|
inauiuumui~*
of the
FLORIDA
MUSEUM OF
NATURAL HISTORY
A CLASSIFICATION AND ORDINATION
OF THE TREE COMMUNITY
OF TIKAL NATIONAL PARK,
PETEN, GUATEMALA
Mark D. Schulze and David F. Whitacre
Volume 41 No. 3, pp. 169-297 1999
UNIVERSITY OF FLORIDA
GAINESVILLE
Numbers of the BULLETIN OF THE FLORIDA MUSEUM OF NATURAL HISTORY are
published at irregular intervals. Volumes contain about 300 pages and are not necessarily completed in
any one calendar year.
JOHN F. EISENBERG, EDITOR
RICHARD FRANZ, CO-EDITOR
RHODA J. BRYANT, MANAGING EDITOR
Communications concerning purchase or exchange of the publications and all manuscripts should be
addressed to: Managing Editor, Bulletin; Florida Museum of Natural History, University of Florida;
P. O. Box 117800, Gainesville FL 32611-7800; U.S.A.
This journal is printed on recycled paper.
ISSN: 0071-6154
CODEN: BF 5BA5
Publication date: March 29, 1999
Price: $ 12.00
A CLASSIFICATION AND ORDINATION OF THE TREE
COMMUNITY OF TIKAL NATIONAL PARK, PETEN,
GUATEMALA
Mark D. Schulze" and David F. Whitacre2
ABSTRACT
We studied tropical forest composition and structure in relation to topographic and edaphic
variation, with the goal of understanding the factors that determine species distributions and the degree
to which species composition can be predicted by local environmental conditions. We sampled the tree
community of Tikal National Park in Pet6n, Guatemala, using 294 sample plots of 0.041 ha each,
totaling 12.1 ha, placed systematically along topographic gradients. In addition to recording tree and
sapling occurrences, we took data on environmental factors, vegetation structure, and disturbance
history. Indirect and direct gradient analyses were performed (using DCA and CCA, respectively) to
investigate variation along a predominant topographic/edaphic gradient and along gradients of
disturbance and light availability. We used TWINSPAN analysis and ranked sorting of samples by
environmental variables to produce a forest type classification practical for field use. Individual
species distributions were examined with respect to edaphic conditions, understory light availability,
and natural forest disturbance. Though topographic relief is moderate in the area, several edaphic
factors varied strongly and predictably, though not monotonically, from hillcrests to low-lying
depressions, and exerted a predominant influence on tree community composition. Most tree species
distributions were strongly correlated with topography and associated edaphic conditions, and
distribution patterns were consistent among widely separated topographic gradients, suggesting that
environmental conditions, rather than historical events, were largely responsible for these patterns.
Many of the less shade-tolerant species also showed positive relationships with natural disturbance
history and degree of canopy opening. Treefall gaps have different functional significance for
regeneration at different points along the topographic/forest type continuum, due to correlated
differences in canopy height, canopy evenness, and light penetration. Hence, many species that appear
to depend on treefall gaps for colonization and recruitment in upland portions of the continuum, are
less confined in their exploitation of lower regions of the gradient, where canopy discontinuities may
provide sufficient light for seedling persistence and recruitment. Thus many highly light-demanding
species were not associated with treefall gaps, and appeared capable of colonizing relatively open
lowland forests. Distribution patterns and disturbance/light responses are discussed for individual
species, which are provisionally placed into ecological guilds based on apparent tolerances for light
levels and edaphic and other environmental factors. In Tikal, much of the natural variation in forest
composition and structure can be related to topography and edaphic conditions.
RESUME
Con el objeto de entender los factors que determinan la distribuci6n de species y el nivel en
que la composici6n de species puede ser predecida por condiciones ambientales locales, nosostros
Current Address: 208 Mueller Lab, Department of Biology, Pennsylvania State University,
University Park PA 16802, U.S.A.
2The Peregrine Fund, 566 West Flying Hawk Lane, Boise ID 83709, U.S.A.
SCHULZE, M. D., and D. F. WHITACRE. 1999. A classification and ordination of the tree
community of Tikal National Park, Pet6n, Guatemala. Bull. Florida Mus. Nat Hist. 41(3): 169-297.
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
estudiamos la composici6n y estructura del bosque tropical en relaci6n a variaciones topogrificas y
edaficas. Muestreamos la comunidad de Arboles del Parque Nacional Tikal en Pet6n, Guatemala,
mediante el uso de 294 sitios de muestreos de 0.041 ha cada uno, totalizando 12.1 ha, colocadas
sisematicamente a lo largo de gradientes topograficos. Ademis de registrar la ocurrencia de Irboles y
renovales, tambi6n recolectamos informaci6n sobre factors ambientales, estructura vegetal e historic
de alteraciones. Se realizaron analisis de gradiente director e indirectos (usando DCA y CCA,
respectivamente) con el objeto de investigar la variaci6n a to largo de los gradientes
topogrificos/edificos predominantes, y a to largo de gradientes de alteraci6n y disponibilidad de luz.
Produjimos un sistema de clasificaci6n de tipos de bosque para uso prictico en el campo mediante el
analisis y ordenamiento de las muestras de acuerdo a variables ambientales y usando TWINSPAN. La
distribuci6n de cada especie fue examinada con respect a las condiciones edificas, a la disponibilidad
de luz, y a las alteraciones naturales del bosque. Aun cuando el relieve topogrifico es moderado en el
Area de studio, various factors edhficos mostraron una fuerte y predecible variaci6n desde las cimas a
los valles, y aunque esta no fue monot6nica, impuso una influencia predominante en la composici6n de
la comunidad arb6rea. La distribuci6n de la mayor parte de las species arb6reas estuvo fuertemente
correlacionada con la topografla y condiciones edaficas asociadas, siendo los patrons de distribuci6n
similares entire gradientes topogrificos separados. Estos resultados sugieren que condiciones
ambientales, mas que events hist6ricos, fueron mayormente las responsables de estos patrons.
Varias de las species menos tolerantes de sombra tambi6n mostraron correlaciones positivas con la
historic de alteraciones naturales y el grado de aperture del dosel. La significaci6n funcional para la
regeneraci6n de los claros producidos por arboles caidos, fue diferente a lo largo del gradiente
topogrifico y del tipo de bosque, debido a diferencias correlacionadas con la altura y homogeneidad
del dosel y con la penetraci6n de luz. Debido a esto, varias species que parecen defender de los
claros de Arboles caidos para la colonizaci6n y desarrollo de renovales en las porciones de tierras altas,
estAn menos restringidas en las regions bajas del gradiente topografico, donde las discontinuidades del
dosel pueden proveer suficiente luz para la persistencia de plAntulas y el desarrollo de renovales. Asi,
muchas species con una alta demand de luz no estuvieron asociadas con los claros de Arboles caidos
y aparecieron capaces de colonizar Areas relativamente abiertas del bosque de tierras bajas. En este
trabajo, los patrons de distribuci6n y de respuesta al disturbio y a la presencia de luz se discuten a
nivel de species, las que son puestas provisionalmente en agrupaciones ecol6gicas basadas en la
tolerancia aparente a los niveles de luz y a caracteres edificos y a otros factors ambientales. En Tikal
la mayor parte de la variaci6n natural de la composici6n y estructura del bosque, puede ser relacionada
a las condiciones topogrificas y edaficas.
Table of Contents
Introduction.............................. ............................................................. .................... ...................... 17
Acknowledgments ................................................... ......................... 17
Study A rea ................................................................................................................................................. 17
M ethods...................................................................................................................................................... 17
Data Analysis........ .............. ...... .. .......... ............................. ............ 17
Ordinations ..... ..................... ............................................... 17
Forest type classification............................. ... .................... .... .. ............. ..... 18
Distribution patterns for individual species..... ...................................................................... 18
Tests of responses to light and natural disturbance........................................................................... 18
Results and D iscussion.............................................................................................................................. 18
The predominant environmental gradient..................................................... 18
Ordination results........................ .................. ..... . ....... .... ..................... 18
Twinspan results compared to nested sorting of samples ............................................................... 18
Forest types ............................................................. ... .................... .............. ........................ 18
Distribution patterns of individual species.............................................................................. 2C
Species responses to canopy disturbance and light intensity......................................................... 2C
Ecological Groups..................................... ..... .... ..... ........... ............... 21
Conclusions .............................. ................. ............................ 21
Literature C ited.................................................................. ............................................... 22
Appendix 1. Species codes and membership in ecological groupings........................................... 22
Appendix 2. Tree species abundances by forest type..................................................... ........................ 23
F i u res........................................................................................................................................................ 2 4
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 171
INTRODUCTION
"If the traveler notices a particular species and wishes to find more like it, he
may often turn his eyes in vain in every direction. Trees of varied forms,
dimensions and colors are around him, but he rarely sees any one of them
repeated. Time after time he goes toward a tree which looks like the one he seeks,
but a closer examination proves it to be distinct He may at length, perhaps, meet
with a second specimen half a mile off, or may fail altogether, til on another
occasion he stumbles into one on accident" (Wallace 1878). As Alfred Russel
Wallace noted, two of the more distinctive characteristics of many tropical and
subtropical forests are high tree species richness and a high ratio of "rare" species
to common species in a given patch of forest (Hubbell 1979; Hubbell and Foster
1986b). One of the fundamental ecological questions pertaining to tropical forests
continues to be-what are the mechanisms that facilitate coexistence of so many
tree species? Ironically, it is the high species diversity itself that has hampered
investigations of the forces structuring tropical plant communities; adequate
sample sizes for the majority of tree species are often hard to obtain, and patterns
of species occurrence are difficult to discern amidst constant variation in species
composition.
The niche diversification paradigm, although invoked in early explanations
of tropical plant diversity (e.g. Ashton 1969), has been largely abandoned for
tropical forests, owing to the belief that small scale environmental variability in
lowland tropical areas is not sufficient to allow niche divergence of such a large
number of species. Hubbell and Foster (1986a) found that the coexistence of tree
species on Barro Colorado Island (BCI) in Panama was best explained by
stochastic processes and argued that adequate niche segregation to permit species
coexistence was highly unlikely in tropical tree communities, because high
species diversity results in largely unpredictable species identity of neighboring
individuals, rendering competitive niche divergence improbable. Stochastic
processes conveniently explain species coexistence, but do not adequately explain
observations of distinct species distribution patterns and patterns of species co-
occurrence. More recently Hubbell and Foster (1990) found evidence of weak
density dependent selection for many of the trees on BCI, suggesting that, as
hypothesized by Janzen (1970) and Connell (1971), increased sapling mortality
rates under conspecifics may play a role in preventing competitive exclusion.
Stochastic processes clearly play a role in shaping tropical tree communities, but
some question whether random processes are the principal forces in community
dynamics (Terborgh et al. 1996).
Other workers have emphasized treefall gaps as a major influence on tree
species composition and diversity in tropical forests (Ricklefs 1977; Denslow
1980, 1984, and 1987; Orians 1982; Pickett 1983; Brokaw 1985; Brandani et al.
1988; Nufiez-Farfin et al. 1988). Much attention has focused, with limited
success, on defining ecological guilds of tropical tree species based on their
modes of exploitation of treefall gaps. The one undisputed dichotomy in gap
exploitation is between pioneer and non-pioneer species. Some authors also
recognize an intermediate group of long-lived pioneer species (Whitmore 1992;
Finegan 1996), or large gap specialists (Denslow 1980). This long-lived pioneer
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
class has not yet been adequately defined with regard to life history attributes, but
is a first step in recognizing the considerable variation represented within the
catch-all 'climax' or 'mature forest' species category (Finegan 1996). The
pioneer/non-pioneer dichotomy essentially serves only to separate a handful of
tree species from the remaining hundreds of species within a given community. It
is increasingly recognized that light environments (Chazdon and Fetcher 1984;
Chazdon 1988; Canham et al. 1990) and species response patterns to light are
more varied than implied by the above dichotomy or even the pioneer/long-lived
pioneer/mature forest species trichotomy (Welden et al. 1991; Clark and Clark
1992).
Topographic and edaphic gradients have long been recognized as important
influences on species composition and structure of forest tree communities (e.g
Whittaker 1956, 1960). However in the tropics, after some early attempts to relate
plant species distributions to edaphic variation (e.g. Lundell 1937, Ashton 1969),
consideration of small-scale topographic and edaphic variation, until recently, was
largely absent from studies of tropical forest composition and community
structure. There is now ample evidence to support the hypothesis that edaphic
mosaics significantly increase the species diversity of tropical forests (Gentry
1981). Furthermore, it is clear that even small-scale variation in topography or
edaphic conditions can have a profound effect on plant species distributions in
many, if not most, tropical forests (Furley 1979; Lescure and Boulet 1985;
Lieberman et al. 1985b; Hubbell and Foster 1986b; Kahn 1987; Basnet 1992;
Johnston 1992; Steege et al. 1993; Oliveira-Filho et al. 1994; ter Tuomisto et al.
1994; Clark et al. 1995). Even with renewed research interest in edaphic
associations of tropical plants, our understanding of community responses to
environmental gradients in the tropics still lags far behind that in the north-
temperate zone, where gradient analysis is a well-developed tradition (e.g. Bray
and Curtis 1957; Whittaker 1967). The majority of studies published to date have
focused on a small subset of the woody flora (e.g. palms-Kahn 1987; Clark et al.
1995), or on non-woody taxa (Tuomisto et al. 1994). There have been few
community-level studies of tree species responses to edaphic gradients in the
tropics, and even less is known about the interactive effects on vegetation of
edaphic and other environmental factors, including variation in light availability,
over topographic gradients.
With an understanding of the patterns of spatial heterogeneity in species
occurrence within tropical forests, and of the processes underlying these patterns,
we would be better able to predict the long-term effects of human activities in
these forests and to inform conservation and management efforts. The ecological
associations of even most economically valuable tree species are not known with
precision, and it is not known whether species responses to disturbance will vary
under different edaphic conditions. In our study area, efforts to maximize
conservation of forest biota in the human-impacted buffer and multiple use zones
within the Maya Biosphere Reserve serve as another example of the need to
understand species responses to natural environmental variation. In this region
agricultural activities are focused in certain portions of the topographic
continuum, and remnant forest patches therefore rarely include these portions of
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 173
the vegetation continuum; this may result in local extirpations of plant and animal
species restricted to these habitats.
In the work reported here we attempt to document tree species responses to
both edaphic conditions and light availability as a means of understanding patterns
of variation in forest structure and composition. We attempt to understand the
degree to which species distribution patterns are predictable and to evaluate
mechanisms potentially contributing to tree species richness. We divide the
vegetation continuum into a functional forest type classification, providing a basis
for quantitative study of variation in plant and animal communities along the
predominant environmental gradient Additionally we present a provisional guild
classification of tree species in the Tikal region, incorporating both light and
edaphic tolerances, and responses to natural disturbance.
ACKNOWLEDGEMENTS
This is a contribution of the "Maya Project," a conservation and basic research effort of The
Peregrine Fund. The Maya Project has received funding from Robert Berry, the Archie W. Grace
Berry Charitable Foundation, Crystal Channel Foundation, Fanwood Foundation, Gold Family
Foundation, KENNETECH/U. S. Windpower, the John D. and Catherine T. MacArthur Foundation,
Mill Pond Press, National Fish and Wildlife Foundation, Norcross Foundation, Henry and Wendy
Paulson, Pew Charitable Trusts, Andres Sada, Joe and Flinda Terteling, the U. S. Agency for
International Development, and U.S. Man and the Biosphere Program/Tropical Ecosystems
Directorate. Preparation of the monograph itself was made possible by an NSF Graduate Fellowship,
and major financial support from Robert Berry and the Wolf Creek Charitable Foundation.
For laboratory soil analyses, we are grateful to Guatemala's Dirreci6n T6cnica de Riego y
Avenamiento, and especially to Licenciada Anabella Menindez Gudiel, formerly chief of that
institution's Laboratorio de Suelos y Aguas. For skilled assistance in field sampling, we are grateful to
Marcial C6rdova Alvarez of El Caoba, Flores, Pet6n. For cordial working conditions in Pet6n and
issuance of necessary permits, we are grateful to several directors each and other personnel of CONAP
and IDAEH, and are especially indebted to past and present administrators of Tikal National Park:
Rogel Chi Ochaeta and Rolando Pernillo. We are also grateful to the many other employees of Tikal
National Park who made us welcome. People in Guatemala who rendered assistance are too numerous
to recount, but special thanks are due to Sofia and Dr. Jorge Paredes for several kindnesses and to Don
Mundo and Dofia Pati Solis of the Jaguar Inn in Tikal and Santa Elena for continual assistance of many
kinds.
Nicholas Brokaw served as an early mentor to M. Schulze and was an inspiration for the research
reported here.
STUDY AREA
Tikal National Park lies in the northeastern corner of the department of
Peten, Guatemala at 170 N latitude. The forest here is classified as Subtropical
Moist (Holdridge 1971) or Tropical Semi-deciduous (Pennington and Sarukhan
1968), with annual rainfall of 1,300-1,500 nun and a pronounced dry season from
February through May. The Tikal area is significantly drier than subtropical moist
forest in northwestern Belize, just 70 miles to the northeast (Brokaw and Mallory
1993). While tree species diversity is considerably lower in northeastern Peten
than in wetter tropical forest at lower latitudes in Central and South America, it is
still extremely high relative to temperate zone forests, and the forest is subject to
similar dynamic processes as those of true rain forests farther south. Brokaw and
Mallory (1993) estimated between 250 and 300 tree species for the Rio Bravo area
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
of northwestern Belize. We believe the figure to be somewhat lower for Tikal
National Park as it represents a smaller area (57,600 vs 82,000 ha) with less
topographic diversity and rainfall than Rio Bravo, but a conservative estimate of
the total tree list for Tikal is over 200 tree species. In our study we sampled 142
tree species and observed 185.
The entire study region may have been deforested at the peak of Mayan
occupation of the area, ca. 1000 years ago, during which time some species may
have been locally extirpated and the relative abundances of the remaining species
may have been temporarily or permanently shifted. Despite this history of large-
scale disturbance, the forest can be considered mature, and has undoubtedly
achieved some form of equilibrium, with respect to structure, composition, and
patterns of species abundance. Although as recently as the 1960s portions of
Tikal National Park were subjected to low intensity logging, the forest areas that
we sampled lacked evidence of logging disturbance.
Topographic variation is not extreme in Tikal National Park, with the lowest
areas being 160 m and the highest 400 m above sea level. With the exception of a
large karst area in the northwestern one-eighth of the park, where slopes
exceeding 250 are common, the hills in Tikal are relatively gentle, often standing
only 30 m above nearby lowland basins. In contrast, the variation in vegetation
character can be drastic even over short distances. Tall-canopied, open-
understoried, palm-rich forest can be found within a few hundred meters of low,
nearly impenetrable, xerophytic scrub swamp forest (Fig. 1).
Soils in Tikal National Park are all limestone-derived clays but vary
considerably in texture from the upper regions of slopes, referred to hereafter as
"upland" sites, to the low-lying depressions known locally as "bajos." In addition
to clay content and texture, other significant differences between the soils are in
the depth, degree of weathering, pH, mineral and organic matter content, and
amount of rock fragments present. Soils in uplands are usually shallow and of
moderate clay content, with a high percentage of limestone fragments, providing
good to excessive drainage. In the lowest lying areas the soil is consistently more
than 70 cm deep (often >140 cm) and clay-rich, with few or no rocks in the upper
70 cm; the combination of nearly pure clay soil and low topographic position
results in poor drainage in the wet season and edaphic drought in the dry season,
as the water present in the soil is bound strongly to clay particles.
In this manner topography and associated edaphic characteristics combine to
create impressive changes in the vegetation over short distances. As soil
characters and topographic position are strongly correlated, vegetation
characteristics are typically similar along the same regions of the topographic
gradient at different sites, tempting designation of "forest types." Development of
a forest type classification system provides a practical means to compare species
distributions, forest structure, and microclimate along what is in reality a
vegetative continuum, and hence is fundamental to any phytosociological
analysis.
174
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 175
METHODS
We used a relatively small plot size (0.041 ha) to adequately sample the tree
community despite patchy species distributions, and to capture the full range of
topographic variation in our study area. This allowed us to distribute plots over a
17 kan by 9 km portion of Tikal National Park. Small sample plots provide more
quantitative data with tighter connections between vegetation samples and
environmental variables than do the less labor-intensive, point-based nearest
individual or presence/absence sampling techniques (Clark et al. 1995). In 1992,
we used circular plots to sample trees >7.5 cm dbh, and belt transects of 1.76 m x
20 m for sampling trees < 7.5 cm dbh, and >1 m tall. Plots were regularly spaced
at predetermined locations along transects that were sited in a stratified random
fashion, and along (50 m from) a series of archeological transects that bisect Tikal
National Park from north to south and east to west. Inter-plot distances ranged
from 100 m to 300 m. A total of 201 plots were sampled in 1992, yielding a total
area of 7.6 ha.
In 1993 a streamlined supplemental sampling was completed on 93 plots
also of 0.041 ha but in a more practical shape, 10 m x 41 m. These plots were
spaced 100 m apart along well-developed topographic gradients that were selected
on a topographic map. Despite differences in plot shape between 1992 and 1993
samples, we detected no differences in the number of stems recorded per plot by
the two methods and therefore treated the samples as equivalent.
Our methods of plot placement were not strictly random. We felt that an
even and widespread sampling of Tikal National Park would be more useful than
a truly random sample, and opted for a stratified random design. Gauch (1982)
makes a strong case that random placement of study plots is rarely achieved by
plant ecologists, and that systematic placement is often a better approach to
sampling. To avoid observer bias in plot selection, we determined plot locations
before entering the field.
For each 0.041 ha plot in 1992, all trees > 7.5 cm in diameter at breast height
(dbh = 1.3 m) were measured, identified to species, surveyed for presence of
aeroids or epiphytes on the trunk and in the canopy, and for vines > 2.5 cm in
diameter (for the palm Cryosophila stauracantha all individuals with a trunk
reaching breast height were recorded, regardless of dbh). Our 7.5 cm cut-off
differs from the traditional minimum diameter limit of 10 cm, but it was
considered more appropriate in our area, since in some vegetation types trees of
>10 cm dbh were only sparsely scattered amongst a sea of saplings and stunted
trees. When an individual could not be identified to species in the field, a sample
was collected if possible.
Three measures of canopy height were taken per plot, using a Haga
clinometer and measuring tape: that of the tallest tree in the plot, modal height of
the upper canopy surface, and the lowest point of the upper canopy surface within
the plot. Evenness of the canopy surface was classified into one of four
categories: even, even with emergents, uneven, or broken. Soil depth was
recorded at three locations 10 m apart in each plot. Two devices were used for
this measurement: a 70 cm long, pointed-tipped rebar staff pounded in with a five
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
pound mallet, and a 140 cm long auger. The maximum soil depth was considered
to be that at which penetration of the rebar probe was halted, or at which the
substrate changed to whitish decomposed limestone in the case of the auger. To
supplement these readings, soil pits were excavated to the level of un-decomposed
limestone fragments (c horizon), or 140 cm if no such layer was encountered, in
80 of the 201 plots. Slope and aspect were recorded for each plot using a
clinometer and compass, and one of eighteen topographic positions was assigned
(Table 1).
We collected soil samples from 15 cm to 30 cm depth at the center of each
sample plot. These were evaluated by a single observer, using the feel method
Table 1. Correspondence between topographic positions, 18 topographic position codes
used to describe them, and 11 recognized forest types and their codes.
Topo
C
graphic Description
:ode ______ *
1 hilltop
2 upper 1/3 of hill
3 rolling upland with Mayan house mounds or rock
outcrops
4 middle 1/3 of hill, or rolling upland without rock
5 lower 1/3 of hill
6 upper 1/3 of slope, with ravine, "shelf" or other
mesic factor
middle 1/3 of slope, with ravine, "shelf' or other
mesic factor
lower 1/3 of slope, with ravine, "shelf" or other
mesic factor
9 upland plain, not adjacent to low-lying area
10 upland plain, adjacent to low-lying area
11 elevated depression surrounded by upland forest
12 base of hill
13 low-lying area 200-300 m from upland area, slope
detectable
14 low-lying area >300 500 m from upland area,
slope detectable
15 low-lying area >400 m from upland area and < 800
m from upland
16 low-lying area > 800 m from upland area (periphery
of Bajo depression)
17 low-lying area, > 1 km from upland area (central
portion of Bajo depression)
18 low-lying area with standing water in dry season
18 ..
Forest
Code
1
1
2
2
Forest Type Name
Dry Upland forest
Standard Upland forest
11 11
3 Mesic Upland forest (1)
3 "
4 Mesic Upland forest (2)
4 "
* Cohune forest
5 Hillbase forest
6 Sabal forest
6 1
7 Transitional forest
8 Tall Scrub Swamp
Low Scrub Swamp
Mesic Bajo Forest
True Swamp
* Due to its rarity at Tikal, Cohune Forest was not given a forest type code.
176
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 177
(Foth 1972) for clay content and texture, and visual inspection for rock content.
We assigned each of the variables a value using an 1 1-point scale (0-10). The
laboratory of la Direcci6n Tecnica de Riego y Avenamiento in Guatemala City
subjected a subset of 72 samples to quantitative analyses. Textural ratings based
on the feel method were highly correlated with percent clay in laboratory analyses
(r2 = 0.84, p = 0.000) and when regressed on lab textures (adj. r2 = 0.71 and p =
0.1 x 1014). Based on this high correlation, we used our manual textural ranking
rather than laboratory results in ordinations, allowing us to use 201 samples rather
than 72.
For each plot all treefall gaps intersecting the plot, and those within 10 m of
the perimeter, were recorded and their length and width estimated. Young gaps
were recognized using Brokaw's (1982) criteria: holes in the forest canopy down
to a minimum height of 2 m. For older gaps (equivalent to "building-phase"
forest) boundaries were judged as the point of discontinuity between the upper
surface of regenerating vegetation and the surrounding intact canopy. To gain
some insight into the relative ages of gaps, the height range and modal height of
regenerating vegetation were estimated, and all fallen trees in the gap were
estimated for diameter and placed into one of five categories based on their degree
of decomposition. We used this same system to record all fallen trees within
sample plots whether or not a canopy gap was currently present.
Disturbance history of each plot was rated with respect to three variables:
disturbance type (natural, human-induced or uncertain), age since disturbance (0
to 2 yrs, 2+ to 5 yrs, 5+ to 10 yrs, and 10+ yrs) and intensity (none, light,
moderate, and heavy, based on the size of disturbance and percent of plot
affected). Determination of past disturbances was based only on structural
evidence (i.e. presence of large diameter logs, stumps, uniform canopy noticeably
lower than surrounding forest indicating regenerating gap, etc.) and not on
floristic composition. This avoided the circularity of using suspected pioneer
species as indicators of past disturbance. Time since disturbance was gauged
using the modal height and diameter of vegetation within the disturbed area, and
degree of decomposition of the fallen trees or stumps. For calibration of this
somewhat subjective aging technique we used personal observations of
regeneration in gaps of known age and the decomposition states of stumps of
known age, in addition to published accounts of regeneration and decomposition
rates (Lang and Knight 1977). Since only structural evidence was admissible in
this classification we were only able to identify disturbances in the range of 0-20
years old. Older disturbances, or smaller disturbances leaving evidence for a
shorter duration, were not recorded, although in many plots that lacked this
concrete evidence species composition indicated that a disturbance had occurred.
The understory subsample was an arm's width (176 cm) transect 20 m long
across the diameter of the circular sample plot, following an east to west compass
line. On this transect all plants > I m tall and < 7.5 cm dbh were identified,
grouped into one of three diameter classes (0-2.49 cm, > 2.5-4.99 cm, > 5.0-7.5
cm), and their height estimated. All vines intersecting the transect (at height < 2
m) were recorded and placed in diameter classes (as above) but were not
identified. Dead standing vegetation was recorded in the same way as live
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
vegetation. Dead hanging vegetation 2 m or lower was recorded in one of the
three diameter classes. Often a mass of dead hanging vegetation was encountered,
rather than one or two easily measurable branches. In this case we estimated the
equivalent number of individuals 25 cm long, of each of the three diameter
classes, that were contained in the mass. We also recorded ground cover density
using a sighting tube with cross-hairs to make 25 presence/absence readings
("hits" on green foliage). These were taken parallel to the transect line, but 1 m to
the side in order to avoid the effects of trampling. To supplement measurements
of ground cover density, we visually estimated the percentage of total ground
cover accounted for by monocotyledonous plants (primarily grasses and sedges).
