Document text
BIRD POPULATIONS IN THE HEMLOCK SERE ON
THE HIGHLANDS PLATEAU, NORTH CAROLINA,
1946 TO 1972
Jane P. Holt
Bird populations were studied in selected successional stages in the hemlock
sere in western North Carolina, in relation to changes that have occurred in
the vegetation over a twenty-five year period. The three study areas represent
mesic shrubland, intermediate hemlock-hardwood forest, and climax virgin
hemlock forest, all occurring at elevations between 3,800 and 4,100 ft. The
study areas were initially mapped and censused for breeding birds in 1946
and 1947 by E. P. Odum (1950). I carried out subsequent studies on the
same study plots in 1959-1960 and 1971-1972. These combined data provide
the basis for this paper.
All of the study areas are located within four mi of Highlands, Macon
County, North Carolina. This plateau region, averaging approximately 4,000
ft in elevation, is located at the southern end of the Blue Ridge Mountains,
in southwestern North Carolina. Many plants and animals typical of areas
farther north are found in this area, and the annual rainfall of more than
80 in supports a luxuriant vegetation. The ecology of the Highlands Plateau
has been described by Oosting and Billings (1939) and Odum (1950).
METHODS
The three study plots were censused for breeding birds by using the spot-mapping
technique (Williams, 1947; Odum, 1950). Five to six censuses were taken in the mornings
on each area in the period 15-30 June 1959, three times on each from 22 May-16 June 1960,
and five to seven times on each from 23 May-16 June 1971 and 23 May-7 June 1972.
In each daily census the locations were recorded of all birds seen or heard (indicating
sex when possible), using individual copies of Odum’s original maps of the study areas.
From composite maps constructed at the end of each year’s study, territories were
delineated and the numbers of pairs of each species on each study area were determined.
When the territory of a pair of birds did not lie totally within the study area, the
appropriate fraction of the territory was estimated from the composite maps and from
field observations. All three study areas were visited in rotation for a combined total of
three to four hr each morning. Afternoon visits were used to check for nest locations.
Population density figures for 1959-1960 and for 1971-1972 were converted to pairs per
100 acres and then averaged to conform to Odum’s (1950) 1946 1947 data. Where study
plots consisted of less than 10 acres, the density of each species was reduced by 10
percent to compensate for the small sample size (Odum, op. cit.).
By 1971, some of the exact boundaries of the original 6.4 acre mesic shrubland tract
had become obliterated. This necessitated a remapping of this tract. The presence of
old logged hemlock stumps interspersed within the secondary growth made it possible
to remain within the original logged hemlock area at all times and to include all of the
397
398
THE \\ ILSON BULLETIN
December 1974
Vol. 86. No. 4
Table 1
Bird Species and Their Densities on the Highlands Plateau the Hemlock Sere
(A.O.U., 1957. Order of Species;
Pairs per 100 acres
Intermediate
Mesic shrubland Hemlock-hardwood Virgin hemlock
1946- 1959- 1971- 1946- 1959- 1971- 1946- 1959-
Species 1947 1960 1972 1947 1960 1972 1947 1960
2
Broad-winged Hawk
Buteo platypterus
Ruffed Grouse
Bonasa umbellus
Common Flicker
Col apt es auratus
Pileated Woodpecker
Dryocopus pileatus
Hairy Woodpecker
Dendrocopos villosus
Great Crested Flycatcher
Myiarchus crinitus
Eastern Phoebe
Sayornis phoebe
Eastern Wood Pewee
C onto pus tirens
Blue Jay
Cyanocitta cristata
Carolina Chickadee
Par us carolinensis
Tufted Titmouse
Par us bicolor
White-breasted Nuthatch
Sitta carolinensis
Red-breasted Nuthatch
Sitta canadensis
Brown Creeper
Certhia jamiliaris
Carolina Wren
Thryothorus ludovicianus
Gray Catbird
Dumetella carolinensis
Brown Thrasher
Toxostoma rujum
American Robin
Turdus migratorius
Wood Thrush
Hylocichla mustelina
6
6 10
6
8
6
7 6 5
13
6
8
42 28 16 25
14 7 3
14 3 2
8 16 42
7
4
8
1
4
10
4
4
4
4
*7
i
4
6
8
17
8
10
20
8
9
7
4
2
9
7
4
8
4
5
7
17
4
4
8
8
7
27
30
16
4
7
10
17
39
20
10
17
Jane P.
