Russell Greenberg ? Vladimir Pravosudov
John Sterling ? Anna Kozlenko ? Vitally Kontorschikov
Divergence in foraging behavior of foliage-gleaning birds
of Canadian and Russian boreal forests
Received: 11 May 1998 /Accepted: 24 June 1999
Abstract We compared foraging behavior of foliage-
gleaning birds of the boreal forest of two Palaearctic
(central Siberia and European Russia) and two Nearctic
(Mackenzie and Ontario, Canada) sites. Using discrimi-
nant function analysis on paired sites we were able to
distinguish foliage-gleaning species from the Nearctic
and Palaearctic with few misclassifications. The two
variables that most consistently distinguished species of
the two avifaunas were the percentage use of conifer
foliage and the percentage use of all foliage. Nearctic
foliage-gleaner assemblages had more species that for-
aged predominantly from coniferous foliage and dis-
played a greater tendency to forage from foliage, both
coniferous and broad-leafed, rather than twigs, branches,
or other substrates. The greater specialization on foliage
and, in particular, conifer foliage by New World canopy
foliage insectivores is consistent with previously pro-
posed hypotheses regarding the role of Pleistocene veg-
etation history on ecological generalization of Eurasian
species. Boreal forest, composed primarily of spruce and
pine, was widespread in eastern North America, whereas
pockets of forest were scattered in Eurasia (mostly the
mountains of southern Europe and Asia). This may have
a?ected the populations of birds directly or indirectly
through reduction in the diversity and abundance
of defoliating outbreak insects. Loss of habitat and
resources may have selected against ecological special-
ization on these habitats and resources.
Key words Ecological convergence ? Pleistocene
glaciations ? Warblers ? Historical ? Outbreak insects
Introduction
Of all the major biomes, boreal forest (or taiga) ?
comprising 11% of the world?s forest in a 10?23? cir-
cumpolar belt (Bonan and Shugart 1989) ? arguably has
the most similar vegetation between the continents. Yet,
the Palaearctic and Nearctic boreal forests support bird
assemblages with largely independent evolutionary his-
tories. The disparity between the similarity of vegetation
and the lack of evolutionary relationship of birds pro-
vides an opportunity to test whether ecologically similar
faunas develop in response to similar habitats. Similar-
ities in patterns in the face of the di?erent evolutionary
histories can result from di?erential filtering of species
colonizing the habitat or actual convergent evolution
(Schluter 1986), but in either case argue for the impor-
tance of similar extant ecological conditions in shaping
bird assemblages.
Avian ecologists have used such intercontinental
studies to search for evidence of trophic convergence,
based primarily upon morphological similarity (Karr
and James 1975; Ricklefs and Travis 1980; Blondel et al.
1984; Niemi 1985; Wiens 1991). Fewer studies have
compared foraging behavior or diet based on data
gathered in the field (Cody 1974; Remsen 1985; Ter-
borgh and Robinson 1985; Holmes and Recher 1986;
Wiens 1991). With respect to boreal forest assemblages,
Haila and Ja
?
rvinen (1990) provided an analysis of spe-
cies richness, faunal similarities and ecological structure
of boreal forest avifaunas. However, we know of no
studies that have directly compared the foraging ecology
of avian species of New World and Old World sites.
The strongest evidence for ecological convergence is
the presence of species-to-species matching between the
Oecologia (1999) 120:451?462 ? Springer-Verlag 1999
R. Greenberg (&) ? J. Sterling
Smithsonian Migratory Center, National Zoological Park,
Washington, DC 20008, USA
e-mail: antbird@erols.com
V. Pravosudov
Department of Zoology,
Ohio State University,
Columbus, OH 43210, USA
A. Kozlenko
4075 Monticello Blvd., Apt. 305,
Cleveland Heights, OH 44121, USA
V. Kontorschikov
State Darwin 57, Vasilova,
Moscow 117292, Russia
comparable assemblages (Cody 1974; Schluter 1986;
Wiens 1991; Ricklefs and Miles 1994). Although com-
parable assemblages often have examples of ecologically
similar species, overall pairing of species has rarely been
demonstrated. Given this finding, an examination of the
overall patterns of dissimilarity, as opposed to conver-
gence, might be more instructive (Wiens 1989). For ex-
ample, a basic question is whether the foraging
characteristics of the species assemblages show an
overall pattern of divergence. If systematic divergence is
detected, then the species-matching hypothesis cannot
be supported and the search must shift to the underlying
explanations for the overall divergence. Therefore, it is
the overall divergence hypothesis we address for boreal
forest bird communities in this paper.
We compare patterns of foraging ecology of birds of
Old World and New World boreal forest, because for-
aging ecology is relatively easily quantified and is gen-
erally assumed to play a central role in habitat use as
well (MacArthur 1958). To provide further focus, we
restricted our study to canopy foliage gleaners ? the
most species rich foraging guild of temperate zone for-
ests. We analyze these data to address the following
question: do foliage-gleaning species in Eurasian and
North American boreal forests have emergent di?eren-
ces in their constituent foraging strategies?
Materials and methods
Study sites
Overview
Boreal forest shows similarity in both floristics and structure. The
dominant trees are of the same genera worldwide (such as Pinus,
Picea, Abies, Larix, Betula, and Populus) and the understory plant
species often have circumpolar distribution. The basic structure of
the forest is also simple and similar (Erskine 1977). Most boreal
forests have a single-layered canopy with a subcanopy of saplings,
shrub, and ground layers of herbs or a spongy bed of mosses and
lichens. Boreal forests also have similar landscape patterns. Rather
than vast tracts of unbroken coniferous forest, boreal forests are
generally a complex mosaic of conifer and broad-leafed elements,
sensitive to variation in soil moisture and disturbance regimes
(Bonan and Shugart 1989). Coniferous or mixed forest dominates
undisturbed upland sites. Broad-leafed vegetation grows on flood
plains of rivers (alder, willow, and poplar) or in upland areas dis-
turbed naturally by fire and anthropogenically by logging and fire.
In contrast to the plants that comprise the boreal flora, the birds
that breed in these woods are often unrelated between the continents
(Haila and Ja
?
rvinen 1990; Helle and Mo
?
nkko
?
nen 1990). Although
certain groups have circumpolar species (wrens) or congeners
(Parus, Regulus, Sitta, and Carpodacus), many taxa, particularly
those composed of tropical migrants have no taxonomic counter-
parts, and boreal forest species are part of largely independent ra-
diations of insectivorous birds. These include wood warblers,
tanagers and tyrant flycatcher in the New World and sylvioid war-
blers and muscicapine thrushes and flycatchers in the Old World.
Regions
We conducted fieldwork in central Siberia in 1993?1995, north-
western Canada in 1994, European Russia in 1995, and in eastern
Canada 1996. We selected four study sites in two comparable pairs.
The first comparative pair comprises ??middle taiga?? sites (Ha
?
met-
Ahti 1981) with highly continental climates: the southern Mack-
enzie District of the Northwest Territories (Mackenzie) and the
middle section of the Yenissey River between the Tunguska and
Angara Rivers (Siberia). Both areas were centered on a major river
system (Yenissey and Mackenzie are the fourth- and seventh-
longest rivers in the world, respectively) and had a similar range of
latitudes (Siberia 60?10??62?15? N and Mackenzie 60?14??
61?52? N). We conducted research at three localities each in
Northwest Territories and central Siberia. The Siberian locations
(Myrnoe, Vorogovo, and Fomka) were a maximum of 295 km
apart and the maximum distance was 340 km for the Mackenzie
locations (Fort Liard, Fort Simpson, and Fort Providence). We
established the second pair of study areas in less continental regions
characterized as ??south taiga.?? Both the European Russian and
eastern Canadian sites were far more disturbed and developed than
the central Siberian and Mackenzie areas. Due to logistical con-
straints, the Russian field work was conducted primarily within
40 km of the Kostromskaya Taiga Field Station located in the clay
belt along the Unja River (58?14?, 55?25? N). The Ontario localities
were primarily in the clay belt, as well, between Gogama, Timmins,
Cochran and Hearst, as well as White River and Marathon with a
maximum intersite distance of 490 km and a latitudinal range of
47?51??49?30? N. The lower latitudes of locations in Ontario reflect
the overall southern distribution of the boreal forest in this region
(Ha
?
met-Ahti 1981).
