Landscape Diversity: The Concept
Applied to Yellowstone Park
William H. Romme and Dennis H. Knight
Changes in landscape patterns may influence a variety of natural features including
wildlife abundance, nutrient flow, and lake productivity. Data suggest that cyclic
changes in landscape diversity occur on areas of 100 km2 in Yellowstone National
Park. When properly managed, large wilderness areas provide the best and probably
the only locale for studying the kind of landscape changes that occurred for
millenia in presettlement times. (Accepted for publication 12 May 1982)
Each successive level of biological organization
has properties that cannot be
predicted from those of less complex
levels (Odum 1971). Thus, populations
have certain attributes distinct from the
characteristics of the individuals of
which they are composed, and communities
have unique properties beyond the
attributes of their component populations.
An important level of organization
that is now receiving more attention is
the landscape, or mosaic of communities
that covers a large land unit such as a
watershed or a physiographic region
(Forman and Godron 1981).
The importance of large-scale landscape
patterns has been widely recognized
(e.g., Bormann and Likens 1979,
Forman 1979, 1982, Forman and Boerner
1981, Forman and Godron 1981, Habeck
1976, Habeck and Mutch 1973, Hansson
1977, Heinselman 1973, Loucks 1970,
Luder 1981, Pickett 1976, Reiners and
Lang 1979, Rowe 1961, Shugart and
West 1981, Sprugel 1976, Sprugel and
Bormann 1981, Swain 1980, White 1979,
Wright 1974, Zachrisson 1977, and others).
A few studies have quantitatively
treated changes in landscape patterns
(e.g., Hett 1971, Johnson 1977, Johnson
and Sharpe 1976, Shugart et al. 1973).
We recently made a detailed analysis of
landscape composition and diversity in a
pristine watershed in Yellowstone National
Park in relation to fire and forest
regrowth following fire (Romme 1982).
In this paper we describe the natural
changes that have occurred in landscape
pattern over a period of 240 years and
the possible consequences of these
changes for certain aspects of ecosystem
structure and function. Although we focus
on Yellowstone in this analysis, the
concepts are applicable to other ecosystems
as well.
The term landscape diversity refers to
the diversity of plant communities making
up the vegetational mosaic of a land
unit. Landscape diversity results from
two superimposed vegetation patterns:
the distribution of species along gradients
of limiting factors, and patterns of
disturbance and recovery within the
communities at each point along the environmental
gradients (Forman and Godron
1981, Reiners and Lang 1979). Both
of these patterns contribute to the vegetational
diversity of the Yellowstone
landscape.
Over the park's 9000 km2, elevation
ranges from about 1800 m along the
Yellowstone River in the northern portion
to over 3000 m on the high peaks of
the east and northwest. As a result, there
are pronounced gradients of temperature
and moisture, with related patterns in
species distribution. The areas at lower
elevations in the north support open
sagebrush (Artemisia tridentata) parks
on drier sites and aspen (Populus tremuloides)
woodlands and Douglas fir (Pseudotsuga
menziesii) forests in more mesic
locations (Despain 1973). On the cooler
subalpine plateaus one finds extensive
upland coniferous forests of lodgepole
pine (Pinus contorta var. latifolia), subalpine
fir (Abies lasiocarpa), Engelmann
spruce (Picea engelmannii), and whitebark
pine (P. albicaulis), broken by occasional
meadows and sagebrush parks
on alluvial and lacustrine soils. The high
peaks are covered by forests of spruce,
fir, and whitebark pine on sheltered
slopes, with alpine or subalpine meadows
and boulder fields on the more exposed
sites. Pollen analysis of pond sediments
indicates that these basic patterns
of species distribution have been relatively
stable during the last 5000 years
(Baker 1970).
However, vegetational patterns related
to the second source of landscape
diversity-perturbation-have undergone
changes during this time. Most of
the changes have been natural, as described
below, but some aspen and sagebrush
communities in northern Yellowstone
appear to have been altered
somewhat by fire suppression during the
last century. Comparisons of 100-yearold
photographs with recent photographs
of the same sites show that forests today
are generally more dense, with an increase
in conifers and a decrease in
aspen, and that many sagebrush parks
now contain more shrubs and fewer
grasses and forbs. Streamside thickets of
willow (Salix spp.) and alder (Alnus spp.)
also appear less extensive and robust
than formerly (Houston 1973).' Some
have attributed these changes to excessive
browsing by elk (Cervus elaphus)
(Beetle 1974, Peek et al. 1967).
A more common explanation appears
to be the virtual elimination of fire in this
area from 1886 to 1975. Houston (1973)
found that fires formerly recurred at average
intervals of 20-25 years in northern
Yellowstone, a disturbance frequency
that probably was essential for the
persistence of plant species and communities
representing early stages of secondary
succession (notably aspen and
herbaceous plants). In the absence of
fire, succession has proceeded unchecked
and other species such as Douglas
fir and sagebrush have become increasingly
predominant. Thus fire
Romme is with the Department of Natural Science,
Eastern Kentucky University, Richmond, KY
40475. Knight is with the Department of Botany,
University of Wyoming, Laramie, WY 82071.
? 1982 American Institute of Biological Sciences.
All rights reserved.
'Houston, D. B. 1976. The Northern Yellowstone
Elk. Parts III and IV. Vegetation and habitat relations.
Unpublished report, Yellowstone National
Park, Wyoming.