Canopy cover was recorded using a spherical densiometer (Lemmon 1957).
Five readings were taken, one at the center of the arms-width transect, and one
each at 5 m and 10 m in both directions along the transect line. Limited sampling
of understory light intensity was conducted using hemispherical photographs at
sites in each topographic position lacking any signs of recent (within 15 years)
canopy disturbance. Photographs were digitized and analyzed using Hemiphot
(ter Steege 1996) for canopy cover and relative light intensity. These measures do
not correspond with vegetation sampling plots, but provide more precise estimates
of average understory light intensity than do densiometer readings.
For the 1993 sampling we omitted many of the measurements taken in 1992,
as the main purpose of this effort was to increase our sample size for trees larger
than 7.5 cm dbh, which were sampled in the same manner as the year before.
Understory vegetation was not sampled. Topographic position, slope and modal
canopy height (visual estimate), were recorded for each plot. A soil sample was
also taken. Disturbance history and gap occurrence were recorded, although sizes
of treefall gaps were not quantified.
DATA ANALYSIS
Plant species data analyzed were the number of individuals per species per
0.041 ha plot (for trees of 7.5 cm diameter or greater), and the number of
individuals per 20 m x 1.76 m transect for trees at least 1 m in height but < 7.5 cm
diameter.
Ordinations
Ordinations were performed using CANOCO (ter Braak 1987a), including
its subroutines for Detrended Correspondence Analysis (DCA, Hill 1979a) and
Canonical Correspondence Analysis (CCA). Vegetation classifications were
produced by a nested sorting of samples by topographic and edaphic variables and
were then compared to a species-based classification using two-way indicator
species analysis (TWINSPAN, Hill 1979b) and to the first axis scores produced
by DCA ordination of samples.
For all ordinations, species with fewer than five individuals encountered in
the sampling were omitted from the ordination and those with fewer than 15 were
made passive, i.e. their positions were plotted in ordination space but did not enter
into the ordination calculations. Several species displayed bimodal distributions,
178
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 179
having abundance peaks along two distinct portions of the predominant
environmental gradient. Such bimodality violates a primary assumption of both
DCA and CCA (Jongman et al. 1987; ter Braak 1987b), namely that species show
unimodal responses across coenoclines. To prevent such bimodal species from
skewing ordinations, we split each such species into two "pseudo-species" for all
full-gradient ordinations (14 species were split into pseudo-species). Ten samples
were eliminated from the primary ordinations as outliers; these samples were all
from highly disturbed areas. The final DCA ordination of the full gradient
included 283 samples, 98 active species and 17 passive species.
Indirect gradient analysis
Our initial analysis utilized Detrended Correspondence Analysis (DCA) for
an unconstrained ordination of samples and species. The purpose of this
ordination was to reveal patterns within the data set using an objective method and
to provide a means of evaluating constrained ordinations produced by CCA and
forest type classifications. Detrending by second-order polynomial and by
segments yielded similar results-we chose to use detrending by segments for the
final ordinations. Species were scaled as weighted mean sample scores and
samples as weighted mean species scores.
Direct gradient analyses
In addition to these unconstrained ordinations (indirect gradient analyses) we
used CCA to directly examine the association between patterns of tree species
composition and patterns of environmental variation at Tikal. Environmental
variables were incorporated into CCA as described below, and were centered and
standardized prior to analyses.
Physical factors.- After initial examination of environmental variables we
decided to use only four such variables for analysis of the large tree data:
topographic position (18-point ordinal scale), soil clay content (11-point ordinal
scale), soil rock fragment content (11-point ordinal scale), and slope (degrees).
Our ordinal ranking of clay content was highly correlated with percent clay
content as determined by lab analysis of 72 samples (Pearson correlation coef. =
0.844, p < 0.001). Although we collected data on soil depth for 201 samples,
comparison of our depth measures with actual depth revealed by soil pits eroded
our confidence in this index. Furthermore, comparison of soil pit depths and soil
texture indicated that soil depth and clay content were closely correlated (Pearson
correlation coef. = 0.854, p < 0.001) with deep, clay-rich soils occurring in low-
lying areas and shallower soils, lower in clay, occurring in upland sites. Slope
aspect had no detectable effect on vegetation and was omitted from consideration.
Natural disturbance history.- Recent human disturbance was not a
significant factor in any of the plots; in this study "disturbance" therefore refers to
natural canopy openings from treefalls. In the initial CCA ordination, data on
disturbance and degree of canopy opening were not included. In later analyses,
variables reflecting disturbance history and degree of canopy opening were used
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
as primary variables, with topography, soil texture, soil rock content, and slope
included as covariables, in order to remove their effects and allow focus on the
primary (disturbance) variables. Using Monte Carlo tests we evaluated the
significance of six variables reflecting canopy disturbance and understory
environment: intensity of canopy disturbance, age of canopy disturbance, number
of gaps per plot, total gap area within the 0.04 ha plot, mean canopy height, and
percent canopy opening (mean). Following forward selection of disturbance and
light intensity variables, we used CCA to examine the relation between
disturbance history and light availability and distributions of juvenile trees (<7.5
cm dbh) and treelets arborescentt plants that do not attain diameters >10 cm). For
adult trees only the intensity and age of canopy disturbances were used as primary
variables, with the four environmental variables as covariables, as current light
environment may have little relation to conditions present during the
establishment and recruitment of adults.
Forest Type Classification
In our preliminary designation of forest types we used a hierarchical sorting
of samples based on four environmental variables in the following order:
topographic position, soil clay content, slope, and soil rock content. All samples
were sorted first by topographic position, and then within each topographic class
by soil clay content, followed by slope and finally rock content. The relative
importance given to these four environmental variables was based on the strength
of correlation of each variable with species data, as determined by CANOCO
(Monte Carlo significance tests and length of lines in biplots). This hierarchical
sorting yielded a sample ordering consistent with that obtained from the first axis
of the CCA ordination. From the hierarchical ordering we grouped samples into
11 forest "types." Breakpoints between these 11 forest types coincided with
breakpoints between the original 18 topographic positions. To a large degree,
forest type designation was merely a condensation of topographic positions, but
information on species composition was also considered in arriving at groupings
of the 18 topographic positions into 11 groups that reflected observable
differences in the forest along the topographic gradient.
The above method of forest type classification is straightforward and relies
primarily on topographic and environmental data, but it does require some
subjectivity in the actual dividing of the sample sequence into forest types. For
this reason we felt it necessary to compare this classification to an independent,
objective grouping of species and samples. To achieve this, we applied
TWINSPAN to the full data set and then to two subsets (upland and low-lying
sites) to classify samples into "pseudo-forest types" based on species composition.
Classifications resulting from TWINSPAN were compared to the hierarchical
classification described above, which was in turn compared to sample ordering
produced by the first axis sample scores of the DCA ordination, to insure
consistency between forest type designations and patterns of sample distribution
in ordination space.
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 181
Distribution Patterns For Individual Species
To supplement the full community ordinations provided by DCA and CCA,
we examined individual species abundance patterns along the topographic
gradient. After testing several methods, we favored grouping samples into the 11
final forest types/topographic zones. Using the mean number of occurrences of
each species per sample (with standard error) in each forest type, we were able to
study distribution patterns of adults and juveniles in greater detail, and adequately
represent bimodal distributions.
Tests of Responses to Light and Natural Disturbance
Our goals in this portion of the study were to: (1) test the hypothesis that the
impact of gaps and other canopy disturbance on understory light environments (as
revealed by tree regeneration patterns) varies with topography, (2) determine
whether there are some species that require high light intensity for regeneration,
but for which most regeneration does not occur in treefall gaps, (3) compare
associations of saplings versus adult trees of a given species with disturbance and
light conditions, to see whether different conditions are required for attainment of
sapling and adult size, (4) assign tree species to response guilds with respect to
disturbance and light environment, and (5) determine whether small-scale
differences in canopy opening (e.g. between <10% and 10-20% canopy opening)
have an effect on species distribution within a habitat type.
Tests of the effects of canopy opening, disturbance history, and topographic
position were accomplished using multiple random permutation procedure Chi-
square tests (MRPP, Berry and Mielke 1986) and analysis of variance, using
SYSTAT (1990, SYSTAT, Inc.) MGLH Model statements. For the analysis of
variance tests of disturbance, those sites receiving a disturbance intensity rank of
zero or "low" were combined into a single "undisturbed" group, and sites with
"medium" or "high" disturbance were combined into a "disturbed" group. For
degree of canopy opening, two cut-points were selected to divide the resulting
distribution into three groups with approximately equal sample sizes (0-10%,
>10-20% and >20% canopy opening); this variable was log-transformed prior to
ANOVA.
Each species distribution was subjected to a Chi-square goodness of fit test,
to test numbers of trees observed in samples of each class of disturbance history
and canopy opening against that expected under the null hypothesis of no
correlation between these variables and species distribution. Juvenile and large
tree (> 7.5 cm dbh) data were tested separately. For chi-square tests of
disturbance and canopy opening effects, we divided samples into three groups
reflecting positions on the dominant topographic gradient, with one group
comprising upland samples, another group transitional/hill-base samples, and the
third group samples in low-lying areas that experienced periodic inundation. We
used the same two disturbance classes as above. For analyses of juvenile
distributions, disturbances of all ages were included, while for large tree analyses
only disturbances with estimated ages of 5 years or more were used.
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
In each portion of the topographic gradient, Chi-square tests utilized two
classes of canopy opening: "relatively open" and "relatively closed." In the
upland and transitional areas, only a few samples had >20% canopy opening;
hence "open" sites were those with >10% opening, while "closed" sites were those
with 10% opening or less. In contrast, in low-lying sites, no samples had less than
10% canopy opening while many exceeded 20%; hence, in this sample group,
"open" sites were those with greater than 20% opening and "closed" sites were
those with 10-20% opening.
Wherever multiple comparisons were made, we maintained table-wise
significance at stated levels by using a simultaneous Bonferroni procedure (Rice
1989) to adjust rejection regions. However, we have discussed results for both
Chi-square and ANOVA tests that were not significant after Bonferroni
adjustments, but were significant prior to such correction (alpha < 0.05). These
marginally significant results are clearly distinguished from statistically
significant results (after Bonferroni) in both tables and text.
RESULTS AND DISCUSSION
The Predominant Environmental Gradient
Although the topographic gradient in Tikal is modest, several physical
factors are strongly correlated with it and with one another, and this complex
factor gradient proves to exert a strong influence on the vegetation. Figure 2
depicts the manner in which several physical factors vary along the topographic
gradient, while Figure 3 depicts intercorrelations among these factors. Percent
clay, organic matter content, and pH were all highly correlated with one another
and with topographic position (linear regression of clay on topographic position
r2=56.4, p = 0.3 1x10'3). Organic matter and pH were positively correlated with
one another (linear regression r2 = 17.5, p = 0.3 1x0"3) and negatively correlated
with clay content (linear regression r2 = 47.9, p = 0.23x10'_ o.m./clay; r 2= 40.1, p
= 0.4x10'8 pH/clay), so that organic matter decreases and pH drops as one moves
into regions of higher clay content. Clay content is strongly related to topographic
position, being low in upland sites and dramatically higher in low-lying areas.
Rockiness is higher in upland than in low-lying sites, while soil depth is shallow
on upland sites (ranging from 20 m to 70 cm) and generally deep in low-lying
sites (often >1.4 m).
The net result is that organic matter and pH are high in well-drained upland
sites, which also feature shallow, rocky soil that is low in clay content; in contrast,
organic matter, pH, and rockiness are low in low-lying sites, which feature deep,
rock-free, clay-rich soils. Observations indicate that drainage is generally good
on upland sites, which do not accumulate standing water. In contrast, low-lying
areas may be inundated for weeks at a time during the rainy season (with some
areas being subject to strong currents of flowing water), while their high clay
contents and visible cracking suggest that edaphic drought may occur during the
dry season.
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 183
In sum, we feel that the environment for plants along the topographic
continuum may be characterized as follows: (1) on the upper reaches of slopes
(Dry Upland Forest), soils and topographic features result in good to excessive
drainage and probably some dry season drought stress for plants, nutrient
availability is probably not limiting, and pH is moderate; (2) in upland areas
occurring lower on the gradient (Standard and Mesic Upland Forests) deeper soils
with relatively low clay and high rock content provide moderate moisture
conditions throughout the year, relatively high nutrient availability, and moderate
pH; (3) low-lying areas near the bases of slopes (Hill-base and Sabal Forests)
appear to experience wet season soil saturation but not excessive waterlogging,
and dry season drought is not extreme, while nutrient availability and pH are
probably not much lower than in upland areas; (4) areas on the edges of lowland
depressions (Transitional Forest) experience some wet season inundation as well
as dry season drought, and soils show signs of gleying and are lower in nutrient
content and pH than soils higher up on the gradient; (5) soils in lowland
depressions (Tall and Low Scrub Swamp) experience extreme wet season
inundation and extreme dry season edaphic drought, are highly gleyed, and may
have nutrient and acidity levels that limit plant growth; (6) in the lowest lying
areas (Mesic Bajo) soil characters are similar to those in Scrub Swamp, but dry
season drought may be ameliorated by the proximity of the dry season water table
to the soil surface.
Ordination Results
The first axis of the full-gradient ordinations of samples and species
produced by DCA (Figs. 4-7) corresponded well with topographic position, with
upland samples grouped at the left and samples farther to the right belonging to
increasingly lower regions of the topographic gradient (passive correlation of
topography, clay content, rock content and slope with first axis = 0.792). The
species plot (Fig. 5) was consistent with this interpretation. Species generally
restricted to xerophytic swamps or transitional forest are located in the right
portion of the graph, while species with abundance peaks in lowland forests but
not common in the extreme xerophytic swamps are located in the lower center
region. Species occurring in Upland Forest but with abundance peaks in more
mesic transitional lowland forests are found in the center of the graph, while at the
left hand extreme appears a group of species occurring most commonly in upland
sites.
The second axis in this ordination (Fig. 4) separated lowland samples into
mesic swamp sites in the upper right of the graph and the xerophytic Scrub
Swamp sites in the lower right. This axis accounted for variation between
extreme lowland sites, hence upland sites are tightly clustered on the left of the
plot. Amplifying the left-hand portion of this plot (Fig. 6) makes it clear that the
DCA ordination did separate upland sites into drier sites occupying the upper
portions of slopes [on left] and wetter upland sites lower on the topographic
gradient and with deeper, more clay-rich soils [on right]). In the biplot of species
centroids (Fig. 7), species to the left ofMalmea depressa were all more abundant
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
in drier upland samples, with species like Gymnanthes lucida, Drypetes lateriflora
and Thouinia paucidentata poorly represented in wet upland samples.
Conversely, the species on the right side were significantly more common in
lower, more mesic areas of the upland topographic gradient.
DCA was also performed on understory samples of treelets and tree
juveniles, which resulted in virtually identical ordering of samples and species to
the DCA of trees >7.5 cm dbh; hence, ordination plots are not pictured here.
Eigenvalues for the first two axes were also similar to those for the ordination of
adults (first axis = 0.735, second axis = 0.471); and passive correlation with
topographic position, clay content, rock content and slope showed a strong
association of the first axis (0.789) with these variables, and the second axis was
weakly correlated (0.286). The percent variance in species composition accounted
for by the first two axes was low (first axis = 9.2%, second axis 5.9%).
The CCA ordination using the four primary environmental variables showed
that topographic position and soil composition were significantly associated with
the variation in species composition between samples (Fig. 8). The first axis was
largely related to topographic position and the second to soil traits. Monte Carlo
significance tests of environmental variables showed topography (p = 0.01) and
clay content (p = 0.03) significant at the 0.05 level. Rock content and slope were
not significant at the 0.05 level after fitting of topographic position and clay
content but were included as they seemed to enhance the ordination. Monte Carlo
significance testing of the overall ordination and the first ordination axis (p = 0.01
both tests), as well as the close correspondence of these direct ordination results
with the indirect ordinations above, indicate the significance of this ordination,
despite the fact that it accounted for a low percentage of the total variance in the
species data. Tree species distribution in our study area is of a characteristically
patchy nature, there are a large number of rare species, and our data set includes a
relatively large number of small samples (mean number of individuals per sample
= 30). All of the above factors produce a large amount of inter-sample species
variation which is not explained by environmental conditions. In addition, much
of the variation in species composition is due to the disturbance history of each
plot. Hence, it should be expected that a constrained ordination such as CCA will
only account for a small percentage of the variance within a large data set, and this
is typical of the technique (Jongman et al. 1987).
Although results are not pictured here, constrained ordination of understory
vegetation (treelets, shrubs and tree saplings) produced an ordering of samples and
species similar to that provided by CCA for trees >7.5 cm dbh This sapling
ordination also showed that species composition was significantly related to
variation in topography and clay content, although the first two axes only
accounted for a fraction of the total species variance (first axis = 6.1%, second
axis = 0.8%). Eigenvalues and species-environment correlations, respectively,
were as follows: first axis = 0.490, second axis = 0.055; first axis = 0.867, second
axis = 0.551. Monte Carlo significance testing of topography (p = 0.01) and clay
(p = 0.04) again showed a significant correlation between these two
environmental variables and patterns of variance in species composition. In the
CCA ordination of juveniles along disturbance and light intensity gradients (Fig.
9), only percent canopy opening (p = 0.01) and total area of gap and building
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 185
phase forest within the plot (p = 0.03) were found by Monte Carlo testing to be
significantly correlated with variance in species composition after fitting the four
environmental covariables. In Figure 9, species associated with high understory
light intensity but not with treefall gaps were clustered in the lower right of the
graph. These species are primarily those that were abundant in the relatively
open, low-lying, or 'bajo' forests, but that were not commonly found in treefall
gaps in Upland Forest. Species clustered in the upper central portion of the plot
showed positive correlations with both treefall gaps and understory light intensity,
while species clustered at the upper left of the graph showed strong positive
correlations with treefall gaps, but not with current light intensity. The remaining
species in the lower left showed no association, or negative correlation, with both
light intensity and treefall gaps.
For trees > 7.5 cm dbh (Fig. 10), the correlation of canopy disturbance with
species composition was weaker in our data set than for juveniles (Monte Carlo
significance testing of environmental variables: disturbance age-p = 0.07,
disturbance intensity-p = 0.10). This was likely due to the fact that many of the
trees sampled were substantially older than the maximum age of canopy
disturbance that we were able to document. Hence, for an older tree the lack of
evidence of canopy disturbance in the vicinity does not preclude the possibility
that the individual became established or recruited to adult size in a treefall gap.
Despite the limitations of our data, the CCA ordination did distinguish species
with strong treefall gap associations from less light-demanding species, with gap-
associated species (including several pioneer species) clustered at the upper right
of the plot (Fig. 10).
TWINSPAN Results Compared to Nested Sorting of Samples
Using TWINSPAN to achieve simultaneous sorting of samples
encompassing the full topographic continuum did not produce meaningful sample
groupings. However, when upland and lowland samples were separately
classified, TWINSPAN recognized two upland and six lowland groups for a total
of eight "forest types." In comparison, our nested ordering of samples by
topography and soil characteristics produced four upland (including the rare
cohune upland) and seven lowland (considering Mesic Bajo and True Swamp as
separate types) sample groups. Of the 152 lowland samples, 92% were placed in
the same groups by TWINSPAN and our nested sorting technique. Of the
lowland samples classified differently by the two techniques, 62% were classified
as either Transitional or Mesic Bajo by the nested sorting technique. We found
that these samples were not consistently separated from one another by
TWINSPAN, nor were they placed into two categories corresponding to those of
the nested sorting method. Rather, both "types" were included together in each of
several TWINSPAN groupings; some samples were allocated to TWINSPAN
groups corresponding to our Sabal and Tall Scrub Swamp categories and
TWINSPAN separated the remaining samples into four small groups.
The above result is not surprising, as there were substantial similarities in
species composition between these samples. The failure of TWINSPAN to
separate these two forest types may result from intrinsic problems in the divisive
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
method of this analysis (van Groenewoud 1992) and from limitations on the
number of indicator species used per analysis. None of the samples classified as
Transitional Forest by the nested sorting occupied portions of the topographic
gradient adjacent to samples classified as Wet Scrub Swamp. Additionally,
several tree species, such as Pithecellobium belizensis and Lonchocarpus
guatemalensis were found in plots we classified as Wet Scrub Swamp but not in
those we classified as Transitional Forest. Therefore, we felt that combining
samples of these two types as suggested by TWINSPAN would lead to less clarity
than that provided by our nested topographic sorting in this case.
The differences in the classification of upland samples by the two techniques
is largely due to the relative degree of splitting resulting from the two methods.
To compare the three primary upland sample groups resulting from nested sorting
with the two groups generated by TWINSPAN, we combined these three into two
groups, based on topographic position. Grouped in this way, 92% of the 141
upland samples were placed in the same groups by both techniques. The indirect
ordination of samples by DCA supported division of upland samples into three
groups (Fig. 6), and the ordering produced by the first axis DCA scores is highly
correlated with that produced by the nested sorting (Spearman correlation coef. =
0.902, p = 0.000).
To summarize, both TWINSPAN and ordinations generally supported the
division into forest "types" resulting from our nested sorting on topographic
position and soil factors, but differed in some details. Our final classification rests
mainly upon the results of the nested sorting, and designates 11 forest types,
including four upland types. Due to a large sample size for Upland Wet forest
sites, for subsequent analyses we have divided this forest type into two subgroups
based on topographic position. In addition, Swamp and Mesic Bajo sites have
been lumped for analyses due to low sample sizes and high similarity in structure
and composition between these forest types. One forest type, Cohune Forest, was
only documented in one small area of Tikal, and has therefore been omitted from
further analyses.
Forest Types
Any classification of forest types is artificial, but a good classification can be
useful in comparing tree species distributions, vegetation structure,
microenvironment, and other attributes along what is actually a continuum of
variation. Moreover, classification is a prerequisite to many kinds of ecological
research, as sampling must take into account the underlying vegetation continuum
in order to facilitate valid choice of replicate sample plots and to define valid
comparison units and scope of inference. Lundell (1937) classified the vegetation
of northern Petdn into three primary forest types,: two upland forest types,
Ramonal and Zapotal (including Caobal association) and lowland wooded swamp
(Akalches). Using quantitative data on forest structure and composition we were
able to recognize several upland and lowland vegetation types. In this section we
summarize the structural variation of vegetation along the topographic gradient in
Tikal and then describe and compare the 11 forest types recognized by our
classification with respect to structural and environmental traits, characteristic
186
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 187
species, patterns of species diversity, and rates and types of dynamic processes
influencing the sub-canopy light environment. The forest types are listed in order
of their occurrence along an idealized topographic gradient (Fig. 11). Wherever
appropriate, we have conformed to the terminology of Brokaw and Mallory
(1993) in naming forest associations.
There was a pronounced decrease in canopy height and basal area along the
topographic gradient, from well-developed Upland Forest through low, dense
Scrub Swamp forest in the low-lying depressions (Table 2, Figs. 12 and 14). The
principal exception to this pattern occurred in the Mesic Bajo and True Swamp
types, which occur lower down on the topographic gradient than Scrub Swamp,
but are taller and have greater basal area than the Scrub Swamp. Canopy opening,
density of dead standing stems and live stem density (>1 m tall, >7.5 cm dbh),
mirrored the above pattern, with the highest values occurring in Low Scrub
Swamp, and decreasing values with increasing elevation along the topographic
gradient (Table 2, Fig. 12). However, the density of stems >10 cm dbh was
virtually the same across the gradient, though small stems accounted for a larger
percentage of total stems in lowland forests than in upland portions of the gradient
(Fig. 13). Conversely, stems >30 cm dbh accounted for up to 68% of the basal
area (of stems >7.5 cm) in Upland Forest, while contributing less than 30% of the
basal area in Scrub Swamp (Fig. 14). Mesic lowland forests (Hill-base, Sabal,
Transitional and Mesic Bajo) displayed intermediate values with a high density of
10-30 cm dbh stems and scattered large individuals (Fig. 14).
Vine density increased from upland to lowland portions of the gradient
(Table 2, Fig. 12). However, the size distribution of vines varied considerably
between lowland forest types, with Scrub Swamp areas dominated by small vines
(<2.5 cm diameter), and Hill-base, Sabal, and Mesic Bajo forest displaying a high
ratio of large to small vines (Table 2). Dead hanging vegetation was dramatically
more abundant in the understory of lowland forest areas than in upland, a result of
the higher densities of live understory vegetation in these forest types (Fig. 12).
Similarly, ground cover (stems <1 m tall) was much higher in lower than upper
regions of the topographic gradient, with the exception of Mesic Bajo and True
Swamp in which standing water for much of the year probably reduces seedling
density and lower light prohibits the growth of grasses and sedges (Table 2, Fig.
12). In Scrub Swamp, sedges accounted for the overwhelming majority of ground
level stems (Table 2, Fig. 12).
Species diversity did not decrease from Upland Forest to lowland areas,
although one might predict such a decrease from the trend of decreasing canopy
height and increasingly broken structure along the topographic gradient, and the
harsh edaphic extremes found in the low-lying depressions. Instead, species
richness (Fig. 15) and dominance (not pictured) were roughly equal between Dry
Upland Forest and Low Scrub Swamp, which are at or near (when Mesic Bajo
occurs below Scrub Swamp) the endpoints of most topographic gradients in Tikal
(Fig. 11). However, the high stem density in Scrub Swamp relative to that typical
of upland forest types might counteract the environmental conditions that promote
low diversity, due to the density effect on species diversity (Denslow 1996). The
highest number of species per unit area was found in the more mesic lowland
Table 2. Structural and environmental variation along a topographic gradient, Tikal, Guatemala. Means and standard errors are given for each often forest types.
