Avian Ecology & Managment II

Contributed Paper
ROOM: CC, Room 15
SESSION NUMBER: 64
 

1:10PM Assessing Avian Use at the First Freshwater Offshore Wind-Energy Project in North America
Jennifer H. Stucker; Jason D. Carlisle; Diem Pham
As the demand for energy increases, wildlife managers often evaluate the effects of energy production on wildlife to meet legal and permitting requirements. To date, land-based wind-energy facilities in the U.S. have been developed generally following a prescribed set of pre- and post-construction wildlife assessments focused on risk and estimating mortality post-construction. Wind development in marine offshore areas followed a tiered pre-construction process focused on avoidance and minimization. Icebreaker Wind, the first freshwater offshore-wind project in North America, has been proposed for Lake Erie, approximately 15 km northwest of Cleveland, Ohio. This DOE supported demonstration project includes six turbines (20.7 megawatts). Our design and survey objectives were to design a robust pre- and post-construction survey approach to characterize avian abundance and distribution within the project area, incorporating a gradient sampling design to detect changes in wildlife distribution post-construction. The surveyed area (5-km buffer of proposed turbines) encompassed 145 km2 within Lake Erie, and we surveyed seven aerial transects through the project area and two opportunistic transects nearer to shore every two weeks from October 2017 – May 2018. We used a double observer mark-recapture distance-sampling approach to estimate avian abundance. Through April 2018 we documented 15 species within the project area, with gull species accounting for 70% of observations. In contrast, off-project transects nearer to shore have included additional species and greater abundance. The average group size by species was 4.3 birds (SD = 16.2) in the project area, and 9.3 birds (SD = 36.0) outside the project area. No eagles or other raptors have been observed during surveys. Pre-construction results suggest that the offshore avian community within the project area differs from the adjoining near-shore community. Data collection and analysis challenges in this Great Lakes environment include low densities of wildlife, reliable distance estimation, and accounting for ice.
1:30PM The Effect of Season of Prescribed Fire on Breeding Bird and Plant Communities in Minnesota Lowland Brush Ecosystems
Annie Hawkinson; Rebecca Montgomery; Lee Frelich; Charlotte Roy; Lindsey Shartell; Lori Knosalla
Lowland brush ecosystems (LBEs) are disturbance-dependent and make up approximately 3.5 million hectares in Minnesota. LBEs support bird communities that are equally or more diverse than nearby forested landscapes and provide habitat for 38 avian Species in Greatest Conservation Need (SGCN) in Minnesota. Previous studies in LBEs have suggested that wildlife diversity is related to a heterogeneous mix of shrub and herbaceous cover. This characteristic patchiness was historically maintained by fire occurring in spring, summer and fall seasons. Mimicking historical disturbance regimes with prescribed fire is difficult due to variability in weather and safety considerations to execute fires effectively. Thus, most prescribed fires are conducted in spring when conditions are most often conducive to burning, and managers rely on additional methods such as mowing and shearing to maintain LBEs in earlier successional states. Yet, prescribed fire in summer and fall may produce unique outcomes that are important for maintaining and enhancing bird and vegetation diversity. We compare bird and plant community responses to spring, summer and fall prescribed fires in LBEs at two 160-ha sites. We sampled birds at 64 point count locations and vegetation at 128 fixed-radius plots in a Before-After-Control-Impact sampling design, with each site divided into 3 fire season treatments and a control. We developed best-fitting models to predict bird species diversity and species-level abundance and occupancy, while accounting for detection probability. We also predicted baseline breeding bird information from plant metrics and modeled shifts in these bird indices. At both community and species levels, breeding birds varied in their responses to fire season treatments, with some species increasing in abundance, some decreasing, and some not responding, as compared to controls. The effect of prescribed fire season is important to incorporate into management planning to address the needs of the full suite of SGCN in Minnesota LBEs.
