Old Dominion UniversityODU Digital Commons

Biological Sciences Faculty Publications Biological Sciences

2005

Effects of Human Disturbances on the Behavior ofWintering DucksMelissa L. PeaseOld Dominion University

Robert K. RoseOld Dominion University, brose@odu.edu

Mark J. ButlerOld Dominion University, mbutler@odu.edu

Follow this and additional works at: https://digitalcommons.odu.edu/biology_fac_pubs

Part of the Behavior and Ethology Commons, Natural Resources and Conservation Commons,and the Ornithology Commons

This Article is brought to you for free and open access by the Biological Sciences at ODU Digital Commons. It has been accepted for inclusion inBiological Sciences Faculty Publications by an authorized administrator of ODU Digital Commons. For more information, please contactdigitalcommons@odu.edu.

Repository CitationPease, Melissa L.; Rose, Robert K.; and Butler, Mark J., "Effects of Human Disturbances on the Behavior of Wintering Ducks" (2005).Biological Sciences Faculty Publications. 228.https://digitalcommons.odu.edu/biology_fac_pubs/228

Original Publication CitationPease, M. L., Rose, R. K., & Butler, M. J. (2005). Effects of human disturbances on the behavior of wintering ducks. Wildlife SocietyBulletin, 33(1), 103-112. doi:10.2193/0091-7648(2005)33[103:eohdot]2.0.co;2

Human presence alone can negatively affectwildlife by causing animals to alter behaviors nec-essary to survival, such as feeding. Because thenumber of Americans participating in outdoorrecreational activities has increased in recentdecades, conflicts between humans and wildlifehave been steadily increasing (Anderson and Keith1980, Boyle and Samson 1985). Furthermore,coastal flyways in the continental United States andassociated wetland habitats upon which migratorywaterbirds depend frequently correspond withareas of largest human development (Purdy et al.1987). Coastal wetland habitat continues to be lostto human development; therefore, it is increasinglyimportant to minimize the effects of human activi-ty on waterbird behavior in remaining habitat.

Ducks rely in part on stored lipids and nutrientsgained while wintering, staging, and migrating to

meet the high energetic costs of reproduction. Forexample, stored lipids may indirectly influenceclutch size in ducks by providing the energy hensneed to forage for the protein food sourcesrequired for egg production (Krapu 1981).Migration between wintering and breedinggrounds also imposes high energetic costs onwaterfowl. From a resource management perspec-tive, it is important to assess human activities thatmay adversely affect energy reserves of wintering,staging, and migrating waterfowl, which, in turn,affect nesting success, fecundity, and survival ofwaterfowl (Ankney and MacInnes 1978, Krapu1981, Havera et al. 1992).

Many natural resource managers are charged bytheir administrators with balancing multiple andoften conflicting uses, which include resource pro-tection and human recreation. Managers are

HUMAN DISTURBANCE OF WINTERING DUCKS 103

Wildlife Society Bulletin 2005, 33(1):103–112 Peer refereed

Address for Melissa L. Pease: Elizabeth River Project, 801 Boush Street, Suite 204, Norfolk, VA 23510, USA, and Department ofBiological Sciences, Old Dominion University, Norfolk, VA 23529-0266, USA; present address: 8731 Woodmere Crossing Lane,Charlotte, NC 28226, USA; e-mail: melissapease@carolina.rr.com. Address for Robert K. Rose and Mark J. Butler: Department ofBiological Sciences, Old Dominion University, Norfolk, VA 23529-0266, USA.

Effects of human disturbances on thebehavior of wintering ducks

Melissa L. Pease, Robert K. Rose, and Mark J. Butler

Abstract Human activity causes wintering waterfowl to expend energy to avoid humans at a timein their annual cycle when energy conservation is important to survival, migration, andbreeding reserves. Understanding the effects of recreational activities on waterfowl isimportant to managing natural resource areas where migratory birds depend on wetlandhabitat for resting and feeding. We investigated responses of 7 species of dabbling ducksto 5 different experimental human activities, (a pedestrian, a bicyclist, a truck traveling at2 different speeds, and an electric passenger tram). Responses of ducks depended ontype of disturbance, species, and distance from disturbances. Most birds responded tothe treatments. People walking and biking disturbed ducks more than vehicles did.Northern pintail (Anas acuta) was the species least sensitive to disturbance, whereasAmerican wigeon (A. americana), green-winged teal (A. crecca), and gadwall (A. strepera)were most sensitive. Ducks were more likely to fly when closer to sources of disturbance.These results will be helpful to managers making decisions about public use that strive tominimize disturbance of dabbling ducks.