D.U. D.U. U.S. U.S. MU.1 MU.1 MU.2 MU.2 H. H. S. S. T. T. T.S.S T.S.S L.S.S. L.S.S M.B. M.B.
avg se avg se avg se avg se avg se avg se avg se avg se avg se avg se
Maximumcanopyhgt(m) 26.8 0.7 26.7 0.8 25.0 0.8 25.8 0.5 24.8 1.0 23.1 1.1 18.6 1.0 18.8 0.9 12.1 0.8 19.0 2.2
Mean canopy hgt(m) 20.6 0.5 20.8 0.7 19.7 0.6 19.9 0.4 17.8 0.8 17.2 0.5 14.5 0.7 12.7 0.9 10.1 0.5 13.3 1.4
Minimum canopy hgt(m) 15.6 0.5 16.9 0.7 15.2 0.7 15.2 0.4 13.0 0.7 12.9 0.6 10.9 0.7 9.1 1.1 8.1 0.3 9.9 1.1
Mean canopy opening (%) 8.9 0.6 7.6 0.6 8.7 0.6 8.5 0.5 10.3 0.8 12.1 0.9 14.1 1.8 19.3 0.7 28.6 2.3 15.2 1.2
Average standard deviation
of canopy opening (%) 1.7 0.2 2.0 0.2 2.1 0.3 2.4 0.3 2.6 0.4 3.2 0.4 4.8 1.1 4.7 0.8 6.4 0.9 3.9 1.1
Basal area (cm2 per plot) 14468 11438 9850 10748 8878 8517 9511 8301 8173 7693
Stem density (>2m tall) 500 474 400 430 398 430 570 675 1331 503
# treefall gaps in plot 0.5 0.2 0.5 0.1 0.5 0.1 0.5 0.1 0.9 0.2 0.7 0.2 1.5 0.3 1.4 0.3 0.2 0.2 1.1 0.5
% of plot covered by gap
or building phase forest 0.2 0.1 0.3 0.1 0.3 0.1 0.2 0.1 0.5 0.1 0.2 0.1 0.4 0.1 0.3 0.1 0.1 0.1 0.3 0.1
Tot area of gap/building
phase forest within 10 m
radius of plot boundary (m) 15.3 6.3 25.8 9.7 28.6 9.2 24.3 6.5 35.1 14.0 37.5 19.4 30.5 8.6 73.4 25.2 4.8 4.6 27.5 12.2
Average size of treefall
gaps (in) 119.3 127 91.9 85.1 109.9 60.9 151.8 169.2 119.3 127.8 156.3 123.8 66.8 94.9 106.9 90.7 43.8 32.5 70.1 96.5
6
# trees leaning > 25 degrees
off vertical per plot
# dead hanging branches
0-2.49 cm diameter *
# dead hanging branches
2.5-4.99 cm diameter *
# dead hanging branches 5
cm+ diameter*
Total # dead hanging
branches *
0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 2.7 0.7 2.9 0.6 2.2 0.6 2.7 0.6 1.5 0.6 6.1 1.0
9.6 2.1 4.9 1.7 6.4 1.7 6.3 1.3 8.7 2.8 11.6 2.9 27.2 13.2 21.5 5.0 52.8 11.7 60.0 24.2
1.0 0.4 0.6 0.4 1.3 0.6 1.5 0.7 1.0 0.5 2.3 0.8 2.4 1.5 3.1 1.7 3.9 1.0 2.8 1.5
0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.2 0.2 0.5 0.4 0.5 0.5 0.0 0.0 0.5 0.5 0.0 0.0
10.7 2.2 5.5 1.7 7.7 1.9 7.9 1.7 9.9 3.2 14.4 3.0 30.1 13.9 24.5 5.5 57.2 11.9 62.8 24.3
D.U. D.U. U.S. U.S. MU.1 MU.i MU.2 MU.2 H. H. S. S. T. T. T.S.S T.S.S L.S.S. L.S.S M.B. M.B.
avg se avg se avg se avg se avg se avg se avg se avg se avg se avg se
# dead standing stems 0-
7.5cmdbh* 1.9
Percentage of ground cover
accounted for by monocots
(grass & sedges) 0.2
Percent ground cover 16.4
# vines 0-2.49 cm diameter -
understory 8.7
# vines 2.5-4.99 cm
diameter- understory 1.3
# vines 5 cm+ diameter -
understory 0.3
Total # of vines* 10.3
# trees per plot with 1-2
vines 0-2.5 cm diameter 4.4
# trees per plot with 3-5
vines 0-2.5 cm diameter 1.8
# trees per plot with 6-10
vines 0-2.5 cm diameter 0.8
# trees per plot with >10
vines 0-2.5 cm diameter 0.1
# trees per plot with 1-2
vines >2.5 cm diameter 0.4
# trees per plot with >2
vines >Z.5 cm diameter 0.0
0.3 1.2 0.4
1.7 0.4 1.5
0.3 0.7 0.2 1.3 0.4 1.6 0.9 3.0 0.7
0.1 0.8
1.3 17.5
1.6 7.1
0.3 0.8
0.1 0.2
1.7 8.1
0.6 4.2
0.3 1.5
0.2 0.4
0.1 0.1
0.1 0.2
3.5 19.1
4.0 26.7
6.6 18.2
1.0 2.5
0.1 0.1
7.5 20.8
0.6 15.9
0.7 3.5
0.8 0.9
0.4 0.8
0.3 1.9
4.7 0.8 2.0 0.4
6.1 99.3
4.0 38.2
4.0 13.3
0.7 1.2
0.1 0.0
4.3 14.5
10.1 3.5
0.9 0.5
0.4 0.0
0.4 0.0
0.5 0.0
0.0 0.1 0.1 0.0 0.0 0.1 0.0 0.3 0.1 0.1 0.1 0.0 0.0 0.2 0.2 0.0 0.0 0.2 0.2
# trees per plot with
small (<20 cm diameter) 0.2 0.1 0.4 0.1 0.2 0.1 0.4 0.1 0.5 0.2 0.3 0.1 1.4 0.5 3.2 0.5 5.5 1.0 2.8 0.8
epiphytes on trunk
D.U. D.U. U.S. U.S. MU.1 MU.1 MU.2 MU.2 H. H. S. S. T. T. T.S.S T.S.S L.S.S. L.S.S M.B. M.B.
avg se avg se avg se avg se avg se avg se avg se avg se avg se avg se
# trees per plot with
aeroids on trunk 0.2 0.1 0.6 0.2 1.0 0.3 0.9 0.2 0.9 0.4 1.2 0.5 0.1 0.1 0.3 0.2 0.0 0.0 0.2 0.1
# trees per plot with
large (>20 cm diameter) 0.2 0.1 0.2 0.1 0.3 0.1 0.2 0.1 0.6 0.2 0.9 0.3 0.9 0.3 2.5 0.6 16.1 1.4 3.4 1.2
epiphytes on trunk
# trees per plot with
large (>20 cm diameter) 7.4 0.8 5.2 0.7 3.3 0.6 3.6 0.4 2.7 0.5 1.9 0.6 3.5 0.7 3.0 0.9 2.4 0.6 4.2 1.9
epiphytes in crown
Degree slope of land 9.0 1.2 7.9 1.6 6.5 1.0 2.9 0.3 1.6 0.5 1.1 0.1 0.5 0.3 0.0 0.0 0.0 0.0 0.1 0.1
Mean soil clay content 4.5 0.5 4.9 0.6 5.7 0.5 5.7 0.3 9.0 0.4 9.9 0.1 10.0 0.0 9.9 0.1 10.0 0.0 10.0 0.0
Mean soil rock content 3.0 0.4 2.3 0.4 2.4 0.4 2.0 0.3 0.4 0.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Soil depth (cm) 40.7 3.4 45.4 5.3 52.2 6.3 54.8 5.4 105.3 21.5 143.0 0.0 143.0 0.0 143.0 0.0 143.0 0.0 143.0 0.0
Forest type codes are as follows: D.U.=Dry Upland Forest; U.S.=Upland Standard Forest; MU. l&MU.2=Mesic Upland Forest; H.=Hillbase Forest; S.=Sabal
Forest; T.=Transitional Forest; T.S.S.=Tall Scrub Swamp; LS.S.=Low Scrub Swamp; M.B.=Mesic Bajo. >
* = unit area of 35.2 mL
All other area based measures are per 0.041 ha.
z
PO
0
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 191
Table 3. Degree of overlap in species composition between forest types, Tikal, Guatemala.*
D.U.F. U.S.F. M.U.1 M.U.2 H.F. S.F. T.F. T.S.S. L.S.S. M.B.
Dry Upland Forest 16 15 15 11 10 8 5 1 3
Upland Standard For. 77 17 17 14 11 10 5 1 4
Mesic Upland Forest 1 76 76 17 14 12 11 5 1 4
Mesic Upland Forest 2 77 78 82 15 13 11 5 1 4
Hillbase Forest 64 70 70 73 15 11 5 1 6
Sabal Forest 64 63 73 75 72 11 6 1 8
Transitional Forest 63 60 68 70 65 69 6 1 8
Tall Scrub Swamp 47 49 46 53 49 52 73 12 9
Low Scrub Swamp 30 31 30 38 39 45 58 68 7
Mesic Bajo 46 48 49 43 53 62 62 58 52
* The upper half of the table gives the number of tree species in common between forest type pairs, out
of the 20 most abundant species per forest type. The bottom portion of the table depicts the percentage
similarity (% of species found in one forest type that were found in both) between total species lists
between forest type pairs.
forests, Sabal and Transitional, which include both upland species that are unable
to persist in more extreme lowland conditions, as well as lowland species that only
rarely colonize Upland Forest. Forest types that were close together on the
topographic continuum had a much greater overlap in species composition than
those that occurred farther apart, as would be predicted by the relative similarity
of environmental conditions in adjacent topographic zones and the greater
probability of seed exchange between spatially contiguous stands (Table 3). At
the extreme, only 30% of the total number of species found in either Dry Upland
or Low Scrub Swamp were recorded in both forest types.
Dry Upland Forest
Environmental factors and vegetation structure
This forest type (Fig. 16) occurs on the higher portions of slopes or in other
rocky upland areas such as those with small Maya structures, and is particularly
distinctive on tall and steep slopes. Dry Upland Forest corresponds roughly with
Lundell's (1937) Ramonal. Here the combination of rapid surface drainage with
shallow, rocky soils underlain by porous limestone, creates relatively dry
conditions compared to other upland areas, and moisture stress is undoubtedly
amplified during the dry season, especially in drought years. The shallow soils
also limit vertical root penetration, favoring species with shallow, spreading root
systems, and in extreme sites may favor subcanopy species over those which
attain large sizes, especially on the steeper slopes. Exposure to higher winds no
doubt exacerbates drought stress for trees and, combined with shallow soils, may
result in higher treefall rates, at least of major windthrows and treefalls involving
more than one individual. However, through conditioning, trees on ridges may be
more resistant to wind than are individuals on lower ground (Basnet et al. 1992).
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
We did not record a high rate of large treefalls (Table 2-gaps), but the largest
disturbance events are typically rare in tropical forests (Sanford et al. 1986,
Hartshorn 1980). However, the only gap that we recorded larger than 500 m2 was
on such a ridgetop in Dry Upland Forest. The aforementioned factors would
explain the somewhat lower canopy height recorded on the more exposed hilltops
and upper slopes (mean = 15.5 m upper slopes, 20.6 m Dry Upland overall), than
in Standard Upland Forest (mean = 20.8 m).
The upper canopy surface in Dry Upland Forest is generally relatively even,
but in the more extreme situations the canopy is thin, with apparently fewer
canopy leaves above a given point on the forest floor than in Upland Forest on
lower, less exposed sites. This appears to result in higher ambient light levels in
the understory of such exposed Dry Upland sites. However, this type of variation
in canopy cover was only moderately well reflected by densiometer readings
(Table 2-canopy opening) and was under-represented in our sampling, as we used
no direct measure of light intensity. Under intact canopy cover, light intensity
was higher in Dry Upland hemispherical photo samples than in Standard Upland
samples (one-way ANOVA, p = 0.0001; est total site factor Dry Upland = 0.2,
Upland Standard = 0.13 ; difference significant in Tukey's pairwise comparison).
Vine density in non-gap-influenced areas was typically equal to or slightly lower
than that of the Standard Upland Forest (Table 2).
Characteristic species
The most readily apparent distinction between this forest type and Standard
Upland or Mesic Upland, is the dominance of Piper psilorrachis in the understory
along with a high density of tree saplings (Fig. 13 ), and the relative rarity of the
understory palm Cryosophila stauracantha (Appendix 2). A few species that
were widespread in all Upland Forest types reached their highest abundance in
Dry Upland Forest, most notably Brosimum alicastrum, Trichilia minutiflora,
Talisia olivaeformis, Pouteria campechiana, Malmea depressa, Manilkara zapota,
and Pouteria reticulata. Nectandra coriacea, uncommon in other upland areas,
was a major component of the subcanopy of Dry Upland Forest (Fig. 30). Several
species that were abundant in other upland areas also occurred in dry forest, but
were consistently less abundant in this topographic range, including Pouteria
amygdalina, Pseudolmedia oxyphyllaria, Wimmeria concolor, Casearia bartlettii,
and Sebastiana longicuspis. In less exposed sites, the points outlined above are
typically the extent of differences between Dry Upland and other Upland forest
types. It is important to note that Brosimum alicastrum and Talisia olivaeformis
are abundant in all dry rocky upland areas and are not limited to ruin sites. The
observation of high densities of these and other species at ruin sites has resulted in
speculation that the Maya favored them for economic purposes (Lundell 1937).
Although these species may have been cultivated by the Maya and may have had a
considerable advantage over other species in the large-scale forest succession that
followed the collapse of the Maya, our findings indicate that for Brosimum and
Talisia current distribution patterns are related to topographic/edaphic conditions,
not directly to historical events.
In the more exposed hilltops and upper slopes, a number of species that
occurred in some low-lying regions of the topographic gradient, but were rare in
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 193
mesic upland sites, were consistently present and often densely distributed,
producing a bimodal distribution pattern across the overall topographic continuum
(Fig. 30). Included in this list are: Gymnanthes lucida, Thouinia paucidentata,
Drypetes lateriflora, Amyris elemifera, Coccoloba acapulcensis, Gaussia maya,
and Talisiafloresii. This bimodal pattern is interesting as it highlights similarities
in edaphic drought conditions and light availability between Dry Upland Forest
and Transitional and Tall Scrub Swamp forest. Ceiba aesculifolia, Diospyros
campechiana, and Bernoullia flammea, although rare throughout Tikal National
Park, were consistently seen only in Dry Upland Forest.
Upland Standard Forest
Environmental factors and vegetation structure
This forest type (Fig. 17) is widespread, covering the majority of rolling
upland country and lower regions of slopes, where drainage is good but not
excessive as it is in Dry Upland Forest. Soils in these areas are on average
slightly deeper than under Dry Upland Forest but much shallower, lower in clay,
and rockier than in the deep, clay-rich soils of low-lying sites (Fig. 2, Table 2).
This is the tallest forest type, with the exception of the rare Cohune palm type, and
on average has the most open understory. Average emergent canopy height was
27 m and average canopy height 21 m. Vine density was usually relatively low,
on average only seven trees per plot (20%) having vines on them, and heavily
laden trees were infrequent (Table 2). Sapling density is moderate (Fig. 13).
Species composition
The family Sapotaceae dominates in this forest type, with Pouteria
reticulata, P. amygdalina, P. campechiana, P. durlandii, Manilkara zapota, and
Mastichodendron foetidessimum. Other common trees are Wimmeria concolor,
Brosimum alicastrum, Pseudolmedia oxyphyllaria, Trichilia minutiflora, T.
moschata, T. pallida, Blomia prisca, Simira salvadorensis, Aspidosperma
cruenta, Pimenta dioica, Acacia dolichostachya, and Protium copal. In some
areas Cedrela mexicana, Spondias mombin, and Lonchocarpus castilloi are
abundant emergents, probably reflecting a large, old disturbance event. Sabal
mauritiiformis is often abundant as a canopy palm but does not dominate or have
a major effect on canopy structure as it does in Sabal Forest.
The primary differences between this forest and Dry Upland described above
are that: (1) the ratio of abundance of common dry upland species such as
Brosimum and Trichilia minutiflora, to mesicc" upland species such as Pouteria
amygdalina and Pseudolmedia is more equal, (2) species such as Gymnanthes
lucida, Drypetes lateriflora and Nectandra coriacea (common in Upland Dry) are
virtually absent, and (3) Cryosophila stauracantha dominates the understory, with
Piper psilorrachis abundant but of secondary importance (compare Figs. 16 and
17).
Mesic Upland Forest
Environmental factors and vegetation structure
The third distinguishable upland association (Fig. 18) occurs in some
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
particularly mesic situations such as (1) ravines or shelves on the lower portions
of slopes, (2) flat areas of elevated upland, and (3) low-lying, yet still upland areas
found in regions of gradual topographic transition from upland to lowland areas.
While forests in these three situations have many shared characteristics and are
combined here under a single forest type, in some of the analyses we found it
more enlightening to distinguish the former from the latter two. In the gently
sloping, low-lying upland areas, and to a lesser degree the flat but more elevated
areas, the vegetation often took on a resemblance to Transitional Forest, although
the species composition was primarily of upland, shade-tolerant species. The soil
in these areas was lower in clay content and higher in rocks and organic matter
than that of Transitional Forest areas, but may still be subjected to periods of
saturation, although not inundation, in the wet season, as occurs in Transitional
forest. In these lower-lying Mesic Upland areas, stem diameter distributions were
also skewed towards smaller classes (10-30 cm dbh) compared to other Upland
forests (Figs. 13 and 14). In contrast, Mesic Upland Forest in the more cove-like
situations did not differ materially in structure from the Standard Upland Forest
type (Table 2). The Zapotal of Lundell (1937) encompasses both Standard and
Mesic Upland Forest.
Species composition
The following patchily distributed species were locally abundant in this
forest type, whereas they were infrequent or rare in drier upland sites: Annona
cherimoya, Cymbopetalum penduliflorum, Ouratea lucens, Exothea paniculata,
Zanthoxylum procerum, Matayba oppositifolia, Hirtella americana, and
Stemmadenia Donnel-smithii. Widely distributed species reaching their peak
abundance in these mesic upland areas included: Pouteria amygdalina, Wimmeria
concolor, Pseudolmedia oxyphyllaria, Casearia barlettii, Sebastiana longicuspis
and Cordia gerascanthus. Other species, such as Pouteria durlandii, Simira
salvadorensis, Protium copal, Ampelocera hottlei and Vatairea lundellii, while
reaching peak abundance in the more mesic Hill-base and Sabal forests, occurred
substantially more frequently in Mesic Upland than in the drier upland types.
Cryosophila stauracantha dominates the understory while Piper psilorrachis is
only sparsely distributed (Fig. 18).
Cohune Palm Upland Forest
Environmental factors and vegetation structure
This forest type, named for the palm that dominates it, Orbignya cohune, is
rare in Tikal; it was encountered only in one location during the sampling.
However, in wetter regions of Petdn and neighboring Belize this forest type is
much more common, and occurs in a number of topographic situations (Brokaw
and Mallory 1993, M. Schulze pers.obs.). The grove of Cohune forest that we
sampled occurs in a shallow depression on a shelf-like area some 30 m above
surrounding lowlands on a pronounced ridge which rose tens of meters above.
The soil in this depression had a higher clay content than that in the adjacent
upland areas and was deeper. However, the soil was more friable and of a darker
brown color than the compacted and seasonally xeric Scrub Swamp soils. The
location of this depression on a relatively high, well-drained ridge presumably
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 195
helps prevent excessive waterlogging of the soil in the wet season. Also, the
location along the side of a tall, steep slope may result in available dry season
water being concentrated here through surface and subsurface flow from the
slopes above.
To summarize, the edaphic conditions here would appear to be highly mesic
for Tikal, but without excessive drought or waterlogging. In the Rio Bravo region
of northern Belize, Brokaw and Mallory (1993) found Orbignya cohune most
abundant on mesic but well-drained soils, and in lower densities in less mesic
sites. A similar pattern can be seen in western Pet6n, which is wetter than both
Tikal and Rio Bravo (M. Schulze, pers.obs). The more limited distribution of
cohune in Tikal is likely due to the drier climate here than in the above locations.
The Cohune palm forest was slightly taller than other Upland forest types,
with modal canopy height averaging 22 m and emergent trees averaging 30 m tall.
Structurally, Cohune Upland has many features in common with Hill-base and
Sabal forest. The canopy is uneven and small light gaps, particularly under
concentrations of canopy palms, are frequent As in Hill-base and Sabal forest,
leaning but living trees occur at a higher rate here than in other upland areas (Fig.
12), probably due to unstable soil conditions; this may contribute to canopy
openings and light availability. The mean number of treefall gaps per sample was
also slightly higher in this forest type than in other upland areas. The understory
in Cohune Upland is typically open and palm-dominated (by Orbignya and Sabal
juveniles, and Cryosophila adults) although vine density can be locally high.
Species composition
In terms of species composition, this forest is closely related to Hill-base and
Sabal forests, with which it shares its most abundant species, such as Trophis
racemosa, Pouteria durlandii, Licaria peckii, Guarea glabra, Quararibea
funebris, Spondias mombin, and Ampelocera hottlei. The main attribute
separating Cohune forest from Hill-base and Sabal forests is the abundance of the
Cohune palm itself.
Hill-base Forest
Environmental factors and vegetation structure
This forest type (Fig. 19) occurs at the bases of larger hills, in a narrow zone
up to 100-150 m from areas of upland relief. The degree of variation from typical
Upland forest depends on the size of the hill and on the surrounding topographic
matrix, i.e. whether the hill base is only a relatively narrow valley between two
upland areas, or is at the transition to a large, low-lying ("Bajo") area. In the first
situation, the forest is only slightly different than Standard Upland Forest, the
most obvious difference being a higher density of vines and a lower, more
irregular, canopy in the Hill-base forest. At the bases of large hills and at hill
bases on the edges of large depressions, or bajos, the vegetation differs strikingly
from Standard Upland Forest in both structure and species composition.
Canopy height averaged 18 m and was often very uneven, while emergent
height averaged 25 m (Fig. 12, Table 2). Vine density in this forest type is high;
most trees have at least a few large woody vines in their canopies, and frequently
as many as half of the canopy trees are vine-laden to the extent that significant
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
crown shading is caused by the foliage of the vines (Table 2). While this
increased density of vines results in a denser understory, the growth of shrubs and
tree saplings usually is not much denser in Hill-base than in Upland Forest (Fig.
19, Table 2).
Several factors may influence the patterns of species composition observed
in Hill-base areas. The soils at hill bases are generally deeper than in upland
areas, and there is little rock debris in the upper layers of the soil. In many cases
the upper soil is only slightly lower in clay content than is soil associated with
Xerophytic Scrub Swamp Forest, which occurs farther down the topographic
gradient. However, as in the Cohune Palm Forest, the soil is darker and more
friable than are the compacted Scrub Swamp soils.
In soil pits dug at hill bases, there was a rocky layer at a depth of 50-60 cm.
Therefore, while the soils at hill bases may drain poorly during periods of
inundation in a similar manner to Scrub Swamp soils, this effect is probably
moderated by the more porous and texturally diverse sub-soil layer. The periods
of inundation are also undoubtedly much shorter and less severe in this
topographic region due simply to higher elevation. The soil textural qualities
mentioned above probably also ameliorate effects of dry season moisture stress.
Low-lying areas at hill bases are subject to more extreme surface and sub-surface
water flow during rainy season flood periods than are upland areas (M. Schulze
pers. obs.). This water flow would be expected to have a destabilizing effect on
trees rooted in these soils, and it is reasonable to expect a higher treefall rate in
these areas (cf. Denslow 1980). We observed slightly more treefall gaps in Hill-
base forest than in Upland forest types. In addition to completely fallen trees in
Hill-base Forest, there were a large number of live trees that are leaning at angles
of at least 25-45, often one-third of the trees in a plot, while in upland areas these
leaning trees are numerically insignificant (Table 2, Fig. 12). Presumably this
high incidence of leaning trees is a result of strong water flows during the rainy
season. These leaning trees often create openings in the canopy up to 30 m2 in
area which are functionally equivalent to small treefall gaps in their effects on the
understory light environment, and contribute greatly to the general unevenness
and low stature of the canopy in Hill-base forest. In Hill-base forest, leaning trees
are usually small, from 10-30 cm dbh, and these individuals presumably will not
reach upper canopy levels. In this way, water flow may affect canopy height and
structure as well as tree size-class distributions, with only scattered individuals
remaining upright to achieve emergent height.
In the low canopy of Hill-base forest small treefall gaps may allow greater
light penetration to the understory than do equal-sized gaps in Upland Forest.
While small gaps in Upland Forest may only be regularly exploited by shade-
tolerant species, small gaps in Hill-base forest may be occupied by light-
demanding species, although not extremely shade-intolerant ones. Thus, for at
least some light-demanding species, there are probably more usable light gaps for
regeneration in Hill-base than in Upland Forest. This helps to explain both the
high density of vines in this forest type, many of which are light-demanding, and
the abundance of trees species such as Trophis racemosa, Swietenia macrophylla,
Sebastiana longicuspis, Simira salvadorensis and Spondias mombin, which
require high light intensity in at least some stage of regeneration.
196
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 197
Characteristic Species
Even in the least pronounced form of Hill-base forest, where overall species
composition does not differ drastically from Upland Forest, species such as
Guarea glabra, Pouteria durlandii and Ampelocera hottlei, which are only
scattered in upland regions, are abundant. In the 'archetypical' Hill-base forest,
species composition is strongly skewed toward species with higher moisture
and/or light requirements than those of characteristic upland species. Included in
this list of species which dominate Hill-base (and Sabal forest), and reach their
abundance peaks in these forest types, are Pouteria durlandii, Protium copal,
Spondias mombin, Sabal mauritiiformis, Vatairea lundellii, Ampelocera hottlei,
Guarea glabra, Tabebuia rosea, Licaria peckii, Trophis racemosa, and Trichilia
moschata. Swietenia macrophylla, while prominent here, reaches its abundance
peak in lower-lying areas. Several species, such as Ocotea sp. and Trichilia
havanensis, were encountered only in Hill-base forest, although they were never
abundant.
Sabal Forest
Environmental factors and vegetation structure
A forest type (Fig. 20) that is closely associated with the Hill-base forest but
which generally occurs farther away from the bottom of major hills is the Sabal
Palm Forest, or Botanal (Lundell 1937). In some catenas, a zone of Sabal Forest
is not distinguishable, for example, in elevated, hilly regions with only small
valleys (100-300 m wide) connecting upland areas, or in areas of gradual
transition from low, flat upland to lowland topography (Fig. 11). However, in
most complete gradients from Upland to Scrub Swamp, this vegetation type is
present, typically in the zone 100-400 m from a hill base.
The canopy was lower on average (17 m) than in Upland Standard or Hill-
base forests (Fig. 12, Table 2), and considerably more broken in structure than
even in Hill-base forest. As in Hill-base forest, heavily leaning trees and treefalls
were frequent, presumably due to the unstable, frequently saturated soils. In
addition to these influences on canopy irregularity, the tall, small-crowned Sabal
mauritiiformis palms, which abound, often create small light gaps around their
periphery. In this manner, areas receiving moderate to high light intensity at the
forest floor are even more frequent in Sabal than in Hill-base Forest, rendering
this forest type yet more hospitable to light-demanding species. Vine density can
be low or high in this forest type and often varies dramatically over short
distances; treefall history seems to be the major factor in this variation.
The soil underlying Sabal forest is similar texturally to that of Hill-base sites,
although in soil pits the subsoil layer of decomposing limestone fragments was
found to be at a depth greater than 70 cm. This last factor, combined with lower
elevation, may result in poorer drainage of these soils during wet season
inundations, which in turn may inhibit root growth. Aside from this difference,
which may make Sabal forest less hospitable for some "upland species," Sabal
forest appears to occupy soils very similar to those underlying Hill-base forest.
Characteristic species
This forest is similar in species composition to Hill-base Forest, but Sabal
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
Palms (Sabal mauritiiformis) dominate the upper canopy, forming a regular but
incomplete cover, with a dicot sub-canopy and some scattered dicot emergents.
Also, species such as Guarea glabra, Vatairea lundellii, Trophis racemosa,
Spondias mombin, and Pouteria durlandii, which first become abundant in Hill-
base forest (as one moves from upper to lower regions of the topographic
gradient), provide a larger proportion of the stems, while many upland species
become infrequent. The understory in Sabal Forest is distinctive for the
abundance of Costus sp., Psychotria sp., Alibertia edulis, Calyptranthes
chytriculia, and several species of branched grasses. Dry season moisture stress
does not seem to be extreme in these regions, as evidenced by the occurrence of
tree species that are typical of wetter climatic regions of Guatemala and Belize:
Pouteria sapota, Quararibeafunebris, Ampelocera hottlei, and Guarea glabra.
Transitional Forest
Environmental factors and vegetation structure
Moving down a typical gradient from upland areas into the low-lying, gently
sloping drainage depressions, the canopy becomes consistently lower and the
understory denser (Fig. 21). Between Sabal Forest and Scrub Swamp Forest is a
zone of vegetation that is transitional, both in species composition and structure,
between the broad-leaved mesophytic Upland and Hill-base/Sabal forest types and
the Xerophytic Scrub Swamp. This may be equivalent to Lundell's (1937)
Escobal. In terms of clay content, the soil in these areas is not significantly
different from that of the Scrub Swamp, which occurs farther down the gradient
on nearly pure clay soils. However the soil appears to have consistently higher
organic content and is typically less compacted than Scrub Swamp soil. Although
periods with standing water occur, there is less severe flooding in the Transitional
Forest zone, and for a less extended period of time than farther down the gradient,
as evidenced by the dark soil color under Transitional Forest compared to the light
gray, heavily gleyed Scrub Swamp soils. Also by being closer to rocky uplands,
the soil, although well over 70 cm deep, may be slightly shallower, with closer
contact to decomposing rock below the surface, than are most Scrub Swamp soils.
