Research Questions:
How does Tillandsia get its water?
What are the pH conditions of the tank water?
How are the seeds dispersed?
How do population densities of Tillandsia compare in primary and secondary forests.

Hypotheses:
1. Water is absorbed through the roots of Tillandsia, an epiphytic Bromeliad.
2  The pH is uniform in the tanks of a Tillandsia
3. The pH of water in tank of a Tillandsia is more acidic than rain water.
4. The distribution of Tillandsia is different in primary and secondary forests.
5. There are no algae in the Tillandsia tanks.
6. The seeds are dispersed by animals.

Procedure:
Hypothesis 1

Part 1
Collect two Tillandsia that are of similar size and habitat. Prepare a dye solution by placing 10 drops of red food color in 300 ml of water. Put one plant in a beaker with only the roots placed in the solution.  Cut the roots off a second plant and coat the cut surface with glycerin and wrap it with parafilm. Fill the tanks of the second plant with the dye solution. Repeat this procedure with two smaller plants. Leave the plants in the solution for 24 hours.  Make cross-sectional and longitudinal slides of the leaves and roots of each plant after 24 hours.  Examine the tissue to see if the dye has been absorbed into the leaf and the root of each specimen.

Part 2
Place two drops of dye in a medium-sized Tillandsia in situ. Monitor the movement of the dye over 30 hours.

Hypothesis 2
Label three similar-sized Tillandsia on the same tree. Remove five mlof water from the center tank and from the tank in row three. Measure the  pH using a pH meter.

Hypothesis 3
Compare the pH of the water in the tanks to that of rainwater.

Hypothesis 4
Observe the distribution of Tillandsia in primary and secondary forests.
Walk into the primary forest and observe two different types of trees that have Tillandsia. Record the size, type, and locations of Tillandsia on each.  Go into the
secondary forest and repeat the procedure.

Hypothesis 5
Insert a pipette into the center tank of a Tillandsia.  Remove 5 ml of liquid from 4 separate plants.  Observe and identify the organisms found in the tanks.

Hypothesis 6
Examine the fruit and seeds of the Tillandsia to see if they are designed for animal or wind pollination.

Data and Analysis

Experiment 1: Water is absorbed through the roots.

In microscopic analysis of both  cross sections and surface views of the leaf, the red dye had  moved into the parenchyma cells of the experimental leaves.  This was not true of  the plant cells that  had their roots submerged in the dye.   The leaf surfaces were covered with minature raised structures which had long hairs or trichomes surroundintg their edges.  There were four pie-shaped cells were at the center of the structure.

Discussion
From our experiment it appears that the lack of movement of dye from the roots to the leaves means  that the roots are not responsible for the uptake of water.   The data reject the hypothesis that roots function in the absorption of water.  Tank bromeliad leaves are covered with trichomes  that act as wicks for the absorption of water  (Tomlinson, 1969).  The leaf rather than the root absorbs water.

Experiment 2: The pH is uniform in all tanks.

Table 1

Sample	pH in center	pH in row 3 tank
1	4.2	5.7
2	4.6	4.0
3	3.8	4.5
4	3.8	4.5
5	5.7	5.4

The data  reject the  hypothesis that the pH is the same in all tanks (Table 1).  We believed that if there were a difference  in pH it would be due to the accumlation of detritus in the older tanks. We expected the pH to decrease as decay occurred. The data indicate that the pH is actually higher in the third row.

Experiment  3:  pH in the tanks is more acidic than rain water.

Table 2

Samples	pH
1	4.2
2	5.7
3	3.8
4	4.5
Average	4.6
Rain	5.4

Discussion
The pH of the water in the tanks was different from that of rainwater (Table 2). The data support the hypothesis that the liquid in the plant is more acidic than rain water. We believe that this is caused by the detritus in the tanks.

Experiment 4: Distribution of Tillandsia is different in primary and secondary forests.

Data
We walked into the primary forest to CEN 300 and located one large and one medium-sized tree that contained Tillandsia.  We described the tree and attempted to count the Tillandsia.  The first story of the forest contained palms approximately 25’ high and there were no bromeliads on our selected trees up to that height.  When trees reached the canopy and branched out, there were some bromeliads, but the density of ephiphytes was still low.  We chose two East-facing trees to examine, one tall, slender tree with a diameter  of .9 meters with no branches until it reached the canopy, and the other an older, buttressed tree with a diameter of  3.4 meters.  There were no bromeliads on the trunk of the smaller tree, and only a few at the canopy level.  The larger tree hosted three large bromeliads (not the variety that we studied) at the level of the palms (25’), and no others until the canopy where the tree branched out.  We counted about 20 bromeliads at that level, but it was difficult to determine which tree in the canopy the bromeliads were actually attached to.
Using the same method, we then moved to the secondary forest at the junction of the STR and CEN trails and located two similar trees with the same eastern orientation .  We examined the pattern of the bark and the leaves to determine that we were using the same types of trees.   We saw a variety of bromeliads of different sizes spaced about a meter apart on the horizontal branches beginning at about 30’ high on both trees.

Discussion
Our observations indicate a need for  improved methods of testing for this hypothesis. It is difficult to count individual plants without climbing into the canopy.  For further study we would make a clinostat to measure height and bring a tape to measure distance from the tree.

Experiment 5

Data
The micro-organisms found in the tanks were  rotifers, copepods, green algae, diatoms, and lots of organic particulate matter.  There appeared to be more organic particulate matter in the bromeliads collected in the forest than those collected  on the lawn.  None of the micro-organisms were counted, but merely surveyed.

Discussion
The micro-organisms were typical of those found in a lake.  The data reject the  hypothesis that there is not algae in the tanks.  Algae is present in the tanks of the bromeliads. Evidently enough light enters the tanks to support aquatic photosynthesis. However, since particulate organic matter was so abundant, it appears to be a detritus-based system rather than phytoplankton-based system.  In order to  make a conclusion about the actual composition of the aquatic community, a more quantitive study needs to be conducted.

Experiment 6

Data
When the fruit of the  Tillandsia was opened it contained tiny seeds that had cotton-like plumes attached to them.

Discussion
The size and structure of the seeds makes them appear to be wind dispersed.  The data rejects the hypothesis that Tillandsia are animal dispersed.

Conclusion
Tillandsia gets its water by absorbing it through trichomes on the leaves. In our samples the pH range varied  from 3.8 to 5.7 and was generally higher in the older tanks.  The pH of the water in the tanks was more acidic than rainwater.  Tillandsia appear to be wind dispersed. We were unable to draw conclusions regarding the population densities in the primary and secondary forests.  This pilot study created more questions than answers.  For future study we would like to characterize the food webs in the tanks and develop a better strategy for studying population densities.

Literature Cited

Cardelus, Catherine, 1999. Personal communication.  La Selva, Costa Rica.

Tomlinson, P.B., 1969. Anatomy of Monocotyledons.   Oxford Press