Research Projects that Utilize Resources of L. Gilbert's Tropical Greenhouse-Insectary System at UT Austin

The following is a brief explanation of how we use living populations of plants and insects for scientific investigations. The live specimens exported from P.N. Corcovado form only a part of our system, which is available for comparative studies to any student or researcher in the world. The detailed studies here greatly enhance what we can learn in the field. Eventually the combination of laboratory, greenhouse, and field studies will give us a deep understanding of how this system maintains its structure and diversity in a place like Sirena. With this understanding I hope we can provide insights that will be of general use in managing the biodiversity of rainforest parks in the neotropics.

L.E. Gilbert's research in P.N. Corcovado on insect-hostplant relationships and population dynamics of Heliconius butterflies also involves detailed laboratory studies of behavior, genetics, host plant use, and physiology which cannot be conducted in the field. For many studies, the numbers of insects or plants required for research greatly exceed the number that can be safely removed from the natural population. Therefore, a few living males and females or eggs or larvae of butterflies and a few seeds or stems of host plants can be collected for cultivation in greenhouses. There, away from natural diseases, parasites and predators, these insects and plants can be grown to large quantities, sufficient for experiments.

Another reason for exporting small amounts of living insects and plants to the research greenhouses at The University of Texas is that living plants and insects can be analyzed with equipment and techniques that cannot be taken to the field. Moreover, some studies involve comparing insects and plants from different countries. For example, our collection of living Passiflora and Heliconius includes plants from Brazil, Trinidad, Panama, Mexico, Guatemala, Texas, Peru, Colombia, Costa Rica, Puerto Rico, and Ecuador and butterflies from Brazil, Bolivia, Mexico, Ecuador, and Costa Rica.

This research facility allows us to investigate many biological questions about the ecosystem and is available to students and researchers from any country. Currently, students from Canada, Brazil, Peru, and the U.S. are working on projects in these greenhouses and contributing to our knowledge of Heliconius and Passiflora.

Projects which utilize some of the living insects and plants originally collected in P.N. Corcovado include

1) Learning behavior in Heliconius. Observations of marked Heliconius adults at Sirena indicate that these insects are able to learn the locations of flowers and nocturnal roost sites. We are trying to verify the extent of learning capacity in these butterflies using small-scale experiments in cages. The results indicate that Heliconius do not remember the specific location of a rewarding flower within a small area. Their ability to learn appears to involve landmarks on a very large scale (many meters).
- Evandro Oliveira

2) Evolution of body size in Heliconius. Within a location such as Corcovado Park, different species of Heliconius are characterized by different average body and wing lengths. Why should this be, and why aren't all species bigger or smaller? How is size determined? To begin to answer some of these questions, we are comparing growth and final adults size in two races of Heliconius erato which differ in size and for which size can be studied in respect to quality of host plants. We are comparing the impact of novel vs. natural hosts on a small race (H. erato petiveranus from P.N. Corcovado) versus a larger race (H. erato phyllis from Sao Paulo, Brazil). The study will make it possible to know where in the life cycle the differences in size develop between these populations. Genetic studies will attempt to estimate the genetic components of the observed differences.
- Mirian Medina

3) Toxic chemicals and host plant specialization in Heliconius. At Sirena, some Heliconius, such as H. pachinus, use many different species of Passiflora for larval food. Others, such as H. melpomene, H. sara, and H. hewitsoni, are highly specialized, placing eggs on only one host plant. We know that H. melpomene is not confined by host chemistry to its only host P. menispermifolia because we can rear H. melpomene on many of the available but unused species. However, what about other specialists?

It appears that the larvae of both H. hewitsoni and H. sara are almost as specialized as their egg-laying mothers. It appears also that both of these specialists are constrained by their mode of detoxifying or counteracting the toxic cyanogenic glycosides of their hosts. Unlike most Heliconius which appear to excrete or metabolize the host plants' cyanogens and later manufacture their own, H. hewitsoni and H. sara store the unique cyanogenic glycosides of their respective hosts (P. pittieri and P. auriculata) and do not manufacture their own cyanogens. The special biochemical mechanisms needed by larvae to store the cyanogenic compounds of their host Passiflora may reduce their capacity to use chemically different Passiflora. See Discovery review of Nature article.
- Helene Engler

4) Adult feeding and palatability in Heliconius. Most Heliconius are unpalatable to birds such as the Jacamar studied by Peng Chai (Taiwan) at Sirena. Heliconiines which do not feed on pollen are relatively edible. At Sirena, Heliconius charitonia does not compete well for pollen and is relatively palatable to birds, as are the non-pollen feeding species of Dryas, Eueides, Philethria, and Dione.

It is possible that the rarity of H. charitonia at Sirena and in P.N. Corcovado results from the fact that adults are more edible to birds and do not live as long as other Heliconius. We are comparing H. charitonia from Sirena (the laboratory population was established in 1982) with those from Veracruz, Mexico, where, with less competition from other species of Heliconius, H. charitonia adults can collect pollen on a regular basis. Our hypothesis is that pollen provides the amino acids that allow adults Heliconius to produce the cyanogenic glycosides, and that these compounds provide protection against birds. We are also testing the hypothesis that in areas, such as Veracruz, where pollen is more available to H. charitonia, this species has evolved the ability to use pollen more effectively for producing distasteful compounds.
- Marcio Cardoso

