Weed Technology. 2004. Volume 18:1400–1402
Symposium Genetic Variation in Melaleuca quinquenervia Affects the Biocontrol Agent Oxyops vitiosa1 F. ALLEN DRAY, JR., BRADLEY C. BENNETT, TED D. CENTER, GREG S. WHEELER, and PAUL T. MADEIRA2 Abstract: Melaleuca was first imported into the United States during 1886, and introduction records suggest that at least six sources have contributed to extant populations in Florida. Allozyme analyses found substantial genetic variation within and among populations, contributing to geographic structuring of melaleuca in southern Florida. The presence and distribution of two chemical phenotypes (chemotypes) contributed to this variation. Performance of the melaleuca snout beetle imported as a biological control agent differed dramatically in laboratory studies depending on which chemotype it was fed, with larval survivorship and growth substantially greater on an (E)-nerolidol chemotype. We are currently investigating whether these differences can be detected in the field. Nomenclature: Melaleuca, Melaleuca quinquenervia (Cav.) Blake; melaleuca snout beetle, Oxyops vitiosa Pascoe. Additional index words: Allozyme analysis, biological control, chemotype, herbivory, invasion history, insect–plant interactions.
INTRODUCTION
Differential response of herbivorous insects to variation among host–plant genotypes is common, as reported in the general insect–plant interaction literature. For example, genotype differences among cottonwood hybrids (Populus angustifolia 3 Populus fremontii) have a strong effect on bud-gall mite (Aceria parapopuli Kiefer: Eriophyidae) population growth rates (McIntyre and Whitham 2003). Such differential responses to plant variation may be particularly prominent among insects selected for use as biological control agents because they are so closely adapted to their host plants. However, relatively few biocontrol projects have investigated this possibility, despite repeated calls for such studies (Crawley 1983; Harley and Forno 1992; Harris 1989; van den Bosch and Messenger 1973). Australian populations of melaleuca differ in the principal terpenoid constituents in their leaves (Ireland et al. 2002), and terpenoids often serve a defensive function 1 Received for publication January 27, 2004, and in revised form May 18, 2004. 2 First author: Ecologist, USDA-ARS Invasive Plant Research Laboratory, Fort Lauderdale, FL 33314, and Ph.D. Candidate, Florida International University, Miami, FL 33199; second author: Professor, Florida International University, Miami, FL; third, fourth, and fifth authors: Entomologist, Entomologist, and Molecular Biologist, respectively, USDA-ARS Invasive Plant Research Laboratory, Fort Lauderdale, FL 33314. Corresponding author’s Email:
[email protected].
in myrtaceous plants (Edwards et al. 1993; Morrow and Fox 1980; Schmidt et al. 2000). It is possible that selective pressures arising from these defensive compounds may have caused herbivorous insects that feed on melaleuca to specialize on subpopulations characterized by different mixtures of terpenoids. The terpenoid compounds that characterize essential oils in melaleuca are genetically controlled (Shen and Dooner 2000; Trapp and Croteau 2001). Thus, importation of the melaleuca snout beetle from Australia as a biological control of this Everglades invader afforded an opportunity to test the hypothesis that biocontrol agent performance can be influenced by the genetics of the target weed in its adventive range (an area where it is not native but has become locally naturalized). INVASION HISTORY
Examination of the invasion history of melaleuca (Dray 2003) showed that populations in Florida derive from multiple introduction events, involving six source populations. However, only two of these were from the native range of melaleuca (Australia, New Caledonia, and southern Papua New Guinea) (Dray 2003). This finding suggested that Florida populations centered around introduction points derived from different sourc-
1400
WEED TECHNOLOGY
Figure 1. Allozyme analysis using cellulose acetate gel electrophoresis showed that melaleuca in Florida harbored significant among-population genetic differences and that a substantial portion of this variation was due to regional differences between Gulf Coast and Atlantic Coast populations.
