Vegetation-Environment Relations
at Whiteface Mountain in the Adirondacks
ae Gary Holway ¢ Jon T. Scott
eee -__ Co-Investigators
Report No. 92
Atmospheric Sciences Research Center
‘State University of New York at Albany
N.S.F. Grant NO. GB-7020
VEGETATION-ENVIRONMENT
(
Peo RELATIONS AT WHITEFACE
|
| MOUNTAIN IN THE ADIRONDACKS
|
|
f
Report No. 1
J. Gary Holway
Jon T. Scott
Co-Investigators
With contributions by A. R. Breisch, J. G. Droppo,
H. L. Hamilton, &. W. Holroyd III, R. F. Kujawski
P. C. Lemon, S. Nicholson and R. Park
' Sponsored by
National Science Foundation
Grant No. GB-7020 and
the Atmospheric Sciences Research Center
Report No. 92
Atmospheric Sciences Research Center
State University of New York at Albany
Albany, N.Y. 12203
August 1969
INTRODUCTORY REMARKS
This report is the first'in a planned series on the ecology of
Whiteface Mountain. . The research included began in 1964 as a summer
program sponsored by the Atmospheric Sciences Research Center. In 1967
further support was provided by the National Science Foundation.
While some of the papers in this report have been completed or nearly
so for periods of one to three years, it was decided to hold their
printing until all of the reports were finished. It was hoped that the
result would be a more complete picture of the Whiteface ecology studies.
Certainly, there are desirable features of having these eight related
studies in one report. Unfortunately, this procedure has delayed
publication of the early results. We apologize for the inconvenience to
the many interested persons who have written to us for reports.
The results presented here represent the early phases of a planned
long-term study of vegetation-environment relations in the Whiteface
Mountain region. The first four papers deal with the nature of the
vegetation, its variation with topography and comparison with other
regions. The next two deal with special aspects of the vegetation;
one with the alpine tundra and the other with an important species..
The last two papers represent beginning studies of the environment.
The investigators are in accordance with the viewpoint that an
important approach to the study of nature lies in understanding the
ecosystem. The ecosystem, no matter how it is defined, cannot be under-
stood except in an environmental framework. Much of our present and
planned future studies deal with measuring the environment and understanding
its relation to the living component of the ecosystem.
ACKNOWLEDGEMENTS
The co-investigators wish to thank the many persons and organizations
who have contributed to the work leading to this report. Many of these
are acknowledged in the individual papers. We wish to express special
thanks to Vincent Schaefer, Director of the Atmospheric Sciences Research
Center and Raymond Falconer, Director of the ASRC Field Station at
Wilmington, N.Y., whose cooperation has been invaluable. We thank the
Natural Sciences Institute sponsored by the Kettering Foundation for the
help provided by many of its student participants and are grateful to
the Whiteface Mountain Authority and New York State Conservation Department
for cooperation on use of the mountain facilities.
Many persons provided useful advice and discussion on various aspects
of the research. These include in particular Richard Arnold, Earl Stone,
and Lee Miller of Cornell University, Edwin Ketchledge of Syracuse University,
Orie Loucks of the University of Wisconsin, Richard Park of Rensselaer
Polytechnic Institute and Stanley Smith of the New York Museum.
We are especially grateful to the Atmospheric Sciences Research Center
and the National Science Foundation (GB-7020) for support of this research.
ii
TABLE OF CONTENTS
Page
INTRODUCTION. 00... cc ceeeseeveeeee srecerecavacorceiete ates SaereReaaeEN ee mums 14
ACKNOWLEDGEMENTS.........eee005 AG @NNEORS 9G TREE NRT TDS CORO RREN ii
TABLE OF CONTENTS........000 WLS EOTRNTSS CRGENG LES COS eeee Ui
Vegetation of the Whiteface Mountain Region of the Adirondacks.
J.G. Holway, J.T. Scott and S. Nicholson......... aiseracqraserslainnerezelwleenesees . 1
Comparison of Topographic and Vegetation Gradients in Forests of
Whiteface Mountain, New York. J.T. Scott and J.G. Holway........ seeeee 44
Multi-dimensional Ordination of Boreal and Hardwood Forests on
Whiteface Mountain. A.R. Breisch, J.T. Scott, R.A. Park and P.C. Lemon 89
A Floristic Comparison of Undisturbed Spruce-fir Forests of the
Adirondacks with Four Other Regions. S. Nicholson, J.G. Holway and
J.T. SCOtt... cece cea e eee A He aS a esaee ee sieceumcwzece wiaieterstese oe weinreNtees HimeTeNNETTErS 136
Slope-aspect Variation in the Vascular Plant Species Composition in the
Treeless Community Near the Summit of Whiteface Mountain, N.Y.
S. Nicholson and J.T. Scott.
Ecological Effectiveness of Yellow Birch in Several Adirondack Forest
Types. R.F. Kujawski and P.C. Lemon........... seeeeeeee eonietesnte ase aceceee 161
The Determination of Vertical Micrometeorological Profiles Through a
Forest Canopy with a Single Set of Sensors. J. Droppo and H. Hamilton. 92
The Prevailing Winds on Whiteface Mountain as Indicated by Flag Trees.
E. Holroyd III....... ccc cece eee eeee 4 Serwewes RUPVISSTS VE DER EEE. veeee 220
iti
VEGETATION OF THE WHITEFACE MOUNTAIN
REGION OF THE ADIRONDACKS
by
J. Gary Holway, Jon T. Scott,
and Stuart Nicholson
VEGETATION OF THE WHITEFACE MOUNTAIN
REGION OF THE ADIRONDACKS
by J. Gary Holway, Jon T. Scott, and Stuart Nicholson
ABSTRACT
One hundred and eighty-two. forest stands located by both selective and
systematic means were sampled for the standard ecologic parameters of
frequency, density and basal area. Under the systematic stand selection
method altitude and slope aspect were the determining factors in site
selection. Selective stand location was used to increase the sample
size of ‘undersampled' association types.
The tree species presence list for all stands includes 30 species.
At least one or the other of the three leadin stand dominant species, red
spruce (Picea rubens) 143 stands, balsam fir (abies balsamea) 120 stands,
and sugar maple (Acer saccharum) 80 stands, occurs in all but 9 of the
stands and dominates 103 of them.
The altitudinal ranges 1500 ft and under, 1500-2500 ft, and 3000 ft
and over are primarily dominated by species associations representative of
the Appalachian oak and pine, the northern mesic-hardwood and the boreal
floristic provinces, respectively. Less expected than the altitudinal
relationships was the apparent effect of slope aspect on composition.
While north facing slopes are typically envisioned as being more boreal, on
Whiteface the west slope was in all stands except those above 4000 ft.
INTRODUCTION
The forest vegetation of the Adirondack Mountains in northern New York
has received little attention from ecologists. This is difficult to under-
stand in view of the floristic diversity of the region and its strategic
location between boreal forests to the north and the Catskill and Appalachian
mountain forests to the south.
This paper is the first in a planned series concerning the relation
between vegetation and environment in the region of Whiteface Mountain in
the northern part of the Adirondacks. It deals primarily with a description
of the study area, its dominant vegetation and how this vegetation is dis-
tributed over the wide range of topographic habitats on Whiteface Mountain.
The present findings are based upon standard ecologic measures from 182
forested stands taken during three summers from 1964-1966. Later papers
will emphasize more detailed statistical treatment of the data including a
comparison of several techniques for obtaining the vegetation gradient Sor
ordination) similar to those discussed by Goff and Cottam (1967), methods
of measuring forest environments and techniques of relating vegetation
and environment gradients.
oe
~3-
THE STUDY AREA
General
One of the high peaks within the protective boundaries of the Adirondack
Forest Preserve seemed a reasonable choice as a site representative of forest
vegetation in the Adirondacks. To be compatible with the research aims, the
mountain selected had also to meet the requisites of floristic and physio-
graphic diversity and preferably be readily accessible without being exces-
sively disturbed. Whiteface Mountain, northernmost of the high peaks of
the Adirondacks more than met these criteria.
Whiteface Mountain (4867 ft) is the fifth highest peak in New York
State. It is located just west of the village of Wilmington, in the Town
of North Elba, Essex County. The base of the mountain, which varies from
about 1008'ft on the east to over 1900 ft on the west, rises quite abruptly
on all sides to a horn-like summit. The east and west sides of the summit
exhibit well developed cirques, with a moderately-developed cirque on
the northern exposure.
The cirques are separated by rather sharp, arete-like ridges which
broaden out or slope up again to form minor peaks at some distance from
the summit (see Figure 1). Compared to the other Adirondack high peaks,
Whiteface is relatively isolated. Except for Mount Esther (4200 ft), and
Lookout Mountain (4000 ft), which are, in effect, sub-peaks, of the Whiteface
massif, the nearest 4000 foot peaks are more than 10 miles to the south.
This isolation allows more critical evaluation of mountain physiographic
effects on microclimate and vegetation distribution because of the reduction
of modifying influences. by nearby peaks.
The flora of Whiteface ranges from a sparsely developed alpine tundra
with extraneous elements of arctic tundra present on the summit, through
dense spruce-fir boreal elements to mixed hardwood and coniferous types
near the base with many species typical of more southerly climates.
In addition, there are abundant oak, pine, and bog (spruce and tamarack)
associations in the vicinity. Indeed, it is not uncommon to find elements
of arctic, boreal forest, and deciduous forest biomes existing within the
same square mile of area.
Accessability was another of the strong attributes of Whiteface. It
is 150 miles from the State University Center at Albany and 35 miles from
additional facilities at State University College at Plattsburgh, New York.
Whiteface Memorial Highway allows ready access to the summit, and the
Whiteface ski area facilities provide similar ease of access to the east and
southeast sub-prominences. In addition, there are hiking trails, and roads
of one form or another (county highways or closed fire-access roads) virtu-
ally encircle the basal perimeter of the mountain. All of these have con-
tributed to the ease of obtaining the vegetation data.
Another attribute of Whiteface that might seem contradictory to the
accessibility of the area is the relative freedom from modern disturbance.
While the summit and roadway show marked effects of heavy tourist impact,
and the ski site development has created severe localized disturbance, the
greater portion of the mountain has been free of significant human distur-
bance since the passing of the "Forever Wild" legislation in 1896.
A final factor which strongly influenced the selection of Whiteface
was an invitation extended by the Atmospheric Sciences Research Center of
the State University of New York to base operations at their Marble Mountain
Field Station located at 1980 feet on the east slope of the Whiteface complex.
Figure 1:
~4-
Map of Whiteface Mountain study area. The area
shown contains 132 of the 182 stands sampled,
“6s
Geology
The Adirondacks are one of the oldest mountain ranges on earth.
Metamorphic rock beds form the bulk of Whiteface and were intruded into
sedimentary rocks of the Grenville series, about 1.1 billion years ago
(Bird 1963). The mountain consists primarily of Whiteface anorthosite.
Only the lower western portion of the mountain is granitic. Erosion since
early Cenozonic has completely removed deposits formed during the Paleozoic
submergences (Miller 1918). The major features evident today are the
result of Pleistocene glaciation which culminated in the Wisconsin age about
10,000 years ago generally, but ended much more recently at the higher ele-
vations. Cirques, U-shaped valleys, and the many lakes and ponds, all bear
testimony to this glacial impress, Glacial moraines are also well dispersed
over the Whiteface terrain. Isachsen (1964), reports that striations on
the highest Adirondack peaks imply a minimum ice thickness of 5000 ft.
Craft thee. comm.), who has been studying the Pleistocene geology of the
Adirondacks, with particular interest centered on Whiteface and the high
peaks area, can present data to support a lower thickness for the continental
ice sheet during the last glacial advance in the Adirondacks. He has
found no evidence of continental glaciation above approximately 4200 ft,
although there is considerable evidence below this elevation throughout
the high peaks area. Craft hypothesizes that near the end of the Wisconsin
stage the peaks above 4200 ft appeared as nunataks sticking up out of the
continental sheet and that the till and debris from earlier advances became
eroded from these peaks. This would help to account for the very shallow
mineral soils to be found on the summits of the higher Adirondack peaks.
Adirondack Forest Soils
There have been few studies of the forest soils in the Adirondacks to
date. Heimburger (1934), emphasized that drainage conditions and the geo-
logical origin of the soil are of primary importance in the distribution
of vegetation in the Adirondacks. In this report he also superficially
describes a number of Adirondack forest associations in relation to soil
types,
Donahue (1940) reported on a forest-site quality investigation, and
McFee and Stone (1965) made a quantitative analysis of the physical and
chemical nature of an Adirondack forest podzol near Paul Smith's, approxi-
mately 20 miles west of Whiteface.
Reilly, (1964), has described broad soil features encountered in a
study of the mosses of Whiteface, but these are primarily qualitative
observations. Z
Our observations show that there is considerable variation in the charac-
teristics of the soil under the various forest associations, Near the summit
the forest is rooted in many cases in thick peaty mats overlying virtually
structureless and poorly differentiated mineral soil horizons containing
many boulders. Conversely, the deciduous forest associations have the
typical brown podzol characteristics with shallow organic layers, well
developed Ay, and well defined structures and mineral horizons. The soils
are primarify acidic even though most are derived from the underlying
anorthosite which is basic.
aps
An intensive study of the soils in the Whiteface region was made by
Witty (1968). His main emphasis was to establish a set of criteria for
classification of the Adirondack forested soils. He identified and
described 18 subgroups including 12 histosols and 6 spodosols for the
Whiteface area. Witty's extensive quantitative data will form the basis
of our-own evaluation of forest soil environment in later studies.
Climate
The Adirondacks are a region of cold snowy winters and cool wet summers.
Under the Koppen-Geiger system the typical lowland Adirondack climate fits a
Dfb, or very nearly a Dfc, the cold-summer, humid continental type. Near
the summits of the higher mountains the climateis colder, more windy and
more moist with a high frequency of cloud caps. Rime icing may occur in
any month of the year.
The Adirondacks receive from 37 to 53 in. of precipitation annually
with an excess precipitation over evapotranspiration of from 25 to 40 in.
The region is the source of the Hudson and other rivers.
The growing season in the Whiteface region is about 80 to 105 days
in the lowlands with much shorter perfods near mountain summits. The
cumulative monthly degrees over 40°F is below 100 in the lowlands. Mean
monthly temperatures near the summit of Whiteface are from 10°F to 15°F
lower than at Lake Placid with summer monthly means of 50°F to 58°F.
Vegetation History
Historical records, filed evidence, and interviews with long-time
residents, indicate some marked differences between many of the original
forests and present cover. Damage to vegetation by Indians and transient
visitors appears to have been negligible prior to permanent settlement in
the 1800's. Wood was cut only for local use until cutting of hardwoods for
charcoal production began east of Lake Placid in 1815, and near Wilmington
in 1832. Charcoal cutting continued in the lowlands until the late 1800's.
These clear-cutting operations were frequently followed by fires of varying
intensities. A severe fire on the sunmit in 1867 (Watson, 1869) is the
only other known disturbance of major proportions before pulp logging began
in 1892. As late as the 1860's, much of the total forest area on Whiteface
Mountain was apparently untouched.
Virtually all of the spruce-fir was first-growth when pulping began.
Street (1869) emphasized the "gloom of the terrific forests" on the east
side of Whiteface, and Stoddard (1879) pictured the coniferous forests on
the north as "dark and thick."
Pulp cutting began on the east side of Whiteface and ceased before 1900,
but not before much of these first growth forests had been completely
decimated. Marketable size spruce and balsam (6" basal diameter) were cut
wherever they grew in profitable numbers. The last trees were taken frofi
the west side of the mountain which was least accessible and now shows the
least disruption. These despoiling practices and the prevalence of fires
throughout the Adirondacks prompted protective forest legislation in 1896.
In 1909 a fire tower was built on the summit of Whiteface. Nonetheless,
reports of fire and disturbance during the last 60 years are fragmentary
and somewhat contradictory. W. C. Petty, District Conservation Officer, .
states that there has been no appreciable fire damage since 1909, but
France and Lemon (1963) mention an extensive fire on the mountain in 1915.
Mr. Rogers, son of the mill owner in charge of the logging recalled that
Whitebrook Valley was burned about that time, and O'Kane (1928) confirms
this. Other long time residents, however, recalled no major fires after 1900.
Aerial photographs and field studies confirm a large burned area of uncertain
age at 2000 to 3000 ft on the northwest side, a large burn over much of the
east side from about 1500 to 3000, and several other smaller burned areas.
COLLECTION OF FIELD DATA
Stand Selection
The selection of stands was primarily based upon a systematic procedure.
It was arbitrarily decided to sample at500 foot altitude intervals over a
range of slope aspects and magnitudes. To reduce the field time promising
sites were “preselected" by examining aerial photographs and topographic
maps with the criterion that the photograph revealed that the site was not
markedly disturbed. If upon reaching the predetermined sites they were
judged to be recently disturbed by conditions such as logging, fire, wind
damage, etc., then the nearest "undisturbed" site was sought and sampled.
If none could be found in the immediate vicinity, the preselected location
was left unsampled,
Preliminary analysis of the first 2 years of stand data showed that
only small samples of certain species which are relatively common in the
study area, particularly at lower elevations, were included. Because white
pine (Pinus strobus)! and hemlock (Tsuga canadensis) were undersampled, these
species could not be well located on an ordination using the methods of
Curtis and McIntosh (1951). No systematic stand selection method easily
resolved this problem. Therefore, stands which contained these species were
sought and sampled. Consequently, our selection procedure was not entirely
systematic. Rather, because of the physiographic characteristics of the
study area, a combination of systematic and selective stand location was
employed.
Vegetation Data
Stands were sampled for density, basal area and frequency at points
20 paces apart. Normally, 10 to 40 points were used depending upon the
vegetation type. For example, it was not deemed worthwhile to obtain a
large number of points in stands containing a few species while larger
samples (30 to 40 points) were used in diverse vegetation.
The sampling points were paced along the slope. No change in course
was made because of the type of vegetation present. However, if the direc-
tion of the slope changed 45 degrees or more from the original point, the
1 - Classification is according to Fernald (1950).
-9-
course was altered. The sampling then proceded 2 points (40 paces) down-
hill and then in the reverse direction to the original course. The same
procedure was used when an area of disturbed vegetation, usually wind-throw,
was encountered. The stand definition was therefore based primarily upon
the topographic features of altitude and slope aspect rather than on
composition.
Measures of the density, frequency and basal area were obtained for
forested stands at elevations from just below 500 ft to above 4500 ft.
In 1964 the quarter method (Cottam & Curtis, 1956) was used. In 1965 and
1966, the quarter method was used for frequency, while Bitterlich prisms
(Grosenbaugh, 1952) were used to calculate basal areas and to establish
circular radius plots for density determinations. The efficiency of this
combination of methods has been cited by Lindsey et. al. (1958).
To ascertain a suitable radius to use for the circular plots, a test
plot was established in a beech-maple stand. There was a complete counting
of all trees and saplings within the test plot. The plot was then sampled
by a quarter method and also by the combination quarter and Bitterlich prism
method using 1/40 and 1/80 hectare plots. The 5, 10, 20, 30, and 50 factor
prisms were compared. Statistical comparisons of the results of these
sampling methods to the actual data for the test plot showed that the most
efficient method was the 1/80 hectare radius plot with the 30 factor prism
for density and basal area combined with the quarter method for determination
of frequency.
Values of sapling density and frequency were obtained by the quarter
method in all stands. Saplings in this. study were considered to be any stems
greater than 1 inch but less than 4 in dbh.
The frequency, density and basal area (dominance) values were relativ-
ized for both trees and saplings in order to determine the relative importance
value in percent for ease of interpretation and interstand comparison.
Ground flora was sampled by 1 square meter quadrats placed at each
quarter point within a stand. Frequency of herbaceous species and tree
seedlings was recorded.
In addition to the collection of quantitative data at each stand
location, the general features of the vegetation were described. Average
height of the canopy, degree of cover, uniformity, or heterogeneity or age
were estimated. Evidences of disturbance such as cut stumps, charcoal,
wind throw, and insect damage were also noted. Any unusual or distinguishing
characteristics of individual tree species were likewise noted.
Supplemental Data
Preliminary measurements of variables pertaining to the substrate
included observations of slope magnitude and direction, and distance to
the nearest ridge and drainage channel. Also measured were depth of litter,
fermentation and humus layers in the organic matter zone and of leached and
accumulation layers in the mineral zone when they could be readily
distinguished. Qualitative observations of soil texture and general descrip-
tion of the overall physiography of the site were made. Any unusual or
distinguishing site characteristics such as the presence of charcoal,
buried horizons, erratics, mounds, and troughs were noted.
-10-
The summit of Whiteface Mountain was used as a Weather Bureau
auxiliary station for a total of 8 years beginning in 1937. The
standard meteorologic variables have been measured year-round by the
ASRC staff at Marble Lodge Station and during the summer months at the
summit and various other locations on or near Whiteface. Since 1964,
net-radiation and solar radiation have been recorded year-round at
Marble Lodge and during the summers at the summit.
Topographic Properties of the Sample
Because of the non-random stand selection procedure there was no
guarantee that a representative sample of the vegetation was obtained.
Because some regions on Whiteface Mountain were more disturbed than
others, certain slope aspects and altitudes may have been undersampled.
A check on the representativeness of the sample was made by
comparing its topographic properties (altitude, slope-aspect and slope
magnitude) against those of a random selection of points on a map of
the study area, Because the area containing the entire sample was
rather large and not well defined only the stands on or in the close
proximity of Whiteface were used for the comparison. Random points
were then plotted on this same area and the topographic properties
determined for these points which did not fall on lakes or roads.
This comparison is given in Table 1 where percentages of the respective
samples are given for the three topographic properties. Data for the
total sample (182 stands), which includes much of the low-lying region
surrounding Whiteface Mountain, are also included in Table 1. The
vegetation sample of Whiteface-proximity area (156 stands) was over
represented by stands at high altitudes and under represented by sites
in the altitude ranges centered on 1500, 2000, and 2500 feet. The
total sample was probably more representative of the region but over-
sampled at the range centered on 1500 feet. The mean altitude of the
stands near Whiteface (2710 ft.) was higher than that of the total
sample (2310 ft.) and of the random points (2200 ft.).
The vegetation samples compared more favorably with the random
points in the slope-aspect property (Table 1) but north facing and
level sites were oversampled. Northeast and northwest sites were under-
sampled. Comparing the slopes of the sample with the random sites shows
an oversampling of steep slopes and undersampling of slopes in the
1 to 15 degree range.
TEST FOR HOMOGENEITY
The primary emphasis in this study was to describe the vegetation
of a mountain with a wide range of environments caused by topographic
variation. The selection of a stand was based upon topography and not
upon the species or groups of species contained within an area. No
effort was made to obtain homogeneous stands, but a test of homogeneity
was desired for later analyses of the data. Therefore, a chi-square
test was applied to all 182 stands. The test was the same as used by
“Te
Table 1: Percent of stands in various topographic groups for different
cases, Data in the first column are for the total sample of 182 stands,
in the second column for only those stands on or in the close proximity
to Whiteface Mountain (156 stands), and the third column for random
points on the map of the region in close proximity to Whiteface.
Altitude Total Whiteface Random
Range Sample Proximity Points
500 1.6 0.0 -
1000 6.5 0.0 5.6
1500 28.4 19.2 14.2
2000 19.7 19.2 38.6
2500 13.0 18.4 22,3
3000 8.7 12.0 9.6
3500 8.7 12.8 6.2
4000 9.8 13.6 3.5
4500 3.3 . 4.8 0.0
T00.0- 100.0 100.0
Mean altitude 2310 2710 2200
Slope-
aspect
Range
N 12.6 16.0 13,0
NE 8.2 7.2 16.1
E 7 8.0 9.9
SE 19.2 17.6 14.1
Ss 11.0 8.8 ted
SW 6.0 11.2 8.3
W 12.6 11.2 10.4
NW 12.6 12.0 19.3
Level 10.4 8.0_ 1.6
00.0 To0.0 Too.0-
Slope
Level 9.3 8.0 5.6
1-5 13.2 5.6 1.7
6-10 19.6 11.2 24.4
11-15 11.5 12.0 23.8
16-20 12.6 15.2 15.2
21-25 16.5 22.4 13.7
26-30 1.0 16.8 2.0
31-35 2.8 4.0 3.6
35 + 3.3 4.8 -
100.0 00.0 700.0
-12-
Curtis and Mc Intosh (1951) and Buel] et. al. (1966). - This is essentially
a test for heterogeneity. A stand was considered "homogeneous" if it did
not pass the chi-square test at the 5% confidence level. The results
showed that 42 stands were "non homogeneous". Inspection of the original
data usually showed that this heterogeneity was caused by clumping of
the major species in the stand although in some cases it was caused by
a change in composition along the transect.
COMPOSITION OF THE VEGETATION SAMPLE
Table 2 lists the 30 species reaching tree size diameters (4"dbh)
in forest stands sampled in this study. Mountain ash (Pyrus decora),
cherry (Prunus pensylvanica), mountain maple (Acer spicatum), striped
maple (A. paneylvanieiny; and ironwood (Ostrya virginiana) are included
in the Tist but seldom reach diameters o n. or more in the study
area and more frequently are less than.4 in. dbh at maturity. Consequently,
if included in discussions of forest dynamics, the high proportion of
"saplings" to "trees" might seem to indicate an increasing importance
of these species. Because field observation does not appear to support
this contention, these 5 species are considered as components of the
understory. ,
The remaining 25 species on the list more typically form trees with
diameters usually exceeding 10-12 in. at maturity. Twenty-two of these
occur 10 or more times in the forest sample, but only 14 of them are
stand dominants.
Three of the stand dominants, big-toothed aspen (Populus grandidentata),
basswood (Tilia americana), and red maple (Acer rubrum) are dominants
only once. The first two of these are not widely distributed in the
area and are found in only 17 and 27 stands, respectively, with mean
RIVs (Relative Importance Values) of 6% and 7% in those stands of occurrence.
Red maple is more widely distributed, being present in 70 stands, but
with a mean RIV of only 6% it is not of great importance in the sample.
Basswood and red maple are frequent associates of many of the mature
vegetative groupings in the area, while big-toothed aspen is indicative
of disturbance conditions in its sites of occurrence. The lack of recent
disturbance of stands in the sample is indicated by the presence of the
latter species in only 5 stands as a sapling with a mean RIY of 3% per
stand of occurrence.
The 7 most common trees in the area are red spruce (Picea rubens),
143 stands; balsam fir (Abies balsamea), 120 stands; yellow birc
(Betula allegheniensis), 86 stands; sugar maple (Acer saccharum), 80
stands; paper birch (Betula papyrifera), 75 stands; American beech
(Fagus grandifolia), 72 stapes and cordate-leaved birch (Betula
papyrifera var. cordifolia), 63 stands. All but one of these, paper
birch, occurs as a stand dominant 4 or more times.
Paper birch is a common species in the Adirondacks and is dominant
over several tracts within the study area. However, this dominance
occurs in stands of obvious disturbance. These were not sampled because
of the disturbance. In the sites which have been long undisturbed, this
species plays only a minor role. Paper birch occurs 75 times in stands
Table 2: Number of stands of occurrence, occurrence as leading dominant, mean RIV (mean relative importance
value) for the total sample and for stands of occurrence for tree species in the 182 stand sample. Numbers
are rounded to the nearest whole unit of percent. T stands for "trace" or less than 0.5% mean RIV.
Tree Species Species No. Stands No. Stands Mean RIV Mean RIV
Abbre- of Occur. as Domin. Tot.Samp. Std. Occ.
viation ree/Sap. Tree/Sap. Tree/Sap. Tree/Sap.
Picea rubens Sarg. ........eseeeeee eee eeeeeeeeee Pr 143/127 30/19 16/11 20/15
Abies balsamea (L.) Mill. ..... g Ab 120/115 45/71 21/30 31/47
Betula alleghaniesis B & B Small. Ba 86/51 10/0 6/2 13/7
Acer saccharum Marsh. ......... As 80/80 28/37 11/2 26/87
Betula papyrifera Marsh. Bp 75/41 0/1 3/2 7/8
Fagus grandifolia Ehrh. Fg 72/75 10/15 5/8 14/18
Acer rubrum L. Ar 70/56 14 2/2 6/8
Betula papyrifera var. cordifolia (Regel) Fern.. Bpc 63/47 4/1 6/3 17/12
Tsuga canadensis (L.) Carr. ......seeeeeeeceeeee Te 49/37 16/12 6/4 23/19
Fraxinus americana L. ..... ‘ Fa 39/19 0/0 Wl 7/6
Pinus strobus L. ..... :. Ps 36/22 10/0 5/1 26/9
Acer pensyivanicum L. ........ Apen *37/81 0/0 V7 4/16
Ostrya virginiana (Mill) K. Koc Ov *32/34 0/0 1/3 5/17 1
Pinus resinosa Ait. ........... . Pres 30/22 15/6 7/2 38/18 @
Thuja occidentalis L. ... : To 28/15 5/3 2/1 15/17 ‘
Pyrus decora (Sarg) Hyland..........-..--0- Pd *29/30 0/0 aval 4/6
Quercus rubra var. borealis (Michx. f) Farw..... Or 27/20 6/3 3/1 17/10
Tilia americana Le ... 2... eee e eee eee as Ta 27/15 1/2 ava 6/6
Prunus serotina Ehrh. . az Pser 20/12 0/0 T/T 43
Populus grandidentata Mich. . Par 17/6 1/0 1/0 6/3
Acer spicatum Lam. .......... Aspic *16/52 0/0 T/3 3/10
Populus tremuloides Michx. oe Pt 15/4 0/0 T/T 3/3
Ulmus americana L. ........ wi Ua 6/1 0/0 0/6 1/3
Populus balsamifera L. . oi Pb 4/0 0/0 0/0 2/0
Fraxinis nigra Marsh. ... Fn Si2 0/0 0/0 2/3
Juniperus virginiana L. . Jv 2/2 0/0 0/0 3/2
Quercus. bicolor Willd. . Qb 2/2 0/0 T/T 8/6
Ulmus rubra.........-.-.----- Ur 1/0 0/0 0/0 1/0
Larix laricina (DuRoi) K. Koch. oe L1 1/2 0/0 0/0 1/73
Prunus pensylvanica Le ..... cess ee ee ence ee eeee Pp *1/3 0/0 O/T T/4
*Typically understory species only occasionally reaching tree diameter (4"dbh) or greater, so not considered
to be leading sapling even if leading in RIV.
-14-
as a tree, but in only 45 as a sapling with RIVs of 7.1% and 7.5%,
respectively. These data support the pioneering nature of the species.
Cordate-leaved birch is a high altitude variety of paper birch.
It dominates. only 4 stands in the sample, yet it covers large tracts
of area on the mountain. This area was not sampled because of
disturbance due to such factors as fire, logging and wind throw. In
these areas there is every indication that cordate-leaved birch is
being replaced by spruce and fir: The four stands in which the species
did dominate did not show these obvious signs of disturbance. Even
here, however, in all four stands balsam fir is the leading sapling and
the birch has a relatively low sapling RIV. Nevertheless, the cordate-
leaved birch does appear to be a permanent minor associate of certain
high altitude sites within the boreal complex. This is perhaps because
the severity of the climate in these sites maintains disturbance-like
conditions which typically favor the species.
Yellow birch shows a marked decline in sapling presence from 86
stands as a tree to 51 as a sapling. Unlike the other birches, it is
a climax species in certain of the mature vegetation associations of
the Whiteface area. As a tree it is extremely tolerant and survives
easily in the deep shade of the mature hardwood forests. Its poor
representation as a sapling is perhaps best explained by its substrate
specific germination requirements. Kujawski and Lemon (1969) found
that seedlings become established only on exposed mineral soil or
mixed humus-mineral soil. Roots of seedling birch do not penetrate
the leaf litter of the hardwood stands. Seeds typically find few
such places suitable for germination in the vigorous, pre-degenerate
forest stands of the area, so saplings are limited even though several
mature, prolific seed-producing trees may be found in the stand.
The data for American beech give the impression that this species
is increasing in its importance. It occurs in 72 stands as a tree and
in 75 stands as a sapling with mean RIVs of 14% and 18%, respectively,
per stand of occurrence. These data may be misleading because beech
reproduces abundantly from root suckers which give.the impression of
vigorous regeneration. However, many of these suffer a high mortality
rate in the "sapling" and young tree stages. Field observations indicate
that beech may be in fact decreasing in importance as a tree in the
majority of stands in which it is found.
Balsam fir, red spruce, and sugar maple are the most important
species in the "undisturbed" sites of the study area. One of these 3
species occurs in all but 9 of the 182 stands sampled and as a dominant
in 103 stands. Balsam fir is first in dominance with 45 stands, red
spruce is second with 30, and sugar maple, which occurs in 65 less
stands than red spruce, is third with 28.
Balsam fir is most common at the higher elevations, sometimes
occurring in pure stands, but more frequently in association with red
spruce. Red spruce assumes its greatest dominance at altitudes
intermittant to those dominated by balsam fir and by sugar maple and
is frequently associated with both of these species. Sugar maple
reaches its highest importance in the 1500-2500 foot elevation range.
Figure 2 shows the relationship of these 3 species in stands
dominated by any 1 of the 3 species. Only 8 of the 103 stands contain
-15-
t Table 3: Numbers of stands of occurrence and stands as a dominant for
the three leading stand dominants by altitude intervals.
Elevational Number of Abies balsamea Picea rubens Acer saccharum
Range Stands Occurs Dom Occurs ~ Dom Occurs — Dom
250-749 3 0 0 0 ) 0 0
750-1249 i 2 0 3 0 3 0
1250-1749 52 30 3 38 6 38 13
i 1750-2249 36 16 1 31 5 26 8
f 2250-2749 24 14 2 22 8 am 7
; 2750-3249 16 16 3 16 8 2 0
' 3250-3749 16 7 12 15 3 0 0
} 3750-4249 18 18 18 14 0 0 0
I 4250-4749 6 6 6 4 0 0 0
; 4750-4868 Oo. 0 0 0 0 0 0
| Totals 182 “T2045 Was 30 80 8
Table 4: Number of occurrences of the various species in groups of
i stands of leading RIV. Arrangement of species is according to the
mean altitude of stand dominants (Figure 4).
Ab Bpc Pr Ba As Fg To Tc Pres Ta Pgr Ar Qr Ps Total
: Ab 64506-6406 (28 (99 C7 4 U5 - - 1 - 2 120
i Bee 37 4 16 2 1 - 2 «2 - - - - os 63
{ Pr 38 3 30 10 16 9 5 16 11 - - - - 4 143
| Ba 7 1 #6 10 20 10 3 15 - lo - 1 - - 86
As - + 8 “7 28 10 - 13 2 1 1 1 5 3 80
Fg - - 5 7 2% 10 - 15 - 191 1 4 - 72
To 1 - 8 2 - T 5 6 3 - - - - 2 28
Te 1 - 10 3 7 3 2 16 5 - - - - 2 49
Pres - - 3 = - + = T 15 - - 1 10 30
Ta - - - 13 2 - 4 T J - - 5 1 27
j Pgr - + 3 - 5 - = = 2 5 1 1 4 1 7
i Ar 6 - 12 4 11 4 4 12 8 - T 1 3 4 70
Qr - - - - 8 2 - 1 6 - 1 - 6 3 27
Ps 1 - 9 - To - 1 2°11 - - - T 10 36
Figure 2:
Figure 3:
Figure 4:
-16-
Association relationships of the three leading
stand dominants (Abbr. explained in Table 2) (page 17).
Numbers of species, stand dominants and stand by
500 foot intervals of altitude (page 18).
Mean altitude of the 14 leading stand dominants
(page 19).
NUMBER
30
pe)
t
fo)
(3)
(I)
|
(ee)
Ge) (ea) UB)
( )= NO STANDS IN
ALTITUDE RANGE
@—e SPECIES
O—d STAND DOMINANTS
tle)
(6)
1000
2000 3000
ALTITUDE (FT.)
4000
-81-
ALTITUDE FEET
-19-
4000
3000
2000
POSITION OF DOMINANT
~20-
all 3 species. The most frequent association, as expected, was. between
balsam fir and red spruce, Red spruce was an associate in 38 of the 45
balsam fir dominated stands, and balsam fir was associated with 28 of
the 30 red spruce stands, Red spruce was found in 16 of the 28 sugar
maple dominated. stands, but sugar maple was only found in 8 of the 30
red spruce stands. The spruce-maple relationship is not surprising
because red spruce ranges well above the elevational limits of sugar
“maple, but is not restricted to higher altitudes, The altitudinal range
of red spruce in the sample is from 4500 feet at the upper limits in
a balsam fir stand to 980 feet at the lower level in a white pine
(P. strobus) stand (Table 3). The maximum altitude at which sugar maple
was found was 3100 feet in an east facing yellow birch stand.
The lowest degree of association of the 3 major species was between
balsam fir and sugar maple. None of the balsam fir stands contained
maple while 8 sugar maple stands contained balsam fir.
The tree to sapling ratios of balsam fir and sugar maple in Table 2
suggests that they both will increase. The ratio of occurrences for
fir is 120/115 but the ratio of mean RIVs is 30/47%. The same ratios
for sugar maple are 80/80 by occurrence and 12/27% by mean RIV. The
increase of dominance of fir with increase in altitude is evident from
Table 3. Field evidence of low mortality for sugar maple saplings
indicates that it is strengthening its position as a dominant in middle
and low altitude forests.
Red spruce occurrences are 143/127, with mean RIVs of 20% and 15%.
Both these data and the field observations seem to indicate that balsam
fir is a better competitor on most sites of mutual occurrence. For
example, balsam fir is the leading sapling in 15 of the 30 red spruce
dominated stands, while red spruce has gained leading sapling status
in only 3 yellow birch and 2 red pine (P. resinosa) stands. This would
indicate that while red spruce may remain an important associate, it
may decrease as a dominant in the study area. On the other hand, field
evidence shows that balsam fir has a high mortality rate as a young
tree and is especially subject to wind and icing damage, while spruce
tends to resist destruction and lives longer than fir. Thus, the
relative dominance of these two species may not be changing as much as
data in Table 2 indicate.
The 5 remaining stand dominants in the sample are hemlock (Tsuga
canadensis) in 16 stands, red pine in 15 stands, white pine in 1
Stands, northern red oak (Quercus rubra) in 6 stands, and eastern white
cedar (Thuja occidentalis) tn 5 stands. The lower number of stands
dominati y these species compared to balsam fir, red spruce, and
sugar maple is probably attributable to the lack of special sites which
these species occupy or to past disturbance. For example, there is a
suggestion from the early literature and from discussion with the older
local residents that hemlock was once much more prevalent than now.
Apparently, clearing of the land for agriculture and lumbering caused
extensive reduction of hemlock in the more accessible sites. Possibly
as a result of disturbance hemlock seems to be split into two major
habitat sites. One is the low level wetlands and stream beds where it
mixes abundantly with red spruce and balsam fir. The more common case
seems to be lowland sites on small rounded drumlins or eskers with sandy
-21-
or gravelly well-drained soils. On the latter sites hemlock often forms
extensive nearly pure stands with abundant regeneration under trees of
all ages and sizes. Hemlock’ is also a common minor associate in hemlock-
yellow birch-sugar maple-beech stands of mesic sites.
White pine is predominantely a species of the low elevations of
the area. It is generally found in nearly pure stands or associated
with red pine on level or gently sloping sites with sandy soils. These
“soils often have a hardpan and are apt to be quite wet in the early
growing season but very dry during the rest of the season. Lowland sites
of this type are not common in the area. Again, logging and land clearing
have probably greatly reduced its abundance compared to former times.
Red pine is found repeatedly on windswept, poorly drained ridges
with shallow soils at the mid-elevations, and on sandy well-drained soils
both on the level and on steep ridges at the lower elevations. It is
without doubt the most drought tolerant of the dominant tree species
of the area.
Northern red oak is also a lower elevation species associated with
gentle to moderate slopes, usually on east facing sites. The soils
in the oak stands are more mesic than those in the pine stands but too
dry or rocky to support the sugar maple, American beech, and yellow
birch forests.
Eastern white cedar has a distribution which is difficult to
interpret. It is found in level wet areas as a dominant, but also appears
on ridge tops up to 2500 feet or more in red pine or red spruce stands
where soils are quite shallow and apparently quite dry. Habeck (1958)
has studied this species in transplant gardens in Wisconsin and concludes
there are two distinct ecotypes based on similar site preferences.
