Forests Too Deer: Edge Effects in
Northern Wisconsin
WILLIAM S. ALVERSON*
Botany Department
Birge Hall
University of Wisconsin
Madison, Wisconsin 53706, U.S.A.
DONALD M. WALLERt
Botany Department
Birge Hall
University of Wisconsin
Madison, WI 53706, U.S.A.
STEPHEN L. SOLHEIM
Botany Department
Birge Hall
University of Wisconsin
Madison, WI 53706, U.S.A.
Abstract: Browsing by white-tailed deer (Odocoileus virginianus)
can profoundly affect the abundance and population
structure of several woody and herbaceous plant species.
Enclosure studies and population surveys reveal that
past and current deer densities as low as 4 deer/km2 may
prevent regeneration of the once common woody species,
Canada yew (Taxus canadensis), eastern hemlock (Tsuga canadensis),
and white cedar (Thuja occidentalis), as well as
several herbaceous species. Prior to European settlement, forests
in northern Wisconsin contained relatively sparse deer
populations (<41km2), but extensive timber cutting in the
late nineteenth century boosted deerpopulations. Continued
habitat fragmentation resulting from scattered timber harvests
and the creation of "wildlife openings" to improve deer
forage maintain these high densities throughout much of the
Northeast
Because deer wander widely, the effects of high deer densities
penetrate deeply into remaining stands of old and mature
forest, greatly modifying their composition. Thus, abundant
early successional and "edge" habitat, and the high deer
densities they engender, represent significant external threats
* Requests for reprints should be addressed to this author.
t This author presented this paper at the first annual meeting of the
Society for Conservation Biology, Bozeman, Montana June 25,
1987.
Resumen: El pastoreo y ramoneo del venado de cola blanca
(Odocoileus virginianus) en la zona nortena del estado de
Wisconsin, puede afectarprofundamente a la abundancia y
estructura de poblacion de varias especies de plantas herbdceas
y lefiosas. Antes de la Coloniag los bosques nortefi
de Wisconsin eran habitat de poblaciones relativamente pequefias
de venados (menos de 4 por Km2) pero las extensas
talas de estos bosques, a finales del siglo 19, dio paso a un
incremento poblacional de estas especies. Estudios recientes
de unidades de exclusion y catastros poblacionales revelan
que, en efecto, una densidad de solo 4 ejemplos por Km2
logra impedir la regeneracion de las especies de Taxus canadensis,
Tsuga candensis y Thuja occidentalis, muy conocidas
otrorag como tambien varias especies herbaceas. Lafragmentacion
de habitats, que ha resultado de la tala dispersa de
arboles y la creacion, de "claros silvestres" con propositos de
incrementar el forraje para los venados, ha aumentado las
poblaciones de estas especies a lo largo de la region del
nordeste.
Debido a que un gran nutmero de venados vagan ampliamente
en la region, los efectos sobre los espacios remanentes
de bosque maduro son considerables modificandose
significativamente su composicion. De esta manera, incipientes
y extensos ecotonos, ademas de las grandes densidades
de poblaciones de venados que estos albergan, representan
una amenaza externa sobre estas comunidades de bosque.
348
Conservation Biology
Volume 2, No. 4, December 1988
This content downloaded from 147.251.87.220 on Tue, 03 Dec 2019 14:19:27 UTC
All use subject to https://about.jstor.org/terms
Alverson et al. Deer Edge Effects 349
to these plant communities. We hypothesize that establishing
large (200-400 km2) continuous areas of maturing forest,
especially in conjunction with increased hunting could reduce
local deer densities and so provide a simple and inexpensive
method for retaining species sensitive to the deleterious
effects of browsing.
[M]obile animals greatly affect plant life, so that a small
virgin forest may appear to be natural when actually it
has been profoundly affected by forces applied to animals,
waters, or climate at points far distant. (Thus the
deer populations determined by laws passed at Lansing,
by hunters ..., and by lumbering operations ... have
apparently exterminated the ground hemlock [yew]
from the "virgin" forest of Mountain Lake.)
A. Leopold (1938)
Ever since the pioneering work of Leopold (1936) on
habitat manipulation, wildlife biologists have strived to
boost populations of game species by creating clearings
and other areas of sharp transition between two or more
types of plant community. Indeed, the traditional meaning
of the term "edge effect" was the local increase in
the diversity and abundance of animal species found
along the boundary between two habitat types (Leopold
1936; Swift 1946; Dahlberg & Guettinger 1956; Yahner
1988). In seeking to explain this phenomenon, a number
of other edge effects have been noted, including
microclimatic changes in temperature, light, and humidity;
altered tree species composition due to increased
colonization by shade-intolerant and exotic plants; invasions
by insects; and increased parasitism, predation,
and competition by "weedy" birds and mammals (Ranney,
Bruner, & Levenson 1981; Matthiae & Stearns
1981; Guntenspergen 1983; Brittingham & Temple
1983; Wilcove 1985; Janzen 1983, 1986; Wilcove,
McLellan, & Dobson 1986; Yahner & Scott 1988).
Because the apparently beneficial effects of habitat
tend to be local (on the order of a few hundred meters
at most), wildlife managers often try to establish small
clearings throughout a forested area. These efforts, the
simultaneous creation of edge via other ongoing human
disturbances, and controls on hunting have resulted in
abundant populations of game and other edge-loving
species. As reflected in the leading quote, however,
some naturalists recognize a darker side to edges.
Within conservation biology, the term "edge effects" is
now usually used to refer to increased predation and
parasitism of vulnerable animals in the vicinity of edges
(e.g., Temple & Cary 1988). We would like to extend
this connotation to include the deleterious effects of
herbivores on sensitive plant species within stands of
mature forest. While younger forest can undoubtedly
buffer older forests against many microclimatic and biological
edge effects (as assumed for western U.S. forests
by Harris 1984), such a matrix might also threaten diNos
permitimos suponer que el establecimiento de grandesy
continuas acreas de bosques en desarrollo (200-400 Kms),
conjuntamente con el incremento en la actividad de cazag
podria reducir localmente la densidad de poblacion de los
venados y de esta forma disponerse de un metodo simple y
poco costoso quepermita conservar estas expecies sujetas al
deterioro por efectos del pastoreo y ramoneo.
versity by facilitating the invasion of successional plants
and animals capable of interfering with species restricted
to older communities (Janzen 1983, 1986).