es might harbor different genotypes and levels of diversity. The earliest introduction occurred during 1886 in Oneco, FL, and preceded by 20 yr the best known introductions before this study (Dray 2003). Other important introduction points in Florida included Miami, Orlando, and Estero (Dray 2003). The principal reason for the importation of melaleuca was as a landscape tree (Andrews 1930, cited in Meskimen 1962; Gifford 1945; Meskimen 1962; Nehrling 1933), but some proponents also sought a timber crop (Gifford 1946). Some founder populations escaped cultivation to become naturalized within two decades, but the principal means of dispersal was through human transport (Dray 2003). GENETIC VARIATION
Allozyme analyses (Dray 2003) using the cellulose acetate gel electrophoresis method confirmed the premise that this invasion history resulted in high levels of genetic diversity in Florida. Measures of allelic richness (A) and heterozygosity (Ho) for melaleuca in Florida fell into the ranges reported for other tropical woody plants (A 5 1.00 to 3.12, Hamrick et al. 1979; Ho 5 0.07 to 0.34, Hamrick and Loveless 1989). Further, the pattern of the introductions, and the subsequent redistribution of progeny, resulted in geographic structuring among the populations (Figure 1). This structuring follows the primary distributional pattern for melaleuca in Florida: populations along the Gulf Coast are genetically distinct from those along the Atlantic Coast. Rates of gene flow were quite low, as would be expected with high amongpopulation gene diversity (Hamrick 1989). Herbivorous insects are unable to detect genetic variation, however. It was thus important to assay phenotypic differences that are recognized by and might influVolume 18, Invasive Weed Symposium 2004
Figure 2. Melaleuca snout beetle survival from neonate to adult was substantially greater when the larvae were raised on melaleuca leaves characterized by high concentrations of (E)-nerolidol as opposed to a mixture of viridiflorol and 1,8-cineole. Gadj-values represent results of two-way classification G-tests of independence used to analyze survivorship data, adjusted using William’s correction due to small sample size.
ence biological control agents. Some of the principal essential oil constituents in melaleuca have been found to improve plants’ resistance to herbivores (Edwards et al. 1993; Morrow and Fox 1980; Schmidt et al. 2000). Florida populations were thus surveyed for the presence of different chemotypes (i.e., chemical phenotypes; Dray 2003) as characterized by essential oil composition. Foliar oils followed similar trends to the allozyme analysis, in that Gulf Coast populations differed from Atlantic Coast populations. For instance, Gulf Coast trees yielded nearly twice as much oil as Atlantic Coast trees when both were grown in a common garden. These differences were partially explained by the predominance of a chemotype very rich in the sesquiterpene (E)-nerolidol in melaleuca trees from the Gulf Coast but rich in a mixture of the monoterpene 1,8-cineole and the sesquiterpene viridiflorol in trees from the Atlantic Coast.
BIOAGENT RESPONSES
This variation among melaleuca populations in Florida provided the foundation for investigating the principal question of this study: whether biocontrol agents might be influenced by host–plant genotype differences. Bioassays showed that melaleuca snout beetle performance differed dramatically depending on the chemotype of the foliage they were fed (Dray 2003). Larval survivorship was fourfold greater on the (E)-nerolidol chemotype (Figure 2). Growth was also greater, with adult melaleuca snout beetles gaining nearly 50% more biomass on the (E)-nerolidol plants than on the second chemotype. 1401
DRAY ET AL.: MELALEUCA VARIATION AFFECTS MELALEUCA SNOUT BEETLE PERFORMANCE
These results offered possible insight as to why a few of the initial melaleuca snout beetle releases in Florida failed to establish persistent populations—releases on the resistant chemotype were more at risk and coupled with abiotic factors to prevent establishment. Further, Wheeler (2003) reported that foliar nitrogen concentrations did not affect melaleuca snout beetle performance as anticipated, and hypothesized that chemotype differences may have confounded the results of his study. Results from our bioassays with melaleuca snout beetle support Wheeler’s hypothesis (Wheeler 2003). IMPLICATIONS
Biological control of weeds is a scientific discipline founded on the evolutionary premise that through geologic time many herbivorous insects have become specialized to breed, feed, and develop on one or a very few plant species (Center et al. 1997; DeBach 1974; Harley and Forno 1992; Huffaker 1971). Results from this study argue that weed biocontrol researchers must be aware that such close host fidelity may cause some biocontrol agents to be adapted to a restricted subset of genotypes within the target weed species. Such extreme specialization may explain some of the ‘‘failures’’ (i.e., no establishment, poor performance) that have occurred in weed biocontrol programs. Remaining vigilant for potential genotype effects could help improve success rates in the future. The implications of this research can be summarized as follows: (1) invasion history offers to biocontrol workers important insights into the genetics and biology of weedy species in their adventive ranges and should be carefully scrutinized as part of the initial stages of a weed biocontrol project, (2) routine examination of weed population genetics would help biocontrol researchers differentiate among genotypes (biotypes, ecotypes, demes) present in a target’s adventive and native ranges, (3) weed biocontrol targets should routinely be screened for plant defensive compounds, which can also be useful indicators of the numbers and kinds of biotypes present, and (4) specialist insects are, by definition, highly adapted to their hosts and are likely to be sensitive to differences among biotypes of their weed targets. Thus, efficacy of biological control programs can be enhanced by screening for such differences early in the process.
