19th Annual Conference of The Wildlife Society
October 13-17, 2012
Oregon Convention Center
Portland, OR
http://wildlifesociety.org/

Researchers from the USGS Western Ecological Research Center present findings and posters at the TWS national conference each year.


PRESENTATIONS

Assessing population structure and subspecific status of the Western Shovel-nosed snake across the Colorado and Sonoran desertscrub valleys of California and Arizona
Sunday, October 14, 2012 | 8:50 a.m. PDT | Room A103-A104
WERC Contributors: Dustin Wood*, Amy Vandergast

There is concern that the Tucson Shovel-nosed snake (Chionactis occipitalis klauberi) has experienced significant declines over the past 25 years, which has lead to a petition to list klauberi under the Endangered Species Act. An important question in evaluating that petition is whether klauberi is a distinct subspecies. Subspecific recognition within the Western Shovel-nosed snake has been based primarily on weak morphological differentiation, and is further complicated by the fact that klauberi purportedly forms a large zone of intergradation with the Colorado Shovel-nosed snake (C. o. annulata) across central Arizona. Results from our previous range-wide analyses, using morphology and mtDNA data, supported six monophyletic units; however, neither genetic nor morphological data were entirely concordant with the traditional subspecies taxonomy. In particular, representative samples of klauberi were marginally differentiated morphologically, yet genetically embedded in a mtDNA clade that included both klauberi and annulata. Given these inconclusive results, we developed a multi-locus strategy, using nuclear microsatellite loci, to provide a robust estimate of the genetic units within shovel-nosed snakes, with particular emphasis on the zone of intergradation between the klauberi and annulata subspecies. We use 11 microsatellite loci genotyped for 264 specimens and two Bayesian clustering methods to infer population structure without a priori subspecies assumptions. Analyses support five genetic clusters across the species range, with admixture between groups generally occurring along contact zones. Importantly, one of the clusters recovered was concordant with the distribution of klauberi, providing genetic support for taxonomic recognition. Finally, significant estimates of population differentiation between clusters are detectable and appear to be driven, at least in part, by isolation-by-distance. However, even after controlling for geographic distance there still appears to be a gene flow barrier between the Tucson Shovel-nosed snake and more western genetic clusters, suggesting factors other than geographic distance may be at play.

Distribution and abundance of American pika in the Sierra Nevada and White Mountain ranges of California
Tuesday, October 16, 2012 | 2:10 p.m. PDT | Room A106
WERC Contributors: Robert Klinger*, Cody Massing

We used a combination of line transect and point count surveys to collect data on distribution, abundance and habitat use of American pika (Ochotona princeps) in the Sierra Nevada and White Mountain ranges of California between 2008 and 2011. In the Sierra Nevada, observations were made each year along a minimum of 9 (2008) and a maximum of 21 (2011) 10-km transects, and at a minimum of 40 (2008) and a maximum of 210 (2011) point count stations. In the White Mountains, six transects 1.5 to 8 km long were sampled each year. Pika occurred widely throughout the alpine and subalpine zones of both ranges, but their distribution was very patchy and occupancy and abundance highly variable. Approximately 37% of the study area was occupied by pika, but variability in occurrence within a given geographic region was as high as variability between regions. Site occupancy expanded between 2008 and 2009 but 50% fewer sites were occupied in 2011 following the second consecutive heavy and extended winter in as many years. Land cover and topography variables had greater predictive value of occupancy than climate variables. Density was highly variable at patch, local, and regional scales; areas in both mountain ranges that had high abundance in one year could have much lower abundance the next year and vice versa. Density in the White Mountain range was generally 3-4 times greater than that in the Sierra Nevada (12.2 km2 ± 1.4 SE vs. 3.5 km2 ± 0.4, respectively), but density decreased by about 50% in both ranges following the winter of 2010/2011. The distribution and abundance of pika in the Sierra Nevada and White Mountain ranges are very dynamic, but even in years when populations are low tens of thousands of individuals occur in the two mountain ranges. 