Key words Anas, Anatidae, behavior, ducks, human disturbance, impoundments, mid-Atlanticregion, recreation, waterfowl

13 (Pease et al).qxp 6/24/2005 12:35 PM Page 103

increasingly required to make decisions abouttypes and amounts of public use that should beallowed without degrading an area’s value towildlife. Such decisions often are controversial andchallenged by the public. In order for managers tomake defensible decisions about recreational usesthat minimize impacts on waterfowl, informationmust be obtained about impacts of various activi-ties and which activities have greater or lessereffects (Pomerantz et al. 1988). Our objective wasto determine which human activities are most dis-turbing to dabbling ducksin the genus Anas ontheir wintering grounds.

Several studies havesought to identify howhuman activities disturbwaterbirds (Burger 1981,Bélanger and Bédard1989, Havera et al. 1992,Klein et al. 1995), but fewwith the exception ofKlein (1993) have usedcontrolled experimentaldisturbances as we havedone in this study.

Study areaThe work was conduct-

ed on the barrier-spit sec-tion of Back Bay NationalWildlife Refuge (BBNWR),located along the AtlanticFlyway in the southeastcorner of Virginia in theCity of Virginia Beach (Fig-ure 1). Like many Nation-al Wildlife Refuges alongthe eastern seaboard,BBNWR was near a largemetropolitan area andconsequently receivedheavy public use (Burger1981,Purdy et al.1987). Aman-made dune systemprevented saltwater fromentering the 356.3-hafreshwater impoundmentsystem and the freshwaterecosystem of Back Bay(United States Depart-

ment of the Interior [USDI] 1996). Historically,Back Bay was known for its large concentrations ofoverwintering waterfowl, which declined partlydue to market and sport hunting and, more recent-ly, to changes in water quality that reduced theabundance of submerged aquatic vegetation, impor-tant foods for dabbling ducks.

Observation blinds for the study were located inthe impoundment system, which was intensivelymanaged in part by water-level manipulation toprovide feeding habitat for migratory and overwin-

104 Wildlife Society Bulletin 2005, 33(1):103–112

Figure 1. Study area: barrier-spit portion of Back Bay National Wildlife Refuge, Virginia Beach,Virginia, 1998–2000.

13 (Pease et al).qxp 6/24/2005 12:35 PM Page 104

tering waterbirds. Dominant species of submergedaquatic vegetation include pondweeds (Potamoge-ton spp.), widgeon-grass (Ruppia maritima),water-milfoils (Myriophyllum spp.), and coontail(Ceratophyllum demersum). The dominant, emer-gent wetland-associated plants include spike-rushes(Eleocharis sp.), bulrushes (Scirpus sp.), and water-hyssop (Bacopa monnieri). Fragments of maritimeforest lay adjacent to the impoundments and onsand mounds within the impoundments.

Two main dikes, called the east and west dikes,were positioned north–south, with individualimpoundments being formed by “cross dikes” posi-tioned east–west (Figure 1). The east and westdikes were the only improved surfaces (gravelroads) by which visitors and employees of theCommonwealth of Virginia could gain access toFalse Cape State Park (FCSP), located south of therefuge. The barrier spit not only contained thehighest-quality habitat for waterfowl on the refugebut also sustained the highest public use, creating aconflict between wildlife habitat management andpublic use objectives.

MethodsThe primary researcher (MLP) conducted obser-

vations in winter months from 8 November–20February 1998–1999 and from 1 November–15February 1999–2000. Five randomly applied exper-imental treatments simulated the most commonhuman activities on the dike roads, and responseswere recorded for the most common dabblingducks (American black duck [Anas rubripes], gad-wall [A.strepera],mallard [A.platyrhynchos],north-ern pintail [A. acuta], American wigeon [A.Americana], northern shoveler [A. clypeata], andgreen-winged teal [A. crecca]). Volunteers in ourstudy performed standardized disturbances, allow-ing the primary researcher to make detailed obser-vations (from a blind). Because we observed ducksfor approximately 30 minutes before each distur-bance event, any change in behavior was easily dis-cernible.

