THE ORIENTATION
OF PASSERINE
NOCTURNAL
MIGRANTS
FOLLOWING
OFFSHORE
DRIFT
KENNETH
P. ABLE
ABSTR•CT.--Nocturnal autumn passerinemigrantsare frequently drifted by wind or carried by
downwind flight from New England to offshore islands. Most individuals in these flights are
immatures.Many recoverthe mainland via reorientednorthward flights. During fall 1972, I made
field observationsand performed orientation cage tests on nocturnal migrants on Block Island,
Rhode Island. Migrants reachedthe island during four southward movementsin following winds.
Migration was very light or absenton most other nights. Reorienteddiurnal flights left from the
island immediately after the arrival of a large nocturnal migration. Northward movementsoccurred on two nightsfollowing daytime reorientedflights. These nocturnalmovementsare tentatively
interpretedas constitutingreorientedflights. About 27% of the 79 individualstestedin orientation
cagesshowedsignificantlydirectionalnocturnal activity. About two-thirds of theseorientedbasically northwestward,correspondingto the reorientedflight of free-flying birds. These and similar
reorientationsof immature birds can be explainedby a simple compassreorientationin responseto
wind drift over the oceanwithout invoking complicatednavigation mechanisms.The remaining
individuals orientedtoward the southeast.Eleven birds exhibited a significanteastward orientation during the first few hoursafter dawn; this is interpretedas a positivephototaxis.Three of four
Blackpoll Warblers orientedsouthward.No correlationexistedbetweenthe quantity of Zugunruhe
and the amount of subcutaneousfat, but fatter birds were significantly more likely to show oriented
nocturnal activity.--Department of Biological Sciences,State University of New York, Albany,
New York 12222. Accepted 3 November 1975.
NOCTURNALpasserinemigrants face the risk of displacementfrom their normal
routes of migration. At any time only a small fraction of the migratory population of
a speciesmay be involved, but natural selectionshould strongly favor individuals
able to recognizeand correctfor theseaccidents.Evidence that migratorypasserines
have this ability comesprimarily from two sources.Homing has been demonstrated
in a few species(Mewaldt and Farner 1957; Perdeck 1958, 1967; Schwartz 1963;
Mewaldt 1964a, 1964b), but these experiments were not designed to clarify the
navigation mechanism.Field observationsby Baird and Nisbet (1960), radar surveillance (Myres 1964, Richardson 1972), and banding recoveries(Evans 1968) indicated
that nocturnal migrants drifted by wind reorient in a manner that compensatesfor
the drift.
Studyingthisphenomenonin thefield is confoundedby practicalproblems.One can
rarely be certain that migrants in flight are performing a compensatorymovement of
sometype or that birds on the ground have been displacedfrom normal migration
routes. As a result, most workers have resorted to artificial displacementsand
examinedthe subsequentcageorientation of the birds (Sauerand Sauer 1959;Hamilton 1962; Potapov 1966;Dolnik and Shumakov 1967; RabOl 1969, 1970, 1972). Birds
are usually transportedunrealisticallygreat distancesto make the predicted shifts in
orientation directions detectable and the results have been ambiguous.
A few localitiesexist where displacedmigrants can be recognizedwith reasonable
confidence and occur in considerablenumbers, and the orientation directions of
compensatingbirds can be predicted. Evans (1968) exploiteda situationof this type
in northeasternBritain where Scandinaviannocturnal migrants occasionallymake a
landfall when drifted by strong southeasterlywinds over the North Sea.
More clear-cut casesof displacementand reorientationoccurregularly in fall along
the New England coast. Radar studies(Drury and Keith 1962, Drury and Nisbet
320
The Auk 94: 320-330. April 1977
April 1977]
Orientationof DriftedMigrants
32 1
1964) have shown that passerinemovementsthrough this regionoccurin two main
directions:(1) Large numbersof birdsmovesouthwestwardon a broadfront moreor
less parallel to the coast;(2) following the passageof cold fronts, large southward
migrationsof passerinesdepart from New England and the Maritime Provincesover
the westernAtlantic Ocean.This southwardmigrationmay be madeup of a number
of species,but only the Blackpoll Warbler (Dendroica striata) has been shown
convincinglyto make a transatlanticflight to northern South America (Nisbet et al.
1963, cf. Murray 1965, Nisbet 1970).
The weather conditionsaccompanyingoffshoreflightsare alsoassociatedwith the
arrival of large numbersof migrantson the coastand offshoreislands.The grounded
migrantsprobablyconsistof individualsfrom the southboundflight as well as birds
•triftedsoutheastward
fromthemovement
paralleling
thecoast.Thenumberof migrants groundedin coastalregionsunder such circumstancesmay be impressively
large (Baird and Nisbet 1960), althoughit probably constitutesonly a small fraction
of the total migrationthat occurredduring the night (Drury and Nisbet 1964;C. J.