664 BioScience Vol. 32 No. 8
This content downloaded from 147.251.87.8 on Thu, 05 Dec 2019 09:44:21 UTC
All use subject to https://about.jstor.org/terms
prevention appears to have modified the
overall composition of the northern Yellowstone
landscape, reducing landscape
diversity by increasing the area covered
by late successional plant communities
at the expense of early successional
communities. The magnitude of this
change is relatively small in the context
of the entire northern Yellowstone landscape,
however, since aspen and herbaceous
communities comprised a small
fraction of the landscape even in presettlement
times (Despain 1973). Similar
changes have also been described in several
other western parks and wilderness
areas following effective fire control (Habeck
1976, Habeck and Mutch 1973,
Kilgore and Taylor 1979, Loope and
Gruell 1973, Lunan and Habeck 1973).
Because a major management goal in the
large national parks is to preserve ecosystems
in their primeval state (Houston
1971), Yellowstone recently instituted a
new fire management policy that allows
lightning-caused fires to burn without
interference if they do not threaten human
life, property, or other values (US
National Park Service 1975).
The situation seems to be different on
the high subalpine plateaus that dominate
most of the central, western, and
southern areas of the park. Because of
inaccessibility, effective fire control was
not accomplished here until about 1950
when fire-fighting equipment and techniques
were greatly improved (US National
Park Service 1975). Moreover, our
research indicated that fire occurs naturally
at very long intervals because of
very slow forest regrowth and fuel accumulation
after fire (Romme 1982). On an
average site, 200 years or more are required
for a fuel complex to develop that
is capable of supporting another destructive
fire. On dry or infertile sites, 300-
400 years may be necessary. Fires ignited
prior to that time are likely to burn a
very small area and have a minimal impact
on the vegetation (Despain and Sellers
1977, Romme 1982). Recent uncontrolled
fires in the park that burned
intensely in 300-year-old forests have
been observed to stop when they
reached a 100-year-old stand, even
though weather conditions remained favorable
for fire (Despain2, Despain and
Sellers 1977). Thus, in an ecosystem
where fire historically occurred at intervals
of 200+ years on any particular site,
suppression during the last 20-30 years
probably has had very little effect on
overall landscape pattern. Any major
changes that have occurred are largely
the result of natural processes that would
have taken place even in man's absence.
Although the subalpine landscape apparently
has not been substantially altered
by man's activities (excluding, of
course, those areas of intensive development
for visitor use), it has by no means
been static during the last 100 years. We
found evidence that major fires occur
cyclically, i.e., thousands of hectares
may burn at intervals of 300-400 years
with relatively few major fires in the
same area during the intervening periods
(Romme 1982). Such a fire cycle can
occur because: geologic substrate, soils,
and vegetation are very similar over
much of the plateau region; forests over
large contiguous areas grow and develop
a fuel complex at approximately the
same rates; and the plateau topography
has low relief and few natural barriers to
fire spread. Thus one extensive fire
tends to be followed by another fire in
the same area some 300-400 years later.
In other parts of the Rocky Mountains
where topographic barriers are more numerous,
where succession occurs more
rapidly, or where fuel characteristics are
different, this particular type of fire cycle
may not occur.
LITTLE FIREHOLE RIVER
WATERSHED
We conducted our study in the Little
Firehole River watershed, which covers
73 km2 on the Madison Plateau, a large
rhyolite lava flow in west-central Yellowstone.
Coniferous forests predominate,
with lodgepole pine occurring
throughout and subalpine fir, Engelmann
spruce, and whitebark pine being found
on more mesic sites. Alluvial deposits in
the central and northern parts of the
watershed support subalpine meadows
or open coniferous forests with rich
shrub and herbaceous understories. The
topography is generally flat or gently
sloping, with an average elevation of
about 2450 m.
Fire history during the last 350 years
was determined using the fire-scar methods
developed by Heinselman (1973) and
Arno and Sneck (1977). Major fires occurred
in 1739, 1755, and 1795 (? 5
years), collectively burning over half of
the upland area. Of the forested areas
that did not burn at that time, nearly all
were located either on topographically
protected sites (ravines, lower northeastfacing
slopes) that burn rarely (Romme
and Knight 1981, Zachrisson 1977), or in
places that had been burned by a moderately
large fire in 1630, less than 200
years earlier, and were covered by
young forests. Since 1795 only three fires
>4 ha have occurred, and all three were
relatively small (<100 ha). The absence
of recent large fires is almost certainly
due to a lack of suitable fuel conditions
over most of the watershed, not to fire
suppression by man. In fact, park records
show that only one fire has been
controlled in this area, a 90-ha burn in
1949. The fire probably would not have
covered a much larger area even without
suppression, since it was surrounded by
young forests and topographically sheltered
sites. Today the areas burned in
the 1700s support lodgepole pine forests
that are all developing more-or-less synchronously;
in another 100-150 years
extensive portions of the watershed will
again have fuel conditions suitable for a
large destructive fire.
Three stages of forest regrowth following
fire (early, middle, and late successional)
can be recognized on upland
sites. Early successional stages are usually
present for about the first 40 years
and are characterized by an abundant
growth of herbs and small shrubs. The
large dead stems of the former forest
remain standing throughout most of this
period, and an even-aged cohort of
lodgepole pine becomes established.
