Survey methods for assessing land cover map accuracy

The increasing availability of digital photographic materials has fueled efforts by agencies and organizations to generate land cover maps for states, regions, and the United States as a whole. Regardless of the information sources and classification methods used, land cover maps are subject to numerous sources of error. In order to understand the quality of the information contained in these maps, it is desirable to generate statistically valid estimates of accuracy rates describing misclassification errors. We explored a full sample survey framework for creating accuracy assessment study designs that balance statistical and operational considerations in relation to study objectives for a regional assessment of GAP land cover maps. We focused not only on appropriate sample designs and estimation approaches, but on aspects of the data collection process, such as gaining cooperation of land owners and using pixel clusters as an observation unit. The approach was tested in a pilot study to assess the accuracy of Iowa GAP land cover maps. A stratified two-stage cluster sampling design addressed sample size requirements for land covers and the need for geographic spread while minimizing operational effort. Recruitment methods used for private land owners yielded high response rates, minimizing a source of nonresponse error. Collecting data for a 9-pixel cluster centered on the sampled pixel was simple to implement, and provided better information on rarer vegetation classes as well as substantial gains in precision relative to observing data at a single-pixel.     

Integrating habitat-masked range maps with quantifications of prevalence to estimate area of occupancy in IUCN assessments

Estimates of species geographic ranges constitute critical input for biodiversity assessments, including those for the International Union for the Conservation of Nature (IUCN) Red List of Threatened Species. Area of occupancy (AOO) is one metric that IUCN uses to quantify a species’ range, but data limitations typically lead to either under- or overestimates (and unnecessarily wide bounds of uncertainty). Fortunately, existing methods in which range maps and land-cover data are used to estimate the area currently holding habitat for a species can be extended to yield an unbiased range of plausible estimates for AOO. Doing so requires estimating the proportion of sites (currently containing habitat) that a species occupies within its range (i.e., prevalence). Multiplying a quantification of habitat area by prevalence yields an estimate of what the species inhabits (i.e., AOO). For species with intense sampling at many sites, presence–absence data sets or occupancy modeling allow calculation of prevalence. For other species, primary biodiversity data (records of a species’ presence at a point in space and time) from citizen-science initiatives and research collections of natural history museums and herbaria could be used. In such cases, estimates of sample prevalence should be corrected by dividing by the species’ detectability. To estimate detectability from these data sources, extensions of inventory-completeness analyses merit development. With investments to increase the quality and availability of online biodiversity data, consideration of prevalence should lead to tighter and more realistic bounds of AOO for many taxonomic groups and geographic regions. By leading to more realistic and representative characterizations of biodiversity, integrating maps of current habitat with estimates of prevalence should empower conservation practitioners and decision makers and thus guide actions and policy worldwide.     

A comparison of land-sharing and land-sparing strategies for plant richness conservation in agricultural landscapes

Strategies for conserving plant diversity in agroecosystems generally focus on either expanding land area in non-crop habitat or enhancing diversity within crop fields through changes in within-field management practices. In this study, we compare effects on landscape-scale species richness from such land-sharing or land-sparing strategies. We collected data in arable field, grassland, pasture, and forest habitat types (1.6 ha sampled per habitat type) across a 100-km2 region of farmland in Lancaster County, Pennsylvania, USA. We fitted species-area relationships (SARs) for each habitat type and then combined extrapolations from the curves with estimates of community overlap to estimate richness in a 314.5-ha landscape. We then modified these baseline estimates by adjusting parameters in the SAR models to compare potential effects of land-sharing and land-sparing conservation practices on landscape richness. We found that species richness of the habitat types showed a strong inverse relationship to the relative land area of each type in the region, with 89 species in arable fields (66.5% of total land area), 153 in pastures (6.7%), 196 in forests (5.2%), and 213 in grasslands (2.9%). Relative to the baseline scenario, major changes in the richness of arable fields produced gains in landscape-scale richness comparable to a conversion of 3.1% of arable field area into grassland habitat. Sensitivity analysis of our model indicated that relative gains from land sparing would be greatest in landscapes with a low amount of non-crop habitat in the baseline scenario, but that in more complex landscapes land sharing would provide greater gains. These results indicate that the majority of plant species in agroecosystems are found in small fragments of non-crop habitat and suggest that, especially in landscapes with little non-crop habitat, richness can be more readily conserved through land-sparing approaches.     

