Eyes on the herd: Quantifying ungulate density from satellite, unmanned aerial systems, and GPScollar data.

Novel approaches to quantifying density and distributions could help biologists adaptively manage wildlife populations, particularly if methods are accurate, consistent, cost-effective, rapid, and sensitive to change. Such approaches may also improve research on interactions between density and processes of interest, such as disease transmission across multiple populations. We assess how satellite imagery, unmanned aerial system (UAS) imagery, and Global Positioning System (GPS) collar data vary in characterizing elk density, distribution, and count patterns across times with and without supplemental feeding at the National Elk Refuge (NER) in the US state of Wyoming. We also present the first comparison of satellite imagery data with traditional counts for ungulates in a temperate system. We further evaluate seven different aggregation metrics to identify the most consistent and sensitive metrics for comparing density and distribution across time and populations. All three data sources detected higher densities and aggregation locations of elk during supplemental feeding than non-feeding at the NER. Kernel density estimates (KDEs), KDE polygon areas, and the first quantile of interelk distances detected differences with the highest sensitivity and were most highly correlated across data sources. Both UAS and satellite imagery provide snapshots of density and distribution patterns of most animals in the area at lower cost than GPS collars. While satellite-based counts were lower than traditional counts, aggregation metrics matched those from UAS and GPS data sources when animals appeared in high contrast to the landscape, including brown elk against new snow in open areas. UAS counts of elk were similar to traditional ground-based counts on feed grounds and are the best data source for assessing changes in small spatial extents. Satellite, UAS, or GPS data can provide appropriate data for assessing density and changes in density from adaptive management actions. For the NER, where high elk densities are beneath controlled airspace, GPS collar data will be most useful for evaluating how management actions, including changes in the dates of supplemental feeding, influence elk density and aggregation across large spatial extents. Using consistent and sensitive measures of density may improve research on the drivers and effects of density within and across a wide range of species.

Predicting attenuation of salinized surface- and groundwater-resources from legacy energy development in the Prairie Pothole Region.

Oil and gas (energy) development in the Williston Basin, which partly underlies the Prairie Pothole Region in central North America, has helped meet U.S. energy demand for decades. Historical handling and disposal practices of saline wastewater co-produced during energy development resulted in salinization of surface and groundwater at numerous legacy energy sites. Thirty years of monitoring (1988-2018) at Goose Lake, which has been producing since the 1960s, documents long-term spatial and temporal changes in water quality from legacy energy development. Surface water quality was highly variable and decoupled from changes in groundwater quality, likely due to annual and regional climatic fluctuations. Therefore, changes in surface water-quality were not considered a reliable indicator of subsurface chloride migration. However, chloride concentrations in monitoring wells near wastewater sources exhibited systematic temporal reductions allowing for estimates of the time required for natural attenuation of groundwater to U.S. Environmental Protection Agency acute and chronic chloride toxicity benchmarks and a local background level. Point attenuation rates differed based on sediment type (outwash vs till) and yielded a range of predicted years when water-quality targets will be reached: acute - 2045 to 2113; chronic - 2069 to 2160; background - 2126 to 2275. Bulk attenuation rates from four separate years of data were used to calculate the distances chloride could migrate downgradient from the largest wastewater source. Potential distances of downgradient migration before dilution to water-quality targets decreased from 1989 to 2018: acute - 949 to 673 m; chronic - 1220 to 922 m; background - 1878 to 1525 m. Several downgradient wetlands are within these distances and will continue to receive saline contaminated groundwater for years. While these results demonstrate chloride attenuation at a legacy energy site, they also highlight the persistence of saline wastewater contamination and the need to mitigate future spills to prevent long-term salinization from energy development.

Effects of brine contamination from energy development on wetland macroinvertebrate community structure in the Prairie Pothole Region.

Wetlands in the Prairie Pothole Region (PPR) of North America support macroinvertebrate communities that are integral to local food webs and important to breeding waterfowl. Macroinvertebrates in PPR wetlands are primarily generalists and well adapted to within and among year changes in water permanence and salinity. The Williston Basin, a major source of U.S. energy production, underlies the southwest portion of the PPR. Development of oil and gas results in the coproduction of large volumes of highly saline, sodium chloride dominated water (brine) and the introduction of brine can alter wetland salinity. To assess potential effects of brine contamination on macroinvertebrate communities, 155 PPR wetlands spanning a range of hydroperiods and salinities were sampled between 2014 and 2016. Brine contamination was documented in 34 wetlands with contaminated wetlands having significantly higher chloride concentrations, specific conductance and percent dominant taxa, and significantly lower taxonomic richness, Shannon diversity, and Pielou evenness scores compared to uncontaminated wetlands. Non-metric multidimensional scaling found significant correlations between several water quality parameters and macroinvertebrate communities. Chloride concentration and specific conductance, which can be elevated in naturally saline wetlands, but are also associated with brine contamination, had the strongest correlations. Five wetland groups were identified from cluster analysis with many of the highly contaminated wetlands located in a single cluster. Low or moderately contaminated wetlands were distributed among the remaining clusters and had macroinvertebrate communities similar to uncontaminated wetlands. While aggregate changes in macroinvertebrate community structure were observed with brine contamination, systematic changes were not evident, likely due to the strong and potentially confounding influence of hydroperiod and natural salinity. Therefore, despite the observed negative response of macroinvertebrate communities to brine contamination, macroinvertebrate community structure alone is likely not the most sensitive indicator of brine contamination in PPR wetlands.

