We question whether the growing popularity of model selection based on information theory (IT) and using the Akaike's Information Criterion (AIC) represent a useful paradigm shift in data analysis or a substitution of 1 statistical ritual for another, which leaves in place long-standing problems in wildlife science. We discuss the relevance of model selection in science, problems in the IT-AIC algorithm, errors of commission and omission in IT-AIC-based studies, and the role of IT-AIC in knowledge accrual. Model selection is just another minor tool in the grand panorama of science. The human mind, not statistical methods, produces scientific breakthroughs. Although IT-AIC might include elements of hypothetico-deductive science, it is arguably a form of sensitivity analysis, magnitude of effects estimation, or simple description as currently applied. Accordingly, it is largely an inductive approach to knowledge accrual and, therefore, subject to the pitfalls of induction. The algorithm tends to over fit data (i.e, use too many variables), resulting in models that contain useless variables and that generalize poorly. Errors of commission in IT-AIC-based papers include hopelessly uninformative lists of encrypted models and imposition of the model-selection approach on studies better executed in a simple, descriptive format. The major error of omission is an almost universal failure to test selected models on independent data. From our perspective, IT-AIC is a harmless human construct that is being ritualistically applied and therefore cannot be expected to correct long-standing problems in the conduct of wildlife science, such as failure to apply the hypothetico-deductive method. We view the growing application of IT-AIC as problematic because that growth might discourage use of the full panoply of available methods of inquiry. Accordingly, we urge colleagues to avail themselves of the rich pageant of available analytical techniques that can be applied in wildlife research under the hypothetico-deductive method and to keep ecology, rather than statistics, in the forefront of wildlife science.

Adaptive and science-based management is widely accepted as necessary to safeguard wildlife and their habitats into the future. However, many of the decisions in this field are still based on unsupported ideas that lack validation with real data and which do not make their analysis available for a public review. Decisions based on soft foundations can be harmful to wildlife, habitat and the survival of both. I suggest a new wildlife management approach founded on scientific databases that has become possible, if not imperative, with improved technology and increasing access to freely available data over the Word Wide Web (WWW). This approach is partly a consequence of the effective implementation of the U.S. Freedom of Information Act and the National Spatial Data Infrastructure. In order to justify management decisions relating to wildlife, habitat, and conservation, the listing, use, and full investigation of all available and relevant databases needs to be implemented, voluntarily or legally, as a prerequisite. Second, similar to International Organization for Standardization (ISO) Standards, each major wildlife management decision needs to add a standard management documentation system as a backup that clearly records what data and research were or were not available at that time and what the recommended research still needs to address in order to complete the data situation and to decrease uncertainty.

Efforts to design monitoring regimes capable of detecting population trends can be thwarted by observational and economic constraints inherent to most biological surveys. Ensuring that limited resources are allocated efficiently requires evaluation of statistical power for alternative survey designs. We simulated the process of data collection on a landscape, where we initiated declines over 3 sample periods in species of varying prevalence and detectability. Changing occupancy levels were estimated using a technique that accounted for effects of false-negative errors on survey data. Declines were identified within a frequentist statistical framework, but the significance level was set at an optimal level rather than adhering to an arbitrary conventional threshold. By varying the number of sites sampled and repeat visits made, we show how managers can design an optimal monitoring regime that maximizes statistical power within fixed budget constraints. Results show that 2 to 3 visits/site are generally sufficient unless occupancy is very high or detectability is low. In both cases, the number of required visits increase. In an example of woodland bird monitoring in the Mt. Lofty Ranges, South Australia, we show that, although the budget required to monitor a relatively rare species of low detectability may be higher than that for a common, easily detectable species, survey design requirements for common species may be more stringent. We discuss implications for multi-species monitoring programs and application of our methods to more complex monitoring problems.

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 km²) 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.

Quail hunting consists of a complex set of behaviors that involve humans, pointing dogs, and wild birds. The recent development of models known as Hunter–Covey Interface (HCI) theory provides an opportunity to analyze how we perceive, understand, and manage quail hunting. There are 2 groups of HCI models: static and dynamic. The static HCI model predicts daily hunting mortality based on the velocity of the hunt and area covered during a hunt. The dynamic HCI models estimate the probability of flushing a covey given a set of circumstances that revolve around the potential rate of which quail learn to avoid hunters. We quantified the variables required to test whether the static and dynamic models of HCI theory provide meaningful results. We used spatial data on hunting velocity and area covered during >100 quail hunts, along with estimates of quail population density, mortality, and movements from 2 areas in south Texas, to evaluate output from HCI models. The static model predicted average daily harvest rates that ranged from 5 to 50 birds. This average is within the range of average daily bag of south Texas quail hunters in groups of 2–4. Output from the dynamic models suggested that quail populations on our study areas were subjected to relatively low hunting intensities with a corresponding low avoidance behavior and learning rate, which is a scenario that matched 1 of 4 predictions by Guthery (2002). We found it difficult to meet basic assumptions of HCI theory. For example, hunting pressure was potentially redundant, coveys were not always randomly distributed in space, and the extent to which quail are naive at the beginning of a hunting season was unknown. However, both the static and dynamic HCI models appeared robust to violation of these assumptions. Application of HCI theory may provide meaningful results that can be used to manage quail hunting pressure, optimize harvest, and sustain populations.