We scheduled 1 treatment every half-hour begin-ning on the quarter-hour closest to official sunriseand ending no later than 1030, so we could con-duct up to 6 treatments in 1 morning. We random-ly determined the order and type of treatments,plus a control. We recorded the response of an indi-vidual bird, the experimental unit, during each dis-turbance event. We scheduled work every other

day to minimize disturbance to wintering birds. Wecancelled study days during times of heavy fog,high winds, and frozen water, and ended them earlywhen the refuge staff conducted biweekly aerialsurveys or when there were insufficient numbersof ducks in the study area.

Experimental human activitiesWe conducted all treatments at a constant speed

along the dike roads except for the control. Thetreatments were as follows:

1. Control. No human activity was performedand the observation period lasted 2 minutes,based on the average time it took for the treat-ments to be conducted.

2. Tram. A white electric tram traveled at itsmaximum speed of ca.21 km/hour. This tram,used to move visitors through the impound-ment system during nonwinter months toFCSP,had a driver’s cart and 2 passenger carts.Because weighted carts bounce less on the

Human disturbance of wintering ducks • Pease et al. 105

The researcher’s blind at the edge of a study area.

13 (Pease et al).qxp 6/24/2005 12:35 PM Page 105

gravel roads and may create less noise thanempty carts, 1 cart was weighted withapproximately 340 kg of gravel-filled bags tosimulate the weight of passengers.

3. Slow truck. A white 1994 ChevroletCheyenne™ (General Motors, Pontiac, Mich.)with dual rear tires and an 8-cylinder diesel-powered engine traveled at ca. 21 km/hour.This speed, the same speed as the tram,allowed a direct comparison between theslow truck and the tram.

4. Fast truck. The same truck used in the slow-truck treatment traveled at 48 km/hour.

5. Bicyclist. One person biked at ca. 11.38km/hour on a dark blue or black bicycle.

6. Pedestrian. One person walked at ca. 3.17km/hour.

Volunteers wore clothes of neutral colors orearth tones while conducting treatments. For coor-dination and standardization purposes, volunteersattended an orientation session at the beginning ofeach season, received protocol instructions viaphone the night before each study morning, andreceived written directions each study morning.

Observations of duck responses Study areas were open areas of shallow water

with minimal emergent vegetation. The primaryresearcher made observations of duck responseswith Pentax® 8 × 42 binoculars (Pentax Corpora-tion,Tokyo, Japan) from 1.22 × 1.22 × 1.68-m cam-ouflage-painted plywood blinds. We placed 4blinds, each with 4 sides and a roof to conceal theresearcher from birds in all directions, along thedike roads at the edges of the impoundments. Wemoved blinds during the season as concentrations

of waterfowl shifted in reaction to managedchanges in water levels of the pools.

The primary researcher entered the blind beforesunrise at least 30 minutes before the first sched-uled treatment. She first drove to within ca. 300 mof the blind, and the remaining approach was madethrough concealing vegetation to minimize distur-bance. This method of approach did not causeducks to fly from the study areas.

We chose individual ducks to examine responsesof different species at different distances and in dif-ferent positions in the flock. To reduce potentialbias, the researcher targeted an individual bird afew minutes before the treatment (if an individualbird was not chosen until the treatment was initiat-ed, one might inadvertently get an indication of thebird’s probable response). We made efforts to avoidobserving the same individuals repeatedly, thoughthis would be unlikely given the thousands ofducks present at the refuge and the movement offlocks southward during times of unusually coldweather. In the rare event that a volunteer did notfollow protocol or an external stimulus corre-sponded with the timing of a treatment (e.g., planeor predatory bird flying overhead), we did notinclude the observation in the analysis, to ensurethat the reactions of birds were in response to treat-ments. Hours of observation totaled 191, althoughthis number was not relevant to the statistical analy-sis of this study.

We assumed that the following responses weregraded indicators of the level of disturbanceimposed on a bird: 1) no observable response orchange in behavior, the least-disturbed state; 2)became alert: the bird raised its head or feeding was

106 Wildlife Society Bulletin 2005, 33(1):103–112

Northern shovelers within 25m of the researcher’s blind.