Habitats sampled
We sampled four types of mature forest habitat in Siberia: riparian
alder (Alnaster fruticosus) and willow (Salix spp.); mixed floodplain
forest, with spruce-fir (Picea obovata-Abies sibirica) and birch
(Betula spp.); upland birch stands with some aspen (Populus tre-
mula); and mixed upland taiga with Siberian pine (Pinus sibirica),
spruce, larch (Larix sibirica), fir, and birch. We surveyed five
habitat types in Mackenzie: black spruce forest (Picea mariana);
mixed floodplain forest, with white spruce (Picea glauca) and bal-
sam poplar (Populus balsamifera); riparian, with balsam poplar,
alder (Alnus tenuifolia) and willow; trembling aspen stands (Populus
tremuloides); black spruce (P. mariana); and jack pine (Pinus
banksiana) forests. In the European Russian site we sampled mixed
upland second-growth taiga, (approximately 60 years old) with
spruce (Picea abiea), aspen, and linden (Tillia cordata); and birch
(Betula pubescens); birch-aspen forest; European pine (Pinus syl-
vestris) forest; and riparian elm-alder (Ulmus-Alnus incana) woods.
In Ontario we sampled aspen stands [aspen, birch (B. papyrifera)
and balsam poplar]; mixed upland taiga [white spruce, balsam fir
(Abies balsamea), jack pine, birch, aspen, and poplar]; black spruce;
and jack pine stands. Characteristics of these habitats, based on
estimates recorded at point count survey circles (Greenberg et al.
1999), are summarized in Table 1.
Methods
Target species
Target species included all foliage-gleaning birds above the herb
and low-shrub level: species that feed predominantly on arthropods
of foliage and twigs above the herb and low shrub level. However,
we include in the analysis only species occurring with an abundance
of >0.02 individuals per point on fixed-diameter (50 m) point
counts from the region (Greenberg et al. 1999). Because guild
classification is based on foraging location and substrate, we have
included birds that are as taxonomically diverse and morphologi-
cally distinct as finches, jays, warblers, tits, and flycatchers. We
included some species that are often not considered in analyses of
foliage-gleaning species, but based on our foraging observations
feed substantially on arthropods of foliage and fine twigs. The non-
traditional species include some that also forage on the ground
452
(Eurasian robin, dark-eyed junco, chipping sparrow, and yellow-
breasted bunting; see Appendix 1 for Latin names). Our observa-
tions indicate that these birds foraged predominantly above the
ground during the period and in the habitats of our study. How-
ever, with ground-foraging birds it is often di?cult to determine
how much of their foraging time they spend on the ground, since
there they are more di?cult to observe. We will probably have to
use more sophisticated technology to resolve this problem. We also
include some cardueline finches (siskins and bullfinches), that are
well known to feed on seeds extensively, even when rearing young
(Newton 1972). In the forest habitats we studied, siskins and
bullfinches forage commonly for foliage arthropods during the
height of the breeding season. We considered flycatchers and other
sit-and-wait predators that use long attack distances provided that
a majority of the foraging attacks were directed towards prey that
were resting or flushed from foliage or fine twigs (in this case,
species for example in the genera Contopus and Muscicapa were
excluded because our data indicated that most prey were flying
arthropods).
Foraging observations
We made foraging observations 1 June?20 July, the peak breeding
season in all study areas. We justify the restriction to a narrow
season by noting that the nesting season is usually a period of
relative stability in resource use and that a number of species show
restricted forest habitat use. In addition, the nesting season is a
period of relatively stable foraging behavior that is followed by a
post-breeding period often characterized by a high degree of for-
aging plasticity.
To make regional comparisons, foraging observation were rel-
atively broad brushed and the sampling consisted of gathering a
small amount of data from a number of individuals spread out over
di?erent habitats. We walked di?erent transects through the focal
habitats (See Table 1) each day to ensure we observed di?erent
individuals. Because we attempted to maximize the sample size of
all target species, we covered the di?erent habitats more or less
equally. In addition, every e?ort was made to gather foraging data
on every individual located ? audibly or visually ? to minimize the
bias against birds that forage high in the trees or in dense foliage.
During the breeding season there is often a bias towards gathering
data on singing males. Males and females often forage at di?erent
heights and may even use di?erent techniques or substrates (Morse
1980). Searching for birds by sight and by ear should minimize this.
In fact, the overall mean percentage of singing males was 26%
(2.9% SE) for North America and 36% (3.3% SE) for Europe with
an overall range from 0 to 70% (these data were not consistently
recorded for Siberia). We do not address this variation in our
analysis but present the percentage of singing males for each species
in Appendix 1.
We quantified foraging attack location and maneuvers, prob-
ably the most easily quantifiable aspects of foraging behavior, while
ignoring interspecific variation in locomotary patterns associated
with searching for prey. Clearly, the work of Robinson and Holmes
(1982) and others shows that, although considering locomotion
provides additional insight into interspecific di?erences in foraging
behavior, locomotary patterns are well correlated with attack be-
havior and distance.
We recorded general location data for each individual for the
first foraging maneuver. These data included an estimation of the
height of the individual and the tree or shrub in which it was
located. At the beginning of each season, all observers practised
estimating foraging heights by calibrating estimates to those
determined by range-finders, or to other known heights. We
recorded the position of the bird in relation to the stem or trunk of
the plant, classifying each sighting into inner, middle, and outer
zone which are assigned the number 1?3 for the analysis. We es-
timated the foliage cover for a 1-m cube around each bird and
assigned one of four classes: 1 = 0?25%, 2 = 26?50%, 3 = 51?
75%, 4 = 76?100%.
We used a modified version of the foraging classification system
of Remsen and Robinson (1990). We preserved the most important
aspect of this system which is the complete separation of foraging
maneuver type and foraging substrate ? the surface on which the
prey item was located at the time a foraging attack is initiated (even
Table 1 Descriptive statistics for the habitats searched for foraging birds in this study. Based on vegetation estimates from 75?250 50-m-
radius point count circles per habitat (Greenberg et al. 1999). Cover and height refer to canopy cover and height
Habitat Height
(m)
Percent
cover
Percent
conifer
Percent
spruce
a
Percent
pine
Percent
poplar
Percent
birch
Percent
alder
Percent
willow
Percent
shrub
Percent
herbs
Percent
moss
Europe
Mixed upland 23 74 45 34 11 20 31 0 0 26 26 24
Birch 22 66 3 3 0 36 60 0 0 35 20 4
Pine 18 65 91 9 82 2 3 0 0 22 10 60
Riparian 19 56 15 10 14 3 4 25 45 30 20 0
Mean 20.5 65.3 38.5 13.5 26.8 15.3 24.5 6.3 11.3 28.3 19 22
Siberia
Mixed upland 25 63 62 31 23 7 27 0 0 20 12 29
Mixed flood 23 67 53 48 4 9 19 11 0 9 27 10
Birch 22 78 10 8 1 23 64 1 0 0 16 6
Riparian 11 63 13 3 0 6 7 50 22 8 30 0
Mean 20.3 67.8 34.5 22.5 7.0 11.3 29.3 15.5 4.4 9.2 21.3 11.3
Ontario
Mixed upland 26 61 57 45 13 27 15 0 0 30 28 48
Aspen 31 64 9 8 0 86 3 0 0 44 28 9
Black spruce 17 45 94 92 2 4 2 0 0 36 14 87
Jack pine 21 56 99 2 96 1 0 0 0 52 15 80
Mean 23.8 56.5 64.8 36.8 27.8 29.5 4.0 0 0 43.0 21.3 56.0
Mackenzie
Mixed 24 56 66 66 0 21 10 1 0 42 26 53
Aspen 24 70 10 7 2 80 8 0 0 70 20 2
Black spruce 12 43 95 88 0 0 3 0 3 47 10 88
Riparian 20 73 13 11 0 50 0 21 13 48 6 1
Mean 20.0 60.5 46.0 43.0 0.4 37.8 5.3 5.5 4.0 51.8 15.5 36.0
a
Spruce includes both spruce and fir
453
if a subsequently flushed prey is captured in the air). In the field, we
recorded a number of di?erent types of prey substrates (e.g., foliage
type, foliage surface, twig, branch, air, curled leaf, dead curled leaf,
infloresence, air), some of which were pooled for analysis.