Ralph, pers. comm.). Immediately after such flights, large numbers of nocturnal
migrants move northward from coastalareas, especiallyfrom the offshoreislands.
This phenomenonand the evidence that it reflects reorientation following drift are
detailed by Baird and Nisbet (1960). Briefly, the bulk of passerinemigrants appears
to avoid the immediate New England coast and offshore islands. Most individuals
that do occur are immatures, presumably becausethey are inexperiencedand less
able to recognizeor correct for such displacements(Baird and Nisbet 1960, Drury
and Keith 1962, Nisbet et al. 1963, Drury and Nisbet 1964).
Becausethe main directionsof movement of passerinemigrants in southernNew
England are reasonablywell-known and involve severaldistinct directions,it is an
idealplaceto conductexperiments
on migratoryorientation.This paperdescribes
some preliminary experimentsdesignedto examine the orientation of autumn nocturnal migrants on Block Island, a circa 3600-ha island 15.5 km off the coast of
Rhode Island. Pertinent data on the simultaneous behavior of free-flying migrants
are included for comparison.
METHODS
Between 23 Septemberand 6 October 1972, I studied the orientationof nocturnal migrants on Block
Island, RhodeIsland. Observationsof free-flyingnocturnalmigrantswere made with portableceilometers
(Gauthreaux 1969, Able and Gauthreaux 1975) at the Block Island Airport. During the first severalhours
after dawn each day I made countsof birds leaving the north end of the island (Baird and Nisbet 1960).
Birds usedin orientation cagetestswere captured in mist nets on the property of F. David Lapham on
the northeastsideof the island. The birds were usuallycaughtin the afternoon,and eachwas weighed,
measured,agedby skull ossification,and banded. The birds were held in individual cellsin a holdingcage
and were exposedto natural photoperiod,but were not allowed to seethe sky. The individualsusedin
orientation cage tests were not a random sample of birds captured in the course of routine banding
operations.Rather, when possible,I selectedthe fattest individuals of nonresidentspecies.
The orientationcageswere "Emlen funnels"(Emlen and Emlen 1966)placedon the groundon an open
bluff alongthe northeastcoastof Block Island. No terrain featuresor horizonglowsshouldhave been
visible to the birds from this point. The cageswere exposedto environmentalsounds,includingthe ocean
to the east. Birds were placed in the cagesat night after all visible traces of light from the settingsun had
disappeared.I checkedthe cagesimmediatelybeforedawn, placednew funnelsin the cagesof individuals
that showedan apparent directional tendencyduring the night, and returned thesehirds to the orientation
cagesfor 3-4 additional hours. All birds were releasedat the end of the tests.
The blotter paper conesbearingthe footprint recordsof the birds' activity were analyzed as described
by Emlen and Emlen (1966). Each cone was divided into 16 sectors(22.5ø each). From the quantified
footprint densitiesin each sectorI computeda mean vector, length of the meanvector(r), and angular
322
KENNETH P. ABLE
TABLE
[Auk, Vol. 94
1
SUMMARY OF VISUAL OBSERVATIONS OF l¾1IGRATION
Nocturnal migration
Date
Magnitude
•
22 Sep.
23 Sep.
24 Sep.
25 Sep.
26 Sep.
27 Sep.
28 Sep.
3000
800
None
Fog
None
Rain
8400
29 Sep.
Fog
30 Sep.
1 Oct.
7000
9700
2 Oct.
300
3 Oct.
4 Oct.
400
None
5 Oct.
No obs.
Wind blowing
Direction
Reorienteddiurnal
toward
flight
SE
NNW
----SW
SE
NNE
NE
NE
NNE
Variable
SW
No observations
Small
None
None
Small
None
None
--
NE
Very small
SE
S
None
None
SSE
SSE
NE
N
Mixed
---
NE
NW
--
Moderate
Large
Moderate-large
None
t Traffic rate in units of birds per mile of front per hour.
deviation(s)for eachbird as describedby Batschelet(1965).To determinewhetherthe activity of the birds
wasconcentrated
in anycompass
direction,I applieda Rayleightestto eachdistribution
(Batschelet
1965,
1972), usingsamplesizescorrectedas describedby Emlen (1969). Emlen'scorrectionswere derivedonly
from observations
of IndigoBuntings,but theyappearto be sufficientlyconservative
to be appliedto
otherspeciesas well. Other statisticaltestswill be describedas appropriate.
RESULTS
The origin of migrantsonBlocklsland.--Table 1 summarizes
the magnitudesand
directionsof nocturnalmigrationover Block Island during the study. Thesedata
agreein generalwith the radar pictureof migrationin southernNew England(Drury
and Keith 1962, Drury and Nisbet 1964).Moderate or large movementsoccurredon
onlyfour nights.In eachcase,the flightswere movingtoward betweensoutheastand
southwestin followingwinds. On a numberof nightsno detectablemigrationoccur-
red. Two nights,23 September
and 2 October,sawveryslightn•)rthward
movements. Both theseflights were in light southerlywinds on nightsfollowinga large
southward migration. Their directions and the fact that the birds were seenwithin 2
hoursafter darknesssuggestthat they initiated migrationon the island.