Agricultural land use may indirectly affect the body size of amphibians by altering the hydroperiods of nearby wetlands and influencing amphibian densities—both factors which can limit the larval and postmetamorphic growth rates of amphibians. We measured postmetamorphic body size for 4 species (Spea multiplicata, S. bombifrons, Bufo cognatus, Ambystoma tigrinum mavortium) and 3 age classes (metamorph, subadult, adult) of amphibians captured at playa wetlands surrounded by one of 2 general land-use types (cultivation, grassland) in the Southern High Plains. Sixteen playas (4 per land-use type in 1999 and 2000) were partially enclosed with drift fence and pitfall traps, and mass and snout-vent length (SVL) were measured from a subsample of captured individuals. Mass and SVL were 10–148% greater for amphibians captured at grassland wetlands than at cropland wetlands for most species and age classes. Mass and SVL also were 3–124% greater in 1999 than in 2000 for most species and age classes. We attribute differences in body size between land-use types to differences in the hydroperiods of the associated wetlands, and potentially to variation in the density of terrestrial conspecifics and aquatic predators. We attribute differences in body size between years to differences in rainfall. Body size is positively related to the probability of survival, reproduction, and evolutionary fitness in amphibians. Thus, if cultivation of landscapes surrounding wetlands negatively influences postmetamorphic body size of amphibians, restoration of native grasslands surrounding playa wetlands may help prevent local amphibian declines.

We compared the influence of clearcut and selective timber harvest treatments on spatial and temporal variability of amphibians and reptiles in an east Texas bottomland hardwood forest. The dataset represented a time series of 5 years post-treatment. A total of 18,645 amphibians and reptiles was captured in 144 pitfall arrays. We used 9 plots (3 clearcut, 3 select cut, and 3 untreated). Each plot had 16 arrays and was bisected by 1 of 3 streams. Pitfall captures represented 46 species (16 amphibians, 30 reptiles). When analyzed with a traditional ANOVA approach, these data suggested an increase in reptile species richness in response to clearcut treatments; amphibian species richness did not respond to treatment. When analyzed as a time series, however, the data revealed fluctuations in site use by species and species groups, and these fluctuations were independent of treatment effects. Exploratory analyses of spatio-temporal dynamics showed that species richness and the relative abundance of common species displayed spatial patterns that remained consistent over time. In control and select cut treatments, spatial patterns of richness and abundance shifted over time and were not necessarily confined to areas adjacent to streams. In clearcuts, stationary habitat refugia were located within riparian management zones.

Leatherback sea turtles (Dermochelys coriacea) nesting in French Guiana have been the focus of a long-term monitoring and metal flipper-tagging program and were subjected to a double-tagging experiment starting in 1994. We developed a new method to estimate the rate of tag loss based on the number of days at liberty after tagging, using maximum likelihood function fitting. Metal tag loss in leatherback sea turtles is characterized by a high rate of loss just after tagging (>30% of turtles lose at least 1 tag within a year) and is followed by higher retention over a longer time period. Using additional data from permanent internal tags that were subsequently applied to turtles in French Guiana, we have shown that the rate of loss of the second metal tag was higher than that of the first metal tag. More accurate estimates of tag loss rates are essential for correctly generating demographic parameters from Capture–Mark–Recapture (CMR) analyses. We have developed a software package to facilitate the use of these models by others researchers in estimating tag loss.

Roads built through or near wetlands cause significant mortality of reptiles and amphibians and create barriers to migration and dispersal. I investigated the number of times turtles and other herpetofauna attempted to cross a 4-lane highway at Lake Jackson, Florida, USA, during a period of severe drought (Feb–Apr 2000). Levels of road mortality were so high that I designed and installed a temporary drift fence system to work with an existing drainage culvert and for the next 2.5 years I evaluated its effectiveness at reducing road mortality and facilitating migration. I monitored roads and fences several times per day for 44 months, during both drought and non-drought conditions. A total of 10,229 reptiles and amphibians of 44 species were found either road killed or alive behind drift fences: 8,842 turtles, 838 frogs, 363 snakes, 152 lizards, 32 alligators, and 2 salamanders. Drift fences combined with intensive monitoring greatly reduced turtle road kills and facilitated the use of an under-highway culvert. Along a 0.7-km section of the highway, turtle mortality before installation of the fence (11.9/km/day) was significantly greater than post-fence mortality (0.09/km/day) and only 84 of 8,475 turtles climbed or penetrated the drift fences. Pre-fence data provided strong evidence that turtles cannot successfully cross all 4 lanes of U.S. Highway 27, as 95% of 343 turtles were killed as they first entered the highway adjacent to the shoulder and the remaining 5% were killed in the first two traffic lanes. According to a probability model, the likelihood of a turtle successfully crossing U.S. Highway 27 decreased from 32% in 1977 to only 2% in 2001 due to a 162% increase in traffic volume. Therefore, at least 98% of turtles diverted by the fences probably would have been killed if fences were not in place. The results of this study represent the highest attempted road-crossing rate ever published for turtles (1,263/km/year). Because of demographic and life history constraints, turtle populations may incur irreversible declines in areas where road mortality is high, especially when mass migrations are triggered by periods of drought.

