Waterfowl Conservation & Management I

Contributed Oral Presentations

Contributed paper sessions will be available on-demand for the duration of the conference, then again at the conclusion of the conference.


Climate-Driven Shifts in Prairie Pothole Wetlands: Assessing Future Impacts to Critical Waterfowl Habitats
Owen P. McKenna; David M. Mushet; Samuel R. Kucia
The North American Prairie Pothole Region (PPR) is an expansive region that covers parts of five Midwestern states and three Canadian provinces. The region contains millions of wetlands that produce between 50-80% of the continent’s waterfowl population each year. Previous modeling efforts indicated that climate change would result in a shift of suitable waterfowl breeding habitat from the central PPR to the southeast portion of the region where over half of wetlands have been drained. The implications of adopting these projections would require a massive investment in wetland restoration in the southeastern PPR to sustain migratory waterfowl populations at harvestable levels. We revisited these projections using a newly developed model for simulating prairie-pothole wetland hydrology in combination with the most up-to-date climate model projections to estimate how future climate may impact the distribution of waterfowl-breeding habitat. We also presented our findings in changes to wet May ponds, which is a metric that is used by managers at the US Fish and Wildlife Service to estimate waterfowl breeding populations to establish harvest regulations. Based on the output of 32 climate models and 2 emission scenarios we found a projected change in wet May pond numbers from -23% to +.02% when comparing the most recent climate period (1989-2018) to the end of the 21st century (2070-2099). We also found no evidence that the distribution of wet May ponds will shift in the future. These results suggest that management and conservation strategies for wetlands in the PPR that focus on areas with the high densities of intact wetland basins support large numbers of breeding duck pairs and will likely be the most successful in maintaining habitats critical to continental waterfowl populations.
Integrating Data from Aerial and Ground Surveys to Estimate Densities of Waterfowl
Beth E. Ross; Nicholas Masto; Christian Roy; Richard Kaminski; Jamie Dozier; Mark McAlister; Joseph Woods
Wildlife monitoring data can be challenging to incorporate into models for statistical and ecological inference because spatio-temporal scales and data quality may vary. Surveys to monitor waterfowl are particularly challenging in that waterfowl are migratory and occur throughout multiple ecoregions during their life cycles. Given their mobile nature, waterfowl are typically monitored based on continent-wide monitoring or surveying efforts (e.g., Waterfowl Breeding Population and Habitat Survey [WBPHS]). Our goal was to combine count data from aerial surveys with ground-based surveys to provide an example of how multiple data sources could be combined to estimate abundance of waterfowl during the WBPHS as well as along the coastal plain of South Carolina. In addition to estimating detection probability for waterfowl during the WBPHS, we also illustrate how data integration models can be used to understand habitat use, spatio-temporal distributions, and population dynamics of waterfowl during the non-breeding season at finer scales than the WBPHS. For the South Carolina data, our data integration model provided more precise estimates of abundance while adjusting for detection probability. When analyzed in a traditional N-mixture model framework, our aerial survey data from South Carolina resulted in low detection probability (p = 0.1) which lead to a large mean abundance with high CV (N = 6200, 95% CI = 3000-16500). When analyzed with ground crew surveys in the data integration model, our estimates of detection probability were larger (p = 0.45) with more precise and biologically reasonable estimates of mean abundance (N = 998, 95% CI = 915-1074). We illustrate how this approach can be scaled up to continent-scale aerial surveys such as the WBPHS and present results.
