November 2001 / Vol. 51 No. 11 · BioScience 967
Articles
One species that typifies the northern forests of the
eastern United States in the minds of many ecologists
and naturalists is the eastern white pine (Pinus strobus L.).Its
range includes New England, the Great Lakes region, New
York,Pennsylvania,and along the Appalachian Mountains to
northern Georgia (Figure 1).The northern limits of its range
include southwestern Ontario,above Lake Superior,to south-
ern Newfoundland. In these forests, white pine is the tallest
tree,with some individuals reaching a height of 150­175 feet
(45­53 m) and a diameter of 40­50 inches (102­127 cm).The
great height of white pine makes it an emergent above the up-
per canopy in many mixed-species forests.
Historically, white pine has been one of the most valuable
tree species because of the large volume of standing timber
and its multiple uses as a lightweight building material with
a straight grain.Intensive logging for the highly preferred white
pine started in New England in the 1700s and moved west-
ward to the Great Lake states by the mid-1800s.Indeed,white
pine has been called"the tree that builtAmerica"and"the tree
that won theAmerican Revolution,"the latter referring to the
outcome when the British Navy could no longer obtain the
trees for their ship masts and planks (Harlow et al.1979).The
amount of white pine logged from the northern forests be-
fore 1900 measures into the hundreds of billions of board feet
(Goodlett 1954, Whitney 1994).
Ironically, white pine was a relatively small component of
the pre-European forests of the northeastern United States,
compared with the dominance of hemlock (Tsuga canaden-
sis),beech (Fagus grandifolia),oak (Quercus),and sugar maple
(Acer saccharum) (Lutz 1930a, Whitney 1994, Abrams and
Ruffner 1995). Thus, white pine's reputation for great dom-
inance in the original forest was based on inflated estimates
to attract settlers as well as on its easy accessibility.Before the
railroad era,the earliest logging was concentrated along river
systems because of the ease of floating or rafting the logs to
sawmills or markets. Early land surveys of the original forest
suggest that high concentrations of white pine were mainly
found on light,sandy soils along rivers (Whitney 1990,1994,
Cogbill 2000). This location, coupled with its great size and
highly valued wood,made white pine an easy target for early
logging.However,white pine also dominated sandy outwash
deposits prone to burning and dry upland areas affected by
frequent blow-down (wind-related tree fall), forming pre-
settlement pineries or pine plains. Various studies (Braun
1950, Whitney 1986, Nowacki and Abrams 1994) suggest
that white pine could occupy a range of soil,moisture,and dis-
turbance conditions even wider than those normally associ-
ated with this species.
White pine is a major component of five forest cover types
recognized by the Society of American Foresters: white pine,
white pine­hemlock,white pine­northern red oak (Q.rubra),
Marc D. Abrams (e-mail: agl@psu.edu) is Steimer Professor of For-
est Science, as well as professor of forest ecology and tree physi-
ology, at the School of Forest Resources, Pennsylvania State Uni-
versity, University Park, PA 16802. His specific research interests
are in the long-term temporal and broad spatial changes that have
occurred in the northern and central hardwood and conifer forests
of the eastern United States.  2001 American Institute of Biological
Sciences.
Eastern White Pine
Versatility in the
Presettlement Forest
MARC D. ABRAMS
THIS EASTERN GIANT EXHIBITED VAST
ECOLOGICAL BREADTH IN THE ORIGINAL
FOREST BUT HAS BEEN ON THE DECLINE
WITH SUBSEQUENT LAND-USE CHANGES
968 BioScience · November 2001 / Vol. 51 No. 11
Articles
white pine­chestnut oak (Q.prinus),and red pine (P.resinosa)
(Burns and Honkala 1990). Ecologically, white pine is best
known for its role in the hemlock­white pine­northern hard-
wood forest that stretches along the northern tier of the east-
ern United States and southern Canada (Figures 1 and 2a;
Nichols 1935,Braun 1950).Not only is white pine widely dis-
tributed and associated with a large variety of species, but it
can also persist in various successional stages within these
forests.Indeed,white pine has been documented as a pioneer
after disturbance,a long-lived successional tree in mature or
climax forests,and a physiographic or fire-maintained climax
type.
The ecology of white pine was the subject of ecological stud-
ies throughout the 20th century,yet these studies did not fully
capture the ecological breadth of white pine's role in the
eastern forest (Lutz 1930b, Nichols 1935, Hough and Forbes
1943,Hibbs 1982).Early ecologists concluded that it became
established after large-scale disturbances, such as agricul-
tural abandonment, fire, and blow-down, and was replaced
by shade-tolerant trees during succession.However,this con-
clusion does not explain its role in highly contrasting forest
types,including the presence of multiple-age white pine in old-
growth forests that have not been intensively disturbed for 300
or more years (Abrams and Orwig 1996,Abrams et al.2000).
Moreover,many of the ideas formulated about the ecology of
white pine during the 20th century lacked dendroecological
(tree ring) data. Dendroecological techniques have become
an important tool in the study of stand dynamics and eco-
logical history.Studying tree-ring chronologies along with age
structure and land-use history can elucidate species recruit-
ment patterns,periodicity and intensity of disturbance,pop-
ulation dynamics,and successional pathways.My students and
I used these techniques to study the dynamics of white pine
in a variety of old-growth forests of the eastern United States.
From these studies, we found a wide array of canopy re-
cruitment patterns and multiple successional pathways for the
species.
Information on species'distribution in the original forest
and contrasting successional roles in old-growth forests sug-
gests that white pine defies easy classification or pigeonhol-
ing.Moreover,white pine has exhibited dramatic changes in
forest dominance from presettlement to the present day,with
most regions showing a decrease for the species (Whitney
1987, Abrams and Ruffner 1995, Abrams and McCay 1996,
Foster et al.1998).In this article,I explore the ecology of white
pine in the pre-European settlement forest by using data
from early land surveys (witness trees) and dendroecology
(tree ring) studies of a variety of old-growth forests. Den-
droecological research of old-growth forests and witness tree
data serve as a window to the past and allow for the recon-
struction of species'distribution and dynamics during a time
that,in general,was under much less anthropogenic pressure
than are present-day second- and third-growth forests. The
results of this approach should increase our understanding
of the complex ecology for this dominant species and shed
light on its past,present,and future status,including reasons
for the postsettlement decline in this giant of the eastern
forest.
Figure 1. Range map for eastern white pine.
Ecophysiological attributes
of white pine
White pine is generally considered to be in-
termediate in shade tolerance and an early-
to-middle successional species (Burns and
Honkala 1990). Consistent with these
ideas,saplings of understory eastern white
pine in northwestern Connecticut had
65% survival over a 5-year period, which
was intermediate among eight hardwood
tree species plus hemlock (Pacala et al.
1996).Black birch (Betula nigra) and hem-
lock had more than 90% survival,whereas
white ash (Fraxinus americana), red oak,
and red maple (Acer rubrum) had
20%­38% survival during the same time
interval. White pine saplings appear to
have very slow radial growth at very low
light levels (less than 1% full sun) but in-
termediate growth in light greater than
10% full sun,as compared with nine other
eastern tree species (Pacala et al. 1996).
Similarly, a greenhouse study comparing
the shoot growth of tree species common
to old fields in the Hudson River valley of
southeastern NewYork reported that white
pine had much lower growth than sugar
maple and red oak in low light (Canham
et al. 1996). All species had greatly in-
creased growth with higher light levels.A
study of suppressed trees over a 19-year
period at Harvard Forest in central Mass-
achusetts rated white pine as tolerant be-
cause it had a high survival rate (59%)
and one of the highest mean diameter
growth rates among nine tree species
(Lorimer 1983).These studies suggest that
white pine has a slow height growth rate
but moderate survival in the understory of
second-growth, mixed species forests.