Changes in harvest rate over the past 3 decades have been shown to be closely related to population growth of greater snow geese (GSG; Chen caerulescens atlantica). We used band-recovery and harvest survey data from 1970 to 2001 to study temporal variations in geographic harvest distribution and composition of GSG in Québec, Canada and the Atlantic Flyway states (AF) in the United States. We sought to determine whether (1) geographic variation in harvest was associated with temporal trends in total harvest rates observed during this period; (2) spatiotemporal distributions of harvest varied with age and sex; and (3) harvest distribution and composition differed between the spring conservation harvest initiated in 1999 and the regular fall hunt in Québec. We detected over time a gradual spreading in the geographic distribution of the fall harvest from the upper St. Lawrence estuary toward southwestern Québec. During winter, a sudden northward shift in the distribution of the United States harvest in the mid-1980s was associated with a high concentration of geese in mid-Atlantic Flyway states (Maryland, Delaware, and New Jersey), possibly due to short-stopping during migration. We argue that this led to a reduction of hunting pressure on GSG and may have contributed to the sudden decline in harvest rate that occurred at that time and ultimately to the ensuing population increase. We observed a decreasing proportion of juveniles in the kill throughout the hunting season within fall staging grounds in Québec and between Québec and the United States. We also found a much higher proportion of adults in the spring harvest than in the fall that is consistent with the conservation goal of increasing the adult harvest. We recommend that management actions focus on increasing harvest in mid-Atlantic Flyway states if further control of population growth is desired.

Unlike most prairie-nesting ducks, the North American population of northern pintails (Anas acuta) has shown limited response to improved wetland conditions on the U.S. and Canadian prairies during the mid to late 1990s. Because adult female survival is a key parameter affecting waterfowl population dynamics, and a large fraction of annual mortality for some dabbling ducks occurs during the nesting season, we used radiotelemetry to estimate nesting season survival and cause-specific mortality of adult female northern pintails that were monitored during an induced renesting study. We conducted our research in a predominantly agricultural landscape in southern Saskatchewan during 1998–2000, an area that is typical of much of the pintail's prairie breeding range. The survival rate of 140 females for a 75-day interval (30 Apr–14 Jul) during the nesting season was 0.806 ± 0.046. Survival did not vary among years or between ages (second-year, after second-year), nor was it related to female body condition (body mass adjusted for structural size, date, and year of capture). We documented 15 mortalities and found that the 75-day mortality rate due to raptors (0.141 ± 0.040) was greater than that due to red fox (Vulpes vulpes; 0.011 ± 0.011), collisions with power lines (0.011 ± 0.011), or unknown factors (0.039 ± 0.027). Our results differ from prior research that indicated red fox was a major predator of female pintails during the nesting season. Raptors were the primary mortality agent for female pintails, perhaps because of limited exposure during incubation in our study, or the use of open habitats by pintails in southern Saskatchewan. Management programs that provide nesting or wetland habitats with overhead concealment, or that reduce perch sites for raptors in areas with high pintail densities, may decrease the risk of avian predation on female pintails.

We estimated variation in nest survival of lesser scaup (Aythya affinis), greater scaup (A. marila), and other common duck species at Minto Flats, Alaska, USA, during 1989–1993 and 2002–2003. Daily survival probability of scaup nests, as well as nests of all other duck species, varied with year, date, and nest habitat. Daily survival probability was unrelated to nest age and distance from the nest to water. Average, year-specific nest survival of all ducks at Minto Flats was 0.11 (95% CI: 0.05 to 0.22), comparable to nest survival of ducks breeding in mid-continent regions (i.e., the prairie pothole region and the Canadian prairie-parklands). Nest survival of scaup was variable among years, ranging from 0.01 (95% CI: 0.00 to 0.06) in 1992 to 0.61 (95% CI: 0.50 to 0.74) in 1993 and was probably related to variation in predation risk and water levels. Scaup production could have been limited by low nest survival during most years of this study; nest survival exceeded 20% only in 1993 and 2002. Because of the high densities of breeding scaup and other waterfowl species at Minto Flats, we recommend management to maintain existing habitat for breeding scaup and other duck species. This management could be most effectively informed by yearly monitoring of production to more accurately understand spatial and temporal variation in recruitment and to identify potential effects of proposed oil and gas exploration on recruitment of ducks at Minto Flats.

We studied duckling survival of lesser (Aythya affinis) and greater (A. marila) scaup at Minto Flats, Alaska during 2002–2003. We captured and marked 42 female scaup during the study, and we subsequently monitored survival of 32 scaup broods with 284 ducklings. Daily Survival Rate (DSR) of ducklings varied between years and was positively related to female body condition and duckling age. Estimated duckling survival to 30 days for both species of scaup was 0.24 (95% CI: 0.16 to 0.36) in 2002 and 0.03 (95% CI: 0.00 to 0.19) in 2003. Because we obtained little support for relationships between environmental covariates and duckling survival, we speculated that most of the variation between years could have been related to high rates of predation by gulls (Larus spp.) during 2003. Estimated recruitment of female scaup to 30 days (clutch size * nest survival * duckling survival * 0.5) was 0.30 (SE = 0.19) in 2002 and 0.01 (SE = 0.01) in 2003. Given the low recruitment we observed at Minto Flats and the current population status of scaup, restriction of harvest of adult female scaup could be useful to support scaup populations at Minto Flats.

