SYSTEMATICS
Phylogenetic Relationships, Species Limits, and Host Specificity of Gall-Forming Fergusonina Flies (Diptera: Fergusoninidae) Feeding on Melaleuca (Myrtaceae) S. J. SCHEFFER,1 R. M. GIBLIN-DAVIS,2 G. S. TAYLOR,3 K. A. DAVIES,3 M. PURCELL,4 M. L. LEWIS,1 J. GOOLSBY,4 AND T. D. CENTER5
Ann. Entomol. Soc. Am. 97(6): 1216Ð1221 (2004)
ABSTRACT Phylogenetic analysis of recently described gall-forming Fergusonina Malloch ßies was performed using DNA sequence data from the mitochondrial cytochrome oxidase I gene. Fifty-three ßies reared from nine species of Melaleuca L. were sequenced. Species boundaries delimited by mitochondrial data conÞrm recent morphological investigation with one exception. Fergusonina turneri Taylor, believed to feed on both Melaleuca quinquenervia (Cav.) S. T. Blake and Melaleuca fluviatilis Barlow, seems to be comprised of two cryptic species, each specialized on one of the two hosts. Because F. turneri is under consideration as a potential biological control agent for invasive M. quinquenervia in the Florida Everglades, understanding cryptic variation and the degree of dietary specialization of this species is critical. KEY WORDS Fergusoninidae, mutualism, gall, plant parasitism, cryptic species
PLANT-FEEDING INSECTS FORM AN extraordinarily diverse assemblage (Ehrlich and Raven 1964, Strong et al. 1984), individual species of which can exhibit a wide range of feeding patterns from broad polyphagy to extreme dietary specialization (Ward and Spalding 1993, Bernays and Chapman 1994). One complication to fully cataloging, describing, and understanding the diversity of phytophagous insects is that sometimes what initially seems to be a single poly- or oligophagous species actually represents a collection of highly specialized, closely related, or even cryptic species (Kulp 1968, Fox and Morrow 1981, Spencer 1990). Molecular systematics can play a critical role in delineating species and providing corroboration for evidence on species limits obtained from other sources, such as morphology, ecology, and/or behavior (Scheffer and Wiegmann 2000). In cases of agricultural pests or potential biological control agents, it is especially important to adequately delineate biologically distinct species and populations (Rosen 1977, Rossman and Miller 1996, Scheffer 2004). 1 USDAÐARS, Systematic Entomology Laboratory, Bldg. 005, Rm. 137, BARC-W, 10300 Baltimore Ave., Beltsville, MD 20705. (e-mail:
[email protected]). 2 University of Florida-Institute of Food and Agricultural Sciences, Fort Lauderdale Research and Education Center, 3205 College Ave., Davie, FL 33314 Ð7799. 3 Centre for Evolutionary Biology and Biodiversity, and School of Agriculture and Wine, Waite Campus, Adelaide University, PMB 1 Glen Osmond, South Australia 5064, Australia. 4 USDA Australian Biological Control Laboratory, PMB 3, 120 Meiers Rd., Indooroopilly, Queensland 4068, Australia. 5 USDAÐARS, Invasive Plant Research Laboratory, 3205 College Ave., Davie, FL 33314 Ð7799.