White pine witness
tree data
Land surveys during the period of early European settle-
ment in the eastern United States involved the use of witness
(or bearing) trees to identify property corners and other
boundaries (Lutz 1930a, Whitney 1994). For many areas,
witness trees represent the only quantitative information on
presettlement forest composition and, therefore, are an in-
valuable resource,despite potential biases in tree selection by
early surveyors (Whitney 1994). Witness tree data are best
viewed as regional rather than stand-level information because
of the small number of trees used within any one forest.
Moreover,in several instances the land surveyors,particularly
in New England and the mid-Atlantic Ridge and Valley
Province, listed pine at the genus level, not at the species
level.Thus,some of the pine from these regions included pitch
pine (P. rigida) or red pine along with white pine.
White pine in the pre-European settlement forests of New
England ranged from 0% to 22% of the total number of wit-
ness trees surveyed (Table 1).It is the first-rank species in only
two of the 11 witness tree surveys reviewed here.White pine
grew mainly in association with oak and chestnut (Castanea
dentata) and to a much lesser extent with beech, maple, and
hemlock.White pine was widely distributed in colonial forests
of central New England, averaging 20% (with a range of
16% to 22%) of the witness trees in the Connecticut River val-
ley, Pelham Hills, central uplands, and eastern lowlands,
where it was an important member of the hemlock­
November 2001 / Vol. 51 No. 11 · BioScience 969
Articles
Figure 2. (a) A remnant, old-growth hemlock­white pine­northern hardwood
forest at Heart's Content in northwest Pennsylvania, (b) wind-snapped old-
growth hemlock (foreground) and white pine in the Cook Forest in northwest
Pennsylvania, (c) even-age cohort of white pine in an old gap in old-growth
mixed-oak­pine forest in eastern West Virginia, (d) 300-year-old white pine
(background) and hemlock in Cook Forest in northwest Pennsylvania,
(e) uneven-age white pine and hemlock on the steep rocky slope of Ice Glen
Natural Area in southwest Massachusetts.
northern hardwood,oak­pine ­hemlock,and oak­pine­chest-
nut forest types (Foster et al. 1998).
In 179 town surveys (48,260 trees) from northern New Eng-
land and northern NewYork, pine witness trees (most being
white pine) are found in 60% of the townships with an av-
erage of only 5% density (Cogbill 2000). White pine was
12% of the presettlement forest in Chittenden County, Ver-
mont,and part of the northern hardwood­hemlock forest that
occurred in ravines, slopes, and sandy terraces (Siccama
1971).White pine represented less than 2% of all trees in pre-
settlement surveys of 115 towns in northern Maine,which is
consistent with previous work in northeastern Maine (Table
1; Lorimer 1977). Cogbill (2000) concludes that white pine
was generally a minor component in the forest of northern
New England. It occurred consistently only in special cir-
cumstances: on sandy outwash, on rocky ridges, or on the
floodplains of large rivers, particularly if they had been
burned. Even in these conditions, it was never abundant ex-
cept on the sandy outwash areas, which were mixed-species
areas designated as"pine plains."Indeed,pine had a maximum
of 57% forest composition in the Merrimack River valley of
New Hampshire,a sandy coastal plain site that most likely had
a predominance of pitch pine (Cogbill 2000).
In the mid-Atlantic region,white pine ranged from 0% to
38% of individuals in the presettlement forests (Table 1). It
was the first-rank species in only one of 18 witness tree data
sets reviewed; it was not listed as a dominant species in nine
of these studies. In contrast, beech and white oak were listed
as the first-rank species in nine and five of these studies, re-
spectively.White pine was typically associated with northern
hardwoods and hemlock in the north and with white oak fur-
ther south. However, beech­sugar maple­hemlock was a
much more common forest association than hemlock­north-
ern hardwood­white pine forest on theAllegheny Plateau.The
highest frequency of white pine (38%) among witness trees
was recorded with beech,hemlock,spruce,and white oak on
the Pocono Plateau of east-central Pennsylvania (Dando
1996).
On theAllegheny High Plateau of northwest Pennsylvania,
white pine represented only 6% of the witness trees, and ex-
tensive stands of the species were mainly limited to coarse-tex-
tured outwash terraces of the Allegheny River and its tribu-
tary creeks (Lutz 1930a,Whitney 1990).Old-growth white pine
was mainly concentrated on the upper,southeast-facing slope
of the Tionesta River valley,growing with hemlock and chest-
nut on shale-derived soils (Hough and Forbes 1943).In Pot-
ter County, white pine grew in"great forests"along many of
the river valleys and creeks to the north and on ridge tops and
hillsides to the south (Goodlett 1954). In the Allegheny
Plateau of western New York, white pine in the original for-
est was concentrated in the southern counties in a dissected
landscape of river drainages, where it formed part of
oak­chestnut­pine and pine­beech­hemlock­white oak
forests (Seischab 1990).In theAllegheny Mountains of Penn-
sylvania,white pine represented about 10% of the witness trees
and dominated cove forests and, to a lesser extent, stream
valleys (Abrams and Ruffner 1995).In eastern WestVirginia,
white pine dominated mountaintops and southeast-facing
slopes; it was not recorded in cove and valley floor forests,sug-
gesting that it occurred at higher elevations at the transition
of the central and southernAppalachian Mountains (Abrams
and McCay 1996). In the Ridge and Valley Province of cen-
tral Pennsylvania, white pine mainly occurred with white
oak on rich mesic valleys,stream valleys,and mountain coves
(Nowacki and Abrams 1992, Abrams and Ruffner 1995).
White pine was rarely a first-rank species in the pre-
European forests in the Great Lakes states. But many forests
in the region had a white pine component of 5%­35% (Table
1).White pine was associated with hemlock­northern hard-
woods or mixed-pine forests on upland sites and with tama-
rack (Larix), cedar (Thuja), and spruce (Picea) on poorly
drained sites. In the Upper Peninsula of Michigan, white
pine constituted 10% of the witness trees recorded in well-
drained regions and 6% of those recorded in poorly drained
regions (Zhang et al. 2000). Hemlock­white pine­northern
hardwood forests in northern and lower Michigan had 3%
white pine,compared with 19% in mixed-pine forests (Whit-
ney 1987). The highest dominance of white pine (35%) was
recorded with red pine, aspen (Populus), and birch on
dry­mesic, sandy­loam forests in north-central Wisconsin
(Nowacki et al.1990).In northern-lower Michigan,white pine
in the original forest occupied a wide range of soils and
drainage classes,but it showed a strong association with loam
to sandy­loam, ice-contact­derived soils (Whitney 1986).
White pine in old-growth forests in Upper Michigan domi-
nated south-facing, excessively drained, sandy­loam soils
and exposed rocky ridges (Pregitzer and Barnes 1984).
According to Braun (1950),the hemlock­white pine­north-
ern hardwood forests existed as an alternation of deciduous,
coniferous, and mixed-forest communities, with white pine
being most characteristic of conifer forests on dry sand plains
and ridges.White pine was a part of the mixed-forest type but
was subordinate to hemlock and northern hardwoods.Thus,
the hemlock­white pine­northern hardwood forest type ex-
isted mainly at the regional rather than at the stand level.
These studies suggest that white pine was widely distrib-
uted in the presettlement forest and that it occurred in a
large range of topographic positions and moisture condi-
tions.White pine occurred mainly on sandy and sandy­loam
soils, but it was not limited to these soil textures.White pine
was rarely a first-rank dominant at the regional level, but it
probably occurred in nearly pure stands in specialized sites
such as river valleys and stream valleys, side slopes and ridge
tops, and glacial outwash. Therefore, its distribution was
somewhat bimodal, with peaks in riparian valleys and dry,
nutrient-poor uplands.Sugar maple and beech probably out-
competed white pine on mesic, rich sites, and white pine
had to compete with hemlock on cool,moist sites.Pitch pine
was probably much more important than white pine on the
coastal plain of New England and the northern mid-Atlantic,
where white pine may have been limited by climate,soils,and
frequent aboriginal burning on the coastal plain (Patterson
970 BioScience · November 2001 / Vol. 51 No. 11
Articles
November 2001 / Vol. 51 No. 11 · BioScience 971
Articles
Table 1. Percentage composition of witness tree species in pre-European settlement forests in the hemlock­white
pine­northern hardwood range.
