Memoirs of the Museum of Victoria 56(2):393-399 (1997) 28 February 1997 https://doi.org/10.24199/j.mmv.1997.56.30 BIODIVERSITY OF PLANTHOPPERS (HEMIPTERA: DELPHACIDAE) ON THE HAWAIIAN SILVERSWORD ALLIANCE: EFFECTS OF HOST PLANT PHYLOGENY AND HYBRIDISATION

2 George K. Roderick 1 and Edward C. Metz

2 Center for Conservation Research and Training, University of , 3050 Maile Way. Gilmore 409. Honolulu, HI 96822, USA -Present address: Scripps Institute of Oceanography, University of California-San Diego, La Jolla, CA 92093, USA

Abstract

Roderick, G.K. and Metz, E.C., 1997. Biodiversity of planthoppers (Hemiptera: Dclpha- cidae) on the Hawaiian silversword alliance: effects of host plant phytogeny and hybridis- ation. Memoirs of the Museum of Victoria 56(2): 393-399. Many hypotheses have been advanced to account for the biodiversity of herbivorous insects. Here we test whether the diversity of a group of planthoppers, genus Nesoxydnc (Hemiptera: Delphacidae), can be explained by the history and patterns of hybridisation of their host plants in the Hawaiian silversword alliance, many of which arc critically endangered or threatened. Molecular data (DNA sequences of mitochondrial cytochrome between these oxidase I) for planthoppers examined to date reveals significant co-speciation species insects and their hosts. Nesosydne planthoppers are highly host specific with each feeding on only one plant species or on closely related species that hybridise. There was no herbivorous evidence to support the "hybrid-bridge" hypothesis as a mechanism by which were centres of insects may switch hosts. Nor was there evidence that plant hybrid zones suggests that insect biodiversity. Rather, patterns of host plant use within plant hybrid zones planthopper diversification follows host plant diversification.

Introduction at the generic and family levels appear to be largely conservative with respect to their host these lineages The current threats to biological diversity have affiliations and many species in (Dethier, 1954; Ehrlich necessitated the understanding of the forces are highly host specific Farrell and Mitter, 1993). responsible for both its generation and demise and Raven, 1964; plant use by herbivor- (Wilson, 1988; 1996). Insects are among the Current patterns of host insects can be explained by one of two most diverse organisms, with over 1 million ous of which can lead to greater described to date and perhaps another 10 to 30 hypotheses, both co-speciation with host million remaining to be discovered in tropical herbivore diversity: plant switching. Co-speciation is regions (Erwin, 1982; 1986). Herbivorous plants and host of speciation events in two lin- species form a major component of these insects the matching for eages, such that the two phylogenies resemble (Strong et al., 1984) and most explanations another (see Brooks, 1979; Mitter and their diversity are linked to the diversity of one 1983; Brooks, 1988; Hafneret al., 1994; plants on which they feed (Mitter et al., 1988; Brooks, 1995a). Co-speciation may be a conse- 1991; Farrell and Mitter, 1993; Thompson, Page, al., quence of co-evolution, but may also arise 1994; Funk et al., 1995; Futuyma et through other mechanisms. By contrast, host- 1995). switching is a change of hosts (see Much of the current insect biodiversity was plant 1983a; 1983b; Thompson, 1994) present even before the radiation of the flower- Futuyma, other than would be predicted by the host phy- ing plants (angiosperms) and appears to be such that the two phylogenies arc no coincident with the earlier diversification of logeny, longer congruent. Compilations of research on seed plants (Labandeira and Sepkoski, 1993). herbivorous insects and their hosts indicates Herbivorous insects may have further diversi- which that while a few insect radiations do appear to be fied with the radiation of angiosperms tightly correlated with radiations of their hosts, required pollinators (Ehrlich and Raven, 1964); most insect radiations show evidence of host- indeed, some radiations of insect species do flowering switching (Mitter et al., 1988; 1991; Farrell and appear to be correlated with those of lineages Mitter, 1993; Funk et al., 1995). plants (Mitter et al., 1988; 1991). Insect

393 .