Northward reorientedflightsleft the islandon oneor moremorningsfollowingthe
arrival of a large nocturnal migration (Table 1). Departures failed to materialize
when no nocturnalmigration occurredover the island for severaldays. On the first
morning after a large migration, the reorientedflights left into a head wind as
describedby Baird and Nisbet(1960),but on subsequent
daysdeparturesalsooccurred with followingwinds.
Basedon theseobservations
and the speciescompositionof the nettedbirds, it is
virtually certainthat mostof the individualstestedin the orientationcagesarrived on
the island during oneof the four southwardmigrationsand thus originatednorth of
the island.Of coursethe exactdatesof the nettedbirds'arrival or their lengthof stay
on the island could not be determined.
Cageorientation.--I tested79 individualsof 23 speciesin orientationcageson 11
nights. The compilationin Table 2 showsthat only 21 (26.5%) of the birds showed
significantorientationat night. As expectedat this locality, immaturesoverwhelmingly predominatedin the nettedsamplesof almostall species,the BlackpollWarbler
beingthe outstandingexception.
April1977]
Orientation
ofDriftedMigrants
A.
B.
323
C.
Fig. 1. Distribution of mean orientationvectorsof individual birds. A, mean vectorsof all individuals
that showedstatisticallysignificantcageorientationat night (excludingdata from BlackpollWarblers).B,
mean vectorsof nocturnalorientationof four Blackpoll Warblers. C, mean vectorsof orientationduring
the first hours after dawn.
Somespeciesand speciesgroupswere consistentlybetterorientationcagesubjects
than others.Gray Catbirds(Dumetellacarolinensis)were notoriouslypoor, but some
of the individuals may have been from the island'sbreedingpopulation. As a group
the warblers(Parulidae)performedwell, and Blackpoll Warblers had an especially
high proportionof orienters.Vireos(Vireonidae)were poorlysuitedto the orientation
cages.They fluttered a great deal and marred the blotter coneswith slashingwing
marks. Also all vireoscharacteristicallyclungto the hardware cloth topsof the cages
and hung upsidedown for long periods.
Solid overcastprevailedthroughonly one night. None of nine birds testedon this
night was oriented.
The data from all individualsshowingstatisticallysignificantnocturnalorientation (RayleighP •< 0.05) are presentedin Table 3. The mean directionsof nocturnal
TABLE
2
SPECIES AND AGE COMPOSITION OF BIRDS TESTED IN ORIENTATION
Number
CAGES
tested
No.
of
Species
Total no.
Adults
Immatures
orienters
Gray Catbird, Dumetella carolinensis
Swainson'sThrush, Catharus ustulatus
Gray-cheeked Thrush, C. minimus
Veery, C. fuscescens
White-eyed Vireo, Vireo griseus
Solitary Vireo, V. solitarius
Red-eyed Vireo, V. olivaceus
Philadelphia Vireo, V. philadelphicus
Warbling Vireo, V. gilvus
Black and White Warbler, Mniotilta varia
Magnolia Warbler, Dendroica magnolia
Yellow-rumped Warbler, D. coronata
Blackpoll Warbler, D. striata
Ovenbird, Seiurus aurocapillus
Northern Waterthrush, S. noveboracensis
Common Yellowthroat, Geothlypis trichas
Wilson's Warbler, Wilsonia pusilla
American Redstart, Setophagaruticilla
Indigo Bunting, Passerinacyanea
Dark-eyed Junco, Junco hyemalis
White-throated Sparrow,
18
2
2
1
2
1
17
3
1
0
0
0
0
0
0
0
0
0
18
2
2
1
2
1
17
3
1
0
0
0
1
1
0
6
0
1
I
1
2
0
1
0
1
1
0
6
4
1
4
2
2
2
0
0
0
0
0
1
3
0
0
1
0
2
2
1
4
2
2
2
I
3
5
1
I
0
0
0
5
1
1
2
0
0
Zonotrichia
albicollis
White-crowned Sparrow, Z. leucophrys
Lincoln's Sparrow, Melospiza lincolnii
5
1
2
1
1
2
0
1
324
K•NNETH P. ASLE
TABLE
[Auk, Vol. 94
3
PARAMETERS OF NOCTURNAL AND DIURNAL CAGE ORIENTATION FOR INDIVIDUALS THAT SHOWED
STATISTICALLY SIGNIFICANT
Nocturnal
DIRECTIONALITY
orientation
Diurnal
Mean dir.
s
Veery
337.5
66.6
0.001
White-eyed Vireo
330.5
71.1
0.025
Red-eyedVireo
352.9
66.9
0.001
136.9
146.2
312.7
339.9
72.8
73.6
72.8
72.0
0.025
0.050
0.010
0.050
Warbling Vireo
Black and Whit• Warbler
Magnolia Warbler
Blackpoll Warbler
Ovenbird
Northern Waterthrush
Common Yellowthroat
Wilson's Warbler
American Redstart
Dark-eyed Junco
White-throated Sparrow
RayleighP• Mean dir.
