Browsing by Author "Haukos, David A."
Now showing 1 - 20 of 22
- Results Per Page
- Sort Options
Item Amphibians of the southern Great Plains: Habitat and the impacts of temperature and pH on survival(2013-05) Kissner, Jessica; Griffis-Kyle, Kerry L.; McIntyre, Nancy E.; Haukos, David A.; Wallace, Mark C.Global climate change has had an increasing role in the decline of amphibian populations due to direct (e.g. critical temperature thresholds) and indirect (e.g. temperature stress, spreading of disease, and habitat alteration and loss) effects of increasing atmospheric temperature. Due to the complexity of habitats needed to complete all of their life stages, amphibians are subject to stressors through a number of systems (e.g., water quality, climatic characteristics, and terrestrial habitat). Furthermore, habitat currently used by amphibians may become inhospitable due to indirect stressing effects of increased temperatures projected for the Southern Great Plains (SGP). To protect and manage for species at risk, we must first understand their basic community assemblages, distributions, habitat associations, and tolerances. Because little amphibian research has come from northeastern New Mexico, collecting data on these parameters will be a useful tool for future research and management in this area. The objectives of this study were to: (1) examine amphibian species presence, richness, an index of reproductive success, and larval growth in northeastern New Mexico and relate them to site-specific and landscape-level habitat characteristics and (2) identify environmental factors that may jeopardize population persistence in the future. After conducting field work, I determined an additional objective based on the current average high temperature (28.3°C) and pH ranges (4.9 to 9.4) I recorded: (3) to determine the effect of a 5.6°C increase in average summer temperature by 2090 (to 33.9°C) due to climate change on embryo and tadpole survival, the effects of pH 4.9 and 9.4 on embryo and tadpole survival, and the effects of the interactions between pH and temperature on embryo and tadpole survival. Over 2011 and 2012, I sampled each playa wetland 5 times over a 17-day sampling period after precipitation of >2cm using call, visual encounter, egg mass, dip net, and funnel trap surveys to examine the amphibian community. I also collected tadpole measurements throughout development to examine larval growth patterns. Water quality was sampled during each survey, and ArcMap was employed to determine landscape-level habitat features. I sampled 14 playa wetlands in northeastern New Mexico, detecting 7 amphibian species with a maximum species richness per playa of 4. I detected reproduction in 4-6 species (as Pelobatid tadpoles are difficult to identify to species), and larval growth in Anaxyrus cognatus, Pseudacris clarkii, and Pelobatid spp. was exponential up to age 17 days (all p < 0.01). I also documented the first New Mexico state record of Pseudacris clarkii (spotted chorus frog). Although I did not document any statistically significant associations between amphibian species presence, richness, or reproductive success and habitat characteristics, I observed a trend of no P. clarkii reproduction at the acidic sites (4.9-6.9) even though adults were present in playa wetlands with pH 4.9-9.4. Based on my field observations about amphibian natural history from 2011, I hypothesized that increased temperature due to global climate change in 2090 may influence the ability of anurans to survive in pH-stressed playas. I tested these hypotheses in the lab in order to strengthen my field-based inferences. In 2012, I performed a fully-factorial laboratory experiment using the 2011 pH extremes (4.9 and 9.4), 2011 current average temperature (28.3°C), and 2090 projected temperature (33.9°C ), with embryos from 3 pairs of wild-caught P. clarkii. The results of my experiment suggest that P. clarkii cannot successfully reproduce in water with high temperature (33.9°C), regardless of the pH conditions. Furthermore, survival to metamorphosis in acidic conditions (pH 4.9), regardless of temperature, is unlikely, findings consistent with both my field observations and laboratory results. Survival to metamorphosis in basic conditions (pH 9.4) is very low (9%), compared to survival at pH 7 (84 %) that other studies have documented. Based on my findings, an increase in summer temperature of 5.6°C over the next 70 years will likely cause mortality and extirpation of some amphibian species in the Southern Great Plains. My results demonstrate the threat to P. clarkii due to projected climate change and suggest that other amphibian species may also be vulnerable, particularly for anuran populations that are at the edge of their range such as Gastrophryne olivaea, a New Mexico state-endangered species. By combining field and laboratory techniques and coming to a similar conclusion, I have been able to provide conclusions about the effect of temperature and pH on amphibian survival and population persistence.Item Behavioral effects of wind farms on wintering Sandhill Cranes (Grus canadensis) on the Texas High Plains(2011-12) Navarrete, Laura; Griffis-Kyle, Kerry L.; Haukos, David A.; Mulligan, KevinTexas has a superior annual capacity for wind power which has led to the erection of multiple wind farms across Texas with many more facilities planned. Wind energy is vital for a shift to carbon-emission free energy, however there has been relatively little research investigating the effects of wind farms as disturbance factors across the landscape. This project examines how wind energy infrastructure affects Sandhill Crane (Grus canadensis) behavior including landscape level habitat uses. Sandhill cranes are known to avoid areas of human activity and wind farms have been shown to render surrounding habitat of up to 1 km unsuitable to other species through direct effects (destroying habitat) and indirect effects on (avoidance). I examined the distribution of cranes at multiple wind farms on the Southern High Plains of Texas. I evaluated the effects wind farms have on roost occupancy, habitat use and crane behavior by comparing areas with wind turbines to those without for presence of cranes at roosting sites and behavior of cranes at foraging sites. I found that cranes within wind farm plots exhibited more vigilant behavior and less resting behavior. The distance to the nearest turbine contributed to 18% of the variation in the percent of cranes seen resting. Crane density and flock size are both negatively correlated with distance to turbines. Cranes in the vicinity of wind farms were found in smaller flocks, in larger habitat patches and closer to the road than cranes in control plots. Cranes also showed different habitat preferences within wind farms plots, utilizing non-foraging habitat which they avoided in control plots. Surveying playas in Texas using occupancy modeling methods resulted in no combination of variables explaining crane presence or absence in playas. To make effective management and conservation decisions, managers must be furnished with tools that help them understand large-scale ecological processes. Evaluating crane risk to wind farms based on behavioral characteristics can be used to predict areas of avoidance and help preserve important crane habitat in a rapidly developing landscape.Item Breading season demographics of a lesser prairie-chicken (Tympanuchus pallidicinctus) population in the northeastern Texas panhandle(2012-12) Holt, Doug; Boal, Clint W.; Gipson, Philip S.; McIntyre, Nancy E.; Haukos, David A.; DeYoung, RandallLesser prairie-chicken (Tympanuchus pallidicinctus) populations have declined across their range since 1900 and are a candidate for listing as Threatened under the Endangered Species Act. The historic lesser prairie-chicken range included all of the Texas panhandle. Currently, lesser prairie-chickens are found in 2 distinct populations in the northeastern and southwestern Texas panhandle. I conducted a 3-year study of breeding season demographics of lesser prairie-chickens on private properties in Gray and Hemphill counties (1 March-31 August, 2008-2010). I estimated male and female breeding bird survival, nest survival, and chick survival. I used demographic data collected during this study to conduct a population viability analysis and estimate time until extinction in the northeastern Texas panhandle. I assessed the efficacy of traditional road-based lek surveys to monitor lesser prairie-chicken populations in the Rolling Plains and High Plains ecoregions of Texas. Male survival differed between seasons and age-classes. Juvenile male survival was 1.00 (SE=0.00) during the lekking season and 0.88 (SE=0.62) during the nesting season. Adult male survival was 0.51 (SE=0.10) during the lekking season and 0.82 (SE=0.08) during the nesting season. Female survival did not differ with respect to season or age and was 0.55 (SE=0.13) for the entire breeding season. Nest survival was 0.36 (SE=0.05) during my study. Chick survival was lower between hatch and 14 days post hatch (0.18; SE=0.01) than 15-63 days post-hatch (0.55; SE=0.16). Results of a population viability analysis indicated low population growth rate (=0.44) and time to extinction was 3.5 years under baseline conditions based on field data. Management practices resulting in higher vital rates across the annual cycle resulted in higher population growth rates and longer times until extinction. The probability of detecting an active lek during road-based lek surveys was affected by wind speed, wind direction, and ecoregion. Current assumptions of detectability of active leks are likely too great and should be adjusted based on environmental conditions. Due to the drastic reduction in lesser prairie-chicken population across their range, listing as threatened is likely. As such, it is important to collect demographic data and establish management plans for species recovery. The results of analyses based on the data collected during my study suggest that the lesser prairie-chicken population in the northeastern Texas panhandle are on the brink of extripation.Item Community response to use of prescribed grazing and tebuthiuron herbicide for restoration of sand shinnery oak communities(2012-05) Zavaleta, Jennifer; Boal, Clint W.; Haukos, David A.; Cox, Robert D.; Perry, GadThe sand shinnery oak (Quercus havardii) mixed-grass community is an isolated, relict habitat located within short-grass prairie of the Southern High Plains. With the introduction of center-pivot agriculture, unmanaged grazing, oil and gas exploration and suppression of the natural fire regime, the vegetation composition of the shinnery oak community has changed during the past century. While some areas have become dominated by a monoculture of shinnery oak, the absolute amount of shinnery oak has been drastically reduced. This is particularly disconcerting not only because native biodiversity is lost, but also because this area is home to a number of species of conservation concern including the lesser prairie-chicken (Tympanuchus pallidicinctus), sand dunes lizard (Sceloporus arenicolus), and Cassin’s sparrow (Aimophia cassinii). Land managers have used herbicides (e.g., tebuthiuron) and a variety of grazing systems as tools to manage shinnery oak. However, very little research has been done to test how these tools can be used to restore an altered shinnery oak-grass community to pre-settlement standards of species composition. This thesis tests the community response in terms of abiotic factors; plant composition, structure and production; and mammal, herptile, and invertebrate