Middle successional stages are marked
by the maturation and dominance of the
even-aged pine cohort, beginning with
canopy closure around 40 years and lasting
until senescence around 250-300
years. Herbaceous biomass and species
diversity are lowest during this period
(Taylor 1973). During late successional
stages (250-300+ years) the even-aged
pine canopy deteriorates with heavy
mortality and is replaced by trees from
the developing understory to produce an
all-aged, usually mixed-species stand,
which then persists until the next destructive
fire.
We used our data on fire history and
on the rates and patterns of forest succession
after fire to reconstruct the sequence
of vegetation mosaics that must
have existed in the Little Firehole River
watershed during the last 240 years. Past
landscape patterns were reproduced by
first making a map showing the age (time
since the last destructive fire) of all homogeneous
forest units in 1978, based on
extensive field sampling and aerial photography.
Then, to reconstruct the landscape
of 1738, for example, we subtracted
240 years from the age of each stand
in 1978 and determined in which successional
stage a stand of that age would2D. G. Despain, personal communication.
September 1982 665
This content downloaded from 147.251.87.8 on Thu, 05 Dec 2019 09:44:21 UTC
All use subject to https://about.jstor.org/terms
have been. Where a fire had occurred
more recently than the date of interest
(e.g., areas that burned in 1739 in the
reconstruction for 1738) we assumed that
the stand was in a fire-susceptible late
successional stage (Romme 1982).
Figure 1 shows the proportions of the
Little Firehole River watershed covered
by early, middle, and late successional
stages at different times since 1738. In
1738 most of the area was covered by
late successional forests, but fires in
1739, 1755, and 1795 greatly reduced the
old-growth forests and replaced them
with early successional stages. Middle
successional stages became most abundant
around 1800 and have dominated
the watershed since. The early successional
stages that were common in the
late 1700s and early 1800s have been
very uncommon since the mid-1800s. A
decrease in middle successional stages
after 1938 and an associated increase in
late successional stages reflect forest
maturation on areas burned in 1739.
To further describe historic patterns in
landscape diversity, we calculated three
diversity indices (similar to those used
for measuring species diversity) and applied
them to our landscape reconstructions
for 1778-1978. We computed a
richness index, based on the number of
community types present; an evenness
index, reflecting the relative amount of
the landscape occupied by each community
type; and a patchiness index, indicating
the size and interspersion of individual
community units as well as the
structural contrast between adjacent
communities (Romme 1982). Figure 2
shows the results of plotting a weighted
average of all three indices, as a measure
of overall landscape diversity, and the
Shannon index (Pielou 1975), which we
calculated by using the proportion of the
watershed covered by a community type
as a measure of abundance. Both indices
reveal a similar pattern: Landscape diversity
was high in the late 1700s and
early 1800s following the extensive fires
of 1739, 1755, and 1795; it fell to a low
point in the late 1800s during a 70-year
period with no major fires; and it increased
again during this century as a
result of two small fires plus some variation
in the rate of forest maturation in
areas burned in 1739 and 1795. This
variation in rates of succession is attributable
to several factors including localized
high densities of the mountain pine
beetle (Dendroctonus ponderosae)
(Romme 1982).
The dramatic changes in landscape
composition and diversity in the Little
Firehole River watershed during the last
240 years (Figures 1 and 2) must have
been associated with significant changes
in ecosystem structure and function, including
net primary productivity, nutrient
cycling, total biomass, species diverMI
DLE60- /
,
50 /
Ui 40
30 - \ LATEI
I- .. .
S20- .
-1755 1834 1905 1949I0
?Y . EARLY
40 % Watershed Areo Burned
1739 1795
1834 1905 1949
1738 '58 '78 '98 1818 '38 '58 '78 '98 1918 '38 '58 '78
YEAR
Figure 1. Percent of watershed area covered by early, middle, and late stages of forest
succession from 1738-1978 in the 73-km2 Little Firehole River watershed, Yellowstone National
Park.
(A) AVERAGE OF RICHNESS, EVENNESS,
100 AND PATCHINESS INDICES
90
80
(B) SHANNON INDEX
.90
.80so
.70
1778 '98 1818 '38 '58 178 .98 1918 38 58 '78
YEAR
TIME OF EFFECTIVE
FIRE CONTROL
Figure 2. Changes in two measures of landscape
diversity in the Little Firehole River
watershed from 1778-1978.
sity, and population dynamics of
individual species. These relationships
cannot be fully quantified at this time,
but speculation based on existing knowledge
is useful.
IMPLICATIONS FOR WILDLIFE
Taylor and Barmore (1980) censused
breeding birds in a series of lodgepole
pine stands representing a gradient from
the earliest successional stages after fire
through late successional stages in the
park. Their data show the pattern of
avifaunal succession in a single homogeneous
stand. In attempting to answer the
question of how breeding bird species
and populations change with time in an
entire subalpine watershed, we used
Taylor and Barmore's (1980) census data
to estimate the number of breeding pairs
in each stand within our reconstructed
vegetation mosaics, summing the estimates
for all to arrive at an estimate of
breeding pairs in the entire watershed.
Figure 3 shows the results for three
representative species and for the total
number of breeding pairs of all species.
Mountain bluebirds (Sialia currucoides)
require open habitats with dead trees for
nesting. Such habitat was most abundant
in the Little Firehole River watershed
during the late 1700s and early 1800s
when 25-50% of the area was covered by
early forest successional stages following
the large fires of the 1700s (Figure 1).