1:50PM Population Estimation, Habitat Attributes, and Nesting Success of Mew Gull Nests on Fort Wainwright, Alaska
Garrett Savory; Emily Richmond
Mew gulls (Larus canus) nest on the Main Cantonment of US Army Garrison Fort Wainwright in interior Alaska every spring and summer in high numbers. They regularly nest on buildings, military vehicles, and other problematic places. Additionally, mew gulls are a serious strike hazard to aircraft using Ladd Army Airfield. The objectives of our study were to estimate the number of mew gull nests occurring on the Main Cantonment, determine which habitat features influenced nesting locations, and estimate nesting success. To estimate nest numbers, we surveyed the Main Cantonment in May of 2016 and 2017 by using double sampling methods. The Main Cantonment was divided into 96 plots of 400 m2. Each year, 48 plots were randomly selected for an initial survey and then 24 of the initial survey plots were randomly selected to be resurveyed. We estimated there were 91 (84-99 95% CI) nests in 2016 and 68 (56-80 95% CI) nests in 2017. Most gull nests were present on buildings, located in areas with human activity, avoided nesting in residential areas, and were present in loose colonies. To estimate nest success we monitored gull nests to determine their fate. Nests were visited two times per week to determine their status until success or failure. A nest was considered successful if at least one egg hatched one chick. We used a nest survival model in Program Mark to evaluate multiple models built from several factors to estimate nesting success probability. From preliminary analysis, we found that gulls nested on buildings and structures had higher nest success than those that nested on the ground. Our results inform natural resource managers how to manage habitat on Main Cantonment of Fort Wainwright to discourage mew gulls from nesting at problematic sites.
2:10PM Translocation of Lesser Prairie-Chickens: Does Lek Presence Limit Dispersal?
Liam Berigan; Daniel Sullins; David Haukos; Kent Fricke; Jonathan Reitz; Liza Rossi; Kraig Schultz
Prairie grouse translocation is routinely complicated by the taxa’s tendency to disperse incredible distances after release. Minimizing post-release dispersal is critical to ensure translocated individuals join the population that managers are attempting to reestablish. One of the primary factors assumed to be limiting the dispersal of translocated prairie grouse is conspecific attraction, specifically the presence of a nearby lek during release. A multi-agency effort seeks to augment and increase lesser prairie-chicken (Tympanuchus pallidicinctus) populations within the Sand Sagebrush Ecoregion of southwestern Kansas and southeastern Colorado. Lesser prairie-chickens were captured in the Shortgrass Prairie/CRP Mosaic Ecoregion of northwestern Kansas, equipped with VHF radio-collars or SAT-PTT GPS transmitters, and translocated to the Cimarron and Comanche National Grasslands. Releasing translocated birds at multiple release sites near several established or recently occupied leks provided us the opportunity to test the effect of lek presence on post-release dispersal. The project began with a fall 2016 release of 26 males onto the National Grasslands to establish or augment several leks. In spring 2017, we released 45 additional males and 38 females onto the National Grasslands. Apparent survival for the six months following release was 40% and 44% of surviving females raised at least one chick to 35 days post-hatch. Despite extensive post-release dispersal (mean = 18 km, sd = 16 km), 75% of individuals stayed within the immediate vicinity of the National Grasslands (≤24.8 km of the release site). Fall and spring-released males attended 10 leks in spring 2017, 5 of which hadn’t been previously occupied. Attendance at these leks has fluctuated widely since, and we are continuing to supplement them with annual male releases. The release effort continued in spring 2018, with >100 individuals translocated. Leks appeared to aid in retaining translocated lesser prairie-chickens by reducing the probability of dispersal following release.
2:30PM Refreshment Break
3:20PM Osprey Habitat Suitability at the Rural-Urban Interface in Central Idaho
Elizabeth Meisman; Tyrell Styhl; Andrew Dorrell; Bryan Krouse; Marc Bechard; Robert Miller; Dusty Perkins
Ospreys (Pandion haliaetus) are fish-eating, top predators of aquatic ecosystems that serve as useful sentinel species for monitoring environmental contaminants and ecosystem health. Ospreys further appear highly adaptable to human-dominated landscapes and readily nest on artificial structures that occur within an array of land use and land cover (LULC) types and human settlement regimes. In Long Valley, Idaho, the abundance of breeding Ospreys has remained relatively stable since the late 1970’s yet the distribution of nests and nest substrate use has changed. To evaluate if changes in nest structure availability, coupled with increasing anthropogenic landscape conversion, could be changing osprey nest site selection and success, we evaluated relationships among nest site characteristics and osprey breeding activities. We used multivariate generalized linear models with model selection procedures to evaluate the relative importance of LULC composition and nest site characteristics associated with nest occupancy and daily survival rates. Preliminary results from 2015 – 2016 demonstrate changes in available nesting substrates compared to historic data. Our models also suggest that distance to water and roads are correlated with patterns of territory occupancy while distance to water and land cover types (water, agricultural and grassland) are correlated with nesting success. Here, we present breeding population monitoring data, trends in available nesting substrates, and output from habitat suitability models for the 2015 – 2017 breeding seasons. Additionally, we discuss the implications of changing land use patterns and substrate availability on osprey breeding ecology and management.