Oil and gas development footprint in the Piceance Basin, western Colorado.

Understanding long-term implications of energy development on ecosystem function requires establishing regional datasets to quantify past development and determine relationships to predict future development. The Piceance Basin in western Colorado has a history of energy production and development is expected to continue into the foreseeable future due to abundant natural gas resources. To facilitate analyses of regional energy development we digitized all well pads in the Colorado portion of the basin, determined the previous land cover of areas converted to well pads over three time periods (2002-2006, 2007-2011, and 2012-2016), and explored the relationship between number of wells per pad and pad area to model future development. We also calculated the area of pads constructed prior to 2002. Over 21million m2 has been converted to well pads with approximately 13million m2 converted since 2002. The largest land conversion since 2002 occurred in shrub/scrub (7.9million m2), evergreen (2.1million m2), and deciduous (1.3million m2) forest environments based on National Land Cover Database classifications. Operational practices have transitioned from single well pads to multi-well pads, increasing the average number of wells per pad from 2.5 prior to 2002, to 9.1 between 2012 and 2016. During the same time period the pad area per well has increased from 2030 m2 to 3504 m2. Kernel density estimation was used to model the relationship between the number of wells per pad and pad area, with these curves exhibiting a lognormal distribution. Therefore, either kernel density estimation or lognormal probability distributions may potentially be used to model land use requirements for future development. Digitized well pad locations in the Piceance Basin contribute to a growing body of spatial data on energy infrastructure and, coupled with study results, will facilitate future regional and national studies assessing the spatial and temporal effects of energy development on ecosystem function.

Evaluating the Effectiveness of Wildlife Detection and Observation Technologies at a Solar Power Tower Facility.

Solar power towers produce electrical energy from sunlight at an industrial scale. Little is known about the effects of this technology on flying animals and few methods exist for automatically detecting or observing wildlife at solar towers and other tall anthropogenic structures. Smoking objects are sometimes observed co-occurring with reflected, concentrated light ("solar flux") in the airspace around solar towers, but the identity and origins of such objects can be difficult to determine. In this observational pilot study at the world's largest solar tower facility, we assessed the efficacy of using radar, surveillance video, and insect trapping to detect and observe animals flying near the towers. During site visits in May and September 2014, we monitored the airspace surrounding towers and observed insects, birds, and bats under a variety of environmental and operational conditions. We detected and broadly differentiated animals or objects moving through the airspace generally using radar and near solar towers using several video imaging methods. Video revealed what appeared to be mostly small insects burning in the solar flux. Also, we occasionally detected birds flying in the solar flux but could not accurately identify birds to species or the types of insects and small objects composing the vast majority of smoking targets. Insect trapping on the ground was somewhat effective at sampling smaller insects around the tower, and presence and abundance of insects in the traps generally trended with radar and video observations. Traps did not tend to sample the larger insects we sometimes observed flying in the solar flux or found dead on the ground beneath the towers. Some of the methods we tested (e.g., video surveillance) could be further assessed and potentially used to automatically detect and observe flying animals in the vicinity of solar towers to advance understanding about their effects on wildlife.

Land cover changes associated with recent energy development in the Williston Basin; Northern Great Plains, USA.

The Williston Basin in the Northern Great Plains has experienced rapid energy development since 2000. To evaluate the land cover changes resulting from recent (2000-2015) development, the area and previous land cover of all well pads (pads) constructed during this time were determined, the amount of disturbed and reclaimed land adjacent to pads was estimated, land cover changes were analyzed over time for three different well types, and the effects from future development were predicted. The previous land cover of the 12,990ha converted to pads was predominately agricultural (49.5%) or prairie (47.4%) with lesser amounts of developed (2.3%), aquatic (0.5%), and forest (0.4%). Additionally, 12,121ha has likely been disturbed and reclaimed. The area required per gas well remained constant through time while the land required per oil well increased initially and then decreased as development first shifted from conventional to unconventional drilling and then to multi-bore pads. For non-oil-and-gas wells (i.e. stratigraphic test wells, water wells, and injection wells), the area per well increased through time likely due to increased produced water disposal requirements. Future land cover change is expected to be 2.7 times greater than recent development with much of the development occurring in five counties in the core Bakken development area. Direct land cover change and disturbance from recent and expected development are predicted to affect 0.4% of the landscape across the basin; however, in the core Bakken development area, 2.3% of the landscape will be affected including 2.1% of the remaining grassland. Although future development will result in significant land cover change, evolving industry practices and proactive siting decisions, such as development along energy corridors and placing pads in areas previously altered by human activity, have the potential to reduce the ecological effects of future energy development in the Williston Basin.