community responses in terms of abundance, dominance, and diversity to a designed restoration effort from 2000-2011. The primary objectives were to (1) determine the variable response to tebuthiuron and grazing treatments, (2) assess the temporal response of the variables to the use of four treatment combinations over a 12-year period, and (3) compare resultant vegetation composition to historical standards. In Roosevelt Co, New Mexico, on the Southern High Plains, 532 ha of private land were treated with tebuthiuron (rate of 0.60 kg/ha with dune avoidance) in 2000. The state of New Mexico owned 518 ha of adjacent land, representing extant shinnery oak community (experimental control). This application rate was approximately 50% of previously recommended rates because the goal was to reduce shinnery oak to historical levels, not to eliminate it. A moderate grazing treatment was designed to take a maximum of 50% of the annual herbaceous production. To allow for inference beyond the study site, the experimental design was a combined completely randomized design with a systematic application of treatments following random assignment of initial treatment combination. The four treatments were treated/grazed, treated/not grazed, not treated/grazed, and not treated/not grazed. There were four replicates for the four treatments, totaling 16 plots. To remove the variation of year-to-year precipitation on the effect of tebuthiuron and grazing, I first ranked precipitation variables (based on two winter, growing and annual precipitation indices) and then used an analysis of variance with precipitation as a covariate. Plant and abiotic variables include soil moisture; line intercept data that was used for percent composition of shrubs, grasses, forbs, litter, bare ground, percent shinnery oak and sand bluestem (Andropogon gerardii) grass; Robel pole heights to measure visual obstruction; angle of obstruction to measure overhead obstruction; herbaceous production; and seed production for dropseed species (Sporobolus spp.), bluestems (Andropogon spp.), sand paspalum (Paspalum maritimum), and gramas (Bouteloua spp.). Small mammals and herptiles were trapped from May to September with 800 trap nights and 960 trap nights per plot per year from 2002-2010. Mammals were trapped with Sherman live traps and herptiles were trapped with pitfall and funnel traps. Abundance, dominance, diversity and the three most abundant mammals - kangaroo rat (Dipidomis ordii), spotted ground squirrel (Spermophilus spilosoma), and pocket mouse species (Perognathus spp.) - and herptiles (prairie lizard (Sceloporus lecontei), Great Plains skink (Eumeces obsoletus), and coachwhip (Masticophis flagellum)) were variables used in analyses. Invertebrates were collected in April and June with a terrestrial vacuum sampler. Invertebrate abundance, biomass, and number of taxonomic families as well as biomass of the three most abundant (grasshoppers (Acrididae), treehoppers (Acrididae), and caterpillars (Lepidoptera larvae) were variables used in analysis. My results show that at relatively low levels of tebuthiuron (0.60 kg/ha) and subsequent moderate grazing system, sand shinnery oak can be reduced and maintained at near historical levels without reapplying tebuthiuron because the tested management approach allowed grasses to remain competitive in the system. Over the ten years, there was 91% less shinnery oak in untreated areas. The removal of shinnery oak made environmental soil moisture more available for grasses and forbs to germinate and grow. Indeed, grasses increased by 149% and forbs increased by 257% in treated areas as compared to untreated areas throughout the study period. In terms of visual obstruction, there was both an herbicide and grazing effect in April such that visual obstruction increased by 30% in treated areas as compared to untreated and decreased by 6.5% in grazed areas as compared to non-grazed areas. Similarly, there was an herbicide and grazing interaction effect such that treated areas had a 14% decrease in overhead obstruction in grazed areas and non-treated areas had only 10% less in grazed areas as compared to non-grazed areas. These changes in plant composition and structure increase biodiversity such that there are more available niches to fill. The results from mammal, herptile, and invertebrate data indicated that species react on an individual basis to herbicide and grazing combinations, so the treatments yielded mixed results. I found no significant herbicide effect of overall abundance of small mammals. However, there was a significant grazing effect such that there was 23% more abundance of small mammals in grazed areas as compared to non-grazed areas, which was likely driven by kangaroo rats. In terms of herptile abundance, there was an interaction effect such that more lizards were found in treated/non-grazed and not treated/grazed areas. This mixed result indicates that species act individually in response to herbicide and grazing. Invertebrates, for the most part, responded positively to herbicide treatment and negligibly in terms of grazing, presumably due to increases in forbs. Areas that were treated with tebuthiuron and had moderate grazing statistically reached historical standards only during one year, but showed trends that were comparable to historical standards throughout the study compared to other treatment combinations. The largest difference between treated areas and historical standards was that treated areas had more forbs. This may not necessarily be interpreted as a bad thing because increase in invertebrates due to forb presence indicates good habitat quality and increases food sources for animals higher on the food chain. With low rates of tebuthiuron (0.60 kg/ha) and moderate levels of grazing, shinnery oak communities can approach historical standards and provide the most plant composition diversity. With moderate levels of grazing, shinnery oak will not need to be re-treated as grasses remain competitive in the system. With the removal of shinnery oak, there is increased soil moisture such that grasses and forbs can establish. The change from a shrub monoculture to a mixed-grass prairie changes the plant composition and structure and provides more niches for invertebrates, mammals and herptiles to fill. Treatment with tebuthiuron at low doses is the first step in managing shinnery oak communities; however, given the incredible, absolute loss of shinnery oak managers should be cautious about treating all shinnery-oak dominated areas.Item Design with innovation: Transforming an urban playa stormwater site(2008-12) Glancy, Candi Lu; Klein, Charles H.; Haukos, David A.; Billing, John C.Stormwater management in the Playa Lakes Region of West Texas historically requires that playa wetlands in city limits be modified to collect runoff from seasonal storms. These urban playas often serve a second function, that of an open space park. Design of these parks frequently lacks innovation in aesthetics and stormwater cleansing functions. Creativity in stormwater solutions extends beyond the typical functional-only design. In cities such as Renton, Washington, Minneapolis, Minnesota, and Gilbert, Arizona, stormwater and wastewater have been innovatively incorporated into the urban culture. This thesis explores an innovative solution for the Playa Lakes Region of West Texas that goes beyond what was discovered in the literature review. Design criteria are developed to transform Lake 87, a Lubbock, Texas urban stormwater management site, into an aesthetically pleasing, artistic, innovative site that integrates the history and nature of the area through native plantings and sculpture without compromising its primary function. The design also resolves increasing erosion through innovative best management practices (BMP) and landforming. This thesis investigates playa wetlands (Lake 87’s origin), urban stormwater function and management practices (Lake 87’s role in the community), aesthetic and artistic design solution case studies, and the role of art in society (Lake 87’s potential). The innovative design solution for Lake 87 has a complexity that appeals to the senses so that different interpretations are possible. The design attempts to engage the participant by becoming part of the harmony and contrast that is present in nature. It is an exploration of an alternative in stormwater management.Item Effects of land use on sediment deposition in playas(1994-05) Lou, Hong-Ren; Smith, Loren M.; Allen, Bonnie L.; Haukos, David A.Item Environmental availability and potential influences of heavy metals on snowy plovers in the southern great plains of Texas, New Mexico, and Oklahoma(2016-08) Ashbaugh, Hannah M.; Conway, Warren C.; Collins, Daniel P.; Comer, Christopher E.; Haukos, David A.Interior snowy plover (Charadrius nivosus) populations are declining due to deteriorating habitat quality, poor juvenile survival, and declining nest success. Snowy plovers are philopatric to saline lake and alkali flat habitats of the Southern Great Plains (SGP) of Texas, Oklahoma, and New Mexico during the breeding season. Artesian spring flow and surface water availability have declined since groundwater mining became widespread in one of the most agriculturally intensified regions in North America; and aquifer withdrawals continue to outpace recharge. Furthermore, as natural artesian freshwater springs associated with most saline lakes dramatically declined, precipitation runoff has become an important contributor to hydroperiod duration and spatial extent, such that saline lake habitats are becoming less stable, reliable, and predictable for regional breeding snowy plover populations. Snowy plover exposure to potential environmental contaminants remains unknown, but may be partially linked to declining populations and deteriorating habitat conditions throughout much of the SGP during the breeding season. Currently, few trace metal concentration, or potential environmental availability data are available for these habitats. The objectives of this research were to quantify V, As, Cd, Pb, and Se within (a) snowy plover blood, feathers, and potential prey (tiger beetles [Cicindela circumpicta and Cicindela togata]), and (b) sediment (active and inactive springs), and water (flowing, standing, and run-off) in snowy plover nesting habitats (saline lakes and alkali flats) of the SGP. Concentrations of Se within sampled biota were (a) most frequently detected as compared to any other heavy metal and (b) frequently quantified at levels exceeding either background or toxicity thresholds. Of quantifiable samples, 98% of snowy plover blood and 22% of feather data were above Se toxicity thresholds for both matrices. Almost all tiger beetle Se concentrations were below reported invertebrate thresholds, but Se concentrations ranged up to nearly 50 ppm. Snowy plover blood Se concentrations were related to tiger beetle Se concentrations, sex, and individual body condition, while primary feather 5 (P5) Se concentrations were related to state, sex, and presence of molting (of other feathers), and tiger beetle Se concentrations were positively related to being collected in Texas saline lakes. Most (> 80%) sediment samples had detectable concentrations of V, As, Cd, and Pb, while Se was detected in only 20% of combined sediment samples. Metal concentrations were detectable in water samples less frequently and varied by spring flow conditions; although, no metal concentration exceeded 0.4 ppm. However, regression models indicated that Se concentrations in study site sediments were positively related to being collected near active artesian