THE RELATION OF FOREST COMPOSITION TO TOPOGRAPHY
Altitude Variation
In an area of physiographic diversity such as the Whiteface Mountain
it is expected that the severity of climatic and edaphic parameters would
be greatest at the extremes of elevational range, and hence, the numbers
of species able to occupy the habitat sites at these extremes would be
reduced. Figure 3 shows this relationship of species by altitude groups
as they occur on Whiteface Mountain and in its vicinity. The 1500 foot
level is the area of maximum diversity for both numbers of species and
numbers of stand types. It is also the altitude of highest sampling
frequency. Much of the basal portion of Whiteface is between 1200 and
2000 feet. The small amount of area partly accounts for the reduction
in number of species and stand dominants at the 500 and 4500 foot ranges,
but the more severe habitat variables to be found at these elevations
must also be considered.
Figure 4 shows the mean altitude of the species occurring as leading
stand dominants. Casual observation of this figure seems to imply a
fairly smooth transition from species to species through the eievational
range of Whiteface Mountain with the hint of a possible natural clumping
Figure 5:
Figure 6:
aD
Association relationships of stand dominants
representing boreal (a), mesic-hardwood (b), and
lowland oak and pine (c) associations (page 23).
Association relationships of coniferous stand
dominants (page 24).
c. LOWLAND
OAK _AND
PINE
-24-
Ab
: FIR-HEMLOCK-PINE
CEDAR
HEMLOCK-
SPRUCE
9 2
To
DAR-PINE
CEDAR-
HEMLOCK-PINE
6 2
Z ‘ Tc Z \ Ps
Pres .
Pr Te
SPRUCE -PINE
HEMLOCK- PINE
4 5 2
Pres Ps Pres Ps
Figure 7: Association relationships of selected coniferous
and hardwood stand dominants (page 26).
a Figure 8: Association relationships of a boreal (yellow birch),
| : a mesic-hardwood (sugar maple) and a lowland hard-
t wood (northern red oak) hardwood stand dominant (page 27).
-27~
~28-
of species groups. Only red pine seems to be out of place. This is
due to its occupation of the dry, windswept ridges in the 1500-2500
foot range on the south side of eastward pointing ridges of the mountain.
Table 4, which has the species ordered primarily by their RIVs by
altitude, shows the constancy of association of stand dominants with
one another. The data in this table seem to further support this idea
of natural clumping of the species into altitude related groups. The
‘strong positive associative values of balsam fir, red spruce, cordate-
leaved birch, and, perhaps, yellow birch suggests a boreal element.
The association between sugar maple, American beech, yellow birch, and
perhaps hemlock indicates a mesic hardwood element. This scheme places
yellow birch in a dual role, but field observation does indicate that
the species is a frequent associate with both groups although its
affinity appears to be somewhat stronger towards the mesic hardwoods.
The red spruce shows a strong affinity to dominants of the mesic-
hardwood range. It occurs in 16 of the 28 sugar maple stands, in 9
of the 10 American beech stands, and in all 16 of the hemlock stands.
The remaining stand dominants are northern red oak and the white pine
and red pine stand types found typically in the lower elevations of
the area. These constitute the dry oak hardwoods and the pine associations
representative of more southerly latitudes.
Frequent associations are interesting and descriptive of vegetative
composition. Equally interesting, however, are cases of vegetative dis-
association. For instance, data in Table 4 show that neither American
beech nor yellow birch were ever found to associate with red pine or
white pine in any of the 45 stands that one or the other of these 4
species dominated. In fact, there is a very weak relationship of the
pines to all hardwoods in the study area. More than elevational gradient
is involved here, because several of the red pine stands are at altitudes
well within the major dominance range of American beech and its common
associate yellow birch. Variations in moisture may be responsible.
A final point of interest from Table 4 is the wide amplitudes of
some of these species, Red maple and sugar maple, for example, occur
with 11 of the other 13 stand dominant species, balsam fir occurs in
10 of the 13, and red spruce and yellow birch in 9 of the 13. Three
of these are the leading stand dominants. One of these, red maple, is
only a stand dominant once and is of relatively minor importance in all
associations in which it occurs, yet it is a very wide ranging species.
This would suggest that something more than site tolerance is involved
in determining the abundance and dominance potential of a species.
Figure 5 shows the association relationships of the major species
of the vegetative groupings identified above as boreal, mesic-hardwood
and the oak and pine associations. Within the boreal element (Figure 5a)
the weakest association is between cordate-leaved birch and spruce in
spruce stands. This is best explained by the occurrence of several red
spruce stands at altitudes below the lower limits of the birch. The
constancy is 100% in the other direction indicating probable succession
of the birch stands toward a balsam fir-red spruce climax.
All 3 of the species representing the mesic-hardwood grouping (Fig. 5b)
show high degree of association with one another, for example, between
sugar maple and yellow birch in American beech stands. There is 100%
-29-
association of these 2 species in the beech dominated stands. Associations
in the other directions are nearly as strong.
From Figure 5c it can be.seen that the northern red oak stands are
primarily independent of white pine and red pine. The only northern red
oak stand which contained the two pines was a gentle east-facing slope
at 1180 feet. Oak, on the other hand, Sppenrs in about one third of
the pine dominated stands showing a somewhat closer affinity to ted pine.
‘This agrees favorably with the site differences of the two pines. Red
pine often is the dominant on the ridges capping the lower east facing
slopes where northern red oak commonly occurs.
Figure 6 shows some of the association relationships of the various
conifers. Using red spruce as the indicator it appears as though the
boreal conifers associate conmonly with the other 4 conifers of the area,
while the pines show little affinity to hemlock and eastern white cedar.
The relationship of hemlock to cedar is stronger than that of either
species to the pines and about equal in both directions.
Figure 7 shows some hardwood-conifer association values. Some associa-
tions are high, yellow birch and sugar maple to hemlock; and some are low,
sugar maple and yellow birch to white pine. The degree of relationship
of hardwood to conifer is likely most closely tied to the soil moisture
requirements of the species. Hemlock does well in the mesic environments
of sugar maple and yellow birch, while these latter species do poorly
in the dry soils on which white pine typically occurs. Balsam fir and
red spruce are seen as very common associates in yellow birch stands,
even when these occur at relatively low elevations.
Figure 8 is a comparison of the associative values of a somewhat
boreal hardwood (yellow birch), a mesic-hardwood (sugar maple), and a
lowland hardwood (northern red oak). There is no association between
yellow birch and northern red oak in any of the 16 stands dominated by
‘them. The association of sugar maple to both yellow birch and northern
red oak is a strong one, as is the association of yellow birch in sugar
maple stands. Affinity of northern red oak in sugar maple stands is
much weaker. All 3 species only occurred together in 4 of the 44 stands
dominated by any one of them. In all cases these were sugar maple
dominated stands.
A closer evaluation of this relationship of the stand dominants
to the altitudinal gradient is afforded by looking at the mean RIVs of
the altitude groups. In Table 5, 10 altitude groups based on a 500
foot interval have been established and the mean RIVs of the 14 stand
dominants for each of the altitude groups is presented, both for trees
and for saplings.
This information reveals the characteristics of the distribution
of the major tree species of the area. Balsam fir is the most important
species of the higher elevations, but also has a considerable range of
importance, and if the sapling data are reliable indicators, is likely
to become even more important in the mid-range elevations. Sugar maple
is the dominant tree of the mid-range elevations and, likewise, by sapling
data seems to be increasing its importance and range.
Red pine appears to be about holding its position, but northern
red oak and white pine ‘seem to be decreasing in importance. In general,
the trends of the rest of the species are evident from Table 5.
Table 5: Mean RIV of leading stand dominants (and number of stands) by altitude groups.
A. Trees
Altitude
250-749
750-1249
1250-1749
1750-2249
2250-2749
2750-3249
3250-3749
3750-4249
4250-4749
4750-4868
Mean all stands
B. Saplings*
Altitude
250-749
750-1249
1250-1749
1750-2249
2250-2749
2750-3249
3250-3749
3750-4249
4250-4749
4750-4868
Mean all stands
Ab Pr
52(3) 42(6)
52(1) 48(5)
43(2) 56(8)
45(3) 51(8)
57(12) 52(3)
8i(ig) --
88(6) --
70(45) 50(30)
Ab. Pr
58(2) --
42(12) 50(4)
54(3) 38(7)
44(6) 39(5)
57(11) 58(2)
72(14) 43(1)
90(17) --
91{6) --
66(71) 43(19)
34(1)
==(0)
As
42(13)
46(8)
61(7)
48(37)
Fg
39(4)
34(6)
36(10)
Fg
51(10)
37(5)
46(15)
58(16)
Te
43(11)
4n(1)
43(12)
50(3)
40(6)
24(3)
20(2)
Ar
40(2)
33(2)
36(4)
=-(0)
*Values will tend to be low because of inclusion of Acer spicatum, A. pensylvanicum, Pyrus decora, Ostrya virginiana,
and Prunus pensylvanicum as’ saplings in sample.
“31-
Slope-aspect Variation
The change to a more boreal vegetation with increase in altitude
is a well-known feature of mountain
slope-aspect variations are Tess obvious. To study the variation in
vegetation composition with slope-aspect, only data from the stands on
or near Whiteface Mountain are used and those below 1250 ft. are omitted.
or hemlock). The selection procedure for stands on or near Whiteface
conforms to a systematic procedure with rare exception (see METHODS section).
constant. Weighting by the actual number of stands in the slope-aspect
range overweights for species of high importance in altitude ranges with
sites followed by west and lowest on east slopes. Red spruce, also a
boreal species, js very high in RIV on west slopes but low on north
A summary of the mean importance of single species may lead to some
inconsistencies due to Some peculiar nature of a given species or because
the sample size is too small. For this reason, the data are also summarized
as "species groups" in Table 6b. The boreal group consists of balsam fir,
red spruce, cordate-leaved birch, and mountain ash. The hemlock-northern
hardwoods group consists of hemlock, Sugar maple, American beech, and
yellow birch. The dry hardwoods group contains northern red oak, basswood,
and white ash, while the pines are red pine and white pine...
Only one important Species, eastern white cedar, is omitted from
the species groups because when it is found in a stand it is usually
by far the most important species. These “cedar swamps" seem to be a
special type and cannot be logically lumped into one of the four groups
even though they have more boreal associates than others. Minor species
containing less than 1%- mean relative importance of the total sample
are also omitted from the species in Table 6b.
-32-
Table 6: Mean RIV of eight major species (a) and groups of species (b)
for the altitude range 1250 to 4749 feet for the four major quadrants
of slope-aspect. Data are means of seven 500 foot ranges in altitude -
with each range weighted equally (see text).
East South West North
(a) Species 45°-134° =: 135°-224° 225°-314° 315°-44°
Abies balsamea 26.1 35.0 38.3 41.5
Picea rubens 17.2 17.3 23,8 14.6
Betula alleghaniensis 3.3 4.4 6.1 9.2
Acer saccharum 16.0 13.9 5.2 9.4
Tsuga canadensis 2.3 0.6 3.8 2.7
Fagus ‘grandifolia 6.9 2uf 4,7 5.8
Quercus rubra 4.2 0.2 -- --
Pinus resinosa -- 3.5 0.5 11
Number of stands 25 26 26 33
(b) Species Group
Boreal 56.1 63.0 69.0 65.0
Hemlock-Northern Hardwoods 28.5 21.6 19.8 27.1
Dry Hardwoods 6.4 1.1 0.3 1.7
Pines -- 4.2 0.8 11
Table 7: Mean RIV of eight major species (a) and groups of species (b)
for altitudes 1250 to 4749 feet grouped by 180 degree ranges in slope aspect.
Data are means of seven 500 foot altitude ranges with each range weighted
equally (see text).
Relative Importance of 180° Slope-Aspect Groupings
East vs. West North vs. South
(a) Species. 1°-180° 181°-360° 271°-90° 91°-270°
Abies balsamea 26.8 40.4 34.6 32.6
Picea rubens 13.8 19.8 14.2 19.4
Betula alleghaniensis 6.7 6.9 7.4 6.2
Acer saccharum 18.1 4.0 12.5 9.6
Tsuga canadensis 2.9 2.2 3.0 2.1
Fagus grandifolia 6.5 6.2 6.3 6.4
Quercus rubra 2.3 -- 1.0 1.3
Pinus resinosa 1.0 4.2 2.2 3.0
Number of stands 54 58 54 58
(b) Species Group
Boreal (WBC) . 53.9 68.7 58.4 64.2
Hemlock-Northern Hardwoods 34.2 19,3 29.2 24.3
Dry Hardwoods 4.0 0.3 Zot 2.2
Pines 1.0 4.6 2.2 3.4
~33-
Table 8: Mean RIV of eight major species (a) and groups of species (b)
for the altitude range 1250 to 2749 feet grouped by four quadrants of
slope aspect plus level stands. Data are means of three 500 foot
altitude ranges with each range weighted equally.
EAST SOUTH WEST NORTH LEVEL
(a) Species 45°-134° 135°-224° 225°-314° 315°-44°
Abies balsamea 0.3 2.8 3.3 8.1 13.0
Picea rubens 2.6 12.6 24.1 13.6 15.1
Betula alleghaniensis 4.4 8.4 13.7 17.8 9.9
Acer saccharum 37.4 32.5 12.2 21.9 22.1
Fagus grandifolia 16.1 6.3 11.1 13.5 4.6
Tsuga canadensis 5.4 1.5 9.0 6.4 1.8
Thuja occidentalis - 0.7 0.4 6.6 36.7
Pinus resinosa - 8.2 1.2 2.6 0.1
Number of stands YW 16 7 14 u
(b) Species Group
Boreal (WBC) 2.9 15.8 27.4 23.5 28.1
Hemlock-Northern
hardwoods 63.3 48.7 46.0 59.6 22.7
Dry hardwoods 15.0 2.9 0.6 4.0 1.0
Pines - 9.8 1.9 2.6 0.8
odds
The reason for the difference between presenting the data for 90°
slope-aspect ranges versus 180° ranges for sugar maple and red spruce
is clarified by looking at Table 10, Here the same data for 4 species
are divided into 90° slope-aspect ranges centered on NE, SE, NW, and
SW. Sugar maple is high only on NE and SE slopes and red spruce only
SW and NW.
Plotted data for species groups in Figure 10 again shows that the
east slopes contain species which indicate a relatively mild environment
compared to the west slopes. The contrast between north and south slopes
is not nearly so great in these lower altitude northern hardwood forests.
Table 8 also indicates data for the level sites. Poorly-drained
level sites contain spruce-fir and cedar stands and well-drained sites
contain sugar maple and other hardwoods. The high relative importance
of boreal species and eastern white cedar may be due to an oversampling
of level wet sites which have not been logged compared to well-drained
level sites which may have been logged for hardwood timber and thus not
sampled when visited.
The forests on the lower slopes of Whiteface contain pines on south
and southwest sites, but not on east facing sites (Tables 8-10). On
the latter there are dry hardwoods including several stands of northern
red oak (Figure 10). The dry hardwoods are not typically found on west-
facing slopes.
Stands above 2750 feet rarely contain species which were not in
the boreal group. Several stands between 2750 and 3500 feet contain
significant amounts of yellow birch and small amounts of mountain maple
and striped maple. Qne northeast facing site at 3100 feet contains
29% importance of sugar maple which is the highest altitude at which
this species was found. An east facing stand at 3350 feet was dominated
by large trees of yellow birch which occurred in scattered amounts up
to about 3450 feet.
Figure 11 is a bar graph of the mean RIVs of boreal species for
4 quadrants and 2 altitude groups. The upper altitude (3750 to 4749 ft.)
shown in Figure 11a contained mostly balsam fir. This species was highest
on north sites followed by west and lowest on east slopes. Red spruce,
cordate-leaved birch, and species listed as "others", including mostly
yellow birch and mountain ash, were least important on west slopes again
indicating the most severe environment is on west slopes.
In summarizing the slope-aspect compositional variation, a significant
feature is that except in the very high altitude range the western slopes
were the most boreal of all aspects indicating the most rigorous environ-
ment. The contrast between east and west sites was much larger than
between north and south. The east facing slopes contained species which
indicate that the mildest environment occurs on those slopes. South and
southwest sites contained relatively high composition of "dry" indicating
species while "mesic" hemlock-northern hardwoods were highest on east
and north facing sites.
~35-
Table 9: Mean RIV of seven "indicator" species (a) and groups of
species (b) for the altitude range 1250 to 2749 feet grouped by
180° ranges in slope aspect. Data are means of three 500 foot
altitude ranges with each range weighted equally.
EAST vs. WEST NORTH vs. SOUTH
(a) Species 1°~180° _181°-360° 271°-90° _91°-270°
Abies balsamea 2.9 6.2 6.1 3.0
Picea rubens 6.3 21.2 11.5 16.0
Betula alleghaniensis 10.1 14.4 12.5 12.0
Acer saccharum 36.7 9.3 23.6 22.4
Tsuga canadensis 6.4 5.3 6.6 6.1
Fagus grandifolia 14.1 14.5 13.6 15.0
Pinus resinosa 2.4 9.9 5.2 71
Number of stands 27 31 26 32
(b) Species Groups,
Boreal (WBC) 9.3 31.2 20.4 20.1
Hemlock-Northern
Hardwoods 67.3 43.3 56.3 54.3
Dry hardwoods 8.8 0.9 4.3 5.4
Pines 2.5 na 5.5 8.1
~36-
Table 10: Mean RIV of four major species for the altitude range 1250 to .
2749 feet grouped by four quadrants of slope aspect. Data are means of _
three 500 foot altitude ranges with each range weighted equally.
NORTH= SOUTH- SOUTH~ NORTH~
EAST EAST WEST WEST
Species 1°-90° 91°-180° 181°-270° 271°-360°
Abies balsamea 4.1 1.8 4.3 8.2
Picea rubens 5.4 7.2 24.7 7.7
Acer saccharum 36.9 36.5 8.4 10.3
Fagus grandifolia 16.6 11.6 18.4 10.6
Number of stands am 16 W7 14
Table 11: Mean RIV of several species (a) and groups of species (b) for
all altitudes grouped according to steepness of slope. Data are means
of stands occurring in slope range regardless of altitude.
Slope Aspect Range in Degrees
(a) Species Less than] Jto10 Jl to 20 20 to 30 30+
Abies balsamea 17.4 20.2 16.8 35.8 61.7
Picea rubens 16.3 10.6 17.0 23.6 23.1
Betula alleghaniensis 6.3 14.5 8.0 5.3 -*
Fagus grandifolia 7.8 11.0 8.1 2.3 -
Tsuga canadensis 0.3 9.9 3.2 0.8 ~*
Acer saccharum W.7 19.5 19.0 6.1 -
Quercus rubra - 0.1 3.6 - -
Pinus resinosa 0.1 3.2 2.9 4.9 -*
Number of stands nv 25 33 45 16
Boreal 33.7 31.3 34.6 60.6 85.6
Hemlock-Northern
hardwoods 26.1 54.9 38.3 12.5 -*
Dry hardwoods 2.1 0.9 6.9 0.9 -
Pines 0.8 3.2 3.3 5.7 *
*occurred in this range
~37~
Data from Table 6b are also plotted as bar graphs in Figure 9a.
West slopes tend to be the most boreal followed by north aspects while
east slopes have the highest importance values for the northern hardwood
species indicating a relatively mild environment on the east aspects.
Pines are most abundant on south facing slopes although they are of
relatively low total importance in this sample because stands below
1250 ft. are omitted. Dry hardwoods are most important on east aspects.
Because of the interesting result that the west to east contrast
appears to be more pronounced than the north to south one, the data are
examined in a different way in Table 7 and Figure 9b. In this case,
the sample of stands on or near Whiteface proper is divided first into
east and west aspects only and then into north and south only. Comparing
the aspects in this way shows a greater contrast between east and west,
but north slopes do not appear to be more boreal than south slopes.
South slopes contain more red spruce but only slightly less balsam fir
(Table 8a). North slopes have more hemlock-northern hardwoods species
than the south, but lower amounts of pines.
The contrast between east and west facing slopes is shown by much
higher importance of sugar maple and lower importance of balsam fir
and red spruce on the east sites. Dry hardwoods are important on the
east and pines on the west slopes. The high contrast between east and
west and lower contrast between north and south sites is not easily
explained. Subjective inspection of stand data shows that there are
not only many protected northeast facing stands of sugar maple and
American beech on Whiteface Mountain but also many stands (even at lower
altitudes) on south-west sites with high amounts of red spruce and
balsam fir.
Casual inspection of the stand data at high altitudes indicates
that the high contrast between east and west sites may be due to
variations at lower elevations. The sample of stands between 1250
and 2749 feet consisting mainly of hemlock-northern hardwood forests
is treated in the same manner as in the previous section. Table 8
presents mean importance values for 7 major species and the 4 species
groups for 4 90° slope-aspect groups plus level stands. Table 9 presents
similar data, but they are divided into 180° groups to compare north
versus south and east versus west. Figure 10 is a plot of the data for
the species groups from Tables 8b and 9b.
The mean RIV of red spruce in Table 8 is much higher for the west
slopes (24%) compared to the east (3%) while it is nearly the same on
north and south slopes. However, in Table 9 when 180 degree slope-
aspect ranges are examined instead. of quadrants the south stands have
higher mean RIVs for red spruce than north sites (16% vs 12%). The high
difference between east and west is still evident (6% vs 21%) with west
facing sites apparently more boreal.
The data for sugar maple in Table 8 shows the reverse effect. When
the summary is by quadrants, the mean RIV of sugar maple is 33% for
south slopes compared to 22% for north slopes, but in Table 9 (summarized
by 180° ranges), they are nearly the same. Similarly for sugar maple the
high variation between east and west was evident in both Tables 8 and 9
with eastern sites having much higher values for this hardwood species.
Figure 9: Mean RIVs of four species groups for (a) four 90°
slope-aspect quadrants (a) and 180° slope-aspect
for the altitude range (b) from 1250 to 4749 feet (page 39).
Figure 10: Mean RIVs of four species groups for four 90° slope-
aspect quadrants (a) and 180° ranges of slope-aspect
(b) for the altitude range from 1250 to 2749 feet (page 40).
Figure 11: Mean RIVs of tree species for four quadrants of
slope-aspect for altitude ranges of 3750-4749 feet
(a) and 2750-3749 feet (b) (page 41).
MEAN REL. IMPORTANCE
MEAN REL. IMPORTANCE
“39+
60-L =
40k - |
20F |: 5
re) ae 1 RSH va eS
EAST SOUTH NORTH
a. SLOPE_ASPECT QUADRANT
= Re] HEMLOCK
“| BOREAL = BY. NO. HARDWOODS
DRY
HARDWoops 4 PINES
60 a 4
40+ 4
20r
ol PS BS
EAST vs WEST
b. 180° SLOPE ASPECT GROUP
—
42.
Variation with Degree of Slope
The variation in composition with degree of slope for the entire
sample is shown in Table 11. These data were not normalized by altitude
and, therefore, the boreal species show high amounts for steep slopes
because the upper altitudes are the most steeply sloping.
Species with maximum RIVs on "moderate" slopes (1 to 10°) are those
in the hemlock-northern hardwoods group including sugar maple, American
beech, and hemlock. Sugar maple occurs in relatively high importance
on higher sloping terrain (1 to 20°), while hemlock is not found either
on slopes greater than 10° or on level sites. Moderate slopes in the
1°-10° range are most favorable for hemlock.
The dry-hardwoods group including northern red oak obtains highest
importance on relatively high slopes ai to 20°). Red pine reaches
maximum importance on steep slopes (above 20°), but also occurs on less
steeply sloping sites.
ACKNOWLEDGEMENTS
We express our special thanks to the numerous individuals who have
contributed to this study. No attempt is made to mention them all by
name, but our thanks to those not identified are no less sincere.
Especial thanks go to Paul Lemon who initiated ecological interest
in the study area and who contributed many valuable suggestions to the
current program, to Stuart Nicholson who did much research on the early
history of the area and has added much to our understanding of the
upper elevation spruce-fir element, to Jesse Craft who has been kind
enough to share with us his findings related to the glacial history of
the Whiteface area, and to John Witty for his contribution to our
knowledge of the soils underlying the various forest associations to
be found on Whiteface Mountain.
We also express special thanks to Harry Hamilton, Desmond Bailéy,
Jim Droppo, and Ron Kujawski, State University of New York at Albaty’,
and to Richard Arnold, Lee Miller, and Earl Stone, Cornell University
for their continued interest and contributions to the total study.
Finally we express our deep gratitude to the Atmospheric Sciences
Research Center of the State Universtiy of New York whose generous support
allowed the initiation and completion of the early phases of this study,
and to the National Science Foundation for continued support of the
basic program under grant number GB-7020.
REFERENCES
Bird, J.M. 1963. Reconnaisance geologic study of the Whiteface Mountain
hee Publ. No. 15. Atmosph. Sci. Res. Center, State Univ. of N.Y.,
any.
Buell, M.F., A.N. Langford, D.W. Davidson, and L.F. Ohmann. 1966. The
upland forest continuum in northern New Jersey. Ecology 47:416-452.
. fOODS
-43-
Cottam, G., and J.T. Curtis. 1956. The use of distance measures in
phytosociological sampling. Ecology 37:451~460.
Curtis, J.T., and.R.P. McIntosh. 1951. An upland forest continuum in the
prairie-forest border region of Wisconsin. Ecol. 32:476-496.
Donahue, R.L. 1940. Forest-site quality studies in the Adirondacks: I. Tree
growth as related to soil morphology. Cornell Univ. Agr. Exp. Sta.
Memoir 229. Ithaca.
Fernald, M.L. 1950. Gray's Manual of Botany, 8th ed. American Book Co.,
N.Y. 1632 pp.
Goff, F.G., and G. Cottam. 1967. Gradient analysis: the use of species and
synthetic indices. Ecology 48:793-806.
Grosenbaugh, L.R. 1952. Plotless timber estimates - new, fast, easy.
J. For. 50:32-37.
Habeck, J.R. 1958. White cedar ecotypes in Wisconsin. Ecol. 39:457-463.
Heimburger, C.C. 1934. Forest-type studies in the Adirondack region.
Cornell Univ. Agr. Exp. Sta. Memoir 165. Ithaca.
Isachsen, Y.W. 1964. The geology of the Adirondacks. Public Lecture,
Atmospheric Sciences Research Center Lecture Series, Wilmington.
Kujawski, R.F., and P. Lemon. 1969. Ecological effectiveness of yellow
birch in several Adirondack forest types. Publ. No. 92. Atmosph. Sci.
Res. Center, State Univ. of N.Y., Albany.
France, 0., and P. Lemon 1963. Preliminary observations on forest tree
ecology of the Whiteface Mountain area. Publ. No. 15. Atmosph. Sci.
Res. Center, State Univ. of N.Y., Albany.
Lindsey, A.A., J.D. Barton, and S.R. Miles. 1958. Field efficiencies of
forest sampling methods. Ecology 39:428-444,
McFee, W.W., and E.L. Stone. 1965. Quantity, distribution, and variability
of organic matter and nutrients in a forest podzol in New York. Soil
Sci. Soc. of Amer. Proc. 29:432-436.
Miller, W.d. 1918, Geology of the Lake Placid Quadrangle. N.Y. Museum .
Publ. Nos. 211, 212. Univ. of the State of New York, Albahy. 106 Pa
O'Kane, W. 1928. Trails and summits of the Adirondacks. Riverside, New York:
Reilly, R.W. 1964. A general ecotogical study 6f the moss flere on
Whiteface Mountain. Publ, No. 21. Atmosph. Sci. Res. Center, State
Univ. of N.Y., Albany. ,
Stoddard, S.R. 1879. The Adirondacks. Van Benthuysen, Albany. p. 64.
Street, A.B. 1869. The Indian Pass. Hurd and Houghton, New York. p. 126,
Watson, W.C. 1869. History of Essex County. Munsell, Albany. p. 170.
Witty, J.W. 1968. Classification and distribytion of soils on Whitéface
Mountain, Essex County, New York. Unpubl. Ph.D. Dissertation, Cornell
University. 291 pp.
COMPARISON OF TOPOGRAPHIC AND VEGETATION
GRADIENTS IN FORESTS OF WHITEFACE MOUNTAIN
NEW YORK
by
Jon T. Scott and J. Gary Holway
Comparison of Topographic and Vegetation
Sradients in Forests of Whiteface Mountain,
New York
by
Jon T. Scott and J. Gary Holway
ABSTRACT
Tree species data for 182 forested stands from on or
near Whiteface Mountain in northern New York was used to
obtain a vegetation gradient. The stand index value (STV)
obtained by a modification of the method of leading dominants
was found to be highly correlated with altitude. A regression
analysis of STV versus altitude for three altitude ranges
and plots of mean STV by 500 ft altitude intervals showed a
non-linear relation, The regression lines and plots showed a
region of steep slope between the upland spruce-fir (Picea
rubens-Abies bhalsamea) and the lowland hardwood forests,
If this “ecotone” between two “associations” is not caused
by environment then there must be reasons for species to
associate into types, If the slope of the curve is due to
environmental variations along the altitude gradient then
the continuum hypothesis is correct and association is due
only to the chance that species overlap in their range of
tolerance.
The STV was found to be higher on east facing than on
west facing sites, but the north to south variation was small,
A possible explanation is that heat balance differences between
the east and west sites governed by wind exposure and diurnal
variation in solar radiation input have greater impact than
the north to south differences in solar radiation, East-west
differences in wind damage are also larger than north to
south differences,
INTRODUCTION
A phytosociologic index provides a framework upon which
ecologists can perform a variety of fruitful investigations
including studies of evolution, succession (Buell, et. al., 1966)
vegetation structure (Goff and Zedler, 1968), consumer
populations (Beals, 1960) and vegetation-environment relations,
This paper deals with the last mentioned of these. Its purpose
is to describe the use of an index based upon data from 182
“AB:
~46~
forested stands located on or near Whiteface Mountain in the
Adirondacks of northern New York, This index will be used
to study the variation of vegetatian along topographic gradients,
that. is, along changes in altitude, slope and slope aspect.
This application to a region of widely divergent vegetation
within a small area leads to some tentative conclusions
concerning the nature of vegetation distribution along environ-
mental gradients,
The application of a phytosociologic (or compositional)
index to vegetation data produces an arrangement of stands
or species which has been given various names such as
"continuum" by Curtis and McIntosh (1951), and "ordination"
by Goodall (1954) and Curtis (1959). The term "gradient
analysis" was used by Whittaker (1956) in a study of the
distribution of many species along topographic gradients.
Goff and Cottam (1968) have also used this term to express the
variation of vegetation across wide ranges in composition and
it is retained here. The placement of stands along a gradient
governed by the species compositional index will be termed
the "vegetation gradient".
The term vegetation gradient lacks the ambivalence of the
others which have been used, The word “continuum" implies
that vegetation is a continuous variable in all cases where~
as it is well known that abrupt spatial variations exist
(i.e. at borders of swamps or lakes, etc.). Also, the word
continuum implies that there are no reasons for species to
associate except that they happen to overlap in their range
of tolerance along an environmental gradient, This may be
so but has not yet been proven, Although these may be trivial
arguments, we believe that the use of the word continuum
has led to some misunderstanding. The word "ordination"
implies an ordering or ranking which may not necessarily be
based upon measured distance along the order. For instance,
we can rank the numbers 0, 3, 4, 9, 18, in the proper order,
but not provide the useful information that the difference
between the last two numbers is equal to that of the first four.
The term "gradient analysis" implies that the variation or
distances along the gradient have been determined.
The vegetation gradient is based upon a species sacio~
logic index which we will term the species index number (SPN).
This is derived from measurable properties of vegetation
such as density, basal area, cover and frequency. Goff and
Cottam (1968) compared six methods of obtaining species
index numbers using the same vegetation sample of upland
stands in southern Wisconsin. They found that all. methods gave
similar results. In our studies at Whiteface Mountain we have
compared three methods and also found that they give essentially
the same information on the first axis, Only the fairly
spa method of Curtis and McIntosh (1951) will be discussed
eres
-47-
STUDY AREA AND VEGETATION SAMPLE
The study area and properties of the vegetation sample
have been described by Holway, et. al. (1969) and will be
briefly summarized here. All of the 182 stands were located
on or within seven miles of the 4867 foot summit of Whiteface
Mountain (Figure 1). This mountain is the most isolated of
| all the Adirondack high peaks being about 15 miles north of
| the so-called “high peaks region".
The climate of the Adirondacks is cool and wet with
precipitation exceeding evapotranspiration by from 25 to 40 in,
The Whiteface region receives about 40 in. of precipitation
with about 30 in. of runoff, July mean temperatures range
from about 60° F in the lowlands to the low 50° F range on
the peaks, Frosts and rime-ice occur during all months of
the year near Whiteface summit.
Bedrock of the region consists mostly of anorthosite
with smaller amounts of granite, The soils above 3500 ft.
: contain little fine mineral material and consist mostly of
j ‘M@8S-covered boulders and peat. Glacial till occupies the
level sites and drainage channels, but the till is usually
very shallow, The lower slopes have deeper soils with podsols
fairly well developed,
| Witty (1968) classified the soils on Whiteface Mountain,
' On the basis of morphological and other Properties, field
i descriptions and results of laboratory analysis he found the
i Spodosol and Histosol families to be common in the ‘region.
The cation exchange capacities were generally low and the
soils strongly acid. The amount of exchangeable bases was
| low except in some soils under hardwoods.
| The summit of Whiteface contains a small region of tree~
t less vegetation with tundra species present. Below this region
ae to about ayte ft. are ues te pure fede of balsam led
Abies balsamea)“, Red spruce Picea xupene is mixed w
the fir down to about 2700 ft. and on wet or exposed sites
| lower on the mountain. Below 2700 ft. northern hardwoods are
the dominant type with maple (A saccharum), beech (Faqus
grandifolia), and yellow birch Caetute alieahaniensis). the
| moe commen oe On Sacre s ee te stro (Pips ft,
| and below, red pine resinosa) and white pine inus
8. us) are common, mall amounts of hemlock (Isuqa canaden-
Sis) are often mixed with the hardwoods but this species occurs
more often at altitudes of 1200 to 1800 ft, in stands where it
is by far the most common species,
There are several finger-like ridges pointing to the
gast on Whiteface Mountain and Stevenson range to the north
(see Figure 1), On the south side of these ridges from
about 1500 to 2400 ft, red pine dominated stands occur repeatedly.
1 Classification is according to Fernald (1950).
Figure 1:
~48-
Map showing the location of Whiteface
Mountain and the region surrounding
it. Of the 162 stands sampled, 132
were taken from the area shown with
the remainder from nearby lowlands.
+i
&
a
mA)
| Oe
LE:
loomce
ay, Seale ube 2 contour interval =250' a
~50-
On the cast-facing slopes of these ridges from about 1000
to 2000 ft are stands containing northern red oak (Quercus
rubra var. borealis) mixed with the northern hardwoods and
small amounts of. fronwood (Qstrya virginiana) and white ash
(Eraxinus americana). These species also occur on other
moderately well-drained sites of the Ausable River Valley
to the east of Whiteface but not om the west side of the
mountain.
Adirondack vegetation received many kinds of disturbances
beginning with settlement in the early 1800's, Charcoal
production for the local iron smelting industry began about 1815.
This clearécutting of lowland hardwoods continued until the
Jate 19th century. Logging of the virgin spruce-fir forests
took place in the late 1800's but ended in 1896 when the state
legislature passed the "Forever Wild” bill to protect the
Adirondacks.
Fires followed much of the leggings The more recent
intensive fires are readily discerned from site inspection
and many are well marked on ch photographs by extensive
areas of nearly pure paper birch (Betula Papyrifera and its
variety 8. papyrifera var. cordifolia). The cordate-leaved
variety may occur in nearly pure stands above 2800 ft but
is also mixed with Picea rubens and Abies balsamea in the
Sspruce-fir forests, The two varieties of birch may be inter=
mixed at altitudes of about 1800 to 2800 ft but the cordate~
leaved variety is rare below 2000 ft.
The 182 stands were selected primarily to obtain samples
from a range of altitudes and slope aspects, It was arbitrarily
decided first to obtain stands at 500 foot altitude intervals
on the four major compass points. Aerial photographs and
maps were studied to select promising sites, If these sites
were judged to be “recently disturbed" then the nearest
“undisturbed" site was sampled if one could be found.
Preliminary analysis of the first two years of data showed
-that several common species were undersampled because
undisturbed stands below 1500 ft were not easy to find,
White pine (Pinus strobus)and hemlock (Isuga canadensis) which
occur on low-lying sites were much more rare in the sample
than appeared to be the case from inspection of the area,
Stands containing these species were sought and sampled.
When an “undisturbed" site was reached it was sampled
for density, frequency and basal area, The quarter method
(Cottam and Curtis, 1986) was used in 1964 but in 1965 and 1966
basal areas were determined with a Bitterlich prism (Grosen-
baugh, 1952), circular plots were used for density and the
quarter method for frequency. This latter method was deemed
ie ao efficient from our studies and from Lindsey, et. al,
(1958).
Trees were defined as stems of 4 in dbh or greater.
Saplings were defined as stems from 1 to 3.9 in dbh, Thece
were sampled for relative density and relative frequency cy
the quarter method. Ground flora frequency was obtained from
one square meter plots at each sample point,
Table 1: Mean relative importance and constancy for the species contained in the vegetation sample of 182
stands at Whiteface Mountain. Species with less than 0.5 percent importance have been omitted. Common
names and abbreviations used in Tables and Figures are also listed.
Abbreviations RIV Constancy
Species and Common Name Tables Figures a %
Abies balsamea (L.) Mill. (balsam fir)............cceeecee serene Ab F 21.0 66
Picea rubens Sarg. (red spruce)...... «Pr Ss 16.2 77
Acer saccharum Marsh. (sugar maple) As M Ted 43
Pinus resinosa Ait. (red pine)................0000 Pres R 6.6 7
Betula alleghaniensis B and B Small (yellow birch). Ba Y. 6.2 47
Tsuga canadensis (L.) Carr. (hemlock)...........--..-.-00 .. Te H 6.1 28
Betula papyrifera var. card. (Regel) Fern. (cordate-leaved birch. Be c 5.7 36
Fagus grandifolia Ehrh. (beech)......-..... 0. cece eee e eee e ee ee eee Fg B 5.6 39
Pinus strobus L. (white pine)...... -. Ps W 5.3 20
Betula papyrifera Marsh. (paper birch)............... B pap P 2.8 4]
Quercus rubra var. borealis (Michx. f) Farw. (red oak) -. Or 0 2.5 14
Acer rubrum L. (red maple)............seceeee eee ee eee -. Ar - 2.3 39
Thuja occidentalis L. (white cedar) To N 2.3 16
Fraxinus americana (white ash). Fa A 1.1 yal
Tilia americana L. (basswood). Ta T 1.0 13
Ostrya virginiana (Mi11.) K. Koch. (ironwood). Ov I 0.8 18
Acer pensylvanicum L. (striped maple).............- Apen - 0.7 20
Populus grandidentata Michx. (large-toothed aspen). -- Pg - 0.6 8
Pyrus decora (Sarg.) Hyland (mountain ash).............ccceeeeeee Pd - 0.6 15
-Lg-
-52-
Sample points in each stand were placed 20 normal paces
apart along the slope. No change in course was made unless
the slope aspect changed from the original point by at least
45 degrees (except for nearly level sites) or when an obviously
disturbed area was encountered. When either of these avents
occurred, the pacing proceeded two points (40 paces) downhill
and then in the reverse direction to the original course along
the slope. No attempt was made to define a homogeneous stand
in the field. That is, the stands were defined by altitude,
slope aspect, and judgment of lack of recent disturbance.
The number of points in a given stand ranged from 10
to 40, The fewest number (usually 20) were used in the high
altitude stands containing only a few species but 30 to 40
points were used in the mixed vegetation at lower altitudes.
The Chi-square test for homogeneity was applied to all 182
stands. The test described by Curtis and McIntosh (1951) was
used which is essentially a test for heterogeneity. A stand
was considered to be homogeneous if it did not pass the
Chiesquare test at the five percent level of confidence.
There were 42 stands which proved to be "“non=homogeneaus".
Inspection of the stand data showed that “heterogeneity” was
caused by either clumping of the major species or by change
in composition along the slope. In the following analysis
all 182 stands will be used because the primary concern
here is not to study only “homogeneous types" but to determine
properties of the vegetation as it varies with topography.
Analyses of vegetation gradients using only: tha homogeneous
stands will be reported at a later date.
The relative importance value and constacy for all of
the species in the Whiteface sample are given in Table l.
Balsam fir was the most important species followed by red
spruce and sugar maple. Red spruce was the most commonly
occurring species, Details regarding the composition ef the
samples and the ecologic relations of the major species were
discussed by Holway, et. al. (1969).
THE VEGETATION GRADIENT
Leading Dominants
A method of determining a vegetation gradient was developed
by Curtis and ficIntosh (1951) whereby the relative placement
of major species of a large number of stands could be obtained
from their degree of association. Buell, et. al, (1966) used a
similar procedure and the Whiteface data were also subjected
to this “method of leading dominants".
