UNIVERSITY OF PUERTO RICO
Focully of Arts and Sciences
INSTITUTE OF CARIBBEAN SCIENCE
STUDIES ON THE VEGETATION OF PUERTO RICO
I. DESCRIPTION AND INTEGRATION OF THE 0
PLANT-COMMUNITIES by Pierre Donsereau
II. ANALYSIS AND MAPPING OF THE ROOSEVELT
ROADS AREA by Peter F. Buell and Pierre Dansereau
SPECIAL PUBLICATION No. I
MAYAGUEZ, PUERTO RICO
STUDIES ON THE VEGETATION OF PUERTO RICO
I. Description and integration of the
by Pierre Dansereau
II. Analysis and mapping of the
by Peter F. Buell and Pierre Dansereau
University of Puerto Rico
Faculty of Arts and Sciences
INSTITUTE OF CARIBBEAN SCIENCE
SPECIAL PUBLICATION No. 1
Mayag'uez, Puerto Rico
/ SAN JUAN,
-.PTA VACIA TALEGA
ANevares Laguna de
SHaystack Piffones a
Rio delay Plat El YunqueALuquil lo
Rio de laMte ritton
S n de oan Cristobal Rio Blanco
STUDIES ON THE VEGETATION OF PUERTO RICO
Explanation of Contents
Map of Puerto Rico (with place names) . . Frontispiece
Studies on the vegetation of Puerto Rico. I. Description and
integration of the plant-communities. Bg Pierre Dansereau. . 3
Introduction . . . . . . . 3
Ia. Littoral Subzone . . . . . 12
I. Lowland Rainforest Zone . . . . 20
II. Seasonal Evergreen Forest Zone . . . .. 26
IIa. Hill Scrub Subzone . . . . . .. 30
III. Semi-deciduous Forest . . . . .. .31
IV. Lower Montane Rainforest Zone . . . . 35
V. Montane Forest Zone . . . . . .. 40
VI. Montane Scrub Zone . . . . . .. 44
Conclusion . . . . . . 45
(Figure 1 presents four transects across Map 1 (zonal
division of the Island). Figures 2 to 7 are line-
drawings indicating relative positions in typical
regional (zonal) landscapes of the vegetation-types
and/or plant-communities described in the text and
enumerated in the Tables.)
Studies on the vegetation of Puerto Rico. II. Analysis and
mapping of the Roosevelt Roads area. By Peter F. Buell and
Pierre Dansereau. . . . . . . 6
Literature Cited in Parts I and II . . . . 56
Appendix A: Terminology . . . . . ... 63
Appendix B: Map errata . . . . . 6
Master-Tables A to J. . . . . . 65
(Keys to the units and symbols used in the Text,
Diagrams, Maps, Figures, Photos, and Tables I-XIV.)
Tables I to VII. . . . . . . 89
(Regional analyses of vegetation-types and their
corresponding site-features: landform, habitat,
Table VIII. . . . . . . .. 100
(All of the species of vascular plants mentioned
in the Text and found in the Relev6s (Tables IX
to XIV) are graded according to the criteria in the
Tables IX to XIV . . . . . . .. 12
(Phytosociological charts which record the Relev6s
used in the construction of Diagrams.)
Photographs 1 to 118. . . . . . ... 180
(In most instances the legend includes (in paren-
theses) the number of the plant-community or vege-
tation-type listed in Master-Table D. In some ex-
amples the exact quadrat where the phytosociological
Releve was made is shown (Tables IX-XIV), and the
number of the corresponding Diagram is given.)
Diagrams 1 to 73 . . .......................... * 212
(Each one was plotted from a Releve bearing the
same number and which is given in full in Tables
IX to XIV. They are drawn according to Dansereau's
(1958) method as modified by Dansereau, Buell, and
Dagon (1966) and shown in Master-Table A. The name
of the community, its corresponding number, and the
related photo references are listed in Master-Table D.)
Maps 2 to 7 (in envelope) show various features and interpretations
of the vegetation of the Roosevelt Roads area.
STUDIES ON THE VEGETATION OF PUERTO RICO
I. DESCRIPTION AND INTEGRATION OF
The New York Botanical Garden
The plant-cover of the Island of Puerto Rico is now entirely
controlled by man. There is no such thing as virgin vegetation, whether
forest, savana, scrub, or grassland. The various "forest reserves" (some
of which are not really or mostly forested?/), however, contain stands that
have never been actually clear-cut and/or have been free of serious inter-
ference for a long enough time to convey to the contemporary observer some
idea of their original composition and quite possibly of their primeval
structure as well.
The present series of papers is offered in part as an experiment
in description and interpretation. I had devised, in 1951, a method of
recording vegetation which has proven to be applicable very widely and which
I have been encouraged to develop and amplify. The Waterways Experiment
Station (Vicksburg, Mississippi) of the United States Army Corps of Engineers
had given me a first contract (DA-22-079-eng-208) in 1957-58 which resulted
in the publication, in 1958, of a revised scheme. A second contract (DA-22-
079-eng-332) extending from 1962 to 1965 was focused on Puerto Rico. Mean-
while, a number of 'field teams, some of them directed by the Waterways
Experiment Station itself, and some of them under contract, applied and
variously modified my 1958 system. A meeting held in Vicksburg in April
1963 brought these various groups together and a summary of the proceedings
was published (U. S. Army Engineer Waterways Experiment Station 1963).
In the years 1958-1965, on the other hand, I have given my best
efforts to the development of a more broadly based system of functional
analysis of vegetation. A preliminary application of this multidimensional
characterization was published in 1961 and a fuller account will soon be
Head of the Department of Ecology at The New York Botanicl Garden,
and Adjunct Professor of Botany and of Geography at Columbia University,
2 See definition of Forest in Master-Table C, page 67.
My 1958 system (as shown in Master-Table A) is adhered to in the
present study for the purpose of drawing the diagrams. However, in a
separate paper, my associates and I have tried out a number of modifications
(Dansereau, Buell, & Dagon 1966).
I am much indebted to the chief of the Area Evaluation Branch,
Dr. Warren Grabau, for innumerable suggestions and enlightening discussion
at all stages of this work. My assistant, Peter F. Buell, did much of the
field work and all of the airphoto interpretation. He and Ronald Dagon
proposed many valuable alterations. My permanent research assistant,
Virginia A. Weadock, provided much help in the compiling of data and edit-
ing the manuscript. John Marks and Ronald Dagon drew the maps and the dia-
grams, and helped in other ways. Charles C. Clare drew the figures. Linda
Marschner assisted in the compilation and preparation of the tables. Bob
Benn, chief of the San Juan area field team of the Waterways Experiment
Station, and the crews under his direction, provided transportation and
other facilities. They collected a large body of field data which, however,
are not used in these first two contributions. Roy Woodbury, of the Puerto
Rico Agricultural Experiment Station, accompanied me on most of my field
expeditions and I could have achieved very little without his identification
of species. Brother Alain Liogier also named a number of plants for us and
revisited several of our sites. I am also grateful to Dr. Gustavo Candelas,
chairman of the Biology Department at the University of Puerto Rico in Rio
Piedras, who invited me to give a course on tropical plant-formations and
ecology, and who provided office space. He and his colleagues offered many
stimulating discussions of vegetation problems.
I must also acknowledge the support given me by the National
Science Foundation (under Grant GB-2101), since the present memoir is a
contribution to a long-range project on functional analysis of vegetation.
My attempt at reconstruction of natural vegetation is based on
extensive travel to all parts of the island. A sharper focus was obtained
by making a number of precise releves, which are transcribed in Tables IX
to XIV, are analyzed in the text, and are summarized in diagrams (based on
Master-Table A). I was also able to make useful observations on the eco-
logical amplitude of many of the indicator species, and feel that I can
suggest correlations of physiographic and edaphic features with vegetation
structures (see Figures 2 to 7). I have been influenced by Gleason and
Cook's (1927) earlier study and by my own experience of analogous situations
elsewhere (southern Florida, eastern Mexico, Hawaii, the Canaries and Azores,
West Africa, coastal Brazil). My interpretation of the tendencies detectable
in vegetation change rests upon the following assumptions:
1) Quantitatively fluctuating assemblages of species having a common migra-
tional and ecological history develop segregating associations upon con-
tact with different habitats. (This is the law of association-segrega-
tion. See E. L. Braun 1935, 1950; also Dansereau 1957a, 1962.Y
2) Within an ecosystem where intensity and quantity of elements vary, there
is a relay of species-groups (and especially of dominance) which follows
a definite order. (This is the law of the continuum. See Curtis 1959,
Dansereau 1962. )
3) Man's interference very frequently results in the confrontation of
individual species (free or associated) with a historically unprece-
dented constellation of environmental features.
I have defined elsewhere and in much more detail (1956a, 1956b,
1957a, 1961a, 1962, 1965) a number of ecological laws, principles, and proe-
esaes which I believe applicable to the origin, composition, structure, and
dynamics of vegetation everywhere.
The following descriptive account is given within the framework of
six principal vegetation zones which are conditioned primarily by climate
and secondarily by either physiography or parent-material. Within each one
of them, different ecosystems are distinguished which present a quantitatively
uniform combination of resources. It is within these ecosystems that vege-
tation units are plotted and defined by their composition and structure and
related (at least tentatively) to each other in a dynamic scheme.
Such an analysis of the landscape proceeds by singling out areas
of decreasing magnitude. The largest unit is the zone, which is bioclimatic.
Following a precedent set by Egler (1950, p. 229), I have used zone for
geographic units and belt for ecological units (see Dansereau 1957a). Map 1
shows the relative location of the six geographic zones (I to VI) and the
two subzones that are given recognition herewith. The lines are largely
extrapolated from physiographic and climatic data, not from actual observation
of vegetation, as will be made amply evident below. A more detailed mapping
on a larger scale of a few selected areas is offered in subsequent contribu-
tions in this series. Table I aligns the principal characteristics of each
gone, upon which a few comments will be made. They seem to deserve recog-
nition because they occupy sizable portions of space on the island and are
readily definable as physiographic and climatic units. It need not be
assumed, however, that they are biogeographically equivalent. In fact, the
littoral subzone, the hill scrub subzone, and quite possibly the montane
scrub zone owe their distinctness to widespread but clearly edaphic conditions.
In other words, the climatic climax below 200 meters is lowland rainforest,
whereas a "littoral complex" is determined by edaphic conditions; in the
Limestone Hills the seasonal-evergreen forest is climax and the hill scrub
occupies emerging spurs. It is not equally clear, however, that the montane
scrub is edaphic, and it may be well to conserve both the montane forest and
the montane scrub as separate bioclimatic zones.
Figure 1 offers four transects across Map 1 that show the main
features of altitude and topography and the contacts between zones, as de-
fined. Subsequent figures will illustrate (on a much larger scale) the more
immediate relation of vegetation to topography.
In view of the generalized disturbance, and in view of the ex-
clusive occurrence of some landforms in each of these eight area units, it
has seemed more useful to maintain the subzones separate in devising a frame-
work for the present descriptive account.
Therefore, in the following chapters (numbered to fit the units of
Map 1, Figurel, and Table I), I will attempt to encompass the principal eco-
C' B' 40 miles
I. Lowland rainforest
Ea. Hill scrub
LT. Lower montane
-7 Montane forest
VI. Montane scrub
Vegefafion zones of Puerto Rico, gener-
alized and extrapolated from physio-
graphic and climatic features (see
Table I). Transects across the island
correspond to profiles in Figure 1.
systems within each landform and to give a preliminary description (or at
least an enumeration) of the principal communities of plants that have been
encountered (Master-Table D).
The landform is the major physiographic unit (lagoon, reef, dune)
which is definable in geomorphological terms, a combination of relief, geo-
logical features, and drainage pattern. It is sometimes sufficiently
differentiated (when observed at closer range) to allow several habitats
(or minor topographic units) to appear. The habitat (pond, shallows, flat),
in turn, generally favors the establishment of a definite ecosystem, although
it may allow more than one to emerge. The ecosystem is the conjunction of
living and non-living elements which, on a given site, operate the con-
version cycles of environmental resources. It often comprises several belts
(not zones!) each one of which is occupied by a distinct community. Com-
munities, in turn, are definable in both structural and floristic terms and
only incompletely so in the one or the other.3/ Moreover, they are usually
related in a dynamic fashion.
Each ecosystem (and therefore all of the communities which it
harbors) is characterized by a regime which is defined according to the kind
of physiological stress that prevails. This has been expressed in a variety
of ways and has led to a rather confused vocabulary. I shall adhere to a
scheme originally formulated by Huguet del Villar (1929) before the emergence
of the theory of the ecosystem, and which I have redefined (1952, 1956b,
1957a, 1961b, 1964). Master-Table B offers a key and an explanation of this
classification and also some examples of typical corresponding ecosystems.
A given regime may extend to several habitats, whether they are contiguous
or not. The regime itself, therefore, is conditioned by a predominant
physiological aptitude which is required of the living members of the eco-
system and which marks their ability to utilize its resources by meeting the
particular stress that it offers (for instance, excessive acidity, cold,
hardness or looseness of substratum, etc.).
Finally, the plant-community (whatever the method whereby it is
defined) is a more or less cohesive group of species, mutually interacting
and characterized by various features of composition, structure, and dynamics.
There are often several plant-communities in a single ecosystem; some may
occur in several ecosystems. In fact, the very object of phytosociological
research greatly concerns the power shown by the community (or by communities
that have compositional, structural, or dynamic features in common) to trans-
gress the barriers of: ecosystem, habitat, landform, zone!
In the present account it is not always possible to refer to actual
plant associations (sensu Braun-Blanquet 1932 and/or sensu Dansereau 1957a,
19^9), which are definable by a large sampling of stands. They would form
the basic unit, of course, but such an inventory is not available and
often the broader and less-defined vegetation-type will be described, in
This has been quite thoroughly discussed elsewhere (Dansereau 1957a,
1961a). It is of special importance in the present context where structure
is going to be emphasized.
which one or more associations may be involved. These units are recorded in
Master-Table D and in the various tables given for each zone. They are also
sketched in on Figures 2 to 7.
In order to insure the clarity and legibility of the present texts
and of the rather considerable illustrative materials, a liminary Explanation
of Contents and Appendix A lend themselves to ready consultation for cross-
reference and definition of terms.
Regional patterns of vegetation, or vegetation zones, are deter-
mined primarily by climatic conditions, which in turn are due to the various
interference of airmasses and physiography. Lobeck's map and study (1922)
and subsequent interpretations by Pic6 (1954, 1962, 1963) and Young (1955)
provide us with an adequate description of the relief and of the landforms.
Likewise, Roberts (1942), Thorp (1941), and Pic6 (1954, 1962) have mapped
and summarized the climatic features. Climatic data and diagrams will not
be presented in this report.
The interactions of airmasses (climatic regimes), of physiography,
and Tertiary migrations of flora have been studied by Beard (1944) who has
presented a unified scheme for tropical America. Asprey and Robbins (1953)
have applied this to Jamaica. It can be said that Puerto Rico repeats the
Jamaican patterns (see their Figures 38 and 40) in a somewhat depauperate
form. On the other hand, Stehl6 (1945, 1946) has emphasized the Lesser
Antilles and made many useful edaphic and climatic correlations which it
will be worth testing further by means of functional analyses such as are
provided by the data offered in the present contribution.
Map 1 gives a much simplified account of the six zones and two sub-
zones of vegetation whose principal characteristics are listed in Table I.
Extrapolated from physiographic and climatic data, the zones correspond, on
the whole, fairly well with most of Pic6's (1962) natural areas, and with
Thornthwaite's climatic types (Thorp 1941). According to this latter defi-
nition, there is almost no arid land and not very much semi-arid either.
The greater part of the driest zone is termed subhumid.
Earlier attempts at mapping of vegetation (Murphy 1916, Pic6 1954)
provide very little help, for they are even sketchier than the present one.
I am afraid I cannot make much, either of Kumme and Briscoe's (1963) appli-
cation of the Holdridge system to the "forests of Puerto Rico."4/ On the other
hand, Little and Wadsworth's (1964) map is based on the same concepts that
have guided me here, and it will be seen that many of our boundaries between
zones are nearly coincident. The "Common trees of Puerto Rico" is based on
many years of field experience and incorporates a full knowledge of forest
types as revealed by foresters' studies, a few of which have been published
in the "Caribbean Forester."
The present contribution is a survey and establishes the framework
within which more detailed studies can be integrated. The following figures,
photographs, and tables will set the stage. The vegetation of each zone will
I should have much to say in both praise and criticism of Holdridge's
definitions and mapping. Sawyer and Lindsey's (1964) recent application of
these norms to Eastern North America certainly point to a lack of equivalence
between bioclimatic zones and vegetation zones. It is reassuring that Hold-
ridge (1965) now calls his areas "life-zones."
S 000 ME TE PS
FIGURE 1. Four transects across the Island as plotted on MAP 1 (which see for explanation of vegetation symbols
A '1000 METERS
C A C'
therefore be described in the order in which it appears in Table I. The
description of the vegetation-types and/or plant-communities will follow the
order indicated in the table which lists the site-characteristics.
The landscape of each zone will be represented, in a figure, by
several transects that show the usual contiguities of communities. The
numbers used are those listed in Master-Table D which carries references to
photos and structure-diagrams. Similarly, each one of the regional (or zonal)
tables (II to VII) will list all of the communities (by their names and
numbers) to be found in the landscape under specified landform, regime (as in
Master-Table B), habitat, and local ecosystem.
The scientific names of plants, with very few exceptions, are
those of Britton and Wilson (1923-30). It has not been possible to incor-
porate all of the numerous changes recently proposed by Liogier (1965).
Table VIII lists all the species of Puerto Rico that are mentioned
in the text and tables, and it grades them according to the various criteria
which serve the purpose of functional analysis and especially of structural
The following descriptions, in most instances, will not be very
detailed so far as floristic notations are concerned. However, since our
field work consisted principally of phytosociological releves made according
to the method defined in my "Biogeography, an ecological perspective" (1957a)
and exemplified in the Appendix of that book, these data have been assembled
in Tables IX to XIV. They have served as the base for the diagrams that bear
corresponding numbers. They have been grouped under the following floristic
A. Mangrove complex (Table IX): principally Rhizophora, Avicennia, and
Laguncularia, but also many saltmarsh species. These are tropical
Atlantic species, very widely distributed.
B. Tropical beach-and-strand complex (Table X). Most of the species in
this group are pan-tropical, whether truly indigenous throughout the
Tropics of Pacific and Indian as well as Atlantic shores, or second-
arily distributed. Such are Ipomoea pes-caprae, Cocos nucifera.
C. Lowland marsh complex (Table XI). This contains a majority of pan-
D. Hygrophytic complex (Table XII). Many of these are strictly American
tropical plants, some of them endemic to the Caribbean, to the Antilles,
or even to Puerto Rico. The species range from the shade-loving, high-
requirement species of the original rainforest to the ecologically un-
stable second-growth and jungle species.
E. Subxerophytic complex (Table XIII). The woodlands, savanas, and scrubs
of the Caribbean have possibly developed an even more characteristic
flora than the rainforest: Bursera simaruba, Plumiera alba are good
examples. There is rather a broad range, in fact, from the seasonal
evergreens characteristic of Zone II to the semi-deciduous plants of
Zone III, all of which are part of the same floristic group.
F. Montane complex (Table XIV). The isolated Caribbean highlands harbor
a flora of their own, ranging from genera of tropical affinity (Eugenia,
Tabebuia) to widespread temperate-rainforest types (Ilex, Rapanea, Podo-
carpus, and possibly Euterpe), but no boreal ones!
G. Tropical exotic complex. This comprises a number of invaders from other
lands that show various degrees of naturalization, from the planted
Eucalyptus globulus (from Tasmania) to the subspontaneous Spathodea
campanulata (from Central and South Africa) and the forest-invading
Erythrina poeppigiana (from Peru).
Some further discussion of these "complexes" will be given in the
course of the description of communities. But it has been found useful to
group the releves together under the above headings, whereas the individual
species have all been graded (in Table VIII) according to floristic element
Throughout the following account, therefore, cross-references will
constantly be made to the following:
community: as listed in Master-Table D where essential facts appear con-
cerning composition and structure (formation-type as defined in Master-
floristic composition (in some instances) is given for one or more stands
in Tables IX to XIV; and these field data, in turn, have served to
structure, as recorded in the diagrams (constructed on the basis of Master-
photos of most of the communities (or vegetation-types) show their par-
ticular position in the landscape which, for each zone, is schematized
figures (2 to 7) that present typical profiles in each one of the zones that
Map 1 (and on Figure 1); each community bears the number given to it in
Ia. LITTORAL SUBZONE
Map 1; Figures 1, 2, 5; Tables I, II, IX, X;
This area, which lies near the sea and is variously prolonged
inland in the flatter areas, is under the influence of saltwater, saltspray,
and sea-winds. It comprises the climatically driest portion of the Island
(near Guanica, in the southwest) but also extends to very rainy territory.