A modeling framework to integrate harvest and habitat management of North American waterfowl: case-study of northern pintail metapopulation dynamics
Tuesday, October 16, 2012 | 2:50 p.m. PDT | Room B113-B114
WERC Contributors: Joseph Fleskes

We developed and evaluated the performance of a metapopulation model designed to enable managers to examine, for the first time, consequences of alternative management strategies involving habitat conditions and hunting on both harvest opportunity and carrying capacity for migratory waterfowl at a continental scale. We focus on the northern pintail (Anas acuta; hereafter, pintail), which serves as a useful model species to examine the potential for integrating waterfowl harvest and habitat management in North America. We developed submodel structure capturing important demographic processes for pintail populations during breeding, fall migration, winter, and spring migration while representing several core breeding and nonbreeding areas. Continental-scale predictions from our baseline parameterization resulted in a carrying capacity of 5.5 million with an equilibrium population size of 2.9 million and harvest rate of 12% at maximum sustained yield. These values are within 10% of those from the pintail harvest model currently used by the U.S. Fish and Wildlife Service. Our model allows estimation of the relative contributions to continental-scale population dynamics of habitat quantity and quality among the core breeding and nonbreeding areas. Ongoing model refinements include: 1) extending submodels with additional parameters to link impacts of habitat management and environmental conditions on biological response parameters; 2) completing a formal sensitivity analysis; and 3) estimating costs of changing biological parameters through habitat management efforts. By integrating habitat and harvest dynamics within this modeling approach, we can predict habitat and harvest management influences on continental-scale population responses while explicitly considering putative effects of climate change. Our model could be readily adapted to address management questions for many habitat-limited species.


Spatial and temporal variability in alpine meadow condition and boundaries in the Sierra Nevada mountain range of California
Tuesday, October 16, 2012 | 4:20 p.m. PDT | Room A106
WERC Contributors: Robert Klinger*, Otto Alvarez, Matthew Brooks, J.R. Matchett

Alpine meadows provide critical habitat for many mammal species but are considered to be particularly vulnerable to climatic shifts. It is generally assumed that the loss or alteration of alpine meadow habitat would be from drying and/or transitions to woody dominated (conifers and/or shrubs) communities, but data on the degree to which this is actually occurring is lacking for the Sierra Nevada mountain range. Therefore, our goal was to use aerial photos and remote sensing data to analyze changes in alpine meadow boundaries and condition (productivity and wetness) across three degrees of latitude and 2650 meters of elevation in the Sierra Nevada. The analysis of the aerial photos indicated that conifer density increased in areas associated with meadow ecotones but there was little encroachment within the meadows themselves. GIS layers on downscaled climate and vegetation productivity (the normalized difference vegetation index, NDVI) were analyzed to evaluate if there was evidence of temporal or spatial change in meadow productivity from 1990-2010. Generalized additive models indicated there was substantial temporal variability in meadow productivity but little overall change or even a slight increase over the last 20 years. Principal Components Analysis indicated that meadows that had decreased in productivity were mostly small ones in the southern part of the range. The results suggest that climatically related changes to vegetation may be ongoing in lower elevation zones in the Sierra Nevada but have not occurred to the same extent in higher elevation zones. Overall, habitat conditions for many mammal species in the alpine zone of the Sierra Nevada are spatially and temporally variable but do not appear to be deteriorating in quality or quantity.


Potential effects of extreme climatic events on California Black Rail breeding habitat
Wednesday, October 17, 2012 | 9:10 a.m. PDT | Room A103-A104
WERC Contributors: Vivian Thuy-Vy Bui*, Karen Thorne, Kevin Buffington, John Takekawa