394 G. K. RODERICK AND E, C METZ

Recently, it has been suggested that the fre- and shape (Carr, 1985), and recently by gen- quent occurrence of hybrids between plant etic data (RAPDs, V. Caraway and C. species may form centres of biodiversity or Morden, unpubl. data; Friar et al., 1996). otherwise play a role in herbivore diversification Recently, Baldwin and colleagues (Baldwin (Floatc and Whitham, 1993; Strauss, 1994; et al., 1991; Baldwin and Robichaux, 1995; Whitham et al.. 1994). Hybridisation between Baldwin, in press) have used molecular genetic host species may also play a role in parasite data to generate a hypothesis of evolutionary diversification. Hybrid hosts not only create relationships among members of the alliance

new niches, but may allow specialised parasites (fig. 1). The alliance appears to be monophyletic an escape from evolutionary dead ends (see and to contain distinct clades within Hawaii. Moran, 1988) — hybrids may provide a Divergence among extant species is likely in the "bridge" to novel host species (Floate and range of 4-6 MY, or no older than the age of

Whitham, 1 993). Whether "hybrid-bridges" are Kauai (Baldwin and Robichaux, 1 995; Baldwin, important in explaining the history and patterns in press). Conflicting evidence from nuclear, of host plant use by herbivorous insects remains karyotype, and cytoplasmic DNA data suggests controversial (see Floate and Whitham, 1993). that some species may be of hybrid origin. The The Nesosydne planthoppers (Homoptera: existence of a phylogenetic hypothesis for plant Delphacidae) in Hawaii are one of several insect species in the silversword alliance is an essential radiations with members associated with the element in the analysis presented here for Hawaiian silversword alliance (Asteraceae) (see several reasons:

Roderick, in press). Here, we investigate factors 1 it provides the basis for tests of co-speciation underlying the biodiversity of these sap-feeding of planthoppers and their host plants,

insects in light of the history and hybridisation 2. it gives insight into the degree of host speci- of their plant hosts. We first test for the import- ficity, and ance of "hybrid-bridges" by comparing phylo- 3. it establishes the relatedness between genetic histories of silversword plants and plan- parental plant species associated with each thoppers. Secondly, we examine whether plant plant hybrid zone. hybrid zones are centres for biodiversity, by examining planthopper species diversity across Nesosydne planthoppers five hybrid zones. Many radiations of insects, including delpha- cid planthoppers, are endemic to Hawaii Methods (Simon, 1987; Howarth and Mull, 1992; Asquith, 1995; Eldredge and Miller, 1995; Silversword A lliance. Miller and Eldredge, 1996). In contrast to other The silversword alliance in Hawaii comprises delphacid planthoppers which are mainly grass 28 species, presumably with one common ances- feeders (Denno and Roderick, 1990; Denno et tor (Baldwin et al., 1991; Baldwin and Robi- al., 1991; Roderick, 1994), these planthoppers in chaux, 1995; Baldwin, in press). This radiation Hawaii feed on a wide variety of plant families is among the most well-studied of all plant lin- (Zimmerman, 1948; Swezey, 1954). The genus eages in Hawaii (see Wagner and Funk, 1995) Nesosydne in Hawaii contains at least 80 species with published works on ecology, physiology, (Zimmerman, 1948). Systematic studies of the systematics, conservation status, and hybridis- genus Nesosydne based on morphological ation (Carr, 1987). Carr (1985; 1990a; 1990b) characters are in progress (Asche, in press). has investigated the extent of hybridisation Nesosydne is now thought to be polyphyletic and between members of the silversword alliance represent several independent colonisations of and has documented that many, if not most, Hawaii with subsequent radiations within the members of the silversword alliance form natu- archipelago (M. Asche, personal communi- ral hybrids in the field. Hybrid zones differ in: cation). At least 15 Nesosydne species are 1. the plant species involved. reported to be found only on plant species in the 2. the relatedness of plant species that hybrid- Hawaiian silversword alliance (Zimmerman, ise, 1 948; Swezey, 1954); whether these species form 3. the range of ecological conditions occupied, a monophyletic group within the Nesosydne has and not been established based on morphological 4. the extent of overlap between the hybrids and characters, although the molecular data pre- one or both parental species. Hybrids and sented here support monophyly. Presently, no potential F's have been identified by leaf size other literature exists on abundance, life stages, PLANTHOPPERS ON THE HAWAIIAN S1LVERSWORD ALLIANCE 395 and seasonality of Nesosydne planthoppers on Planthopper history these different host plant species. A 441 base pair piece of cytochrome oxidase 1 was amplified using primers Cl-J-1751 'Ron' by R. Host plants associations and Cl-N-2191 'Nancy' (designed we Planthoppers were collected using an aspir- Harrison lab, Simon et al., 1994) To date, Nesosydne ator and sweeping on members of the silver- have examined individuals in six trans- sword alliance including five extensive hybrid species. The frequency of transitions and for Nesosydne species zones (see fig. 1). For these collections, R. versions was examined uncorrected Robichaux provided information on the likeli- using several genetic distances: Kimura (1980) 2- hood that individual plants were hybrids, pairwise percent divergence, (1993). A phylo- although in each hybrid zone there is a con- parameter, and Tamura-Nei both parsimony tinuum of backcrosses between likely parental geny was reconstructed using neighbor-joining species. Insects were frozen at — 80°C shortly (PAUP, Swofford, 1993) and 1993; MEGA, Kumar et after collection. (Phylip, Felsenstein,

