Search

Sequoyah National Wildlife Refuge manages 2,200 acres of farmland annually. Twenty-five percent of the crops grown on this farmland is left as standing grain/green browse for wintering waterfowl. This protocol provides the methods used to quantitatively determine the productivity (yield/acre) of the unharvested corn on the Refuge. The primary purpose of this monitoring effort is to document the energetic potential of the corn yields that are left standing on the Refuge. This value can be used to calculate potential duck-use-days (DUDs) and can inform the Refuge on the quantity and type of agriculture needed to meet its waterfowl objectives.

Monitoring waterfowl numbers during primary migration and winter periods is essential to developing effective habitat conservation and management strategies. The relationship between available habitat and waterfowl numbers obtained from aerial survey transects is not well studied. To determine these relationships, multispectral SPOT-5 satellite imagery acquired for Sequoyah National Wildlife Refuge close to the time of waterfowl surveys was used to map habitat conditions. Robust Random Forest classification trees were used to model 16 land cover types using 416 reference locations collected in the field or derived from aerial photos close to or during waterfowl survey dates. The normalized difference vegetation index (NDVI), normalized difference water index (NDWI) and a simple ratio (SR) of red and near infrared bands were used to enhance classification accuracy for key habitat areas and abundance of water. Terrain variables such as slope, solar illumination and cosine transformed aspect derived from a digital elevation model (DEM) were also used to enhance habitat classification. Random Forest (RF) models were also compared to support vector machines (SVM) and cforest (CF) conditional inference trees. We used error matrices and the Kappa agreement statistic (K) to compare model results from each classifier. Results indicated that a tuned RF classifier showed better performance (K=0.73) than SVM (K=0.65) and unbiased cforest trees (K=0.63). Overall class agreement between similar RF and cforest models, designed to reduce predictor variable selection bias, was also relatively low (K=0.47). A final tuned RF model was selected resulting in 75% accuracy overall and was used to map habitat types for the refuge and surrounding landscape. We found that elevation and minimum noise fraction (MNF) bands were the most important predictor variables. MNF bands can help to reduce the number of correlated variables entering into a classification model when a larger number of correlated spectral bands are used. Similar forest types such as riverine, bottomland hardwood, and floodplain forest showed the greatest misclassification error. Overall, the RF model and SPOT-5 leaf-off imagery generated accurate land cover data for assessing habitat conditions during waterfowl surveys

A primary focus of Sequoyah National Wildlife Refuge (SNWR) is to provide habitat and foraging opportunities for waterfowl and other wetland dependent species. SNWR currently meets this objective by providing wetland and agricultural habitat. Wetland restoration and enhancement projects have occurred in recent years, with the intent to sustain healthier and more diverse waterfowl and wildlife populations. However, the relationship between habitat and food availability relative to waterfowl presence has yet to be quantified. This information is necessary to determine how SNWR can best manage refuge habitats for wintering waterfowl populations to ensure the available foraging resources on the refuge or within the surrounding landscape are sufficient to meet the energetic demands of waterfowl populations. Specifically, our goal was determine current carrying capacity for wintering waterfowl provided by SNWR relative to the surrounding landscape within the foraging flight distance of focal species: mallard, gadwall, and snow geese. Models included in this study emphasized efficient land use and maximization of resources within the SNWR refuge boundary in consideration of food resources available with the FFD of wintering waterfowl (e.g. Johnson et al. 2014). Specifically, we calculated the current energetic carrying capacity within Sequoyah and the surrounding FFD to determine to contributions of on-refuge agriculture relative to native wetland habitat types. We then determined the energetic effects of reductions in the area farmed on the refuge, changes to crop type and proportion (i.e., green browse versus standing grain) planted on the refuge, changes to the timing and proportion of standing crops mowed during the winter period, and hypothetical reductions in crop yield that could result from implementation of National Wildlife Refuge System policy eliminating the use of neonicotinoid coated and/or genetically modified seeds. To offset energetic losses associated with reductions or changes to on-refuge agricultural production, we determined the energetic gains associated with the creation of one passive and one actively managed wetland. Our model outcomes can be used to guide land management and restoration decisions and be incorporated into Habitat Management and Comprehensive Conservation Plans.

White-tailed deer (Odocoileus virginianus; hereafter, deer) are a common species found throughout North America. Because deer management occurs at a localized scale, reliable estimates of population sex and age ratios are necessary to set hunt quotas and track deer population trends over time. The most widely-used deer survey techniques have been criticized in recent years for being biased. One alternative is the use of infrared camera traps. This protocol was developed to assist Tishomingo National Wildlife Refuge conduct camera trap surveys using two analysis methods (Jacobson et al. 1997, Weckel et al. 2011) to estimate local deer populations. These surveys will provide baseline information and assist with deer harvest management decisions. The protocol calls for a camera density of 1 camera per 65 hectares (ha) and locations are selected from a grid approach. This survey is to be conducted for 14 days during the late summer and early fall on a biennial rotation. Initial startup costs are estimated at $14,209, and follow-up survey costs are estimated to be $10,014 including staff time. Surveys are expected to take approximately 0.10 FTE. Because photographic success can vary yearly and within camera locations, actual sex ratios and fawn to doe ratios have potential to be biased. Although camera surveys for white-tailed deer are not perfect, this protocol with included templates and application tools provide biologists and land managers confidence to conduct surveys and interpret results to effectively monitor long term trends and guide deer management at the local scale.

Management of wildlife, whether to mitigate damage, enhance safety, or effect conservation goals, requires identification of hazards posed by or to members of a particular species population or guild, and prioritization of management goals. We examined the special problem of managing birds to reduce hazards to aviation, particularly those species known to cause structural damage to aircraft when struck, as well as posing problems to airport facilities. Our objectives were to synthesize sampling theory and methods to provide airport biologists with 1) means to design and implement an avian survey on an airport that will maximize accuracy in quantifying avian hazards; 2) an understanding of bias and precision, and their influences on quantification of avian hazards; 3) suggestions on how to quantify avian hazards and use these data to estimate relative risk posed to aviation safety by a particular species or guild by time period and habitat type; and 4) knowledge of how data can be used to prioritize management goals. Our recommendations are intended to compliment U.S. Federal Aviation Administration procedures for Wildlife Hazard Assessments and subsequent management on airports. We stress the need for survey data to be ecologically relevant and accurate, such that management guidelines are defensible. However, we recognize that “real world” issues, such as regulatory, labor, and financial constraints, as well as the dynamics of airport environments, inevitably influence survey methods. Though we do not advocate use of naive count data in estimating relative abundance or habitat use, we recognize that animal observations obtained by airport biologists outside of a standardized sampling protocol are critical for identifying potential hazards to aviation safety. We recommend field testing of our suggestions and development of training materials for airport biologists that distill the information that we will present in light of constraints affecting survey design and conduct.