Because foraging location and tree type generally do not change
within short foraging bouts, the previously mentioned location
variables were recorded once per individual. However, because
foraging maneuver type changes frequently during short bouts, we
recorded foraging attack behavior for up to five maneuvers per
bird. We recorded the type of maneuver the bird used to capture
prey in a manner consistent with Remsen and Robinson (1990). In
addition, we made a further separation of foraging maneuvers into
primary or secondary. A primary maneuver type is based on the
body position and action of wings used to attempt to capture prey
(e.g., down strike, upward strike, lunge, leap, stand, hang up, hang
down). A secondary maneuver is added to the initial attack. For
example, a strike maneuver is often followed by hovering in place
or hanging on to the substrate or a flush pursuit, and standing is
often accompanied by the bird stretching its legs and neck in a
??reach??. In all, we recorded 10,415 primary foraging maneuvers
from 6227 individuals ? 3401 maneuvers from 2774 individuals of
27 New World species and 7014 maneuvers from 3453 individuals
of 31 Old World species.
Analysis
The analyses are based on mean values for foraging variables for
each species. Based on the field observations, the following forag-
ing variables were calculated for each species: percentage of for-
aging maneuvers directed towards foliage versus non-foliage
substrates, the percentage of foliage-foraging maneuvers directed
towards conifer needles versus leaves, the mean foraging zone
(based on integer classes 1?3 corresponding to inner, middle, and
outer), percentage of foraging maneuvers that were aerial, mean
attack distance (cm) of all aerial maneuvers, mean foraging height
(m), mean relative foraging height (bird height divided by tree
height), percentage of foraging maneuvers with hovering as a sec-
ondary maneuver, percentage of foraging attacks with hanging as a
primary or secondary maneuver, the percentage of foraging ma-
neuvers with reaching as a secondary maneuver. We transformed
variables for which the samples showed significant deviation from a
normal distribution (Kolmogorov-Smirnof test; Statsoft 1995). We
used arcsine, logarithmic, or square root transformation depending
upon which gave the smallest d-value for the Kolmogorov-Smirnof
test for departure from normality of the sample distribution. Arc-
sine transformation was used for percentage conifer use, foliage
use, relative height, and aerial attack frequency; log transformation
was used for attack distance, hang and hover frequency; square
root transformation was used for reach frequency.
Because species-typical foraging patterns arise from the com-
bination of a variety of individual characteristics, such as foraging
attack method and foraging site (Holmes et al. 1979), we relied
largely upon multivariate analysis. Specifically, we employ forward
stepwise discriminant function analysis (DFA, Statsoft 1995; F to
include = 1.00) to distinguish foraging characteristics of Old
World and New World species using a combination of a minimal
number of variables for each of the paired habitat comparisons
(i.e., middle taiga sites and south taiga sites). We examine the ef-
ficacy of the classification functions by examining the a posteriori
classification table. In addition, we examine the phylogenetic pat-
tern by plotting the group classifications on a phylogenetic tree
derived from Sibley and Ahlquist (1990) using MacClade software
(Maddison and Maddison 1992).
Results
Overall di?erences in breeding foraging variables
The DFA of foraging data from middle taiga sites
(Appendix 1) produced a significant function (Wilks?
lamda = 0.18, F
8,23
= 13.3, P < 0.0001) with the
following variables included in the analysis: use of fo-
liage (0.29 factor loading), use of coniferous foliage
(0.24), cover (0.007), hover frequency ()0.06), attack
frequency (0.05), reach frequency (0.18), hang fre-
quency (0.006), and foraging height (0.06). Of these
only percent conifer (0.24 factor loading), percent use
of foliage (0.29), and percent of maneuvers using
??reach?? (0.18) had loadings greater than 0.1 and can be
considered important in contributing to the discrimi-
nation. The DFA of south taiga sites also produced a
significant function (Wilks? lamda = 0.51, F
4,39
=
9.42, P < 0.0001) with the following variables: foliage
frequency ()0.59) conifer frequency ()0.38), cover
(0.15), and attack frequency (0.02) with only the first
three variables having loadings with absolute values
greater than 0.1. The a posteriori classification for the
middle taiga sites was correct for 97% of the species,
with only the coal tit incorrectly classified as a New
World species. The classification for the south taiga
sites was correct for 85% of the species: black-capped
chickadee, philadelphia vireo, chestnut-sided warbler,
and American redstart were misclassified as Old World
species and the icterine warbler, wood warbler, and
goldcrest were misclassified as New World species.
The Palaearctic and Nearctic assemblages di?ered
consistently between both pairs of sites in percentage
use of conifer versus broad-leafed vegetation and
percentage use of foliage (Fig. 1). The average Cana-
dian species foraged from conifer needles for about
50% of the maneuvers directed toward foliage (47.4%
Ontario, 52.3% Mackenzie), whereas this value was
26% for the Russian species (28.4% Siberia, 24.7%
Kostroma). The variation between sites was significant
Fig. 1 Percentage of foraging maneuvers directed towards foliage
(needles or leaves) versus percentage of foliage maneuvers directed
towards conifer needles. Each point represents a species or
population
454
110 
90 
n, i 70 
m Ol 
CVJ ? 50 
a 
Z     K - 
-10 
!??      . i n i     ,  ..."  
0   )                          I       ? 
i 1 i #  
? 
? 
]             t             f     ? 
;o 
V 
o 
0 
? 
? 
1*                              HP        *m   ii 
O:                               .                             i   * 
io       o 
*    m?       w        m 
20 40 80 100 
0     Palearctic 
?    Nearctic 
Percent foliage foraging 
(F
3,72
= 4.12, P = 0.009) and the pairwise compari-
sons between Siberia and Mackenzie and Kostroma
and Ontario were significant using planned compari-
sons (F
1,72
= 9.2, P = 0.006 and F
1,72
= 5.9, P =
0.017, respectively). Ten of the 17 Mackenzie species
(59%) foraged from conifer foliage for more than 50%
of the observations, whereas only 3 of the 17 Siberian
species (17%) did so. Similarly, 9 of the 21 Ontario
species (43%) foraged from conifer foliage a majority
of the time versus 4 of 23 (17%) for the European site.
Canadian species foraged from foliage (as opposed to
twigs, branches, and other substrates) on average 79%
(75.5% Ontario, 81.4% Mackenzie) of the time versus
61% for Russian species (59.8% Kostroma, 62.1%
Siberia). The overall variation between sites was sig-
nificant (F
3,72
= 8.3, P = 0.0005) as was the di?er-
ence between paired sites (F
1,72
= 13.03. P = 0.005,
and F
1,72
= 13.1, P = 0.005 for southern and middle
taiga sites, respectively).
Selective use of conifer foliage
Because use of conifer foliage is one of the consistent
variables distinguishing the species assemblages in sites
on the two continents, we conducted one additional
analysis. We examined the degree to which foliage
gleaners selected either broad-leaved or coniferous veg-
etation, as they could be selecting this foliage rather than
responding passively to di?erences in relative availabil-
ity. We calculated the di?erence between percentage
conifer-foliage foraging versus the mean estimated per-
centage conifer foliage within 25 m of the focal bird
(these data were not gathered for the Siberian samples).
Plotting this percentage deviation against the percentage
conifer-foliage foraging (Fig. 2) revealed a strong rela-
tionship between the two variables (r
2
= 0.69). This
indicates that species that have a high proportional use
of conifer foliage also use this foliage out of proportion
to its estimated relative abundance (and the same applies
to broad-leafed foliage).
Phylogenetic pattern
We have plotted the classification New World type
and Old World foraging type onto phylogenetic trees
(Fig. 3) based on the MacClade analytical program
(Maddison and Maddison 1992) using the phyloge-
netic inference of Sibley and Ahlquist (1990). We use
the classification of the DFA even where this resulted
in a misclassification of the actual continent of origin.
Inspection of the tree reveals that foraging type is not
restricted to a single clade in each continent: the New
World type occurs predominantly in the distally lo-
cated Emberizinae and the Old World type occurs in
the Muscicapidae, Sylviodae and basal groups of the
Fringillidae. However, the New World type dominates
in the basal Corvidae, Vireonidae, and Tyrannidae
clades as well.
Discussion
Di?erences between Russian and Canadian
boreal forest foliage gleaners
The data show overall divergence in foraging behavior
of foliage-gleaning birds of the Palaearctic and Nearctic
boreal forests. The strongest and most consistent dis-
criminating variables were degree of specialization on
foliage and the relative use of conifer foliage. The
Palaearctic assemblage comprised more generalists spe-
cies that commonly use substrates other than foliage,
and within foliage insectivory, more species of Nearctic
foliage gleaners concentrated on coniferous versus
broad-leafed foliage.