The above factors apparently result in less severe edaphic extremes than in Scrub
Swamp soils, and may allow root contact with the mineral-rich decomposing
limestone subsoil layer. However, in soil qualities, as in vegetation, this
topographic region is intermediate between the mesic Hill-base/Sabal forests and
xerophytic Scrub Swamp, and dry season edaphic moisture stress is no doubt a
factor for plants occurring here.
The average height of emergent trees in the Transitional Forest was 19 nm,
average canopy height 14.5 m, and average lower canopy 11 m. The canopy in
Transitional Forest was more open on average than in Upland Forest, although
denser than in true Scrub Swamp (Fig. 12, Table 2). The understory is dense, due
to abundant vines, shrubs, and tree saplings. However, trees frequently achieve
diameters of 20-30 cm and occasionally larger, resulting in a more structurally
diverse canopy and less dense understory than in Scrub Swamp (Figs. 13and 14).
Canopy shading may still prevent regeneration of some particularly light-
demanding species typical of the Scrub Swamp in the absence of treefall gaps or
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK
other disturbance; however, we have observed a full range of size classes of a
number of relatively shade-intolerant species, such as Matayba oppositifolia and
Croton pyramidalis, beneath intact canopy in Transitional Forest, indicating
successful regeneration under such conditions. As in Hill-base forest and all other
forest types of low-lying areas in Tikal, partially fallen and leaning trees are
abundant. While the number of treefall gaps per sample was not significantly
higher for Transitional than for Upland Forest types, the abundant small canopy
gaps caused by leaning trees probably allow enough light penetration in this
already low, sparse-canopied forest for germination of even highly light-
demanding seedlings.
Characteristic species
A group of species reaches its peak abundance in Transitional Forest,
including Matayba oppositifolia, Terminalia amazonia, and Bursera simaruba.
The edaphic extremes of Transitional Forest, while not intolerable for most upland
tree species, are probably suboptimal. A number of species that are most
abundant in upland areas occur in very low densities in Transitional Forest. Such
species include Pouteria reticulata, P. amygdalina, P. campechiana, Blomia
prisca, and Pseudolmedia oxyphyllaria. A number of species that achieve their
highest densities in Hill-base or Sabal Forest remain relatively abundant in
Transitional Forest; Trophis racemosa, Vatairea lundellii, Pouteria durlandii, and
Simira salvadorensis are conspicuous members of this group. Species that we
consider Scrub Swamp specialists at Tikal, such as Croton pyramidalis,
Coccoloba acapulcensis, Coccoloba reflexiflora, and Metopium brownei, first
become abundant (when moving down-slope) in Transitional Forest, as light
conditions become more favorable for them and edaphic conditions less favorable
for their more upland-adapted competitors.
Tall and Low (Xerophytic) Scrub Swamp Forest
Environmental conditions and vegetation structure
In Tikal National Park there are three major lowland depressions or drainage
systems, in addition to numerous smaller, related depressions (Fig. 1). In the wet
season these depressions fill with water up to I m or more in depth, which may
cover the ground in the lowest areas for months before gradual drainage and the
onset of dry weather removes it. The three major depressions are drained by a
network of seasonal streams or arroyos which eventually connect with perennial
streams or large permanent swamp areas well outside the park boundaries. Soil in
the central regions of these depressions is virtually pure, heavily gleyed and
compact clay, which binds strongly to water molecules (Eyre 1964), suggesting
that plants in Scrub Swamp experience extreme edaphic drought as these soils
become desiccated and deeply fissured in the dry season.
Typical Scrub Swamp Forest is a thicket of stunted, gnarled trees of a
uniform, low height, and corresponds to the forest types locally known as "bajo,"
including the particular type of bajo forest known as "Tintal" due to the
predominance of tinto (Haematoxylum campechianum). As one moves from the
periphery to the interior of the large basins where this forest type occurs, the
199
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
vegetation increases in density and decreases in height, from taller, more open
Scrub Swamp to a virtually impenetrable thicket of sapling-sized trees, vines, and
sawsedge (Scleria bracteata). Despite the continuous nature of this variation, it is
convenient to divide this forest into two categories, "tall" (Fig. 22) and "low" (Fig.
23), corresponding to the topographically higher (peripheral) and lower (central)
regions of the depressions in which Scrub Swamp Forest occurs, and named so as
to reflect the respective canopy heights of these two forest "types." Tall Scrub
Swamp Forest typically occurs between 600 m and I km from the nearest upland
relief, and Low Scrub Swamp Forest at distances greater than 800 m from such
relief. There are a number of exceptions to this rule, as humps of higher ground or
even rocky upland islands occur seemingly out of place within Scrub Swamp
depressions. These formations may support Tall Scrub Swamp or even an unusual
type of Transitional Forest, though the surrounding forest is Low Scrub Swamp.
Between one drainage system and another, there is a high degree of variability in
the relative locations of these forest types.
Sixteen plots totaling 0.61 ha were located in Low Scrub Swamp and 17
totaling 0.65 ha in Tall Scrub Swamp Forest. Duration of inundation increases
from the peripheries to central portions of low-lying basins, and soil characters
appear to vary in like fashion, with soils in central regions appearing higher in
clay and lower in organic matter than those of peripheral areas. These patterns
suggest that edaphic extremes are more pronounced for trees in Low than for
those in Tall Scrub Swamp Forest
Structural and floristic differences between Tall and Low Scrub Swamp
Forest appear to be related to the differences in soil and drainage regime described
above. The average emergent height in Tall Scrub Swamp Forest plots was 18 m,
average modal canopy height 12.7 m, and average lowest canopy height 9 m,
compared to 12 m, 10.1 m and 8.1 m respectively for Low Scrub Swamp. Vine
density was slightly higher in Tall than in Low Scrub Swamp (Fig. 12, Table 2).
Average canopy opening was significantly higher in Low Scrub Swamp, which is
evidenced by the higher density of sedges in the understory (Fig. 12, Table 2).
However, light availability does not seem to be a limiting factor for regeneration
of trees in either of the Scrub Swamp forest types, where canopy opening
averaged 2-3 times greater than in Upland forest types (Fig. 12). Stem density
was higher on average in Low than in Tall Scrub Swamp forest, but basal area was
higher in Tall Scrub Swamp (Figs. 13 and 14).
In Tall Scrub Swamp, light-demanding trees such as Bucida and Swietenia
are often present as 16-20 m tall emergents over a uniform 9-12 m canopy.
While these species also occur in the more severe Low Scrub Swamp, emergents
of these and other species are much less frequent and lower in stature in the latter
forest type.
Characteristic species
Tree species common to both types of Scrub Swamp include Manilkara
zapota, Croton pyramidalis, Croton reflexifolia, Ouratea nitida, Coccoloba
reflexiflora, Coccoloba acapulcensis, Coccoloba cozumelensis, Swietenia
macrophylla, Gliricidia sepium, Metopium brownei, Haematoxylum
campechianum, Bucida buceras, and Amyris elemifera. The following species
200
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 201
occur commonly in Tall Scrub Swamp but are infrequent or absent from Low
Scrub Swamp: Vitex guameri, Bursera simaruba, Gymnanthes lucida, Matayba
oppositifolia, Lonchocarpus castilloi, Guettarda combsii, and Terminalia
amazonia. Conversely, several species are infrequent in Tall Scrub Swamp, but
are important components of Low Scrub Swamp, for example Camareria latifolia,
Erythroxylon guatemalensis, Rapanea guianensis, and Plumeria obtusa. Several
tree species that attain relatively large sizes (>20 cm dbh) in other topographic
regions appear to occur only as stunted treelets in the Low Scrub Swamp (e.g.
Calophyllum brasiliense, Chrysophyllum mexicanum, Amyris elemifera, Ouratea
lucens). One of these species, Ouratea lucens, commonly fruits at a height of
three meters in the Scrub Swamp.
Mesic Bajo (including Arroyo Forest) and True Swamp
Both of these forest "types" correspond to position 18 of our topographic
position code. However, floristic and structural comparison of samples in this
topographic range support recognition of two forest types.
Mesic Bajo (including Arroyo Forest)
Environmental conditions and vegetation structure
The larger lowland depression systems in Tikal generally are drained by one
or more seasonal streams (arroyos). Along the margins of these waterways a
distinct type of vegetation, Arroyo Forest (Fig. 24), occurs, apparently resulting
from longer periods of inundation and soil saturation and shorter and less extreme
periods of dry season edaphic drought than those experienced by the xerophytic
Scrub Swamps occurring slightly higher on the topographic continuum. In the
central, lowest-lying portion of one of these drainage systems we encountered an
area of low swamp, below the Low Scrub Swamp forest, in an area where
standing water was present even at the height of the dry season. We have termed
this forest type "Mesic Bajo" (Fig. 25). It is likely that a similar type of
permanent low swamp occurs in the other two major drainage basins in Tikal,
although we did not intersect any such area in our sampling.
The vegetation occurring in these two zones (arroyos and permanent
swamps) within the xerophytic Scrub Swamp-dominated depressions is similar in
species composition and structure, and for the purposes of this study is considered
under one forest type-Mesic Bajo Forest.
Mesic Bajo and Xerophytic Scrub Swamp differ more in structure than they
do in species composition. While the average canopy heights for Mesic Bajo are
virtually identical to those of the Tall Scrub Swamp, the average canopy opening
is significantly less for Mesic Bajo than for Tall Scrub Swamp and drastically less
than for Low Scrub Swamp (Table 2). In addition, the canopy in Mesic Bajo is
much more irregular than in Scrub Swamp (Fig. 25); in areas without natural
disturbance the canopy is high and quite dense, but adjacent to these patches of
tall forest are areas consisting of a 14 m tall tangle of vines, bamboo, and saplings.
Because canopy density is often high in Mesic Bajo, treefall gaps play a more
important role in this forest type than in Xerophytic Scrub Swamp, and light,
although available in high enough intensity for light-demanding tree species to
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
dominate, is patchily distributed, in contrast to the universally high availability of
light in the Scrub Swamp. Vine density is also higher in Mesic Bajo than in
Xerophytic Scrub Swamp, contributing to canopy closure and presumably to tree
mortality through crown-shading and increased susceptibility to toppling in the
unstable soil conditions. Basal area was higher in Mesic Bajo than in neighboring
Scrub Swamp forest, as trees here achieve greater size (Fig. 14). In contrast,
sapling and treelet stem density is lower in Mesic Bajo, presumably due to
shading by the dense canopy, combined with prolonged inundation.
Characteristic species:
As the Mesic Bajo is included within the Scrub Swamp spatial matrix, many
of the tree species common in Scrub Swamp also occur here. The two most
important are Bucida buceras, which often dominates tall, relatively open groves,
which are interspersed with impenetrable bamboo tangles, and Haematoxylum
campechianum, which forms almost pure stands (known as "tintales") in some
areas with excessive inundation. Other Scrub Swamp specialists such as Croton
pyramidalis occur in Mesic Bajo but are less common here than in Scrub Swamp.
A number of species that are common in riparian forests in other areas of Petdn
and northern Belize, in Tikal were found only in Mesic Bajo and True Swamp,
discussed below. Of these, Pithecellobium belizensis, Lonchocarpus
guatemalensis, and Casearia corymbosa were particularly abundant.
Margaritaria nobilis and Amyris elemifera, although occurring in Scrub Swamp
forest in stunted treelet form, achieved their greatest densities and sizes in Mesic
Bajo.
True Swamp
Environmental conditions and vegetation structure
Two forests classified as True Swamp were encountered at Tikal. Both were
in relatively narrow valleys, 300-600 m wide, in low-lying areas with standing
water in the dry season, but surrounded by upland topography, and at higher
elevations than the large, Scrub Swamp-dominated depressions described above.
Hill-base and Sabal forest were present in the transitional areas between the True
Swamp and surrounding upland areas, but no Transitional or Scrub Swamp forest
occurred adjacent to the swamp. While areas supporting True Swamp were
edaphically similar to those supporting Mesic Bajo, and these forests shared
important structural characters and species, they were sufficiently distinctive to
warrant differentiation.
The two True Swamps were encountered during streamlined 1993 sampling,
in which understory and most structural variables were not recorded. This
precluded most quantitative structural comparisons of this with other forest types.
Qualitatively, however, the True Swamp and Mesic Bajo seemed similar, both
having high vine density, low stem density, and areas of tall, dense-canopied,
open-understoried forest interspersed with low, dense vine tangles hardly meriting
the term forest. As these two forest types are ecologically and structurally similar,
they were combined for many of the analyses. The True Swamp is taller on
average than Mesic Bajo, probably resulting in lower light intensity in a gap of
any given size in this forest than in Mesic bajo. Hence there may be fewer
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 203
suitable gaps in True Swamp for regeneration of bajo specialists than in Mesic
Bajo.
Characteristic species
Many of the floristic differences between the Mesic Bajo and True Swamp
appeared to be related to the forest matrix in which each occurs: Scrub Swamp
forest versus upland/Sabal forest, respectively. The distance of the True Swamp
from Scrub Swamp seed sources reduces the probability of colonization of True
Swamp by many species that frequently occur in Mesic Bajo but are largely
limited to Mesic Bajo and Scrub Swamp. Conversely, tree species that are
abundant in Sabal or Hill-base forest and tolerant of prolonged inundation should
have an advantage in dispersal to gaps in the neighboring True Swamp due to their
proximity. Trophis racemosa, Spondias mombin, Tabebuia rosea, Simira
salvadorensis, and Sapium are the more distinctive examples of Sabal Forest
'colonizers' of the True Swamp. These species also seem to be more successful in
colonizing small gaps than are many more light demanding species typical of
Scrub Swamp.
Common species shared by True Swamp and Mesic Bajo are Lonchocarpus
guatemalensis, Pithecellobium belizensis, and Casearia corymbosa. In addition to
these species, the following wet-tolerant species were found at Tikal only in True
Swamp: Inga edulis, Pachira aquatica, and Cassia grandis. In some regions of
these swamps the terrestrial bromeliad Aechmea sp. forms a dense ground cover
and gives a distinctive appearance to the swamp.
Distribution Patterns of Individual Species
From the multivariate analyses presented above it is clear that a significant
portion of tree species distribution patterns in Tikal can be explained by
topographic position and associated microenvironmental conditions. The
topographic/edaphic gradient in Tikal is not, however, simply a monotonic
progression from dry upland sites to wet low-lying sites. Rather, conditions for
plant life are affected not only by mean moisture availability, but also by seasonal
extremes, including drought and standing water. Soil depth, texture, and resultant
moisture holding and drainage characteristics, as well as pH and organic matter
content, interact with both general position on the topographic sequence and with
local site topography to produce a complex and non-linear environmental
continuum.
To understand species-environment responses, it is therefore important to
consider the distribution of each species along this entire environmental gradient,
rather than simply peaks of abundance curves which are the principal currency of
ordinations. This is particularly true for the numerous species displaying bimodal
abundance curves. In addition, the abundance curves of canopy trees and saplings
of a given species suggest the relative success in seedling survival to canopy
stature at different points along the predominant environmental gradient. While it
is unlikely that any two tree species have precisely the same response to this
environmental gradient, species may be grouped into several general response
patterns. Here we present graphs of adult and sapling abundance patterns for
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
species representative of the major response types we were able to recognize.
Memberships of 155 species in these ten response types are indicated in Appendix
1 under "Topographic Distribution Group."
Distribution patterns revealed three groups of species with abundance peaks
in upland forest areas: (1) a group of species most common in the drier upper
portions of the gradient (termed Xeric Upland species, Fig. 26); (2) species
reaching maximum frequencies and densities in moister, lower-lying upland areas
(Mesic Upland species, Fig. 27); and (3) a large group of species with relatively
uniform abundance throughout the upland regions of the topographic gradient
(Upland Generalists, Fig. 28). Species within these three groups were not
restricted to upland regions of the gradient, but decreased in abundance with
increased distance from upland areas and were virtually absent from the Scrub
Swamp forests in the interiors of the lowland depressions.
A group of species showed an affinity for mesic topographic positions
including wetter upland areas, true swamps, and particularly the moist areas at the
bases of hills and up to 300 meters into lowland depressions (Mesic Forest
species, Fig. 29). This group is interesting in that it included both shade-tolerant
and light-demanding species that displayed virtually identical abundance patterns.
Bimodal Xeric species (Fig. 30) showed virtually the opposite of Mesic forest
species distributions, with peak abundances in the areas experiencing the most
extreme xeric conditions-rocky upper slopes and the central regions of lowland
depressions.
A large number of species appeared specialized for conditions existing in the
lowland portions of the gradient. One group consisted of highly light-demanding
species that were ubiquitous in Scrub Swamp forests but absent or significantly
less frequent in more mesic lowland forests (Scrub Swamp Specialists, Fig. 31).
Some of these species were also common in second growth forests. Transitional
Forest Species (Fig. 32) displayed similar distribution to Scrub Swamp
Specialists, but appeared susceptible to the most extreme drought conditions
found in the Low Scrub Swamp Forest, and reached peak densities in Transitional
or High Scrub Swamp Forest. Another group, (Swamp Obligates, Fig. 33)
showed the reverse distribution pattern, occupying primarily those areas with
seasonally waterlogged soils, but without the extreme drought of the Scrub
Swamp areas. Finally, some species were uncommon in upland forest but
occurred with uniformly high frequencies in lowland areas (Lowland Generalists,
Fig. 34). A few species (Topographic Generalists, Fig. 35) were common
throughout the topographic gradient.
Species Responses to Canopy Disturbance and Light Intensity
Ambient light levels in the forests of Tikal vary between regions of the
topographic continuum, due to differences in canopy height, canopy cover, and
evenness of the canopy's upper surface, and in the frequency of treefalls (Table 2,
Fig. 12). Canopy height is an important determinant of the amount of light
reaching the forest floor in canopy gaps, as a gap surrounded by tall-canopied
forest will receive direct light for a lower percentage of the day than a similarly
204
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 205
sized gap bordered by lower forest. Hence, the light environments within treefall
gaps (and under other canopy openings) vary significantly with forest type along
the dominant topographic gradient (Table 2 ), due simply to the pattern of
decreasing average canopy height from upland to lowland portions of the gradient.
This fact is well illustrated by comparison of light environments in human-made
logging gaps in forest adjacent to Tikal. Light intensity was significantly higher
in lowland forest logging gaps than in upland forest (ANOVA of total site factor,
calculated from hemispherical photographs in 86 logging gaps, by forest type, p =
0.007), even though gap size did not differ significantly between forest types
(Schulze et al. unpubl. data).
The density of intact forest canopy affects the light environment of seedlings
and saplings occurring in undisturbed areas. In general, canopy cover decreased
from upland to lowland portions of the topographic gradient, with two primary
exceptions. In Dry Upland Forest on hilltops and upper slopes, the understory
beneath intact canopy received significantly higher average light intensity (as
calculated from hemispherical photographs; Fig. 36) than did Standard Upland
understories (see Dry Upland Forest description above). Mesic Bajo typically is a
mosaic of canopy gaps and patches of dense, vine-laden canopy. Beneath these
patches of intact canopy, light intensities were typically only slightly higher than
under Upland forests (est. total site factor from hemispherical photographs = 0.26,
s.d. = 0.12 Mesic Bajo; range in undisturbed upland areas = 0.08-0.29), while in
the intervening open areas light penetration was such that sedges were common as
ground cover (Table 2). In the extreme case-Low Scrub Swamp-canopy cover
was so sparse that light intensity was as high under intact canopy in this forest
type as in logging gaps in Upland Forest (est. total site factor = 0.50, s.d. = 0.1
logging gaps; 0.45, s.d. = 0.08 Scrub Swamp), and treefalls in Low Scrub Swamp
did not typically leave a discernible canopy gap. In some regions of the gradient,
particularly Hill-base and Sabal forests, canopy cover can be locally quite dense,
but overall canopy structure is broken, with numerous small openings created by
processes other than treefall gaps (e.g. leaning trees, canopy palms, discontinuities
in canopy cover). Canopy unevenness therefore results in higher understory light
intensities in these areas, even in the absence of treefall gaps. Earlier we
presented evidence that treefalls are more frequent in lowland forests than in
upland (Table 2), resulting in more situations of high light intensity.
On top of the background variation in understory light environments among
forest types, light levels within a given forest type or topographic range vary due
to natural or anthropogenic disturbances and small-scale variation in canopy
cover. Our multivariate analyses suggested that variation in light availability
exerts a major influence on distribution of many tree species at Tikal (Figs. 9 and
10). Although the technique used to produce these ordinations (CCA) did not
allow for significance testing of species associations with disturbance factors,
plots of species centroids in ordination space indicated that for 16 species juvenile
abundances were positively correlated with canopy opening, and 22 and 30
species showed associations between juvenile and adult occurrence, respectively,
and natural canopy disturbance. To explicitly test the significance of these
apparent associations with light and disturbance, we conducted a series of
additional analyses, described under DATA ANALYSIS and reported below. We
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
obtained large enough sample sizes (n > 20) for 63 tree species to include them in
one or more of the nine analyses of associations of adults and juveniles with
natural disturbance and understory light intensity. Results for all nine analyses
are summarized in Table 4.
The most powerful technique for analysis of tree juvenile and treelet
associations with light intensity and disturbance proved to be two-way ANOVA
of species abundance with topographic position (8 values), natural disturbance,
and degree of canopy opening (<10%, 10-20%, >20%) as categorical variables.
Of the 56 species tested, 26 did not display significant associations with either
natural disturbance or canopy opening: 12 of the 26 species (4 after Bonferroni)
were significantly associated with forest type/topographic position, while
ANOVA models were non-significant for the remaining 14 of 26 species. Thirty
species had one or more effects of disturbance or canopy opening that were
significant before Bonferroni correction (Table 4), and 15 species retained
significant effects after Bonferroni adjustment. Nine species had significantly
higher abundances in samples with high understory light intensity, with two
additional species showing marginal associations with high light (p < 0.05, but n.s.
under Bonferroni). Three species (Haematoxylum campechianum, Ouratea
lucens, Myrtaceae spl), showed significant positive associations with the
interaction of forest type and canopy opening-i.e. they were most abundant in
lowland sites with open canopies. Four species (Nectandra coriacea, Gymnanthes
lucida, Drypetes lateriflora, Laetia thamnia) had highest abundances in sites with
medium light intensity, although none of these models were significant under
Bonferroni. Cryosophila stauracantha and Piper psilorrachis were significantly,
and three other species marginally, associated with sites with dense canopy cover.
The majority of species tested did not show any association with disturbance.
Three species (Lonchocarpus latifolius, Thevetia ahouai, Pithecellobium
belizensis) were significantly more abundant in disturbed than undisturbed plots,
with seven species marginally associated with disturbance. One species was
significantly, and one marginally, more abundant in undisturbed plots.
For Upland forest samples, we performed Chi-square analyses on data for 35
species to test associations of tree juveniles and treelets with disturbance and light
intensity (Table 4). Of these, two species displayed significant (under Bonferroni)
positive associations with disturbance (Chamaedorea pacaya, and Piper
sempervirens, both treelets), while distribution of one species, Gymnanthes lucida,
was negatively associated with disturbance. Fourteen of the remaining species
displayed what we consider a convincing lack of correlation with natural
disturbance (p > 0.5), suggesting the ability of seedlings to persist and recruit to
sapling size classes in the absence of treefall gaps. Two species (Piper
psilorrachis, Pseudolmedia oxyphyllaria) showed significant positive associations
with low understory light intensity (<10% canopy opening). In upland forest, no
tree species showed significant associations with high light intensity (>10%
canopy opening), but one such species association was significant prior to
correction (Eugenia sp.), and another two displayed non-significant tendencies to
occur under high light conditions (Gymnanthes lucida, Pouteria durlandii).
In transitional forest areas (Hill-base, Sabal, Transitional forests) we tested
light and disturbance associations of 19 species using MRPP Chi-square tests
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK
(Table 4). For one treelet species (Piper cf. aduncum) and one non-woody
understory dominant (Costus sp.) associations with past canopy disturbance were
significant, but no tree juveniles were positively associated with disturbance. One
understory treelet, Calyptranthes chytriculia, was significantly associated with
transitional forest areas that had not been subject to recent canopy disturbance.
Eight species showed a convincing lack of relation between natural disturbance
and juvenile abundances (p > 0.5). Croton pyramidalis, Trichilia moschata, and
Ouratea lucens were significantly associated with sites under relatively open
canopies (>20% opening), although they did not show an association with treefall
gaps. Gymnanthes lucida and Costus sp. also displayed positive associations with
high light intensity, but these were only significant prior to Bonferroni correction.
Piper psilorrachis was significantly associated with low light levels, and 6 species
showed a convincing lack of correlation with understory light levels (p > 0.5).
We were able to test sixteen species associations in lowland sites (Mesic
Bajo, True Swamp, Tall and Low Scrub Swamp). Four species (Xylosma sp.,
Haematoxylum campechianum, Croton pyramidalis, Coccoloba reflexiflora)
displayed significant associations with high understory light intensity (canopy
opening >20%), while two species were significantly associated with low light
conditions (Trichilia moschata, Cryosophila stauracantha; Table 4). Four species
displayed a convincing lack of association with canopy opening in lowland forest.
Interpretation of Chi-square tests of species associations with disturbance in
lowland sites was complicated by the lack of influence of treefalls on canopy
structure in many Scrub Swamp areas-canopy gaps were rarely recorded in plots
from these forest types, even though treefall rates appeared to be equal to or
higher than rates in other forests. Observed associations with disturbance for
many species merely reflect associations with either Mesic Bajo and True Swamp
(positive associations), or with Scrub Swamp (negative associations). An
illustration of this confounding factor is the apparent association of the understory
palm Cryosophila stauracantha with natural disturbance within these lowland
sites. Any observer of this species at Tikal would conclude that it is shade-
tolerant, and indeed our other Chi-square tests demonstrated that it is not
positively associated with disturbance or high light intensity, and in fact displays
an affinity for undisturbed, low light sites (Table 4). We do not, therefore place
much confidence in results of disturbance correlations at lowland sites.
Adults of eight species were significantly more abundant in Upland forest
plots with old treefall gaps than in undisturbed plots, with two more species
displaying marginal associations with disturbance (Table 4). For one species,
Brosimum alicastrum, distribution was significantly correlated with undisturbed
forest, and two more species (Pouteria reticulata, Cryosophila stauracantha)
displayed marginal associations with undisturbed forest. Fifteen species displayed
a compelling lack of association with natural canopy disturbance (i.e. p > 0.5). In
lowland forest (Hill-base through Mesic Bajo) one species, Sebastiana
longicuspis, was significantly associated with undisturbed areas, with Protium
copal and Pouteria reticulata marginally associated with undisturbed and
disturbed areas, respectively. This last species was the most commonly
encountered subcanopy or canopy tree species in the understory and is clearly
shade-tolerant. Therefore we can only interpret this result as a statistical artifact.
207
Table 4. Results of chi-square and ANOVA testing of juvenile and adult tree associations with natural disturbance and canopy opening.
Juveniles sten 7.5 cm dbh Juveniles
Chi-square Chi-square ANOVA
Upland sites Transitional sites Lowland sites Upland Transitional Full
n=114 n=46 n=35 n=114 n=46 gradient
Canopy Canopy Canopy Canopy opening
Species opening Disturbance opening Disturbance openmg Disturbance Disturbance Disturbance /disturbance
Acacia cookii 0.482 0.847 0.659 0.556
Amyris elemifera
Ardisia densiflora
Ardisia paschalis
Aspidosperma cruenta
Aspidosperma megalocarpon
Bactris sp.
Bamboo *
Blomia prisca
Brosimum alicastrum
Bursera simaruba
Calyptranthes chytraculia
Casearia bartlettii
Cecropia peltata
Chamaedorea pacaya
Chamaedorea sp2
Chamaedorea spl
Coccoloba reflexifolia
Costus sp. *
Croton pyramidalis
Cryosophila stauracantha
0.001 H*
0.0003 H*
0.168
0.393
0.124 L
0.590
0.393
0.235
0.115 H 0.499
0.702
0.558
0.0003 D*
0.009 L* 0.283
0.531 0.894
0.542 0.761
0.835
0.001 U
0.0006 D
0.276
0.002 D
0.656 0.00005 U
0.209 0.00002 D
0.737 0.386
0.921 0.336
0.802 0.896
0.0001 H 0.006 U*
0.008 H*
0.00004 H
0.200 0.128 U 0.761
0.2xl0- D
0.788
0.187
0.12x10-" H 0.057 U
0.004 L 0.7x10lsD 0.023 U*
0.467
0.023 D*
0.0002 H
0.715 0.16x10 H
0.403 0.19x10-' U&L
Cupania belizensis
Dendropanax arboreus
Drypetes lateriflora
Eugenia sp.
