Early Journal Content on JSTOR, Free to Anyone in the World
This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in
the world by JSTOR.
Known as the Early Journal Content, this set of works include research articles, news, letters, and other
writings published in more than 200 of the oldest leading academic journals. The works date from the
mid-seventeenth to the early twentieth centuries.
We encourage people to read and share the Early Journal Content openly and to tell others that this
resource exists. People may post this content online or redistribute in any way for non-commercial
Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early-
JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people
discover, use, and build upon a wide range of content through a powerful research and teaching
platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit
organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please
Vol. 41 No. 4
TORREY BOTANICAL CLUB
The " Pocosin " of Pike County, Alabama, and its bearing on certain
problems of succession
Roland M. Harper
(with four text figures)
In eastern North Carolina, where the geographical term
" pocosin" is used more frequently than in all the rest of the world,
it means a flat wet place with peaty soil, usually remote from large
streams, with a scattered growth of trees, mostly Pinus serotina,
and a dense shrubby undergrowth.* In South Carolina and
Georgia this term is almost unknown, f but it reappears in the
coastal plain of Alabama, with a very different meaning.
In a chapter contributed by E. Q. Thornton to Prof. Tuomey's
second report on the geology of Alabama, published in 1858 (p.
244), we find the following interesting bit of information:
"East of this place [Troy] is a small scope of country of peculiar appearance,
known as the Pocosson settlement. The vallies, which have the rank luxuriant
growth of a swamp, are surrounded on three sides by a ridge of snow-white sand,
which seems to have been heaped up by the tides on a sea-shore. The only occupant
of the hills is a scrubby oak covered with long moss. The soil is exceedingly un-
productive, and soon wears out by cultivation."
*See Bull. Torrey Club 34: 361-363. 1907; C. A. Davis, N. C. Geol. Surv.
Econ. Paper 15: 149-150. 1910; L. W. Stephenson, N. C. Geol. Surv. Vol. 3: 280.
pi. 24B. 1913.
t Miss E. F. Andrews (see Torreya 13: 64-66. 1913) wrote me early in 1913 that
she had heard the term applied to certain "large flowery swamps" in the vicinity of
Albany, Georgia, where she used to live; and the government soil survey of Dougherty
County, Georgia, published in October, 1913. indicates a "Percosin Creek" a few
miles west of Albany, but says nothing about any application of the name to vegeta-
tion. Still more recently (March, 1914) I have heard of an occurrence of the term
in West Florida.
[The Bulletin for March (41: 137-208. pi. 1-6) was issued 22 Ap 1914]
210 Harper: The "Pocosin" of Pike County, Ala.
In Dr. Eugene A. Smith's description of Pike County in his
report on the agricultural features of Alabama (Tenth Census
U. S. 6: 151; Geol. Surv. Ala., Report for 1881 and 1882, p. 522.
1884) these observations of Thornton's are briefly referred to.
About the same time, in the 9th volume of the Tenth Census
(p. 528), Dr. Charles Mohr published some more definite informa-
tion about the vegetation of the same place, as follows :
"Pike County. — On the broad ridges which form the divide between the waters
of the Pea and Conecuh rivers, upon a purely sandy soil, are found, within the forest
of long-leaved pine, tracts with strictly-defined outlines from a half mile to several
miles in width, covered with a dense vegetation of small trees and shrubs peculiar to
the perpetually moist and cool hummocks* of the coast. The soil covered with this
growth presents no unusual features; it is as poor and arid as that covering the rest
of these heights. Surrounded on all sides by pine forests, not a single pine tree is
seen within the limits of these glades, called by the inhabitants 'pogosines,' an Indian
name the meaning of which I was unable to learn.
"The trees are of small growth, the willow oak, the water oak, beech, red maple,
and black gum rarely rising to a height of more than 30 feet among the sourwoods,
junipers, hornbeams, hollies, papaws, fringe trees, red bays, and other trees of the
coast. These glades verge upon deep ravines from which issue large springs, and
from this fact I conclude that, below their sandy, porous soil, strata must exist
perpetually moistened by subterranean waters near enough to the surface to supply
the moisture necessary to support such a luxuriant vegetation."
There seems to be no reference to this interesting place in
Dr. Smith's report on the geology of the coastal plain of Alabama
(1895), or in Dr. Mohr's Plant Life of Alabama (1901). In the
summer of 1906 Dr. Smith and the writer were in Pike County
together for a short time, and heard some accounts of the "pocosin,"
which led Dr. Smith to make his first visit to it a little later.