Region, location Presettlement forest composition Reference
New England
Northeast Maine Spruce (20%), birch (17%), beech (15%), balsam fir (14%), cedar (12%), pine (1%) Lorimer 1977
Western Maine Spruce (37%), birch (20%), fir (11%), maple (9%), cedar (6%), pine (0%) Cogbill 2000
Central New Hampshire Beech (33%), hemlock (22%), spruce (16%), maple (11%), birch (5%), pine (5%) Cogbill 2000
Northern Vermont Beech (30%), spruce (16%), maple (15%), birch (11%), hemlock (11%), pine (1%) Siccama 1971
Northern Vermont Maple (24%), birch (16%), pine (12%), hemlock (11%) Siccama 1971
Western Vermont Beech (34%), maple (20%), ironwood (9%), pine (8%) Cogbill 2000
Northeast Massachusetts White oak (27%), black oak (26%), pine (18%), hickory (9%) Whitney and Davis 1986
North-central Massachusetts Pine (22%), white oak (17%), chestnut (11%), black oak (9%) Whitney 1994
North-central Massachusetts
Upland Pine (22%), oak (21%), hemlock (14%), maple (10%) Foster et al. 1998
Lowland Oak (59%), pine (17%), chestnut (6%) Foster et al. 1998
Northwest Connecticut
Mesic Beech (40%), hemlock (25%), maple (9%), birch (6%), white pine (2%) Winer 1955
Dry­mesic Beech (20%), chestnut (14%), hemlock (13%), maple (13%), oak (13%) Winer 1955
Dry Oak (28%), chestnut (10%), white pine (10%), walnut (10%), maple (9%) Winer 1955
Mid-Atlantic region
Allegheny Plateau
Western New York Beech (19%), oak (19%), sugar maple (11%), hemlock (11%), white pine (11%) Seischab 1990
Western New York Beech (47%), sugar maple (22%), white pine (2%) Marks and Gardescu 1992
Western New York, swamps Black ash (47%), hemlock (9%), tamarack (7%), white pine (1%) Marks and Gardescu 1992
Northwest Pennsylvania Beech (31%), hemlock (27%), sugar maple (8%), white pine (6%) Lutz 1930a
Northwest Pennsylvania Beech (43%), hemlock (20%), birch (6%), white pine (3%) Whitney 1990
North-central Pennsylvania Beech (50%), sugar maple (17%), hemlock (17%), birch (8%), white pine (2%) Goodlett 1954
North-central Pennsylvania Beech (33%), hemlock (32%), sugar maple (8%), white pine (2%) Abrams and Ruffner 1995
Allegheny Mountains
Central Pennsylvania White oak (19%), maple (15%), chestnut (14%), white pine (11%) Abrams and Ruffner 1995
Southern West Virginia White oak (24%), chestnut (11%), hickory (9%), white pine (5%) Abrams et al. 1995
Eastern West Virginia Maple (19%), beech (13%), hemlock (10%), basswood (9%), pine (0%) Abrams and McCay 1996
Ridge and Valley
Southeast New York White oak (47%), black oak (16%), hickory (12%), pine (3%) Glitzenstein et al. 1990
Central Pennsylvania, valley White oak (41%), hickory (12%), pine (13%), black oak (9%) Nowacki and Abrams 1992
Central Pennsylvania, Oak (33%, pine (27%), hemlock (17%), chestnut (7%) Nowacki and Abrams 1992
mountain cove
Eastern Pennsylvania Pine (28%), white oak (23%), hemlock (9%), chestnut (7%), maple (6%) Dando 1996
Eastern West Virginia White oak (33%), pine (16), chestnut oak (8%) Abrams and McCay 1996
Pocono Plateau, Pennsylvania Pine (38%), beech (12%), hemlock (9%), spruce (7%), white oak (7%) Dando 1996
Till plain, western New York Beech (32%), sugar maple (18%), basswood (12%), white oak (11%), white pine (2%) Seischab 1990
Catskill Mountains, eastern NY Beech (49%), hemlock (20%), sugar maple (13%), birch (7%), white pine (< 1%) McIntosh 1962
Lake States
North-central Wisconsin
Mesic Hemlock (27%), birch (25%), maple (17%), white pine (5%) Nowacki et al. 1990
Dry­mesic White pine (35%), red pine (14%), aspen (12%), birch (10%) Nowacki et al. 1990
Northern-lower Michigan
Mesic Beech (51%), hemlock (20%), sugar maple (17%), white pine (3%) Whitney 1987
Dry­mesic Red pine (41%), white pine (19%), jack pine (18%), white oak (9%) Whitney 1987
Swamps Tamarack (30%), cedar (25%), spruce (11%), white pine (8%) Whitney 1987
Mesic Hemlock (44%), beech (22%), white pine (11%) Palik and Pregitzer 1992
Dry­mesic Red pine (45%), jack pine (17%), white pine (15%) Palik and Pregitzer 1992
Northern-lower Michigan
Mesic Hemlock (32%), white pine (26%), red pine (26%) Kilburn 1960
Dry­mesic Red pine (40%), white oak (19%), white pine (15%), aspen (12%) Kilburn 1960
Xeric Jack pine (61%), red pine (27%), white pine (4%) Kilburn 1960
Northern Michigan, swamp Cedar (41%), tamarack (18%), spruce (9%), birch (8%), pine (1%) Deelen et al. 1996
Northern Michigan
Well-drained Beech (14%), sugar maple (14%), white pine (10%), hemlock (10%) Zhang et al. 2000
Poorly drained Tamarack (23%), white cedar (16%), spruce (14%), hemlock (9%), white pine (6%) Zhang et al. 2000
Eastern Wisconsin Sugar maple (45%), beech (29%), hemlock (15%), basswood (12%), pine (1%) Ward 1956
Eastern Wisconsin Beech (36%), sugar maple (20%), hemlock (10%), white pine (10%) Ward 1956
Northern Minnesota Jack pine (22%), aspen (21%), red pine (19%), white pine (13%) Spurr 1954
and Sassaman 1988).White pine may have reached its great-
est distribution on coarse-textured glacial outwash soils in the
Great Lake states, which were probably drier and more fre-
quently burned than their counterparts in New England
(Whitney 1994, Clark and Royall 1996).
Presettlement versus present-day forests. Compar-
isons of presettlement versus present-day forest composition
of white pine indicate that an overall decline has occurred in
the species in northeastern and Great Lakes states forests
(Table 2). All eight case studies from the Great Lake region
show a decline in white pine percentages from past to pres-
ent, whereas four of the nine case studies in the Northeast
show an increase. In most cases, however, the magnitude of
the white pine increases is rather modest. One exception in-
volves the pine plains of Concord, Massachusetts, in which
presettlement forests were probably dominated by pitch pine
(Whitney and Davis 1986). The dramatic post-European
settlement increase in white pine at Concord has been at-
tributed to its prolific seed production and ability to invade
abandoned pastureland. Past increases in white pine in the
Hudson River valley of New York have also been explained
by its invasion of abandoned agricultural lands (Glitzen-
stein et al. 1990).
The frequency and magnitude of white pine decreases
from presettlement to the present day are rather striking. A
35% decrease in white pine composition occurred on dry­
mesic sites in north-central Wisconsin (Nowacki et al.1990).
White pine in the Ridge and Valley Provinces of Pennsylva-
nia and West Virginia experienced a 23% and 17% decline,
respectively (Abrams and Ruffner 1995,Abrams and McCay
1996).In several cases,white pine has been nearly extirpated.