Here, we review evidence that herbivory can profoundly
alter plant community composition in the
Northeast. We became concerned that such efforts
could be widespread after learning of Hough's (1965)
20-year field study in the Allegheny National Forest in
Pennsylvania, where he found the understory of a large
(1650 ha) tract of virgin hemlock-hardwood forest to be
severely damaged by deer browsing. Because herbivorous
mammals wander widely and can invade even areas
of ostensibly unfavorable habitat, such edge effects penetrate
much farther than those previously reported for
the region. This raises the important policy question of
how plant species diversity is to be retained in areas
subject to regional management for high deer densities.
For concreteness and relevance, we primarily discuss
interactions between white-tailed deer (Odocoileus virginianus)
and various plants in northern Wisconsin
(scientific names in Table 1). Policy decisions are now
pending regarding this issue in two National Forests located
there (Task Force 1986; U.S. Forest Service 1986).
The Context: Northern Wisconsin
Land survey records and detailed analyses of remnant
forest stands allow a reasonable reconstruction of presettlement
forest conditions (Curtis 1959; Finley 1976).
Upland mesic forest habitats were predominantly oldgrowth
(200-300 years old), with only 17-25% of their
area occupied by successional communities (Canham &
Loucks 1984). This relationship has now been reversed,
with small patches of old and mature growth occupying
less than 5% of the forest within a matrix of younger
successional communities. The pre-Columbian forest
contained about 64% "hardwood" by area and consisted
mostly of the hemlock-hardwood community type. Intact,
mature examples of this forest type are now relegated
to token occurrence, primarily in existing or proposed
"Research Natural Areas" 12 to 260 ha in size.
Because white pine, once a major component of
northern Wisconsin forests, was preferred by early loggers,
its abundance was drastically reduced. Hemlocks
were then cut preferentially to service the tanning
trade. Finally, with increasing demand for hardwood
Conservation Biology
Volume 2, No. 4, December 1988
This content downloaded from 147.251.87.220 on Tue, 03 Dec 2019 14:19:27 UTC
All use subject to https://about.jstor.org/terms
350 Deer Edge Effects Alverson et al.
Table 1. Common and scientific names of organisms mentioned
in the text.
Aspen Populus tremuloides Michx.
Black ash Fraxinus nigra Marsh.
Blunt-leaved orchid Habenaria obtusata (Pursh)
Richards.
Buckthorn Rhamnus cathartica L.
Flowering dogwood Cornusflorida L.
Hemlock, eastern Tsuga canadensis (L.) Carr.
hemlock
Honeysuckle Lonicera tatarica L., L. morrowi
Gray, and their hybrid, L. X
bella Zabel
Indian cucumber-root Medeola virginiana L.
Large-flowered trillium Trillium grandiflorum (Michx.)
Salisb.
Leatherwood Dirca palustris L.
Purple fringed orchid Habenaria psycodes (L.) Spreng.
Redbud Cercis canadensis L.
Showy lady's-slipper Cypripedium reginae Walt.
orchid
Sugar maple Acer saccharum Marsh.
Tall northern bog Habenaria hyperborea (L.) R. Br.
orchid
Yellow birch Betula alleghaniensis Britton
Yew, Canada yew, Taxus canadensis Marsh.
ground hemlock
Yew-tree Taxus baccata L.
White cedar Thuja occidentalis L.
White oak Quercus alba L.
White pine Pinus strobus L.
Wood sorrel Oxalis acetosella L.
Yellow lady's-slipper Cypripedium calceolus L.
orchid
Brainworm Parelaphostrongylus tenuis
Canadian lynx Lynx canadensis Kerr
Deer, white-tailed deer Odocoileus virginianus
(Zimmerman)
Deer tick Ixodes
Elk, American elk Cervus elaphus Linnaeus
Moose Alces alces (Linneaus)
Mountain lion Felis concolor Linnaeus
Timber wolf Canis lupus Linnaeus
Wolverine Gulo gulo (Linnaeus)
Woodland caribou Rangifer tarandus (Linnaeus)
during and after World War I, most of the area was
clearcut. Wisconsin's two National Forests, the Chequamegon
and Nicolet, were created in the 1920s and
30s and now occupy 3,420 and 2,650 kM2, respectively
(Fig. 1).
Aspen is the preeminent early successional tree species
of the region. Its wind-dispersed seeds, clonal propagation,
and fast growth allow it to quickly occupy large
areas. Partly because it freely root-sprouts following fire
or cutting, it has increased from about 1 % on these
National Forests presettlement to about 26% now (U.S.
Forest Service 1986c). Other disturbance and edgeadapted
species were originally rather scarce and limited
to tree-fall gaps, riparian habitats, and areas of forest
recently blown down or burned. They are now quite
0 KM 300
SUPERIOR 1
CHEQUAm
ANISTEE FINGER
OF LAKES
Figure 1. The US. National forests of the Great
Lakes region Stippled areas represent the Great
Lakes and solid lines the boundaries between
states. The Chequamegon and Nicolet National
Forests of northern Wisconsin are 3,420 and 2,650
km2, respectively. Source: Modified from U.S.D.A.
Forest Service map.
common throughout these forests. Many weedy, exotic
plant species have colonized heavily disturbed habitats,
but the less disturbed habitats have not yet been seriously
invaded (unlike the situation in southern Wisconsin,
where buckthorn and honeysuckle have invaded
even "intact" forests [Barnes & Cottam, 1974]).
Not surprisingly, major changes in Wisconsin's fauna
have also occurred during the last century. Moose, elk,
and woodland caribou, as well as predators like the wolverine,
have all been extirpated. Forest disturbance does
not fully explain why these species were lost, but is
certainly a contributing factor (Jackson 1961; Gates,
Clarke, & Harris 1983). Timber wolf and Canadian lynx
continue to occupy sections of both forests but are
quite scarce, largely due to human activity. Although it
was assumed to be extirpated, there have been several
recent sightings of the mountain lion (Lewis & Craven
1987).
Population Densities of Deer
Severe winters and wide expanses of virgin timber lacking
undergrowth originally produced marginal habitat
for the white-tailed deer in the northern Great Lakes
region (Swift 1946; Schorger 1953; Dahlberg &
Guettinger 1956, Blouch 1984; but see Habeck & Curtis
1959). As stated by the U.S. Forest Service, "Species
associated with aspen and other early successional
stages were present but in low numbers. Early settler's
notes indicate few deer and other game animals in
northern Wisconsin" (1986c, p. D5).