A. Silversword Alliance B. Nesosydne Planthoppers

P.marginata

- Calif. Tarweeds N. koae N. naenae - Wilkesia spp. D.p. N. naenae D. paleata 62 K N. naenae 100 D.r. D. raillardioides N. naenae

D. plantaginea N. sp. D.I. D. knudsenii N. sp. 100

93 •c D. laxa N. chambers! 95 D.c. N. chambers'! D. imbricata 81 N. chambersi 80 D. laevigata D.s. N. chambersi 54 D. sherffiana 100 N. raillardiicola D. arborea 100 100 H N. raillardiicola 92 ID. ciliolata N. raillardiicola 58 scabra H D.m. N. raillardiicola D. 87 30 d N. raillardiicola D. platyphylla M N. raillardiicola 45 [D. menziesii M N. raillardiicola 2%Seq.Var \A. sandwicense N. argyroziphii A.s. 100 HE A. caliginis N. argyroziphii 99 N. eeke A.c. I &]

plant spec.es in the Hawaiian hypothesis of evolutionary relationships among selected Fieure l A Phvlogenetic total number of species in and Rob.chaux, 1 995). The sword aSnce showing major clades (after Baldwin ^ ve tarweed species only in the past shown), which likely evolved from a single California thehe radTatradiationon is "s_o (notnuua „ > ,, between pairs of taxa examined in the current study are

hlwToyTh^ parsimony of the his ory ol branches. B. Phylogenetic reconstruction by for rrdcular clades are shown below silversword alliance using DNA sequence variation m species found on members of the Hawaiian ^vovSw Numbers below branches denote bootstrap Branch lengths are proportional to sequence variation. mt DNZ£oT name. Boxes show host individuals of the same species are shown with the same rU D^orV(%) for nodes Multiple points corresponding to co-specat ,on between left. On both reconstructions, pZ affiliations with names to the a-f. planthoppers and their host plants are lettered \9t G. K. RODI-.KK K AND K C. ME 1/

