Summary
This study is a part of a long-range investigation of the variation in resin composition in the tropical leguminous genus Hymenaea and the possible ecological role of these resins within tropical ecosystems. Although the stem, pod and leaf resins are being studied comparatively, emphasis has been placed on the leaf resin system because of the relative simplicity of the chemistry, opportunities for experimental studies on seedling plants under controlled environmental conditions and expected heavy selection pressures on seedlings in the field. The leaf resins from all species of Hymenaea contain 17-18 sesquiterpene hydrocarbons which vary quantitatively such that six compositional types have been recognized. Previous studies of leaf resin composition of 11 species of Hymenaea and 53 populations of the widespread H. courbaril have shown centers of diversity in composition to occur in Amazonia and in southern Brazil in contrast to very little diversity in central and northern South America. Previous investigations of the effects of temperature, light intensity and extreme differences in photoperiod have indicated the low degree of phenotypic plasticity of leaf resin composition in response to physical factors of the environment. Thus it appears that there is strong genetic control over this resin composition. Genetically controlled patterns in resin composition might be due either to random genetic variation or to genetic variation under selective pressures from various environmental conditions. Although there has been considerable controversy regarding this question in attempts to explain floral diversity generally in tropical rainforests (where only differences in adaphic conditions among physical factors of the environment seem to vary sufficiently to exert significant selective pressures), the role of predators or pathogens as selective agents is increasingly being suggested. Also it has been pointed out that an effective way for plants to meet the challenge of predation is by alterning their chemistry. These are changes which are transmitted genetically. The more favorable the physical environment for predators or parasites, the more frequently in evolutionary time the chemical defenses of the plant will have to be modified through genetic changes if the plant is to be successful in the community. Thus one could conclude that recombination systems in higher plants in tropical rainforests and species are subject to directional pressures from herbivores and pathogens in an "evolutionary arms race". These agents of selection favor improvement of defensive posture on the part of the plant hosts which will be balanced in turn by counter adaptation by the predator or parasite. Tropical floras appear to be better endowed than temperate ones with secondary chemicals Which may function defensively. The rate at which chemical resistance to herbivores is achieved is dependent upon a variety of conditions, such as the number of genes controlling the chemistry, dominance of the traits, relative general fitness, the tolerance of the plant, size of the herbivore population, etc. Gene substitution at one or few loci may alter the secondary metabolism of a plant in such a way that it becomes resistant to a given herbivore, and alleles for resistance have been found to be dominant over those for susceptibility. Perhaps too much emphasis in the past has been given to the role of insect predators of plants because the chemical defenses of most plants probably will be revolutionary with a variety of predators and parasites. However, relatively little is known at present about the relative importance of selective pressures of different kinds of herbivores in evolution of the defenses of plants. Also recent research has indicated that chemical defenses of plants may have far-reaching affects in natural ecosystems, i.e., chemical defense systems against herbivores and pathogens may very as a function of plant persistance and abundance, which in turn affects the ecology of herbivores and their plant foods in different ecosystems. In this study leaf resin composition of Hymenaea was compared between parent trees and their progeny in three species: H. intermedia and H. parvifolia from the rainforest ecosystem at Reserva Ducke near Manaus, Amazonas, Brazil and H. courbaril from the savanna ecosystem at Laguna de Los Patos adjacent to the Estacion Biologica de Los Llanos near Calabozo, Guarico, Venezuela. Gas liquid chromatographic analyses were made of resins from leaves collected from seedlings and saplings growing under the canopy of the tree as well as from saplings grown in the greenhouse from seeds originating from some of the same parent trees. Age of the seedlings collected in the field are unknown but specimens from H. intermedia and H. parvifolia are known from phenological data to be at least three years old. Plants of H. parvifolia grown in the UCSC greenhouse vary in age from 6-10 years. Three chemical types were represented among the adult trees but only one type appeared among the progeny under each parent tree. For example, the parent tree of H. parvifolia have leaves with Type I resin composition and all of the surviving seedlings have Type V. However, greenhouse grown saplings of H. parvifolia from seed collected from the same parent tree had three types (I, II and V). An anticipated less variable pattern occurred in H. courbaril from the savanna ecosystem. Both parent tree had leaves of Type II composition with a large number of both seedlings with leaves also of Type II. One tree also had saplings, as well as seedlings, which were also Type II. This predominance of Type II composition of seedlings agrees with previous extensive analysis of greenhouse grown seedlings from various populations of H. courbaril in Central and northern South America. Six other adult trees were also characterized by Type II composition. It is not known whether leaf resin composition changes with ontogeny of the tree from seedling to adult. In some cases, the only type found at various stages of ontogeny has been Type II, which is the most common and generally the most variable type. In any event, whether or not chemistry changes with ontogenetic changes, results from this study indicate that under tropical rainforest conditions there is striking difference in leaf resin composition between parent trees and their progeny. It is in this habitat that the greatest diversity of predators and parasites would be expected. Considerably less variability occurs between parent tree and progeny in the savanna conditions in Venezuela. Here it might be assumed that diversity of herbivores and pathogens would be less than the rainforest. Research in several areas in the tropics has indicated that both distance — and density — responsive predators will tend to eliminate progeny under the canopy of the parent tree or close to it. Therefore it has been assumed that any strategy that enables a tropical tree to disperse its seed away from the parent is likely to help ensure the success of the progeny. On tropical islands, where there are fewer predators and parasites than on mainland sites, seedlings and saplings are often prolific under and adjacent to parent trees. In light of these observations, it is interesting that Hymenaea seedlings are successful under parent trees in both mainland sites studied in this paper, and even under the assumed peak diversity of predators and parasites in the Amazonian rainforest. The question then arises as to why seedlings of these species of Hymenaea are successful under their parent trees, albeit in varying degrees in the two tropical ecosystems. The preliminary results from this study suggest evolution of a successful chemical defense by the plant in response to predation (probably both distance — and density-responsive). Variability in leaf resin composition of seedlings may be an adaptive mechanism enabling them to grow close to or immediately under the canopy of the parent tree in Amazonia. Also, if the seedlings of a different chemical type are successful in becoming established near the parent (and do not change ontogenetically to the type of the parent), this will lead to diversity of types among the adult trees. Diversity of leaf resin composition among the adults probably is of less significance than among the seedlings since the adults can withstand predation (except repeated defoliation) better than younger plants. Less variability in leaf resin types in the llanos might possibly indicate less predator pressures there than in Amazonia. However, this might well be a relative phenomenon because the chemical variability has been measured in terms of chemical types. These types are characterized primarily by the major compounds, whereas certain kinds of herbivores and parasites might be deterred either by the presence or particular concentration of a compound (even in small quantities) or synergistic effects of several compounds. From previous studies it is known that bulk leaf resins from Hymenaea can affect the pupal weight, length of time to pupation and mortality due to viral infection of the widespread legume generalist, beet army worm. Analyses of the effects of single components or particular combinations of them have not been attempted as yet. This preliminary study opens many doors for exploration. Direct study of the genetic control of sesquiterpene synthesis is infeasible in such trees as Hymenaea. Our increasing knowledge of the patterns of variation in leaf resin composition is aiding our understanding of the biochemical pathways through analysis of co-occurrence of the sesquiterpenes. Perhaps by analogy we can later speculate as to whether certain compounds are under single gene regulation and possible dominance relationships. In subsequent papers in this series we plan to include another Amazonian species (H. courharil v. subsessilis) and a more detailed examination of the important central Amazonian species H. intermedia and compare them with two species in southern Brazil (H. stigonocarpa in the cerrado and H. courbaril v. stilbocarpa in riparian forests).