The sample was first divided into groups based upon the
leading dominant. The mean relative importance value (RIV)
of all the species in each group was determined, The groups
were then arranged in the "most symmetric" pattern as shown
in Table 2 using both mean RIV and constancy. The same arrange~
ment procedure was used for the data expressed as “normalized”
Table 2: Percent importance value and constancy (lower figures) for groups of stands based upon
the species of leading importance value and arranged symmetrically. Abbreviations are given in
Table 1.
No. of Stands (44) (4) (29) (5) (10) (15) (10) = (28) (6) (11) (15)
Species Ab Be Pr To Ba Te Fg As Qr Ps Pres
Abies balsamea 67 21 16 7 8 3 2 2 -- 4 1
(100) (100) (93) (100) (90) (60) (40) (25) -- (27) (40)
Betula pap. var. cord. 12 52 8 3 4 -- - 1 ata sii we
(85) (100) (62) (40) (20) (4) -- ee aa aie
Picea rubens 7 18 49 13 7 9 8 3 -- 4 6
(82) (100) (100) (100) (100) (100) (90) (57) -- (45) (73)
Thuja occidentalis -- -- 1 63 1 1 a = -- 3 3
-- -- (28) (100) (30) 3) + oo (27) (33)
Betula alleghaniensis 1 8 6 2 37 9 16 6 -- -- -- Pay
(14) (75) (65) (60) (100) (93) (100) (72) -- - =~ Oa
Tsuga canadensis -- -- 3 1 2 57 4 2 _ 1 2
; (2) -- (35) (40) (30) (100) (60) (25) -- (1g) (33)
Fagus grandifolia -- -- 7 -- 10 6 36 14 1 -- --
se 07) (70) (93) (oo) (oo) (67) -- =<
Acer saccharum -- -- 2 -- 13 5 24 48 30 1 =-
-- -e (28) (80) (0) (100) Goo) @3) (27) (7)
Quercus rubra var. bor. -- -- -- -- = ae = 5 45 1 2
“oe -- -- -- 7) 0) (29) (oo) (9) (20)
Pinus strobus -- -- 2 -- -- -- -- -- 2 61 14
-- == 28) 0) -- 20) -- -- (17) (00) = (80)
Pinus resinosa -- -- 1 -- -- 1 -- a 3 15 65
-- -- (10) -- -- 7) -- = 7) (9) (100)
~54-
RIV in Table 3, In the latter case each RIV in Table 2 was
multiplied by the ratio of the mean RIV of the most important
species (balsam fir) divided by the mean RIV of the particular
species, This procedure was an attempt to avoid overweighting
the importance of common species in the sample but gave the
Same arrangement as Table 2,
The arrangements in Tables 2 and 3 give a species ordering
with the high altitude spruce-fir (cool and wet) sites on
one end and lowland pine (warm and dry) stands at the other
with mesic hardwoods stands in the center. The wet to dry
sequence is indicated but not proven, There is reason to
believe from field evidence that white cedar (To) and hemlock
(Tc) occur on drier sites than their position shown in Table 2,
Strip Method
To obtain a more precise positioning of the species along a
vegetation gradient a method similar to that of Curtis and
McIntosh (1951) was applied to the sample of 182 stands,
The RIV of each species in each stand was plotted on a 100 cm
long strip of graph paper, Each species was represented by a
different colored dot. The 182 strips were then arranged so
@ to produce “solid normal curves" of the colored dots. In
the first such arrangement the authors were guided by the order
of the major species in Table 2. The only rule applied was
to attempt to find the smoothest possible curve for each species,
To obtain a value for ranking the species, the mid-points for
each species in the order was found by adding the species RIV
from left to right until the sum reached 50% of the total
sample RIV of that species, The positions of the species
mid=points were then normalized to a scale from 0.00 to 1.00
and rounded to the nearest 0,05 units, These numbers are
here called SPN (species index number) and are analagous to
the "climax adaptation values" of Curtis and McIntosh (1951).
The final determination of these numbers is given in Table 4,
Obviously, this procedure is not without bias because the
arrangement may have been affected by the personal judgment
of the authors who had knowledge of the site characteristics,
For example, if the strips were to be arranged according to a
wet to dry sequence balsam fir stands would be placed at the
Opposite end from red pine stands, and the arrangement. would
no doubt be similar to the one based upon the order in Table 2,
The question was asked as to how the strips would be arranged
without knowledge of Tables 2 and 3 and without any field
knowledge of the species involved. To answer this question,
the strips were thoroughly mixed .in bundles and given to
students to arrange. The students were completely unaware
as to what the symbols represented on the strips, and even
if they had known, had no familiarity with the vegetation of
of the region, The only instructions given were to arrange
the strips so as to produce what seemed to them to be the most
reasonable series of "solid normal distribution" curves,
~55-
Table 3: Adjusted mean RIV for groups of stands having one species of
leading RIV, Values are those in Table 2 multiplied by the ratio of
the mean RIV of Abies balsamea to that of the particular species.
Abbreviations are given in Table 1.
(44) (4) (29) (5) (10) (15) (10) (28)_(6) (11) (15)
Ab Bc Te Ps
)
Pr To Ba c Fg As Qr
Pres
Ab 67 aT T6 7 [re 2 20 == 4 T
Bc %% 189 28 12 13 -- -- 3 ow -- -
Pd 49 “% 40 -- 23 -- -- -- ee -- --
Pr 21 24 «64 17 22 nN n 4 -- 5 8
To -- -- TS 572 6 5 -- eo 6 24
Ba 4 28 «620 ~9F 124 30 55 20 -- -- --
Apen 3 -- 40 9 “TF 62 57 9 3 -- 3
Te -- -- 9 2 7 197 14 7 -- 5 6
Fg -- -- 2 -- 40 “23 136 52.4 -- --
Ar 4 -- 23,41 25 38 “38 16 8 40 17
As -- -- 30 -- 23 9 44 86 54 2 --
Ta -- -- ad -- 4 -- 87 69 -- --
Bpap 1 -- 36 0 22 7 13 7 46 26 23 34
Fa -- -- 4 14 10 2 12 83 (135 -- --
Ov -- -- 30 -- -- 18 5 39 276 89645 8
Qr -- -- wee -- 1 1 38 374 n 4
Ps - -- 7 1 -- 2 -- -- “TO 241 56
~56-
Table 4: Listing of the species index numbers (SPN) for the
Whiteface Mountain vegetation sample and the "smoothed" SPN
(see text for explanation).
Tree Species "Smoothed"
SPN SPN
Abies balsamed.....sececeeeeseeeereeeee 0,00 0.00
Betula papyrifera var. cord. .......... 0.10 0.10
Pyrus decora.....seseccsecverseceeeees O.10 0.10
Picea rubens......secceeseeereene + 0.25 0.25
Thuja occidentalis... . 0,40 0.35
Acer spicatum....... . 0.45 0.55
Acer pensylvanicum.... - 0.45 0.45
Betula alleghaniensis. » 0.45 0.50
Tsuga canadensis.... . 0.50 0.50
Acer rubrum.......++ ~ 0,55 0.60.
Fagus grandifolia..... . 0,60 0.65
Acer saccharum...... + 0.70 0.75
Betula papyrifera... + 0.75 0.75
Fraxinus americana.... ve O75 0.80
Tilia americana..... + 0.75 0.85
Populus gradidentata....... . 0,80 0.95
Ostrya virginiana........ . 0.85 0.90
Quercus rubra var. bor. . « 0.85 0.90
Pinus strobus........+. . 0,90 1.00
Pinus resinosa........eseee seveee 1,00 1.00
j
i
I
-57-
Three such independent arrangements by students were obtained.
In all three cases balsam fir was placed at the extreme left
and red pine at the extreme right, This is the same as the
sequence arrived at based on field evidence and on Tables 2
and 3. Further, the order of other species between the twa
terminal ends was nearly the same, but the students found it
difficult to place the stands dominated by hemlock and white
cedar,
Because of the difficulty of placing the hemlock stands
in the strip order, a student was asked to arrange the strips
without the stands having hemlock leading in RIV. The SPN were
then determined as before, In this case the stands containing
hemlock (not as the leading species) were placed by the student
according to the species with which hemlock associates. The
SPN of hemlock turned out to be 0.50 by this method.
The stands with hemlock leading in RIV were then examined
for the species associating with hemlock. The weighted mean
SPN of all species in hemlock stands (not including hemlock)
was found to be 0,52. Thus the SPN of hemlock was determined
by three different methods. The first was an arrangement
of the strips based upon the most symetric placement of the
leading dominants. The second was based upon a determination
of the species hemlock associates with and the third upon
the species associating with hemlock when it is the most
important species in a stand. All three methods produced
essentially the same result.
The above procedure was repeated for white cedar with
similar agreement between methods. Such a procedure might
also be recommended for obtaining the SPN for all species,
but it is doubtful that the results would change the numbers
given in Table 4,
The students with no knowledge of the meaning of the
dots on the strips seemed to place strips predominantly
according to the leading few species. This caused large
variation in the SPN of the rare species for the several trials
used. Rare species could be more reliably placed in the order
by computing a "stand index value" (STV) and then ranking the
stands by this value, The STV was obtained by multiplying
the SPN (Table 4) by the RIV for all species in the stand and
summing the products. The range of STV was therefore from 0
to 100.
After ranking the stands according to their STV the
species mid-points were again computed to obtain the new
"smoothed" SPN. These numbers are shown in the second column
of Table 4,
The SPNs obtained here have no relevance to the vegetation
in any other area except perhaps in relative placement of many
of the species, The SPNs in fact are truly relevant only
to the sample of 182 stands used in the analysis although
nearly the same numbers would be obtained from a completely
different sample taken in the Whiteface area, The relative
Figure 2:
-58-
Scatter diagram of STV versus altitude
and plots of regression lines, All
stands are included, Numbers refer
to the columns for the regression
data in Table 6. Abbreviations are
given in Table 1,
0 59 -
t. Fe oy <a Eg
/: ae
1O}- Fe os’ ed a ha Fst
sf sr cs BORA
| FY S Kefsc SC
20 FS dy /
> SF Ss J
| nH 3O0- ce gp
| og Gy Boyd ow
=“ 40- WE vi vse ya Why
8 MF ag Jf
a es 7 pM
He
>< 50{- " ins feo
a ; oMiiMe_B ME
a 60 wn va YB
z /- MB MB
| = | MA ie Me MPa,
t 70 og Cn
RN MT rR
| a ae
= WR RS
re
RW
SOLWAY we Non Re WR "
wrRw RW P
. a RW en
100) | | | | | |
500 1500 2500 3500 4500
ALTITUDE FEET
~60-
placement of the SPN from the present analysis is similar to
those obtained by Curtis and McIntosh (1951), Brown and Curtis
(1952) and Buell, et. al. (1966), but the earlier studies
probably did not include as wide a range of site characteristics
as in the Whiteface sample. If the stands from higher than
2500 ft were ‘removed from the Whiteface sample then the SPN
would perhaps be in closer agreement to those found in
Wisconsin and northern New Jersey.
Although the SPNs have ecologic significance it is the
STVs which are useful for relating the vegetation to environment.
More rigorous techniques are available for placing stands
along a vegetation gradient than the simple procedure followed
here. In the present study the STV will be related only to
simple measures of environment and therefore, only one “axis"
of the vegetation gradient is desired. When more sophisticated
environmental data is obtained techniques will be used which
extract ‘more information from the sample. Goff and Cottam
(1968) have shown’ that all standard methods produce nearly
the same results for the first axis.
VARIATION OF STV WITH TOPOGRAPHY
The major difficulty in a study of vegetation-environ-
ment relations is in measuring the environment, Paraphrasing
the rigorous definition of Mason and Langenheim (1987),
environment ideally consists of all phenomena which directly
affect an organism sometime during its life cycle. ; They
emphasize that environment acts at the organism level,,
Therefore; the concept of "stand environment" is nebulous
at best. It consists of the integrated effects of all
phenomena which affect individual plants and animals. However,
many physical parameters may be measured at the stand level
which are closely related to environmental phenomena and
. therefore influence the composition of vegetation, We can
“never hope to measure the "operational environment" as de-
fined by Mason and Langenheim (1957), unless we have sensors
which are organisms themselves, Approximations to the environ-
ment would -consist‘of measurements of such things as energy,
moisture, and nutrient balances and smaller elements of these. ;
It is the-goal of the study at Whiteface Mountain to obtain
better measures df these quantities in order to relate them
to the vegetation gradient. Because this work is not yet
complete, only simple approximations of the environment can
be used. In this paper the vegetation gradient (STV) will
be related only to topographic variables,
The Topographic Groups
The STV can be related to topography by plotting it in
scatter diagrams against each feature (altitude, slope-aspect,
‘or slope) or the various topographic features can be broken
-into groups (or ranges) and mean STVs. compared for each feature.
blz
Both techniques will be used here, but problems arise with
each method. Looking strictly at mean values does not consider
the kind and amount of variance and looking only at plots
of individual stands may influence the result by overweighting
for sub-groups: within the sample which may have been "over-
sampled".
The difficulties involved with "oversampling" of certain
topographic ranges are illustrated by the data for all stands
summarized in Table 5. Mean STV is given for topographic
groups broken down first by 500 foot ranges of altitude (250
to 750 ft etc.).then by slope-aspect quadrants (e.g. east
includes stands with slope-aspect ranging from 45° to 134°)
and last by 10° intervals of slope. The number of stands
in each group is also given. Note that certain ranges may
include several stands while others were not.even sampled.
Therefore, if a scatter plot and a correlation were made,
for example, of STV versus slope-aspect the result may be
strongly influenced by the fact that some altitude ranges
contained either a high or low number of stands with certain
slope-aspect ranges. For example, in the 2500 ft range there
were eleven south facing stands but only three each facing
east and west. The mean STV for this range will be influenced
by the fact that the STV for south-facing stands was higher
than other aspects if stands are weighted equally.
Other methods of weighting to compute mean STV can be used.
For example, when computing means for altitude ranges one can
first compute means for slope-aspect. groups by weighting the
slope ranges equally and then obtain altitude means by weighting
the slope-aspect ranges equally. Many combinations of weighting
exist and the resulting. STV means may be quite different than
those in Table 5. Some of these are discussed in the following
sections.
Altitude Variation of STV
It is expected that many physical phenomena related to
the environment of an organism are highly correlated with
altitude, These include many constituents of the heat, mois-
ture, and nutrient balances, Many authors have found that
the mean value of measures of species importance form “bell-
shaped" curves when plotted against altitude indicating that
the species has a range of tolerance with some optimum at a
particular altitude. This work is well represented by Whittaker
(1956, 1960), and Whittaker and Niering (1965). (See also
Figure 7.)
Plotting of the stand index value (STV) rather than a
species importance value has the advantage that the STV may
result from the integration of more environmental phenomena
than the species value. It is, therefore, a method of relating
the total vegetation with environment. The STV is plotted
against altitude in Figure 2. For each plot the one or two
species of leading importance value are shown by the appropriate
letters.
Table 5: Mean stand index values (STV) and number of stands by various groupings of altitude,
slope-aspect, and slope. Weighting is at stand level (all stands weighted equally).
Slope Range Slope-
aspect Altitude
Altitude Slope More
10°tol9° ~=—20°to29° ~—s than 30° ~=— means. Range
Range Aspect Li
iy
<
Ly
we
°
ct
fo
iJ
0
Range NSW N SIV N STV WN STV N STV N STV NSW
E Saar Re TW6 = oe To =
s -- 1895 - - 2 = = 189.5 - -
500 W - eee -- - - -- _- --
N mon =: & Sl in - - -- -- --
Means 184.4 189.5 190.6 - - -- 290.0 3 88.2
E -- 186.7 - - 2h. & #3 186.7 - -
Ss a a oe ee <a me ae aes
1000 W - - 5701 197.30 - - ss 674.6 - -
N “2 190.0 - - — 190.0 - -
Means 391.2 672.8 293.6 - - -- 878.0 11 81.6
E - - 861.9 565.1 177.4 9+ - Wed - -
s - + 757.6 267.5 - - z 3 959.8 - -
1500 W - - 459.3 6493 274.0 -- 1256.7 - -
N - - 3409 152.0 271.2 - - 652.8 - -
Means 11 37.5 2257.1 1457.8 573.5 - - 4159.3 5254.7
E - - 463.3 5657 - - a 964.6 - -
s - - 361.4 3609 2655 - - 862.3 ° - -
2000 W - - 565.4 3506 141.1 -- 957.8 - -
N - - 4558 - - 171.9 - - 559.0 - -
Means 5 52.016 61.7 1160.3 461.0 - - 3161.1 36 59.8
E oe me -29.0 2668 - - 354.2 - -
s - - 3691 3726 5493 132.7 W610 - -
2500 w - = 156.1 - - 136.5 - - 341.8 - -
N “om 323.2 339.2 - - 631.2 - -
Means 140.4 465.8 745.1 1148.6 132.7 2349.9 24 49.4
-29-
Table 5 (Continued)
Altitude Slope
Range Aspect
Range
3000 W
N
3500 W
4000 W
4500 W
N
Means
Means of slope
groups
Slope Range
Level 3°to9° 10°tol9° ~—-20°t029°
is ws wsty “NSIV
- = - = -- 217.0
DI ofl awe 6:
PIoll ole 413.8
[I ll 315.9 4272
22 ll 5156 10 19:8
ee ee DB 218.0
Sas 2 8.8 310.0
>If l 485) 3 72
TI ll 478 810.9
-2. 233 -- 2 10.0
TI fT" 435 3 29
rr rae 1 5.1
72 ll 2 os 61 38
720233 3107 7 54
wee em me 1 3.6
PIOol — 1 0.0
2. ee fe 2 1.9
51.0 5159.8 4746.3 47 31.3
21
Awr1asny Pr inn
Pinas
16
bu SR wo &
ML OWN TDWtow
wn
7.7
it
Slope-
aspect
means
$
2 17.0
2 15.2
4 13.8
8 21.3
6 18.1
5 13.5
2 9.8
5 9.5
4 7.2
6 10.2
6 6.8
5 2.6
1 5.1
6 2.6
8 4.1
2 4.4
1 3.4
2 0.8
1 0.0
6 2.3
Altitude
Prigd
a
meen
-¢9-
~64-
The leading species in Figure 2 tend to "clump" in regions
of the altitude- STV plot. For example, the fir and spruce
stands are concentrated. in the upper right corner, the hemlock
in the left center, and pines on the lower left, This indicates
thatan STV-altitude “classifination" system might be fruitful
because the stand index expresses stand composition. When
plotted on an STV axis, the dominant species would, therefore,
tend to be found in regions close to the species index number
of the particular species, For example, hemlock should "clump"
near 50 STV and balsam fir near zero on the STV scale. Yellow
birch and red spruce are more widely scattered in Figure 2
indicating that they have wide ranges of tolerance.
The scatter plot in Figure 2 shows that STV correlates
with altitude. The linear regression coefficient is 0.79 with a
standard error of 18,0 STV units (Table 6). The scatter is
fairly small for high altitude stands and increases with decrease
in altitude down to about 1200 ft. Below 1200 ft the sample
contained only a few stands because of the small amount of
area below this altitude. The scatter would presumably con-
tinue to increase if more low-lying stands were included,
Computing the regression coefficient for the entire range
of altitude of Figure 2 may be misleading because upon inspec-
tion the slope of the plot seems to change at about 3000 ft,
Regressions were performed on smaller altitude ranges and the
results are given in Table 6. The slope and zero altitude
intercepts are given by A, and Ag respectively, The regression
equations can be obtained From Table 6 and these are also
drawn for the four cases in Figure 2.
The regression line. for the entire sample falls below
the stand plots for high altitudes and above at low altitudes.
When the regression lines for the three smaller ranges are
plotted, (columns 2, 3, and 4 in Table 6) the slopes of the
lines change with altitude and the lines seem to fit the
data more closely, The correlation is low for the 1250 to 2750
foot range (r = -0,11) and the scatter is great (S = 20.2)
with significance only at the 10% level of confidence,
Breaking the scatter plot in Figure 2 into smaller groups
indicates that there may be a real change in slope of STV
with altitude, Perhaps it would be better to fit the data to a
more complicated function (e.g. sigmoid curve) but this would
not improve the correlation because the scatter in the middle
altitude range is vary great,
The changes in slope of the regression lines in Figure 2
may lead to the interpretation that the steeper portion of
the plot between 2500 and 3000 ft represents an “ecotone”
between the two more gradually sloping hardwoods and spruce-
Fir "associations". If a straight line results when vegeta-
tion is plotted against environment the assumptions would
be that there is no tendency for essociation between species
or that vegetation is a “continuum”. Because environment is
not necessarily a linear function of altitude the interpreta-
tion that “ecotones" and “associations” exist must be viewed
~65-
Table 6: Regression analysis data for STV versus altitude for several
ranges in altitude. Data show the number of stands in the range (N),
correlation coefficient (y), standard error of the regression in STV
units (S), the regression constant or zero altitude intercept (A.) and
the regression coefficient or slope of the regression (A,). The°three
groups on the right marked with an asterisk (*) contain no Tevel or
pine-dominated stands.
1 2 3 4 5 6 7
250- T250- 2250- 2750- 250- T250- 2250-
Variable 4750 2750 3250 4750 4750* 2750* — 3250*
N 182 12 40 56 139 83 36
Y ~0.79 -0.11 ~0.67 -0.69 -0,86 -0.41 ~0.70
Ss 18.0 20.2 18.5 6.2 13.0 15.0 14.1
Ay 99.8 66.3 193.5 §2.5 97.6 86.8 164.3
Ay -0.024 -0.006 -0.058 -0.012 ~-0.024 -0.017 | -0.048
Signif
Level 1% 10% 1% % 1% % 1%
Table 7: Mean STV for altitude groups centered at 500 foot intervals
with various weightings for slope-aspect and slope groups and certain
stands selectively eliminated. In column (1) are means including the
entire sample with each stand weighted equally. In column (2) stands
are weighted equally but stands dominated by pine species and those on
level sites have been removed. In column (3) pine and level stands
have been removed but the means are obtained by weighting each slope-
aspect group (E, S, W, N) equally. Column (4) is the same as (3)
but each slope group is weighted equally to compute slope-aspect means
before computing altitude means. Column (5) is the same as (4) except
that stands not on Whiteface Mountain and adjacent peaks or ridges
have been removed. Column labeled s is standard deviation,
Center of.
Altitude 1 2 3 4 5
Range NST $s Nsv s WN SV WSIV WS
500 388.2 - - - - - moe -
1000 V1 81.6 17.3 247,7— - - - -
1500 52 54.7 18.5 3757.0 14.1 456.6 458.5 3 60.3
2000 3659.8 18.8 26 56.6 14.1 456.8 457.7 4 55.6
2500 24 49.4 23.1 2042.4 17.7 442.5 443.4 4 48.2
3000 1617.9 9.3 1617.9 9.3 416.8 417.8 417.8
3500 1610.2 4.4 1610.2 44 4100 49.8 4 9.8
4000 18 4.5 3.3 18 4.5 3.3 4 4.3 4 4.2 4 4.2
4500 6 2.3 2.6 6 2.3 2.6 4 2.2 4 2.2 4 2.2
Figure 33
Figure 4:
Figure 5:
-66-
Scatter diagram of STV versus altitude and
plots of regression lines. Sample does
not include pine-dominated and level stands,
Numbers refer to columns for regression
data in Table 6, Abbreviations are given
in Table 1 (page 67).
Plots of mean STV by 500 foot ranges of
altitude. Numbers refer to columns of
mean STV in Table 7 (page 68).
Scatter plot of STV versus altitude and
plots of mean STV for 500 foot altitude
ranges, Data do not include pine-dominated
or level stands. Numbers refer to columns
in Table 7 (page 69).
STAND INDEX VALUE (STV)
40
~67-
50he
60
70
80
90
100
© 1250 - 2750
@) 2250-3250
~ N#83,r=-04
ALTITUDE - FEET
e
© 0% N=36,r=-0.70 7
ogee @ 2750-4750
| e N=56,r=-069 —
° © 250-4750
Z N=139,r=-0.86 _|
| | | | | |
1500 - 2500 3500
4500
STAND INDEX VALUE (STV)
o a BS a Ded
fe) 52) :2) (2) fe)
N“
(e)
1000 2000 3000 4000
ALTITUDE - FEET
-69-
90}-
N he) ¢ 8B o ie
(oe) eae OG
I500 2000 2500 3000 3500 4000 4500
ALTITUDE - FEET
-70-
with caution. In fact, there are reasons to believe that
environment may. also vary with altitude in the same manner
as does the vegetation in Figure 2,
The scatter plot in Figure 2 c@uld perhaps be changed
by elimination of stands which are jocated on sites having
special physiographic conditions. For example, there are
several stands in the 1500 foot range which have low STV.
These are lowland poorly-drained spruce-fir stands on level
sites, Also in the range centered at 2000 ft there are
several pine-dominated stands with high STV, These are located
on south-facing sides of ridges or on the ridge tops and are
very well-drained sites. Many of these could have received
moderate burning. It seems likely that these special cases
of topographic and physiographic conditions caused much of
the wide scatter in Figure 2. Perhaps they could legitimately
be removed from the sample.
The scatter plot with all pine-dominated and level stands
removed is given in Figure 3 and the appropriate regression
analysis data in the last three columns of Table 6... The scatter
has been reduced compared to Figure 2, The regression for
the 1250 to 2750 foot range now begomes significant at the 1%
level, When the regression lines are drawn the interpretation
from Figure 3 is the same as from Figure 2; the changing
slopes may indicate an “ecotone" between two "associations".
Another method of examining the STV-altitude relationship
is to plot the mean STV for altitude groups. This has been
done in Table 7 and Figures 4 and b. The danger of considering
mean values lies with the large variance within groups and
with the manner in which the means are computed.
The mean STV for 500 foot ranges of altitude (250. to 749
ft etc.) for all stands was first computed by giving equal
weight to each. stand regardless of its slopeeaspect or slope.
The result is column 1 of Table 7 and plot number 1 of
Figure 4. For this case, the mean STV at 2000 ft is lower
than at 1500 ft. The plot shows a fairly gradual change
from 4500 ft down to 3000 ft, a rapid change between 3000
and 2500 ft, and then again a gradual slope from 2500 to 1500 ft.
These slope changes are almost the same as those found by
examining the scatter plots in Figures 2 and 3 and lead to
the same tentative interpretation.
The question which should first be answered is whether
or not the computation of mean STV flor altitude groups is valid.
A good simple answer cannot be obtained but several simpli-
fications can be made. The first simplification is to remove
the pine-dominated and level stands, as before, because these
may represent anomalous physiographic conditions.
With the level and pine-dominated stands.removed from the
sample, the resulting mean STV is smpwn in column 2 of Table 7
and plotted in Figure 5. The curvaifor the mean STV still has a
shape similar to the plot in Figure’4 but the values at 2000
“71«
and 1500 ft have changed significantly, Also given in Figure 5
is the scatter plot of stands above 1250 ft with level and
pine stands removed, f
A further simplification can be made by using a different
method of weighting for the mean STV. In column 3 of Table 7
and plot 3 of Figure 4 the mean STV was obtained by first
obtaining means for the four quadrants of slope-aspect (E, N,
S, W). and then weighting these equally regardless of how many
stands ware contained in the slope-aspect groups. In this
case, the pine and level stands have bean removed. This
same type of weighting before removing pine and level stands
is not shown but is similar to column 1 of Table 7, In column
of Table 7 the mean STV for the four slope groups were weighted
equally to obtain slope-aspect means before weighting these
equally to obtain altitude means. Although the type of
weighting did not always affect the result it is probably
best to give equal weight to the groups rather than to indivi-
dual stands because there was often large differences in the
number of samples from one group to another (see Table 5 for
the number of stands in each category).
The last column in Table 7 is weighted ao in column 4
but only stands in the immediate vicinity of Whiteface Mountain
have been included. These data are plotted in both Figures 4
and 5 for comparison, This last result shows the smoothest
change of mean STV. with altitude but the two gradually sloping
regions at high and middle altitudes are still present with a
region of rapid change between. It seems likely that this
relation of STV with altitude is real for the particular
sample, Interpretation must be made with full knowledge of
many possible causes, Some of these may be physiographic,
some ecologic, and some may involve flaws in the sampling
or analysis, The many possibilities will be considered in
the discussion section below,
S
Slope~aspect Variation
Tha slope~aspect variation of such quantities as solar
radiation, wind speed and direction, rain catch and rime or
fog drip collection should cause enough differences in environ-
ment on Whiteface Mountain to cause differences in vagetation.
The. slops-aspect variations in other regions have been well»
documented in the literature, for example, by Whittaker (1956),
Ayyad and Dix (1964), Whittaker and Niering {i96s), Buell, et.
al, (1966) and Mowbray and Oosting (1968).
The problem of determining slope-aspect variation of STV
was similar to the case for altitude in that the various
groupings of topographic features contained varying numbers of
stands (Table 5). Data in Table 5 seem to show that there
is a tendency for higher STV on east and south facing stands
than on the north and west,
Figure 6s
Figure 7:
-72-
Scatter plots of STV versus slope-aspect
for two different altitude ranges.
Abbreviations are given in Table 1 (page 73),
Relative mean species importance value
for 500 foot ranges of altitude for the
dominant species on Whiteface Mountain,
Each curve is normalized by expressing
each mean RIV‘'s for each altitude range
as 4 percent of the range of maximum
mean RIV (page 74).
STAND INDEX VALUE
a. STANDS AT 1250 TO 2249 ft.
b. STANDS AT 2750 TO 4750 ft.
100 50
RW eR
Rw WR OR YM
Wey, Ol
80 WOR 40F
aviow
me Mur aM
60 - Ls 40+ sp
BM
uy Heyer HRMS YB s
40 SM SPE SN 20b YC aan
SH Ns SC go SF SC SF
SY sc.OUS SSF oe
20 IOb <BR ra
rs Be Fo fBS FS
fe a BS |
_ FS Foe Fo
O l l l OLE Fil er Rie
ON E S W N 0 90 180 270 360
re) 90 180 270 360
SLOPE ASPECT
SLOPE ASPECT
~74-
1334 - S0NLILIV
OO0O0V athe 0002 (elefe))
Ald NVAW
GAZIIVWYON
aS
The mean STV for the four major aspect quadrants and
all altitude ranges was computed by various methods of weighting
and restrictions on the sample. four of these are given in
Table 8, In the first column each stand is given equal weight.
This is probably a poor method because both STV and the number
of stands varied with altitude, In column 2 weighting is
First by slope groupings and then by altitude. The results
are quite different and show the highest mean STV for east
slopes followed by south with west and north having the lowest
values, Secause all slope-aspects are not represented at
altitudes below 1250 ft, these altitude ranges were removed
in columns 3 and 4 of Table 8 with the weighting the same
as column 2, Data in column 3 contain stands from the total
sample but column 4 is restricted to the sample on Whiteface
Mountain or its very close proximity. These latter two cases
probably give the best representation of slope-aspect variation.
The east Side appears to contain more species with high index
numbers while the west appears to have larger representation
of species near the low end of the index scale, Holway, et.
al. (1969) reported a similar species to slope-aspect relation
for the same sample, but using species importance values,
They found that contrast between north and south slopes was
not as great as between east and west slopes which is also
the case in Table 8,
The individual STV are plotted against slope-aspect in
Figure 6 for two altitude ranges. The scatter plot for the
lower altitude range (1250 to 2249 ft) shows a tendency for
decrease in STV from the east or northeast quadrants through
the south with lowest on the northwest, Because the mean
value of STV for the same altitude range as Figure 6a also
showed a maximum in the NE range with a minimum in the NW, a
regression of STV versus slope-aspect was run. Using slope-~
aspect directly gave r = 0.37 significant at the 1% level
but a large standard error of 13.4 due to the high degree
of scatter. The regression was also run on STV versus slope~
aspect by subtracting 45° From gach slope-aspect so that NE
facing sites were given a value of 0 and NW sites a value
of 360, This did not improve the results,
The same procedure was used for_the high altitude range
of Figure 5b. Before subtracting 45°, the regression was only
significant at the 10% level with r = -0.14 and standard
error of 18.9. Adjustment of the slope-aspect value did
not improve the result.
The mean STV for four major quadrants (£, N, S, W) and
for the quadrants centered on 45° from the major compass
points (NE, SE, SW, NW) are given in Table 9 for the same
altitude ranges as plotted in Figure 6, When all stands are
weighted equally (column 1) the result is somewhat different
than when weighting is first by four slope groups and then
by altitude. The interpretation for these two altitude ranges
is again that the "boreal" species are most common on west
-76-
Table 8: Mean STV for slope-aspect groups by various weightings.
In column (1) are means for the entire sample with each stand
weighted equally. In column (2) means for the entire sample are
first computed by weighting each slope group equally to compute
altitude means which are in turn weighted equally to compute the
four slope-aspect means. Column (3) is the same as (2) except
stands below 1250 feet have been removed. Column (4) are means
computed in the same way as (2) but stands not on Whiteface
Mountain or adjacent peaks and ridges have been removed.
Slope-Aspect 1 2 3 4
Group N Siv NST N stv oN STV
East 43 46.0 9 44.3 7 31.6 7 35.1
South 39 47.8 8 38.2 7 30.9 7 28.7
West 42 44.8 8 34.4 7 27.4 7 24.3
North 3734.2 8 34.3 7 26.3 7 26.9
Table 9: Mean STV for the same two altitude ranges as in Figure 5.
The breakdown of slope-aspect groups is either by four major quadrants
(N, S, E, W) or by quadrants centered on the 45 degree compass points
(NE, SE, SW, NW). In column (1) all stands are given equal weight. In
column (2) the four slope means are first weighted equally to obtain
altitude means which are then weighted equally to obtain the mean STV
by slope-aspect groups.
1250 to 2249 feet (72 stands) 2750 to 4749 feet (79 stands)
Slope-
aspect
1 2 1 2 ;
NE 65.7 65.5 T2-3(6.7)* T0.8(6.3)*
E 64.2 67.0 10.0 9.9
SE 62.6 59.4 9.0 6.2
Ss 60.9 62.5 6.6 7.8
SW 58.5 62.3 9.8 8.4
W 57.1 60.5 9.1 7.0
NW 56.8 61.6 8.3 9.1
N 55.6 58.5 11.3(7.8)* 8.9(6.1)*
*Omitting two stands (see text)
-77-
Table 10: Mean STV for slope groupings for various selections of stands
and various weighting of altitude and slope-aspect groups. Column (1)
contains means for the entire sample with each stand given equal weight.
Column (2) is the same as {} except that means of altitude groups are
weighted equally. Column (3) is the same as (2) but stands not on
Whiteface Mountain or nearby peaks and ridges have been removed.
Column (4) contains means for the altitude range 2500 to 4000 feet for
three slope ranges with altitude groups weighted equally. Column (5)
contains means for the altitude range 1500 to 2500 feet for three slope
ranges with altitude groups weighted equally.
Slope ] 2 3 4 5
Range VN siv W stv Ww Stv WN STV Vv sv
Level 21 51.0 5 61.1 3 39.2 - + -
3° to 9° 51 59.8 6 58.5 4 43.2 - + 3 56.6
10° to 19° 47 46.3 8 46.4 6 30,0 4 16.8 3 51.3
20° to 29° 47 31.3 7 31.6 7 29.6 4 21.1 3 56.3
> 30° 16 7.7 5 12.4 5 12.5 4 14.8 -
-78-
and north sites and the hardwoods species with higher index
numbers are more common on east and south facing slopes.
In Figure 6b there are two stands plotted which contained
sugar maple, both on north-east facing sites. These were
found at 3100 ft and were unusual sites for this altitude.
Sugar maple was never found at a higher altitude. The high
STV for these two stands greatly influenced the mean values
in Table 9 which gives the mean for the appropriate slope-
aspect group both with and without these maple stands.
The one or two leading species are indicated by the
appropriate letters in the scatter plot in Figure 6, The
tendency for “clumping" of species is not as obvious as in
the altitude plots in Figures 2 and 5. Red oak and sugar
maple tend to be more common in the range from northeast to
south and hemlock and yellow birch more prominent on west and
north sites, but this is not obvious,
Slope Variation
Changes in drainage caused by slope steepness differences
should influence vegetation. The problem of detecting such
variations in the Whiteface sample is difficult because the
steepness of the slope is highly correlated with altitude
(r = .67). Table 10 lists several methods of obtaining the
mean STV by five slope ranges. Column 1 gives the data with
no weighting. Data in Table 5 reveals that this procedure is
not valid because high altitude ranges did not include many
stands of low slope. Civing 500 foot altitude groups equal
weight still does not resolve the problem (columns 2 and 3)
for the same reason. In bath cases STV decreases markedly
with increase of slope, but only because of the high negative
correlation of STV with altitude and positive correlation
of altitude with slope steepness.
The last two columns in Table 10 give comparisons of mean
STV for three slope-steepness ranges which were included in
all of the altitude ranges for the particular column. For the
2500 to 4000 foot range in column 4 the interpretation is that
the most “mesic” sites are on slopes ranging from 20° to 299
and the most boreal on slopes greater than 30° but it is un-
‘likely that the sample was large enough for.this to be proven.
In column 5 for the range of 1500 to 2500 ft the interpreta-
tion is that there is no variation with slope steepness but
again the sample may be too small,
DISCUSSION
The Concepts of the Continuum and Association
The argument has not been settled as to whether natural
vegetation is composed of discrete units or rather that it
consists of a distribution of individuals based on chance and
regulated only by environment. The argument is not whether
-79-
Figure 8: Hypothetical plots of a measure of mean
species importance versus a measure of
environment for three cases (page 80).
Figure 9: Hypothetical plots of a measure ofa
vegetation gradient (i.e, stand index
f
value) versus a measure of environment for
| four cases (page 81).
MEASURE OF SPECIES IMPORTANCE
~80-
c DIFFICULT TO INTERPRET
b ASSOCIATIONS ("DISCRETE TYPES")
a CONTINUOUS (LINEAR)
MEASURE OF ENVIRONMENT
~8i-
100
LINEAR AND CONTINUOUS
("CONTINUUM")
re)
fe}{e)
NON-LINEAR AND CONTINUOU:
(TENDENCY FOR ASSOCIATION)
NON-LINEAR WITH "PLATEAUS" -
(DISCRETE TYPES
WITH “ECOTONES")
GRADIENT (RELATIVE)
fe)
MEASURE OF VEGETATION
Blo
DISCONTINUOUS j
d.
"(DISCRETE TYPES
— WITH DISCONTINUITIES).
LINEARIZED MEASURE OF
~ ENVIRONMENT
~82~
it is most appropriate to use either classification or ordina-
tion techniques in vegetation studies because as Lambert and
Dale (1964) point out both methods are useful when properly 3
applied to the purpose for which they were designed.
The problem is whether or not species tend to clump into
types (or “associations") along gradients of uniformly
changing environment. Daubenmire (1966) has presented argu-
ments in favor of the association point of view, This hypothe-
sis implies that vegetation dynamics produced by such things
as biologic competition or tendency of certain species ta
coexist to produce mutual benefit causes clumping along
the gradient of environment,
The individualistic hypothesis originally proposed by ‘
Gleason (1926) views vegetation as a collection of species ‘
which associate merely because they have overlapping ranges
of tolerance along environment gradients, Tha existence of
classifiable types would therefore result from chance factors
such as accidents of seed dispersal or similarities in genetic
history. Vegetation is viewed as a linear function of environ~
ment or a continuum by Curtis and McIntosh (1951) and Curtis *
(1959), Evidence in favor of this hypothesis is reviewed by
McIntosh wae and the subject is treated in detail by
McIntosh (1967).
If these two concepts have been properly defined then it
seems obvious that they cannot both be correct. The debate :
has been largely one of semantics due to differences in approach :
and data treatment, As Lambert and Dale (1964) point out, #
much of the argument has resulted from the use of inappropriate
methods, For example, ecotones or discontinuities may appear
when the environment has not been adequately measured or defined.
Ecotenes may be regions where the environment as well as the
vegetation undergo rapid change. On the other hand, ecotones
and regions of clumping may be rather subtle and could be
meeked by not specifically looking for the effect in the data By
analysis, .
Most investigators who have interpreted vegetation as a
“continuum” have examined measures of species importance along
- 80me measurable quantity such as space, time, temperature and
moisture. McIntosh (1963) gives many examples of ‘these and
Figure 7 shows the result for the Whiteface Mountain sample.
From this kind of analysis it would be difficult to tell
whether or not ecotones and clumping exist. The problem is
illustrated for three hypothetical cases in Figure 8 If the
distributions appear as case (a) the interpretation would be
that vegetation is a continuous linear function of environment
although this cannot be proven by the method. Case (b) is
one where there is obvious clumping of species curves jalong
the environment gradient and the interpretation that types
or associations exist would be proper. In case (c) there may
be some tendency for clumping but it is difficult to prove
this interpretation from the approach used, :
A better approach than plotting measures of species ‘
importance is to relate a measure of the vegetation gradient
ws
-83-
such as the stand index value to the measure of environment.