0.601 0.177 0.747 0.162
0.415
0.029
H*
0.503
0.629
Guettarda combsii
Gymnanthes lucida 0.111 H 0.0002 U
Haematoxylum campechianum
Laetia thamnia
Lonchocarpus latifolius
Malmea depressa
Malpighia glabra
Manilkara zapota
Margaritaria nobilis
Myrtaceae sp (bajo #1)
Nectandra coriacea
Nectandra salicifolia
OPPLVI (Myrtaceae sp.)
Ouratea lucens
Pimenta dioica
Piper cf. aduncum
Piper psilorrachis
Piper sempervirens
Pithecellobium belizensis
Pouteria amygdalina
Pouteria campechiana
Pouteria durlandii
0.169
0.534
0.727
0.01 H*
0.394 0.738
0.001 H 0 0046 1 *
0.374
0.0006 D
0.636
0.0001 D
0.00004 D
0.745
0.777
0.571
0.338
0.558
0.528 0.799 0.545
0.939 0.883
0.07 L
0.869
0.002 L
0.056 L
0.555
0.703
0112 H
0.188
0.086 D
0.842
0.0002 D
0.840
0.330
0 181
0.0003 H
0.084 H
0.626
0.001 L
0.332
0.235
0.001 D
0.564
0.069 H 0.4x10-' U
0.052 H 0.261
0.875 0.045 D*
0.013 D*
0.374
0.447
0.743 0.004 D
0.378
0.245
0.813 0.515 0.069 L 0.041 D* 0.520
0.827
0.005 M*
0.05 M*
0.25 xl 0-H(F)
0.09 M
0.0002 H &D
0.026 U*
0.2 x10' H
0.16 xlO' H
0.00014 H
0.029 M*
0.15 xlOs H(F)
0.001 H(F)
0.035 D*
0.99 x10-" L
0.014 D & L*
0.00078 D
0.17 D
0.520
.otei .ul~i 1 8
Juveniles stemn 7.5 cm dbh Juveniles
Chi-square Chi-square ANOVA
Upland sites Transitional sites Lowland sites Upland Transitional Full gradient
n=114 n=46 n=35 n=114 n=46 n=201
Canopy Canopy Canopy Canopy opening
Species opening Distug Distuopen Disturbance opening Disturbance Disturbance Disturbance / disturbance
Pouteria reticulata 0.341 0.555 .084 H 0.648 0.080 L 0.046 D* 0.020 U* 0.005 D*
Protium copal 0.567 0.152 U 0.263 0.849 0.694 0.041 U*
Pseudolmedia oxyphyllaria 0.004 L 0.184 0.647 0.074 U 0.063 L
Phyllanthus nobilis 0.33x10- H
Sabal mauritiiformis 0.304 0.445 0.438 0.546 0.881 0.896
Sebastiana longicuspis 0.227 0.517 0.887 0.2xlO- U 0.00008 D 0.0005 U 0.07 L
Simira salvadorensis 0.233 0.118 D 0.727 0.712 0.012 D*
Spondias mombin 0.003 D
Stemmadenia donnell-smithii 0.004 D
Talisia olivaeformis 0.753 0.262 0.581
Thevetia ahouai 0.31xl0s H&D
Trichilia minutiflora 0.482 0.073 U 0.347 0.145 0.363 0.582
Trichilia moschata 0.657 0.694 0.004 H 0.069 U 0.0002 L 0.213 0.061 D 0.587
Trichiliapallida 0.401 0.629
Wimmeria concolor 0.708 0.128 D 0.539 0.028 D*
Xylosmasp. 0.7x10' H 0.0001 U 0.0003 H
All p-values are reported for Chi-square tests. For ANOVA, p-values are reported for only those models that were significant or nearly significant. Values in bold
indicate results that were significant following Bonferroni correction. Values followed by an indicate results that were significant at p=0.05 prior to Bonferroni
correction but not afterward. Letter codes indicate direction of association: D = + correlation with canopy disturbance; U = correlation with canopy
disturbance; L = associated with low light intensity (0-10% canopy opening); M = associated with moderate canopy opening (10-20%); H = associated with
high light intensity ( >20%); (F) = positively associated with the interaction of forest type and light intensity (i.e. most abundant in high-light, lowland forest).
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 211
The results of the above statistical analyses confirm the trends evidenced in
the CCA ordinations of adults and juveniles with disturbance and canopy opening
included as variables. Overall, 13 species displayed significant associations with
canopy opening >20% (plus four marginal associations), and distributions of 13
species were significantly correlated with canopy disturbance (either juveniles or
adults, with eight additional marginal associations). These numbers are only
slightly lower than would be predicted from the CCA ordinations, and the
discrepancy is due in part to the fact that several species showing associations in
CCA were not included in Chi-square or ANOVA tests, due to small sample sizes.
Furthermore, these results demonstrate that species responses to light
intensity and canopy disturbance vary within the tree community (Table 4). For
some species, juveniles did not show a positive association with canopy opening
or gap disturbance, but adult distribution was highly correlated with old
disturbances, suggesting that these species were able to maintain seedling and
sapling stocks under intact canopy, but required gaps for recruitment to pole and
adult size classes.
A number of light-demanding species were significantly associated as adults
with old treefall gaps in Upland forest within our sample, while juveniles were
virtually absent from understory samples. This pattern suggests that such species
require relatively large disturbances in order to colonize upland forest, but recent
gaps (<5 years of age), where juveniles would be found, were much less abundant
than older gaps (5-20' years) in our sample. It should be noted that for a number
of long-lived, light-demanding species, such as Pseudobombax ellipticum,
Swietenia macrophylla, and Cedrela mexicana, our limited ability to objectively
record the presence of disturbance events older than ca. 20 years, restricted our
ability to examine associations of adults with canopy disturbance. Size class
distributions for these species were skewed towards diameters >45 cm; even
assuming a very high growth rate of 2 cm per year, these individuals would be
older than twenty years, and more likely they were closer to fifty years of age.
Relatively small differences in canopy opening appeared to be significant for
saplings of some species. For example Gymnanthes lucida seedlings and saplings
were abundant under light, but intact canopy in dry upland forest, but were not
common under the denser canopy of more mesic upland forest.
As noted above, in Scrub Swamp forests, treefall gaps do not significantly
modify the understory light environment, and species reaching peak abundance in
these forests showed no association with canopy disturbance, although both
saplings and adults were positively associated with light intensity. In contrast, in
upland areas and mesic lowland forests canopy disturbance does appear to be an
important source of regeneration opportunities, with a number of tree species
positively associated with treefall disturbances in juvenile (recent gaps) and/or
adult classes (building phase forest). Hence, as in other tropical forest areas
(Denslow 1987, Swaine et al. 1987), gap dynamics are important in determining
local species composition, but only in some portions of the topographic gradient.
Knowing the manner in which light and edaphic conditions vary along the
topographic gradient in Tikal, it is possible to infer the relative importance of
edaphic factors and light availability in determining patterns of individual species
occurrence, through comparison of juvenile and adult abundance peaks along the
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
gradient. If a species is well represented by both juveniles and adults in Upland
forest, and does not display a positive association with canopy disturbance, the
logical assumption would be that light is not generally a limiting factor and that
the species distribution pattern across the gradient is more likely influenced by
edaphic conditions. A large number of species in our sample fit this model.
Conversely, for species with juveniles and adults largely restricted to low-lying
portions of the gradient (particularly Scrub Swamp), but with adults occasionally
occupying upland areas, we infer that light availability is a more critical limiting
factor in colonization and recruitment than are edaphic factors. Similarly, species
entirely restricted to Scrub Swamp areas are probably light limited, as they are
clearly able to tolerate extreme edaphic drought as well as inundation, and hence
might be expected to occupy a greater range based solely on edaphic
considerations.
The majority of tree species appear to fall in between the two extremes of
species limited primarily by edaphic conditions and those limited chiefly by light
availability. A number of species had juvenile abundance peaks in lowland forest
even though adults were most common in upland forest. The majority of these
species displayed positive associations with either canopy opening or disturbance
in the analyses above. We hypothesize that there is a trade-off for these species
between the optimal light regime and tolerable edaphic conditions. High intensity
light environments are most common in Scrub Swamp regions of the continuum,
yet these areas are also subject to the most severe edaphic extremes. Hence, while
germination and early survival rates may be high for these species due to light
availability, ultimately survival and recruitment to adult size appears to be quite
low in Scrub Swamp. Few seedlings and saplings may establish in upland
regions, but those that do so appear to have a much higher probability of survival
here than in Scrub Swamp portions of the topographic gradient. For species with
juvenile and adult abundance peaks in Hill-base through Transitional forests we
can infer little, from distribution patterns alone, about the relative influence of
light and edaphic conditions on survival and recruitment rates, as conditions in
these areas are acceptable to both drought-sensitive and light-demanding species.
As mentioned above, numerous species were virtually absent from
understory samples, with adults positively associated with past disturbance. In
establishment and/or recruitment of seedlings these species likely are limited by
light. In addition many of these species are also absent from the high-light
environment of Scrub Swamp regions of the gradient, indicating that edaphic
tolerances also play a role in determining their distribution patterns.
Using the results of the Chi-square and ANOVA analyses, as well as
comparison of adult and juvenile distributions, we were able to recognize seven
light/disturbance response patterns. The response types are presented in Fig. 37,
ordered with respect to three variables: the association of adults with past canopy
disturbance; the association of saplings with high understory light intensity; and
the association of juveniles with treefall gaps. In Appendix 1, each of the species
studied is assigned to one of these seven response types or guilds. For species
with low sample sizes (indicated by an asterisk in Appendix 1) we based our
classification on qualitative observations of light and disturbance associations
during five years of field work in the region. Descriptions of light/disturbance
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 213
response types (guilds) follow in order of increasing shade tolerance:
1. Highly shade intolerant, non-gap species.-This species group (Table
4, Appendix 1) is composed of species positively correlated with degrees of
canopy opening of 20% or greater, but not correlated with intermediate (i.e. 10-
20%) canopy opening or with disturbance history (treefall gaps). These species
are poorly represented in Upland Forest and are most abundant in (often limited
to) Scrub Swamp, Mesic Bajo and Transitional Forest, where treefall gaps are
either insignificant modifiers of understory light environments, or are rendered
suboptimal by edaphic factors for colonization by fast-growing pioneer species.
Most of these species appear to be relatively slow growing, and probably do not
compete well with pioneer species in the colonization of gaps in upland habitats.
2. Large-gap colonizers.-In this group (Table 4, Appendix 1), adults were
positively correlated with evidence of older disturbance (five years or more of
age), but saplings were too rare in samples to allow statistical tests. Most recent
gaps recorded by our sampling were small or medium-sized, whereas the older
disturbances that were detectable were primarily large ones. The observed
association between adult trees of guild 2 members and large, old disturbance
events, and the rarity of saplings in the prevalent small and medium-sized gaps,
suggest that these species require large treefall gaps for successful regeneration.
Indeed, several of the species in this group have been classed as pioneers in
previous studies (e.g. Brokaw 1987; Popma and Bongers 1988; Whitmore 1989).
The consistent absence of saplings, but occasional presence of seedlings, of this
group in recent treefall gaps and the presence of larger saplings in older gaps,
indicate that these species are regenerating in gaps from seed rather than from pre-
established saplings.
3. Gap generalists.-This guild includes species with saplings (or adults in
the case of treelets) positively associated with disturbance. These species show
similar regeneration requirements to those of 'large-gap colonizers', but appear
more able to colonize small and medium-sized treefall gaps, explaining the higher
frequency of small individuals of these species in our sampling than of those in
Guild 2. These species may colonize gaps in lowland forests as advance
regeneration.
4. Moderately light demanding species.-This group includes species in
which saplings are associated with relatively open (>10%) canopy, and adult trees,
but not saplings, are associated with past disturbance. Species in this group do not
appear to require treefall gaps for germination and establishment of saplings.
However, moderate to high light intensity does seem to favor either the
establishment or persistence of saplings. These light conditions can be found in
the absence of treefalls or other disturbance, even beneath the relatively closed
canopy of upland forest. Guild members Gymnanthes lucida and Nectandra
coriacea are dominants in drier upland forests on upper slopes, where the canopy
is typically uniform and thin; in such environs, saplings of these species were
often abundant beneath the canopy, while adults were conspicuously linked to
older canopy gaps. Other guild members such as Trophis racemosa and Zuelania
guidonia occur commonly as saplings in the absence of treefall gaps in Hill-base
and Sabal forests where a high frequency of leaning trees and relatively broken
canopy combine to create numerous light patches smaller than those created by
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
treefall gaps. While gaps are not required for sapling establishment by members
of this guild, gap occurrence appears to strongly favor growth from sapling to
adult stature.
5. Gap recruiters.-This guild contains species in which sapling
occurrence does not show a correlation with disturbance or canopy opening, but
adult occurrence shows strong positive correlation with disturbance history.
While only one species in our sampling significantly displayed this pattern, our
field observations indicate that this is a common response type. Species in this
group do not appear to require treefall gaps for germination and sapling
establishment, but they do seem to require some canopy disturbance (at least in
upland regions) in order to attain adult size. The main difference between the
requirements of these species and those of members of Guild 7 (shade-tolerant
species) is most likely that Guild 5 members require larger disturbances in order
to reach adult stature. While we did not determine sapling survival rates, it is
likely that members of Guild 5 have lower sapling survival rates under closed
Upland forest canopies than do Guild 7 species. Saplings of Guild 5 species were
less abundant on average than those of Guild 7 species in upland forests where
light availability under intact canopies is generally low (Appendix 2).
6. Semi-shade-tolerant species.-Saplings of these species are positively
correlated with relatively open canopy (>10% open) in upland and/or transitional
regions of the topographic gradient, but uncorrelated with disturbance history.
Some species in this group are abundant on upper slopes, where the canopy is
typically uniform but thin, while others are more common in the hill-base and
Sabal regions of the topographic gradient, where the canopy is typically more
broken. While members of this guild did not show strong correlations with
disturbance, it is likely that they successfully colonize gaps as advance
regeneration. One guild member, Nectandra salicifolia, was among the more
consistently dominant species in adult size classes in 12-year-old logging gaps in
Upland Forest near Tikal (Schulze, unpubl. data), indicating that this species is
well adapted to colonization of canopy gaps.
7. Shade-tolerant species.-Members of this guild show a convincing lack
of correlation with canopy opening and disturbance, a significant negative
correlation with disturbance (Malpighia glabra), or a significant negative
association with degree of canopy opening (Piper psilorrachis, Cryosophila
stauracantha, Pseudolmedia oxyphyllaria, Calyptranthes chytriculia). This group
comprises shade-tolerant species. All members of the group can establish and
maintain saplings under closed canopy, though they may not reach adult size there
without some kind of opportunity for increased light In contrast to members of
more light-demanding guilds, these species probably can utilize small canopy
openings such as those created by a fallen limb or uneven horizontal spacing of
tree crowns, to reach adult sizes; they do not depend on treefalls. Not
surprisingly, the species in this group account for an overwhelming majority of
stems in a given patch of mature Upland Forest.
Ecological Groups
Up to this point we have shown that tree species composition varies
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK
predictably along the predominant topographic/edaphic gradient, and that more or
less distinct distribution patterns can be identified for individual species. Natural
disturbance, in the form of treefall gaps, and variation in vegetation structure,
particularly canopy height and cover, play significant roles in determining species
composition at any given point in space. Species-specific response patterns to
light and disturbance regimes can also be identified. However, species with
similar distribution patterns along the topographic gradient do not necessarily
display the same responses to disturbance, and species with apparently similar
light requirements often display different distribution patterns with respect to the
topographic gradient Hence, topographic/edaphic associations and light/
disturbance response patterns must be considered simultaneously in order to create
a meaningful grouping of tree species based on ecological characters.
By considering topographic distributions and disturbance/light response
guilds jointly for each tree species we have been able to identify 19 ecological
groups for trees at Tikal (Fig. 38). These groups are provisional; further research
will likely show that even within a given "ecological group" species differ
significantly in their tolerances to edaphic and light conditions. Moreover, a
number of additional factors no doubt affect species distributions including:
pathogens, processes and patterns of seed dispersal, seed predation and stochastic
processes. In addition, edaphic and light tolerances are not clear cut and are
probably reflections of interactions among numerous factors, both abiotic and
biotic, which combine to confer reduced fitness and increased mortality on a
species in a given portion of the topographic gradient. We were unable to classify
many species due to small sample sizes. Despite these limitations, the following
classification appears to be a useful translation of the patterns we have recorded in
this sampling into ecologically meaningful species groups, and is more accurate
than simple pioneer/climax species distinctions. In some cases where sample
sizes were small, we have relied on understanding provided by qualitative
observations over several years of field work, to aid in tentatively placing species
into ecological groupings.
Figure 38 depicts the manner in which seven light/disturbance response
guilds and 10 topographic distribution patterns combine to result in 19 ecological
groupings of tree species at Tikal. Appendix 1 shows our tentative allocation of
155 tree species with regard to the three means of categorization.
We distinguish five groups of shade-tolerant tree species (groups 1-5, Fig.
38, Appendix 1). These species groups differ primarily in their success under
edaphic extremes, especially in the ability to tolerate seasonal drought and soil
saturation. The first three species groups appear to be favored by conditions in
upland forest, with group 1 species favored by moderate conditions (Standard and
Mesic Upland), group 2 species reaching highest densities in Dry Upland forest,
and group 3 species able to exploit all upland areas with equal success. Group 4
species achieve peak abundances under mesic conditions, from mesic upland to
Sabal lowland, and group 5 species show no significant topographic associations.
Eight species groups are classified as highly light-demanding (groups 12-19,
Fig. 38, Appendix 1). These species differ in their ability to exploit gap
disturbances to colonize upland forest areas and in their tolerance of drought and
inundation. One extreme is exemplified by group 12 and 14 species, which
215
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
successfully colonize gaps in upland forest, but appear to be highly susceptible to
both flooding and seasonal drought. In contrast, group 17 species are highly
tolerant of both extreme inundation and drought, but do not appear to compete
successfully in gaps in Upland forest, and may even be at a competitive
disadvantage in high-light, seasonally flooded, but not drought-impacted Swamp
and Mesic Bajo. Group 18 and group 19 species also appear to have very low
success in exploiting high-light situations in Upland forest and are thus largely
restricted to lowlands, with group 19 species further restricted to mesic lowland
forests (Mesic Bajo and True Swamp) by intolerance of drought. Species of
groups 13-15 appear to be more successful in colonizing large gaps in Upland
forest and particularly intermediate forests (Transitional, Sabal, Hill-base), but
differ substantially in their edaphic tolerances (Fig. 38). Group 12 species are
equivalent with pioneer species-they are adapted for high success in colonizing
medium to large gaps in upland areas, but appear to be sensitive to extremes in
both drought and inundation and are therefore uncommon in Transitional, Scrub
Swamp, True Swamp and Mesic Bajo.
Moderately light-demanding species (light/disturbance guilds 4-6) can also
be divided based on tolerance of environmental extremes. These species all show
some capacity for seed germination and early survival under low or intermediate
light conditions (canopy opening 10-20%). Species with intermediate shade
tolerance appear to exploit areas with thin or broken canopies for recruitment of
juveniles, and hence are present in treefall gaps as advanced regeneration in a
manner similar to shade-tolerant species. These areas of intermediate light
intensity are more common in some regions of the topographic continuum than
others (e.g. hilltops, upper slopes, and Hill-base/Sabal forest), and distributions of
these species appear partially determined by the availability of appropriate light
conditions across the topographic gradient, and partially by the species' edaphic
tolerances. Species with moderate shade-tolerance, high tolerance of inundation
and low tolerance of drought make up group 7, and are found primarily in True
Swamp and Mesic Bajo. At the other extreme are group 10 species, which are
highly tolerant of excessive soil drainage and drought, but intolerant of
inundation, and occur primarily in Dry Upland. Members of groups 11 and 8 are
tolerant of both drought and inundation, but the former are better able to persist in
the areas producing the most extreme edaphic stress (i.e. Low Scrub Swamp) than
are group 8 species. Group 8 species are particularly abundant on hilltops and
upper slopes, where the thin canopy appears to facilitate recruitment of seedlings
to sapling size classes, allowing these species to exploit canopy openings as
advanced regeneration to reach the canopy, and in Transitional and Tall Scrub
Swamp forests, where treefall gaps do not appear to be necessary for recruitment
during any stage of the life-cycle. Species in groups 6 and 9 show similar
sensitivities to drought, with the former apparently better able to exploit small to
large gaps in Upland forest from seedlings or saplings, and the latter more tolerant
of inundation (evidenced by relatively high densities in True Swamp) and less
successful in competing with advanced regeneration of shade-tolerant species and
pioneers in upland forest gaps.
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 217
CONCLUSIONS
Tropical forests in the lowlands of Pet6n are at most 11,000 years old
(Leyden 1984), and over much of the region age is probably 1,000 years or less.
Hence, the current canopy trees might represent only the third or fourth generation
since disturbance. The question of whether Peten forests have reach a steady-state
equilibrium is a fascinating problem in the study of current forest dynamics.
While it is still not possible to answer this question, our results may contribute to
this discussion. That variation in forest structure and composition (i.e. forest
types) along topographic/edaphic gradients is highly predictable, is indicative of
some degree of stabilization. Throughout the topographic gradient, the dominant
species in the canopy are also abundant in the understory, and appear to be
regenerating successfully, although regeneration strategies vary considerably.
Within a given forest type, many species do not display anything approaching a
reverse-J diameter distribution; some species appear to regenerate successfully in
treefall gaps and seem assured of representation in future generations, while others
are so poorly represented by juveniles, even in the largest gap areas, that
competitive exclusion is a plausible future scenario. However, even for the latter
set of species, healthy populations in adjacent forest types would seem to preclude
local extirpation of the species, and should provide sources for recolonization.
Judgment of whether the forests of Peten are in a steady-state will depend on
the scale considered. In Tikal, it is relatively common to encounter patches of
upland forest up to several hectares in size that are dominated by light-demanding
emergent species, such as Swietenia, Vitex, Pseudobombax, and Cedrela. In these
areas, the understory and subcanopy is dominated by the shade-tolerant species
typical of upland forest, and clearly species composition will shift in the future. It
has been hypothesized that such large assemblages of light-demanding emergents
reflect large disturbances, such as hurricanes (Snook 1992). Given the absence of
these species from small to medium gaps in upland forest, large disturbances do
seem the best explanation for dense upland stands of emergents like Swietenia.
Without the recurrence of a major disturbance event, Swietenia and similar species
will likely decline through time in these upland areas. Nonetheless, in lower
portions of the gradient, Swietenia is able to regenerate in the absence of major
disturbances; Swietenia populations may expand and contract around lowland
sources at intervals of several hundred years, and thus may be stable over large
areas and long time periods. As discussed earlier, we believe that species
associations with Maya ruins are more likely related to edaphic conditions than
directly to activities of the ancient Maya. The evidence we have recorded at Tikal
suggests that the vegetation is in some form of steady-state equilibrium, but not a
static climax formation. This view does not discount the impact that land-use and
forest clearing by the Maya may have had on the forests of Pet6n. In fact, human
disturbance probably altered forest composition substantially, and we likely still
observe the results today. At a regional scale, however, forest composition
appears stable.
Our results suggest that topography (and related edaphic conditions),
ambient light availability, and natural canopy disturbance all play major roles in
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
determining patterns of tree species occurrence in Tikal. Most tree species
showed strong abundance peaks in one or more parts of the topographic
continuum, and many were rare in, or absent from, certain portions of the
continuum. Not all species exhibited unimodal abundance patterns across the
predominant environmental gradient; several showed two distinct abundance
peaks in different portions of the continuum, suggesting that these positions were
similar in at least some environmental characteristic. Responses of trees to the
topographic continuum were as pronounced as were species associations with
certain light and disturbance conditions. Indeed, we found cases in which
incompatibility with local soil/environmental conditions appeared to prevent the
occurrence of a species in spite of appropriate light conditions for regeneration.
Some workers have emphasized the role of specialization for different
regeneration conditions in determining patterns within tropical tree communities
(e.g. Denslow 1980; Pickett 1983; Whitmore 1985; Brandani 1988). Many studies
have focused on treefall gaps as the primary source of variation in light
availability and consequently in species composition. We find this approach
oversimplified, not only in failing to recognize the importance of topographic/
edaphic responses, but also with respect to the variety of light sources and tree
responses to them. Levels of light availability in tropical forests often are
intermediate between those of treefall gaps and continuous canopy, and light often
reaches the understory by means of many small-scale sources (Chazdon 1988;
Lieberman and Lieberman 1989; Clark et al. 1993). Moreover, relatively small
differences in light intensity can have a significant impact on the establishment
and survival of individual trees (Clark and Clark 1992; Kobe et al. 1995).
Our study provides additional evidence that light availability other than from
major treefall gaps can be important to some tree species. We found that ambient
light levels, in addition to being influenced by incidence and size of treefall gaps,
varied across the topographic gradient, largely in concert with, and no doubt as a
result of, canopy height and structural characteristics. In particular, many small
gaps were provided by limb-falls, leaning trees, uneven canopy, and canopy
discontinuities around the periphery of palm crowns. Small variations in light that
did not depend on treefalls appeared to provide important recruitment
opportunities for a number of species having intermediate levels of shade
intolerance. Some highly light-demanding species did not appear to be adapted
primarily for colonizing large treefall gaps but rather appeared to be associated
mainly with areas of thinner canopy and overall higher levels of ambient light in
the understory. Other species that relied on treefall gaps for regeneration in
upland regions regenerated beneath the canopy in portions of the topographic
continuum supporting thin-canopied forests. In lower, more open-canopied
regions of the topographic continuum treefalls did not seem to significantly
determine local species composition.
We suspect that greater diversity may exist in species regeneration strategies
than has been commonly recognized (e.g. Kennedy and Swaine 1992; Lieberman
et al. 1995). For example, some species that primarily colonize gaps as advanced
regeneration (e.g. Nectandra salicifolia, Sebastiana longicuspis) rather than by
seed as do pioneer species, seem to be adapted for more rapid growth in gaps than
are most shade-tolerant species. Even within our shade-tolerant species groups,
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 219
we have noted substantial variation in species responses to increased light
availability.
We found that species guilds based on responses to topographic position and
associated moisture and edaphic factors were recognizable, as were guilds based
upon regeneration requirements with regard to light However, neither of these
methods of categorization was in itself adequate to describe observed patterns; in
some cases two or more species reached abundance peaks in the same portions of
the topographic gradient, but apparently for different reasons. Simultaneous
consideration of species tolerances and optima with respect to both light and
topographic/edaphic conditions leads to a better understanding of community
composition than does either perspective by itself. This species categorization is
based on specific hypotheses about relative growth and survival rates under
varying light and edaphic conditions. These hypotheses should be tested
experimentally; this would no doubt modify guild designations and increase our
understanding of how these environmental factors structure tropical tree
communities.
Physiological tolerance limits and optima with respect to topographically
related environmental conditions only determine the broad boundaries within
which a species can occur. The distribution patterns we recorded are almost
certainly the result of a complex interaction of chance, species attributes, and
many abiotic and biotic factors. In addition to those described above, such factors
may include: soil and litter disturbance requirements for seedling establishment
(cf. Putz 1983); variation in pathogen susceptibility both between species and
between habitats for a single species (cf. Augsperger 1984b, c); seed dispersal
methods and resultant seed shadows; seed predation (Janzen 1970); numerous
stochastic factors (cf. Hubbell and Foster 1986a); and additional factors we have
not mentioned.