al., 1993). Bootstrapping (n = 500) was used to holme, 1985). Transitions were approximately provide a level of confidence associated with double transversions for the range of genetic dis- each branch. Outgroups included another tances encompassing the Nesosydne planthop- Hawaiian Nesosydne planthopper. A', koae, pcrs feeding on the silversword alliance and both which feeds on Acacia Aoa(seeO'Connell, 1991) transitions and transversions increased linearly and Prakelisia marginata, a delphacid from the over this range of genetic distances. The linear continental United States (see Roderick, 1987; increase indicates that cytochrome oxidase I is a Denno et al., in press). good candidate for the evolutionary relation- ships in vestigated here, and that both transitions Tests hypotheses of for herbivore diversity and transversions contain useful information. If The hybrid-bridge hypothesis predicts that island age can be used as a rough time frame, host switching has occurred in the history of these genetic distances correspond to approxi- relationships between plants and their hosts". To mately 2 percent per million years (for island test this hypothesis, the phylogeny of the plan- ages see Carson and Claguc, 1995). Parsimony, thoppers was compared to that of the silver- neighbor-joining, and maximum likelihood sword alliance. The planthopper phylogeny was gave identical trees with similar bootstrap mapped onto Baldwin and Robiclniux's (1995) values; only the parsimony tree is shown here phylogeny for the silversword alliance based on (fig. IB). While there may be other species not sequences of nuclear ribosomal DNA. Events of examined here that would fall within those feed- hybrid switching and co-speciation, were recon- ing on the silversword radiation, this group was structed using TreeMap (Page, 1995b). A ran- supported by multiple synapomorphics that dis- domisation test (by "randomising" the plan- tinguished it significantly from N. koae. The thopper tree using TreeMap. Page, 1995b) was time frame suggested here for the diversification used to test significance of the observed level of of Nesosydne species that feed on the silver- co-speciation between planthoppcrs and their sword alliance is consistent with a single origin plant hosts. of these insects in Hawaii and corresponds to the To test of whether plant hybrid zones are age of the silversword alliance. centres for insect diversity, the number of plan- thopper species was examined across five hybrid /ones and compared to taxa and locations where hybridisation does not occur. Tests of hypotheses for herbivore diversity The reconstruction of planthopper and host Results plant phylogcnies resulted in 6 co-speciation events and no host-switching (fig. 1 ). It should be Host plains associations noted that the base of the tree for both host Nesosydne planthoppcrs have now been col- plants and planthoppers was unresolved. The lected on 13 out of 28 members of the silver- randomisation test shows that this number of sword alliance. Based on observations of both co-speciation events is significant (p < 0.01). adults and developing nymphs, many earlier Note that the method identifies a "co-specia- recorded associations are either spurious, site- tion" events for the different planthopper haplo- specific, or no longer occur (Roderick, unpubl. types collected on D. rai/Iardioides and I), pal- data). On five plant species no planthoppcrs eata. and D. ci/io/ata and D. scahra, even though were found. On the other plant species, only one the planthoppers collected on each species pair planthopper species occurred in sympatrv on the are identified as the same species (see fig. IB). same host plant species. Where a planthopper The reconstructed co-phylogenics provide no species occurred on more than one host species, evidence of recent host-switching. Research is the hosts were closely related and/or hybridise' ongoing to determine whether this pattern of These host records indicate that planthopper cospeciation is supported when all species of species in this group are highly host-specific to Hawaiian Nesosydne are included. either single plant species or closely related The reconstructed co-phylogenies provide species. no evidence to suggest that host-switching is important in the host associations documented Planthopper history here, as would be predicted by the "hybrid- The 441 base pair piece of cytochrome oxi- bridge" hypothesis. dase I amplified was one codon insertion longer Planthoppers were collected over 5 extensive than Drosophila yakuba (Clary and Wolsten- hybrid zones (fig. 2). Baldwin and Robichaux's PLANTHOPPERS ON THE HAWAIIAN SILVERSWORD ALLIANCE 397