Mid-air collisions with black vultures (Coragyps atratus) and turkey vultures (Cathartes aura) regularly cause substantial damage to military and civilian aircraft. Information concerning the flight behavior of black and turkey vultures potentially could improve predictive models designed to reduce bird strikes by aircraft. We examined the flight behavior of black and turkey vultures at the Savannah River Site (SRS) in South Carolina, USA, and determined whether flight characteristics were predictable with respect to weather and time variables. We captured birds at their primary roost and subsequently relocated them via aerial telemetry from 11 February 2002 through 29 January 2003. One hundred eighty of 326 locations (55%) for 8 black vultures and 129 of 206 locations (63%) for 5 turkey vultures were of flying birds. Black vultures flew at an average altitude of 169 ± 115 (SD) m above ground level, whereas turkey vulture flight altitude averaged 163 ± 92 m. Our results contrast with those of previous studies that reported less frequent and lower altitude flights. The flight behavior of both species appeared to be influenced minimally by weather and time variables. However, we were unable to construct useful models predicting aspects of flight behavior using the variables we measured (all models had R² or pseudo R² values <0.10). We suggest that other factors, such as food availability, inter- and intra-specific interactions, and physiological demands play a larger role in vulture flight behavior than the variables we measured. Our results suggest that the development of bird avoidance strategies by aircraft operators should consider the variability of flight behaviors of black and turkey vultures across their ranges. Future research emphases should shift from examinations of the effects of local conditions on flight behavior to the elucidation of factors contributing to differences in flight behavior among regions.

We studied the summer foraging ecology of resident and migrant bald eagles (Haliaeetus leucocephalus) along the lower Hudson River, New York, from 1998 to 2001. In this area the Hudson is a freshwater tidal river with 1- to 2-m tidal ranges. Eagles foraged most often in the open channel (35%), where success was lowest (68% capture rate). When compared to landscape availability, eagles foraged in tidal mudflats devoid of aquatic vegetation more often than expected, and they avoided areas of deep water (>3 m). Eagles foraged more often during ebb tides with foraging activity peaking just before low tide. Eagles avoided areas of high human activity but also preferentially selected areas of low to moderate activity. Fish were the most important source of food and comprised 91% of prey identified. Over 50% of the observed prey captures consisted of 3 species: American eel (Anguilla rostrata), gizzard shad (Dorosoma cepedianum), and white catfish (Ictalurus catus). Our data indicate that unvegetated tidal mudflats that were isolated from intensive human activity provided the highest quality foraging habitat. Future loss of tidal mudflats through exotic plant invasions or shoreline development may limit eagle foraging opportunities and population growth.

We studied Mexican spotted owl (Strix occidentalis lucida) diets and the relative abundance and habitat associations of major prey species in a ponderosa pine (Pinus ponderosa)–Gambel oak (Quercus gambelii) forest in north-central Arizona, USA, from 1990 to 1993. The owl's diet was comprised of 94% mammals by biomass and consisted of primarily the deer mouse (Peromyscus maniculatus), brush mouse (P. boylii), Mexican woodrat (Neotoma mexicana), and Mexican vole (Microtus mexicanus). Spotted owl prey in our study area were smaller on average than prey in other locations, and the total biomass of potential prey was less than that reported in other areas within the owl's geographic range. Although all prey populations exhibited seasonal fluctuations in relative abundance, only the deer mouse exhibited significant temporal variation in population abundance. The general pattern was for prey populations to rise during spring, peak during summer, decline in fall, and reach a winter low. Deer mice exhibited the greatest amplitude in population change as evidenced by the shift from a high of 12.2 mice/ha (SE = 2.3) during summer 1991 to a low of 3.3 mice/ha (SE = 0.7) during winter 1991–1992. Woodrats and brush mice used areas on slopes >20°with relatively more rocks and shrub cover than found in other areas. In contrast, deer mice were found in forests with relatively open understories and little Gambel oak. Conservation measures for the Mexican spotted owl must include management directed at sustaining or increasing prey numbers rather than assuming that managing for owl nesting and roosting habitat will provide favorable conditions for the prey as well. Management practices that increase and sustain shrub and herbaceous vegetation should receive the highest priority. This can be accomplished by thinning small diameter trees, using prescribed fire, and managing grazing pressures.

The newly described Gunnison sage-grouse (Centrocercus minimus) is a species of concern for management because of marked declines in distribution and abundance due to the loss and fragmentation of sagebrush habitat. This has caused remaining populations to be unusually small and isolated. We utilized mitochondrial DNA sequence data and data from 8 nuclear microsatellites to assess the extent of population subdivision among Gunnison sage-grouse populations in southwestern Colorado and southeastern Utah, USA. We found a high degree of population structure and low amounts of gene flow among all pairs of populations except the geographically adjacent Gunnison and Curecanti populations. Population structure for Gunnison sage-grouse was significantly higher than has been reported for greater sage-grouse (C. urophasianus). Further, we documented low levels of genetic diversity in some populations (particularly Dove Creek/Monticello and Piñon Mesa with an average of only 3.00 and 2.13 alleles per locus respectively) indicating that translocations from larger, more genetically diverse populations may be warranted. Bayesian analysis identified 3 potential migrants (involving San Miguel, Dove Creek/Monticello, Crawford, and Curecanti). Further, this analysis showed that 4 individuals from Cerro/Cimarron were more closely related to birds from San Miguel than to its geographically closer neighbors Gunnison and Curecanti. This suggests the Cerro/Cimarron area may act as a stepping stone for gene flow between San Miguel and Gunnison and that habitat restoration and protection in areas between these 2 basins should be a priority in an attempt to facilitate natural movement among these populations. Conservation plans should include monitoring and maintaining genetic diversity, preventing future habitat loss and fragmentation, enhancing existing habitat, and restoring converted sagebrush communities.

