Chick Behavior, Habitat Use, and Reproductive Success of Piping Plovers at Goosewing Beach, Rhode Island

J. Field Ornithol., 69(2):228-234

CHICK BEHAVIOR, HABITAT USE, AND REPRODUCTIVE SUCCESS OF PIPING PLOVERS AT GOOSEWING BEACH, RHODE ISLAND

MER• R. GOLDIN• •D JONATHANV. REGOSIN The Nature Conservancy 45 SouthAngell Street Providence, Rhode Island 02906 USA

Abstract.--We studied Piping Plovers (Charadrius melodus)at GoosewingBeach, Rhode Is- land during 1993 and 1994. Broodswith accessto salt-pondmudflat habitat experienced higher fledging success(3.0 fledglings/brood) than broodslimited to oceanbeachfront habitat (1.4 fledglings/brood). This difference may be attributed to greater survivorshipamong chickswith accessto mudflat habitat. Broodsusing pondshoremudflat habitat spent less time respondingto human disturbance(1.6%) than chicksusing ocean beachfront habitat (17.0%). The differencein time spent feeding betweenpondshore (77.5%) and ocean beachfront(51.2%) habitatsapproached statistical significance. Salt pond waterlevels were artificiallymanipulated to increaseavailability of mudflat habitatfor plover chicks,though the effect of such manipulation on Piping Plover reproductivesuccess remains unclear.

CONDUCTA DE PICHONES, USO DE HABITAT Y •ITO DE ANIDAMIENTO DE INo DIVIDUOS DE CHARADRIUS MELODUS DE LA PLAYA GOOSEWING, RHODE ISLAND Sinopsis.--Duranteel 1993 y 1994 se estudiaronindividuos de Charadriusmelodus en la playa Goosewing,Rhode Island. Camadas con accesoa lodazales de charcasde sal fueron mJs exitosas(3.0 volantones/camada)que aquellasque estuvieronlimitadas a frentes de playas (1.4 volantones/camada).Esta diferencia puede ser atribuida a una mayorsupervivencia de pichonesque tuvieron accesoa lodazales.Las camadasque utilizaron este tipo de habitat tomaronmenor tiempopara respondera disturbiospor parte de humanos(1.6%) que los pichonesque utilizaron frentes de playas(17%). La diferenciaen tiempo utilizado para alimentarseen lodazales(77.5%) y frentesde playas(51.2%) se acercaa diferenciasestad- isticassignificativas. Los nivelesde agua de las charcasde sal fueron manipuladosartificial- mente para incrementar la disponibilidadde lodazalespara los polluelos,aunque el efecto de dicha manipulaci6nen el •xito reproductivode la especieno queda del todo claro. The Piping Plover, a threatened shorebird, breeds on open beaches and sandflats on North America's northern Great Plains, Great Lakes, and Atlantic Coast (Haig 1992). Numbers of Piping Plovershave declined over the past 50 years (U.S. Fish and Wildlife Service 1988), resulting in the placement of this specieson the U.S. Endangered and Threatened SpeciesList (U.S. Fish and Wildlife Service 1985). Severalfactors includ- ing direct human disturbanceon the breeding grounds (Burger 1987, Cairns 1982, Elias-Gerkin 1994, Flemming et al. 1988, Goldin 1993, Strauss1990), habitat destructionthrough development,water level reg- ulation, and predation have been suggestedas contributing factorsto the decline of this ground-nestingspecies (Haig 1992). Atlantic Coast Piping Ploversforage in a number of habitatsincluding ocean beach, ocean intertidal zone, bay beach intertidal zone, shoreline of ephemeralpools and saltponds, overwash area, wrack, and dunes(Elias- Gerkin 1994, Goldin 1993, Hoopes 1993, Loegering and Fraser 1995).