Linking Long Term Resource Monitoring Vegetation Data with Bioenergetics Needs of Waterfowl to Inform Wildlife Habitat Management
Kirsten Schmidt; Jacob Straub; Benjamin Sedinger; Stephen Winter
The Upper Mississippi River is a large and diverse ecosystem with impounded areas ideally suited for the growth of aquatic vegetation, especially wild celery (Vallisneria americana). Due to its high energy content wild celery winter buds are an excellent food source for migrating waterfowl, with some species like canvasbacks (Aythya valisineria) selecting for wild celery. Since 1998 the Upper Mississippi River Restoration Program, Long Term Resource Monitoring (LTRM) element has monitored aquatic vegetation by conducting rake samples. During rake sampling aquatic vegetation species are given a quantitative value based on species relative presence/abundance on a rake. There remain questions however, on how data collected by the LTRM translates to waterfowl habitat quality and bioenergetics. This study aims to discover if a relationship exists between LTRM rake scores and winter bud biomass estimates from substrate core samples in impounded areas of pools 4, 8, and 13 of the Mississippi River, and estimate kilocalories of wild celery winter buds available to waterfowl in areas open and closed to hunting. LTRM collected rake samples in summer 2018 in pools 4, 8 and 13. Substrate cores were collected at the same locations as LTRM rake sample sites before the annual waterfowl migration in fall 2018. LTRM rake scores for wild celery were positively related to wild celery winter bud counts for Pool 8 (P < 0.05, R2 = 0.38). Substrate core winter bud counts and documented bud weight and caloric values were used to estimate the total kilocalories in the sampling areas. In fall 2018 the Pool 8 open area to hunting had an estimated 16,800 kilocalories per acre compared to 28,000 kilocalories per acre in the closed area to hunting. These numbers translate to 42 canvasbacks use-days per acre in the open area, and 70 use-days per acre in the closed area.
Behavior-Specific Habitat Selection by Raccoons during the Waterfowl Nesting Season in the Prairie Potholes Region of Manitoba
Charlotte R. Milling; Stanley D. Gehrt
Logistical and analytical constraints of traditional analyses of habitat selection have required the assumption that purpose is homogeneous across contexts. Such an assumption can result in underestimation of the strength of selection or failure to observe selection, rendering subsequent strategies to manage populations ineffective. In the Prairie Potholes Region of Manitoba, raccoons (Procyon lotor) are an important nest predator of waterfowl, but it is unclear how raccoons use this landscape during the waterfowl nesting season. Our objective was to use high-resolution GPS telemetry to differentiate among behaviors by raccoons, evaluate behavior-specific habitat selection during the waterfowl nesting season, and describe the characteristics of wetlands used for different purposes. We collected ~ 35,000 locations from 22 animals during the 2018 and 2019 nesting seasons, amounting to 475 animal-night’s worth of movement data. We used hidden Markov models (HMM) to fit 3-state models to the movement trajectories, classified observations into discrete behaviors, and fit behavior-specific Random Forest resource selection functions to evaluate the relative importance of proximity to landscape features on habitat selection for resting, exploratory, and travelling behaviors. Proximity to a wetland was the most important variable contributing to selection for the resting and exploratory states, and the HMM predicted increasing probability of switching from the resting and exploratory states to a travelling state with increasing distance from a wetland. Even though half of our observations were classified as resting, raccoons used far fewer wetlands for resting than for exploratory behavior. Wetlands used for both purposes were similar in size and proximity to other landscape features. These results illuminate the complexity of habitat selection by an important predator in this patchy landscape, allowing managers to focus on landscape features where raccoons are having a disproportionate impact during the waterfowl nesting season.
Towards Developing a Methodology to Monitor Scoter Breeding Grounds
Amelia Cox; Scott Gilliland; Christian Roy
Scoter (Melanitta sp.) populations across North America have declined substantially since the 1980s for unknown reasons. Identifying causes of this decline proves challenging in part because scoters, like many sea ducks, are poorly monitored by long-running waterfowl surveys. Continental monitoring programs for waterfowl in North America were designed for dabbling ducks and so miss much of scoter species’ breeding ranges and are conducted too early to capture these late breeding birds. Using both helicopter plot and fixed-wing aircraft line-transect surveys along the border of Labrador and Quebec, Canada in mid-June, we assessed the feasibility of conducting surveys for late nesting waterfowl like scoters. The fixed-wing component of the survey used a distance sampling scheme sampling methods to calculate and correct for imperfect detection rates. We used a Bayesian hierarchical framework to analyze the data.. In mid-June, we detected approximately equal sex ratios and predominantly flocks of two, suggesting the timing of the survey likely occurred during egg laying, as intended. From the fixed-wing survey, we estimated the combined detection rate for the Black (M. americana), Surf (M. perspicillata), and White-winged Scoters (M. deglandi) at 38% (95% credible interval: 33-45%) with a combined breeding density of 0.11 indicated pairs per km2 (0.08-0.14). Due to the low detection rates, implementing distance sampling methods improved the density estimates by approximately doubling the estimated density compared to raw counts for all waterfowl species. Although accounting for detection improved density estimates for all species, final density estimates from the fixed-wing aircraft component were 25 to 80% lower than estimates from the helicopter component. Operating fixed-wing aircraft in remote areas is considerably more cost effective than helicopters; the next steps will be to reconcile and integrate data from the two survey platforms to provide a cost effective monitoring program for scoters and other late-nesting waterfowl.
