Raptor Predation on Wintering Dunlins in Relation to the Tidal Cycle
Author(s): Dick Dekker and Ron Ydenberg
Source: The Condor, Vol. 106, No. 2 (May, 2004), pp. 415-419
Published by: Cooper Ornithological Society
Stable URL: http://www.jstor.org/stable/1370651
Accessed: 19-05-2015 18:31 UTC
REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/1370651?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.
Cooper Ornithological Society is collaborating with JSTOR to digitize, preserve and extend access to The Condor.
http://www.jstor.org
This content downloaded from 142.58.26.133 on Tue, 19 May 2015 18:31:34 UTC
All use subject to JSTOR Terms and Conditions
SHORTCOMMUNICATIONS 415
1999. Temporal and geographic variaABOOKIRE.
tion in fish communities of lower Cook Inlet,
Alaska. Fishery Bulletin 97:962-977.
Hobart,Tasmania.
2000. Benthic and peTREMBLAY,
Y., ANDY. CHEREL.
lagic dives: a new foraging behaviourin RockhopperPenguins.MarineEcology ProgressSeries
204:257-267.
8.1. SAS InstituteInc., Cary, NC.
I. 1995. Marineecological processes.SpringSEALY, S. G. 1973. Adaptive significance of post- VALIELA,
er, New York.
hatchingdevelopmentalpatternsand growthrates
in the Alcidae. Ornis Scandinavica4:113-121.
ZAR, J. H. 1999. Biostatisticalanalysis. Simon and
LTD.2000. Echoview v. 2.1.
SONARDATAPROPRIETARY
Schuster,UpperSaddle River,NJ.
SAS INSTITUTE.
2000. SAS/STAT user's guide, version
The Condor 106:415-419
? The Cooper Ornithological Society 2004
RAPTOR PREDATION ON WINTERING DUNLINS IN RELATION TO THE TIDAL CYCLE
DICK DEKKER1,3AND RON YDENBERG2
13819-112 A Street, Edmonton, AB T6J 1K4, Canada
2Department of Biological Sciences, Simon Fraser University, Burnaby, BC V5A 1S6, Canada
Abstract. At Boundary Bay, British Columbia, 94 ejemplaresde Calidris alpina en 652 horas. Los
Canada, Peregrine Falcons (Falco peregrinus) cap- dos m6todosprincipalesde caza fueronataquesabiertured 94 Dunlins (Calidris alpina) in 652 hunts. The tos sobreindividuosque estabanvolando (62%)y atatwo mainhuntingmethodswere open attackson flying ques encubiertossobreindividuosque estabanposados
Dunlins (62%) and stealthattackson roostingor for- o forrajeando(35%).F. peregrinuscaz6 a lo largo del
aging Dunlins (35%). Peregrineshunted throughout dia, pero la tasa de matanzapor hora de observaci6n
the day, yet the kill rateper observationhourdropped disminuy6 1-2 hr antes de la pleamary alcanz6 un
1-2 hr before high tide and peaked 1-2 hr after high maximo 1-2 hr despu6sde la pleamar.La caida en la
tide. The dropin kill ratecoincidedwith the departure tasa de matanzacoincidi6 con la partidaen masa de
of the mass of Dunlins for over-oceanflights lasting C. alpina pararealizarvuelos sobreel oc6anoque du2-4 hr.The peak in kill rateoccurredjust afterthe tide raron 2-4 hr. El pico en la tasa de matanzaocurri6
began to ebb and the Dunlinsreturnedto foragein the justo despu6sde que la mareacomenz6 a menguary
shore zone. The hypothesisthatcloseness to shoreline de que los individuosde C. alpina regresarona forravegetation is dangerousfor Dunlins is supportedby jear a la zona de playa.La hip6tesisde que la cercania
three converginglines of evidence: (1) the high suc- de la vegetaci6na la linea de playa es peligrosapara
cess rate (44%)of peregrinehuntsover the shorezone C. alpina es apoyadapor tres lineas convergentesde
comparedto the rate (11%) over tide flats and ocean; evidencia:(1) la alta tasa de 6xito (44%) de las cace(2) the high kill rateper observationhourat high tide; rias de F. peregrinussobrela zona de playacomparada
and (3) the positive correlationof kill rate with the con la tasa (11%) de las cacerias sobre los planos de
heightof the tides. Seven of 13 Dunlinskilled by Mer- la mareay el oc6ano; (2) la alta tasa de matanzapor
lins (Falco columbarius)and all five Dunlinskilled by hora de observaci6ndurantela pleamar;y (3) la corNorthernHarriers(Circuscyaneus)were also captured relaci6n positiva de la tasa de matanzacon la altura
de las mareas. Siete de 13 individuos de C. alpina
in the shore zone.