The selective logging of the highly prized white pine in the
original forests,resulting in the loss of a local seed source fol-
lowed at times by catastrophic wildfire, has undoubtedly
contributed to the decrease in white pine dominance in many
areas (Whitney 1987, Nowacki et al. 1990, Palik and Pregit-
zer 1992, Frelich 1995). During the 20th century, white pine
was faced with a decrease in cutting and fire and an increase
in deer browsing, which further exacerbated the decline in
white pine dominance in many areas (Abrams and Ruffner
1995, Frelich 1995).
Multiple pathways of white pine
recruitment in old-growth forests
During succession,the ecological role of species can be highly
variable.Variations in disturbance type and intensity,coupled
with complex species reactions and chance events,can result
in a multitude of pathways in species recruitment, forest de-
velopment, and successional changes.
Early and mid-successional white pine. White pine
is probably best known for its role in the early successional de-
velopment of forests.However,the ecophysiological attributes
of white pine allow it to exploit periodic disturbances and re-
cruit into the middle successional stages of some forests.
CookForest,Pennsylvania. Dendroecological techniques
were used to examine the patterns of canopy recruitment in
relation to disturbance history for white pine and hemlock at
Cook Forest Cathedral Pines, located on the upper slope of
972 BioScience · November 2001 / Vol. 51 No. 11
Articles
Table 2. Pre-European settlement versus modern forest composition of white pine.a
Species Location Presettlement (%) Modern (%) Reference
Northeast
Pine Champlain Valley, VT 8.5 11.6 Siccama 1971
Pine Green Mountains, VT 1.4 7.6 Siccama 1971
Pitch pine Concord, MA 18.5 -- Whitney and Davis 1986
White pine Concord, MA 0.3 23.3 Whitney and Davis 1986
Pine Central Massachusetts 19.9 15.2 Foster et al. 1998
Pine Hudson Valley, New York 3.5 9.2 Glitzenstein et al. 1990
White pine Allegheny Plateau, Pennsylvania 3.1 0.4 Whitney 1990
White pine Allegheny Mountains, Pennsylvania 10.5 0.8 Abrams and Ruffner 1995
Pine Ridge and Valley, Pennsylvania 25.6 -- Abrams and Ruffner 1995
White pine Ridge and Valley, Pennsylvania -- 2.6 Abrams and Ruffner 1995
Pine Ridge and Valley, West Virginia 18.2 -- Abrams and McCay 1996
White pine Ridge and Valley, West Virginia -- 1.4 Abrams and McCay 1996
Lake States
White pine Roscommon County, Michigan 18.0 3.0 Whitney 1987
White pine Huron, MI 15.3 0.0 Palik and Pregitzer 1992
White pine Upper Peninsula, Michigan 10.3 3.5 Zhang et al. 2000
White pine Upper Peninsula, Michigan, swamp 6.0 1.5 Zhang et al. 2000
White pine Eastern Wisconsin 10.3 3.7 Ward 1956
White pine North-central Wisconsin
Mesic 4.6 0 Nowacki et al. 1990
Dry­mesic 35.1 0 Nowacki et al. 1990
Red pine/ Lake County, Minnesota 29.5 5.9 Frelich 1995
white pine
a. The pine designation in the presettlement surveys most likely indicates a majority of white pine.
a river valley on theAllegheny Plateau in northwestern Penn-
sylvania (Abrams and Orwig 1996).This 300-year-old primary
stand growing on a stony loam soil is dominated by hemlock
(53%), white pine (26%), beech (10%), and yellow birch
(Betula alleghaniensis, 8%). Hemlock had an extremely high
density,averaging more than 11 trees per 0.02-ha plot.In con-
trast,white pine averaged only two trees per plot but had a high
importance value because of the large size of almost all of the
existing trees (Figure 2d).White pine recruitment into the tree-
size class occurred throughout the initial 110-year period
(1690­1800) of the stand history,but it dominated during the
first 40 years (Figure 3).Most recruitment of hemlock occurred
between 1740 and 1800,although it began as early as 1705 and
continued until 1900. Most of the younger trees belonged to
several northern hardwood species.
This stand probably experienced a catastrophic distur-
bance,such as tornado or fire,in the 1690s.Evidence for this
disturbance is the absence of any hemlock trees older than 284
years, even though they can exceed 400 years (Hough and
Forbes 1943). Furthermore, there was a pulse of white pine
recruitment between 1690 and 1730,consistent with many old-
growth forest studies in which white pine existed as even-age
cohorts that became established soon after large-scale dis-
turbances (Lutz 1930b, Cline and Spurr 1942, Heinselman
1973, Foster 1988a). However, the recruitment pattern of
white pine in this study is not consistent with the idea of strictly
early-successional,even-age populations,because its initial re-
cruitment was followed by an additional 70 years of concur-
rent white pine and hemlock recruitment.
White pine also exhibited huge variation in annual tree-ring
width.Radial growth in young white pine entering the stand
around 1700 grew as much as 4­5 mm annually (Abrams and
Orwig 1996). This growth rate dropped to a few tenths of a
millimeter in mature trees during decades of suppression,fol-
lowed by prolonged growth increases up to 2­3 mm after dis-
turbance.This degree of growth plasticity enables white pine
to respond to a wide variation in growth conditions during
its long life in an old-growth forest.
None of the existing white pine recruited in Cook Forest
after 1800 (Figure 3). These results concur with those re-
ported for a former virgin stand of white pine­hemlock­
northern hardwoods in the same region,in which white pine
recruitment peaked during the initial 80 years of postdistur-
bance stand development,continued at a lower level for an ad-
ditional 80 years, and then stopped completely (Hough and
Forbes 1943). Thus, white pine can apparently play the role
of both early and middle successional species in forests on the
Allegheny Plateau. The initial domination of tree recruit-
ment by white pine reflects its faster height growth relative to
hemlock rather than differential establishment dates between
the two species (Abrams and Orwig 1996). The forest expe-
rienced increased hardwood invasion in the early 20th
century. However, since 1930 recruitment for all species was
practically nonexistent, despite severe storms in the 1950s.
Presently, the stand is virtually devoid of an understory be-
cause of intensive deer browsing.
Old-growth, old-field white pine. White pine frequently
invades abandoned agricultural fields and pastures (called old
fields) throughout its range,playing the role of an early-suc-
cessional, pioneer tree species (Hibbs 1982, Whitney and
Davis 1986, Burns and Honkala 1990). Because of the early
settlement through much of New England, some old fields
have been abandoned for 2 centuries or more. The original
white pine that invaded these long-since abandoned fields is
now considered to be old growth. In the case of Cathedral
Pines near Cornwall, Connecticut, the oldest white pine
formed two age classes that were established about 1780 and
1840 (Patterson and Foster 1990).It was considered one of the
last remaining stands of old-growth white pine in southern
New England. White pine in other forests in central New
England dates back to agricultural and pastured fields aban-
doned in the mid-1800s (Foster et al. 1998). In the case of
Cathedral Pines in Connecticut, a severe windstorm in 1989
snapped off or uprooted 90% of the old pine, starting a new
cycle of forest succession (Patterson and Foster 1990).
Nonetheless,these sites serve as an interesting example of early
successional, old-field white pine that formed old-growth
forests, given adequate time, and demonstrate the potential
effect of windstorms on the dynamics of white pine forest.
Even-age cohorts of white pine in old-growth
forests.The range of ages for individual trees of a species can
be broad or narrow. A narrow age range often indicates that
the trees became established following a marked disturbance
event. The age structure of these individuals is generally re-
ferred to as an even-age cohort, a condition common to
white pine in many forests.
November 2001 / Vol. 51 No. 11 · BioScience 973
Articles
Figure 3. Number of individuals recruiting into the 1.5-m
height category by species and decade in an old-growth
hemlock­white pine forest at Cook Forest Cathedral Pines in
northwest Pennsylvania. Figure adapted from Abrams and
Orwig (1996).