Planthopper Use of Hybrids: EZ3 of planthoppers and members of the silversword alliance examined to date share parallel or co- D.paleata Hybrid (Fl) U.raillardioides <»**i phylogenies. Unfortunately, this result does not identify the processes that underlie the pattern (see Farrell and Mitter, 1993; Funk et al., 1995; Price, 1996). For example, co-phylogenies may may |nci processes that . DlatilhoDDers \ I arise though a number of include vicariance on one or both players, co- D.menziesii player in n< piantriappers evolution, or adaptation by one 3 1 response to the other. It is possible that major D.menziesii • .','. ,'.-.-, volcano for- .:. I > ... . . | —vwv.v , X] XXXXKOO0OO0 vicariant events, such as island and mation, have shaped both planthopper and Figure 2. Patterns of planthopper host plant use and plant phylogenies concurrently, and that plan- characteristics of five hybrid zones between members thopper adaptation to hosts may not be import- of the Hawaiian silversword alliance. Range of hybrid ant in explaining significant co-speciation. noted by shaded rec- use for each planthopper species Some biological observations can shed light tangles. In two hybrid zones, parental plant species on this issue. Research on other delphacid plan- were found with no planthoppers. See text for islands thoppers suggests that planthoppers can adapt to and degree of relatedness between plant taxa. closely related novel hosts. For example, numer- ous studies have shown that the rice brown plan- thopper, Nilaparvata higens, can overcome new "resistance genes" in rice in only a few genera- data provide information on the related- (1995) tions, but that host-switching is limited (for species involved in each of these ness of the plant review Roderick, 1994). Investigation of hybrid In two hybrid zones between hybrid zones. zones presented here demonstrate that the same plant species a single planthopper closely related planthopper species is not found on distantly each entire hybrid zone: D. pal- species spanned related hosts, despite sympatry and the existence (Kauai) and D. cilio- eata and D. raillardioides of intermediate hybrids. This pattern of host (Hawaii). In two other hybrid lata and D. scabra association indicates that limits exist to host relatives, a single zones between close plant adaptation by planthoppers. That single species species was associated with only planthopper feed on some closely related hosts but not on in one of these, the one parental plant species; others suggests that some closely related plant all identifiable planthopper also occurred on species have not diverged sufficiently to limit planthopper's hybrids while in the other, the planthopper distribution. These observations the parental species: D. range was restricted to are consistent with the hypothesis that diversity (Hawaii) and D. ciliolata and D. arborea of Nesosydne planthoppers parallels and likely (Maui). On inter- menziesii and D. platyphylla follows the diversity generated in the silver- sandwiceme and D. generic hybrids between A. sword alliance. species menziesii (Maui) both parental plant Data presented here do not yet adequately each and all supported one planthopper species address the role of hybridisation in the host species associated apparent hybrids shared the species in causing differentiation of planthopper of plan- with D. menziesii. No greater diversity populations and species. For the taxa examined in these hybrid thopper species was observed to date, hybrid bridges and host-switching do and locations zones than on silversword taxa not explain current patterns of host use by plant- four of five where hybridisation does not occur: hoppers; patterns of host use can be predicted planthopper hybrid zones supported only one entirely on the basis of co-speciation of the species. species and one supported only two planthoppers and hosts. Planthoppers are no the hypoth- Thus, the evidence does not support more diverse in number of species than plant- centres for esis that silversword hybrid zones are hoppers; collected on non-hybrids. Whether planthopper biodiversity. genetic diversity within planthopper species is greater in hybrid zones remains to be tested. Discussion More rigorous tests for the role of plant hybrid- the diversification of Nesosydne plan- The Hawaiian Nesosydne planthoppers are isation in on thoppers are now underway and include recipro- highly host specific, with each species feeding assess- signifi- cal transplant studies, population genetic one or a few closely related hosts. The ments of planthoppers across plant hybrid cance of co-speciation demonstrates that species 398 G. K.. RODERICK AND E. C. METZ

zones, and a more complete molecular analysis Carr. G.D., 1985. Monograph of the Hawaiian Madii- of Nesosydne planthoppers on members of the nae (Asteraceae): . , and 1-123. silversword alliance and other host plants. Wilkesia. Allertonia 4: Carr, G.D., 1987. Beggar's ticks and tarweeds: masters of . Trends in Ecology and Evol- Acknowledgments ution 2: 192-195.