springs (mostly in Texas saline lakes), and water Se concentrations were negatively correlated with run-off water. Although neither sediment nor water Se concentrations were quantified at known levels of concern for avifauna, nesting snowy plovers are clearly at risk for Se exposure, particularly in Texas saline lakes, as regional snowy plover occupancy, presence, and nesting success is linked to presence of surface water, near active artesian springs – precisely the locations within these habitats that provide the greatest risk of exposure. Similar to most other shorebird toxicity data, few data on regional contaminants exist; and this research provides some of the first estimates of trace metal exposure risks and identifies potential trophic pathways by which snowy plovers may be chronically exposed to Se (and to a lesser degree As). Future efforts will need to focus upon identifying specific uptake pathways during the breeding season, as well as Se exposure risks during the non-breeding season, as snowy plovers breeding in the SGP appear to be potentially exposed to Se throughout the annual cycle.Item Influence of climate change and land use on Lesser Prairie-Chicken (Tympanuchus pallidicinctus) populations persistence in the Sand Sagebrush and Short-Grass Prairies(2016-08-11) Godar, Alixandra J.; Grisham, Blake A.; Haukos, David A.; Boal, Clint W.; Kahl, Samantha S.Observed long-term, range wide declines in Lesser Prairie-Chicken (Tympanuchus pallidicinctus) numbers are primarily attributed with decreases in habitat quality and quantity in the Great Plains, with weather driving short term trends. The influence of land cover patterns and anthropogenic structures on lek attendance and the influence of weather on Lesser Prairie-Chicken vital rates remain unexplored in the Sand Sagebrush and Short-Grass Prairies in the northern extent of the species’ range. I assessed the influence of weather on survival rates by using the known-fate model and nest survival model in Program MARK to assess effect sizes of biologically relevant weather variables on adult survival and nest and brood survival, respectively, for each ecoregion. To assess the relationship between land cover and lek attendance, I used an anthropogenic dispersion index to evaluate the influence of structures within 4.8 km of the lek on lek attendance using a generalized linear model. Finally, I developed an integrated population model (IPM) to assess the effect of contemporary weather patterns on lambda (rate of population change), female survival, nest survival and chicks fledged to facilitate the development of forecasting long-term population trends, given climate change forecasts for the Great Plains. In my assessments of the influence of weather on survival rates, female survival in the Short-Grass Prairie was low (0.41, SE = 0.07; 95% CI = 0.29, 0.56) but small effect sizes indicate survival was not driven by weather. The nest survival model of hot days during incubation + cold days during incubation + total precipitation during incubation had the most support in the Short-Grass (AICC = 498.27, AICC wi = 0.99) and Sand Sagebrush ecoregions (AICC = 889.44, AICC wi = 0.99). My results of the influence of land use on lek persistence, suggest that lek attendance is affected by anthropogenic structure dispersion in the Sand Sagebrush Prairie Ecoregion but not in the Short-Grass Prairie Ecoregion. Significant parameters in the Sand Sagebrush Prairie Ecoregion include well dispersion (0.43, SE = 0.18; 95% CI = 0.06, 0.79) and transmission line dispersion (-1.93 SE = 0.22; 95% CI = -2.37, -1.49). In addition, there was a small negative effect of shrub patch size on lek attendance (-0.64, SE = 0.27; 95% CI = -1.17, -0.11) in the Sand Sagebrush Prairie Ecoregion while all confidence intervals overlapped zero in the Short-Grass Prairie Ecoregion. I speculate that responses varied due to the variation in landscape characteristics between ecoregions with the large tracts of grasslands in the Short-Grass Prairie Ecoregion and the highly developed Sand Sagebrush Prairie Ecoregion. The IPMs predicted either exponential growth (Short-Grass) or extinction (Sand Sagebrush). When weather projections varied from the current range of conditions, rates of change quickly became illogical in the Short-Grass Prairie. The IPM requires additional information to improve performance, either through additional parameters (Short-Grass) or increased sample size (Sand Sagebrush). My results suggest that weather is not the primary driver of vital rates and there is support for the influence of landscape metrics on Lesser Prairie-Chicken lek attendance in Sand Sagebrush Prairie Ecoregion, but not the Short-Grass Prairie Ecoregion. In the Sand Sagebrush Prairie Ecoregion IPM, landscape metrics should be incorporated while the Short-Grass Prairie Ecoregion requires additional research to determine which parameters should be incorporated into the model.Item Interactive effects of severe drought and grazing on the life history cycle of a bioindicator species(2018) Fritts, Sarah Rebecah; Grisham, Blake A. (TTU); Cox, Robert D. (TTU); Boal, Clint W. (TTU); Haukos, David A.; McDaniel, Patricia; Hagen, Christian A.; Greene, Daniel U.We used the lesser prairie-chicken (Tympanuchus pallidicinctus), an iconic grouse species that exhibits a boom–bust life history strategy, on the Southern High Plains, USA, as a bioindicator of main and interactive effects of severe drought and grazing. This region experienced the worst drought on record in 2011. We surveyed lesser prairie-chicken leks (i.e., communal breeding grounds) across 12 years that represented 7 years before the 2011 drought (predrought) and 4 years during and following the 2011 drought (postdrought). Grazing was annually managed with the objective of achieving ≤50% utilization of aboveground vegetation biomass. We used lek (n = 49) count data and covariates of weather and managed grazing to: (a) estimate long-term lesser prairie-chicken abundance and compare abundance predrought and postdrought; (b) examine the influence of annual and seasonal drought (modified Palmer drought index), temperature, and precipitation on long-term lesser prairie-chicken survival and recruitment; and (c) assess and compare the influence of grazing on lesser prairie-chicken population predrought and postdrought. Lesser prairie-chicken abundance was nearly seven times greater predrought than postdrought, and population declines were attributed to decreased survival and recruitment. The number of days with temperature >90th percentile had the greatest effect, particularly on recruitment. The population exhibited a substantial bust during 2011 and 2012 without a boom to recover in four postdrought years. Adaptive grazing positively influenced the population predrought, but had no effects postdrought. Results suggest that the severe drought in 2011 may have been beyond the range of environmental conditions to which lesser prairie-chickens, and likely other species, have adapted. Land management practices, such as grazing, should remain adaptive to ensure potential negative influences to all species are avoided. Increasing habitat quantity and quality by reducing habitat loss and fragmentation likely will increase resiliency of the ecosystem and individual species.Item Lesser prairie-chicken habitat selection across varying land use practices in Eastern New Mexico and West Texas(2013-08) Borsdorf, Philip K. K.; Boal, Clint W.; Haukos, David A.; Wallace, Mark C.; Boydston, KathyThe primary focus of this thesis was to assess gender-specific lesser prairie-chicken breeding season home range size and habitat selection. Data were collected on two different study areas occurring in sand-shinnery oak grassland communities on the Southern High Plains. From 2008-2012, on a study area occurring in west Texas (Chapter 2), male lesser prairie-chickens demonstrated selection for vegetative cover-types comprised of sand-shinnery oak, but dominated by native prairie-grasses, while females used the cover-types in proportion to availability. On experimentally tebuthiuron treated (0.6 kg/ha) and rotationally grazed plots in eastern New Mexico (Chapter 3), results suggested selection by male lesser prairie-chickens for areas not treated with tebuthiuron in comparison to tebuthiuron treated areas. Female lesser prairie-chickens, in New Mexico, used the not-treated and not-grazed areas in greater proportion to availability, the treated and not-grazed, and not-treated and grazed areas in proportion to availability, and the treated and grazed areas less than were proportionally available. Gender-specific home range size was assessed for lesser prairie-chickens from 2008-2012 on the west Texas study area (Chapter 2). Three different home range estimators were used; 95% fixed kernel density estimates using two different bandwidths (LSCV and Plug-in) and a 100% Minimum Convex Polygon were implemented to estimate home range sizes. For male lesser prairie-chickens, both 95% fixed kernel density estimates (LSCV and Plug-in) were significantly larger than 100% MCP home ranges. For female lesser prairie-chickens, only the 95% fixed kernel density (LSCV) estimates were significantly larger than 100% MCP home ranges. Despite the home range estimator, all female home ranges were approximately twice the size of male home ranges. Nocturnal roosting site selection was assessed in relation to microhabitat and microclimate in west Texas during the breeding seasons of 2011 and 2012 (Chapter 4). There were no differences in visual obstruction readings between roost and paired random sites. Lesser prairie-chickens roosted directly on bareground with no overhead vegetative coverage. In 2011, there was significantly more bareground and less humidity (% relative humidity) at roosting site point center compared to paired random sites. In 2012, roost and random sites were equally humid, but there was more litter coverage surrounding roosting sites compared to paired random sites. Temperature did not appear to differ between roost and random sites in either 2011 or 2012. Results may show evidence for the selection of nocturnal roosting sites based on tradeoffs between visual and/or scent concealment from nocturnal predators as related to the available microclimate and microhabitat variables during the course of this study.Item Local and landscape factor influences on avian community composition in playas of the Southern High Plains(Texas Tech University, 2007-12) Tsai, Jo-Szu; Smith, Loren M.; Haukos, David A.; Martin, Clyde F.; McIntyre, Nancy E.; McMurry, Scott T.The more than 25,000 playas on the Southern High Plains (SHP) are essential to the maintenance of regional biodiversity. The SHP landscape has been highly fragmented by intensive cultivation since the 1930s. Cultivation has degraded the condition of playas through sedimentation, resulting in a loss of playa volume in cropland watersheds and shortened hydroperiod. These effects should influence the avian communities that depend on playas for breeding, overwintering, or migratory stopover sites. However, information about the influence of land-use change and other factors on avian community composition in SHP playas is limited. Previous studies have focused on the relationship between game species and local variables during winter; however, systematic documentation of playa variables that are important to birds throughout the year is lacking. To fully understand the importance of playas to different avian communities, information on avian use during the wet and dry phases of playas throughout the year is needed. Therefore, I examined influences