| | ||||
| | | | | |
| BACK TO PROJECT LIST | |||||
Classroom
Applications To Data Page
To 50 Questions Page
|
Barbara Dorritie,
Cambridge Rindge & Latin, Cambridge MA Nancy Freese, Noble HS, Berwick ME Mary Frieze, Arrowhead
School, Livingston MT Robert Kuhn, Centennial HS, Roswell GA |
 |
AbstractThe rainforest on La Selva contains a wide diversity of epiphytes, including bromeliads. These plants provide a wet place for many canopy dwellers including moisture sensitive amphibians. The rainforest at La Selva is fragmented into many areas including primary growth forest, secondary growth forest, landscaped, and converted pasture to name a few. The human impact of fragmentation may influence the role of bromeliads in the ecosystem in terms of decreased water absorption and altered water quality within the bromeliad microhabitat. By examining bromeliads from primary growth forest, secondary growth forest, landscaped forest, and converted pasture, the quality of microhabitat according to degree of human impact may be measured.
Bromeliads are a diverse group of epiphytes that occur throughout neotropical
forests (Kricher, 1997). Bromeliads account for over 70% of epiphytes at La
Selva, and flourish as a product of a limited dry season and high rainfall (40
cm/yr) (McDade et al., 1994). Bromeliad structure consists of a rosette
of leaves with a center resembling a cistern or tank, which often collects water
and detritus during rain (Kricher, 1997). The collected water creates a microhabitat for tree
frogs, mosquitoes, worms, snails, and salamanders (Zahl, 1975; Wilson, 1997-in
McDade, 1994). Arguably the most sensitive group that uses the bromeliad aquatic microhabitat is the amphibians. Amphibians may
be undergoing a widespread worldwide decline in population (McDade et al.,
1994). Perhaps the greatest detriments to amphibian reproduction are predation
and moisture (Janzen, 1983). Bromeliads provide both a refuge from predators and
a source of moisture. At La Selva, two frog species use bromeliads as a
deposition place for their terrestrial eggs. Newly hatched tadpoles are
transported by Dendrobates pumilio and
Pyllobates
lugubris to the bromeliad microhabitat where they develop (McDade et al.,
1994). Fragmentation of the forests at La Selva creates variable abiotic
conditions, which may affect the aquatic microhabitat contained within the
bromeliad. Primary forest bromeliads are usually found
high in the canopy, but are sheltered from drying and some direct
sunlight by the forest canopy. The secondary forest canopy is not as developed and younger trees are
susceptible to windfalls. These windfalls alter bromeliad habitats by bringing
them closer to the ground and placing them in direct sunlight
created
by light gaps. Converted pastureland and landscaped forests like the
Holdridge Arboretum contain large light gaps 
where
bromeliads reside closer to the ground and live in greater direct sunlight.
Therefore, areas of human impact may create changes in the abiotic aquatic
microhabitat contained in the bromeliads, thus affecting the potential for
sensitive groups like amphibians to breed.
The hypothesis is that bromeliads in more disturbed habitats will be more
hot and dry than those in primary forest, and will show greater evaporation
rates. By measuring temperature, pH
and evaporation rate in the different forest habitats described above, the
affect of habitat disturbance on the bromeliad aquatic micohabitat may been
seen.
We selected sites representing a range of levels of disturbance of bromeliad habitat. The most highly disturbed was along the driveway into La Selva in an area with cleared trees, mowed grass, and no understory layer. Other study site areas included the Holdridge Arboretum, secondary growth forest along the Sendero Tres Rios (STR) path, and finally primary growth forest along the Sendero Sura trail (SUR). The Holdridge Arboretum is also an area of cleared trees with little understory. The STR site was on a tree fall in an area of old plantation growth. The primary growth forest study area was also near a tree fall site. At each site, whenever possible, we identified a minimum of six bromeliads using the criteria described below.
Sample Selection
The bromeliads identified for testing had between 8 and 14
leaves in the rosette and a center tank large enough
to insert the CBL probes. We
specifically looked for samples that were less than 2 m above the ground level.
If a bromeliad was attached to a fallen branch and was lying upright on
the ground surface it was used as an acceptable sample. If
at least 6 bromeliads were identified at the site we randomly paired them and
then randomly selected one of the pair (coin toss) and discarded the second for
testing. The selected sample was
then tagged with a code number for the site.