Nesting habitat degradation and its negative effect on nesting success might contribute to the recent population and distributional declines of greater sage-grouse (Centrocercus urophasianus) throughout North America. We used radiotelemetry to locate greater sage-grouse nests in 7 different areas of central and southwestern Wyoming between 1994 and 2002; we studied each area for 2 to 4 years. Using binary logistic regression, we compared microsite vegetal data collected at nests (n = 457) and random (n = 563) sites and successful (n = 211) and unsuccessful (n = 238) nests to test hypotheses concerning greater sage-grouse nesting habitat selection and vegetal conditions associated with nesting success. We used Akaike's Information Criterion (AIC_c) and model averaging to make inference about the weighted support for the importance of individual habitat variables through the comparison of sets of competing models. Selected nest sites were located in areas with increased total shrub canopy cover (relative importance [RI] = 1.00), residual grass cover (RI = 0.47), and residual grass height (RI = 0.77) compared to random sites. Successful nests had increased residual grass cover (RI = 0.43) and height (RI = 0.48) relative to unsuccessful nests. Additionally, annual nest success rates (i.e., above vs. below our study's average) were related to the preceding year's spring (Apr–May; RI = 0.44) and winter–early spring (Jan–Jun) precipitation (RI = 0.32). Correct classification rates for weighted average models that we derived through the 3 comparisons were between 60 and 70%, suggesting the variables adequately differentiated between plot types. However, high model selection uncertainty (i.e., the total number of models included in the sets of AIC_c-selected models) suggested that nest site selection and nesting success may be influenced by factors not considered in the modeling process. Management strategies that protect dense sagebrush stands and enhance residual grass cover and height within those stands should be used to maintain nesting habitat and increase nesting success of greater sage-grouse.

The introduction of pen-reared Attwater's prairie-chickens (APC, Tympanuchus cupido attwateri) into the wild to supplement existing populations has met with marginal success. Flight characteristics, predator avoidance behavior, and rearing methods may contribute to post-release mortality of pen-reared birds. We compared flight characteristics and predator avoidance behaviors of pen-reared APC released onto the Attwater Prairie Chicken National Wildlife Refuge with those of a sister subspecies, wild greater prairie-chickens (GPC, T. c. pinnatus) in Minnesota and Kansas. There was no difference (P = 0.134) in flight speed of pen-reared APC and wild GPC. However, wild GPC flew further (P <0.001) in response to an approaching human than did pen-reared APC. Wild GPC flushed at a greater (P <0.001) distance from an approaching human than did pen-reared juvenile APC that had lived in the wild for less than 3 months. A trained dog was able to approach closer (P <0.001) to APC than to GPC before birds flushed, and APC did not fly as far as GPC after being flushed by a dog. The deficiencies in flight endurance and predator avoidance behavior of pen-reared APC compared to wild GPC could explain the high mortality of APC in the wild. Flight conditioning in circular flight pens using dogs to scare and harass captive birds at an early age could improve birds' predator avoidance behavior and increase flight endurance by allowing them to exercise their flight muscles. If the recovery of the APC is to succeed, the behavioral deficiencies of pen-reared birds should be addressed.

We propose a metric for estimating the quantity (ha) of usable space (suitable permanent cover) for wildlife on an area of interest with use-availability data. The metric and derivations from it provide information for better understanding the field behavior of wildlife, planning habitat management, and estimating population response to habitat management. We illustrate the metric with radiotelemetry (use) and Geographic Information System (GIS; availability) data on northern bobwhites (Colinus virginianus) from Texas and Arkansas, USA. In Texas, a 796-ha area provided 225 ± 1.8 ha SE of usable space (n = 9,288 radiolocations of 217 bobwhites), and we found an avoided cover type provided more usable space (i.e., was more important) than a selected type. During the breeding season of 2001 in Arkansas, a 5,560-ha area provided 1,461 ± 65.1 ha SE of usable space (n = 300 radiolocations of 16 bobwhites). A derivation from the metric provides a method for estimating the contribution of weakly selected, randomly used, and avoided cover types to population welfare in the sense of usable space. The derivation also reduces concern over arbitrary human classification of cover types because it provides quantitative information on the effects of those classifications.