Fergusonina Malloch (Diptera: Fergusoninidae) ßies and Fergusobia Currie (Tylenchida: Neotylenchidae) nematodes together form galls on plants in the Myrtaceae in what seems to be a unique and obligate mutualism found primarily in Australasia (Currie 1937, Giblin-Davis et al. 2004b). All female Fergusonina ßies carry reproductive nematodes within their abdomens. During ßy oviposition, both ßy eggs and juvenile nematodes are deposited on or near undifferentiated meristematic tissues of the plant host (Currie 1937; GiblinDavis et al. 2001, 2004b, c). The resulting galls develop on shoot buds, inßorescence buds, ßower buds, leaves, or stems, depending on which species are involved (Currie 1937, Giblin-Davis et al. 2004a). Fergusonina/ Fergusobia galls have been recorded from ⬎65 species of myrtaceous host plant species, primarily within the Leptospermoideae, e.g., Eucalyptus, Corymbia, Angophora, and Melaleuca (Currie 1937, Giblin-Davis et al. 2004b, c). Presently, 21 Fergusonina ßy species have been described in association with Eucalyptus (Tonnoir 1937), one species has been described from Syzygium (Harris 1982), and seven species have been described from Melaleuca (Taylor 2004), although it is likely that tens or even hundreds of Fergusonina species on myrtaceous hosts remain to be discovered. Fergusonina that feed on Melaleuca have recently become the subject of intensive study with the goal of identifying a potential biological control agent for the invasive paperbark tree, Melaleuca quinquenervia (Cav.) S. T. Blake (Goolsby et al. 2000). This plant is a highly invasive weedy species currently occupying ⬎500,000 acres in southern Florida and constituting a severe threat to the Everglades ecosystem (Laroche
November 2004 Table 1.
SCHEFFER ET AL.: PHYLOGENETIC RELATIONSHIPS OF GALL-FORMING Fergusonina
1217
Plant host and location information for Fergusonina specimens used in this study
Fly
n
F. burrowsi Taylor F. centeri Taylor F. goolsbyi Taylor F. makinsoni Taylor F. purcelli Taylor F. schefferae Taylor F. turneri Taylor
F. sp. 2 (Taylor 2004)
2 2 3 2 3 2 6 1 1 5 1 1 2 3 1 2 2 4 3 5 2
Plant Host M. viridiflora M. leucadendra M. nervosa M. dealbata M. cajuputi M. nervosa M. quinquenervia
M. fluviatilis M. stenostachya
Location
Ref.
Gall type
MacKay, QLD Cairns, QLD Linear Park, QLD Mareeba, QLD Cairns, QLD Linear Park, QLD Daintree Ferry, QLD Gilmore Rd., QLD Mareeba, QLD Arrawarra, NSW Ballina, NSW Nambucca Heads, NSW Pt. Macquarie, NSW Woodburn, NSW Maryborough, QLD MorayÞeld, QLD Rainbow Beach, QLD Stradbrooke Island, QLD Home Hill, QLD Townsville, QLD Mareeba, QLD
V L L N D D C N N Q1 Q6 Q4 Q3 Q2 Q9 Q7 Q8 Q5 F10 F11 S
Terminal bud Terminal/axial bud Terminal/axial bud “basal rosette” axial stem Terminal/axial bud Terminal/axial bud Terminal/axial bud Terminal/axial bud Terminal/axial bud Terminal/axial bud Terminal/axial bud Terminal/axial bud Terminal/axial bud Terminal/axial bud Terminal/axial bud Terminal/axial bud Terminal/axial bud Terminal/axial bud Terminal/axial bud Terminal/axial bud Axial bud gall, “leaf gall variant”
Ref. indicates a collection reference letter/no. corresponding to collection locations (NSW, New South Wales; QLD, Queensland) indicated in Fig. 1. Gall type follows terminology used in Giblin-Davis et al. (2004a).