Holt
HEMLOCK SERE BIRD POPULATIONS
399
Table 1
(continued)
Pairs per
100 acres
Intermediate
Mesic shrubland
Hemlock-hardwood
Virgin hemlock
1946- 1959-
Species 1947 1960
1971-
1972
1946-
1947
1959-
1960
1971-
1972
1946-
1947
1959-
1960
Veery
8
9
7
9
Catharus fuscescens
Golden-crowned Kinglet
Regulus sat r a pa
Cedar Waxwing
Bombycilla cedorum
White-eyed Vireo 7
7
7
Vireo griseus
Solitary Vireo 7
9
30
27
24
25
13
Vireo solitarius
Red-eyed Vireo
16
3
14
4
4
Vireo olivaceus
Black and White Warbler
9
27
24
17
12
13
Mniotilta varia
Golden-winged Warbler 7
6
Vermivora chrysoptera
Northern Parula Warbler 4
6
3
12
13
17
17
Par id a americanus
Black-throated Blue Warbler
16
40
44
24
75
42
Dendroica caerulescens
Blackburnian Warbler
3
10
7
4
49
8
Dendroica jusca
Chestnut-sided Warbler 49 49
31
Dendroica pensylvanica
Ovenbird 11
16
5
12
7
21
Seiurus aurocapillus
Northern Yellowthroat 7
6
Geothlypis trichas
Yellow-breasted Chat 28
3
Icteria virens
Hooded Warbler
6
2
4
8
Wilsonia citrina
Canada Warbler
6
7
4
33
8
Wdlsonia canadensis
Brown-headed Cowbird
3
7
Molothrus ater
Scarlet Tanager 7
3
7
7
4
27
8
Piranga olivacea
Cardinal 7
6
3
7
10
Cardinalis cardinalis
400
THE WILSON BULLETIN
December 1974
Vol. 86, No. 4
Table 1 (continued)
Pairs per 100 acres
Intermediate
Mesic shrubland Hemlock-hardwood Virgin hemlock
1946- 1959- 1971- 1946- 1959- 1971- 1946- 1959-
Species 1947 1960 1972 1947 1960 1972 1947 1960
Rose-breasted Grosbeak
Pheucticus ludovicianus
Indigo Bunting
Pas senna cyanea
American Goldfinch
Spinus tristis
Rufous-sided Towhee
Pipilo erythrophthalmus
Dark-eyed Junco
Junco hyemalis
Chipping Sparrow
Spizella passerina
Field Sparrow
Spizella pusilla
Song Sparrow
Melospiza melodia
Total pairs
Total species
4
3
17
14
9
14
7
6
14
28
19
9
8
7
3
22
35
6
126
28
9
288
321
296
270
8
22
35
21
14 24 2 2
4 17
20 30 2
351
349
331
230
28
30
19
23
1946-1947 tract during the process of remapping. The remapped mesic shrubland tract
censused in 1971-1972 covered an area of 16 acres, and this acreage figure was used in
computing pairs per 100 acres for this period. Although the increase in the number of
species from 1959-1960 to 1971-1972 could have been affected by the increase in size
of the mesic shrubland tract, I consider the increase primarily a result of change in habitat
due to further plant succession. An examination of Table 1 shows the species added in
1971-1972 to be birds typical of woodland habitats.