of local and landscape factors on avian species richness, diversity, and density in playas and determined how avian community composition was influenced by water-level fluctuation as a disturbance. My objectives were to 1) determine how local variables within a playa (e.g., vegetation cover and playa area) influence richness, density, and diversity of avian communities; 2) evaluate the effect of landscape variables in the wetland complex (i.e., playas and the adjoining habitat patches in 3 defined diameters) on richness, density, and diversity of avian communities; 3) examine the effect of water-level fluctuations as a disturbance on avian community composition; and 4) examine relationships between local variables and percent composition of different avian groups and dominant species during summer. After significant precipitation events in June 2003 and 2004, I selected 80 playas (40 playas each year) across the SHP containing water with watersheds ranging from 100% native grassland to 100% cropland. Wet playas were surveyed biweekly by counting all birds present within the playa boundary for as long as playas retained water, from June 2003 to May 2004 and June 2004 to May 2005. When playas dried, I conducted surveys monthly using distance sampling techniques. Water depth was measured concurrently with each bird survey and percent vegetation cover, vegetation structure, and plant species richness were measured 3 times during each summer. Eight landscape variables (i.e., number of playas, percent cover of playas, interspersion and juxtaposition index, landscape shape index, Shannon evenness index, diversity index of land use, edge density, and mean edge density) were obtained using ESRI® ArcGIS and FRAGSTATS*ARC® for each study playa at 3 distance scales (i.e., 1, 5, and 10 km radii from the study playas, representing a gradient of local to landscape-level influences). I recorded 127 bird species during 1487 surveys over 2 years. Seventy-seven were wetland-dependent species and 50 were non-wetland species. In wet playas, area was a positive predictor for total species richness and diversity in all seasons. Water depth was a consistent predictor across different seasons for total species richness, and species richness was greatest at an intermediate level of water depth (ranging from 40 to 100 cm). Tilled index ([tilled landscape ¡V untilled landscape]/[tilled landscape + untilled landscape]) positively influenced total species richness and native species richness, but I also found a positive relationship between percentage of exotic bird species in the avian community and tilled index, indicating that more exotic bird species (fewer native species) were found in playas surrounded by a greater percentage of cultivated land. Models predicting avian species richness were variable among seasons, and diversity of land use within 1 km had a positive influence on total species composition. In general, variables in the best-fit models for wetland-dependent species richness, diversity, and density were similar to models predicting total species. Models predicting species richness, diversity, and bird density for wetland-dependent species explained more variation than predictive models for total species richness. This is likely because wetland-dependent species outnumbered non-wetland species and were important in driving avian community composition in wet playas. Playa area had a positive influence on waterfowl species richness and density, whereas water depth had a polynomial relationship with waterfowl species richness and density, showing a positive linear and a negative quadratic term. Waterfowl species richness and density peaked when water depth ranged from 40-100 cm. Percent vegetation cover had a polynomial relationship with waterfowl species richness, indicating that waterfowl most frequently occupied wetlands with intermediate levels of vegetation. Landscape variables were important in predicting waterfowl species richness and density, but the appearance of landscape variables in best-fit models varied among seasons, which suggests different habitat requirements for waterfowl in different portions of the annual cycle. Playa area was positively correlated with shorebird species richness but there was a negative relationship between shorebird density and playa area. There was a polynomial relationship, showing a positive linear and a negative quadratic term between water depth and shorebird species richness and density, which indicates that shorebirds tended to occupy wetlands with intermediate levels of water depth (30-80 cm). Percent vegetation cover was negatively correlated with shorebird density. Playa area was a consistent positive variable for predicting wading bird species richness and density. Because of high evaporation rates in summer and unpredictable precipitation, playa hydroperiods are unpredictable and the number of playas that contain water varies among seasons and years. Therefore, the natural wet-dry cycle is ideal for studying the influences of disturbance on wetland biota. The intermediate disturbance hypothesis (IDH) states that the highest levels of species diversity will occur at intermediate levels of disturbance, whereas diversity will be lower at higher and lower levels of disturbance, owing to effects of abiotic limitations and competitive exclusion, respectively. I tested the predictions of the IDH by comparing species richness and similarity index (changes of avian composition; Sorenson¡¦s similarity coefficient) among 3 disturbance regimes (considered different combinations of disturbance intensity and frequency and starting moisture condition). Results did not support the IDH, however, as the highest species richness did not occur at intermediate levels of disturbance. These results suggest that moisture condition itself may be more important in influencing species richness than a change in moisture condition (i.e., disturbance). Because wet playas have more species than dry playas in this semi arid landscape, playas associated with flooded conditions may have higher species richness regardless of the degree of disturbance. Playas with a higher degree of disturbance were associated with higher tilled indices in summer. The increased degree of disturbance due to sedimentation and lack of protection from the cultivated watershed had a negative impact on avian community composition. I used canonical correspondence analysis to examine the influence of local environmental variables (i.e., playa area, tilled index, water depth, percent vegetation cover, vegetation structure, vegetation height) on percent composition of the avian assemblage. Waterfowl, shorebirds, and wading bird were analyzed as groups, along with individual analyses on blue-winged teal (Anas discors), mallard (Anas platyrhynchos), American avocet (Recurvirostra americana), killdeer (Charadrius vociferus), long-billed curlew (Numenius americana), upland sandpiper (Bartramia longicauda), mourning dove (Zenaida macroura), and red-winged blackbird (Agelaius phoeniceus). Water depth, vegetation structure, and percent vegetation cover were consistently the top 3 variables influencing percent composition of birds throughout the summer. In general, percent composition of waterfowl as a group peaked when water depth ranged from 40-70 cm. Percent composition of shorebirds as a group was negatively correlated to water depth and vegetation-related variables, indicating that shorebirds occupied playas with shallower water and sparse vegetation. Percent composition of long-billed curlew was negatively correlated with tilled index, suggesting that playas within grassland watersheds are important for this species. In contrast, percent composition of mourning dove and red-winged blackbirds were positively associated with vegetation structure and percent vegetation cover. Watershed management surrounding playas should be given a high priority depending on the goals in the conservation plans, as increasing cultivation in playa watersheds increases total species richness but also favors exotic avian species and increases the degree of disturbance (i.e., water level fluctuation). With shortened hydroperiod of sedimented playas, the value of playas for avian species is decreased. Larger playas should also receive higher emphasis when other conditions (e.g., land use, vegetation, and landscape variables) are the same, since conserving larger playas is likely to maximize species richness. At a local scale, maintaining an intermediate level of water depth (40-100 cm) and 26-50% of vegetation is likely to meet the needs for the highest number of species and for most of the bird groups. Variables at a landscape level should be considered in conservation plans, as landscape-level variables appeared in many of the top models. Considering a group of playas as a complex instead of dealing with individual playas should provide diverse habitat for different groups of birds.Item MULTI-SCALE HABITAT SELECTION BY WINTERING WATERFOWL ON ANAHUAC NATIONAL WILDLIFE REFUGE(2017-11-15) Lane, Tiffany C.; Grisham, Blake A.; Moon, Jena Alyce; Conway, Warren C.; Haukos, David A.Waterfowl conservation and management focuses on conserving important habitats across their range. To understand waterfowl habitat selection, studies have described resource selection by individuals and populations of single dabbling duck species at multiple scales, ranging from geographic areas across North America to small-scale selection of food resources. Management efforts targeting selected habitat variables affect multiple species, especially during winter when waterfowl communities aggregate to share resources. Overlap in resource selection and utilization among similar species of dabbling ducks creates a community of wintering waterfowl. This community can be defined as a guild of waterfowl migrating in the Central Flyway, wintering in marshes along the Gulf coast and foraging mainly on plant matter. When multiple species share similar life history characteristics, Johnson’s (1980) 4 orders of habitat selection can be modified to reflect the process of disproportionate use by a community of waterfowl rather than an individual species. The goal of this study was to assess habitat selection by waterfowl communities wintering on Anahuac National Wildlife Refuge (NWR) to develop effective conservation plans targeting multiple species of dabbling ducks. Specifically, I quantified the relative influence of habitat covariates on wintering waterfowl habitat selection at multiple spatial and temporal scales, on Anahuac NWR in the Upper Chenier Plain of the Texas Coast. I observed mixed-flocks of waterfowl species that I defined as a guild of migratory, shallow-water dabblers that winter on the upper Texas Coast. The most abundant species include gadwall (Mareca strepera), Northern shoveler (Spatula clypeata), Northern pintail (Anas acuta), and green-winged teal (Anas crecca). Aerial waterfowl surveys were conducted in a grid pattern across the entire refuge to track waterfowl monthly habitat use from October to February 2014–2016. I visited geographically referenced sites where >500 birds were observed by aerial surveys in an aggregated flock and recorded water salinity, water temperature, water depth, submerged aquatic vegetation (SAV), and aquatic invertebrates at each used location. I collected these data at a random location within each management unit to represent available resources. I developed a resource selection function using used/available data with no marked individuals in a generalized linear model framework to assess waterfowl community habitat selection at multiple