Northern bobwhites (Colinus virginianus) are captured routinely for scientific study. Extreme exertion during restraint and transport can lead to acidosis and free radical production that result in muscle tissue damage. This condition, termed capture myopathy, can result in dyspnea, hyperthermia, weakness, muscle rigidity, and collapse. These complications could lead to increased susceptibility of individuals to predation or contribute to death weeks or months after capture. Death caused by capture could negatively bias mortality rate estimates if mortalities occur after traditional censor periods. We hypothesize that muscular damage incurred during capture handicaps northern bobwhites. We evaluated our hypothesis by comparing survival of northern bobwhite that were treated for muscular damage with an injection of vitamin E and selenium, a solution recommended in veterinary avian medicine, with bobwhites that were not treated. We captured northern bobwhite during 2002 and 2003 and injected half of the birds with 0.1 ml of a vitamin E and selenium solution and the other half with a saline control. We then transported these birds to our study site and monitored them using radiotelemetry. We estimated survival rates using the staggered-entry Kaplan-Meier Product Limit Estimator. The survival of treated birds was greater than the survival of control birds in 2002 at days 45 (58% vs. 29%, Z = 1.98, P = 0.05) and 66 (53% vs. 29%, Z = 1.79, P = 0.07), a period before and after which sample sizes were very small. Treated birds exhibited a greater (χ² = 7.11, P <0.008) survival curve than control birds during 2003 and when treatments were pooled by year (χ² = 4.19, P = 0.04). Thus, our data suggest that injection of wild northern bobwhite with vitamin E and selenium increased survival of transplanted birds as compared to control birds injected with saline.

Knowledge of the length of time spent at migratory stopover sites, or stopover duration, can help biologists estimate the total number of birds passing through a site, which in turn can be used to estimate population size. We estimated the stopover duration for 106 radiomarked pectoral sandpipers (Calidris melanotos) in the Lower Mississippi River Alluvial Valley (MAV) during fall migrations in 2001 and 2002. We estimated time-at-site-after-capture by tracking radiomarked birds daily to determine their time of departure from stopover sites. We used length-biased sampling and program DISTANCE to estimate stopover duration from time-at-site-after-capture. Our estimate of stopover duration for pectoral sandpipers in the MAV over all sites and both years was 10.0 days (95% CI = 8.2–11.7). Aerial telemetry relocations from 2002 indicated that many pectoral sandpipers used multiple stopover sites during their migration through the MAV, which suggested that the total time spent by pectoral sandpipers in the MAV during fall migration, or turnover rate, was longer than the stopover duration. Future research should focus on estimating the number of stopover sites used by pectoral sandpipers and determine whether other shore-bird species use multiple sites as well. If shorebirds migrating through the MAV stay in the region for a period longer than the 10 days suggested by the Lower Mississippi Valley Joint Venture (LMVJV) Migratory Bird Science Team in their modeling of shorebird energetic requirements, then more habitat may be required by shorebirds during fall migration than the 2,000 ha the group calculated for the MAV.

Populations of shrubland birds in eastern North America have consistently declined since the 1960s, but conservation is hampered by an inadequate understanding of the area requirements of most species. We examined the sensitivity of shrubland specialists to (a) the area of shrub stands and (b) proximity to mature-forest edges, and we evaluated whether habitat characteristics, food resources, or productivity of bird populations could have caused the relationships we identified. In 2002–2003, we used constant-effort mist-netting on 6 small (4–8 ha) and 6 large (13–16 ha) regenerating clearcuts that were 4–6 years post-harvest in southern Ohio, USA. We placed 3 nets at 20, 50, and 80 m from the mature-forest edge (n = 9 nets per site), and we sampled vegetation, fruit, and arthropods at each net. Seven of 8 shrubland specialists, particularly blue-winged warbler, prairie warbler, yellow-breasted chat, indigo bunting, and field sparrow, avoided mature-forest edges, with twice as many birds caught 80 m from edges compared to 20 m. Abundances of most species, especially yellow-breasted chats, were positively correlated with area, though the combined area effect was not statistically significant. We found no evidence of reduced avian productivity in small stands. Neither area nor edge was associated with habitat characteristics, fruit abundance, or arthropod biomass. Our results suggest shrubland birds avoid habitat edges. Thus, small or narrow cuts may not provide optimal habitat for this suite of declining species, and managers should consider options to minimize edge and provide larger patches of shrubland habitats in landscape-scale planning efforts.

We examined response of breeding bird communities to forest harvest that removed varying amounts of tree basal area from riparian buffers on a 2- to 4-m-wide stream in northern Minnesota, USA. We compared bird species and communities in 30-m-wide riparian buffers along the stream. Buffers were established within plots in which upland forests were clear-cut (basal area 2 m²/ha) according to standard local forest management practice. Buffers had 4 treatments (3 plots/treatment): (1) no harvest (riparian control); (2) reduction of basal area to an average of 7–10 m²/ha; (3) reduction of basal area to an average of 2 m²/ha (defined as a clear-cut); and (4) control (no harvest in either riparian buffer or adjacent upland). Bird surveys were conducted 1 year prior to harvest and for 4 years after harvest. Results revealed a significant response of the bird community to varying amounts of tree basal area retention in the riparian area. Univariate (analysis of variance) and multivariate (principal response curves [PRC]) analyses showed that in the first year after harvest, bird community composition in the riparian buffers changed in all 3 treatments relative to the control plots, and continued to diverge over time. More species and individuals, primarily those species associated with edge or early-successional habitats, colonized the harvested riparian buffers after treatment. In contrast, the number of birds and species that inhabit interior forests declined in the riparian buffers. Results suggest that any amount of harvest in riparian buffers next to clear-cut upland forest will affect breeding bird communities along small headwater streams. Because individual bird species are differentially affected by riparian forest harvest, management should consider the desired future condition of the forest and choose a harvest prescription to benefit the desired avifauna community.