Carr, G.D., 1990a. 1 1. Asteraceae. 9. Argyroxiphium We thank U. Bergstrom, A. E. Isagawa, Metz, DC Silversword. Pp. 258-262 in: Wagner, W.L., and B. Thorsby for technical assistance and B. Herbst, D.R. and Sohmer, S.H. (eds). Manual of Baldwin, R. Gillespie, S. Palumbi, R. the flowering plants of Hawaii. Vol. 1. University Robichaux, and F. Villablanca for helpful dis- of Hawaii Press: Honolulu. cussions and suggestions. For assistance in col- Carr. G.D., 1990b. 11. Asteraceae. 26. Dubautia lecting, we are grateful to R. Gillespie, R. Bart- Gaud. Na'ena'e, kupaoa. Pp. 292-308 in: lett (Maui Land and Pineapple), and R. Robi- Wagner, W.L., Herbst. D.R. and Sohmer, S.H. (eds). Manual the flowering plants Hawaii. chaux. This research was supported by the of of Vol. 1. University of Hawaii Press: Honolulu. University Research Council of the University Carson, H.L. and Clague, D.A., 1995. Geology and of Hawaii and the United States Fish and biogeography of the Hawaiian Islands. Pp. 14-29 Wildlife Service. Additional laboratory support in: Wagner, W.L. and Funk, V.A. (eds). Hawaiian was provided by grants from United States biogeography. Evolution on a hot spot archipelago. Department of Agriculture, California Depart- Smithsonian Institution Press: Washington. ment of Food and Agriculture, International D.C.

Rice Research Institute, Rockefeller Foun- Clary, D.O. and Wolstenholme, D.R., 1 985. The mito- dation, and Pacific Biomedical Research Center chondrial DNA molecule of Drosophila yakuba: of the University of Hawaii. nucleotide sequence, gene organization, and gen- etic code. Journal ofMolecular Evolution 22: 252- 271. References Denno, R.F. and Roderick, G.K., 1990. Population biology of planthoppers. Annual Review of Ento- Asche, M.. in press. Systematics and evolution of mology 35: 489-520. Hawaiian Fulgoroidea (Hemiptera). Pacific Denno. R.F.. Roderick. G.K.. Olmstead. K. Science. and Dobel, H.G., 1991. Density-related migration in Asquith, A., 1995. Evolution of Sarona (Heteroptera, planthoppers: the role of habitat persistence. Miridae). Pp. 90-1 20 in: Wagner, W.L. and Funk, American Naturalist 138: 1513-1541. V.A. (eds.), Hawaiian biogeography. Smithsonian Denno, R.F., Roderick, G.K., Peterson, Institution Press: Washington, D.C. M.A., Huberty, A.F., Dobel, H.G., Eubanks, M.D., Baldwin. B. and Robichaux. R., 1995. Historical bio- Losey, J.E. and Langellotto. G.A., in press. Habi- geography and ecology of the Hawaiian silver- tat persistence and dispersal in planthoppers. sword alliance (Compositae): new molecular phy- Ecological Monographs. logenctic perspectives. Pp. 259-287 in: Wagner, Dethier, V.G., 1954. Evolution of feeding W. and Funk, V. (eds), Hawaiian biogeographv. preferences in phytophagous insects. Evolution Smithsonian Institution Press: Washington 8: 833-854. DC. Ehrlich, P.R. and Raven, P.H., 1964. Butterflies and plants: a study in coevolution. Baldwin, B.G., in press. Adaptive radiation of the Evolution 18: 586- Hawaiian silversword alliance: congruence and 608. conflict of phylogenetic evidence from molecular Eldredge, L.G. and Miller, S.E., 1995. How many- and non-molecular investigations. In: Givnish, T. species are there in Hawaii? Bishop Museum and Sytsma, K.J. (eds), Molecular evolution and Occasional Papers 41: 3-18. adaptive radiation. Cambridge University Press: Erwin. T.L., 1982. Tropical forests: their richness in Cambridge. Coleoptera and other arthropod species. Coleop- Baldwin, B.G., Kyhos, D.W., Dvorak, J. and Carr, terists Bulletin 36: 74-75. G.D.. 1991. Chloroplast DNA evidence for a Erwin. T.L., 1986. The tropical forest canopy: the North American origin of the Hawaiian silver- heart of biotic diversity. Pp. 123-129 in: Wilson, sword alliance (Asteraceae). Proceedings of the E.O. (ed.). Biodiversity. National Academy of National Academy of Sciences (USA) 88: 1840- Sciences: Washington, D.C. 1843. Farrell, B.D. and Mitter, C, 1 993. Phylogenetic deter- Brooks. D.R., 1979. Testing the context and extent of minants of insect/plant community diversity in: host-parasite coevolution. Systematic Zoology 28- Ricklefs, R.E. and Schluter, D. (eds). Species 299-307. diversity in ecological communities. UniversiU of Brooks, D.R.. 1988. Macroevolutionary comparisons Chicago Press: Chicago. of host and parasite phylogenies. Annual Review Felsenslein, J., 1993. Phylip. University of Washing- of Ecology and Systematics 19: 235-259. ton. PLANTHOPPERS ON THE HAWAIIAN SILVERSWORD ALLIANCE 399