Systematic di?erences in the foraging behavior of
foliage-gleaning birds between the continents could
result from historical and contemporary di?erences in
boreal forest habitats between the continents. Alterna-
tively, because the Palaearctic and Nearctic assemblages
are dominated by species of di?erent major phylogenetic
groups, there may be explanations associated with un-
known phylogenetic constraints. We reject an overriding
phylogenetic-constraints argument for two reasons. The
first is based on natural history considerations. We have
examined foraging behavior in this study because the
general lability of behaviors minimizes the fixed e?ect
of design and phylogeny on morphological features
(Terborgh and Robinson 1985). The characteristics
most consistently distinguishing the two avifaunas ? use
of foliage and use of conifer foliage ? are particularly
Fig. 2 The mean di?erence between the percentage of foliage
foraging maneuvers directed towards conifer needles and the
percentage conifer foliage within 25 m of focal bird versus the
estimated percentage conifer canopy cover (conifer cover/coni-
fer+broad-leafed cover). Each point represents a species from
Europe, Ontario, or MacKenzie, sites where cover data were
recorded. The horizontal line indicates no selection
455
30 50 70 
Percent conifer foraging 
? Europe 
?    Ontario 
? MacKenzie 
110 
flexible. The second is based on heuristic analysis of the
phylogenetic distribution of New World and Old World
foraging types. Although the New World foraging types,
as determined by the DFA, are concentrated in the
distally located Emberizinae clade, they involve species
of groups (Corvidae, Tyrannidae) that are basal to the
Passeriform phylogenetic tree. It is also noteworthy that
two members of the Old world Sylviid warblers were
classified in the New World foraging type, and three
members of the New World vireos and wood warblers
were classified in the Old World foraging type (Fig. 3).
Because simple phylogenetic divergence cannot explain
the di?erences we have documented, we will focus
the rest of discussion on possible historical ecological
factors that could explain the overall pattern of inter-
continental di?erences.
Di?erences in present-day habitat
When specific plots are not matched between regions, as
in this study, comparisons are confounded by di?erences
in availability of particular habitat components. In
particular, di?erences in availability of conifer foliage
might help explain the di?erence in the number of spe-
cies that forage predominantly in conifer foliage. It
should be noted that the arguments for di?erence in
availability do not easily explain the other feature that
distinguishes Palaearctic and Nearctic foliage-gleaning
birds documented here. For example, it would be di?-
cult to argue that a systematic di?erence in the structure
of the vegetation between the continents explains the
greater number of species that are specialized on foliage
versus twig and branch gleaning at the Nearctic sites.
However, the potential e?ect of the di?erential
abundance of conifers in our Palaearctic and Nearctic
sites is clearer. The vegetation data show that conifers,
particularly the ??dark conifer?? spruce-fir element, were
more prevalent in most habitats in the New World than
Old World sites (Table 1). The standardized average
proportion of conifer crown cover was 58% across our
Nearctic versus 37% in the Palaearctic habitats. This
di?erence is even greater (40 vs 22%) for spruce and firs.
Therefore, the greater number of conifer-arthropod
foraging species could simply reflect di?erences in rela-
tive tree abundance at the study sites rather than either
an active selection of one foliage type over another or a
general continental di?erence in preference.
Fig. 3 Distribution of the two foraging classes distinguished by a
discriminant function analysis (DFA) contrasting species from
comparable sites in Canada and Russia: Mackenzie versus Siberia
(A)and Ontario versus Kostroma (B). Foraging type classification is
indicated by shading. Palaearctic taxa (actual distribution) are
indicated in bold. The MacClade program displays the inferred
ancestral character based on consistency of the character state in the
taxa of a particular clade. Where the character state is not consistent it
is displayed as equivocal. Taxa containing species that were
misclassified to continent in the DFA are indicated with an asterisk
b
456
Foraging 
H^l New World Type 
I       I Old World Type 
RSS3 Both Types 
rTTTTH equivocal 
\MEmpidonax 
IB Vireo* 
P Ficedula 
fc\Erithacus 
f^Acrocephalus 
Hippolais* 
BPhylloscopus* 
|[]Sy/wa 
Regulus (Old World)* 
MRegulus (New World) 
Aegithelos 
Parus (Old World) 
EParus (New World)* 
fflPyrrhula 
IQ Spinus 
10 Carpodacus 
foFringilla 
Junco 
Spizella 
f^Setophaga* 
\?Dendroica* 
Wilsonia 
Vermivora 
\MEmpidonax 
Perisoreus 
0 Ficedula 
? Phoenicurus 
^Phylloscopus 
flAcrocephalus 
0 Sylvia 
Regulus 
OAegithelos 
^Parus (Old World)* 
UParus (New World) 
QPyrrhula 
UFringilla 
U Junco 
Spizella 
|Q Emberiza 
Piranga 
USetophaga 
MDendroica 
Vermivora 
Although the overall greater relative abundance of
conifers at the specific New World sites used in this
study may provide a partial explanation, the relation-
ship between percent conifer foliage use and proportion
of canopy cover (Fig. 2) suggests that a substantial de-
gree of specialization results from true preference and
active selection. Furthermore, consideration of what has
been published of the natural history of boreal forest
birds further suggests that the specialization on conif-
erous foliage by more Nearctic species is not restricted
to the sites of this study and their particular habitat
configuration. The close association between a number
of New World warblers and dark coniferous vegetation
is both well known and well studied (MacArthur 1958;
Rabenold 1978; Morse 1980; Sabo and Holmes 1983)
and to a large degree independent of the particular site
of study. Parrish (1995) showed that the preference of
populations of black-throated green warblers for co-
niferous or broad-leafed vegetation was largely intrinsic
and unrelated to relative abundance of foliage types. To
our knowledge, few Palaearctic foliage-gleaning birds
have been reported to have a strong tendency to selec-
tively forage in coniferous trees. For a broad overview
of how ornithologists perceive the use of conifer versus
broad-leafed foliage in the species considered in this
study, we examined the breeding habitat classifications
provided by Ehrlich et al. (1988, 1994). These accounts
are based largely upon information summarized by the
life history series of A.C. Bent and Cramp et al. (see
Ehrlich et al. 1988, 1994 for full citations). For three
species of Siberian birds (dusky, pallas?, and yellow-
browed warblers), we referred directly to Cramp and
Brooks (1992). Based on these accounts, we find that
species for which conifer-dominated habitats are listed
as the first or primary habitat correspond closely to
those species showing preference for conifer foliage in
this study (Appendix 1) and that 58% of the Nearctic
versus 16% of the Palaearctic species show primary
association with coniferous vegetation. Note that these
figures refer to habitat and not foraging preference;
however, there appears to be a high correlation between
these two (for example, Fig. 2).
The greater relative abundance of conifers, particu-
larly spruce and fir, at our New World sites is probably
not idiosyncratic, but represents a real continent-level
di?erence. One factor that contributes to the greater
abundance of the dark conifer element in the Nearctic
communities is di?erences in the natural re-establish-
ment of dark conifer species after disturbance. Both
natural fire or logging (without reseeding and vegetation
management ? important in some regions of the boreal
forest) favor the re-establishment of broad leafed stands
(birch or aspen) in the boreal forest (Kozlenko 1987;
Welsh 1987). Since conifer stands can take 100?
500 years to become re-established after the initial dis-
turbance, an increase in logging will shift the natural
balance between coniferous and broad-leaved stand to
primarily broad-leafed stands. Re-establishment of dark
coniferous forests appears to take longer in the Palae-
arctic than the Nearctic boreal zone (Walter 1979;
Kozlenko 1987; Scott 1995).
Large-scale conversion of the landscape from dark
conifer to deciduous-pine has already occurred, for ex-
ample, in the boreal zone of European Russia where
areas of the once dominant spruce forest are restricted to
small remnants (K. Preobzazhenskaya, personal com-
munication) due to human activities primarily within the
past 50?80 years. The lack of many species of foliage-
gleaning species strongly associated with conifers in
Palaearctic forests compared to the Nearctic forests
suggests that the e?ects of the conversion of boreal
forest on bird diversity may be more pronounced in the
New World.
Divergent histories of boreal forests
Nevertheless, the degree of divergence found between
the Palaearctic and Nearctic avifaunas appears to be too
great to be easily explained by existing ecological con-
ditions. First, the degree of specialization on foliage
cannot be explained by di?erences in habitat availabili-
ty. In addition, the strikingly low species richness of
conifer-specialized foliage gleaners is di?cult to under-
stand given the abundance of coniferous vegetation in
Palaearctic boreal communities. Therefore, we now look
to historical di?erences to explain the divergence in
foraging behavior.