3:40PM Developing a Species Distribution Model for Loggerhead Shrike to Prioritize a Citizen Science Survey Effort.
Stephen F. Spear; Rose Caplan; Amy Chabot
The Loggerhead shrike (Lanius ludovicianus) is a carnivorous songbird currently in sharp decline, particularly in the northeastern portion of the United States and Canada. The species relies on grassland habitat with interspersed shrubs and trees. The Loggerhead Shrike Working Group was formed to help reverse the trend of shrike declines through surveys, habitat improvement, and conservation breeding. A key strategy for the working group is to assemble a “Shrike Force” of citizen scientists to identify undocumented populations of shrike. Our objective was to develop a model of likely occurrence of shrike within the northeastern United States that could be used to set survey routes for the Shrike Force. We used the modeling platform MaxEnt, which is commonly used for predicting species occurrence and only requires presence points. Due to the small number of reliable observations in most northeast U.S. states, we first built a model based on intensive surveys in Ontario and projected that model to several northeastern U.S. states within the range of loggerhead shrike. In these models, we also tested a number of variables related to NDVI and habitat neighborhood. This initial model was an accurate prediction of shrike habitat in Ontario, but did not provide a consistent projection to the U.S. states. The primary reason for the poor projection was a high influence of climatic variables in which the range within Canada was outside the range within the U.S. projection. As a result, we developed a final model that incorporated both observations from Ontario as well as from sites within the U.S. with consistent survey effort, and that largely excluded climatic variables in favor of those related more directly to physical habitat and productivity. Our next step is to select sampling sites for the Shrike Force, which will also serve as validation for our final model.
4:00PM Harlequin Duck Occupancy and Landscape-Scale Habitat Associations on National Forest Lands in Washington.
Peter Singleton; Lexine Long
Harlequin ducks (Histrionica histrionica) are a migratory sea duck that is listed as a “Sensitive Species” in their breeding range on federally managed forest lands in Oregon and Washington. The US Forest Service conducted systematic surveys for harlequin ducks in Washington in 2013 and 2014. We used a combination of species distribution modeling and occupancy analysis to evaluate the distribution and occupancy rates of harlequin ducks on National Forest lands in Washington. Our goals were to; 1) identify landscape-scale habitat characteristics correlated with harlequin duck detection locations, 2) estimate harlequin duck occupancy and identify correlated habitat characteristics, and 3) evaluate the effect of survey timing on harlequin duck detection probability. A total of 353 km of streams were surveyed in 2013 and 119 km in 2014. Compared to random points taken from surveyed reaches, detections were: at lower elevations; closer to open roads; surrounded by less medium-large tree conifer forest, more small tree conifer forest, and more broadleaf forest; surrounded by more canyons and steep valleys; and surrounded by less area with moderate solar radiation and more area of cold solar radiation exposure. Classification tree analysis showed that elevation, stream flow, and area of medium-large tree conifer forest within 500m were best for distinguishing between harlequin duck detection locations and from random points in the surveyed area. Our occupancy modeling highlighted that detection probability was generally higher during breeding season (April 15 to June 1) surveys than during post-fledging (July 1 to Aug 31) surveys, and declined over the summer. Probability of harlequin duck occupancy was higher in sample units with more gentle stream gradient, moderate elevation, more canyon landform, less medium-large tree conifer forest, more small tree conifer forest, and moderate stream flow.