Modeling Opportunity Costs of Conservation in Transitional Landscapes

Abstract:  Conservation scientists recognize the urgency of incorporating opportunity costs into conservation planning. Despite this, applications to date have been limited, perhaps partly because of the difficulty in determining costs in regions with limited data on land prices and ownership. We present methods for estimating opportunity costs of land preservation in landscapes or ecoregions that are a changing mix of agriculture and natural habitat. Our approach derives from the literature on estimating land values as opportunity costs of alternate land uses and takes advantage of general availability of necessary data, even in relatively data-poor regions. The methods integrate probabilities of habitat conversion with region-wide estimates of economic benefits from agricultural land uses and estimate land values with a discount rate to convert annual values into net present values. We applied our method in a landscape undergoing agricultural conversion in Paraguay. Our model of opportunity costs predicted an independent data set of land values and was consistent with implicit discount rates of 15–25%. Model-generated land values were strongly correlated with actual land values even after correcting for the effect of property size and proportion of property that was forested. We used the model to produce a map of opportunity costs and to estimate the costs of conserving forest within two proposed corridors in the landscape. This method can be applied to conservation planning in situations where natural habitat is currently being converted to market-oriented land uses. Incorporating not only biological attributes but also socioeconomic data can help in the design of efficient networks of protected areas that represent biodiversity at minimum costs.     

Effect of land use and land cover change on soil erosion and the spatio-temporal variation in Liupan Mountain Region, southern Ningxia, China

The Liupan Mountains are located in the southern Ningxia Hui Autonomous Region of China, that forms an important divide between landforms and biogeographic regions. The populated part of the Liupan Mountain Region has suffered tremendous ecological damage over time due to population pressure, excessive demand and inappropriate use of agricultural land resources. To present the relationship between land use/cover change and spatio-temporal variation of soil erosion, data sets of land use between the late 1980s and 2000 were obtained from Landsat Thematic Mapper (TM) imagery, and spatial models were used to characterize landscape and soil erosion conditions. Also, soil erosion in response to land use and land cover change were quantified and analyzed using data from geographical information systems and remote sensing. Soil erosion by water was the dominant mode of soil loss, while soil erosion by wind was only present on a relatively small area. The degree of soil erosion was classified into five severity classes: slight, light, moderate, severe, and very severe. Soil erosion in the Liupan Mountain Region increased between the late 1980s and 2000, both in terms of acreage and severity. Moderate, severe, and very severe eroded areas accounted for 54.86% of the total land area. The lightly eroded area decreased, while the moderately eroded area increased by 368817 ha (22%) followed by severe erosion with 146552 ha (8.8%), and very severe erosion by 97067.6 ha (5.8%). Soil loss on sloping cropland increased with slope gradients. About 90% of the cropland was located on slopes less than 15°. Most of the increase in soil erosion on cropland was due to conversion of steep slopes to cropland and degradation of grassland and increased activities. Soil erosion was severe on grassland with a moderate or low grass cover and on dry land. Human activities, cultivation on steep slopes, and overgrazing of pastures were the main reasons for the increase in erosion severity.     

Economics of Grassland Conversion to Cropland in the Prairie Pothole Region

Abstract: Much of the remaining grassland, particularly in North America, is privately owned, and its conversion to cultivated cropland is largely driven by economics. An understanding of why landowners convert grassland to cropland could facilitate more effective design of grassland‐conservation programs. We built an empirical model of land‐use change in the Prairie Pothole Region (north‐central United States) to estimate the probability of grassland conversion to alternative agricultural land uses, including cultivated crops. Conversion was largely driven by landscape characteristics and the economic returns of alternative uses. Our estimate of the probability of grassland conversion to cultivated crops (1.33% on average from 1979 to 1997) was higher than past estimates (0.4%). Our model also predicted that grassland‐conversion probabilities will increase if agricultural commodity prices continue to follow the trends observed from 2001 to 2006 (0.93% probability of grassland conversion to cultivated crops in 2006 to 1.5% in 2011). Thus, nearly 121 million ha (30 million acres) of grassland could be converted by 2011. Conversion probabilities, however, are spatially heterogeneous (range 0.2% to 3%), depending on characteristics of a parcel (e.g., soil quality and economic returns). Grassland parcels with relatively high‐quality land for agricultural production are more likely to be converted to cultivated crops than lower‐quality parcels and are more responsive to changes in the economic returns on alternative agricultural land uses (i.e., conversion probability increases by a larger magnitude for high‐quality parcels when economics returns to alternative uses increase). Our results suggest that grassland conservation programs could be proactively targeted toward high‐risk parcels by anticipating changes in economic returns, such as could occur if a new biofuel processing plant were to be built in an area.     