The Pleistocene history of vegetation was quite dif-
ferent between Eurasia and North America (Frenzel
1968; Frenzel et al. 1992; Webb et al. 1993; Adams and
Faure 1995) and this is believed to explain the greater
degree of habitat specialization and regional diversity in
the eastern Nearctic compared to the western Palaearctic
avifaunas (Mo
?
nkko
?
nen 1994; Blondel and Mourer-
Chauvire 1998). In addition, major di?erences in the
distribution of forest during the glacial maxima, par-
ticularly boreal coniferous forest, conform to the major
distinctions we report here between the foliage-gleaning
birds of the two continents. The greater diversity of
canopy species, particularly those associated with dark
coniferous tree species, in the Nearctic can be related to
the large aerial extension of spruce-pine forest-steppe in
eastern North America versus the small pockets of
birch-pine forest which may have persisted near the
Italian Alps, the Carpathian, and the Caucasus Moun-
tains during the glacial maxima (Adams and Faure
1995). The distribution of boreal forest in Siberia is more
poorly known, but at best, a small belt of boreal vege-
tation may have persisted in the extreme south, as well
as the Far East.
Because of their great impact on forest vegetation,
glaciations in Eurasia may have eliminated bird species
most specialized on the forest canopy, particularly the
canopy of dark coniferous (spruce and fir) vegetation,
leaving more ecologically plastic species, an argument
made by Mo
?
nkko
?
nen (1994) more generally for habitat
selection. Even in the absence of such di?erential
457
extinction, Pleistocene vegetation changes have created a
selective environment that favors ecological plasticity in
species. The hypothesis that loss and fragmentation of
boreal forest during the Pleistocene has contributed to
the lack of specialization of foliage-gleaning birds needs
to be investigated for the remaining boreal forest re-
gions, including the south Asian mountains and the
Russian far east. With respect to southern Asia, it ap-
pears that there were no significant boreal forest refugia
(Adams and Faure 1995), and it is therefore consistent
with the hypothesis that Price (1991) found only two of
nine species of sylviids in the mountains of Kashmir to
show a preference for coniferous vegetation.
As an alternative to long-term climate-based changes
in vegetation, the greater degree of more recent human
impact on the forests of the western Palaearctic than the
Nearctic has been posited as a possible reason for the
greater apparent degree of ecological generalization in
European forest birds (Scmiegelow et al. 1997). However,
it should be noted that we found qualitatively similar
di?erences between the birds of the relatively pristine
central Siberian and northwesternCanadian study sites as
we did with the more disturbed European and eastern
Canadian sites. Furthermore, the northern Russian bo-
real forest, although highly altered, is not nearly so
fragmented as the forests of western Europe. Finally, al-
though more distant in time, the degree of habitat loss
imposed upon the boreal forest by the advance of the
glaciers must have far exceeded that caused by more re-
cent human activities. Mo
?
nkko
?
nen and Welsh (1994)
argued that the present-day European avifauna is little
a?ected by fragmentation because it consists of species
that survived the vegetation changes of the Pleistocene.
Biogeography of the arthropod prey base
The pulse of food (arthropod) abundance has been re-
lated to the species richness of migratory foliage-glean-
ing birds in Nearctic boreal forests (Rabenold 1978).
However, the high abundance of migratory canopy-fo-
liage-gleaning birds in the European boreal site
(Greenberg et al. 1999) does not support the idea that
lower food abundance explains the lower richness of
canopy foliage gleaners in the Palaearctic sites. How-
ever, it is possible that a relatively low abundance of
arthropods in Old World conifers, if documented, might
account for the relatively low number of conifer-forag-
ing species. We know of no data to examine this possi-
bility. The work of Bryant et al. (1989), mostly on
deciduous boreal trees, at least suggests that general
di?erences in anti-herbivore chemical defenses can occur
within species or genera at the broad geographic level in
the boreal forest, and this regional variation could un-
derlie foraging di?erences in birds foraging in habitats
with taxonomically similar floras. Helle and Mo
?
nkko
?
nen
(1990) hypothesized that the diversity of the prey base
(arthropods) could have been more greatly reduced
during Pleistocene glaciations in Eurasia because of their
great impact on forest vegetation and floras (Grubb
1987; Latham and Ricklefs 1993; but see Currie and
Paquin 1987). However, as far as we know there is no
direct support for this. Niemela
?
et al. (1994) found
??remarkable similarity?? in species richness of carabid
beetles in mature forests of Finland and eastern Canada,
and Coope (1987) suggested that Pleistocene events
caused few global extinctions in arthropod assemblages.
Even if an impact on arthropod diversity were estab-
lished, its relationship to abundance is unclear.
Perhaps more important than a greater loss of ar-
thropod species in Palaearctic forests would be the loss
of keystone outbreak species that resulted from the
Pleistocene forest loss. By keystone outbreak species, we
refer to defoliating species that occur predictably, over
large enough regions, and achieve high enough local
densities to impact bird diversity at a regional or con-
tinental scale. In particular, the eastern spruce budworm
(Chirostoneura fumiferana) appears to play a funda-
mental role in the diversity of bird species in the Nearctic
boreal forest (MacArthur 1958; Morse 1980). For ex-
ample, at least five of the Nearctic species (Tennessee,
bay-breasted, Cape May, and blackburnian warblers
and chipping sparrow; Erskine 1977) have been thought
to have a strong association with budworm gradations,
and populations of many other species (Erskine 1977)
are found most abundantly in the presence of high
populations of this outbreak lepidopteran. Spruce bud-
worm gradations have a high local impact, often spread
over huge areas of Nearctic boreal forest, and appear to
have long, but relatively predictable cycles (Royama
1984). The intensity and scope of these cycles may ex-
plain the apparent support for increased bird diversity.
Much of the eastern boreal forest experienced spruce
budworm outbreaks in the late 1970s and early 1980s
and budworm populations were peaking in the western
boreal forest in the early 1990s. In fact, compilation of
cumulative defoliation in this period depicts a wave
moving through the boreal forest (Natural Resources
Canada 1996; also references in Royama 1984). The
global population of a number of species, as measured at
netting stations or Breeding Bird Survey routes in a large
area, is correlated with the occurrence of large-scale
budworm outbreaks in eastern Canada (Hussell et al.
1992). In contrast, our review of the Russian and Fenno-
Scandinavian literature and the field work conducted for
over 20 years at the Yenissey site (O. Boursky, personal
communication) showed that local outbreaks of Lepi-
doptera larvae have been documented, but they are local
and do not change the species composition of the fo-
liage-gleaning guild. The one exception is the Siberian
silkworm (Dendrolimus sibericus) which can a?ect up to
10% of the dark coniferous forest of southern Siberia
each year (Kondakov 1974); however, these caterpillars
are well defended and consumed by few birds (Frolov
1938; Rozhkov and Reymers 1958). We are not familiar
with any reports of bird populations on a continental
scale correlated with the epidemic of a specific defoliat-
ing insect.
458
The lack of a similar Eurasian system seems to be
another fundamental di?erence between Eurasian and
Canadian coniferous forests that can be related to
forest history. Consider that the spread of spruce
budworms is facilitated by its occurrence in large
topographically uniform and climatologically and
ecologically similar areas (Royama 1984). With high-
intensity herbivory at the very high population levels
experienced during some outbreaks, local forest patches
may be severely damaged and many trees killed out-
right. This should cause periodic extinction of local
budworm populations (Royama 1984) except where
adults can readily disperse to other favorable sites
which would be more likely in larger areas of appro-
priate habitat. From a geographic perspective, depen-
dence on large uniform tracts of habitat may explain
the common observation that outbreak insects are
thought to be more important and assert a greater
impact on bird density and species richness in the
northern boreal, as opposed to the western Cordilleran
(Wiens 1975) and Appalachian (Rabenold 1978) forests
in the New World. From an historical perspective,
large-scale defoliating systems may also have been dif-
ferentially impacted by the disparate degrees of forest
loss and fragmentation in Nearctic and Palaearctic
forests during Pleistocene glaciations. While dark co-
niferous vegetation is widespread in North American,
European and southern Siberian boreal forest, tracts of
such forest were only common in North America dur-
ing the Pleistocene and were greatly restricted (to
southern Siberia) in Eurasia, where extreme fragmen-
tation would not have favored forest outbreak species.
In summary, loss of keystone irruption species, exem-
plified presently in the Nearctic boreal forest by the
spruce budworm, could be another result of the ex-
treme isolation of remnant boreal forest in the Palae-
arctic during the ice age peaks.