Consequently, bluebirds may have been
very numerous at that time. However, as
forests matured bluebird populations
666 Bioscience Vol. 32 No. 8
This content downloaded from 147.251.87.8 on Thu, 05 Dec 2019 09:44:21 UTC
All use subject to https://about.jstor.org/terms
probably dropped dramatically (Figure
3). Today bluebirds are uncommon in the
watershed except in the 90-ha area that
burned in 1949. Note that this probable
population decline was a perfectly natural
event, occurring at a time when European
man had not yet entered the area.
In contrast to the bluebird, rubycrowned
kinglets (Regulus calendula)
prefer mature forests. Thus, kinglets
were less common when bluebirds were
most abundant (Figure 3). The yellowrumped
warbler (Dendroica auduboni)
breeds successfully in a variety of habitats
and as a result the population of this
species probably has fluctuated little
during the last 240 years despite the
major landscape changes that have occurred
(Figure 3). Figure 3 also shows
that the total number of breeding pairs of
all species has probably fluctuated greatly
in the last few centuries. The highest
numbers apparently were in the late
1700s and early 1800s when landscape
diversity was also greatest (Figure 2).
The population estimates shown in
Figure 3 can be challenged easily on the
basis that they were derived solely from
habitat availability, i.e., the number of
hectares of forest present in each age
class. Of necessity we have ignored other
critical determinants of population
MOUNTAIN 8LUESIRD
2000
1500
1000
500
o 0
x RUSY-CROWNED KINGLET
In 2000
1500
3--
S500
V.
YELLOW-RUMPED WAR8LER
Z 1000
w 500
0ALL
81RD SPECIES16000
12000
8000
4000
O
1738 '58 '78 '98 11818 '38 '58 '78 '98 1918 '38 '58 '78
YEAR
Figure 3. Estimated population sizes of
breeding birds in upland forests of the Little
Firehole River watershed, based on data from
Taylor and Barmore (1980) and the trends
shown in Figure 1. Populations in meadows
and riparian forests, which cover approximately
16% of the watershed, are not included
because appropriate population density data
are not available for these habitats.
Table 1. Relative values of plant communities and successional stages for el
habitat in the Little Firehole River watershed.*
DISTANCE COEFFICIENT
Potential
forage DISTANCE (m) TO COVER OR WATER
Plant community type value 0-320 320-800 800-1500 1500+
Alluvial woodland adjacent
to moist meadow 10 1.0 0.9 0.7 0.5
Meadow 9 1.0 0.9 0.7 0.5
Upland forest; early
successional stages 7 1.0 0.9 0.7 0.5
Upland forest; late
successional stages 3 1.0 0.9 0.7 0.5
Upland forest; middle
successional stages 2 1.0 0.9 0.7 0.5
*Based on models and discussions by Asherin 1973,
Leege 1976, Lonner 1976, Lyon 1971, Marcum 19
1976, Thomas et al. 1976, and Winn 1976.
density. Nevertheless, the overall patterns
are valid to the extent that they
show the constraints of habitat on potential
populations.
We were also able to consider the
effect of landscape change on elk. Using
a model much like that developed by
Thomas et al. (1976), we examined
changes in three critical habitat features
during the last 200 years, namely, forage
quantity and palatability, shelter (or cover),
and water. The forage and shelter
provided by an individual forest stand
change greatly during postfire succession.
Early successional stages usually
have the best forage whereas middle and
late successional stages provide the best
shelter. However, because elk use several
different kinds of habitat, the distribution
and interspersion of plant communities
and successional stages is critical.
Thus the center of a large meadow or
recently burned area may receive little
elk use, despite abundant forage, if it is
too distant from shelter or water, and the
potential shelter of very extensive tracts
of mature forest may be largely ignored if
little forage is available (Black et al.
1976, Hershey and Leege 1976, Marcum
1975, Reynolds 1966, Stelfox et al. 1976,
Thomas et al. 1976, Winn 1976).
We developed a relative ranking system
by which every type of plant community
and successional stage in the
Little Firehole River watershed was assigned
a value from 0-10 to indicate
potential forage value (Table 1). These
values were subjective, based on published
literature and our own observations
in the study area. We then divided
the watershed into 1429 units of 5 ha
each, identified the dominant vegetation
type within each unit, and assigned appropriate
values to each. Every value
was multipled by a distance coefficient
reflecting the distance to the nearest
shelter or water if those features were
not present within the unit itself (Table
1), the product being our elk habitat
index. In this manner we analyzed our
reconstructed vegetation mosaics for
1778, 1878, and 1978.
Figure 4 (a, b, and c) shows the results
for three 5-ha units having different histories
of fire and forest regrowth. As a
result of changes in stand structure, the
quality of elk habitat has varied greatly.
However, when we averaged the values
for all 1429 individual 5-ha units to obtain
an estimate of elk habitat quality for the
watershed as a whole, we found much
less difference among the landscapes of
1778, 1878, and 1978 (Figure 4d). There
are probably two main reasons for this
result. First, temporary increases in habitat
quality in one part of the watershed
(due primarily to the great improvement
in forage after fire) have been balanced
by decreases resulting from forest maturation
on other areas burned earlier. Second,
and probably more important, the
best habitat is in and around moist meadows
where forage, shelter, and water all
occur in close proximity. In fact, our
model may underestimate the habitat
quality of subalpine meadows in the
park, since we reduced our elk hr
index in the centers of large meadow N to
reflect the distance to shelter. However,
the shelter requirement apparently is
much less critical for elk populations that
are not hunted by man, and elk in the
park are frequently observed feeding in
the centers of large meadows.3 We were
unable to determine whether the large
fires in the surrounding uplands had
burned the meadows and adjacent allu-
3L. Irwin, personal communication.