cazados
por F. columbariusy todos 5 individuosde
words:
Calidris
Falco
Key
alpina,Dunlin,
peregrinus, Peregrine Falcon, raptor predation, tidal cycle.
Depredaci6n de Calidris alpina por Rapaces
durante el Periodo Invernal con Relaci6n al
Ciclo de la Marea
Resumen. En la Bahia Boundary,Columbia Brit~nica, Canada,halcones Falco peregrinuscapturaron
Manuscriptreceived29 May 2003; accepted12 November2003.
3E-mail:tj.dick-dekker@hotmail.com
C. alpina cazados por Circus cyaneus tambi6n fueron
atrapadosen la zona de playa.
Predationrisk has been implicatedby many researchers as an importantdeterminantin the feedingbehavior
of a wide variety of prey species (Lima et al. 1985,
Milinski 1986). Accordingto theory, avian prey species balance predationrisk with foraging needs. For
instance, in a trade-offbetween relative safety from
predatorsand optimal caloric gain, forest passerines
tend to forage close to the protectivecover of trees
and bushes, whereas open-country birds stay well
This content downloaded from 142.58.26.133 on Tue, 19 May 2015 18:31:34 UTC
All use subject to JSTOR Terms and Conditions
416
SHORT COMMUNICATIONS
away from vegetationthat could conceal raptors(Valone andLima 1987, Lima 1988). In estuarinehabitats,
shorebirdsfrequentopen expanses of mudflatswhich
allow for the timely discoveryof approachingraptors.
Raptorsthathuntshorebirds,such as the PeregrineFalcon (Falco peregrinus) and the Merlin (Falco colum-
barius), commonly use stealth methods to take their
prey by surprise(Page and Whitacre 1975, Dekker
1980, 1988, Palmer1988, Cresswell 1996, Whiteet al.
2002). If rising tides inundatethe mudflatsand force
the shorebirdsclose to the vegetated high-tide line,
most leave andfly to roostingsites in adjacentcountry.
In some locations,shorebirdsroost in habitatsthey do
not normallyfrequent,such as agriculturallands(Butler 1994) or wave-sweptbeaches(Buchanan1996). On
the northwest Pacific coast of North America, and
probablyelsewhere in its extensive winteringrange,
Dunlin flocks fly out over the ocean and remain in
flightfor 2-4 hr until the tide turns.This behaviorwas
strategyby Breninterpretedas a possible antipredator
nan et al. (1985) and firstdocumentedas suchby Dekker (1998). Hotker(2000) saw the same phenomenon
on the northcoast of Germanyandtermedit "airborne
roosting."Stayingin flightover the waterfar offshore
makes sense if roosting sites close to the high-tide
markare dangerous.Ydenberget al. (2002) foundthat
the choice of stopover sites by migrating Western
Sandpipers(Calidrismauri)in BritishColumbia,Canada, could best be explained by a hypothesis that
sandpipersare more vulnerableto raptorpredationon
small feeding sites than on wide expanses of open
mudflats.
In this article, we examine a large data set of observed huntsand kills by PeregrineFalconsto test the
hypothesis that proximity to shoreline vegetation is
dangerousfor Dunlins,and thatDunlinstend to avoid
risk when they are satiated,but are more willing to
take risk when hungry.As additionalevidencefor that
hypothesis,we also reporton Dunlin kills by Merlins
(Falco columbarius) and Northern Harriers (Circus
cyaneus).