spatial and temporal scales. I defined the largest spatial scale (e.g., modified 2nd order) as Anahuac NWR, the intermediate spatial scale (e.g., modified 3rd order) as 8 individual management units within the refuge and the fine scale (e.g., modified 4th order) as flock site location (250 m flock buffer). The 3 temporal scales were year, early (October–November) and late (December–February) winter, and monthly. I extrapolated these data to the refuge and used habitat metrics best supported by the model set to build predictive habitat use maps for large flocks of dabbling ducks on the refuge. These results highlight the importance of considering multiple spatiotemporal scales in winter waterfowl habitat selection studies, particularly when animals of the same guild aggregate around shared-resources. At the 2nd order, waterfowl were influenced by multiple factors of the wetland landscape, water depth across the refuge acts as an initial selection cue to migrating waterfowl. Habitat selection within the refuge (3rd order) was positively influenced by a decrease in water salinity and waterfowl concentrated in Roberts-Mueller unit, a structurally managed intermediate marsh. Fourth order results suggest invertebrate density is a strong attractor to foraging waterfowl. In summary, waterfowl most benefit from multiple wetlands that provide a heterogeneous habitat composition of natural wetlands, allows access and visibility to open-water, has an average salinity of < 3ppt or below, and maximizes foraging by limiting water depth to <30 cm and supporting an abundance of food resources. Using the guild concept in lieu of single species is likely a more efficient and cost-efficient strategy for waterfowl habitat management.Item Occurrence, function, and conservation of playa wetlands: The key to biodiversity of the southern great plains(2011-05) Johnson, Lacrecia A.; Haukos, David A.; Smith, Loren M.; McMurry, Scott T.; McIntyre, Nancy E.; Mulligan, Kevin; Barnes, Melanie A. W.Playas form the primary wetland system in the High Plains portion of the Southern Great Plains (SGP) and provide valuable ecosystem services and functions including being key sites for biodiversity. Current estimates of the number of playas within the SGP (Texas, New Mexico, Oklahoma, southwestern Kansas, southeastern Colorado) from historical soil surveys (pre-1970s), topographic maps, and field checks exceed 25,000. This number often gives the potentially mistaken impression that there are numerous, adequately functioning playas in the region that continue to meet ecological and societal needs. In addition, these historical estimates are used to generate samples of playas for a variety of natural resource survey and research efforts, which depend on the occurrence of functional playas to generate sound inferential results. In January 2001, the U.S. Supreme Court ruled on Solid Waste Agency of Northern Cook County (SWANCC) v. United States Army of Corps of Engineers, which eliminated the provisions of the Clean Water Act available for the protection of playas and other isolated wetlands. The 1985 Food Security Act enacted the Federal “Swampbuster” provision for conservation of wetlands, which has been maintained in subsequent Farm Bills and is currently the only remaining potential incentive for the conservation playas throughout the SGP. For a playa to be determined as a wetland under Swampbuster, there must be a prevalence of hydric soil. Hydric soils were used not only to characterize and define playa locations, but also as a necessary criterion for determination of wetland status. However, since 1994, soils of the SHP associated with depressions, including playas, have been subject to reclassification and subsequent remapping by the United States Department of Agriculture (USDA). Anthropogenic factors and lack of regulatory protection have negatively impacted soil reclassifications, accelerated wetland loss, and impaired delivery of ecosystem services in many playas. Revised estimates of playa number, location, level of function, landscape connectivity, and capability for restoration are needed to effectively implement conservation of playas and species that rely on them. Playas occur in perhaps the most agriculturally impacted landscape in North America, with numerous prominent landscape alterations resulting from changing cultivation and grazing practices since the late 1800s. As a consequence, sediment accumulation and other contaminants are negatively impacting playas and leading to both physical and functional loss. One method to reduce the impacts of altered watershed condition upon adjacent playas is to provide a surrounding buffer of native vegetation. Wetland buffers reduce the adverse impacts of adjacent land uses by preventing or reducing sediments, contaminants, and excess nutrients (e.g., fertilizer or manure) from reaching the wetland. Conservation agencies and organizations have installed a very limited number of playa buffers during the past 15 years, and Farm Bill Programs have served to incidentally (e.g., Conservation Reserve Program; CRP) and intentionally (e.g., CP23a) buffer playas. However, a field-based study is needed to quantitatively evaluate the effectiveness of this technique. Collectively, playa wetlands are the primary sites for biodiversity and the primary wetland habitat for numerous wetland-dependent species that breed, migrate through, and winter in the High Plains. The loss of biodiversity in the Great Plains and elsewhere is driven by habitat loss, degradation, and increasing fragmentation, with future biodiversity also subject to changes as a result of climate change. Combined, these ecological stressors will have a strong influence on habitat and ecosystem connectivity and serious implications for the conservation of future biodiversity. Habitat network analysis is a promising tool for conservation planning and management in fragmented landscapes facing climate change impacts. Using this approach it is possible to model the impacts of landscape change on connectivity and processes supported by connectivity. To estimate playa loss, determine scale of function reduction, and evaluate conservation strategies for playa wetlands, my study objectives were to (1) estimate physical loss and modifications of playas as a function of anthropogenic impacts, (2) utilize these data to estimate a rate of playa loss since 1970 and the number of playas remaining on the landscape in the Southern Great Plains, (3) develop the framework of a functional assessment for Great Plains playa wetlands, (4) present the data collected to characterize the chosen functions and apply the resulting Function Assessment to playa wetlands of Texas, Oklahoma, and New Mexico to assess the level of function for playas within the SGP, (5) evaluate impacts of the potential change in area and number of hydric locations (i.e., historical playas) as a result of the USDA soil reclassification and subsequent remapping of upland and depressional soils in the SHP of Texas, (6) evaluate implications of remapping for natural resource managers and other scientists involved in conservation of playas, (7) address the effectiveness and impact of vegetative buffers of different widths and vegetation structure around playa wetlands on concentrations of metals, nutrients, and sediment in precipitation runoff and total volume of water entering playas, and (8) develop a framework for network analysis of playa systems. To assess physical loss and modifications, a combination of GIS and field data were utilized. A Geographic Information Systems (GIS) representation of playas historically and currently present on the SGP landscape was developed by combining the Playa Lakes Digital Database for the Texas Portion of the Playa Lakes Joint Venture Region (PLDD) and National Wetland Inventory (NWI) spatial data. Two percent of the wetland polygons were randomly selected for assessment of water feature type, anthropogenic impacts, and physical presence. USDA National Agricultural Imagery Program (NAIP) and Google Earth imagery were used to categorize the current state of each polygon as a playa, other waterbody, or no visible water feature. This categorization of the water features within the combined layer allowed for an estimate of the number of historical and current playas represented in the combined layer. When a polygon was categorized as playa, anthropogenic modifications and impacts were evaluated. Anthropogenic modifications and landcover was described within a 500-m diameter surrounding the playa. Historical playa locations with no depression visible in recent imagery and/or with loss of ≥ 100% of the original hydric-soil defined volume due to sediment accumulation were defined as physically lost. Sediment loads were determined through field visits to randomly selected dry playas. Using sediment depth, percent volume loss due to sediment accumulation was calculated. The percent of playas with ≥ 100% loss of the original hydric soil defined volume was determined. The combined dataset contained 34,512 polygons representing potential water features in the SGP across 51 counties in three states. The status of 740 polygons (2% per county) was assessed and categorized. Six hundred and two were categorized as historical playas, 125 as other waterbodies, and 13 as no visible water feature. Only 1 playa (0.2%) had no playa or watershed modification. Twenty-eight playas (4.7%) had no playa modification. Three playas (0.5%) had no watershed modification. Across all landcover types, 51.5% of playas were tilled at the time the imagery was taken or showed signs of being tilled in the past. Mean volume loss within native grassland and CRP were <100% of the original hydric volume in Texas, New Mexico, and Oklahoma. A playa confirmation rate of 81.4% resulted in an estimated 28,092 depicted historical playa locations in Texas, Oklahoma, and New Mexico. During the time period of 1970-2008 an estimated 17% of playas have been physically lost from the SGP landscape. Including the impacts of sediment accumulation, playa wetlands were reduced by an estimated 60% (1.6% per year). These estimates are conservative in that cultivation of the playa floor is not included as a loss. However, during cultivation of the playa, the characteristic hydric soil layer is altered. The inclusion of past and present tilling as a physical loss increases this estimate to 85.7% and a rate of 2.3% per year, during the 38 years under study. The cumulative influence of these losses is unknown until the spatial distribution of lost playas is evaluated in relation to existing playas. The framework for a functional assessment was created through the use of field data, literature searches, and best professional judgment. It consists of 15 key elements or predictors for playa wetlands (i.e., model variables) and their associated measurement or condition important in evaluating the level of an individual playa’s function based on the three ecological functions or services of interest: hydrology/recharge, water quality, and biodiversity. Based on the outcome from this matrix, a continuous function scale was populated, producing a distribution of sampled playas. Cut-off points were established within this continuous function scale to place playas into one of five categories (i.e., fully functional, partially functional and restorable, partially functional and non-restorable due to cost, partially functional and non-restorable, and non-functional). In 2008-2009, playas were stratified by playa density/county and randomly selected. These playas were used to create the function matrix, adjust values within the