The tram transporting visitors through Back Bay NationalWildlife Refuge along the East Dike.

13 (Pease et al).qxp 6/24/2005 12:35 PM Page 106

interrupted, but it did not move away from thesource of human activity; 3) swam away: the birdswam any distance away from source of humanactivity; and 4) flew away: the bird flew any dis-tance away from source of human activity.

Distance from source of disturbanceThe primary researcher recorded distance from

the source of disturbance (dike road) to the bird’spretreatment location. Distance categories were0–50 m, 51–100 m, 101–150 m, 151–200 m,201–250 m, and 251–300 m. To improve accuracyof the distance estimates, we placed stakes made ofnatural wood and white polyvinyl chloride (PVC)pipes in the impoundments at 50-m intervals meas-ured from the edge of the road. Other white PVCpipes had been placed in the impoundments byrefuge staff many years before this study, so therewas little reason to believe they affected birdbehavior.

Statistical analysisWe used a multidimensional log-linear (contin-

gency table) analysis to test the null hypothesis thatfrequencies of bird responses (4 levels) were inde-pendent of disturbance treatments (6 levels),species (7 levels), and distances from the source of

disturbance (5 levels) at α=0.05. The 201–150-mand the 251–300-m categories were combined dueto small sample sizes. The 4-way and 3-way interac-tions were not significant. The P-value was adjust-ed to α=0.013 for the 2-way interactions to accountfor multiple tests. The 2-way analysis of duckresponses as a function of disturbance treatmentswas subdivided (Zar 1999) to find differencesamong treatments at α=0.006.

Because we chose treatments randomly, theirsample sizes are similar:1) control,n=62;2) tram,n=59; 3) slow truck, n=73; 4) fast truck, n=57; 5)bicyclist, n=62; 6) pedestrian, n=69. Sample sizesfor the different species and distance categorieswere not equal. It is typical to have unequal samplesizes in contingency table analyses and this has nobearing on the results.

ResultsResponses of ducks depended on the type of

human disturbance (likelihood ratio χ2=222.68, df= 15, P < 0.001, n = 382) at α = 0.013 (Figure 2).Pedestrians and bicyclists caused the highest pro-portions of ducks to fly (Figure 2). There were nodifferences among the vehicular treatments in thenumber of ducks that flew, and no difference

Human disturbance of wintering ducks • Pease et al. 107

Figure 2. Proportions of duck responses exhibited for each human disturbance; includes all 7 species observed. From χ2 analy-sis of response as a function of disturbance treatment, likelihood ratio χ2 = 222.68, df = 15, P < 0.001, n = 382, Back Bay NationalWildlife Refuge, Virginia Beach, Virginia, 1998–2000.

13 (Pease et al).qxp 6/24/2005 12:35 PM Page 107

among nonvehicular treatments (pedestrians andbicyclists) in the number of ducks that flew.However, there was a difference between vehicular

and nonvehicular treatments in the number ofducks that flew (likelihood ratio χ2=12.68, df=1,P<0.001, n=320) at α =0.006 (Figure 2). Ninety

108 Wildlife Society Bulletin 2005, 33(1):103–112

Figure 3. Proportions of duck responses within each species; includes all forms of human disturbance tested. From χ2 analysis ofresponse as a function of species, likelihood ratio χ2 = 41.75, df = 18, P < 0.001, n = 382, Back Bay National Wildlife Refuge,Virginia Beach, Virginia, 1998–2000.

13 (Pease et al).qxp 6/24/2005 12:35 PM Page 108

percent of the birds showed an observableresponse (alerted, swam, or flew) to the humanactivities, of which 43.1% flew. All but 2 responsesto the control treatments were “no response.”

Responses of ducks also depended on species(likelihood ratio χ2 = 41.75, df = 18, P < 0.001, n =382) at α=0.013 (Figure 3). Northern pintail hadthe lowest proportion of flight response and, con-versely,demonstrated the highest proportion of “noresponse.” American wigeon, northern shoveler,green-winged teal, and gadwall demonstrated highproportions of flight responses. Conversely, green-winged teal and gadwall demonstrated the lowestproportions of “no response” (Figure 3).