September 1982 667
This content downloaded from 147.251.87.8 on Thu, 05 Dec 2019 09:44:21 UTC
All use subject to https://about.jstor.org/terms
vial woodlands. We assumed that the
fires in these areas were of low intensity
and produced little change in community
structure or elk habitat. Although our
results suggest that fires may not greatly
influence the overall quality of elk summer
range on the high plateaus of Yellowstone,
elk are attracted to recently
burned areas (Davis 1977), and over
much of the subalpine zone, moist meadows
are less common than in the Little
Firehole River watershed. Where meadows
are less common, summer elk populations
may fluctuate in response to
changes in the upland landscape.
IMPLICATIONS FOR
AQUATIC ECOSYSTEMS
One of the most interesting and attractive
features of the park is Yellowstone
Lake. This virtually unpolluted subalpine
lake covers 354 km2 and contains
populations of the native cutthroat trout
(Salmo clarkii). The trout support a complex
food chain including pelicans, ospreys,
otters, and bears. Some evidence
indicates that the lake's net primary productivity
has declined during the last
century, as has its carrying capacity for
trout and associated top predators
(Shero 1977, US National Park Service
1975). Because the period of apparent
decline coincides with attempts at fire
control, some have suggested that the
cause is reduced nutrient input to the
lake due to biotic immobilization by forests.
As noted earlier, however, our research
indicates that the natural fire regime
has not been greatly altered by
man's activities in the Yellowstone subalpine
zone, particularly in the very remote
areas that drain into Yellowstone
Lake.
Rather than attribute the cause to fire
suppression, we favor the hypothesis
that lake productivity is to some extent
synchronized with the long-term fire cycle
that seems to prevail in the watershed
of Yellowstone Lake. A variety of evidence
supports this hypothesis. For example,
experiments in the Rocky Mountains
have shown that removal of mature
forest from 40% of a subalpine watershed
results in an increase in total water
discharge of 25% or more (Leaf 1975).
The increase is due to several factors
related to the distribution and melting of
the winter snowpack. Albin (1979) compared
two small tributary streams of
Yellowstone Lake; about 20% of one
watershed was burned by fires 36 and 45
years previously, whereas the other watershed
was unburned. The burned watershed
had greater seasonal variation in
streamflow and greater total water discharge
per hectare. If a large portion of a
subalpine watershed burns at intervals of
approximately 300 years, as seems to
occur in the Little Firehole River watershed,
then streamflow also may exhibit a
long-term cycle over and above yearly
and seasonal fluctuations. During the
high-discharge portion of the cycle, especially
in years of high snowfall, debris
is washed out of stream channels, new
channels are cut, and new alluvial deposits
are created. Such events influence
habitat for fish as well as for floodplain
species like willow and alder, which in
turn are important browse species for elk
and other terrestrial animals (Houston
1973).
But more important to the question of
Yellowstone Lake is the nutrient content
of stream water. Immediately after
deforestation by fire or cutting there
often is an increase in dissolved minerals
due to erosion, reduced plant uptake,
increasd microbial activity, increased
leaching, and the release of elements
from organic matter by fire (Bormann
and Likens 1979, McColl and Grigal
1975, Wright 1976). The increase is usually
short-lived, lasting several years at
most (Albin 1979, Bormann and Likens
1979), but it may be important as a
periodic nutrient subsidy (Odum et al.
1979) to oligotrophic aquatic ecosystems.
As young forests become established,
biotic immobilization is so effective
that nutrient concentrations in
stream water fall to very low levels (Bormann
and Likens 1979, Marks and Bormann
1972, Vitousek and Reiners 1975).
Thus a watershed dominated by early
and middle forest successional stages
(e.g., the Little Firehole River watershe
during the 1800s) would produce relatively
nutrient-poor water. As forests
reach late successional stages, tree
growth and net primary productivity decrease,
nutrient uptake is less, and consequently
the leachate is richer in dissolved
minerals (Bormann and Likens
1979, Vitousek and Reiners 1975).4
Thus, although the possible connection
between fire suppression and reduced
productivity in Yellowstone Lake
is plausible, an equally attractive alternative
hypothesis is that extensive fires in
the watershed about 100 years ago replaced
many late successional forests
with early successional stages. As young
forests over much of the watershed began
utilizing soil nutrients more efficiently,
the total amount leached into stream
water feeding the lake was reduced accordingly.
If this is true, any recent
decline in lake productivity may be a
natural phenomenon that has occurred
many times in the past and will be alleviated
as forests in the watershed mature.
Of course, the Yellowstone Lake watershed
is very large (ca. 2600 km2), and
landscape patterns over this large area
may be in a state of dynamic equilibrium,
or what Bormann and Likens (1979) have
referred to as a shifting mosaic steady
(a) 5-HA (b) 5-HA (c) 5-HA (d) AVERAGE
UNIT UNIT UNIT FOR
BURNED BURNED BURNED ENTIRE
5.0 1739 1834 1949 WATERSHE D
-4.0 -
3.0
2.0-
.0 - . --.
CD ) w 0 wYR O
YEAR
Figure 4. Elk habitat index (see text) for three representative 5-ha units and for the entire Little
Firehole River watershed (d) in 1778, 1878, and 1978.