and Ferriter 1992, Turner et al. 1998). A survey of pests and pathogens attacking native Australian populations of this species uncovered an inßorescence bud-galling fergusoninid as a possible candidate for use in biological control (Balciunas et al. 1994, 1995; Goolsby et al. 2000). However, surveys of other Melaleuca species indicated that very similar ßies attacked these plants, raising questions of species limits and host speciÞcity of the potential biological control agent. Subsequent morphological study delineated 12 fergusoninid species (but only assigned names to seven) from the nine Melaleuca species known to host Fergusonina ßies (Taylor 2004), but little is known concerning relationships among species and patterns of host use evolution. Additionally, morphological investigation found that Fergusonina turneri Taylor, the potential biological control agent for M. quinquenervia, possesses subtle morphological variation across its range (Taylor 2004) and also seems to feed on Melaleuca fluviatilis Barlow, a degree of oligophagy not recorded in the other Melaleuca feeders. The purpose of this study was to use phylogenetic analysis of mitochondrial sequence data to investigate species limits, host speciÞcity, and host use evolution in Melaleuca feeding fergusoninids. Particular attention was paid to geographic and host plant-associated variation in F. turneri, the potential biological control agent feeding on M. quinquenervia and M. fluviatilis. SpeciÞcally, we asked how variation within this species is structured and whether some populations seem to be more suitable than others for use as a biological control agent. Materials and Methods Fergusonina specimens were obtained from Melaleuca hosts in a variety of locations in Queensland Australia (Table 1; Fig. 1). Larval, pupal, and/or adult
specimens were preserved for study in 95% ethanol and stored at ⫺80⬚C. Voucher specimens of most species have been deposited at the National Museum of Natural History in Washington, DC. When voucher material consists of a portion of an extracted specimen, the voucher carries the specimen extraction code given both in Fig. 2 and in GenBank records. Extraction of total nucleic acids from single specimens (or portions thereof) was accomplished using the DNeasy insect protocol B (QIAGEN, Valencia, CA). Polymerase chain reaction (PCR) ampliÞcation of most of cytochrome oxidase I (COI) was carried out using a Mastercycler Gradient thermocycler (Eppendorf ScientiÞc, Inc., Westbury, NY) with a touchdown ampliÞcation program: initial denaturation at 92⬚C for 2 min, followed by two touchdown cycles from 58 to 46⬚C (10 s at 92 C, 10 s at 58 Ð 46⬚C, 2 min at 72⬚C), 29 cycles of 10 s at 92⬚C, 10 s at 45⬚C, 2 min at 72⬚C, and a Þnal extension step for 10 min at 72⬚C. A single fragment of 1533 bp resulted from PCR ampliÞcation with primers C1-J-1535 (5⬘-ATTGGAACTTTATATTTTATATTTGG-3⬘) and TL-N-3017 (5⬘-CTTAAATCCATTGCACTAATCTGCCATA-3⬘) (primer names follow the system of Simon et al. (1994). PCR product was puriÞed using the QIAquick PCR puriÞcation kits (QIAGEN). We sequenced an ⬇800-bp portion of the COI gene located at the 3⬘ end of the original ampliÞcation product. Complimentary strands of this region were sequenced using the internal primers C1-J-2183 (5⬘-CAACATTTATTTTGATTTTTTGG-3⬘), C1-J-2441 (5⬘-CCTACAGGAATTAAAATTTTTAG TTGATTAGC-3⬘), and C1-N-2508 (5⬘-CTCCAGTTAATCCTCCAACTGTAAAT-3⬘) as well as the external primer TL-N-3017. Sequences were deposited in GenBank under accession numbers AY687933ÐAY687987. ABI Big Dye Terminator sequencing kits (Applied Biosystems, Foster City, CA) were used for all se-
1218
ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA
Vol. 97, no. 6
made to remove identical haplotypes from these analyses; sequence data from all ingroup specimens was used. Outgroups used in the analysis of this dataset were the congeners Fergusonina lockharti Tonnoir, reared from a terminal leaf gall on Eucalyptus camaldulensis Dehnh, and Fergusonina nicholsoni Tonnoir, reared from a ßower gall on Eucalyptus macrorhyncha F. Muell. ex Benth. These eucalypt-feeding species were found to be phylogenetically outside of the Melaleuca feeders in an analysis of the entire genus (Scheffer et al., unpublished data). Results
Fig. 1. Collection locations of specimens used in the study. Letters in the Þgure refer to Melaleuca species as follows: C, cajuputi; D, dealbata; F, fluviatilis; L, leucadendra; N, nervosa; Q, quinquenervia; S, stenostachya; and V, viridiflora. Collections locations for F. turneri (on M. quinquenervia and M. fluviatilis) have been given numbers so that the phylogenetic placement of specimens from these locations can be observed (Fig. 2).