Greenbrier, West Virginia. The species compo-
sition,tree ages,diameter structure,and radial growth
were studied in a virgin white pine­mixed-oak for-
est,located on a silt loam,valley floor in the Monon-
gahela National Forest in southern West Virginia
(Abrams et al.1995).The patterns of white pine dy-
namics in a mixed-oak forest and the effect of fire
protection during the last half-century were also
evaluated. This uneven-age forest is dominated by
white pine (34%), red maple (24%), white oak (Q.
alba;13%),northern red oak (7%),and black oak (Q.
velutina; 6%).White oak and white pine represented
most of the oldest and largest trees in the forest,
whereas maple, beech, and hemlock were the
youngest and smallest (Figure 4). Maximum age
was 295 years for white oak and 231 years for white
pine.Recruitment of white oak trees was continuous
from 1700 to 1900. Most of the white pine recruit-
ment occurred in two pulses (1830­1850 and
1875­1885), both of which were associated with
peaks in radial growth of the white pine tree-ring
chronology (Figure 4). The peak recruitment of
northern red oak and black oak from 1880 to 1900
suggests the possible facilitation of these species by
white pine. After the cessation of pine and oak re-
cruitment in 1900, the abundance of maple, beech,
and hemlock increased, particularly for red maple.
The results of this study suggest that white pine re-
cruited as even-age cohorts after moderate-scale
disturbances (Figure 2c). Similarly, Hibbs (1982)
concluded that white pine persists in mixed hard-
wood forests in New England by group reproduction
in small and medium-size gaps. Moreover, white
pine in portions of the original forest on the Al-
legheny Plateau in northwest Pennsylvania was re-
stricted to small,defined areas,having reproduced in
even-age cohorts after disturbance (Lutz 1930b).
White pine and oak were probably maintained in the
Monongahela Forest during the 18th and 19th cen-
turies by periodic disturbance,in particular fire and
wind-throw. The lack of recruitment for these species after
1900 and the subsequent increase in later successional tree
species indicate the transitional nature of this forest in the ab-
sence of fire.Future canopy gaps formed in the forest will most
likely be filled with late successional tree species,including red
maple, sugar maple, beech, and hemlock, according to their
dominance of the overtopped and intermediate canopy classes
(Abrams et al. 1995).
Schall's Gap, Pennsylvania. A dendroecological study of
an old-growth hemlock forest at Schall's Gap in central Penn-
sylvania is another example of the disturbance-dependent na-
ture of white pine.The forest is 80%-dominated by hemlock,
and it has a small component of white pine and yellow birch.
The white pine trees in the forest are strictly even-age, hav-
ing recruited in a narrow period between 1855 and 1865
(Figure 5). Most of the hemlock was recruited during and
shortly after this time.This finding corresponds to a major dis-
turbance in the forest, which increased the hemlock tree-
ring growth by 200% shortly after 1855. However, a distur-
bance of similar magnitude around 1810 apparently did not
stimulate a great deal of pine or hemlock recruitment,based
on current age structure.The total lack of recruitment for any
tree species after 1890 is most likely due to intensive deer
browsing, which is characteristic of central Pennsylvania.
Selective logging in old-growth forests. A number of
studies reported the formation of even-age cohorts of tree
species, including white pine, after selective logging in old-
growth forests. For example, partial logging around 1855 in
an old-growth hemlock­white pine­hardwoods forest on
the Allegheny Plateau in northwest Pennsylvania stimulated
a major recruitment period for white pine,red oak,beech,and
hemlock (Orwig and Abrams 1999). A similar finding was
974 BioScience · November 2001 / Vol. 51 No. 11
Articles
Figure 4. Age-diameter relationships for all cored trees and the mean
standardized tree ring index for the oldest Quercus alba and Pinus
strobus in an old-growth forest in eastern West Virginia. PIST, Pinus
strobus; QUAL, Quercus alba; QURU, Q. rubra; QUVE, Q. velutina;
ACRU, Acer rubrum; ACSA, A. saccharum; TSCA, Tsuga canadensis;
FAGR, Fagus grandifolia; CAGL, Carya glabra; NYSY, Nyssa sylvatica;
QUPR, Q. prinus. Figure adapted from Abrams et al. (1995).
reported for the old-growth forests at Bear Meadows Bog in
central Pennsylvania,in which selective logging around 1890
stimulated the establishment and growth of early,middle,and
late successional tree species,including white pine,red maple,
yellow birch, black spruce (P. mariana), and hemlock in the
subsequent half-century (Abrams et al. 2001). Selective and
more complete logging in old-growth forests at the Alan
Seeger NaturalArea in central Pennsylvania in the early 1800s
greatly increased even-age cohorts of oak, birch, white pine,
red maple,and hemlock (Nowacki and Abrams 1994).These
studies suggest that tree species representing both gap op-
portunist,including white pine,and late successional species
exhibit large increases in recruitment in canopy openings
created by selective logging.
Uneven-age white pine in old-growth hemlock:
Ice Glen Natural Area, Massachusetts. The dynam-
ics of an old-growth hemlock­white pine­northern hardwood
forest on the extreme slope (65%) at Ice Glen Natural Area
in the Berkshire Hills of southwest Massachusetts was stud-
ied (Figure 2e;Abrams et al.2000).The forest grows on a stony,
silt loam soil and is dominated by hemlock (60%), white
pine (12%), and red maple (10%). Hemlock is the oldest
species in the forest, with maximum ages of 305­321 years
(Figure 6).The maximum ages for white pine and several hard-
wood species are 170­200 years old.Recruitment of hemlock
trees was continuous from 1677 to 1948. All of the existing
white pine recruited from 1800 to 1880, when it formed an
uneven-age population within the uneven-age hemlock
canopy. This recruitment seemed to be facilitated by a series
of moderate disturbances to the stand in the early to mid-1800s
(Figure 6). Nearly all of the hardwood species also recruited
between 1800 and 1880, but after 1900 recruitment for all
species declined dramatically.
The apparent cessation of white pine recruitment after
1880 at Ice Glen may be attributable to the relative intolerance
of this species to old-growth understory conditions, includ-
ing low light, organic seedbed, and high root competition
(Cline and Spurr 1942, Abrams et al. 1995). The tree-ring
chronology indicates the absence of major standwide dis-
turbances between 1860 and 1945, which may have retarded
recruitment of white pine in favor of the later successional
hemlock.A lack of tree regeneration and limited overstory re-
November 2001 / Vol. 51 No. 11 · BioScience 975
Articles
Figure 5. Age-diameter relationships for all cored trees
and the mean standardized tree ring index for Tsuga
canadensis in an old-growth forest at Schall's Gap in
central Pennsylvania. Figure from Black and Abrams,
School of Forest Resources, Pennsylvania State
University.
Figure 6. Age-diameter relationships for all cored trees
and the mean standardized tree ring index for the oldest
20 Tsuga canadensis and 18 Pinus strobus in an old-
growth hemlock­white pine­northern hardwood forest
at the Ice Glen Natural Area, southwest Massachusetts.
Figure adapted from Abrams et al. (2000).
cruitment during the 20th century in this forest may be
attributable to intensive deer browsing, as observed in many
of these old-growth forests.
Scattered individuals of white pine in old-growth
oak. Most forests are a mixture of dominant and subordi-
nate species. In addition, many forests contain just a few
scattered individuals of a species that appear at random mo-
ments in the history of a forest. White pine is known to ex-
ist as scattered individuals in forests for many reasons, in-
cluding severe edaphic or ecological constraints.