The hypothetical results of this kind of analysis are shown
in Figure 9 for four cases,
This kind of study requires a stochastic model because
the data from stand composition will give a large scatter when
plotted (see Figures 2 and 3). However, the analogous analytic
function can be used once the statistical relation is determined.
Thus, case (a) would be a linear continuous function and would
indicate that vegetation is a “continuum”. In case (b) the
function can be continuous but non-linear. Here the slope of
the curve changes indicating ecotones and regions of clumping.
This tendency for association could be due to “vegetation
dynamics", If case (c) is the result the plateaus where the
Function is constant would represent definable types which would
be easy to classify, Case (d) represents discontinuities of
vegetation along a uniformly changing environment and classi-
fication is the best procedure.
Such a study as described in Figure 9 can only be accom~
plished by the techniques which obtain the vegetation gradient.
Classification would be inappropriate for case (a) but would
be proper in all of the others, It should not be assumed that
vegetation is not a linear function of environment before
it is proven. The methods described by Daubenmire (1966)
provide only circumstantial evidence that “discontinuities"
exist. The difficulty is in measuring the environment. To
the knowledge of these authors the only published work which
has attempted to resolve the problem described in Figure 9 is
by Loucks (1962). He related a synthesized ordination of
several environmental measurements to a vegetation ordination.
His interpretation is that the vegetation in his study area
is a linear function of environment (i.e. a continuum).
Loucks did not find the subtle kind of variation which may
have led to the interpretation that ecotones exist as in case
(b) of Figure 9,
At first glance the results for the Whiteface vegetation
sample could be interpreted as case (b) of Figure 9 (see
Figures 2 through 5). There seems to be an “ecotone" in the
altitude range of 2500 to 3000 ft between the hardwood stands
and the high altitude spruce-fir communities, Unfortunately,
it cannot be proven that environment is-a linear function of
altitude. There are reasons to believe that many environmentally
related parameters also change rapidly between 2500 and 3000 ft.
For example, the heat balance of the high altitudes may be
affected by the fact that Whiteface protrudes above the lower
peaks which are not generally higher than. about 3000 ft.
The soils above 3000 ft are thinner and more rocky than below
where thick layers of glacial parent material are more common.
A cloud cap covers Whiteface Mountain above 3000 ft on about 50%
of the growing season days. The fog drip from this cloud :
influences the moisture balance of only the high altitude regions.
The cloud cap causes higher absorption of solar radiation for
the stands above 3000 ft compared to lower stands.
~84-
The proper approach to be used to prove whether or not
the Whiteface sample can be interpreted as case (b) of |
Figure 9 would be to obtain independent measures of environment
which are not arbitrary properties of the site such as
altitude. The independent measures from a large number of
stands. can either be combined statistically or related to
the vegetation separately. The problem of combining several
environmental measures to obtain a “linear” result has-been
treated extensively by Loucks (1962). Because a large number
of stands are required many of the environment measures may
have to be obtained by using analytic techniques, Lettau
(1952) has treated the problem of synthesizing many of .the
climatic variables. If a large sample is used, the errors
in determining the stand environmental measures can be smoothed
and subtle changes in the plot of vegetation versus environment
may be revealed. “
Possible Causes of Slope-Aspect Variations of STV
The interpretation of the slope-aspect variation of STV
from Tables 8 and 9 and Figure 6 is that there is greater
contrast between east and west facing sites than between -north
and south. The east sites are the most "mesic" for all
altitude ranges. The west sites are apparently the most
“boreal” followed by north and then south, but the small
difference between the latter three quadrants of slope-
aspect may not be significant at least from the Whiteface
sample.
If the east-west contrast is indeed larger than the north-
south contrast then certain interpretations about the environ-
ment can be made. Secause the north to south contrast in
solar radiation is large compared to any east to west contrast
which may exist then the total input of solar radiation to a
stand must be ruled out as an important cause of vegetatian
differences, It is rather the way in which the solar radiation
input is divided which may be important. For example, the
solar energy on south slopes may be used for evaporating
available moisture so that the temperature does not rise much
higher than on north and west sites. fecause only the east-
facing sites are protected from strong winds (prevailing
direction SSW) then more of the solar input can be used to
warm the stand compared to the other three aspects, The pro-
per approach to an environmental analysis would be to examine
the heat and moisture balances of a large number of stands,
Another contrasting feature of east versus west~facing
sites involves the timing of maximum solar radiation input,
For clear skies east sites receive their maximum input during
the morning and west sites during the afternoon but the moun~
tains cause cloudiness to vary diurnally. Measured solar
radiation at the Whiteface Mountain Field Station over two
summers was found to be 15% lower from local ncon to sunset
than from sunrise to noon. Thus west sites received less
radiation. The afternoon radiation also contains more of the
~85-
diffuse (non-directional) component which is absorbed eqully
well by east and west sites. The contrast in solar input. from
west to east would, therefore, be mora than 15%,
The diurnal variation of the wind speed may influence
the manner in which solar radiation affects east and west sites,
The maximum solar input.on east sites eccurs during the morning
when the wind speed is relatively lew, The radiation is, there-
fore, utilized in warming the stand. The wind speed-is
climatically higher in the afternoon when the west slapes
receive their maximum input, The energy would, therefore, be
rapidly lost by the stand as evaporation or sensible heat flux,
This effect is qualitatively observable by the fact that it
is more comfortable to work on the east side of the mountain
on cool windy days.
Another contrast between east and west sites which is
not so severe for the north-south case exists when rime icing
occurs following major storms in the westerlies, These frost
conditions are most common in late spring and late summer but
occur in all growing season months, The rime icing is most
severe on west sites followad by north and then south, On
several occasions icing was observed to be serious on the
west side of Whiteface down to altitudes as low as 2700 ft
while there was no icing on the east. On these days, icing
on northwest and southwest sites was not as severe as on the
west but southeast and northeast sites were nearly free of
timing.
a Because the winds are far more severe on the west~facing
180° of slope-aspect wind disturbance is also more obvious.
Wind-throw patterns on high altitude west-facing sites seem
to be a regularly occurring phenomenon while only scattered
blow-down is found on east sites, The wind-throw may be a
factor which selects for balsam fir and thus stands of low STV,
The blow-down is followed by rapid reproduction of balsam fir.
Thus, repeated blow-down at intervals of about 50 to 70 years
would help to maintain nearly pure balsam fir stands.
+86-
REFERENCES
Ayyad, M. A. Ge and Rw L. Dix, 1964, An analysis of a
vegetation-microenvironmental complex on prairie slopes
of Saskatchewan. Ecol. Monog. 34: 421-442,
Beals, E, 1960. Forest bird communities in the Apostle Islands
of Wisconsin, Wilson Bull, 72: 156-161.
Brown, R. T. and J. T, Curtis, 1952. The upland conifer
hardwood forests of northern Wisconsin. Ecol. Monog. 223
217-234,
Buell, M. F., A. No Langford, 0. W. Davidson, and L. F. Ohmann,
1966, The upland forest continuum in northern New Jersey.
Ecol. 47: 416-432,
Cottam, G. and Je Ts Curtis, 1956, The use of distance measures.
in phytosociological sampling. Ecol, 37: 451-460,
Curtis, J. Ts, 1959, The vegetation of Wisconsin, Univ. of
Wisconsin Press, Madison, Wis,
» and Re P. McIntosh, 1951. An upland forest continuum
in the prairie-forest border region of Wisconsin.
Ecol, 22:°. 476-496,
Daubenmire, R., 1966. . Vegetation: identification of typal
communities, Science 151: 291-298,
Fernald, M, L., 1950, Gray's manual of botany, Sth edition.
American book co., New York, 1632 pp.
, Gleason, H, A., 1926, The individualistic concept of the
Plant association. Bull, Torrey Bot, Club 53: 7-27,
Goff, F. Ge and P. H, Zedler, 1968, Structural gradient
analysis in the Western Great Lakes area, Ecol,
Monog. 38: 65-86,
Goff, F. G. ard G. Cottam, 1968. Gradient analysis: the use of
species and synthetic indices, Ecol. 48: 793-806,
Goodall, D. W., 1954, Objective methods for the classification
: of vegetation. III. An essay in the use of factor
analysis. Aust. J. Bot, 2: 304-324,
‘Grasenbaugh, L. R., 1952, Plotless timber estimates ~ new,
fast, easy. J. For. 50: 32-37,
~87-
Holway, J. Ge, Je Te. Scott, and S, Nicholson, 1969, Vegetation
of the Whiteface Mountain region of the Adirondacks,
Publication No. 92, Atmosph, Sci. Res, Center, State
Unive NeYe at Albany, Albany, N. Y. 12203,
Lambert, J. M. and M, B. Dale, 1964, The use of statistics
in phytosociology, In: “Advances of Ecological
Research", Academic Press, Inc., New York, Vol. 2,
55-99.
Lettau, H., 1952, Synthetische klitatologie. Ber. des sent.
Wetterdienster in der U. S. Zone. 38: 127-136.
Lindsey, A. Av, J. D, Barton, and S. R. Miles, 1958, ° Field
efficiencies of forest sampling methods, Ecology 39:
428-444,
Loucks, 0. L., 1962, Ordinating forest communities by means
of environmental scalars and phytosociological indices.
Ecol. Monog. 32: 137-166.
Mason, He J. and J. Langenheim, 1957. Language analysis and
the concept Environment. Ecol, 38: 325-339,
‘MeIntosh, R. P., 1963. Ecosystems, evolution, and relational
patterns of living organisms. Amer. Sci. Sl: 390-395,
McIntosh, R. P.,1967, The continuum concept of vegetation,
Bot. Rev. 33: . 130-187,
Mowbray, T. B. and H. J. Gosting, 1968, Vegetation gradients
in the Southern Blue Ridge Gorge. Ecol. Monog. 38:
309-344,
Whittaker, R. H., 1956, The vegetation of the Great Smoky
Mountains. Ecol. Monog. 26: 1-80,
’ 1960. ‘Vegetation of the Siskiyou Mountains, Oregon
and California. Ecol. Monog. 30: 279-338,
and W. A. Niering, 1965, Yegetation of the Santa
Catalina Mountains, Arizona: gradient analysis of the
south slope. Ecol. 46: 429-451,
Witty, J. £2, 1968, Classification and distribution of soils
: on Whiteface Mountain, Essex County, New York. Ph.D,
thesis. Dept. of Agronomy. Cornell Univ. Ithaca,
New York,
MULTI-DIMENSIONAL ORDINATION
OF BOREAL AND HARDWOOD FORESTS
ON WHITEFACE MOUNTAIN
By
Alvin R. Breisch, Jon T. Scott
Richard A. Park and Paul C. Lemon
~92-
Altitudinal changes of almost 4000 ft within a distance
of three miles, coupled with the rugged topography of a
mountainous area make the environment and therefore the
vegetation quite heterogeneous, The effects of logging and
fire in lower elevations add greatly to the diversity of this
region. The growing season for this general area is reported
to be from 90 to 104 days (Ferree & Hagar, 1956), Heimburger
(1934) considered the strong winds and low temperatures to be
limiting in the upper elevations, while soil moisture is the
major factor controlling disttibution of species in the lower
elevations.
The nomenclature of the vascular species in this report
follows that of Gray's Manual of Botany, 8th ed. (Fernald,
1950) with the exception of yellow birch which is here referred
to as Betula alleghaniensis. A complete list of species names,
common names and symbols used in figures is given in Appendix I.
The nomenclature of bryophytes follows Ketchledge (1957).
Heimburger (1934) describes three ve etational series in
the Adirondacks - a high altitude series ?esubel ine") and
tuo low altitude series (“western" and menstorn”)« These
series are composed of twenty-two forest types, nine of which
are considered in this study, The forest types are named for
the most characteristic ground flora species in the associa-
tions, This is done to enable one to recognize successional
stages before the characteristic shrubs and trees have become
established.
The subalpine series is equivalent to the type designated
as spruce-fir by Ferree and Hagar (1956). In addition to
the dominants Abies balsamea and Picea rubens, Betula papyri-
fara var, cotdifolia and Pyrus dgcora are common associates,
Tn this forest type, bryophytes (especially the feathery
mosses Ptilium crista-bpastrensis, Hylocomium s lendens, and
Pleurozium schreberi) often dominate the ground flora,
Cornus canadensis, Oxalis montana, Dryopteris Spinulosa, Maian~
themum canadense, Clintonia borealis, Solidago macrophylla,
and Aster acuminatus are the most abundant vascular species
included in the ground flora.
The "western" and "eastern" series of Heimburger (1934)
correspond respectively to the northern hardwood and mixed
oak-northern hardwood types of Ferree and Hagar (1956), The
dominant trees in the western series are Acer saccharum,
A. rubrum, Fagus grandifolia, Betula alleghaniensis and B.
papyrifera with an admixture of Abies and Picea. The eastern
series contains Acer saccharum, Tilia americana, Fraxinus
americana, Quercus rubra var. borealis and Ostrya virginiana.
Bryophytes do not assume a dominant role in this series,
Ferree and Hagar (1954) also mention a pioneer forest type
composed of 8. papyrifera or B, papyrifera var. cordifolia.
Heimburger (1934) considers this as an early successional
stage and classifies it according to the ground flora.
-93-
In addition to these forest types several vegetation associations
exist on Whiteface which were not included in these studies,
One such type is the alpine zone which lies above treeline.
Another is the exposed ridges dominated by Vaccinium species.
Witty (1968) gave detailed descriptions of the soils on
Whiteface Mountain. Two major groups have been described,
spodosols (mineral soils) and histosols (organic soils).
The spodosols are moderately deep to deep, well-drained soils
that occur mainly at lower elevations on gently to steeply
sloping terrain. The histosols are divided into two groups;
(1) very shallow to moderately deep, moderately to well-
drained on slightly to very steep slopes, and (2) moderately
deep, poorly to moderately well-drained, on level to moderately
sloping soils.
The spodosols were located mainly in northern hard-
wood areas but also included pine stands, open blueberry
ridges, some above timberline areas and well-drained spruce-
fir stands. These high elevation mineral soils were moderately
high in organic content. The histosols occurred mostly in
spruce-fir stands and at least in the poorly drained areas
were usually associated with sphagnum moss, which is noted
for a tremendous water=holding capacity and low pH (Grout,
1903).
The organic soils are found chiefly at the higher eleva-
tions of Whiteface Mountain and the surrounding associated
peaks. Arms of these histosols extend along the major ridges
from the summit. The boreal zone = Picea, Abies and associated
ground flora species ~ occupies most of this area. In addition,
on the east side of the mountain, which is composed of a
series of east-west ridges separated by steep-walled cirques,
the boreal zone extends to between 2500 and 4000 ft. This
area has high organic content mineral soils,
The lower areas to the east side of Whiteface are
characterized by mineral soils with relatively thin "0"
horizons. This area is chiefly Acer saccharum with admixtures
of A. rubrum, Quercus and Tilia. The stands with southern
exposures in the low elevations are also on mineral soil
but here the "0" horizon is ganerally much thicker, The
slope is not as severe on this side where the dominants are
sugar maple, beech and yellow birch.
FIELD METHODS AND DATA ANALYSIS
Criteria for stand selection
The stands used in this study were selected from 182
stands sampled from 1964 to 1966. A stand was defined as a
sample of vegetation from an area of uniform slope, aspect,
and altitude, Details of stand selection were reported by
Holway and Scott (1969), The techniques used in any ordination
-94-
model require only that a wide range in stand composition
be sampled (Greig-Smith, 1964) and therefore no attempt was
made to obtain a random sample.
In addition to the field criteria for stand selection in
the field, stands were eliminated from further consideration
if there were not at least twenty points sampled (with excep-
tion of two stands at 4500 ft where 15 points were used).
Since the stands used in this report were originally sampled
as part of a forest tree study (Scott & Holway, 1969; Holway
& Scott, 1969) stands in which the ground flora data was
scanty or missing were also not used. Since the major consi-
deration of this study was an ordination from the subalpine
spruce-fir zone to the lowland northern hardwood-sugar maple~
beech forests, stands that were present as a result of pri-
marily edaphic conditions (Isuga, Thuja and Pinus) or remnants
of earlier forests (Pinus) were eliminated.
The final criterion for accepting a stand was a test for
homogeneity (Greig-Smith, 1964), The quadrats in each stand
were combined into four groups and the data for the dominant
trees in these groups were tested for. homogeneity using the
Chi-square test. A stand was judged as homogeneous if the
variance among the groups was less than could be expected
by chance alone. All heterogeneous stands were eliminated
from the study.
| Field methods
The field work for this study took place during the
summers of 1964 to 1966 and has been described in detail
by Holway and Scott (1969), The vegetation was recorded in
three size classes, Trees were considered to he woody stems
over 4" dbh; saplings from 1" to 4" dbh; and ground flora as
all vascular plants under 1" dbh. Non-vascular plants
were not included due to difficulty of identification.
Figure 1 shows the location of most of the selected stands,
the others are located outside of the mapped area.
Although the same information was derived for gach stand
the method was altered during the second summer of sampling
to obtain more efficiency in the field without loss in accuracy
of measurements (Holway & Scott, 1969). The ground flora
was always sampled in a series of one meter square quadrats,
All species rooted within the quadrat were recorded. In 1964
(Nicholson, 1965) the trees and saplings were sampled by the
point-centered quarter method developed by Cottam and Curtis
(1956). In 1965 a modified forester's prism method using a
#30 prism was used to determine basal area. Density was
recorded from a 1/80 hectare circular plot. The quarter
method was still used to obtain frequency. Efficiencies and
techniques of the quarter and prism methods are discussed by
Lindsey, Barton and.Miles (1958) and Lemon (1962), The quarter
method should be used if dbh size class information is needed.
‘
a ’ )
: or Wr,
> f SCALE IN MILES oN ®~Stand number .
¢ contour interval = 250!
FIGURE |, | Map of study area showing location of stands.
-96-
However, the prism method is more efficient in the field
and laboratory if only mean basal area is needed, as in
computing importance value.
Analysis of stand data
From the stand data for trees, relative density, relative
frequency, and relative dominance (basal area) were combined
to give the importance value (IV), Formulas for calculating
these parameters and a discussion of their relationship to
each other can be found in Curtis and MeIntosh (1950). Since
basal area was not recorded for saplings, their importance
value was based only on relative frequency and relative
density, The importance values for both trees and saplings
were converted to percentages to give the relative importance
value (RIV), The only measurement recorded for the ground
flora was frequency. As Goodall (1952) pointed out, frequency
measurements alone ara not a reliable measure of relative
quantity, but frequency is useful in determining distributions
af species, Bray and Curtis (1957) were able to construct
meaningful gradients using only frequency measurements for
both herb and shrub layers.
Seventy-one stands were used for the final analysis,
The twenty-one most common tree spacies were included in the
study. After eliminating all ground flora species of question=
able identity (immature plants, seedlings, etc.) 82 species
remained, Species that occurred in less than 10% of the
stands or which contributed less than 0,5% to the total rela~
tive frequency were eliminated from the calculations. This
was done for three reasons: (1) to eliminate the possibility
of rare or accidental species from biasing the data, (2) fre=
quency measurements tend to overestimate rare species (Goff
& Cottam, 1967), (3) to simplify calculations and save computer
time. Most of the species eliminated appeared in only one
stand as a single record in a series of 20 to 40 square meter
quadrats. The 27 species finally selected account for 81%
of the total frequency. Of these, twelve are seedlings of
tree species,
The ordination model
Ordination models are based on a variety of indices.
Goff and Cottam (1967) compared six methods of gradient
analysis and found that the results of the Index of Similarity
method (Bray and Curtis, 1957) showed the highest degree: of
correlation (r = .93) with the other methods, This method is
not only easy to visualize but is relatively simple to cal-
culate, It is based on Sorenson's Index, 1 = (2w/(atb)),
where "a" and “b" are RIV values of a species in any two
stands while “w" is the amount of the species RIV the two stands
have in common, summed for all species in the stands. Since
the sum of species RIV for each stand is equal to 100% this
-97-
Stand C is placed Z distance from stand A,
| The distance Z is calculated by the formulas
2 2 2
where Wa dissimilarity between endpoints,
X = dissimilarity between endpoint A and stand C,
Y » dissimilarity between endpoint B and stand C,
Figure 2: Geometric methad of determing position of
stands along ordination axis.
~98-
formula reduces to I =Xw/100, where “w" is the sum of -
the smaller RIV values for each species being considered.
The index used in this study is the Index of Dissimilarity
and is equal to 1.00 ~ (= w/100) expressed as a percent.
The assumptions of ordination are that the samples cover
a range of values along a gradient and that each sample shows
some degree of similarity with another sample, That is, no
stand is tatally dissimilar from every other stand. A series
of samples which show no similarities would not show gradients
and therefore would not fit in the ordination model of gradi-
ent analysis, Ideally, no stand should be completely dissimi-
lar from any other stand, since a dissimilarity greater than
100% cannot be determined between two stands,
Another important consideration is that the ecological
nature of a species is relatively constant throughout the
area being considered (Goff & Cottam, 1967). This would elimi-
nate areas in which the species exhibit obvious ecotypes or
intraspecific clines that are not separable except in physio-~
logical terms (Goff &-Cottam, 1967). The results would be bi-
modal species-abundance curves in the first case and platykur~
tic curves in the second example. A species or character
that shows no difference along the gradient also adds nothing
to interpretation.
The selection of end points for each axis in an ordination
can be accomplished by a variety of methods, A subjective
method is to select end points which represent extremes
of the gradients in question, such as open canopy to dense
shade, xeric to mesic, or early successional to late succes=
sional, Since two simultaneous ordinations were being per-
formed on two separate groups of data for the same series of
stands it was decided that an objective method in which the
computer program selected end points for three axes would
be preferable, This would remove any subjective bias permit-
ting comparison on statistical criteria, The “standard
‘deviation criterion" devised by F. G. Goff (Park, 1968) was
applied. In this method the stand which differs most from
the mean stand composition is selected for the first end point
and the stand which differs most from it is selected as the i
Opposite end point. The distance between the end points is
equal to the dissimilarity between the two stands selected as
end points. The placement of stands between the end points
is determined by the amount of dissimilarity between the
particular stand and the two stands chosen as end points.
The method is shown schematically in Figure 2. Two stands
which occur close together on this axis are not necessarily
similar in composition or environmental requirements (Bray
& Curtis, 1957), Their only factor in common may be that they
have equal dissimilarities from the two end points, This dif-
ference becomes obvious when the second or third axes are used.
In constructing additional axes the end points are selected so
as to reduce as much of the remaining dissimilarity as possible
and at the same time keep the axes orthogonal.
=99-
Cluster analysis
The second method.used to obtain an understanding of the
vegetation on Whiteface Mountain was cluster analysis, This
method can be used to group associated species or stands.
The grouping of species should be similar to those representing
a vegetation type whereas the grouping of stands would be
similar to combining the stands that represent a vegetation type.
According to Harman (1960) the grouping by cluster analy-
sis is based on the theory that "the variables of a group
identifying a factor have higher intercorrelations than with
the other variables of the total set." As in ordination,
similarity between the stands was determined for cluster
analysis using Sorenson's Index of Similarity (1). Each stand
is coupled to the stand most similar to itself, The stands
are ordered in the dendrogram by the following procedure,
The stand coupled to the stand which shows the least similarity
to all other stands in the sample is used as the beginning
point. The stand most similar to the first was clustered
with it. The stand which had the highest sum of similarity
indices with the first two stands was added to the cluster.
Successive stands were added to the cluster based on the
sum of the similarity indices a stand had with all other stands
previously included in the cluster. The dendrograms for the
clusters based on the tree and ground flora data are shown
in Appendix II. The percent similarity used to separate the
cluster was determined subjectively, but was not based on
any prior knowledge of stands or species in the clusters.
In nearly all samples of vegetation, a small number of stands
do not seem to fit any cluster.
The relationship or distance between each cluster was
not obvious from these calculations. By superimposing the
results of the cluster analysis on the ordination model, these
distances can be determined. The discreteness of the clusters
or vegetation types can be inferred by the degree of separation
of the clusters when plotted on the ordination axes, A cluster
occupies a segment of a gradient, which may be overlapping
another cluster when only one dimension is viewed. In this
study, each cluster occupied a separate volume in the three-
dimensional space defined by the first three axes of the ordina~
tion. model,
RESULTS
Distribution of species along topographic gradients
The change in community composition from lowland northern
hardwood forests to subalpine spruce-fir forests is obvious
in all synusiae of the forests, The differences between the
ridges and valleys and the variations related to slope aspect
-100-
1005
754
50
254
* °
i Betulo pepyrifera Mersh. Betula @leghanionsis Britt,
3 so4 ¥. Gor ditette (Regel) Fern, 4
a.
Relative
2000 3000 4000 2000 3000
ALTITUDE (FEET)
—— TREES (34"ébh) in importance Value
sesnsnen SAPLINGS (1% 4" dba) in Importance Value
=~ SEEDLINGS (<I"dbh) in Absolute Frequency
FIGURE 3A. Distribution of seedlings, saplings and
trees for six major species against altitude.
Abundance
Relative
-101-
—— TREES (1V.)
Pyrus decora (Sarg.) Hyland Acer pensyivanicum L.
304 ’ a
s.
’ \
tf \
4
, \
\
204
\
\
oN \
r \
\
105 / \ \
/ \ \
\
/
’ \ \
SA
os T f T t SFT
97 Acer spicatum Lam. Fraxinus americana L.
a
\
/ ss
104 / \ \
‘ \ \
/ \ \
4 \ \
7 ” \ \
/ -—" —S
o- => T T T
cer ¢ubfum L. Ostrya virginiana (MIIL)K.Koch
“sou. \
\ \
\ \
x NY
‘ a
“SY —_ ss
o_- : T T T T
2000 3000 4000 2000 3000
ALTITUDE (FEET)
--- SEEDLINGS (Frequency)
FIGURE 3B. Distribution of seedlings and trees
for six minor species against altitude.
~102-
q | obo | | \
2000 3000 4 2000 3000 4000
ALTITUDE (FEET)
‘Aa-— Aster gcyminatus Michx. Om——Qxalis montana Raf.
An—Aralia nudicoulis L. Sm-—— Solidago macrophylla Pursh
Cb—Clintonia borealis (Ait.) Raf, Sr—— Streptopus roseus Michx.
Co— Cornus canadensis L. , Th-— Triontalis borealis Raf.
Co—Coptis groenta groeniandica (Oeder) Fern. To—— Tiarelta cordifolia L.
Ds-—Dryopteris js spinulosa(O.FMuel) Watt =. Uv— Uvutaria seselitalia L.
Li—Lycopedium lucidulum Michx, Va—— Viburnum gtoifolium Marsh.
Mc—Melanthemum canadense Desf.
FIGURE 3C. Distribution of ground flora. (percent
frequency) against altitude.
-103~
are not quite so evident. A fairly detailed analysis of the
distribution of trees, saplings and some of the ground flora
species with respect to altitude and slope aspect can be
Found in the reports by Scott and Nicholson (1964) and Nichol~
son (1965), Scott and Holway (1969) and Holway and Scott (1969),
Altitude is closely correlated with changes in vegetation
and is therefore the major gradient considered here, However,
because climatic, edaphic, and biotic gradients such as tempera
ture, biomass, soil characters and light are affected by
altitude, it should be considered a “complex gradient" (Whit-
taker, 1956).
The abundance measurements are averaged for 500-foot
altitude intervals and plotted in Figures 3A, 38 and 3C,
Tree and sapling abundance is graphed according to RIV, The
seedlings and ground flora are measured in frequency. In
Figure 3A data for trees, saplings and seedlings of the six
dominant tree species are plotted, In Figure 3B the data for
trees and seedlings for six minor tree species that were included
in both ordinations are plotted and in Figure 3C the frequency
of the fourteen ground flora species included in the ordination
and cluster analysis are plotted,
The curves for the trees, saplings and seedlings of each
species are in agreement with each other except for Abies and
Betula papyrifera var. cordifolia. for these species the
Seedlings reach a maximum at elevations different from the
saplings and trees, In Abies the seedling maximum is at 3500 ft
as opposed to 4500 ft for the other size classes, This could
be because of recent unfavorable conditions at high altitudes
which limit reproduction in this species or to low seedling
success at all times in established sites at high elevation.
8. papyrifera var. cordifolia seedlings have their maximum
abundance at 3500 to 4000 ft which is higher than for the
larger size classes, Since this is mainly an early successional
or disturbance species, the high frequency of seeditage at
the upper elevations may be the result of re-occurring recent
disturbances caused by the severe weather conditions or may
represent the general failure of seedlings to survive beyond
this stage at these higher elevations. This species is also
being replaced in some of the lower areas in which it had
become established following the logging and fires in the
latter part of the last century.
The curves of abundance along a gradient give the environ-
ment tolerance (i.e. ecological amplitude) of each species,
Ecological amplitude implies that a species inhabits a sector
of the gradient rather than just a point (Daubenmire, 1968).
They approximate a normal distribution along the gradient but
have three distinct shapes for the tree species. The species
most closely approximating a normal distribution have their
maxima somewhere toward the center of the altitudinal range
dropping off to near zero at both ends, This type of curve
is exemplified by Picea, Betula alleqhaniensis, Acer spicatum
and Pyrus. Abies-is typical of the type which reaches its
~104-
T84 “
rn)
oe
2
3
>
$07
<
2
=
2 10
s
2
3
28-4
oe)
4
x
< Wi *20 3
(2
1% e
40
° v Tv T T
8
78-1
$6
+66
Jo
3 44. 61 3
s 164 Gh 88 152
$07 ws
Eid +38 “sy
a 53 2h 136 162
s 3 ae a6
3 ” * decreasing altitude ss ely
< 10 42 :
= a 0 oY 71
2 age “15 6331
S& |? ann as vee ” 39 ‘no
Wag 28
28- 3 Tig Ht 58 37 po”
” ;
a 38 ae x 59 55 go
a
Pi
<x 67
cal
9 u T U
o sf? . so 75 100
Axis |—— Ordination Value
FIGURE 4, Distribution of stands in ground flora
ordination,
-105-
maximum near the upper limit of altitude and thus appears as
only the left-hand side of a normal distribution because the
entire range of environmental tolerance for the species does
not exist in the area. Acer saccharum and Fraxinus have their
maximum RIV near the lower end of the gradient, thus appearing
as the right-hand side of a bell-shaped curve. The other
species exhibit varying degrees of truncated normal curves.
The shapes of most of the curves (species abundance
versus altitude) for the ground flora species in Figure 3C
are not as nearly expressed as are those for the tree species.
The curves for Tiarella cordifolia, Viburnum alnifolium and
Uvularia sessilifolia appear bell-shaped, but curves for
the other species have quite irregular shapes. AS a group,
their altitudinal distribution is not as distinct as is that
of the trees. The average frequency of Maianthemum canadense
is almost constant for a range of 2000 ft. At 3500 ft it
shows a sharp decline in frequency, then rises to twice its
previous value. Several species show bi-modal distributions,
including most notably Oryopteris spinulosa. for this species,
bi-modality would be due to ecotypic variation along the
altitude gradient, if the two ecotypes have different altitu-
dinal tolerances.
Species diversity also varies with altitude, For sap~
lings and trees there is a marked decrease of diversity with
altitude (Nicholson, 1965). Table 1 shows that the ground
flora layer has the same relationship of species as the trees;
that is, the number of species decreases with increase of
altitude for altitude classes from 1750 to 4250 ft. The
average number of species per stand is considerably higher in
the highest altitude class than in the one below it and also
has the greatest number of species per quadrat of any altitude
class. Since stands at upper elevations in general were
sampled with fewer quadrats a correction factor is included
to account for variation due to size of sample area, for
this correction factor the species diversity index (SDI) is
given bys
SDI, = (Wep/standg) x Ra/standnax ()
Wg/stand,
where A = stands of a particular altitude class,
max = altitude class with maximum value
(1750 to 2250 foot class),
N. p = average number of species, and
RG = average number of quadrats
The species diversity indices derived from this calcula-
tion show that the highest altitudes have potentially the
most diverse ground flora for a given area, even though the
lowest altitude class has the greatest number of different
species, This could be due to more frequent open areas in
the canopy, exposed boulders, or other phenomena that increase
the variety of microhabitats. This may also be related to
-106~
Table 1: Diversity of ground flora species in 500-foot
altitude classes.
Alt. No. of Total Ave. No. Ava. Now Species
Class Stands No. of Species Species Diversity
(Feet) Species Per Per Index
in Alt. Quadrat Stand (Eq. 1)
Class
4500 6 29 a. 768 16.83 31.14
4250
4250 16 28 4.01 10.37 18.76
3750
3750 15 37 - 4,02 15,13 19.85
3250
3250 16 57 4,44 17.56 22.86
2750
275016 72 3.93 19.88 21.42
2250
2250 ai 76 4.42 24,71 24.71
1750
1750 22 82 3.67 19,23 22.94
1250
TABLE 2
Correlation coefficients between topographic variables and the first three
axes of ground flora and tree ordinations.
ground | ground | ground
attitude | slope | stone | cris | axis 2 | axes] atl | at | ans’
altitude 1.00 | 60 |.663 +850} -.923 |}-458|-831 | -'5° 398
slope aspect 1.00 | i5s2 | -219 | -2 |-.0z9 |-.273] .o8 | -028
slope 100 |-629| 024 | -!55_ |-505) 20 |.394
tree axis | 1.00 | -.030 | .344|.858)| ou |-549 .
tree axis 2 1.00 | .202 | .292) .086 | .360 2
tree oxis 3 1.00 |.499] 403) -'05
ground flora | 1.00 | 13 -.208
ground flora 2 . 1.00 -025
ground flora 3 1.00
69. degrees of freedom
LH — P> .05
.280 —P<.05
.350 —P<.Ol
.500 — P< .0O!
-108-
the change in percent of species in each life form with
latitude and altitude, Raunkiaer (Kershaw, 1964) describes
vegetational trends in which the ground flora layer becomes
more predominant and the tree layer less pronounced at higher
latitudes and altitudes.
Three-dimensional ordination
The location of the stands along the first three axes
of the ordinations are plotted on Figure 4 for ground flora
and Figures 5A and 5B for trees. The stands are numbered
sequentially with respect to altitude, stand 1 is located
at 4500 ft and stand 71 is located at 1370 ft. These two
figures are similar in several respects and can be explained
with an understanding of the diversity of the vegetation and
the assumptions of ordination. The first axis should account
for the greatest proportion of dissimilarity existing in the
total sample, The amount of dissimilarity explained with
each axis varies with the method used to select end points.
Using the stand with the greatest variance (standard deviation
criteria) does not account for the maximum dissimilarity
possible. A subjective method, such as pre-selected end
points, is the most efficient method of explaining the
dissimilarity (Park, 1968). The second (third, fourth.,.nth)
axis is picked to explain as much of the remaining dissimilarity
as possible, According to Park, reduction of about 60 to 80
percent in dissimilarity is possible for three axes.
The interpretation of the first three axes for both
the tree and ground flora ordinations is that they correspond
to similar gradients, Axis 1 ordinates the stands along
the major environmental gradient which is related to altitude,
The low elevation stands are separated from each other along
axis 2, This is because the greatest dissimilarity remaining
after axis 1 is computed that can be explained with an axis
orthogonal to the first is in the low elevation stands which
have more species than the high altitude stands. This groups
the high altitude stands near the center of axis 2 as shown
in Figure 5A because they are equally dissimilar from both
low altitude end points (Figure 2). for similar reasons the
high altitude stands are separated in Figure 5B along axis 3
which groups the low altitude stands near its center. A plot
of axis 2 versus axis 3 would therefore not be meaningful
in this study.
Correlation coefficients were computed for all possible
combinations of the first three axes of the ground flora and
tree ordinations and the three topographic measurements -
altitude, slope aspect, and slope. The significant correlations
(P <.05) are listed in Table 2,
Altitude is significantiy correlated with axes 1 and 3
of both the ground flora and tree ordination. The approximate
direction of the altitudinal gradient is drawn on Figures 4,
5A and 5B, The degree of slope is significantly correlated
30
75-1
3 7”
=
s
2504
£
2
6
is -
Le
ea be wn”
2 Za
=
k ila
“esta at
a -
"0a? -20
2s
28
-2e
ik
° T —# T ]
° 25 50 75 100
Axis | —— Ordination Value
FIGURE 5B. Distribution of stands in tree ordination (axis | versus axis 2).
-601-
37
65
TS
52
s a
3
> as a
% Oe
ae 55 MP
3 * 44 60 OF
£ 2 2 70
Skee os decreasing altitude > ri
° 3232 22 2
| 53°28 38 wt 30 a a
N os
20
° oe 4
x O5—
225 zz +8 58
3 59.51
37
36 “4
6
2 63
0 T T rt J
te) 25 50 75 bo
Axis | — Ordination Vatue
FIGURE 5A. pistribution of stands in tree ordination (axis | vs. axis 2).
-OLL-
-111-
to axis 1 of the tree ordination and axes 1 and 3 of the
ground flora ordination. However, this may be a nonsense
correlation because slope is significantly correlated to
altitude, the higher elevations having steeper slopes, .
The firgt axes of the two ordinations show a correlation
with each other that is even higher than the correlations
with altitude, This indicates that the similarity existing
between the ground flora and trees is more than can be
explained by an altitudinal gradient, The use of a sub jective
method of end-point selection in the ordination model may
yield even greater similarities between tree and ground flora
ordinations. .
Even though the second axis defines a low altitude gradient
and the third axis defines a high altitude gradient in both
ordinations, the correlation coefficients comparing the
second and third axes of one with the second and third axes
of the other are not significant, Distinct similarities do
i exist and will be discussed later in relation to cluster
analysis. -
The great dissimilarity existing between the high and
| low elevation stands results in a clumping of dissimilar stands
| when axes 2 and 3 are considered. This is especially notice-
t able in Figure 5A where separation of the low altitude stands
i is good along axis 2, The result is that the high altitude
| . stands on the left-hand side of the Figure are very poorly
}
\
|
t
1
|
|
|
separated, For this reason comparisons of axes other than
the corresponding axes of the two ordinations may be ecologi-
: cally insignificant even though the correlation coefficients
i are statistically significant. for the same reason the
t graphs depicting the distribution of clusters in Figure 9 and
the leading dominants in Figures 11 and 13 on the tree
ordination are composites of axis 1 versus 2 and axis 1
versus 3,
i : A scatter diagram comparing the ordination values of
the stands using ground flora and tree data is shown in Figure 6,
In Figure 7 the average ordination values for both ordinations
are plotted for 500-foot interval altitude classes.
In Figure 6 the stands are chiefly distributed along a
diagonal from.the lower left-hand corner to the upper right=
hand corner. This. diagonal consists of two groupings, one
at each end of the. diagonal. The high altitude stands are
located mainly in the lower left-hand corner while the low
altitude stands are grouped in the upper right-hand corner,
The diagonal therefore represents an altitudinal gradient,
The relative lack of stands near the center of this diagonal
indicates that even though enough similarity exists in these
stands to construct a meaningful vegetational gradient from
the lowland northern hardwoods to the higher altitude spruce-
fir, these two groups have distinct characteristics,
The slope of the line representing average ordination
value against altitude in Figure 7 is very steep between 2500
10
04 70 1004
aw
a Son 47.
5754 5 so | 8 754
2 M gg 60-13
< 5 66° 55
2 65 37 62 68
° -¥8 $9, o
5%
BE oF 49-57 w
Zz 33 2
a « wiz
sS0- = * >507
-6
= 30 3 Shed " z
3 392 “~ 63 2
a 38 <
=z
a +20 =
3 36 a 4
0254 us wie 35-28 0257
& 220 wit =
“ we -40
7 «al 23 28
a -27 2s
$0, 53
ee” 24
Mg! to
°% 25 50 75 eo ° a .
2000 3000 4000
TREE ORDINATION VALUES ALTITUDE in Feet
FIGURE 6. Comparison of first axis of ground flora FIGURE 7. First axis of
and tree ordination.
tree(>——) and ground flora(---)
ordination vs. altitude.
“2lt-
,
t
i
-113-
and 3500 ft indicating either that a rapidly changing environ-
mental gradient with altitude exists or that an ecotone
occurs along a uniformly-changing environmental gradient.
Above and below these altitudes the slope of the line is not
as steep indicating the existance of spruce-fir and northern
hardwood groups, respectively, The need of two separate axes
to separate the high (axis 3) and low (axis 2) altitude stands
emphasizes this difference.
Relationship of clusters and leading dominants
to the ordination axes
The distribution of clusters on the two ordinations are
illustrated in Figures 6 and 9. Similarities exist between
the ground flora and tree ordinations, They are both composed
of three main clusters (see Appendix III for a list of stands
in each cluster), one high altitude and two low altitude.