Unfortunately, an accurate census was not possible in the virgin hemlock forest tract
in 1971-1972. Only a trail around one side of this study tract remained open by that time.
All remaining trails were completely obliterated by impenetrable tangles of Rhododendron ,
and my single crossing of the tract involved a great effort and progression on hands and
knees. Although an accurate census was impossible, it was possible to walk along the
open trail and identify the calls of birds breeding in approximately one-fourth of the
study tract. This was done oil eight mornings in 1971-1972.
Although more refined statistical methods of analyzing bird species diversity exist today,
the breeding bird density (in pairs per 100 acres) of one plot is compared with that of
another plot according to the ‘'percentage difference” method of Odum (1950). This
allows the data from the 1959-1960 and 1971-1972 studies to be treated by the same
method as the 1946-1947 data. For example, in comparing avian density in the mesic
shrubland with that in the intermediate hemlock in 1959-1960 (Table 1), the species-
Jane P.
Holt
HEMLOCK SERE BIRD POPULATIONS
401
by-species difference in pairs per 100 acres are first calculated: e.g. for the Chestnut-sided
Warbler the value is 49 and for Gray Catbird it is one (scientific names of birds are
given in Table 1). All of the differences are then added, and that total difference is
divided by the total number of pairs per hundred acres in the two tracts. The percentage
difference is a rough index of difference (or similarity) in species composition and in
density. A percentage difference at the low end of the scale indicates a close similarity,
while one at the higher end shows that the populations are markedly different.
THE STUDY PLOTS
Mesic Shrubland
The example of this primary successional stage of the hemlock sere is
located four mi east of Highlands, North Carolina, on U.S. Highway 64,
downstream from the Virgin Hemlock Forest study area. In 1946-1947,
Odum (1950) described the vegetation as consisting . . of shrubs and
small trees chiefly alders, willows, and brambles ( Rubus ) which form dense
thickets interspersed with more open areas dominated by grasses and herbs.
Huge stumps of hemlock and thickets of laurel ( Kalrnia ) and rhododendron
remain from the original hemlock forest which occupied the area. The plot
has never been cultivated, but has apparently been subjected to some grazing
in the past.”
Of the three study plots, naturally the mesic shrubland has undergone the
greatest change during the twenty-five years since 1947. Approximately one
acre of this tract has continued under intermittent grazing or cultivation. An
additional acre appears to have been disturbed by 1971-1972. The size of
the tract censused in 1946—1947 and 1959—1960 was 6.4 acres. As explained
earlier, the censused area was enlarged in 1971-1972 to 16 acres, including
the two which had been disturbed in the past.
In the disturbed area, blackberry ( Rubus sp.) forms a dense cover, with
herbs and grasses covering the more open areas. A dense stand of sassafras
(Sassafras albidum ) has grown up at one end. The stream runs through a
portion of the disturbed area. The remaining part of the study tract is
covered with maples ( Acer spp.), oaks ( Quercus spp.), birches (Betula spp.),
tulip poplar ( Liriodendron tulipifera) , cherry { Prunus sp.), sourwood ( Oxy -
dendrurn arboreum) , Fraser magnolia ( Magnolia fraseri ), locust ( Robinia
spp.), dogwood ( Cornus spp.), and hemlock ( Tsuga spp.). Thickets of
rhododendron and laurel persist along the stream. Stumps of the old logged
hemlocks may still be seen. In some cases hemlock sprouts from these stumps
have produced trees more than twenty ft in height. The forest canopy aver¬
ages approximately 20 ft in height, with some individual maples and hemlocks
30 to 40 ft high. The plot is in an intermediate stage of succession, but will
continue to be termed the “mesic shrubland” for purposes of comparison.
402
THE WILSON BULLETIN
December 1974
Vol. 86, No. 4
Intermediate Hemlock-hardwood Forest
I he example of this secondary successional stage covers an area of 15 acres,
located around the Highlands Biological Station but excluding the clearings
for the buildings, etc. Most of the trees are hemlock, white pine {Pinus stro-
bus ). birch, and maple, with evergreens comprising approximately 50 per¬
cent of the total. Most of these trees comprise second growth, but some
are quite large and may represent remnants of the original virgin forest.