Deer populations in northern Wisconsin were originally
less than 4/kM2 of range, and probably as low as
Conservation Biology
Volume 2, No. 4, December 1988
This content downloaded from 147.251.87.220 on Tue, 03 Dec 2019 14:19:27 UTC
All use subject to https://about.jstor.org/terms
Alverson et al. Deer Edge Effects 351
2/km2 over large areas of range (total surface area minus
the area of lakes, rivers, urban areas, and large farms).
Populations began to rise in the mid-1850s and peaked
in the Forest in the 1930s and 1940s at about 14/kM2
due to extensive favorable habitat and protective hunting
laws (Swift 1946; Dahlberg & Guettinger 1956; McCaffery
1986). During the last 25 years, densities in
northern Wisconsin have ranged from 5 to 12/kM2 (McCaffery
1986) and are now estimated at 2 to 9 deer/kM2
in the northern units of Wisconsin's National Forests (F.
Haberland, personal communication, data for 1985 and
1986). The stated goals of the Forest Service for deer
production are much higher, calling for sufficient habitat
to support 31,952 deer in the Chequamegon, or 9.3
deer/km2 (U.S. Forest Service 1986a, p. B4). Goals of the
Wisconsin Department of Natural Resources are for 4 to
8 deer/kM2 overwintering in the same area (F. Haberland,
personal communication).
Habitat management for deer in the Upper Great
Lakes region entails establishing winter range, young
aspen growth, oaks for acorns, and openings in the forest
to supply grasses and other pasturage (McCaffery
1984, 1986). In keeping with these traditional management
practices, the management plans for the national
forests intersperse small (<20 ha) timber cuts and
openings designed to boost deer and other game populations
throughout both Forests (U.S. Forest Service
1986). Collectively, these comprise at least 14% of the
Forests' areas and should result in a uniform, abundant
distribution of deer, a goal explicitly embraced by the
Forest Service (Task Force 1986).
Effects of Deer Browsing on Woody Plants
The damage deer do to crop and natural vegetation has
been extensively studied, and depends on deer density.
High deer populations slow the regeneration of several
commercial species, causing significant economic losses
(Graham 1954; Marquis 1981; Redding 1987). However,
as deer densities in northern Wisconsin have declined
to levels below that which threatens commercial
forestry or deer range per se, concern for deer damage
to vegetation has virtually disappeared (McCaffery
1986).
Partial lists of preferred deer foods in Wisconsin all
agree that Canada yew, eastern hemlock, and white cedar
are highly preferred by deer during winter months
(DeBoer 1947; Swift 1948; Cottam & Curtis 1956; Dahlberg
& Guettinger 1956; Beals, Cottam, & Vogl 1960).
Yew is severely damaged by deer because it is both
sought out and does not recover well after browsing.
For some time, there has been little or no reproduction
of yew in most of the region (Stearns 1951; Curtis 1959)
with many populations now lost from known sites of
prior occurrence. Surviving populations exist on rocky
outcrops that are inaccessible to deer, or as scattered
individuals browsed nearly to ground level. The only
other area in Wisconsin where yew populations are
known to be extensive and healthy is in the tribal lands
of the Menominee Reservation (Waller et al., in prep.)
where year-round hunting limits deer densities (see
below). On the nearby Apostle Islands in Lake Superior,
Beals, Cottam, & Vogl (1960) found yew common on
islands with few or no deer, yet yew were lacking in
mainland forests. Leuthold (1980) has shown a similar
decline of the related yew-tree in Switzerland and predicts
extirpation of the species unless active protection
occurs. Besides direct browsing losses, Canada yew suffers
indirectly via a novel mechanism that has only recently
been recognized: browsing skews floral sex ratios
which, in turn, limit the availability of pollen to the
point where it becomes limiting and reproduction is
impaired (Allison 1987).
Like yew, eastern hemlock and white cedar are quite
sensitive to deer browsing. Although these trees are capable
of growing tall enough to escape browsing, deer
can severely impair reproduction by preventing seedling
and sapling recruitment, particularly since slowgrowing
seedlings and saplings of these species are vulnerable
to browsing for decades (Hough 1965; Rogers
1978). Browsing is particularly conspicuous within winter
deer yards in Wisconsin where hemlock and white
cedar are reproducing poorly or not at all (Task Force
1986).
Deer enclosure studies carried out in Wisconsin,
through other parts of the Northeast, and elsewhere
show dramatic differences in survival and reproduction
of hemlock, yew, white cedar, and other species within
fenced enclosures compared to exposed individuals
outside the protected areas (Graham 1954; Dahlberg &
Guettinger 1956; Stoeckeler, Strothman, & Krefting
1957; Marquis 1974; Blewett 1976; Kroll, Goodrum, &
Behrman 1986; Tilghman, in press; Fig. 2). These observations
all support the results of Goff (1967), Anderson
& Loucks (1979), and Waller et al. (in prep.), in
which hemlock only exhibits a healthy population structure
within Wisconsin on certain islands, in the Menominee
Reservation, and within deer enclosures. Similar
results hold for white cedar (Blewett 1976 and references
therein), with striking differences in stem height
and density within and outside enclosures.
The enclosure illustrated contains a population of
hemlock with hundreds of individuals representing all
seedling and sapling age classes, yet the surrounding
area outside the enclosure contains only a few individuals,
which either show signs of recent browsing or are
shorter than the winter snow cover. Nearby, a showcase
grove of old-growth hemlocks is virtually devoid of recent
hemlock reproduction, despite falling within a deer
management unit with a reported population density of
only 2 deer/km2. Its understory is composed mainly of
Conservation Biology
Volume 2, No. 4, December 1988
This content downloaded from 147.251.87.220 on Tue, 03 Dec 2019 14:19:27 UTC
All use subject to https://about.jstor.org/terms
352 Deer Edge Effects Alverson et al.
.A,
x
Figure 2. Deer enclosure at Fould's Creek; Chequamegon
National Forest, Wisconsin, viewed
from the top of the 4 m high fence, which bisects
the photograph. Vigorous growth of hemlock can
be seen within the forty-year-old enclosure (left
side of photograph). Source: W. S. Alverson, February,
1988.
stunted, gnarled sugar maple seedlings and saplings
bearing the distinctive mark of heavy browsing by deer
(cf. Fig. 1 in Switzenberg, Nelson, & Jenkins 1985; Fig. 1
in Stoeckeler, Strothman, & Krefting 1957).