The high and low altitude clusters can be thought of as
representing the spruce-fir and northern hardwood types,
respectively, The northern hardwood type contains two main
clusters in both the ground flora and tree ordinations.
The clusters are formed in such a way so that stands with
similar species are grouped together. In ordination, if
appropriate axes are used, the stands are placed so that the
ones close together are more Similar to each other than to
stands further away. The stands which occur in the two
clusters near the center of the right-hand side of Figure 9
are more similar to each other than they are to stands within
the same cluster near the end points. Even though the stands
have to be considered as parts of separate vegetation types
by the cluster analysis and therefore as "discrete" units,
they represent two very similar segments of a vegetation
gradient in the ordination model,
The similarity of dominant trees and ground flora in the
corresponding clusters of the two analyses is represented in
Table III, Index of Similarity (1) was calculated for the
trees by summing the RIV for all species with a value greater
than 20% for all stands in the cluster. for the ground flora,
all species were included with a relative frequency greater
than 15%. The clusters depicted on Figure 8 were formed by
analysis of ground flora with the dominant trees being super~
imposed after the clusters were formed, The dominant tree
and ground flora species for each stand were used as a method
of comparison between the two cluster analyses. The similari-
ties are due to the environmental gradients that control the
distribution of the ground flora and associated trees,
Figures 10, 11, 12 and 13 show the distribution of leading
dominants on the ordination axes (symbols for species listed
in Appendix I). The distribution of species is similar for
both ordinations, especially for the most common species.
The one major exception is Acer rubrum which clusters with
-114-
Table 3: Comparison of species in tree and ground flora
cluster analysis.
Tres Ang
Major species
High Altitude
Trees
Abies balsamea
ysis 9
Minor species Major species
Ground Flera Analysis IT
1
Minor species %
Betula Abies Pyrus
Betula papyrifera Picea Betula 93
papyrifera Betula Betula alleghaniensis
var. cordifolia alleghaniensis papyrifera
Picea rubens Pyrus decora var, cordifolia
Gro Flora
bies Acer Abies Acer
Oxalis pensylvanicum Dryopteris pensylvanicum
Oryopteris Acer rubrum Oxalis Acer rubrum
Betula Aster Betula Aster 95
papyrifera Viburnum papyrifera Viburnum
var. cordifolia var. cordifolia
Maianthemum MaLanthemum
Clintonia Clintonia
Coptis Coptis
Cornus Cornus
Picea Picea
Low Altitude 1
Trees
Acer saccharum Betula Acer saccharum Picea
Quercus papyrifera Quercus Tilia 77
Populus Populus
grandifolia grandifolia
Tilia Betula papyrifera
Ground Flore
Acer saccharum Maianthemum Acer rubrum Maianthemum
Acer spicatum Aster Acer saccharum Acer rubrum
Aralia Betula Acer Picea
Fraxinus alleghaniensis pensylvanicum Uvularia
Ostrya Dryopteris Aralia 73
Streptopus Fagus Fraxinus
Uvularia Trientalis Ostrya
Streptopus
Trientalis
Joer cent similarity of species composition based on Sorenson's
Index.
ma jon species ~ present most frequently or in high abundance
Minor species = present as. a subordinate species or exclusive
to one cluster
~115-
Table 3 (Continued)
Tree Analysis ,
Major species. Minor species
Low Altitude IT
eas
Acer saccharum Picea
Fagus
Betula
alleghaniensis
Acer rubrum
E
Acer saccharum Acer
Dryopteris pensylvanicum
Fagus Acer rubrum
Lycopodium Acer spicatum
Maianthemum
Oxalis
Picea
Solidago
Streptopus
Tiarella
Trientalis
Ground Flora Analysis I
Major species Minor species %
Acer saccharum Quercus
Fagus Tilia 78
Betula
alleghaniensis
Acer saccharum Acer
Dryopteris pensylvanicum
Fagus Viburnum
Lycopodium Uvularia 71
Betula
alleghaniensis
-116-
ONDA UOLJOUIPIO —— @ SIX
-uoljourpso 9844 uo suapsnjo JO uoINqisia §=°G ZYNOIS
anjpA woNDUIpso —— I SIXV
FOS
psd
OnjDA uo}}DUIp40
754
'S
fo}
ve
Axis 2——Ordination Valu,
gy
x
@ High altitude cluster
O Low altitude chister
X Non-cluster
o: 2
0 25 50 75 100
Axis! —— Ordination Value
FIGURE 8. pistribution of clusters on ground flora ordination.
wLil-
meP
757
ma
MG
gp MR s
MR u
qo
s §
s0- s ma
3 s § “oe
S$ sy
> F F MT
< Fes sve
q FF F
s |rc® Sev EM
5 rr _s ma ™
rE s ye YM 8 ™
F S Fs s FS 6M
% 28- FOF Scr le - u
2 6M
4 F c MT
BM 8M
aw 8M
uM
oO T T BM T 1
° 25 50 75 100
Axis | Ordination Value
FIGURE 10.
Distribution of leading dominants on ground
flora ordination
“BLL-
s
M
754 7S
s M ‘i .
3 mq M if 3
$s veut .
<807 é BM M 50.9
2 mM BM $
= ; * uc MoM 2
: 5 Fes oh 6
bs oe YM vm
" 85 BM mR .
2 F ¢ Fs 3
g2stre F BM rs
fe >.
c s Ye
F s YB
Ecy sys
° C)
0 25 50 50 75 100
Axis | Ordiaction Value
FIGURE
Distribution of leading dominants on tree ordination.
“6LL~
-120-
Betula allechaniensts and Fagus in the tree ordination (low
altitude cluster Il) and with Acer saccharum and Quercus in
the ground flora ordination (low altitude cluster I).
Vegetation groups: also exist for the ground flora (Table
3). These groups are not as distinct, mainly because of
the wider distribution of individual species, However, the
similarity of species between corresponding clusters of the
two models is at least 70% (low altitude cluster II, ground
ere and is as high as 95% (high altitude cluster, ground
flora).
The tree species in the three clusters are essentially
the same as reported by Heimburger (1934) as belonging ‘to
three vegetational series (Table 4). The high altitude
cluster corresponds to Heimburger's subalpine series. Low
altitude cluster Il, dominated by Fagus, Betula alleghaniensis,
Acer saccharum and A. rubrum, is equivalent to his western
series, Low altitude cluster I which contains the other
northern hardwood species corresponds to the eastern series.
The eastern series contains the drier habitat species, where=-
as the species of the western series are adapted to more mesic
conditions requiring a better developed soil and are, in
general, more shade tolerant. —
If the stands are arranged by increasing ordination
values for any axis, a distribution of species RIV results.
The midpoint of occurrence of the RIV for each species along
the ordination axis can be calculated. When the species mid-
point values are ordered and placed on a scale of 0.0 to 10.0,
a number results which is equivalent to the “climax adaptation
number” of Curtis and McIntosh (1951), In this study, Acer
rubrum was found to have a value of 10.0 and Abies a value
of 0.0. The midpoints for the 14 most common tree species
are plotted in Figure 14. The species composing the high
altitude cluster are widely separated from the others, The
species in low altitude cluster I have values from 7.8
to 8.9, while the species of low altitude cluster II have
values from 8.8 to 10.0. The species midpoint value assigned
a species has meaning only with regard to the gradient it
was derived from -. in this case altitude.
DISCUSSION
Gradients and the n-dimensional niche
No one can debate the existance of gradients of environ-
mental factors both on a large and small scale. However,
these gradients cannot be considered linear with respect to
a measured topographic feature or with each other. As a
result, a complex mosaic pattern develops which is far too
complex to be described by a “discrete community" theory or
even a one-dimension ordination model.
-121-
Table 4: Classification of Adirondack vegetation
series* according to Heimburger (1934),
Subalpine Series
Trees Ground Flora
Pic ubens pyaesond um splendens
bies baisamea gurozium schreberi
seyary sordifolie Conus cenadenais
vare 2 Cornus jansis
Pyrus decora Oxalis
oaky a alleohaniensis into porealts
stunted Dryepteris spinulosa
sajansbenun 3
olidago macro
Aster acuminatus
Eastern Series
Acer saccharum Aralia nudicaulis
ame As unin
Fraxinus ame na zr 3 spinulosa
uvercus rubra Clintonia borealis
var. bore 8 aianthemum canadense
Gstrya ia Oxalis wen Ano
entalis borealis
cer pensylvanicum
Acer spicatum
Western Series
Acer 8 2. Gxalis montana
8 and sis ‘0 ue Ey 8
8 grandifolia borealis
Acer rubrum Oryopteris 2)
Picea rubens Viburnum alnifoliui
Betu a eae Acer pensylvanicum
Abies Aser spicatum
Tsuga canadensis Trientalis boraalis
Majanthemum canadense
L
Pale auaetie
7 2 t eerie
Lrmees ry sessilifolia
*Types selected to correspond toitypea used in this study.
m tb
754
m on
m stw
sst
me stmer tb sst mater
; Mme st m gt gr mus
s
cme h
507 cb st mime
se
3 f me fs mstsr mst
om me f om me
5 fr om aa “an mst
- fme cb me fst
3 om cb
£ fds ome fo8 mb 4
2° m Iss x
6 tb coaa | ds om mus mb by
fd om fe stil dsm : mel
Is m st
ds om some tds mb
Zo54 — tdsom "Fon “ian dsm mb
* fds ome me?
ft ds om dsm mi va mbil
ds om te
dsm
° 7 ds m
25 Z 100:
FIGURE 12. Distribution of ground flora species on ground flora ordination.
Axis I
Ordination Value
dsm
sst
m st sr
754 P75
$s “fs
3 : mus sits m 3
eh mb 3
s mus ae >
390] mMDY gsm mb P50 €
a ac st m mest mst.srm st = :
2 m st w me" $ r=
f st mb =
fc ts, °
ome cs
tr ds m dsm ow
= cBecon emet ds ome dsom tc mer tb]
2254 simemed fdsom L2sg
< fom aa dsm
bill
fom fme sst moi,
ds om mil va
‘ds om om me mst
ds om
cb me “as st il
0. 7 gr stsm 0
fe) 25 50 50 75 . 100
Axis f Ordination Value
FIGURE 13. Distribution of ground flora on tree ordindtion. —
-126-
synergistic effects of all these climatic factors during the
past life history of each stand is an impossible problem.
However, the responses of the vegetation and the development
of the soil horizons have already integrated the effect of
all environmental variables (Rowe, 1956).
The only gradient plotted on the ordination model is
altitude (Figures 4, 5A and 5B). This, however, is a topo-
graphic gradient which is actually a composite of many environ~
mental gradients, as are the changes related to slope and
slope aspect which occur with altitude. The topography
influences environmental gradients in three major categories -
biotic, edaphic and climatic. The biotic gradient, which
may be either successional or compositional, can be considered
to be a produce of the other two,
Edaphic gradients have been analyzed by many workers,
Most of these are soil moisture gradients (Loucks, 1962;
Whittaker, 1956; Whittaker & Niering, 1965), but at least
One worker has described a gradient related to calcium ion
concentration (Monk, 1965),
The description of a soil moisture gradient for Whiteface
Mountain is somewhat questionable at this stage because it
would have to be based upon the location of each stand on the
soil map compiled by John Witty (1968). Values for the soil
characters would then have to be interpolated for each
stand. About one-third of the stands used in this study are
located outside of the area mapped by Witty. However, a mois-
ture gradient does exist which approximates the altitudinal
gradient, Another soil moisture gradient may exist for the
low elevation stands (the right half of Figure 11) parallel
to axis 2 with soil moisture increasing inversely with ordina-
tion number, so that the beech-birch-maple end is mesic while
the paper birch-oak=basswood end is drier, This is based on
less severe slope and greater depth of “O" horizon with an
accompanying increase in water-holding capacity.
In such mountainous areas as Whiteface Mountain, where
shallow soils and rock outcrops exist, the drainage patterns
and soil development can vary considerably in distances
which are quite small in comparison to the size of the average
sampling unit, For this reason, vegetational and edaphic
gradients will be difficult to measure on a small scale.
Ground flora data which can be recorded on a smaller scale
than tree data, may be better suited to describing such
gradients. The flora which surrounds the slightly raised
base of a tree or the downhill side of a rock outcrop is, in
general, quite different from the typical vegetation associates:
It seems that neither importance value in the case of trees :
nor frequency, in the case of ground flora, will be sufficient
to measure the edaphic gradients which are being sought at the
microhabitat level,
The differences which exist between the tree and ground
Flora ordinations may be due in part to the response time of
these two layers, It could take from 30 to 70 years or more
-127-
for a tree to obtain a diameter of 4", Any noticeable change
in species composition with the exception of catastrophic
destruction of existing trees would take a long time to occur.
The ground flora respond to climatic changes much more quickly
and are therefore indicative of the recent environmental
history.
Interpretation of the ordination model
Ordination has been defined by Orloci (1966) as “a sum-
marization of the information content of a matrix whose
elements, distance, or angles define the spatial relationships
between ecological entities." The methods of construction
of matrices and their interpretation differ with investigators,
Arguments have developed concerning the mathematical or
statistical validity of various methods, Although Orloci states
that the method of Bray and Curtis is not statistically suffi-
cient to ordinate a group of samples, Goff and Cottam (1967)
have shown that the results of the Bray and Curtis method
closely correlated with the other methods used.
The results of the three-dimensional ordination model
with superimposed clusters indicate a large degree of dissimi-
larity among the samples. In such a case, it would be pos~
sible to subdivide the original sample according to the
results of the first axis and ordinate each part separately.
A similar technique was used by Ream (1963) in the Wasatch
Mountains where the gradient ran from semi-desert to alpine
through four distinct vegetation types. The composite three-
dimensional ordination would be more meaningful and easier to
interpret. It could be viewed as a group of three-dimensional
ordinations (hypervolumes ) within a larger ordination (hyper-
space), The clusters can be thought of as the hypervolume
occupied by a community type or a group of closely related
communities situated somewhere in a hyperspace.
The use of the standard deviation criterion as a method
of end-point extraction can be questioned for several reasons.
It was used in this study so that correlations could be
computed between the two ordinations without biasing end-point
selection, According to Austin and Orloci (1966), the standard
deviation criterion which picks the extremes is not the
most efficient method. Such a method emphasizes the unusual,
accidental or rare sample. As a result, samples dissimilar
from the unusual sample used as an end point but not necessarily
similar to each other are clumped.
A compositional index based purely on relative values
(ise. RIV) results in a loss of information. Goodall (1952)
points out that when all samples have the same total value
(i.e. 100%) each one is given equal importance, The species
in each stand are also given equal weight - an ubiquitous
species being considered as important as a species with a
very narrow ecological amplitude. It might be best to express
each species in a stand by standard deviation or weighted
standard deviation. :
~128-
SUMMARY AND CONCLUSIONS
The purpose of this paper was to apply the techniques of
ordination and cluster analysis to selected stand data in
order to describe quantitatively environmental gradients
controlling the distribution and associations of vegetation
on Whiteface Mountain. A comparison of the three-dimensional
ordination and cluster analysis constructed using frequency
measurements for ground flora and IV for trees was made.
Seventy-one forest stands were included in this study.
The stands were composed of vegetation that is typical of
either the boreal spruce-fir or northern hardwood forests
or any degree of combination of these two types.
From this study we can conclude that the measurement of
frequency for ground flora can be used in an-ordination of
these forest stands to construct gradients with as much
confidence as the more complex relative importance value which
was used for the tree species. Some of the advantages of
using ground flora frequency measurements are as Follows
(1) it is easier to obtain in the field} (2) it is easier to
calculate; (3) Smaller areas tan be sampled, making it a
better indicator of local environment; (4) it is more indica-
tive of recent environmental changes; (5) it is a better
indicator of sucéessional trends. The usefulness of the
ground flora is due to their small size and rapid response
time to environmental changes.
The first axis of the tree and ground flora ordinations
was closely related to an altitudinal gradient, The first
axis was also significantly correlated to slope and was
indicative of a soil moisture gradient. A vegetational
gradient from the high altitude spruce-fir forests to the
low altitude northern hardwood forests was described by
the placement of stands along axis 1.
The low altitude stands were separated from each other
by the second axis in both ordinations, The vegetational
gradient described was one from the xeric, subclimax species -
Quercus, Tilia and Fraxinus - to the mesic, shade-tolerant
Climax northern hardwoods ~ Fagus, Betula alleghaniensis and
Acer saccharum. The third axis was used to separate the high
altitude stands from each other, The early successional
Betula papyrifera var. cordifolia stands were at one end
of the gradient and the Picea~Abies stands at the other.
The species composition of the clusters obtained from
the analysis of the ground flora and trees were very similar
and corresponded to the three vegetational series described
by Heimburger (1934) for the Adirondacks. The high altitude
cluster composed of Abies, Picea, Pyrus and Betula papyrifera
var, cordifolia was equivalent to the “Subalpine series”.
Low altitude cluster I with Quercus, Tilia, Ostrya and Acer
saccharum as dominants was analogous to the “eastern series",
Low altitude cluster II, which was dominated by Fagus, Betula
alleghaniensis, Acer saccharum ard A. rubrum, corresponded
to Heimburger's "wésterh series". .
-129-
LITERATURE CITED
Austin, M.P, and L, Orloci. 1966, Gaometric models in
ecology IIs An evaluation of some ordination
techniques. J. Ecol, 54:217=238,
Bray, JoR. and J.T. Curtis, 1957. An ordination of the
upland forest communities of southern Wisconsin,
Ecol. Monographs 27:325=349,
Clements, F.E. 1936. The nature and structure of the
climax, J. Ecol, 24:252~284,
Cottam, G.uand J. T. Curtis. 1956. The use of distance
measurements in phytosociclogical sampling.
Ecol, 37:451-460,
Curtis, J.T. and R.P, McIntosh, 1950, The inter-
relations of certain analytic and synthetic phyto-
sociological characters, Ecol, 31:434-455,
Curtis, J.T. and R.P,. McIntosh, 1951, An upland
forest continuum in the prairde-forest border
region of Wisconsin, Ecol, 32:476-496,
Daubenmire, RFs 1959, Plants and environment: a
textbook of plant autecology. 2nd ed, John Wiley
& Sons, New York. 422p,
Daubenmire, RF. 1968, Plant communities: a textbook
of plant synecology. Harper and Row, New York. 300p,
Fernald, M.L. 1950, Gray's manual of botany. 8th ed.
American Book Company, New York. 1632 p.
Ferree, M.J. and R.Ke Hagar. 1956, Timber growth rates
for natural forest stands in New York State. State
University of New York College of Forestry, Syracuse, 5S6p/
Goff, F.G. and G. Cottam, 1967, Gradient analysis:
the use of species and synthetic indices, Ecol 48:
793-804,
Goff, F.G. and PH, Zedler. 1968, Structural gradient
analysis of upland forests in the western Great
Lakes area, Ecol, Monographs 36:65~-86.
Goodall, D.W, 1952, Quantitative aspects of plant
distribution. Biol. Rev. 27:194-245,
Greig-Smith, P. 1964, Quantitative plant ecology.
Butterworths, Washington. 256p,
Grout, A.J, 1903, Mosses with hand-lens and micro--
scope. Eric Lundberg, Ashton, Maryland. 416 p,
Harman, H.H. 1960. Modern factor analysis, University
of Chicago Press, Chicago. 6582p.
Heimburger, C.C. 1934, Forest-type studies in the
Adirondack region, Cornell University Agricultural
Experimental Station, Memoir 165, 185p,
Holway, G. and J.T. Scott, 1969. Vegetation of the
Whiteface Mountain region of the Adirondacks.
Atmospheric Sciences Research Center Publ. 92.
State University of New York, Albany.
-130-
Hutchinson, GE. 1965, Tha ecological theatre and tha
evolutionary plays Yale University Prass, New
Havan, Conn, 139p-.
Kershaw, KA, 1964, Quantitative and dynamic ecology.
American Eleevier Publishing Co., Inc., New York, 1683p.
Ketchledge, E,H. 1957. Checklist of the mosses of
New York State. New York State Dept. of Education,
Albany. 55p,
Knight, DeH. 1965. A gradient analysis of Wisconsin
prairie vegetation on the basis of plant structure
and function. Ecol,’ 46:744-747,
Lemon, P.C. 1962, Field and laboratory guide for
scoloays Burgess Co., Minneapolis, 1680p.
Lindsey, A.vA.s, JeD. Barton, Jr. and S,R. Miles, 1958,
Field efficiencies of forest sampling methods,
Ecol, 391:428-444,
Loucks, O.L, 1962, Ordinating forest communities by
means of environmental scalars and phytosociological
indices, Ecol, Monographs 32:137-166.
Monk, C.D. 1965, Southern mixed forests of north
central Florida. Ecol, Monographs 35:335~354,
Nicholson, 5. 1965, Altitudinal and exposure
variations of the spruce-fir forest on Whiteface
Mountain. M.S. thesis, State University of New
York at Albany, 61p.
Orloci, L. 1966. Geometric models in ecology Ts
the theory and application of some ordination
methods, J. Ecol, 54:193~215,
Park, R.A. 1968. Paleocology of Venericardia sensu*
lato (Pelecypoda) in the Atlantic and Gulf Coastal
province: an application of paleosynecologic
methods. J. Paleontology 42:955-986,
Ream, RoR. 1963, The vegetation of the Wasatch
Mountains, Utah and Idaho, Ph.D, thesis, Wisconsin
University, Madison, Wisconsin, 178p.
Rowe, J.S. 1956, Uses of undergrowth plant species in
forestry, Ecol, 37:461-473,
Scott, J.T. and J.G. Holway, 1969, Comparison of
topographic and vegetation gradients in forests of
Whiteface Mountain, New York, Atmospheric
Sciences Reszarch Center Publ, 92, State
University of New York, Albany. :
Scott, J.T. and 5S. Nicholsan, 1964, Some character~
istics of vegetation of Whiteface Mountain and
implications concerning their use in studies of
microclimate. Atmospheric Sciences Research
Center Publ. 22, State University of New York,
Albany. pp. 190-232,
Whittaker, RH. 1956. The vegetation of the Great
Smoky Mountains. Ecol, Monographs 26:1-80,
Whittaker, R.H. and WA. Niering. 1965. Vegetation of
the Santa Catalina Mountains, Arizona: a gradient
analysis of the south slope, Ecol, 46:429-452, ,
Witty, J.E. 1968, Classification and distribution of
soils on Whiteface Mountain, Essex County, N. Y.
Ph.D. thesis, Cornell Intversity, Ithaca, New York.
-131-
APPENDIX I
List of Species
Tree Species:
symbol
wo
2
Gren OU<asSsaD4
scientific name
feor anevtvgnisun Ue
oF 8 i un Marsh
Betula nsis Britt,
Betu. 2. gatas
Betula a Mars
vare (Regal) Fern.
8
lia Ehrh,
cana L,
A virginia (Mill.) K.Koch
@ pubens Sarg
Pines Eertrsea Ait.
Pinus strobus L.
Populus jentata Michx.
corre Prunus se Ehrh
Byrus decora (Sarg.)" Hyland
Quercus boleaite
Vare ea borealie (nichx.t. )Farw,
ate 2 : a Pas cerry
amsricana
pie Sere (L.) Carr,
Ground Floras
igelie qudicauls Michx,
nu iis L.
tralia foams t9 ia borealis (Ait.) Raf.
ry ganadonsis sl,
common name
*balsam fir
*striped maple
*red maple
*sugar maple
*mountain maple
*yellow birch
paper birch
*cordate-leaved birch
*beech
*white ash
*Lronwood
*red spruce
red pine
white pine
big-toothed aspen
black cherry
*mountain=ash
northern red oak
white cedar
basswood
hemlock
o fests Cootis apventandice (Oeder) Fern,
Bevont eris spinulosa (0.F. Muell,) Watt,
qugangdium dy ucidulum Michx
fesanthg mum canadense Desf.
Oxalis tiontana Raf,
Solidaga macrophylla Pursh
‘pique oes rosaus seb ee
iare ie c cordifolia
Trie nta is borealis fer
Uvularia sessilifolia Le
Viburnum alnifolium Marsh,
*Seedlings of this species also used in
Symbol used for seedling is lower case
’
round flora ordination, '
etter,
<numbers
Stand
BUSSEY spur BET STATES
j
-132-
APPENDIX IL
Dendrogram of clusters based on ground
flora data,
bi
i
5 f =
|
4 uaa i
i =
———————
——————|
——————
Hy
Aa
3
=|
Site
i
BV
EE=]Low altitude cluster T
CODD ow altitude cluster IT
CDHigh altitude cluster
-133-
Dendrog:am of clusters bdsed on tree data,
TR RTE :
MI y
at
tH
i ‘cca
: i Mi me
———————
fat
E=——| Low altitude cluster I
MOM) tow altitude cluster
ti High altitude cluster
-134-
APPENDIX ILI
Tepographic-and Vegetational Characteristics of Stands
Stand Altitude Slope Slope Dominant Dominant Cluster! 2
No, (Feet) Agpeot Trees Ground Flora I, il iit i"
4500 9 3 F om mc tY9
2 4500 332 27 F om ch tg
3 4500 108 36 F f me tg
4 4000 200 28 F f ds om tg
5 4060 273 26 F f ds omc tg
6 4000 120 34 FS f om tq
7 3970 $5 7 F f ds om tg
8 3960 20 30 F f ds om tg
9 3800 24 30 F f cb cc aa tg
10 3780 59 28 Fe F om tg
11 3500 248 27 FS fF om to
12 3500 320 17 F f ome tg
13 3500 273 10 F f ds om tg
14 3500 350 22 F da om tg
15 3480 113 24 $ Fo tg
16 3450 104 37 Ss c me tg '
17 3450 124 32 F f ds omc tg
18 3350 107 26 FC Y ds om tg
19 3110 229 28 S fes tg
20 3100 33 21 BM ds om tc t Q
21 3075 357 22 F f om aa tg
22 3000 248 23 FS f ds tg
23 3000 327 dl Ss f ds om tg
24 3000 204 16 Ss f ds ome tg
25 3000 339 23 ) ds om tg
26 2900 74 24 Ss fs tg
27 2875 355 41 $ f me tg
28 2860 99 24 c f ds om tg
29 2650 29 16 F cb mc tg
30 2625 189 27 s 8 st t 9
31 2600 153 5 Ym ds m ta
32 2600 185 28 FCS f st t g
33 2560 164 28 Ss s st 9 t
34 2550 243 7 Ym dm tg
35 2525 353 17 s cb st tg
36 2520 342 1l Ss Y om me 9 t
37 2500 145 17 m ds m t 9
38 2410 264 32 Ss fs t 9
39 2400 84 21 BM mst tg
40 2320 248 24 $ ds om 9
-135-
Stand Altitude Slope Slope Dominant Dominant Cluster!
jo. Fee Aspe. Trees Ou i I I
4 4 ™ mst tg
42 2200 240 170 OS Y ds s t
43 2150 152 13 Bm mb tg
44 2045 249 22 SP me st t 9
45 2000 125 15 mQ mmo h tg
46 2000 268 15 S YB st sm t
47 2000 4 12 m mb tg
48 1950 126 6 mp m tb to
49 1950 161 5 YB m st tg
50 1900 310 8 Bm bil tg
51 1900 106 13 BM ds m tg
52 1850 151 14 m m us tg
§3 1800 Go F fr tg
54 1750 130 11 mQ m us t 9
55 1750 121 8 mT mby t 9
56 1730 51 26 mQ man tg
57 1710 36 2 BM m1l va to
58 1710 181 8 8 M ds m tg
59 1680 56 5 BM dsm ; tg
60 1670 281 4 m mst sr tg
61 1660 131 6 mR me r tb Q t
62 1630 337 22 mT m st tg
63 1620 248 11 Y8 st ll tg
64 1575 148 8 Q mime st tg
65 1550 152 6 m mst sr tg
66 1530 120 14 MG mst w tg
67 1530 131 5 ™ ds m t 9
68 1525 143 6 m R mst sr g t
69 1500 10 Bm mb tg
70 1470 141 19 Bm mb tg
71 1370 37 14 m mb t 9
Joiuster I -- High altitude cluster (spruce-fir)
Cluster II ~~ Low altitude cluster I (beech=-birch-maple)
Cluster III -- Low altitude cluster II (oakebeabucad)
2y a= non-cluster
34 -- tree analysis
4 =-- ground flora analysis
A FLORISTIC COMPARISON OF UNDISTURBED SPRUCE-FIR FORESTS
OF THE ADIRONDACKS WITH FOUR OTHER REGIONS
By
Stuart Nicholson, J. Gary Holway, and Jon T. Scott
A FLORISTIC COMPARISON OF UNDISTURBED SPRUCE-FIR FORESTS
OF THE ADIRONDACKS WITH FOUR OTHER REGIONS
By
Stuart Nicholson, J. Gary Holway, and Jon T. Scott
ABSTRACT
The relationship of Adirondack boreal spruce-fir to the spruce-fir
vegetation of other geographic regions reported in the literature is ex-
amined based on number of species shared in common. Two methods of evalua-
tion are employed. One involves a simple determination of percentage of
species shared in common. The second employs the standard 2w/atb Index of
Similarity using constancy values for values to represent a, b, and 2w.
The Adirondack sample is divided into two sub-samples, the Whiteface
uplands (3000-3500 ft) and the Whiteface highlands (4000-4500 ft). Simi-
larity decreases with geographic distance from Whiteface Mountain. The
Smoky Mountain and Wisconsin forests are furthest and least similar to
both types of Whiteface stands, the Catskill Mountain forests are closest:
and most similar.
INTRODUCTION
Extensive areas of undisturbed spruce-fir forests are found in the
Adirondack Mountains of New York State. Much of this forest lies within
the Adirondack Forest Preserve which provides a substantial measure of
protection from human disturbance. There have been a few ecological stud-
jes in the Adirondacks and no comprehensive phytosociological study has
‘been published to date. Oosting and Billings (1951) and McIntosh and
Hurley (1964) have emphasized the need for quantitative ecological studies
of the spruce-fir forests.
The floristic relationship fo the Adirondack spruce-fir to spruce-fir
in other regions of eastern North America has been variously interpreted
by different authors. Braun (1950) treats it as a variant of the eastern
deciduous forest separate from the boreal forest to. the north. Curtis (1959)
stated that there was a close relationship between the Adirondack spruce-fir
and the Wisconsin lake forest, as evidenced by the data of Heimburger (1934).
-137-
~138~
McIntosh and Hurley (1964) considered the spruce-fir of the Adirondacks,
Catskills, and White Mountains to be similar. Crandall (1958) pointed
out basic floristic and physiognomic similarities between spruce-fir of
the Great Smoky Mountains and. the Adirondacks. Oosting and Billings (1951)
and McIntosh and Hurley (1964) suggested that the spruce-fir forest is
essentially continuous from the Great Smoky Mountains to the northern
Appalachians.
These references to Adirondack spruce-fir are based primarily on
qualitative data. The prupose of this paper is to present a comparison
of the Adirondack spruce-fir forest composition to the above mentioned
regions based on quantitative stand data collected at Whiteface Mountain
(Lat. 44° 20' N., Long. 77° 55' W.) in the northern Adirondacks.
METHODS
Data on the Adirondack spruce-fir forest were obtained from that being
collected for a larger study concerning the relation of vegetation and en-
vironment at Whiteface Mountain. Stands with obvious signs of disturbance
were excluded. Trees and saplings were recorded using the quarter method
(Cottam and Curtis, 1956) and ground flora from square meter quadrants.
Nomenclature follows Fernald (1950).
Of the 56 stands sampled during the summer of 1964, 17 were selected
as being typical of well-developed, undisturbed spruce-fir forest. Of these
In order to compare species found in different regions of the spruce-
fir forest, constancy data taken from Curtis (1959) for the Wisconsin Lake
forest, and ours for the Adirondacks were appended to a table comparing
spruce-fir stands in the Appalachians presented by McIntosh and Hurley
(1964). The comparisons made in the present paper are by two methods.
The first is by the direct examination of the percent of species common
between regions. The second method compares regions based on similarity
using the common 2w/atb Index of Similarity. The latter method involved
a ranking of regions by trees only, shrubs only, ground flora only, all
three components combined, and by rank values determined by order of po-
sition of each region by each of the four preceeding rankings.
RESULTS AND DISCUSSIONS
Constancies of vascular herbs, shrubs, and trees from representative
regions of the spruce-fir forest are listed in Table 1. Within this sample
the Great Smoky Mountain flora is most distinct in that it has the highest
-139-
Table 1: Percent presence of trees, shrubs, and herbs in spruce-fir stands
from the Catskills, White Mountains, Great Smokies, Adirondacks,
and Wisconsin.
Great Cat- White- White-
Smoky skill c face White face
Species Mts@ Mts Wisc. upland Mts@ high
Trees:
Picea rubens ......seg 100 100 - 100 100 100
Betula cordifolia ..... 0 75 - 100 0 100
PYPUS SP. veeceeeeeeree 100 75 56 88 100 56
I Acer spicatum .......- 89 50 82 38 100 22
Hy Abies fraser? .......-. 100 0 - 0 0 0
i Prunus penslyvanicum .. 45 50 33 25 0 0
! Ilex monticola ......+- 1 0 - 0 0 0
: Abies balsamea ........ 0 100 100 100 100 100
Fagus grandifolia...... an} 25 - 0 0 0
Betula papyrifera..... 0 0 87 0 100 0
Tsuga canadensis......- 0 25 4) 0 25 0
| Prunus serotina......++ 0 0 23 oe 25 0
Acer rubrum ... one 0 25 77 25 0 0
Larix laricin < - - 5 0 - VV
| Acer saccharum.... . - - 54 oe - 0
Betula alleghaniensis.. 100 75 4) 50 75 0
Acer penslyvanicum....+ - 75 - 25 75 0
Shrubs:
Viburnum alnifolium--+- 100 50 - oe 100 0
Viburnum cassinoides --» 11 0 - oe 25 0
Vaccinium erythrocarpum 100 0 - 0 0 0
Rubus canadensis....... 100 0 - 13 0 oe
Sambueens pubens....... 89 25 - 25 0 0
Ribes rotundifolia..... 22 0 - 0 0 0
Rhododendron catawbiense 11 0 - 0 0 0
Vaccinium pallidium.... 11 0 - 0 0 0
Nemopanthus mucronata .. 0 25 - 0 50 22
Vaccinium angustifolium 0 0 44 25 25 22
Lonicera canadensis .... 0 25 97 0 0
RubUS SP. seeceeeeeeree 0 50 {f} 25 25 0
“RibeS Sp. seseee os 0 25 f) 25 0 67
Rhododendron roseum .... 0 25 - 0 0 0
Diervilla lonicera ..... - - 80 13 - 0
Ledum groenlandicum .... - - - 0 - 22.
Alnus crispa .. wear - - - 0 - 33
Vaccinium Sp. ..ssseeeee 0 25 (f) 13 1) oe
apata from Oosting and Billings (1951) summarized by McIntosh and Hurley (1964).
bpata from McIntosh and Hurley (1964).
Data from Maycock (1956) in Curtis (1959).
ddesignated in text as B. papyrifera var. cordifolia.
ePresent in other stands.
present.
-140-
“Table 1: (Continued)
Great Cat- White- White-
Smoky skill face White * face
Species (Herbs) Mts Mts Wisc. upland Mts high
Aster acuminatus ......... 100 100 - 38 100 44
Dryopteris dilatata ...... 100 100 - 09 100 09
Oxalis montana ..... 100 75 - 100 88 100 «
Clintonia borealis . + 100 100 95 75 100 100
Monotropa uniflora ....... 66 0 - 38 50 in
Trillium undulatum . é 55 25 - 25 * 60 0
Carex flexuosa os 55 50 - - 50 -
Lycopodium lucidulum ..... 33 25 - 63 75 of
Cinna latifolia ...... oD 33 ) - 25 25 oe
Viola rotundifolia ....... - 22 50 - 0 0 0
Dryopteris intermedia . iW 25 100 0 (f) 0
Streptopus roseus .....0.. WV. 50 80 oe 0 38
Senecio rugelia ....c.ceee 89 0 - 0 0 0
Houstonia serpyllifolia .. 44 0 - 0 0 0
Solidago glomerulata ..... 44 0 - 0 0 0
Aster divarcatus ......... .. 33. 0 - 0 0 0
Impatiens pallida ........ 33 0 - 0 0 0
Chelone lyont ........ce0e 33 0 - 0 Q- 0
Arisaema quinatum ........ °°. - 22 0 - 0 0 QO. .
Circea alpina .......s0.0. 22 0 36 13 0 0
Stachys clingmanii ....... Ww 0 - 0 0 0
Maianthemum canadense .... 0 100 100 63 100 100
Cornus canadensis .. 0 75 97 63 100 100
Avalia nudicaulis .. 100 50 95 75 0 22
Solidago macrophylla ..... 0 0 - 75 100 89
Coptis trifolia .......o0, 0 75, 62 50 50 89
Trientalis americana ..... 0 75 97 25 50 56
Dryopteris hexagonoptera . ‘0 0 - 0 50 0
Chiogenes hispidula ...... 0 0 - 0 25 44
Solidago sp. ..sccsosecees 0 25 (f) 0 0 0
Dennstaedtia puncttoba oe 0 25 - 0 0
Linnea borealis .......666 - - 74 0 - 44
Lycopodium annotinum....... 0 = - - 0 - 89
Pyvola SP. .cscesccseseees 7 - (f) 13 - 0
Tiavella cordifloia ...... - - 3 13 - 0
Smilacina racemosa ....6.. 5 0 = - 39 13 - 0 w
Athyrium filix-femina .....00 = - 59 oe - -
Cyperidium acuale ........ - - - 13 - 0
Pteridium aquilinum ...... - - 77 13 - 0 *
Aster macrophylius ....... - - 90 25 - 0
Lycopodium obscurum ...... - - 62 oe - 0
Trillium crectum ......068 - - - 13 - 0
Osmunda Claytonia ........ - - - 13 - -
Polygonatum biflorum ..... -. ‘e - 25 - 22
Thelypteris phegopteris .. - - 36. 38 - 67
Potentilla tridentata .... - - - 0 - VY *
Arenaria groentandica .... - - “ 0 - VV
Veratrum viride .......008 - - - 0 - 56
SDropteris spinulosa had a percent occurrence of 100 in both Whiteface stand groups. -
“141-
number of species (16) not listed for any of the other ‘areas. The Whiteface
combined list has only six species not contained on the species lists for
the other regions,
Within the woody species group only mountain maple (Acer spicatum) is
| recorded for all six regions. Three additional genera (Abies, fons and
t Betula) are found in all regions. At least one species of spruce and one
i r occurs in each region except Wisconsin where no Picea is listed.
Several other tree species were found in at least half the areas considered.
These are Betula alleganiensis, B. papyrifera var cordifolia, Prunus pen-
sylvanicum, Acer rubrum and Tsuga canadensis. Regions which include the
1 atter species are envisioned as being of a different type of spruce-fir
I association than on Whiteface. Spruce, fir, and hemlock do occur in stands
t together in the Whiteface region, but these are low altitude stands of a
distinct non-boreal character.
I No shrub occurred in more than four of the six regions. Species
i Present in at least three regions were: Vaccinium angustifolium, Nemopanthus
\ mucronata, Sambucus pubens, and Viburnum alnifolium. Genera such as Vaccinium,
i Ribes, and Rubus were present in most regions, but are difficult to interpret
| because each contains many species.
Only one (Clintonia borealis) of the 46 herbaceous species listed was
} reported for all six regions, although several species were found in five
,- of the six (e.g. Maianthemum canadense, Cornus canadensis, Coptis trifolia,
f Tiarella americana, Aster acuminatus, Dryopteris dilatata, and Oxalts montana).
Several other species were confined to only one region. Most of these fall
into one of three classes: (1) those native to the undisturbed spruce-fir
forest of the Great Smokies (e.g. Aster divaricatus, Senecio rugelia, and
Solidago glomerulata), (2) species characteristic of alpine tundra which
jivaded hiterace high altitude spruce-fir stands (e.g. Potentilla tridentata .
i and Arenaria groenlandica), and (3) species more characteristic of deciduous
or cont fer-fardwbed forests which may indicate écotonal conditions in some
1 Spruce-fir stands (e.g. Trillium erectum, Athyrium filix-femina, and Osmunda
| claytonia).
To obtain more generalized comparisons than cursory inspection of
‘Table 1 allows, a summary of region relationships based on percentages of
common occurrence of species is presented in Table 2. In the table the
row values following a region designation are species that the region has
‘in common with those regions indicated in the column headings. For example,
59% of the Great Smoky species occur in the Catskills, 52% in the White
Mountains, etc., while only 50% of the Catskill species occur in the Great
Smokies and 69% occur in the White Mountains.