Dense stands of rhododendron, laurel, witch hazel (Hamamelis sp.), and
azalea (Rhododendron spp.) form the understory. Little change has occurred
in this tract since 1946-1947, except an increase has occurred in the size of
the trees and the density of the understory in some areas. The size of the
tract was decreased slightly with clearing for the erection of some new build¬
ings at the Highlands Biological Station.
Virgin Hemlock Forest
The final successional stage studied is the virgin hemlock forest, a 12
acre tract located four mi east of Highlands on U.S. Highway 64 (opposite
the Mesic Shrubland). In 1946-1947, Odum (1950) stated that three-fourths
of the stand w as comprised of hemlock trees three to four ft in diameter and
400 or more years old. This virgin forest is a remnant of a once extensive
stand of hemlock known as the “Primeval Forest” or “Ravenel’s Woods.”
Large black birch (Betula lenta), red maple (Acer rub rum ), Fraser magnolia,
red oak, and other deciduous trees are widely scattered through the stand.
Rhododendron makes up the dense understory.
This plot has been subjected to one man-made alteration since 1946, the
addition of a small pond (approximately 30 by 50 ft), close to the center
of the study area. Observations of the species nesting there in 1959-1960
indicated that the pond did not affect species or numbers of breeding birds
during that period. At that time access to the interior of this forest was
gained along primitive roads around part of the perimeter and paths to the
center. By 1971-1972 these had become obliterated by rhododendron.
RESULTS
Table 1 shows species composition and densities of birds in the three study
areas of the hemlock sere. Bird names are listed in A.O.U. Check-list order
(1957; Eisenmann et al., 1973). In the interval from 1946-1947 to 1959—
1960, 14 additional species were added in the mesic shrubland, eight in the
intermediate hemlock-hardwood tract, and eight in the virgin forest. One
species which occurred in the intermediate hemlock-hardwood forest in 1946-
1947 was not found in 1959-1960 and four species were absent from the
virgin forest. In the 1971 1972 census, fourteen additional species were
Jane P.
Holt
HEMLOCK SERE BIRD POPULATIONS
403
Table 2
Percentage Difference in the Breeding Bird Populations of the Three Study Plots
Compared with Each Other
Census year
1946-1947
1959-1960 1971-1972
Mesic shrubland and virgin hemlock
97.9
83.5
Mesic shrubland and intermediate hemlock
95.8
70.2 50.1
Virgin hemlock and intermediate hemlock
50.5
33.2
recorded in the mesic shrublancl and one species found in 1959—1960 was
absent, while five species were added in the intermediate hemlock-hardwood
tract and three species were absent. The number of species common to all
three tracts increased from one in 1946-1947 to seven in 1959-1960.
Percentage differences (Table 2j showed marked declines. The decrease is
particularly striking when the mesic shrubland is compared with the inter¬
mediate hemlock-hardwood forest, the value shifting from 95.8 percent in
1946-1947 to 50.1 percent in 1971-1972.
Table 3 shows the species of birds common to both mesic shrubland and
intermediate hemlock-hardwood tracts increased from 12.5 percent in 1946-
1947 to 65.7 percent in 1971-1972. The greatest change in bird species rich¬
ness on any of the tracts during the twenty-five year period occurred in the
mesic shrubland, where the number of species increased from eight in 1946-
1947, 22 in 1959-1960, and 35 in 1971-1972.
Although the virgin tract was too dense for accurate census in 1971-1972,
two additional species were recorded during visits to this tract. The Golden-
crowned Kinglet was consistently present both years and a nesting Rufous¬
sided Towhee was noted in 1971.
