The Auk 113(3):626-635, 1996
BIRD
MIGRATION
EASTERN
AT DIFFERENT
NORTH
LATITUDES
IN
AMERICA
I. NEWTON • AND L. C. DALE
Instituteof Terrestrial
Ecology,
MonksWood,Abbots
Ripton,Huntingdon,
Cambridgeshire
PE172LS,UnitedKingdom
ABSTRACr.--We
attempt to quantify the relationshipbetween migration and latitude in the
avifaunaof easternNorth America.Progressingnorthwardup the eastsideof the continent,
the proportionof breedingspeciesthat movessouthfor the winter increasessteadily,from
12% at 25øN (southern Florida) to 87% at 80øN (Ellesmere Island), a mean increase of 1.4% for
every degreeof latitude. Conversely,the proportionof wintering speciesthat movesnorth
for the summer decreases with latitude, from 52% at 25øN to none at 70øN, a mean decrease
of 1.1%for every degreeof latitude. Theserelationshipshold despitethe fact that 24%of all
speciesbreedingin easternNorth Americaleavethe areacompletelyin fall to winter south
of 25øN,mostlyin Central and SouthAmerica(including the CaribbeanIslands).Thesetrends
are similar to those in western Europe, but at any given latitude an average of 17% more
breeding speciesleave for the winter in easternNorth America,and 10%more wintering
speciesleave for the summer.This differenceis attributed to climate,in that at any given
latitude temperaturesare cooler in eastern North America than in Europe. We argue that
relationshipsbetweenmigrationand latitude existbecauselatitude is a goodsurrogatemeasure of factorslikely to more directly influence migration, such as climate and daylength,
which in turn control the amplitude of seasonalchangesin food supplies.Received
2 August
1995,accepted
3 November1995.
WE EXAMINE
THEPROPORTIONS
of all bird speciesbreeding at successivelatitudes in eastern
Using mapsof the breedingand wintering rangesof
easternNorth Americanbirdsgiven by Godfrey (1966),
Scott (1987), Root (1988), and Price et al. (1995), we
recorded
the distributionof relevantspecies,summer
We also examine the proportionsof all species
North
America
that move south for the winter.
winteringat successive
latitudesthat movenorth
to breed. The study follows a similar analysis
for the avifauna of western Europe (Newton
and Dale 1996),enabling a comparisonbetween
the two regions.The relationshipbetweenmigration and latitude has been previouslyexamined only in fairly general terms (Lack 1954,
Slud 1976, Alerstam 1991), or for particular
groupsof birds(MacArthur 1959,Willson 1976,
Herrera 1978). These earlier studies do not re-
veal the broad latitudinal picture that we discuss for the entire avifauna, summer and win-
ter. The main relevanceof our paper is in understandingthe role of migration in influencing the distributionpatternsof birds, and the
resulting diversity gradients.
METHODS
We adoptedthe sameprocedureas in our earlier
analysisfor Europeanbirds(Newton and Dale 1996).
and winter, in an approximately1,000-kmwide band
spanning60øof latitude (25øto 85øN)lying along the
easternedge of North America and adjacentcoastal
areas. The band extended
from
Florida
to Ellesmere
Island (Fig. 1).
Within this band, we calculated for each successive
5øof latitude the numbersof bird speciespresent:(1)
year-round, (2) in summer only, and (3) in winter
only, including only speciesthat occurredwithin 1ø
on eithersideof eachlatitudeline. Fromthesefigures,
we examined changesin the latitudinal distribution
of birds in eastern North
America
between
summer
and winter. We alsocalculatedthe proportion of species breeding at each latitude that moved south for
the winter, and the proportionof specieswintering
at each latitude
that moved
north
for the summer.
Some374 regularly occurringspecieswere included
in the analysis(not vagrants),the criterionfor acceptancebeing a rangemap in Scott(1987).The list includeda few introducedspecies,suchasHouse Sparrow (Passer
domesticus)
and EuropeanStarling (Sturnus
vulgaffs),
that have becomeestablishedover wide areas, but excluded those found only in restricted locations.