Not too many littoral communities seem to be affected by this difference in
climate, although some variants are to be observed and also a difference in
rate of development and replacement. Therefore, it is in contact, on its
upper limits (see Fig. 1), mostly with Zone I, but also with Zone III, and
occasionally with Zone II (see Photo 81). Figure 2 and Table II show the
following major divisions: open sea, lagoon, reef, intertidal and supratidal
belts, dunes, and estuaries. These are in turn under the impact of six eco-
systematic regimes (seven if one considers man's influence). I propose to
use these as the framework within which to describe the plant-communities.
Saltwater ecosystems (halohydrophytia)
These consist of the open sea (Atlantic or Caribbean), of partly
submerged coral reefs, either isolated or forming fringes that enclose
shallows, and of lagoons (embraced or closed by strands). The vegetation of
the sea, of the reef talus (or scarp), and of the permanently submerged part
of the lagoon will not be considered here: it consists almost exclusively
of algae. Its communities can rarely be accounted for without considerable
attention to the animal components, especially on rocky shores (see Dansereau
I have no knowledge at all of this marine vegetationE5/ and will
make no attempt to include it in the present account, except to indicate its
points of contact with palustrine formations. The kelp tangle and the sea-
lettuce meadow do not have such contact. But the turtle-grass meadows that
occupy the shallows do. The somewhat succulent linear-leaved, grasslike
Thalassia testudinum grows in dense stands especially where a layer of sand
covers the underlying rock. These plants have vigorous vegetative propaga-
tion. They are periodically uprooted by the sea currents or exceptionally
high tides and deposited in great masses on the beach, on an otherwise
sterile strip. Here they can offer conditions of germination and ephemeral
growth to such halophytes as Cakile lanceolata which, however, does not
become established. Moreover, these matted debris dry up and are blown
inward to the upper beach where they increase the organic content of the
soil. This increment of organic (and mineral) elements can be of consider-
able importance and it would seem to operate in a variety of conditions
(e.g., in the Arctic: see Dansereau 1954).
SThis is being studied excellently by Diaz Piferrer and his collabo-
SEA INTERTIDAL BELT SUPRATIDAL BELT UPLAND 24
REEF (ROCKY) (ROC KY) 224
L AG OON ,,- ,i 13.4 20.4 OO
LAGOON 147 159 137 16 4
6.4 5.4 6.4 14.8 14.6 15.4 V/
1.54.0 2.6 46 V4.6 V Vvv v .
halo-" halo- halo- helohalophytia petrohalophytia V psammohalophyia paranth
hydro- phytia hydro- phyt
SHOAL (SANDY) (SANDY) DUNE UPLAND 4
20.4 4 3.7 (43.1
19.6 21.9 18.5 20.4
18.7 17.6 .\Il 22.4 26.4
17.8 10.6 17.6 17.5 oo.: .CD
halohydrophytia helohalophyfia i psammohalophytia meso- h
0 phytia i
0I 4 0
REEF (SILTY) (SILTY) ESTUARY
5.1 6.1 7.1 6.1 34.5
4.0 7.4 4.0 3.6 96 12.5 7.4 12.5 29.6 30.5
halohydrophyfia helohalophytia helophyia limno- hlo
FIGURE 2. Littoral subzone (la), showing landform, habitat, and plant-cormmunities, listed by numbers
corresponding to MASTER-TABLE D and to TABLE II. (See also FIGURE 6.)
Tidal saltwater ecosystems (helohalophytia)
I cannot offer a full account of the complex donation of the tidal
area, at this time, but will outline the principal formations' response to
Whereas the sandy beaches are quite sterile in the intertidal zone,
the rocky shores, especially where the surf beats strongly, are covered, at
that level, with Fucus and especially with Turbinaria turbinata. These
fucoid beds occupy the lower part of the intertidal belt (Photos 1 and 2),
being exposed only a few hours, whereas the algal crust dominated by Entero-
morpha more often thrives on surf-spattered rocks higher up. Rawitscher
(194I described an analogous ecosystem in southeastern Brazil.
The reef itself, where it barely emerges above the high-tide level,
often harbors bits and pieces of mangrove. These are isolated individuals of
Rhizophora mangle, well anchored in the brittle crevices of the crumbling
coral. Where the reef platform is somewhat higher and inundated only by
storm tides, very large, old Avicennia can be found (e.g., on La Gata, Photo 15).
Colonies of Rhizophora are known to pioneer also on a sandy shore very much
exposed to wind action (Photo $).
The mangrove complex occupies various positions in the tropical
world (see map and bibliography in West 1956). In Indonesia, the Philippines,
and Colombia it reaches great spatial development. Even more interesting,
however, in a longer perspective, is the great floristic richness of the
Indo-Pacific area to which Cuatrecasas (1958b) ascribed its origin. This
same hypothesis was revived by van Steenis (1962), who listed 17 genera and
43 species in Indonesia, whereas Atlantic America has only $ genera and 10
species, and Atlantic Africa, the same $ genera with only 7 species. This
migration radiation suggests to me an application of E. L. Braun's (1935)
association-segregate mechanism. Both the lag in migration of genera and
species well suited to the mangrove and their relative inability to cross
certain climatic barriers are indicated.
In Puerto Rico mangrove covers a good deal of territory on fine
sand and silt and forms a broken ring around the island. It exhibits a
number of patterns that are determined by the following features:
1) Rhizophora mangle, Avicennia nitida,J/ Laguncularia racemosa, Conocarpus
erecta can withstand flooding by saltwater, in the order mentioned.
Therefore, here as elsewhere (in Brazil, for instance: see Dansereau 1947a),
it can be expected that Rhizophora will be nearest the sea and Laguncularia
and Conocarpus farthest inland.
2) Avicennia is more tolerant of stagnant water than either Rhizophora or
Laguncularia. This partly explains the position of the latter two at the
opposite ends of the soil-moisture gradient (where water movement is
greater) and the exclusive presence of Avicennia over large tracts of the
6 This species should now be called A. germinans (L.) L., according to
Compare (1963) and Liogier (1965).
3) There are many variations of structure in the mangrove as shown in
Diagrams 1 to 12: height to 20 meters (Photos 6 and 7), and canopy
coverage of 85 per cent, to height of 8 meters and canopy coverage of
50 per cent (Photo 9). It does not involve any other species but the
four trees (and usually not Conocarpus at that). Coconuts are regu-
larly washed in and are able to germinate in certain stands, but cannot
make much growth, even less maintain themselves. Where the mangrove
has been exploited (especially Avicennia), it tends to branch very low,
and many shrubby stands can be observed (Photo 11). (See Wadsworth 1959.)
4) In some areas, the back-waters favor a megaphorbia of Acrostichum aureum
rather than a forest, and in other areas, on the contrary, the occasion-
ally flooded higher ground tends to concentrate salt, by evaporation,
and allows the development of herbaceous flats (with Sesuviura portula-
castrum mats (Photo 19) and/or low Batis maritima scrub). Many of these
lands are reinvaded by Avicennia and a savana may prevail (Photo 17).
$) Nowhere have I seen a scrub or forest of Hibiscus tiliaceus or Thespesia
populnea, although both of these species are present in Puerto Rico.
Some mangroves lie behind a strand. Where the vegetation of the
latter is dense (a coconut or an Australian pine screen (Photo 46) or a well-
consolidated sea-grape scrub), very little windblown sand is deposited on
the emerging pneumatophores of the black-mangrove or the white-mangrove.
But where the levee is very open, the sand drifts in. It does so rather
spectacularly where the levees are exploited by man for building purposes.
If the sand ridge is low, in fact, the high tides will wash in a sheet of
beach sand, eventually a foot or more in depth and the mangrove trees will
be killed. This has been observed on a large scale at Punta Maldonado.
The literature on mangroves is extremely abundant, and the physi-
ology of mangrove plants as well as their ecology have received a good deal
of attention. Walter (1962) has brought these data together, on a world
basis, and re-focused it in a very comprehensive way. In Puerto Rico itself,
Holdridge estimated in 1940 some 16,000 acres and reported on changes induced
by man. From the silvicultural point of view, Wadsworth (1959) has compiled
the growth rates of white and black mangroves under various treatments.
Golley, Odum, and Wilson (1962) have produced a pioneering study of the
metabolism of a red mangrove forest.
Supratidal belts (halophytia)
This area is under the influence of salt spray and in unprotected
exposures it can extend inland. This is much in evidence on the hard lime-
stone slabs of the Guanica coast where salt-wind shearing affects the
scrubby vegetation many hundreds of yards from the shoreline.
Rocky shore (petrohalophytia)
Along this gradient, the occasionally surf-splashed rock has small
pockets of sand, some of it formed in situ by the weathering of the rock
(San Juan sandstone in the North, Ponce limestone in the South) or blown in
from neighboring beaches. Here, individual tufts of Euphorbia buxifolia and
Fimbristylis spadicea (Photos 2h and 25) are to be seen, as well as seedlings
of Suriana maritima, Borrichia arborescens (Photo 22), Gundlachia corymbosa,
in the form of a steppe or desert. Where more sand is available, Spartina
patens is also found (Photo 38). The community that attains the highest de-
velopment here, however, is a tight, low meadow of Fimbristylis spadicea
Higher up, on the hard rock, firmly anchored in the fissures, a
low, often succulent scrub develops, dominated by Gundlachia, Borrichia, or
Suriana (Photo 21), almost always accompanied by Conocarpus erecta. The
latter here is matted so densely as to exclude invaders (Photos 23 and 28).
It is apparently very long-lived (Photo 29).
On the south shore, above this often disturbed area, subject to
equinoctial and storm tides, the jointed and pitted slabs of limestone
harbor a desert-like spiny and succulent vegetation, beginning with such
low plants as Cactus intortus (Photo 26), and then the taller Opuntia
rubescens, Cephalocereus royeni (Photos 27 and 30), Lemaireocereus hystrix
(Photo 31), and Anthacanthus spinosus (Photo 32). The shearing effect
becomes more visible where the scrub has somewhat consolidated, especially
where the button-mangrove (Conocarpus erecta) is still involved. This plant,
however, gives way to various xerophytic shrubs, mostly to the sea-grape.
Photo 23 illustrates the salt-sedge to Conocarpus to Coccoloba gradation.
At the same level on the north shore, there is most likely to be
a strong overlapping of sand, and the beach communities take over.
Sandy shore (psammophytia)
This sector comprises a beach and a levee habitat. The outer beach
is sterile, as a rule, under the combined impacts of equinoctial and other
very high tides, storms, and winds (Photos 33, 3h, hO, and h6). The inner
beach commonly exhibits two belts. Nearest the ocean, the pioneering streamers
of Sporobolus virginicus, Paspalum vaginatum, or Spartina patens extend pre-
cariously, sometimes accompanied by tufts of the annual Cakile lanceolata
(Photos 3h and 46), or else the latter alone is present. The less often
disturbed upper part also frequently receives and holds vast quantities of
driftwood (Photo 39). Here the three grasses and the crucifer mentioned
above may form a lush meadow or prairie which is reinforced by two vines,
Ipomoea pes-caprae (Photo 34) and Canavalia maritima. A further step is the
establishment of the succulent-leaved shrubs, Scaevola plumieri, Suriana
maritima, Borrichia arborescens (Photos 35 and 36), and an eventual invasion
by sea-grape (Photos 39 and hO).
The levee is almost everywhere planted to coconuts (Cocos nucifera)
and also to Australian pines (Casuarina equisetifolia) (Photos h and h5).
But the dominant of the primeval scrub, Coccoloba uvifera, is rarely absent.
This extremely aggressive plant is most obviously adapted to a variety of
coastal habitats, ranging from the hard surfaces of sparsely-fissured schist
or limestone (Photo 23) to the loose sand of dunes (Photo hO) where it
responds to burying and to uncovering alike. It also takes the pruning of
windshear (Photos 39 and hl). As a result, it may branch low or high, and
the size and even the texture of its leaves will vary. On the typical levee,
it is low-branching and forms a closed canopy; it grows alone or is over-
whelmingly more abundant than any other woody species. Table X has nine
releves taken in this community (Releves and Diagrams 29 to 37). The flo-
ristic and structural variations of the prevalent type of coastal scrub
show more persistence of beach plants gradually shaded out and invasion of
more sciophilous species of the snake-bark scrub as the total mass and
stratification increase. Inland, it is frequently replaced by maray-maray
(Dalbergia ecastophyllum) scrub.
In the lee of the levee, or in otherwise protected spots, a much
taller, structurally complex, and floristically rich xerophytic scrub de-
velops. This is Beard's (1944) "littoral woodland." Snake-bark (Colubrina
ferruginosa) is most conspicuous. Many lianas are present and there can be
a fair development of lower strata too. For instance, some stands have a
great deal of Bromelia pinguin. Four releves and diagrams (39 to 42) repre-
sent this association. The formation of this thicket opens the way to an
invasion by the members of what I have called above the subxerophytic flo-
ristic complex (Table XIII) which plays a major role in the consolidation of
vegetation in the semi-deciduous forest (Zone III) and the seasonal-evergreen
forest and scrub (Zone II).
The dune ecosystem (psammophytia)
Wind rather than salt is the influential factor on the dunes, which,
however, do not occur inland in Puerto Rico. They largely share the flora of
the levee (there are some high levees and some low dunes...) and some of the
communities are quite identical if somewhat less stable. Thus the sharp,
shoreward, steep, eroding face of the dune (top-heavy sand flowing down, wave-
cutting undermining) commonly has nothing but hanging shreds of one or more
of the three beach grasses (mostly Spartina as in Photo 37) and vines (mostly
Ipomoea), whereas it is capped by a tenacious Coccoloba scrub.
The interdunal areas (which are not very highly developed) will
harbor a snake-bark scrub or sometimes a Malabar-almond tall scrub which
looks spontaneous even though Terminalia catappa appears to have been intro-
duced. Beach-grass (Sporobolus virginicus) forms low, irregular stands on
both sides of the dune and can be very lush in the intervale.
The influence of man has deeply modified the littoral region. The
mangrove has been drained in many places and reclaimed for sugarcane (back-
ground of Photo 47); the strands are nearly all planted to coconut, under
which are occasionally grown sugarcane (Mayaguez area) or pineapple (Jobos
area)(Photo 45). Moreover, the ablation on beaches, levees, and dunes of
sand for construction purposes has caused mangroves to be covered by several
feet of sand and thereby destroyed (as has been mentioned above).
Estuary (limnophytia and helophytia)
Several rivers discharge into lagoons, mangroves, or breaks in the
strand. Some of them are not too badly polluted, in that ultimate sector,
to harbor several aquatic (limnophytic) and marsh (helophytic) communities
that display characteristic zonation (Photos h7 and 48).
Floating mats, sometimes precariously secured to a shallow bottom
or a protruding shore, respond to slightly different conditions: the water-
hyacinth (Eichhornia crassipes, Photo h9) is very mobile, spreads rapidly,
and is not carried away by running water so readily as the water-lettuce
(Pistia stratiotes) would be. In fact, the latter does better in backwaters
and in ponds than in rivers. The para-carib-grass mat (Photo h8) with its
rhizomatous grasses (Panicum purpurascens, Eriochloa polystachya) has to
grow outward from a bank, and waterlily pads are anchored in an organic
bottom ooze in slow-flowing water.
It is a question of how brackish the substratum of the estuarine and
other coastal marshes can be. Solid stands of cattail (Typha domingensis) and
of reed-grass (Phragmites communis) here as elsewhere can certainly endure a
small amount of salt. I know very little in this respect concerning the spike-
rushes (Eleocharis mutata, E. interstincta) of Puerto Rico. The marshes
dominated by them, which are analyzed in Table XI and Diagrams h5 and h6, are
(or were) well developed, forming rather a firm substratum- and tending to
evolve to a fen structure by the invasion of Chrysobalanus icaco (and some-
times Drepanocarpus lunatus), associated with sedges (Cladium, Scirpus, Carex)
and ferns (Blechnum occidentale). (The sawgrass, Cladium jamaicense, so im-
portant in the Florida Everglades, does not seem to dominate to any extent
here.) The icaco scrub can be very dense and extensive. It is interrupted
here and there by emergent pond-apples (Annona glabra) and button-mangrove
(Conocarpus erecta). In fact, the latter form a swamp-woodland of rather
irregular structure, riddled with pools of different depths, harboring true
aquatics (such as Pistia, Nymphaea, Utricularia) and large sedge tussocks.
Three other helophytic communities seem to lie beyond brackish in-
fluence. The riparian trompetilla-grass marsh dominated by Panicum aquaticum
and Hymenachne amplexicaulis is very similar in structure to the para-carib-
grass mat, but is probably more firmly rooted in the mud, below constantly
flowing although fluctuating water.
The rivercane brake is a gigantic grassy screen perched on the upper
edge of the floodplain: Gynerium sagittatum (a relative of the pampas-grass)
grows in very dense, canelike formations and marks off the edge of sugarcane
plantations, often alternating with planted bamboo (Photo 68).
The Pterocarpus officinalis swamp forest must once have occupied the
upper plain of many of the estuaries, at least on the N, E, and SE of Puerto
Rico. There may be no more than two stands left today. The one at Dorado is
probably the best example of its kind, with its streaming and twisting but-
tresses. This association, a well-developed forest of high dimensions,
typically lies behind the mangrove on salt-free soil. Stehle (1945, Fig. 7)
offers a good profile-diagram that also shows contact with the Acrostichum
and Drepanocarpus communities. It was very well illustrated by Gleason and
Cook (1927, Plate 17), whose study are yielded no living stand of Pterocarpus
in 1963. It is interesting to compare the structure of this swamp-forest
with one in Tahiti (Papy 1955) where the dominant Inocarpus edulis is physi-
The littoral subzone is, of course, not distinct, climatically,
from the lowland rainforest zone. It is the influence of the sea that serves
to define it: salt in air, water, and soil; offshore winds; tidal effect
(salt- and freshwater). In fact, some of its plant-communities (especially
the marshes) extend considerably inland. Many of the upland littoral vege-
tation units also show a contact with one or more of the lowland rainforest
communities. Thus Table X reveals the presence in coastal scrubs of some
of the species of both the subxerophytic complex (Table XIII) and the hygro-
phytic complex (Table XII). This may well indicate a successional sequence
as already suggested by Gleason and Cook (1927).
Transformation by man of the littoral subzone (as here defined)
1) sugarcane plantations on mangrove or marsh sites (Photos 67 to 69);
2) pineapple plantations on rich upland (Photos 63 to 65);
3) exploitation of sand and rock which are removed from strands and puntas;
4) port and marina installations, on various topographies;
5) salinas on mangrove or saltmarsh sites;
6) pasturing and horticulture;
7) stripping of riparian donation (Photo 51);
8) construction of wharves, buildings, roads, airports, gardens, cities.
I. LOWLAND RAINFOREST ZONE
Map 1; Figures 1 and 3; Tables I, III, XI;
Photos $1, 62 to 69
Even at the time (1926) of Gleason and Cook's investigation, ex-
tremely little was left of the primeval lowland rainforest. Nowadays such
stands have been depleted even more drastically. The guesses which can be
made concerning the composition and structure of the original plant-cover
of the well-drained uplands are based on ecological laws and principles
(which are themselves none too securely established: see Dansereau 1962,
1965) and on analogues in other areas.
The geological nature of the substratum, the topography, relief and
natural drainage patterns, the known levels and fluctuations of temperature
and amounts of rainfall all point to the existence, at an earlier period, of
some kind of tropical rainforest. The structure of this vegetation is most
likely to have conformed to a widespread pattern of stratification such as has
been recorded in climatically similar places elsewhere and as summarized by
Richards (1952). It would have consisted of a number of very stout and tall
trees, unbranched in the lower two-thirds, some of them possibly overtopping
the general canopy. Their shade would preclude the development of inter-
mediate layers of any great density; lianas would reach high into the canopy
and remain fairly close to the larger trees (not forming tangles in the lower
strata). Epiphytes would be abundant on the higher branches of trees, less
so lower down. There would be no palms or tree-ferns. The very thin ground
layer would consist primarily of woody plants, many of them seedlings of the
Such a "classical" structure is hardly to be observed anywhere in
Puerto Rico at the present time. At least, I have not seen any stand that
nearly approximates the above description. The fairly mature stands I have
seen (at Dorado and on the lower slopes of the Luquillo Mountains) have very
few of the really big trees and the density of their lower layers suggests
a state of unbalance or progress or "healing" which places the stand very
much on this side of climax structure.