5) Phylogenetic studies of the 10 Heliconiiti genera. Larvae and adults collected in the field and produced in the greenhouse were used to gather morphological and molecular characters to understand the phylogenetic relationships among the 10 presently recognized genera of Passionvine butterflies (sub-tribe Heliconiiti). The objectives of this study are 1) to investigate the relationship between the passionvine butterflies and their relatives within the subfamily Heliconiinae, and 2) understand the relatedness among the members of this sub-tribe with special emphasis on high-level relationships. Knowledge of the high-level phylogeny for the passionvine butterflies will help understanding presumably adaptive patterns observed today such as host-plant associations, chemical defense, adult feeding behavior (nectar supplemented with pollen), and adult mating behavior (males are attracted to female adults or pupae).
- Carla Penz

6) Adult feeding behavior. Adult Heliconius and Laparus butterflies supplement their nectar diet with pollen gathered mainly from Psiguria (Cucurbitaceae) and Lantana (Verbenaceae) flowers. In order to understand the evolution of this key innovation, proboscis movements during flower visit and pollen-processing were recorded with the aid of a video camera for several species of Heliconius and their relatives of the genera Agraulis and Dryas that do not use pollen in their diets. Morphological studies revealed no unique structures to the species that use pollen in their diets. However, a distinct combination of features exists, and can be functionally interpreted in the context of collection and processing of pollen. These results will be studied in a phylogenetic context so to recognize the association of morphological features, behavioral, and physiological traits within and between linneages of passionvine butterflies.
- Harald Krenn and Carla Penz

7) Mating tactics: pupal vs non-pupal mating. Approximately one-half of Heliconius species possess a unique mating behavior: males locate female pupae and initiate mating before the female has eclosed. Our laboratory works on three aspects of pupal mating. First we study the behavior of competing males and the evolutionary consequences of mate competition on pupae. Second we study the role of chemical signals in the discovery and evaluation of pupae and in the enforcement of female monogamy. Third we are studying the impact of pupal mating males on other species of Heliconius in the community. Greenhouse work is important for those species whose mating is rarely observed in the field and for studies of chemical signals.
- Erika Deinert

8) Evolutionary genetics of the Heliconius cydno group. Heliconius species vary dramatically in wing pattern from region to region. In Costa Rica a dramatic color pattern difference occurs between the related species H. cydno on the Atlantic side and H. pachinus on the Pacific side of Costa Rica. Genetic studies involving crosses of these species reveal that in spite of striking color pattern distinction, these species have few other differences in life history, ecology, behavior or genetics. Indeed, they are fully interfertile and would form a broad hybrid zone if a mountain barrier did not separate their populations. Studies in progress investigate the mating preferences of males and females of each "species" and the developmental genetics of color pattern. Stocks of H. pachinus derive from females collected at Sirena. H. cydno can be obtained from butterfly farms on the Atlantic side of Costa Rica.
- L.E. Gilbert

9) Eudulaphasia invaria is a small but conspicuous day-flying moth which ranges from Brazil to Mexico. In Corcovado I found that it places eggs in spider webs near host plants, Spigelia anthelemia (Loganiaceae). It has been cultivated in Austin for 18 generations. In laboratory experiments, spiders and preying mantids attack but reject these moths. They are highly repellent to insect predators. Research projects planned include

a) finding the chemical responsible for the repellent qualities of E. invaria,

b) studies on the learning ability of salticid spiders (we use Texas spiders),

c) testing for the ability to sequester host plant chemicals (Costa Rican moths are fed plants from Brazil, Florida and Texas in addition to natural hosts) by looking for variation in compounds stored by these moths that follows variation in host plant chemistry.

- L.E. Gilbert

10) Origin and evolutionary relationships of the cultivated passion fruit. Passiflora (Passifloraceae) is a genus composed of approximately 475 species of climbing herbs and woody lianas. It is primarily a Neotropical group, however, a few species are restricted to the tropical and sub-tropical regions of Africa, New Zealand and Australia. Members of this genus are widely cultivated for their ornamental flowers and edible fruit. Although most species are cultivated on a regional scale, P. edulis f. edulis and P. edulis f. flavicarpa are produced world-wide. In addition to its economic importance, Passiflora is of significant evolutionary interest as a model system illustrating coevolutionary relationships between plants and their specialist herbivores. The last comprehensive tratment of this genus was completed in 1928 by Killip. Although this was a phenomenal piece of work, both in scope and detail, many of the subgenera and sectional circumscriptions are clearly artificial and poorly defined. In addition, many new species have been described since this revision was pubished that have not been incorporated in Killip's framework due to the lack of clear sectional delimitations. A phylogeny has never been proposed for this group. The focus of this study is the X=9 group which is composed of subgenera Passiflora, Manicata, Tacsonia, Rathea, Dysosmia, Calopathanthus, Distephana, Tacsonioides, and Granadillastrum. These subgenera contain all of the species that are cultivated for their fruit on a local or world-wide scale. Molecular techniques including restriction site analysis of cpDNA and sequencing of the ITS region of the nuclear ribosomal repeat will be utilized to generate markers and analyzed in order to propose a phylogeny of the X=9 group. This phylogeny will then be utilized to investigate the hypothesis of a hybrid origin for P. edulis f. flavicarpa. The information yielded through this study will provide important phylogenetic information for a historically understudied group of significant economic, ecological and evolutionary interest. This project will clarify the relationships of the X=9 group providing the necessary phylogenetic framework for testing hypotheses about the origin of one of the most widely cultivated species and for investigations into the coevolutionary relationship between Passiflora and its specialist herbivore, the Heliconius butterfly.
- Katie Hansen

Back to
Gilbert Lab
Home Page
Butterfly/Plant Research Facilities Graduate
PostDocs, etc. Teaching Bibliography

Back to UT Austin home page

25 Oct 00
For further information, contact:
This site created and maintained by Sharon S. Bramblett.