Evaluating the Physiological Response of Sub-Lethal Infections of Trematodes in Captive Lesser Scaup
Cheyenne R. Beach; Rebecca A. Cole; Joseph D. Lancaster; Aaron P. Yetter; Auriel M. V. Fournier; Heath M. Hagy; Christopher N. Jacques
Since 2000, thousands of lesser scaup (Aythya affinis) die annually during migration through the Upper Midwest, USA from Cyathocotyle bushiensis (Cb) and Sphaeridiotrema spp. (Ss) intestinal infections after consuming infected exotic faucet snails (Bithynia tentaculata). Faucet snails have since reached Pool 13 of the Mississippi River raising concern of further spread to Pool 19, a critical mid-latitude stopover area for lesser scaup. Although not all trematodiasis infections result in mortality, we hypothesize that sub-lethal infection may affect subsequent migration and fitness through decreased body condition, indexed by a wide range of blood metabolites (e.g., triglycerides, glucose, albumin), white blood cells (WBCs), and total body measures (e.g., weight and temperature). We experimentally tested these physiological parameters in captivity with wild-caught and captive-reared lesser scaup. In July 2019, 21 wild-caught females received a single sub-lethal dose (x̄ = 96 Ss and x̄ = 169 Cb) of metacercariae while 16 individuals served as controls. In December 2019, 16 male and female captive-reared lesser scaup received a single sub-lethal dose (x̄ = 293 Ss and x̄ = 124 Cb) of metacercariae while 8 individuals served as controls. We collected blood, feces, body temperature and weight from all individuals prior to dosing (i.e., day 0), on day 5, and on day 10 when all birds were euthanized and necropsied. Preliminary results suggest that infected individuals had decreased blood biochemical concentrations of triglycerides, blood urea nitrogen, bilirubin, albumin, PCV, and overall decreased body condition, but increased concentrations of non-esterified fatty acids, eosinophils, basophils, and WBCs as well as elevated body temperature. We witnessed mixed results between trials in β-hydroxybutyrate and glucose concentrations. Infected individuals experienced a predicted physiological reaction to the treatment, but those in better body condition (i.e., captive-reared birds) experienced a less severe reaction.
Female Mallard Harvests, Harvest Rates, and Survival Before and after Changes in Daily Harvest Limits in Michigan
Dave Luukkonen; Connor Smeader; Trey McClinton; Scott Winterstein; Barbara Avers
Breeding population estimates for mallards (Anas platyrynchos) in Michigan increased from 1991 through 1998, reaching a peak of 523,000 birds. Mallards then began to decline after 1998, reaching a low of 200,000 birds in 2008, and have remained at low abundance in recent years. The Michigan mallard decline was nearly coincident with initiation of 60-day duck hunting season frameworks associated with adaptive harvest management (AHM) of mid-continent mallards. Michigan maintained a 1-hen daily limit since AHM began in 2005 through 2013 despite more liberal Federal Frameworks which allowed 2 birds daily. Previous research found that Michigan mallard abundance was following declining wetland hydrological conditions and there were no increases in mallard harvest rates or declines in survival associated with 60-day AHM seasons as expected if excessive harvest was suppressing the mallard population; so the State adopted a 2-hen limit in 2014 and has maintained those regulations since. We investigated changes in Federal mallard estimates for the 5-year periods before (2009-2013) and after (2014-2018) changes in regulations allowing 2 hen mallards daily. There was most support for a linear model indicating cohort-specific changes in harvest (AICc model weight = 0.77): harvest of male mallards remained relatively constant while female mallard harvest increased by about 24% after initiation of the 2 hen daily limit. Also, the annual estimates of the number of females harvested per male in the recent time period (range: 0.60 – 0.72 females/male) increased after a long-term declining trend (from 1965-2013) and estimates since 2014 have been the highest since 1998. We will update estimates of harvest rates and survival to test if there have been any detectable changes in these parameters since regulations changes initiated in Michigan in 2014. This study will help inform future waterfowl hunting regulations in Michigan.