The clues that are to be gleaned from composition provide further
insight. There is no atlas of the distribution of species in Puerto Rico.
But Britton and Wilson's (1923-30) flora, and additional unpublished herbarium
and sight records made by Roy Woodbury and Alain Liogier, might allow us to
line up a number of species that are never found (in Puerto Rico) above
approximately 30 meters. Many of these, of course, belong to the littoral
subzone (Rhizophora mangle, Acrostichum aureum, Pterocarpus officinalis,
Coccoloba uvifera, Canavalia maritima, etc.). Still others are definitely
not forest species and are confined to sunny situations, are intolerant of
shade, etc. (e.g., Piper aduncumn, Ochroma pyramidale, Casearia guianensis).
In other words, acceptable primeval upland forest indicators must be well-
defined mesophytes and sciophytes, plants that can fulfill their cycle under
sylvatic conditions, and this on a well-drained soil (which may or may not
quite show the zonal profile). Such are Manilkara bidentata, Lucuma multi-
flora, Diospyros ebenaster, Stenostemum obtusifolium, Ixora ferrea, Cassi-
pourea elliptica, Faramea occidentalis, Hernandia sonora, Petitia domingensis.
It is, of course, not excluded that some species (such as Zanthoxylum martini-
cense and Bucida buceras) which have a broader range of ecological and/or
climatic valences (including shade tolerance) could have been authentic
members of some "mixed mesophytic" complex.
This notion of a floristic cohort of species that have very closely
overlapping ecological ranges, first formulated by E. Lucy Braun in 1935, was
later elaborated by her (1950). I made an initial application of it to
tropical vegetation in Brazil (1947b) and later attempted to re-formulate it
(1956a, 1965) and to extend its application to the origin and growth of plant-
communities in general (1961a). I am convinced that this approach, if
critically followed, will yield the most valuable clues. Our improving
knowledge (Berlioz 1965) of geographic relationships in the Antilles will
soon allow us to discern the affinities of what I have called (above) the
hygrophytic complex from those of the subxerophytic complex. Aubreville
(19h9) and others have gone to some length in exploring a similar relation-
ship in Africa.
Our criteria for delimiting this geographical zone at present are
therefore largely floristic. Its boundaries will coincide with the upper
limits of a certain number of species (mostly the trees mentioned above);
with the lower limits of others (for instance Dacryodes excelsa), and this
is roughly 350 meters. To be sure, ravines, crests, scarps, nature of parent-
material make such a line waver upwards and downwards.
At all events, such a zonal delimitation based upon the presumed
extension (on average upland topography) of a climatic climax association has
more theoretical than practical value and surely no descriptive value at all,
at this time, in Puerto Rico. Other forms of vegetation now occupy virtually
all of the surface and must be considered in some detail. Figure 2 and
Table III give us our bearings in this respect. Table XII, on the other hand,
refers exclusively to Zone IV, the lower montane.
Lake and river ecosystems (limnophytia)
Rivers and lakes prolong the estuarine conditions described above.
They have floating mats of Eichhornia crassipes which are sometimes dense
enough to seal in the water surface altogether. In ponds or very still
water the duckweed (Lemna perpusilla) performs this function; also, the
water-lettuce (Pistia stratiotes). Panicum purpurascens often comes next
in the shoreward donation, forming a mat which is attached, however, to the
bank by a rooted hinge. This grass grows to a height of about 60 centimeters.
Behind this, marsh conditions prevail, rather than truly aquatic ones.
Riparian ecosystems (helophytia)
A marsh ecosystem of varying complexity is determined by the
drainage pattern of the stream or lake border. Cattail communities (Typha
domingensis, dominant) alternate with the taller reeds (Phragmites communis).
helophytia 33.5 I
J J OC(
The principal structural variations of vegetation in the Lowland Rainforest Zone (I) at present and
They commonly consist of pure populations of these species, with only a
scattering of other plants, such as Eleocharis spp., Cladium jamaicense,
Sagittaria lancifolia. On rather steep banks, the river-cane (Gynerium
sagittatum) may rise as high as 15 meters. At present it is usually seen
to occupy rather a narrow strip. However, it is evident that with only a
slight improvement in drainage the rich muck of the lowlands has been con-
verted to the cultivation of sugarcane.
The torrential streams, especially in the southwest, have periodi-
cally emersed river-gravels and flats (Photo 51) that mostly carry ephemeral
vegetation. The spider-flower (Cleome spinosa), an annual, is often the most
conspicuous member of such a community, where it grows in clumps.
Upland vegetation (hygrophytia)
Spontaneous and subspontaneous vegetation occupy a relatively small
portion of the well-drained lands below 350 meters. Figure 3 shows the prin-
cipal communities that are to be found on former tropical rainforest sites.
I have already mentioned the difficulty of describing the lowland rainforest
and even the hazards of identifying its principal floristic components, in
the absence of extensive and well-conserved relicts. Lists of species given
by Stehl& (1945, 1946), Beard (1946, 1949), Asprey and Robbins (1953) else-
where in the Caribbean indicate the persistence of many second-growth species
in otherwise well-structured stands. It follows that it is also difficult
to distinguish it from the more mature rainforests with which it is in con-
tact, namely the seasonal-evergreen and the lowland montane, for both of
which I have more data (see Table XII and Relev6s 50-53 in Table XIII). The
more mature second-growth has also been recorded better in Zones II and IV
(which see). Virtually all forest stands in the lowlands are second-growth,
where the presence of such species as trumpet-wood (Cecropia peltata), Didymo-
panax morototoni, Casearia spp., Andira inermis, testify to an opening-up of
the canopy without which these species are not likely to have germinated and
to have gone through their early growth stages. Some of the stands, of course,
may never have been clear-cut and they testify to the "healing" of a disturbed
rainforest. There are certainly many types of forest in this category, re-
sulting from various kinds of interference, all the way from the natural, and
more or less cyclical, effect of hurricanes to selective cutting and planned
Jungle, in the proper sense of the term, is the result of either
degradation or recovery, probably more often of the former than the latter.
It is vegetation of difficult access_/ because of its encumbrances in the
lower layers. This is mostly caused by woody and/or herbaceous vines. Much
jungle is of shrub height or consists of scattered trees overtopping tangled
shrubs and themselves draped in vines, forming a savana or eventually a wood-
SIt is quite remarkable that the etymology and folklore of vegetation
terms repeatedly lead to this anthropocentric and peregrine origin of the
meaning of words. Tundra, for instance, means "hostile" land, as does also
land. Very rarely, if a true forest structure develops, does the lianescent
element retain much prominence.
There are not too many examples of this vegetation type in the
Puertorican lowlands: for the most part, either the forest is exploited as
such or is cleared for agricultural purposes. The most conspicuous jungles
lie at the foot of the limestone hills, in the north, in valleys and ravines
on shallow soils. (These areas will be described below: see Zone II.)
Controlled vegetation (paranthropophytia)
It seems useful to distinguish three levels of human interference:
1) managed land, which involves old fields in various stages of subspontaneous
or spontaneous revegetation on the one hand and pastures on the other; 2)
cultivated land, which is geared more or less intensively to production of crops;
3) occupied land, which is not strictly productive but is used for gardens,
dwellings, transport, recreation, commerce, etc._/
1) Managed land. The scrubs and savanas are mostly derived from
agricultural land. These communities are better developed in the lower montane
zone than in the lowlands and will be considered in a later section. It may
be worth noting, however, that the guava scrub is probably more aggressive
and successful in this (warmer) zone.
As for the pastures, it is not impossible that some of the succes-
sional grasslands near sea-level were developed subsequently to the downward
migration of certain species such as Andropogon bicornis. Garcia Molinari
(1952) has given an account of many of the introductions that have been attempted
and of their relative success. As Figure 3 illustrates very schematically, a
fairly dense low sward of St. Augustine-grass (Stenotaphrum secundatum) occupies
deep soils or rolling topography, with rather a high water table, whereas the
carpet-grass (Axonopus compressus) meadow is in shallow depressions. The wiry-
beardgrass (Andropogon gracilis) dominates on dry limestone ridges and crests.
The cerillo-grass (Sporobolus indicus) is a very coarse plant of low require-
ments and is quick to invade even somewhat salt soils.
2) Cultivated land. This occupies by far most of the surface. The
principal crops are sugarcane (Photo 68), pineapple (Photos 63-65), and coconuts
(Photo hh); there are also orchards (mango, citrus, bananas, plantains). Pico
(1963), recently reviewing his earlier land-use and crop distribution maps,
shows us the extent of these principal crops, and their regional correlations.
The vegetation structures of croplands undergo a number of important
variations. Whereas the orchards maintain their savana structure over long
periods of time and the coconut groves their woodland or forest formation-type,
sugarcane and pineapple do not. Photo 69 shows the conspicuous differences at
8 There is, of course, a larger gamut of human interference (see Dansereau
1997a, Chapter V, and Dansereau 1965), but this is not involved in the present
discussion. Obviously these criteria are paramount in land-use classification
and have been fully applied by Pic6 (195h, 1962, 1963).
harvest time between a tall prairie of cane in the alluvial fields of the
northeast, and Photo 70 the low prairie structure on the drier calcareous
hills of the southwest. Likewise, pineapple starts with a steppe-like
structure, which it maintains for some time, and ends up as a low prairie.
Photo 45 shows pineapple grown under coconut, and thereby forming a herb
layer in a forest.
3) Occupied land. Roadside and garden vegetation mostly involve
exotics. Some trees are regularly planted in rows along the highways and
streets. The most common are: Spathodea campanulata, Albizzia lebbeck and
A.procera, Poinciana regia, Gliricidia sepium. Others are rather scattered
or even quite local, such as Artocarpus incisa, Ochroma pyramidale, Montezuma
speciosissima, Adenanthera pavonina, Bucida buceras, Ceiba pentandra, Euca-
lyptus globulus, Cordia sulcata, Ficus nitida.
Wolcott (1945), considering the various species that do well in
this habitat, warned against some of the entomological perils involved, such
as the pink bollworm of cotton (Pectinophora gossypiella) which breeds on
Montezuma speciosissima, a Puertorican endemic.
II. SEASONAL-EVERGREEN FOREST ZONE
Map 1; Figures 1 and h; Tables I, IV, XIII;
Photos 54, 55, 57-59, 72, 74, 78
The range of limestone hills that extends from the extreme west
almost to the extreme east of the Island, in the north, is in part immersed
in the preceding Zone I (Map 1) or else it comes in contact with the foothills
of the Central Range and even with Zone IV to the south. It also extends, in
places, all the way to the coast and contacts Zone la, as shown in Figure 1.
Rivers cut through the range and are bordered by alluvial plains (Photo 66)
and by gentle slopes that accumulate the eroding materials. The altitudes of
the hilltops (Zone IIa, which see) are not very great (at most, and quite
exceptionally, 600 meters), so that there are no appreciably low temperatures.
The precipitation throughout runs quite high, but decreases somewhat to the
The differences in vegetation (at the same altitudes) seem princi-
pally due to topographic and edaphic features. There are two principal types
of topography depending upon the extent of the ridge-tops: the sugarloaf
type (or "mogote") is compact, has a domed plateau, few spurs and exposed
surfaces; the haystack type (or "pepino"), on the contrary (Photos 54, 80),
offers many exposures. The lesser the asperities, the more dense is the vege-
tation and the higher it rises on the hills. However, some coastal areas
(Photos 81, 82) and a few of the interior uplands harbor quite a bit of scrub.
This geographical region is characterized by the periodicity of its
vegetation. It may seem that the terms "seasonal-evergreen" (used for Zone II)
and "semi-deciduous" (used for Zone III in the southwestern region) are six of
one and half-a-dozen of the other. They correspond respectively to Beard's
(1944) "evergreen-seasonal forest" and to his "semi-evergreen seasonal forest";
and also probably to Asprey and Robbins' (1953) "wet limestone forest" and
"dry limestone scrub forest." Inasmuch as some of the most important tree
species are common to both (for instance, Bursera simaruba and Bucida'buceras)
such an impression is reinforced, and the same "subxerophytic complex" can be
assumed to provide the raw materials for these closely related vegetations.
Actually, a difference in emphasis is implied which is supported by other
characteristics of the vegetation.
An assignment of Puertorican tree species to the three categories
"deciduous," "semi-deciduous," and "evergreen" (Dansereau 1951, 1958, and
Master-Table A) is not very satisfactory. It seems, for instance (fide Frank
Wadsworth and Roy Woodbury), that dormancy and leaf-shedding in a number of
woody species responds to short-day periods. Some species, such as Plumiera
alba, Bursera simaruba, and Poinciana regia, lose their leaves whether or not
precipitation falls below a critical level. Others, such as Bucida buceras,
shed them sooner or later and more or less completely (or even not at all!)
according to the intensity of precipitation during the short-day (winter)
period. Beard's (1942) study in Trinidad makes these essential physiological
distinctions and uses them to characterize his evergreen semi-monsoon forest
which, maybe, is similar to the gateado forest of Puerto Rico. Needless to
say, we do not have systematic records for many species of this potential vs.
actual periodicity. Neither Britton and Wilson (1923-30) nor Little and
Wadsworth (1964) provide information beyond the statement of whether the tree
is deciduous or evergreen.
Whatever the cause of deciduousness, I have assigned a category to
each species listed in Table VIII. It remains that the seasonal-evergreen
forest of the northern limestone hills (Zone II) is in full leaf almost all
year and does not suffer, at any time, as drastic a depletion of foliage as
can be witnessed every year in the southwest (Zone III). Table IV and
Figure 4 show the correlations of plant-communities.
Intervales and lower slopes (hygrophytia)
The lowland rainforest reaches various altitudes on the flanks of
the hills and is particularly well-developed in the ravines where they are
not too steep; it must earlier have occupied much of the intervale surface.
At this time, however, the intervales and lower slopes are cultivated to
sugarcane (Photo 66) or are occupied by pasture. The latter have been de-
scribed in the previous section.
Trumpet-wood forest (listed here under No. 43.1, and shown in Photos
54 and 57) certainly includes many plant-communities belonging to the sub-
climax of lowland rainforest, of seasonal-evergreen forest, and of lower mon-
tane forest. Cecropia peltata (Photo 56) plays an important role in all of
them. These forests contain trees (ecologically like our boreal birches and
pines) that do not readily regenerate in the shade. Acrocomia aculeata
(Photo 58), Ochroma pyramidale (Photo 55), and Montezuma speciosissima are
among them. But they also contain some of the truly sciophilous species.
Moreover, their multistoried structure is not unlike that of the climax and
does not usually involve the heavy masses in the lower layers that charac-
The latter is nowhere better represented than in this zone and in
its various structural aspects, from forest (44.1) to scrub (44.4). Persist-
ing banana plants and large Zingiberaceous forbs, a great tangle of many
lianas, woody and herbaceous, occasional clumps of bamboo, make this vegeta-
tion very difficult to penetrate and quite useless for grazing.
Lower slopes (mesoxerophytia)
On slightly more drained topography, where soil is less mature and
the surface more eroded, the lowland rainforest has not fully developed, and
a somewhat more xerophytic vegetation, the gateado forest, takes over the
whole middle portion of the hills. Where they are more or less flat-topped
and of the "sugarloaf" type, it can extend to the top. The largest indi-
viduals of Bucida buceras are to be encountered there. Other abundant trees
42.1 ^- '
I N T E R V A L E AND S L O P E
pa ran t h r op op h y i a
FIGURE h. Vegetation of the Limestone-hill Seasonal-evergreen Forest Zone (II) (and the Hill-scrub'.
are Coccoloba laurifolia and C. diversifolia, Quararibaea turbinata, Torrubia
fragrans, Zanthoxylum martinicense (Photo 72). There are no tree-ferns, but
quite a few terrestrial and epiphytic ones: Blechnum occidentale, Tectaria
heracleifolia, Cyclopeltis semicordata, Adiantum tenerum. Lianas and epi-
phytes are not at all conspicuous, and play a very minor role. The trees are
straight-trunked, very infrequently buttressed (Ficus laevigata). The leaf-
litter is not so rapidly decomposed as in the lowland rainforest and can form
a more or less persistent mulch.
This forest, of course, has been very heavily damaged and the dis-
proportion in diameter between some very large surviving individuals of
Bucida buceras and Hernandia sonora and the associated trees testify to their
relict status. 9/
Abandoned fields (paranthropophytia)
Cut-over land, once pastured and/or plowed, has often reverted to
firebrush (Photo 62) and more often to pepper scrub which is mostly Piper
aduncum, an extremely aggressive steeple-shaped shrub (Photo 60) with slightly
drooping large leaves that catch the light, and is most conspicuous. This
serves as a nurse-shrub for the development of many taller woody plants,
indigenous (Randia aculeata, Didymopanax morototoni) or introduced (Spathodea
campanulata), that eventually form a savana. This particular line of succes-
sion does not inevitably involve masses of lianas (and therefore does not
necessarily lead to jungle formation). With the advent of Roystonea borin-
quena, Acrocomia aculeata (Photos 57, $8), Ochroma pyramidale (Photo 55),
and Cecropia peltata (Photo $6), the much more hygrophytic trumpet-wood
forest takes over.
This reminded me of a New Caledonian hill forest where Hernandia,
Garcinia, Kentia, Calophyllum, and Manilkara are all represented.
IIa. HILL SCRUB SUBZONE
Map 1; Figures 1 and 4; Tables I, XV, XII;
Photos 77, 79-82
In the upper part of the hills (and yet sometimes extending to low
altitudes, see Figure 1), the xerophytic character is more accentuated.
The somewhat shaded walls in the forest are typically hung with the
stilt-roots of potential stranglers such as the balsamfig (Clusia rosea) and
Ficus laevigata. They send downwards considerable lengths of aerial roots
that fasten themselves laterally with great solidity. Many shrubs (such as
Gesneria alba) grow in the minor ledges and pockets and sometimes a great
abundance of Anthurium acaule and/or Pitcairnia angustifolia (Figure 4 and
Photos 77, 78).
Plateaus and gentle slopes (mesoxerophytia)
The crabwood scrub (never quite a forest in height, but often quite
dense) is the highest type of vegetation to develop and shows little evidence
of being displaced. Gymnanthes lucida is often dominant. It is frequently
branched at mid-height or slightly above. Ground layers are very poorly de-
veloped. Epiphytes are not abundant.
Wherever the tops of the limestone hills are narrow enough to receive
light and be exposed to wind on all sides, and where the original forest cover
has been stripped well beyond immediate recovery, the sciophilous species,
without all being eliminated, no longer have any advantage over the heliophilous
ones. Moreover, the rapidly eroding, karstic limestone allows a complete run-
off, and drought-resistant species are favored (Photos 79, 80). Some rather
typical southwestern (Zone III) species are here: Bursera simaruba, for
instance, is a dominant of scrubs and savanas. Some indeed are only here in
the northern part of the island, such as Plumiera alba (Photo 82). Others are
on the hilltops and also in the coastal scrub (Anthacanthus spinosus). Yet
others are fairly widespread but particularly prominent in this position
(Comocladia glabra). Finally, a small group are strictly confined to this
habitat: the palm Gaussia attenuata (Photo 80) offers the most noteworthy
example. It can reach a height of 10 meters, its plumose rosette overtopping
the canopy at the end of a very slender trunk.
This subxerophytic formation approximately six meters in height,
varying from scrub to savana, is fairly open. It sometimes consists of as many
as five layers, is rich in epiphytes. This is not unlike Asprey and Robbins'
(1953) "dry limestone scrub forest."
The roble-prieto scrub (Photos 81, 82) has a greater development of
herbs, mostly grasses, and is possibly best developed in areas of contact with
the littoral subzone.
III. SEMI-DECIDUOUS FOREST
Map 1; Figures 1 and 5; Tables I, V, XIII;
Photos 70, 71, 73, 75, 76, 83-88, 100
The southwestern part of the island is very much the driest. The
prevailing vegetation has often been called a deciduous forest. It seems to
me that it more closely corresponds to Beard's (1944) "semi-evergreen sea-
sonal forest" than to his "deciduous seasonal forest."