E-mail:[email protected]
Only speciesthat abandoneda given latitude corn626
July1996]
Migration
inRelation
toLatitude
627
Proportion of breeding
specieswhich migrate
south for the winter
80ø
75 ø
70 ø
65 ø
60 ø
55ø
'::• Winter
sea-ice
45 ø
25 ø
FIG.1. Map of easternNorth Americashowingproportions
of breedingbirdsat differentlatitudesthat
migrate south for the winter.
pletely for the winter or summerwere classedas migrantsat that latitude, whether they movedshortdistanceswithin North America or longer distanceselsewhere. Speciesthat at any given latitude were partial
migrantswere classedas residentat that latitude becausethey could be found there year-round (though
somemay have moved from inland to coastor from
mountain to valley). The breeding latitudes for seabirds were based on locations of nesting colonies,
recognizingthat somepelagicspeciesmight forage
hundreds of kilometers from their colonies during
the breeding season.Eachspecieswas placed in one
of three categoriesaccordingto whether its main win-
RESULTS
Of the 374speciesfound regularlywithin the
area considered,69 (18%) are present throughout their latitudinal rangeyear-round,and the
remaining 305 (82%)are completelymigratory
in part or all of their respectiveranges(Table
1). Eighty-three(22%)speciesbreed in eastern
North America and leave the area completely
for the winter, mostlymoving either to Central
or SouthAmericaand nearbycoastalareas(including the CaribbeanIslands;78 species)or to
ter habitat was land, freshwater, or coastal marine.
western North America (Yellow-billed Loon
This last categoryincluded both shorebirdsand sea-
[Gaviaadamsii],PacificLoon [G. pacifica],Thayer's Gull [Larusthayeri]) or Africa (Northern
birds, and will hereafter be referred to as coastal.
Some speciesin this categoryoccupieda different
habitat in summer,most shorebird speciesmoving to
inland terrestrial habitats and most sea duck species
to freshwater.They were classedas coastalfor the
analysesbecausethe emphasiswas on winter distribution.
Wheatear [Oenantheoenanthe],Common Ringed
Plover [Charadriushiaticula]).Conversely, 22
speciesthatdo not breedin the areaconsidered
move in for the winter, either from farther west
or northwest on the continent (17 species),from
628
NEWTON
ANDDALE
[Auk,Vol. 113
TABLE1. Migratory statusof 374 bird speciesin easternNorth America.
Coastal
Migratory status
species
species
Land species
69
6
5
58
2
5
36
0
48
62
3
95
0
218
Resident a
Breeding onlyb
Winteringc
Migrating within North Americaa
Summeringbut not breeding'
Total number of species
Freshwater
All species
81
22
191
11
374
17
14
60
11
108
aPresentover full latitudinal range year-round,but may include somepartial migration in which part of
population moveswithin year-round range.
bBreeds in eastern North America, but entire population winters elsewhere.
cWinters in easternNorth America, but entire population breedselsewhere.
a Presentin partsof easternNorth Americayear-round,but latitudinaldistributionchangesbetweensummer
and winter.
• Seabirdsthat breed outside area consideredhere but are present off easternNorth America in summer
(see Table 2).
Iceland and northwestern Europe (Great Skua
[Catharactaskua],LesserBlack-backedGull ILar-
vs. 60øN) and the Atlantic Puffin (Fraterculaarctica; 80ø vs. 55øN).
usfuscus],
Black-headed
Gull [L. ridibundus],
Eurasian Wigeon [Anaspenelope]),
or from Greenland (Dovekie [Alle alle]). The populations of
several speciesthat both breed and winter in
The low speciesnumbersin Florida may be
partly attributable to the peninsula effect (in
which speciesnumbersdecline from baseto tip)
andpartlyto the lackof montanehabitats,which
occurover much of the latitudinal range farther
north. In quadraticregressionanalyses,latitude
easternNorth America are augmentedin winter
by birds from other breedingareasin Greenland or western Canada. Finally, 11 seabirds
occur regularly off easternNorth America in
summer but do not breed there; seven species
breed in neighboringor more distantareasin
the Northern Hemisphere (including Manx
accounted
for 59% of the variance
in overall
speciesnumbersin summer(comparedwith 90%
in western Europe), 57% for landbirds alone,
77%for freshwaterbirds, and only 6% for coastal birds (Table 3). The decline in speciesnumShearwater[Puffinus
puffinus],
which alsobreeds bers from 40øN northward is apparent within
in small numbers in eastern North America),
each major vegetation zone, including boreal
coniferous forest and tundra.