METHODS
The study areawas at BoundaryBay, on the southern
edge of the FraserRiver estuary(49005'N, 123000'W)
in BritishColumbia,Canada.The bay is 16 km across
and the intertidalzone is roughly 4 km wide at the
lowest ebb. The tidal rhythmincludestwo flood tides,
one usually higher than the other, per 24-hr period.
During winter the highest tides almost always occur
duringdaylighthours and inundateall intertidalmudflats and most of the narrowstripof saltmarsh,which
is covered with low vegetation.A dyke protectslowlying agriculturalfields inland.BoundaryBay is a major stopoverfor migratorywaterbirdsand a wintering
locationfor circa 50 000 Dunlinsand 1000 Black-bellied Plovers (Pluvialis squatarola). Birds of prey are
common (Butler and Campbell 1987). In winter,the
bay is huntedover by at least six peregrinesand one
or more Merlins(Dekker2003).
BetweenearlyNovemberandearlyFebruary,19942003, DD spentpartor all of 151 days (940 hr) in the
study area, walking the dyke or sitting in a parked
vehicle. Flocks of Dunlins were monitoredfor alarm
behavior such as sudden flushing. Hunting raptors
were also discoveredby frequentlyscanningthe area
through8x wide-anglebinoculars.Perchedperegrines
and Merlinswere often kept undersurveillancefor periods of up to 2 hr in the hope of seeing them hunt.
We use the term "hunt"to mean a completedattack
of which the outcome was known. A hunt could include one or more passes or swoops at the same Dunlin. An attackon a flock and subsequentpursuitof a
single Dunlin fleeing that flock were counted as one
hunt.However,if the falcon abandonedthe pursuitand
again attackedthe same or a different flock, it was
tallied as anotherhunt. This definitionof a hunt was
also used by Dekker (1980, 1988, 2003) and is the
equivalent of the term "attack" as formulatedby
Cresswell (1996). In this paper,both terms are used
interchangeably.
Field data were recordedin diary form and entered
into an annotatedtable of huntsand kills, dividedover
three zones. Zone 1 representedthe saltmarshshore
includinga 5-10 m stripof wrackand sparselyvegetated mud beyond the raggedmarshedge. To investigate whetherperegrinehuntingsuccess was influenced
by distance from shore, we arbitrarilysplit the intertidal zone in two: zone 2 extended roughly 0.5 km
from the saltmarsh;zone 3 lay beyond zone 2. Dependingon the tide height,zones 2 and 3 could consist
of mudflatsand ocean. Although a hunt or pursuit
might cross over from one zone into the next, the position of the prey at the startof the attackdefinedthe
zone in which the hunt was consideredto have taken
place. In a few borderlinecases, the choice amounted
to a judgmentcall or best guess. Details on Peregrine
Falcon characteristicsand huntingmethods,and Dunlin behaviorwhen avoidingraptorswere given in Dekker (2003).
We recordedthe zone (1, 2, or 3) in which each
hunt took place and the time until or since the nearest
high tide. Fromthis summary,we derivedthe kill rate
per observationhour in relation to the time of day
(intervalsof 1 hr) and the time of the nearesthigh tide,
based on the tide tablespublishedfor PointAtkinson,
BritishColumbia.In a separatedataanalysis,RY compared the kill rate to the height of the tide (grouped
into 40-cm intervals)at the time of the kill, using the
algorithmavailableat XTide (Flater1998). The result
was tested for significanceby standardregressionprocedures. The peregrinehunting success rate over the
shore zone (zone 1) was compared statistically to
zones 2 and 3 by a G-testof independence(Sokal and
Rohlf 1981).
No attemptwas madeto recordthe numberof Dunlins present in the attack zone when hunts and kills
took place. Large numbersof Dunlins do not necessarily translateinto a high kill rate. On the contrary,
singletons or small, isolated flocks of prey were reportedto be more vulnerableto stealthattacksby peregrines than large or multiple flocks spaced out over
a wide area (Dekker 1980, 1998, Thiollay 1982).
RESULTS
A total of 652 peregrinehuntsdirectedat Dunlinswas
observed. Of these, 94 ended in kills. Huntsand kills
took place throughoutdaylighthours at averagerates
This content downloaded from 142.58.26.133 on Tue, 19 May 2015 18:31:34 UTC
All use subject to JSTOR Terms and Conditions
SHORTCOMMUNICATIONS 417
0.20a.
