Wildlife agencies have carefully managed wood duck Aix sponsa populations in part through harvest regulations since the early 1900s. However, unlike many other waterfowl species in North America, waterfowl managers largely do not know breeding population size. Lincoln–Petersen methods based on harvest and band recovery data are an attractive alternative to air or ground count-based surveys for estimating populations of wood ducks by state and federal agencies that are tasked with sustainably managing harvest opportunities for this species. We used banding and recovery data to estimate annual survival rates, harvest rates, and population size in late summer (August–September) using Lincoln–Petersen methods for wood ducks banded within Ohio from 1990 to 2017. Sex, age, and daily bag limits best explained survival rates of wood ducks banded in Ohio, with lower survival rates in years with more liberal bag limits. Lincoln–Petersen estimates of population size ranged from 116,992 to 632,462 annually, and we detected a significant declining trend in population size through time. Mean harvest rates of wood ducks banded in Ohio ranged from 0.069 (adult females) to 0.121 (hatch-year males), and we detected a significant increasing trend in harvest rate through time for adult male, hatch-year male, and hatch-year female wood duck cohorts. Aerial surveys in other Great Lakes states provide comparable population estimates with our Lincoln–Petersen estimates for Ohio and also show a declining trend in population size. We recommend continued investigation into the use of Lincoln–Petersen techniques for estimating wood duck population size throughout the Great Lakes region. If declining population trends are not unique to Ohio, waterfowl managers may need to further assess the potential impact of increased harvest rates on population size to ensure sustainable harvest into the future.
Wood ducks Aix sponsa are consistently one of the most abundant duck species harvested by waterfowl hunters in the Mississippi Flyway, ranking from fifth to second annually relative to other species in terms of overall contribution to the harvest (Fronczak 2018). Waterfowl managers in the Mississippi Flyway liberalized hunting regulations in 2008 by increasing daily bag limits from two to three birds since there was evidence the eastern population (Central, Mississippi, and Atlantic flyways) could sustain additional harvest (Garrettson 2007). However, wood ducks present some unique challenges from a population monitoring perspective. Aerial surveys are conducted in spring across the northern United States and Canada to monitor prairie-, boreal-, and arctic-nesting waterfowl populations (Smith 1995; Williams et al. 1996; Johnson et al. 2015; USFWS 2018). However, the breeding distribution of wood ducks lies outside the traditional waterfowl population survey area, although some Great Lakes states (Michigan, Minnesota, Wisconsin) do conduct aerial breeding waterfowl surveys that include wood ducks. Nevertheless, wood ducks nest in tree cavities and inhabit forested wetlands with limited visibility, so aerial surveys of breeding populations are likely biased (Sherman et al. 1992; Kelley 1996). Consequently, managers use a separate strategy based on kill rates to inform daily bag limits for wood ducks within the Mississippi and Atlantic flyways (Garrettson 2007).
The Ohio Division of Wildlife has banded over 44,000 wood ducks since 1990 providing a robust dataset to estimate survival rates and population trends. Several recent investigations have used Lincoln–Petersen estimators based on harvest rates obtained from banding data and estimates of total harvest to estimate waterfowl abundance in late summer (Lincoln 1930, Alisauskas et al. 2009, 2014; McAlister et al. 2017). This methodology is a promising technique for agencies tasked with managing wood duck populations since no range-wide estimates of breeding population size are available. Consequently, we developed this project with three major objectives: 1) to estimate annual survival and harvest rates of wood ducks banded in Ohio by sex and age cohorts, 2) to use Lincoln–Petersen methods to estimate annual wood duck population size in Ohio in late summer, and 3) to determine the possible effects of liberalized harvest regulations on survival, harvest, and population trends of wood ducks. We hypothesized that hatch-year birds likely have higher harvest rates than adults as they are more susceptible to harvest (Johnson et al. 1986). We also hypothesized that since wood ducks are sexually dimorphic, hunters would preferentially target males over females resulting in higher harvest rates for male cohorts relative to female cohorts (Metz and Ankney 1991).