Hubbell (1979) and Hubbell and Foster (1986a) have made a strong case
against viewing niche partitioning as the primary factor allowing coexistence of so
many tree species in tropical forest, and have shown that ecologically equivalent
species may coexist for long periods without competitive exclusion, through the
stochastic nature of regeneration. While we believe we have shown strong
evidence of important levels of niche diversification among tree species at Tikal
with respect to edaphic and light environments, we do not propose that this is the
only, or even the dominant, mechanism permitting species coexistence there.
Hubbell and Foster (1986a) further argued that high species diversity in tropical
trees results in largely unpredictable species identity of neighboring individuals,
rendering competitive niche divergence improbable. Unpredictability of neighbor
identity amounts simply to diffuse competition; while such unpredictability may
render unlikely fine-tuned divergence of species characteristics, it does not
preclude specialization for some subset of the broad range of abiotic conditions
existing at Tikal.
There is perhaps a need to refine the question of what mechanisms permit
coexistence of so many tree species in tropical forests, to explicitly consider
coexistence across obvious environmental gradients such as catenas, which
contribute to beta diversity, versus coexistence within a more narrowly defined
portion of the landscape (e.g. well-drained upland sites), amounting to alpha
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
diversity. Presumably, much of the discussion to date surrounding the question of
tropical tree species coexistence has implicitly considered coexistence in
restricted, homogeneous areas, and hence dealt with alpha diversity. In that
regard, the demonstration that many species respond strongly to a fairly
pronounced topographic gradient at Tikal may perhaps be viewed as contributing
little to the debate. However, on closer inspection, it becomes clear that
specialization for some portion of the regionally available habitats may indeed
have important outcomes for competitive and stochastic processes at the level of
alpha diversity, as argued below.
The apparent fact that species are specialized to some degree with respect to
soil moisture and perhaps other edaphic factors, renders the identity of neighbors
more predictable than would otherwise be the case, such that a given species no
doubt interacts more strongly with one subset of potential competitors than with
others. The degree of diffuse or 'background' light penetration to the forest floor
(i.e. not that caused by treefall gaps) is largely determined by structural characters
that are strongly correlated with topographic position, and that presumably
resulted from environmental conditions along the gradient. Hence, the light
environment is in part an emergent property determined indirectly by
topographic/edaphic factors. These two factors-apparent differences in species
optima with respect to environmental factors, and predictable covariation of these
factors (e.g. of diffuse light penetration and topographic/edaphic factors)-
promote interaction among a certain set of species and admit more scope for niche
divergence than generally has been recognized. Furthermore, the fact that most
species have one or more abundance peaks somewhere along the gradient in effect
may give them a refuge from which to send propagules to a broader range of the
environmental gradient In this manner upland regions may be colonized
sporadically by species that are at a competitive disadvantage in those
environmental conditions, but that maintain stable populations in neighboring
lowland areas. A similar phenomenon appears to occur on a regional scale, as
many species with very restricted ranges and low densities in Tikal, are much
better represented in adjacent wetter forests in Belize and western Petdn.
Qualitative observation of forests in tropical Guatemala and Belize lead us to
conclude that it is relatively common for 'rare' species in one locale to be
substantially more abundant in another geographic region, with slightly different
climatic or soil conditions. It is possible that regional 'source' populations in
areas with optimal environmental conditions, provide a critical subsidy, in the
form of genetic exchange and occasional seed input, for marginal populations of
these species in Tikal.
Topographic gradients as pronounced as those in Tikal do not occur in all
tropical areas. However, even in the relatively flat Amazon basin, regional and
local variation in soil and moisture conditions can be substantial (Sombroek 1968,
Tuomisto et al. 1995). Alpha diversity in many tropical areas is considerably
higher than in Tikal, and in such areas there may be more reason to assign a
portion of species richness to stochastic factors. However, in recognizing that
niche divergence along light/disturbance axes is far from the only mode of
divergence available to tropical tree species, it becomes clear that the potential for
niche divergence is indeed large. While stochastic factors no doubt are important,
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK 221
they are played out upon an environmental stage which is more varied and
influential than has generally been recognized.
LITERATURE CITED
Ashton, P.S. 1969. Speciation among tropical forest trees: some deductions in the light of recent
evidence. Biol. J. Linn. Soc. 1:155-196.
Augspurger, C.K. 1984a. Light requirements of neotropical tree seedlings: a comparative study of
growth and survival. J. Ecol. 72:777-795.
. 1984b. Seedling survival of tropical tree species: interactions of dispersal distance, light-gaps
and pathogens. Ecology 65:1705-1712.
. 1984c. Pathogen mortality of tropical tree seedlings: experimental studies of the effects of
dispersal distance, seedling density, and light conditions. Oecologia 61:211-217.
Basnet, K. 1992. Effect of topography on the pattern of trees in a tabonuco (Dacryoides excelsa)
dominated forest of Puerto Rico. Biotropica 24:31-42.
Berry, K.J., and P.W. Mielke, Jr. 1986. R by C chi-square analyses with small expected cell
frequencies. Educ. Psych. Measure. 46:169-173.
Brandani, A., G.S. Hartshorn, and G.H. Orians. 1988. Internal heterogeneity of gaps and species
richness in Costa Rican tropical wet forest J. Trop. Ecol. 4:99-119.
Bray, J.R., and J.T. Curtis. 1957. An ordination of the upland forest communities of southern
Wisconsin. Ecol. Monogr. 27:325-349.
Brokaw, N.V.L. 1982. The definition oftreefall gap and its effect on measures of forest dynamics.
Biotropica 14:158-160.
1985. Treefalls, regrowth, and community structure in tropical forests. Pp. 53-69 in The
ecology of natural disturbance and patch dynamics, S.T.A. Pickett and P.S. White (eds.).
Academic press, New York.
1987. Gap-phase regeneration of three pioneer tree species in a tropical forest. J. Ecol. 75:9-19.
. 1989. Species composition in gaps and structure of a tropical forest. Ecology 70:538-541.
and E.P. Mallory. 1993. Vegetation of the Rio Bravo Conservation and Management Area,
Belize. Manomet Bird Observatory, Manomet, Massachusetts.
Canham, C.D., J.S. Denslow, W.J. Platt, J.R. Runkle, T.A. Spies, and P.S. White. 1990. Light
regimes beneath closed canopies and treefall gaps in temperate and tropical forests. Canadian J.
For. Res. 20:620-631.
Chazdon, R.L. 1988. Sunflecks and their importance to understory plants. Adv. Ecol. Res. 18:1-63.
and N. Fetcher. 1984. Photosynthetic light environments in a lowland tropical rain forest in
Costa Rica. J. Ecol. 72:553-564.
Clark, D.B., and D.A. Clark. 1992. Life history diversity of canopy and emergent trees in a
neotropical rain forest. Ecol. Monogr. 62:315-344.
Clark, D.B., D.A. Clark, and P.M. Rich. 1993. Comparative analysis of microhabitat utilization by
saplings of nine tree species in neotropical rain forest. Biotropica 25:397-407.
Clark, D.A., D.B. Clark, R. Sandoval M., and M.V. Castro C. 1995. Edaphic and human effects on
landscape-scale distributions of tropical rain forest palms. Ecology 76:2581-2594.
Connell, J.M. 1971. On the role of natural enemies in preventing competitive exclusion in some
marine animals and in rain forest trees. Pp. 298-312 in Dynamics of Numbers in Populations,
P.J. den Boer and G.R. Gradwell (eds.). Proc. Advanced Study Inst., Center for Agricultural
Publishing and Documentation (Pudoc), Osterbeek Wageningen, The Netherlands.
Denslow, J.S. 1980. Gap partitioning among tropical rainforest trees. Biotropica 12 (Suppl.):47-55.
1984. Influence of disturbance on species diversity: reply to T.C. Whitmore. Biotropica
16:240.
1987. Tropical rainforest gaps and tree species diversity. Ann. Rev. Ecol. Syst. 18:431-451.
1996. 1995. Disturbance and diversity in tropical rain forests: the density effect Ecol. Applic.
5:962-968.
Eyre, S.R. 1964. Vegetation and Soils, a World Picture, 2'd ed. Aldine Pub. Co., Chicago, Illinois.
Finegan, B. 1996. Pattern and process in neotropical secondary rain forests: the first 100 years of
succession. Trends Ecol. Evol. 11(3): 119-124.
Foth, H.D., and L.M. Turk. 1972. Fundamentals of Soil Science. Wiley, New York.
Furley, P.A., and W.W. Newey. 1979. Variations in plant communities with topography over tropical
limestone soils. J. Biogeogr. 6:1-15.
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
Gauch, H.G. Jr. 1982. Multivariate analysis in community ecology. Cambridge University Press,
Cambridge.
Gentry, A.H. 1981. Distributional patterns and an additional species of the Passiflora vitifolia
complex: Amazonia species diversity due to edaphically differentiated communities. Plant Syst.
Evol. 137:95-105.
Hartshorn, G.S. 1980. Neotropical forest dynamics. Biotropica supplement: Tropical Succession: 23-
30.
Hill, M.O. 1979a. DECORANA-a FORTRAN program for Detrended correspondence analysis and
reciprocal averaging. Ecology and Systematics, Comell University, Ithaca, New York.
1979b. TWINSPAN-a FORTRAN program for arranging multivariate data in an ordered two-
way table by classification of the individuals and attributes. Ecology and Systematics, Cornell
University, Ithaca, New York.
Holdrige, L.R., W.C. Genke, W.H. Hatheway, T. Liang, and J.A. Tosi Jr. 1971. Forest Environments
in Tropical Life Zones: a Pilot Study. Pergamon Press, Oxford.
Hubbell, S.P. 1979. Tree dispersion, abundance and diversity in a tropical dry forest Science
203:1299-1309.
and R.B. Foster. 1986a. Biology, chance, and history and the structure of tropical rain forest tree
communities. Pp. 314-329 in Community Ecology, J. Diamond and T.J. Chase (eds.). Harper
and Row, New York.
and R.B. Foster. 1986b. Commonness and rarity in a neotropical forest- implications for tropical
tree conservation. In Conservation biology: the science of scarcity and diversity, M.E. Soule
(ed.). Sinauer, Sunderland, MA.
and R.B. Foster. 1990. Structure, dynamics and equilibrium status of old-growth forest on Barro
Colorado Island. Pp. 522-541 in Four Neotropical Rainforests, A.H. Gentry (ed.). Yale
University Press, New Haven, CT.
Janzen, D.H. 1970. Herbivores and the number of tree species in tropical forests. Amer. Nat.
104:501-528.
Johnston, M.H. 1992. Soil-vegetation relationships in a tabonuco forest community in the Luquillo
Mountains of Puerto Rico. J. Trop. Ecol. 8:253-263.
Jongman, R.H.G., C.J.F. ter Braak, and O.F.R. van Tongeren (eds.). 1987. Data analysis in
community and landscape ecology. Center for Agricultural Publishing and Documentation
(Pudoc), Wageningen, The Netherlands.
Kahn, F. 1987. The distribution of palms as a function of local topography in Amazonian terra-firme
forests. Experentia 43:251-259.
Kennedy, D.N., and M.D. Swaine. 1992. Germination and growth of colonizing species in artificial
gaps of different sizes in dipterocarp rain forest. Phil. Trans. Royal Soc. London B 335:357-
366.
Kobe, R.K., S.W. Pacala, J.A. Silander, and C.D. Canham. 1995. Juvenile tree survivorship as a
component of shade tolerance. Ecol. Applic. 5:517-531.
Lang, G.E., and D.H. Knight 1979. Decay rates for boles of tropical trees in Panama. Biotropica
11:316-317.
Lemmon, P.E. 1957. A new instrument for meauring forest overstory density. J. Forestry 55:667-
668.
Lescure, J.P., and R. Boulet. 1985. Relationships between soil and vegetation in a tropical rain forest
in French Guiana. Biotropica 17:155-164.
Leyden, B. 1984. Guatemalan forest synthesis after Pleistocene aridity. Proc. Natl. Acad. Sci.
81:4856-59.
Lieberman, D., M. Lieberman, R. Peralta, and G.S. Hartshorn. 1985a. Mortality patterns and stand
turnover rates in a wet tropical forest in Costa Rica. J. Ecol. 73:915-924.
Lieberman, D., M. Lieberman, R. Peralta, and G.S. Hartshorn. 1995. Canopy closure and the
distribution of tropical forest tree species at La Selva, Costa Rica. J. Trop. Ecol. 11:161-178.
Lieberman, M., D. Lieberman, G.S. Hartshorn, and R. Peralta. 1985b. Small-scale altitudinal
variation in lowland wet tropical forest vegetation. J. Ecol. 73:505-516.
Lieberman, M., D. Lieberman, and R. Peralta. 1989. Forests are not just swiss cheese: canopy
stereogeometry of non-gaps in tropical forests. Ecology 70:550-552.
Lundell, C.L. 1937. The Vegetation of Pet6n. Carnegie Inst. Washington, Publ. 478. Washington,
D.C.
Nfiiez-Farfan, J., and R. Dirzo. 1988. Within-gap spatial heterogeneity and seedling performance in a
Mexican tropical forest. Oikos 51:274-284.
SCHULZE & WHITACRE: TREE COMMUNITY OF TIKAL NATIONAL PARK
Oliviera-Filho, A.T., E.A. Vilela, D.A. Carvalho, and M.L. Gavilanes. 1994. Effects of soils and
topography on the distribution of tree species in a tropical riverine forest in south-eastern Brazil.
J. Trop. Ecol. 10:483-508.
Orians, G.H. 1982. The influence of tree-falls in tropical forests on tree species diversity. Trop. Ecol.
23:255-279.
Pennington, T.D., and J. Sarukhan. 1968. Arboles tropicales de Mexico. Inst. Nac. Investig.
Forestales, Mexico, D.F., Mexico.
Pickett, S.T.A. 1983. Differential adaptation of tropical tree species to canopy gaps and its role in
community dynamics. Trop. Ecol. 24:69-84.
Popma, J., and F. Bongers. 1988. The effect of canopy gaps on growth and morphology of seedlings
of rain forest species. Oecologia 75:625-632.
Putz, F.E. 1983. Treefall pits and mounds, buried seeds, and the importance of soil disturbance to
pioneer trees on Barro Colorado Island, Panama. Ecology 64:1069-74.
Rice, W.R. 1989. Analyzing tables of statistical tests. Evolution 43:223-225.
Ricklefs, R.E. 1977. Environmental heterogeneity and plant species diversity: a hypothesis. Amer.
Nat. 111:376-381.
Sanford, R. L., Jr., H.E. Braker, and G.S. Hartshorn. 1986. Canopy openings in a primary neotropical
lowland forest. J. Trop. Ecol. 2:277-282.
Snook, L.C. 1992. Stand dynamics of mahogany (Swietenia macrophylla) and associated species after
fire and hurricane in the tropical forests of Quintana Roo, Mexico. Ph.D. dissertation, Yale
School of Forestry and Environmental Studies, New Haven CT. 138 pp.
Sombroek, W.G. 1968. Amazon Soils, a reconnaissance of the soils of the Brazilian Amazon region.
Center for Agricultural Publishing and Documentation (Pudoc), Wageningen, The Netherlands.
Swaine, M.D., D. Lieberman, and F.E. Putz. 1987. The dynamics of tree populations in tropical
forest. a review. J. Trop. Ecol. 3:359-366.
SYSTAT, Inc. 1990. SYSTAT: The system for statistics. SYSTAT, Inc., Evanston, Illinois.
Terborgh, J., R.B. Foster, and P. Nufiez V. 1996. Tropical tree communities: a test of the non-
equilibrium hypothesis. Ecology 77:561-567.
ter Braak, C.J.F. 1987a. CANOCO-a FORTRAN program for canonical community ordination by
partial detrended canonical correspondence analysis, principal component analysis and
redundancy analysis. TNO, Wageningen, The Netherlands.
1987a. Unimodal models to relate species to environment AC Wageningen, The Netherlands.
ter Steege, H. 1996. HEMIPHOT:a program to analyze vegetation indices, light and light quality
from hemispherical photographs. Tropenbos.
, V.G. Jetten, A.M. Polak, and M.J.A. Werger. 1993. Tropical rain forest types and soil factors in
a watershed area in Guyana. J. Veget. Sci. 4:705-716.
Tuomisto, H., and K. Ruokolainen. 1994. Distribution of Pteridophyta and Melastomataceae along an
edaphic gradient in an Amazonian rain forest J. Veget Sci. 5:25-34.
Tuomisto, H., K. Ruokolainen, R. Kalliola, A. Linna, W. Danjoy, and Z. Rodriguez. 1995. Dissecting
Amazonian diversity. Science 269:63-66.
Uhl, C., K. Clark, N. Dezzeo, and P. Maquirino. 1988. Vegetation dynamics in Amazonian treefall
gaps. Ecology 69:751-763.
Van Groenewoud. 1992. The robustness of correspondence, detrended correspondence, and
TWINSPAN analysis. J. Veget Sci. 3:239-246.
Wallace, A.R. 1878. Tropical nature and other essays. Macmillan and Co., London.
Welden, C.W., S.W. Hewett, S.P. Hubbell, and R.B. Foster. 1991. Sapling survival, growth and
recruitment-relationship to canopy height in a neotropical forest. Ecology 72:35-50.
Whitmore, T.C. 1984. Gap size and species richness in tropical rain forests. Biotropica 16:239.
. 1989. Canopy gaps and the two major groups of forest trees. Ecology 70:536-538.
. 1991. Tropical rainforest dynamics and its implications for management Pp. 67-89 in Rain
Forest Regeneration and Management, A. Gomez-Pompa, T.C. Whitmore and M. Hadley (eds.).
UNESCO, Paris.
Whittaker, R.H. 1956. Vegetation of the Great Smoky Mountains. Ecol. Monogr. 26:1-80.
. 1960. Vegetation of the Siskiyou Mountains, Oregon and California. Ecol. Monogr. 30:279-
338.
_ 1967. Gradient analysis of vegetation. Biol. Rev. 42:207-264.
223
224 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
SCHULZE & WHITACRE: TREE COMMUNITY IN TIKAL NATIONAL PARK
Appendix 1
Tree, treelet, and understory plant species
codes and membership in light/disturbance,
distribution, and ecological groupings
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
Abbreviation Species
acac 'sp Acacia sp.
acaccook Acacia cookii
acacdoli Acacia dolichostachya
acacglom Acacia glomerosa
aegimons Aegiphila monstrosa
alibedul Alibertia edulis
allocomi Allophylus coming
alseyuca Alseis yucatanensis
ampehott Ampelocera hottlei
amyrelim Amyris elemifera
anno 'sp Annona sp.
ardidens Ardisia densiflora
ardipasc Ardisia paschalis
aspicrue Aspidosperma cruenta
aspimega Aspidosperma megalocarpon
astrgrav Astronium graveolens
bactmajo Bactris major
bernflam Bernoullia flammea
blompris Blomia prisca
brosalic Brosimum alicastrum
bucibuse Bucida buceras
burssima Bursera simaruba
byrsbuci Byrsonima bucidaefolia
calobras Calophyllum brasiliense
calychyt Calyptranthes chytraculia
camelati Cameraria latifolia
casebart Casearia bartlettii
casecory Casearia corymbosa
casselip Cassipourea guianensis
cassgran Cassia grandis
castelas Castilla elastica
cecrpelt Cecropia peltata
cedrmexi Cedrela mexicana
ceibpent Ceiba pentandra
celttrin Celtis trinerva
chain 'sp Chamaedorea sp I
cham 'sp2 Chamaedorea sp2
Light /
Disturbance
Response Group
?
4*
?
?
2*
7*
1*
?
7*
1
2*
1
3*
7
7
4*
?
2*
7
7
1*
2
1*
5*
7
1*
7
1*
1*
?
5*
2
2*
2*
2*
7
7
Topographic
Distribution
Group
?
2
?
?
3
2
8
1
2
5
1
7
6
1
3
2
6
4
3
4
8
6
7
10
9
7
1
9
2
9
2
3
3
?
4
3
1
Ecological
Group
?
9
?
?
12
4
18
1 or 6
4
15
12
17
16
1
1
9
16
14
3
2
18
16
17
11
7
17
1
19
13
7
9
12
12
?
14
3
1
SCHULZE & WHITACRE: TREE COMMUNITY IN TIKAL NATIONAL PARK
Abbreviation Species
cham 'sp4 Chamaedorea sp4
champaca Chamaedorea pacaya
chryicac Chrysobalanus icaco
chrymexi Chrysophyllum mexicanum
clus 'sp Clusia sp.
coccacap Coccoloba acapulcensis
cocccozu Coccoloba cozumelensis
coccrefl Coccoloba reflexiflora
cordgera Cordia gerascanthus
cordseba Cordia sebastena
costs 'sp Costus sp.
cousolig Coussapoa oligocephala
crotpyra Croton pyramidalis
crotrefl Croton reflexifolia
cryostau Cryosophila stauracantha
cupabeli Cupania belizensis
cymbpend Cymbopetalum penduliflorum
dendarbo Dendropanax arboreus
dioscamp Diospyros campechiana
dryplate Drypetes laterifolia
erytguat Erythroxylon guatemalensis
euge 'sp Eugenia sp.
exotpani Exothea paniculata
ficuglab Ficus glabra
forctrif Forchammeria trifoliata
glirsepi Gliricidia sepium
guarglab Guarea glabra
gausmaya Gaussia maya
guetcomb Guettarda combsii
gymnluci Gymnanthes lucida
haemcamp Haematoxylum campechianum
hamptril Hampea trilobata
hirtamer Hirtella americana
ingaedul Inga edulis
jacq 'sp Jacquinia sp.
krugferr Krugiodendron ferreum
laettham Laetia thamnia
licapeck Licaria peckii
Light /
Disturbance
Response Group
7
3
1*
1*
hemiepiphyte
1
1
1
4*
1*
3
hemiepiphyte
1
1
7
72*
2*
2*
7
6
1 *
6
?
2*
7
1 *
7*
7*
2
4
1
2*
?
1*
?
6
4*
Topographic
Distribution
Group
9
3
7
2
5
6
7
7
1
7
2
3
7
7
1
2
1
3
4
5
7
5
2
4
7
2
5
7
5
7
7
1
9
6
5
5
6
Ecological
Group
7
12
17
13
15
16
17
17
6
17
13
?
17
17
1
4
12
12
2
8
17
8
?
13
2
17
4
8
17
8
17
17
6?
19
16
8
8
16
227
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
Abbreviation Species
lonccast Lonchocarpus castilloi
loncguat Lonchocarpus guatemalensis
lonclati Lonchocarpus latifolius
loncrugo Lonchocarpus rugosus
malmdepr Malmea depressa
malpglab Malpighia glabra
manizapo Manilkara zapota
margnobi Margaritaria nobilis
mastfoet Mastichodendron foetidesimum
mataoppo Matayba oppositifolia
metobrow Metopium brownei
myriceri Myrica cerifera
myrt 'sp Myrtaceae sp. (pimim)
myrt 'sp Myrtaceae sp.
nectcori Nectandra coriacea
nectsali Nectandra salicifolia
ocot 'sp Ocotea sp.
opplvl OPPLV1 (Myrtaceae sp)
orbicohu Orbignya cohune
oreo 'sp Oreopanax sp.
oureluce Ouratea lucens
pachaqua Pachira aquatica
phylnobi Phyllanthus nobilis
pimedioi Pimenta dioica
pipe 'ad Piper cf aduncum
pipe'ps Piper psilorrachis
pipesemp Piper sempervirens
pithbeli Pithecellobium belizensis
plumobtu Plumeria obtusa
poutamyg Pouteria amygdalina
poutcamp Pouteria campechiana
poutdurl Pouteria durlandii
poutreti Pouteria reticulata
poutsapo Pouteria sapota
protcopa Protium copal
pseuelip Pseudobombax elliptica
pseuoxyp Pseudolmedia oxyphyllaria
psyc 'sp Psychotria sp.
Light/
Disturbance
Response Group
2
?
3
1
7
7
7
1
7*
1*
2*
1
?
4
6*
6*
1
7*
?
1
1'
7
3
7
3
3
1 *
7
6-7*
7
7
7*
6*
2*
7
7
Topographic
Distribution
Group
6
9
6
7
4
3
10
8
3
6
5
7
7
?
4
1
2
7
11
?.
6
9
7
3
2
4
3
9
7
1
4
2
3
2
2
5
1
2
Ecological
Group
16
19
16
17
2
3
5
18
3
16
15
17
17
10
6
9
17
4
16
19
17
3
13
2
12
19
17
1
2
4
3
4
9
15
1
4
SCHULZE & WHITACRE: TREE COMMUNITY IN TIKAL NATIONAL PARK 229
Abbreviation Species
quarfune Quararibeafunebris
quinschi Quiina schippii
rapaguia Rapanea guianensis
rehepenn Rehdera penninervia
sabamaur Sabal mauritiiformis
sapiniti Sapium nitidum
sebalong Sebastiana longicuspis
simaglau Simarouba glauca
simisalv Simira salvadorensis
sponmomb Spondias mombin
stemdonn Stemmadenia donnell-smithit
swarcube Swarizia cubensis
swiemacr Swietenia macrophylla
tabe 'sp Tabebuia sp.
taliflor Talisiafloresii
talioliv Talisia olivaeformis
termnamaz Terminalia amazonia
terntepa Ternstroemia tepazapote
thevahou Thevetia ahouai
thoupauc Thouinia paucidentata
trichava Trichilia havanensis
tricminu Trichilia minutiflora
tricmosc Trichilia moschata
tricpall Trichilia pallida
troprace Trophis racemosa
vatalund Vatairea lundellii
viteguam Vitex guameri
wimmconc Wimmeria concolor
xylo 'sp Xylosma sp.
xylofrut Xylopiafrutescens
zantcari Zanthoxylum caribeum
zantproc Zanthoxylum procerum
zuelguid Zuelania guidonia
Light /
Disturbance
Response Group
7*
1*
2*
7
1*
5
5*
3
2
2
5*
2*
6*
7
7
2*
?
3
4*
1*
7
6*
7
5*
6*
2*
3
1
?
?
5*
4*
Topographic
Distribution
Group
2
?
7
2
2
9
1
2
2
2
1
3
8
2
5
4
6
?
6
4
3
4
10
3
2
2
6
3
7
?
9
1
2
Ecological
Group
4
?
17
13
4
19
6
9
13
13
12
6
18
9
8
2
16
?
16
10
12
2
11
3
9
9
16
6
17
?
19
6
9
* = Assignment to light/disturbance response type was based on qualitative observations rather than
statistical tests due to small sample size for that species.
Numbers refer to patterns described in text. Names of light/disturbance response groups appear in
Figure 37, those of topographic distribution groups in Figures 26-35, ecological groups described in
Figure 38.
230 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
SCHULZE & WHITACRE: TREE COMMUNITY IN TIKAL NATIONAL PARK
Appendix 2
Tree species abundances by forest type. Values
given are mean number of individuals per 0.041 ha
plot for trees > 7.5 cm dbh, and number per 1.7 x
20 m belt transect for juveniles. Means and std
errors are based on 294 sample plots; 'n' gives the
number of individuals detected overall.
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
Dry Upland Mesicl Mesic2
Upland Stand. Upland Upland
Family Species n avg. s.e. avg.. s.e. avg. s.e. avg. s.e.