Ecology of the Pacific Black Duck and Habitat Quality on the Island of Aunu’u, American Samoa
Marissa Kaminski; Flor Hernández; Joshua I. Brown; Mark MacDonald; Philip Lavretsky
Three subspecies of the Pacific Black Duck, Anas superciliosa, range throughout the Australasian Pacific. Among locations, those populations at the periphery of their distributions are least known about. Here, we study the ecology and assess habitat availability of the Pacific Black Duck in Aunu’u, American Samoa, where little is known about the species. First, time-budget analyses to determine daily life history events were coupled with habitat assessments to determine food quality and quantity. In short, biomass estimates were used to determine if there were ample plants preferred by dabbling ducks on the landscape. In one wetland, plant biomass estimates showed sufficient amounts of sedges while low biomass in other categories. In a second wetland, there was low plant biomass between all categories, but especially in plants most utilized as food by dabbling ducks. Results showed that all wetlands contained scores on the lower end for macro-invertebrate abundance, diversity, richness, and biomass. With these results, we conclude that Aunu’u does not provide an abundance of food resources that dabbling ducks would typically consume. Conducting time-budget analyses revealed that these Pacific Black Ducks adapted to hunt and consume small to medium sized tilapia. Attached backpack-style satellite transmitters on three Pacific Black Ducks further demonstrated that birds spent considerable time in the agricultural wetland and wetland containing large quantities of tilapia. Together, the lack of abundant calories may explain why these birds were 200-300 grams smaller than their counterparts in mainland Australia. Finally, comparing genetic variation from Pacific Black Ducks on America Samoa to references elsewhere in the Australasian pacific revealed American Samoan birds to be more closely related to A. s. rogersi. This is in contrast to expectations of these being A. s. pelewensis subspecies. Work is underway to further add important information to better conservation of this uniquely adapted duck.
Evaluating Duck Brood Use of Wetlands in Agriculturally-Dominated Landscapes in the Iowa and Minnesota Prairie Pothole Region
Blake Mitchell; Adam Janke; Kaylan Kemink
Alterations to the Prairie Pothole Region (PPR) through wetland drainage, grassland conversion, and agricultural expansion have changed the capacity of the landscape to support breeding waterfowl. While understanding these changes has been the focus of waterfowl research for decades, few studies have examined the extent to which intensively cropped landscapes support and sustain duck production. Our research aims to assess wetland use and selection by duck broods in agriculturally-dominated landscapes in the Iowa and Minnesota PPR. We conducted aerial brood surveys using thermal and RGB sensors attached to a quadcopter drone. Drones increased our detection rates substantially over standard ground-based brood surveys, and as such, this technology allowed us to conduct repeat-visit brood surveys at the landscape-scale with high confidence in our estimates of brood abundance. During the 2019 field season, initial detection probabilities were 0.80 and 0.87 for the first and second round of surveys, respectively. We also sampled a random subset of wetlands for aquatic invertebrates and water chemistry after brood surveys were conducted. Preliminary t-test findings regarding aquatic invertebrates suggested that invertebrate biomass was greater in occupied wetlands. However, non-metric multidimensional scaling (NMDS) ordination analyses indicated that invertebrate communities did not differ between occupied and unoccupied wetlands. Our next steps will involve extended occupancy modeling with brood, wetland, and landscape-scale metrics, which will improve our understanding of the mechanisms driving brood habitat use in agriculturally-dominated landscapes. Final results from this project will yield a better understanding of wetlands in agriculturally-dominated landscapes and help conservation organizations inform strategies for restoring or acquiring productive wetlands for ducks.