4Pearson, J. A., D. H. Knight, and T. J. Fahey.
Unpublished ms. Net ecosystem production and
nutrient accumulation during stand development in
lodgepole pine forest, Wyoming.
668 BioScience Vol. 32 No. 8
This content downloaded from 147.251.87.8 on Thu, 05 Dec 2019 09:44:21 UTC
All use subject to https://about.jstor.org/terms
state. If this is found to be true for the
Yellowstone Lake watershed, then total
nutrient input to the lake should be about
the same from year to year (though the
source would vary), and some other explanation
for the decline in lake productivity
will be required.
CONCLUSIONS
After a century of ecological research
that focused largely on species or individual
communities or ecosystems, there
now is a growing interest in still higher
levels of organization such as the landscape
and biosphere. Changes in landscape
patterns influence a variety of natural
features including wildlife, water
and nutrient flow, and the probability of
different kinds of natural disturbances.
Given a sufficiently large area and a
natural disturbance regime, various measures
of landscape pattern may remain
fairly constant over time despite dramatic
cyclic changes in localized areas such
as a small watershed. Such "steady
states" have been demonstrated or hypothesized
for a Swedish boreal forest
(Zachrisson 1977), high-elevation fir forests
in New England and elsewhere
(Sprugel 1976, Sprugel and Bormann
1981), primeval northern hardwood forests
of North America (Bormann and
Likens 1979), and mesic deciduous forests
of the southern Appalachians (Shugart
and West 1981). Our results suggest
that strong cyclic changes occur on areas
of at least 100 km2 in Yellowstone National
Park, but more research is needed
to determine if the landscape patterns in
the park as a whole are in a state of
equilibrium. Large wilderness areas,
when protected from pollutants and
managed so that natural perturbations
can continue, provide the best and probably
the only locale for studying the kind
of landscape changes that occurred for
millenia in presettlement times.
ACKNOWLEDGMENTS
This research was supported by grants
from the University of Wyoming---National
Park Service Research Center. We
thank D. G. Despain and D. B. Houston
for sharing their observations and records
of natural fires in Yellowstone; L.
Irwin for advice on our elk habitat model;
N. Stanton and M. Boyce for guidance
in estimating diversity; R. Levinson
and R. Marrs for assistance with aerial
photograph interpretations; M. Cook for
computer programming assistance; R.
Levinson, L. van Dusen, K. White, and
P. White for field assistance; D. G. Despain,
L. Irwin, W. H. Martin, W. G.
Van der Kloot, and two anonymous reviewers
for helpful comments on the
manuscript; and K. Diem, M. Meagher,
J. Donaldson, and the staff of the Old
Faithful Ranger Station for administrative
and logistical support.
REFERENCES CITED
Albin, D. P. 1979. Fire and stream ecology in
some Yellowstone Lake tributaries. Calif.
Fish Game 65: 216-238.
Arno, S. F. and K. M. Sneck. 1977. A method
for determining fire history in coniferous
forests of the mountain west. US For. Serv.
Gen. Tech. Rep. INT-42.
Asherin, D. A. 1973. Prescribed burning effects
on nutrition, production and big game
use of key northern Idaho browse species.
Doctoral Thesis, University of Idaho, Moscow,
ID.
Baker, R. G. 1970. Pollen sequence from late
quaternary sediments in Yellowstone Park.
Science 168: 1449-1450.
Basile, J. V. and C. E. Jensen. 1971. Grazing
potential on lodgepole pine clearcuts in
Montana. US For. Serv. Res. Pap. INT-98.
Beetle, A. A. 1974. Range survey in Teton
County, Wyoming. Part IV: quaking aspen.
Univ. of Wyom. Agri. Exp. Station Publ.
SM 27.
Black, H., R. J. Scherzinger, and J. W.
Thomas. 1976. Relationships of Rocky
Mountain elk and Rocky Mountain mule
deer habitat to timber management in the
Blue Mountains of Oregon and Washington.
Pages 11-31 in Proceedings of the ElkLogging-Roads
Symposium. Forest, Wildlife,
and Range Experiment Station,
University of Idaho, Moscow, ID.
Bormann, F. H. and G. E. Likens. 1979. Pattern
and Process in a Forested Ecosystem.
Springer-Verlag, New York.
Davis, P. R. 1977. Cervid response to forest
fire and clearcutting in southeastern Wyoming.
J. Wildl. Manage. 41: 785-788.
Despain, D. G. 1973. Major vegetation zones
of Yellowstone National Park. Information
Paper Number 19, Yellowstone National
Park, WY.
Despain, D. G. and R. E. Sellers. 1977. Natural
fire in Yellowstone National Park. West.
Wildlands 4: 20-24.
Forman, R. T. T., ed. 1979. Pine Barrens:
Ecosystem and Landscape. Academic
Press, New York.
Forman, R. T. T. 1982. Interaction among
landscape elements: a core of landscape
ecology. In Perspectives in landscape ecology.
Proceedings of the 1981 Symposium of
the Netherlands Society for Landscape
Ecology, Veldhoven, Pudoc, Wegeningen,
the Netherlands, in press.
Forman, R. T. T. and R. E. Boerner. 1981.
Fire frequency and the Pine Barrens of
New Jersey. Bull. Torrey Bot. Club 108:
34-50.
Forman, R. T. T. and M. Godron. 1981.
Patches and structural components for a
landscape ecology. BioScience 31: 733-
740.