Coastal salt marshes are projected to be disproportionately impacted by climate change, including increased storm frequency and intensity. The San Francisco Bay estuary is one of the largest remaining tidal marsh complexes in California and contains important habitat for federal and state-listed wildlife species. Our objective was to evaluate the potential impact of storms on California Black Rail (Laterallus jamaicensis coturniculus) breeding habitat. In 2005-2006, we calculated home ranges of individual rails through radio telemetry. In 2010, we conducted elevation and vegetation surveys in these same areas to characterize rail habitat. Water level loggers were deployed to capture local hydrograph information and then hindcasted to 2005- 2006 to create site-specific tidal datums. By combining tidal information, habitat data, and rail home ranges, we were able to measure inundation risk of rail habitat during the breeding season. We found the frequency of stormy days in April to be increasing in our study area over the past few decades (R2=0.22), indicating that storm frequency and overall tide levels may continue to rise during times of rail nesting efforts. A storm in April 2006 was of sufficient intensity to inundate dominant marsh vegetation by an additional 20%, reduce overall availability of the marsh to rails, and increase the frequency of tide levels above mean rail nest height when compared to the previous year. Water level monitoring in 2010 and 2011 at three additional sites indicate that there was a 7-fold increase in the percent of habitat inundated during the March 2011 storm compared to periods of non-storm high water. In the face of changing climate patterns and resulting increased inundation of coastal salt marsh habitats, land managers will need to take into consideration availability of adjacent upland refugia during the breeding season when developing conservation strategies for threatened salt marsh species.

Reconstructing time-specific diet composition in greater sage-grouse chicks using feather stable isotopes
Wednesday, October 17, 2012 | 1:30 p.m. PDT | Room A103-A104
WERC Contributors: Erik Blomberg*

Diet during early stages of growth can have profound direct effects on young birds, as well as carryover effects to later life stages. Many diet studies for precocial birds rely on sampling crop contents; an approach that is limited because it is lethal to the animal and only provides a snapshot of diet that cannot be connected to other values of interest (e.g., survival). We developed a novel, non-lethal method for analyzing diets of greater sage-grouse (Centrocercus urophasianus) chicks using stable isotope composition of feather tissue, which allowed us to quantify contributions of diet items and reconstruct a post-hatch dietary timeline. We collected secondary feathers from greater sage-grouse chicks in Eureka County, Nevada, at 28 days of age. Feathers were sectioned into subsamples that corresponded to sequential multiday periods, isotopic composition of carbon (δ13C) and nitrogen (δ15N) was analyzed for each subsample, and Bayesian mixing models were used to estimate the relative contributions of invertebrate versus plant materials to diet during 1-week periods. We found δ15N to be a robust predictor of diet composition, whereas results for δ13C were more ambiguous. Bayesian mixing models using δ15N estimated the mean contribution of invertebrates to chick diet as 33 ±6% for week 1, 23 ±3% for week 2, 17 ±3% for week 3, and 14 ±0.3% for week 4. These results are consistent with previous studies that suggest a shift to a greater herbivory as chicks aged. We also show chicks that maintained a more intermediate diet were larger at 28 days, compared to individuals that consumed greater proportions of either plants or invertebrates throughout growth. These methods are well-suited to dietary assessment for grouse, and provide a new tool for evaluating sage-grouse response to management that can complement studies of habitat use and survival.

POSTERS

Survival rates of Pacific Coast band-tailed pigeons determined using VHF and satellite telemetry
Monday, October 15, 2012 | 8:00 a.m. - 12:00 p.m. PDT | Hall A | Poster 28
WERC Contributors: Michael Casazza, Peter Coates, Cory Overton, Brian Halstead

The Pacific Coast population of band tailed pigeons (Patagioenas fasciata monilis) has been in decline since the mid 1960’s. Currently, the only population estimates for survival rates were derived from banding efforts primarily during the 1960’s and 1970’s. The highly migratory nature of this species makes conventional telemetry alone impractical for assessing annual mortality rates. The use of solar powered satellite platform terminal transmitters hereafter PTT’s, on pigeons marked during the breeding season in California, Oregon, Washington and British Columbia, (n=31, 2006-2008) allowed for both seasonal and annual estimates of survival. We compare those estimates to estimates obtained from VHF radio-marked individuals in California (n=54, 1999-2000). We developed an a priori list of candidate models with covariates suspected of affecting band-tailed pigeon survival including body condition, season, sex, and age. No mortalities were observed during a 13-week breeding period and no indication of hunting related mortality was observed. Using ‘RMark’ in Program R we calculated an annual survival rate of 67.8% ± 12.0 (41.0 - 85.4) for PTT marked band-tailed pigeons during 2006-2008. Weekly survival rate was 99.2% ± 0.3 (98.2 - 99.7). The model that included a breeding period covariate was the top model of the a priori set. Derived survival estimates for breeding, fall migration, winter, and spring migration periods were 100%, 87.3%, 91.0%, and 87.1%, respectively for the PTT marked birds. Of 54 VHF monitored band-tailed pigeons, we observed 3 mortalities and found the survival rate during the breeding period (15 June - 07 September) of 1999 and 2000 was 95.0% ± 3.5 (82.0 - 98.7). Lower survival during the winter, and fall and spring migration periods is likely contributing to declining pigeon abundance and future research should focus on those periods.