TABLE 1. Success rates of PeregrineFalcons hunting Dunlinsat BoundaryBay, BritishColumbia,Canada. Zone 1 represents the ocean shore-saltmarsh
edge; zone 2 extends 0.5 km beyond zone 1; zone 3
begins 0.5 km from the saltmarsh.Dependingon tide
height, zones 2 and 3 could consist of mudflatsand
ocean.
0.15
0.10
0.05
08:00
10:00
12:00
14:00
Zone
16:00
Time
ofday
c-b.
0.25
.
Zone 1
Zone 2
Zone 3
Total
0.20
C)
.0.
0- 0.10
00.05
-6 -5 4 -3 -2 -1 0 +1 +2 +3 +4 +5 +6
)
from
Hours
hightide
C
C
0.20
0.15
0
0.10
,
0.05
2
3
4
5
Tideheight(m)
FIGURE1. Temporaland tidal patternsof Peregrine
Falcon predationon Dunlins wintering at Boundary
Bay, BritishColumbia,Canada.(a) Kill rate (no. capturesper observationhr) in relationto the time of day.
Numberof kills and observationtime (hr) in the successive intervalswere 1/49, 11/95, 15/117, 19/128, 13/
136, 15/126, 13/127, 3/110, 4/52. (b) Kill raterelative
to the crest of the nearesthigh tide. Numberof kills
and observationtime (hr) in the successive intervals
were 2/49, 3/55, 6/62, 5/75, 4/87, 12/100, 13/106, 25/
104, 18/90, 3/79, 1/69, 2/58. (c) Correlation between
kill rateandheightof the tide. The regressionequation
is kill rate = 0.068 X tide height (m) - 0.16 (r2 =
0.86, F.5 = 29.8, P < 0.001). Number of kills and
observationtime (hr) for the successivepoints were 0/
25, 1/58, 8/104, 16/224, 26/279, 36/199, 7/50.
No. hunts
No. kills
Success
rate (%)
75
299
278
652
33
33
28
94
44
11
10
14
of 0.69 huntsper hr and 0.10 kills per hr.The kill rate
per observationhour was quite consistentthroughout
the day except for declines in early morning(08:0009:00) and late afternoon(15:00-16:00; Fig. la).
The kill rate showed a strong associationwith the
tidal cycle, rising with the flooding tide and falling
with the ebbing tide (Fig. lb). The kill rate peaked 12 hr after high tide. The only anomalyin this pattern
was a markedlow in kill rate beginning 2 hr before
high tide. The rateduringthis 1-hrintervalwas nearly
one-half of the previousintervaland one-thirdof the
next. The kill rates also showed a strong and significant linear relationwith the height of the tide, rising
steadily to 0.14-0.18 kills per hr at tide heights over
4.2 m (Fig. Ic).
The success rate of hunts differedsubstantiallybetween zones and apparentlyas a function of the type
of attack.In zone 1 all attackswere aimed at Dunlins
roostingor feeding nearthe saltmarsh,and peregrines
always attackedusing stealth, approachingvery low
(<1 m) over the shoreline vegetation or rushing up
over the dyke. Surprisecould be near complete, resultingin the captureof a Dunlinthe momentthe flock
flushed in alarm.In zone 1, peregrineshad a success
rateof 44% (33 kills in 75 hunts;Table 1). If a stealth
huntin this zone was not immediatelysuccessful,peregrines rarelygave chase.
The success rate of all hunts in zone 2 was nearly
the same as in zone 3 (11% vs. 10%).Combined,the
success rate of stealth hunts in these two zones was
14% (21 kills in 154 hunts), significantlylower than
in zone 1 (G = 14.2, P < 0.001). Most (70%) hunts
in zones 2 and 3 did not use stealth, but were open
attackson Dunlinsin flightor of birdsthathad flushed
well aheadof the falcon. These open attackson flying
Dunlins had a success rate of 9% (37 kills in 406
hunts). The success rate of all peregrineattacksover
zones 2 and 3 (11%) was significantlylower than all
huntsover zone 1 (44%;G = 29.4, P < 0.001).
