and water sources within hillside urban areas. Because the Turner et al. (2004) model was developed using data primarily from this atypical subpopulation, the model has low external validity and is unlikely to accurately predict habitat selection by other bighorn sheep subpopulations in the Peninsular Ranges. Furthermore, with the NW subpopulation used in model development now excluded from urban areas, the Turner et al. (2004) model is unlikely to accurately predict habitat selection patterns of even this subpopulation. We suggest the Turner et al. (2004) model is at best only applicable to this subpopulation between the years 1994–1998.

In the absence of a reproducible validated critical-habitat delineation for Nelson's bighorn sheep (Ovis canadensis nelsoni) in the northern Santa Rosa Mountains, California, Turner et al. (2004) developed a quantitative habitat model based on biotic and abiotic habitat parameters comprised of the primary constituent elements allegedly used in the United States Fish and Wildlife Service's (USFWS) habitat model (USFWS 2000, 2001). Bighorn distribution and location (observation) data used by the USFWS in its model were 96% explained on the basis of the Turner et al. (2004) model; no validation or prediction could be made from the USFWS model (USFWS 2000, 2001). Criticisms by Ostermann et al. (2005) of the Turner et al. (2004) model emphasize a basic miscomprehension by Ostermann et al. (2005) of the differences between habitat characterization and habitat utilization. Most of the Ostermann et al. (2005) critique relates to data that were not provided to Turner et al. (2004). Failure of Ostermann et al. (2005) to fully grasp the sampling strategy used by Turner et al. (2004) relates to an incomplete comprehension of the statistical methodology employed rather than with the statistical procedures themselves. Claims by Ostermann et al. (2005) that Turner et al. (2004) made extrapolations to misdirect bighorn recovery efforts, drew inferences to other bighorn populations, alleged the northern population was normal, and equated sheep density (frequency of observation) to habitat quality are without substantiation, example, or merit. Contentions of problematic water source data are unsupported by the literature cited by Ostermann et al. (2005). The permissive attitude of resource agencies allowing management and conservation efforts to be directed by special interests largely in the absence of independent scientific scrutiny, casts a question of legitimacy over this Endangered Species Act listing, recovery strategy, and critical habitat delineation.