and four speciesbreed in the Southern Hemispherein the australsummerand migratenorth
In summer,the largestnumberof coastalspecies also occurs in mid latitudes, at 45 ø to 65øN.
to spend the northern summer (= australwinter) in the North Atlantic (Table 2). The main This is mainly due to the peakat theselatitudes
distributional patterns found among eastern in number of seabirdspecies(rather than shoreNorth American birds are elaboratedin the Ap- birds), which coincideswith a similar peak in
pendix.
zooplankton biomassat >500 mg - m-3 (Food
Latitudinal trendsin speciesnumbers.--The and Agricultural Organization 1982). Presumnumber of breeding speciesin eastern North ably, it is zooplanktonon which seabirdnumAmerica increasesprogressivelywith latitude bersultimately depend.
from 110 at 25øN (in southern Florida) to 210 at
In winter, the latitudinal trend in species
40øN(near ChesapeakeBay),and then declines numbers is more marked than in summer, and
progressively
to 30at 80øN(on EllesmereIsland; approximately linear, as speciesnumbers deFig. 2A). Baffinand Ellesmereislandsextend cline progressivelyfrom 203 at 25øNto 4 at 80øN
northward from latitude 70øN and are so large (Fig. 2B). Of the four speciesthat can be found
and close to other land areas that their isolation
in winter at 80øN,the Rock Ptarmigan (Lagopus
is unlikely to accountfor much of the decline mutus)is the sameas in Europe at that latitude.
in speciesnumbersat high latitudes.Indeed, The other three--Gyrfalcon (Falco rusticolus),
speciesnumberson Baffin Island are no differ- Snowy Owl (Nyctea scandiaca),and Common
ent than those on the nearby mainland. How- Raven(Corvuscorax)--areabsentaltogetherfrom
ever, someseabirdspeciesextend much farther Svalbard, which covers this latitude in western
north in western Greenland
than in eastern
Europe.In contrastto Europe, no seabirdsoccur
Canada,notably the Razorbill (Alca torda;70ø in winter at 80øN in eastern North America,
July1996]
Migration
inRelation
toLatitude
presumablybecauseof the greaterextentof pack
ice in North
America.
In winter, latitude alone accounts for 95% of
the variancein overall speciesnumbers,and for
92 to 95% of the variance
in the numbers
of
speciesof different habitat categories(Table 3).
Unlike the situation in summer, the latitudinal
trend in winter is apparentin every biome,with
speciesnumbers at successive5ø latitudes declining from 203 to 194 within the southern
evergreenzone,from 156to 86 within the eastern deciduous forest, from 58 to 32 within
real conifer forest, and from 19 to 4 within
bothe
tundra zone. The rate of decline in speciesnumberswith latitude is approximatelythe samein
landbirds, freshwater birds, and coastal birds,
although very few land and freshwaterbirds
occur north of 50øN in winter.
Seabirds that
extend in winter north beyond 55øN include
the Razorbill, Black Guillemot (Cepphus
grylle),
Dovekie, and Ivory Gull (Pagophilaeburnea).
Thesespeciesalsowinter farthestnorth among
seabirdsin Europe, making use of gaps in the
pack ice, but it is uncertain to what extent they
movesouthduring the period of completedarkness.
The relative
contributions
made
to the total
avifauna by birds of different habitatschange
markedly with latitude. In particular, coastal
birds increasein representationfrom 20 to 67%
of total breeding speciesin summer over the
latitudinal range 25ø to 80øN, and from 27 to
67%of total speciesnumbersin winter over the
range25øto 65øN,and landbirdsand freshwater
birds decline in proportion accordingly.In the
northern parts of their winter range, where
freshwater is frozen, aquatic speciesare more
or less confined to the coast.Examplesinclude
Ring-neckedDuck (Aythyacollaris),LesserScaup
(A. affinis),and RuddyDuck (Oxyurajamaicensis),
all of which
also occur inland
farther
south.
Comparison of the summer and winter species distributions (Fig. 2) reflects the marked
southwardshift of many speciesbetweensummer and winter. In the south, at 25ø to 35øN,
more speciesare presentin winter than in summer, but from 40ø northward, summer species
increasinglyoutnumberwinter ones.The same
situation holds in western Europe, the switch
again occurringat 35øto 40øN.At 70øNin eastern North America,summerspeciesoutnumber
winter ones by more than six to one, and at
80øNby more than sevento one.