Other raptorsalso attackedDunlins. Harriersfrequently attemptedto approachroosting Dunlins by
stealth. Of an estimated300 harrierattacksonly five
were successful and all of these took place in zone 1.
Probably reflecting the low risk posed by harriers,
Dunlins showed minimal avoidance response. If
flushed by harriers,the Dunlins returnedto the same
place as soon as the raptorhadpassedby. Merlinswere
This content downloaded from 142.58.26.133 on Tue, 19 May 2015 18:31:34 UTC
All use subject to JSTOR Terms and Conditions
418
SHORTCOMMUNICATIONS
The peak in kill rate immediatelyafterhigh tide is
probablydue to the Dunlins' need to compensatefor
energy expendedduring2-4 hr of over-oceanflying.
ReturningDunlinsare likely to be hungry,moreintent
on foragingandless vigilantthanat othertimesduring
the tidalcycle. Consequently,the trade-offtemporarily
shifts towardincreasedrisk.
A second and complementaryreason for the high
kill rate just after high tide is that some peregrines,
particularlythe adults,spendmost of the day perching
and may not starthuntinguntil flocksof Dunlinsbegin
to congregatenearthe saltmarsh.Adult peregrinesare
DISCUSSION
significantlymore successfulin the use of stealththan
The hypothesis that raptor predationrisk for small juveniles. Only 25% of all hunts were by adults, yet
shorebirdsincreases with closeness to vegetation is they accountedfor 47% of kills (Dekker2003).
supportedin this study by three converging lines of PREYSELECTION
evidence:(1) the high success rateof peregrineshunting over the saltmarshzone; (2) the relativelyhigh kill Based on the examinationof prey remains,a relatively
rateper observationhourwhen Dunlinsare in the salt- high percentageof shorebirdscaught by raptorsare
marshzone; and (3) the positive correlationof kill rate known to be juveniles (Kus et al. 1984, Whitfield
with the height of the rising tide. The results of this 1985, Warnock1994). The mechanicsof age-related
study also lend additionalsupportfor the hypothesis prey selectionseem simple if we assumethatjuveniles
that the over-oceanflocking of Dunlins during high are on the outsides of flocks, either on the groundor
wintertides is an antipredatorstrategy(Dekker 1998, in flight (Ydenbergand Prins 1984, Ruiz et al. 1989,
Hotker2000). Flying far from shore, at varying alti- Newton 1998). In this study,at least 69% of captured
tudes dependingon weather conditions, the Dunlins Dunlins were taken by peregrinesdirectly from the
are safe fromthe most dangeroustype of raptorattack: outside or tail end of flocks (Dekker 2003). Furthermore, the percentage of juveniles may be high in
a stealthapproachconcealedbehindvegetation.
The overall percentageof stealth hunts by pere- small,isolatedflocksthatrenderthemselvesvulnerable
to stealthattacks
the high tide. Such risky begrines in this study (35%) is nearly equivalentto the haviorincludes during
(1) late departureon over-oceanflock36% reportedin 233 shorebirdhunts on the coast of
Scotland(Cresswell1996). By contrast,the percentage ing flights;(2) persistencein roostingor foragingalong
of stealthflights in 569 shorebirdhuntsrecordedat a the high-tideline; (3) early returnto shore afteroverocean flocking; and (4) a switch to inland roosts or
large marshylake in Albertawas 77% (Dekker1988).
sites. The proportionof juveniles in flocks in
feeding
The explanationfor these dissimilarvalues is that the
habitatin these two areas was quite different.At the such high-risksituations,and theirphysical condition
to
that flock over the ocean,
lake, reedy shorelinescreatedsuitablecover for stealth compared conspecifics
an
avenue for furtherremight
present
interesting
hunts. Open attackson flying shorebirdswere uncommon at the lake except when droughthad caused the search.
shallows to recede well away from shore (Dekker
This
was
financedby DD. During
1991, 1999). By the same token, the high proportion the last studytravelprivatelywere
year,
expenses
partiallyreimbursed
of open hunts at BoundaryBay (62%) reflectedthe
lack of opportunitiesfor surpriseover zones 2 and 3. by the Centre for Wildlife Ecology at Simon Fraser
Once the Dunlinswere >10 m from the saltmarsh(the University, British Columbia. Accommodation in
2000-2003 was providedby D. Leach. Occasionalcoboundarybetween zones 1 and 2), the distancefrom observerswere I.