1 Current address:P O. Box 4423, PagoPago, American Samoa 96799 USA.

228 Vol.69, No. 2 PipingPlover Hath'tat Use [229

However,Piping Ploverswith accessto mudflat, bay-beach,and ephemeral pool habitatsmay disproportionatelyuse thesehabitats, and enjoyrepro- ductive advantages(Elias-Gerkin 1994, Hoopes 1993, Loegering and Fra- ser 1995, Pattersonet al. 1991). Hypothesizedsources of theseadvantages include decreased human disturbance rates (Elias-Gerkin 1994, Goldin 1993), or a particularlyrich food resourcefor Piping Plover chicks(Elias- Gerkin 1994, Hoopes 1993, Loegering and Fraser 1995). Becauseincreased availability of mudflat habitat may benefit Piping Ploversby providing high-qualityforaging habitat and decreasedhuman disturbance,the Atlantic CoastPopulation Piping Plover RevisedRecov- ery Plan (U.S. Fish and Wildlife Service1995) encouragesthe draw-down of coastalponds, in some cases,to make more feeding habitat available. However,no studiesdirectly demonstrateincreased reproductive success for broods with accessto coastalpond mudflat habitat. This studytests the hypothesisthat broods with accessto coastalpond mudflats experi- ence greater survivorshipand fiedging successthan broods that are restricted to ocean beachfront habitat. In addition, we expected broods using mudflats to spend lesstime respondingto human disturbanceand more time feeding than broods restricted to the ocean beachfront.

METHODS We studied Piping Plovers at GoosewingBeach in Little Compton, Rhode Island from April-August 1993 and 1994, as part of The Nature Conservancy'songoing Piping Plover and Least Tern (Sterna albifrons) protection program. Goosewingis a cobbly barrier beach, 1.4-km long, bordered to the north by the shallow,158-ha Quicksand Pond. The beach typicallybreaches, forming an outlet to the ocean, as pond water levels rise. In addition, the beach wasbreached mechanically as a management measureonce during each nesting season.Initially, breaching lowerswa- ter levelswithin the pond and allowsfor water exchangewith the ocean. The breach usually closeswithin 1-2 wk. Thus, throughout the spring and summer,varying amountsof sandymudfiats were exposedalong the pond-shore.Piping plover chicksand adultsforaged on these mudfiats. Eight and nine pairs of Piping Ploversnested on GoosewingBeach in 1993 and 1994, respectively,accounting for approximately25% of Rhode Island's Piping Plover population. We monitored plovers daily from a distanceof 50-75 m with spottingscopes and binoculars.We surrounded nestscontaining two or more eggswith a net-topped,wire meshexclosure to reduce depredation (Melvin et al. 1992). Clutch size,hatching success, and fiedging successwere recorded for each brood. Chicks observedin flight or reaching 25 daysof age (whichevercame first) were considered fledged (U.S. Fish and Wildlife Service 1988). We attempted to count all chicksfrom all broodsdaily and to map the location of the brood when counted. Because there is a moratorium on banding Piping Ploversto minimize disturbance,birds were unbanded with few exceptions.Therefore, we noted distinctiveneckbands, head- bands,and other plumage characteristics.These features,along with plo- 230] •4. R. Goldinand J. V. Regosin J.Field Ornithol. Spring 1998 ver territoriality and asynchronoushatching of broods,made it possible to identify breeding pairs and their associatedbroods. Becausechicks were usuallycounted daily,we could determine chick mortality to within 2 days (approx. 36 h) in 14 of 26 (54%) cases.However, not all chicks were counted each day,and mortalitycould only be determined to within 3 daysin eight (31%) cases,and to within 4 daysin four (15%) cases. During 1993 we conductedbehavioral observationson plover chicks from 15 June-3 Augustbetween 0830-1700 h EST. A subsetof broods were randomly selectedfor samplingeach day and sampledin random order. We grouped chick behaviorsinto four classes:(1) feeding (peck, probe, swallow,walk between pecksor probes); (2) disturbance(freeze, run or walk in apparent response to disturbance); (3) maintenance (bathe, preen, stand,loaf); and (4) other (walk or run in the absenceof feeding or disturbance,agonistic behavior, brooding). The first chicken- counteredin a givenbrood servedas the focal individualand its behavior was continuouslyrecorded for up to 15 min (Martin and Bateson1986). If the chick disappearedfrom view, observationwas halted, and, if the chick did not reappear within 5 rain, the samplingsession was aborted. Observation sessionswere subsequentlyconducted for all other chicks present. If no other chickswere encounteredwithin 10 rain, data collec- tion began on the next randomlyselected brood. Only observationses- sionsthat lasteda full 15 min were included for analysis. For each behavioral observation we recorded date, time, brood number, chick age, and habitat. For analysis,we divided habitats into two classes:(1) mudflat (wet, sandy,partially organic substrateadjacent to Quicksand Pond) and (2) ocean beach (intertidal zone, beachfront in- cluding broad "overwash"areas, and dunes). In addition to recording the actual habitat used by broods during behavioral observations,all broodswere divided into two classes:(1) those that were observedusing mudflat habitat during the prefledging period, and (2) those that were never observedusing mudflatsand spent time exclusivelyon beach habitat. Becausemultiple behavioralobservations from individual broods lack statisticalindependence, for each brood we computed the mean time/ observationperiod spentin each behavioras the unit of analysis.For one brood for which behavioraldata wasavailable from both habitat types,we computedmeans for each habitat.We used two samplet-tests (Sokal and Rohlf 1981) to test for differencesin mean time spent feeding and respondingto human disturbanceby broods using mudflat versusocean beach habitats.