We tested 3 bromeliads at each site.
We measured the distance in meters of the sample above ground surface. Calipers were used to measure the diameter of the base of the plant. We collected data on several abiotic factors such as core temperature of the bromeliad and pH of the core of the bromeliad were measured using CBL probes whenever possible. If the sample plant contained water in the tank it was poured out into a sample collection bag and pipetted for a volumetric measurement. Then we added a 5 mL volume of water to the bromeliad to record water loss over time. After approximately a 24-hour period we returned and measured the temperature and pH of the
Bromeliads were tested for in situ temperature, pH, and
water volume loss. Temperature was lowest in the
primary forest, and varied by 7.3 degrees Celsius from the lowest temperature in
the primary forest to the highest in the secondary forest site.
The evaporation rate was highest in the arboretum, where all water
disappeared, and lowest in the primary forest.
The pH ranged from 3.83 to 6 with lowest average pH found in the most
disturbed habitat, and the highest in the secondary forest.
pH in all bromeliads was lower than our control, rainwater, which had a
pH of 6.0.
Discussion
No conclusion can be drawn from our results about differences in bromeliad microclimate in sites with varying levels of disturbance. Bromeliad tank temperature was similar in all sites, although it was slightly lower in the primary forest site. The pH was slightly lower than the control (rainwater, 6.0) in all sites, with the exception of a single measurement of 3.83 in one bromeliad, which we believe, is an outlier resulting from contamination by the acidic pH probe buffer. In many bromeliads all water had evaporated after the 24-hour testing period, so no measure of pH was possible. The amount of water that evaporated from bromeliads was greatest at the arboretum, a relatively protected site, but evaporation rate was lowest in the primary forest, which is what we would have predicted. All of these results indicate that more sampling is needed.
There are
numerous sources of error that may have affected our results.
First of all, we were unable to use all of the same species of bromeliad,
because specimens of the morphotype we selected were not found in all sites. We substituted a structurally similar type, but have no way
of knowing whether these morphospecies offer similar habitat opportunities for
animals. Rainfall was another
uncontrolled factor. The secondary
forest site had to be tested on a different day than the other sites, and while
there was no rain during the first day of testing, there was 0.5 inches of rain
on the second day. The amount of
captured soil contained in each bromeliad was not measured or controlled for,
and certainly could have resulted in absorption of water rather than the
evaporation we were trying to measure. 
Our study also suffered from another kind of design flaw, identified as sin #5 (“Not knowing your species”) in a list of the twenty most common censusing sins in the book Ecological Census Techniques: “Knowing your species is essential for considering biases and understanding the data (Sutherland, 1998).” We discovered in the course of trying to find bromeliads to sample that bromeliads are almost entirely absent from secondary forests except when emergent trees are present. Bromeliads are only found on the ground in either primary or secondary forests when a treefall or broken branch has brought them down from the canopy where they live. This brings up another issue. Since bromeliads are only found at ground level when there is a break of some sort in the canopy, it is unclear how much our sites really differed in “level of disturbance” since they were all somewhat disturbed (as were the investigators themselves). Even fallen bromeliads may provide significant habitat for animals, but it would be interesting to compare these with primary forest habitat provided by canopy-dwelling bromeliads.
In future research on bromeliad microhabitat, evaporation rate could be measured more accurately by placing water in a vial equal to the amount placed in each bromeliad. This would serve as a control for the absorption of water by the captured soil in the bromeliad tank or by the plant itself. Bromeliads could also be tested after removing captured soil from their tanks. In order to better establish the amount of disturbance in each test site, humidity, light intensity and amount of canopy cover lost due to tree fall should be measured. Finally, accurate identification of the bromeliad species and choice of a species that is available to test in all sites would allow for a more fair comparison.
The most valuable results gained from this research project were the insights into experimental design gained through discussions following experimentation. Often students engaged in experiments are not given this same chance to reflect. We have recognized an essential component of inquiry-based learning is allowing students to discuss their results, refine their methodology and collect new data.
core. We then poured the water present in the rosette into a Ziploc collection bag and used a 10mL pipette to measure the amount present after the 24-hour evaporation period.
Tom
Langen, Jim Cole, and John Cozza for help in the topic selection
Melanie
Phelps for expertise in photographic documentation