Conclusions
We found the Nearctic foliage-gleaning bird assem-
blages to consist of more species that are specialized on
the foliage substrate, and strongly associated with co-
niferous vegetation. These di?erences are consistent with
a boreal forest biota that has been less severely elimi-
nated, fragmented, and floristically modified during
ongoing ice ages than its Palaearctic equivalent.
Acknowledgements We would like to thank the Institute of Evo-
lutionary Morphology and Animal Ecology of the Russian Acad-
emy of Science, in particular Katya Preobzazhenskaya and Oleg
Boursky, for providing housing and logistical support for work in
Myrnoe and Kostromomskaya Field Stations. Anwar Kiremov
hosted us at the Zvinograd field station. Lee Ann Hayek provided
statistical consultation. Kevin Omland assisted in preparing the
phylogenetic graphics. The following people assisted in gathering
data for this project: Mark Amberle, Sam Droege, Todd Easterla,
Jennifer Gammon, Gjon Hazard, Richard Hoyer, and Olga Ro-
manenko. Valuable comments were provided by Stuart Pimm, Je?
Parrish and an anonymous reviewer. A Pew Fellowship for Biodi-
versity and the Environment provided funding to the senior author.
Appendix 1 Mean values for foraging variables for boreal forest birds
a
Species Site n CON
(%)
FOL
(%)
ATT
(%)
ATDI
(cm)
HA
(%)
HO
(%)
RE
(%)
HT
(%)
RHT
(%)
ZON COV Male
(%)
North America
Yellow-bellied flycatcher
(Empidonax flaviventris)
O 17, 26 28.0 69.0 96.2 120.0 0.0 20.0 0.0 9.1 65.0 2.19 1.38 59
Alder flycatcher
(Empidonax alnorum)
M 17, 20 18.0 61.0 100.0 108.8 0.0 16.7 0.0 8.5 66.0 2.41 1.38 6
Least flycatcher O 92, 104 10.0 77.0 99.0 92.1 1.9 24.0 0.0 10.2 64.0 1.99 1.47 30
(Empidomax minimus) M 61, 74 7.8 85.0 99.0 98.1 0.0 16.2 0.0 12.3 60.4 2.14 1.45 15
Gray jay*
(Perisoreus canadensis)
M 60, 75 97.4 63.0 1.3 20.0 6.4 0.0 25.3 6.9 51.0 2.26 2.01 0
Black-capped chickadee
(Parus atricapillus)
O 106, 118 34.0 53.0 20.5 41.6 48.3 5.0 10.2 9.2 60.0 2.07 1.63 12
Boreal chickadee* O 23, 39 96.0 66.0 5.7 22.7 20.5 3.0 10.3 11.7 68.0 2.30 1.89 0
(Parus hudsonicus) M 80, 106 90.5 71.0 20.3 35.0 41.5 1.0 17.9 12.7 64.7 2.10 1.94 0
Golden-crowned kinglet*
(Regulus satrapa)
O 49, 65 96.0 66.0 54.4 21.9 13.2 25.0 0.0 10.3 59.0 2.26 1.98 31
Ruby-crowned kinglet* O 41, 53 91.0 89.5 50.0 17.5 10.8 17.9 17.1 10.6 60.0 2.28 2.02 7
(Regulus calendula) M 81, 89 92.0 86.0 58.0 13.0 8.0 28.5 18.0 8.2 67.0 2.61 1.77
Warbling vireo
(Vireo gilvus)
M 59, 73 5.7 90.0 72.5 38.9 8.2 24.7 13.7 10.1 72.0 2.30 1.91 20
Philadelphia vireo
(Vireo philadelphicus)
O 32, 35 0.0 83.0 68.6 49.0 8.5 19.0 11.4 13.7 81.0 2.60 2.06 44
Red-eyed vireo O 112, 121 9.0 90.0 70.7 64.7 14.0 14.1 14.0 10.6 65.0 2.23 1.90 39
(Vireo olivaceus) M 59, 83 26.0 92.0 73.3 67.5 2.4 21.7 7.3 10.5 71.5 2.22 1.98 22
Tennessee warbler* O 78, 80 31.0 80.0 12.4 33.0 7.5 24.0 41.3 9.6 72.0 2.49 2.01 18
(Vermivora peregrina) M 162, 210 55.4 90.0 11.3 39.8 14.7 8.1 35.2 8.3 70.0 2.56 1.90 13
Nashville warbler*
(Vermivora ruficapilla)
O 108, 125 28.0 82.0 16.0 18.8 8.0 5.9 25.6 8.9 72.0 2.50 1.90 39
459
Appendix 1 (Contd.)
Species Site n CON
(%)
FOL
(%)
ATT
(%)
ATDI
(cm)
HA
(%)
HO
(%)
RE
(%)
HT
(%)
RHT
(%)
ZON COV Male
(%)
Yellow warbler
(Dendroica petechia)
M 49, 71 5.1 80.0 46.4 14.7 8.5 1.5 28.2 5.1 73.0 2.40 2.03 20
Chestnut-sided warbler
(Dendroica pensylvanica)
O 61, 70 4.0 70.0 47.1 16.6 2.8 15.8 17.1 8.4 69.0 2.14 1.88 52
Magnolia warbler* O 115, 139 47.0 83.0 49.5 24.9 1.4 13.0 15.1 7.4 60.0 2.32 1.94 55
(Dendroica magnolia) M 95, 111 28.0 86.0 54.0 16.6 1.8 16.2 28.0 5.4 53.5 2.26 1.84 42
Cape May warbler*
(Dendroica tigrina)
O 41, 41 95.0 79.0 17.6 30.0 3.4 0.0 13.9 15.3 78.0 2.89 2.17 31
Yellow-rumped warbler* O 143, 167 69.0 60.0 30.8 38.3 4.2 14.4 15.0 9.4 63.0 2.19 1.71
(Dendroica coronata) M 212, 272 88.8 86.0 44.3 30.0 5.2 14.7 18.3 10.4 71.0 2.56 2.03 19
Black-throated green
warbler*
(Dendroica virens)
O 34, 40 26.0 94.0 46.3 21.7 2.0 31.0 22.1 11.3 73.0 2.50 1.95 38
Blackburnian warbler*
(Dendroica fusca)
O 114, 147 57.0 83.0 40.1 32.4 4.8 22.0 18.4 13.6 72.0 2.54 2.22 43
Bay-breasted warbler* O 98, 118 51.0 86.0 35.3 53.3 1.7 12.0 11.0 10.1 64.0 2.54 1.95 45
(Dendroica castanea) M 15, 20 80.0 83.0 30.0 23.5 11.0 5.6 11.0 9.3 62.7 2.45 2.44 42
American redstart O 103, 119 13.0 71.0 78.7 43.3 0.8 20.0 3.3 8.7 62.0 2.10 1.72 47
(Setophaga ruticilla) M 106, 132 20.2 84.0 79.8 64.8 2.3 16.7 9.0 5.9 59.5 2.30 1.68 13
Canada warbler
(Wilsonia canadensis)
O 20, 28 15.0 92.0 82.1 33.4 0.0 30.7 0.0 4.3 58.0 2.11 1.57 50
Western tanager*
(Piranga ludoviciana)
M 58, 86 53.2 87.0 55.8 73.9 3.5 11.6 9.3 12.7 65.0 2.45 2.07 10
Chipping sparrow* O 109, 116 68.0 69.0 26.1 20.2 1.4 3.5 12.1 7.2 54.0 2.46 1.89 29
(Spizella pusilla) M 184, 211 94.0 91.0 11.6 19.2 1.0 3.3 32.0 10.2 69.3 2.61 2.01 4
Dark-eyed junco* O 28, 28 92.0 42.0 29.8 22.7 0.0 10.0 10.7 6.2 45.0 2.14 1.50 36
(Junco hyemalis) M 43, 47 82.1 66.0 26.1 35.0 6.4 0.0 8.5 3.8 47.0 2.50 1.95 2
Eurasia
Eurasian robin
(Erithacus rubecula)
E 45, 68 11.1 40.0 81.8 94.1 0.0 13.2 9.9 5.1 50.1 2.00 1.48 24
Eurasian redstart
(Phoenicurus phoenicurus)
S 34, 50 18.7 38.0 95.0 100.7 0.0 16.0 0.0 7.3 54.0 2.12 1.57
Blythe?s reed warbler E 108, 201 0.0 76.8 38.5 35.7 5.0 5.5 27.0 3.0 54.5 2.27 1.96 33
(Acrocephalus dumetorum) S 57, 98 9.0 74.0 29.7 34.3 1.1 5.1 20.4 3.5 52.0 2.48 2.21