Study Site and Methods
Ohio Division of Wildlife staff banded wood ducks in 66 of Ohio's 88 counties during 1990 to 2017 (Figure 1). Total bandings by county ranged from a single wood duck in Butler and Cuyahoga counties to 3,664 in Mahoning county. Banding distribution closely followed estimated breeding wood duck distribution based on the North American Breeding Bird Survey and Ohio Breeding Bird Atlas (Rodewald et al. 2016). Bandings were concentrated in Ohio's remaining coastal wetlands in northwest Ohio and more isolated, forested wetlands in northeast and east-central Ohio (Figure 1, Rodewald et. al. 2016).
We obtained all banding and recovery data from Gamebirds (USGS 2019). We used all wood ducks banded during the pre–hunting season banding period in Ohio (June–August) with standard federal leg bands (reward bands excluded) and that were reported as either hunter-harvested or dead during the hunting season to estimate annual survival rates (Johnson et al. 1986). We excluded wood ducks with unknown sex (n = 320) or age information (n = 12) from analyses as we assumed both sex and age might be important covariates influencing survival and harvest rates. We used a total of 44,664 bandings and 4,901 recoveries from 1990 to 2017 to estimate survival rates in program MARK (Data S1, Supplemental Material, Table 1; White and Burnham 1999).
We developed a list of 10 a priori Brownie dead-recovery models based on biological relevance and our interest in harvest regulation changes to explain survival and recovery rates of wood ducks banded in Ohio (Table 2; Brownie et al. 1985). We also included a null model that assumed a constant survival rate regardless of sex, age, season length, or daily bag limit changes. We ranked models using Akiake's Information Criterion adjusted for small sample size (Burnham and Anderson 2002), allowing survival rates to vary by sex and age. We also investigated the potential impacts of changing harvest regulations on wood duck survival rates, so we allowed survival to vary with changes in daily bag limits, changes in season length, and changes in both variables simultaneously. Daily bag limit and season length were time-dependent covariates with bag limit changing only once (2008) and season lengths changing three times: in 1994, 1995, and 1997 (Table 2). Because daily bag limit and season length were both time dependent, we could not include them with a fully time-varying model, so we did not include time as a covariate when estimating survival parameters (White and Burnham 1999; Heller 2010; Shirkey et. al. 2018).
We estimated annual population size of wood ducks in Ohio in late summer by using Lincoln's method of dividing total estimated harvest by estimated harvest rate (Lincoln 1930; Alisauskas et al. 2014). Several fundamental assumptions that must be met to obtain unbiased estimates of abundance using Lincoln's estimator include random sampling, population closure, independent band recovery probabilities, no influence of bands on survival, and no loss of bands, and that hunters report all bands. Likely the most difficult assumption to address for this analysis was obtaining valid estimates of total harvest and harvest rate from the same closed population. Few, if any, wild bird populations meet the assumption of complete population closure. However, we assumed, based on migration chronology, that summer breeding populations, when considered at the spatial extent of an entire state like Ohio, remain closed until the onset of autumn migration. Furthermore, if we assume that banded individuals are a random sample of the entire population, then we could apply an index of immigration and emigration rates of banded individuals to the wood duck population at large to investigate the assumption of population closure. We determined based on previous research most large-scale wood duck migratory movements begin in late October and early November in Ohio (Rodewald et al. 2016). Consequently, we assumed the population was closed until October 31. We examined wood duck band recovery data (1996–2017) and found that 808 of 843 (95%) of all direct recoveries in October of wood ducks banded in Ohio were recovered within the state, suggesting limited emigration out of the Ohio population before October 31. Furthermore, we found 819 of 940 (87%) direct recoveries in Ohio through October 31 came from birds banded in Ohio, suggesting some immigration into the population before 31 October.
After investigating population closure, we began the process of estimating abundance using Lincoln methods by first estimating harvest rate from what we assumed was a closed population. We restricted our use of Lincoln methods to estimate population size from 1996 to 2017 because reporting rates increased significantly after the introduction of toll-free bands in 1996. This likely biased harvest rate estimates in years prior to 1996 (Garrettson et al. 2014). We censored band recovery data to include only birds banded in Ohio and harvested before October 31. We then calculated an October harvest rate by dividing annual October direct recovery rates by band reporting rates (Table 3). We defined direct recoveries as recoveries of banded wood ducks harvested in October the same year they were banded. We took band reporting rates from reward-band studies and previously published Lincoln estimates with a band reporting rate of 0.491 in 1996, 0.620 in 1997, and 0.736 from 1998 to 2017 (Garrettson 2007; Garrettson et al. 2014, Alisauskas et al. 2014).