quencing reactions with the modiÞcation that volume of all reaction components was reduced to 25% of that recommended by the manufacturer. Sequence data were obtained by analyzing samples on an ABI 377 Automated DNA sequencer. Contig assembly, as well as the Þnal alignment of consensus sequences, was accomplished using the program Sequencher (Gene Codes Corp., Ann Arbor, MI). Maximum parsimony analysis of the dataset was performed using the heuristic search feature of PAUP* 4.0b8 (Swofford 2001) with 50 random addition replicates. Bootstrapping was performed with 500 pseudoreplications of the dataset. No effort was
For molecular analyses, we were able to obtain eight of the 12 species delineated by Taylor (2004). The Þnal aligned dataset consisted of sequences from 53 ingroup specimens and two outgroup taxa and was 745 bp in length. Uncorrected pairwise distances ranged from 0 to 14.8% within the ingroup and from 14.1 to 17.2% from the ingroup to the outgroups (Table 2). Parsimony analysis resulted in four equally parsimonious trees, one of which is shown in Fig. 2. The four equally parsimonious trees differed only in intraspeciÞc arrangements; the branches recovered in the strict consensus are indicated in bold in Fig. 2. All leaf bud-galling fergusoninids on Melaleuca formed a monophyletic group with strong bootstrap support. Fergusonina goolsbyi Taylor, which forms a morphologically distinct stem gall on Melaleuca nervosa (Lindl) Cheel, exhibited relatively high levels of divergence from the other Melaleuca feeders (Table 2). Excluding F. goolsbyi from the analysis lowered the maximum pairwise distance within the ingroup from 14.8 to 7.1%. All named Melaleuca-feeding fergusoninid species formed distinct clusters with generally high bootstrap support (Fig. 2), and all such species also were recovered in the strict consensus. Although in some cases sampling was limited, with the exception of F. turneri, the potential biological control agent, all named species exhibited maximum uncorrected pairwise distances of 1.1% or less. In contrast, uncorrected pairwise distances within F. turneri ranged from 0 to 5.5%. Inspection of the four equally parsimonious phylograms (Fig. 2) and/or the strict consensus (data not shown) indicates that this named species contains three distinct lineages: one clade, FTQ, feeding only on M. quinquenervia, and two sister clades, FTF-A and FTF-B, feeding only on M. fluviatilis. The maximum uncorrected pairwise distance within the M. quinquenervia-feeding clade FTQ was 2.0%, within both FTF-A and FTF-B clades on M. fluviatilis was 0%, and between the two M. fluviatilis clades was 3.0%. Discussion Molecular analysis of species limits in the Melaleucafeeding fergusoninids largely corroborates previous morphological study (Taylor 2004). All of the host-
November 2004
SCHEFFER ET AL.: PHYLOGENETIC RELATIONSHIPS OF GALL-FORMING Fergusonina
1219
Fig. 2. One of four equally parsimonious phylograms resulting from parsimony analysis of mitochondrial COI sequence data from Fergusoninia ßies. Estimated branch lengths shown above branches, bootstrap values (based on 500 pseudoreplicates) shown below. Branches recovered in the strict consensus are indicated in bold. Host plant from which specimen was reared given after specimen code. Within F. turneri, the collection location reference number is given in parentheses; the geographic distribution of collections is shown in Fig. 1.
speciÞc species delimited here by molecular analysis of mitochondrial variation correspond to species deÞned by morphological analysis. Most of the species are bud gallers that are similar both morphologically
and molecularly. The stem galler on M. nervosa, F. goolsbyi, is distinct in both morphological (Taylor 2004) and molecular characters (Table 2). This species may represent an independent colonization of
