Search

Long-billed Curlews ( Numenius americanus ) are declining throughout North America, and the loss of grassland breeding habitat is one of the primary threats to the species. Intermountain West, in particular, has been identified as the most important region in North America for breeding curlews. Nevertheless, the density and abundance of Long-billed Curlews in this region is not well understood. Lands managed for military training can provide habitat for wildlife species of conservation concern, and increasingly these lands are becoming relevant to sustaining biodiversity. We conducted point count surveys of Long-billed Curlews on Department of Defense lands in the Columbia Basin near Boardman, Oregon, USA during two consecutive breeding seasons. We used multinomial-Poisson mixture models to estimate detection probability and density of curlews and to investigate environmental correlates of those metrics. Mean detection probability at a distance of 400 m was 0.45 and 0.61 in 2015 and 2016, respectively. In 2015, the clarity of skies increased detection probability, but in 2016, none of the variables we measured influenced detection probability. Mean predicted density was 3.3 (95% confidence interval: 2.4–4.7) and 1.8 (1.2–2.7) curlews/km² in 2015 and 2016, respectively. In both years, curlew density was higher in lower-elevation or topographically smoother areas. Estimated abundance of curlews in the study area was 639 (456–912) and 350 (237–520) birds in 2015 and 2016, respectively. The number of curlews appeared to fluctuate across the two years of our study, a demographic trend that may have been influenced by a wildfire in our study area in June 2015. The results of our study indicate that federal grasslands, including areas where military operations are conducted, can provide conservation benefit to breeding Long-billed Curlews.

Abstract Accurately predicting behavioural modes of animals in response to environmental features is important for ecology and conservation. Supervised learning (SL) methods are increasingly common in animal movement ecology for classifying behavioural modes. However, few examples exist of applying SL to classify polytomous animal behaviour from environmental features especially in the context of millions of animal observations. We review SL methods (weighted k‐nearest neighbours; neural nets; random forests; and boosted classification trees with XGBoost) for classifying polytomous animal behaviour from environmental predictors. We also describe tuning parameter selection and assessment strategies, approaches for visualizing relationships between predictors and class outputs, and computational considerations. We demonstrate these methods by predicting three categories of risk to bald eagles from colliding with wind turbines using, as predictors, 12 environmental state features associated with 1.7 million GPS telemetry data points from 57 eagles. Of the SL methods we considered, XGBoost yielded the most accurate model with 86.2% classification accuracy and pairwise‐averaged area under the ROC curve of 90.6. Computational time of XGBoost scaled better to large data than any other SL method. We also show how SHAP values integrated in the R package (xgboost) facilitate investigation of variable relationships and importance. For big data applications, XGBoost appears to provide superior classification accuracy and computational efficiency. Our results suggest XGBoost should be considered as an early modelling option in situations where the intent is to classify millions of animal behaviour observations from environmental predictors and to understand relationships between those predictors and movement behaviours. We also offer a tutorial to assist researchers in implementing this method.

Abstract Sophisticated animal‐borne sensor systems are increasingly providing novel insight into how animals behave and move. Despite their widespread use in ecology, the diversity and expanding quality and quantity of data they produce have created a need for robust analytical methods for biological interpretation. Machine learning tools are often used to meet this need. However, their relative effectiveness is not well known and, in the case of unsupervised tools, given that they do not use validation data, their accuracy can be difficult to assess. We evaluated the effectiveness of supervised (n = 6), semi‐supervised (n = 1), and unsupervised (n = 2) approaches to analyzing accelerometry data collected from critically endangered California condors (Gymnogyps californianus). Unsupervised K‐means and EM (expectation–maximization) clustering approaches performed poorly, with adequate classification accuracies of 0.81. Kappa statistics were also highest for RF and kNN, in most cases substantially greater than for other modeling approaches. Unsupervised modeling, which is commonly used for the classification of a priori‐defined behaviors in telemetry data, can provide useful information but likely is instead better suited to post hoc definition of generalized behavioral states. This work also shows the potential for substantial variation in classification accuracy among different machine learning approaches and among different metrics of accuracy. As such, when analyzing biotelemetry data, best practices appear to call for the evaluation of several machine learning techniques and several measures of accuracy for each dataset under consideration.

