Please
Click here to return to main page
Comparative
Distribution of the Naked Indian Tree, Bursera
simaruba,
in Primary Highland, Primary Lowland, and
Secondary Lowland Dry Forest
at Palo Verde National Park, Costa Rica
By Dean A. Birdsall, Wayne
Cokeley, Catherine Roberts, Christine
A. Royce and Linda Strauss
Woodrow Wilson Foundation
Fellows
Introduction
The Naked Indian or Gumbo Limbo Tree, Bursera simaruba, is a well known and
widely distributed tree of the order Burseraceae which is widespread in the
Neotropics. The tree is one of several trees with photosynthetic bark that is
found in the dry forests of Palo Verde.
B. simaruba is used medicinally by the people of the region for
treatment of stomach ailments. Bursera
is commonly used for “living fenceposts” in Costa Rica and the tree is
apparently a patch colonizing species.
Development of the tropical dry
forest in Costa Rica has typically replaced the Bursera tree’s habitat with open fields for cattle grazing. In the Palo Verde area, the tropical dry
forest is restricted to rugged terrain which is unsuitable for ranching, such
as ridges, steep slopes, and cliffs. These ridges provide the last undisturbed
habitat for the species. In this investigation we will examine the distribution
of the species in long term primary forest on both elevated and lowland
terrains and in secondary growth forest which was once cattle pasture land.
Methods
Populations
of Bursera were sampled in three
separate areas. The plot referred to as the highland primary forest is located
on the ridge immediately behind the OTS field station at Palo Verde. This area extends from the rear of the station
grounds up to the top of the rocky outcrop at the top of the ridge along the
Mirador Guayacan Trail. This trail starts at approximately 12 meters elevation
and extends uphill to over 117 meters in elevation. The plot called secondary
lowland forest runs along the Sendero La Martilla Trail which is parallel to
and adjacent to the road from the OTS station to the picnic area at Palo Verde.
This land was once part of cattle pastures. The plot called primary lowland
forest is approximately 5 kilometers west of the OTS station and was accessible by truck. This region has not
been subjected to cattle grazing in historical times according to the OTS
staff.
Sampling methods included
measurement of all trees visible within eyeshot of the walking trails passing
through the properties. Only trees with
the characteristic orange and green papery bark were counted. Seedlings and
younger trees with diameters too small to be easily recognized were not counted
in this study. Location of samples
within the plot was determined by pacing off distances along the trail equal to
twenty Dean Steps which is equivalent to thirteen meters. A Dean Step is defined for this study as a
standard size pace taken by research team member, Dean Birdsall. All distances at all sample sites were paced
off by Dean to maintain consistency of distance. These paced distances were
reconciled as best possible with maps of the area to obtain elevation
information. Each of the three sites sampled had sample paths of the same paced
length and relative width.
All trees seen were measured for
diameter at breast height (DBH) in centimeters and data was recorded in the
field by paced position and DBH.
Data
Field measurements for the three survey sites can be
found in the Appendix of this report
Discussion
The Naked Indian Tree appears to be widely
distributed throughout the forests of Palo Verde. We observed the tree in lowland areas with heavy brush growth, as
well as growing out of the limestone rock at the tops of the highest ridges.
The transect taken through the
highland primary forest shows the numbers and size distribution up the ridge in
the figure below. Zero elevation is the
beginning of the transect (OTS station).
As the plot shows, the
majority of the trees measured had DBH measurements of less than 30 cm. There
were a limited number of larger diameters (DBH measurements) in the sample,
most of these being found at about 320-360 Dean Steps elevation on the
trail. It is possible that these trees
represent the longest lived specimens of the population which are growing in
the optimal area for the species on the ridge. The slightly smaller individuals
growing at higher elevations on the ridge might represent individuals of
similar age but growing in less desirable conditions and therefore with poorer
DBH growth. The smaller individuals at lower elevations on the slope probably
represent recovery trees which colonized the slopes since the time of the
formation of Palo Verde National Park in 1979.
It is likely that the lower altitude plants are under more grazing
pressure from the population of lowland grazing mammals such as White Tailed
Deer, Cottontail Rabbits, Ctenosaurs, and the occasional errant cow from the
old ranching days.
The greatest numbers of individuals
in this section occurred at 250 Dean Steps elevation. The figure below shows
the relative numerical distribution of the species through the transect. Note
that the largest numbers of trees are found immediately downslope from the
largest individuals. It is likely that these trees are the offspring of the
larger tree uphill from them and somehow colonized a local gap at some time in
the past, hence the clumped distribution.