Anacardiaceae Astronium graveolens 23 0.03 0.03 0.05 0.04 0.00 0.00 0.14 0.05
juveniles
Metopium brownei
juveniles
Spondias mombin
juveniles
Annonaceae Annona sp.
juveniles
Cymbopetalum
penduliflorum
juveniles
Malmea depressa
juveniles
Xylopia frutescens
juveniles
Apocynaceae Aspidosperma cruenta
juveniles
Aspidosperma megalocarpon
juveniles
Cameraria latifolia
juveniles
Plumeria obtusa
juveniles
Stemmadenia donnell-smithii
juveniles
Thevetia ahouai'
Arecaceae Bactris major '
(Palmae) Chamaedorea pacaya'
Chamaedorea spl 1
Chamaedorea sp2 1
Chamaedorea sp4 I
Cryosophila stauracantha
juveniles
Gaussia maya
juveniles
0.07 0.05 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.03 0.03 0.05
0.00 0.00 0.00
0.00 0.00 0.05
0.00 0.00 0.00
0.00 0.00 0.00
1 0.00 0.00 0.06
45 0.53 0.17 0.32
32 0.15 0.07 0.24
1 0.00 0.00 0.00
0 0.00 0.00 0.00
320 0.35 0.14 0.73
48 0.11 0.06 0.06
66 0.21 0.07 0.14
24 0.19 0.07 0.35
20 0.00 0.00 0.00
7 0.00 0.00 0.00
1 0.00 0.00 0.00
6 0.00 0.00 0.00
39 0.09 0.05 0.14
5 0.04 0.04 0.00
26 0.00 0.00 0.00
24 0.00 0.00 0.00
31 0.19 0.18 0.00
330 2.33 0.52 2.88
12 0.04 0.04 0.06
30 0.26 0.16 0.24
4223 6.3 1.0 14.6
1022 5.22 0.77 5.71
16 0.06 0.04 0.23
3 0.04 0.04 0.06
0.00 0.04
0.00 0.00
0.00 0.00
0.04 0.16
0.00 0.04
0.04 0.10
0.00 0.00
0.00 0.13
0.04 0.05 0.04
0.00 0.00 0.00
0.00 0.00 0.00
0.08 0.14 0.05
0.04 0.00 0.00
0.05 0.11 0.05
0.00 0.00 0.00
0.09 0.05 0.03
0.06 0.00 0.00 0.00 0.00
0.13 0.03 0.03 0.19 0.07
0.10 0.13 0.07 0.38 0.12
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.27 1.19 0.18 2.09 0.22
0.06 0.30 0.11 0.38 0.10
0.10 0.13 0.06 0.26 0.08
0.18 0.30 0.13 0.11 0.05
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.07 0.06 0.04 0.18 0.07
0.00 0.00 0.00 0.00 0.00
0.00 0.04 0.04 0.05 0.04
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.11 0.08
0.44 1.57 0.38 3.05 0.50
0.06 0.09 0.06 0.14 0.07
0.16 0.17 0.10 0.19 0.14
1.9 17.7 1.5 20.3 1.1
0.85 6.91 0.64 7.00 0.55
0.16 0.10 0.05 0.09 0.04
0.06 0.04 0.04 0.00 0.00
SCHULZE & WHITACRE: TREE COMMUNITY IN TIKAL NATIONAL PARK
Tall Scrub Low Scrub Mesic bajo/
Hillbase Sabal Transitional Swamp Swamp Swamp
avg. s.e. avg.. s.e. avg. s.e. avg. s.e. avg.. s.e. avg. s.e.
0.00 0.0
0.00 0.0
2.12 0.5
0.64 0.1
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.00
0.05
0.00
0.00
0.00
0.91
0.40
0.32
0.07
0.00
0.00
0.00
0.00
0.18
0.00
0.20
0.00
0.00
1.27
0.00
0.00
16.32
6.13
0.00
0.00
0.00
0.09
0.07
0.00
0.00
0.68
0.13
0.26
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.47
0.40
0.00
1.20
0.00
0.00
28.15
4.07
0.00
0.00
0.00 0.00
0.00 0.00
0.05 0.04
0.00 0.00
0.50 0.18
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00
0.04
0.20
0.04
0.00
2.52
0.30
0.30
0.10
0.00
0.00
0.00
0.00
0.00
0.00
0.20
0.40
0.00
0.30
0.00
0.00
14.48
3.30
0.00
0.00
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
Dry Upland Mesicl Mesic2
Upland Stand. Upland Upland
Family Species n
Orbignya cohune 4
juveniles 21
Sabal mauritiiformis 777
juveniles 128
Araliaceae Dendropanax arboreus 25
juveniles 2
Oreopanax sp. 2
juveniles 0
Bignoniaceae Tabebuia sp. 17
juveniles 3
Bombacaceae Bernoullia flammea 3
juveniles 0
Ceiba pentandra 1
juveniles 0
Pachira aquatica 4
juveniles 0
Pseudobombax elliptica 6
juveniles 0
Quararibea funebris 9
juveniles 3
Boraginaceae Cordia gerascanthus 13
juveniles 0
Cordia sebastena 2
juveniles 0
Burseraceae Bursera simaruba 58
juveniles 0
Protium copal 168
juveniles 47
Capparidaceae Forchammeria trifoliata' 9
Celastraceae Wimmeria concolor 150
juveniles 21
Chrysobalanaceae Chrysobalanus icaco 1
juveniles 0
Hirtella americana 18
juveniles 3
Clusiaceae Calophyllum brasiliense 67
juveniles 16
avg. s.e. avg..
0.00 0.00 0.00
0.00 0.00 0.00
0.35 0.12 0.86
0.48 0.20 0.41
0.12 0.06 0.00
0.00 0.00 0.06
0.03 0.03 0.00
0.00 0.00 0.00
0.03 0.03 0.00
0.00 0.00 0.00
0.00 0.00 0.05
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.03 0.03 0.09
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.05
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.06 0.04 0.09
0.00 0.00 0.00
0.38 0.11 0.59
0.26 0.08 0.12
0.19 0.07 0.00
0.62 0.14 0.95
0.11 0.06 0.12
0.00 0.00 0.00
0.00 0.00 0.00
0.03 0.03 0.14
0.00 0.00 0.00
0.00 0.00 0.23
0.00 0.00 0.00
s.e. avg. s.e. avg. s.e.
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.26 1.10 0.28 2.14 0.31
0.12 0.30 0.11 0.62 0.14
0.00 0.19 0.07 0.19 0.09
0.06 0.04 0.04 0.00 0.00
0.00 0.00 0.00 0.02 0.02
0.00 0.00 0.00 0.00 0.00
0.00 0.03 0.03 0.00 0.00
0.00 0.00 0.00 0.03 0.03
0.04 0.00 0.00 0.04 0.03
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.09 0.00 0.00 0.02 0.02
0.00 0.00 0.00 0.00 0.00
0.00 0.06 0.04 0.04 0.02
0.00 0.04 0.04 0.00 0.00
0.04 0.03 0.03 0.16 0.05
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.06 0.06 0.04 0.14 0.05
0.00 0.00 0.00 0.00 0.00
0.15 0.55 0.14 0.49 0.09
0.08 0.13 0.07 0.30 0.08
0.00 0.00 0.00 0.08 0.04
0.20 1.06 0.25 1.12 0.18
0.08 0.22 0.11 0.24 0.10
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.07 0.10 0.05 0.16 0.05
0.00 0.04 0.04 0.05 0.04
0.13 0.10 0.10 0.16 0.05
0.00 0.00 0.00 0.03 0.03
SCHULZE & WHITACRE: TREE COMMUNITY IN TIKAL NATIONAL PARK 235
Tall Scrub Low Scrub Mesic bajo/
Hillbase Sabal Transitional Swamp Swamp Swamp
avg. s.e. avg.. s.e. avg. s.e. avg. s.e. avg.. s.e. avg. s.e.
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
4.50 0.83 10.03 0.84 3.52 0.56 0.81 0.30 0.12 0.11 1.36 0.42
1.47 0.42 1.20 0.25 0.90 0.28 0.58 0.26 0.00 0.00 0.90 0.36
0.09 0.06 0.00 0.00 0.04 0.04 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.09 0.06 0.21 0.07 0.00 0.00 0.00 0.00 0.00 0.00 0.27 0.09
0.00 0.00 0.07 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.03 0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.18 0.18
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.04 0.04 0.06 0.06 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.05 0.04 0.09 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.05 0.04
0.00 0.00 0.07 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.03 0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.04 0.04 0.00 0.00 0.06 0.06 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.14 0.10 0.06 0.04 0.30 0.13 0.69 0.28 0.24 0.10 0.14 0.10
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.50 0.20 1.35 0.31 1.09 0.28 0.25 0.19 0.00 0.00 0.18 0.18
0.13 0.09 0.67 0.24 0.40 0.15 0.17 0.11 0.00 0.00 0.00 0.00
0.00 0.00 0.07 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.09 0.06 0.03 0.03 0.26 0.15 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.06 0.02 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00- 0.00
0.00 0.00 0.00 0.00 0.09 0.09 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.09 0.06 0.21 0.08 0.48 0.14 1.06 0.37 0.29 0.16 0.36 0.16
0.07 0.06 0.07 0.06 0.20 0.12 0.92 0.45 0.00 0.00 0.00 0.00
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
Dry Upland Mesici Mesic2
Upland Stand. Upland Upland
Family Species n
Clusia sp. 21
juveniles 4
Combretaceae Bucida buceras 55
Ebenaceae
juveniles
Terminalia amazonia
juveniles
Diospyros campechiana
juveniles
Erythroxylaceae Erythroxylon
guatemalensis
juveniles
Euphorbiaceae Croton pyramidalis
juveniles
Croton reflexifolia
juveniles
Drypetes lateriflora
juveniles
Gymnanthes lucida
juveniles
Margaritaria nobilis
juveniles
Phyllanthus nobilis
juveniles
Fabaceae
(Caesalpinioideae)
(Papilionoideae)
Sapium nitidum
juveniles
Sebastiana longicuspis
juveniles
Cassia grandis
juveniles
Haematoxylum
campechianum
juveniles
Swartzia cubensis
juveniles
Gliricidia sepium
juveniles
Lonchocarpus castilloi
juveniles
Lonchocarpus
avg. s.e. avg..
0.12 0.06 0.05
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.09 0.05 0.00
0.00 0.00 0.06
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.05
0.04 0.04 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.76 0.21 0.18
0.33 0.12 0.06
1.03 0.56 0.14
1.11 0.60 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.12 0.06 0.68
0.04 0.04 0.47
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.03 0.03 0.00
0.00 0.00 0.06
0.00 0.00 0.00
0.00 0.00 0.00
0.06 0.04 0.00
0.00 0.00 0.00
0.00 0.00 0.00
Dry Upland Mesicl Mesic2
Upland Stand. Upland Upland
g
uatemaenss
s.e. avg. s.e. avg. s.e.
0.04 0.00 0.00 0.09 0.06
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.03 0.03 0.04 0.02
0.00 0.00 0.00 0.00 0.00
0.00 0.03 0.03 0.00 0.00
0.06 0.00 0.00 0.03 0.03
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.04 0.03 0.03 0.00 0.00
0.00 0.00 0.00 0.05 0.04
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.08 0.06 0.04 0.14 0.05
0.06 0.13 0.07 0.08 0.04
0.13 0.00 0.00 0.02 0.02
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.04 0.02
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.29 0.77 0.42 1.79 0.46
0.35 0.04 0.04 0.22 0.08
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.13 0.06 0.05 0.03
0.06 0.04 0.04 0.11 0.05
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.10 0.05 0.07 0.03
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.02 0.02
SCHULZE & WHITACRE: TREE COMMUNITY IN TIKAL NATIONAL PARK
Tall Scrub
Hillbase Sabal Transitional Swamp
avg. s.e. avg.. s.e. avg. s.e. avg. s.e.
Low Scrub
Swamp
avg.. s.e.
0.29 0.18
0.29 0.16
1.00 0.31
0.14 0.09
0.12 0.08
0.00 0.00
0.00 0.00
0.00 0.00
0.35 0.14
0.14 0.09
9.41 0.92
13.00 2.39
0.47 0.28
0.29 0.10
0.00 0.00
0.00 0.00
0.29 0.23
0.50 0.35
0.06 0.06
3.93 0.84
0.65 0.20
0.50 0.17
0.00 0.00
0.00 0.00
0.12 0.11
0.00 0.00
0.00 0.00
0.00 0.00
6.59 0.87
Mesic bajo/
Swamp
avg. s.e.
0.00 0.00
0.00 0.00
1.23 0.40
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.10 0.09
1.18 0.33
1.00 0.65
0.05 0.04
0.00 0.00
0.00 0.00
0.00 0.00
0.05 0.04
0.00 0.00
1.32 0.63
5.40 3.58
0.32 0.19
0.10 0.09
0.14 0.10
0.00 0.00
1.55 0.68
0.00 0.00
0.27 0.18
0.00 0.00
2.27 1.72
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
Dry Upland Mesici Mesic2
Upland Stand. Upland Upland
Family Species
juveniles
Lonchocarpus latifolius
juveniles
Lonchocarpus rugosus
juveniles
Platymiscium yucatanum
juveniles
Vatairea lundellii
juveniles
(Mimosoideae) Acacia cookii
juveniles
Acacia dolichostachya
juveniles
Acacia glomerosa
juveniles
Acacia sp.
juveniles
Inga edulis
juveniles
Pithecellobium c.f. pachypus
juveniles
Pithecellobium belizensis
juveniles
Flacourtiaceae Casearia bartlettii
juveniles
Casearia corymbosa
juveniles
Laetia thamnia
juveniles
Xylosma sp.
juveniles
Zuelania guidonia
juveniles
Lauraceae Licaria peckii
juveniles
Nectandra coriacea
juveniles
n
avg. s.e. avg..
0.00 0.00 0.00
0.00 0.00 0.05
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.24 0.08 0.36
0.11 0.06 0.18
0.03 0.03 0.05
0.00 0.00 0.00
0.03 0.03 0.05
0.00 0.00 0.00
0.03 0.03 0.14
0.00 0.00 0.06
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.09 0.06 0.09
0.07 0.05 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.21 0.09 0.18
0.07 0.05 0.00
0.03 0.03 0.00
0.00 0.00 0.00
0.00 0.00 0.05
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.06
0.94 0.29 0.14
129 2.67 0.67 0.88 0.31 0.65 0.29 0.22 0.09
Dry Upland Mesicl Mesic2
Upland Stand. Upland Upland
s.e. avg. s.e. avg. s.e.
0.00 0.00 0.00 0.00 0.00
0.04 0.03 0.03 0.02 0.02
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.03 0.03 0.05 0.03
0.00 0.00 0.00 0.00 0.00
0.15 0.42 0.13 0.42 0.11
0.09 0.17 0.10 0.11 0.05
0.04 0.00 0.00 0.05 0.03
0.00 0.00 0.00 0.00 0.00
0.04 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.07 0.00 0.00 0.05 0.03
0.06 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.06 0.23 0.11 0.49 0.11
0.00 0.09 0.06 0.08 0.04
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.10 0.29 0.14 0.18 0.06
0.00 0.04 0.04 0.11 0.05
0.00 0.03 0.03 0.04 0.02
0.00 0.00 0.00 0.00 0.00
0.04 0.03 0.03 0.07 0.03
0.00 0.00 0.00 0.00 0.00
0.00 0.13 0.08 0.19 0.06
0.06 0.04 0.04 0.19 0.06
0.10 0.32 0.24 0.07 0.03
238
SCHULZE & WHITACRE: TREE COMMUNITY IN TIKAL NATIONAL PARK
Tall Scrub Low Scrub Mesic bajo/
Sabal Transitional Swamp Swamp Swamp
s.e. avg.. s.e. avg. s.e. avg. s.e. avg..
s.e.
Hillbase
avg.
0.00 (
0.09 (
0.00 (
0.00 (
0.00 (
0.00 (
0.00 (
0.18 t
0.07 (
0.55 (
0.20 (
0.09 (
0.00 (
0.00 (
0.00 (
0.00 (
0.07 (
0.00 (
0.00 (
0.00 (
0.00 t
0.14 (
0.53 C
0.09 (
0.00 C
0.00 C
0.00 (
0.05 (
0.07 C
0.00 (
0.27 C
0.09 C
0.13 (
0.09 C
0.00 C
0.00 6
0.00 6
s.e. avg.
0.00 0.00
0.00 0.77
0.00 0.60
0.00 0.09
0.24 0.00
0.00 0.00
0.00 0.00
0.00 0.23
0.00 0.00
0.00 0.27
0.21 0.10
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.24 0.32
0.00 0.00
0.00 0.91
0.00 0.00
0.06 0.09
0.07 0.20
0.11 3.41
0.00 1.40
0.00 0.00
0.00 0.00
0.00 0.59
0.00 0.30
0.06 0.00
0.22 0.00
0.08 0.05
0.69 0.00
0.06 0.05
0.00 0.00
0.00 0.14
0.00 0.10
0.00 0.00
0.00 0.00
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
Dry Upland Mesici Mesic2
Upland Stand. Upland Upland
Family Species n
Nectandra salicifolia 17
juveniles 41
Ocotea sp. 8
juveniles 3
Malphigiaceae Byrsonima bucidaefolia 23
juveniles 7
Malpighia glabra 1
juveniles 26
Malvaceae Hampea trilobata 12
juveniles 12
Meliaceae Cedrela mexicana 12
juveniles 1
Guarea glabra 34
juveniles 9
Swietenia macrophylla 47
juveniles 6
Trichilia havanensis 1
juveniles 3
Trichilia minutiflora 275
juveniles 166
Trichilia moschata 239
juveniles 117
Trichilia pallida 2
juveniles 42
Moraceae Brosimum alicastrum 366
juveniles 30
Castilla elastica 3
juveniles 0
Cecropia peltata 32
juveniles 0
Coussapoa oligocephala 17
juveniles 0
Ficus glabra 6
juveniles 0
Ficus salicifolia 1
juveniles 0
Ficus sp (free standing) 6
avg. s.e. avg..
0.00 0.00 0.00
0.00 0.00 0.18
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.33 0.16 0.12
0.00 0.00 0.00
0.00 0.00 0.00
0.06 0.04 0.05
0.00 0.00 0.00
0.03 0.03 0.05
0.04 0.04 0.06
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
2.24 0.39 1.91
1.52 0.37 1.12
1.74 0.48 0.86
0.74 0.48 0.06
0.00 0.00 0.00
0.19 0.09 0.59
3.50 0.47 2.32
0.30 0.11 0.06
0.00 0.00 0.00
0.00 0.00 0.00
0.09 0.06 0.09
0.00 0.00 0.00
0.12 0.07 0.09
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
Dry Upland Mesicl Mesic2
Upland Stand. Upland Upland
s.e. avg. s.e. avg. s.e.
0.00 0.00 0.00 0.05 0.04
0.12 0.13 0.06 0.26 0.12
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.02 0.02
0.08 0.22 0.09 0.08 0.04
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.04 0.06 0.04 0.02 0.02
0.00 0.00 0.00 0.00 0.00
0.04 0.13 0.06 0.05 0.03
0.06 0.00 0.00 0.03 0.03
0.00 0.00 0.00 0.05 0.03
0.00 0.00 0.00 0.03 0.03
0.00 0.00 0.00 0.00 0.00
0.00 0.04 0.04 0.00 0.00
0.56 0.77 0.19 0.98 0.17
0.30 1.43 0.40 0.97 0.24
0.53 0.77 0.40 0.53 0.15
0.06 0.35 0.16 0.41 0.13
0.00 0.00 0.00 0.02 0.02
0.19 0.35 0.12 0.35 0.09
0.28 1.16 0.23 0.82 0.13
0.06 0.22 0.11 0.22 0.09
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.06 0.13 0.08 0.12 0.06
0.00 0.00 0.00 0.00 0.00
0.09 0.06 0.04 0.04 0.02
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
SCHULZE & WHITACRE: TREE COMMUNITY IN TIKAL NATIONAL PARK
Hillbase
avg.
0.40 0
0.14 (
0.32
0.07
0.00
0.00
0.00 t
0.00 (
0.00 (
0.07 t
0.05 t
0.00 0
0.55 0
0.07 C
0.09 6
0.07 6
0.00 6
0.00 6
1.27 6
1.13 0
1.23 0
1.40 0
0.00 0
0.13 0
1.41 0
0.13 0
0.05 0
0.00 0
0.05 0
0.00 0
0.05 0
0.00 0
0.05 0
0.00 0
0.00 0
0.00 0
0.05 0
Mesic bajo/
Swamp
s.e.
1.32
1.07
1.15
1.06
1.00
1.00
.00
1.00
1.00
1.06
1.04
.00
1.17
.06
1.06
1.06
.00
.00
.40
.41
.34
.59
.00
).09
.33
.09
.04
.00
.04
.00
.04
.00
.04
.00
.00
.00
.04
Sabal
avg..
0.33
0.12
0.03
0.20
0.00
0.00
0.00
0.07
0.00
0.00
0.06
0.07
0.18
0.07
0.18
0.00
0.03
0.00
0.44
0.20
0.38
0.67
0.03
0.27
1.44
0.13
0.06
0.00
0.06
0.00
0.12
0.00
0.09
0.00
0.00
0.00
0.00
Transitional
avg. s.e.
0.00 0.00
0.13 0.09
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.10 0.09
0.04 0.04
0.10 0.09
0.00 0.00
0.00 0.00
0.00 0.00
0.10 0.09
0.52 0.15
0.10 0.09
0.00 0.00
0.00 0.00
0.26 0.11
0.00 0.00
0.48 0.35
1.10 0.90
0.00 0.00
0.00 0.00
0.00 0.00
0.10 0.09
0.00 0.00
0.00 0.00
0.04 0.04
0.00 0.00
0.09 0.06
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
Tall Scrub
Swamp
avg. s.,
0.00 0.1
0.00 0.1
0.00 0.(
0.00 0.O
0.38 0.J
0.17 0.j
0.00 0.(
0.00 0.1
0.50 0.,
0.33 O.i
0.00 0.(
0.00 0.(1
0.00 0.f
0.00 0.1
0.19 0.J
0.00 0.(
0.00 0.1
0.00 0.1
0.06 0.C
0.00 0O1
0.00 0.C
0.00 0.(
0.00 O.(
0.00 0O1
0.00 0.C
0.00 0.C
0.00 0.C
0.00 0.C
0.00 0.C
0.00 0.C
0.00 0.6
0.00 0.6
0.00 0.6
0.00 0.6
0.06 0.6
0.00 0.6
0.13 0.6
Low Scrub
Swamp
avg.. s.e
0.00 0.6
0.00 0.6
0.00 0.0
0.00 0.0
1.00 0.2
0.36 0.1
0.00 0.0
0.00 0.0
0.12 0.0
0.43 0.1
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.47 0.1
0.14 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.01
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.01
0.00 0.Oi
0.00 0.Oi
0.00 0.0i
0.00 0.Oi
0.00 0.Oi
0.00 0.Oi
0.00 0.aO
0.00 0.01
241
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
Dry Upland Mesicl Mesic2
Upland Stand. Upland Upland
Family Species
juveniles
Ficus sp (strangler)
juveniles
Pseudolmedia oxyphyllaria
juveniles
Trophis racemosa
juveniles
Myricaceae Myrica cerifera
juveniles
Myrsinaceae Ardisia densiflora
juveniles
Ardisia paschalis 1
Rapanea guianensis
juveniles
Myrtaceae Calyptranthes chytraculia
juveniles
Calyptranthes sp.
juveniles
Eugenia sp.
juveniles
Myrtaceae sp.
juveniles
Myrtaceae sp. (pimim)
juveniles
OPPLVI
juveniles
Pimenta dioica
juveniles
Ochnaceae Ouratea lucens
juveniles
Piperaceae Piper cf. aduncum '
Piper psilorrachis 1
Piper sempervirens'
Polygonaceae Coccoloba acapulcensis
juveniles
Coccoloba cozumelensis
juveniles
n
avg. s.e. avg..
0.00 0.00 0.00
0.06 0.04 0.00
0.00 0.00 0.00
0.71 0.19 0.95
0.22 0.10 0.88
0.00 0.00 0.00
0.04 0.04 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.22 0.15 0.06
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.11 0.11 0.00
0.15 0.06 0.00
0.78 0.25 0.24
0.26 0.08 0.00
0.03 0.03 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.59 0.13 0.64
0.48 0.15 0.35
0.00 0.00 0.00
0.00 0.00 0.06
0.15 0.15 0.06
11.0 1.45 8.59
7
0.11 0.06 0.88
0.06 0.04 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
Dry Upland Mesicl Mesic2
Upland Stand. Upland Upland
s.e. avg. s.e. avg. s.e.
0.00 0.00 0.00 0.00 0.00
0.00 0.06 0.04 0.04 0.02
0.00 0.00 0.00 0.00 0.00
0.30 1.81 0.36 3.42 0.48
0.63 0.52 0.21 1.03 0.27
0.00 0.00 0.00 0.05 0.03
0.00 0.04 0.04 0.03 0.03
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.02 0.02
0.00 0.00 0.00 0.00 0.00
0.06 0.52 0.25 0.27 0.11
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.03 0.03
0.00 0.00 0.00 0.09 0.05
0.13 0.04 0.04 0.16 0.07
0.00 0.22 0.11 0.22 0.08
0.00 0.03 0.03 0.02 0.02
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.21 0.32 0.12 0.40 0.11
0.18 0.52 0.17 0.62 0.15
0.00 0.00 0.00 0.19 0.06
0.06 0.09 0.06 0.11 0.06
0.06 0.09 0.06 0.51 0.15
1.23 5.61 0.70 4.92 0.59
0.74 0.04 0.04 0.19 0.11
0.00 0.00 0.00 0.02 0.02
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
SCHULZE & WHITACRE: TREE COMMUNITY IN TIKAL NATIONAL PARK 243
Hillbase Sabal
Tall Scrub Low Scrub
Transitional Swamp Swamp
avg. s.e. avg. s.e. avg.. s.e.
0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00
1.48 0.52 0.00 0.00 0.00 0.00
0.10 0.09 0.00 0.00 0.00 0.00
0.13 0.07 0.00 0.00 0.06 0.06
0.10 0.09 0.08 0.08 0.00 0.00
0.00 0.00 0.00 0.00 0.06 0.06
0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.19 0.13 0.24 0.13
0.00 0.00 0.17 0.11 0.64 0.28
0.80 0.30 0.17 0.11 0.21 0.21
0.00 0.00 0.00 0.00 0.24 0.10
0.00 0.00 0.17 0.11 0.36 0.19
0.00 0.00 0.00 0.00 0.06 0.06
0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00
0.10 0.09 0.33 0.19 0.00 0.00
0.09 0.06 0.00 0.00 0.00 0.00
0.20 0.12 0.25 0.13 0.00 0.00
1.10 0.44 2.00 0.71 3.93 1.40
0.26 0.13 0.19 0.10 0.47 0.17
0.00 0.00 0.00 0.00 0.06 0.06
0.10 0.09 2.25 2.25 3.36 1.42
0.09 0.09 0.63 0.23 1.76 0.42
0.20 0.18 1.08 0.38 2.57 0.60
0.43 0.15 0.19 0.10 0.00 0.00
0.50 0.28 1.42 0.47 0.00 0.00
0.26 0.15 0.00 0.00 0.06 0.06
2.50 0.99 3.83 0.78 1.50 0.30
0.10 0.09 0.00 0.00 0.00 0.00
0.20 0.12 0.00 0.00 0.00 0.00
Mesic bajo/
Swamp
avg. s.e.
0.00 0.00
0.23 0.13
0.00 0.00
0.00 0.00
0.00 0.00
0.32 0.15
0.10 0.09
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
1.09 0.56
1.10 0.74
0.05 0.04
0.50 0.47
0.14 0.13
0.20 0.13
0.40 0.21
0.00 0.00
0.00 0.00
0.00 0.00
0.36 0.16
0.50 0.29
0.05 0.04
0.00 0.00
0.00 0.00
2.50 1.02
0.00 0.00
0.00 0.00
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
Dry Upland Mesicl Mesic2
Upland Stand. Upland Upland
Family Species
Coccoloba reflexiflora
juveniles
Quinaceae Quiina schippii
juveniles
Rhamnaceae Krugiodendron ferreum
juveniles
Rhizophoraceae Cassipourea guianensis
juveniles
Rubiaceae Alibertia edulis 1
Alseis yucatanensis
juveniles
Guettarda combsii
juveniles
Rubiaceae sp (ruins)
juveniles
Rubiaceae sp.
juveniles
Simira salvadorensis
Rutaceae
juveniles
Amyris elemifera
juveniles
Zanthoxylum caribaeum
juveniles
Zanthoxylum procerum
juveniles
Sapindaceae Allophyllus coming
juveniles
Blomia prisca
juveniles
Cupania belizensis
juveniles
Exothea paniculata
juveniles
Matayba oppositifolia
juveniles
Talisia floresii
juveniles
n avg. s.e. avg..