Abstract Addressing ongoing biodiversity loss requires collaboration between conservation scientists and practitioners. However, such collaboration has proved challenging. Despite the potential importance of tracking animal movements for conservation, reviews of the tracking literature have identified a gap between the academic discipline of movement ecology and its application to biodiversity conservation. Through structured conversations with movement ecologists and conservation practitioners, we aimed to understand whether the identified gap is also perceived in practice, and if so, what factors hamper collaboration and how these factors can be remediated. We found that both groups are motivated and willing to collaborate. However, because their motivations differ, there is potential for misunderstandings and miscommunications. In addition, external factors such as funder requirements, academic metrics, and journal scopes may limit the applicability of scientific results in a conservation setting. Potential solutions we identified included improved communication and better presentation of results, acknowledging each other's motivations and desired outputs, and adjustment of funder priorities. Addressing gaps between science and implementation can enhance collaboration and support conservation action to address the global biodiversity crisis more effectively.

Abstract There is increasing pressure on wind energy facilities to manage or mitigate for wildlife collisions. However, little information exists regarding spatial and temporal variation in collision rates, meaning that mitigation is most often a blanket prescription. To address this knowledge gap, we evaluated variation among turbines and months in an aspect of collision risk—probability of entry by an eagle into a rotor‐swept zone (hereafter, “probability of entry”). We examined 10,222 eagle flight paths identified and recorded by an automated bird monitoring system at a wind energy facility in Wyoming, USA. Probabilities of entry per turbine–month combination were 4.03 times greater in some months than others, ranging 0.15 to 0.62. The overall probability of entry for the riskiest turbine (i.e., the one with the greatest probability of entry) was 2.39 times greater than the least‐risky turbine. Our methodology describes large variation across turbines and months in the probability of entry. If subsequently combined with information on other sources of variation (i.e., weather, topography), this approach can identify risky versus safe situations for eagles under which cost of management, curtailment prescriptions, and collision risk can be simultaneously minimized.

Abstract Automated curtailment is potentially a powerful technique to reduce collision mortality of wildlife with wind turbines. Previously, we used a before–after–control–impact framework to demonstrate that eagle fatalities declined after automated curtailment was implemented with the IdentiFlight system at a wind power facility in Wyoming, USA. We received substantial interest and feedback regarding our study and, here, we implement several analytical suggestions and include more recent data that strengthen the inference we draw from our results. The five main analytical suggestions we received were to (1) exclude from analysis data that were collected during the period when automated curtailment was only partially implemented; (2) only analyse data from a single make and model of turbine; (3) evaluate changes in the rate of fatality, instead of the yearly numbers of fatalities that result from fluctuations around that rate; (4) calculate a standard measure determining effects of a treatment in a before–after–control–impact study and (5) examine yearly fluctuations of the fatality rate during the before period. After incorporating these suggestions and including additional data collected since the prior paper was published, our results confirm prior work. We demonstrate that eagle fatalities were reduced by 85% (95% highest density interval = 12%, 100%) after implementation of automated curtailment. Rate of fatalities declined by 2.85 eagles per year (−0.67, 5.70) between before and after periods at the treatment site and increased by 2.26 eagles per year (−1.77, 7.37) at the control site. Overall, the fatality rate declined by 4.91 (−0.27, 11.27) more eagles per year at the treatment site than at the control site. The probability that the fatality rate declined at the treatment site relative to the control site was 0.97. Our re‐analysis strengthens our inference by using more robust analyses and data to support the conclusions of the prior study suggesting that automated curtailment was effective at reducing eagle fatalities at our treatment site. Because of the site‐ and species‐specific nature of our work, future research should examine the efficacy of automated curtailment at other sites, with other species, and under different curtailment regimes.

Abstract A central theme for conservation is understanding how animals differentially use, and are affected by change in, the landscapes they inhabit. However, it has been challenging to develop conservation schemes for habitat‐specific behaviors. Here we use behavioral change point analysis to identify behavioral states of golden eagles (Aquila chrysaetos) in the Sonoran and Mojave Deserts of the southwestern United States, and we identify, for each behavioral state, conservation‐relevant habitat associations. We modeled behavior using 186,859 GPS points from 48 eagles and identified 2,851 distinct segments comprising four behavioral states. Altitude above ground level (AGL) best differentiated behavioral states, with two clusters of short‐distance movement behaviors characterized by low AGL (state 1 AGL = 14 m (median); state 2 AGL = 11 m) and two associated with longer‐distance movement behaviors and characterized by higher AGL (state 3 AGL = 108 m; state 4 AGL = 450 m). Behaviors such as perching and low‐altitude hunting were associated with short‐distance movements in updraft‐poor environments, at higher elevations, and over steeper and more north‐facing terrain. In contrast, medium‐distance movements such as hunting and transiting were over gentle and south‐facing slopes. Long‐distance transiting occurred over the desert habitats that generate the best updraft. This information can guide management of this species, and our approach provides a template for behavior‐specific habitat associations for other species of management concern.