In 1910a soil survey of Pike County was made by W. E. Tharp,
of the U. S. Department of Agriculture, and W. L. Lett and W. E.
Wilkinson, representing the State; and on the map accompanying
their report, published in December, 191 1, the location of the
pocosin is shown, probably for the first time, but there is not a
word about it in the text. Stranger still, its soil is not differentiated
on the map from that of the surrounding country (" Norfolk
coarse sand"), although it does differ in at least one important
* Dr. Mohr doubtless wrote "hammocks," as he did in his Plant Life of Alabama
17 years later, but it was evidently changed to "hummocks" in Washington, as it was
throughout the 5th and 6th volumes of the same series, except on the maps, and one
or two places in the text that escaped the proof-readers. (See Geol. Surv. Ala.
Monog. 8: 83. 1913O
Harper: The "Pocosin" of Pike County, Ala. 211
Guided by the map just mentioned, the writer visited the spot
on November 6, 1912, and March 27, 1913, and made a rough quan-
titative study of its vegetation. An incessant rain at the time of
the first visit interfered somewhat with note-taking and made
photography out of the question, but conditions were more satis-
factory the second time, when the views here published were
taken. A brief mention of it, with the only photograph of its
vegetation hitherto published, appeared last summer in my report
on the forests of Alabama.*
The so-called pocosin (Dr. Mohr's description above quoted
implies the existence of several such areas, but I know of only one)
embraces a hundred acres or more, mostly in Section 7, T. 9 N.,
R. 22 E., in the midst of a few square miles of undulating sandy
country on the east side of Walnut Creek, about half way between
Troy and Brundidge, and about 50 miles south of the fall-line
and 100 miles from the Gulf coast. The pocosin itself is practically
untouched by civilization, except for having one or two little-used
roads through it, but the surrounding sandy country is partly
under cultivation. As observed by Thornton, however, the sand
is much less productive than the loamy soils which prevail else-
where in Pike County and other parts of the southern red hills or
Eocene region of the coastal plain. Where not cultivated it
bears a vegetation much like that of the sand-hills of Georgia,!
but the soil is evidently a little richer in mineral plant food than
that of the average sand-hill, as shown by the prevalence of Pinus
echinata and the scarcity of Pinus palustris.%
In the following list of sand-hill plants growing around the
pocosin the trees, shrubs and herbs are arranged as nearly as
possible in order of abundance, but the data are not sufficient for
assigning percentages to them. Evergreens are indicated by
Pinus echinata Hicoria glabra
Crataegus Michauxii? Quercus stellata
Quercus Catesbaei Quercus marylandica
Quercus cinerea Nyssa sylvatica
Quercus Margaretta Pinus palustris
* Geol. Surv. Ala. Monog. 8: 99-100, 160-161. June, 1913.
t See Ann. N. Y. Acad. Sci. 17: 82-89. 1906.
% This is in a narrow belt of the coastal plain where the long-leaf pine is very rare.
See Geol. Surv. Ala. Monog. 8: 99. 1913.
212 Harper: The "Pocosin" of Pike County, Ala.
Shrubs and Vines
Batodendron arboreum Osmanthus americana
Ceanothus americanus Chionanthus virginica
Gaylussacia dumosa Gelsemium sempervirens
Opuntia vulgaris Cracca virginiana
Pteridium aquilinum Chrysopsis aspera?
Afzelia pectinata Coreopsis major Oemleri
Kuhnistera pinnata* Tillandsia usneoides
Solidago odora Sorghastrum secundumf
Stenophyllus ciliatifolius* Siphonychia sp.
Polypodium polypodioides Pitcheria galactioides*
Iris verna Astragalus villosus*
Triplasis americana* Warea cuneifoliaj
This vegetation, like that on typical sand-hills, is not very
dense, as shown by one of the accompanying illustrations (fig. i).
Apparently about one third of the woody plants (counting individ-
uals, not species) are evergreen.