We collected data on 212 wood thrush (Hylocichla mustelina) nests in central New York from 1998 to 2000 to determine the factors that most strongly influence nest success. We used an information–theoretic approach to assess and rank 9 models that examined the relationship between nest success (i.e., the probability that a nest would successfully fledge at least 1 wood thrush offspring) and habitat conditions at different spatial scales. We found that 4 variables were significant predictors of nesting success for wood thrushes: (1) total core habitat within 5 km of a study site, (2) distance to forest–field edge, (3) total forest cover within 5 km of the study site, and (4) density and variation in diameter of trees and shrubs surrounding the nest. The coefficients of these predictors were all positive. Of the 9 models evaluated, amount of core habitat in the 5-km landscape was the best-fit model, but the vegetation structure model (i.e., the density of trees and stems surrounding a nest) was also supported by the data. Based on AIC weights, enhancement of core area is likely to be a more effective management option than any other habitat-management options explored in this study. Bootstrap analysis generally confirmed these results; core and vegetation structure models were ranked 1, 2, or 3 in over 50% of 1,000 bootstrap trials. However, bootstrap results did not point to a decisive model, which suggests that multiple habitat factors are influencing wood thrush nesting success. Due to model uncertainty, we used a model averaging approach to predict the success or failure of each nest in our dataset. This averaged model was able to correctly predict 61.1% of nest outcomes.

We investigated habitat selection and home-range characteristics of American martens (Martes americana) that occupied home ranges with partially harvested stands characterized by basal area of trees <18 m²/ha and canopy closure <30%. During the leaf-on season (1 May–31 Oct), martens selected second-growth (80–140-years-old, >9-m tree height) forest stands (deciduous, coniferous, and mixed coniferous-deciduous) and mixed stands that were partially harvested (x̄ = 13 m²/ha residual basal area, >9-m tree height), and they selected against forests regenerating after clearcutting (≤6-m tree height, cuts ≤24-years-old). Marten home ranges included a greater proportion of partially harvested stands during the leaf-on season (maximum = 73%) than during leaf-off (1 Nov–30 Apr; maximum = 34%). Higher use of partially harvested stands during the leaf-on season coincided with greater canopy closure, higher use of small mammals, and greater relative densities of small mammals. During the leaf-off season, martens exhibited reduced relative selection for partially harvested and regenerating stands and increased selection for second-growth forest types. Partially harvested and regenerating clearcut stands had canopy closure <30% and basal area of trees >9-m tall of <13 m²/ha; both were below published thresholds required by martens. Coincidentally, home-range areas of martens increased during the leaf-off season to include a greater proportion of second-growth forest and less partially harvested forest. Further, martens with partial harvesting in their home ranges used areas almost twice as large during the leaf-off season as martens with no partial harvesting. Snowshoe hares (Lepus americanus) were prevalent prey for martens during the leaf-off season, and partially harvested stands had the lowest density of hares among all forest overstory types. Our findings suggest that the combination of insufficient basal area and overhead canopy closure, subnivean behavior of small mammals, increased reliance on hares, and reduced density of snowshoe hares relative to second-growth forest types reduced habitat quality in partially harvested stands during the leaf-off season. We suggest land managers retain basal areas >18 m²/ha and canopy closure >30% during winter to maximize use by martens in stands where partial harvesting is practiced.

I used estimates of energetic requirements of pumas (Puma concolor) to estimate prey requirements for mule deer (Odocoileus hemionus). This estimate might help explain the impact pumas may have on deer populations. Accepted daily energy demands of pumas result in estimates of annual kill rates per puma of 44 to >100 mule deer/yr. However, recent studies of puma activity indicate that the methods used to calculate these earlier values may have been flawed. To test this possibility, I used a more recently developed allometric equation to calculate energy demand of pumas based on movement data from a radio telemetry study in south-central Idaho and northwestern Utah. I calculated energy demand for 20 adult female and 8 adult male pumas based on their movements during 95 monitoring blocks of 24 hrs. I then calculated prey requirements for individuals and for low and high population estimates from the study area. I estimated that daily energy demands for male pumas (3,143.7 ± 120.1 kcal; 13,203.5 kilojoules; kJ) and female pumas with kittens (2705.4 ± 57.0 kcal; 11,362.7 kJ) were substantially lower than previously reported estimates of approximately 5,500 kcal/day (23,100 kJ/day) for males and females with an average (2.6 kittens) litter size. The resulting estimated annual kill rates (19.4 deer/yr for males, 39.6 deer/yr for females with kittens) were also substantially lower than previous estimates (44 deer/yr and 73 deer/yr respectively). I demonstrated that the earlier energy demand estimates were based on overestimations of puma activity for these 2 social groups and that my estimates of activity, and thus energy demand, are more biologically reasonable for pumas. I discuss the need for a reassessment of the impact pumas can have on deer populations in light of these new energy calculations.