144 0.00 0.00 0.00
66 0.00 0.00 0.00
5 0.00 0.00 0.05
0 0.00 0.00 0.00
2 0.06 0.04 0.00
0 0.00 0.00 0.00
22 0.00 0.00 0.00
1 0.00 0.00 0.00
16 0.00 0.00 0.00
19 0.00 0.00 0.00
1 0.00 0.00 0.00
40 0.00 0.00 0.00
4 0.00 0.00 0.00
3 0.09 0.09 0.00
6 0.00 0.00 0.00
38 0.03 0.03 0.05
123 0.11 0.06 0.12
94 0.12 0.08 0.18
46 0.11 0.06 0.12
16 0.03 0.03 0.00
35 0.41 0.12 0.24
1 0.00 0.00 0.00
0 0.00 0.00 0.00
22 0.09 0.06 0.00
9 0.11 0.06 0.06
7 0.00 0.00 0.00
3 0.00 0.00 0.00
274 1.76 0.42 1.50
30 0.19 0.07 0.35
57 0.03 0.03 0.23
63 0.15 0.09 0.59
10 0.09 0.05 0.00
13 0.07 0.07 0.18
70 0.00 0.00 0.05
21 0.00 0.00 0.06
16 0.00 0.00 0.00
1 0.00 0.00 0.00
s.e. avg. s.e. avg. s.e.
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.04 0.00 0.00 0.02 0.02
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.35 0.19 0.09 0.04
0.00 0.00 0.00 0.03 0.03
0.00 0.03 0.03 0.04 0.02
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.03 0.03
0.04 0.19 0.10 0.16 0.06
0.08 0.17 0.08 0.24 0.10
0.08 0.16 0.07 0.09 0.06
0.08 0.30 0.18 0.16 0.06
0.00 0.00 0.00 0.00 0.00
0.13 0.00 0.00 0.05 0.04
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.10 0.07 0.21 0.08
0.06 0.00 0.00 0.05 0.04
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.29 1.39 0.31 1.86 0.33
0.14 0.35 0.12 0.16 0.06
0.13 0.16 0.08 0.28 0.06
0.32 0.39 0.15 0.38 0.10
0.00 0.06 0.04 0.07 0.03
0.09 0.00 0.00 0.19 0.11
0.04 0.06 0.04 0.18 0.08
0.06 0.00 0.00 0.05 0.04
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
244
SCHULZE & WHITACRE: TREE COMMUNITY IN TIKAL NATIONAL PARK 245
Tall Scrub Low Scrub Mesic bajo/
Hillbase Sabal Transitional Swamp Swamp Swamp
avg s.e. avg. .e avg. .e. avgg s.e. avg. s.e. avg. i.e.
0.00 0.00 0.03 0.03 0.48 0.17 1.56 0.34 6.06 1.00 0.18 0.10
0.00 0.00 0.00 0.00 0.09 0.02 0.63 0.04 3.06 1.20 0.09 0.02
0.05 0.04 0.06 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.47 0.17 0.13 0.09 0.13 0.12 0.00 0.00 0.05 0.04
0.00 0.00 0.07 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.07 0.06 0.93 0.59 0.00 0.00 0.00 0.00 0.00 0.00 0.10 0.09
0.00 0.00 0.00 0.00 0.04 0.04 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.03 0.03 0.30 0.10 0.25 0.11 0.53 0.25 0.00 0.00
0.00 0.00 0.07 0.06 0.20 0.12 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.33 0.26 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.18 0.08 0.29 0.11 0.09 0.06 0.06 0.06 0.00 0.00 0.00 0.00
0.33 0.26 1.60 1.08 1.10 0.55 1.50 0.53 0.79 0.32 3.30 1.23
0.36 0.14 0.82 0.15 0.48 0.17 0.31 0.19 0.06 0.06 0.82 0.27
0.20 0.14 0.27 0.11 0.60 0.24 0.67 0.31 0.00 0.00 0.50 0.29
0.09 0.09 0.06 0.04 0.04 0.04 0.31 0.19 0.00 0.00 0.23 0.13
0.20 0.14 0.00 0.00 0.10 0.09 0.00 0.00 0.86 0.28 0.20 0.13
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.05 0.04
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.09 0.06 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.20 0.12 0.08 0.08 0.00 0.00 0.00 0.00
0.05 0.04 0.00 0.00 0.04 0.04 0.06 0.06 0.00 0.00 0.14 0.13
0.07 0.06 0.00 0.00 0.00 0.00 0.17 0.11 0.00 0.00 0.00 0.00
0.50 0.17 0.18 0.08 0.00 0.00 0.00 0.00 0.00 0.00 0.05 0.04
0.20 0.10 0.07 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.55 0.17 0.24 0.10 0.13 0.07 0.06 0.06 0.00 0.00 0.00 0.00
0.20 0.10 0.93 0.39 0.50 0.15 0.08 0.08 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.04 0.04 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.09 0.06 1.22 0.46 1.50 0.51 0.06 0.06 0.00 0.00
0.00 0.00 0.20 0.10 0.20 0.12 1.08 0.56 0.00 0.00 0.00 0.00
0.05 0.04 0.09 0.05 0.17 0.12 0.13 0.08 0.35 0.17 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.07 0.07 0.00 0.00
BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
Dry Upland Mesicl Mesic2
Upland Stand Upland Upland
Family Species n
Talisia olivaeformis 45
juveniles 30
Thouinia paucidentata 16
juveniles 2
Sapotaceae Chrysophyllum mexicanum 7
juveniles 1
Manilkara zapota 229
juveniles 82
Mastichodendron 12
juveniles (foetidesimum) 0
Pouteria amygdalina 173
juveniles 69
Pouteria campechiana 100
juveniles 26
Pouteria durlandii 257
juveniles 84
Pouteria reticulata 955
juveniles 425
Pouteria sapota 1
juveniles 0
Simaroubaceae Simarouba glauca 12
juveniles 10
Solanaceae Cestrum racemosum I
juveniles 0
Theophrastaceae Jacquinia sp 3
juveniles 11
Ulmaceae Ampelocera hottlei 13
juveniles 11
Celtis trinerva 13
juveniles 2
Verbenaceae Aegiphila monstrosa 12
juveniles 1
Rehdera penninervia 12
juveniles 7
Vitex guameri 27
juveniles 0
avg. s.e. avg..
0.47 0.14 0.45
0.30 0.13 0.24
0.35 0.12 0.09
0.00 0.00 0.00
0.03 0.03 0.00
0.00 0.00 0.00
1.29 0.27 0.59
0.22 0.10 0.29
0.09 0.05 0.14
0.00 0.00 0.00
0.21 0.08 0.68
0.37 0.14 0.47
0.74 0.13 0.73
0.11 0.08 0.06
0.09 0.05 0.18
0.11 0.08 0.12
4.76 0.52 5.09
3.44 0.58 3.65
0.00 0.00 0.00
0.00 0.00 0.00
0.03 0.03 0.05
0.07 0.07 0.12
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.05
0.11 0.08 0.00
0.00 0.00 0.00
0.00 0.00 0.06
0.21 0.09 0.09
0.04 0.04 0.00
0.00 0.00 0.05
0.00 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.06
0.03 0.03 0.05
0.00 0.00 0.00
Dry Upland Mesicl Mesic2
Upland Stand Upland Upland
' = treelet or shrub species no individuals > 7.5 cm dbh
s.e. avg. s.e. avg. s.e.
0.15 0.19 0.10 0.05 0.03
0.10 0.17 0.10 0.16 0.08
0.06 0.00 0.00 0.02 0.02
0.00 0.00 0.00 0.05 0.04
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.14 1.13 0.27 0.63 0.13
0.14 0.17 0.08 0.27 0.08
0.07 0.06 0.04 0.02 0.02
0.00 0.00 0.00 0.00 0.00
0.23 1.10 0.23 1.56 0.18
0.15 1.00 0.25 0.57 0.15
0.17 0.61 0.17 0.49 0.09
0.06 0.30 0.11 0.19 0.06
0.14 0.61 0.25 0.70 0.12
0.08 0.17 0.08 0.32 0.14
0.61 4.71 0.56 3.98 0.35
0.68 2.48 0.51 1.97 0.24
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.04 0.06 0.04 0.07 0.03
0.08 0.00 0.00 0.05 0.04
0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.04 0.00 0.00 0.02 0.02
0.00 0.00 0.00 0.00 0.00
0.00 0.03 0.03 0.02 0.02
0.06 0.00 0.00 0.14 0.06
0.06 0.06 0.04 0.04 0.02
0.00 0.00 0.00 0.03 0.03
0.04 0.10 0.07 0.00 0.00
0.00 0.00 0.00 0.00 0.00
0.00 0.06 0.04 0.07 0.04
0.06 0.00 0.00 0.08 0.04
0.04 0.13 0.06 0.05 0.03
0.00 0.00 0.00 0.00 0.00
SCHULZE & WHITACRE: TREE COMMUNITY IN TIKAL NATIONAL PARK
Hillbase
avg.
0.00 (
0.00 (
0.00 (
0.00 (
0.00 (
0.00 (
0.09 t
0.13 (
0.09 (
0.00 6
0.32 6
0.33 6
0.27 6
0.07 6
1.91 6
0.93 6
4.05 6
3.13 6
0.00 6
0.00 6
0.00 6
0.00 6
0.05 6
0.00 0
0.00 6
0.00 0
0.18 6
0.20 6
0.00 0
0.00 0
0.00 0
0.00 0
0.00 0
0.07 0
0.05 0
0.00 0
Sabal
avg..
0.18
0.13
0.00
0.00
0.15
0.00
0.32
0.07
0.00
0.00
0.12
0.07
0.03
0.20
2.62
1.00
2.15
1.60
0.03
0.00
0.09
0.00
0.00
0.00
0.00
0.00
0.12
0.00
0.00
0.00
0.03
0.00
0.09
0.00
0.09
0.00
Transitional
avg. s.e.
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.83 0.24
1.00 0.52
0.00 0.00
0.00 0.00
0.43 0.21
0.10 0.09
0.13 0.09
0.30 0.14
2.00 0.56
1.10 0.72
3.65 0.68
1.90 0.44
0.00 0.00
0.00 0.00
0.00 0.00
0.20 0.12
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.04 0.04
0.20 0.12
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.04 0.04
0.10 0.09
0.35 0.12
0.00 0.00
Tall Scrub
Swamp
avg. s.i
0.00 0.1
0.00 0.1
0.00 0.(
0.00 0.(
0.00 0.(
0.00 0.(
1.63 0.,
1.42 0.4
0.00 0.(
0.00 0.(
0.19 O.J
0.00 0.(
0.00 0.(
0.08 0.1
0.00 0.(
1.08 0.
1.00 0.4
1.67 0.4
0.00 0.6
0.00 0.(
0.00 0.6
0.00 0.(
0.00 0.6
0.00 0.(
0.06 0.6
0.17 0.1
0.00 0.6
0.00 0.6
0.00 0.6
0.00 0.6
0.00 0.6
0.00 0.6
0.06 0.6
0.08 0.6
0.31 0.1
0.00 0.6
Low Scrub
Swamp
avg.. s.e
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.06 0.0
0.07 0.0
1.59 0.3
1.50 0.2
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.00 0.0
0.14 0.1
0.00 0.0
0.00 0.0
0.00 0.0
0.36 0.1
0.00 0.0
0.00 0.0
0.00 0.0O
0.00 0.0
0.00 0.0O
0.00 0.01
0.00 0.0O
0.00 0.0
0.06 0.01
0.00 0.O0
Mesic bajo/
Swamp
avg. s.e.
0.05 0.04
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.36 0.14
0.60 0.29
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.23 0.16
0.50 0.38
0.32 0.19
0.90 0.33
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.10 0.09
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.05 0.04
0.00 0.00
0.00 0.00
0.00 0.00
247
248 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 41(3)
Figure 1. An illustration of the predominant influence of topographic position on forest structure at Tikal, Guatemala; tall, upland forest on a slight
ridge extending into a low-lying basin dominated by Low Scrub Swamp Forest
Mean soil depth across topographic gradient
)-
II I
Soil clay content across topographic gradient
- : 4 X: *
Topographic Position/Forest Type
2 Soil rock content across topographic gradient
S 4 -
0 3
2-
I'--
O- *
Mean degree slope by forest type
10 -
!
S6-
S4 -
2 2-
I I I I I I I
Topographic position/Forest Type
Figure 2. Variation in physical characteristics across the dominant topographic gradient, Tikal, Guatemala. Values are presented as means per forest
type (= topographic position), 1 standard error.
SCHULZE & WHITACRE: TREE COMMUNITY IN TIKAL NATIONAL PARK
1.0
0.5 Soil pH
/ Soil Organic Matter
Topographic Position
Co 0.0 -
Clay Content
Soil Rock Content
-0.5 Slope
-1.0 I I
-1.0 -0.5 0.0 0.5 1.0
PCA axis 1
Figure 3. Ordination of six environmental characteristics along first two axes of a Principal
Components Analysis; data from vegetation sample plots, Tikal, Guatemala. The first axis (eigenvalue
= 3.62) explained 60.3 % of the variance and the second axis (eigenvlaue = 0.84) explained 14.0 % of
the variance. Topographic Position takes values from 1 (hilltop) to 18 (low-lying areas with standing
water in dry season); Soil Rock Content, based on visual assessment of soil samples, ranges from 0-10;
Slope is measured substrate angle; remaining variables are from laboratory soil analyses.
tThis portion of the gradient
shown in figure 6
*
* * #* *
.* t.
2
1
4
2
DCA axis 1
Forest type
o upland dry
+ upland standard
o upland wet
hillbase
sabal
transitional
tall scrub swamp
low scrub swamp
mesic bajo/swamp
0
-1 -
0
1 I
*
Figure 4. Position of sample plots at Tikal, Guatemala along first two axes of a Detrended Correspondence Analysis (283 active samples, 98 active
species) of mature tree (> 7.5 cm d.b.h.) data. Samples are coded according to forest classes identified mainly on the basis of topographic and edaphic
conditions; forest codes are as in Table 1. Not all samples are visible in plot due to high degree of overlap in ordination space. Upland sites on the
left of the plot are detailed in Fig. 6.
Ordination Axes
Ordination diagnostics Total
1 2 3 4 inertia
Eigenvalues 0.756 0.645 0.382 0.258 7.483 '-
Cumulative % of variance in
species data accounted for 10.1 18.7 23.8 27.3
z
z
Pt
A Euge 'sp
: outside range of fig.
A spp outside graph range (dist #9) -
Ingaedul, Loncguat, Cassgran,
Calypchyt, "Tabe 'sp2", "Troprace2",
"Sponmomb2"
Pachaqua
Distribution code
. Pithgiga
.Tabesp
Lonccast
Sapiniti Calobras (dist 10)
Casecory
o +o Allocomi
I" o- Margnobi
Vitegaurn Amyrelem
Mataop o. Termamaz Swiemacr
Lonclati BurssimaCoccacap .Buciburs
Coccacap Taliflor uManizapo Guetcombo
Gymnlucio o
This portion of gradient
shown in figure 7
unlabelled dist 7 spp =
Phylnobi, Hamptril,
Opplv1, Ardidens,
Metobrow, Glirsepi
o Haemcamp
Coccrefl
o Erytguat
Crotrefi Cocccozu Byrsbuci Camelati
DCA axis 1
Figure 5. Positions of tree species centroids along first two axes of a Detrended Correspondence Analysis (283 active samples, 98 active species) of
mature tree (> 7.5 cm d.b.h.) data. Species are coded by distribution type as identified through examination of individual species distributions along
a topographic/edaphic gradient (refer to topographic distribution groups in Appendix 1 and section on species distribution patterns). Not all species
are labelled due to high degree of overlap in ordination space. Full species names corresponding to abbreviations are given in Appendix 1. Upland
sites on the left of the plot are detailed in Fig. 7.
0 -
-1 -
0.10 -
0.02
-0.44 -0.36 -0.28 -0.20
DCA axis 1
Figure 6. Positions of upland samples along first two axes of a Detrended Correspondence Analysis of mature tree (> 7.5 cm d.b.h.) data. Plots are
coded according to forest type as in Fig. 4. Not all samples are visible in figure due to overlap in ordination space.
Forest type
o upland dry
+ upland standard
o upland wet
hillbase
sabal
-0.06
-0.14 -
-0.60
-0.52
I I I I .
0.2
o Gymnluci Cecrpelt
+ Quarfune
0.1- Thoupauc Cymbpend, Casebart (dist 1) Distribution code
Dryplate oliv o Annonas Dioscamp Sabamaur 0 1
T O. Laettham ,' +Guarglab + 2
unlabelled x Sonmomb 0 Sebalong
W ap s t4= ^ Q Sponmomb Sebalong 2
< 0.0 rin -4 x + + + Poutdurl Aspimega +Zuelguid x 3
Brosalic -c Poutret" + +
Tricminu -Nectsal Acaccook 4
Malmdepr e .o ,' Vatalund
Poutcamp ..ble Poutamyg (dist2) + Simisalv o 5
dist 3 unlabelled Manizapo (dist 10) a -iisv 0
Mastfoet spp: dist 2 = Aspicruea p
-0.1 Wimmconc Stemdonn imedioc (dist 3Casselip
Blompris Cordgera Pimedioc (dist 3),
Cedrmexi Pseuoxyp
dist 2 = Zantproc
Hirtamer a unlabelled dist 2 spp = Astrgrav, Troprace,
dist 4 = Protcopa, Cupabeli, Ampehott, Cryostau
-0.2 -Gausmaya
-0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0
DCA axis 1
Figure 7. Species centroids for tree species with abundance peaks in upland areas; ordination space defined by first two axes of a Detrended
Correspondence Analysis of mature tree (> 7.5 cm d.b.h.) data. Species centroids are coded by distribution type as in Fig. 5 and Appendix 1; species O
names corresponding to abbreviations are given in Appendix 1. .
SCHULZE & WHITACRE: TREE COMMUNITY IN TIKAL NATIONAL PARK 257
unlabelled spp
unlabelled sp dist 7=
dist 9 = Metobrow o Gausmaya
Margnobi ,' Camelati
Casecory Haemcamp o Annonasp
Pithbeli Erytguat Distribution code
Calychyt Glirsepi Casebart
Hamptril o o Laettham
Cassgian, Ingaedul Arddens Pseuoxy ymbpend Gymnluci +
Poachaqua Coccrefl Dendarbo x Mastfoet + 2
Loncgujit Alocomi Hirtamero ROCKS
_ Cocccozu 0o x 3
oArdiden Crotrefl Poutamyg Exotpani_. Thoupauc
Sn o Blompris
Buciburs Phylnobi x Cecr elt x Manizapo 4
Crotpyr Zuelguid o Sebslong. Poutcamp
S Guetcomalobras Aspicr Stemdonr Taliv 5
8o Astrgrav + SLOPE 'Zantproc Wimmconc Nectcori
Sapiniti- -mlu- P Burssima Quarfune o 6
-. --. -- ymn c. -_ - - .. - - -- ... ... .. Tri-cr nin- u
-,TOPO.-o TermamazLo g a Brosalic o 7
T afr" L.reCrdgea ^ Malmde- r .Dryplate
unlabelled .T fo L Guardlab-.-dg oa Cedrmexi *M .prCelttrin 8
spp dist 5= Tabe sp atalund outdurl, Sponm.om
Dlae 2 2L ncati + Sabamaur unlabelled near origin: dist 2= Protcopa, < 9
LAelhm m + Sirhisalv Acaccook dist 1= Aspimeg, Nectsali dist 3=
dist 10= CLAY + Cassgula Poutreti, Acacdoli dist 5= Euge 'sp
Manizap Troprace Ampehott
ansap Troprace+ m - - spp outside range of fig = Cryostau, Dioscamp,
Coccacap I + s- i-i Orbvcohu (all dist 2) =
Mataoppo -1 0 1 2
CCA axis 1
Figure 8. Positions of tree species centroids along first two axes of a Canonical Correspondence Analysis (283 active samples, 98 active species) of
mature tree (> 7.5 cm d.b.h.) data, incorporating four environmental variables: topographic position, slope, soil clay content and soil rockiness. Lines
indicate relation of environmental variables with first two canonical axes. Species are coded according to distribution classes identified through
examination of individual species distributions along the predominant topographic gradient. Distribution codes and full species names are given in
Appendix 1. Complete lists of species corresponding to each distribution code are presented in Figs. 26-35.
Ordination diagnostics
Eigenvalues
Species-environment correlations
Cum. % variance of species data accounted for
Cumulative % of spp-environment relation
Sum of all unconstrained eigenvalues (after fitting covariables)
Sum of all canonical eigenvalues (after fitting covariables)
1
0.425
0.827
5.7
81.0
Ordination Axes
2 3
0.068 0.020
0.533 0.372
6.6 6.9
93.9 97.8
Total
4 inertia
0.012 7.483
0.314
7.0
100.0
7.483
0.525
Monte Carlo test of significance of environmental variables (99 permutations under full model):
(1) topographic position: F = 16.79; p = 0.01; (2) soil clay content: F = 2.3; p = 0.01; soil rock content and slope n.s. after fitting of topographic
position and clay content.
Mico'sp' + Glirsepi
Cost'sp'
outside range of fig.
Cymbpend Acac'sp'
Tabe sp Laettham, Sebalong, Gymnluci,
Nectsali Calobras Dryplate, Forctrif, Coccacap,
+ Ouraluce, Euge'sp
Aegimons GAP AREA 4Exotpani
Anon'sp'+ +Pipe'ad'
Allocomt \ Gausmaya
Poutcamp Rehdpinh Astrgra
Poutdurl Pithbeli. r"
Protcopa AN\ O n+Hamptril
Wimmconc CANOPY
Simisalv + + + OPENING +Phylnobi
Bactmajo ++ + Guetcomb Crotpyra
Manizapo Swiemacr + Jacq'sp' Pisiplsi
+ +, ++ Mataoppo Xyl Rapaguia
Ampehott Poutreti Amyrelem Coccrefl +Haemcamp
Casecory + + + *+Plumobtu
Pipepsil+ + + rsbuciMetobrow
Orbicohu + Tricminu Clus sp Ca
+ eCryostau outside
Hirtamer
-1 0 1 2
CCA axis 1
Ardidens
melati
e range of fig.
(-2
Figure 9. Position of species centroids along first two axes of a Canonical Correspondence Analysis of data on tree juveniles and understory treelets -
(< 7.5 cm d.b.h.; 196 active samples, 126 active species) with percent canopy opening and percentage of plot area in gap or building phase included
as environmental variables. Effects of topography, slope, and soil clay and rock content were removed by incorporating these as covariables in the
analysis. Lines indicate relation of environmental variables with first two canonical axes. Unlabelled species at center of plot showed low correlations
between juvenile occurrence and both light availability and canopy disturbance; unlabelled species are as follows (see Appendix 1 for full names):
Sabamaur, Poutamyg, Ardipasc, Manlmdepr, Zantproc, Cupabeli, Casebart, Aspicrue, Aspimega, Stemdonn, Simaglau, Blompris, Talioliv, Qaurfune,
Swarcube, Bucibuce, Termamaz, Dioscamp, Erytguat, Lonccast, Lonclati, Alibedul, Pseuoxyp, Pimedioi, Calychyt, Acaccook, Tricpall, Tricmosc,
Licapeck, Vatalund, Troprace, Brosalic, Zuelguid, Sponmomb.
Ordination Axes Total
Ordination diagnostics 1 2 3 4 inertia
Eigenvalues 0.238 0.040 0.260 0.213 4.750
Species-environment correlations 0.776 0.511 0.000 0.000
Cum. % variance in species data accounted for 5.6 6.5 12.6 17.6
Cum. % of spp-environment relation 85.6 100.0 0.0 0.0
Sum of all unconstrained eigenvalues (after fitting covariables) 4.271
Sum of all canonical eigenvalues (after fitting covariables) 0.278 z
Monte Carlo test summaries (99 permutations under full model):
Test of significance of first canonical axis: eigenvalue = 0.24; F = 11.13; p = 0 .01. Z
Overall test: Trace = 0.28; F = 6.57, p = 0.01.
Test of significance of environmental variables: (1) canopy opening: F = 10.88; p = 0.01; (2) gap area: F = 2.19; p = 0.03.
ATermamaz
outside range of fig.
+ Zantproc
Pseuoxyp
Protcopa
Ouraluce
Sabamaur
Alseyuca
- Gymnlu
+ Coccacap
DISTURBANCE AGE
Licapeck R
+Ampehott + Rehd:
S / DISTURBANCE
ectsali ./ INTENSITY + Cymbpend
,+ isalv
Thoupau Loncguat + Amyrelmi
'Exotpa Stemdonn n Guetcomb
.+ / ~ + Cedrmexi
Sponmomb + Dendarbo
-- Cecrpelt
+ + + ++ +Sebalong L
++ ++ + Nectcori +Zuelguid
+ + + + Simaglau
Cordgera
Guargla, + Vitegaum + Burssima
c~j
co0
enn
onclati
+ Gausmaya
Mataoppo, Poutcamp, Swiemacr, outside range ofs
Dryplate, Acaccook, Hirtamer,
Malmdepr, Pimedioc, Cupabeli, -----
Celttrin, Tricmosc
S I I
CCA axis 1
+ Troprace +Calobras
+ Quarfune
Figure 10. Positions of tree species centroids along first two axes of a Canonical Correspondence Analysis of mature tree (> 7.5 cm d.b.h.) data from
upland sites (149 active samples, 62 active species) with natural treefall canopy disturbance intensity and disturbance age as environmental variables.
Effects of topography, soil clay and rock content and substrate slope were removed by incorporating these as covariables in the analysis. Lines indicate
relation of environmental variables with first two canonical axes. Unlabelled species at center of plot showed low correlation with canopy disturbance,
and are as follows (see Appendix 1 for full names): Poutreti, Poutamyg, Pouldurl, Manizapo, Mastfoet, Astrgrav, Casebart, Aspicrue, Aspimega,
Swarcube, Wimmconc, Laettham, Lonccast, Vatalund, Tricminu, Acacdoli, Brosalic, Talioliv, Blompris, Vitegaum, Cryostau.
Ordination diagnostics
Eigenvalues
Species-environment correlations
Cum. % variance of species data accounted for
Cum. % of spp-environment relation
Sum of all unconstrained eigenvalues (after fitting covariables)
Sum of all canonical eigenvalues (after fitting covariables)
Ordination Axes
1 2 3
0.041 0.025 0.308
0.593 0.603 0.000
1.6 2.6 14.5
62.0 100.0 0.0
Monte Carlo test summaries (99 permutations under full model):
Test of significance of first canonical axis: eigenvalue = 0.04; F = 1.71; p = 0.06;
Overall test: Trace = 0.07; F = 1.39; p = 0.05;
Test of significance of environmental variables: (1) disturbance age: F = 1.46; p = 0.10;
(2) disturbance intensity: F = 1.32; p = 0.07.
4
0.189
0.000
21.8
0.0
Total
inertia
2.839
2.578
0.066
Figure 11. Graphic representation of topographic gradients in Tikal National Park, Guatemala, with associated forest types: a) complete gradient from
a dry upland hilltop to the center of a lowland depression; b) a narrow lowland depression between two upland areas; c) illustration of the dramatic
influence exerted on vegetation by subtle topographic changes within lowland depressions.
SCHULZE & WHITACRE: TREE COMMUNITY IN TIKAL NATIONAL PARK 265
Mean canopy height by forest
22
201- I
18-
-.1
30 -- Mean canopy opening by forest
25 -
20-
3 I I 3
I I
I I I I I i 1
type
I I
ID
I~I
I Il
st type
cz
I I
II
I -------- i I
Frequency of trees leaning >25 degrees off vertical
6 -
5-
4 -
3--
0 FI
16000 -
14000 -
12000 -
10000 -
8000 -
6000 -
4000 -
2000
0
1200
1000
800
2 600
400
200
0
Average basal area of stems >7.5cm
IlIlllllli
Mean stem density per hectare by forest type
_ stems >7.5cm dbh [
\zzza stems>10cm dbh n ii
l
& l l . . .
I I I I D w ,
I I
|