Proportionof migrantsat differentlatitudes.The proportion of breeding speciesthat moves
629
630
NmrroualqDD^I,E
[Auk,Vol.113
T^BI,œ3. Regressionrelationshipsbetween speciesnumber (or percent migrants) and latitude. Presenceof
a quadraticterm meansthat relationship is significantlybetter fitted by a curve than a line, though in all
such casesa linear model also gives a significantfit.a
Independent
variables
Latitude
(x)
Dependentvariable
Latitude2
(x2)
Variance
explained
(100r2)
All birds
-2.63
-12.18
4.33
-3.22
Speciesnumber in summer
Speciesnumber in winter
Percent migrants among summer species
Percentmigrantsamongwinter species
59
O.O8
-0.03
0.03
95
98
87
Landbirds
-2.23
Speciesnumber in summer
Speciesnumber in winter
Percent migrants among summer species
Percent migrants among winter species
Freshwater
57
-6.39
5.42
- 1.75
0.04
-0.05
93
96
91
birds
Speciesnumber in summer
Speciesnumber in winter
Percent migrants among summer species
Percentmigrants among winter species
-0.45
-1.63
3.05
-0.26
77
0.05
-1.23
1.48
-0.55
06
95
86
26
92
90
15
Coastal birds
Speciesnumber in summer
Speciesnumber in winter
Percentmigrants among summer species
Percentmigrantsamongwinter species
Analysesomit SouthernHemisphereseabirdsthat passthe australwinter (• northern summer)in North Americanwaters.
south
for the winter
increases
from
about
12%
at 25øN to 75% at 60øN (treeline) and 87% at
80øN, a mean increase of 1.4% of speciesfor
every degree of latitude (Fig. 3, Table 3). Over
most of the latitudinal range, the relationship
is approximatelylinear, but the rate of increase
reducesbeyond 45øN (Fig. 3). This changeof
slope is attributable entirely to the pattern in
landbirds,among which the proportion of migrants among breeding speciesdoes not increasemuch beyond 45øN (Fig. 3). This is in
and the shallowest
for landbirds.
The converse
of these relationshipsis shown in Figure 3 as
the proportionsof all bird specieswintering at
different
latitudes
that move north for the sum-
mer. This proportion is greatestin the south,
affecting52%of specieswintering at 25øN,and
declines linearly northward, affecting 21% of
speciesat 60øN (mostly seabirds)and none at
70 ø to 80øN, a mean decline of about 1.1% of
breeding area differed significantly between
speciesper degree of latitude to 70øN.Latitude
againaccountsfor a largeproportionof the variance,but not in all groups(on linear regression
analysis,87%overall and 91%amonglandbirds,
decreasingto 26%in coastalbirds and only 15%
among freshwater birds; Table 3). These relationships held despite the fact that 24% of all
speciesbreedingin easternNorth Americaleave
the area completely to winter south of 25øN.
The relationship between the proportion of
speciesmigrating north to breed and the latitude of the wintering areadiffered significantly
birds of different habitats (landbirds vs. fresh-
between birds of land and freshwater habitats,
water birds, F = 25.4, df = 1 and 16, P < 0.001;
landbirds vs. coastal birds, F = 8.1, df = 1 and
20, P < 0.01; freshwater birds vs. coastal birds,
with a steeperregressionslopefor landbirds(F
contrast to freshwater
and coastal birds in which
the proportion of migrants continues to increase to the northern
limits of distribution.
On
a quadratic regression analysis, latitude accounted for 98% of the variance in this proportion overall, and for 86 to 96% among the birds
of different habitats(Table 3). The relationship
between the proportion of speciesmigrating
south
for
the
winter
and
the
latitude
of the
F = 13.6, df = 1 and 16, P < 0.01), with the
steepestregressionslope for freshwater birds
= 13.9, df = 1 and 8, P < 0.01).
Comparison
with westernEurope.--There are
some striking parallels between eastern North
America and western Europe,notably the pro-
July1996]
Migration
inRelation
toLatitude
631
225
A.
Terrestrial
Summer
Freshwater
200
Coastal
175.
150
125.
100
75.
50
25.
25
30
35
40
45
50
55
60
65
2251
B.Winter
70
75
80
Terrestrial
Freshwater
•Coastal
200.
175.
150.
125.
10075.
50.