Use of artificial floating islands for nesting and refuge by the endangered California clapper rail
Monday, October 15, 2012 | 1:00 - 5:00 p.m. PDT | Hall A | Poster 22
WERC Contributors: Cory Overton, Vivian Thuy-Vy Bui, John Takekawa

Projected changes in sea-level indicate increasing risks to obligate tidal marsh species such as the endangered California clapper rail (Rallus longirostris obsoletus). Increased inundation of primary habitats threatens to lower both survival and reproduction through intermittent loss of refuge cover and flooding of nesting substrate. We provided artificial refuge cover and nesting substrate for California clapper rails using floating islands and assessed use with respect to tidal inundation patterns and availability of natural refuge cover. Artificial refuge islands (1.5x2 m) were deployed at 10 locations in Arrowhead Marsh in Oakland, CA during the winter of 2010-11 and 2011-12. Artificial nesting islands (0.5x0.66 m) were deployed at 25 locations within each of 5 marshes in South San Francisco Bay from February 2012 to August 2012. All islands were constructed of recycled plastic materials with woven palm screens attached to rigid support poles providing cover. Waterproof, digital cameras mounted on each island recorded time elapsed images to monitor use. Nearly three million photos showed that rails regularly occupied all of the artificial refuge islands. Use of refuge islands by clapper rails increased slowly with tide level until tides reached 1.5 meters. Above 1.5 meters the odds of island use were 23 times the odds of use below 1.5 meters. Island use was also strongly diurnal. Odds of daytime use were 320 times the odds of use during the evening. Artificial nesting islands were used both for refuge during tidal inundation and as nesting substrate. Our results suggest that clapper rails in San Francisco Bay are strongly attracted to artificial habitats which provide protection from tidal inundation. Artificial islands for California clapper rails may be a timely and effective method for providing relief from adverse impacts of limited refuge and nesting habitat and could inform future habitat management under sea level rise scenarios. 


Effects of wetland management on carrying capacity of duck and shorebird benthivores in a coastal estuary
Monday, October 15, 2012 | 1:00 - 5:00 p.m. PDT | Hall A | Poster 29
WERC Contributors: Arrianna Brand, John Takekawa, Joel Shinn, Tanya Graham, Kevin Buffington, Sarah Spring, Keith Miles

With loss of natural wetlands, managed wetlands are increasingly relied upon to support energy requirements for wintering shorebirds and waterfowl. Despite substantial research effort on energetics in freshwater systems, less is known about invertebrate prey for waterbirds in coastal estuaries such as managed ponds in San Francisco Bay that form part of the largest wetland restoration in western North America. We assessed energy requirements for 8 shorebird and 4 diving duck species during the winters of 2007-2010 in 4 types of managed ponds - seasonal and circulation ponds each in two salinity classes - based on bird abundances, days in the season, and allometric equations that incorporate body mass and assimilation efficiency. We estimated energy available to birds on the basis of preferred prey abundances, biomass, energetic content, and prey accessibility. We then applied a daily-ration model to evaluate the carrying capacity of managed ponds for waterbirds as potential bird-days that a site can support. An average of 35,450 ± 1,559 (mean ± SE) and peak of 45,458 ± 1,653 diving ducks, and an average of 64,253 ± 14,838 and peak of 108,171 ± 4,854 shorebirds, used the study area at high tide through the winter. Greatest shorebird abundances were in seasonal ponds and greatest diving duck abundances were in circulation ponds. Prey energy varied by pond type, with mesohaline (< 30 psu) circulation ponds providing the greatest energy per unit area and total energy available for diving ducks and shorebirds, despite only 11% of the area accessible for shorebirds. Potential bird-days per ha was highest in mesohaline circulation ponds followed by low-hyperhaline (30-80 psu) circulation ponds and then medium-hyperhaline (80-150 psu) seasonal ponds, although bird-days were highly variable in the latter. Our results suggest that pond management for depth and salinity could substantially enhance energy available for these waterbirds. 