Table 3
Comparison of Bird Species Composition of Mesic Shrubland Tract with
Intermediate Hemlock-hardwood Forest Tract
Census year
Total species
mesic shrubland
Species common
to both tracts
Percentage of
species common
to both tracts
1946-1947
8
1
12.5
1959-1960
22
12
54.5
1971-1972
35
23
65.7
404
THE WILSON BULLETIN
December 1974
Yol. 86, No, 4
DISCUSSION
In the 1946-1947 census, only the Gray Catbird was found in all three
study areas. By 1959-1960, the number of species common to all three study
areas had increased to seven. Calculations of the percentage difference of
the three study plots also confirm greater similarity. The difference between
the mesic shrubland and intermediate hemlock-hardwood tracts decreased
from 95.8 percent in 1946-1947 to 70.2 percent in 1959-1960 and 50.1 per¬
cent in 1971-1972.
The species in common between mesic shrubland and intermediate hem¬
lock-hardwood tracts over the twenty-five year period increased from 12.5 to
65.7 percent. The similarity of species in the three plots would have been
even greater in 1971-1972, if the entire mesic shrubland plot had been left
undisturbed. Disturbance served to maintain prime habitat for several species
characteristic of early stages of succession. These include such birds as the
Song and Field Sparrows and the very high concentration of Chestnut-sided
Warblers. The American Goldfinch, Indigo Bunting, Yellow-breasted Chat,
Chipping Sparrow, Golden-winged Warbler, and Northern Yellow-throat also
occurred in the area. These nine species, together with birds characteristic
of the later successional stages, account for the high figures of 35 species
and 296 pairs listed for this period.
The former mesic shrubland is now entering the intermediate hemlock-
hardwood stage, as is evidenced by the presence of the Blackburnian Warbler
which now breeds there. Bent (1953) describes this bird as nesting to heights
of over 80 ft above ground in coniferous trees, with males singing from tall
trees. The habitat preference of this bird and its presence in this study tract
certainly reflect the degree of plant succession which has occurred.
The number of species represented within all serai stages in the 25 year
period has increased markedly. The most dramatic increase shown in this
study is in the mesic shrubland where the number increased from eight to 35
species. The increases in the intermediate hemlock-hardwood forest from
21 to 30 and in the virgin forest from 19 to 23 species (in the first thirteen
years of this study) are less striking.
Kricher (1972) found that bird species richness at all seasons increased
with the age of the serai stages. When the Highlands area serai stages are
compared with each other, however, species richness peaked as the former
mesic shrubland entered the intermediate hemlock-hardwood stage (1971—
1972), rather than in the climax virgin forest. Mesic seres are known to
have peaks in avian diversity in the middle rather than at the end (Adams,
1908; Aldrich, 1943).
While the number of species increased within each serai stage, the number
of individuals in the virgin hemlock tract declined markedly from 331 pairs
Jane F.
Holt
HEMLOCK SERE BIRD POPULATIONS
405
per 100 acres in 1016-1947 to 230 in 1959-1960. This decline probably can
be attributed to continued growth of the rhododendron understory, which
characteristically is poor bird habitat.
Several observations concerning the distribution of particular species are
of interest. Odum (1950) noted in 1946-1947 that the Highlands area lacked
such northern species as Golden-crowned Kinglet, Magnolia Warbler, and
Black-throated Green Warbler. Recently, Johnston (1964) lists the Golden-
crowned Kinglet as an uncommon permanent resident of the area, and a nest
was found in Highlands on 11 June 1951. In 1971-1972, I recorded this
species consistently as a breeding bird of both the intermediate hemlock-
hardwood and virgin hemlock forests. Johnston (1964) classifies the Black-
throated Green Warbler as a summer visitor of the Highlands area, but with
no definite evidence of breeding. I observed this species in the virgin hemlock
forest tract once during the census of 1959 and again in I960. As it was
not observed more than once, it was not included in the breeding bird list;
nevertheless, its presence during the breeding season may indicate that further
study will prove it to be a member of this breeding avifauna. The Magnolia
Warbler, although a regular migrant, has not been recorded as a summering
or breeding species.