Responses of ducks depended on distances the

ducks were located from the dike road (likelihoodratio χ2 = 87.18, df = 12, P < 0.001, n = 382) at α =0.013 (Figure 4). The proportion of ducks that flewwas greatest in the 0–50-m and 51–100-m cate-gories (Figure 4). We observed a predictable grad-ed response with respect to distance as ducks near-er the source of disturbance showed higher pro-portions of responses and higher levels of flightresponses than ducks located farther away: 1) forthe 0–50-m distance category, 84.7% of the birdsresponded (alerted, swam, or flew); 2) 51–100-m,78.9% responded; 3) 101–150-m, 69.8% responded;4) 151–200-m,74.2% responded;and 5) 201–300-m,59.0% responded. Conversely, the incidence of “noresponse” increased with distance (Figure 4).

Human disturbance of wintering ducks • Pease et al. 109

Figure 4. Proportions of duck responses exhibited at each distance; includes all 7 species observed. From χ2 analysis of responseas a function of distance the ducks were from the source of disturbance, likelihood ratio χ2 = 87.18, df = 12, P < 0.001, n = 382,Back Bay National Wildlife Refuge, Virginia Beach, Virginia, 1998–2000.

13 (Pease et al).qxp 6/24/2005 12:35 PM Page 109

DiscussionBecause people walking and biking caused more

ducks to fly, we judged these activities to be moredisturbing than vehicles to the 7 species of dab-bling ducks studied. Flight response is of greatestinterest of those studied because it requires thehighest energetic output for birds. These resultswere consistent with Klein’s (1993) conclusionthat “out-of-vehicle-activity” (people getting out ofvehicles to observe wildlife) was more disturbingthan vehicular traffic. We suspected the electrictram would be less disturbing to waterfowl thanother vehicles because it is considered inherentlyquieter than a combustion engine; however, wefound the tram to be no less disturbing than a dieseltruck traveling the same speed.

Because the 3-way interaction between theresponse variables, the independent variables, andspecies was not significant, the analysis does notallow individual species to be discernible in the 2-way analysis of duck responses as a function of dis-turbance treatments (Figure 2). Although we treat-ed all species together in this case, the results indi-cating which human activities were more or lessdisturbing to ducks in the genus Anas are valuablebecause the ducks are closely related. Klein et al.(1995) found dabbling ducks, as a group, to be sen-sitive to lower rates of vehicular traffic than someother water birds.

A person walking was highly disturbing to ducksin this study, making it unclear how birdwatcherscould be minimally disturbing to birds, as reportedby Burger (1981). Burger (1981) reported that bird-watchers walking slowly as they search for birdsand people digging clams in shallow water usuallydid not cause birds to flush. People walking in thisstudy were moving at a steady pace, unlike bird-watchers whose pace typically involves periods ofstanding still.

It is important to understand differences amongspecies when evaluating impacts of recreationaluse (Vaske et al. 1983). We considered proportionsof flight response and “no response,” indicating sen-sitivity and tolerance respectively, in judging rela-tive sensitivity between species. Northern pintailclearly was the least-sensitive species. Americanwigeon, green-winged teal, and gadwall were themost sensitive. Although it may appear that north-ern shovelers were relatively sensitive because oftheir high incidence of flight response, this speciesflew more because they were more often located

closest to the source of disturbance (Pease 2001).Another indication that they are not among themost sensitive species is their high proportion of“no response” indicating relative tolerance to dis-turbance (Figure 3).

Birds were more severely affected by human dis-turbance the closer they were to the road, thesource of disturbance. Although lower percentagesof birds flew the farther they were from the road,they responded in other ways that included swim-ming and alerting. Very few birds showed noresponse to human activity.

It was not possible to determine the extent towhich ducks were indirectly responding to eachother,a product of the realistic field-based design ofthis study. We speculate the results are conservativeto some degree because as each morning pro-gressed, a more and more tolerant group of ducksremained. This resulted from observations wemade of the most sensitive birds leaving the studyarea altogether due to earlier treatments and fewnew birds moving into study areas.