Some researchers question whether deer browsing
alone has caused the conspicuous changes in hemlock
reproduction in the upper Midwest (Webb, King, & Patric
1956; Tierson, Patric, & Behrend 1966). Stearns
(1951) suggested that changes in climate or catastrophic
storms allowed greater hemlock regeneration
during certain periods in the past (reviewed by Eckstein
1980). This seems unlikely, however, both from the enclosure
studies cited above and because cycles of hemlock
reproduction are asynchronous between noncontiguous
stands and appear to be governed by internal
stand dynamics (Hett & Loucks 1976).
Even more definitively, Frelich & Lorimer (1985)
documented changes in size-class distributions and extensive
browse damage to young hemlock that appear
directly attributable to deer browsing in the Porcupine
Mountains of Michigan's Upper Peninsula. A forest near
the Lake Superior shore with an estimated deer density
of 10/kmn2 (winter) suffered almost complete annihilation
in some size classes, while inland sites with lower
winter deer densities (2/km2) exhibited unimpaired reproduction.
They rejected the hypothesis that climate
was responsible for the differences in hemlock reproduction
by demonstrating that herbivory was the causal
factor. A model they constructed predicts eventual exclusion
of hemlock by hardwood species in the coastal
sites within 200 years if deer densities remain high.
In a study designed specificaily to test whether high
deer densities prevent reproduction in these species,
we compared hemlock's population structure within
the Menominee Reservation to its structure within the
adjacent Nicolet National Forest (Wailer et al., in prep.).
Because the Menominee Reservation allows hunting
year-round, deer densities are lower than in surrounding
areas (ca. 1-21km2; Morehouse & Becker 1966; 0.
Rongstad, personal communication). While almost half
of the stands within the Menominee showed substantial
hemlock reproduction, less than 6% of those in the Forest
did. The Nicolet stands exhibit drastically reduced
seedling abundance, especially relative to the number of
adult trees (Fig. 3).
Other Species Affected by High Deer Densities
Deer can affect the composition of entire communities
and not just individual woody plant species. For example,
in Great Smoky Mountains National Park, areas subject
to intensive deer browsing close to openings lost
a) 1.2
W- MENOMINEE
[ NICOLET
w
0.
06.6 -
0.4-
0.2
< 15 15-24 25-49 50-74 75-99 100-149
HEIGHT (cm)
b) 6
5-
3 -
(1)
0-
1-5 6-10 11-20 21-40 41-60 61-80
DIAMETER (cm)
Figure 3. Size distributions of eastern hemlock
seedlings (a) and adult trees (b) in the Menominee
Reservation and the adjacent Nicolet National
Forest Deer densities are much lower in the
Menominee Reservation due to year-round hunting.
Source: Data from Waller, Judziewicz, Alverson,
& Solheim (in prep.).
Conservation Biology
Volume 2, No. 4, December 1988
This content downloaded from 147.251.87.220 on Tue, 03 Dec 2019 14:19:27 UTC
All use subject to https://about.jstor.org/terms
Alverson et al. Deer Edge Effects 353
more than a quarter of their total species richness compared
to control areas (Bratton 1979). White oak, redbud,
and flowering dogwood all appeared to be significantly
affected by deer browsing. Not just plant species
are affected, either. Moose are thought to be excluded
from many areas by infection with brainworm, a parasite
carried by deer. Decreases in deer abundance could enhance
the chances for immigrant moose now drifting
into northern Wisconsin to become successfully reestablished,
as recently occurred in New York (Hicks &
Stumvoll 1985). Lyme disease, carried primarily by deer
ticks, also appears to be increasing in many areas in
response to increased deer abundance, but no effects on
animal communities are yet evident.
Many herbaceous species are also favored by deer,
including the showy and yellow lady's slipper orchids,
the blunt-leaf orchid, the tall northern bog orchid, and
the purple fringed orchid. Many of these could be experiencing
reduced reproductive success and/or local
extirpation due to intensified deer herbivory (Cottam &
Curtis 1956; L. Lipsey, personal communication; personal
observation). Enclosure studies in the Allegheny
National Forest of northwestern Pennsylvania demonstrated
that deer populations of 4/kM2 caused significant
reductions in the abundance of Indian cucumber-root
and large-flowered trillium, both of which also occur in
northern Wisconsin (Tilghman, in press). Like yew,
herbs are highly susceptible to herbivory by deer because
they never outgrow the zone of accessibility (approximately
2 m).
Changes in canopy composition could also result in
changes in community composition in other forest
strata. For example, observed and predicted losses of
mature hemlock and other northern hardwood canopies
in the Forest could cause the eventual loss of shrubs and
herbs like leatherwood and wood sorrel that tend to be
restricted to these habitats (Stearns 1951; personal observation).
As evergreen conifers lose dominance to deciduous
species like sugar maple and black ash, light
regimes change drastically and understory species can
be expected to respond. Furthermore, as remaining habitats
of older hemlock and white cedar become smaller
and more isolated, further gradual but inexorable losses
of the many species restricted to these habitats are expected
via "relaxation" (MacArthur & Wilson 1967).
How quickly this occurs obviously depends on speciesspecific
characteristics, but analogous losses of herb
species have been documented for isolated forests in
southern Wisconsin (Hoehne 1981).
Recommended Deer Densities
Assuming that deer browsing has caused the observed
reductions in reproduction of hemlock, yew, and white
cedar, it becomes important to determine that density
of deer below which successful regeneration is possible.
While few studies directly addressing this question exist,
existing work allows us to infer which densities are
clearly incompatible with successful reproduction. For
example, studies of deer carrying capacity routinely suggest
that densities of 8 deer/km2 are compatible with
good range management. However, studies of carrying
capacity are normally undertaken out of primary concern
for healthy populations of deer or commercial timber
species, especially their ability to fulfill their auxiliary
role as "cover" or "browse" for deer. Such studies
provide little or no assurance that the impact of deer is
uniformly benign. For example, Tubbs, Jacobs, & Cutler
( 1983) combine data for hemlock with numerous other
species composing the "northern hardwood types," and
obscure the problematic relationship with deer: "The
northern hardwood types can support relatively high
populations of deer without serious injury; damage will
be minimal if management practices favor dense reproduction
and vigorous shoot growth (Jacobs 1969)" (p.