At this time, in fact, there is rather little forest of any kind
to be seen, if forest is to be defined (as in Master-Table C) as taller than
eight meters with a canopy of more than 60 per cent coverage. As for de-
ciduousness, many patterns are to be observed (see Section II, above). The
determinants are: light-period, vegetative-sexual antagonism, drought. The
responses are: A) leaf-shed just before flowering and rapid leafing during
and after flowering (this usually coincides with the short-day period); B)
leaf-shed with or without pronounced drought during the short-day period,
with very rapid recovery of leafing activity; C) as in B but with one to
three months dormancy; D) leaf-shed never complete, but very pronounced
depletion of leaf-mass during dry period; E) complete evergreenness, but
new growth somewhat in evidence at the beginning of the rainy season. No one
has provided us, so far, with a rating of the woody species of Puerto Rico in
this respect. Britton and Wilson's flora (1923-30) offers no such informa-
tion. Nor do we have a record of systematic observations through the years
that would permit a partial unraveling of the internal-external forces
responsible for leaf-shedding. Inhabitants of the island, however, do remem-
ber the typical Guanica area as remaining green throughout some of the wetter
years and being virtually bare in the driest. In 1963 the contrast between
the rather green late-February and the much grayer late-March aspects was
quite notable. Table V and Figure 5 show the principal features of vegetation
in this zone.
Gravelly slopes and plateaus (mesoxerophytia and subxerophytia)
Bucaro forest seems to be the regional climax where a relatively
deep soil has had time to develop. It can be fairly dense (coverage to 75
per cent in the canopy). The species associated with Bucida buceras are
Savia sessiliflora, Coccoloba laurifolia, Gymnanthes lucida, Schaefferia
frutescens, Thyana portoricensis, and occasionally Bursera simaruba. Lianas
and epiphytes are not conspicuous or abundant.
This forest shows many floristic affinities with the limestone-hill
vegetation, especially with the communities of the higher and drier places.
It also has a good deal in common with the hammocks of the Florida Everglades
The bucaro does not often reach the girth or height that it attains
in the gateado forest in spite of the fact that it often forms a woodland
S67.5 68.5 69.6
ZONE Ia ZONE II
15.4a 4v v SCHIST
14.6.V vv..V, Vv .v V V--%
14.7 -. vv v Vvvv v v vVVparanthropophytia
V "V* v V hyperxerophytia /
petro-halophy+ia 64.3 ,, ",, ,"
/ ^.\ L : ',, LIMESTONE -,
16.4 64 AN1D SCHIST
16.4 V v -
v' V V V V V V' # V V
v V V V / V
v ,,v" v, ",' subxerophytia 73.
V v vv, I
VV Y VV v
hyperxerophytia 67.5 51.6 70.7
V..-.-. .hyperxerophy t i a
FIGURE c. Hillside semi-deciduous ve-etation in the Guanica re-ion 'Zone III' and art of th L"
rather than a forest. The more open the canopy the greater the role played
by gumbolimbo (Bursera simaruba). Such thinning-out of the upper layer and
increased xerophytism occurs on ridges where the soil is thinner and on
slopes where the rubble is coarser, and in areas where the original bucaro
forest has been disturbed.
Gumbolimbo savana is commonly found at mid-elevations, beyond the
coastal (or at least the marine) influence, but possibly at a level of
greater condensation of humidity, judging by the increased abundance of epi-
phytes (Photos 83, 84). Here the trees can be quite tall (to 20 meters),
but are widely spaced. The cactus Cephalocereus royeni probably reaches its
optimum under these conditions (Photo 85). A number of spiny or otherwise
xeromorphic shrubs are present: Plumiera alba, Croton lucidus, Pictetia
aculeata, Comocladia dodonaea. This savana appears to be stable, although
it is possible that it would be replaced, in time, by the Bucida forest. In
some of the valleys where a windy corridor allows even greater condensation
of moisture, it becomes a woodland and carries a heavy investment of epi-
phytes, mostly Tillandsia spp.
Limestone pavement (hyperxerophytia)
Hard, pitted, and fractured limestone (see Littoral Subzone Ia,
above) shows alternating areas of bare rock with sprigs of Krameria ixina,
Comocladia dodonaea, Croton discolor, or Cactus intortus emerging from the
small pits (Photo 26) or mats of Aristida adscensionis or of Bouteloua
heterostega and of very dense thickets of sebucan-tachuelo (Cephalocereus
royeni-Pictetia aculeata) (Photos 27, 30, 31). These thickets accumulate a
good deal of leaf litter and harbor virtually no undergrowth at all. It is
most likely in this biotope that Bursera simaruba (which eventually emerges)
gets its start. A mosaic involving those several communities occupies a good
deal of the south-facing territory (Photo 27). This can be compared with the
"thorn-woodland to cactus-scrub" complex outlined by Curtis (1947) in Haiti.
Wire-grass steppe (Uniola virgata) (Photo 86) is rather an unusual
type of grassland (Garcia-Molinari 1952). It is a bunch-grass or tussock-
grass type, high and coarse, and tends to be invaded by the surrounding
spiny scrub, forming a grassy maquis.
Screes and gravels (paranthropophytia)
Three grassland communities in this area that have structural
analogies show definite soil preferences. According to Garcia-Molinari (1952)
the lanilla meadow is restricted to the lateritic plateau and the rabo-de-
gato prairie occurs on serpentine. As for the Guinea-grass prairie, it is
compatible with a number of substrata and has a wide range in Puerto Rico,
as it has in many other tropical areas.
Flats, sandy and silty (hyperxerophytia)
The lowermost block in Figure $ shows the contact between the lit-
toral zone (la) and the extension inland of flats. At the extreme left, the
salt-grass steppe is in contact with grama-grass steppe and with other plant-
communities that could either be quite spontaneous or maintained by man.
This flat area is the domain of mesquite savana which is, however,
quite restricted. It does not extend east of Salinas and is really important
only in the Lajas Valley (Photo 88) and in the vicinity of Coqui. It requires
fine-textured soils. At least, it does not occur on the limestone pavement
or on the gravelly rubble. A study made at Jauca (Photo 87) suggests con-
siderable stability. Prosopis juliflora of all age-classes is observed, and
the grass layer (mostly of Chloris inflata) alternates with patches of semi-
woody Achyranthes indica. Other woody plants are Parkinsonia aculeata and
Capparis flexuosa. Other grasses are Andropogon annulatus and Eragrostis
plumosa. Pasturing probably maintains the present structure. In sandier
areas where woody growth could be more favored, the stands do not have such
big trees and the other thorny species, mostly Cephalocereus royeni and
Anthacanthus spinosus, tend to take over, much as they do in Haiti (Curtis
On the contrary, where the pressure of grazing is not so great,
bucaro (Bucida buceras) comes in, but I can find no indication that it would
consolidate to forest, although woodland might develop.
The complex lithology of this area, mentioned above, makes all early
stages of vegetation development potentially very different. Quite apart from
the halophytic influence (considered in Zone Ia), the limestone-schist-
serpentine-laterite-sand alternatives allow more than ten grassland communities
to develop. Prairie, steppe, and meadow are the three structures involved:
the steppes' angleton-grass, grama-grass, and wire-grass are the driest and
are highly susceptible to invasion by woody plants. The meadows may or may
not be distinct from the herbaceous layer that normally forms part of the mes-
quite savana. This is not so of the Guinea-grass but most likely is of the
parag'uito and of the rabo-de-gato, and maybe of the lanilla.
As for the cerrillo-grass, it belongs on low-fertility, depleted
sites, and it also often forms a transition from the halophytic zone.
IV. LOWER MONTANE RAINFOREST ZONE
Map 1; Figures 1 and 6; Tables I, VI, XII;
Photos 50, $2, 53, 56, 60, 61, 89-95, 99, 101
There is a great deal of forest at present in this zone, which is
very extensive (see Map 1). Most of it, of course, is second-growth or
planted. The relief varies a good deal, consisting in foothills of the
Luquillo Mountains, of the Guavate or the Maricao Highlands, and of the
Cordillera Central. (The latter is the principal coffee area: see Photos
90, 91.) The east-central region also offers a particular topography: the
"llanos y honduras" consist of very deep canyons cut into plateaus that
are level or gently rolling. (This is the tobacco country: see Photos 99,
101.) Figure 6 provides a sketch of the topographic variation and of the
Rainfore st (hygrophytia)
Tabonuco forest differs quite substantially, in structure, from
the lowland rainforest which is the "rainforest" of Beard (1944), whereas
this corresponds to his "lower montane forest." The presence of tree-ferns,
for one thing, is quite important, since they are absent in the lowlands
(and indeed not to be expected in typical tropical rainforest anywhere).
There are also rather more herbs, in greater abundance, than are found lower
down. This is Wadsworth's (1952a,b) "tabonuco type."
The releves in Table XII are not of virgin stands, which would
hardly contain such taxa as Cecropia peltata and the two species of Inga.
However, these stands have a comparatively mature structure and probably
contain most of the climax characteristics.
The climax (what we can know of it) and some of the subclimax
forests of this zone in some respects do not differ very markedly from those
of the lowland zone. However, quite a few of the lowland species (Mammea
americana, Diospyros ebenaster, for instance) do not extend above 350 meters
and are therefore absent from the lower montane rainforest; some of the
others take on rather a more important role at this level: Zanthoxylum
martinicense, Sloanea berteriana, Lucurna multiflora, Cordia sulcata; and
finally, some species make their appearance here that are not to be found
lower down, for instance, Dacryodes excelsa (Photo 89), Magnolia splendens,
Buchenavia capitata, and Cyathea arborea. A host of others could be men-
tioned also that properly belong to the montane forest and that can be found
as stragglers or in "privileged" positions within the lower montane zone:
no doubt the sierra palm (Euterpe globosa) is the most notable of these.
It is quite possible that altitudinal limits in other Antillean islands are
different, as witness the composition of tabonuco forest in Guadeloupe
(Stehl6 19L6, Tables 89-111).
Thus the development of an association-segregate (Braun 1935) out
of a rich "undifferentiated" matrix seems to follow the same pattern as in
UPLAND (Ilanos) ,
pa ran h r opophy i a
FIGURE 6. Vegetation and relief in the Lower Montane Zone (IV) and its
characteristic vegetation. Numbers of communities correlated to
MASTER-TABLE D and to TABLE VI.
Appalachia (Braun 1950) and, to reach for a nearer analogy, in southeastern
Brazil (Dansereau 1947b). Asprey and Robbins (1953, Fig. 38) indicate
just such a derivation, in Jamaica, of both "lower montane rain forest" and
"wet limestone forest" from "rain forest."
This phenomenon extends to subclimax and/or second-growth vegetation
as well. For instance, the trumpet-wood (Photos 52, 56) forest shows an ad-
mixture of the endemic royal palm (Roystonea borinquena) and of Buchenavia
capitata. It is most abundant in ravines and on the more shaded slopes
Managed land (paranthropophytia)
However, most of the forest is managed for the cultivation of coffee
(Photos 90, 91). It was found that Inga vera and Erythrina poeppigiana (which
were introduced from Peru for the purpose) provided excellent shade. These
two trees are now very abundant since they have freely naturalized. Wadsworth
(1945) reports that several other species were tried, but with a good deal
less success; he also points out the abusive exploitation of many of the
steeper slopes and suggests means of restoring the forest cover. The indige-
nous Inga laurina is also frequently found on coffee lands. And, since many
of these plots were never clear-cut, most of the primeval forest trees are
present as scattered individuals, and occasionally in good numbers: Lucuma,
Dacryodes, Sloanea, Zanthoxylum, Cordia, etc. The tree-ferns Cyathea arborea
and Hemitelia horrida are occasiona-l more frequent on the forest edge). It
is to be expected that in spite of a fairly unmodified structure the coffee
forest will have been managed mostly in its shrub layer where the coffee plants
are substituted for both native shrubs and transgressing trees.
Modifications in structure and composition of forest, here as in
other zones, can be brought about by hurricanes. Wadsworth and Englerth (1959)
have given us tables that show the susceptibility to windthrow and to breakage
of various species. In fact, they point out that depth of soil is not nearly
so important as the density of the stand itself. Other silvical characteristics
had been established earlier by Wadsworth (1952a,b), such as: maximum size,
insect and disease resistance, decay resistance, strength of wood, tolerance,
growth-rate, and ease of reproduction. These features, as estimated, were
then cumulated in the form of an index which resulted in very high economic
values for such species as Andira inermis, Ocotea leucoxylon, Byrsonima
coriacea, Linociera domingensis, Inga laurina, Cupania americana, and much
lower ones for Didymopanax morototoni, Nectandra coriacea, Alchorneopsis porto-
Abandoned land and rough pasture (paranthropophytia)
The Inga-coffee forest shows the effects of many forms of management,
overmanagement, mismanagement, neglect, and abandonment. Therefore, it ranges
from very orderly forest or woodland all the way to liana-infested jungle.
Tree-ferns, especially Cyathea arborea and Hemitelia horrida, are
the prime indicators of lower montane conditions. Released by lumbering and
clearing from their subordinate position in the forest stratification, they
cominate two rather distinct communities. The first is a tall scrub which is
quite closed at the top by their contacting crowns (Photo 96). It is a promi-
nent feature of open ravines and hillsides and appears able to maintain itself
for many years. The second is more frequent on very steep slopes that have
been stripped of forest: there the tree-ferns are isolated or in small patches,
forming a savana whose lower layers are a continuous and cascading carpet of
Dicranopteris nervosa, D. bifida, sometimes containing an admixture of Odonto-
soria aculeata and Lycopodium cernuum (Photos 97, 98). Where even this has
been destroyed, the African molasses-grass (Melinis minutiflora) often invades.
Big rosettes of Furcraea gigantea are also present.
Orderly pasture is not common, but some meadows of carpet-grass
(Axonopus compressus) are seen.
A bluestem (Andropogon bicorne) prairie can maintain itself under
light grazing (Photo 94), but will revert to scrub, involving roble-prieto
(Tabebuia heterophylla, Photo 95) or pepper (Piper aduncum, Photo 61),if over-
grazed or abandoned.
The latter, in turn, are soon invaded by Didymopanax morototoni,
Dendropanax arboreum, Casearia spp., Randia mitis, and other heliophilous tall
shrubs and small trees (Photo 60). But the introduced rose-apple or pomarrosa
(Eugenia jambos) is even more aggressive, to the point of dominance in many
places (Photos 60 and 93). As it has been used for posts and fuel, and has a
tremendous growth-rate, Wadsworth (1943) has recommended its discriminate use
under careful control.
Cultivated and occupied land (paranthropophytia)
Plantain and banana plantations are in optimum conditions at this
level and can be seen to occupy small and large areas. They probably do best
on northeast slopes, even steep ones. Individual plants or groups are seen
to persist under all kinds of conditions.
This is the tobacco zone as well, mostly in the eastern sector where
it does best on the plateaus (llanos). This high-yielding crop is managed
with much care (Photos, 99, 100, 101).
Very few if any of the species grown in orchards, gardens, along
roadsides, etc., at lower altitudes are eliminated at the lower montane level.
The "llanos y honduras" landscape marks a very sharp contrast: the
deep canyons (see Figures 1 and 6) that abruptly break these hills and pla-
teaus have vertical cliffs (with waterfalls and ponded waters).
Watercourses (limnophytia and helophytia)
The pools (limnophytia) may harbor waterlilies, but more often con-
tain a thick growth of green algae (sometimes sealed in by Lemna) and are
bordered by luxuriant Colocasia antiquorum (Photo 0SO). There is little de-
yelopment of marginal marsh, although a uyperus iS fairly abundant and stands
of cattail are well established. The riverbed vegetation is not very well
developed: such characteristic plants as Jussiaea are only occasional. It
must be added that there is virtually no unpolluted free water anywhere in
the lowland and lower-montane zones and that this alone accounts for most of
the destruction of spontaneous aquatic and palustrine communities.
The cliffs are clothed with algae and mosses under the waterfalls
or where the seepage is fairly constant. On the dry faces, thinner crusts
of algae, mosses, and lichens are formed; if the surface is somewhat pitted
and where small ledges are available the balsamfig screen (Clusia-Anthurium)
described above takes over. Clusia rosea roots can dangle down some 60 meters!
The scree that forms at the foot of the canyon cliff consists of
coarse gravel and is more or less stable due to periodical fall of new ma-
terials and rise of water level. Where it is best consolidated the tree-fern
scrub will tend to take hold and straggling woody plants, such as the rose-
apple and several of the savana shrubs and trees, will be mixed in.
V. MONTANE FOREST ZONE
Map 1; Figures 1 and 7; Tables I, VII, XIV;
Photos 96-98, 102-116
This tier is characterized by the montane floristic complex of which
the sierra palm (Euterpe globosa) is typical. It contains a number of species
not found at lower levels (such as: Cyathea portoricensis, Cyathea pubescens,
Podocarpus coriaceus, Didymopanax gleasoni, Clusia minor, Ocotea portoricensis,
Icacorea luquillensis), many of them endemic, and some of them very narrowly
so. This is Beard's (1944) montanee rain forest" which is quite widespread in
Caribbean lands. There are four areas (above 750 meters) where montane vege-
tation is developed, from east to west: Luquillo, Guavate, Toro Negro, and
Maricao (see Map 1). Table VII and Figure 7 show typical features and rela-
tionships of the vegetation.
Watercourses (limnophytia, helophytia)
The topography being rather steep, there is not much aquatic, even
less palustrine, vegetation. A few pools (some of them artificially dammed)
hold waterlily or pondweed communities. There is a more promising development
of algal and mossy growth on seeping rocks and in waterfalls.
The temperate character of this zone being very much in question, on
the grounds of climate, flora, and vegetation, one does not expect true bog to
develop, but rather fen, marsh, or swamp. The presence of Sphagnum is not
enough, of course. And there do not seem to be any species that are restricted
to this habitat. The melastome Nepsera aquatica is also in non-acid wetlands.
Altogether I have seen only one fragmentary stand of this vegetation and must
refer to it as a very dubious bog.
I have named four variants of the montane rainforest: they most
likely form a continuum that involves the entire subhygrophytic montane com-
plex, which extends to the scrub as well. Table XIV has nine releves that
fairly indicate this.
Stehl6 (19h6) has drawn attention to the prevalence of the sclero-
phyll and microphyll character of these communities. White (1963) has measured
the changes in structure and composition that take place from the lower montane
to the summit in the Luquillo sector: thus many wide-ranging species emerge
as dominants between much narrower limits. Such are, at the lower altitudes:
Calycogonium squamulosum, Micropholis garcinifolia; and at higher altitudes:
Tabebuia rigida and Ocotea spathulata.
Vegetation at the Montane Forest level (Zone V) and at the Montane Scrub level (Zone VI).
Numbers of communities correlated to MASTER-TABLE D and to TABLE VII.
Palm forest undergoes several variations (Photos 107, 112, 113, ll.,
11). In its optimum climatic conditions (in the Luquillo massif, where the
rainfall is highest), it forms pure stands (Beard's (194h) "palm brake") and
shows evidence of enough regeneration to maintain its effective. However,
it would not seem able to keep out the more tolerant broadleaved evergreens
(Clusia minor, Calycogonium squamulosum, Tabebuia rigida) (Photos 112, 115).
It is quite possible that-not unlike Pinus strobus in New England-it owes
its ecological prevalence to hurricanes. On El Yunque the palm dominates in
a canopy that rises to 15 meters. Its very straight trunk contrasts (espe-
cially on the steep mountain sides) with the arcuating stems of its associ-
ates. In the Cordillera Central it is more abundant in ravines (Photo 107);
in Guavate (which is the lowest of the montane areas) it is scattered and
marginal (Photo 89).
The broadleaved trees gain a decided advantage in the drier montane
areas of the west. This is essentially Wadsworth's (1952a,b) "Colorado type."
In Maricao, for instance, the sierra palm hardly ever attains dominance,
whereas Clusia minor usually does, whether forming a scrub (Photo 104) or a
true forest (Photos 104, 105). The presence of Podocarpus coriaceus at both
ends of the island is rather a special situation. Although this species shows
perfect vitality, reproducing abundantly wherever it grows, it does not
actually dominate and most often does not quite reach canopy size.10/
The areas of dense cloud condensation show another variant--the
"mist forest" of Asprey and Robbins (1953)-in which Tabebuia rigid is very
prominent, but where mosses play a conspicuous role. Like the moss-forests
of the Pacific (Japan, Philippines, Borneo), the low-growing trees exhibit
a definite "moss-line" consisting of a continuous sheath of bryophytes en-
veloping branches and trunk completely, and thinning out both downward and
Wadsworth and Bonnet (1951) have analyzed the soil of such a "mon-
tane thicket" and found it to differ sharply from that of the lower montane
forest: greater accumulation of organic matter near the surface and conse-
quent fuller swing of the podzolization regime account for both floristic and
structural differences. In other words, the climatically-induced edaphic
development is more important than the direct effect of increased rainfall
and exposure to wind of the woody plants themselves.