Abstract Movements by animals can serve different functions and occur over a variety of spatial and temporal scales. Routine movement types, such as residency (localized movements) and migration, have been well studied. However, nonroutine movement types, such as dispersal, prospecting, and nomadism, are less well understood. Documenting these rarely detected events requires tracking large numbers of individuals across all age classes. We studied >500 golden eagles (Aquila chrysaetos) tracked by telemetry over a 10‐year period in western North America, of which 160 engaged in nonroutine, long‐distance (>300 km) movements. We identified spatial and temporal correlates of those movements at both small and large scales, and we quantified movement timing and direction. We further tested which age and sex classes of eagles were more likely to engage in these movements. Our analysis of 88,093 daily tracks suggested that distances traveled by eagles were responsive to the updraft potential of the spatial and temporal landscape they encountered. Tracks covered longer distances at locations and times of higher updraft potential, and older birds traveled farther than younger birds. By contrast, after decomposing daily tracks into 563 nonroutine, long‐distance movements measured at a multiday scale, only the duration of travel was responsive to environmental conditions encountered by eagles. Multiday trips that were longer were those initiated in open and warm landscapes and those that ended in mountainous regions. Finally, long‐distance movements were more frequently made in seasons other than winter, in north–south directions, and by young birds. We documented clear correlates of nonroutine, long‐distance movements by golden eagles at small, local scales but found little evidence of such correlates at larger, regional scales. Most long‐distance movements we documented fit patterns associated with traditional definitions of prospecting and nomadism but not migration. Our study is the first to describe these movement types by golden eagles, and as such provides a foundation for subsequent study into the movement ecology of other species.

Renewable energy production can kill individual birds, but little is known about how it affects avian populations. We assessed the vulnerability of populations for 23 priority bird species killed at wind and solar facilities in California, USA. Bayesian hierarchical models suggested that 48% of these species were vulnerable to population-level effects from added fatalities caused by renewables and other sources. Effects of renewables extended far beyond the location of energy production to impact bird populations in distant regions across continental migration networks. Populations of species associated with grasslands where turbines were located were most vulnerable to wind. Populations of nocturnal migrant species were most vulnerable to solar, despite not typically being associated with deserts where the solar facilities we evaluated were located. Our findings indicate that addressing declines of North American bird populations requires consideration of the effects of renewables and other anthropogenic threats on both nearby and distant populations of vulnerable species.

Abstract Recent advances in digital data collection have spurred accumulation of immense quantities of data that have potential to lead to remarkable ecological insight, but that also present analytic challenges. In the case of biologging data from birds, common analytical approaches to classifying movement behaviors are largely inappropriate for these massive data sets. We apply a framework for using K‐means clustering to classify bird behavior using points from short time interval GPS tracks. K‐means clustering is a well‐known and computationally efficient statistical tool that has been used in animal movement studies primarily for clustering segments of consecutive points. To illustrate the utility of our approach, we apply K‐means clustering to six focal variables derived from GPS data collected at 1–11 s intervals from free‐flying bald eagles (Haliaeetus leucocephalus) throughout the state of Iowa, USA. We illustrate how these data can be used to identify behaviors and life‐stage‐ and age‐related variation in behavior. After filtering for data quality, the K‐means algorithm identified four clusters in >2 million GPS telemetry data points. These four clusters corresponded to three movement states: ascending, flapping, and gliding flight; and one non‐moving state: perching. Mapping these states illustrated how they corresponded tightly to expectations derived from natural history observations; for example, long periods of ascending flight were often followed by long gliding descents, birds alternated between flapping and gliding flight. The K‐means clustering approach we applied is both an efficient and effective mechanism to classify and interpret short‐interval biologging data to understand movement behaviors. Furthermore, because it can apply to an abundance of very short, irregular, and high‐dimensional movement data, it provides insight into small‐scale variation in behavior that would not be possible with many other analytical approaches.

Energy infrastructure, particularly for wind power, is rapidly expanding in Africa, creating the potential for conflict with at-risk wildlife populations. Raptor populations are especially susceptible to negative impacts of fatalities from wind energy because individuals tend to be long-lived and reproduce slowly. A major determinant of risk of collision between flying birds and wind turbines is the altitude above ground at which a bird flies. We examine 18,710 observations of flying raptors recorded in southern Africa and we evaluate, for 49 species, the frequency with which they were observed to fly at the general height of a wind turbine rotor-swept zone (50–150 m). Threatened species, especially vultures, were more likely to be observed at turbine height than were other species, suggesting that these raptors are most likely to be affected by wind power development across southern Africa. Our results highlight that threatened raptor species, particularly vultures, might be especially impacted by expanded wind energy infrastructure across southern Africa.