RESULTS Fledgingsuccess and chicksurvivorship.mMean fledging success(number of chicksfledged/brood) was 3.0 (SD = 1.6, n -- 5) and 3.0 (SD = 1.3, n = 6) for broods with accessto mudflats and 1.3 (SD = 1.0, n -- 4) and 1.7 (SD = 1.5, n = 3) for broods limited to the ocean beach and dunes in 1993 and 1994, respectively.Because all pairs hatched chicks Vol.69, No. 2 PipingPlover Hath'tat Use [231

ß Beach [] Mudflat

0 I 2 3 4 5 No. Chicks Fledged

F[c;um•1. Fledging successfor Piping Plover broods with and without accessto mudfiat habitat at GoosewingBeach, Rhode Island. Brood of five chicks resulted from an interpair adoption (see Methods). and no pairs fledged chicksfrom more than one brood, these resultsare equivalentto the mean number of chicksfledged/breeding pair, the stan- dard measureof productivity(U.S. Fish and Wildlife Service1988, 1995). Combining data from both years,mean fledging successwas higher for broods with accessto mudflats (k = 3.0, SD -- 1.3, n = 11) than broods limited to beach and dunes (k = 1.4, SD = 1.1, n = 7; t -- 2.56, df = 16, P -- 0.021; Fig. 1). Neither hatching success(eggs hatched/eggs laid; mudflat: 97.7%, n = 43 eggs;beach: 96.4%, n = 28 eggs;X 2= 0.10) nor initial brood size (mudflat: • = 3.7; n = 11; SD = 1.0; beach: • = 4.0; n = 7; SD = 0.8; t = 0.59; df = 16; P = 0.55) differed for broods from the two habitats.Thus, the difference in fledging successby habitat can be attributed to differencesin chick survivorship,with 33 of 41 (80.5%) chickswith accessto mudflat habitat survivingto fledging, comparedto only 10 of 28 (35.7%) chickslimited to beach habitat (Fig. 2). Behavioralobservations.--Fifty-seven observation sessions were complet- ed on nine broods (one pair was double-brooded).Because we were in- terested in comparing chick behavior between the mudflat and beach habitats,it wasimportant to samplechicks of comparableages from the two habitats.Chicks in beach habitat ranged from 1-22-d old when sam- pied, while chicksin mudflat habitat ranged from 4-27-d old. Becauseof two interpair chick adoptions,not all chick agescould be preciselydeter- mined at the time theywere sampled.Nonetheless, the maximumpossible age of chicksin mudflat habitat (• -- 15.7 d, n = 46, SD = 6.6) did not differ significantlyfrom the minimum possibleage for chicksusing beach habitat (• = 13.3 d, n = 11, SD = 5.6; t -- 1.14; df = 55; P = 0.256). In mudflat habitat, broods experiencedhuman disturbanceevents at an averagerate of 1.0/h, as comparedwith a rate of 7.3/h for ocean beach habitat (t -- 3.87, df = 8, P = 0.005). All human disturbances involved pedestriansor joggers. Gulls (Larus spp.), terns (Sternaspp.), Mute Swans( Cygnusolor), Osprey (Pandion haliaetus), and AmericanOys- 232] M. R. Goldin and J. V. Regosin J. Field Ornithol. Spring 1998