Estimates of minimum patch size depend on the method of estimation and the condition of the habitat

Minimum patch size for a viable population can be estimated in several ways. The density-area method estimates minimum patch size as the smallest area in which no new individuals are encountered as one extends the arbitrary boundaries of a study area outward. The density-area method eliminates the assumption of no variation in density with size of habitat area that accompanies other methods, but it is untested in situations in which habitat loss has confined populations to small areas. We used a variant of the density area method to study the minimum patch size for the gopher tortoise (Gopherus polyphemus) in Florida, USA, where this keystone species is being confined to ever smaller habitat fragments. The variant was based on the premise that individuals within populations are likely to occur at unusually high densities when confined to small areas, and it estimated minimum patch size as the smallest area beyond which density plateaus. The data for our study came from detailed surveys of 38 populations of the tortoise. For all 38 populations, the areas occupied were determined empirically, and for 19 of them, duplicate surveys were undertaken about a decade apart. We found that a consistent inverse density area relationship was present over smaller areas. The minimum patch size estimated from the density-area relationship was at least 100 ha, which is substantially larger than previous estimates. The relative abundance of juveniles was inversely related to population density for sites with relatively poor habitat quality, indicating that the estimated minimum patch size could represent an extinction threshold. We concluded that a negative density area relationship may be an inevitable consequence of excessive habitat loss. We also concluded that any detrimental effects of an inverse density area relationship may be exacerbated by the deterioration in habitat quality that often accompanies habitat loss. Finally, we concluded that the value of any estimate of minimum patch size as a conservation tool is compromised by excessive habitat loss.     

Toward estimation of map accuracy without a probability test sample

The time and effort required of probability sampling for accuracy assessment of large-scale land cover maps often means that probability test samples are not collected. Yet, map usefulness is substantially reduced without reliable accuracy estimates. In this article, we introduce a method of estimating the accuracy of a classified map that does not utilize a test sample in the usual sense, but instead estimates the probability of correct classification for each map unit using only the classification rule and the map unit covariates. We argue that the method is an improvement over conventional estimators, though it does not eliminate the need for probability sampling. The method also provides a new and simple method of constructing accuracy maps. We illustrate some of problems associated with accuracy assessment of broad-scale land cover maps, and our method, with a set of nine Landsat Thematic Mapper satellite image-based land cover maps from Montana and Wyoming, USA.     

Ease vs. noise: Long-run changes in the value of transport (dis)amenities

For a complete cost-benefit analysis of durable infrastructures, it is important to understand how the value of non-market goods such as transit time and environmental quality changes as incomes rise in the long-run. We use difference-in-differences and spatial differencing to estimate the land price capitalization effects of metro rail in Berlin, Germany today and a century ago. Over this period, the negative implicit hedonic price of rail noise tripled. Our results imply income elasticities of the value of noise reduction and transport access of 2.2 and 1.4, substantially exceeding cross-sectional contingent valuation estimates.     

The importance of incorporating functional habitats into conservation planning for highly mobile species in dynamic systems

The distribution of mobile species in dynamic systems can vary greatly over time and space. Estimating their population size and geographic range can be problematic and affect the accuracy of conservation assessments. Scarce data on mobile species and the resources they need can also limit the type of analytical approaches available to derive such estimates. We quantified change in availability and use of key ecological resources required for breeding for a critically endangered nomadic habitat specialist, the Swift Parrot (Lathamus discolor). We compared estimates of occupied habitat derived from dynamic presence‐background (i.e., presence‐only data) climatic models with estimates derived from dynamic occupancy models that included a direct measure of food availability. We then compared estimates that incorporate fine‐resolution spatial data on the availability of key ecological resources (i.e., functional habitats) with more common approaches that focus on broader climatic suitability or vegetation cover (due to the absence of fine‐resolution data). The occupancy models produced significantly (P < 0.001) smaller (up to an order of magnitude) and more spatially discrete estimates of the total occupied area than climate‐based models. The spatial location and extent of the total area occupied with the occupancy models was highly variable between years (131 and 1498 km²). Estimates accounting for the area of functional habitats were significantly smaller (2–58% [SD 16]) than estimates based only on the total area occupied. An increase or decrease in the area of one functional habitat (foraging or nesting) did not necessarily correspond to an increase or decrease in the other. Thus, an increase in the extent of occupied area may not equate to improved habitat quality or function. We argue these patterns are typical for mobile resource specialists but often go unnoticed because of limited data over relevant spatial and temporal scales and lack of spatial data on the availability of key resources. Understanding changes in the relative availability of functional habitats is crucial to informing conservation planning and accurately assessing extinction risk for mobile resource specialists.     