From this table we see that the stands from Wisconsin and the Great
Smokies share the lowest percentages of common species. The data further
imply that the two most similar regions are the Catskills and the White
Mountains. This is because the region having the greatest percent of
species occurring in the White Mouhtains is the Catskills (63%) and the
one having the greatest percent of Catskill species was the White Mountains
-142-
(69%). This degree of similarity exceeds even that for the two Whiteface
stands which are separated by less than a mile as compared to the 150 mi.
or more that separate the Catskills and the White Mountains. This can be
largely explained by the general paucity of flora that is typically en-
countered as habitat extremes are approached such as they are when you
approach timberline.
Data in Table 2 fail to provide a distinct indication of the relation-
ship of the Adirondack stands to the other regions. The Whiteface Mountain
upland stands are least similar to the Great Smoky Mountain stands (42%
upland species in Smokies and 43% Smokies species in upland), but are simi-
larly related to the other regions (63-65% upland species found in other
four regions). Highland stands were also least similar to the Great Smoky
region (24% highland species in Smokies and 31% Smokies species in highland).
Values for highland species in the other regions range from 45-55%,
An additional way to evaluate the degree of similarity or dissimilarity
of the Whiteface upland and highland stand types with other regions is by
comparison of similarity index values. Table 3 shows the relationship of
each region to every other region using the percent constance values from
Table 1 to determine the a, b and 2w values for computation of the Index
of Similarity.
Examination of the degree of similarity of each region on the basis
of trees only, shrubs only, and ground flora only, as well as the relation-
ship based on combined values further clarifies the position of the Whiteface
spruce-fir in relation to other regions. However, before looking at specific
relationships, some general comments about the general relationships seem in
order. For example, tree association values run consistantly higher than
either shrub or ground flora values with one exception, the White Mountain-
Whiteface highland types. In this case the ground flora value was slightly
higher, but the shrub value is still considerably lower. In fact, the shrub
association values consistantly have the lowest similarity values through-
out. The maximum value is .45 (1.00 denotes absolute similarity) in the
Catskill-Whiteface upland shrubs, while the Great Smokies-Wisconsin, and
Great Smokies-Whiteface highland stands had no shrubs listed in common.
The ordering of the region pairs in Table 3 is probably most indicative
of the overall relationship of the spruce-fir forests of the various regions
to each other. The idea that the Catskills and White Mountains are most
similar floristically as implied in Table 2 is further supported by Table 3.
The Catskill-White Mountain regions show the highest value of .67 based on
total composition. Once again, this relationship is even closer than that
between the Whiteface upland and highland stand types. The value of simi-
larity between these is .62. Only on the basis of trees alone do the
Whiteface stands show more similarity to each other than the Catskill-White
Mountain ones do, although even in this regard the Catskills and Whiteface
uplands show a closer similarity than do the two Whiteface types.
-143-
In general the Whiteface Mountain stands show closest relationships
to the Catskills, White Mountains, Wisconsin and Great Smoky spruce-fir
forests in that order. In all. cases the upland stands show a closer af-
finity to each of the other régions than do the highland stands.
The Wisconsin stands, in general, displayed a weak relationship to
each of the other regions. The highest value based on total stand compo-
sition was .48 between Wisconsin and Great Smoky stands. By the rank
total method the Wisconsin-Catskill association ranks highest of the
Wisconsin associations but is still among the weaker relationships in
the table. Data from both Table 2 and Table 3 seem to contradict the
strong floristic similarity claimed by Curtis (1959) to exist between
Wisconsin and Adirondack spruce-fir forests.
In summary it would seem that while all of these regions may well
be spoken of as having spruce-fir forests, each is distinct. The dis-
tinct character is attributed to the differences in understory species,
particularly the shrub layer, although in a few cases even the tree
species composition is greatly dissimilar. While it has not been sta-
tistically analyzed at this point, there appears to be a general cor-
relation between distances separating regions and values of similarity.
For example, Whiteface stands were least similar to those furthest from
them, and most similar to those closest.
Only two species were found in all six areas, but ecological equi-
valents of many species are present. For example, balsam fir is a domi-
nant in the five northern regions and Fraser fir replaces it as a tree
dominant in the southern Appalachians. It is likely that the presence
of these ecological equivalents, as well as the species shared in common,
accounts for the concept of a basic similarity in spruce-fir forests in
the eastern United States as reported by Crandall (1958) and Curtis (1959).
In the final analysis it may turn out that the upland and highland
spruce-fir forests of the Adirondacks are most closely related to those
of the boreal forest of southeastern Canada. Both Cape Breton Island
(Collins, 1951) and higher elevations in the Adirondacks are dominated
by the same three tree types (balsam fir, paper birch and spruce), and
lack northern hardwoods characteristic of ecotonal areas such as the
pe wondack lowlands, Catskills, White Mountains, and Wisconsin lake
‘orest.
Table 2: Percentages of species in common between representative areas of
spruce-fir forests.
Region and Tot. Great Catskill White Whiteface Whiteface
No. of Spp. Smokies Mts. Mts. Upland Highland
Gt. Smokies (38) - 59 52 43 31
Catskills a 50 - 69 54 52
White Mts. (26 40 63 - 51 $7
Wf. Upland 2 42 63 66 - 66
Wf, High A. (28 24 47 55 49 -
Wisconsin (30) 24 50 45 54 48
Wisc.
Table 3: Comparisons of. Spruce-Fir Stands from Whiteface Mountain, the Catskills, the White Mountains, Wisconsin,
and the Great Smokies Based on the 2w/atb Index of Similarity.
Areas
Catskills - White Mountains
Whiteface Upland - Whiteface Highlands
Catskills - Whiteface Uplands
Catskills - Whiteface Highlands
Whiteface Uplands - White Mountains
Great Smokies - White Mountains
Catskills - Great Smokies
White Mountains - Whiteface Highlands
Catskills - Wisconsin
Wisconsin - Whiteface Uplands
Great Smokies - Whiteface Uplands
Wisconsin - White Mountains
Great Smokies - Wisconsin
Wisconsin - Whiteface Highlands
Great Smokies - Whiteface Highlands
The order was determined by combining the
rank numbers obtained when ordering the areas on values obtained from tree data only, shrub data only,
ground flora data only, and from a combination of the tree, shrub and ground flora data.
Trees Shrubs Ground Flora Total (T,S,GF) Rank
- Val. Rank - Val. Rank Ind. Val. Rank Ind. Val. Rank Total
73 3 42 4 69 1 67 1 7
80 2 31 4 58 3 62 2 aa
-84 1 245 1 51 8 61 3 13
66 4 ~23 5 255 5 55 5 19
64 5 -03 13 61 2 59 4 24
58 8 +33 3 52 7 51 7 25
58 7 22 6 -53 6 49 8 27
51 11 22 7 58 4 +52 6 28
4 10 HW 1 48 9 45 10 40
-50 12 21 8 43 10 43 ia] 4]
254 9 +13 9 42 Ft A 12 “4
64 6 HW 10 33 12 -40 14 42
37 14 -00 15 20 15 48 9 53
37 15 ae 12 -2200«14 4 13 54
38 13 .00 14 26 13 225 15 55
-145-
ACKNOWLEGEMENTS
The authors thank the Atmospheric Sciences Research Center, State
University of New York at Albany which provided support for this study.
Special thanks also go to Stanley Smith of the New York State
Museum and to Orie Loucks of the University of Wisconsin for verification
of species identification.
REFERENCES
Braun, E, L. 1950. Deciduous forests of eastern North America. McGraw-
Hill, New York.
Collins, E. H. 1951, A study of the boreal forest in northern Cape Breton
Island. M. A. Thesis, Acadia University, Wolfville, Nova Scotia.
Cottam, G. and J. T. Curtis. 1956. The use of distance measures in phyto-
sociological sampling. Ecology 37: 451-460.
Crandall, D. L. 1958. Ground vegetation patterns of the spruce-fir areas
of Great Smoky Mountains National Park. Ecol. Monog. 28: 337-360.
Curtis, J. T. 1959. The vegetation of Wisconsin. University of Wisconsin
Press, Madison,
Fernald, M. L. 1950. Gray's manual of botany. 8th ed. American Book,
New York.
Heimburger, C. C, 1934. Forest-type studies in the Adirondack region.
Cornell University, Agric. Expt. Sta. Memoir 165. Ithaca.
McIntosh, R. P. and R. T. Hurley. 1964. The spruce-fir forests of the
Catskill Mountains. Ecology 45: 314-326.
Oosting, H. J. and W. OD. Billings. 1951: A comparison of virgin spruce-
fir in the northern and southern Appalachians system, Ecology
32: 84-103,
SLOPE ASPECT VARIATION IN THE VASCULAR PLANT SPECIES
COMPOSITION IN THE TREELESS COMMUNITY
NEAR THE SUMMIT OF WHITEFACE MOUNTAIN, N.Y.
By
Stuart Nicholson and Jon T. Scott
(
i
i
Slope Aspect Variation in the Vascular Plant Species
Composition in the Treeless Community
Near the Summit of Whiteface Mountain, N. Y.
By
Stuart Nicholson and Jon T. Scott
INTRODUCTION
The treeless community near the summit of the 4867 foot high
Whiteface Mountain, N.Y. (Lat. 44° 20') endures a rigorous physical
environment as well as considerable human disturbance. It has been
interpreted as being true alpine tundra by Smith (1964) and as spruce-
fir-tundra ecotone by France and Lemon (1963).
The microclimate at the summit is one of the coldest in the eastern
U.S. Only a few higher, or more northerly peaks such as Mt. Washington,
N.H, (6288 ft., Lat. 44° 16') or Mt. Katahdin, Me. (5268 ft., Lat. 45° 55')
may have colder temperature regimes. Mean monthly temperatures at the
summit during the summer range from 48 to 58° F. Mean January temperature
js about 8° F.
Mean annual precipitation at Lake Placid (4 mi. S.W. at 1860 ft.) is
approximately 39 inches. Total precipitation at the summit exceeds this
by about 10 inches due to orographic lifting, but long term records are
lacking. Average relative humidity is high, exceeding 70% three-fourths
of the days in the normal year (Falconer, 1963, 1964). Condensation on
trees during the presence of a cap cloud could yield several inches of
precipitation each year. The cool temperatures, abundant precipitation,
and high relative humidity result in a high precipitation to evaporation
ratio. The sunmit is capped by clouds on about 40% of summer days and
50 to 70% of days in fall and spring adding to the cool-moist nature of
‘the climate. Strong winds at the summit average 17 to 20 m.p.h. in
summer and nearly double this in winter. Frost or rime icing can occur
in all months of the year which would indicate a tundra climate but from
the temperature data available the summit area is classified as a Dfd
type (called humid continental) in the Koppen system. Growing season
length near the summit may average about 50 to 70 days compared to 80
to 105 days for the Northeastern Adirondacks reported by Stout (1956)
and Feuer et. al. (1963).
Soils in the Whiteface area were classified by Witty (1968). For
the treeless summit he found two units for mapping purposes. One unit
included most of the north and east and part of the west-facing slopes
of the treeless area. This unit was composed mainly of organic material
and was classified as HistosOls of either the frigid Sysleptist or frigid
Dyssaprist Humodic types. On the south and part of the west slopes of
-147-
-148-
b. South
c. West d. North
Figure 1: Pictures of the typical vegetation on the four aspects of the
treeless area near the summit of Whiteface Mountain.
~149-
the treeless area Witty found a mineral soil which he classified as a
Spodosol. He called it a frigid Typic Haplorthod. These soils were
found on benches between nearly vertical rock exposures. Most of this
south and west area is rock outcrop while the north facing unit was nearly
completely covered with vegetation. Pictures of the typical vegetation
of the four aspects are shown in Figure 1.
Disturbance by man in the treeless community was probably negligible
until the summit became a popular hiking spot in the 1870's (Wallace, 1896),
but Watson (1869) mentions a fire of undetermined origin which consumed
virtually all the organic matter around the summit during the summer of
1867. Logging operations may have extended up the east side and to near
the summit during the 1890's (Rogers, 1964), but the effect of this
activity on the treeless community is unknown. Disturbance from visitors
increased markedly after 1938 when the paved highway which extends nearly
to the summit was completed. Hiking trails, an observation building,
and other human impact have resulted in a removal of much of the treeless
community. Distribution of exotics such as dandelion, milfoil, plantain,
and various introduced grasses appears to closely coincide with the
location of regularly disturbed areas, most of which are above 4640 ft.
The purpose of this paper is to report the present vascular plant
composition of the Whiteface Mountain treeless community as it is
represented on the four major slope-aspects. further, an attempt will
be made to explain the ecological basis of the compositional variations
evident on the different aspects.
METHODS
Frequepcies of vascular plants were recorded from 40 systematically
located 1 m* quadrats on each of the four major slope-aspects (N,E,S,W)
during the summer of 1964. Many of the specimens were verified by s
Stanley Smith of the New York State Museum. Nomenclature follows Fernald (1950).
Areas regularly disturbed by man were not included in the sampling.
‘These were recognized from trampling, growth habit of plants, presence
of exotics, and observations on the movements of tourists. On the east
face, where tree cover extends nearly to the summit, sampling was restricted
to treeless areas.
Descriptive data such as slope, slope-aspect, altitude, substrate
characteristics, ground cover, and general comments were noted and are
filed with the vegetation data.
“180° A. borealis
@ A. balsamea e
o C. deflexa © V. angustifolium
a. 4 B. cordifolia ~ c. 2 S. Cutleri
Oo L. groenlandicum x P. tridentata 7
* V. uliginosum
30 x I5r 4
RS 3s
o ia Xx
= 207 7 = 1OrF 4
& 10h i. 4 @& 5+ 7
(e) (0)
E S W N E Ss W N
b. ° Cornus canadensis q.* C. Houghtonii
° S. macrophy lia
2 15 2 15r 5
o os
2 10r © {Or
iv ire
& 5- © 5+ 4
o (0)
Ss W N E Ss W N
Figure 2: Relative frequency plotted against aspect for the most frequent
species on the (a) north (b) east (c) south and (d) west facing
aspects of the treeless area near the summit of Whiteface Mountain.
-151-
RESULTS
Relative frequencies of vascular plant species at each aspect and
for all aspects summarized are given in Table 1, The alpine bilberry
i (Vaccinium uliginosum) was the most prevalent species overall and was
the most frequent species on the north and west aspects. Three-toothed
| cinquefoil (Potentilla tridentata) was second in overall frequency and
H was the most frequent species on the south aspect. Next in overall
frequencies were bunchberry (Cornus canadensis), Labrador tea (Ledum
t= roenlandicum), and an alpine’ goldenrod (Solidago Cutleri). AlT these
species are considered to be true tundra species (Woodfin, 1959). Bunch-
berry and blue-joint grass (Calamagrostis canadensis) were co-dominants
on the east aspect. Other important species were balsam fir seedlings
(Abies balsamea), a goldenrod (Solidago macrophylla) common in nearby
spruce-fir forests, a northern bentgrass (Agrostis borealis), and mountain
sandwort (Arenaria groenlandica).
Although all species occurring in the treeless community are not
tundra species per se, many may be represented by tundra ecotypes (e.g.
Calamagrostis canadensis and Rubus idaeus) according to Stanley Smith
| (Pers. comm). “Discontinuous distributions of several of these species
on Whiteface Mountain has been discussed by Nicholson (1965) Holway
et. al. (1969) and Breisch et. al. (1969).
A detailed floristic comparison with other tundra communities is
i beyond the scope of this paper, but data are presented with this purpose
i in mind,
i
Relative frequency of the twelve most common species is plotted
against aspect in Figure 2. These data include all of the species which
were found in at least three of the four aspects. Several distribution
patterns may be recognized according to relative locations of maxima
and minima and slopes of the curves joining relative frequencies. All
but three of the twelve species had its highest frequency on north (5 spp.)
or south (4 spp.) aspects. Two species had frequency maxima on the east
aspect and only one on the west.
The five species (group I) with frequency maxima on the north (Vaccinium
uliginosum, Betula papyrifera van cordifolia, Ledum groenlandicum, Abies
balsamea, and Carex der rexa) all had similar distribution patterns (FTgure 2a).
AIT except V. uliginosum were least frequent on the south and west, respéctively.
The latter was Teast frequent on the east.
Four widely tolerant species (Agrostis borealis, Potentilla tridentata,
Vaccinium angustifolim, and Solidago Cutleri) were most frequent on the
t south aspect (group IT). AlT except V. angustifolium had minima on the
if east and distribution patterns of all four species were quite similar
i (Figure 2c). The two major species with maxima on the east (Cornus
canadensis and Calamagrostis canadensis) also had similar distributions
(Figure 2b). The distributional pattern of Carex Houghtonii, the only
species with frequency maximum on the west aspect was not common on any
other aspect (Figure 2d). ‘
-152-
Table 1: Relative frequencies of vascular plant species on the four major
aspects and for all aspects combined in the treeless community of Whiteface
Mountain, N.Y.
ATT
Species East North West South Aspects
Vaccinium uliginosum adie § 1.2 27.2 (16.8 W107 14.4
Potentilla tridentata 1.6 5,8 11.5 19.8 9.6
Cornus canadensis .. W7 10.9 3.8 8.1 8.8
Ledum groenlandicum 4.8 17.0 11.5 2.0 8.3
Solidago Cutleri .....ceeseeveeeeeeees - 6.4 6.2 13.2 6.1
Abies balsamea ....... 2.8 13.6 6.7 1.5 5.6
Agrostis borealis .. « 1d 1.4 2.9 10.2 4.0
Solidago macrophylla .....+++ . 10.0 2.7 - 0.5 4.0
Arenaria groenlandica ........- _ ot - 7.2 6.6 3.6
Betula papyrifera v. cordifolia . . 2.4 5.8 5.3 1.5 3.6
Vaccinium myrtilloides .......+ 5 Ol - - 2.5 3.2
Calamogrostis candensis ... 17 - - - 3.2
Rubus idaeus .....eeeeeeeee » 8.5 - - 1.0 2.9
duncus trifidus ... - - 1.0 10.2 2.8
Salix Uva-ursi .... - - 5.7 4.6 (ae)
Carex Houghtonii .. 0.4 0.7 8.6 - 2.6
Lycopodium Selago ......- - - 8.6 2.3
Vaccinium angustifolium . 0.8 - 1.9 5.1 2.0
Gentiana linearis .,.... 4.4 2.7 - 1.9
Dropteris spinulosa 4.8 - - - 1.5
Scirpus caespitosus ..... 2.8 - 0.5 - 1.0
Clintonia borealis .......+ 2.8 - - - 0.9
Carex deflexa (7) ...sseeee 0.4 2.4 1.0 - 0.8
Spiraea latifolia ..... 1.2 1.4 - - 0.6
Picea rubens’.......+.+ - 1.4 1.0 - 0.5
Dryopteris Phegopteris 1.6 - - - 0.5
Aster acuminatus .....- 1.2 - - - 0.4
Fragraria virginiana .. 1.2 - - - 0.4
Coptis groenlandica . 1.2 - - - 0.4
Pyrus decora ...... . 0.8 - - - 0.3
Veratrum viride ... seeeee - 0.8 - - - 0.3
Gaum macrophyllum ......eeeeeeee - 0.8 - - - 0.3
Achillea millefolium ..... _ ot 1.4 - - 0.3
Prenanthes Bootii .......s.++- . - - 1.0 0.3
Maianthemum canadense ..... 0.8 - - - 0:3
Gramineae sp. ......- Puen 0.8 - - - 0.3
Alnus crispa .....+.seeeee wane 0.4 - - - 0.1
Aralia nudicaulis .,...-+.6 0.4 - - - 0.1
Linnea borealis . ous en shaiaaisvas 6. O84 - - - 0.1
Amelanchier sp. .....+- 0.4 - - + 0.1
Prenanthes trifoliata .... 0.4 - - - 0.1
Carex Bigelowii ......+. wee 0.4 - - - 0.1
Poa palustris ...seeesceeeeeeseeenes « «8 0.7 - - 0.1
VW hought to be a hybrid of D. spinulosa and another Dryopteris species
(Smith, pers. comm. )
{
I
I
i
Table 2:
of Whiteface Mountain, N.Y.
-153-
Relative frequency and number of species in vascular plant families
for the four aspects and for all aspects combined in the treeless community
Family Wo hte. NOTRE. noMShur, NooUR. Nol tRlE.
Lycopodiaceae ........ - - - - 1 8.6 = - 1 2.6
Polypodiaceae ........ 2 6.5 - - - - - - 2 2.3
Pinaceae ... | 2.8 2 14.9 2 7.7 04 1.5 2 6.7
duncaceae .. ot - - - 1 1.0 1 10.3 1 3.1
Cyperaceae . 3 1.2 2 2.7 2 9.6 - - 3 3.9
Gramineae .. 4 14.9 2 2.0 2 3.3 1 10.3 5 9.6
Liliaceae .... 3 44 - - - - - - 3 1.6
Carophyl laceae - - - - 1 7.201 6.6 1 4.0
Salicaceae ...... - - - - i] 5.3 1 4.5 1 3.0
Corylaceae .. 2 3.5 1 5.4 1 5.3 1 1.5 2 4.1
Ranunculceae .. 1 1.20 + - - - - - 1 0.4
Saxifragaceae . 1 7.70 = - - - 1 0.5 1 2.8
Roxaceae .... 7 14.5 2 6.9 1 11.5 2 20.8 7 18.8
Araliaceae . 1 04 - — = - - - - 1 0.1
Cornaceae 1 11.7) 1 «10.9 1 3.8 1 8.1 1 9.8
Ericaceae .. 4 14.9 2 44.1 3 30.1 4 21.3 4 29.6
Gentianaceae .. aga Ol 4.4] 27° - - - - 1 2.1
Caprifoliaceae ....... 1 0.4 - - - - - - 1 0.1
Compositae ........... 3 12.1 3 10.2 1 6.2 3 14.7 5 11.5
-154-
All but one of the major group I species with maxima on the north
were woody perennials, and only one woody species (V. angustifolium) was
not most frequent on the north. This would suggest that conditions for
woody shrub growth are best on this aspect. The lack of Vaccinium
uliginosum on the east, and prevalence on south, north and west aspects
ndicates a wide tolerance to contrasting microenvironments, but an
inability to compete with the many species growing on the east facing
slopes. Species in group I show an affinity for cold, moist sites,
based on their distribution patterns over the aspect gradient.
The three species in group II (Figure 2c) with essentially similar
distribution patterns are all considered to be true tundra species
(Woodin, 1959); the fourth, Vaccinium angustifolium, is not. This group
includes all tundra forbs which occurred on every aspect. Species in
group II withstand the greatest extremes of temperature, transpirational
stress, and lowest moisture levels, yet are apparently least able to
compete on the densely covered east and north aspects.
Figure 3a includes two light-intolerant forbs which normally are
found in the understory of the spruce-fir forest (Nicholson, 1965).
These are most prevalent on the east and north aspects and least so on
the south and west ones. The latter is apparently also least favorable.
for spruce-fir forest species.
that this species has a low competative abi ut can withstand the
rigorous windswept climate of the west exposure. On the other hand, the
many species confined to the east (Table 1) apparently have little or
no tolerance for the extreme condtions found on the south and west aspects.
Many of these species are characteristic of the spruce-fir forest.
The distribution pattern of Carex Houghtonii (Figure 2d) suggests
ty,
Of the 43 species recorded in the treeless community, 25 (58%) were
most frequent on the east. In all, 34 species were found on the east
aspect, at least twice the number found on any other aspect. In contrast
to the east, only four of the 43 species (9%) were most frequent on the
west aspect.
Twenty-one species were restricted to only one aspect. Of these,
17 were on the east aspect. The other four were: Poa palustris and
Achillea millefolium (north), Prenanthes Bottii (south), and Louped i
SeTago (west). ave been observed on the east, but were not foun
within the quadrants sampled.
The absence of Poa palustris and Achillea millefolium on the south
and west aspects, where bare soil areas are common, may be indicative
of more severe growth conditions than on the east and north facing slopes.
These are adventive species whose distribution does not necessarily
reflect natural microenvironmental conditions, since they are often
confined to disturbed areas. All 17 species which were restricted to
the east aspect have strongest affinities to the spruce-fir or other
forest types in the Whiteface Mountain area (Nicholson, 1965).
-155-
Figure 3: Area-based species diversity of the four aspects (page 156).
Figure 4: Average number of species per family on the four
! major aspects (page 157, left).
Figure 5: Average frequency per species in the 40 quadrant
| samples of the four aspects (page 157, right).
~156-
WN
AI
AI
ANI
Average Frequency Per Species in 40 Meter Q
-NUDRUOANMWOSO—NW
T T ' T T T T T T T T T
AI
ANI
NI
ANI
7£St-
-158-
The relative importance of plant families estimated by summing the
relative frequencies of component species is shown in Table 2. The
Ericaceae was the dominant family in the treeless conmunity with total
relative frequency of 29.6%. -It was the most important family on all
four aspects. Its prevalence in tundra has often been reported (Oosting, 1958).
Rosaceae was second in total relative frequency (15.8%), and was. also
common on every aspect but the north one. Three other families exceeding
7% in overall relative frequency were the Compositae (12.4%), the
Graminea (9.6%), and the Cornaceae (9.8%). Cornaceae is only represented
by one species, Cornus canadensis, while each of the other families which
exceeded 4% in overall relative frequency were represented by at least
two species.
The high prevalence of Rosaceae is not typical of most alpine tundra
regions. The occurrence of several spruce-fir species on the east and
Potentilla tridentata elsewhere accounts for its high value. Compositae,
Gramineae, and Caryophyllaceae are frequently more characteristic
families in tundra regions (Oosting, 1958).
Interesting relationships between family phylogenetic standing
(after Fernald, 1950) and their distribution over the aspect gradient
were indicated by the data. All 10 of the most advanced families recorded
from the treeless community of Whiteface were present on the east aspect.
Seven of the nine most advanced families were also most frequent on the
east slope. In contrast five of the ten primitive families had maxima
on the west, while none of the nine advanced families were most frequent
on the west. This suggests that the more primitive vascular plant groups
are tolerant of more rigorous environments such as are found on the west
and south, but are intolerant of competition from the many species which
can grow in the more favorable environment on the east aspect.
The total number of species for the four aspects are given in Figure 3.
It is evident that from this area-based measure that species diversity
was greatest on the east with little variation on the other three aspects.
The east also had the highest species diversity when using a measure
based upon number of species per family shown in Figure 4. Average number
of species per family was lowest on the west and south. When the four
‘aspects were compared for diversity expressed as total frequency of
species divided by the total number of species, the east was. again the
most diverse as shown in Figure 5, followed in order by the north, south
and west aspects.
DISCUSSION
From the evidence presented in Figures 3 thru 5 and also Tables 1
and 2, it is clear that the vascular flora of the east aspect is more
diverse than that of any other aspect inthe treeless community of
Whiteface Mountain. Floristically the east aspect must be regarded as
an ecotone, while that on the other three aspects may be more properly
-159~
considered as true tundra. The species associations and diversity measure
suggest spruce-fir grades into tundra from east-north-south-west. This
is in close agreement with forest-slope aspect relationships described
by Holway et. al. (1969), ;
On Whiteface Mountain the best developed soils are found on the
east aspect. Soils on the west and south are thin and rocky. These
unfavorable substrate regimes could account in part for the lack of
vegetation and species found on these aspects. However, it is still
probable that a rigorous microclimate permits the survival of only tundra
species. The south and west aspects should have greater temperature
extremes than either the east or the north aspects, and should also be
drier. Thin soils and exposure to strong prevailing winds should result
in more transpirational stress on south and west sites than on east and
north aspects.
In summary, the less favorable microclimatic conditions prevailing
on the south and west aspects is thought to be a major cause of species
composition in the treeless community of Whiteface. Milder environmental
conditions on the east aspect allow for the development of a more diverse
flora which is “ecotonal" between spruce-fir and tundra. Increasing
environmental stress with change in aspect results in a shift to more
tundra-like vegetation which reaches its maximum development on the west
aspect.
REFERENCES
Breisch, A.R., J.T. Scott, R.A. Park and P.C. Lemon. 1969. Multi-
Dimensional ordination of Boreal and Hardwood Forests on Whiteface
Mountain. Publ. No. 92. Atmosph. Sci. Res. Center, State Univ.
of N.Y. at Albany, Albany, N.Y. 12203.
Falconer, R. 1963. Weather summary for Whiteface Mountain, Wilmington,
New York. Report of Whiteface Mountain activities of the Atmospheric
Sciences Research Center, Publ. 15; 46-59.
» 1964, Weather summary for Whiteface Mountain, Wilmington,
‘ Wew York. Atmospheric Sciences Research Center Research Reports
1964, No. 33: 67-72.
Fernald, M.L. 1950. Gray's manual of botany, 8th Ed. American Book Co.,
New York. 1632 pp.
Feuer, R., C.C. Lowe, H.B. Hartwig, and M. Peech, 1963. New York
agrteul tune at a glance. N.Y.S. Coll. of Agriculture, Ithaca, N.Y.
pp. 5-6.
France, 0, and P.C. Lemon, 1963, Preliminary observations on forest
tree ecology of the Whiteface Mountain area. Report of the
Atmospheric Sciences Research Center Summer 1962, Publ. 15: 74-112.
Holway, J.G., J.T. Scott and S. Nicholson. 1969. Vegetation of the
Whiteface Mountain region of the Adirondacks. Publ. No. 92, Atmosph.
Sci. Res. Center, State Univ. of N.Y. at Albany, Albany, N.Y. 12203.
-160-
Isachsen, Y.W. 1964. The geology of the Adirondacks. Public Lecture,
ASRC Series, Summer 1964, Wilmington, N.Y.
Nicholson, Stuart, 1965. Altitudinal and exposure variations of the
spruce-fir forest of Whiteface Mountain, N.Y. M.S. Thesis, State
University of New York at Albany, Albany, N.Y.
Oosting, H.J. 1958. The study of plant communities. W.H. Freeman
San Francisco.
Reilly, R.W. 1964. A general ecological study of the moss flora of
wid betace Mountain. Whiteface Mountain Summer Program 1963, Publ. No.
~83.
Rogers, H.J. 1964. Personal Communication, Ausable Forks, N.Y.
Smith, S.J. 1964. Personal Communication, Wilmington, N.Y.
Stout, N.J. 1956. Atlas of forestry of New York State. State u. of N.Y.
Coll. of Forestry, Syracuse, N.Y.
Wallace, E.R. 1896. Wallace's guide to the Adirondacks. Syracuse, N.Y.
Watson, W.C. 1869. History of Essex County. Munsell, Albany. P. 170.
Woodin, H.E. 1959. Establishment of a permanent vegetational transect
above timberline on Mt. Marcy, New York. Ecology 40: 320-322.
i
i
ECOLOGICAL EFFECTIVENESS OF YELLOW BIRCH
IN SEVERAL ADIRONDACK FOREST TYPES
By
Ronald F. Kujawski and Paul C. Lemon
ECOLOGICAL EFFECTIVENESS OF YELLOW BIRCH
IN SEVERAL ADIRONDACK FOREST TYPES
By
Ronald F. Kujawski and Paul C. Lemon
ABSTRACT
The ecology of yellow birch (Betula alleghaniensis) in certain
forest types of the Adirondacks was studied by means of vegetational
analysis of forested stands. Sampling data provided information on
the size, distribution and reproduction of yellow birch in these forest
types, Optimal ecological effectiveness was demonstrated in those
forest types characterized by moist sites. In forests of the red
spruce-yellow birch type, a competitive advantage was shown by yellow
birch as a result of successful establishment of seedlings.
INTRODUCTION
Yellow birch (Betula alleghaniensis Britton), a species of the
northern forests, has received passing notice in numerous investigations
but has been of primary concern only recently. Special attention is
justified by its economic importance and because it appears to be in
decline in many areas, Yellow birch is considered the third most
important hardwood, economically, in New York, first in importance in >
New Hampshire (Gilber 1960) and first in the Great Lakes area (Jacobs 1960),
Nevertheless, this species has been decreasing in numbers in many areas
due to "birch dieback" (Leak 1961) and poor growth or poor regeneration
following cutting (Fraser 1956, Godman and Krefting 1960, Jarvis 1957).
Much research is presently being conducted in the Great Lakes Region
+ and in northeastern United States including the Adirondacks. Although
there is some current literature concerning yellow birch, mainly
published by the U.S. Forest Service, the emphasis is often placed on
studies of economic value and management. Ecologists, attempting to
define the environmental niche of this species, have disagreed as to
the physical, chemical and biological parameters determining its require~
ments for optimum success (Godman and Krefting 1960, Fraser 1956). It
is important to obtain precise information on the regeneration, germination
and growth characteristics of yellow birch. fi
Ecologically oriented research has been done by Winget, Cottam and
Kozlowski 1965) who have studied species association and seedbed
conditions for germination of yellow birch seed in Wisconsin. Stearns
(1951) examined the role of yellow birch in a climax sugar maple-hemlock-
yellow birch forest. These investigators conclude that yellow birch,
-162-
i
I
'
i
-163-
although appearing in the climax, is a true gap-phase species normally
dependent upon disturbance in the forest for regeneration. Godman and
Krefting (1960), working in Michigan, studied some of the factors
important to yellow birch establishment. Winget and Kozlowski (1965)
also studied germination and seedling growth of yellow birch. All agreed
that germination and establishment were poorest on undecomposed organic
material and leaf litter, Godman and Krefting found that the best
seedbed was one of mixed organic and mineral soil ,while Winget and
Kozlowski found that peak germination and establishment occurred on intact
humus over mineral soil. Neither offered specific reasons to explain
these observations. Tubbs and Oberg (1966), finding that height growth
of yellow birch was best in a mixture of organic and mineral soil ina
1:1 ratio, concluded that this type of substrate offered the most favorable
moisture conditions.
Statements by the above authors appear to apply to conditions
existing around the Great Lakes, but studies and observations of yellow
birch in the Adirondacks and much of the northeast are not always in
agreement with these midwestern studies. Studies of the vegetation of
the Adirondack region (France and Lemon 1963, Scott and Nicholson 1964,
Kujawski and Lemon 1967) have indicated-that yellow birch is indeed a
strong competitor and seems to have a more important role in the climax
than attributed by the authors of the Great Lakes studies.
The study described here represents an effort to examine some of
the features of sites where yellow birch is growing and to determine
the ecological effectiveness of yellow birch in certain forest types of
the northern forest and of the Adirondack region in particular. The
importance of some environmental factors determining effectiveness is
briefly discussed.
The ecological effectiveness of yellow birch is assessed from the
importance values calculated for it in the forest type concerned and
from data on regeneration. The figures included were obtained from two
regions of the Adirondacks. The first was in the area of Whiteface
Mountain, typical of the rugged and mountainous portion of the Adirondack
. region (Figure 1). The second was around Cranberry Lake, less mountainous,
but hilly and in the western portion of the region.
GEOGRAPHY
The Adirondacks exist over underlying formations of igneous and
metamorphic rocks (Cressey 1966). The upland areas of the Adirondacks
consist of a domed Pre-Cambrian-erosion surface with the erosional
remnants, monadnocks, forming high and rugged peaks such as Mount Marcy
and Whiteface Mountain. The prevailing rock base is crystalline,
resembling that of the Canadian shield, As a result of intense glaciation,
most of the original soil has been removed and land surfaces smoothed
4
Cranberry
Lake
Location of. study areas.
Figure 1:
Mt.
Whiteface
“9l-
-165-
filling valleys with gravel, sand and silt. Some of the eroded material
has clogged the preglacial valleys, upsetting drainage patterns and
producing numerous lakes. Lacustrine deposits occur around the margins
of the upland. The area around Cranberry Lake is in part covered by
sandy lacustrine deposits laid down in glacial lakes.
Soils are typically shallow and acid on glacial till over steep
terrain (De Laubenfels 1966). The effect of the glaciation has been to
create stony conditions and to influence drainage patterns often creating
poor drainage in many areas. The overall result is soils with extreme
acid conditions, shallow profiles, rockiness and poor drainage.
High in elevation, the general climatology of the Adirondacks is
cool and moist (Carter 1966). The usual number of frost-free days during
a year ranges from 85 to 140 days. Annual precipitation is about 40 to
44 inches (1 to 1.1 m.), the wettest portion of the state. Snowfall
ranges from 90 to 165 inches (2.3 to 4.2 m.) per year. The winters are
cold with a mean January temperature of about 15 degrees F. (-9 degrees C.)
and summers cool with a mean temperature of 65 degrees F. (18 degrees C.).
VEGETATION
The Adirondacks are in the vegetational domain referred to as the
Hemlock-White Pine-Northern Hardwoods Region according to Braun (1950).
She divides the vegetation of the Adirondacks into four major types
defined by different topographical situations. These are the swamp
forests, spruce flats or mixed woood, the hardwood forest and the spruce
slope. Swamp forest occurs in the wet lands, partially filled depressions
and low spots around streams and lakes. The major species occupying this
type are red spruce, yellow birch and balsam fir. The second type inhabits
moist lower slopes and rolling flats in valleys and around lakes. Spruce
and yellow birch are the dominant species. The hardwoods occur on well
drained benches and gentle slopes over deep soil. Again, red spruce
dominates with beech, sugar maple and yellow birch high in density. The
spruce slope is characterized by steep and rugged mountain slopes with
“thin soil. Spruce, yellow birch and beech are the leading species.
Overall, Braun estimates yellow birch to be the second leading species
in total density behind red spruce throughout the Adirondack region.
Further consideration will be given to the observations of Braun and to
these forest types later.
METHOD
There were two principal steps in the collection of data. The first
phase was concerned with estimating stand structure or composition. A
combination of methods was employed in gathering data. A 30 Basal Area
Factor (BAF) forester's prism was used to estimate basal area (Bruce 1955).
-166-
Density data were obtained from 1/80 hectare circular plots. Frequency
data were obtained by the quarter method since it was felt better
resolution of frequency of species could be obtained by this method over
that of the plot method. (Lindsey, Barton and Miles 1958). The field
procedure was to locate stands in which yellow birch appeared to be a
prominent member. A starting point of sampling was arbitrarily selected
by tossing a rod over a shoulder, the point for sampling being where the
rod landed. A11. subsequent sampling points were at 20 paces about 15 m.)
from the previous point following a compass line so that a grid of points
was established.
Basal area, density and frequency were converted to relative values
following the procedure of Curtis and McIntosh (1951). The sum of
relative basal area, relative density and relative frequency thus gave
a value called importance value (IV) of a given species in the stand
under study. Importance values here are relativized (RIV) and thus
presented on a basis of: 100%.
Using RIV as a measure of predominance, each stand, after careful
examination, was then placed under the appropriate forest type. The
forest types used for this grouping are defined by the Society of
American Foresters for Eastern North America (1954). A total of 40 |
stands were grouped in this manner. The following criteria were used
for accepting or rejecting a stand for analysis. First, yellow birch
must have a relative density of at least 10% in the stand. This is to
eliminate stands were yellow birch may have occurred only as an incidental.
Secondly, the stands must be fairly homogeneous without signs of recent
or gross disturbances.
The 40 stands finally accepted were grouped under four forest types
which seemed to include all these stands. These were types 5, 24, 25,
and 30.
Type 5 is dominated by balsam fir with yellow birch as one of the
associates on moist upland but flat sites. Only one stand was found to
fit into this type and care must be taken when interpreting results.
prep in the zone just below timberline, this type is considered to be
subcl imax.
Type 24 is dominated by hemlock and yellow birch with sugar maple
and beech among the associates. In the Adirondack region this type is i
found most often on moist flats or on north slopes. Stands found to
fit this type were al] on north facing gentle slopes. Considered to be
a subclimax type, over long periods of time it may give way to hemlock .
or: to sugar maple-beech-yellow birch on drier sites. It is in such :
situations that yellow birch has been found by many investigators to
have achieved maximum development (Winget et al. 1965, Stearns 1951).
Four stands were placed in this type.
Type 25, the sugar maple~beech-yellow birch type, included most of
-167-
the stands sampled as might be expected sinte this is the most frequent
type. in the Adirondack region. Twenty-five stands seemed to fit into
this type. The major associates were red maple and red spruce which
occasionally reached high importance values. These stands, usually on
more moist sites, may be considered as variant types though not acutally
fitting any of the designated types given by the Society of American
Foresters. These two species may be viewed as indicators of moist
situations or areas of shallow soil (Braun 1950). Yellow birch does not
seem to be affected in any consistent manner by increase or decrease in
RIV of either or both of these species. Type 25, at least on more mesic
sites, may be considered as climax.
Type 30, the red spruce-yellow birch predominant stands, have
balsam fir, red maple, beech and sugar maple as associates. Found on
lower slopes and moist well-drained flats, this type is probably climax
on moist sites. This type may often be referred to as a yellow birch
type, but the decline of yellow birch has been so great that pure yellow
birch stands are rare and are no longer recognized as a separate type.