0.6
73.8
0.025
194.1
67.2
0.050
3.4
147.7
207.4
65.6
55.2
58.0
0.050
0.001
0.010
168.7
292.9
293.4
71.9
67.4
70.2
75.5
70.8
0.010
0.050
0.050
0.001
272.4
6.6
63.6
69.7
0.050
0.001
155.2
332.9
67.9
71.4
0.010
0.010
130.2
73.6
0.050
orientation
s
RayleighP•
347.1
70.8
0.025
54.4
72.4
0.100
60.7
62.7
0.001
28.9
58.2
112.7
65.6
71.7
67.0
0.010
0.100
0.010
142.3
69.2
0.010
165.9
99.3
36.2
60.6
0.001
0.001
140.3
137.3
69.4
70.3
0.001
0.100
I Rayleighstatisticprobabilitybasedon samplesizecorrectedaccordingto E mlen (1969). SeeMethods.
activity for eachindividualare plottedin Figure 1A. ExcludingBlackpollWarbler
points,the distributionof meanactivityvectorsappearsto be bimodal,with points
concentratedin the northwestand southeastsectors.I testedthe null hypothesisthat
this samplewas drawn from a uniformdistributionusingRao'stest (Batschelet
1972),whichis powerfulwhenappliedto multimodalcases.The distributionof
mean orientation directions of the 17 birds differed significantly from uniform
(U = 169.4,P < 0.05). The five pointsin the southeast
quadrantare closelyclustered around their mean directionof 147.4ø (s = 13.6ø, Z = 4.7, P < 0.01). More
scatterexistedamongthe 12 pointsin the northhalf of the circle,but if treatedasa
separate population, they are significantly oriented about their mean
(mean= 333.0ø,s = 37.3ø, Z = 7.5,P < 0.001).Thereisreasonto hypothesize
that
two populations(in a statisticalsense)exist basedon their nocturnalorientation
directions.Making thisassumption,
applicationof theMardis-Watson-Wheeler
two
sampletest(Batschelet1972)showsthat a significantdifferenceexistsbetweenthese
two groupsof samplepoints (P < 0.001).
About twice as many individualsorientedtoward the northwestas toward the
southeast.
Nonetheless
the groupof birdsorientedsoutheastward
consisted
of five
species
testedonseveralnightsandit seems
unlikelythat theobserved
distribution
is
an artifact.Apparently
twonocturnal
directional
preferences
existamonggrounded
fall migrantson Block Island.
Because
BlackpollWarblersare believedto embarkona southwardtransatlantic
flightto SouthAmerica(Nisbet1970),I analyzedseparately
thesmallamountof data
obtainedfromthisspecies.
Four of sixBlackpollstestedshowedsignificantnocturnal
April 1977]
325
Orientation of Drifted Migrants
TABLE
RELATIONSHIP
OF CAGE ORIENTATION
4
TO AMOUNT
OF SUBCUTANEOUS FAT DEPOSITS
Fat class
0
Number of nonorienters
Number of orienters
Total individuals
Proportion orienting
1
2
3
4
10
2
12
15
5
20
16
6
22
10
6
16
3
2
5
0.165
0.250
0.273
0.375
0.400
orientation (Fig. lB). Three of these birds (two adults, one immature) oriented
southward, while the fourth (an adult) showeda weak northward orientation. Although no conclusionscan be drawn from such a small sample, most Blackpoll
Warblers testedshowedthe predictedorientationdirection, which differed from that
of most of the other individuals
tested.
Becausea large portion of the northward exodusof nocturnal migrants from Block
Island occursduring the early morning hours, I examinedthe behavior of orienting
nocturnal migrants during the first hours of daylight. Of the 21 individuals that
showedsignificantnocturnalorientation,6 (32%) alsoexhibiteda directionalpreference during the early morning. Five additional birds that failed to yield statistically
significantnocturnaldirectionalitydid orient in daylight. The data on diurnal orientation are presentedin Table 3, and the meansare plotted in Figure 1C. With only
one exception, the resultant vector of orientation for each of the 11 birds was in the
eastward half of the circle. The distribution is significantly different from uniform
(P < 0.025, Rayleigh test)and no differenceexistsbetweenthe directionschosenby
individuals that showedsignificantorientation during the night and thosethat did
not. A comparison of the nocturnal and diurnal orientation shows that most of the
birds that showedorientedhoppingboth day and night shiftedtheir orientationfrom
northwestward
to eastward or southeastward at dawn.