D. Hancock,R. Swanston,
shorehad no significantbearingon the huntingsuccess and P. Thomas. G.Dekker,
Court did the G-tests. I. Gordon
rateof peregrines(i.e., kill ratesfor zones 2 and3 were
assistedwith the dataanalysisfor Figureic. R. Dekker
nearlyequivalent).
the graphs.RefereesJ. Buchananand R. ButWhile peregrinesevidently hunt and captureprey prepared
ler made helpful comments on the first draft of this
throughoutthe day, the fact thatthey killed much less paper.
often just before and much more often just after high
tide cannot be relatedto habitat.If that were so, the LITERATURECITED
kill rate per observationhour 1-2 hr before and 1-2
hr afterhigh tide shouldbe the same, which is clearly BRENNAN, L. A., J. B. BUCHANAN, S. G. HERMAN, AND
not the case. The most plausible reason for the obT. M. JOHNSON.
1985. Interhabitat movements of
serveddifferenceis relatedto the behaviorof the DunwinteringDunlinsin westernWashington.Murrelins. Well before the cresting tide, often when the
let 66:11-16.
floodwatersare still >50 m from the saltmarsh,the BUCHANAN,J. 1996. A comparison of behavior and
success rates of Merlins and PeregrineFalcons
great majorityof Dunlins departon their over-ocean
when hunting Dunlins in two coastal habitats.
flights, while othersfly inlandespeciallyduringor afJournalof RaptorResearch30:93-98.
ter heavy rain.The drop in kill rate duringthis period
reflectsthe decreasedvulnerabilityof the birdsduring BUTLER,R. W. 1994. Distribution and abundance of
Western Sandpipers,Dunlins, and Black-bellied
over-oceanflights.(Thekill rateat inlandroostingsites
Plovers in the FraserRiver estuary,p. 18-23. In
outside the study areawas not recorded.)
far less commonthanharriers,but more adepthunters
of Dunlins;7 of 23 stealthattackson flocks roosting
in zone 1 were successful (30%).Merlinsalso hunted
over zones 2 and 3, using stealthas the initialstrategy
in 28 attacks on flocks. After the stealth approach
failed, six captureswere made by persistentpursuitof
single Dunlinsthathad left the flock (21%).
Attacks by Merlins and peregrinesalways caused
Dunlin flocks to move to anotherlocation or to begin
theirover-oceanflights,which generallystarted1-2 hr
before the rising tide inundatedall mudflathabitat.
This content downloaded from 142.58.26.133 on Tue, 19 May 2015 18:31:34 UTC
All use subject to JSTOR Terms and Conditions
SHORTCOMMUNICATIONS 419
R. W. Butler and K. Vermeer[EDS.],The abundance and distributionof estuarinebirds in the
Strait of Georgia, British Columbia. Canadian
Wildlife Service OccasionalPaperNo. 83, Ottawa, ON, Canada.
MILINSKI,M.
1986. Constraintsplacedby predatorson
feeding behaviour,p. 236-252. In T. L. Pitcher
[ED.], The behaviour of teleost fishes. Croom
Helm Ltd., London.
NEWTON, I. 1998. Populationlimitationin birds. Aca1987. The birds
BUTLER,R. W., ANDR. W. CAMPBELL.
demic Press, San Diego, CA.
of the FraserRiverdelta:populations,ecology and PAGE,
1975. Raptorpredation
G., ANDD. E WHITACRE.
internationalsignificance.CanadianWildlife Seron winteringshorebirds.Condor77:73-83.
Occasional
No.
Cavice
65, Ottawa,ON,
Paper
PALMER, R. S. 1988. Handbookof North American
nada.
birds. Vol. 5. Diurnal raptors,part 2. Yale UniW. 1996. Surprise as a winter hunting
CRESSWELL,
versity Press, New Haven,CT.
strategy in Sparrowhawks,Peregrines,and MerANDE A.