Presence and abundance of non-native plant species associated with recent energy development in the Williston Basin.

The Williston Basin, located in the Northern Great Plains, is experiencing rapid energy development with North Dakota and Montana being the epicenter of current and projected development in the USA. The average single-bore well pad is 5 acres with an estimated 58,485 wells in North Dakota alone. This landscape-level disturbance may provide a pathway for the establishment of non-native plants. To evaluate potential influences of energy development on the presence and abundance of non-native species, vegetation surveys were conducted at 30 oil well sites (14 ten-year-old and 16 five-year-old wells) and 14 control sites in native prairie environments across the Williston Basin. Non-native species richness and cover were recorded in four quadrats, located at equal distances, along four transects for a total of 16 quadrats per site. Non-natives were recorded at all 44 sites and ranged from 5 to 13 species, 7 to 15 species, and 2 to 8 species at the 10-year, 5-year, and control sites, respectively. Respective non-native cover ranged from 1 to 69, 16 to 76, and 2 to 82%. Total, forb, and graminoid non-native species richness and non-native forb cover were significantly greater at oil well sites compared to control sites. At oil well sites, non-native species richness and forb cover were significantly greater adjacent to the well pads and decreased with distance to values similar to control sites. Finally, non-native species whose presence and/or abundance were significantly greater at oil well sites relative to control sites were identified to aid management efforts.

Risk assessment of brine contamination to aquatic resources from energy development in glacial drift deposits: Williston Basin, USA.

Contamination to aquatic resources from co-produced water (brine) associated with energy development has been documented in the northeastern portion of the Williston Basin; an area mantled by glacial drift. The presence and magnitude of brine contamination can be determined using the contamination index (CI) value from water samples. Recently, the U.S. Geological Survey published a section (~2.59 km(2)) level risk assessment of brine contamination to aquatic resources for Sheridan County, Montana, using oilfield and hydrogeological parameters. Our goal was to improve the Sheridan County assessment (SCA) and evaluate the use of this new Williston Basin assessment (WBA) across 31 counties mantled by glacial drift in the Williston Basin. To determine if the WBA model improved the SCA model, results from both assessments were compared to CI values from 37 surface and groundwater samples collected to evaluate the SCA. The WBA (R(2)=0.65) outperformed the SCA (R(2)=0.52) indicating improved model performance. Applicability across the Williston Basin was evaluated by comparing WBA results to CI values from 123 surface water samples collected from 97 sections. Based on the WBA, the majority (83.5%) of sections lacked an oil well and had minimal risk. Sections with one or more oil wells comprised low (8.4%), moderate (6.5%), or high (1.7%) risk areas. The percentage of contaminated water samples, percentage of sections with at least one contaminated sample, and the average CI value of contaminated samples increased from low to high risk indicating applicability across the Williston Basin. Furthermore, the WBA performed better compared to only the contaminated samples (R(2)=0.62) versus all samples (R(2)=0.38). This demonstrates that the WBA was successful at identifying sections, but not individual aquatic resources, with an increased risk of contamination; therefore, WBA results can prioritize future sampling within areas of increased risk.

A GIS-based vulnerability assessment of brine contamination to aquatic resources from oil and gas development in eastern Sheridan County, Montana.

Water (brine) co-produced with oil in the Williston Basin is some of the most saline in the nation. The Prairie Pothole Region (PPR), characterized by glacial sediments and numerous wetlands, covers the northern and eastern portion of the Williston Basin. Sheridan County, Montana, lies within the PPR and has a documented history of brine contamination. Surface water and shallow groundwater in the PPR are saline and sulfate dominated while the deeper brines are much more saline and chloride dominated. A Contamination Index (CI), defined as the ratio of chloride concentration to specific conductance in a water sample, was developed by the Montana Bureau of Mines and Geology to delineate the magnitude of brine contamination in Sheridan County. Values >0.035 indicate contamination. Recently, the U.S. Geological Survey completed a county level geographic information system (GIS)-based vulnerability assessment of brine contamination to aquatic resources in the PPR of the Williston Basin based on the age and density of oil wells, number of wetlands, and stream length per county. To validate and better define this assessment, a similar approach was applied in eastern Sheridan County at a greater level of detail (the 2.59 km(2) Public Land Survey System section grid) and included surficial geology. Vulnerability assessment scores were calculated for the 780 modeled sections and these scores were divided into ten equal interval bins representing similar probabilities of contamination. Two surface water and two groundwater samples were collected from the section with the greatest acreage of Federal land in each bin. Nineteen of the forty water samples, and at least one water sample from seven of the ten selected sections, had CI values indicating contamination. Additionally, CI values generally increased with increasing vulnerability assessment score, with a stronger correlation for groundwater samples (R(2)=0.78) than surface water samples (R(2)=0.53).