An eastward direction doesnot correspondwith any observeddiurnal migratory
movement on Block Island, nor is it intuitively reasonable for numbers of birds to
move eastward from New England. I think the most reasonableexplanation for these
resultsin orientation cagesis that the activity of the birds was not the product of
migratory motivation and that the observeddirection was the result of a positive
phototacticresponseto the sun. The resultscan be adequatelyexplainedas reflecting
simple escapebehavior.
Zugunruhe, orientation, and subcutaneousfat.--I tried to selectthe fattest available birds for orientation cage tests, but the limited numbers of birds available on
somedays forcedme to useleaner individuals. I examinedthe relationshipbetween
the amount of migratory fat and the birds' cagebehavior. I expected,a priori, that a
positive relationship existed between the amount of subcutaneousfat and the inten-
sity of migratory motivation of an individual bird. Thus it surprisedme to find no
correlationbetweenthe amountof nocturnalactivity(basedon thefootprintrecords,
quantified accordingto Emlen and Emlen 1966) and the visual fat classesof the migrants (Helms 1959). For the birds that showedsignificant orientation, r = -0.129;
for all birds combined, r = -0.088. Nor did the amount of nocturnal activity differ
between orienters and nonorienters within a given fat class.
A significant relationship did exist between the amount of fat and the likelihood
that a given individual would showorientedZugunruhe during the night (Table 4).
326
K•ETa
P. ABLE
[Auk, Vol. 94
The proportionof individualsthat showedorientedZugunruheincreasedsignifi-
cantly
withincreases
insubcutaneous
fatdeposits
(arcsinx/proportion
oforienters
-- 24.8 + 3.8 (fat class), r -- 0.979, F = 68.3, P < 0.001). The proportion of
orientersincreasedrather sharply betweenfat classes2 and 3, suggestinga threshold
of responsein this relationship.
DISCUSSION
The absenceof a correlationbetweenZugunruhe and the amount of depot fat in
the birds I tested is contrary to prevailing theories on the relationship between
energeticsand migratory motivation (e.g. Dolnik and Blyumental 1967, Helms 1968,
Evans 1968). Becausemy data were obtainedfrom a number of speciescaughtunder
special circumstances,other factors may have obscuredthe expectedrelationship.
Somecautionmay be in order when quantitatively equatingZugunruhe with migratory motivation, even at the heightof the migratory season.OrientedZugunruhe has
not generally been used, but may provide a much better indicator.
My field observationsare in complete agreement with what is known of fall
migration in southernNew England (Baird and Nisbet 1960, Drury and Keith 1962,
Drury and Nisbet 1964). Large migrationsmoved in southerlydirections(between
south-southeastand southwest)in generally following winds. During periodsof opposingwinds very little nocturnal migration occurred. Nearly all the migrants netted
on Block Island thus originatedon the mainland to the north. Although thosebirds
arriving from the northwestmay have experiencedsomesoutheastwarddrift, all the
birds showeda basicallysouthward orientation during the flight and brought them to
the island.
After arriving from the mainland, migrants left by day from the northern end of
the island, as described by Baird and Nisbet (1960). On two occasionsnocturnal
northward migrations occurred, both times on the night following a i'najor southward flight and the arrival of numbersof migrants. The important questionarises
whether these reversed nocturnal flights representedthe same type of reoriented
movement seenduring the day. I cannot be certain that they do. In both caseswinds
had shifted back to southerly during the day preceding the reversed flights, so the
birds were not leaving for the mainland into a head wind as is often the case in the
early morning flights. I believe it reasonable to hypothesize that these were
reorientedflights on the following bases:(1) The flights occurredimmediately following a large southward migration; (2) they occurredin very light and variable winds,
rather than in a strong southerly air flow when broad front reversed migrations are
most likely to occur (Drury and Keith 1962, Drury and Nisbet 1964, Nisbet and
Drury 1968); (3) the temporal pattern of the movementsindicated that the birds
initiated migration on Block Island; (4) Baird and Nisbet (1960) describeda northward nocturnal movement in opposingwinds following a large southward flight on
the previous night.
The majority of birds in orientation cagesheaded northwesterlyat night, probably
in a cage analog of the reorientednorthward flights performedby free-flying birds.
The birds are capable of selectingthe reorientation direction on the basis of cues
available in an orientation cage. This doesnot mean that the processof reorientation
occurred in the cagesor was based only on cues available to the birds while in the
orientation cages. The integration of information leading to the reorientation may
have occurred prior to the birds' capture, either during the flight that carried them
April 1977]
Orientation
ofDriftedMigrants
327
offshoreor on the islanditself.The resultspresentedheresupportthe conclusion
that'
this informationcan at leastbe transferredto a cuesystemavailableto the birdsin
funnel cages.The failure of birdsto orientunder overcastskiessuggests
that stellar
cues were of primary importance.