function matrix, and to assess function delivered by the playas remaining on the landscape. For the function assessment, empirical data were combined with GIS-derived information, forming 15 model variables. Nine model variables were directly evaluated via either field sampling or GIS-based techniques: (1) Physical Modifications, (2) Reduction in Original Volume, (3) Sediment Depth, (4) Agricultural Impacts, (5) Ratio of Native to Non-Native Plant Species, (6) Buffer Zone, (7) Density of Wetlands, (8) Landcover in Watershed, and (9) Upland Land Use. The remaining 6 variables were assessed using a subset of playas, due to both time and financial constraints, through the use of field data, basic statistics and modeling: (1) Average Duration of Surface Water, (2) Measured Aluminum, (3) Measured Chromium, (4) Measured Iron, (5) Measured Total Suspended Solids, and (6) Biomass. Through the application of the function matrix, none of the sampled playas were estimated to function at full functional capacity in the SGP. Seventy-three (47%) of playas were estimated to be partially functional and restorable. Partially functional and non-restorable due to cost playas were estimated at 12.9% or 20 playas, and 61 (39.4%) playas were partially functional and non-restorable because effective restoration techniques do not exist. Lastly, 1 (0.01%) of the extant surveyed playas was estimated to be non-functional. For eight counties in Texas, I compared the occurrence of playas, as indicated by soils designated as hydric in original soil surveys, to designations in USDA-remapped soil surveys of upland and depressional soils. I estimate a 65% decrease in playa numbers and 50% decrease in playa area as defined by the presence of a hydric soil following soil remapping. An estimated 80% of small playas (<5 ha in area) will potentially lose protection due to soil remapping. Playas embedded in both grassland and cropland watershed cover types were remapped; however, the mean (±SE) size of playas where soil type changed was 7±0.40 ha and 3±0.06 ha, respectively, for each cover type. Excessive sediment accumulation and other anthropogenic factors, resulting in an altered hydrology and masking of hydric soil characters, are proposed as primary factors responsible for differences in playa numbers and area following soil remapping. Other factors potentially impacting the remapping results include current USDA methodology and correction of historical survey errors. Therefore, any playa in the SHP being considered for inclusion under USDA conservation programs must be individually and independently assessed on-site for wetland criteria, rather than relying on information provided by revised USDA-NRCS Soil Survey maps. Further, during on-site evaluations, effects of anthropogenic alterations on the ability of the playa soil to develop and maintain hydric characteristics must be considered. Finally, until the remapping effort is complete, the online USDA-NRCS Soil Survey maps will be comprised of a mixture of historical soil surveys and revised classification of historical surveys, which will cause confusion during interpretation by those unfamiliar with the status of soil survey reports for the Texas SHP. In 2008 and 2009, the effect of buffers surrounding playa wetlands on water quality was evaluated as functions of buffer width and vegetation cover. Precipitation runoff was collected from 36 playa buffers (Conservation Reserve Program=7, Fallow Cover=18, Native Grassland=11) using I-CHEM storm water samplers (n=238) placed at the outer edge of the buffer adjacent to cultivated watersheds and each 10 m to the playa floor. Samples were analyzed for total dissolved solids (TDS), total suspended solids (TSS), specific conductance (SC), pH, nitrate, phosphorus, and metals (i.e., Al, As, Ba, Ca, Cd, Cr, Cu, Fe, K, Mg, Mn, Ni, Sr, V, Na, Zn). Also evaluated was the influence of buffers on hydrology, using the NRCS-USDA Curve Number Method. TDS and TSS reached a combined maximum removal at 50 m, 49% and 72% respectively. Nitrate and phosphorus reached a combined maximum removal at a distance of 20 m, 49% and 33% respectively. Maximum removal of metals occurred at 40 m. Estimated percent reduction in runoff reaching the playas due to the presence of a buffer was greatest for the native CRP cover type (-5.8%). A minimum buffer width of 40-50 m is necessary to maximize contaminant removal from runoff entering playa wetlands. However, use of buffers does not completely negate the impacts of watershed cultivation on playa wetlands. Therefore, additional conservation practices are necessary in adjacent cultivated watershed (e.g., contour plowing, no-till agricultural) to minimize movement of contaminants from watersheds into playas. The proposed framework for an initial network analysis of the Great Plains playa system includes use of metrics and algorithms outside of currently available software and incorporates physical and functional loss data. For playa locations, I recommend the combined spatial data developed for use in assessing physical and functional loss of playas; for surrounding land use I recommend the National Landcover Database (http://www.mrlc.gov/). Indices for playa quality are probability of inundation and level of function. The utilization of this framework will address how physical and functional loss of playas, in concert with climate change, is currently influencing or going to influence connectivity and biodiversity.Item Recruitment of mottled ducks (Anas fulvigula) on the upper Texas gulf coast(2008-12) Rigby, Elizabeth Anne; Haukos, David A.; Smith, Loren M.; McIntyre, Nancy E.Surveys of the mottled duck (Anas fulvigula) population in Texas have shown steep declines over the last 20 years. The causes of the population decline are not well understood, but low recruitment has recently been hypothesized. Female survival during the breeding season has not been widely discussed as a contributing factor, even though it plays an important role in population dynamics. I investigated mottled duck recruitment as well as female survival during the breeding season and gathered information on habitat use to guide management decisions. My objectives were to (1) estimate weekly survival rates of mottled duck females throughout the breeding season, (2) estimate the daily movement and home range of mottled duck females during the breeding season, (3) estimate daily survival rates of ducklings and estimate mottled duck recruitment, and (4) estimate daily movements and home range of mottled duck broods. My study area was the Texas Chenier Plain National Wildlife Refuge (NWR) Complex, which was comprised of Anahuac, McFaddin, Texas Point, and Moody NWRs, with most research occurring on Anahuac NWR. I caught female mottled ducks in decoy traps and baited swim-in traps on Anahuac NWR from 1 February to mid-April or early May, 2006-2008. Females were fitted with back-mounted radio transmitters and tracked through the end of the breeding season, 30 June. I used 2 tracking methods: remote locations, where I used observer coordinates and bearings to the female to triangulate female locations, and visual locations where I tracked the female until seen. I captured 15 females in 2006, 12 in 2007, and 20 in 2008. Of these, 2 died in each of the first 2 years and 3 in the third year due to subsequent trapping stress and were removed from analyses. Survival of a female in each week was documented either with a visual location or a remote location that indicated obvious movement since the last location. I estimated weekly female survival using the Known Fate procedure of Program MARK (White and Burnham 1999) and tested 5 a priori models with a body condition index (body mass (g) / wing chord (mm) ) included as a covariate. Models were evaluated based on Akaike’s information criterion corrected for small sample size (AICc) values. Female survival estimates for the entire 19-week breeding season were 75.5% for the top model, range 63.3% -87.2% for other models with ÄAICc less than 4. Mottled duck nesting propensity was similar for 2006 and 2007 but increased in the third year. There is some evidence that breeding season survival of female mottled ducks decreased when nesting effort increased. Females showed the highest nesting propensity (62.5%) in 2008, the year with the lowest female survival rate. For all years, weekly survival for females was the same during the breeding season as it was for the remainder of the year. Although nesting may carry increased risk, as signaled by the decrease in female survival when nesting effort increased, nesting rates were so low that the weekly survival rate was not reduced significantly during the breeding season. Low nesting propensity appeared to be linked with poor habitat conditions such as drought and high marsh salinity. The links among nesting propensity, female survival, and habitat suitability, suggests that a “trade-off” is at play for female mottled ducks. I also used visual and remote locations of mottled duck females to calculate minimum daily movements. Large dispersal movements (> 10 km) were excluded from daily movement and home range calculations. I calculated home ranges for females using 95% adjusted-kernel density estimates. In 2006, 7 females dispersed > 10 km from the study area, moving an average of 31.9 km. Dispersing females moved to inland freshwater impoundments in a year of extreme drought, indicating dispersal is influenced by weather and habitat conditions. Dispersal was precluded during drought in 2008 by presence of new local refugia. Females that nested moved less than females that did not. Home range size did not differ between breeding versus non-breeding females or among years, suggesting female home ranges were limited by habitat patch size. To estimate duckling and brood survival, movement, home range, and habitat use, I tracked radio-transmittered ducklings and radio-transmittered females with broods. I caught mottled duck broods from airboats in 2006, 2007, and 2008. I randomly selected 2 ducklings from each brood and fitted them with radiotransmitters. I captured 7 ducklings from 4 broods in 2006, 43 ducklings from 23 broods in 2007, and 9 ducklings from 5 broods in 2008. I used duckling visual and remote location data to estimate daily duckling and brood survival using the Known Fate procedure of Program MARK (White and Burnham 1999). The top models for both brood and duckling survival were constant survival among days and years. My duckling survival estimate was greater than expected (66.9% for the 30-day period) and was high compared to a previous estimate of 41.1% in Texas, as well as compared to estimates of duckling survival for closely related species. Although duckling survival was high in my study, recruitment estimates were low. To maintain a stable population, I estimated that a recruitment rate of R = 0.91 is needed. When recruitment was calculated using existing nest success estimates, all but 1 recruitment estimate were smaller than R = 0.91, implying negative population growth. By comparing mottled duck recruitment parameters to mallard parameters, I confirmed that low hen success, determined by nest success and nesting propensity, was largely responsible for low mottled duck recruitment. Duckling locations and female with brood locations were also used to calculate minimum daily brood movements and brood home ranges using 95% adjusted-kernel density estimates. Brood movements differed among years, but brood home range size did not because habitat patch size was not limited by water level. Instead, human created barriers (roads, levees, bayous), dry zones, and