Abstract A central challenge in applied ecology is understanding the effect of anthropogenic fatalities on wildlife populations and predicting which populations may be particularly vulnerable and in greatest need of management attention. We used three approaches to investigate the potential effects of fatalities from collisions with wind turbines on 14 raptor species for both current (106 GW) and anticipated future (241 GW) levels of installed wind energy capacity in the United States. Our goals were to identify species at relatively high vs low risk of experiencing population declines from turbine collisions and to also compare results generated from these approaches. Two of the approaches used a calculated turbine‐caused mortality rate to decrement population growth, where population trends were derived either from the North American Breeding Bird Survey or from a matrix model parameterized from literature‐derived demographic values. The third approach was potential biological removal, which estimates the number of fatalities that allow a population to reach and maintain its optimal sustainable population set by management objectives. Different results among the methods reveal substantial gaps in knowledge and uncertainty in both demographic parameters and species‐specific estimates of fatalities from wind turbines. Our results suggest that, of the 14 species studied, those with relatively higher potential of population‐level impacts from wind turbine collisions included barn owl, ferruginous hawk, golden eagle, American kestrel, and red‐tailed hawk. Burrowing owl, Cooper’s hawk, great horned owl, northern harrier, turkey vulture, and osprey had a relatively lower potential for population impacts, and results were not easily interpretable for merlin, prairie falcon, and Swainson’s hawk. Projections of current levels of fatalities to future wind energy scenarios at 241 GW of installed capacity suggest some species could experience population declines because of turbine collisions. Populations of those species may benefit from research to identify tools to prevent or reduce raptor collisions with wind turbines.

Abstract Illegal killing of nongame wildlife is a global yet poorly documented problem. The prevalence and ecological consequences of illegal killing are often underestimated or completely unknown. We review the practice of legal recreational shooting and present data gathered from telemetry, surveys, and observations on its association with illegal killing of wildlife (birds and snakes) within conservation areas in Idaho, USA. In total, 33% of telemetered long‐billed curlews (Numenius americanus) and 59% of other bird carcasses found with known cause of death (or 32% of total) were illegally shot. Analysis of spatial distributions of illegal and legal shooting is consistent with birds being shot illegally in the course of otherwise legal recreational shooting, but snakes being intentionally sought out and targeted elsewhere, in locations where they congregate. Preliminary public surveys indicate that most recreational shooters find abhorrent the practice of illegal killing of wildlife. Viewed through this lens, our data may imply only a small fraction of recreational shooters is responsible for this activity. This study highlights a poorly known conservation problem that could have broad implications for some species and populations of wildlife.

Oil and gas development in the Arctic Coastal Plain, Alaska, may pose threats to wildlife. Management guidelines within the National Petroleum Reserve-Alaska dictate buffer zones for coastal wildlife habitat and for breeding and foraging sites of yellow-billed loons (YBLOs; Gavia adamsii), a species of conservation concern. However, few data exist to assess the value of these buffer zones for YBLOs or for sympatrically breeding red-throated loons (RTLOs; G. stellata) and Pacific loons (PALOs; G. pacifica). We evaluated movements of these loons on the Arctic Coastal Plain to understand: 1) seasonal timing; 2) marine habitat use; and 3) spatial characteristics of breeding home ranges. We then used those findings to evaluate current management strategies in the context of loon biology. Compared to RTLOs and PALOs, YBLOs nested on larger lakes, stayed on them longer, had larger home ranges, but less frequently used the marine environment in summer. During autumn, YBLOs and PALOs were located farther from shore than during spring or summer, whereas RTLOs used the marine environment similarly during all three seasons. Management buffers established for YBLO nesting areas corresponded well with nest site space use of all three species, suggesting cross-taxa applicability of current management guidelines for nesting habitat. However, the majority of loon locations on the Arctic Coastal Plain were outside of coastal buffer zones established to protect fish and wildlife habitat within the National Petroleum Reserve-Alaska. These findings can inform assessment of conservation buffer zones to mitigate anthropogenic disturbance to loons in Alaska.