Someindividualsorientedtoward the south. This is the expecteddirectionfor
Blackpoll Warblers basedon Nisbet's (1970) hypothesis.The severalother species
that showedsouthwardnocturnalorientationare not believedto makelongoverwater flights from New England. These birds may representa portion of the migrant
populationthat is unableto correctfor displacement,althoughotherexplanationsare
possible.
Becausemostof the northwarddepartureof migrantsapparentlyoccursduringthe
day, I predictedthat birds showingnocturnalorientationmight continueto show
orientedhoppingafter dawn. The observedorientationafter sunriseis not readily
explicablein a migratorycontext.It is mostsuggestiveof a phototacticresponse.In
part this may be an artifact of the type of cageused.The funnel cagesallow the birds
to seethe sun continuouslyonceit is high enoughin the sky. Wiltschkoand HiSck
(1972)reportedthat the directionsselectedby EuropeanRobins(Erithacusrubecula)
during early morninghourscorresponded
to the migratorydirection,but only when
the sameindividualsshowedorientedrestlessness
during the previousnight. Their
cagesprecludeddirect sightof the sun and birds in thesecagesdo not orient consistently toward either the sun or moon even when these are directly visible to them
(Wiltschko, pers. comm.).
Few other empirical data exist with which my resultscan be compareddirectly.
Evans (1968) obtained orientation cage and band recovery data from Scandinavian
migrantspresumablydrifted to the coastof northeasternBritain. The majority of
birds (most were juveniles) orientedbetweensoutheastand south, eastwardfrom
their presumednormal migratoryheadingprior to drift. If the normalmigratory
headingwastowardthe south-southwest
(cf. Nisbet 1969,Rab011969),theyseemto
have reorientedin a directionthat wouldintersecttheir normalmigratoryroutenorth
of the winteringarea. The directionschosenby Evans' birds werevery similarto the
main reorientation direction noted on radars around the British Isles (data from a
variety of placessummarizedby Myres 1964).
Based on several series of artificial displacement experiments with sylviid
warblers, Rab01(1969, 1970, 1972) proposeda migratory navigationsystembased
on a hypothetical"goal area" that movesin spacealong the migratoryroute of the
speciesin a time-programmedmanner.Basedon thishypothesis,the goalarea(sensu
Rab01) of a bird drifted offshorefrom New England could conceivablylie to the
northwestor north. Indeed Rab01(1969) interpretedthe observationsof Baird and
Nisbet (1960) in supportof his hypothesis.
Three explanationsare possiblefor the observations
describedin this paper, by
Evans(1968)and by Rab01(1969, 1970, 1972):(1) The birdsmay recognizethe georaphical displacementand employ navigation(Type III orientation)to recoverthe
normal migration route or reachthe migratorygoal, as Evans proposed;(2) the birds
may navigate toward an intermediategoal area as postulatedby Rab01;and (3) the
birds may make an automaticresponsethat will reorient them toward land should
they find themselvesover the ocean.
The classicstudiesof Perdeck(1958, 1967)indicatedthat juvenile birds making
their first autumnmigrationemployedonly compassdirectionand distancecomponents. In contrast to adults, they seemedunable to compensatefor longitudinal
328
KENNETH P. ABLE
[Auk, Vol. 94
displacements.Similar resultshave been obtained with nonpasserinespecies(Bellrose1958),but both Evans(1968)and RabOl(1969, 1970, 1972)interpreteddata from
immature birds as supporting navigation at least back toward the normal route of
migration.
Data from both free-flying migrants and birds testedin orientation cagesshow that
someimmature migrants reaching Block Island are capable of reorienting in a proper
direction to correct for a departure from normal routes of migration. If a navigational
responseis involved, immature birds must be able to perform it in flight over the sea
well out of sight of land (Baird and Nisbet 1960, Richardson 1972, Ferren pers.
comm.). I believe all the currently available data on thesemigrants can be explained
by a simple compass-basedreorientation. It seemspremature to attribute navigational ability to immature fall migrants, either in North America or in Europe.
A nocturnal migrant from an easternNorth American breedingpopulationfinding
itself over the ocean at dawn could most easily recover the coast by flying northwestward. Southwestwardor westwardflight would often be lessreliable or involve
greater distancebecauseof the orientation of the coastline, especiallysouth of New
England. Substantialevidencesummarizedby Baird and Nisbet (1960) indicatesthat
the migratory goalsof the birds displacedto Block Island lie to the southwest.Birds
could depart in that direction and reach Long Island, New York via an overwater
flight of ca. 24.5 kin, only ca. 9 km further than the distanceto the mainland to the
north. Both are visible from Block Island on clear days and a southwestward flight
would often be accompanied by at least partially following winds, but neither freeflying birds nor orientation cage subjectsoriented in that direction.