Management implicationsAs in most refuges managed by the United States

Fish and Wildlife Service, providing habitat formigratory waterbirds is a primary objective atBBNWR. Therefore, management activities andresources of the refuge are focused on theimpoundment system, which is formed by dikesthat also function as access roads to FCSP. Findinga solution to the dilemma of ensuring adequateaccess to the state park while not compromisingthe primary objective of the refuge was the subjectof a Final Environmental Assessment (FEA) datedSeptember 1996 (USDI 1996). Part of the solutionto facilitate visitation to the park was the tram sys-tem, which is why the tram was of interest in thisstudy. Our research was initiated as a result of theFEA process.

The essence of the refuge’s solution is spatial andseasonal control of public access (USDI 1996). Atspecific times of the year, people are routed awayfrom areas where migratory bird concentrations arehighest, and from 1 November–31 March theimpoundment system is closed to the public alto-gether. Our study emphasizes the need for contin-ued seasonal closure of the impoundment systemin order to protect overwintering dabbling ducksfrom disturbance at a time when fat deposition andenergy conservation are important. It is important

110 Wildlife Society Bulletin 2005, 33(1):103–112

13 (Pease et al).qxp 6/24/2005 12:35 PM Page 110

that human disturbance levels on the dike roads donot limit the use of the impoundment system bythe birds for which it was created. Although theBack Bay watershed currently is rural in nature,development is steadily encroaching (United StatesFish and Wildlife Service 1993). As this trend con-tinues, public-use pressures are likely to increase atBBNWR.

A tram or bus system is an access method thatmay reduce the impact on birds and other wildlifewhile allowing the public to see wildlife in naturalareas. Trams or buses not only reduce the rate ofdisturbances but also potentially eliminate the mostdisruptive disturbances: humans walking or biking(Pease 2001) and humans approaching birds (Klein1993). A bus or tram system enhances wildlifeviewing, especially with knowledgeable driversproviding environmental interpretation. Also,manypeople watching together increases the chance ofspotting wildlife and more wildlife may be presentdue to lower disturbance rates.

The adverse effects of human disturbance arecompounded by other anthropogenic stresses onbird populations, such as degradation of wetlandhabitat through development, pollution, and inva-sive species. Impacts of human disturbance alsoare likely compounded for wintering and migratingbirds during times of unusual weather patterns orstorm events that can significantly alter food sup-plies (Owens 1977). Because waterfowl face con-tinued loss of habitat, the need to minimize humanrecreational impacts in remaining habitat is impera-tive.

Acknowledgments. We are grateful to J.P.Stasko,Refuge Manager, BBNWR, and all other refuge stafffor their exemplary cooperation. R. M. Erwin ofthe University of Virginia offered advice in studydesign and field implementation. We thank the vol-unteers who performed the human disturbancetreatments; several were members of the VirginiaBeach Chapter of the National Audubon Society.We appreciate A. W. Pease for help with buildingblinds and formatting tables. Special thanks aredue M. M. Mayfield, Executive Director of theElizabeth River Project, for a flexible work sched-ule that enabled M. L. Pease to conduct this study.D. K. Emminger, graphic designer, Old DominionUniversity, created Figure 1. L. Johnson served asresearch assistant in the first season of the study.This study was partially funded through aChallenge Grant from the National Fish andWildlife Foundation to BBNWR.

LLiitteerraattuurree cciitteeddANDERSON, D.W.,AND J. O. KEITH. 1980. The human influences of

seabird nesting success: conservation implications.Biological Conservation 18:65–80.

ANKNEY, C. D., AND C. D. MACINNES. 1978. Nutrient reserves andreproductive performance of female lesser snow geese. TheAuk 95: 459–471.

BÉLANGER L., AND J. BÉDARD. 1989. Responses of staging greatersnow geese to human disturbance. Journal of WildlifeManagement 53: 713–719.

BOYLE, S.A., AND F. B. SAMSON. 1985. Effects of nonconsumptiverecreation on wildlife: a review. Wildlife Society Bulletin 13:110–116.

BURGER, J. 1981. The effect of human activity on birds at a coastalbay. Biological Conservation 21: 231–241.

HAVERA S. P., L. R. BOENS, M. M. GEORGI, AND R. T. SHEALY. 1992.Human disturbance of waterfowl on Keokuk Pool,Mississippi River. Wildlife Society Bulletin 20: 290–298.

KLEIN, M. L. 1993. Waterbird behavioral responses to human dis-turbances. Wildlife Society Bulletin 21: 31–39.