122). Yet Jacobs considers only the ability of sugar maple
to survive under such deer densities, not hemlock,
yellow birch, or yew, all important components of the
northern hardwood forest. Furthermore, it is sugar maple
that replaces hemlock in this region as the latter is
browsed (Anderson & Loucks 1979; Frelich & Lorimer
1985).
Densities of 8/km2 appear far too high if maintaining
the diversity of all plants and animals is the management
objective, as reviewed above. Instead, deer densities approximating
presettlement conditions for substantial periods
of time appear necessary to ensure the survival
and healthy reproduction of hemlock, yew, and other
sensitive plant species. Existing meager data suggest this
density to be less than 4 deer/kMn2, and possibly as low
as 1-2 deer/km2. Precise figures cannot yet be stated
because of the lack of thorough, species-specific studies
in our region. A wildlife biologist currently studying
deer movements in the area suggests that young hemlock
occur in areas where deer densities approach
2/km2 (0. Rongstad, personal communication).
Discussion
Browsing by elevated populations of white-tailed deer
appears to constitute a major edge effect in the forests of
northern Wisconsin and perhaps other parts of the
Northeast. Deer affect forest composition through direct,
well-documented negative effects on several
woody plant species and through direct and indirect
damage to many herbaceous species. Failure to acknowledge
these ecological interactions and plans to
maintain dense populations of deer (8-9.3/km2) by
state and federal land stewards work directly against the
preservation of these components of natural diversity.
Conservation Biology
Volume 2, No. 4, December 1988
This content downloaded from 147.251.87.220 on Tue, 03 Dec 2019 14:19:27 UTC
All use subject to https://about.jstor.org/terms
354 Deer Edge Effects Alverson et al.
What steps could be taken to protect the viability of
species sensitive to browsing at high deer densities?
Habitats suitable for deer-sensitive species could be created
in at least three ways: enclosures, increased hunting,
or habitat management to reduce deer densities.
Other means of controlling deer density exist (such as
deer repellents and birth control), but these are unlikely
to be viable solutions (Redding 1987; Marquis 1987).
Enclosures are now being used in northwestern Pennsylvania
for regenerating commercial forest stands (Marquis
1987), but are extremely expensive to construct
and maintain (Kochel & Brenneman 1987; U.S. Forest
Service 1986c, p. F63). Alternatively, commercial seedling
caps can be used to protect individual seedlings,
also at great cost. Both enclosures and seedling protectors
appear best suited to regenerate small local stands
of a single target species such as hemlock or yew. However,
such a solution, unless extended to a complete set
of other sensitive plant species (many of which must
still be unknown), offers no general relief. The same
problem applies to the use of silvicultural techniques
aimed at regenerating single species (Eckstein 1980;
Johnson & Booker 1983; Tubbs, Jacobs, & Cutler 1983;
Wendel et al. 1983; Marquis 1987). Such methods for
hemlock require soil scarification, removal of litter,
fencing and/or partial canopy removal while reducing
the area's attractiveness to deer. These intrusive management
techniques are prohibitively expensive on a
large scale and could still cause or permit damage to
other species sharing the habitat. It would also be an
obvious mistake to assume that protecting hemlock (or
any other particular species) somehow protects the
overall diversity of the hemlock-hardwood forest community.
At present, there exists neither the knowledge
nor the will to create active programs of species-specific
management for all deer-sensitive species in these com-
munities.
Increased hunting pressure can also decrease deer
populations locally (Morehouse & Becker 1966; Creed
et al. 1984). While some uncertainty exists as to the
relative importance of hunting versus deer behavior and
habitat quality in determining deer population levels
(McCaffery 1986), few doubt that increased hunting
pressure would reduce browsing, especially if coupled
with habitat alteration. Whether hunting alone could
reduce deer densities to 2/km2 is unclear, however, particularly
since most hunters prefer to hunt in areas of
known high deer density.
Species sensitive to high deer densities could also be
protected by habitat management if vegetation capable
of supporting only reduced deer densities could be established.
This would involve running conventional
game management practices in reverse. Instead of increasing
edge habitat and young browse, large blocks of
forest would be allowed to mature naturally to the point
where they become inferior deer habitat (Fig. 4). Such
0)
Z
GS 00
Figure 4. Theoretical model of the influence of
habitat on the population densities of white-tailed
deer in northern Wisconsin. Units of deer density,
originally given in deer/mi2, have been converted
to deer/km2; density figures are comparable to
those given in the text. One purpose of Diversity
Maintenance Areas would be to create large
blocks of habitat corresponding to the lowermost
surface of the slope. Source: Modified with permission
from McCaffery (1986).
areas would have to be large and continuous enough to
create core areas relatively free of the edge effects produced
by deer and could be created by redistributing
management activities in public (and perhaps private)
forests. For example, efforts to harvest timber and improve
deer habitat could be confined to 80% of each
National Forest, leaving 20% in one or two large, contiguous
blocks that would eventually become, through
natural succession, habitats unfavorable to dense deer
populations. This, in fact, represents the actual recommendation
made in a formal appeal process involving
Wisconsin's National Forests (Task Force 1986). The
proposed biotic reserves were termed "Diversity Maintenance
Areas."
The crucial question concerning this final alternative
lies with size: How large must a block of old forest be to
effectively reduce deer densities? The literature on deer
movements is extensive but insufficient by itself to resolve
the size issue (e.g., Tierson et al. 1985). Winter
ranges of individual deer tend to be less than 480 ha in
Minnesota (Rongstad & Tester 1969), with summer
home ranges somewhat larger. Winter to summer range
movements for adult deer averaged 5.6 km in a Wisconsin
study, with 90% of the deer moving 12 km or less
(Dahilbert & Guettinger, 1956). An eight-year study in
Conservation Biology
Volume 2, No. 4, December 1988
This content downloaded from 147.251.87.220 on Tue, 03 Dec 2019 14:19:27 UTC
All use subject to https://about.jstor.org/terms
Alverson et al. Deer Edge Efects 355
Michigan's Upper Peninsula found that the mean annual
dispersal distance between winter yards and the following
November kill site was 13.8 km for hunter-killed
deer; yearly mean distances ranged from 10.9 to 20.2
km (Verme 1973). Bratton (1979) concluded that intensive
deer impacts did not extend beyond 1 km away
from openings in the Cades Cove region of Great Smoky
Mountains National Park. Fortunately, a study on deer
movements within the Forest is now underway (O.
Rongstad, personal communication).