As offshore drift occursfrequently one would expect natural selection to have
established a more or less automatic northwestward reorientation response. The
occurrence of this kind of reorientation could be accomplished by simple compass
orientation (see also Emlen 1975). It is probably not coincidental that a similar
responseoccursamong unrelated specieson the eastern side of the Atlantic and that
the reorientation directions observed there are again such as to return the drifted
migrants to a coast over the shortestpossibledistance(Myres 1964). Vagrant migrantspassover the Farallon Islands, California, in a northeasterlydirectionfollowing morning reorientation over the sea southwest of the island. Immature Blackpoll
Warblers captured on the island also tended to orient in that direction in funnel
orientationcages(DeSante 1973).That a simplereorientationmechanismof this type
may be of widespread occurrenceis suggestedby observationsof migratory bats
(probably Lasiurus borealis) and Monarch Butterflies (Danaus 1•lexi1•1•us)
flying
northwest into the wind 90 km off the Rhode Island coast in autumn (Ferren pers.
comm.).
The preponderance of immatures among grounded nocturnal migrants near the
Atlantic and Pacific coastshas beenwell documented(Baird and Nisbet 1960, Nisbet
et al. 1963, Murray 1966, Ralph 1971). Ralph (1971) summarizedvarious hypotheses
to accountfor this phenomenonand concludedthat geneticallyinducednavigational
errors were the most likely explanation for it. As adult birds are rare near the coast,
most of the iramatures flying offshore must either perish or learn something that
enablesthem to avoid the coastal belt during subsequentautumn migrations.
The distributionsof adult and immature migrants clearly indicate somedifference
in the migration of thesetwo age classes.Severalcausesare possible.Although most
of the birds reaching the islands offshorefrom New England have flown on tracks
from the north or northwest, they may be substantiallydrifted (Baird and Nisbet
April 1977]
Orientationof Drifted Migrants
32 9
1960, Able 1974). Thus the heading directionsof adults and iramatures could be the
same, but iramaturesmight be lessable to recognizeor to correctfor lateral drift in
flight. Alternatively, adults might be more selective of the weather conditionsfor
starting migration and thereby avoid winds most likely to drift them offshore. No
data now exist to distinguishbetweenthe severalalternatives. The fact that migrants
fly offshore does not necessarilyimply that they have faulty basic orientation
mechanisms.Evans' (1968) data and the resultsdescribedhere imply that at least
someof theseinexperiencedbirds recognizeand correctfor offshoredrift.
ACKNOWLEDGMENTS
I am greatlyindebtedto Mr. and Mrs. F. David Laphamfor their gracioushospitalityand for allowing
me freeuseof their property.Mrs. Lapham'sefficientbandingoperationprovidedthe birdsusedin these
experiments.J. Baird, S. T. Emlen, S. A. Gauthreaux, B. G. Murray, and C. J. Ralph read and
commentedon a preliminary draft of the manuscript.I. C. T. Nisbet kindly made a copyof the DeSante
thesisavailable to me. This work was supportedby the ResearchFoundation of the State University of
New York.
LITERATURE
CITED
ABLE, K.P.
1974. Wind, track, heading and the flight orientationof migrating songbirds.Pp. 331357 {n Proc. Conf. Biol. Aspectsof the Bird-Aircraft CollisionProblem(S. A. Gauthreaux, Ed.).
, ANDS. A. OAUTHREAUX.1975. Quantificationof nocturnalpassefinemigrationwith a portable ceilometer. Condor 77: 92-95.
BAIRD,J., ANDI. C. T. NISBET. 1960. Northwardfall migrationon theAtlanticcoastandits relation
to offshore drift. Auk 77: 119-149.
BATSCHELET,E. 1965. Statisticalmethodsfor the analysisof problemsin animal orientationand
certainbiologicalrhythms.Washington,D.C., Amer. Inst. Biol. Sci.
1972. Recent statisticalmethodsfor orientation data. Pp. 61-91 in Animal orientationand
navigation(S. R. Galler, K. Schmidt-Koenig,
G. J. Jacobs,and R. E. Belleville,Eds.). Washington,
D.C., N.A.S.A. SP-262.
BELLROSE,
F.C.
DESANTE,D.F.
1958. The orientationof displacedwaterfowlin migration.WilsonBull. 70: 20--40.
1973. An analysisof the fall occurrences
and nocturnalorientationsof vagrantwood
warblers (Parulidae) in California. Unpubl. Ph.D. thesis, Palo Alto, California, Stanford Univ.
DOLNIK,V. R. ANDT. I. BLYUMENTAL. 1967. Autumnalpremigratoryand migratoryperiodsin the
Chaffinch(Fringilla coelebs)and someothertemperate-zone
passefinebirds. Condor69: 435-468.