25.
25
30
35
40
45
50
55
60
65
70
75
80
Latitude øN
FIG. 2.
Number of bird speciesbreeding (A) and wintering (B) at successivelatitudes in eastern North
gmerca.
gressiveincreasein summermigrants, and decreasein winter migrants, with increasinglatitude. On both continents, the proportion of
migrantsamongbreeding speciesincreaseswith
latitude (Fig. 4). The slopesof the regression
lines do not differ significantly between continents but the interceptsdo (F = 27.5, df = 1
and 19, P < 0.001). At any one latitude, an average of about 17%more of the local breeding
speciesleavesfor the winter in North America
than in Europe. This difference reflectsthe climatic difference
between
ern sides of the Atlantic
the eastern and west-
avifaunadecreaseslinearly with latitude (Fig.
4). Again, the slopesof the regressionlines do
not differ significantlybetweencontinents,but
the interceptsdo (F = 9.9, df = I and 16, P <
0.01). Throughout the latitudinal range, the
proportion of wintering speciesthat leavesfor
the summer averages around 10% greater in
easternNorth America than in western Europe.
DISCUSSION
in that, over most of
the latitudinal range,winters at any given latitude are colder in eastern
winter, when on both continents the proportion of migrant speciesin the local wintering
North
America
than
in western Europe. A similar pattern holds in
The trend for species numbers to decrease
with increasinglatitude is well establishedin
many types of organisms(Fischer 1960,Pianka
632
NœWTON
ANDDArE
100-]
A.Allbirds
[Auk, Vol. 113
1007B. Freshwater
birds
'• 80
• 70
• 60
80
70
60•,......',.. / ...-....
:s5ø
I........
- '"-.. ....
30• •"
5o•
'-./,..'
',
30
•• 20•/
•
-'"'...,,.,...-,,'
......
20
•o
o
25 30 35 40 45 50 55 60 65 70 75 80
100-C. Land
birds
9o•
'• 80-
80•
60-
•.•
ß
• 50-.,,•o• 20-/
•
•
10]
25 30 35 4'0 4'5 •0 •5 dO d5 ½0 •5 8'0
1001
D. Coastal
birds
• 90-
•
o
....
,.,/
60•
50• / /
',,,
........ -.
,•
/,--
/' ',
,•,
•
/
,,,'
..."
.........-'-
20
1
025
3'03'54'04'55'05'56'06'57'07'58'0 253035404550556065707580
Latitude øN
Latitude
FIG.3. Proportions
of birdspecies
present
at successive
5ølatitudes
in summer
thatmigratesouthfor
winter(continuous
line), or in winter that migratenorthto breed(dottedline).The two linesarenot mirror
images
of eachotherbecause
some
birdsleaveeastern
NorthAmerica
completely
forsummer
orwinter.
1966,Stevens1989),including birds in North
much emphasison these differences,because
America (Cook 1969, MacArthur 1972, Tramer
1974).In easternNorth Americain summer,the
categorization
wasbasedon winter (vs. sum-
mer) habitatsand was inevitably somewhatarclassed
ascoastalon
numbersof breedingspeciesincreasebetween bitrary.Many shorebirds
25øand40øN,andonly then declinenorthward. their winter habitats would have been more
appropriately
classed
aslandbirdsin summer.
Floridapeninsula,
andif countshadbeenmade It is the overall pattern in the avifaunaas a
whole that is mostmeaningful.The basisfor
in the wider land areas farther west (i.e. northern Mexico),they would have shown greater the latitudinal trend in migrationwasdiscussed
species
numbersin thislatitudinalzoneanda in our earlier paper,and attributedto a northprogressive
declinefrom25øNnorthward(Mac- ward increasein the amplitude of seasonal
Arthur 1972). The decline in eastern North changein foodsupply(NewtonandDale1996),
Americafrom 40ønorthward is not apparentin followingMacArthur(1959),Rabenold(1993),
all bird taxa,andforestpasserines,
in particular, and others.
The increase between 25øand 30øN refers to the
show the reverse trend (Rabenold 1993).