The pocosin itself seems to center around the precipitous heads
of a few small tributaries of Walnut Creek, but its vegetation,
except in the bottoms of the ravines, is not at all of a swamp
character, the statements of Thornton and Mohr to the contrary
notwithstanding. Its soil was doubtless originally the same as
that of the surrounding sandy country, but it is now covered and
more or less mixed with a thin layer of humus, derived from the
trees and protected from desiccation and oxidation by their shade,
as in other dense forests the world over. In this forest there is a
characteristic faint odor of sour humus (corresponding approxi-
mately with the raw humus of Warming§ or more closely with the
upland peat of Coville||), as in the mountains of North Carolina
* These five species do not seem to have been found so far inland in Alabama
before. Three of them are Leguminosae.
f Apparently not previously reported from Alabama. (See Ann. N. Y. Acad.
Sci. 17: 300. 1906.) But in September, 1912, I found it in considerable abundance
in the pine-barrens of Clarke, Monroe and Baldwin Counties.
% Another addition to the Alabama flora, its previously known range being from
South Carolina to Florida.
§ Oecology of Plants, 62-63. 1909.
|| U. S. Dept. Agr. Bur. Plant Industry Bull. 193: 32-34. F 191 1; Jour.
Wash. Acad. Sci. 3: 84-86. F 1913.
Harper: The "Pocosin" of Pike County, Ala. 213
and many other places where humus accumulates on a soil that
is poor in animal life or not intrinsically very fertile. The bound-
ary between the pocosin and sand-hill vegetation is not very sharp,
but it is comparatively easy to fix it definitely enough for statis-
tical purposes. In attempting to make a quantitative analysis
of the vegetation, however, one encounters a difficulty in that it is
a many-storied forest, in which there seem to be all gradations in
Fig. i. Sand-hill vegetation near the pocosin, with Pinus echinata, Quercus
Catesbaei, Crataegus, Batodendron, etc.
size between large trees and shrubs, and it is not exactly right to
include both trees and shrubs in the same list for the purpose of
calculating percentages. The following list is divided into more
classes than usual, and the relative position of the species within
each class is approximately correct, but it gives no idea of the
relative abundance of species in different classes. The illustra-
tions, however, partly make up for this. (Figs. 3, 4.)
214 Harper: The "Pocosin" of Pike County, Ala.
Solidago odora (toward edges)
Crataegus Michauxii? (toward edges)
Quercus Margaretta (toward edges)
Kalmia latif olia
Bryophytes and Thallophytes
It is interesting to note that among 57 species here listed as
growing in the pocosin there is only one fern, and such large
families as Gramineae, Cyperaceae, Leguminosae and Com-
* An unidentified — and probably undescribed — oak, somew T hat similar in ap-
pearance to Q. marylandica Muench. and Q. arkansana Sarg., but apparently more
closely related to Q. nigra L. (Q. aquatica Walt.), Q. myrtifolia Willd. (as that species
is commonly interpreted), and Q. microcarya Small. I have never seen anything like
it elsewhere. It grows mostly toward the edges of the pocosin, but not in the sand-
hill vegetation. It is one of the commoner species there, as indicated by its position
in the list, and it is difficult to understand how Dr. Mohr overlooked it if he visited
this same spot, unless he was there only in winter, when it was leafless. (Fig. 3.)
Harper: The "Pocosin" of Pike County, Ala. 215
positae are also represented by only one species each. There are
five oaks (the same number as in the surrounding sand-hill area)
three representatives of the Ericales, and three colorless sapro-
phytes or root-parasites, namely, Monotropa, Conopholis and
Although there are no quantitative figures to show it, just
about half of this vegetation is evergreen. It is very similar to
Fig. 2. Scene on sandy road just outside of the pocosin. Flowering branches of
Quercus stellata in upper left corner; Quercus Catesbaei at right, bearing many dead
leaves of the preceding season.
Fig. 3. Trunk of the undescribed Quercus (34 inches in circumference, breast-
high, and about 30 feet tall) in the pocosin near its eastern edge.
that of some of the sandy hammocks of Georgia and Florida;*
and it may be more or less of an accident that the name pocosin
was applied to this place by the early settlers instead of hammock .f
* See Ann. N. Y. Acad. Sci. 17: 98-102. 1906; Bull. Torrey Club 38: 515-517.
t If the early immigrants had come from the southeastward instead of north-
eastward they would doubtless have been familiar with hammocks, which are more
prevalent nearer the coast.
216 Harper: The "Pocosin" of Pike County, Ala.
The proportion of evergreens is usually somewhat larger in the
hammocks nearer the coast, however. The more remote relation-
ships of this vegetation may be indicated by saying that it is
intermediate in character between the deciduous forests that are
common on rich uplands in the moderately humid parts of the
North Temperate zone, and the " sclerophyllous forests" of
Warming and other European ecologists.