Another species which appears to have extended its breeding range into
the Highlands plateau is the Brown-headed Cowbird. This bird was not
recorded in the pre-1972 breeding bird counts, and Johnston (1964) classified
it as uncommon—with only one observation appearing on the Highlands
Biological Station records. In 1972 this cowbird appeared in both the inter¬
mediate hemlock-hardwood forest around the Highlands Biological Station
and in the mesic shrubland, four mi east of Highlands. Both adults and young
were commonly noted in May and June and its population density should
be monitored in the future.
SUMMARY
Breeding bird population studies were carried out in the hemlock sere at elevations of
3,800 to 4,100 ft on the Highlands Plateau, western North Carolina, over a twenty-five
year period. Populations of three carefully selected stages of succession in the sere (mesic
shrubland, intermediate hemlock-hardwood, and virgin forest) were studied using the
spot-mapping technique. Breeding bird population analyses in 1959-1960 and 1971-1972
were compared with the initial survey of these study areas by Odum (1950) in 1946-1947,
and population density and species composition were compared with changes in the
vegetation. Increases in density and species richness were correlated with the age of
the serai stages, as evidenced by comparing percentage differences in bird populations
of the study areas in the different census years. Both avian density and species richness
showed an increase in the intermediate stage and a decline in the climax stage of the
hemlock sere.
406
THE WILSON BULLETIN'
December 1974
Yol. 86, No. 4
ACKNOWLEDGMENTS
This study was supported by N.S.F. grants made to the Highlands Biological Station,
Highlands, North Carolina. 1 wish to express my thanks to Dr. E. P. Odum who suggested
this study in 1959, to Mr. Henry Wright who owns the virgin hemlock forest, to the late
Mr. Toliver Crunkleton for valuable aid in the initial stages of this study, and to Dr.
Thelma Howell, former director of the Highlands Biological Station for her continued
guidance. Dr. F. W. Harrison and Dr. Paul R. Ramsey of Presbyterian College offered
advice and encouragement during the course of study and the preparation of this paper.
LITERATURE CITED
Adams. C. C. 1908. The ecological succession of birds. Auk, 25:109-153.
Aldrich, J. W. 1943. Biological survey of the bogs and swamps in northeastern Ohio.
Amer. Midland Nat., 30:346-402.
A.O.U. 1957. Check-list of North American birds. 5th ed. American Ornithologists''
Union
Eisenmvnn, E. (chairman). 1973. Thirty-second supplement to the American Ornithol¬
ogists’ Union check-list of North American birds. Auk, 90:411-419.
Bent, A. C. 1953. Life histories of North American wood warblers. U.S. Natl. Mus.
Bull. 203.
Johnston, D. W. 1964. The birds of Highlands, North Carolina, with a preliminary list
from Cashiers and nearby gorges. Jour. Elisha Mitch. Sci. Soc., 80:29-38.
Kricher, J. C. 1972. Bird species diversity: the effect of species richness and equita-
bility on the diversity index. Ecology, 53:278-282.
Odcm, E. P. 1950. Bird populations of the Highlands (North Carolina) Plateau in
relation to plant succession and avian invasion. Ecology, 31:587-605.
Oostinc. H. J., and W. D. Billings. 1939. Edapho-vegetational relations in RavenePs
Woods, a virgin hemlock forest near Highlands, North Carolina. Amer. Midland Nat..
22:333-350.
Williams. A. B. 1947. Breeding-bird census: climax beech-maple forest with some
hemlock (15 year summary). Audubon Field Notes, 1:205-210.
DEPARTMENT OF BIOLOGY, PRESBYTERIAN COLLEGE, CLINTON, SOUTH CAROLINA
29325. ACCEPTED 1 AUGUST 1974.