Icterine warbler
(Hippolais icterina)
E 56, 104 6.7 81.3 53.8 40.6 0.7 10.6 23.0 8.2 64.0 2.29 1.82 62
Garden warbler
(Sylvia borin)
E 145, 223 1.1 64.8 40.7 30.0 5.4 8.5 22.4 5.3 60.0 2.19 1.96 44
Lesser whitethroat
(Sylvia curruca)
S 131, 260 23.0 71.0 37.5 83.3 5.0 3.1 10.8 7.5 59.0 2.12 1.93
Whitethroat
(Sylvia communis)
E 62, 108 2.5 64.5 36.5 28.7 2.8 8.3 34.3 3.7 59.2 2.20 2.08 47
Blackcap
(Sylvia atricapilla)
E 62, 112 0.0 80.7 48.2 63.5 5.4 4.5 16.1 7.1 62.2 2.31 1.96 55
Wood warbler
(Phylloscopus sibilitrax)
E 160, 304 14.7 68.2 83.3 48.6 1.6 22.4 6.6 7.8 59.3 2.26 1.41 62
Willow warbler
(Phylloscopus trochilus)
E 269, 432 8.0 69.5 56.9 29.2 5.1 15.5 18.3 7.8 64.7 2.50 2.00 30
Chi?cha? E 70, 136 28.6 70.0 50.5 58.7 2.8 19.1 14.0 9.7 68.3 2.28 1.97 47
(Phylloscopus collybita) S 81, 309 9.0 74.0 42.7 64.7 3.7 37.8 16.4 7.3 70.0 2.39 1.78
Dusky warbler
(Phylloscopus fuscatus)
S 44, 190 0.0 63.4 34.8 65.6 3.7 8.4 12.1 2.8 66.0 2.20 2.00
Arctic warbler
(Phylloscopus borealis)
S 108, 399 2.0 72.7 66.7 66.2 1.1 25.3 4.0 5.5 69.0 2.20 2.00
Greenish warbler
(Phylloscopus trochiloides)
E 176, 330 13.1 73.9 72.4 26.0 3.6 14.9 9.0 9.7 64.7 2.37 1.89 62
Yellow-browed warbler
(Phylloscopus inornatus)
S 281, 411 5.0 72.0 51.8 45.6 4.1 17.3 4.7 6.1 54.0 2.12 1.74
Goldcrest*
(Regulus regulus)
E 136, 242 85.7 66.9 53.7 15.5 17.4 27.3 14.5 11.3 66.2 2.44 1.93 38
Pied flycatcher
(Ficedula hypoleuca)
E 79, 118 12.5 60.8 84.4 176.0 1.7 6.8 2.5 9.0 53.4 1.79 1.41 39
Red-breasted flycatcher E 44, 94 12.5 36.4 85.2 166.6 2.1 9.6 3.0 8.3 55.9 1.81 1.27 70
(Ficedula parva) S 33, 82 25.0 12.5 95.6 111.4 2.4 2.4 1.2 4.9 43.0 2.26 1.18
Willow tit E 261, 481 30.0 41.8 30.1 23.4 46.6 3.1 10.0 9.9 61.2 2.10 1.66
(Parus montanus) S 59, 159 62.1 41.0 5.5 45.0 49.1 7.1 12.0 8.7 59.0 1.19 1.83
460
References
Adams JM, Faure H (1995) Review and atlas of paleovegetation:
preliminary land ecosystem maps of the world since the last
glacial maxima. www.soton.ac.uk/~tgms
Blondel J, Mourer-Chauvre (1998) Evolution and history of the
western Palaearctic avifauna. Trends Ecol Evol 13:488?492
Blondel J, Vuilleumier F, Marcus LF, Terouanne E (1984) Is there
ecomorphological convergence among Mediterranean bird
communities of Chile, California, and France? Evol Biol
18:141?213
Bonan GB, Shugart HH (1989) Environmental factors and eco-
logical processes in boreal forests. Annu Rev Ecol Syst 20:1?28
Bryant J, Tahvanainen J, Sulkinoja M, Julkunen-Tiitto R, Rei-
chardt PB, Green T (1989) Biogeographic evidence for the
evolution of chemical defense against mammal browsing by
boreal birch and willow. Am Nat 134:20?34
Cody ML (1974) Competition and the structure of bird commu-
nities. Princeton University Press, Princeton, NJ
Coope GR (1987) The response of late Quaternary insect com-
munities to sudden climatic changes. In: Gee JHR, Giller PS
(eds) Organization of communities past and present. Blackwell,
Oxford, pp 421?438
Cramp S, Brooks DJ (1992) Handbook of the birds of Europe,
the Middle East, and North Africa: the birds of the Western
Palaearctic, vol VI. Warblers. Oxford University Press, Ox-
ford
Currie DJ, Paquin V (1987) Large-scale biogeographical patterns
of species richness of trees. Nature 329:326?327
Ehrlich PR, Dobkins DS, Wheye D (1988) The birders handbook:
a field guide to the natural history of North American birds.
Simon Schuster, New York
Ehrlich PR, Dobkin DS, Wheye D, Pimm SL (1994) The bird-
watcher?s handbook: a guide to the natural history of the birds
of Britain and Europe. Oxford University Press, Oxford
Erskine AJ (1977) Birds in boreal Canada. Can Wildl Serv Rep Ser
41
Frenzel B (1968) The Pleistocene vegetation of northern Eurasia.
Science 161:637?649
Frenzel B, Pecsi B, Velichko AA (eds) (1992) Atlas of paleoclimates
and paleoenvironments of the northern hemisphere. INQUA/
Hungarian Academy of Sciences, Budapest
Frolov DI (1938) Siberian silkworm in eastern Siberia. Irkust
University, Irkutsk
Greenberg R, Pravosudov V, Sterling J, Kozlenko A, Ko-
ntorschikog V (1999) Warblers, finches, and tits: structure of
boreal forest foliage-gleaning assemblages. Condor 101:299?
310
Grubb PJ (1987) Global trends in species-richness in terrestrial
vegetation: a view from the northern hemisphere. In: Gee JHR,
Giller PS (eds) Organization of communities past and present.
Blackwell, Oxford, pp 99?118
Haila
?
Y, Jarvinen O (1990) Northern conifer forests and their bird
species assemblages. In: Keast A (ed) Biogeography and ecol-
ogy of forest bird communities. SPB, The Hague, pp 61?97
Ha
?
met-Ahti L (1981) The boreal zone and its biotic subdivisions.
Fennia 159:69?75
Helle P, Mo
?
nkko
?
nen M (1990) Forest successions and bird com-
munities: theoretical aspects and practical applications. In:
Keast A (ed) Biogeography and ecology of forest bird com-
munities. SPB, The Hague, pp 299?318
Holmes RT, Recher HF (1986) Determinants of guild structure in
forest bird communities: an intercontinental comparison.
Condor 88:427?439
Holmes RT, Bonney RE, Pacala SW (1979) Guild structure of the
Hubbard Brook bird community: a multivariate approach.
Ecology 60:512?520
Hussell DJT, Mather MH, Sinclair PH (1992) Trends in numbers
of tropical- and temperate-wintering migrant landbirds in mi-
gration at Long Point, Ontario, 1961?1988. In: Hagan J,
Johnston D (eds) Ecology and conservation of Neotropical
Appendix 1 (Contd.)