I had observed the same relatively uneasy and incomplete adaptation of
Podocarpus in the cloudforest of the Sierra Madre oriental in Mexico. Both
the Northeast-Mexican and the Puertorican podocarps are hardly to be compared
with the great trees that arise in the "bosque andino" of Colombia (see
Cuatrecasas 1958b, Plate XIV-2), a more fully developed example of this type
of vegetation. The speciation and behavior of Podocarpus is of great plant-
geographical interest since this genus holds the largest common denominator
for all temperate rainforest development in the world today (see map in
There may not be a considerable array of xerophytic and/or heliophi-
lous shrubs at this level. The ever-abundant moisture and low evaporation
allow rapid healing of the forest where it has been disturbed and the forest
species mentioned above, especially ferns (Photo 98) and palms, probably
reinvade very readily.
At least this seems to be the usual pattern in the Luquillo massif.
The Cordillera Central, however, and even more the Maricao area are more ex-
posed to drought effects, for three reasons: 1) the mass of the montane
zone is lesser; 2) the rainfall is reduced; and 3) serpentines and later-
ites are present in the westernmost sector.
I have ventured to distinguish two scrub communities. The montane
maquis (Photo 102) is quite sclerophyll (Didymopanax gleasoni) and contains
some spiny plants (not succulents). It does appear to be limited to serpen-
tine rock and possibly it can be invaded by the montane broadleaf scrub
(Photos 103, 104). The latter in turn differs structurally but not flo-
ristically from the sierra broadleaf forest of which it could be a variant
on a poorer site.
VI. MONTANE SCRUB ZONE
Map 1; Figures 1 and 7; Tables I, VII, XIV;
Photos 117, 118
At the highest levels, below the outcropping and isolated rocks,
the palm drops out almost completely and so-called elfin forest develops
(Photo 117). This formation is less than eight meters high-therefore a
scrub by my present standard (Master-Table C). Like the elfin forest-or the
"elfin woodland" of Beard (1944)-it may have a moss-line at mid-height. The
almost sarmentose trees are twisted into a very complex network.
It is tempting to consider this montane vegetation as temperate in-
stead of tropical. The simplicity of its composition, the well-marked dominance
of a few species, the importance of mosses, all liken it to woody formations in
Madeira or New Zealand. Several of the genera are widespread temperate rain-
forest characteristics: Podocarpus, Ardisia, Ocotea, Alsophila, Cyathea,
Hemitelia, Myrica, Magnolia, Ilex, Symplocos (see Dansereau 1957b). Another
feature is the accumulation of humus. One of these stands (near Cerro de Punta)
had a great depth (35 cm. or so) of fibrous organic matter, a dry peat of sorts.
The rock outcrops at the tops (where they have not been stripped of
their natural vegetation for military or other purposes) have a low, wind-
cropped, even-topped scrub, consisting of some (but not all) of the montane
scrub evergreen woody species: Micropholis garcinifolia, Eugenia borinquensis,
Tabebuia rigida, Daphnopsis philippiana (Photo 118).
The foregoing description is based upon a concept of vegetation
dynamics that gives recognition to composition and structure in plant-com-
munities and ascribes their relative position to variously weighted determi-
nants in the non-living environment. There is a resultant ordination of
vegetation units according to their magnitude: the zone (Map 1, Figure 1,
Table I) is under the influence of physiography and climate and is charac-
terized by climatic formation-classes (structurally defined). The mosaic
of vegetation-types or communities (for which a sampling of releves made in
selected localities is available in Tables IX to XIV) is assembled, within
each of the six zones, according to its fitness to the physiographic, edaphic,
and historical determinants that have created the present ecosystematic matrix.
A coordinated series of figures, tables, photos, diagrams, and
releves is commented upon in the text, which is intended as a unified descrip-
tion of Puertorican vegetation.
As for interpretation, the master-tables offer a series of graded
criteria which have been applied to all the species involved in the sampling.
Table VIII fully displays this grading, upon which the diagrams also have
been based. It will remain to make a functional analysis of the releves in
Tables IX to XIV by applying an already-tested procedure (Dansereau 1961a)
which has brought out significant differences between stands (and, by ex-
tension, between whole communities) that mere floristic differences, as such,
do not reveal.
In spite of much care for detail, there no doubt are some errors
and misapprehensions in the present account. It is hoped that they will be
pointed out and corrected. It is also hoped that the present framework will
prove useful for more detailed ecological research and experiment, such as
phytosociological surveys, and population studies of both plant and animal
A comparison also remains to be made with the vegetation of other
Caribbean areas. The excellent work of Cuatrecasas (1934), of Beard (1942,
1944, 1946, 1949, 1953), of Stehle (1945, 1946, 1947), of Asprey and Robbins
(1953) provide abundant materials. There are also valuable indications in
Verdoorn's (1945) compendium, and in numerous other papers, such as Shreve
(1914), Marie-Victorin and Leon (1942, 1944, 1956), Seifriz (1943), Howard
(1952), Lindeman (1953), Lindeman and Moolenaar (1959), Taylor (1959),
Heyligers (1963), Espinal and Montenegro (1963), Espinal (1964), Harris
Continued floristic exploration (Liogier 1965), especially where it
is geared to ecological surveys (Duke 1965), will provide new insights in the
next few years. Also, comprehensive agricultural surveys (Koenig 1953) and
silvicultural reports and plans (as published throughout the span of the Carib-
bean Forester's existence (1939-1965) by Frank Wadsworth and his collaborators)
will certainly allow a better understanding of both natural and induced vege-
STUDIES ON THE VEGETATION OF PUERTO RICO.
II. ANALYSIS AND MAPPING OF THE ROOSEVELT
Peter F. Buell and Pierre Dansereaull/
The present study of a very small area is offered principally as
an experiment. It consists in airphoto readings which are coded and trans-
ferred to a map at a scale of 1:20,000 (Map 2). A part of this has then
been enlarged to 1:6,667 (Map 3); it is this map which is further subjected
to various interpretations that concern the ecological features of the land-
The area mapped is approximately eighteen square miles (7L x 21
miles) and is noteworthy principally for the extent and diversity of its man-
grove swamps (about 20 per cent of the map area). It is situated at the
eastern end of Puerto Rico in the townships of Ceiba and Naguabo, and most of
it lies within the boundaries of Roosevelt Roads Naval Base (U. S. Navy).
Aside from the extensive tidal lands dominated by the mangroves there are
level, alluvial valleys which are predominantly under sugarcane cultivation,
and low hills (less than 150 meters above sea level) which are vegetated with
grassland (unimproved and abandoned pasture), various scrubs, and successional
stages between the two. Areas of considerable extent are taken up by military
installations and residential complexes.
The airphotos which were available to us in 1962-64 were a series
made by the U. S. Air Force in January 1951 at a scale of approximately 1:15,000.
We knew of some instances where there had been change in the vegetation and in
other aspects of the landscape, of course, but could use only the documentation
at our disposal at the time. Ultimately (in 1965) we obtained more recent air-
photos (January 1964, scale approximately 1:20,000) and were able to record
these changes (compare Maps 3 and 7).
We read our airphoto-types by concentrating on the structural fea-
tures which are defined in Master-Tables A and C. The structural character-
istics recorded in Master-Table E, therefore, concern: 1) the habit-form (as
Respectively, research assistant and head of the Department of Ecology,
The New York Botanical Garden.
in Master-Table A) of the dominant plants, with their coverage and height;
2) the formation-type (Master-Table C) of the vegetation as a whole. Only
ten such formation-types are recognized here (see Dansereau 1958).
Identification of plants to species or even to genus was not pos-
sible in all instances. Therefore, floristic data are very unevenly recorded.
For this reason, and because we do not have a complete ground-check on all
areas mapped, our correlation of airphoto-types with the vegetation-types
and/or plant-communities is not always secure and many question marks must
The site features can usually be read fairly well by combining
airphoto and topographic-map information. The substratum, therefore, can be
correlated to the landforms listed in Table II. As for the ecosystematic
regimes, they are defined in Master-Table B and have been discussed in Part I.
In the initial mapping of the area, boundaries were drawn wherever
there appeared to be vegetation differences in the photoimage. At the same
time as these boundaries were delimited, whatever structural and floristic
characteristics of the vegetation-types and their environments could be in-
terpreted from the photos (in the light of limited ground information) were
noted. Both during the progress of the mapping and after the mapping was
essentially completed, it was apparent from the initial type-descriptions
that some of the types were quite similar. The photoimages of these types
were re-examined and carefully compared, and in many instances the types were
equated. However, since this procedure was applied very conservatively, there
are probably several instances where what is essentially one type may be mapped
as two or more because variation in such factors as wind velocity and angle of
lighting may result in variations in the photoimage which are independent of
any variation in the vegetation. It is hoped that the converse situations
(two or more vegetation-types mapped as one because of their similar photo-
images) have been eliminated, but this possible error, unfortunately, is
inherent to airphoto interpretation.12/
Mapping was started on acetate overlays of the airphotos. This
method was later abandoned in favor of mapping on Xerox copies of the photos.
The procedure for this method is as follows (examples drawn from the Roosevelt
12 The basic principles and techniques of airphoto interpretation are
discussed in the American Society of Photogrammetry's Manual of Photographic
Interpretation (1960). The Society's Manual of Photogrammetry (1966) presents
a comprehensive treatment of the subject of photogrammetry.
1) Make a Xerox copy of each of the airphotos to be used in the mapping.
(Xerox seemed to be the most satisfactory for this purpose of any of the
photo-copying processes experimented with.)
2) Under a mirror stereoscope map the central region of each Xerox copy
using the copy and an adjacent photo as a stereopair.
Assume a sequence of nine-inch by nine-inch airphotos, with 60 per
cent overlap between photos and 30 per cent overlap between adjacent flight
lines, covering a portion of the map area, is numbered 149, 150, 151, etc.
(from right to left or east to west). The eastern edge of the map area is
assumed to fall approximately in the center of photo 149. Place the Xerox
copy of photo 149 under the right mirror of the stereoscope and photo 150
under the left mirror. (This gives a readable, though somewhat coarse and
blurry, stereoimage.) Map Xerox 149 from the center to about 21 inches from
the left (west) edge and to about 1l inches from the top (north) and bottom
(south). If at any point greater clarity of detail is needed for interpreta-
tion, photo 149 can be placed temporarily over Xerox 149, without disturbing
the position of the Xerox, to give a normal stereoimage.
Replace Xerox 149 with photo i49 and photo 150 with Xerox 150, and
map the right half of Xerox 150 over to the line where mapping was stopped on
Replace photo 149 with Xerox 150, place photo 151 under the left
mirror, and map the left half of Xerox 150 (see Figure 8).
Continue in this manner until the sequence has been mapped. Follow
the same procedure to map adjacent series (to the north and/or south) of photos.
Using a red pencil for the mapping helps considerably in avoiding
confusion of vegetation boundary lines with linear features on the photos and,
at a later stage in this procedure, with contours, roads, streams, etc., on
the topographic map.
3) Make photostatic enlargements (or reductions) of a topographic map of the
area to a sequence of scales which spans the range of scales found on the air-
As an example: for the Roosevelt Roads area, a topographic map at
a scale of 1:20,000 was used (U. S. Geological Survey 74 minute topographic
series). The scale on the photos ranged from about 1:14,500 to 1:17,500.
Enlargements were made to the scales of 1:14,500, 1:15,000, 1:15,500, 1:16,100,
1:16,800, and 1:17,500. A greater number of enlargements made at smaller
intervals through the range of scale variation would have given greater
accuracy. Also, to achieve a given degree of accuracy the number of topo-
graphic enlargements required increases with the relief of the map area and
the scale of the airphotos.
4) At this point it is necessary to work on a light table. Place one of the
Xerox copies (on which mapping has been done) and one of the adjacent photos
under the stereoscope as was done for the mapping (see step 2), except that
FIGURE 8. Mapping with Xerox copy and adjacent photo used as a stereopair.
1) photo 151; 2) Xerox copy of photo 150; 3) photo 150;
4) 25-watt spotlight.
Tracing vegetation boundaries from Xerox copy to topographic-map enlargement.
photo 151; 2) cardboard divider; 3) Xerox 150; 4) portion of a
topographic-map enlargement (1:17,500); $) 25-watt spotlight.
now the Xerox copy is illuminated only by the light transmitted through it
from the light table, whereas the adjacent photo is illuminated by a small
(e.g., 25 watt) spotlight. (An ordinary desk lamp may be used if a vertical
cardboard divider is placed between the photo and the Xerox copy.) (See
5) Place one of the topographic-map enlargements over the Xerox copy, and,
within a small area of reasonably uniform scale, attempt to match the contours,
streams, roads, buildings, etc., on the enlargement with the ridges, hilltops,
drainageways, streams, roads, buildings, etc., on the stereoimage. Unless a
nearly perfect match can be made, replace the enlargement with the next smaller
or larger (depending on whether the scale of the stereoimage seems smaller or
larger than the scale of the first-tried enlargement), and again attempt to
match. Repeat until that enlargement which best fits the stereoimage in the
small area under consideration is determined, and, with this enlargement in
place over the Xerox, trace onto the enlargement any vegetation boundaries
which lie in this area of fit. Select another small area, and repeat the
above procedure until all the boundaries on all the Xerox copies have been
transferred to one or another of the topographic-map enlargements.
Bear in mind throughout this operation that the scale at any point
on a vertical airphoto varies from the datum-plane scale of the photo depending
on the difference in elevation between that point and the datum plane, on the
degree of slope of the ground at that point, and on the aspect (with respect
to the principal point) of the slope at that point.
Any direct lighting on the map enlargement (overlying the Xerox copy)
tends to decrease the readability of the Xerox copy--hence the spotlight and/or
the cardboard divider. The readability of the Xerox copy is increased if the
photostats of the topographic maps are made lighter (with less contrast) than
is customary in photostating. Also, the photostats should be made on matte
or semi-matte (non-glossy) paper.
6) Reduce (or enlarge) photostatically all of the topographic-map enlargements
to the scale of the original map, thereby reducing all the vegetation boundaries
to a common scale.
7) The original topographic map is now used as a base map on which to draw the
vegetation map. Superimpose the topographic map successively on each of the
enlarged-and-reduced topographic-vegetation maps, and trace (using the light
table) the vegetation boundaries onto the base map.
At this point it will be obvious that, while the enlargements of the
topographic map need not have been made exactly to the predetermined scales,
knowing exactly how much they were enlarged facilitates their reduction to
exactly the original scale.
The resulting map is likely to have a confusing abundance of infor-
mation on it and must be given some further cartographic treatment (such as
deleting some or all of the topographic and cultural markings) in order to
produce a useful map. (See Map 2.)
The map made from the 1964 photos (Map 7) was done in much the same
manner, except that the final version at 1:20,000 scale was enlarged to 1:6,667
for ease of comparison with Map 3.
It is felt that, in the absence of sophisticated photogrammetric
equipment, this system is remarkably accurage, rapid, and inexpensive. It
would probably not be so satisfactory in areas of greater relief or in areas
where good topographic maps at a scale close to that of the airphotos were
The code-numbers for airphoto-types, which appear in Master-Table E,
have the following meaning:
First digit: formation-type (as in Master-Table C): 0 9
Second digit: strength of the prevailing ph biological stress of the ecosystem-
atic regime (see Master-Table B): 1 9j3/
Third digit: vegetation zone or subzone (see Map 1): even numbers (0, 2, 4,
6, 8) = lowland rainforest zone (I); odd numbers (1, 3, 5, 7, 9) = littoral
These code-numbers appear on Maps 2, 3, and 7, whereas they are correlated with
other information on Master-Table E.
On Maps 2, 3, and 7, heavy lines indicate a land-water boundary.
Broken lines show the approximate location of a boundary where the transition
from one vegetation-type to another takes place over an appreciable distance
and the exact point of change cannot be defined.
Maps 4, 5, and 6, on the other hand, present a visual analysis of
the airphoto-types in Map 3, according to several criteria. The mosaic will
consist, generally, of fewer pieces and of correspondingly larger areal units.
The procedure adopted here, actually, follows the model proposed earlier (see
Dansereau 1961b) in a schematic form. In that example, the basic units (plant-
associations) had not been presented in any actual spatial situation, whereas
in the present instance they are.
Map 4 shows the distribution of vegetation-types or associations as
described in Part I and as consigned in Master-Table D. A comparison with
13 This is a relative value only, and the values have comparability only
within a given ecosystem. For example: of the three vegetation types 121, 141,
and 560, the prevailing physiological stress in 141 is stronger than (but not
necessarily twice as strong as) it is in 121. However, $60 is in a different
ecosystem from 121 and 141, and therefore the stress cannot be said to be
stronger here than it is in 141 or even in 121.
Map 3 shows that several airphoto-types can sometimes be referred to a single
vegetation-type. It also points to the uncertainty of cross-reference already
in evidence on Master-Table E.
Map 5 is even more simplified than the preceding one since there
are only ten possibilities (see Master-Table C) and since, moreover, two of
the formation-types (tundra and crust) do not happen to materialize in this
particular area. This map, being exclusively structural, gives an excellent
view of differential biomasses: the vast expanse and almost unbroken conti-
nuity of scrub is remarkable, as is the continuity of forest. Woodland and
meadow, possibly because of their instability, are very patchy.
Map 6 shows the distribution of the controlling ecosystematic re-
gimes. The 25 categories listed in Master-Table B follow Huguet del Villar's
(1929) scheme. The symbols assigned to them on this map permit a number of
combinations, such as halohydrophytia, halohelophytia, psammohalophytia,
psammohelophytia, chasmohalophytia. The other regimes represented in this
map are: subxerophytia, mesoxerophytia, paranthropophytia, helophytia. This
makes a total of only nine units.
General trends in vegetation changes in the Roosevelt Roads area,
based on a comparison of the 1951 photos (Map 3) with the 196h photos (Map 7),
are as follows. The structural shifts are easily read by direct reference to
the first digit: thus 5 (prairie) is frequently replaced by h(scrub) and even
by 3 (savana), indicating a progressive succession. Area in scrub is rapidly
expanding at the expense of grassland. This generally occurs by way of a
savana-like stage and is probably due to a relatively recent discontinuance
of pasturing on the hilly land. The most rapidly expanding upland scrub com-
munity seems to be the thicket dominated by Leucaena glauca.
This leguminous plant is a relatively recent invader. In other
places throughout the Tropics it has been deliberately introduced and its
spread encouraged (Fosberg 1960). Whether or not this is so in the Roosevelt
Roads area we do not know, but it does seem to be favored, at least initially,
over the other scrub types. It generally starts in a savana-like distribution
of individuals and small clumps, usually associated with a dense growth of
vines. As the clumps consolidate into a thicket the vines greatly decrease
in importance. There are some indications that the Leucaena thickets are
later replaced by other scrub communities, but as this evidence is meager the
possibility that the thickets are stable for a long period of time and give
way slowly to a forest community cannot be discounted.
Mangrove seems to be expanding slowly seaward and consolidating its
outliers. This process is occurring more rapidly in the shallow, sheltered
lagoons and at the mouth of the principal river course. In the inner mangrove,
where the trees have been cut for firewood and poles in the past, scrubs and
savanas are rapidly becoming forest with the cessation of cutting (see type
323 on Map 3, and type 325 on Map 7, west edge). Some mangrove scrubs in the
outer mangrove belt (e.g., the extensive area in the northeast corner of Maps
3 and 7) seem to be relatively stable as such even though their edges are
expanding into the tidal flats. It may be that the less firm substrate of
the outer mangrove is the cause of this contrast.
More specific changes are as follows. In the northwest corner of
Map 3, airphoto-type 3l1 (Avicennia-Batis savana) shows a sharp boundary
with the surrounding inner mangrove forest. On the basis of photo-interpre-
tation and of field examination of types similar to this elsewhere in Puerto
Rico, it was assumed that this type had developed on an old beach ridge
which, although now surrounded by silt deposits, remained slightly higher
and sandier than the adjacent swamp. Furthermore, that as long as the dif-
ference in substrate remained, the site would continue to support this quite
different vegetation-type. However, 13 years later (Map 7), the area of
type 3l1 has diminished considerably. Part of the original 3)1 area is now
indistinguishable from the inner mangrove forest, whereas other parts of it
are woodland (type 231) which, it may be assumed, is in the process of con-
solidating into a forest.
In another place (north-central part of Maps 3 and 7), a strip of
type hh3 (Batis scrub) has become type 341. This gives us a logical succes-
sional series: Batis scrub (4h3) invaded by mangrove to become Avicennia-
Batis savana (3h7Tl, savana becoming woodland (231) by consolidation of the
mangroves and decrease in the importance of Batis, and finally, woodland be-
coming forest (121) by further consolidation of the mangroves and disappearance
of Batis. There are at least two possible explanations for this succession.