1220 Table 2.
FTQ FTF-A FTF-B F. makinsoni F. schefferae F. centeri F. sp. 2 F. purcelli F. burrowsi F. goolsbyi
ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA
Vol. 97, no. 6
Uncorrected pairwise distances within and between Fergusonina species F. turneri FTQ
F. turneri FTF-A
F. turneri FTF-B
0.0Ð2.0
4.6Ð5.1 0.0
3.2Ð4.6 3.0 0.0
F. makinsoni
F. schefferae
F. centeri
F. sp. 2
F. purcelli
F. burrowsi
F. goolsbyi
3.5Ð5.1 4.3Ð4.4 3.8Ð3.9 0.0Ð0.67
4.0Ð5.5 4.8 4.3Ð4.8 3.0Ð3.6 1.1
5.8Ð7.0 6.4Ð6.7 6.3Ð6.6 5.6Ð5.9 5.7Ð6.3 0.0Ð0.54
5.2Ð5.9 5.5 5.8 5.4Ð5.5 5.1Ð5.6 5.9Ð6.0 0.0
5.9Ð6.7 6.6Ð6.8 6.6Ð6.7 5.9Ð6.3 5.9Ð6.8 6.7Ð7.0 5.5Ð5.8 0.0Ð0.27
6.0Ð6.4 6.6 7.0 6.0Ð6.4 6.0Ð6.8 6.8Ð7.1 5.5 3.0Ð3.2 0.0
14.4Ð14.8 14.8 14.1 15.4Ð15.8 15.0Ð15.3 16.2Ð16.5 15.2 15.0Ð15.2 15.6 0.0
Within F. turneri, phylogenetic analysis resulted in three clades (see Fig. 2), here designated FTQ for those on M. quinquenervia, and FTF-A and FTF-B for two groups on M. fluviatilis.
Melaleuca and will be considered more fully elsewhere (Scheffer et al., unpublished data). Most species exhibited little intraspeciÞc mitochondrial variation, although sampling from some species was limited. F. turneri, the only species described as feeding on two hosts, exhibited substantial mitochondrial variation structured by both host and geography (Figs. 1 and 2). All F. turneri from southern locations were reared from M. quinquenervia and formed one well-supported clade, whereas those from northern locations were reared from M. fluviatilis and formed another well-supported clade (Figs. 1 and 2). Given the uncorrected pairwise distances (Table 2) between the M. quinquenervia ßies and the M. fluviatilis ßies of 3.2Ð5.1%, it seems likely that the differences observed between them represent species level divergence, but it remains possible that ßies on the two hosts represent a single geographically structured species. Even within the M. quinquenervia-feeding clade, geographic structure is evident, with more southern populations forming a distinct group (Figs. 1 and 2). With morphological characters there were some differences observed between ßies collected from M. fluviatilis (northern Queensland) compared with ßies from M. quinquenervia (southern Queensland and northern New South Wales), and some minor, overlapping character differences between ßies from northern and southern localities in the southern portion of its range on M. quinquenervia (Taylor 2004). Additional sampling of ßies from M. quinquenervia and M. fluviatilis from intervening geographic regions and from sympatric populations is necessary to fully resolve questions of species boundaries. Similar issues arise when considering mitochondrial variation observed within M. fluviatilis-feeding ßies; the species status of the two clades FTF-A and FTF-B cannot be determined with the present data. Even though the haplotypes detected in this study differ by 3.0%, a value within the range of interspeciÞc divergences previously observed between sister species of ßies (Scheffer and Wiegmann 2000), we cannot determine from this single gene region whether more than one species is present. Additional sampling with special attention given to biological features of M. fluviatilis gallers and collection of data from addi-
tional genes will be required to determine whether one or two species are present on M. fluviatilis. That the ßies attacking M. quinquenervia seem to form a distinct host-speciÞc species is important information for consideration of F. turneri as a biological control agent of this plant. Most importantly, it suggests that all populations of F. turneri are not equal in their potential for use as a biological control agent for M. quinquenervia. F. turneri reared from M. fluviatilis (FTF-A and FTF-B) may represent a distinct hostspeciÞc species and would be unlikely to Þnd M. quinquenervia suitable for oviposition and/or development. Biological control efforts should focus on F. turneri reared from M. quinquenervia (FTQ). Although behavioral tests of oviposition response to a number of plant species are necessary when considering a plant-feeding insect for release as a potential biological control agent (Wapshere 1974, Cullen 1990, Blossey 1995), historical information from a phylogenetic