One stand did have a high RIV for yellow birch, 61%, but since red spruce
was of minor importance, 1%, the stand had to be typed with the sugar
maple-beech-yellow birch type where sugar maple and beech had RIVs of
8% and 14%, respectively, justifying classification as type 25. This
may well be a transition between type 25 and type 30. Braun who further
subdivided this type into several spruce types (mentioned earlier)
discusses McCarthy's (1920) view that the demarcation between mixed wood
type (type 30) and hardwood type (type 25) may be where beech appears
and balsam stops. This was taken into consideration when judging into:
which types sampled stands were to be placed.
To understand development of yellow birch in the various types both
as individuals and in terms of total development per unit area, data
are presented in terms of basal area per tree and basal area per acre.
Regeneration characteristics are examined by calculating the relative
frequency of saplings and seedlings in each stand, Sapling data were
collected by the quarter method whereas seedling counts came from square
meter plots.
. The second major aspect of this study was the collection of data
on the environmental factors that might be determining the importance
and effectiveness of yellow birch on different sites. To assess the
importance of these factors, two stands, varying in composition, but
adjacent to each other were selected for this comparative study.
The stands selected are in a small valley south of Whiteface Mountain.
The first site is at an elevation of 1950 ft. (594m). The slope of the
land is slight with the valley running in a south-west to north-east
aspect. Slightly higher, 2044 ft. (623 m) on the average, the second
site is situated on a slope of about 10°. The slope aspect is north to
north-west. On either side of the two sites are small ridges.
~168-
The appearance of the topography suggested that this valley may have
been a river valley, the river having been cut off by the continental
glacier, As there are many large, rounded boulders and a profusion of -
shallow streams and of dry stream beds both on and under the surface of
the forest floor, it would seem reasonable to arrive at this conclusion,
The second site does not have as many boulders or streams, but lies south
of the first site almost in line with the proposed direction of the river
(Craft 1966, personal communication). It may be that some of this area
had been filled with glacial till. :
The first site was quite flat and poorly drained as standing water
was frequently observed in areas of the stand. The second site, on the
other hand, appeared to have good drainage.
The nature of vegetation in these stands was calculated from field
sampling data. One tenth acre square plots were set up, five in each
stand, These figures gave values for the basal area per acre and density
per acre for the tree species. Trees were taken to a minimum dbh
(diameter at breast height) of 4 in. as was true of the sampling of the
sampling of stands for forest type analysis. Sapling data and seedling
data were collected by the procedure given above.
In each stand an instrument shelter was installed containing a
hygrothermograph, maximum-minimum thermometer and two alcohol thermometers.
These were for the purpose of observing not only ranges of temperature and
humidity during the short term of the study, but also to observe any
differences in these factors between the two sites. Instruments were
frequently checked for accuracy with a sling psychrometer. Within the
same stand there may be differences in temperature and humidity due to
variations in crown density or canopy. Major differences between the
two sites therefore would have more meaning when considering any
discrepancy in microclimate. Humidity and temperature readings were taken
throughout each of the stands in order to note the degree of variation
within a stand.
Selected characteristics of the soil in each stand were examined.
Soil moisture was measured using a lawn moisture meter (electrical
‘meter). The sensitivity of the meter was increased by connecting a
potentiometer to the simple circuit and replacing the original meter
with a more sensitive ammeter. Because jonic features of the soi]
would have an effect on the readings obtained by a meter of this sort,
only relative measurements are obtained. Actual soil moisture was
measured by oven drying samples of soil collected at each site. Soil
temperatures were taken using dial thermometers inserted to a depth of
20 cm. Samples of readings were taken at each of the sites.
Soil texture was measured by sifting oven dried soil samples
through a series of soil sieves. The mesh diameters of the sieves were
checked using a stereo zoom scope. The mesh diameters were found to be
Imm., .7mm,, .25mm., and .12 mm. Soil particles were classified as follows:
-169-
Table 1: Summary of importance values for major species in each stand.
}
I
|
[.
P. Stand Yellow Sugar Beech Red Hemlock Spruce Balsam
| Number Birch Maple Maple Fir
| TYPE 5: Balsam Fir :
| ¥4 21 1 22 47
| TYPE 24: Hemlock ~ Yellow Birch
i T23 13 4 12 5 56 8 1
i 173 18 7 10 1 48 a]
i 125 19 9 18 49 5
172 29 5 8 38 Wo
| Mean 18 7 A q Li 9
TYPE 30: Red Spruce - Yellow Birch
[ 83 19 20 4 4] 1
} 135 20 8 13 18 29 10
i 85 21 4 5 2 51
15 26 9 34 9 18
106 27 1 6 45 10
164 33 31 12
128 32 1 3 12 8 26 13
14 33 5 VW 37 6 5 1
107 41 10 10 3 22 7
_3 40 6 29 ] 22
Mean 29 6 |e Ts > 2 “Ss
TYPE 25: Sugar Maple - Beech - Yellow Birch
i q9 16 31 29 1 7 14 1
| 97 19 12 22 9 5 20 8
| 19 20.35 32 5 3 5
| 156 20 25 13 38 1
W 22 16 32 27 1 2
18 20 39 34 3 2
4 23 12 48 4 13
87 23 4) 21 13
159 24 10 40 3 8 9 3
9 28 10 27 20 14
5 30 27 32 2 2 1
I 71 30 23 30 1 9
; 12 29 16 23 21 7 2
i 109 32 15 33 15 2
i 7 33 33 30 2
| 1 34 34 25 4
105 35 20 25 19 1
131 37 16 35 9 1
84 37 30 12 7
6 42 1 35 1 3 7 1
2 41 35 16 1
16 46 i 18 13 2 5
130 52 13 12 10 6 5
mara 61 10 14 12 2 1
Mean 3T 2 a “7 ~ “7 TT
-170-
Table 2: Summary of basal area data of yellow birch and topographic
data in each stand.
Stand Sq. In. Sq. Ft. Elevation Slope Slope
Number Per Tree Per Acre (feet) (degrees) Aspect
TYPE 5: Balsam fir
182 53.4 30.0 2650 16 29
TYPE 24: Hemlock - Yellow Birch
23
133.4 22.5 1720 4 200
173 209.7 49.5 1575 3 257
125 166.7 30.0 1780 4 126
72 172.6 33.0 1615 3 76
Mean 70.6 Kay
TYPE 30: Red Spruce - Yellow Birch
83 57.2 18.0 2560 28 164
135 147.4 42.0 1640 1 29
85 109.1 27.0 2320 24 248
15 155.5 46.8 1550 15 300
106 92.1 » 43.5 2200 7 240
164 211.9 81.0 2520 W 125
128 116.2 54.0 1500 5 321
14 90.7 45.0 1500 8 302
107 107.3 70.0 1980 9 246
_3 56.2 41.8 1500 6 278
Mean 14.4 46.9
TYPE 25: Sugar Maple - Beech - Yellow Birch
49 228.6 36.0 1710 2 36
97 138.7 39.0 2000 15 268
19 92.2 20.3 1650 12 329
156 76.2 24.0 1660 6 131
an 161.3 34.5 1500 8 180
18 77.8 18.0 1700 6 142
4 73.4 26.2 1500 8 207
, 87 175.3 34.5 2550 7 243
159 317.6 50.0 1620 11 248
9 128.2 42.2 1700 13 327
5 100.8 36.0 1500 12 309
71 154.7 51.0 1920 10 54
12 118.1 45.7 1500 7 124
109 171.2 68.0 1950 5 161
7 190.1 45.0 1450 19 16
1 105.5 47.8 1480 ‘ 18 8
105 126.0 68.0 2180 22 182
131 146.4 67.5 2044 8 316
84 239.3 73.2 2450 25 18
6 123.8 61.4 1500 7 270
2 57.6 39.7 1450 9 259
16 70.6 50.2 1500 19 22
130 106.5 71.3 1954 2 ---
V7 70.6 62.9 1600 9 n
33.6 %.3
-171-
Figure 2: Mean importance values of major tree species in
forest types 24, 25 and 30 with the range and
standard deviation given for type 25 (page 172).
| : Figure 3: Comparison of basal area of yellow birch in
forest types (page 173).
70
60
50
40
20
10
25
30
24
-W2-
IMPORTANCE VALUES
( Sugar Maple - Beech
Yellow Birch Comm. )
—24
25
25
= - 25: 25
—30 24 pr 8° 4
25 2s
Y.B.
BEECH RED M. BAL. FIR
S.M. SPRUCE HEMLOCK
-173-
Curbs) a3uL “SAV JO Vau"v IWSVa
So
$24 °
& 8 2 8
im T T T T T T T T T v t T T T T t T 7
3 :
a © F
go 2a.
ran N eX
zé nN,
S86 E\ 2
2 4
oe
a VLLLEE
\0A
2
l
Loa L 1 1 1 1 n it 1 L 1 1 l 1 1 1 1
° fo} fo) °
@ © vt N
(4a°bS) vauv IvsvE
24 25 30
FOREST TYPE
5
=174-
1 mm and greater ..... gravel
-70 mm-1.0mm ........coarse sand Li
.25 mm-,70mm ........medium sand
-12 mm-.25nm ........fine sand
less than .12 mm ...,.silt and clay
A soil borer was used to take cores for the purpose of noting the
depth of organic material in each stand. Soi] samples were collected
and placed in small vials for transport to the field station where pH
analyses were made. Difficulty with electrical pH meters resulted in F]
the use of pH paper to determine the pH. A mixture of 50% soil and 50%
distilled water was made for each sample and the pH recorded to a range
of .5 (e.g. 4.0-4.5).
RESULTS
Stand Composition:
Tables 1 and 2 present a summary of the stands sampled and
used in this analysis. The RIVs for the major species in each stand
are given as well as the mean elevation, slope and slope aspect. Basal
area in terms of mean basal area per tree and the mean basal area per
acre are given for each stand. Figure 2 represents the mean RIVs of the
major species of stands grouped under their respective forest type.
Only the species which achieved a RIV of 10% in one or more of the stands .
are presented. It is of interest to see whether there is appreciable
variation in RIV of yellow birch from one type to another. Of the three .
main types considered, it seems as if yellow birch attains its highest
mean RIV in type 25 and also the highest individual RIV as noted from
the range (Table 3). It is somewhat surprising that the mean RIV of
yellow birch in the spruce-yellow birch type is very similar to that of
the maple type and is not very high in the hemlock-yellow birch. The
range of RIV for yellow birch in type 25 is greater than that for the
spruce type. The implications will be discussed later.
. Basal Area:
Defining ecological effectiveness in terms of basal area may
add more meaning to RIVs presented for yellow birch and may help resolve
the question of optimal effectiveness. The data, as summarized in
Figure 3 or from Table 2, indicate total basal area per acre is greatest
in the maple and spruce types, but the best average development per tree
occurred in the hemlock-yellow birch as has been observed by many te
investigators (Winget et al. 1965, Stearns 1951). The low basal area
per acre for yellow birch in the hemlock type indicates a low density
for yellow birch. Development of individual trees in spruce types is “
quite good though not the best.
Regeneration:
Tables 4 and 5 summarize the data on sapling and seedling
relative frequencies for each stand under the appropriate forest type. .
~175-
Table 3: Summary of importance values for yellow birch in four forest
types (Soc. Amer. Foresters classification).
Forest Type S.A.F. No. Mean Range Standard
Code Stands RIV Deviation
Sugar Maple-Beech-
Yellow Birch 25 25 31 16-61 n
Red Spruce-Yellow
Birch 30 10 29 19-41 8
Hemlock-Yellow Birch 24 4 18 13-20 3
Balsam Fir 5 1 10 --- --
~176-
Table 4: Summary of % frequency of saplings for major tree species in
stands sampled.
Stand Yellow Spruce Beech — Sugar Red Hemlock Balsam
Number Birch Maple Maple Fir
TYPE 5: Balsam Fir
T82 10 7 2 47
TYPE 24; Hemlock - Yellow Birch
T23 3 15
18 3 40
173 8 8 6 38 6
125 26 37 18 2
172 3 8 24 8 13
Mean T T acy oe Ky 2 oo
TYPE 30: Red Spruce - Yellow Birch
33 9 24 22 4 20
135 2 34 7 3 5 3 25
85 16 12 W7
15 15 19 2 2 29
106 i 27 2 5 2
164 9 20 2 2 7
128 12 35 6 6 6 29
14 15 16 30 18 15
107 9 12 7 8
_3 A? 7 iv 18 27 _
Mean 2 ee “7 “Ss “9 wT
TYPE 25: Sugar Maple - Beech - Yellow Birch
49 18 62 16
97 10 15 23 18 5 2 8
19 8 47 41
156 5 33 20 2 2
3] 5 4 46 15 6
18 6 42 44
4 7 3 47 33
87 14 23 37 2
159 4 14 25 9 3 4 8
9 10 16 39 20 7
5 2 4 46 23
71 2 13 18 48
12 8 3 37 33 5
109 4 28 33
7 2 2 24 57
1 8 27 38
105 7 15 22 22 1 3
131 9 2 23 36
84 2 2 8 29
6 15 15 36 7 1
2 5 32 43
16 V7 34 34 2
130 13 20 13 21 6 3 2
iv ral 6 24 33 1 _ 1
Mean “7 “7 32 30 z x
-177-
Table 5: Summary of % frequency of seedlings for major tree species
in stands sampled.
Stand Yellow Sugar Beech Red Hemlock Spruce Balsam
Number Birch Maple Maple Fir
TYPE 5: Balsam Fir
T82 18
TYPE 24: Hemlock - Yellow Birch
123 13 13 16 8 13
173 32 4 i 15 4 4
125 7 14 31 4 10
172 44 3 6 3 6
Mean oT oO Te 7 ov 3
TYPE 30: Red Spruce - Yellow Birch
83 13 19 6 23 2
135 4 4 4 62 10 4
85 12 15 19
15 25 10 20 25
106 14 9 13 22 16
164 8 16 VN 8
128 10 13 13 5 27 3
14 5 7 24 24 3 5
107 3 48 3 3 20
_3 37 16 3 26 _ 5
Mean 3 7 7 F 4q TO >
TYPE 25: Sugar Maple - Beech - Yellow Birch
149 45 24
97 4 23 9 12 8 9
19 20 30 22 5
156 4 22 7 3]
a] 19 22 15 20
18 15 33 26 3
4 V7 27 37 1
87 51 15 3
159 13 23 8 19 4
9 31 10 i 14 1
5 23 24 27 1
7 8 35 29 4
12 28 24 13 19
109 8 78 3
7 29 29 iW 1
1 + 25 28 15 1
105 4 43 8 3 10
131 3 44 19 6 6
84 41
6 31 24 18 5 2
2 20 55 14
16 8 29 21 13
5 34 14 19 5
5 33 WwW 14
Mean 3 oF Te a) ~ T T
a
Table 6: Summary of tree sampling in stand 1.
~178~
Species Stems per Sq. Ft. Sq. In. % &
acre per acre per tree Density BA
Betula
al leghaniensis. 126 97.9 112 52.1 65.9
(YeTlow Birch)
Fagus grandifolia 40 20.2 73 16.5 13.6
Beech
Acer saccharum 36 15.6 62 14,9 10.5
(Sugar Mapte)
Picea rubens 10 4.2 60 4.1 2.8
(Spruce)
Acer rubrum 18 6.3 50 7.4 4.2
(Red Maple)
Betula papyrifera 8 3.1 55 3.3 2.1
(Paper Birchy
Abies balsamea 1.0 72 0.8 0.7
(Balsam Fir)
0.4 28 0.8 0.3
Pyrus decora
(hours fn Ash)
Table 7: Summary of tree sampling in stand 2.
Species Stems per Sq. Ft. Sq. In. % %
7 acre per acre per tree Density BA
Fagus grandifolia 80 41,8 75 44.0 22.4
(Beech
Betula
alleghaniensis 44 77.5 254 24.2 41.7
(YeTTow Birchy ~
Acer saccharum 46 58.2 182 25.3 31.3
(Sugar Maple)
Picea rubens 6 3.3 80 3,3 1.8
spruce)
Betula papyrifera 2 2.6 187 1 1.4
(Waper Birch) ;
Acer rubrum 2 2.4 75 11 1:3
(Red Maple)
Abies balsamea 2 0.2 13 La 0.1
(Balsam Fir)
~179-
An estimate of effectiveness and-of future trends may be obtained by
examining sapling and seedling relative frequencies. Seedling relative
frequencies for yellow birch are the same in the maple and spruce types
thus continuing the similarity of effectiveness of yellow birch in these
types. Regeneration under the canopy of hemlock-yellow birch was
apparently neqligible as no seedlings were counted in these stands.
Table 4 indicates that the relative frequency of saplings was a little
greater for yellow birch in the spruce-yellow birch type. This suggests
perhaps less competition or better conditions for growth from seedling
to sapling stage in the spruce-yellow birch type.
Site Comparison:
Tables 6 and 7 summarize the results of the tree sampling in
the two stands under comparison. In stand 1, a flat and moist site,
yellow birch was by far the dominant species. With a relative density
of 52% and comprising 66% of the total basal area of the stand, it would
appear that this site was well suited to the growth of yellow birch.
Beech and sugar maple were the next most frequently appearina species.
Other trees found but not as frequent were spruce and red maple with some
paper birch, balsam fir and mountain ash.
Examination of size of the various species indicates that in terms
of basal area per stem yellow birch was by far the largest with an average
basal area per tree of 112 square inches. Beech with an average of 73
square inches per tree and sugar maple with 62 square inches per tree
were the largest of the remaining species.
A further breakdown of trees into diameter classes indicates a
broad distribution for yellow birch (Tables 8 and 9), The majority of
yellow birch occurs in smaller diameter classes, but the range extends
beyond the 16 in. class with a maximum at 33.3 in. Sixteen percent of
the 63 yellow birch sampled had diameters larger than 16 in. Beech is
more evenly distributed than the yellow birch yet the range of size is
not as great with the maximum at 15.3 in. Sugar maple trees sampled do
not exceed 14.5 in. in diameter and are concentrated in the smaller
diameter classes indicating, as with yellow birch, that there are many
.young trees in this stand. It is interesting that in the case of red
maple, there is an absence of trees in the range 10 to 16 in,
In stand 2, the table indicates that yellow birch is not the most
common species, beech and sugar maple occur in larger numbers per acre,
with beech comprising 44% of the total density and sugar maple 25%. In
basal area per acre, yellow birch was calculated to have 78 square feet
per acre or 31% of the total. Although the number of stems per acre is
greater for many of the species in stand 2, yellow birch, spruce and red
maple are substantially less.
On the average it appears that trees are much larger in this stand
than in the first. Beech has a greater average basal area per tree.
Examination of the breakdown of sampled species, with respect to diameter
-180-
classes, indicates that beech is again fairly well represented in each
class, with a majority in smaller diameter classes. Yellow birch has
an average basal area of 254 sq. in. per tree, larger than that of the
other species and much greater than that in the first stand. Of yellow
birch sampled on this site, 36% had diameters exceeding 16 in. with
the maximum at 33.2 in. Sugar maples were quite large and the majority
of those sampled are more than 16 in. in diameter. The largest listed
for sugar maple was 24.3 in. Most of the other trees sampled were quite
large, with few or no small stems. This might indicate that they are
incidentals or the last of a previous successional stage. The appearance
of fewer red maple might suggest that this is less moist than the first.
Sapling data for each of the stands are presented in Tables 10 and 11.
Sugar maple is the most frequently occurring species in either stand, but
with a higher percent density and frequency in stand 2 where it was found
to be the second leading dominant tree species. Spruce is well represented
in stand 1 as a sapling but there are few mature spruce. There are fewer
yellow birch saplings in stand 2. It may be that yellow birch reproduction
is not as good in this stand or that the seedlings are not surviving or do
not become established as readily.
Results of ground flora sampling do not necessarily give a qood
indication of the reproductive activities of the tree species due to
variations in the number of seeds produced from year to year. However
such data are important in identifying the plants associated with the
types of stands under study. Very often these plants give to the trained
ecologist an idea of the nature of the environment.
The ground flora data (Tables 12 and 13) indicate that sugar maple
was by far the leading tree species sampled. Apparently a profuse
seeding of maple occurred the previous autumn, The variation in the
degree of occurrence of each species should be noted for the different
stands. The fewer number of such species as red maple and Dryopteris.
spinulosa in the second stand may be related to a decrease in so
moisture as well. Yellow birch surprisingly did not vary in occurrence
from one site to the other.
It is interesting to note the fewer number of species of trees,
saplings and ground flora in stand 2 with respect to the first site.
The influence of climate upon vegetation is quite well understood
at least in general terms. Yet the influence of small variations in
climate due to topographic or other features of a qiven locale upon the
vegetation of that area is not well understood. It might be suspected
that subtle changes in local climate or microclimate may affect variations
of vegetation in adjacent areas such as the two sites under observation
in the Whiteface Mountain region.
Tables 14 and 15 summarize temperature and relative humidity data
collected during the summer at each of the sites. Unfortunately much
-181-
Table 8: Distribution of trees in various diameter classes by percent
in each class for stand 1. Diameter classes are in inches.
4-6 6-8 8-10 10-12 12-14 14-16 > 16 Max.
Species
Betula alleghaniensis 19 33 22 6 2 2 16 33.3
(Yellow Birch)” :
Fagus grandifolia 20 «2025 18 10° «10 15.3
TBeech -—
Acer saccharum 22 7 33 V7 6 6 14.5
(Sugar MapTe) . '
Acer rubrum 44-33 v 11 16.5
i (Red MapTe)
t Picea rubens 40 40 20 10.6
i (Spruce) : .
\ Betula papyrifera 25 25 50 10.1
Abies balsamea 100 9.5
(Batsam Fir)
| Pyrus decora 100 5:9
} lountain Ash)
Table 9: Distribution of trees by percent in each diameter class for stand 2.
4-6 6-8 8-10 _10-12__12-14 14-16 > 16 Max.
Species. .
Fagus grandifolia | 33. 13 18 13 15 8 3 16.7
Thaechy °
Betula alleghaniensis 18 18 9 5 9 5 36 33.2
(Wellow Birch :
Acer saccharum ~ 22 9 W7 13 39 24,3
[Sugar MapTe) ‘
. Picea rubens . 33 67 11.3
t Spruce)
Betula papyrifera 100 15.3
(Paper Bivehy _
Acer rubrum 100 14.8
(Red Maptey)
Abies balsamea. 100 41
(Batsam Fir)
-182-
Table 10: Summary of sapling data from stand 1. Data collected by
quarter method using 40 points. Number refers to the number of stems
counted and occurrence to the number of points at which each species
was found.
Species Number Occurrence %Density %Frequency
Acer saccharum 40 23 25.0 22.1
(Sugar MapTe) —
Picea rubens 28 2) 17.5 20.2
(Spruce)
Tsuga canadensis 4 3 2.5 2.9
Tentonk)
Acer pensylvanicum 26 16 16.3 15.4
(Striped Mapte)
Betula alleghaniensis 19 13 11.9 12.5
(Yellow Birch)
Fagus grandifolia 19 13 1.9 12.5
‘Beech
Acer spicatum 13 7 8.1 6.7
(Mountain Maple)
Acer rubrum 9 6 5.6 5.8
(Red Mapiey
Abies balsamea 2 2 1.3 1.9
(Batsam Firy
Table 11: Summary of sapling data from stand 2. Data collected by
quarter method using 20 points.
Species Number Occurrence Density Frequency
Acer saccharum 33 16 41.3 36.4
(Sugar Maple)
Acer pensylvanicum 21 VN 26.3 25.0
(Striped Mapte
Fagus grandifolia 14 10 17.5 22.7
Beech
Betula alleghaniensis 7 4 8.8 9.1
(YeTTow Birch)
Acer spicatum 2 2 2.5 4.5
‘(Mountatn Maple)
Picea rubens 3 1 3.8 2.3
Spruce
-183-
Table 12: Summary of ground flora data collected in stand 1. One
i . meter square plots were used in the sampling in which 40 points were
H : taken.
Species Occurrence dFrequency
Acer saccharum 25 15
vita teris spinulosa 22 13
Viburnum alnifolium 21 13
S montana 19 a
. Acar Yubrum 14 8
; Matanthemum canadense 1
Taal a cordifolia — 10
Trientalis borealis
Betula a a Taatensts
Abies balsamea
Picea rubens
Streptopus roseus
teal ja nudicaulis
Wrtchet ta se repens
Rubus hispidus
Hass N PEON
HSH SVN NNNYWREDDO
Table 13: Summary of ground flora sampling in stand 2. Twenty sampling
plots were used in this stand.
Species _ a Occurrence aFrequency
Acer saccharum 16 21
Lycopodium lucidulum 15 20
Viburnum Wburmum avn foT fun cy 12
in)
Dryopteris spfnulose
Fagus Fagus. arand ifolia
Acer
r pensylvanicum
OxaTis montana
Picea rubens
Acer rubrum
Betula alleghaniensis.
Acer spicatum
CTintonia borealis
Matanthemum canadense
SH SS SH NON OATNOO
Boe Hs wwn)]8L
~184-
of the data presented for temperature and humidity is incomplete due to
difficulties with the hygrothermographs. Still some comparison can be
made though the difference between the two sites does not seem to be
significant. There was a slightly higher mean temperature in stand 2
than stand 1. In terms of daily average, maximum and minimum temperatures ,
stand 2 was almost always a degree or two higher. Such a small difference
may be the result of error in the instrument. Interestingly there was
not much variation in the weekly temperature summaries listed for the
period of eight weeks.
Since daily maximum humidity readings were almost always 100% or
near that mark, it is not of much value to present a summary of this
data. Frequently fog could be observed during the early morning hours
in the valley where the two sites were located. It would seem that
maximum humidity would not be of much importance in affecting a difference
in vegetation at the two sites. The summary of minimum humidity means
for the weeks of study does indicate considerable differences existing
between stand 1 and 2. As before, the degree to which this difference
was a result of instrument error and site selection for the shelters is
not certain. With a more dense cover of vegetation and a more moist
soil it would seem reasonable to expect a more humid environment as in
the first stand.
Wind is an obvious and important factor affecting humidity and
transpiration rates of the various plants of a particular environment.
At the sites described here, it seemed that the wind played a major role
in affecting the nature of the community. Though no sampling or record
was kept of the wind velocity in either stand, it was observed on many
occasions that the valley was not well protected from stronq winds. Many
fallen logs were noticed throughout either site and it was not uncommon
to find trees fallen that had been observed standing the day before.
A result of this was the opening of large areas of the canopy and the
penetration of more light to the forest floor or the lower strata.
Frequently, yellow birch seedlings were observed growing on the older
windthrows and in the exposed mineral soil turned up by fallen trees.
It was only on such sites and decaying stumps, and stream beds free of
leaf litter, that birch seedlings were seen growing. At no time were
* seedlings observed to becoming established where leaf litter covered
the substrate.
One of the most important features of a forest site is the underlying
soil. Vegetation along with climate and th> geology of a site develop
the characteristics of the soil, but also it must be realized that the
soil has a role in determining the type of vegetation that will occur
at a given site. It is for this reason that an examination of the soil
characteristics in each of the stands was undertaken, Unfortunately,
a lack of equipment prevented a more sophisticated approach to the study
of soil conditions than would have been desired. Nevertheless, a summary
of data is presented in Table 16.
Table 14: Weekly summaries of
-185-
daily average, maximum and minimum
temperatures for stations in stand 1 and 2 from July 6, 1966 to
August 28, 1966.
Week Ending: 7/10 7/17
Station
1 61* 61
2
1 74* 75
2
1 48* 44
2
*Data incomplete
Table 15: Weekly summaries of
in stands 1 and 2 from July 6,
Week Ending: 7/10 7/17
Station
1 46* 42*
a 4ax
*Data incomplete
7/24 7/31 8/7 8/14 8/21 8/28
Average°F
58 61 62 63 62 58
61* 62 66* 63 59
Maximum? F
73 76 78 81 75 70
73* 79 82 76 70
Minimum?F
4] 44 51 44 47 48
42* 50 46 49 48
the mean minimum humidity for stations
1966 to August 28, 1966.
7/24 7/31 8/7) 8/14 8/21 8/28
Mean of Minimum Humidity %
50 52 55 64* 54 56
55 59* 49 49
-186~
Table 16: Summary of data collected for various soi] characteristics.
Soil Characteristic Stand_1 Stand 2
Soil Texture: greater than Imm. 6% 11%
.7mm.- Imm. 12% 20%
.25mm.~. 7mm. 29% 34%
12mm. -. 25mm. 32% 23%
less than .12mm. 21% 13%
Soil Classification: Sand Coarse Sand
Depth of Organic
Material: Litter 4cm. 2cm.
Fermentation 3cm. 1.5em,.
: Humus Scm. 6em.
Average of Soil Temperature
Sampling: 70°C 10°C
Soil Moisture: Average of Meter
Readings (milliamps ) 41ma. 18ma.
Oven Drying Method 49% 35%
Soil pH: 4.5 4.0-4.5
-187-
As might be expected, considering the. topography of stand 1, the
soil texture was much finer consisting mostly of fine sand and silt.
This soil was not deep, being located between and over the many boulders
located at the site, and having been sifted by running water that still
exists in some areas. The soil of stand 2 was more coarse in texture
not having been recently affected by water as has been the soil in stand 1.
There appeared to be more leaf debris in stand 1, but the depth of
the humus layer was about the same for each site.
The amount of soil moisture is not accurately represented by the
data for stand 1 at least. It was not uncommon to see saturated soils
in this stand and the 49% moisture given is indicative of surplus soil
moisture. The soil reaction (pH) was quite variable, by our crude method.
There may well be a correlation between amount of soil moisture
or amount of clay particles and pH of each site, but this remains unproven.
Soil .temperatures were taken on several occasions, but only an average
is shown in order to make preliminary comparisons. Soil temperature
fluctuates too greatly to be measured in absolute terms with anything
other than a continuous recorder. So far, it seems that temperatures at
the two sites did not vary importantly.
DISCUSSION
In terms of basal area yellow birch shows strong values in the total
cross section area of stems as well as fairly large individual trees in
the hardwood and the spruce types (Figure 3). The results of the two
site comparisons indicate that yellow birch has a more even diameter
distribution on the more moist site, which has a larger number of spruce
and red maple than the second site, an obvious type 25 stand. The even
distribution of trees in the different diameter classes reflects the
even age distribution of yellow birch suggesting vigorous growth, Stands
of the hemlock-birch type contain a number of very large and old trees.
It would seem that yellow birch is declining and will lose its prominence
‘in the hemlock-birch types.
A similar study made in Wisconsin (Winget et al., 1965) indicated
that yellow birch reached its highest importance values in association
with hemlock and with sugar maple. In our study it appears that yellow
birch is most effective in the spruce-yellow birch type, although it is
only slightly less effective in the sugar maple-beech yellow birch type.
In terms of density, yellow birch was more abundant in the spruce
type but with smaller trees. The total basal area per acre as noted
was about equivalent in the hardwood type. It would seem as if the yellow
birch in spruce-birch stands were younger and more vigorous in growth
judging from regeneration data and density considerations than in the
40
ol
fe}
% FREQUENCY
N)
°o
-98t-
saplings by forest type.
Y.B. SAPLING Y.B. SEEDLING
r % FREQUENCY % FREQUENCY
30 25 24 30 25
FOREST TYPE
Figure 4: Comparison of frequency (%) of yellow birch seedlings and
i
i
-189-
other types. It would also appear that it was exerting its maximum effect
on the stands of the spruce-birch type.
The site comparison data may aid in explaining further the greater
effectiveness of yellow birch in type 30 stands. Hoyle (1965) discusses
the fact that rootlet development of yellow birch is best in humus due
perhaps more to the availability of essential nutrients in this soil
layer than to availability of moisture, The data indicate a greater
depth of humus at the first site, which is more analogous to a type 30
stand. The high moisture content of the soils in the first stand is
however important in terms of birch effectiveness. Yelenosky (1961)
reported that yellow birch seeds will tolerate excessive moisture and
will even germinate in a free water medium. This certainly offers a
competitive advantage over other species in all habitats suitable for
yellow birch.
Figures. 2 and 3 provide only indirect information concerning the
successional status of yellow birch. Consequently we need to consider
data in Figure 4 which presents material on regeneration. Type 30,
except on moist sites, is probably a subclimax stage giving way
eventually to hardwoods. The larger size of yellow birch, decreased
density and regeneration in type 25, a climax type, support this view.
It appears that the common mode of yellow birch establishment, on
elevated micro-sites, may give it a competitive advantage. Tubbs (1963)
found that there is a more marked increase in germination and height growth
of this species on naturally occurring mounds than on adjacent flats.
Decaying stumps and windthrows are a frequent occurrence in spruce-yellow
birch stands. Also, decreased canopy cover in this type, especially
on moist sites, seems to be reflected in higher birch sapling frequency.
In the maple-birch type, yellow birch may be considered as a remnant
of the most immediate seral stage. However the species does manage to
persist for a very long period of time and will regenerate but not with
the vigor displayed in spruce-birch forests. The ultimate effect would
probably be complete elimination of yellow birch from the sugar maple-
beech forest if it were not for frequent minor disturbances favoring
“yellow birch regeneration, The greater range of importance values of
yellow birch in the hardwood types may be a reflection of the various
directions of succession leading to the type 25 climax.
ACKNOWLEDGMENTS.
We wish to express our special gratitude to Dr. Jon T. Scott for
his advice and guidance throughout this research project. Funds for the
project were from grants to Drs. Lemon and Scott from the Atmospheric
Science Research Center of the State University of New York at Albany»
and the National Science Foundation.
We are greatly indebted to Dr. Gary Holway, Raymond Falconer,
Paul Harney, James Houghton, Robert Meyer, and Stuart Nicholson for
their assistance in gathering data.
-190-
LITERATURE CITED
Braun, E. Lucy. 1950, Deciduous Forests of Horth America. The
Blakiston Co. Philadelphia, Pa. 596 pp.
Bruce, David. 1955. A new way to look at trees. Journal of Forestry
53:163-167.
Carter, Douglas B. 1966. Climate. Geography of New York State.
John H. Thompson, Ed. Syracuse Univ. Press. Syracuse, New York
pp. 54-78.
Craft, Jesse. 1966. Personnal communication. Atmospheric Science
Research Center. Whiteface Mt., N.Y.
Cressey, George B. 1966. Land Forms. Geography of New York State.
John H. Thompson, Ed. Syracuse Univ. Press. Syracuse, N.Y. pp. 19-53.
Curtis, J.T. and R.P. McIntosh. 1951. An upland forest continuum in
the prairie-forest border region of Wisconsin. Ecol. Monog. 6 :233-268.
De Laubenfels, David J. 1966. Soil. Geography of New York State.
John H. Thompson, Ed. Syracuse Univ. Press. Syracuse, N.Y. pp. 104-112,
Fernald, Merritt L. 1950. Gray's Manual of Botany, Eighth Edition.
American Book Company, N.Y.
France, Owen and Paul C. Lemon. 1963. Preliminary observations on forest
tree ecology of the Whiteface Mountain area. Atmospheric Science
Research Center, State Univ. of New York pub. #15. pp. 75-112.
Fraser, Donald A. 1956. Ecological Studies of forest trees at Chalk River,
Ontario, Canada. II, Ecological conditions and radial increment.
Ecology 37:777-789.
Gilbert, Adrian M. 1960, Silvical characteristics of yellow birch.
Northeastern Forest Expt. Sta. 134. pp. 1-18 (Burlington, Vermont).
Godman, Richard M. and Laurits W. Krefting. 1960. Factors important to
yellow birch establishment in upper Michigan. Ecology 41:18-28.
Hoyle, M.C. 1965, Variation in foliage composition and diameter growth
of yellow birch with season, soil and tree size. Soil Science
Society of America Proceedings. 29:475-480.
Jacobs, Rodney D. 1960. Top-dying of yellow birch, upper Michigan 1955-59.
Tech. Note No. 585. Lake States Forest Exp. Sta. Forest Service.
Jarvis, J.M. 1957. Cutting and seedbed preparation to regenerate yellow
birch, Haliburton County, Ontario, Can. Can. Dept. Nor. Aff. and
Nat. Res., For. Res. Div. Tech. Note 53. 17 pp.
Kujawski, Ronald F. and Paul C. Lemon. 1967. Stand Composition of yellow
birch sites as related to physical environment. Unpublished ASRC
research report.
Leak, William B. 1961. Yellow birch grows better in mixed-wood stands
than in northern hardwood old-growth stands. Northeast. For. Exp.
Sta., Forest Res. Note No. 122, U.S. Forest Service.
Lindsey, Alton A., James D. Barton Jr., and S.R. Miles. 1958. Field
efficiencies of forest sampling methods. Ecology 39:428-444,
McCarthy, E.F. and H.C. Belyea. 1920. Yellow birch and its relation to
the Adirondack Forest. N.Y.S. Coll. For. Tech. Pub. 12. 50 pp.
Scott, Jon T. and Stuart Nicholson. 1964, Some characteristics of the
vegetation of Whiteface Mt. and implications concerning their use
in studies of microclimate. Atmospheric Sciences Research Report
No. 33. 190-231.
~191-
Society of American Foresters. 1954. Forest Cover Types of North America.
Soc. of Amer. For., Washington, D.C. 67 pp.
Stearns, Forest. 1951. The composition of the suqar maple-hemlock-yellow
birch association in northern Wisconsin. Ecology 32:245-265.
Tubbs, C.H. 1963. Artifically constructed mounds show promise in yellow
birch regeneration. Lake States For. Exp. Sta. Res. Note LS-32.
Tubbs, Carl H. and Robert R. Oberg. 1966. Growth response of Seedling
yellow birch to humus-mineral soil mixtures. North Central Forest
Experiment Station, Forest Service. Research Note NC-6.
Winget, C.H., G. Cottam, and T.T. Kozlowski. 1965. Species association
and stand structure of yellow birch in Wisconsin. Forest Science 11:
369-383.
Winget, C.H. and T.T. Kozlowski. 1965, Yellow birch germination and
seedling growth. Forest Science, 11:386.
Yelenosky, George. 1961. Birch seeds will germinate under a water-light
treatment without prechilling. Forest Res. Note. Northeastern
Forest Expt. Sta. Note No. 124. 5 pp. (Laconia, N.H.).
APPENDIX
Scientific Name Colloquial Name
Abies balsamea Mill. Balsam fir
Acer pensylvanicum L. Striped maple
Acer rubrum L. Red maple
Sugar maple
Mountain maple
Yellow birch
Paper birch
Beech
Red spruce
Mountain ash
Hemlock
THE DETERMINATION OF
VERTICAL MICROMETEOROLOGICAL PROFILES
THROUGH A FOREST CANOPY
WITH A SINGLE SET OF SENSORS
James G. Droppo and Harry L. Hamilton
THE DETERMINATION OF VERTICAL
MICROMETEOROLOGICAL PROFILES
THROUGH A FOREST CANOPY
WITH A SINGLE SET OF SENSORS
By
James G. Droppo and Harry L. Hamilton
ABSTRACT
A vertically-sweeping sensor system for the determination of forest
microclimates has been designed, constructed, and successfully operated
jn a deciduous stand. The reproducibility of the results and their
consistency with other published results indicates that the goal of
developing a portable, low cost, forest microclimate sensing system
that is easy to operate has been achieved,
The energy balance model for the determination of diffusivities
and energy fluxes is discussed and applied to data obtained from the
system. The time rate or change of latent heat storage term, which
js usually not considered in previously reported forest studies, is
found to have the same order of magnitude as the ground heat flux term.
The calculated energy fluxes generally agree with other published
results; however, on several occasions, the eneray balance model fails
to give physically meaningful results in the lower region of the canopy.
The general applicability of the model to forests is discussed in light
of the empirical results.
INTRODUCTION
The microclimate of a forested region is a major factor in
determining the plant associations that will occur in the region.
While forest microclimates in various locations have been the subject
of many studies (Geiger, 1959; Munn, 1966), the interaction between
the microclimate and the vegetation in a region is a complex process
that has been neither understood well nor widely studied.
The microclimate of a forest depends on the energy exchanges that
occur within the canopy. Under conditions of full foliage, the distri-
butions of temperature, moisture, and carbon dioxide are determined
primarily by the magnitudes of the net radiation energy and the turbulent
fluxes of sensible and latent heat. The energy used in the physiological
~193-
-194-
processes of the vegetation is of secondary importance. Under good
growing conditions the energy used in photosynthesis is about five
percent of the solar radiation (Lemon, 1960). The ground surface is
another surface of active exchange, especially when the leaves are off
of the trees.