1402
LITERATURE CITED Andrews, A. H. 1930. The Fort Myers Press. March 31. Center, T. D., J. H. Frank, and F. A. Dray, Jr. 1997. Biological control for managing invasive nonindigenous species. In D. Simberloff, D. Schmitz, and T. Brown, eds. Strangers in Paradise: Impact and Control of Florida’s Nonindigenous Species. Covelo, CA: Island. Pp. 347–373. Crawley, M. J. 1983. Herbivory: The Dynamics of Animal-Plant Interactions. Berkeley: University of California Press. Pp. 204–206, 216–222, 343. DeBach, P. 1974. Biological Control by Natural Enemies. London: Cambridge University Press. P. 46. Dray, F. A., Jr. 2003. Ecological Genetics of Melaleuca quinquenervia (Myrtaceae): Population Variation in Florida and its Influence on Performance of the Biological Control Agent Oxyops vitiosa (Coleoptera: Curculionidae). Ph.D. dissertation. Florida International University, Miami, FL. 161 p. Edwards, P. B., W. J. Wanjura, and W. V. Brown. 1993. Selective herbivory by Christmas beetles in response to intraspecific variation in Eucalyptus terpenoids. Oecologia 95:551–557. Gifford, J. C. 1945. Living by the Land. Coral Gables, FL: Glade House. 139 p. Gifford, J. C. 1946. Ten Trustworthy Tropical Trees. Emmaus, PA: Organic Gardening. 91 p. Hamrick, J. L. 1989. Isozymes and the analysis of genetic structure in plant populations. In D. E. Soltis and P. S. Soltis, eds. Isozymes in Plant Biology. Portland, OR: Dioscorides. Pp. 87–105. Hamrick, J. L. and M. D. Loveless. 1989. The genetic structure of tropical tree populations: associations with reproductive biology. In J. H. Bock and Y. B. Linhart, eds. Plant Evolutionary Ecology. Boulder, CO: Westview. Pp. 131–146. Hamrick, J. L., Y. B. Linhart, and J. B. Mitton. 1979. Relationships between life history characteristics and electrophoretically detectable genetic variation in plants. Annu. Rev. Ecol. Syst. 10:173–200. Harley, K.L.S. and I. W. Forno. 1992. Biological Control of Weeds: A Handbook for Practitioners and Students. Melbourne: Inkata. Pp. 24, 36. Harris, P. 1989. Practical considerations in a classical biocontrol of weeds program. International Symposium on Biological Control Implementation: Proceedings and Abstracts (McAllen, Texas, April 4–6, 1989). NAPPO Bull. 6:23–32. Huffaker, C. B. 1971. Biological Control. New York: Plenum. Pp. 51–52, 160–162. Ireland, B. F., D. B. Hibbert, R. J. Glodsack, J. C. Doran, and J. J. Brophy. 2002. Chemical variation in the leaf essential oil of Melaleuca quinquenervia (Cav.) S.T. Blake. Biochem. Syst. Ecol. 30:457–470. McIntyre, P. J. and T. G. Whitham. 2003. Plant genotype affects long-term herbivore population dynamics and extinction: conservation implications. Ecology 84:311–322. Meskimen, G. F. 1962. A Silvical Study of the Melaleuca Tree in South Florida. M.S. thesis. University of Florida, Gainesville, FL. 178 p. Morrow, P. A. and L. R. Fox. 1980. Effects of variation in Eucalyptus oil yield on insect growth and grazing damage. Oecologia 45:209–219. Nehrling, H. 1933. The Plant World in Florida (collected and edited by A. Kay and E. Kay). New York: Macmillan. 304 p. Schmidt, S., G. H. Walter, and C. J. Moore. 2000. Host plant adaptations in mytaceous-feeding Pergid sawflies: essential oils and the morphology and behaviour of Pergagrapta larvae (Hymenoptera, Symphyta, Pergidae). Biol. J. Linn. Soc. 70:15–26. Shen, B. and H. K. Dooner. 2000. Ac tagging and characterization of a terpenoid cyclase gene induced by herbivore damage. Maize Gen. Conf. Abs. 42:T31, P132. Trapp, S. C. and R. B. Croteau. 2001. Genomic organization of plant terpene synthases and molecular evolutionary implications. Genetics 158:811– 832. van den Bosch, R. and P. S. Messenger. 1973. Biological Control. New York: Intext Educational. P. 79. Wheeler, G. S. 2003. Minimal increase in larval and adult performance of the biological control agent Oxyops vitiosa when fed Melaleuca quinquenervia leaves of different nitrogen levels. Biol. Control 26:109–116.
Volume 18, Invasive Weed Symposium 2004