Duck Nest Survival in the Western Boreal Forest of Alberta: Landscape Effects at Multiple Spatial Scales
Matthew E. Dyson; Stuart M. Slattery; Bradley C. Fedy
The western boreal forest (WBF) is an important breeding area for North American ducks, second only to the Canadian Prairies. The WBF is under intensive industrial development, causing habitat loss and fragmentation. Land use change can have profound effects on predator-prey interactions and influence population dynamics. In most avian species, nest success is critical to population persistence. Therefore, species are under intense selective pressure in choosing a safe nest site. Currently, we have limited knowledge of duck nesting ecology in the WBF, including the influence of changing land use practices on nest survival. We evaluated the effect of micro and macro habitat features on nest survival of ground nesting ducks in the WBF of Alberta at multiple spatial scales using the logistic exposure model. We located 167 duck nests of 8 different species between 2016 and 2018 by nest searching across a gradient of industrial development. We hypothesized nest survival associated with land cover and land use variables would vary with spatial scale. In addition, we hypothesized that industrial development increased predation and expected ducks to experience greater mortality risk in highly disturbed habitats. We will compare and contrast our nest survival results with previous results from our analysis of nest-site selection and present a predictive surface of nest mortality risk across our study area. Our results provide new information about the nesting ecology of boreal ducks and will be used to support conservation and management decision making in the boreal forest.
Relative Abundance and Timing of Waterfowl Use of Key Staging Areas in Michigan
Trey McClinton; David R. Luukkonen; Daniel B. Hayes
Michigan is located in the center of the migratory funnel that is the Great Lakes Region. As such, numerous state and federally operated refuges are located throughout, and serve as major staging areas for migrating waterfowl. Since the 1970s managers of eight areas scattered across the Lower Peninsula have conducted weekly refuge counts to document waterfowl abundance through seasonal progression. These data serve as an index of yearly migration characteristics. Though periodic reviews have helped inform hunting regulations, no comprehensive assessments have been completed. We obtained and analyzed 30 years of count data with the objectives of (1) evaluating trends in the relative abundance and timing of waterfowl migrating through key areas and (2) examining how regional and local factors influence waterfowl congregation on key areas. We ran a general linear model that explained numerical week of peak abundance by year block, unique area, and the interaction between the two. We found that the main effect of area and the interaction between year block and area were significant, while main effect of year block was not. This indicates that a coarse general trend of later migration is not apparent across all sites, but certain areas routinely observe later peak abundances and that there are unique trends in the timing of observed peaks across areas. These insights have implications for establishing zones and season frameworks, as they show that waterfowl move through the various migratory corridors differently. Additionally, we observed a precipitous decline in waterfowl abundances documented when averaged across sites. This could be an indication of changing migratory corridors. Further analyses will help us understand the characteristics of observed migration and the factors influencing the congregation of migrating birds on key staging areas.
American Black Duck Nesting Ecology in North Carolina
Daniel Lawson; Chris Williams
North Carolina is the southernmost extent of the American black duck’s (Anas rubripes) breeding range; however, little is known about their nesting ecology in this region. Therefore, we located and monitored 140 nesting black ducks over two years (2017-18) to quantify preferred nesting habitat and assess nesting productivity within coastal North Carolina, USA. We located nests in brackish marshes (75%) and man-made dredge spoil-islands (25%) at a density of 1 nest per 22 ha. Specifically, black ducks selected high-marsh habitats where nests were located an average of 21.81 m from open water at a mean elevation of 1.36 m. In these habitats, visual obstruction readings were 0.50 m with a maximum mean vegetation height of 0.81 m consisting largely of grasses (84.6%). Apparent nest success rates varied from 31% (2017) to 63% (2018) across years. The majority (73%) of variability in nest success was best predicted by nest location, vegetation density, maximum vegetation height, and elevation. Management for breeding black ducks in coastal North Carolina should focus on promoting selected nesting habitat and reducing nest predators


Contributed Oral Presentations
Location: Virtual Date: Time: -