Detweiler Run,Pennsylvania. The ecological history of a
326-year-old forest on a steep talus slope was studied in cen-
tral Pennsylvania (Ruffner and Abrams 1998). The forest is
dominated by chestnut oak (39%), red maple (17%), black
birch (13%),black gum (Nyssa sylvatica; 12%),red oak (9%),
and white pine (7%). The witness tree record and wood re-
mains on the site indicate that chestnut used to be an im-
portant component of this forest. The forest had fairly con-
tinuous recruitment of chestnut oak from 1660 to 1950
(Figure 7).Most of the red maple and red oak recruitment oc-
curred in the early 1900s and was probably associated with
chestnut blight mortality. White pine exists as scattered in-
dividuals of ages ranging from 30 to 245 years. Three of the
most recent recruits at the site are white pine, which may be
in response to gypsy moth defoliation and the resulting se-
lective mortality of the oak overstory. White pine is likely to
find opportunities for gap capture and will persist as scattered
individuals into the foreseeable future,but it is unlikely to be-
come a dominant component on this talus slope forest.
Whitepineintheoakunderstory. Substantial white pine
regeneration can form under an oak canopy throughout
much of the northern forest.White pine saplings had a den-
sity of 80­130 stems per ha on transitional mesic and
dry­mesic red oak sites in north-centralWisconsin (Nowacki
et al. 1990). White pine was the principal sapling species,
with 117 stems per ha, in mixed-oak, upland slope forest at
theAlan Seeger NaturalArea in central Pennsylvania (Nowacki
and Abrams 1994).According to Henry David Thoreau's ob-
servations of forest succession in Concord,Massachusetts,in
the 19th century, white pine invaded the understory of oak
forests and oak invaded the understory of white pine forests,
allowing for reciprocal replacement of the two species (Whit-
ney and Davis 1986). Hibbs (1982) concluded that white
pine could survive and grow through the understory of
mixed-hardwood forests in southern New England by group
reproduction. Moreover, in mixed-hardwood forests, white
pine has the advantage of being an evergreen among decid-
uous species: It may utilize favorable growth conditions in the
nonsummer months when the hardwoods are leafless (cf.
Lassoie et al.1983) and compete with some tree species of sim-
ilar or lesser understory tolerance (e.g.,oak,cherry [Prunus],
and birch). Nonetheless, for white pine trees to persist over
long time periods and to move into the oak forest canopy,they
probably need to be released in gaps because of the likelihood
that the white pine's shade tolerance will diminish with age.
White pine as a climax type. The combination
hemlock­white pine­northern hardwoods is listed as a climax
association in eastern North America by Braun (1950). She
questioned the climax status of white pine for much of the
range, but she did suggest that it is a climax dominant in the
westernmost part of the area,where hemlock is absent.Indeed,
a virgin,postfire forest on a lakeshore in west-central Ontario
had white pine present in all age classes up to 190 years (Holla
and Knowles 1988). Seedlings younger than 10 years had a
63.5% mortality rate per decade. By the time white pine
reached the age of 40 years,its mortality dropped to less than
6%. A fire of moderate intensity in the early 1900s con-
tributed to an increase in white pine in that age class. Simi-
larly, white pine was reported to be self-replacing, based on
its abundance across all diameter and canopy classes, in old-
growth forests on steep, rugged escarpments in east-central
Ontario (Quinby 1991).The recruitment of white pine in these
forests was most likely facilitated by periodic surface fire,
wind-throw, and death of larger individuals. The results of
these studies suggest that white pine near its northern limit
is self-maintaining,but it certainly benefits from disturbance
events. In addition, white pine was reported to be a physio-
graphic climax on ridge tops in New England (Cline and
Spurr 1942). However, on better-quality sites outside the
northern and western limits of its range, white pine is not a
climax species in the traditional sense because of its low-to-
moderate shade tolerance.Its recruitment tends to be episodic
and linked to disturbance, especially in old-growth forests.
976 BioScience · November 2001 / Vol. 51 No. 11
Articles
Figure 7. Age-diameter relationships for all cored trees in
an old-growth chestnut oak forest on a talus slope at
Detweiler Run in central Pennsylvania. QUPR, Quercus
prinus; PIST, Pinus strobus; BELE, Betula lenta; PIRI, P.
rigida; ACRU, Acer rubrum; QURU, Q. rubra; NYSY,
Nyssa sylvatica. Figure adapted from Ruffner and
Abrams (1998).
Role of fire and wind. Most studies dealing with the
ecology of white pine in the original forest discuss the im-
portance of fire or wind or both in the life history of the
species. White pine is known to be particularly sensitive to
wind-throw because of its tall stature, often growing as an
emergent in eastern forests,and its shallow rooting (Cline and
Spurr 1942, Foster 1988b, Patterson and Foster 1990, Foster
and Boose 1992).Moreover,ice and snow load may cause fre-
quent canopy pruning and top snapping of white pine.
Nichols (1935) concluded that the origin of white pine in the
climax forest characteristically dates back to a forest fire or
some other calamity.Fire and wind were considered to be com-
mon disturbances in the virgin pine­hemlock forests of
southwest New Hampshire (Cline and Spurr 1942): Old-
growth white pine forests at the Pisgah tract in southwestern
New Hampshire formed after a fire in the 1660s and were lev-
eled by a hurricane in 1938; another forest in New England--
Cathedral Pine, Connecticut--was blown down by severe
winds in 1989 (Henry and Swan 1974, Foster 1988a, Patter-
son and Foster 1990). Moreover, white pine needs a combi-
nation of both extensive wind-throw and fire for successful
establishment; wind-throw by itself may not be adequate
(Foster 1988a). Indeed, much of the white pine in the origi-
nal forests of northwest Pennsylvania can be traced back to
major fires and tornadoes (Lutz 1930b, Hough and Forbes
1943).
Evidence of fire or blow-down is consistently noted in the
early land surveys of northern forests containing white pine
(Lutz 1930a, Seischab and Orwig 1991, Marks and Gardescu
1992,Zhang et al.1999).In the presettlement forest of north-
eastern Wisconsin, complete canopy wind-throw averaged
4828 ha annually, with an estimated return interval of 1210
years (Canham and Loucks 1984).In northern-lower Michi-
gan, presettlement forests of mixed-pine, pine­oak, and
hemlock­white pine­northern hardwoods burned at inter-
vals of 129­258 years, 172­342 years, and 1389­2778 years,
respectively (Whitney 1987). However, virgin mixed-pine
forests in Minnesota burned at intervals of 10­40 years (Spurr
1954, Heinselman 1973).
Researchers believe that infrequent fire about every 100­200
years maintained oak, white pine, chestnut, and hickory
(Carya) on upland sites in the precolonial forests of central
New England (Foster et al. 1998). White pine and red pine
forests in northern Minnesota had a natural fire rotation of
about 100 years (Heinselman 1973), whereas white pine
forests along a historic travel corridor in southern Ontario
burned on average every 15 years between 1721 and 1937 (Dey
and Guyette 1996). Paleoecological data also suggest a link-
age between fire and white pine perpetuation across the
northeastern United States (Clark and Royall 1996).
Conclusion
A review of the witness trees and dendroecological data from
old-growth forests indicates a high level of versatility,or eco-
logical breadth,for eastern white pine in the pre-European set-
tlement forest. White pine was widely distributed in north-
ern forests of the eastern United States and grew in densities
ranging from nearly pure stands to scattered individuals. At
the regional level, its dominance probably did not match
that of beech, sugar maple, hemlock, or oak. In present-day
old-growth forests, white pine is much more likely to grow
with hemlock or oak than with beech or sugar maple. This
compatibility may also be true for the presettlement forest,but
witness tree data are generally not discernible at that fine a
scale. White pine occupied a wide range of sites, including
swamps, river valleys, sideslopes, sand plains, and dry, rocky
ridges.White pine is also capable of growing on nutrient-poor
soils and may be outcompeted by sugar maple and beech on
nutrient-rich mesic sites (Burns and Honkala 1990). Thus,
white pine grew in some of the wetter and drier nutrient-poor
sites in the northern forest, resulting in a quasi-bimodal dis-
tribution.
Fire and blow-down are two common disturbing agents as-
sociated with white pine in eastern forests.White pine is one
of the most susceptible trees to wind-throw because of its
height and shallow rooting (Foster and Boose 1992). Its fo-
liage is flammable in both green and dry, brown conditions.