Improving inferences about private land conservation by accounting for incomplete reporting

Private lands provide key habitat for imperiled species and are core components of function protectected area networks; yet, their incorporation into national and regional conservation planning has been challenging. Identifying locations where private landowners are likely to participate in conservation initiatives can help avoid conflict and clarify trade-offs between ecological benefits and sociopolitical costs. Empirical, spatially explicit assessment of the factors associated with conservation on private land is an emerging tool for identifying future conservation opportunities. However, most data on private land conservation are voluntarily reported and incomplete, which complicates these assessments. We used a novel application of occupancy models to analyze the occurrence of conservation easements on private land. We compared multiple formulations of occupancy models with a logistic regression model to predict the locations of conservation easements based on a spatially explicit social–ecological systems framework. We combined a simulation experiment with a case study of easement data in Idaho and Montana (United States) to illustrate the utility of the occupancy framework for modeling conservation on private land. Occupancy models that explicitly accounted for variation in reporting produced estimates of predictors that were substantially less biased than estimates produced by logistic regression under all simulated conditions. Occupancy models produced estimates for the 6 predictors we evaluated in our case study that were larger in magnitude, but less certain than those produced by logistic regression. These results suggest that occupancy models result in qualitatively different inferences regarding the effects of predictors on conservation easement occurrence than logistic regression and highlight the importance of integrating variable and incomplete reporting of participation in empirical analysis of conservation initiatives. Failure to do so can lead to emphasizing the wrong social, institutional, and environmental factors that enable conservation and underestimating conservation opportunities in landscapes where social norms or institutional constraints inhibit reporting.     

基于InVEST模型的三峡库区(重庆段)生境质量时空演变分析

为识别生境质量优劣的具体区域,揭示土地利用变化下生境质量时空演变特征,对集大农村、大库区于一体的三峡库区(重庆段)开展生境质量研究.以2000、2010及2020年土地利用数据为基础,运用InVEST模型对三峡库区(重庆段)2000—2020年生境质量进行定量评价,并结合转移矩阵方法挖掘了生境质量退化的具体区域.结果表...     

平原城市河流面源污染研究范围及方法的选择与效果比较

由于平原城市地貌起伏不大,较难划分城市河流的流域范围,使得与河道面源污染相关的研究范围的边界确定存在一定的困难。针对这一问题,本文对比了基于缓冲区和排水区的两种研究范围及方法。根据遥感影像,利用GIS提取研究区内土地利用信息并与河流水质进行回归分析,对比两种范围及方法所得出的结果,试图为城市河道污染物的研究及管理提供方法支持和理论依据。本文选取了天津滨海地区11条河流进行研究,结果表明：（1）选取城市排水区作为土地利用数据的方法能够与河流中污染物的发生,迁移的真实情况相符合,能够综合考虑不同的景观格局分布对于水质的影响,结合平原城市的特殊性反映土地利用与河流污染物之间的相关关系,在机理严谨性上优于使用缓冲区范围,在整体性上更为完整;（2）基于缓冲区的研究方法能够在一定程度上解释在地表径流作用下,河流近岸不同距离的土地利用类型对河流污染物的影响,如耕地对污染物的影响,河流近岸100 m缓冲区的联合显著性为0.848,300 m缓冲区的仅为0.165;（3）土地利用类型对河流中污染物类型的贡献和消减趋势基本一致,城市居民点与工矿用地是河流当中重金属污染物如铅、汞的主要输入源;耕地对重金属等污染物有一定的截留消减作用,但在面积较大时增加了河流的有机污染物如氨氮;林地、园地、草地和水体对于城市河流污染物有一定的净化作用。两种方法结合起来可以得出更为详尽、客观的研究结果。     