Our next step was to estimate total harvest from the same October population of wood ducks. We obtained estimates of total harvest for wood ducks in Ohio from the Mississippi Flyway Council data book (Fronczak 2018); however, this estimate required a series of corrections to adjust it to one that represented our closed population of wood ducks in October (Data S2, Supplemental Material). First, we accounted for likely bias within the U.S. Fish and Wildlife Service's harvest estimates by multiplying the estimated harvest from the Mississippi Flyway Council data book by 0.73 (Padding and Royle 2012).We then adjusted the resulting product by the annual estimated proportion of Ohio's harvest of wood ducks that occurs during the month of October (Data S2). We did this by dividing the direct recoveries of wood ducks banded in Ohio and harvested in October by the total number of direct recoveries, and then multiplying the previously calculated harvest estimate adjusted by the Padding and Royle correction factor by this proportion (Data S2). Finally, to account for immigration of birds not from Ohio prior to October 31, we multiplied our harvest estimate, now corrected by the Padding and Royle correction factor and by October harvest, by the proportion of direct recoveries reported in October in Ohio that were also banded in Ohio. We calculated this immigration rate by dividing the number of direct recoveries of wood ducks banded in Ohio and harvested in October by the total number of wood duck direct recoveries in Ohio in October (Data S2). These calculations assume that banded birds were representative of the population. We then divided our estimate of Ohio harvest in October by the harvest rate of wood ducks in October to estimate summer abundance using Lincoln's methods and calculated variance for the Lincoln estimates by applying the delta method (Table 3; Lincoln 1930, Alisauskas et al. 2014).
We calculated Ohio wood duck harvest rates by sex and age cohorts from 1996 to 2017. We did this by dividing annual direct recovery rates by the estimated band reporting rate (Alisauskas et al. 2014; Data S3, Supplemental Material). Consequently, we calculated these harvest rate estimates independently from program MARK. Unlike the harvest rates we used for Lincoln–Petersen calculations, we did not censor direct band recoveries to include only the month of October because it was not necessary to meet the assumption of population closure (Data S3). We defined direct recoveries as recoveries of banded wood ducks harvested during the hunting season immediately following banding each year. For example, hunters would have to harvest wood ducks banded in July of 2015 from September of 2015 through January of 2016 for us to consider them a direct recovery. Consequently, we included all direct recoveries of wood ducks banded in Ohio and reported as either shot or found dead during the hunting season allowing for a harvest rate estimate more directly comparable to other published studies than the October harvest rate we calculated for Lincoln estimates. We calculated the variance on the harvest rate estimates by applying the delta method, which incorporates the variance from both the band reporting rate estimates and direct recovery rate estimates (Alisauskas et al. 2014).
The highest ranked model for estimating survival rates of wood ducks banded in Ohio included the effects of sex, age, and daily bag limits (Table 2). The highest ranked model included 36 parameters and was strongly supported over all other models (wi = 0.96). Estimated survival rates were higher among all wood duck sex and age cohorts during years with two-bird daily bag limits compared to years with three-bird daily bag limits and were significantly higher in years with two-bird daily bag limits in both the adult female and adult male cohorts (Figure 2). Under two-bird daily bag limits survival was highest within the hatch-year male cohort (ŝ = 0.67, 95% CI = 0.616–0.725), followed by adult males (ŝ = 0.63, 95% CI = 0.620–0.651), adult females (ŝ = 0.522, 95% CI = 0.501–0.544), and hatch-year females (ŝ = 0.453, 95% CI = 0.407–0.499 [Figure 2]). Survival under three-bird daily bag limits was highest within the adult male cohort (ŝ = 0.545, 95% CI = 0.473–0.616), followed by hatch-year males (ŝ = 0.522, 95% CI = 0.527–0.576), adult females (ŝ = 0.435, 95% CI = 0.402–0.467), and finally hatch-year females (ŝ = 0.408, 95% CI = 0.339–0.471[Figure 2]).