In both western Europe and eastern North
America,roughly one-fourth of the breeding
ward decline in speciesnumbers is apparent species(23 and 24%,respectively)leave the reIn winter in eastern North America, a north-
over almost the whole latitudinal range, par-
gion completelyfor the winter, mostly for
ticularlybeyond35øN.The seasonaldifference regionsto the south--CentralandSouthAmer-
in distributionpatternsreflectsthe withdrawal ica,or Africa, asthe casemay be. Although most
migratewithin theNew
of manyspeciesfrom northernregionsin win- NorthAmericanspecies
ter. Althoughthe precisepatternsdiffer be- World, two species(Northern Wheatearand
tween birds of land, freshwater, and coastal CommonRingedPlover)that breedin the Cahabitats, it would be misleading to place too
nadian Arctic cross the Atlantic to winter in the
July1996]
Migration
in Relation
toLatitude
100.
Old World. The assumptionis that thesespecies
80
70
tering range.This long migrationwould be unlikely to persistunlessit were favored by selection.The sameholds for severalspeciesthat
60.
50.
40.
Alaska from Asia and still return
30-
to wintering areasin the Old World. In contrast,
several speciesthat breed in the Old World,
such as EurasianWigeon, Black-headedGull,
and LesserBlack-backedGull, now winter regularly in smallnumbersin easternNorth America,perhapsrepresentingthe startof a new migrationrouteor colonizationof North America.
One of the puzzling featuresin easternNorth
America is that, among landbirds,the proporß
tion of total migrantsamongbreeding species
does not increasebeyond 50øN. This latitude
roughly correspondswith the boundary be- e)
ß
tween
the eastern deciduous
in most winters.
20.
100
25
100.
This is true of at least
half of all passefinespeciesthat breedbetween
50 ø and 60øN.
o
3'0 3'5 4'0 4'5 5'0 5'5 6• 6'5 7'0 7'5 s'o
North
forsummer
t
50
30
and boreal conif-
erousforests.Many boreal specieslive on tree
seedsand can stay in some numbers at high
latitudes
South for winter
90.
colonized North America from the Old World,
via Greenland, and retained their ancestral win-
have colonized
633
10
Eastern
North
Amerma
tern
'"",
-'
25 30 35 40 45 50 55 60 65 70 75 8'0
Latitude
øN
F•G.4. Proportion of breeding bird species(y) at
different latitudes (x) in eastern North America and
Over mostof the latitudinal range of eastern
western Europethat migrate south for the winter (upNorth America (as in Europe), some species
per) or north for the summer(lower). For southward
migration in easternNorth America, y = -75.05 +
number of different speciesfrom farther north 4.33x - 0.03x2;r2 = 0.98. In western Europe,y = 41.49
movesin. Above35øN,the fall emigrantsexceed - 1.03x + 0.02x•;r• = 0.97. For northward migration
the immigrantssothat,at anyonelatitude,total in easternNorth America,y = 123.72- 3.22x+ 0.03x2;
speciesnumbersare lower in winter than in r• = 0.87. In western Europe,y = 55.65 - 0.66x;r2 =
summer. Below 35øN, the reverse holds and
0.81. In the first three relationships,a quadratic equawinter speciesoutnumbersummerones.Still, tion gave a significantly better fit than a linear one.
move south for the winter, while a smaller
however, more speciesleave eastern North
Americacompletelyfor the winter than for the
summer. Some of those that emigrate totally
spendat leastpart of the winter south of the
equator where the seasonsare reversed (Terborgh 1989). Thus, they gain the benefitsof
summerfood suppliesat both endsof their migration. The same is true for the speciesthat
breed in the SouthernHemisphereduring the
austral summerand then spend the nonbreeding period in the Northern Hemisphere (althoughstill in the tropics).
Consideringthat about24%of easternNorth
American breeding speciesleave the region
completelyfor the winter, many reachingareas
southof the equator,it is strikingthat only four
species(all seabirds)make sucha long reverse
journey,breedingin the SouthernHemisphere
and moving north as far as North America for
the southern winter (= northern summer). The
sameis true in EuropeßThis differencebetween
north and south, both in the Old World and the
New World, may have arisenbecausethe land
areasin the Northern Hemisphereare so much
greater than those in the south. Birds from a
wide longitudinal span in northern North
America move southwardby migration into
progressivelynarrowingland areas,a process
of compression
that could forcethem far to the
south.In contrast,birds migrating north from
southern South America face ever-widening
land areas,so perhaps they can all be accommodatednearer their breeding areas.The reversesituationholdsfor pelagicseabirds,which
have greaterseaareasin the south.Thus, perhapsit is no surprisethat the only four Southern
Hemispherespeciesthat extend in large numbersto North American latitudesduring their
nonbreedingperiod are pelagic.