If we apply the average travel distance for deer of 8
km used by the Forest Service in its management plan
(Task Force 1986), blocks of unfavorable habitat with
radii of 8 km, constituting areas of ca. 200 km2, might
serve to halve deer densities at their center points. If
unfavorable habitat like old growth is found to reduce
travel distances, smaller areas might suffice. Prudence
would dictate the conservative course of first designating
larger areas, then reducing them if penetrations
were found to be of shorter range. If the mean dispersal
distance were reduced to 2 km in a circular block of
unfavorable habitat, the block would still need to have a
radius of 7 km to have a 1:1 ratio of edge to interior
habitat. Historical patterns of movement and other particular
features of deer behavior also clearly influence
how such habitat blocks would function (Tierson et al.
1985). Such information is limited, making it difficult to
predict a priori exactly how large the blocks of unfavorable
habitat need to be to protect sensitive species.
Mature or old-growth forest blocks of this scale are
much larger than any existing old-growth areas in northern
Wisconsin. The two congressionally designated wilderness
areas in the Chequamegon National Forest are
1,710 and 2,660 ha in size, comparable in size to
Hough's (1965) 1,650 ha severely damaged study area in
Pennsylvania. These wilderness areas are imbedded in a
matrix of young forest containing extensive deer habitat.
As more forest lands in northern Wisconsin mature,
deer densities should decline slowly (McCaffery 1986).
However, the National Forest plans call for increased
timber harvests, which all convert nearly 50% of their
area into new successional habitats during the next 50
years. This makes it unlikely that the wilderness areas
will experience consistently low herbivory by deer over
the foreseeable future, even when the wilderness areas
themselves become old. In fact, the scattered, shifting
pattern of timber harvests and the creation of additional
'"ildlife openings" proposed by the Forest Service instead
promotes a Forest-wide homogenization of habitat
via deer edge effects.
Conclusions
Our understanding of edge effects is still in its infancy.
Edge effects on the scale of several kmn resemble those
already suggested for other forests (Janzen 1983, 1986),
but remain politically controversial due to their management
implications. Studies are now under way that
should eventually allow us to tailor the size and shape of
reserves specifically to retain a full complement of plant
and animal species. Such areas, if they prove unnecessarily
large, can always be reduced in size at a later time,
but cannot be expanded without losing decades of forest
growth.
Maintaining the proposed 200 to 400 km2 reserves of
contiguous habitat within the National Forests to retain
species sensitive to deer browsing or otherwise dependent
on forest interior habitats Would be simple and
inexpensive (Task Force 1986). Such areas would also
be freely available for a wide variety of other uses, including
hunting (intensified for deer, if possible), fishing,
snowmobiling, skiing, hiking, camping, and smallscale
wood removal. They would, however, exclude
commercial-scale timber harvests, "wildlife openings,"
and new road construction, all of which create large
amounts of edge habitat. Such management would probably
not reduce the total deer populations of the area,
but would alter the spatial distribution of deer, allowing
local reductions in deer abundance and consequent survival
of sensitive species. Encouragingly, the staff of the
Chequamegon National Forest concluded that such areas
could be created without losing jobs or sacrificing
timber production or other outputs of their Forest (J.
Wolter, personal communication 1986). Disappointingly,
the regional office of the U.S. Forest Service, perhaps
concerned about the precedent such areas would
establish, reversed this decision without any formal scientific
or economic review. Currently these issues are
being considered by the chief of the U.S. Forest Service.
Acknowledgments
We thank 0. Rongstad, C. Lorimer, L. Tyrrell, J. Kotar, E.
Judziewicz, K McCaffery, F. Haberland, and D. Wilcove
for their comments on drafts of this paper or for making
unpublished data available. The E.K and O.N. Allen
Fund of the University of Wisconsin Herbarium provided
funds for publication.
Literature Cited
Allison, T. D. 1987. The reproductive biology of Canada yew
(Taxus canadensis) and its modification by herbivory: Implications
for wind pollination. Ph.D. dissertation, University of
Minnesota, Minneapolis.
Anderson, R. C., and 0. L. Loucks. 1979. White-tail deer
(Odocoileus virginianus) influence on the structure and
composition of Tsuga canadensis forests. Journal of Applied
Ecology 16:855-861.
Conservation Biology
Volume 2, No. 4, December 1988
This content downloaded from 147.251.87.220 on Tue, 03 Dec 2019 14:19:27 UTC
All use subject to https://about.jstor.org/terms
356 Deer Edge Effects Alverson et al.
Barnes, W.J., and G. Cottam. 1974. Some autecological studies
of the Lonicera x bella complex. Ecology 55:40-50.
Beals, E. W., G. Cottam, and R.J. Vogl. 1960. Influence of deer
on the vegetation of the Apostle Islands, Wisconsin. Journal of
Wildlife Management 24(1):68-80.
Blewett, T.J. 1976. Structure and dynamics of the McDougal
Springs lowland forest. M.S. thesis, University of Wisconsin,
Madison.
Blouch, R. I. 1984. [Deer in the] Northern Great Lakes States
and Ontario forests. Pages 391-410 in L. K Halls, editor. The
White-tailed Deer: Ecology and Management. Stackpole Books,
Harrisburg, Pennsylvania.
Bratton, S. P. 1979. Impacts of white-tailed deer on the vegetation
of Cades Cove, Great Smoky Mountains National Park.
Proceedings of the Annual Conference of the Southeastern
Association of Fish & Wildlife Agencies 33:305-312.
Brittingham, M. C., and S.A. Temple. 1983. Have cowbirds
caused forest songbirds to decline? Bioscience 33:31-35.
Canham, C. D., and 0. L. Loucks. 1984. Catastrophic
windthrow of the presettlement forests of Wisconsin. Ecology
65(3):803-809.
Cottam, G., and J. T. Curtis. 1956. The effect on deer of the
botanical composition of deer yards and a method for measuring
it. Unpublished interim report on project 15-933 to
Wisconsin Conservation Department of Natural Resources.
Botany Department, University of Wisconsin, Madison.
Creed, W. A., F. Haberland, B. E. Kohn, and K R. McCaffery.
1984. Harvest management: The Wisconsin experience. Pages
243-260 in L. K Halls, editor. The White-tailed Deer: Ecology
and Management. Stackpole Books, Harrisburg Pennsylva-
nia
Curtis, J. T. 1959. The Vegetation of Wisconsin. University of
Wisconsin Press, Madison. 657 p.