One is that these areas were formerly in mangrove which was cut. The harder,
drier substrate has retarded the natural reforestation rate in comparison to
that on the surrounding, siltier substrate more favorable to mangrove. The
other possibility is that these areas have never before been mangrove-covered
because of the unfavorable substrate; but because of gradual silting-over of
the sandy beach ridge the substrate is becoming soft enough and wet enough to
support mangroves. Considering the rate at which the change has taken place,
the former explanation seems more likely, except that in the sites examined
on the ground the lack of old mangrove stumps is somewhat perplexing.
The most striking changes in the character of the vegetation are
those resulting from man's activities. Prominent among these is the altera-
tion of the predominantly scrubby area in the southeast corner of Map 3 to a
large expanse of meadow on Map 7. This is the result of the erection of a
housing development here, and it can be assumed that the "meadow" (lawns)
will remain essentially unchanged for an indefinite period of time. On the
other hand, the topography was considerably smoothed off by bulldozers before
construction started and some of the rubble from this operation was dumped
into two small mangrove swamps (airphoto-type 121) to the south and east.
In the southern one the formation-type has been changed from forest to steppe
(or perhaps desert immediately after the filling operation), the ecosystem
has been changed from helohalophytia to paranthropophytia and is probably
approaching meso- or subxerophytia, and the community has been changed from
a relatively stable one to a rapidly changing late pioneer or consolidation
stage. Barring further disturbance this will become a scrub (probably
Leucaena), later a forest, and eventually a lowland rainforest, but it
will never revert to mangrove. This "reclamation" process is occurring
in other places in the map area.
The fill-material for some of these reclaimed areas seems to be
coming from two rather extensive quarries or "borrow pits, only parts of
which are in the northeast corner of Maps 3 and 7. There has been con-
siderable expansion of these pits between 1951 and 1964, but evidence that
in some parts they have been worked more recently than in others is evident
from the vegetation patterns. In the current and most recent workings the
ground appears virtually bare. Slightly older workings have a desert or
sparse steppe cover (type 750). In the oldest, a Leucaena savana (354)
and eventually a Leucaena thicket (352 or 354) develops. Whether or not
this will eventually give way to some other scrub type cannot be told from
the photos. (See discussion of Leucaena glauca, page 53.)
LITERATURE CITED IN PARTS I AND II
AMERICAN SOCIETY OF PHOTOGRAMMETRY, 1960. Manual of photographic interpreta-
tion. Amer. Soc. Photogrammetry, Washington, D.C., xv + 868 pp.
AMERICAN SOCIETY OF PHOTOGRAMMETRY, 1966. Manual of photogrammetry (third
edition). Amer. Soc. Photogrammetry, Falls Church, Virginia, 2 Volumes,
xx + 1192 pp.
ASPREY, G. F., and R. G. ROBBINS, 1953. The vegetation of Jamaica. Ecol.
AUBRtVILLE, A., 1949. Climats, forts et desertification de 1'Afrique tropi-
cale. Soc. Edit. Geogr., Maritimes et Colon., Paris, 351 pp.
BEARD, J. S., 1942. The use of the term "deciduous" as applied to forest
types in Trinidad, B.W.I. Empire Forestry Jour., 21(l):12-17.
BEARD, J. S., 1944. Climax vegetation in tropical America. Ecology, 25(2):
BEARD, J. S., 1946. The Mora forests of Trinidad, British West Indies.
Jour. Ecol., 33(2):173-192.
BEARD, J. S., 1949. The natural vegetation of the Windward and Leeward
Islands. Clarendon Press, Oxford, 192 pp.
BEARD, J. S., 1953. The savanna vegetation of northern tropical America.
Ecol. Monogr., 23(2):149-215.
BERLIOZ, J. (ed.), 1965. Contributions a la biogeographie des Antilles.
C. R. Soc. Biog6ogr., N0 Special (Nos. 308-313, 317-319, 323-325):1-64.
BRAUN, E. Lucy, 1935. The undifferentiated deciduous forest climax and the
association-segregate. Ecology, 16:51l4-19.
BRAUN, E. Lucy, 1950. Deciduous forests of eastern North America. The
Blakiston Co., Philadelphia, xiv + 596 pp.
BRAUN-BLANQUET, J., 1932. Plant sociology (translated by H. S. Conard and
G. D. Fuller). McGraw-Hill Book Co., Inc., New York, xviii + 439 PP.
BRITTON, N. L., and Percy WILSON, 1923-30. Botany of Porto Rico and the
Virgin Islands. Spermatophyta and Pteridophyta. New York Acad. Sci.,
Scientific Survey of Porto Rico and the Virgin Islands, 5(1-4):1-626;
COMPERE, P., 1963. The correct name of the Afro-American black mangrove.
CUATRECASAS, Jose, 1934. Observaciones geobotanicas en Colombia. Trab.
Museo Nac. Cienc. Nat., Madrid, Serie Bot., 27:1-144.
CUATRECASAS, Jose, 1958a. Introduccion al studio de los manglares. Bol.
Soc. Bot. Mexico, 23:84-98.
CUATRECASAS, Jose, 1958b. Aspectos de la vegetacion natural de Colombia.
Rev. Acad. Colomb. Cienc. Exactas, Fisicas y Nat., 10(40):221-268.
CURTIS, John T., 1947. The palo verde forest type near Gonaives, Haiti,
and its relation to the surrounding vegetation. Carib. Forester, 8(1):l-12.
CURTIS, John T., 1959. The vegetation of Wisconsin. Univ. Wis. Press,
Madison, xi + 657 pP.
DANSEREAU, Pierre, 1945. Essai de correlation sociologique entire les plants
superieures et les poissons de la beine du Lac St-Louis. Rev. Canad.
Biol., h(3):369-h17; Contrib. Inst. Biol. Univ. Montreal, 16:369-hl7.
DANSEREAU, Pierre, 1947a. Zonation et succession sur la resting de Rio de
Janeiro. I. Halosere. Rev. Canad. Biol., 6(3):448-477; Contrib.
Inst. Biol. Gen. et Zool. Univ. Montreal, 20:448-477.
DANSEREAU, Pierre, 1947b. Notas sobre a biogeografia de uma parte da Serra
do Mar. Rev. Brasil. Geogr., 9(4):h97-520.
DANSEREAU, Pierre, 1951. Description and recording of vegetation upon a
structural basis. Ecology, 32(2):172-229; Bull. Serv. Biogeogr.,
DANSEREAU, Pierre, 1952. The varieties of evolutionary opportunity.
Rev. Canad. Biol., 11(4):305-388.
DANSEREAU, Pierre, 1954. Studies on Central Baffin vegetation. I. Bray
Island. Vegetatio, 5-6:329-339.
DANSEREAU, Pierre, 1956a. Le coincement, un processus ecologique. Acta
DANSEREAU, Pierre, 1956b. Le regime climatique regional de la vegetation et
les controles &daphiques. Rev. Canad. Biol., 15(l):1-71.
DANSEREAU, Pierre, 1957a. Biogeography: an ecological perspective. The
Ronald Press Co., New York, xiii + 394 pp.
DANSEREAU, Pierre, 1957b. A preliminary note on the structure variations of
temperate rainforest. Proc. 8th Pac. Sci. Congr., IV(Botany):407-436.
DANSEREAU, Pierre, 1958. A universal system for recording vegetation.
Contrib. Inst. Bot. Univ. Montreal, 72:1-58.
DANSEREAU, Pierre, 1959. Phytogeographia laurentiana. II. The principal
plant associations of the Saint Lawrence Valley. Contrib. Inst. Bot.
Univ. Montreal, 75:1-147.
DANSEREAU, Pierre, 1961a. The origin and growth of plant communities. In:
"Growth in Living Systems," Proc. Syrup. on Growth, Purdue Univ., June 1960;
ed. by M. X. Zarrow; Basic Books, New York, pp. 567-603.
DANSEREAU, Pierre, 1961b. Essais de representation cartographique des laments
structuraux de la vegetation. In: "Methodes de la Cartographie de la
Vegetation," Colloques Internationaux du Centre National de la Recherche
Scientifique, Paris, 97:233-255.
DANSEREAU, Pierre, 1962. An application of ecological laws to woodlots.
Proc. Lockwood Conf. on Suburban Forest and Ecology, March 1962, New Haven;
Conn. Agric. Exp. Sta., Bull. 642, pp. 45-56; Sarracenia No. 7, 14 pp.
DANSEREAU, Pierre, 1964. The future of ecology. BioScience (July), 14(7):20-23.
DANSEREAU, Pierre, 1965. Ecological impact and human ecology. In: Symposium
on "The Future Environments of North America," Conservation Foundation,
New York, mimeogr. version, 77 pp.
DANSEREAU, Pierre, Peter F. BUELL, and Ronald DAGON, 1966. A universal system
for recording vegetation II. A methodological critique and an experiment.
Sarracenia No. 10, 64 pp.
DANSEREAU, Pierre, and Kornelius LWES, 1957. The grading of dispersal types
in plant communities and their ecological significance. Contrib. Inst.
Bot. Univ. Montreal, 71:1-52.
DUKE, James A., 1965. Keys for the identification of seedlings of some promi-
nent woody species in eight forest types in Puerto Rico. Ann. Missouri
Bot. Gard., 52(3):314-350.
EGLER, Frank E., 1950. Southeast saline Everglades vegetation, Florida, and
its management. Vegetatio, 3(h-5) :213-265.
ESPINAL T., Luis Sigifredo, 1964. Formaciones vegetables del Departamento de
Antioquia. Rev. Fac. Nac. Agron., 24(60):1-84.
ESPINAL T., Luis Sigifredo, y Elmo MONTENEGRO M., 1963. Formaciones vegetables
de Colombia. Republica de Colombia, Inst. Geogr. "Agustin Codazzi,"
Dept. Agrologico, 201 pp. + mapa ecol6gico.
FOSBERG, F. R., 1960. The vegetation of Micronesia. 1. General descriptions,
the vegetation of the Marianas Islands, and a detailed consideration of the
vegetation of Guam. Bull. Amer. Mus. Nat. Hist., 119(1):1-76.
GARCIA-MCLINARI, Ovidio, 1952. Grasslands and grasses of Puerto Rico. Univ.
Puerto Rico, Agric. Exp. Sta., Bull. 102, 167 pp.
GLEASON, H. A., and Mel T. COOK, 1927. Plant ecology of Porto Rico. New
York Acad. Sci., Scientific Survey of Porto Rico and the Virgin Islands,
GOLLEY, Frank, H. T. ODUM, and R. F. WILSON, 1962. The structure and metabo-
lism of a Puerto Rican red mangrove forest in May. Ecology, 43(l):9-19.
HARRIS, David R., 1965. Plants, animals, and man in the Outer Leeward Is-
lands, West Indies. An ecological study of Antigua, Barbuda, and
Anguilla. Univ. Calif. Publ. Geography, 18: ix + 164 pp.
HEYLIGERS, P. C., 1963. Vegetation and soil of a white-sand savanna in
Suriname. Koninkl. Nederl. Akad. Wetenschappen, afd. Natuurkunde,
Tweede Reeks, 54(3):1-148.
HOLDRIDGE, Leslie R., 1940. Some notes on the mangrove swamps of Puerto Rico.
Carib. Forester, l(4):19-29.
HOLDRIDGE, Leslie R., 1965. The tropics, a misunderstood ecosystem. In:
Symposium on "Human Populations in Relation to Food Supply, AIBS Ann.
Meeting, Urbana, Illinois (August 1965), 12 pp. mimeogr. + figures.
HOWARD, Richard A., 1952. The vegetation of the Grenadines, Windward Islands,
British West Indies. Contrib. Gray Herb. Harvard Univ., No. 174, 129 pp.
KOENIG, Nathan, 1953. A comprehensive agricultural program for Puerto Rico.
U. S. Dept. Agric., 299 pp.
KTJME, K. W. 0., and C. B. BRISCOE, 1963. Forest formations of Puerto Rico.
Carib. Forester, 24(2):57-66.
LINDEMAN, J. C., 1953. The vegetation of the coastal region of Suriname.
Uitgevers-Maatschappij v/h Kemink en Zoon N.V., Utrecht, x + 135 pp. + map.
LINDEMAN, J. C., and S. P. MOOLENAAR, 1959. The Vegetation of Suriname,
Vol. I, Part 2: Preliminary survey of the vegetation types of northern
Suriname. Van Eedenfonds, Amsterdam, 45 pp. + maps.
LIOGIER, Alain, 196$. Nomenclatural changes and additions to Britton and
Wilson's "Flora of Porto Rico and the Virgin Islands." Rhodora, 67(772):
LITTLE, E. L., and F. H. WADSWORTH, 1964. Common trees of Puerto Rico and
the Virgin Islands. U. S. Dept. Agric., Agric. Handbook 249, $58 pp.
LOBECK, A. K., 1922. The physiography of Porto Rico. New York Acad. Sci.,
Scientific Survey of Porto Rico and the Virgin Islands, l(4):301-384.
MARIE-VICTORIN, Frere, et Frere LEON, 19h2. Itineraires botaniques dans
l'lle de Cuba (premiere serie). Contrib. Inst. Bot. Univ. Montreal, No.
Wl, h96 pp. + map.
MARIE-VICTORIN, Frere, et Frere LEON, 1944. Itineraires botaniques dans
l'Ile de Cuba (deuxieme serie). Contrib. Inst. Bot. Univ. Montreal, No.
50, o10 pp.
MARIE-VICTORIN, Frere, et Frere LEON, 1956. Itineraires botaniques dans
1'lle de Cuba (troisieme serie). Contrib. Inst. Bot. Univ. Montreal,
No. 68, 227 pp.
MURPHY, Louis S., 1916. Forests of Porto Rico; past, present, and future,
and their physical and economic environment. U. S. Dept. Agric. Forest
Service Bull. 354:1-99.
PAPY, H. Ren6, 1955. Tahiti et les miles voisines. La vegetation des lies de
la Societ6 et de Makatea (Ocpanie Frangaise). 2e parties. Tray. Lab.
For. Toulouse, Tome V, IIe Sect., 1(3):163-386.
PIC6, Rafael, 1954. Geografia de Puerto Rico. Parte I. Geografia fisica.
Editorial Universitaria, Rio Piedras, Puerto Rico, xiii + 2)43 pp.
PIC6, Rafael, 1962. Puerto Rico: planificaci6n y acci6n. Banco Gubernam.
Fomento Puerto Rico, San Juan, xv + 312 pp.
PICO, Rafael, 1963. The commonwealth of Puerto Rico. Focus, 14(2):1-6.
RAUNKIAER, C., 1934. The life forms of plants and statistical plant geography.
Clarendon Press, Oxford, xvi + 632 pp.
RAWITSCHER, Felix K., 1944. Algumas nogoes s8bre a vegetagao do litoral
brasileiro. Bol. Assoc. Ge6gr. Bras., N0 5:13-28.
RICHARDS, P. W., 1952. The tropical rain forest. An ecological study.
Cambridge Univ. Press, xviii + 450 pp.
ROBERTS, R. C., 1942. Soil survey of Puerto Rico. U. S. Dept. Agric., Bur.
Plant Ind., Soil Survey Series 1936, No. 8, 503 pp.
SAWYER, John 0., Jr., and Alton A. LINDSEY, 1964. The Holdridge bioclimatic
formations of the Eastern and Central United States. Proc. Indiana Acad.
SEIFRIZ, William, 1943. The plant life of Cuba. Ecol. Monogr., 13:375-426.
SHREVE, Forrest, 1914. A montane rain-forest. A contribution to the physio-
logical plant geography of Jamaica. Carnegie Inst. Washington, Publ.
199, 110 pp.
STEENIS, C. G. G. J. van, 1962. The distribution of mangrove plant genera and
its significance for palaeogeography, Proc. Koninkl. Nederl. Akad.
Wetenschappen, Ser. C 65(2):164-169.
STEHLE, Henri, 1945. Forest types of the Caribbean Islands. Carib. Forester,
STEHLE, Henri, 1946. Les types forestiers des Iles Caralbes. Carib.
Forester, 7(Suppl.) :337-709.
STEHLE, Henri, 1947. La vegetation sylvatique de l'Archipel caraibe. Etude
d'tco-phytosociologie. These, Universite de Montpellier, 548 pp.
TAYLOR, B. W., 1959. Estudios ecologicos para el aprovechamiento de la
tierra en Nicaragua. Min. Econ., Inst. Fomento Nac., y ONU FAO, Vol. I,
xv + 338 pp.
THORP, James, 1941. Climate and settlement in Puerto Rico and the Hawaiian
Islands. In: Climate and Man, Yearbook of Agriculture, 1941, pp. 217-226.
U. S. ARMY ENGINEER WATERWAYS EXPERIMENT STATION (ed.), 1963. Military evalu-
ation of geographic areas, reports on activities to April 1963. Corps of
Engineers, Vicksburg, Mississippi, Misc. Paper No. 3-610, December 1963,
vi + 237 PP.
VERDOORN, Frans (ed.), 1945. Plants and plant science in Latin America.
Chronica Botanica Co., Waltham, Mass., xxxvii + 383 pp.
VILLAR, E. Huguet del, 1929. Geobotanica. Editorial Labor, Barcelona-
Buenos Aires, 339 pp.
WADSWORTH, Frank H., 1943. Pomarrosa, Jambosa Jambos (L.) Millsp. and its
place in Puerto Rico. Carib. Forester, 4(hT)Tl-19~.
WADSWORTH, Frank H., 1945. Forestry in the coffee region of Puerto Rico.
Carib. Forester, 6(2):71-75.
WADSWORTH, Frank H., 1952a. Forest management in the Luquillo Mountains, II.
Carib. Forester, 13(2):49-61.
WADSWORTH, Frank H., 1952b. Forest management in the Luquillo Mountains, III.
Carib. Forester, 13(3):93-119.
WADSWORTH, Frank H., 1959. Growth and regeneration of white mangrove in
Puerto Rico. Carib. Forester, 20(3-L):59-71.
WADSWORTH, Frank H., and Juan Amedee BONNET, 1951. Soil as a factor in the
occurrence of two types of montane forest in Puerto Rico. Carib. Forester,
WADSWORTH, Frank H., and George H. ENGLERTH, 1959. Effects of the 1956 hurri-
cane on forests in Puerto Rico. Carib. Forester, 20(1-2):38-51.
WALTER, Heirich, 1962. Die Vegetation der Erde in 1kologischer Betrachtung.
Band I: Die tropischen und subtropischen Zonen. VEB Gustav Fischer
Verlag, Jena, xv + $38 pp.
WEST, Robert C., 1956. Mangrove swamps of the Pacific Coast of Colombia.
Ann. Assoc. Amer. Geogr., [6(1):98-121.
WHITE, H. H., Jr., 1963. Variation of stand structure correlated with
altitude in the Luquillo Mountains. Carib. Forester, 24(1):46-52.
WOLCOTT, George N., 1945. Trees for roadside planting in Puerto Rico.
Carib. Forester, 6(3):115-120.
YOUNG, Robert N., 1955. A geographic classification of the landforms of
Puerto Rico. Chapter II In: C. F. Jones and Rafael Pic6 (eds.),
"Symposium on the Geography of Puerto Rico," Univ. Puerto Rico Press,
Rio Piedras, pp. 27-46.
APPENDIX A. -- Terminology
Most of the vocabulary used here is defined in the glossary (pages
317-336) of Dansereau's "Biogeography: an ecological perspective" (1957a).
Additional terms are to be found in Dansereau, Buell, and Dagon (1966).
Airphoto-type: a vegetation unit and its habitat as determined chiefly by
airphoto image interpretation and whose description often lacks some of
the definitive elements, notably the floristic, included in vegetation-
type descriptions (see Master-Table E). It is the basic unit used here
in large-scale mapping (see Maps 2, 3, 7).
Belt: a vegetation-type or a plant-community that occupies a definite position
within a habitat or an ecosystem. It is usually poised on a gradient, such
as a floodplain. (Not to be confused with zone, which see.)
Ecosystem: the conjunction of living and non-living elements which, on a given
site, operate the conversion cycles of environmental resources.
Ecosystematic regime: the prevalent physiological stress that dominates an
ecosystem and determines the behavior of its living members (see Master-
Formation-type: the purely structural definition of a stand of vegetation,
based on height and coverage of woody and/or herbaceous elements. In the
present context, only ten of these are recognized (Master-Table C).
Habitat: the physiographic and edaphic unit within which an ecosystem develops.
Plant-community: a phytosociological unit described in terms of at least
summary floristic composition and outstanding structural features.
Regime: regular and often rhythmic behavior, e.g., climatic regime (mediter-
ranean, monsoon), soil regime (podzolization, laterization), ecosystematic
regime (helophytia, tropophytia).
