Introduction

 The majority of plants in tropical ecosystems depend on animal pollinators for survival.  Animals seek out flowers as a food source, ingesting nectar and, in some cases, pollen.  By travelling from plant to plant, animals disperse pollen, making cross-pollination efficient and insuring reproduction of the plants (Kricher 1997).  This type of mutualism with plants is found in birds, bats, and insects.  Among insects, bees, flies, beetles, butterflies, and moths (Prance 1985) accomplish pollination.  Plants benefit by producing flowers that would attract these types of insects because they fly long distances, ensuring effective cross-pollination between widely separated plants (Janzen 1975).  In the tropics there is diversity of flower color, smell, shape, and size, each combination of characteristics being attractive to specific pollinators.   This kind of specialization that results from a high degree of competition between species causes high species diversity (Kricher 1997).  If a plant were visited by a generalist species, one, which visits a variety of species, the plant would have little chance for its pollen to land on another member of its species.  Conversely, if a plant attracts a single pollinator species, the animal will pollinate many members of that plant species.  For example, hummingbird-pollinated flowers tend to have long tubes and are colored red, orange, purple, or yellow.   Bat-pollinated flowers are often white because they are easy to locate in the dark and they have a musty odor (Kricher 1997).   We wanted to find out whether or not specific combinations of flower color and smell attract specific pollinators, supporting the need for biodiversity in tropical ecosystems.  Our hypotheses were:

1. Hummingbirds will prefer red flowers.
2. Butterflies would frequent red and yellow flowers.
3. Other insects would have no color preference.
4. Hummingbirds would be attracted to a sweet odor.
5. Butterflies would be attracted to a sweet odor.
6. Other insects would have no preference for odor.

Methods and Materials
 

Table 1: Matrix for Placement of Samples Blue Honey Red Syrup Clear Meat Yellow Neutral Clear Honey Red Neutral Yellow Honey Red Meat Clear Neutral Yellow Syrup Blue Neutral Blue Syrup Red Honey Blue Meat Clear Syrup Yellow Meat

A tally sheet was prepared for recording the visitors to each site, see appendix 1. Two members of the team recorded observations for thirty minutes. The observers stood outside the matrix and recorded all visitations for ten minutes, then walked to each pole and made closer observations. This procedure was repeated by another team of two for each thirty-minute interval from ten a.m. to five p.m. On day two, individual team members recorded observations from six am to noon following the same procedure.

Description of site

The site selected was in the cemetery garden about five meters from the grave marker. The slope was gradual about 10 degrees from top to bottom. The site was open and exposed to sun.

Results

Table 2: Visitation to Each Flower

	Blue	Red	Yellow	Clear
Honey	3	21	18	7	Honey Total	49
Syrup	7	5	0	9	Syrup Total	21
Meat	4	9	4	20	Meat Total	37
Neutral	3	0	22	1	Neutral Total	6
Color Totals	17	35	24	37

In our two days of observations only one hummingbird and no butterflies visited our site.
The artificial flowers attracted insects which were mostly bees and flies.  Among the honey flowers, red and yellow were the most preferred (Table 2).  Among the four meat flowers more visitations were made to the clear flower.

Table 3: Chi Square

	x^2	P
Chi sq. for smells across blue	0.647059	0.72359
Chi sq. for smells across red	12.88571	0.001592
Chi sq for smells across yellow	46.33333	8.69E-11
Chi sq for smells across clear	20.18919	4.13E-05
Chi sq for smells across all colors	138.4358	8.69E-31
Chi sq for colors across honey	18.18367	0.000403
Chi sq for colors across syrup	19.65306	0.0002
Chi sq for colors across meat	16.87755	0.000749
Chi sq for colors across neutral	37.32653	3.92E-08
Chi sq for colors across sweet	7.485714	0.057927
Chi sq for colors across all smells	9.442478	0.023951

The Chi Square data above reject the null hypothesis that there is no significant difference between color and odor.
 