Size distribution of trees
in the Secondary Lowland area appeared to be somewhat random. The figure below
shows a plot of these individuals.
Note the general absence of
trees from the first 200 Dean Steps of the transect. This represents the
movement of the sampling team in from a field edge and most likely the absence
of trees is indicative of one of three possibilities. The first possibility is
the loss of seedlings in this high traffic edge area to various herbivores. The
edge environment sampled is home to White Tailed Deer, Ctenosaurs, numerous
rodents and was the edge of the cattle pasture up until the formation of the
National Park in 1979 . Seeds and seedlings would be far less likely to survive
in this environment than on the less traveled slopes at higher elevations.
The second
possibility is a lack of suitable habitat for seed germination. While the adult
trees appear to do well in the bright dry edges of cow pastures it is not known
whether the conditions for germination can be found in this habitat. The
exposed, hot and relatively arid conditions of the open field may be
detrimental to Bursera germination.
The third possibility is that the seeds of Bursera simply do not disperse into the
open fields below. The tree produces drupes, which are fed upon and dispersed
by forest birds. These avian species tend not to venture into the exposed
grazing lands below and therefore the seeds would not be expected to travel
with them past the shady confines of the forest.
The following figure confirms this interior centered
distribution of Bursera on the secondary growth lands. Note that the greatest number of trees were located between 250
and 450 Dean Steps into the plot.
In the lowland primary
forest sampled, the trees appeared to be more or less randomly distributed as
seen in the diagram below. There does not appear to be any set pattern of
distribution. Most of the trees are in the 10-20 cm DBH size range and they
seem to occur as clumps. It is quite likely that these groupings represent
fillings of treefall gaps. The largest trees of the study were seen in this
plot and these are most likely the oldest trees of the area, an indication of
undisturbed primary forest.
A comparison of population composition by DBH for
the highland primary, lowland secondary, and lowland primary forests is seen in
the next figure. All three study areas show populations with greater numbers of
smaller trees than larger, older specimens. The relative absence of seedlings
is no doubt due to the bias of sampling technique and does not indicate a real
absence from the environment.
As expected, the primary habitat plots were the ones
with the oldest and largest trees. The secondary area is apparently showing
colonization of Bursera and would be
expected to eventually come to resemble the primary plot curves.
Bursera
appears to eagerly colonize all the environments examined but its proliferation
is probably limited by herbivore predation on the seedlings. The species may
have a competitive edge over its neighbors by being able to continue
photosynthesis during the leafless season through use of its photosynthetic bark
(Osterling and Barnhorst, 1995). Hence the tree appears in all of the habitats
examined except for the most exposed, grazed areas.
Conclusions
Bursera simaruba is a widely distributed,
colonizing species which is well known throughout the Palo Verde Dry
Forest. It occurs at all elevations and
in both primary and secondary forest plots.
It occurs in a diversity of soil conditions including brushy edge,
forest floor and bare rock conditions. The species only seems limited in area
with high exposure to herbivory and excessively bright conditions.
Examination of the plant communities
with in tropical deciduous forest is useful in predicting the changes that may
occur with global environmental changes.
The successional forest as compared to undisturbed forest allows us to
see the possible changes of B. simaruba as
land use changes are made. The highland
forest as compared to the lowland forest indicates the changes that may come
about if there is a climatic shift along an altitudinal gradient.
Classroom
Applications
This study has shown that teachers can collect a
tremendous amount of data, analyze it, and observe trends without sophisticated
quadrant measurements. The use of park
trails as sampling transects can serve as an easy and useful means of data
collection. Local successional plots
can be used to compare species’ requirements within community areas over a
period of time.
The ad hoc field survey using “Dean Steps” and
simple DBH measuring techniques could easily be used in an Environmental
Science curriculum to survey woodlots anywhere in the world. The comparison of
forest distribution to the occurrences in disturbed areas demonstrates the
relative change incurred in secondary succession and the loss of diversity thereof. This simple method of a field survey could
be utilized with students in one wood lot where different teams of students
examine different species in the same areas thus providing a more in-depth look
at the biodiversity of an area and the results of succession on those areas.
References
Janzen D. H. (1983). Costa Rican history. Chicago, IL: University of Chicago Press.
Osterling, W. L., & Barnhorst,
A. V. (1995). Distribution of Bursera simaruba over three habitats
of varying moisture availability. Dartmouth studies in tropical ecology.
October, 25-26.
Quesada, F. J. (1997). Arboles de la peninsula de Osa. Costa Rica:
INBIO.
Please Click here to return to main page