KLEIN,M.L.,S.R.HUMPHREY,AND H.F.PERCIVAL. 1995. Effects of eco-tourism on distribution of waterbirds in a wildlife refuge.Conservation Biology 9: 1454–1465.

KRAPU,G.L. 1981. The role of nutrient reserves in mallard repro-duction. The Auk 98: 28–38.

OWENS, N.W. 1977. Responses of wintering brent (sic) geese tohuman disturbance.Wildfowl 28: 5–14.

PEASE, M. L. 2001. Effects of human disturbances on the behav-ior of dabbling ducks. Thesis, Old Dominion University,Norfolk,Virginia, USA.

POMERANTZ, G.A., D. J. DECKER, G. R. GOFF, AND K. G. PURDY. 1988.Assessing impact of recreation on wildlife: a classificationscheme. Wildlife Society Bulletin 16: 58–62.

PURDY, K. G., G. R. GOFF, D. J. DECKER, G. A. POMERANTZ, AND N. A.CONNELLY. 1987. A guide to managing human activity onnational wildlife refuges. Human Dimensions Research Unit,Cornell University, Ithaca, New York, USA.

UNITED STATES DEPARTMENT OF THE INTERIOR. 1996. Final environ-mental assessment: proposal to resolve access issues withFalse Cape State Park at Back Bay National Wildlife Refuge,Virginia Beach, Virginia. United States Fish and WildlifeService Region 5, Hadley, Massachusetts, USA.

UNITED STATES FISH AND WILDLIFE SERVICE. 1993. Station manage-ment plan:Back Bay National Wildlife Refuge. Virginia Beach,Virginia, USA.

VASKE, J. J.,A. R. GRAEFE,AND F. R. KUSS. 1983. Recreation impacts:a synthesis of ecological and social research. Pages 96–107in Transactions of the 48th Conference on North AmericanWildlife and Natural Resources. Wildlife ManagementInstitute, 19 March–24 March 1983, Kansas City, Missouri,USA.

ZAR, J. H. 1999. Contingency tables. Pages 486–515 in TeresaRyu, editor. Biostatistical analysis. Prentice Hall, UpperSaddle River, New Jersey, USA.

Melissa L. Pease (photo) earned a B.S. from the United StatesNaval Academy and an M.S. from Old Dominion University.She spent five years planning and directing several upland andwetland habitat restoration projects in Virginia. While workingfor a nonprofit river restoration organization, she provided tech-nical consultation, project management, and public outreachservices for businesses and citizens voluntarily undertaking

Human disturbance of wintering ducks • Pease et al. 111

13 (Pease et al).qxp 6/24/2005 12:35 PM Page 111

wildlife habitat enhancement projects using native plants. Shealso led a community advocacy group whose work resulted inthe City of Virginia Beach, Virginia purchasing 1,000 acres offorested wetlands for open space and wildlife habitat conserva-tion. Ms. Pease’s current position is the mindful caregiver forher two young daughters. Robert K. (Bob) Rose is ProfessorEmeritus of the Department of Biological Sciences at OldDominion University in Norfolk, Virginia. He attended theUniversity of Wisconsin, Madison (B.S.), University of Virginia(M.S.), and University of Kansas (Ph.D.). Major previous posi-tions were at the University of Kansas, University of Virginia,and Georgetown University. He is a past member of TheWildlife Society and former contributor to The Journal ofWildlife Management. Dr. Rose’s research interests focus onnon-game bird and mammal wildlife, especially behavior andpopulation–community ecology. Mark J. Butler is currently aprofessor in the Department of Biological Sciences at OldDominion University (ODU) in Norfolk, Virginia, where he hasbeen on the faculty since 1989. Prior to arriving at ODU, heheld postdoctoral research positions at the Center for Limnologyat the University of Wisconsin, Madison and at Florida StateUniversity. He received his Ph.D. from Florida State University,an M.S. degree from Ohio State University, and a B.A. degreefrom Wittenberg University. Dr. Butler’s research interests are inmarine ecology and biostatistics, with an emphasis on tropicalsystems.

Associate editor: Applegate

112 Wildlife Society Bulletin 2005, 33(1):103–112

13 (Pease et al).qxp 6/24/2005 12:35 PM Page 112