areas dominated by heavy, impassable vegetation formed physical barriers to duckling movement. There was only 1 observed brood movement between habitat patches. The isolation of broods within habitat patches may increase risk to broods in years of severe drought. Movements were greatest in 2007, the year with the greatest rainfall and the only year without a drought period. Movements and home range were similar for broods in moist soil units (MSUs) and broods in coastal marsh. The cornerstone of mottled duck management must be the maintenance of a trio of habitats necessary for their life cycle: nesting habitat, brood rearing habitat, and adult habitat. Low observed nesting propensity and previously low reported nest success indicate increasing nesting habitat area and quality should be a prominent goal of habitat management. Location of nesting areas is as important as habitat quality. Nesting areas need to be adjacent to areas suitable for broods. Predator control in mottled duck nesting areas has not been widely pursued on the gulf coast but should be carefully considered. Current recommendations for desired brood habitat characteristics, such as disturbed, broken marsh, were largely supported by habitat use of mottled duck broods in my study. Managers should encourage connectivity of brood areas to give broods the widest possibilities for habitat selection. In managing habitats for stages of the mottled duck life cycle, managers should be particularly attuned to the effects of drought. Local refugia during drought are critical as adult mottled duck habitat and can reduce the need for dispersal of adult mottled ducks from coastal refuges.Item Remote sensing of vegetation biomass in playa wetlands(1994-05) Comer, Gary L.; Fish, Ernest B.; Haukos, David A.; Smith, Loren M.; Zartman, Richard E.Item Reproductive success and annual habitat preference by Red-shouldered Hawks in southern Texas(2011-05) Strobel, Bradley N.; Boal, Clint W.; Andersen, David E.; Haukos, David A.; Griffis-Kyle, Kerry L.; Blankenship, Terry L.Identifying the vegetation characteristics resulting in habitat use by nesting raptors provides information important to species conservation and management. I examined nesting habitat preference of red-shouldered hawks (Buteo lineatus) using conditional logistic regression on characteristics measured at 27 occupied nest sites and 68 randomly selected unused sites in 2005–2009 in southern Texas. I measured vegetation characteristics of individual trees (nest trees and unused trees) and corresponding surrounding 0.04 ha plots. I evaluated the importance of tree and plot characteristics to nesting-habitat selection by comparing a priori tree-specific and plot-specific models using Akaike‘s Information Criterion corrected for small sample size (AICc). Models with only plot variables carried 14% more weight than models with only center-tree variables. The model-averaged odds ratios indicated red-shouldered hawks selected to nest in taller trees and in areas with higher average diameter at breast height (dbh) than randomly available within the forest stand. Relative to randomly selected areas, each 1-m increase in nest-tree height and 1-cm increase in the plot-average dbh increased the probability of selection by 85% and 10%, respectively. My results indicate characteristics of individual trees and the structure of the 0.04 ha area surrounding the tree, were associated with red-shouldered hawk nesting habitat selection. My results show red-shouldered hawks preferred taller nest trees (average 16.4 m) surrounded by larger diameter trees (average 25.4 cm) relative to the surrounding forest in southern Texas. Preference for taller nest trees by raptors is well documented and suggests an association with increased fitness. Further study is required to understand the mechanism responsible for this preference. Widespread alteration of forest communities combined with the well-documented preference for mature forests by breeding red-shouldered hawks (Buteo lineatus) has caused concern over the conservation of the species. Although habitat preferences of breeding red-shouldered hawks have been well documented, few data describe non-breeding-period habitat selection. I studied the seasonal habitat preferences of 12 adult red-shouldered hawks in southern Texas. I used radio-telemetry to collect >1,800 locations during 3 discrete phenological periods (winter, breeding-spring, and non-breeding-spring). I constructed phenological stage-specific discrete-choice models to compare the proportion of a 0.5-ha circle around each location (used and available) comprised of mature forests, young forests, brush, grass, wetland, and open-water cover types. I used general linear models to assess whether the amount of mature forest covertypes within used areas was dependent upon the sex of the individual, its breeding status, or the time of year the data were collected. During the breeding-spring-period, red-shouldered hawks preferred areas with greater amounts of mature forest cover types (selection ratio = 1.40, 95% CI = 1.16 - 1.69) and higher cover-type richness (selection ratio = 1.10, 95% CI = 1.04 - 1.21). Characteristics associated with habitat use during the breeding-spring-period were different than those associated with winter-period or non-breeding-spring-period habitat selection. Habitat use during the winter was associated with greener vegetation measured using the Normalized Difference Vegetation Index (NDVI), and avoidance of grasslands, brushlands, and wetlands. My data indicate the amount of forested cover types preferred by individual red-shouldered hawks depends upon the sex of the individual as well as its breeding stage (i.e., breeding, non-breeding). Habitat quality is often evaluated based on relative resource preferences, usually by examining disproportionate use of available resources. Alternatively, the Ideal Despotic Distribution hypothesis predicts that given a suite of territories to select from, an individual will select the highest quality territory in regards to its perceived fitness benefit. Presumably, the long-term occupancy rates of specific territories will be positively correlated with their quality. However, the predictions of the ideal despotic distribution hypothesis hinge upon the individual‘s ability to accurately assess territory quality (perfect knowledge), a condition that is difficult to meet under temporally and spatially variant environments. Using a 5-year data set on a resident population of red-shouldered hawks in southern Texas, I examined the predictions of the ideal despotic distribution hypothesis. My results were consistent with the prediction of the ideal despotic distribution hypothesis that the proportion of years a territory supported a nesting attempt was negatively related to the relative nest initiation dates. However, in contrast to the predictions of the ideal despotic distribution hypothesis, my data show that proportion of years a territory supported a nesting attempt was unrelated to the average number of eggs laid and negatively related to the average number of young fledged annually. Whereas the proportion of years a territory supported a nesting attempt varied across territories, the annual reproductive success of breeding red-shouldered hawks varied markedly from year to year and was positively associated with the number of breeding pairs observed. Overall, my results suggest that red-shouldered hawk reproductive success is influenced heavily by stochastic characteristics that violate the ideal despotic distribution hypothesis‘ assumption of perfect knowledge by individuals.Item Response of nesting grassland birds to sand shinnery oak communities treated with tebuthiuron and grazing in eastern New Mexico(Texas Tech University, 2006-05) Smythe, Lindsay Allison; Haukos, David A.; Smith, Loren M.; McIntyre, Nancy E.Many grassland bird species have exhibited population declines during the past three decades. Major components of grassland bird habitat in eastern New Mexico include sand shinnery oak (Quercus havardii) communities, many of which have been subject to unmanaged livestock grazing and other impacts for decades. As a result, these sand shinnery oak communities tend to deviate from the historical grass/shrub vegetation mix in favor of shrubs. These communities are frequently managed with livestock grazing and herbicide application for shrub control. These techniques can be used for potential restoration of historical community structure. However, responses of spring migrating and nesting grassland birds to these management practices in shinnery oak communities are not well understood because of variation in species response. Additionally, most of the research in these communities has focused on game species. My study objectives, in the context of restoring shinnery oak communities, were to: 1) determine if tebuthiuron herbicide application or short-duration grazing had negative impacts on avian community structure (i.e., density, species richness, diversity, and evenness), 2) determine if the vegetation changes caused by tebuthiuron application and short-duration grazing negatively impacted avian reproduction (nest density and success), and 3) develop recommendations to benefit grassland birds when using herbicides and livestock grazing in sand shinnery oak communities. My study site in eastern New Mexico consisted of 1,040 ha divided into 16 plots of 65 ha each. The plots consisted of 2 treatments arranged in 4 combinations: tebuthiuron with grazing; tebuthiuron without grazing; no tebuthiuron with grazing; and a control of no tebuthiuron or current grazing. I performed biweekly point transects on these plots from February through July 2004 and 2005. Density of bird species was estimated using program DISTANCE. I also searched 4-ha subplots in each treatment plot for nests from April-June 2004 and 2005. I monitored nests to estimate daily survival rates and recorded vertical density and overhead cover measurements at each nest site to determine if vegetation structure affected nest daily survival or success. In 2005, I placed 64 artificial nests to supplement information from real nests. This study occurred over years of highly variable precipitation: 2003 represented the end of a 15-year drought with below-average precipitation, but during 2004, the area received 3 times the average amount of precipitation, the second highest amount ever recorded in the region. The above-average precipitation affected habitat conditions and likely impacted my results, but precipitation returned to near average in 2005, so these conditions may not persist. Density of all avian species did not differ between grazed and ungrazed plots. Tebuthiuron-treated plots had a higher density of all species than untreated plots. There was a higher density of all species during spring 2005 than 2004 but density was similar during the breeding season of both years. These trends were predominantly influenced by Cassin’s sparrows (Aimophila cassinii) and, in 2005, grasshopper sparrows (Ammodramus savannarum). Tebuthiuron-treated plots had a higher density of Cassin’s sparrows in both years. Grasshopper sparrows were not recorded in 2004 but were present in greater numbers on tebuthiuron-treated plots in 2005. The density of resident species (meadowlarks [Sturnella spp.] and loggerhead shrikes [Lanius ludovicianus]) exhibited little response to tebuthiuron or grazing treatments. Species richness was not affected by tebuthiuron or grazing, but was higher in the spring of 2005 than 2004 because of greater numbers of migratory sparrows. Decreases in both evenness and diversity in 2005 