perspective can provide important insights. In addition to providing evidence on species boundaries, a phylogenetic approach can identify actual host shifts that have occurred in the evolutionary history of a particular lineage. For example, F. turneri belongs to a lineage that has fed only on Melaleuca for a considerable length of time. Assuming the Melaleuca bud gallers form a monophyletic group (supported by phylogenetic analysis of the entire genus (Scheffer et al., unpublished data)), we can use a generalized molecular clock estimate for insect mitochondrial DNA of 2.3% per million years (Brower 1994) to roughly indicate the length of time that this group of ßies has fed exclusively on Melaleuca. The maximum pairwise divergence within the Melaleuca bud gallers of 7.1% suggests an estimate of 3.1 million years since these ßies colonized Melaleuca. Although past performance is no guarantee of future behavior, especially when insects are placed in a new environment, knowing that F. turneri belongs to a long-standing Melaleuca-feeding lineage rather than a lineage prone to major host shifts suggests that it should remain speciÞc to Melaleuca. Because there are no native Melaleuca species in Florida, where introduction of F. turneri is under consideration, the data presented here suggest that the risk
November 2004
SCHEFFER ET AL.: PHYLOGENETIC RELATIONSHIPS OF GALL-FORMING Fergusonina
of acquisition of new hosts by introduced F. turneri would be small. Acknowledgments We thank Jeff Makinson for technical assistance during this study. Ron Ochoa, Kevin Omland, and two anonymous reviewers provided useful comments on the manuscript. This project was funded, in part, by USDAÐARS SpeciÞc Cooperative Agreement No. 58-6629-9-004 from the USDA Invasive Plants Research Laboratory, an University of Florida Invasive Plants Working Group mini-grant, and a USDA Special Grant in Tropical and Subtropical Agriculture CRSR03-34135Ð8478.
References Cited Balciunas, J. K., D. W. Burrows, and M. F. Purcell. 1994. Insects to control melaleuca I: Status of research in Australia. Aquatics 16: 10Ð13. Balciunas, J. K., D. W. Burrows, and M. F. Purcell. 1995. Insects to control Melaleuca. II: Prospects for additional agents from Australia. Aquatics 17: 16Ð21. Bernays, E. A., and R. F. Chapman. 1994. Host-plant selection by phytophagous insects. Chapman & Hall, New York Blossey, B. 1995. Host speciÞcity screening of insect biological control agents as part of an environmental risk assessment, pp. 84Ð89. In H.M.T. Hokkanen and J. M. Lynch [eds.], Biological control: beneÞts and risks. Cambridge University Press, Cambridge, United Kingdom. Brower, A.V.Z. 1994. Rapid morphological radiation and convergence among races of the butterßy Heliconius erato inferred from patterns of mitochondrial DNA evolution. Proc. Natl. Acad. Sci. U.S.A. 91: 6491Ð6495. Cullen, J. M. 1990. Current problems in host-speciÞcity screening, pp. 27Ð36. In E. S. Delfosse [ed.], Proceedings of the VII International Symposium on Biological Control of Weeds, Istituto Sperimentale per la Patologia Vegetale, Rome, Italy, 1988. Currie, G. A. 1937. Galls on Eucalyptus trees. A new type of association between ßies and nematodes. Proc. Linn. Soc. NSW 62: 147Ð174. Ehrlich, P. R., and P. H. Raven. 1964. Butterßies and plants: a study in coevolution. Evolution 18: 586Ð608. Fox, L. R., and P. A. Morrow. 1981. Specialization: species property or local phenomenon? Science (Wash. DC) 211: 887Ð893. Giblin-Davis, R. M., J. Makinson, B. J. Center, K. A. Davies, M. Purcell, G. S. Taylor, S. Scheffer, J. Goolsby, and T. D. Center. 2001. Fergusobia/Fergusonina-induced shoot bud gall development on Melaleuca quinquenervia. J. Nematol. 33: 239Ð247. Giblin-Davis, R. M., B. J. Center, K. A. Davies, S. J. Scheffer, G. S. Taylor, J. Goolsby, and T. D. Center. 2004a. Histological comparisons of Fergusobia/Fergusonina-induced galls on different myrtaceous hosts. J. Nematol. (in press). Giblin-Davis, R. M., K. A. Davies, G. S. Taylor, and W. K. Thomas. 2004b. Entomophilic nematode models for studying biodiversity and cospeciation. In Z. X. Chen, S. Y. Chen, and D. W. Dickson [eds.], Nematology, advances and perspectives. Tsinghua University Press/ CABI Publishing, New York.