A list of the characteristic plants of the ravines will complete
the description of the pocosin vegetation. They are about as
follows (the arrangement being the same as in the two preceding
Evergreens are decidedly in the majority here.
Previous visitors to this place have been more or less mystified
by the occurrence of such luxuriant vegetation in such sandy soil.
The explanation is doubtless the same as for other sandy hammocks,
and is very simple to one familiar with conditions in South Georgia
The pocosin area, exclusive of the ravines, was presumably at
some time in the past covered with the same sort of sand-hill
vegetation that now partly surrounds it. Vegetation of a denser,
more " climactic "* type must have gradually spread from the creek
valley up the ravines to their heads, and then out across the
* This adjective, derived from climax, is rarely if ever seen in botanical litera-
ture, but some such word seems to be needed. Some ecologists have been using
climatic, an entirely different word, to convey essentially the meaning here intended;
while many use climax and mesophytic more or less interchangeably, which seems
to be a perversion of the original meaning of the latter.
Harper: The "Pocosin" of Pike County, Ala. 217
sand, making more humus all the time, and crowding out the
intolerant sand-hill plants, as in normal succession the world over;
and it is probably still spreading slowly but surely. The pocosin
vegetation is simply the climax for that type of soil, while the sand-
hill vegetation is the pioneer.
In explaining this succession there is one environmental factor
that must not be lost sight of. Most of the plants in the pocosin,
Fig. 4. Typical pocosin vegetation. Magnolia grandiflora (the largest trunk),
Osmanthus, Batodendron, Prunus caroliniana, etc. Ground covered with undecayed
as in other hammocks, are very sensitive to fire. Fire often
sweeps through the upland pine-oak-hickory woods that are
common on the better soils of the same region, but the sand-hill
vegetation is ordinarily too sparse and open to carry a ground
fire any distance, so that it protects the pocosin vegetation on
three sides from any fires which may originate in the surrounding
country. Protection on the remaining side is afforded by the
swamps and bottoms of Walnut Creek, or the broken topography.
The hammocks of the Altamaha Grit region of Georgia are similarly
protected from fire by the creek swamp on one side of them and
the sand-hills on the other, and this protection has been an essential
factor in determining their present vegetation.*
Finally this Alabama pocosin throws valuable light on the
* See Bull. Torrey Club 38: 524-525. 191 1-
218 Harper: The "Pocosin" of Pike County, Ala.
relation of evergreens to soil and succession. Although a great
deal has been written about the phenomena of succession in
various parts of the world, no one yet seems to have published
any satisfactory short definition of pioneer and climax vegetation
by which they may be recognized wherever found, without knowing
the species; and consequently every ecologist has his own views
on these points, and perhaps no two of them agree exactly. It
is probably pretty generally agreed that normal succession (biotic
succession of Cowles*) involves enrichment of the soil in some
way; and as evergreens, in temperate climates at least, are well
known to be most abundant in the poorer soils,f I supposed until
about two years ago that no typical climax forest could contain
any evergreens ; and that view has found expression in my writings.
Closely connected with that belief was one which seems to be still
very generally held, namely, that for every (climatic) region
there is one climax type of vegetation toward which all others are
tending. I am now pretty well satisfied, however, that almost
every type of soil has its own characteristic pioneer and climax
vegetation, more or less distinct from those of other types.
The proportion of evergreens in a given habitat or region seems
to be correlated with the amount of available potash (and perhaps
other minerals) in the soil,t and therefore should not be affected
much by the sort of succession whose essential feature is the
accumulation of humus, or the progress of nitrification. Although
an admixture of humus is indeed believed to increase the availa-
bility of the mineral plant foods in soil, a soil totally lacking in
potash or lime would gain none of these ingredients from humus
formed in place. It is even possible that in soils deficient in
soluble minerals — as is the case with most sandy and peaty soils —
and protected from fire like our pocosin, a considerable proportion
of the potash within reach of the roots of trees is locked up for
several years at a time in the dead leaves which lie on the ground
undecayed, and thus the proportion of evergreens may actually
increase with succession, as it does in the present case, where the
ravines have the most evergreens and the sand-hills the fewest.