Stand: an actual area of continuous and homogeneous vegetation, e.g., a
stand of cattails, a stand of mountain palms.
Vegetation-type: a purely descriptive, sometimes very inclusive, characteri-
zation of the aspect of vegetation; it may contain some floristic, some
structural, and some site notations. (Many examples in text and in Master-
Zone: a geographical area that combines certain features of topography and
climate (see Map 1 and Table I).
APPENDIX B: Map Errata
Map 2. 1) Unlabeled area in the north-central part, bordered on the west by
the head of a bay and on the northeast by type 341: should be
2) Legend: for 1952 read 1951.
Map 5. Narrow strip in the southeast part-bordered on the north by Scrub,
on the west by Forest, and on the south and east by the sea, and
corresponding to type 781 on Map 3- should be Steppe.
Map 6. 1) Area in the southeast part-bordered on the north and west by
subxerophytia, on the east by chasmohalophytia, and on the south
by the sea, and corresponding to type 121 on Map 3-is indicated
as "helophytia" but should be "helohalophytia."
2) Two areas in the northwest part-bordered on the north, east, and
south by subxerophytia, and on the west by helohalophytia, and
corresponding to type 534 on Map 3-are indicated as "helohalo-
phytia" but should be "helophytia."
Map 7. 1) Unlabeled area in north-central part-bordered on the east and
west by type 456, on the north by type 364, and on the south by
type 362-should be continuous with type 362 to the south.
2) Large, very irregularly shaped area just southeast of the center
is labeled 448, but should be 456.
3) Area in the southeast part-bordered on the north, east, and west
by type 644, and on the south by the sea--is labeled 450 but should
4) Legend: for 1963 read 1964.
MASTER-TABLE A. Categories and symbols used to draw structure
diagrams (Dansereau 1958, Dansereau, Buell, & Dagon 1966).
W 0 erect woody plants
L climbing or decumbent woody
A epiphytes and non-rooted
H V herbs
M ( bryoids and crusts
7 more than 25 meters
6 10 25 meters
5 8 10 meters
4 2- 8 meters
3 0.5- 2 meters
2 0.1 0.5 meters
1 0.0 0.1 meters
b barren or very sparse
i interrupted, discontinuous
p in patches, tufts, clumps
d ] deciduous or ephemeral
?. LEAF SHAa -AND SIZM
n <0 needle, spine, scale, subulate
g 0 graminoid
a > medium or small
v ) compound
6. LEF TEXTURE
f 0 filmy
k succulent or fungoid
MASTER-TABLE B. Huguet del Villar's (1929) ecological, non-geographical classification of the physiological regimes that characterize ecosystems
Harmony of factors
partly aquatic Dominant
NATURE OF CONTROL
Biotic Mephitic accumulations
Harmony of factors MESOPHYTIA
Harmony of factors siuostratu
v sub stratum
Harmoxy of factors
of the envirori-ent
20. Biogenophytia (s.str.)
25. E ndobi ophytia
QUALITY OF COiROL
Subtropical and temperate
Monsoon and deciduous forest
Bogs and needle-leaf forest
---- ---- ---- --- -- --- -- D unes -
Buildings, yards, railways
Logs under water
Bark of trees, sheaths of
Intestines of animals,
Lakes, ponds, streams
Marshes, temporary ponds
Sea, salt lakes
Arctic seas, ice, snow
MASTER-TABLE C. Formation-type. The height and coverage
of the ten formation-types: the height, in meters, is
average (not always maximum); the "class" refers to the
stratification and coverage classes of Table III.
Slightly modified from Dansereau 1958.
WOODY PLANTS (w, L) ERBACEOUS PLANTS (H, E, M)
SYMBOL CEATION- Height Coverage Height Coverage
meters class % class meters class % class
1 FOREST +8 7-5 +60 c var. var. var, var,
2 WODLAND -+8 7-5 25-60 i-c var. var. var. var.
3 SAVANA 2-10 5-4 10-25 i-p 0-2 3-1 25-100 i-p-c
4 SCRUB 0.1-8 4-2 25-IO i-c var. var. var. var.
5 PRAIRJI 0.5-2 3 50-100 c
6 MEAW 0.0-0.5 2-1 50-100 c
7 STEPE 0.1-2 3-2 0-25 b-1 0.0-2 3-2 10-50 i-p
8 DESERT 0.0-10 5-1 0-10 b 0.0-0.5 3-1 0-10 b
9 TUNDRA 0.0-0.25 2-1 10-60 b-c 0.0-0.25 2-1 0-20 b-p
10 CRUST 0.0-0.1 1 50-100 p-c
MASTER-TABLE D. List of the vegetation-types and plant communities of Puerto Rico.
The last digit (at the left) refers to formation-type (as in Master-Table C).
Photo numbers show stands of vegetation; if underlined,they display a whole
landscape; if in parentheses, an individual dominant or characteristic plant.
Vegetation-type Zone (Map 1 and Table I) Composition Dia-
or Plant-community gram Photo
S-- -------- and num-
u Name Ia I 1 III IV V VI Dominants Other species relev6 ber
0.0 no vegetation X X X X X X X 34
1.5 kelp tangle X Laminaria sp.
2.6 algal meadow X Ulva lactuca
3,6 turtle-grass Thalassia
bed testudinum 3
4.0 marines algal Enteromorpha sp.
4.0 fucoid mat X Turbinaria 1
red-mangrove Rhizophorangl (7)
5.1 forest A mangle (9)
5.4 red-mangrove X Rhizophora Avicennia 2 4
scrub mangle nitida 3
6.1 black-mangrove Avicennia LagumIularia 10
6.1 forest nitida racemosa 4 13
black-mangrove Avi cennia 11
4 scrub. nitida 5
Bhizophora mangle 6
7.1 mixed mangrove Avicennia nitida 7
forest Laguncularia 8
7.3 mangrove savana X Rhizophora mangle Batis maritima 10 17
.... -----------Avicea nitida -
7 mixed mangrove Laguncularia 11 (15)
scrub racemosa 12
9.6 glasswort mat X Saicornia
sea-purslane Sesuvium 18
10.6 mat X portulacastrum 19
___ __ (20)
11.4 saltwort scrub X Batis maritima Avicennia nitida
12.5 salt-fern aureum
Name Ia I l III IV VI Dominants Other species Dia- Photo
ber Ha Igram
bay-cedar- Suriana maritima 21
13.4 horse-bush X 13Gundlachia 2
scrub corymbosa _____
ba7 -cedar X Suriana maitima Borrichia (22)
13"7 steppe Saarbcrescens
14.6 salt-sedge x Fimbristylis (A)
meadow spadicea 27
14.7 spurge-sedge x Euphorbia 14
14. spurge desert Fimbristylis 15
14.8 spurge desert X spadicea 16
high button- Conocarpus erecta 17 28
15.4 mangrove- X Cephalocereus --
cactus scrub royeni
low button- Conocarpus erecta 26
15.-a mangrove- Cactus intortus -
15.9 button-mangrove Conocarpus erecta 23
matted scrub Conocapus erect 2
Opuntia rubescens (2
sebucan- Conocarpus erecta Lemaireocereus
16.4 tachuelo X Cephalocereus hystrix 31)
thornscrub royeni Plumiera alba 32
beach-grass- Ipomoea- Canavalia 35
17.5 mornirg-glory X Sporobolus- maritima 19 36
17.6 mornirg-glory x Ipomoea pes-caprae 20
17.8 sea-rocket X Cakile lanceolata 46
185 salt-grass X 21 37
Spartina patens 22
7 steppe x 24 (38)
beach-grass Sporobolus Paspalum 2
svard virginicus vaginatum 28
20.4 seagrape X Coccoloba uvifera 33
23. wood land
29.6 waterlily mat
30.5 grass marsh
31- 5 grass marsh
32.5 cattail marsh
335 reed-grass x I
...... ^.. ~ Leucaena glauca
.Anris elem ifera
34- cspike-rush x interstincta Sagittaria 45
34" marsh Eleocharis lancifolia 46
35.5 sedge marsh X X Carex AZ
rivercane X X Gynerium
36.5 brake sagittatum
37.4 icaco scrub X Chrysobalaus Blechum indicum 47
pond-apple Annona glabra Acrostichum 48
38*2 swamp X Conocarpus erecta aureum
Imln 11 ":,l
m Name la I a III IV V VI Dominants Other species ga Photo
39. PterocapusPterocarpus 49
391 swamp officinalis
40.0 duckweed crust X X Lemna perpusilla 50
41.7 spider-flower X Cleome spinosa 51
42.3 lwlanrest X X Mammea americana Nemranace
rainfore st IManilkara nembranacea
x- K x ,aculeata 55
43.1 trumpet-wood X X X Cecropia peltata Erythrina 61A
44.2 jungle X X X Piper
45.4 guava scrub X Psidium guava
46.4 pepper scrub X X X Piper aduncum 66
47.4 firebrush scrub X X Croton lucidus 62
48.2 woodland Randia aculeaa Spahodea
----- ---- --- Didymopanax campanlata
4 second-growth x x x morototoni capan laa 60
48*3 savana -
49.6 St. Augustine Stenotaphrum
4 grass sward X () secundatum
50.6 carpet-grass Afonopus
S sward compresses
51.6 cerrillo-grass X x Sporobolus
51-6 sward indicus
52.6 wiry-beardgrass Andropogon
53.7 field X Ananas sativa 6
54.5 canefield X X (X) officinarum 69
ber Name la I II III IV V VI Dominants Other species Dia- Photo
55.3 orchard X X Mangifera indica
58.3 gardens X (X) (X) X (ornamentals)
59.2 screens X (X) (X) X equisetifolia Termiralia
Poinciana regia catappa
59.4 hedges X (x) X Allamanda
Stachy tarpheta 73
59.7 roadsides X (X) (X) X jamaicensis Ceiba pentandra 74
Sida carpinifolia 75
60.1 gateado Coccoloba Torrubia fragrans 51 (72)
forest X laurifolia Faramea 52
60.4 crabwood scrub X GymnanthesEugenia monticola 54
lucida, Egeia nonticola 54
61 balsamfig Clusia rosea angustifolia 77
61"4 screen a Anthurium acaule Dipholis 55 (78)
62.3 lme-umbo-aa X Gaussia Sideroxylon 56 79
limbo s a attenuata foetidi ssimum
Bursera Ilex nitida
62.4 1l im-gumbo- X simaruba Conocladia glabra 57 80
6. roble prieto Tabebuia Plumiera alba 38 81
" scrub X heterophylla Elaeodendron 58 (82)
64.1 bucaro forest X Bucid sessiliflora
... buceras Krugiodendron
64.2 bucaro woodland X ferreum 59
64.3 savana X Bursera simaruba Thama 18
65.7 wirepeass X Uniola virgata 86
S grama-grass Bouteloua
667 steppe heterostega
67.5 pa arie X X X X Panicum maximum
bem Name la I IIa III IV V VI Dominants Other species Dia Photo
rabo-de-gato X confinis
*5 pasture Andropogon
69.6 lanilla pasture X Aristida
70.7 sward annulatus
71.6 paradito X Pennisetum
72 msite osopis Leonotis 88
savana juliflora nepetaefolia
Bucida buceras aculeaata
72.3a bucaro-mesquite X Prosopis chranthes 6 87
73-5 aroid-sedge X antiquorum 50
belt Cyperus spp.
73.5a yerbade- X Jussiaea repens
tabonuco excelsa Philodendron 62
74.1 forest Lucuma krebsii 63 89
multiflora Ruellia coccinea
Inga vera Eryhrina 0
75.1 Inga-coffee X Inga laurina Erythrina 61
forest Coffea arabica poeppigiana
76.4 ros -aple X Eugenia jambos
Cyathea arborea (97)
77*3 fern savna X Dicranopteris spp. 98
774 fern scrb x Cyathea arborea
77*4 fern scrub X Hemitelia horrida
78.5 field X minutiflora
S-bluestem X bicornis 64
79-5 prairie Sporobolus r 94
80.5 tobacco field X Nicotiana 100
81.5 bananaT Musa sp.
u 1.5 plantation
Num-... II IV II Dia-.
ber Name a I a III IV V VI Dominants Other species gram Photo
2 cliff algal
02*0 crust x
83*5 ptangle x Potamogeton sp.?
84.0 algal carpet
85.5 sedge marsh
86.4 bog scrub
87.4 montane marquis
88.ic sierra moss-
88.4a elfin forest
Eugenia Ilex sintenisii 72
88.4b high montane X borinquensis Ocotea spathulata 118
scrub Tabebuia rigid Icacorea 73
MASTER-TABLE E. Airphoto-types of the Roosevelt Roads area shown on Maps 2, 3, and 7 (n.a. = non-applicable; em. = emulsion or mixture of two format
(5) (6) (7)
1 895 0.0 desert n.a. n.a. n.a. barren n.a. wave-ashed, sandy beach (Psao
may have occasional
tidal flats; silty
... .. high(> 2 m.), in- outer edge of outer
3 151 5.1 forest trees 12(15) 70-80 kn sti s;hizophora gggl mangrove; islets;
diameter: 15 cm.al tdl chnn
low (<1 rm.), in-
scrub 2-10 terlockireg stilts;
desert) shrubs 2- (50-80 small (ca. 5 m. Rhizophra dom. (occ.) tidal flats or very
4 441 5.4 serat) w .hnt), widely d.. (oc) shallow ter
em clumps) scattered clumps
low ( < m.), in-
scrub20-50 terlocking stilts;
5 431 5.4 desert) shrubs 5 ( 0 al d cate 5umps Rhizophora mangle dom. (occ.) tidal flats
em.) clumps) spaced 5-15 m.
5.4-5.1 (forest) (trees) 6-8(10) 7C-90 stilts tidal cba l margins
6 4a9 Rhizophora male outer mangrove;
5.4 scrub shrubs 2-3(5) 50-80 low (< 1 m.-) in- generally not so firm
terlocking stilts as in 141
7 427 5.4 scrub shrubs 2-3(5) 50-80 low (< 1 n.) in- Rho ra generally not so firm
terlockdng stilts a....g.geneas lyn o sfi
(1) (2) (3) (4)
(5) (6) (7)
TRhizophora mangle doio.?
low (<1 m.) in- inner side of outer
8 423 5.4 scrub shrubs 3-4? 65-80 terlocking stilts L-crlaria occ.- mangrove; may be
7.4? racesaF comm. drier than 141
Avicennia nitida? rare
Rhizophora mangle dom.; ab.
stilts abundant Laguncularia comm.
9 141 7.1 forest trees 15(18) 60-80 and usually racemosa (dom.; alb.) outer mangrove
Avicennia nitida occ.
woodland 30-65 clumps and groves; Rhizo phora le dom.
0 225 5.1-4.0 scb) trees (5)8- (70-90 25-50/o non- L-.ucularia
10 25 5.1-4.0 ) (shrubs) 10 wi thin vegetated (?) Launculari
cluhaps) tidal flat racemosa occt
shrubs 3-4(8) 6c-85 Bhizera mangle dom.?
Lvegycularia recently cut-over;
11 323 7*3 savana very unevracemosa in areas normally 141
323 73 (scrub) trees 8-15 5-15 height "-- 0" occ. or 121
Bhi zophora mangl e
Avicennia nitida dom.
12 121 6.1 forest trees 10-15 65-85 (reproduction) Laguncularia inner mangrove
slight racemosa occ.
trees 8-20 2 ------___-_--------
Avicennia nitida? rare-occ.
Upland tree spp. occ.
13 319 7.3 savana shrubs 4-8 10 Upland tree spp. occ. probably cut-over
6 axria rrare-occ. areas
stems dense and racemosa
shrubs 2-4 20 low-branching Avicenna nitid rare-occ
Upland tree spp.? occ.
herbs < 2 75-90 Similar to 528 similarto 528
(1) (2) (3) (4)
(5) (6) (7)
m.) locally dense
Avicennia nitida dom.
inner mangrove; slight-
15 443 11.4 scrub shrubs 0.5 40-80 dense and tangled Bats maritima dom. ly raised, slightly
Salicornia ? sady areas
inner mangrove; slightly
shrubs 1-4(5) 20-50 Avicennia nitida occ. raised, slightly sandy
16 341 7.3 savana ___ ____ areas but slightly
shrubs 0.5 30-70 dense and tangled Batis maritima dom. lower and less sandy
-- than 443
Avicennia nitida ab.
similar to 341
sub I 2/-5 Probably halo-
shrubs 1(2?) 2-5 phytic; Batis? occ.
1.4 savana restricted to vicinity HELOAL
18 355 14, (meadow) Beach grasses? of 831 (q.v.); along (Helop
17 scrub) herbs 0.5 60-70 I o ab ? shallow drainage ways (Mesox
scrub) herbs 0.5 60-70 Halophytic ab,
Shrubs probably restricted to vicinity
stems dense and halophytic; of 831 (g.v.); low, HALOPHY
15.4a? relatively slen- halophycHeloh
15.4a? relatively slen- one or two spe- flat areas; may be Helo
19 475 13.4 scrub shrubs 3-4(6) 65-85 der, possibly ies probably subject to occasional Halox
abundant multiple overwhelmingly tidal (perhaps fresh Subxe
stems and low dominant water) inundations
2 dense canopy; Halophytic do
1(4 80-90 height may vary shrubs;
(4) considerably with Coccoloba?
20 451 20.4 scrub shrubs -topographic ir- lower slopes facing sea H
20 451 20.4 sregularities and (Halo
<0.1- 0-15 exposure to salt- Opuntia dillenii patches
0.5 bearing winds veaff---- s
SBeach grasses? dore o general
meadow herbs? (O.1?) 60-100 very low and sedges? vines? v t of
21 5 17 prairie) 5I? quite dense-; level, low areas Peatr
21 535 7.5 savana) evenly spaced; generally adjacent to Petr
scrub) shrubs 1-2(3) <1 even height rare oc a Subx4
(1) (2) (3) (4) (5) (6) (7)
(0.5)2- ,. -70 canopy; stems sclerophyllous windward hillsides
2 471 20.4? scrub shrubs 4(5) probably closely halophytes dom. facing ocean; never HALCPH
47 13.47 ?. .. spaced and often ...... --- more than 50 meters H
0.1- 0-15 multiple; wind Opuntia dillenii from shore
0.5 and/or salt pruned ve a~7 ff
(0.57) all shrubs and
steppe shrubs I?-2 0-5 most herbs growing
(23 7e8 13.7
S(desert) herbs <0.1 2-10 and on small
dense to dense probably pre-
shrubs 1-3(8) 50-80 in elongate dominantly dom.
24 445 13.4- scrub strips parallel sclerophyllous beacridges PS
(24.2) (savana) to old shoreline
(8)1i0- Cocos nucifera or
trees (8)10- <2 6--fEe-r-_Te-es occ.
shrubs 1-2(3) <1 halophytes? occ.-rare
restricted to vicinity
10.6? steppe halophytic? of 831 (q.v.); ;S
25 765 17.8?7 scrub) graminoids; patches slightly higher ground (EHal
17.5? (meadow) herbs (40.5 30-50 halophytic and occ. than most of 831 (Halc
halophytic; restricted to vicinity PS.bWC
17.87 steppe herbs graminoids? patches of 831 (q.v.); (Heic
26 77 meaow ) (shrubs?) 5 0 and/or and occ. along shallow Eelc
meadow) succulents drainage-ways Halc
halophytic? of one fairly extensive
herbs (0.1?) graminoids? area (with inclusions)
ers 0.5 5 and/or occ.-rare a inner edge of a man-
17.8 succulents? grove swamp; level to PSAMM
27 831 10.6? ---- -------.-- very gently sloping; (Psa
11.4? either sandy or dry (Hy
shrubs 1-2 (1 halophytes rare saline flats; may be
by storm tides
pubescens T occ*
grasses may be Panicu maxi filed estuary heads--
28 528 67.5 prairie herbs 2 85-100 planted by man vel aff. domn. usually adjacent to HELO P
for cattle halophytes? 119, or 319
forage or hay .y.121.