Improving multi-site benefit functions via Bayesian model averaging: A new approach to benefit transfer

A benefit function transfer obtains estimates of willingness-to-pay (WTP) for the evaluation of a given policy at a site by combining existing information from different study sites. This has the advantage that more efficient estimates are obtained, but it relies on the assumption that the heterogeneity between sites is appropriately captured in the benefit transfer model. A more expensive alternative to estimate WTP is to analyze only data from the policy site in question while ignoring information from other sites. We make use of the fact that these two choices can be viewed as a model selection problem and extend the set of models to allow for the hypothesis that the benefit function is only applicable to a subset of sites. We show how Bayesian model averaging (BMA) techniques can be used to optimally combine information from all models.The Bayesian algorithm searches for the set of sites that can form the basis for estimating a benefit function and reveals whether such information can be transferred to new sites for which only a small data set is available. We illustrate the method with a sample of 42 forests from U.K. and Ireland. We find that BMA benefit function transfer produces reliable estimates and can increase about 8 times the information content of a small sample when the forest is ‘poolable’.     

Composite likelihood approach to the regression analysis of spatial multivariate ordinal data and spatial compositional data with exact zero values

In many environmental and ecological studies, it is of interest to model compositional data. One approach is to consider positive random vectors that are subject to a unit-sum constraint. In landscape ecological studies, it is common that compositional data are also sampled in space with some elements of the composition absent at certain sampling sites. In this paper, we first propose a practical spatial multivariate ordered probit model for multivariate ordinal data, where the response variables can be viewed as the discretized non-negative compositions without the unit-sum constraint. We then propose a novel two-stage spatial mixture Dirichlet regression model. The first stage models the spatial dependence and the presence of exact zero values, and the second stage models all the non-zero compositional data. A maximum composite likelihood approach is developed for parameter estimation and inference in both the spatial multivariate ordered probit model and the two-stage spatial mixture Dirichlet regression model. The standard errors of the parameter estimates are computed by an estimate of the Godambe information matrix. A simulation study is conducted to evaluate the performance of the proposed models and methods. A land cover data example in landscape ecology further illustrates that accounting for spatial dependence can improve the accuracy in the prediction of presence/absence of different land covers as well as the magnitude of land cover compositions.     

Costs of expanding the network of protected areas as a response to climate change in the Cape Floristic Region

The expansion of protected areas is a critical component of strategies to promote the continued existence of biodiversity (i.e., life at all levels of biological organization) as climate changes, but scientific, social, and economic uncertainties associated with climate change are some of the major obstacles preventing such expansion. New models of climate change and species distribution and new methods of conservation planning now make it possible to explore the uncertainties associated with climate changes and species responses. Yet few reliable estimates of the costs of expanding protected areas and methods for determining these costs exist, largely because of the many (and uncertain) determinants of these costs. We developed a cost-accounting model to estimate the range in costs of various options for expanding protected areas and to explore the variables that drive these costs. Model development was informed by an existing plan to expand protected areas in the Cape Floristic Region of South Africa to address species conservation under a scenario of climate change. The 50-year present value of total costs varied from US$260 million ($1077/ha) for an off-reserve option that involves agreements with landowners and no compensation of forgone production and associated revenue to $1020 million ($4228/ha) for an on-reserve option that involves land acquisition and protection. The costs of acquiring land or compensating landowners for forgone production and development opportunities were the major drivers of the total costs across all options because most of the area identified in the protected-area expansion plan consisted of urban and high-quality agricultural lands. Total costs were also affected by changes in protected area extent and discount rate. Model-generated outputs such as these may be useful for informing implementation strategies and the allocation of future efforts in monitoring, data collection, and model development.     

Carbon sequestration of black locust forests in the Yellow River Delta region,China

The Yellow River Delta region in China is a land area of 1,200,000 ha with rich natural resources. Adverse environmental conditions, such as low rainfall and high salinity, promote the dominance of black locust trees for afforestation. With the increase of CO₂ in the atmosphere, this forest and others throughout the world have become valued for their ability to sequester and store carbon. Forests store carbon in aboveground biomass (i.e. trees), belowground biomass (i.e. roots), soils and standing litter crop (i.e. forest floor and coarse woody debris). There are well-developed methods to sample forest ecosystems, including tree inventories that are used to quantify carbon in aboveground tree biomass. Such inventories are used to estimate the types of roundwood products removed from the forest during harvesting. Based on standard plot inventories and stem analyses, carbon sequestration estimates of trees were 222.41 t ha^?1 for the Yellow River Delta region accounted for 67.12% of the whole forest. Similarly, carbon storage by herbaceous matter and soil was 0.50 and 50.34 t ha^?1, respectively. The results suggest that carbon sequestration in the forest ecosystem was performed by most of the forest, which plays an increasingly important role in sequestering carbon as the stand grows.     