-ANDM. E. SHUMAKOV. 1967. [Testingthe navigationalabilitiesof birds]. Bionica 1967:500507.
DRURY,W. H. ANDJ. H. KEITH.
1962. Radar studiesof songbirdmigrationin coastalNew England.
Ibis 104: 449-489.
--
ANDI. C. T. NISBET. 1964. Radar studiesof orientationof songbirdmigrantsin southeastern
New England.Bird-Banding35: 69-119.
EMLEN,S.T. 1969. The development
of migratoryorientationin youngIndigoBuntings.Living Bird
1969: 113-126.
1975. Migration:Orientationand navigation.Pp. 129-219in Arian biology,vol. 5 (D. S.
--
Farner and J. R. King, Eds.). New York, Academic Press.
ANDJ. T. EMLEN. 1966. A techniquefor recordingmigratoryorientationof captivebirds. Auk
83: 361-367.
EVANS,P.R.
1968. Reorientationof passefinenight migrantsafter displacementby the wind. Brit.
Birds 61: 281-303.
GAUTHREAUX,
S.A. 1969. A portableceilometertechniquefor studyinglow-levelnocturnalmigration. Bird-Banding40: 309-320.
HAMILTON,W. J., III. 1962. Bobolinkmigratorypathwaysand their experimentalanalysisunder
night skies. Auk 79: 208-233.
HELMS,C.W.
Wilson
Bull.
1959. SongandTree Sparrowweightand fat beforeand aftera nightof migration.
71: 244-253.
330
KENNETH P. ASLE
[Auk, Vol. 94
1968. Food, fat and feathers. Amer. Zool. 8: 151-167.
MEWALDT, L.R.
1964a. California "crowned"sparrowsreturn from Maryland. WesternBird-Bander
39:
--
1-2.
1964b. California sparrowsreturn from displacementto Maryland. Science146: 941-942.
AND D. S. FARNER. 1957. Translocated Golden-crownedSparrows return to winter range.
Condor
59: 268-269.
MURRAY,B.G.
1965. On the autumn migrationof the Blackpoll Warbler. WilsonBull. 77: 122-133.
1966. Migrationof ageand sexclasses
of passerines
on the Atlanticcoastin autumn.Auk 83:
352-360.
MYRES,M.T.
1964. Dawn ascentand re-orientationof Scandinavianthrushes(Turdus spp.) migrating at night over the northeasternAtlantic Ocean in autumn. Ibis 106: 7-51.
SISBET, I. C.T.
1969. Review of Evans, 1968. Bird-Banding 40: 55.
1970. Autumn migration of the Blackpoll Warbler: Evidence for long flight provided by
regional survey. Bird-Banding 41: 207-240.
-AND W. H. DRURY. 1968. Short-term effectsof weather on bird migration: a field study using
multivariate
statistics. Anim. Behav.
16: 496-530.
, W. H. DRURY,AND J. BAIRD. 1963. Weight-lossduring migration. Bird-Banding 34: 107138.
PERDECK,A. C. 1958. Two types of orientation in migrating Starlings, Sturnus vulgaris L., and
Chaffinches,Fringilla coelebsL., as revealedby displacementexperiments.Ardea 46: 1-37.
1967. Orientation of Starlingsafter displacementto Spain. Ardea 55: 194-202.
POTA?OV,
R.L.
1966. [On the navigationalability of somepasserine
birds].Dokl. Akad. Nauk SSSR
171: 226-228.
R•HOL, J. 1969. Orientationof autumnmigratingGarden Warblers(Sylvia borin) after displacement
from westernDenmark (Bl&vand)to easternSweden(Ottenby). A preliminaryexperiment.Dansk.
Ornithol.
Foren.
Tids.
63: 93-104.
1970. Displacement and phaseshift experimentswith night-migrating passerines.Ornis
Scandinavica
1: 27-43.
1972. Displacementexperimentswith night-migratingpasserines(1970). Z. Tierpsychol.30:
14-25.
RALPH,C.J. 1971. An age differentialof migrantsin coastalCalifornia. Condor 73: 243-246.
RICHARDSON,
W.J.
1972. Autumn migrationand weather in easternCanada:A radar study.Amer.
Birds 26: 10-17.
SAUER,E.G. F. ANDE. SAUER. 1959. NiichtlicheZugorientierung
europaischer
Vbgelin Sudwestafrika. Vogelwarte 20: 4-31.
SCHWARTZ,
P. 1963. Orientationexperimentswith Northern Waterthrusheswintering in Venezuela.
Proc. 13th Intern. Ornithol. Congr.: 481-484.
WILTSCHKO,W. AND H. HOCK. 1972. Orientationbehaviorof night-migratingbirds(EuropeanRobins) during late afternoonand early morning hours. Wilson Bull. 84: 149-163.