1.0•1

0.9

0.8

0.7 o Mudflat ß Beach 0.6- 0.51

0.4-

0.3 0 5 10 15 20 25 Day of Nestling Period FIGURE2. Survivorshipcurves of prefledging Piping Plover chickswith accessto mudflat and ocean beach habitat at GoosewingBeach, Rhode Island. Error bars (d) are shown only for mortality events that could not be dated to within two consecutivecalendar days. tercatchers (Haematopuspalliatus) also elicited disturbance responses from Piping Plover chicks.Some, but not all of thesespecies are potential predators of plover chicks.The rates of nonhuman disturbancein mudflat (5.8/h) versusocean beachfront (2.0/h) habitat did not differ sig- nificantly (t = 1.75, df = 8, P = 0.117). The duration of human disturbances (.• = 62 s, SD = 69, n = 28) was greater than the duration of nonhuman disturbances (.• = 25 s, SD = 27, n = 31; t = 2.75, df = 57, P = 0.008). Broodswith accessto mudflats spent lesstime respondingto human disturbances than broods limited to beach habitat (t = 2.44, df = 8, P = 0.040, Table 1). The observeddifference in time spentfeeding in the two habitats (Table 1) approachesstatistical significance (! = 2.06, df = 8, P = 0.073).

DISCUSSION Chick survivorshipand fledging successwere higher for broods with accessto salt-pondmudflats, supporting the Piping PloverRecovery Team

TABLE1. Mean percentagetime spent in variousactivities for Piping Plover broodsusing mudflat and ocean beach habitats, GoosewingBeach, Rhode Island.

Mudflat (n = 6) Beach (n = 4) Activity Mean SD Mean SD P• Feeding 77.5 15.8 51.2 25.0 0.073 Maintenance 16.9 13.9 13.2 8.3 Disturbance (human) 1.6 1.7 17.0 15.8 0.040 Disturbance (nonhuman) 0.8 0.8 3.7 2.6 Other 3.1 3.2 14.9 14.1

t-testsperformed on feedingand human disturbancebehaviors only (seeMethods). Vol.69, No. '• PipingPlover Haln'tat Use [233 recommendation to considercoastal pond draw-downas a Piping Plover managementtool (U.S. Fish and Wildlife Service1995). Fledging success at Goosewingbeach continued to be higher for mudflat habitat in 1995 (mudflat: • = 2.4, SD -- 1.5, n = 5; beach: • = 1.3, SD = 1.5, n = 4; Campelloneand Hadjian, unpubl. data) and 1996 (mudflat: • = 3.0, SD = 0.0, n = 3; beach:• = 0.7, SD -- 1.0, n = 7; Fontes,unpubl. data). Further evidencethat mudflats are, in fact, preferred brood-rearinghab- itat is offered by observationsof breeding pairs at GoosewingBeach mov- ing their broods over considerabledistances through densevegetation in order to gain accessto the pondshore. Findings from this study are consistentwith other research that has shownhigher chick survivorshipfor chicksusing bay-beach island interior, and ephemeral pool habitats (Elias-Gerkin1994, Loegering and Fraser 1995). Furthermore, we found that chicks utilizing pondshore habitat spentless time respondingto human disturbance,and possiblymore time foraging. Though we were unable to draw a direct connectionbetween increased human disturbance and decreasedforaging time, human disturbance has been found to decreasetime chicks devoted to feeding at other sites (Fleming et al. 1988, Strauss1990). The observed reduction in human disturbance associated with salt- pond habitat may have contributed to the observed increase in reproductive success. However, other studies have shown that mudflat habitat may be a particularly rich food resourcefor Piping Plover chicks (Elias- Gerkin 1994, Hoopes 1993, Loegering and Fraser 1995). While the cause of the observed increase in chick survivorshipassociated with mudflats remainsunclear, the magnitudeof the effect suggeststhat augmentation of mudflats through control of water levels may be a powerful management technique at a number of salt pond breeding sites,especially in southeasternMassachusetts, Rhode Island, and Long Island. Artificial breaching of salt ponds may also benefit migrating shorebirdsand some anadromousfish speciessuch as Alewife (Alosapseudoharengus). However, two cautions are in order. First, while water levels were ma- nipulated on a limited basisat Quicksand Pond, it is unclear the extent to which artificially increased availability of mudflat habitat, over and above the mudflats that would have been available under natural condi- tions, contributed to increasedsurvivorship. Second, salt pond manipulation through artificial breaching can have multiple affectson salt pond ecology (Lee 1980). Changesin water level and salinitymay affect sub- merged aquaticvegetation as well as a wide varietyof animal species.The establishmentof permanent breachwaysin many of Rhode Island'slarger salt pondshas causedmajor, unforeseenchanges in theseecosystems (Lee 1980). Sincethe potential long term affectsof regular artificialbreaching are not well understood,managers would be well advisedto considerall potential long term costsand benefits before undertaking salt pond manipulation to benefit Piping Plovers. 234] M. R. GoldinandJ. V. Regosin J.Field Ornithol. Spring 1998