4:20PM Land Cover Types Associated with Kestrel Occupancy of Nest Boxes in Southeast Ohio
Lucille M. Williams; Donald P. Althoff; Robert L. Hopkins II; Henry J. Barrows III
The American kestrel (Falco sparverius), a secondary cavity nester like many obligate cavity-nesting songbird species, has benefitted from nest boxes installed at locations assumed to meet all other habitat requirements. Nest boxes mounted on billboard posts and the backs of road signs—especially along 4-lane highways—are becoming more common place because of the perceived cover mix associated with these landscapes. Based on monitoring kestrel occupancy of nest boxes (n = 37) from 2015-2017 in southeast Ohio, we were able to create a database of occupied vs. unoccupied sites, as well as documenting the presence / absence of European starlings (Sturnus vulgaris) – an apparent competitor for nesting cavities. Using this database, we conducted a GIS-based analysis of land cover across multiple scales for each site, with the objective of identifying a suite of landscape cover features that might enable deployment of nest boxes in the region with greater usage by kestrels. Using principal components analysis coupled with k-means clustering we identified two landscape niche clusters, which differed significantly along the PC1 axis (p = 0.000). One cluster was occupied by kestrels and starlings, the other was occupied by starlings only. The kestrel-occupied cluster had 10-40% herbaceous land and < 15% cropland across all spatial scales, with a high percentage of forest being important at the largest scale. A secondary predictor within the kestrel-occupied cluster was whether starlings established occupancy before kestrels in a given year. The starling-only cluster had a high percentage of cropland at all buffer scales and less forest cover. Future efforts should take these trends into consideration to effectively deploy nest boxes intended for kestrels and minimize occupancy by European starlings.
4:40PM Comparative Analysis of Parental Effects on Developmental Period Length in Anseriform Birds
Daniel C. Barton
Explaining variation among species in the expression of life history traits is a key goal of evolutionary ecology that provides insight into conservation strategies and limits to population growth. A wide range of life history variation is represented within waterfowl (Anseriformes) including more than an order of magnitude in body and egg size variation, and nearly a month of variation in developmental period length. Developmental period length in altricial birds is a key trait thought to explain offspring quality, and is also thought correlated with numerous other traits including size, longevity, predation risk, and parental attentiveness, yet these ideas are largely untested across the wide diversity of precocial waterfowl. Here I use data on waterfowl compiled from ca. 470 sources to test several alternative explanations for among-species variation in developmental period length in a phylogenetic-comparative context. Developmental period length, as found using phylogenetic least squares regression, was related to body size, nest predation length, and attentiveness across a wide diversity of waterfowl taxa. Cavity-nesting waterfowl had longer incubation periods independent of their body size, and incubation period length increased in species with lower attentiveness. These results are consistent with the hypotheses that developmental period lengths evolve in concert with both predation risk and parental attentiveness, after controlling for the influence of body size. I found ambiguous support for the hypothesis that large eggs may provide advance provisioning for egg neglect. Further directions for research on the causes and consequences of variation in waterfowl developmental period lengths include applied and basic research into the consequences of the maternal environment for offspring quality within and among species.
12:50PM Fragmentation Affects the Demography of an Exploited Ground-Dwelling Bird
Paige E. Howell; Seth Wood; Theron M. Terhune; James A. Martin
Habitat fragmentation is often implicated as a major factor leading to increased extinction risk. However for many prey species, increasing fragmentation may actually decrease extinction risk by reducing predation risk through increasing availability of prey refugia. We tested the hypothesis that fragmentation positively influenced population viability using the northern bobwhite (Colinus virginianus), an imperiled facultative grassland species. We quantified the influence of habitat fragmentation on demography by experimentally manipulating the level of habitat fragmentation (none, moderate, and high) of bobwhite habitat patches. We randomly applied these treatments to each patch by mowing patches during October (post-breeding season) each year. From October 2013 until September 2016, we tracked radio-tagged bobwhite weekly, and recorded their fate (alive, dead). We estimated breeding (April 15-September 30) and non-breeding (October 1-April 14) season survival using an integrated joint mark-recovery, mark-recapture model that allowed us to estimate survival, detection of live individuals and recovery of recently dead individuals in the same modeling framework. Non-breeding season survival was almost twice as high in our control patches as in our medium and high fragmentation patches. However, breeding season survival was higher in the fragmented patches compared to control patches, and autumn bobwhite densities were similar among treatments. These results suggest that lower survival rates in more fragmented patches during the non-breeding season, were compensated for by higher survival rates in these same patches during the breeding season. One plausible explanation is that survival is density dependent; as population density declines during the non-breeding season, survival probability of the remaining individuals is higher. As a result, there is little difference in population density among treatments following the breeding season. Under these conditions, the population trajectory was resilient to fragmentation and demonstrates the importance of capturing the entire life-cycle of a species to avoid spurious conclusions.

 

Contributed Paper
Location: Cleveland CC Date: October 10, 2018 Time: 12:50 pm - 5:00 pm