Species Site n CON
(%)
FOL
(%)
ATT
(%)
ATDI
(cm)
HA
(%)
HO
(%)
RE
(%)
HT
(%)
RHT
(%)
ZON COV Male
(%)
Pallas Warbler*
(Phylloscopus proregulus)
S 35, 115 82.4 65.4 50.0 59.7 7.0 12.2 6.1 18.8 77.0 2.25 23.18
Crested tit*
(Parus cristatus)
E 80, 119 86.2 36.3 33.0 16.5 32.8 10.1 13.0 9.2 61.8 2.21 1.53 27
Blue tit
(Parus caerulus)
E 27, 56 25.0 44.4 16.4 48.6 48.2 5.4 3.0 9.1 67.9 2.19 1.50
Great tit
(Parus major)
E 74, 111 11.8 46.0 34.3 37.2 34.2 9.9 5.4 7.2 60.1 2.15 1.79
Coal tit*(Parus ater) E 19, 27 71.4 73.7 48.0 30.0 29.6 18.5 0.0 13.6 72.6 2.38 2.19
S 26, 41 81.2 68.0 37.0 32.0 41.5 12.2 6.0 14.0 64.0 2.36 1.92
Long-tailed tit E 24, 32 0.0 41.7 25.0 9.3 65.6 3.2 3.0 7.8 60.4 2.30 1.60 22
(Aegilithos caudatus) S 21, 57 0.0 53.0 15.0 28.0 71.9 3.2 1.7 5.9 69.0 2.50 1.60
Cha?nch E 363, 572 10.0 71.9 36.9 58.0 2.6 15.6 10.0 9.5 64.3 2.29 1.90 30
(Fringilla coelebs) S 37, 71 27.0 79.0 47.3 78.9 2.8 19.7 7.0 9.0 67.0 2.37 2.06
Brambling
(Fringilla montifringilla)
S 139, 335 22.0 83.0 31.7 67.4 2.6 11.0 6.5 9.6 67.0 2.19 1.99
Eurasian siskin*
(Carduelis spinus)
E 98, 148 42.2 45.9 8.3 14.9 38.5 0.7 17.6 10.3 70.5 2.59 1.70 13
Bullfinch* E 44, 82 50.0 54.6 50.2 62.8 4.9 36.6 1.2 8.1 58.7 2.23 1.70 9
(Pyrrhula pyrrhula) S 44, 87 47.0 60.5 38.0 41.4 2.3 21.8 16.0 10.4 64.0 2.66 2.21
Scarlet rosefinch
(Carpodacus erythrinus)
E 75, 99 0.0 66.1 16.6 16.2 3.0 1.5 23.5 4.4 72.6 2.40 2.06 27
Yellow-breasted bunting
(Emberiza aureola)
S 73, 151 1.0 84.9 18.1 57.4 0.0 4.6 9.9 4.6 75.0 2.44 2.06
a
Abbreviations for variables are as follows: site (O Ontario, M Mackenzie, E Europe, S Siberia), n sample size (individuals, foraging
maneuvers), CON percent conifer, FOL percent foliage, ATT percent aerial attacks, ATDI mean distance of aerial attack, HA percent
hanging, HO percent hovering, RE percent Reach, HT bird height, RHT bird height/tree height ? 100, ZON mean horizontal zone, COV
mean cover class, Male percent singing males. * Species that is reported to be primarily associated with coniferous vegetation by Ehrlich
et al. (1988, 1994), n is individuals, maneuvers
461
migrant landbirds. Smithsonian Institution Press, Washington,
DC, pp 101?114
Karr JR, James FC (1975) Ecomorphological configurations and
convergent evolution in species and communities. In: Cody ML,
Diamond J (eds) Ecology and evolution of communities.
Belknap, Cambridge, Mass, pp 258?291
Kondakov YP (1974) The regularities of Siberian silkworm mass
reproduction. Novosivirsk Sci 1974:206?265
Kozlenko A (1987) Post-fire succession and bird population
changes in central Siberia. PhD thesis, Institute of Comparative
Ecology, Moscow
Latham RE, Ricklefs RE (1993) Continental comparisons of tem-
perate-zone tree species diversity. In: Ricklefs RE, Schluter D
(eds) Species diversity in ecological communities: historical and
geographical perspectives. University Chicago Press, Chicago,
pp 294?314
MacArthur RH (1958) Population ecology of some warblers of
northeastern coniferous forests. Ecology 39:599?619
Maddison WP, Maddison DR (1992) MacClade: analysis of phy-
logeny and character evolution. 3.04. Sinauer, Sunderland,Mass
Mo
?
nkko
?
nen M (1994) Diversity patterns in Palaearctic and Ne-
arctic forest bird assemblages. J Biogeogr 21:183?195
Mo
?
nkko
?
nen M, Welsh D (1994) A biogeographical hypothesis on
the e?ects of human-caused landscape changes on the forest
bird communities of Europe and North America. Ann Zool
Fenn 31:61?70
Morse DH (1980) Foraging and coexistence of spruce-woods
warblers. Living Bird 18:7?25
Natural Resources Canada (1996) GeoAccess spatial data pilot
project: cumulative defoliation of conifer forests by spruce
budworm 1980?1995. www.nais.ccm.emr.ca/~lacasse/spatial-
data/01homee.htm
Newton I (1972) Finches. Collins, London
Niemela
?
J, Langor D, Spence JR (1994) Logging and boreal
ground-beetle assemblages on two continents: implications for
conservation. In: Gaston KJ, Samways M, New T (eds) Per-
spectives in insect conservation. Intercept, Andover, pp 132?142
Niemi GJ (1985) Patterns of morphological evolution in bird
genera of New World and Old World peatlands. Ecology
66:1215?1228
Parrish J (1995) Experimental evidence for intrinsic microhabitat
preferences in the black-throated green warbler. Condor
97:935?943
Price T (1991) Morphology and ecology of breeding warblers along
an altitudinal gradient in Kashmir, India. J Anim Ecol 60:643?
664
Rabenold KN (1978) Foraging strategies, diversity, and seasonality
in bird communities of Appalachian spruce-fir forests. Ecol
Monogr 48:397?424
Remsen JV (1985) Community organization and ecology of birds
of high elevation humid forests of the Bolivian Andes. Ornithol
Monogr 36:733?753
Remsen JV, Robinson SK (1990) A classification scheme for for-
aging behavior of birds in terrestrial habitats. Stud Avian Biol
13:133?160
Ricklefs RE, Miles DB (1994) Ecological and evolutionary infer-
ences from morphology: an ecological perspective. In: Wain-
right PC, Reilly SM (eds) Ecological morphology: integrative
organismal biology. University Chicago Press, Chicago, pp 13?
41
Ricklefs RE, Travis J (1980) A morphological approach to the
study of avian community organization. Auk 97:321?338
Robinson SK, Holmes RT (1982) Foraging behavior of forest
birds: the relationship among search tactics, diet, and habitat
structure. Ecology 63:1918?1931
Royama T (1984) Population dynamics of the spruce budworm
Choristoneura fumiferana. Ecol Monogr 54:429?462
Rozhkov AS, Reymers NF (1958) Data on birds feeding on
Siberian silkworms (Dendrolimus superans sibiricus Tschtev.) in
the areas of mass reproduction of the pest. Zool J 37:1749?
1750
Sabo SR, Holmes RT (1983) Foraging niches and the structure of
forest bird communities in contrasting montane habitats.
Condor 85:1210138
Schluter D (1986) Tests for similarity and convergence of finch
communities. Ecology 67:1073?1085
Scmiegelow FKA, Machtans CS, Hannon SJ (1997) Are boreal
birds resilient to forest fragmentation? An experimental study
of short-term community responses. Ecology 78:1914?1932
Scott GAJ (1995) Canada?s vegetation: a world perspective.
McGill-Queen?s University Press, Montreal
Sibley CG, Ahlquist JE (1990) Phylogeny and classification of
birds: a study of molecular evolution. Yale University Press,
New Haven
Statsoft (1995) STATISTICA for Windows (volume III): Statistics
II, 2nd edn. Tulsa, Okla
Terborgh J, Robinson S (1985) Guilds and their utility in ecology.
In: Kikkawa J, Anderson DJ (eds) Community ecology: pattern
and process. Blackwell, Oxford, pp 65?90
Walter H (1979) Vegetation of the earth and ecological systems of
the geo-biosphere. Springer, Berlin
Webb T III, Bartlein PJ, Harrison SP, Anderson KJ (1993) Vege-
tation, lake levels and climate in eastern North America for the
past 18,000 years. In: Wright HE Jr, Kutzbach JE, Webb T III,
Ruddiman WF, Street-Perrott A, Bartlein PJ (eds) Global cli-
mates since the last glacial maximum. University of Minnesota
Press, Minneapolis, pp 415?467
Welsh DA (1987) The influence of forest harvesting on mixed co-
niferous-deciduous boreal bird communities in Ontario. Acta
Oecol 8:247?252
Wiens JA (1975) Avian communities, energetics, and functions in
coniferous forest habitats. In: Smith DR (ed) Proceedings of the
symposium on management of forest and range habitat for
non-game birds. General technical report WO-I. US Forest
Service, Washington, DC, pp 226?265
Wiens JA (1989) The ecology of bird communities,vol. 1. Cam-
bridge University Press, Cambridge, UK
Wiens JA (1991) Ecomorphological comparisons of the shrub-
desert avifaunas of Australia and North America. Oikos
60:50?63
462