Forests dominated by white pine are susceptible to both
crown and understory fires, depending on composition,
structure, stand age, and climate conditions. Large blow-
down areas of pine and hemlock were prone to subsequent
burning, a set of conditions that can facilitate a new genera-
tion of white pine (Cline and Spurr 1942, Henry and Swan
1974, Foster 1988a). Because young white pine has a thin
bark and older individuals have a very thick bark,its tolerance
to understory burning will range from low to high, depend-
ing on tree age and size.
White pine recruited in a wide range of ecological condi-
tions in old-growth forests. This article has examined exam-
ples of white pine that dominated the early and middle stages
of succession,forming even-age or uneven-age cohorts or scat-
tered individuals after small to moderate-scale disturbances
in uneven-age hemlock and oak canopies, and were self-
maintaining in nearly pure stands at the northern limit of its
range. This variety of successional pathways gave white pine
plenty of opportunities to thrive in the presettlement forest.
Once established in the overstory, white pine can live up to
300 years (and occasionally up to 400 years), during which
time it can produce large quantities of seed for regenerating
in surrounding disturbed areas, both within and outside the
immediate stand. Presently, many of the younger white pine
stands in eastern forests are the result of the trees'invasion of
old fields after agricultural abandonment.
Even though white pine is often a component of late-
successional forests,it is not a climax species in the traditional
sense. Rather, it is classified as a disturbance-dependent
species (Lutz 1930b,Cline and Spurr 1942,Hough and Forbes
1943).In most forests,the amount and duration of white pine
recruitment is often directly proportional to the intensity of
disturbance. Typical climax species, such as hemlock, sugar
maple, and beech, exhibit continuous, long-term recruit-
ment into an uneven-age forest canopy without periodic
November 2001 / Vol. 51 No. 11 · BioScience 977
Articles
disturbance.White pine does not have this ability,with the pos-
sible exception of stands with extreme latitude or edaphic con-
ditions; however, even these stands are typically subjected to
disturbance.Dendroecological studies indicate that the pres-
ence of white pine in late-successional hemlock forests is re-
lated to its being a long-lived species that established after large-
scale disturbance or to its recruitment in the gap phase.
Moreover, white pine can be successful in most subclimax
forests and disturbed areas across a broad range of soils and
sites. However, white pine and oak will most likely be re-
placed by later successional species in the absence of periodic
understory fire, as happened at Greenbrier, West Virginia
(Abrams et al. 1995).
White pine was only occasionally listed as a dominant in
witness tree records. A major limitation of white pine in the
original hemlock­white pine­hardwood forest was the rela-
tively low frequency of fire and catastrophic blow-down in
these northern forests.Return intervals of 800­2000 years for
large-scale disturbance were not adequate to sustain contin-
uous, large-scale white pine recruitment, and they probably
limited its presence to small and medium-size gaps in mature
forests (Lorimer 1977, Canham and Loucks 1984, Whitney
1987, Clark and Royall 1996). In contrast, hemlock, sugar
maple, and beech prospered in these forests without cata-
strophic disturbance. Large, persistent areas of white pine in
the original forests may have been mainly limited to coarse-
textured,fire-prone areas in the Lake States and New England,
which were not conducive to late-successional, mesophytic,
nutrient-demanding,or fire-sensitive species (Whitney 1994,
Cogbill 2000).
Very little of the original eastern forests remain because of
intensive cutting over the last two or three centuries. The
few extant tracts of presettlement-origin white pine are reach-
ing their maximum longevity. Most comparisons of preset-
tlement versus present-day forest composition indicate an
over-all decline in white pine dominance in eastern forests
(Whitney 1987, Abrams and Ruffner 1995, Frelich 1995,
Abrams and McCay 1996, Foster et al. 1998, Zhang et al.
2000). Some areas saw an increase in white pine because of
its invasion of abandoned farmlands. However, the losses in
white pine from the original forests are not being fully com-
pensated for by its increase in abandoned pastures or agri-
cultural areas.
The studies reviewed here indicate that most old-growth
hemlock­white pine forests have exhibited very little re-
cruitment of new trees over the past 50­100 years, probably
because of intensive deer browsing and lack of fire.Large in-
creases in red maple may have further exacerbated the decline
in white pine and are likely to continue into the foreseeable
future (Abrams 1998). Therefore, these stands are being se-
riously affected,both directly and indirectly,by anthropogenic
factors. If current forest management practices--less cut-
ting and less fire--and high deer pressure continue,the result
may well be a continued decline in white pine.
The approach taken in this study--that is,coupling witness
tree records with the dendroecology of old-growth forests and
land-use history--can be applied to other species and re-
gions in North America. Moreover, the study of old-growth
forests in the United States should have high priority, for
ecologists and tree-ring scientists have much to learn from the
past,and they must do so before the pre-European settlement
information contained in the oldest trees is lost.
Acknowledgments
Thanks go to Bryan Black, who reviewed an earlier draft of
this article.
References cited
Abrams MD. 1998. The red maple paradox. BioScience 48: 355­364.
Abrams MD, McCay DM. 1996. Vegetation-site relationships of witness
trees (1780­1856) in the presettlement forests of eastern West Virginia.
Canadian Journal of Forest Research 26: 217­224.
Abrams MD,Orwig DA.1996.A 300-year history of disturbance and canopy
recruitment for co-occurring white pine and hemlock on the Allegheny
Plateau, USA. Journal of Ecology 84: 353­363.
Abrams MD, Ruffner CM. 1995. Physiographic analysis of witness tree dis-
tribution (1765­1798) and present forest cover through north-central
Pennsylvania. Canadian Journal of Forest Research 25: 659­668.
Abrams MD, Orwig DA, DeMeo TE. 1995. Dendroecological analysis of
successional dynamics for a presettlement-origin white pine­mixed oak
forest in the southernAppalachians,USA.Journal of Ecology 83:123­133.
Abrams MD,van de Gevel S,Dodson RC,Copenheaver CA.2000.The den-
droecology and climatic impacts for old-growth white pine and hemlock
on the extreme slopes of the Berkshire Hills,Massachusetts,USA.Cana-
dian Journal of Botany 78: 851­861.
Abrams MD,Copenheaver CA,Black BA,van de Gevel S.2001.Dendroecology
and climatic impacts for a relict,old-growth,bog forest in the Ridge and
Valley Province of central Pennsylvania, USA. Canadian Journal of
Botany 79: 58­69.
Braun EL. 1950. Deciduous Forests of Eastern North America. New York:
Hafner Press.
Burns RM,Honkala BH.1990.Silvics of NorthAmerica,Vol.1: Conifers,Vol.
2: Hardwoods. Washington (DC): USDA. USDA Agricultural Hand-
book 654.
Canham CD,Loucks OL.1984.Catastrophic windthrow in the presettlement
forests of Wisconsin. Ecology 65: 803­809.
Canham CD,BerkowitzAR,KellyVR,Lovett GM,Ollinger SV,Schnurr J.1996.
Biomass allocation and multiple resource limitation in tree seedlings.
Canadian Journal of Forest Research 26: 1521­1530.
Clark JS, Royall D. 1996. Local and regional sediment charcoal evidence for
fire regimes in presettlement north-eastern North America. Journal of
Ecology 84: 365­382.
Cline AC,Spurr SH.1942.The virgin upland forest of central New England.
Harvard Forest Bulletin 21: 1­58.
Cogbill C.2000.Vegetation of the presettlement forest of northern New Eng-
land and New York. Rhodora 102: 250­276.
DandoW.1996.Reconstruction of presettlement forests of northeastern Penn-
sylvania using original land survey records. Master's thesis. Pennsylva-
nia State University, University Park, PA.
Deelen TR, Pregitzer KS, Haufler JB. 1996. A comparison of presettlement
and present-day forests in two northern Michigan deer yards.American
Midland Naturalist 135: 181­194.