Assessing effects of land use on landscape connectivity: loss and fragmentation of western U.S. forests

Effects of land-use change on the conservation of biodiversity have become a concern to conservation scientists and land managers, who have identified loss and fragmentation of natural areas as a high-priority issue. Despite urgent calls to inform national, regional, and state planning efforts, there remains a critical need to develop practical approaches to identify where important lands are for landscape connectivity (i.e., linkages), where land use constrains connectivity, and which linkages are most important to maintain network-wide connectivity extents. Our overall goal in this paper was to develop an approach that provides comprehensive, quantitative estimates of the effects of land-use change on landscape connectivity and illustrate its use on a broad, regional expanse of the western United States. We quantified loss of habitat and landscape connectivity for western forested systems due to land uses associated with residential development, roads, and highway traffic. We examined how these land-use changes likely increase the resistance to movement of forest species in non-forested land cover types and, therefore, reduce the connectivity among forested habitat patches. To do so, we applied a graph-theoretic approach that incorporates ecological aspects within a geographic representation of a network. We found that roughly one-quarter of the forested lands in the western United States were integral to a network of forested patches, though the lands outside of patches remain critical for habitat and overall connectivity. Using remotely sensed land cover data (ca. 2000), we found 1.7 million km2 of forested lands. We estimate that land uses associated with residential development, roads, and highway traffic have caused roughly a 4.5% loss in area (20 000 km2) of these forested patches, and continued expansion of residential land will likely reduce forested patches by another 1.2% by 2030. We also identify linkages among forest patches that are critical for landscape connectivity. Our approach can be readily modified to examine connectivity for other habitats/ecological systems and for other geographic areas, as well as to address more specific requirements for particular conservation planning applications.     

An evaluation of three statistical methods used to model resource selection

The performance of statistical methods for modeling resource selection by animals is difficult to evaluate with field data because true selection patterns are unknown. Simulated data based on a known probability distribution, though, can be used to evaluate statistical methods. Models should estimate true selection patterns if they are to be useful in analyzing and interpreting field data. We used simulation techniques to evaluate the effectiveness of three statistical methods used in modeling resource selection. We generated 25 use locations per animal and included 10, 20, 40, or 80 animals in samples of use locations. To simulate species of different mobility, we generated use locations at four levels according to a known probability distribution across DeSoto National Wildlife Refuge (DNWR) in eastern Nebraska and western Iowa, USA. We either generated 5 random locations per use location or 10,000 random locations (total) within 4 predetermined areas around use locations to determine how the definition of availability and the number of random locations affected results. We analyzed simulated data using discrete choice, logistic-regression, and a maximum entropy method (Maxent). We used a simple linear regression of estimated and known probability distributions and area under receiver operating characteristic curves (AUC) to evaluate the performance of each method. Each statistical method was affected differently by number of animals and random locations used in analyses, level at which selection of resources occurred, and area considered available. Discrete-choice modeling resulted in precise and accurate estimates of the true probability distribution when the area in which use locations were generated was ≥ the area defined to be available. Logistic-regression models were unbiased and precise when the area in which use locations were generated and the area defined to be available were the same size; the fit of these models improved with increased numbers of random locations. Maxent resulted in unbiased and precise estimates of the known probability distribution when the area in which use locations were generated was small (home-range level) and the area defined to be available was large (study area). Based on AUC analyses, all models estimated the selection distribution better than random chance. Results from AUC analyses, however, often contradicted results of the linear regression method used to evaluate model performance. Discrete-choice modeling was best able to estimate the known selection distribution in our study area regardless of sample size or number of random locations used in the analyses, but we recommend further studies using simulated data over different landscapes and different resource metrics to confirm our results. Our study offers an approach and guidance for others interested in assessing the utility of techniques for modeling resource selection in their study area.