Habeck, J. R. 1976. Forests, fuels and fire in
the Selway-Bitterroot Wilderness, Idaho.
Pages 305-352, in E. V. Komarek, general
chairman Proceedings of the Montana Tall
Timbers Fire Ecology Conference and Fire
and Land Management Symposium Number
14, 1974. Tall Timbers Research Station,
Tallahassee, FL.
Habeck, J. R. and R. W. Mutch. 1973. Firedependent
forests in the northern Rocky
Mountains. Quat. Res. (NY) 3: 408-424.
Hansson, L. 1977. Landscape ecology and
stability of populations. Landscape Planning
4: 85-93.
Heinselman, M. L. 1973. Fire in the virgin
forests of the Boundary Waters Canoe
Area, Minnesota. Quat. Res. (NY) 3: 329-
382.
Hershey, T. J., and T. A. Leege. 1976. Influences
of logging on elk on summer range in
north-central Idaho. Pages 73-80 in Proceedings
of the Elk-Logging-Roads Symposium;
Forest, Wildlife, and Range Experiment
Station, University of Idaho,
Moscow, ID.
Hett, J. M. 1971. Land-use changes in east
Tennessee and a simulation model which
describes these changes for three counties.
ORNL-IBP-71-8. Oak Ridge National Laboratory,
Ecological Sciences Division, Oak
Ridge, TN.
Houston, D. B. 1971. Ecosystems of National
Parks. Science 172: 648-651.
. 1973. Wildfires in northern Yellowstone
National Park. Ecology 54: 1111-
1117.
Johnson, W. C. 1977. A mathematical model
of forest succession and land use for the
North Carolina Piedmont. Bull. Torrey Bot.
Club 104: 334-346.
Johnson, W. C. and D. M. Sharpe. 1976. An
analysis of forest dynamics in the northern
Georgia Piedmont. Forest Science 22: 307-
322.
Kilgore, B. M. and D. Taylor. 1979. Fire
history of a sequoia-mixed conifer forest.
Ecology 60: 129-142.
Leaf, C. F. 1975. Watershed management in
the central and southern Rocky Mountains:
A summary of the status of our knowledge
by vegetation types. US For. Serv. Res.
Pap. RM-142.
Lonner, T. N. 1976. Elk use-habitat type relationships
on summer and fall range in Long
Tom Creek, southwestern Montana. Pages
101-109 in Proceedings of the Elk-LoggingRoads
Symposium. Forest, Wildlife, and
Range Experiment Station, University of
Idaho, Moscow, ID.
Loope, L. L. and G. E. Gruell. 1973. The
ecological role of fire in the Jackson Hole
area, northwestern Wyoming. Quat. Res.
(NY) 3: 425-443.
Loucks, O. L. 1970. Evolution of diversity,
efficiency, and community stability. Am.
Zool. 10: 17-25.
September 1982 669
This content downloaded from 147.251.87.8 on Thu, 05 Dec 2019 09:44:21 UTC
All use subject to https://about.jstor.org/terms
Special Intro ductory .Pice
?A.ooto
A i s
Inspire Stud.nts,'
Teachers and
*
. . . .... .. ..?:
THE JOYS OF RESEARCHI
T E I )"I.
Walter Shropshire- Jr., Editor
. u
All too often,, science has an imperfonaI qual
hd : e... - t ii is: :.
ty, and the motivations of individual s.iei.ists
-the o ofr?esar :i.c.h-t t o l i
s ii n
thmare rrl6iie hsclecino ei
niscences and recollections by eight prominent
.. .0
contemporary scientists (including four Nobel
Prize winners), however, is an excellent step
t .d remedygt so g
loinotehman- s-ide fsiec4inldn
the wrong turns and frustrations that can be
suffered and the fulfillment that can be enjyed9
in the pursuit of knowled . The discussions
.houl dbe eas i er
.. ul d:::b'e d :: :: :?::
reoader and m--ay inspire ol?der. te-ens contemplatg
science-reladted professions."
-Booklcsts
Please send The joys of Research at the introductory
price of $5.56.
Name
Title
Organization
Address
City
State, Zip
uL My check/money order is enclosed.
[E Purchase Order is attached.
*Only $5.56 in paper (20% off the retail price of $6.95)
Tii
% N
...
? . ?.
gESEARCH ..
N.,
180 pages 65 illus. 9 i : 93 7
ISBN - -874748-7- paper W95IS-N 0-87478585 cth $17.50
120 357 93
P e a w d ieY
Luder, P. 1981. The diversity of landscape
ecology. Definition and attempt at empirical
identification. Angew. Botanik 55: 321-
329.
Lunan, J. S. and J. R. Habeck. 1973. The
effects of fire exclusion on ponderosa pine
communities in Glacier National Park,
Montana. Can. J. For. Res. 3: 574-579.
Lyon, L. J. 1971. Vegetal development following
prescribed burning of Douglas fir in
south-central Idaho. US For. Serv. Res.
Pap. INT-105.
Marcum, C. L. 1975. Summer-fall habitat selection
and use by a western Montana elk
herd. Doctoral Thesis, University of Montana,
Missoula, MT.
_ .1976. Habitat selection and use during
summer and fall months by a western Montana
elk herd. Pages 91-96 in Proceedings
of the Elk-Logging-Roads Symposium.
Forest, Wildlife, and Range Experiment
Station, University of Idaho, Moscow, ID.