ACKNOWLEDGMENTS

We thank The Nature Conservancy,the town of Little Compton, E. Lawrence, the Oliver S. and Jennie R. DonaldsonCharitable Trust, and numerousvolunteer wardensfor protect- ing Piping Ploverson GoosewingBeach and making this researchpossible. Special thanks to E. Rowland,E. Stephens,and T. Santiagofor assistancein the field, and to T. Bear, R. Berkowitz,V. Carpenter, A. Leviton, G. Venator, and four anonymousreviewers for their encouragetnentand cotmnentson earlier versionsof the tnanuscript.

LITERATURE CITED

BURGER,J. 1987. Physicaland socialdeterminants of nest site selectionin Piping Ploverin New Jersey.Condor 89:881-918. C^m_•s,W. E. 1982. Biology and behavior of breeding Piping Plovers.Wilson Bull. 94:531- 545. ELI^S-GE•O•N,S. 1994. Piping Plover habitat suitabilityof central Long Island, New York, Barrier Islands. M.S. thesis.Virginia PolytechnicInstitute and State University,Blacks- burg, Virginia. 247 pp. FLEMMING,S. P., R. D. CHIASSON,P. C. SMITH,P.J. AUSTIN-SMITH,AND R. P. BANCROFT.1988. Piping Ploverstatus in Nova Scotiarelated to reproductiveand behavioralresponses to human disturbance.J. Field Ornithol. 59:321-330. GOLDIN,m. R. 1993. Effectsof hutnan disturbanceand off-road vehicleson Piping Plover (Charadrius melodus)reproductive success and behavior at Breezy Point, GatewayNa- tional Recreation Area, New York. M.S. thesis. University of Massachusetts,Amherst, Massachusetts.143 pp. H•IG, S. M. 1992. Piping Plover (Charadrius melodus).No. 2, in A. Poole, P. Stettenheim, and E Gill, eds.The birds of North America.Academy of Natural Sciences,Philadelphia and American Ornithologists'Union, Washington,D.C. 18 pp. HOOPES,E. M. 1993. Relationshipsbetween hutnan recreation and Piping Plover foraging ecologyand chick survival.M.S. thesis,Univ. of Massachusetts,Amherst, Massachusetts. 106 pp. LEE,V. 1980. An elusivecompromise: Rhode Island coastalponds and their people. Coastal ResourcesCenter, Marine TechnicalReport 73. Universityof Rhode Island, Narragan- sett,Rhode Island.82 pp. LOEGERING,J.P., •X•NOj. D. F•$ER 1995. Factors affecting Piping Plover chick survivalin different brood-rearinghabitats. J. Wildl. Manage. 59:646-655. MARTIN,P., ANDP. BATESON.1986. Measuringbehaviour. Cambridge Univ. Press,Cambridge, United Kingdom.200 pp. MELVIN,S. M., L. H. M^CIVOR,AND C. R. GRIFFIN.1992. Predator exclosures:a technique to reduce predation at Piping Plovernests. Wildl. Soc.Bull. 20:143-148. PATTERSON,M. E.,J. D. F•$ER, ^NDJ. W. ROGGENBUCK.1991. Factorsaffecting Piping Plover productivityon AssateagueIsland. J. Wildl. Manage.55:525-531. SOKAL,R. R., ANDF.J. ROHLF.1981. Biometry.W. H. Freeman and Co., New York. 859 pp. STyUSS,E. 1990. Reproductivesuccess, life historypatterns, and behavioralvariation in a population of Piping Ploverssubjected to hutnan disturbance.Ph.D. diss.Tufts Univer- sity,Medford, Massachusetts.143 pp. U.S. FISHAND WILDLIFE SERVICE. 1985. Determination of endangeredand threatened status for the Piping Plover.Federal Register50:50720-34. 1988. Atlantic coast Piping Plover recoveryplan. U.S. Fish and Wildlife Service, Newton Corner, Massachusetts.77 pp. --. 1995. Piping Plover (Charadriusmelodus) Atlantic coastpopulation revised recovery plan. U.S. Fish and Wildlife Service,Hadley, Massachusetts. Received8 Jan. 1997; accepted28 Apr. 1997.