compared to 2004 were influenced by large flocks of chestnut-collared longspurs (Calcarius ornatus) and lark buntings (Calamospiza melanocorys). Diversity was also lower on ungrazed plots in February and March. Nest density of all species was similar among tebuthiuron and grazing treatments but was greater in 2005 than 2004. Nests of migratory species were more numerous in 2005 than 2004, whereas equal numbers of resident species nests were found in both years. Daily nest survival rates were similar between years and between incubation and the nestling periods, but varied among species and treatments. Daily nest survival rates during incubation were 6.3% higher in untreated plots than in tebuthiuron-treated plots, but during the nestling period were 17.3% higher in tebuthiuron-treated plots than in untreated plots. Depredation caused the majority of nest failures. Vertical density of vegetation differed among treatments and was greater in 2005 than 2004 but did not differ between nest sites and associated random points, nor between hatched and failed nests. Vegetation overhead cover did not vary among treatments but was greater in 2005 than 2004 and greater at nest sites than at associated random points. Overhead cover did not differ between nests that hatched and nests that failed. Unusually high rainfall in 2004 likely influenced results, but neither grazing nor tebuthiuron treatment had a substantial impact on resident birds. Migratory birds responded positively to tebuthiuron treatment, but overall density and nest success in this community (regardless of treatment) were extremely low.Item Seasonal survival of adult female mottled ducks(2017) Moon, Jena A.; Haukos, David A.; Conway, Warren C. (TTU)The mottled duck (Anas fulgivula) is a non-migratory duck dependent on coastal habitats to meet all of its life cycle requirements in the Western Gulf Coast (WGC) of Texas and Louisiana, USA. This population of mottled ducks has experienced a moderate decline during the past 2 decades. Adult survival has been identified as an important factor influencing population demography. Previous work based on band-recovery data has provided only annual estimates of survival. We assessed seasonal patterns of female mottled duck survival from 2009 to 2012 using individuals marked with satellite platform transmitter terminals (PTTs). We used temperature and movement sensors within each PTT to indicate potential mortality events. We estimated cumulative weekly survival and ranked factors influential in patterns of mortality using known-fate modeling in Program MARK. Models included 4 predictors: week; hunting and non-hunting periods; biological periods defined as breeding, brooding, molt, and pairing; and mass at time of capture. Models containing hunt periods, during and outside the mottled duck season, comprised essentially 100% of model weights where both legal and illegal harvest had a negative influence on mottled duck survival. Survival rates were low during 2009–2011 (12–38% annual rate of survival), when compared with the long-term banding average of 53% annual survival. During 2011, survival of female mottled ducks was the lowest annual rate (12%) ever documented and coincided with extreme drought. Management actions maximizing the availability of wetlands and associated upland habitats during hunting seasons and drought conditions may increase adult female mottled duck survival. © 2017 The Wildlife Society.Item The influence of environmental landscape variables on lesser prairie-chickens in the Sand Shinnery Oak Prairie ecoregion of Texas and New Mexico and the Mixed-Grass Prairie ecoregion of Oklahoma and Kansas(2016-08-10) Griffin, Cody; Grisham, Blake A.; Boal, Clint W.; Haukos, David A.; Kahl, Samantha S.Lesser prairie-chickens (Tympanuchus pallidicinctus) have experienced a decline in population numbers and distribution throughout the central and southern Great Plains, which led to their status as a threatened species under the United States Endangered Species Act of 1973. However, the listing was vacated in September 2015 by a federal judge in Texas. Regardless, the lesser prairie-chicken remains a species of conservation concern and ecological data are pertinent to understanding mechanisms driving population demography. Populations in the Sand Shinnery Oak (Quercus havardii) Prairie are located in the southwestern-most portion of their distribution, isolated from their core distribution in the Mixed-Grass Prairie, Sand Sagebrush (Artemisia filifolia), and Short-Grass Prairie ecoregions. The climate within the species’ distribution is characterized by frequent droughts that negatively affect population numbers, but the effects of weather variables and future climate change on lesser prairie-chicken populations between isolated and core populations are unknown. Additionally, populations are subject to the effects of anthropogenic-driven landscapes, but effects of land use and anthropogenic structures among and between both regions are unknown. To address these concerns, I modeled survival of males, females, broods, and nest from the Sand Shinnery Oak Prairie and Mixed-Grass Prairie ecoregions as a function of a priori models composed of biologically relevant temperature and precipitation parameters. I then modeled the effect of land cover classifications such as croplands, bare ground, shrublands, grasslands, Conservation Reserve Program (CRP) lands, and trees on lesser prairie-chicken lek attendance. In addition, I modeled the effect of anthropogenic structures such as transmission lines, oil and gas wells, improved and unimproved roads, wind turbines, and oil and gas wells with a five-year time lag on lek attendance. Finally, I created an Integrated Population Model (IPM) to assess the potential effect of climate change given different Representative Control Pathways on lesser prairie-chickens populations. My findings suggest that weather variables associated with extreme temperatures and precipitation influence lesser prairie-chicken vital rates in both ecoregions. I modeled the effect of weather parameters on male, female, nest, and brood survival in the Sand Shinnery Oak Prairie Ecoregion (1999–2012) and the Mixed-Grass Prairie Ecoregion (1999–2015). When modeled as a function of weather variables, the monthly probability of male survival during the study period (March 1 to September 12) was 0.80 (SE = 0.004, 95% CI: 0.75–0.84) in both ecoregions. Female survival differed among ecoregions, and the probability of females surviving two-week intervals during the study period was 0.77 (SE = 0.05, 95% CI: 0.71–0.81) in the Sand Shinnery Oak Prairie Ecoregion and 0.96 in the Mixed-Grass Prairie Ecoregion. Likewise, nest survival differed between the two ecoregions, and the probability of a nest surviving the 28-day incubation period was 0.32 (SE = 0.004, 95% CI: 0.28–0.43), and 0.24 (SE = 0.004) in the Sand Shinnery Oak Prairie and Mixed-Grass Prairie ecoregions, respectively. Finally, brood survival differed between ecoregions, and the probability of broods surviving the 30-day brooding period was 0.16 (SE = 0.01, 95% CI: 0.04–0.29 and 0.74 (SE = 0.005, 95% CI: 0.55–1.00) in the Sand Shinnery Oak Prairie and Mixed-Grass Prairie ecoregions, respectively. All demographic groups exhibited positive associations with breeding season extreme temperature variables and non-breeding season temperature variables. Although weather had no significant effect on brood survival in the Sand Shinnery Oak Prairie Ecoregion, broods in the Mixed-Grass Prairie responded positively to extreme hot temperatures the first two weeks of brooding. The response of lek counts to anthropogenic structure dispersion differed between ecoregions. I found no significant relationships between lek counts and anthropogenic structure dispersion in the Sand Shinnery Oak Prairie Ecoregion; however, lek counts exhibited a negative relationship with transmission lines (≥ 69-kV; β = -1.93; 95% CI: -3.29–-0.71) and unimproved roads (β = -0.92; 95% CI: -1.43–-0.42) in the Mixed-Grass Prairie Ecoregion. The response of lek counts to land cover patches differed across the ecoregions with negative responses primarily associated with increasing monocultures of shrublands (β = -0.20; 95% CI: -0.30–-0.11) in the Sand Shinnery Oak Prairie Ecoregion and CRP lands (β = -0.83; U; 95% CI: -1.23–-0.43) in the Mixed-Grass Prairie Ecoregion. Although, the IPM can be improved with more informative parameters, I found that populations in the Sand Shinnery Oak Prairie Ecoregion were predicted to no longer persist after 2036 while populations in the Mixed-Grass Prairie Ecoregions were predicted to decline through 2100. Lambda values for populations in the Sand Shinnery Oak Prairie Ecoregion given all emission scenarios were 0.50 from 1996–1997, spiked to above one in 1999, and stayed at 0.50 from 2000–2009. In 2011, lambda trajectories dropped to < 0.3 and populations became extinct between 2012–2036. Although populations were not projected to become extinct in the Mixed-Grass Prairie Ecoregion, lambda values remained below one from 1995–2100 indicating that the population would continue to decline given every emission scenario. My results indicate that outlook for lesser prairie-chicken populations in both ecoregions will not be sustainable or cease to persist within the 21st century. To ensure population persistence, management efforts should focus on maintaining sufficient vegetation cover at the landscape level while ensuring landscape heterogeneity to curb the effects of more frequent and intense drought events given future climate change.Item The Predicted Influence of Climate Change on Lesser Prairie-Chicken Reproductive Parameters(2013) Grisham, Blake A. (TTU); Boal, Clint W. (TTU); Haukos, David A.; Davis, Dawn M.; Boydston, Kathy K.; Dixon, Charles; Heck, Willard R.The Southern High Plains is anticipated to experience significant changes in temperature and precipitation due to climate change. These changes may influence the lesser prairie-chicken (Tympanuchus pallidicinctus) in positive or negative ways. We assessed the potential changes in clutch size, incubation start date, and nest survival for lesser prairie-chickens for the years 2050 and 2080 based on modeled predictions of climate change and reproductive data for lesser prairie-chickens from 2001-2011 on the Southern High Plains of Texas and New Mexico. We developed 9 a priori models to assess the relationship between reproductive parameters and biologically relevant weather conditions. We selected weather variable(s) with the most model support and then obtained future predicted values from climatewizard.org. We conducted 1,000 simulations using each reproductive parameter's linear equation obtained from regression calculations, and the future predicted value for each weather variable to predict future reproductive parameter values for lesser prairie-chickens. There was a high degree of model uncertainty for each reproductive value. Winter temperature had the greatest effect size for all three parameters, suggesting a negative relationship between above-average winter temperature and reproductive output. The above-average winter temperatures are correlated to La Niña events, which negatively affect lesser prairie-chickens through resulting drought conditions. By 2050 and 2080, nest survival was predicted to be below levels considered viable for population persistence; however, our assessment did not consider annual survival of adults, chick survival, or the positive benefit of habitat management and conservation, which may ultimately offset the potentially negative effect of drought on nest survival.