Greater sage-grouse and energy development: a decision support analysis for guiding the size and location of “no surface occupancy” zones
Monday, October 15, 2012 | 1:00 - 5:00 p.m. PDT | Hall A | Poster 35
WERC Contributors: Peter Coates, Michael Casazza, Cory Overton, Brian Halstead, Brianne Brussee

The development of anthropogenic features, especially those related to energy resources, in sagebrush ecosystems is an important concern to greater sage-grouse (Centrocercus urophasianus) populations among developers, conservationists, and land managers. Sage-grouse are highly dependent on different aspects of the sagebrush ecosystem to meet seasonal life-phase requirements. Research indicates that anthropogenic structures adversely affect sage-grouse within seasonal use areas. Land management agencies have implemented “no surface occupancy” zones of development at lek sites (breeding grounds) to prevent unfavorable effects. However, rationale for their decisions related to zone size and location are often challenged. To help inform this issue our study objective was twofold. First, we conducted an iterative spatial analysis to estimate optimal size of a zone centered on lek sites using seasonal utilization distributions of 228 sage-grouse (12,852 telemetry locations) across five populations within Mono County, California, during 2002 - 2009. We found season and site variation in fractional exponential curves that represented relationships between zone size and amount of utilization distribution. We summarized the diminishing returns of increasing zone size, which allows flexibility in land management decisions. The optimal size of a zone was 8,170 ha (20,188 acres), which equated to a buffer distance of 4.1 km (2.5 miles) based on lek configuration. We then evaluated whether or not leks represent the best locations for these zones. We compared values derived from iterations at leks with those from sites assigned randomly in other seasonal use areas. Centering zones at leks maximized the amount of utilization distribution by zone size seasonally and annually. Although energy development is not currently considered the primary threat to sage-grouse in Mono County, these analyses offer empirical support for decisions by developers and land managers that share an interest in meeting energy demands while conserving sage-grouse populations within the Great Basin.


An evaluation of trap modifications to improve capture success of giant gartersnakes
Tuesday, October 16, 2012 | 8:00 a.m. - 12:00 p.m. PDT | Hall A | Poster 24
WERC Contributors: Glenn Wylie, Brian Halstead, Michael Casazza

Giant Gartersnakes (Thamnophis gigas) are endemic to the Central Valley of California and are federally and state listed as threatened because of extensive loss of their wetland habitat. To inform recovery actions for this species we developed capture methodologies including floating inverted cone traps, but these yielded low capture probabilities. To improve capture success we evaluated eight different trap modifications related to construction material (galvanized or vinyl coated metal mesh), floating depths, funnel modifications, and baiting of traps. From 18 April to 8 September 2011 we examined the effects of trap modifications on snake captures using four transects of 24 traps each at a site known to have a population of T. gigas. We placed two transects in canals and two in a restored wetland to sample habitats representative of the majority of sampling conditions. Within each transect we deployed three replicates of each trap type placed in random sequence. We captured 72 T. gigas in 13,076 trap days. We analyzed the number of T. gigas captures in each trap type using Bayesian analysis of log-linear models. Unmodified vinyl and galvanized traps, and galvanized traps floating lower in the water with cable ties, captured fewer giant gartersnakes than predicted under the null model. Galvanized traps with cable ties had 9.13 (2.47 - 25.98) times as many captures as unmodified vinyl traps and 5.29 (1.92 - 12.73) times as many captures as unmodified galvanized traps. Appropriately modified traps can improve capture probabilities for T. gigas and, perhaps, other aquatic snakes. Improved probability of capture and detection will result in more accurate estimates of occupancy, abundance, survival, recruitment, and other demographic parameters to better inform recovery action planning for T. gigas. 