Lincoln estimates of wood duck population size in late summer ranged from 632,462 in 1998 to 116,992 in 2015 (Figure 3). We detected a significant declining trend (P = 0.015) in wood duck population estimates with a simple linear model predicting a decline of 12,168 wood ducks each year from 1996 to 2017 (Figure 3). Lincoln estimates since 2008, when Ohio moved from a two-bird daily bag limit to a three-bird daily bag limit, only exceeded the mean Lincoln estimate from 1996 to 2007 ( = 427,325) one time (441,514 in 2017 [Figure 3]).
We detected an increasing trend in harvest rates through time for the hatch-year male cohort (P < 0.001), adult male cohort (P = 0.019), and hatch-year female cohort (P < 0.001). Hatch-year male wood ducks had the highest mean harvest rate during 1996 to 2017 ( = 0.123, 95% CI = 0.056–0.190) and ranged from a low of 0.063 in 1998 to a high of 0.220 in 2016 (Figure 4). Adult male wood ducks had the next highest mean harvest rate ( = 0.097, 95% CI = 0.029–0.169) and ranged from a low of 0.058 in 2000 to a high of 0.170 in 2017 (Figure 4). Hatch-year females had a mean harvest rate of 0.090 (95% CI = 0.047–0.138) followed by adult females with a mean harvest rate of 0.071 (95% CI = 0.022–0.119). Hatch-year female harvest rates peaked in 2014 (0.137) and adult female harvest rates peaked in 2010 (0.124) with lows of 0.048 and 0.025 in 1996 for hatch-year females and adult females, respectively (Figure 4).
We found the highest ranked model explaining survival of wood ducks in Ohio was strongly supported (wi = 0.96) over all other models in the candidate set and included an effect of sex, age, and daily bag limits on survival rates. We found estimates of survival were higher among all four wood duck cohorts during years with two-bird daily bag limits than in years with three-bird daily bag limits. Mean annual survival rates (0.478–0.659) were higher than those reported by Johnson et al. (1986) for the northeastern section of the Atlantic Flyway where survival was highest within the adult male cohort (ŝ = 0.520), followed by adult females (ŝ = 0.476), then hatch-year females (ŝ = 0.382), and finally hatch-year males (ŝ = 0.368). In contrast, we found survival was highest in the male cohorts as compared to the female cohorts with minimal difference observed based on age. Survival rates of adult females were slightly lower than those estimated by Hepp et al. (1987) in South Carolina and Dugger et al. (1999) in southeast Missouri (ŝ = 0.55 and 0.63, respectively). Declining survival rates of wood ducks in Ohio appears to at least be correlated with adoption of more liberal daily bag limits; however, further experimental studies such as those suggested by Anderson et al. (1987) would be required to adequately understand the impacts of harvest rates on wood duck survival (Sedinger and Rexstad 1994).
We found harvest rates of wood ducks in Ohio were similar to other published studies. Balkom et al. (2010) estimated harvest rates within the Atlantic and Mississippi flyways that varied from 0.04 to 0.11. All cohorts in our analysis fell within that range except the hatch-year male cohort, which had a slightly higher estimated harvest rate ( = 0.121) than the range reported by Balkom et al (2010). Our results supported the prediction by Balkom et al. (2010) that harvest rates would increase across the Mississippi and Atlantic flyways after increasing the daily bag limit from two to three. Much of the increase in harvest rates occurred after 2007 when Mississippi Flyway states increased potential harvest from a two-bird daily bag limit to a three-bird daily bag limit. We found males had slightly higher harvest rates than females, and although not statistically significant, adult males and adult females had slightly lower harvest rates than their hatch-year counterparts as we hypothesized and other published studies have found (Johnson et al. 1986).