The micrometeorologist studying forest energy balances has at
least two options as to the approach to use. One method is to use
profiles of temperature and moisture to calculate a coefficient of
turbulent diffusion, a parameter used in the determination of the
energy balance. Another is to derive the coefficient on the basis of
wind profiles. The variety of micrometeorological studies that have
been conducted in plant communities are described in books by Sutton
(1953), Priestley (1959), and Munn (1966). Munn suggests that the
former method is the better approach, when he states that "... an
assumption about the ratio Kp/ky is probably preferable to one about
the value of K,/Ky...", where Ky» Kyo and K, are the diffusivities of
sensible heat, hatent heat, and momentum , respectively. He cites the
fact that the determination of a coefficient from wind profiles requires
a value for the friction velocity, which cannot reliably be obtained on
a regular basis using present techniques. In addition, a conceptual
problem arises, since the friction velocity would be a function of
position within the canopy.
Denmead (1964) has calculated the magnitudes of the major terms
of the energy budget using temperature and moisture measurements, He
obtained profiles of temperature, humidity, and net radiation from
sensors at six levels on a vertical tower located in a low pine forest
in Australia. The energy exchanges between each layer were calculated
from these profiles.
For the study reported here it was considered desirable to deter-
mine forest energy balances by this temperature-moisture profile method.
Since future work would call for the determination of the energy balance
in a number of forest stands, the system to be used was required to be
reliable, portable, and relatively inexpensive. The study included
_ the design, construction, and testing of a measuring and recording system,
and the use of the system in a small deciduous forest.
EQUATIONS
The equation for the energy balance of a volume of air located
between the ground surface and a horizontal plane at height z is
R(z) = LE(z) + H(z) + &-(z) + G,(z) + 6, + Bz), (1)
where R(z) is the net radiative flux through the upper surface
-195-
(positive when downward), LE(z) is the latent heat flux through the
upper surface (positive when upward), H(z) is the sensible heat flux
through the upper surface (positive when upward), Ge(z) is the rate
of thermal energy storage in the foliage below height z, Ga(z) is the
rate of energy storage in the atmosphere below height a, G, is the
flux of heat into the ground through the earth's surface, 3nd B(z)
represents the flux of other energy sources and sinks within the
volume. All energy storage terms are defined as positive when storage
occurs. :
The assumption is made that there is horizontal uniformity of
all energy sources and sinks, such that there is no net horizontal
transfer of energy. Thus, only vertical fluxes through a horizontal
unit area are considered. The terms are shown schematically in
igure 1.
The rate of energy storage in the forest air may be considered
to be composed of two components: sensible heat storage, Ga}, and
latent heat storage, G.,. For a volume of forest with unit area and
vertical height z, thead terms are expressed as
Gyre eptz (2)
and
Gay =e baz, (3)
where p is air density, c, is specific heat of air at constant pressure,
L is latent heat of vapor? zation for water, T is the time rate of change
of air temperature, and q is the time rate of change of specific
humidity. .
The rate of energy storage in the foliage is given by a similar
expression:
Ge = hog Cpr Tyz (4)
where h is the areal foliage density, p¢ is foliage density, c, is
foliage specific heat, and Te is the time rate of change of foliage
temperature.
Energy flow at the forest floor may be calculated from temperature
gradient measurements, but in this study direct readings were obtained
from a commercial soil heat flux instrument. Other sources and sinks
to be considered in a forest are suspended particulates, animals, and
-196-
photosynthesis in the vegetation.
For convenience, in the following development all storage terms
will be considered as one cumulative term given by
G(z) = Ga(z) + Gy (z) + Ge + G + B(z). (5)
The flux of sensible heat in air, H(z), is given by
=. aT
H(z) = -o cy K,(z) Ge +r), (6)
where kK, (2) is the diffusivity for sensible heat at height z, -
is the environmental gradient of temperature, and r is the dry
adiabatic lapse rate. The flux density of latent heat is similarly
given by
=~ 3g.
LE(z) = -p L Ky(z) 22, (7)
where *K, (z) is the diffusivity for water vapor at height z and a
is the vertical gradient of water vapor. Substituting G(z), H(z),
and LE(z) from Equations (5), (6), and (7) into the energy balance
Equation (1) gives
R(z) = -9 ey Kylz) GE +r) - 0 L Ky(z) 2+ Giz). (8)
For actual calculations this equation must be put into finite-
difference form. The forest floor is considered to be the lower
surface of a volume under consideration. The upper surface is con-
sidered to be the middle of a layer of finite thickness, Az, as shown
in Figure 1. This layer is bounded by two horizontal planes at
heights z, and z The vertical change in temperature and specific
humidity betweeh these levels is aT and Aq, respectively. Equation (8)
becomes
(Ry + Ry)/2 = R(z) = -cy KA(z) (AL + rn) ~ 0 L Ky(z) 42+ G(z).
(9)
-197-
Figure 1: Energy Budget Model. This represents a vertical
cross section through a forest as shown on the
1° left. The energy fluxes are represented by arrows
which point in the direction defined as a positive
flux. The brackets enclose the height through
which the storage terms refer (page 198).
Figure 2: Tower with Sensor Package Drive System (page 199).
~198-
LE H
A /\
TOF
\.
} ze £ T
Canopy 7, t [RZ
‘ ( )
| Nad St
q Le ony Gon } Vv Ry
Trunk space ) | Sai
{ Ui »
t 7 “VY
Earth I Gg
-199-
+ :
N Wa eA pulley
N \ WY
4
N o——————- top line tab
y
es 4 4 a
To r Y J+—_——— level stop tab
i guide line
IS
2 a
K E yer =S
—— $3,S8
NS x) ———diive ‘motor
t An —F |
V\\ Aa VAVNA
scale + =('
~200-
Adopting the assumption, also used by Denmead, that the diffusivities
for water vapor and sensible heat are equal, and solving for the "apparent"
diffusivity gives for each level, then
. R(z) - G
K(z) = ant Yr a EUR (10)
This layer analysis was used in the present study. In summary,
these formuli are based on the following assumptions: (1) steady state
conditions prevail, (2) the diffusivities of water vapor and sensible
heat are equal, (3) all energy terms not explicitly used in the final
equations are unimportant in the analysis, (4) the vertical distribution
of energy sources and sinks is horizontally uniform. The first assumption
is best approximated by taking data near the meteorological noontime.
The assumption that the diffusivities of sensible and latent heat are
equal has considerable support in the literature, although there is
discussion on this point. The assumption seems to hold for stable
conditions, Under stable conditions there is evidence that Ky >» K
(Priestley, 1959), but the difference, if real, appears to be’ not
enough to significantly effect the accuracy desired in the present
study. The last two assumptions have not yet been verified for forests,
but previous studies suggest that they do hold for relatively uniform
forests.
Vv
This heat balance approach, although perhaps limited by the above
assumptions, provides a consistent model for diffusion in the present
study.
METHODS
Instrumentation
A single sensor package was constructed to obtain readings at all
“levels in the forest. Better relative accuracy between levels is one
of the several advantages of using one sensor rather than many, as has
been discussed by Hamilton (1964). The instrumentation used in previous
studies of microclimate is described by Long (1957). With the develop-
ment of smal] net-radiometers, it has recently become possible to include
net-radiation profiles in forest studies (Funk, 1959, 1962). The sensor
package was mounted on a vertical pulley system which was attached to
the tower as shown in Figure 2. The top of the tower was about three
meters above the canopy top in a wooded area near the campus of the
State University of New York at Albany.
Two methods were used for data sampling: (1) data recorded every
eight seconds as the package moved vertically at the rate of about
~201-
eight centimeters per second, and (2) data recorded at discrete levels
with the sensor package stopped at each level. Manually controlled
switches were used to obtain data in the first method. An electrical
switching circuit was constructed to control automatically the drive-
motor and recording cycle in the latter method. The switching system
is shown schematically in Figure 3.
Except for at the highest and lowest positions, the package was
stopped for readings by the level-control microswitch, S,, when
triggered by the level-control stop-tabs, illustrated in Figure 2.
The highest and lowest positions of the package were set by the top and
bottom stop-tabs as the stop-tabs hit microswitches S, and S,, ;
respectively. These switches, besides stopping the package, ‘also,
switched the motor wiring circuit such that the package would next move
i jin the opposite direction. When the sensor package stopped at each
level, readings from all sensors were recorded. When the recording
cycle was complete, the sensor package moved to the next level and the
process was repeated. The use of remote, automatically controlled
switches not only assured that the package always stopped at exactly
: the same level, but also allowed these levels to be quickly and easily
selected in the field. The length of time the sensor package remained
at each level could be varied by changing the timing cams.
: The circuit for recording is shown in Figure 3. The recording
| cycle began with the recording of the parameter connected to the first
position of a stepping relay, which in the present case was a zero
reference. The relay was then advanced to step two by a cam (not shown) ,
where the second parameter was recorded. This process was repeated
until the last variable was recorded, when the stepping relay automatically
reset. The number of parameters recorded could easily be varied.
Figure 4 is a photograph of the sensor package before an outer
aluminized plastic shield was attached. This package contained the
temperature and humidity sensors, which were ventilated at four meters
per second by an electric fan. The temperature and the wet-bulb depression
were each measured with five copper-constantan thermocouples wired in
_ series. Also shown in this picture are the thermocouple net radiometer
and its mount. A commercially available version of Funk's net radiometer
was used. A commercial soil heat flux plate was used to measure the
ground heat storage. A thermistor circuit was used to measure the
- temperature of the reference bath. Signals from each of the above
were amplified and then recorded by a digital voltmeter-printer system.
When the sensor package stopped at specific levels the profile parameters
were recorded in the order of their increasing lag times. There was
a lag-error in the data taken by the moving package which will be dis-
cussed later. An event-recorder was used to monitor the motion and
position of the sensor package.
i
I
i
|
Figure 3:
Figure 4:
~202- »
Flow chart of the control circuit. The actual
circuit used was a much more complex combination
of switches, cams and relays (page 203, top).
Sensor Arrangement and Recording Circuit. The
dotted boxes indicate the location at which the
various parameters were measured. t is temperature,
Ry net radiation, G heat flux into the ground,
and T,, wet bulb temperature (page 203, bottom).
~203-
Ss
MOTOR OFF
Dp e Oren S. 7 reverse direction
(package moving) Si DRIVE MOTOR OFF.
REMOTE
SWITCHES
[COMPLETION OF| STEPPING RELAY | RECORDING
ee RESETS CYCLE ON
Stepping Relay
poco tme roo
§ Sensor Package {Reference Bath @
} Ty 1 Te
1 Ra I I
I i t
' Ty ‘
! I H t
(Cat |i
r------4 = i
‘ Soil Surface 4 2 1
H c L} q
pepe
CAM SWITCH
-204-
Figure 5. Sensor Package (without outer radiation shield).
-205-
Procedures
The cost of a data acquisition system has been greatly reduced by
the use of a single set of sensors. Also, since moveable sensors yield
more accurate vertical gradients than multiple fixed-sensors, the latter
would have had to be more accurate in order to obtain comparable gradient
results. The relative simplicity of the instrumentation made installation
simple compared to mounting sensors at many levels.
The diffusivity for each level was determined from Equation (10).
The sensible flux was obtained from Equation (6). The gradients were
determined by taking differences across the finite layer, 2) - Zo. The
Jatent heat flux was then obtained from Equation (1).
Heat flux into the ground was obtained directly from the recorded
heat flux plate. The other storage terms were found by use of a graphical
i technique. For a given time-period, the initial and final profiles of
i temperature and water vapor pressure were plotted and then the graphical
areas within the various layers were determined. These areas, in units
| of °C-feet and mb-feet, were then substituted into special forms of
1 Equations (2) and (3) to give values of sensible heat and latent heat
[ storage as a function of height. These values of °C-feet were used as
an approximation to the change in temperature of the foliage in Equation (4).
| RESULTS
Profiles
Profiles of various parameters were examined and found to be
physically reasonable based upon previously published results, Examples
of detailed profiles are given in Figures 6, 7, and 8 for temperature,
net-radiation and water vapor, respectively. These values were recorded
at intervals of two feet without stopping the package as it moved
' vertically through the trees. The data were taken in late afternoon
after an entire day of clear and calm conditions so that significant
-trends should occur. The profiles show the expected trends. For
| example, in Figure 6, the temperature falls about 0.7°C between the two
i runs twenty-four minutes apart. In Figure 7, the net-radiation also
i show$ a significant decrease between the two sets of profiles. The
slope of the individual profiles also were in agreement with expected
results.
The results of the individual sweeps of the sensor package as
shown in Figures 6 through 8 show large vertical fluctuations which
conceivably would cause serious difficulties for the computation of
heat balance. However, when the values for several sweeps are averaged
considerable smoothing results as shown in Figure 9 where 21 sweeps are
averaged. During the period covered in Figure 9 the individual profiles
exhibited a great deal of vertical and temporal variations. The winds
|
i
i
'
-206-
Height,
feet 65
Recording Times: ~
(1) 14:31 E.S.T.
(1) 14:39 EST.
~ (2) 14:55 EST.
15:03 E.S.T.
fe) | | | | |
17.0 18.0 19.0 20.0 21.0 22.0
Air Temperature °C
Fig. 6 Temperature Profiles ,. 5/25/67
Height,
feet 65
~207-
Recording Times:
(1) 1415 E.S,T.
14:23 E.S.T.
(2)
| |
0.0 0.1 0.2 0.3 04
Net Radiation, ly /min.
Fig. 7 Net Radiation Profiles, 5/25/67.
~208-
Height,
feet 65
| ‘| i | |
15.50 1600 1650 17.00 17.50
Pressure, mb
Fig.8 Water Vapor Pressure Profile, 14:23.-14:31 EST,
5/25/67.
-209-
Height,
feet
| |
20.0 21.0
Temperature °C
Fig.9 Mean Air Temperature Profile, (one and a half hour
period ),
~210-
were variable, up to about seven miles per hour maximum at ten feet.
The sky was about one-quarter covered with cumulus. Under such variable
conditions the smoothness of this mean profile gives strong support to
the use of representative mean gradients between five or six levels.
The average vertical profiles of temperature, water vapor pressure,
and net radiation for two dates are plotted in Figures 10 and 11.
These profiles indicate reasonable physical processes based on the pre-
vailing meteorological conditions, which are summarized in Table 1,
The profiles in Figure 10 clearly show a transition of conditions which
agree with the observed meteorological changes. There were no leaves
on the trees on this date. The first sampling period was characterized
by clear and calm conditions, which had existed all morning. By the
third period the wind had risen and the sky had become overcast. The
recorded transition to lower and more uniform values of downward net
radiation, temperature, and moisture are those which would be expected
as a result of the rising winds and decrease of incoming energy. In
Figure 11 are presented profiles for a situation with leaves on the
trees. During the period 9:30 to 12:30 EST there was a change in the
profiles typical of warm spring mornings. The middle profile shows
the continuity of change between the first and third periods. This day
was clear with light winds. The fact that the dominant warming and
evaporation occurred both in the upper canopy and at the earth's surface
is evident from the profiles of temperature and vapor pressure in
Figure 11. The net radiation profiles on this date show a peculiar
maximum at the forty-five foot level. This maximum in the canopy region ©
was also evident in the individual profile values. The effect was even
more pronounced in profiles taken in dune. This unexpected variation
was also observed in the profiles taken every few feet under similar
conditions of strong incoming solar radiation and the foliage on the
trees. In the latter case, the net radiation increased as the sensor
package came down into the upper part of the canopy and was observed
to reach a peak value at the level of one particular tree branch near
the top of the canopy. The radiation then decreased below this level
as should be expected because of shading. This maximum may be the
result of increased downward radiation from leaves above the radiometer
_or the result of decreased upward radiation from below the sensor, or
a’ combination of these two effects.
The most consistent of the results included in Figures 6 through
Wi were those of the temperature and moisture profiles, while the net
radiation profiles were less consistent. This may be attributed to
horizontal variations in the net radiation flux. Fortunately, the
calculation of the heat budget requires the mean radiation through a
layer rather than a gradient. This means that vertical variations in
the net radiation will have only a small effect on the determination
of energy fluxes of sensible and latent heat.
|) ———12:00-13:00 ELST.
+--=- [3:00-14:00 E.ST.
-—- 14:00-15:00 E.ST.
Net Radiation, ly/min.
: . +1
ft.
TO
co- | ji
top of ee ee, ! io Ff _t I
canoPysot fl i
aol i; ii \
i Vi \ &
30-- j : VY; \ r
i } ry \
209- jj Mt \
i 4 ty }
10-— j % 1 f
f l l ‘
15.0 16.0 17.0
Air Temperature,°C
13.0 14.0 15.0
Water Vapor Pressure,mb.
Fig. 10 Vertical Profiles through the Canopy at the Mohawk Campus, 4/9 /67.
— 9:30 -1hOOEST.
77+ 10215 -1h45 EST.
“> 11:00 -12°30 E.sT. xe :
‘ ESh Net Radiation, ly /min.
oO +05 ~~ +10
Height, | |
~ 70K
a 4
topof OV AN ft A _ ee cece ee
canopy | - aN wv \
SON \
a | H if &
is ij =
30;- if 4
if
20/- i
i \
MOON , .
10K \ %, fi \,
e) | | |
160 170 18.0 13.0 14.0 150 16.0
Air Temperature (°C) - Water Vapor Pressure (mb)
Fig. 1] Mean Vertical Profiles through the Canopy at the Mohawk Campus, 5/23/67.
Table 1: Meteorological conditions during sampling periods.
Date Time Mean Air Sky Mean Wind Comments
Interval Temperature Cover Speed at 10'
EST at 8°
4/9/67 12:00- 15°C Clear 5 mph Forest floor covered with fairly
13:00 dry dead leaves, last precipitation
was light snow two days ago, no
leaves on trees, no precipitation
13:00- 16°C Variable, 7 mph during sampling period.
14:00 sky hazy
1
i)
@
n
14:00- 17°C Overcast 9 mph
15:00 thick stratus
5/23/67 9:30- 17°C Clear 4.2 mph Forest floor flora and trees are
11:00 foliaged.
11:00- 18°C Clear 6.7 mph
13:30
-214-
5i- —
oll J | Jy
-02 0 02 04 06 O8
Energy Flux, ly min7!,
Fig. 12 Energy Flux Profiles, lO:I5- 11:45 EST,
5/23/67.
-215-
Energy Balance Calculations
The energy balance has been completely computed for the conditions
of May 23, 1967. The results agree with those of Denmead (1965) with
regard to the vertical distribution of sources and sinks of energy in
the forest.
The terms of the enerqy budget as measured and calculated are
summarized in Table 2 and plotted in Figure 12. The flux of latent
heat is greater than the sensible heat flux everywhere except at the
lower layer. The latent heat flux profile shows the canopy to be the
major source of water vapor on this date. Both the canopy and the
surface of the earth are sources of sensible heat flux. An encouraging
aspect of these calculations is that the net vertical flux of energy
between any two horizontal levels is very close to zero.
Some storage terms which Denmead (1965) neglected were found to
be significant for the conditions in the present study. Figure 12 shows
the vertical variation of the three storage terms as a function of
height. All three terms have similar profiles. The qreatest storage
per unit vertical height occurs near the earth's surface. A secondary
maximum occurs in the lower portion of the canopy and the storage is
less in the upper parts of the canopy where the maximum attenuation of
incoming radiation is occurring. This may be a result of the effect
of turbulence being greater in the upper canopy,
The values of the fluxes at the twenty-foot levels are uncertain.
The actual gradient of. temperature and water vapor pressure is positive
at this level indicating that there should be downward fluxes. Sub-
stitution of the data for this level into Equation (10) results ina
negative diffusivity, which when used in the flux Equations (6, 7) gives
a net upward (positive) flux. The physical significance of a negative
diffusivity is that there is an active diffusion against the normal
gradient diffusion. The values in Figure 11 at the twenty-foot level
do not seem to represent real physical energy fluxes. Assuming that
the profiles are correct, then the storage term, G(z), would have to
be larger to obtain a positive diffusivity. This means that at least
0.33 ly/min must be accounted for. Since this is inconceivable con-
sidering the order of magnitude of the calculated storage terms, there
must be an energy sink such as horizontal advection which is not being
considered in the present model. Another possibility is that penetrative
convection occurred in this layer as a result of the warmer surface layer
below, resulting in the breakdown of the energy balance model, which is
based on gradient transport mechanisms. Preliminary calculations of
energy budgets have been done on other occasions and this failure of
the model occurs several times in the same lower layer of the canopy.
DISCUSSION
The equipment designed and constructed in this study has fulfilled
Height,
feet f
80/;- ---—Latent heat
Sensible heat
sess Foliage
601-
40
\
20+ wu
oL_l_ TTP I
0.0 1.0 2.0 3.0 4.0
Storage, ly / min./ft. (x 104)
Fig.13 Atmospheric and Foliage energy storage per unit height, 10:15 -
145 E.S.T., 5/23/67.
“912-
Table 2:
Height Diffusivity Sensible
Feet 2 Heat Flux
cm /sec ly/min
55 20,900 0.199
35 18,400 -0.011
20 -8,320 0.032
5,320 0.182
7.5
Terms of the energy budget, 5/23/67.
Latent
Heat Flux
ly/min
0.358
0.532
0.304
0.064
Net Radiation
ly/min
0.612
0.567
0.376
0.248
Storage Rate
ly/min
0.054
~Lle-
0.046
0.041
0.038
~218-
original expectations. The installation and maintenance of a single
package of sensors was easier and safer than those for sensors at fixed
Jevels on the tower in that all work was done at the base of the tower
after the upper pulley had been installed. The use of the remote
microswitches proved to be an accurate and reliable method of controlling
the sequencing of data levels. The economic advantage of the equipment
is shown by the fact that the cost of constructing the instrumentation
and control system without the recording instruments was about $1,500,
including labor. This is about the same as the cost of net-radiometers
alone for a five-level fixed-sensor system.
The only limitation to recording extended periods of data in the
present work was the size of fuel supply on the portable electric
generator, Analysis of data revealed interesting changes in gradients
during the averaging periods. An increase in the speed of movement of
the sensor package to about 15 cm/sec is recommended in order to allow
use of shorter averaging periods.
A comparison of profiles with those from a fixed-level system is
not yet available. Five hours of temperature data from clear and fairly
calm conditions were studied for the temporal variation of gradients.
The results under these conditions were encouraging because: (1) the
gradients between levels were clearly defined; (2) the gradients usually
showed reversal either as the result of general trends or when conditions
were nearly isothermal during the entire period; and (3) the unequal
spacing of data in time for the different levels did not appear to have
any effect on the results. The agreement between the calculated profiles
and observed conditions, gives strong support to the use of a single
sensor package system in forest studies.
In general, the energy balance calculations were reasonable for
the observed conditions and were comparable to the results of Denmead (1965).
The reason for the failure of the model in the lower canopy is not under-
stood, although it is strongly suspected to be the result of horizontal
variation of net radiation. This points out the need for extending
the model to account for horizontal variations as has been suggested by
many authors. The present equipment could easily be used for measuring
horizontal variations. The fact that the method of recording would allow
the use of a single set of recording instruments from several sensor
packages is an added advantage. Hence, the instrumentation developed
promises to be ideal for future research on the horizontal as well as
the vertical processes in forests.
CONCLUSION
It is concluded that the single sensor package system of data
acquisition in forests is economically and operationally superior to
a fixed level sensor system for determination of several days of data.
~219-
Although no absolute check on the accuracies of the computed means
are available, the physical consistency of the results can be taken as
an indication that the system is giving accurate results. The comparison
of results with a standard tower will be necessary before a statement
can be made regarding the potential of the system.
REFERENCES
Denmead, D.T. (1964) Evaporation sources and apparent diffusivities in
a forest canopy. J. Appl. Met. 3, pp. 383-389.
| Funk, J.P. (1959) Improved polythene-shielded net radiometer. J. Sc.
| Instr., 36, pp. 267-270.
Funk, J.P. (1962) A net radiometer desiqned for optimum sensitivity and
a ribbon thermopile used in a miniaturized version. J. Geophys. Res.,
67, pp. 2753-2760.
Geiger, R. (1959) "The Climate Near the Ground," 494 pp. (English
translation from German by Harvard Univ. Press, Cambridqe, Mass.)
Hamilton, H.L. Jr. (1964) The use of a sweeping boom mechanism in a
studies of low level radiation flux divergence.
Lemon, E.R. (1960) Photosynthesis under field conditions II. An
aerodynamic method for determining the turbulent carbon dioxide
exchange between the atmosphere and a corn field. Agron. J. 52,
' pp. 697-703.
Long, I.F. (1957) Instruments for micro-meteorology, Quart. J.R. Met.
Soc., 83, pp. 202-214.
Munn, R.E. (1966) "Descriptive Micrometeorology," 245 p. Acad. Press,
New York, New York.
Philip, J.R. (1964) Sources and transfer Processes in the air layers
occupied by vegetation. J. Appl. Met., 3, pp. 390-395.
Priestley, C.H.B. (1959) "Turbulent Transfer in the Lower Atmosphere,"
130 p. Univ. of Chicago Press, Chicago, Illinois.
Sutton, 0.G. (1953) "Micrometeorology," 333 p., McGraw-Hill, New York.
ACKNOWLEDGEMENTS
This research was supported by a grant from the National Science
. Foundation (No. GB-7020) and by the Atmospheric Sciences Research Center
I of State University of New York at Albany.
THE PREVAILING WINDS ON WHITEFACE MOUNTAIN
AS INDICATED BY FLAG TREES
Edmond W. Holroyd, IIT
The Prevailing Winds on Whiteface Mountain
as Indicated by Flag Trees
By
Edmond W. Holroyd, III
ABSTRACT
The flag trees on Whiteface Mountain in New York's Adirondack Mountains
were studied to determine the direction of the prevailing winds. Using both
the direction of branch growth and the position of reaction wood in the
trunk tops a very complex wind pattern was found. Wind instruments placed
at various locations on the mountain during summer recorded the same pre-
vailing winds as indicated by the trees.
INTRODUCTION
A flag tree or banner tree, as it is sometimes called, is a tree with
a conspicuously asymmetric crown, having the longest branches on the opposite
side if the trunk from the shortest. The asymmetric shape may be related
to the wind direction through a number of possible mechanisms. Trees exposed
to severe winds near the summit of a high mountain may lack branches on their
windward side due to the killing of the buds by desiccation and by abrasion
of ice particles, as discussed by Daubenmire (1943, 1947). During a particular
storm branches on the windward side of the tree may be removed by breakage
due to an excessive accumulation of glaze (Lawrence, 1939). A fourth method
of flagging is by a constant, strong wind pressure from one direction during
the growing season; this results in the bending of nearly all branches,
especially those on the windward side, toward the leeward side of the tree
(Warming, 1909, Lawrence, 1939, and Daubenmire, 1947). These mechanisms have
their effect during different seasons: branch bending is an expression of
winds during the summer months; abrasion by ice particles and sand occurs
when these are present and airborne; desiccation of buds acts during the
winter months; breakage by glaze and rime results from the winds during the
seasons when glazing occurs or from particular storms which produce the damage.
The trunk of the tree may be influenced by the prevailing winds during
the growing season. There is an asymmetric growth of the annual rings in
conifers, with most growth on the leeward side due to the formation of
reaction wood (Daubenmire 1947, and Sinnott 1952).
Phototropism frequently produces asymmetries in trees, but the directions
in which the branches point are random. A tree flagged by the wind can be
distinguished from these because they point in the same direction as all the
other flag trees in the neighborhood. Only wind-flagged trees were considered
in the work which follows.
~221-
~222-
PROCEDURE
The trees on Whiteface Mountain were examined to determine the causes
of flagging. There was rarely evidence of breakage, even on trees near the
peak. This indicates that accumulation of glaze and rime ice has only a
small role, if any. A few dwarfed trees near the summit were cut about two
feet from their tops to determine the direction of the summertime prevailing
winds from the position of reaction wood. This was found to coincide with
the direction of flagging, produced by summertime branch bending and winter-
time desiccation of the buds. Since desiccation has the greater effect on
these trees, this indicates that the directions of the summer and winter
prevailing winds near the summit are approximately the same. Trees lower
on the mountain were flagged by branch bending only and so record the
direction of the prevailing winds during the summer.
During the summer months of 1963 and 1965, records were made of the
direction of flagging of the trees on Whiteface Mountain in the northeastern
Adirondack Mountains of New York. The directions were measured by aligning
the edge of a Silva magnetic compass with the flagging direction. For each
tree true direction rounded to the nearest ten deqrees was recorded along
with species, an identifying number, and an indication of the point from
which the tree was observed. Thé last identification was required since most
trees were examined from a distance. The flag trees in the valleys were most
easily seen from a nearby ridge, and often such a vantage point was required.
Whenever the trees were observed from distances greater than about one hundred
yards, binoculars were used to determine direction of flagging. It is believed
that the error in distant observations is within +10° for the first half
mile, #20° for the next half mile, and +30° from there up to two miles. The
direction of flagging was often checked in conifers by observing the position
in the trunk of reaction wood, which is on the leeward side of the tree.
The locations of the trees and the observation points were recorded on
enlarged topographic maps.
Different species growing together flag different amounts. The order
of sensitivity on Whiteface Mountain is white pine, red pine, red spruce,
balsam fir; other conifers, scotch pine, hemlock, and white cedar, which were
encountered less frequently, show good flagging, but their relative degrees
of flagging could not be determined. The deciduous trees do not show good
flagging, but of this group the birches and aspens are best. The conifers
were preferred as indicators; hardwoods were used only if conifers were absent
ane when the trees could be examined from a distance of less than one hundred
eet.
The prevailing winds at a few locations were measured for a comparison
with the flagging directions. Several wind instruments have been set out
by the Atmospheric Sciences Research Center during the summer months for
various purposes. The months and duration of these observations are shown
in Table 1. Some instruments were well exposed to the winds; others were
not and indicate very light winds. The data from these instruments were
analyzed to determine the average velocity from each direction and the vector
~223-
Table 1: Times and durations of wind measurements on Whiteface Mountain.
The wind data are plotted in Figure 6.
Total
Station Operation Operation
: No. Dates Hours.
| 1 July 1966 365
| 2 July, August 1965 1214
i 3 July 1965 239
4 September 1962 1218
I 5 August 1965 308
6 July, August 1964 23
7 duly 1965 422
~224~
average of all winds. The vector average of the winds was determined by
reducing the winds to north and east components, adding the components for
all winds, dividing by the number of observations comprising the sum, and
converting the average components back to direction and velocity. This
computation gives the averaged resultant of the flow of air. Its direction
is most important in this study. The velocity of the vector average indicates
its significance. A low velocity would result from light winds or winds
having a highly variable direction.
RESULTS
Figures 1-5 are maps showing the direction of the prevailing winds as
indicated by the flag trees. Fig 1 shows the indicated wind patterns
around the entire mountain. Figur... 2-5 are enlargements of sections of
Figure 1 and show the details of the wind patterns, Short lines indicate
the directions shown by the flag trees at nearly two thousand points on the
mountain. Each point may represent up to a hundred trees, especially near
the summit, or only one tree in regions where conifers are scarce, Long arrows
are interpolated to show the general flow of air. The length and spacing
of the arrows have no significance in any of the figures. Segmented arrows
show the direction of the air flow as it passes into the upper air at ridges.
Areas of no flagging can be assumed to be areas of light or turbulent winds
or even calm. They are generally in places sheltered from the prevailing
-winds by nearby ridges or cliffs.
The polar coordinate diagrams of Figure 6 show average velocity of winds
at the several points studied, together with their vector average. Only
those directions having a frequency of occurrence of five percent or greater
have been plotted. During the ten day period of operation of the wind
station north of the peak of Whiteface Mountain there was an abnormally
large amount of east and southeast wind. This has considerably distorted
the wind rose for this station. Although these are summertime data and of
short duration, comparison with adjacent flag tree directions shows that the
direction of the vector average is nearly the same as that of the flagging
of nearby trees.
While the observations of the flagging directions were being collected,
it was noticed that usually the wind actually blew in the direction indicated
by the trees, particularly along the ridges. The eastern valleys were the
only places where the winds were flowing in the opposite direction when the
observations were made. This reverse flow of air was due to the heating of
the eastern slopes of the mountain by the morning sun, causing up-valley
breezes. These winds sometimes reached velocities in excess of twenty knots,
but they were mostly in the range of five to ten knots. During the night
in these same valleys there were strong drainage winds of a much higher
velocity and parallelling that indicated by the flag trees. In these localities
the night drainage winds apparently had more of an effect on the trees than
the daytime up-valley winds.
¥4 direction of wind:
surface ———>
aloft -—-—>
wind shift front:
area of no flagging: «77>
Scale t— Yamin |
Figure 1: The wind patterns on Whiteface Mountain as indicated by flag trees.
~226-
As shown in Figures 1-5, ridges are often the positions of wind-shift
fronts. Streams of air from the two sides of a ridge meet at the top at an
angle or sometimes head-on. In such cases, one wind is usually stronger than
the other. By observing cloud fragments in these streams of air, the flow
pattern can be determined in such regions, Along the northeast ridge of
Whiteface Mountain above 4400 feet the air coming from the northwest rides
over the air coming from the south and both go into the upper air. This
pattern is repeated on Esther Mountain and at many other places in the
Whiteface area.
DISCUSSION
The eastern ridges and valleys shown in Figure 3 have a complex flow
pattern. The general wind direction for that region is from the southwest,
as indicated by the winds in the Ausable River Valley below and on the tops
of the ridges. However, there are drainage winds in the bottoms of the
mountain valleys whose heads meet the northwest winds of the plateau near
Lookout Mountain. The air flowing over the mountain travels down these
valleys beneath the air coming around the mountain from the southeast.
Figure 7 shows the vertical cross section from Little Whiteface Mountain to
Marble Mountain of the air and ridges to illustrate this pattern.
The White Brook Valley east of Esther Mountain (Figure 2) is the location
of very strong night, drainage winds, averaging more than fifteen knots with
gusts often reaching forty to fifty knots. These winds were instrumental
in the closing of the ski slopes on the north side of Marble Mountain since
they blew the snow off the trails and into the woods shortly after it fell.
The steep sides of this glacial valley channel the winds into one large stream
of high velocity air. Some air flows up the west side of the northwest ridge
of Esther Mountain. The rest goes around the ridge and up the east side and
forms an opposing wind at the top. The westerly winds are stronger on that
ridge.
Little data could be obtained from the northwest side of Whiteface
Mountain below the toll road (Figures 1 and 4), a region of few tall conifers,
due to logging and fires about a half century ago.
The large area on the south side of Whiteface Mountain in the Whiteface
Brook Valley is one of no detectable flagging, indicating a region of calm
or light winds. Neither are there any of the blowdown areas which are so
frequent elsewhere on the mountain. The lack of strong winds in this region
may be due to a blockage of the winds by mountains on the northwest side of
Lake Placid, but it is more likely caused by a damming of the air by the
south ridge of Whiteface Mountain. The south ridge, when seen on a relief
model, seems to have a shape which could produce and hold a pool of relatively
stagnant air in the Whiteface Brook Valley over which the prevailing wind
rides. It was observed that there are trees on the south ridge which were
flagged by a west wind, but due to poor weather and a lack of time, this
ridge has not yet been examined in detail. The cause of the area of no
~227-
Figure 2: The wind patterns around Esther Mountain and in
the White Brook Valley as indicated by flag
trees (page 228).
Figure 3: The wind patterns on the eastern slopes of Whiteface
Mountain as indicated by flag trees (page 229).
Figure 4: The wind patterns on the northwest side of Whiteface
i Mountain as indicated by flag trees (page 230).
Figure 5: The wind patterns around the peak of Whiteface
Mountain as indicated by flag trees (page 231).
direction of wind: ca /
point = ge
Vv
surface ——~ &
aloft —-+ . oO
location of wind a, ”
instrument: o S
wind shift front
no flagging: <2 ~
Scale -——., —
: oe mi ~° *
—"
I
“822-
~229-
“tw %, -——
[DIS
Cet :6u1660)) a,
24UOdy 4J1YS Pulm
@ tyueurnd sul
PUIM JO U01}D90}
~~ yoyo [| Dig
<—— . e0DJns
— —___jutod |
:PUIM 40 UOLoeIp
~230-
direction of wind:
point “2
surface ——+
aloft, red
location of wind
__ instrument: o
wind shift front:
no flagging: = <22>>
Scale +—— Va mi.
~231-
direction of wind:
point _
@ |surface ——-
location of wind
instrument: o
wind shift front: —— 400 |
no flagging: =< |
Scale
-232-
flagging may be resolved once this ridge has been accurately mapped for
flagging directions.
The most complex pattern is found on the peak of Whiteface Mountain
shown in Figure 5. Flagging indicates that air is rising upwards towards
the peak from every direction, and every ridge produces some anomaly in the
wind pattern. The vortex just east of the peak is shown by the behavior of
cloud fragments as well as by the tree crown.
Looking at the mountain as a whole in Figure 1, one sees that air tends
to flow over this mountain rather than around it. This resulting orographic
uplift is the cause of the frequent cap clouds which obscure the top of the
mountain and the view from it. These clouds are an important source of
moisture for the peak, depositing water on all exposed objects in the form
of fog drip, or as rime ice in the winter.
The method presented in this paper for obtaining detailed wind direction
information over.a large area does not require the establishment of wind
recording stations or other elaborate equipment. Using binoculars, a compass ,
and a map the many years of wind data recorded in flag trees may be studied
ina few days or weeks, depending on the area investigated. The detail of
such a tree survey depends on flag tree density and on the analyst's ability
to concentrate the data for regions of high density on a map. The disadvantages
of this method become evident in regions of low wind velocity. The slight
asymmetries in the flag trees makes measurement difficult, and errors of
+20° are common.
The ultimate proof that flag trees are good indicators of the direction
of the prevailing winds will come from long, detailed measurements of the
prevailing winds. This would require instrumentation of the region under
study, which for Whiteface Mountain with its complex wind patterns, would
be expensive in either time or money.
The above method of examining trees gives the average wind direction
during the growing season very easily. However, to obtain wind velocity
the degree of flagging might be calibrated to give a rough value for the
velocity. Since different species flag different amounts for the same winds,
it would require a calibration of each species of flag trees to obtain
velocities for a region like Whiteface Mountain.
APPLICATIONS
Even without velocities, tree flagging indications of the wind should
be very useful. When combined with a topographic map, the direction of the
wind will show regions of rising or subsiding air. The regions of uplift
will generally be more moist due to orographic condensation and precipitation,
while areas of subsidence will generally be drier. Tree flagging can show
areas of exposure to strong winds and the drying associated with them. Flag
trees can warn loggers of potential blowdown and help them pattern their
~233-
Figure 6: The average velocity from each direction and vector
average of measured winds on Whiteface Mountain.
The duration and months of the measurements are
shown in Table 1. The center of each diagram is
the instrument's location. The length of the
lines shows velocity. Flag tree directions are
plotted for comparison. For two stations the winds
were split into up-valley and down-valley winds
for the computation of two vector averages for
each station; for a third the winds were split
according to the valley from which they came (page 234).
Figure 7: A vertical cross section of the air flow in the
eastern valleys from Little Whiteface Mountain to
Marble Mountain showing the overlap of the streams
of air. The vertical scale is unexaggerated. The
depth of view is two-thirds of a mile, and the
horizontal extent of the sketch is two and one
half miles (page 235).
-234-
a6piaaD
40}90A 4yds 103 Kuopunoq---- abpseap10joaa—o Ajpojen abpseaD —+
r ww
:
G
:
-235-
winds aloft
ae Marble Mt.
ae
=4 SSW iets
ss ae Ine
Little Whiteface Mi.
SW
drainage winds
-236~
cutting according to the local prevailing winds (Alexander, 1964). A map
of these winds would aid in predicting fire behavior and movement, and the
dispersal of seeds and pollen. A close observation of flag trees would also
aid in the location of ski runs. Pockets of calm air would be ideal, and
failures due to high winds across the runs, such as occurred on Marble Mountain,
could be avoided.
ACKNOWLEDGEMENTS
This work was sponsored by a grant from the Charles F. Kettering
Foundation under the direction of Dr. Vincent J. Schaefer. Numerous text
revisions were suggested by Drs. Earl L. Stone and Jon T. Scott.
REFERENCES
Alexander, R.A., 1964, "Minimizing Windfall Around Clear Cuttings in Spruce-
Fir Forests", Forest Science, 10, pp. 130-142.
Daubenmire, R.F., 1943, "Vegetational Zonation in the Rocky Mountains",
Botanical Review, 9, pp. 325-394.
7 947, Plants and Environment, John Wiley and Sons, Inc., New York,
pp. 285-6.
Lawrence, D.B., 1939, “Some Features of the Vegetation of the Columbia River
Gorge with Special Reference to Asymmetry in Forest Trees", Ecological
Monographs 9, pp. 217-257.
Sinnott, E.W., 1952, "Reaction Wood and the Regulation of Tree Form", American
Journ. of Botany, 39, pp. 69-78.
Warming, E., 1909, Oecology of Plants, Lowe and Brydone, Printers, Ltd.,
London, pp. 37-39.
°
CC BY 4.0