Effects of current and historic habitat conditions on the occurrence of the giant gartersnake in the Sacramento Valley of California
Tuesday, October 16, 2012 | 8:00 a.m. - 12:00 p.m. PDT | Hall A | Poster 56
WERC Contributors: Brian Halstead, Glenn Wylie, Peter Coates, Michael Casazza

We examined the effects of current and historic habitat on the distribution of the giant gartersnake (Thamnophis gigas) in the rice-growing regions of the Sacramento Valley, California. We used a removal design and Bayesian analysis, with and without informative priors on detection, to fit and evaluate single-season occupancy models. We detected the giant gartersnake at 6 of 17 sampled sites. Detection probabilities were very low, but were higher with increased water temperature, with increased sampling effort, and when sampling occurred earlier in the giant gartersnake active season. The probability of occurrence of the giant gartersnake in 25 km2 cells in the Sacramento Valley was approximately 64% using uninformative priors on detection probability and approximately 49% using priors for detection probability informed by previous surveys conducted from 2003 through 2009. Models with informative priors indicated that the giant gartersnake was more likely to occur near historic tule marsh habitat, suggesting that limited dispersal from historic locations might affect the distribution of the giant gartersnake.


The role of birth synchrony as a mechanism for an Allee effect in a translocated pronghorn population
Tuesday, October 16, 2012 | 1:00 - 5:00 p.m. PDT | Hall A | Poster 36
WERC Contributors: Diego Johnson, Kathleen Longshore, Chris Lowrey 

The translocated population of pronghorn (Antilocapra americana) on the Carrizo Plain National Monument in California has experienced a 20-year history of population decline, where roughly 13% of the original number of individuals currently remains within the monument boundaries. Previous research in other regions indicates that pronghorn population trends may be strongly affected by high postnatal fawn mortality due to predation. Although pronghorn normally exhibit birth synchrony to offset high predation, in small populations this strategy may be ineffective and could produce an Allee effect. An Allee effect is a positive correlation between population growth rate and population density, where decline below a critical threshold greatly increases the probability of extinction. Here, we quantify fawn survival and test for conditions that may produce an Allee effect by examining relationships between birth synchrony and predation. During 3 field seasons we outfitted 20 newborn fawns (approximately 50% of all fawns born) with lightweight breakaway GPS collars. Our analysis showed no evidence of birth synchrony and no relationship between dates of birth and survivorship. Fawn mortality was relatively high and ranged from 70-83%. Sixty-four percent of necropsies performed indicated that fawns died due to predation, 18% to internal infection, and 18% of unknown causes. In addition to addressing the role of birth synchrony as a mechanism for an Allee effect, we further assessed factors affecting fawn survival by examining whether differences in the behavior of individual fawns (i.e. movement patterns and habitat selection) influenced survivorship. Wildlife translocations are a critical component of repopulation efforts and our results provide information about the importance of population size on successful reintroductions. 


Habitat interaction on a sky-island between two species of chipmunk in the Basin and Range Province of Nevada
Wednesday, October 17, 2012 | 8:00 a.m. to 12:00 p.m. PDT | Hall A | Poster 25
WERC Contributors: Chris Lowrey, Kathleen Longshore

Interspecies interactions can affect how species are distributed, put constraints on habitat expansion, and reduce the fundamental niche of the affected species. Using logistic regression, we analyzed and compared 174 Tamias palmeri and 94 T. panamintinus within an isolated mountain range of the Basin and Range Province of southern Nevada. Tamias panamintinus was more likely to use Pinyon/Ponderosa/Fir mixed forests than Pinyon alone when compared to random sites. In the presence of T. palmeri, however, interaction analyses indicated T. panamintinus was less likely to occupy the mixed forests. This species by habitat interaction data suggests Tamias palmeri excludes T. panamintinus from areas of potentially suitable habitat. Climate change may adversely affect species of restricted distribution. Habitat isolation and species interactions in this region may thus increase the survival risks of a major prey source as climate temperatures rise.