Zimmerman et al. (2015) combined data from the Atlantic Flyway Breeding Waterfowl Survey and North American Breeding Bird Survey and estimated approximately 1.3 million wood ducks for the U.S. portion of the Atlantic Flyway. We are unaware of any other Lincoln estimates using harvest data for wood ducks other than Zimmerman et al. (2015) and Bowers and Martin (1975), but considering continental spring breeding population estimates for several other common waterfowl species including mallards Anas platyrhynchos and blue-winged teal Anas discors reach well into the millions, state-wide estimates of several hundred thousand wood ducks seem plausible. Alisauskas et al. (2014) found that Lincoln estimates were nearly two to four times greater than population estimates from count-based surveys for mallards. They suggested that the differences were due to negative bias caused by the timing, incomplete coverage, and systematic underestimation of detection probabilities of continental-scale aerial surveys of breeding mallards. We have no independent count-based estimates of the breeding population of wood ducks in Ohio, but nearby states (Michigan and Wisconsin) conduct state-wide aerial surveys of breeding wood ducks. Our annual Lincoln estimates of wood duck abundance in Ohio in late summer were above the long-term average (1991–2013) of estimates from state-wide aerial surveys of wood ducks during spring in Michigan (143,401) and Wisconsin (112,541; Soulliere et al. 2017). Accounting for summer recruitment (∼2 juveniles/adult female; Zimmerman et al. 2017), places the aerial survey estimates (286,802 for Michigan, 225,082 for Wisconsin) closer to the long-term (1996–2017) average of 363,683 wood ducks that we estimated with Lincoln methods for Ohio in late summer. By converting abundance estimates both from spring surveys and from our Lincoln estimates to densities in an attempt to provide a more direct comparison among states, we found Ohio had the highest estimated wood duck densities (3.05 wood ducks/km2), followed by Wisconsin (1.33 wood ducks/km2), and then Michigan (1.13 wood ducks/km2). Although these estimates are not directly comparable, we would expect these states to have similar densities of breeding wood ducks. Interestingly we found Ohio's estimated density to be approximately three times greater than breeding densities in Michigan and Wisconsin, a finding that corresponds with the Alisauskas et al. (2014) results indicating Lincoln estimates of breeding mallards were on average two to four times greater than aerial survey–based population estimates.
Another source for comparison is the Fleming et al. (2019) derivation of regional nonbreeding duck population abundance from North American Waterfowl Management Plan long-term averages (NAWMP 2012). Fleming et al. (2019) used harvest and parts collection data to step down county-level estimates of nonbreeding wood duck population. We totaled the Fleming et al. estimates across all 88 Ohio counties to arrive at an estimate of 160,849 nonbreeding wood ducks in Ohio, which was at the low end of the range of annual Lincoln estimates that we calculated. Although Lincoln estimates, aerial surveys, and the Fleming et al. (2019) stepdown approach all likely have limitations for wood ducks, it is encouraging that all 3 methods yielded comparable results.
We also detected a declining trend (P < 0.015) in Lincoln estimates through the study period. This trend was consistent with increasing harvest rates of wood ducks as total estimated harvest within Ohio has been decreasing (Fronczak 2018; Data S2). This trend is also consistent with aerial surveys of breeding wood ducks from Michigan, Minnesota, and Wisconsin, which all indicate a declining wood duck population since the late 1990s, although Wisconsin's population has rebounded in recent years (Soulliere et al. 2017). There is some potential that changes in either banding effort or band reporting rates could bias harvest rates. However, banding effort has remained consistent in Ohio for the duration of the study (Table 1). Furthermore, we think it unlikely that band reporting rates have decreased significantly since 1995, which would bias harvest rates high and produce lower Lincoln estimates of abundance. Error could still be introduced into the Lincoln estimates if total harvest estimates are increasingly biased low. However, we are uncertain as to why this would be the case as it is commonly believed harvest estimates are biased high (Padding and Royle 2012). We recognize that the observed rate of emigration could result in a high bias of Lincoln estimates; harvest rates are biased low as these individuals are not accounted for in the harvest; however, this should have no impact on the overall trend observed in Lincoln estimates through time. Unpublished data suggest that band reporting rates have likely increased substantially in recent years from the 0.736 that we used in our analysis (Pam Garrettson, U.S. Fish and Wildlife Service, personal communication), and we recognize this could result in Lincoln estimates that are biased low during later years of our study. We did allow Lincoln estimates to change with hypothetical band reporting rates as high as 0.90 in later years of our study, and although Lincoln estimates did increase under those scenarios, there was still a declining trend through time in the abundance estimate (see Data S2).