Ilomoea spp. patches
tree species dom* extreme inner edge of MESE
29 117 5.1 forest trees ( 5 75-90 relatively free treejspecies inner mangrove; very (HeloS
10(15) of undergrowrbh Laguncularia slightly saline? (Helo
sila t 1 level to gently sloping
1 ,r 4.1? (7)8-l0 -, relatively free similar to 117floor of smll valley
30 11 42? forest trees 15) 75-90 of undergrowh ut lacking dom. (cove) adjacent to MESOX
mangrove inner mangrove
ation in tree Prosopis
trees 6-8 20-50 density; similar juliflora? or dom. wind-sheltered coastal
31 342 72.3? savana (shrubs) to 444; quite free Pitheceifobium? cove; only one oc- MESOKE
of undergrowth; currency in map area (Subx
herbs 0.5-1 60-80 graminoids may be graminoids dom.
gently to moderately
not exposed to wind- MESCKE
32 232 48.2? woodland trees 15-20 40-60 dom. blown salt; open canopy (Subx
may result from selec-
tive cutting of forest
sedge dom. lower slopes (ca. 259/
33 534 35.5 prairie herbs 0.5-1 70-90 tangled, coarse ------- grade); may be MESX
bunc sedge pp app. occ.- ephemeral seep area
bunch sedge Ipomoea spp. patches
beach grasses? restricted to one area;
34 529 18.5 prairie om, band between 560 and ESC
7.3? .savana) 151; low, very gently Helo
shrubs 1-3 2-5 mangrove ooc. (< 57 grade) sloping (Helo
may be some salt-
spray damage to at least two spe- relatively steep wind- SUBXE
35 465 13.4 scrub shrubs 2-5 80-90 the extent of in- cies of probably dward slopes facing ocean (
s scleropyllous war slopes facing (Hye
ness of shrubs
(1) (2) (3) (4) (5) (6) (7)
36 44 22.4 scrub shrubs 5-10 65-95 undergrowth very versity of usually steep hillsides SUXE~
(forest) (trees) dense species and ravines
near edges of
37 44 72.3? scrub shrubs 342; widely spa-o low level area between
37 444 73.4? scru (trees) 4-5(8?) 60-80 ced stems ulfo- r dom. 52 and 443; only sxER
probably quite te obium occurrence in (elo
free of under- and Leucaena occurrence in map
resembles 465 in
leeward (generally W-
facing) sides of broad,
shallow ravines on
steep hillsides; within
465 areas; possibly
slightly influenced by
commonly on, but not
39 452 23.4 scrub shrubs 3-4(5) 75-90 Leucaena gauca dom. restricted to, hill-
sides and crests
low-branching; may occur on any well-
40 454 23.4 scrub shrubs 2-3(4) 75-90 small stems Leucaena glauca dom. drained site except
closely spaced ----- immediately adjacent
to salt water
dense stem spac-
ing; most common
ca. 3 cm.
dense stem spac-
42 456 44.4 scrub shrubs 2-5 50-70 ing multiple
stems and low-
shrubs 325 10-70 occ.-dom.
43 360 44.3 savana rough pasture
herbs (O.1) 40-80 grasses & forbs om. commonly on hillsides
0.5-1 ------------ and crests
bunch grasses occ. ab.
(1) (;) (3)
(4) (5) (6) (7)
vines (2)4-5 5-50
shrubs 8) 1-10 occ.
44 48.3 savaa rough pasture dom.
44 362 48.3 (prairie) (.1) grasses & forbsdo.
herbs 0.5-i 75-90
bunch grasses occo.-ab.
--1. 4- L 4 4 4.
grasses & forbs
______________ F -
h0 thes occ.-dom.
creeping Ipoea spp.7
grasses & forbs
commonly on hillsides
hillsides and crests;
commonly peripheral to
or in association
sites similar and
adjacent to 232;
probably results from
shrubs <2 5-20 stem spacing and occ. restricted to 831 area
48 770 13.7? steppe overall density oc. restricted to 831 area SUBXEO
48 770 13.70.1- intermediate be- occ.-dom. in and dry
herbs 0.5(1?) 10-30 tween 831 & 466 patches
shrubs locally tal-
along drainage low, level to gently
49 466 47.4? scrub shrubs 1(2) 70-80 most coon stem Croton do. higng ighut S RC
diameter 3 cm.; adjacent to, 831
lar to 45Z
t + + 4 F F 4 +
...... .. . .... .. .. .. S B X E R O
50 560 67.5? prairie herbs 0.5-1 75-95 forbs occ. and crests ara
bunch grasses occ.-ab.
very low and
evenly distribu- restricted to vicinity
ted (except for beach grasses?; of 831 (q.v.); gentle
51 572 17.5 meadow herbs (0.1) 75-85 faint, fine sedges?; dom. (2- grade) slopes; SUBER
(prairie) <0.5 linear distribu- creeping vines? sandy?; very shallow, (Psamm
tion parallel fine gullying
in and on excavations
shrubs 2(3) 5-15 pioneer rare-occ. and construction sites SUJBXERC
52 75C 51.6? steppe shrubs 2() and early not recently worked; (Xerop
herbs 0.1-0.5 10-40 successional occ. (ab. soil generally very (Paran
(52.6? weed) species in patches) poor (topsoil removed)
desert shrubs apparently SUBXER
53 786 51.6? steppe shrubs (0.5) 2-10 bare (may have rare-occ. Psamm
savana; 1(27) sparse graminoid (Para
A10 42 spaced (usually
trees 0 12 40-60 evenly) in grid Cocos nucifera dom.
54 245 24.2 woodland 2..pattern _....... ..... casionally other level, PARANT
undergrowth sandy areas near coast (Psamm
shrubs ( .5)-( 5)10 usually kept occ,
3(4?) open by cutting
spaced ( > 25 m.
to 560 with ad-
dition of trees
low hillsides near coast
trees 10-15 (2)5-10 spaced ( > 25 m. Cocos nucifera occ.
56 332 (24.2)- savana Saccharum valley bottoms and low,
54-5 (<.1) o fficinarum dom. gentle hills near coast
herbs (<-4 ) 0-95 similar to 532 offorbcc.
(1) (2) (3) (4) (5) (6) (7)
-r I1 I____I__I___I__________I___I
(< 1 acre) fields
planted to garden
al fruit trees
similar to 352,
shade and orna-
shade and fruit
m i ii i i 4i i4
height and cover
Qoff m arum
uneven height and or forage and/or valley bottoms only;
60 536 54.5 prairie herbs 2-4 85-95 spacing; may be hay grass or dom. probably poorly (PARe
poor crop of some other drained ground (Hel
sugar care tall coarse
patue low, gentle slopes;
61 622 49.6-50.6- meadow 90-100 dense even turf pasture? dom. commonly situated be-
51.6 (prairie) rb (1 90-10) grasses tween 532 and 560
in general vicinity of
6 644 49.6-50.6- 0.1 dwellings, other occu-
62 644 51.6 meadow herbs (0.5) 75-100 dense, even turf 'lawn" grasses dom. pied buildings, amou-
nition dumps, roadsides
63 648 49.6-50.6 meadow herbs (< 1 60-85 irregularly dense ~lawn" grasses dm.hes aroundoccupied
S51.6 (0.5) turf patches ing
64 862 52.6? desert herbs <0.1 <10? ground apparently "weeds" rare-occ. buildings and on
(0.5) bare recent excavations
65 894 0.0 desert n.a. n.a. n.a. ground bare n.a. n.a. ar ad gravel
MASTER-TABLE E( 1). Guide to symbols used in Master-Table E and on
Maps 2, 3, and 7.
Airphoto-type Key numbers on
Key numbers on
MASTER-TABLE F. Life-forms according to Braun-Blanquet (1932), to
Raunkiaer (1934), and to Dansereau (1945, 1957a, 1959). Two examples
from Eastern North America are given in each case: the first one a
native plant, the second a cultivated one; the third example is
from Puerto Rico.
P PHANERCPHYTES: trees, or woody plants, with regenerating buds high (at least
50 cm.) above ground
Pg meaphaneroplhtes: tall trees, more than 25 m. high
Ex.: Ulmus aimercana, Picea abies, Pacryodes excelsa
Pm mesophanerophytes: small trees, 10-25 m.
Ex.: Ostrya virginiana, Morus alba, Avicennia nitida
Pp micr ohanerphtes: tall woody plants, 2-10 m.
Ex.: Corlus ornuta, Sggavulgaris, Annona glabra
Pn nanophaerophtes: low shrubs, 0.5-2 m.
Ex.: Syiraea latifolia, Berberis vulgaris, Croton poecilanthus
Ps climbing phanerqopkhtes
x. JdelastruT s scandens, Wisteria sinensis, Rhabdadenia biflora
Ch CHXA4EPHYTES: plants with regenerating buds not much above soil level
(less than 50 cm.)
Ex.: LycoDodium annotinum, Vinca minor, Cactus intortus
Chm moss chamag2hIeps: bryophytes and lichens
H BEICRYTOPITHYTES: plants that die down to soil level during the unfavorable
(cold or dry) season
Hs scapose hemicry tophytes: with a well-developed stem, branching quite
Ex.: Eu0atorium rugosum, Campanula media, Pilea krugii
Hr rosette hemicryptoptes: with a crown of leaves at soil level
Ex.: -Saxifraa viriniensis, Primula floribunda, Bromelia Rigyi
Hc cespitose hemicryptophytes: with many tufts or short branches arising
at ground level
Ex.: Danthonia spicata, Poa pratensis Bouteloua americana
Hg decumbent or climbing heicryptophytes: with a weak plagiotropic stem
Ex.: Polygonum cilinode, Iomoea coerulea, Cominelina elegans
MASTER-TABLE F (continued)
G GOPHYES: plants with regenerating buds well under soil level
Gb bulb- o tes
Ex.: Allirm tricoccum, Tulipa gesneriana, Hymenocallis declinata
Gr rhizome-. eopttes
Ex.: Smilacina racemosa, Iris gerrianica, Canna glauca
Gg root e jyhtes
Ex.: Ascl as syriaca Dahlia variabilis, Alpinia ntillarum
Th TEROPHYTES (annuals)
Ex.: Iiatiens ca ensis, sAlsaum rsaxatile, Cakile lanceolata
Th2 THEROPHYTES biennialss)
Ex.: Oenothera biennis, Althaea rosea, Leornrus sibiricus
E EPIPHYTES: plants that rest upon the boughs or stems of other plants
Ex.: Usnea longissimra and many other lichens, Cat labiata,
HH HYDROPHYTES: true aquatics (subdivisions in this category are according to
HHs free-living, submerged, or emerged plants
Ex.: Ceratophyllumr demersum, Salvinia natans, Pistia stratiotes
HHf emerged aquatics with broad leaves
Ex.: Pontederia cordata, Nelumbo lutea, Sagittaria lancifolia
HHj emerged aquatics with thin (reed-like) stems and/or leaves
Ex.: Scirpus acutus, y u paprus .u -caiXs
HHn floating-leaved rooted aquatics
Ex.: Nm scidsacunrosum, mhaea zanzibarica, Ny I nha .21a
HHv submerged aquatics with ribbon-like leaves and/or stems
Ex.: Vallisneria american, Potogetoncrisps, us7 lssia testudinum
HHr submersed rosette plants
Ex.: Lobelia dortmanna, Isoetes j stris, Cymodocea mamatorum
HHt submersed annuals or .ydrotherophytes
Ex.: Najas flexilis, Najas ua dalupensis, Najas marina
HHa eiphytic or epilithic aquatics
Ex.: Fontinaris spp., Podostemon spp., Helianthium tenellum
MASTER-TABLE G. Leaf-size classes (Raunkiaer 1934).
Symbo Area Categog
1 1 less than 25 mm.2 L cPTOPHLL
2 n 25 225 mm.2 NANOILL
3 m 225 2,025 mm.2 MICRPHYLL
4 M 2,025 18,225 mm.2 MESOPHYLL
5 mW 18,225 164,025 mm.2 MACROPHYLL
6 IMM more than 164,025 mm.2 MEGAPHYLL
MASTER-TABLE H. Floristic elements in Puerto Rico,
based on geographical distribution.
E Endemic: only in Puerto Rico (R~ostonea borinquena)
A Antillean: also found in other parts of the West Indies (Gundlachia corymbosa)
C Caribbean: Antilles and ContinentalCentral America, Florida (Bursera simaruba)
G 1NortYern Neotro2ical: Caribbean and Northern South America (Casearia guianensis)
B Neotropical: tropical America (Acacia westiana)
P Pantropical: tropics of old -and new world (Albizzia procera)
W Cosmopolitan: ( omineis)
N Naturalized: (Albizzia lebbeck)
X Planted: (Castilla elastic)
MASTER-TABLE J. Diaspore types, as proposed by Dansereau & Lems (1957).
( 'D" means diaspore.)
D. does not disarticulate from the parent plant before being deposited at the
site of further development. (Parent pl-nt may be dead or alive at
that time.) . . . . . . AUXOCHCRE 1
D. disarticulates from the parent plant before the dispersal phase.
D. very voluminous in relation to the actual reproductive part, consisting
of a loose spherical framework. . . .... CYCLOCHRE 2
D. consisting for the main part of the reproductive organ.
D. with scarious or stiff appendages.
Appendages thin, light, often flexible.
Appendages scarious, wing-like, or saccate .. .... PTEROCHORE 3
Appendages long, hairlike, or plumose. . POGONOCHORE 4
Appendages short, stiff, spiny, or glandular,
adhering to rough surfaces. . .. DESXCHORE 5
D. without appendages (except for arils).
D. with juicy or fleshy outer layers . . SARCOCHO 6
D. with hard outer layer.
D. small or light enough to be carried by a breeze.
......... SPOROCHORE 7
D. too heavy to be carried by breeze.
Parent pliant without mechanism of expulsion.
D. light enough to be carried by wind.
.. SCLROCORE 8
D. very heavy. . . . BAROCHORE 9
Parent plant with mechanism of expulsion. BALLOCHORE 10
TABLE I. The vegetation zones of Puerto Rico and their principal correlations and characteristics.
VI MONTANE SCRUB 9(p) 10(p) 1000 + peaks
V MONTANE FOREST
IV LOWER MONTANE
Ila FILL SCRUOB
tops of hills,
cool, very moist,
warm to cool,
very warm with
plains, hills, entsvery warm, helophytia fo svaa
aL(WLA25-lspsediments, moist subxerophytia
RATkNFREST cla I5(p) 25- 200 0 u. orcrops BA' r paranthropophytia scrub, prairie
2c _____________ 2__.___.__ ........
2a b 4c(p) very warm, helophytia
Ia LT Agenerally moist to dry halohydrophytia desert, meadow
la LITTORAL 50 flat, shoreline rock, sand BAr CA' psamophytia steppe, scrub
subzone and bays; silt, clay CA'r DAId lithophytia savana, forest
some cliffs EA' d chasmophytia
TABLE II. Characteristics and plant communities of the littoral subzone (la).
ter Table C)
1.5 kelp tangle
POND saltwater pond 2.6 algal meadow Ulva lactuca
SHALO s shallows 3 trtle-grass bed meatowestudin
SHALLOWi saltwater shallows 3.6 turtle-grass bed Thalassia testudinum
4.0 marine algal crust
b.4 black-mangrove scrub
4.6 fucoid bed
5.4 red-mangrove scrub
b.4 black-mangrove scrub Avicennia nitida
0.0 (rone) (none)
5.4 red-mangrove scrub Rhizophora mangle
6.4 black-mangrove scrub
7.4 mnixd-mangrove scrub
7.3 mangrove savana
5.1 red-mangrove forest
.1 black-mangrove forest Avicennia nitida
7.1 mixed-mangrove forest
9.6 glasswort mat
10.6 sea-purslane at
Rhi zophora mangle
scrub 11.4 saltwort scrub Avicerma nitida
12.5 salt-fern prairie
Acrostichum aureu n
13.4 bay.-cedar -~
13.4 bay-cedar --
13.7 bay-cedar steppe
14.7 spurge-sedge mat
14.8 spurge desert
14.6 salt-sedge meadow
-' raatted scrub
15.4 cactus scrub
1 high button-mangrove-
15.4a cactus scrub
17.8 sea-rocket strip
lO~ sea-.uurslane ma~t
17.6 morning-glory festoon
i8.5 salt-grass prairie
18.7 salt-grass steppe
meadow 19.b beach-grass sward Sporobolus virginicus
20.4 sea-grape scrub Coccoloba uvifera
scrub 21.4 maray-maray scrub Dalbergia ecastophyllum
22.4 snake-bark scrub Colubrina ferruginosa
24.2 coconut grove
25.2 Australian-pine screen
Ir I. -- Ir
PS Y O0-
31.5 trompetilla-grasv marsh
32.5 cattail marsh
Pani cmn purpurascens
33.5 reed-grass marsh Phragmites communis
34.5 spike-rush marsh Eleocharis spp.
35.5 sedge marsh
36.5 rivercane brake
37.4 icaco scrub
Cladium, Scirpus, Carex
woodland 3 .2 pond-apple camp oncarpus erecta
22.4 srnake-bark scrub
23.4 Leucaena thicket
2b.4 Malabar almond scrub
27.b water-hyacinth mat
28.6 water.lettuce mat
29.6 waterlily mat
30.5 para-car ib-grass mat
17.b morrning-glory festoon Ipomoea pes-caorae
18.5 salt-grass prairie
18.7 salt-lrass steppe
19.6 beach-grass sxard Sporobolus virginicus
20.4 sea-grape scrub Coccoloba uvifera
39-1 Pterocax-ons swamp
TABLE III. Characteristics and communities of the lowland rainforest zone (I).
ter Table C)
27.6 water-hyacinth mat
28.6 water-lettuce mat
29.6 waterlily mat
30.5 para-carib-grass mat
crust 40.0 duckweed crust
31.5 trompetilla-grass marsh
Pani cum purpurascens
Pani cum aquaticum
32.5 cattail marsh Typha domingensis
33-5 reed-grass marsh Phragmites communism
34.5 spike-rush marsh Eleocharis spp.
36.5 rivercane brake
39.1 Pterocarpus swamp
river-flat steppe 41.7 spider-flower flat Cleon~ spinosa
41-7 spider-f Imer f lat
42.1 lowland rainforest
43.1 trumpet-wood forest
45.4 guava scrub
46.4 pepper scrub
47.4 firebush scrub
23.4 Leucaena scrub
23.2 Leucaena wociland
woodland ---- ------ -
48.2 second-growth woodland
48.3 second-growth savana
49.6 St. Augustine grass sward
50.6 carpet-grass swrard
51.6 cerrillo-grass sward Sporobolus indicus
52.6 viry-beardgrass sward Andropogon gracilis
53.7 pineapple field Ananas sativa
54.5 canefield Saccharxan officinarum
24.2 coconut plantation Cocos nucifera
25.2 Australian pine screen
5b.3 gardens & parks
J 414. 'L -
--- --- I
TABLE IV. Characteristics and co'rrun ities of the seasonal-evergreen forest zone (II), and the hill-scrub subzone (IIa).
ter Table C)
42.1 lowland rainforest
43.1 tnrupet-wood forest Cecropia peltata
44.1 jugle Andira inermis
60.1 glateado forest
49.6 St. Augustine grass sward Stenotaphrum secundatum
50.6 carpet-grass sward Axonopus compressus
pasture meadow 51.6 cerrillo-grass sward Sporobolus indicus
52.6 wiry-beardgrass sward Andropogon gracilis
46.4 pepper scrub Piper aduncum
47.4 firebrush scrub
48.2 second-growth woodland
48.3 second-growth savana
___________ + I
CLIFFS CHASMOPHYTIA STEEP ROCK shaded cliff face scrub bl.4 balsamnfig screen Anthurium acaule
savana 62.3 llume-gumbolimbo savana Gaussia atteuata
SUBERO- NARROW 62.4 llme-gumbolimbo scrub Bursera simaruba
SPRS STO limestone scrub
crest scrub Tabebuia heterophylla
PHYTIA RIDGE 63.4 roble prieto scrub Plumiera alba
TABLE V. Characteristics and communities of the semi-deciduous forest zone (III).
REGII FO RMATION- VEGETATION-TYPE
LANDFCH (Master- HABITAT EC0MYSTM TYE (iMas- or PLArT-OMFLTY DOMINANT SPECIES
Table B) ter Table C) (Master-Table D)
open savana savana 64.3 gumbolimbo savana Bursera sinaruba
scrub 1 sebucan-tachuelo Cephalocereus royeni
*a thornscrub Pictetia aculeata
65.7 ire-grass steppe Uniola virgata
b6.7 grama-grass steppe Bouteloua heterostega
meadow 51.6 cerrillo-grass sward Sporobolus indicus
67.5 Guinea-grass prairie Panicum maximum
.prair5 rieo-de-gato waste Arundinella confinis
o.5 roAndropogon semiberbis
pasture meadow 9.6 lan!illa pasture Leptocoryphium lanatum
steppe 66.7 grama-grass steppe Bouteloua heterostega
prairie 67.5 Guinea-grass prairie Panicum maximum
meadow 51.6 cerrillo-grass sward Sporobolus indicus
steppe 70.7 angleton-grass sward Andropogon armulatus
rjadow 71t.6 pr to meadow Pennisetmn ciliare
72.3 mesquite savana
72.3a bucaro-mesquite savana
13.. --* -, -