Ocelots (Leopardus pardalis) are listed as endangered federally and by the state of Texas. Preference for closed canopy habitat has been shown in previous studies, but preference for patch size has not been reported. Geographic Information Systems (GIS) and satellite imagery were used to compare areas in south Texas used by radio-collared ocelots to areas with no known use. We hypothesized that ocelots would prefer large patches of closed canopy habitat and avoid large patches of unsuitable habitat. Areas used by ocelots had a greater degree of fragmentation (i.e., larger number of patches, smaller size, and more edge) than did those not used. Further investigation revealed that ocelots preferred patches of closed canopy over other types of land cover and that this land cover type exhibited a greater degree of fragmentation. Results of this study were used to designate areas for conservation of ocelot habitat and can be applied to the management of other threatened or endangered wildlife.

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 Gaspé 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.

The decline of aspen (Populus tremuloides) in the western United States is well known but the role that pocket gophers (Thomomys spp.) may play in that decline is not known. We investigated the effects of pocket gophers on aspen regeneration in Utah through a series of 4 treatments comprised of baited and fenced plots. The treatments were control (no treatment), baited (pocket gopher removal), fenced (ungulate exclusion), and fenced and baited combined. Fencing reduced browsing rates and increased the height and growth of aspen suckers but did not affect their density. Pocket gopher removal had no effect. Aspen suckers on control plots were the only ones that showed no increase in growth between spring and fall. Overall, these results indicate that the effect of pocket gophers on aspen regeneration is minimal compared to the effects of browsing by ungulates and livestock.

During the last 2 decades, European brown hare (Lepus europaeus) populations have declined considerably. We evaluated whether this decline could have been associated with diminished overall fitness due to reduced genetic variability. We also estimated the extent of population differentiation. We typed 307 hares from 21 localities for 5 microsatellite loci and the mitochondrial d-loop. Analysis of molecular variance revealed a high degree of genetic variability, matrilinearily structured populations, male-biased gene flow, and lack of inbreeding. We did not identify any geographical or anthropogenic barriers to gene flow. Because extant populations are small and susceptible to random genetic drift, we recommend changes to current population management practices and periodic genetic surveys.

We studied the population dynamics of snowshoe hare in 2 areas of the southern boreal forest of eastern North America that differed with respect to furbearer harvest. We predicted that hare density and survival would be higher in the area with trapping (TRAP) than in the protected area (PROT) as a response to differential predator abundance. Indices of predator density suggested that predation risk was slightly higher in PROT than TRAP, particularly during the first year of our 3-year study. Predators killed 86% of the radiomarked hares (body mass >900 g) that died during the study (n = 71). Hare density was higher in TRAP (56 ± 4 hares/100 ha) than PROT (30 ± 5 hares/100 ha). Furthermore, hare densities were relatively stable in both areas between 1998 and 2000. Additional data from the previous 2 years in TRAP indicated that hare density had remained relatively unchanged for 5 years. These findings were in agreement with regional trends in hunter harvests which were low and stable throughout the 1990s. Annual survival of radiomarked hares (>900 g) was high in both areas but decreased during the study. Survival exhibited a sex*area interaction, and was lower for females in TRAP than PROT (46% vs. 76%) but higher for males in TRAP than PROT (64% vs. 37%); hare survival did not differ when sexes were combined. Differential survival of older hares could therefore not explain the difference in hare densities in the 2 areas. Gestation rates were similar in both areas, but the numbers of young hares (300–900 g) entering the population and young: female ratios were higher in TRAP than PROT. Survival of young hares during their first summer was variable between time periods, and tended to be higher in PROT than TRAP. Thus, it is likely that improved survival of juveniles (<300 g) soon after birth explains the increased hare density in TRAP. Study results show that managers cannot assume the existence of cyclicity in all hare populations, and must be aware that complex trophic interactions can generate unexpected responses to perturbations.

Blackbrush acacia (Acacia rigidula) and guajillo (Acacia berlandieri) are major browse species in diets of white-tailed deer (Odocoileus virginianus) in southern Texas and northern Mexico. We conducted both field and greenhouse experiments to test the hypotheses that (1) fertilization reduces tannins and spinescence in black-brush acacia and guajillo seedlings, (2) thorn removal increases utilization of shrub seedlings by deer, and (3) fertilization alters seedling nutrient content, chemical composition, and growth characteristics. Guajillo seedlings grown in a greenhouse with high soil fertility were lower (P < 0.05) in protein-precipitating tannin content relative to guajillo seedlings grown with low fertility. In the field, tannin content and spinescence of fertilized and unfertilized seedlings were similar (P > 0.05). Fertilization increased the levels of nitrogen and digestible protein in shrub seedlings of both species grown in the greenhouse (P < 0.05), but low rainfall during the field study may have precluded a similar response. Intensity of browsing was similar on fertilized and unfertilized seedlings. Seedlings with thorns removed were browsed more heavily than those with thorns (P < 0.05) in the field, where several species of herbivores were present. However, in feeding trials with tame deer, thorns did not affect browsing intensity. Thorns may be more important than tannins in defending seedlings against herbivory, and may be more important in defending against browsing by lagomorphs and small mammals than against browsing by white-tailed deer. Transplanted seedlings of both species should be protected from herbivores during early stages of establishment.