* Bot. Gaz. 51: 171-180. 1911; Ann. Assoc. Am. Geogrs. 1: 12-19. 1912.
f See Rep. Mich. Acad. Sci. 15: 197. 1914.
% See Geol. Surv. Ala. Monog. 8: 28-29; Torreya 13: 140, 141, 143; Bull.
Torrey Club 40: 398. 1913.
Harper: The "Pocosin" of Pike County, Ala. 219
(In the pine-barrens frequent fires return the potash, etc., quickly
to the soil, while in calcareous or clayey soils there are usually
many earthworms, centipedes, snails, bacteria or other soil
organisms which assist greatly in converting the leaves into
humus.*) Another successional factor which doubtless tends to
diminish the availability of mineral nutrients in the soil is the
denser shade of the climax vegetation ; for this lowers the temper-
ature of the soil in summer and probably makes the ground-
water level more constant, besides diminishing evaporation and
Although no chemical analysis of the pocosin soil has been
made, it is evident from a casual inspection that it consists mostly
of siliceous sand, and must be rather poor in soluble minerals,
like the sandy hammocks of Florida. The climax vegetation of the
more clayey soils in the same region is found on river-banks, bluffs,
and sides of ravines, where fire is kept away pretty well by the
topography; and it differs considerably from that of the pocosin.
The following list, of trees only, is a generalized one for the upland
climax forests of the whole southern red hill region of Alabama.
The species are arranged approximately in order of abundance,
Liquidambar Styraciflua Ostrya virginiana
Pinus Taeda Oxydendron arboreum
Liriodendron Tulipifera Acer floridanum
Fagus grandifolia Fraxinus americana
Quercus alba Tilia heterophylla?
Cornus florida Quercus nigra
Magnolia grandiflora Cercis canadensis
Pinus glabra Halesia diptera
Hicoria alba Magnolia acuminata
Magnolia macrophylla Quercus rubra
Ilex opaca Symplocos tinctoria
Nyssa sylvatica Magnolia pyramidata
Most of the species in this list are common also to the pocosin,
but their relative abundance is different, and the proportion of
evergreens is considerably less, being probably not over one third
or one fourth. Evergreens naturally differ somewhat among
themselves in their soil requirements, and it happens that the
* Mr. Coville's valuable paper on the formation of leaf-mold (Jour. Wash. Acad.
Sci. 3: 77-89. 1913) should be consulted in this connection.
220 Harper: The "Pocosin" of Pike County, Ala.
two pines here listed prefer richer soils than all the other southern
pines and some deciduous trees.
Although it is digressing a little from the scope of this paper,
it is interesting to note that the hammocks and river-bluffs of the
Altamaha Grit region of Georgia* differ from each other in much
the same way that the pocosin does from the bluffs, etc., just
mentioned. Both hammocks and river-bluffs are covered with
climax forests, but the soil is sandy in one case and loamy in the
other, which makes considerable difference in the vegetation.
From the frequency numbers which have been published in the
work cited, it appears that 79.6 per cent, of the trees and 30.5
per cent, of the shrubs in the hammocks are evergreen; while the
corresponding figures for river-bluffs are 37 and 15; a difference
whose significance was not suspected until long after these Georgia
lists were published. The sand-hills which border the Georgia
hammocks and protect them from fire have been estimated in the
same way to have 28.2 per cent, of their trees and 48.3 per cent,
of their shrubs evergreen.
This so-called pocosin (which has little in common with the
typical pocosins of North Carolina) is a many-storied climax forest
of a type characteristic of dry sandy soils in the coastal plain of
Alabama, Georgia, and northern Florida. The accumulation of
humus — and consequently the development of climax vegetation —
began in ravines, and has been favored by the protection from fire
afforded by the sparseness of the surrounding pioneer vegetation.
Something like half of the woody plants of the pocosin are
evergreen, which is evidently a larger proportion of evergreens
than in the case of the pioneer vegetation of the same soil and of
the climax vegetation of soils richer in mineral plant food in the
same region and elsewhere.
Each fundamentally different type of soil seems to have its
own characteristic pioneer and climax vegetation; and the pro-
portion of evergreens in this case — if not in many others — increases
with normal succession, owing probably to the lowering of soil
temperatures during the growing season, and to the locking up of
plant food in undecayed leaves or sour humus, among other
♦Ann. N. Y. Acad. Sci. 17: 98-109. 1906.