Hypotheses	Results
Hummingbirds will prefer red flowers	Data is inclusive
Butterflies would frequent red and yellow flowers	Data is inclusive
Other insects would have no color preference	Data rejects hypothesis
Hummingbirds would be attracted to a sweet odor	Data is inclusive
Butterflies would be attracted to a sweet odor	Data is inclusive
Other insects would have no preference for odor	Data reject hypothesis

Discussion

Our observations indicate that meat attracted flies and carnivorous wasps regardless of the color of the flower.  The sweet smells attracted mostly bees that appeared to show a preference for the red and yellow flowers. We question if our simple floral design can compete with the more elaborate designs and smells of surrounding real flowers in this tropical environment which display such a great diversity of specialized flowers and pollinators.  Further study is needed to answer the question. If man continues to simplify the floral environment by deforestation and mono-cultures, does this mean that there will be a corresponding simplification of the animal pollinators?

Originally the experiment was designed to see the relationship between colors and odors as attractant for pollinators.  Based on our experimental design, we could not conclude the floral preferences of butterflies and hummingbirds, since only one hummingbird visited our site.  Koning (1994) reports that pollinators need visual as well as olfactory clues such as flowers with bull’s- eyes and color contrast to attract pollinators.  In examining the flowers of Wilson Gardens it was noted that they usually displayed color contrast.  Our simple design addressed color and odor preference for bees and flies, but did not address the floral design or pattern preferences.  According to A Pollination Primer prepared by Las Cruces Biological Station, shapes as well as red coloration are important to such pollinators as hummingbirds.

Recommendation for Future Studies

One major flaw with our project was the sampling technique. We realize that observations varied with the individual taking the data. We recommend the following protocol for data acquisition:

1. All team members should record individual observations for ten minutes while standing outside the matrix. After these distance observations the group should walk to each pole and  again make  individual observations. There should be no discussion of data until all have completed this procedure.
2. Compare all data and look for agreement. Discuss discrepant data and come to consensus with the group with regard to how data should be recorded.
3. Redefine what visitation means and how it will be recorded.
4. Repeat the above steps until group observations are replicable regardless of observer.
5. Always conduct the observations in teams of two. After each ten-minute interval discuss your data and come to consensus as to the data should be recorded.

Jansen, D. H. 1975. Ecology of Plants in the Tropics. London, Edward Arnold

Koning, R. E., 1994, Pollination Adaptations, http://www.ecsu.ctstateu.edu/plants_human/pollenadapt.html

Kricher, J., 1997. A Neotropical Companion, Princeton University Press, 451 p.

Prance,G.T. 1985. The Pollination of Amazonia, G.T. Prance and Lovejoy, eds. Oxford: Pergaman Press

A Pollination Primer. Las Cruces Biological Station pamphlet.

Teachers  Applications

This experiment could be use exactly as it is written. It ties in an understanding of biodiversity and a need for global conservation while learning valuable skills that are necessary to run a conclusive experiment.
Students can use this activity to learn the importance of careful experimental design ,sampling techniques and data analysis. Group observations and consensus will improve their ability to make scientific observations.

In biology class this activity could be used to illustrate the coevolution of plants and their pollinators as well as in the unit on plant reproduction

In a chemistry class this activity can be modify to test if varying the molarity of sugars will affect the attraction of pollinators.

This type of experiment can also be used as a beginning exercise in a professional development course for educators. The participants could go through the challenges of design and data collection as we did in this study and ideally develop strategies to apply to their own classrooms.

This activity created additional questions for future inquiry lessons such as:

• Would variation in floral shape attraction of more specialized pollinators?
• Do flowers with multiple colors attract more pollinators than mono-color flowers?
• Do flowers with landing strips and bull’s-eyes attract different pollinators?
• Since our experimental design attracted many spiders, how do they detect the potential of these artificial flowers as a site for insect predation?
• Would the diversity of pollinators increase in a more complex ecosystem such as in a forest??
• Are visits of organisms affected by time of day?
• What will happened if we use other artificial smells?