It is important to recognize that we calculated the Lincoln population estimates in this analysis only from a small subset of the data that we used to calculate harvest and survival rates. Despite this, the results of both our survival and harvest rate analyses support the declining trend in Lincoln estimates. Although harvest and survival are not necessarily directly linked (Sedinger and Rexstad 1994), data from wood ducks in Ohio suggest that survival decreased in years with more liberal daily bag limits, harvest increased in concurrence with increased daily bag limits, and over that same time period Lincoln estimates of population size declined. Other factors like declines in reproductive success or increased non–hunting-related mortality could undoubtedly cause declines in population size. For example, estimated harvest rates in our analyses were highest in the male cohorts; however, survival was also highest in the male cohorts, suggesting higher mortality not related to hunting within female wood ducks. Nest predation or other nesting season events that only decrease female survival rates could cause this apparent inconsistency between survival and harvest. Regardless of the cause, the trend in population size is alarming, especially given that estimates of survival have decreased during the same time period.
We believe Lincoln estimates are a viable alternative for tracking populations' trends through time for wood ducks in northern states. The significant, declining trend in Lincoln population estimates of Ohio wood ducks is a concern, especially considering we found higher harvest rates and lower survival rates during the same time period, both of which support the observed trend in population size. We recommend continued banding effort allowing for further investigation of Lincoln estimators and an analysis of Lincoln estimates across the upper Great Lakes region to see if declining trends are unique to Ohio. If declining trends continue or researchers discover declines across the Great Lakes region, researchers may need to conduct further analysis aimed at assessing the impact of harvest on survival and ultimately population size to ensure continued sustainable harvest of wood ducks in the Great Lakes region of the United States.
Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding author.
Data S1. Microsoft Excel file containing the vertical encounter history for wood ducks Aix sponsa banded during the preseason banding period (June–September) in Ohio from 1990 to 2017 and recovered during the hunting season (October–January) 1990 to 2017. We used data to estimate survival in program MARK.
Found at DOI: https://doi.org/10.3996/082019-JFWM-070.S1 (70 KB XLSX).
Data S2. Microsoft Excel file containing annual bandings and direct recoveries used to estimate October harvest rates used in our Lincoln estimates of wood duck Aix sponsa population size in Ohio. The table also contains estimates of total harvest and the adjustments we used to correct for biases present in the total estimate of total harvest. The final harvest estimate we used for our Lincoln calculation is under column heading “Harvest corrected for immigration.” On worksheet titled “Unpublished band reporting rate,” we estimate abundance using Lincoln methods, but allow band reporting rate to increase to 0.80 in 2010 to 2011, 0.85 in 2012 to 2013, and 0.90 from 2014 to 2017.
Found at DOI: https://doi.org/10.3996/082019-JFWM-070.S2 (79 KB XLSX).
Data S3. Microsoft Excel file containing the annual number of wood ducks Aix sponsa banded in Ohio and direct recoveries by sex and age cohorts. The table also shows the calculations we used to estimate harvest rate from 1996 to 2017.
Found at DOI: https://doi.org/10.3996/082019-JFWM-070.S3 (101 KB XLSX).
Reference S1. Fleming KK, Mitchell MK, Brasher MG, Colluccy JM, James JD, Petrie MJ, Humburg DD, Soulliere GJ. 2019. Derivation of regional, non-breeding duck population abundance objectives to inform conservation planning in North America. 2019 revision. Falls Church, Virginia: North American Waterfowl Management Plan Science Support Team Technical Report No. 2019-01.
Found at DOI: https://doi.org/10.3996/082019-JFWM-070.S4 (2.67 MB PDF); also available at https://www.fws.gov/migratorybirds/pdf/management/NAWMP/DerivationofNon-breedingDuckPopulationAbundanceObjectives.pdf
Reference S2. Fronczak D. 2018. Waterfowl harvest and population survey data. Bloomington, Minnesota: U.S. Fish and Wildlife Service.