Marks, P. L. and F. H. Bormann. 1972. Revegetation
following forest cutting: Mechanisms
for return to steady-state nutrient
cycling. Science 176: 914-915.
McColl, J. G. and D. F. Grigal. 1975. Forest
fire: effects on phosphorus movement to
lakes. Science 188: 1109-1111.
Odum, E. P. 1971. Fundamentals of Ecology.
Third edition. W. B. Saunders Co, Philadelphia,
PA.
Odum, E. P., J. T. Finn, and E. H. Franz.
1979. Perturbation theory and the subsidystress
gradient. BioScience 29: 349-352.
Peek, J. M., A. L. Lovaas, and R. A. Rouse.
1967. Population changes within the Gallatin
elk herd, 1932-65. J. Wildl. Manage. 31:
304-316.
Pickett, S. T. A. 1976. Succession: an evolutionary
interpretation. Am. Nat. 110: 107-
119.
Pengelly, W. L. 1963. Timberlands and deer
in the Northern Rockies. J. For. 61: 734-
740.
Pielou, E. C. 1975. Ecological Diversity. Wiley-Interscience,
New York.
Reiners, W. A. and G. E. Lang. 1979. Vegetational
patterns and processes in the Balsam
fir zone, White Mountains, New
Hampshire. Ecology 60: 403-417.
Reynolds, H. G. 1966. Use of openings in
spruce-fir forests of Arizona by elk, deer,
and cattle. US For. Serv. Res. Note RM-66.
Romme, W. H. 1982. Fire and landscape diversity
in subalpine forests of Yellowstone
National Park. Ecol. Monogr. 52: 199-221.
Romme, W. H. and D. H. Knight. 1981. Fire
frequency and subalpine forest succession
along a topographic gradient in Wyoming.
Ecology 62: 319-326.
Rowe, J. S. 1961. Critique of some vegetational
concepts as applied to forests of
northwestern Alberta. Can. J. Bot. 39:
1007-1017.
Shero, B. R. 1977. An interpretation of temporal
and spatial variations in the abundance
of diatom taxa in sediments from
Yellowstone Lake, Wyoming. Doctoral
Dissertation, Univ. of Wyoming, Laramie.
Shugart, H. H., Jr., and D. C. West. 1981.
Long-term dynamics of forest ecosystems.
Am. Sci. 69: 647-652.
Shugart, H. H., Jr., T. R. Crow, and J. M.
Hett. 1973. Forest succession models: a
rationale and methodology for modeling
forest succession over large regions. For
Sci. 19: 203-212.
Sprugel, D. G. 1976. Dynamic structure of
wave-regenerated Abies balsamea forests
in the north-eastern United States. J. Ecol.
64: 889-911.
Sprugel, D. G. and F. H. Bormann. 1981.
Natural disturbance and the steady state in
high-altitude balsam fir forests. Science
211: 390-393.
Stelfox, J. G., G. M. Lynch, and J. R. McGillis.
1976. Effects of clearcut logging on wild
ungulates in the central Albertan foothills.
For. Chron. 52: 65-70.
Swain, A. M. 1980. Landscape patterns and
forest history in the Boundary Waters Canoe
Area, Minnesota: A pollen study from
Hug Lake. Ecology 61: 747-754.
Taylor, D. L. 1973. Some ecological implications
of forest fire control in Yellowstone
National Park, Wyoming. Ecology 54:
1394-13%.
Taylor, D. L. and W. J. Barmore. 1980. Postfire
succession of avifauna in coniferous
forests of Yellowstone and Grand Teton
National Parks, Wyoming. Pages 130-145
in R. M. DeGraff, Technical Coordinator,
Proceedings of the Workshop on Management
of Western Forests and Grasslands
for Nongame Birds. US For. Serv. Gen.
Tech. Rep. INT-86.
Thomas, J. W., R. J. Miller, H. Black, J. E.
Rodiek, and C. Maser. 1976. Guidelines for
maintaining and enhancing wildlife habitat
in forest management in the Blue Mountains
of Oregon and Washington. Pages
452-476 in Transactions of the Forty-first
North American Wildlife and Natural Resources
Conference, Washington, DC.
US National Park Service. 1975. The Natural
Role of Fire: A Fire Management Plan for
Yellowstone National Park. Unpublished
report, Yellowstone National Park, WY.
Vitousek, P. M. and W. A. Reiners. 1975.
Ecosystem succession and nutrient retention:
a hypothesis. BioScience 25: 376-381.
White, P. S. 1979. Pattern, process, and natural
disturbance in vegetation. Bot. Rev. 45:
229-299.
Winn, D. S. 1976. Terrestrial vertebrate fauna
and selected coniferous forest habitat types
on the north slope of the Uinta Mountains.
United States Department of Agriculture,
Wasatch National Forest, Region 4.
Wright, H. E., Jr. 1974. Landscape development,
forest fires, and wilderness management.
Science 186: 487-495.
Wright, R. F. 1976. The impact of forest fire
on the nutrient influxes to small lakes in
northeastern Minnesota. Ecology 57: 649-
663.
Zachrisson, O. 1977. Influence of forest fires
on the north Swedish boreal forest. Oikos
29: 22-32.
CIRCLE NO. 127 ON THE READER'S SERVICE CARD BioScience Vol. 32 No. 8
This content downloaded from 147.251.87.8 on Thu, 05 Dec 2019 09:44:21 UTC
All use subject to https://about.jstor.org/terms