634
NEWTON
ANt)DALE
Severalpelagicseabirdsthat breedin western
Europecan be seenregularly off easternNorth
America in summer.They are presumablyprebreeders, as confirmed
for Manx Shearwaters
by band recoveriesof second-yearbirds (Perrins et al. 1973). If the equivalent dispersaloccurred in reverse, with North American
breed-
[Auk,VoL 113
FISCHER,
A.G. 1960. Latitudinalvariationin organic
diversity. Evolution 14:64-81.
FOOD AND AGRICULTURE ORGANIZATIOIq.
1982.
Atlas
of the living resourcesof the seas.Foodand Agriculture Organization, Rome.
GOt)TREY,W.E. 1966. The birds of Canada. National
Museum of Canada Bulletin No. 203, Ottawa.
HERRERA,
C.M.
1978. On the breeding distribution
ing birds reachingEuropeanwatersin summer,
pattern of Europeanmigrant birds:MacArthur's
theme re-examined.
Auk 95:496-509.
it could be detectedonly by banding or telemetry, for no pelagic speciesbreed in eastern LACK,D. 1954. The natural regulation of animal
numbers. Clarendon Press, Oxford.
North Americathat do not alsobreed in Europe.
R. H. 1959. On the breeding distriAny birds seen in European waters are thus MACARTHUR,
bution
pattern
of North Americanmigrantbirds.
presumed to derive entirely from European
Auk 76:318-325.
breeding colonies.
R.H. 1972. Geographicalecology.PatThe total number of speciescomprisingthe MACARTHUR,
local avifauna (summer and winter) and the
proportion of migrants (summer and winter)
were correlated with latitude. In fact, in most
terns in the distributionof species.Harper and
Row, New York.
NEWTON, I. 1995. The contribution
of some recent
researchon birds to ecologicalunderstanding.
Journalof Animal Ecology64:675-696.
NEWTO,q,I., ANt) L. DALE. 1996. Relationship between migration and latitude among west Europeanbirds. Journalof Animal Ecology65:137-
analyses,the correlation with latitude was so
strongthat no other factorincluded in the same
regressionmodel could have accountedstatistically for much of the variancein speciesnum146.
bers. This is not to imply that latitude as such
PERRL•S,C. M., M.P. HARRIS,AND C. K. BRITTON. 1973.
hasany direct influence on speciesnumbersor
Survival of Manx ShearwatersPuffinuspuffinus.
on the proportion of migrants. Nevertheless,
Ibis 115:535-548.
latitude clearly is a useful surrogatemeasure PIANKA,E.R. 1966. Latitudinalgradientsin species
that integrates day length, climatic variables,
diversity: A review of concepts.American Nat-
and other variables, all of which influence di-
uralist
100:33-46.
rectly the seasonalchangesin avian food supplies on which migration depends(Lack 1954,
MacArthur 1959,Herrera 1978,Terborgh 1989,
PRIcE,J., S. DROEGE,AND A. PRICE. 1995. The summer
atlas of North American birds. Academic Press,
Newton and Dale 1996). The ultimate cause of
RABENOLD,
K.N. 1993. Latitudinal gradientsin avian
speciesdiversity and the role of long-distance
migration. Current Ornithology 10:247-274.
ROOT,T. 1988. Atlas of wintering North American
birds. University of ChicagoPress,Chicago.
ScoTT,S. L. (Ed.). 1987. Field guide to the birds of
North America,2nd ed. National GeographicSociety, Washington, D.C.
SLuD,P. 1976. Geographicand climaticrelationships
the latitudinal variation in migration almost
certainly is the amplitude of the seasonalvariationin foodavailability,and the extentto which
the abundant food suppliesof summer are reduced by winter severity. In any one region,
the food typesavailablein winter clearlyinfluence which speciesstay and which ones leave
(Newton 1995).
London.
We are grateful to Peter Rotheryfor statisticalad-
of avifaunaswith specialreferenceto comparative distributionin the Neotropics.Smithsonian
Contributionsto Zoology No. 212.