Dahlberg, B. L., and R. C. Guettinger. 1956. The White-tailed
Deer in Wisconsin. Wisconsin Conservation Department,
Madison.
DeBoer, S. G. 1947. The deer damage to forest reproduction
survey. Wisconsin Conservation Bulletin 12(10):1-23.
Eckstein, R. G. 1980. Eastern hemlock (Tsuga canadensis) in
north-central Wisconsin. Research Report No. 104. Wisconsin
Department of Natural Resources, Rhinelander.
Finley, R. W. 1976. Original vegetation cover in Wisconsin.
North Central Forest Experiment Station, U.S.D.A. Forest Service,
St. Paul, Minnesota.
Frelich, L. E., and C. G. Lorimer. 1985. Current and predicted
long-term effects of deer browsing in hemlock forests in Michigan,
U.S.A. Biological Conservation 34:99-120.
Gates, D. M., C. H. D. Clarke, and J. T. Harris. 1983. Wildlife in
a changing environment. Pages 52-80 in S. L. Flader, editor.
The Great Lakes Forest: An Environmental and Social History.
University of Minnesota Press, Minneapolis.
Goff, F. G. 1967. Upland vegetation. Pages 60-90 in C. J. Milfred,
G. W. Olson, and F. D. Hole, editors. Soil Resources and
Forest Ecology of Menominee County, Wisconsin. University
of Wisconsin Geology and Natural History Survey Bulletin No.
85, Soil Series No. 60. University of Wisconsin Extension, Mad-
ison.
Graham, S. A. 1954. Changes in northern Michigan forests from
browsing by deer. Transactions of the Nineteenth North American
Wildlife Conference 19:526-533.
Guntenspergen, G. 1983. The minimum size for nature preserves:
Evidence from southern Wisconsin forests. Natural Areas
Journal 3(4):38-46.
Habeck, J. R., and J. T. Curtis. 1959. Forest cover and deer
population densities in early northern Wisconsin. Transactions
Wisconsin Academy Sciences, Arts and Letters 48:49-56.
Harris, L. D. 1984. The Fragmented Forest: Island Biogeography
Theory and the Preservation of Biotic Diversity. University
of Chicago Press.
Hett, J. M., and 0. L. Loucks. 1976. Age structure models of
balsam fir and eastern hemlock. Journal of Ecology
64(3):1029-1044.
Hicks, A., and R. Stumvoll. 1985. An old friend returns. The
Conservationist (New York) 39(4):9-13.
Hoehne, L. M. 1981. The groundlayer vegetation of forest islands
in an urban-suburban matrix. Pages 41-54 in R. L. Burgess
and D. M. Sharpe, editors. Forest Island Dynamics in Mandominated
Landscapes. Springer-Verlag, New York.
Hough, A. F. 1965. A twenty-year record of understory vegetational
change in a virgin Pennsylvania forest. Ecology
46(3):370-373.
Jackson, H. T. 1961. Mammals of Wisconsin. University of Wisconsin
Press, Madison.
Jacobs, R. D. 1969. Growth and development of deer-browsed
sugar maple seedlings. Journal of Forestry 67:870-874.
Janzen, D. H. 1983. No park in an island: Increased interference
from outside as park size decreases. Oikos 41:402-410.
Janzen, D. H. 1986. The eternal external threat. Pages 286-303
in M. E. Soule, editor. Conservation Biology. Sinauer Associates,
Sunderland, Massachusetts.
Johnson, W. F., and R. G. Booker. 1983. Northern white cedar.
Pages 105-108 in R. M. Burns, editor. Silvicultural Systems for
the Major Forest Types of the United States. U.S.D.A. Forest
Service Agriculture Handbook No. 445. U.S. Department of
Agriculture, Washington, D.C.
Kochel, J., and R. Brenneman. 1987. Use and effectiveness of
electric fencing in protecting clearcuts from deer browsing.
Pages 118-123 in Proceedings of the Symposium on Deer,
Forestry and Agriculture: Interactions and Strategies for Management.
Allegheny Society of American Foresters, Warren,
Pennsylvania.
Conservation Biology
Volume 2, No. 4, December 1988
This content downloaded from 147.251.87.220 on Tue, 03 Dec 2019 14:19:27 UTC
All use subject to https://about.jstor.org/terms
Alverson et al. Deer Edge Effects 357
Kroll, J. C., W. D. Goodrum, and P.J. Behrman. 1986. Twentyseven
years of over-browsing: Implications to white-tailed
deer management on wilderness areas. Pages 294-303 in D. L.
Kulhavy and R. N. Conner, editors. Wilderness and Natural
Areas in the Eastern United States: A Management Challenge.
Center for Applied Sciences, School of Forestry, Stephen F.
Austin University, Nacogdoches, Texas.
Leopold, A. 1936. Game Management. Charles Scribner's Sons,
New York.
Leopold, A. 1938. Private report to the Huron Mountain Club.
18 p.
Leuthold, C. von. 1980. The ecological and phytosociological
situation of the yew-tree in Switzerland. Veroffentlichungen
des geobotanischen Institutes ETH, Stiftung Rubel, Zurich:
67:1-217.
Lewis, T. L., and S. R. Craven. 1987. Mountain lions in Wisconsin?
Maybe. Wisconsin Natural Resources. 11(1):21-25.
MacArthur, R. H., and E. 0. Wilson. 1967. The Theory of Island
Biogeography. Princeton University Press, Princeton, NewJer-
sey.
Marquis, D. A. 1974. The impact of deer browsing on Allegheny
hardwood regeneration. Research Paper NE-308. U.S.D.A.
Forest Service, Northeastern Forest Experiment Station, Upper
Darby, Pennsylvania.
Marquis, D.A. 1981. Effect of deer browsing on timber production
in Allegheny hardwood forests in northwestern Pennsylvania.
Research Paper NE-475. U.S.D.A. Forest Service,
Northeastern Forest Experiment Station, Broomall, Pennsylva-
nia.
Marquis, D. A. 1987. Silvicultural techniques for circumventing
deer damage. Pages 125-136 in Proceedings of the Symposium
on Deer, Forestry and Agriculture: Interactions and
Strategies for Management. Allegheny Society of American Foresters,
Warren, Pennsylvania.
Matthiae, P. E., and F. Stearns. 1981. Mammals in forest islands
in southeastern Wisconsin. Pages 55-66 in R. L. Burgess and
D. M. Sharpe, editors. Forest Island Dynamics in Mandominated
Landscapes. Springer-Verlag, New York.