Found at DOI: https://doi.org/10.3996/082019-JFWM-070.S5 (1.19 MB PDF); also available at https://www.fws.gov/migratorybirds/pdf/surveys-and-data/DataBooks/MississippiFlywayDatabook.pdf
Reference S3. Garrettson PR. 2007. A proposed assessment and decision-making framework to inform harvest management of wood ducks Aix sponsa in the Atlantic and Mississippi flyways. Laurel, Maryland: U.S. Fish and Wildlife Service Division of Migratory Bird Management.
Found at DOI: https://doi.org/10.3996/082019-JFWM-070.S6 (575 KB PDF); also available at https://www.researchgate.net/publication/254470754_A_proposed_assessment_and_decision_making_framework_to_inform_harvest_management_of_wood_ducks_in_the_Atlantic_and_Mississippi_Flyways
Reference S4. Heller BJ. 2010. Analysis of giant Canada goose Branta canadensis band recovery data in Iowa and the Mississippi Flyway. Master's thesis. Ames: Iowa State University.
Reference S5. Lincoln FC. 1930. Calculating waterfowl abundance on the basis of banding returns. Washington, D.C.: U.S. Department of Agriculture Circular No. 118.
Found at DOI: https://doi.org/10.3996/082019-JFWM-070.S8 (456 KB PDF); also available at https://archive.org/stream/calculatingwater118linc?ref=ol#mode/2up
Reference S6. [NAWMP] North American Waterfowl Management Plan Committee. 2012. North American Waterfowl Management Plan 2012: people conserving waterfowl and wetlands. Washington, D.C.: U.S. Department of the Interior, U.S. Fish and Wildlife Service; Ottawa, Ontario, Canada: Environment Canada, Canadian Wildlife Service; Mexico City, Mexico: Secretaria de Medio Ambiente y Recursos Naturales.
Found at DOI: https://doi.org/10.3996/082019-JFWM-070.S9 (2.67 MB PDF); also available at: https://www.fws.gov/migratorybirds/pdf/management/NAWMP/2012NAWMP.pdf
Reference S7. Smith GW. 1995. A critical review of aerial and ground surveys of breeding waterfowl in North America. Laurel, Maryland: U.S. Fish and Wildlife Service Technical Report ADA322667.
Found at DOI: https://doi.org/10.3996/082019-JFWM-070.S10 (13.56 MB PDF); also available at http://www.michigandnr.com/FTP/wildlife/LuukkonenD/Breeding%20waterfowl%20survey%20project/Smith%201995.pdf
Reference S8. Soulliere GJ, Al-Saffar MA, Coluccy JM, Gates RJ, Hagy HM, Simpson JW, Straub JL, Pierce RL, Eicholz MW, Luukkonen DR. 2017. Upper Mississippi River and Great Lakes Region Joint Venture Waterfowl Habitat Conservation Strategy. Bloomington, Minnesota: U.S. Fish and Wildlife Service.
Found at DOI: https://doi.org/10.3996/082019-JFWM-070.S11 (6.79 MB PDF); also available at https://umgljv.org/docs/2017JVWaterfowlStrategy.pdf
Reference S9. [USFWS] U.S. Fish and Wildlife Service. 2018. Waterfowl population and status, 2018. Washington D.C.: U.S. Department of Interior.
Found at DOI: https://doi.org/10.3996/082019-JFWM-070.S12 (2.52 MB PDF); also available at https://www.fws.gov/migratorybirds/pdf/surveys-and-data/Population-status/Waterfowl/WaterfowlPopulationStatusReport18.pdf
First and foremost, we thank all Ohio Division of Wildlife staff who have assisted with wood duck banding since 1990 that helped make this project possible. We thank M. Reynolds and M. Ervin for their assistance in starting this project. We thank J. Simpson and D. McClain whose comments improved this manuscript. We would also like to thank the Associate Editor and two anonymous reviewers for their comments and assistance throughout the peer review process that greatly benefited this manuscript.
Any use of trade, product, website, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. Government.
Citation: Shirkey BT, Gates RJ. 2020. Survival, harvest, and Lincoln estimates of wood ducks banded in Ohio. Journal of Fish and Wildlife Management 11(1):185–195; e1944-687X. https://doi.org/10.3996/082019-JFWM-070
The findings and conclusions in this article are those of the author(s) and do not necessarily represent the views of the U.S. Fish and Wildlife Service.