RuIz, G. M., P. G. CONNORS, S. E. GRIFFIN,
lins. Ibis 138:684-692.
PITELKA. 1989. Structureof a wintering Dunlin
D. 1980. Huntingsuccess rates,foraginghabDEKKER,
population.Condor91:562-570.
its, and prey selection of PeregrineFalcons migrating throughcentral Alberta.CanadianField- SOKAL,R. R., ANDE J. ROHLF.1981. Biometry. Freeman and Co., New York.
Naturalist94:371-382.
D. 1988. PeregrineFalcon and Merlin pre- THIOLLAY,
J. M. 1982. Les ressourcesalimentairesfacDEKKER,
dation on small shorebirdsand passerinesin Alteur limitantla reproductiond'une populationinberta.CanadianJournalof Zoology 66:925-928.
sulaire de Faucun Pelerins (Falco peregrinus
D. 1991. Prairiewater:wildlife at Beaverhills
DEKKER,
brookei).Alauda50:16-44.
Lake, Alberta. University of Alberta Press, Ed- VALONE, T. J., AND S. L. LIMA. 1987. Carryingfood
Canada.
monton,AB,
items to cover for consumption:the behaviorof
D. 1998. Over-ocean flocking by Dunlins
DEKKER,
ten bird species feeding underthe risk of preda(Calidrisalpina) andthe effect of raptorpredation
tion. Oecologia 71:286-294.
at Boundary Bay, British Columbia. Canadian
WARNOCK, N. D. 1994. Biotic and abiotic factors afField-Naturalist112:694-697.
fecting the distributionand abundanceof a winD.
DEKKER, 1999. Bolt fromthe blue-wild peregrines
tering populationof Dunlin. Ph.D. dissertation,
on the hunt. Hancock House Publishing,Surrey,
Universityof California,Berkeley,CA.
BC, Canada.
D. 2003. PeregrineFalconpredationon Dun- WHITE, C. M., N. J. CLUM,T. J. CADE, AND W. G.
DEKKER,
HUNT.2002. Peregrine Falcon (Falco peregrinus).
lins and ducks and kleptoparasiticinterference
from Bald Eagles winteringin British Columbia.
In A. Poole and E Gill [EDS.],The birdsof North
Journalof RaptorResearch37:91-97.
America, No. 660. The Birds of North America,
FLATER,D. [ONLINE]. 1998. Xtide: harmonictide clock
Inc., Philadelphia,PA.
and tide predictor.<http://www.flaterco.com/xti- WHITFIELD,D. P. 1985. Raptorpredationon wintering
de/> (1 December2003).
wadersin southeastScotland.Ibis 127:544-558.
HOTKER,H. 2000. When do Dunlins spend high tide YDENBERG, R. C., R. W. BUTLER, D. B. LANK, C. G.
in flight?Waterbirds23:482-485.
GUGLIELMO, M. LEMON, AND N. WOLF. 2002.
Kus, B. E., P. ASHMAN,G. W. PAGE, AND L. E. STENTrade-offs, condition dependence and stopover
ZEL. 1984. Age-relatedmortalityin a wintering
site selection by migratingsandpipers.Journalof
populationof Dunlin. Auk 101:69-73.
Avian Biology 33:47-55.
LIMA,S. L. 1988. Vigilance and diet selection:a simR. C., ANDH. H. T. PRINS.1984. Why do
ple example in the Dark-eyed Junco. Canadian YDENBERG,
birds roost communally?,p. 123-239. In P. R.
Journalof Zoology 66:593-596.
Evans, J. D. Goss-Custard,and W. G. Hale [EDS.],
LIMA,S. L., T. L. VALONE,ANDT. CARACO.1985. ForCoastal waders and wildfowl in winter. Camaging efficiency--predationrisk trade-off in the
grey squirrel.Animal Behavior33:155-165.
bridge UniversityPress, Cambridge,UK.
This content downloaded from 142.58.26.133 on Tue, 19 May 2015 18:31:34 UTC
All use subject to JSTOR Terms and Conditions