McCaffery, K 1984. Fat deer laugh at winter. Wisconsin Natural
Resources 8(6):17-19.
McCaffery, K 1986. On deer carrying capacity in northern
Wisconsin. Pages 54-69 in R. J. Regan and S. R. Darling, compilers.
Transactions of the Twenty-second Northeast Deer
Technical Committee. Vermont Fish and Wildlife Department,
Waterbury, Vermont.
Morehouse, M., and R.J. Becker. 1966. Menominee County
deer. Wisconsin Conservation Bulletin 31(4):20-21.
Ranney, J. W., M. C. Bruner, and J. B. Levenson. 1981. The importance
of edge in the structure and dynamics of forest islands.
Pages 67-96 in R. L. Burgess and D. M. Sharpe, editors.
Forest Island Dynamics in Man-dominated Landscapes.
Springer-Verlag, New York.
Redding, J. C. 1987. Impact of deer on forest vegetation and
timber production in northern Pennsylvania. Pages 23-32 in
Proceedings of the Symposium on Deer, Forestry and Agriculture:
Interactions and Strategies for Management. Allegheny
Society of American Foresters, Warren, Pennsylvania.
Rogers, R. S. 1978. Forests dominated by hemlock (Tsuga canadensis):
Distribution as related to site and postsettlement
history. Canadian Journal of Botany 56:843-854.
Rongstad, O.J., and J. R. Tester. 1969. Movements and habitat
use of white-tailed deer in Minnesota. Journal of Wildlife Management
33(2):366-379.
Schorger, A. W. 1953. The white-tailed deer in early Wisconsin.
Transactions of the Wisconsin Academy Sciences, Arts and
Letters 42:197-247.
Stearns, F. W. 1951. The composition of the sugar maplehemlock-yellow
birch association in northern Wisconsin.
Ecology 32:245-265.
Stoeckeler, J. H., R. 0. Strothman, and L. W. Krefting. 1957.
Effect of deer browsing on reproduction in the northern hardwood-hemlock
type in northeastern Wisconsin. Journal of
Wildlife Management 21:75-80.
Swift, E. 1946. A History of Wisconsin Deer. Bulletin No. 323,
Wisconsin Conservation Department, Madison.
Swift, E. 1948. Wisconsin Deer Damage to Forest Reproduction
Survey: Final Report. Publication 347, Wisconsin Conservation
Department, Madison.
Switzenberg, D. F., T. C. Nelson, and B. C. Jenkins. 1955. Effect
of deer browsing on quality of hardwood timber in northern
Wisconsin. Forest Science 1:61-67.
Task Force. 1986. Statement of Reasons in Appeal of the
U.S.D.A. Forest Service Record of Decision of August 11, 1986
Approving the Land and Resource Management Plan and Final
Environmental Impact Statement for the Chequamegon National
Forest. Sierra Club and Wisconsin Forest Conservation
Task Force, Madison, Wisconsin. (Contains several U.S.D.A.
Forest Service documents obtained under the Freedom of Information
Act. Copies available at cost from Walter Kuhlman,
of Boardman, Suhr, Curry, and Field. P.O. Box 927, Madison,
WI 53701-0927.)
Temple, S. A., and J. R. Carey. 1988. Modelling dynamics of
habitat interior bird populations in fragmented landscapes.
Conservation Biology 2 (this issue).
Tierson, W. C., E. F. Patric, and D. F. Behrend. 1966. Influence
of white-tailed deer on the logged northern hardwood forest.
Journal of Forestry 64:801-805.
Tierson, W. C., G. F. Mattfeld, R. W. Sage, Jr., and D. F. Behrend.
1985. Seasonal movements and home ranges of white-tailed
deer in the Adirondacks. Journal of Wildlife Management
49(3):760-769.
Tilghman, N. G. In Press. Impacts of several densities of whitetailed
deer on regeneration of forests in northwestern Pennsylvania.
Journal of Wildlife Management.
Conservation Biology
Volume 2, No. 4, December 1988
This content downloaded from 147.251.87.220 on Tue, 03 Dec 2019 14:19:27 UTC
All use subject to https://about.jstor.org/terms
358 Deer Edge Effects Alverson et !.
Tubbs, C. H., R. D. Jacobs, and D. Cutler. 1983. Northern hardwoods.
Pages 121-127 in R. M. Burns, editor. Silvicultural Systems
for the Major Forest Types of the United States. U.S.D.A.
Forest Service Agriculture Handbook No. 445, U.S. Department
of Agriculture, Washington, D.C.
U.S. Forest Service. 1986. a b, c, d Chequamegon National
Forest Management Plan and associated documents: a Land
and Resource Management Plan; b. Final Environmental Impact
Statement; c Appendix to Final Environmental Impact
Statement; d Record of Decision. Chequamegon National Forest
Headquarters, U.S.DA. Forest Service. Park Falls, Wiscon-
sin.
Verme, L.J. 1973. Movements of white-tailed deer in upper
Michigan. Journal of Wildlife Management 37(4):545-552.
Webb, W. L, R. T. King, and E. F. Patric. 1956. Effect of whitetailed
deer on a mature northern hardwood forest. Journal of
Forestry 54:391-398.
Wendel. G. W., L. Delia-Bianca, J. Russeli, and K F. Lancaster.
1983. Eastern white pine including eastern hemlock. Pages
131-134 in R. M. Burns, editor. Silvicultural Systems for the
Major Forest Types of the United States. U.S.DA. Forest Service
Agriculture Handbook No. 445. U.S. Department of Agriculture,
Washington, D.C.
Wilcove, D. S. 1985. Nest predation in forest tracts and the
decline of migratory songbirds. Ecology 66:1211-1214.
Wilcove, D. S., C. H. McLellan, and A. P. Dobson. 1986. Habitat
fragmentation in the temerate zone. Pages 237-256 in M. E.
Soule, editor. Conservation Biology. Sinauer Associates, Sunderland,
Massachusetts.
Yahner, R. H. 1988. Changes in wildlife communities near
edges. Conservation Biology 2 (this issue).
Yahner, R. H., and D. P. Scott. 1988. Effects of forest fragmentation
on depredation of artfficial nests. Journal of Wildlife
Management 52:158-161.
A
Conservation Biology
Volume 2, No. 4, December 1988
This content downloaded from 147.251.87.220 on Tue, 03 Dec 2019 14:19:27 UTC
All use subject to https://about.jstor.org/terms