STEVI•S,G.C. 1989. The latitudinal gradient in geographicalrange:How so many speciescoexistin
vice, and to Sidney Gauthreaux,Jr.,TheodoreSimons,
and an anonymousreferee for helpful commentson
the manuscript.
the tropics. American Naturalist. 133:240-256.
TERBORGH,
J. 1989. Where have all the birds gone?
Princeton University Press,Princeton, New Jer-
ACKNOWLEDGMENTS
sey.
ALERSTAM,
T. 1991. Bird migration. Cambridge University Press,Cambridge.
CooK, R. E. 1969. Variation in speciesdensity of
North American birds. SystematicZoology 18:
63-84.
TRAMEll,E.J. 1974. On latitudinal gradientsin avian
speciesdiversity. Condor 76:123-130.
WILLSON,M. F. 1976. The breeding distribution of
migrant birds: A critique of MacArthur (1959).
Wilson
Bulletin
88:582-587.
July 1996]
Migrationin Relationto Latitude
635
APPENDIX.Typesof latitudinal distributionpatternsfound amongeasternNorth Americanbirds.
R. Presentyear-roundthroughoutwhole latitudinal range.Examples:Ruffed Grouse(Bonasa
umbellus),
Limpkin (Aramus
guarauna),
GreatHorned Owl (Bubovirginianus),
CommonRaven(Corvus
corax),Carolina
Wren (Thryothorusludovicianus).
B-R.Presentonly during summerbreedingseasonin the north of range,year-roundin the south.Exampies:TurkeyVulture (Cathartes
aura),Red-tailedHawk (Buteo
jamaicensis),
CommonMoorhen(Gallinula
chloropus),
Killdeer (Charadrius
vociferus),
BlackSkimmer (Rynchops
niger),Common Grackle (Quiscalus
quiscula).
R-W. Presentyear-roundin the north of range,only during winter in the south.Example:EveningGrosbeak (Coccothraustes
vespertinus).
B-R-W.Presentonly during the summerbreedingseasonin the north of range,year-roundat intermediate
latitudes,and during winter only in the south.Examples:
AmericanBlackDuck (Anasrubripes),
Cooper's
Hawk (Accipitercooperii),
Sharp-shinnedHawk (Accipiterstriatus),Cedar Waxwing (Bombycilla
cedrorum),
Dark-eyed Junco(Juncohyemalis),Song Sparrow (Melospizamelodia).
B-W. Summerbreedingrange beginning immediatelyto the north of winter range.Examples:Red-breasted
Merganser (Mergusserrator),House Wren (Troglodytes
aedon),Rusty Blackbird(Euphagus
carolinus).
B-P-W.Summerbreedingrange separatedgeographicallyfrom winter rangeby a gap in which species
occursonly on migration. Examples:Tundra Swan (Cygnuscolumbianus),
Black-belliedPlover (Pluvialis
squatarola),
AmericanTree Sparrow(Spizella
arborea),
LaplandLongspur(Calcarius
lapponicus).
B. Presentin summerbreedingseasononly. Examples:
Swallow-tailedKite (Elanoides
forficatus),
Chimney
Swift (Chaetura
pelagica),
Long-tailedJaeger(Stercorarius
longicaudus),
BarnSwallow (Hirundorustica),Redeyed Vireo (Vireoolivaceus).
W. Presentonly in winter. Examples:GreatSkua(Catharacta
skua),Black-headedGull (Larusrid•bundus),
LesserBlack-backed
Gull (Larusfuscus),Ivory Gull (Pagophila
eburnea),
EurasianWigeon (Anaspenelope).
S. Breedsin the SouthernHemisphereduring the australsummer,and movesto the Northern Hemisphere
during the australwinter (= northernsummer).Examples:
GreatShearwater(Puffinus
gravis),Sooty
Shearwater(Puffinus
griseus),
Wilson'sStorm-Petrel(Oceanites
oceanicus).
P. Occursonly on passagemigration,breedingand wintering elsewhere.Example:HudsonianGodwit (Limosa haemastica).
V. Vagrantonly, occurringmainly or entirely during migrationperiods.Examples:
Curlew Sandpiper(Calidrisferruginea),
Northern Lapwing(Vanellus
vanellus),
Fieldfare(Turduspilaris).
R = resident(found year-round);B = breeding,W = wintering, P = on passage,or migration;S =
SouthernHemispherespecies"overwintering" in the Northern Hemisphereduring the northern summer; V = vagrant.