1221
Giblin-Davis, R. M., S. J. Scheffer, K. A. Davies, G. S. Taylor, J. Curole, T. D. Center, J. Goolsby, and W. K. Thomas. 2004c. Coevolution between Fergusobia and Fergusonina mutualists. Nematol. Monogr. Perspect. 2: 1Ð11. Goolsby, J. A., J. R. Makinson, and M. F. Purcell. 2000. Seasonal phenology of the gall-making ßy Fergusonina sp. (Diptera: Fergusoninidae) and its implications for biological control of Melaleuca quinquenervia. Aust. J. Entomol. 39: 336 Ð343. Harris, K. M. 1982. First record of Fergusoninidae (Diptera: Schizophora) outside Australia: a new species of Fergusonina on Syzygium in India. Syst. Entomol. 7: 211Ð216. Kulp, L. A. 1968. The taxonomic status of dipterous holly leaf miners (Diptera: Agromyzidae). Univ. Maryland Agric. Exp. Stat. Bull. A. 155: 1Ð 42. Laroche, F. B., and A. P. Ferriter. 1992. Estimating expansion rates of Melaleuca in south Florida. J. Aquat. Plant Manage. 30: 62Ð 65. Rosen, D. 1977. The importance of cryptic species and species identiÞcations as related to biological control, pp. 23Ð35. In J. A. Romberger [ed.], Biosystematics in agriculture. Allanheld, Osmun & Co., Montclair, NJ. Rossman, A. Y., and D. R. Miller. 1996. Systematics solves problems in agriculture and forestry. Ann. Missouri Bot. Garden 83: 17Ð28. Scheffer, S. J. 2004. Invasive Diptera: using molecular markers to investigate cryptic species and the global spread of ßies, pp xx-xx. In B. M. Wiegmann and D. K. Yeates [eds.], The evolutionary biology of ßies. Columbia University Press. New York. Scheffer, S. J., and B. M. Wiegmann. 2000. Molecular phylogenetics of the holly leafminers (Diptera: Agromyzidae: Phytomyza): species limits, speciation, and dietary specialization. Mol. Phylogenet. Evol. 17: 244 Ð255. Simon, C., F. Frati, A. Beckenbach, B. Crespi, H. Liu, and P. Flook. 1994. Evolution, weighting, and phylogenetic utility of mitochondrial gene sequences and a compilation of conserved polymerase chain reaction primers. Ann. Entomol. Soc. Am. 87: 651Ð701. Spencer, K. A. 1990. Host specialization in the world Agromyzidae (Diptera). Kluver Academic Publishers, Dordrecht, The Netherlands. Strong, D. R., J. H. Lawton, and R. Southwood. 1984. Insects on plants, community patterns and mechanisms. Harvard University Press, Cambridge, MA. Swofford, D. L. 2001. PAUP*. Phylogenetic analysis using parsimony (*and other methods). Sinauer, Sunderland, MA. Taylor, G. S. 2004. Revision of Fergusonina Malloch gall ßies (Diptera: Fergusoninidae) from Melaleuca (Myrtaceae). Invertebr. Syst. 18: 251Ð290. Tonnoir, A. L. 1937. Revision of the genus Fergusonina Mall. Proc. Linn. Soc. NSW 62: 126 Ð146. Turner, C. E., T. D. Center, D. W. Burrows, and G. R. Buckingham. 1998. Ecology and management of Melaleuca quinquenervia, an invader of wetlands in Florida, USA. Wetlands Ecol. Manage. 5: 165Ð178. Wapshere, A. J. 1974. A strategy for evaluating the safety of organisms for biological control of weeds. Ann. Appl. Biol. 77: 201Ð211. Ward, L. K., and D. F. Spalding. 1993. Phytophagous British insects and mites and their food-plant families: total numbers and polyphagy. Biol. J. Linn. Soc. 49: 257Ð276. Received 8 January 2004; accepted 25 August 2004.