Dey DC,Guyette RP.1996.Fire history near an historic travel corridor in On-
tario. Sault St. Marie (Canada): Ontario Ministry of Natural Resources.
Forest Research Report No. 140.
Foster DR. 1988a. Disturbance history, community organization and vege-
tation dynamics of the old-growth Pisgah Forest, southwestern New
Hampshire, USA. Journal of Ecology 76: 105­134.
------. 1988b. Species and stand response to catastrophic wind in central
New England, USA. Journal of Ecology 76: 135­151.
978 BioScience · November 2001 / Vol. 51 No. 11
Articles
Foster DR, Boose ER. 1992. Patterns of forest damage resulting from cata-
strophic wind in central New England, USA. Journal of Ecology 80:
79­98.
Foster DR,Motzkin G,Slater B.1998.Land-use history as long-term broad-
scale disturbance: Regional forest dynamics in central New England.
Ecosystems 1: 96­119.
Frelich LE.1995.Old forest in the Lake States today and before European set-
tlement. Natural Areas Journal 15: 157­167.
Glitzenstein JC,Canham CD,McDonnell MJ,Streng DR.1990.Effects of en-
vironment and land-use history on upland forests of the Cary Arbore-
tum,HudsonValley,NewYork.Bulletin of the Torrey Botanical Club 117:
106­122.
Goodlett JC. 1954.Vegetation adjacent to the border of the Wisconsin Drift
in Potter County, Pennsylvania. Harvard Forest Bulletin No. 25: 93.
HarlowWM,Harrar ES,White FM.1979.Textbook of Dendrology.NewYork:
McGraw-Hill.
Heinselman MA.1973.Fire in the virgin forests of the Boundary Waters Ca-
noe Area, Minnesota. Quaternary Research 3: 329­382.
Henry JD, Swan JMA. 1974. Reconstructing forest history from live and
dead plant material--An approach to the study of forest succession in
southwest New Hampshire. Ecology 55: 772­783.
Hibbs DE.1982.White pine in the transition hardwood forest.Canadian Jour-
nal of Botany 60: 2046­2053.
Holla TA, Knowles P. 1988. Age structure of a virgin white pine, Pinus
strobus, population. Canadian Field-Naturalist 102: 221­226.
Hough AF, Forbes RD. 1943. The ecology and silvics of forests in the high
plateau of Pennsylvania. Ecological Monographs 13: 300­320.
Kilburn PD.1960.Effects of logging and fire on xerophytic forests in north-
ern Michigan. Bulletin of the Torrey Botanical Club 87: 42­45.
Lassoie JP,Dougherty PM,Reich PB,Hinckley TM,Metcalf CM,Dina SJ.1983.
Ecophysiological investigations of understory eastern red cedar in cen-
tral Missouri. Ecology 64: 1355­1366.
Lorimer CG. 1977. The presettlement forest and natural disturbance cycle
of northeastern Maine. Ecology 58: 139­148.
------.1983.A test of the accuracy of shade-tolerance classifications based
on physiognomic and reproductive traits. Canadian Journal of Botany
61: 1595­1598.
Lutz HJ. 1930a. Original forest composition in northwestern Pennsylvania
as indicated by early land survey notes.Journal of Forestry.28:1098­1103.
------. 1930b. The vegetation of Heart's Content, a virgin forest in north-
west Pennsylvania. Ecology 11: 1­29.
Marks PL, Gardescu S. 1992. Vegetation of the central Finger Lakes Region
of NewYork in the 1790s.NewYork State Museum Bulletin No.484:1­33.
McIntosh RP. 1962. The forest cover of the Catskill Mountain Region, New
York, as indicated by land survey records. American Midland Natural-
ist 68: 409­423.
Nichols GE. 1935. The hemlock­white pine­northern hardwood region of
eastern North America. Ecology 16: 403­422.
Nowacki GJ,Abrams MD. 1992. Community, edaphic and historical analy-
sis of mixed oak forests of the Ridge and Valley Province in central
Pennsylvania. Canadian Journal of Forest Research 22: 790­800.
------. 1994. Forest composition, structure, and disturbance of the Alan
Seeger Natural Area,Huntingdon County,Pennsylvania.Bulletin of the
Torrey Botanical Club 121: 277­291.
Nowacki GJ, Abrams MD, Lorimer CG. 1990. Composition, structure, and
historical development of northern red oak stands along an edaphic gra-
dient in northcentral Wisconsin. Forest Science 36: 276­292.
Orwig DA, Abrams MD. 1999. Impacts of early selective logging on the
dendroecology of an old-growth,bottomland hemlock­white pine­north-
ern hardwood forest on the Allegheny Plateau. Journal of the Torrey
Botanical Society 126: 234­244.
Pacala SW,Canham CD,Saponara J,Silander JA,Kobe RK,Ribbens E.1996.
Forest models defined by field measurements: Estimation, error analy-
sis and dynamics. Ecological Monographs 66: 1­43.
Palik BJ,Pregitzer KS.1992.A comparison of presettlement and present-day
forests on two bigtooth aspen-dominated landscapes in northern lower
Michigan. American Midland Naturalist 127: 327­338.
Patterson WA III,Foster DR.1990.Tabernacle pines.Journal of Forestry 88:
23­25.
Patterson WA III, Sassaman KE. 1988. Indian fires in the prehistory of New
England. Pages 107­135 in Nicholas GP, ed. Holocene Human Ecology
in Northeastern North America. New York: Plenum Press.
Pregitzer KS, Barnes BV. 1984. Classification and comparison of upland
hardwood and conifer ecosystems of the Cyrus H. McCormick Experi-
mental Forest,upper Michigan.Canadian Journal of Forest Research 14:
362­375.
Quinby PA. 1991. Self-replacement in old-growth white pine forests of
Temagami, Ontario. Forest Ecology and Management 41: 95­109.
Ruffner CM,Abrams MD.1998.Relating land-use history and climate to the
dendroecology of a 326-year-old Quercus prinus talus slope forest.Cana-
dian Journal of Forest Research 28: 347­358.
Seischab FK.1990.Presettlement forests of the Phelps and Gorham purchase
in western New York. Bulletin of the Torrey Botanical Club 117: 27­38.
Seischab FK,Orwig DA.1991.Catastrophic disturbances in the presettlement
forests of western New York. Bulletin of the Torrey Botanical Club 118:
117­122.
Siccama TG. 1971. Presettlement and present forest vegetation in northern
Vermont with special reference to Chittenden County. American Mid-
land Naturalist 85: 153­172.
Spurr SH. 1954. The forests of Itasca in the nineteenth century as related to
fire. Ecology 35: 21­25.
Ward RT. 1956. The beech forests of Wisconsin--Changes in forest com-
position and the nature of the beech border. Ecology 37: 407­419.
Whitney GG. 1986. Relation of Michigan's presettlement pine forests to
substrate and disturbance history. Ecology 67: 1548­1559.
------. 1987 An ecological history of the Great Lakes forest of Michigan.
Journal of Ecology 75: 667­684.
------. 1990. The history and status of the hemlock­hardwood forests of
the Allegheny Plateau. Journal of Ecology 78: 443­458.
------. 1994. From Coastal Wilderness to Fruited Plain. Cambridge (UK):
Cambridge University Press.
Whitney GG, Davis WC. 1986. Thoreau and the forest history of Concord,
Massachusetts. Journal of Forest History 30: 70­81.
Winer HI. 1955. History of the Great Mountain Forest, Litchfield County,
Connecticut. Master's thesis.Yale University, New Haven, CT.
Zhang Q,Pregitzer KS,Reed DR.1999.Catastrophic disturbance in the pre-
settlement forests of the upper peninsula of Michigan. Canadian Jour-
nal of Forest Research 29: 106­114.
------.2000.Historical changes in the forests of the Luce District of the up-
per peninsula of Michigan. American Midland Naturalist 143: 94­110.
November 2001 / Vol. 51 No. 11 · BioScience 979
Articles