Abstract Human activity influences wildlife. However, the ecological and conservation significances of these influences are difficult to predict and depend on their population‐level consequences. This difficulty arises partly because of information gaps, and partly because the data on stressors are usually collected in a count‐based manner (e.g., number of dead animals) that is difficult to translate into rate‐based estimates important to infer population‐level consequences (e.g., changes in mortality or population growth rates). However, ongoing methodological developments can provide information to make this transition. Here, we synthesize tools from multiple fields of study to propose an overarching, spatially explicit framework to assess population‐level consequences of anthropogenic stressors on terrestrial wildlife. A key component of this process is using ecological information from affected animals to upscale from count‐based field data on individuals to rate‐based demographic inference. The five steps to this framework are (1) framing the problem to identify species, populations, and assessment parameters; (2) field‐based measurement of the effect of the stressor on individuals; (3) characterizing the location and size of the populations of interest; (4) demographic modeling for those populations; and (5) assessing the significance of stressor‐induced changes in demographic rates. The tools required for each of these steps are well developed, and some have been used in conjunction with each other, but the entire group has not previously been unified together as we do in this framework. We detail these steps and then illustrate their application for two species affected by different anthropogenic stressors. In our examples, we use stable hydrogen isotope data to infer a catchment area describing the geographic origins of affected individuals, as the basis to estimate population size for that area. These examples reveal unexpectedly greater potential risks from stressors for the more common and widely distributed species. This work illustrates key strengths of the framework but also important areas for subsequent theoretical and technical development to make it still more broadly applicable.

The use of animal-born sensors for location-based tracking and bio-logging in terrestrial systems has expanded dramatically in the past 10 years. This rapid expansion has generated new data on how animals interact with and respond to variation in their environment, resulting in important ecological, physiological, and evolutionary insights. Although understanding the finer details of animal locations has important management relevance, applied studies are not prominent in the movement ecology literature. This is despite the long history of applied studies of animal movement and the urgent and growing need for evidence-based conservation guidance, especially in the challenging field of human-wildlife interactions. The goal of this review is to evaluate the realized contribution of tracking-based animal movement ecology to solving specific conservation problems, and to identify barriers that may hinder expansion of that contribution. To do this, we (a) briefly review the history and technologies used in animal tracking and bio-logging, (b) use a series of literature searches to evaluate the frequency with which movement ecology studies are designed to solve specific conservation problems, and (c) use this information to identify challenges that may limit the applied relevance of the field of movement ecology, and to propose pathways to expand that applied relevance. Our literature review quantifies the limited extent to which research in the field of movement ecology is designed to solve specific conservation problems, but also the fact that such studies are slowly becoming more prevalent. We discuss how barriers that limit application of these principles are likely due to constraints imposed by the types of data used commonly in the field. Problems of scale mismatch, error compounding, and data paucity all create challenges that are relevant to the field of movement ecology but may be especially pertinent in applied situations. Finding solutions to these problems will create new opportunity for movement ecologists to contribute to conservation science.

Abstract Background Management requires a robust understanding of between- and within-species genetic variability, however such data are still lacking in many species. For example, although multiple population genetics studies of the peregrine falcon (Falco peregrinus) have been conducted, no similar studies have been done of the closely-related prairie falcon (F. mexicanus) and it is unclear how much genetic variation and population structure exists across the species’ range. Furthermore, the phylogenetic relationship of F. mexicanus relative to other falcon species is contested. We utilized a genomics approach (i.e., genome sequencing and assembly followed by single nucleotide polymorphism genotyping) to rapidly address these gaps in knowledge. Results We sequenced the genome of a single female prairie falcon and generated a 1.17 Gb (gigabases) draft genome assembly. We generated maximum likelihood phylogenetic trees using complete mitochondrial genomes as well as nuclear protein-coding genes. This process provided evidence that F. mexicanus is an outgroup to the clade that includes the peregrine falcon and members of the subgenus Hierofalco. We annotated > 16,000 genes and almost 600,000 high-quality single nucleotide polymorphisms (SNPs) in the nuclear genome, providing the raw material for a SNP assay design featuring > 140 gene-associated markers and a molecular-sexing marker. We subsequently genotyped ~ 100 individuals from California (including the San Francisco East Bay Area, Pinnacles National Park and the Mojave Desert) and Idaho (Snake River Birds of Prey National Conservation Area). We tested for population structure and found evidence that individuals sampled in California and Idaho represent a single panmictic population. Conclusions Our study illustrates how genomic resources can rapidly shed light on genetic variability in understudied species and resolve phylogenetic relationships. Furthermore, we found evidence of a single, randomly mating population of prairie falcons across our sampling locations. Prairie falcons are highly mobile and relatively rare long-distance dispersal events may promote gene flow throughout the range. As such, California’s prairie falcons might be managed as a single population, indicating that management actions undertaken to benefit the species at the local level have the potential to influence the species as a whole.