Floate. K.D. and Whitham, T.G., 1993. The 'hybrid O'Connell, C, 1991. Acoustic communication and sex- bridge' hypothesis: host shifting via plant hybrid ual selection in a group ofHawaiian planthoppers swarms. American Naturalist 141: 651-662. (Homoptera: Delphacidae). Masters thesis. Uni- Friar, E.A., Robichaux, R.H. and Mount, D.W., 1966. versity of Hawaii, Manoa. Molecular genetic diversity following a popu- Page, R.D.M., 1995a. Parallel phylogenies: recon- lation crash in the Endangered Mauna Kea silver- structing the history of host-parasite assmeblages. sword. Argyroxiphium sandwicense spp. sandwi- Cladistics 10: 155-173. cense (Asterceae). Molecular Exology 5; 687-691. Page, R.D.M., 1995b. TreeMap. University of Oxford: Funk, D.J., Futuyma, D.J.. Ort, G. and Meyer, A., Oxford. 1995. A history of host associations and evol- Price, P.W., 1996. Biological evolution. Saunders utionary diversification for Ophrella (Coleoptera: College Publishing: Fort Worth, Texas. Chrysomelidae): new evidence from mitochon- Roderick, G.K., 1987. Ecology and evolution ofdisper- drial DNA. Evolution 49: 1008-1017. sal of a Pacific salt marsh insect. Ph.D. thesis. Futuyma, D.J., 1983a. Evolutionary interactions University of California, Berkeley. among herbivorous insects and plants. Pp. 207- Roderick, G.K.. 1994. Genetics of host plant adap- 231 in: Futuyma, D.J. and Slatkin, M. (eds.), tation in delphacid planthoppers. Pp. 55 1-570 in: Coevolution. Sinauer: Sunderland, Mass. Denno. R.F. and Perfect, T.J. (eds). Planthoppers: Futuyma, D.J., 1983b. Selective factors in the their ecology and management. Chapman and evolution of host choice by phytophagous insects. Hall: London. Pp. 227-244 in: Ahmad, S. (ed.). Herbivorous Roderick, G.K.. in press. Herbivorous insects and the insects. Academic Press: New York. Hawaiian silversword alliance: co-evolution or Futuyma, D.J., Reese, M.C. and Funk, D.J., 1995. co-speciation? Pacific Science. Genetic constraints on macroevolution: the evol- Simon, C, 1987. Hawaiian evolutionary biology: an ution of host affiliation in the leaf beetle genus introduction. Trends in Ecology and Evolution 2: Ophraella, Evolution 49: 797-809. 175-178. Hafner, M.S.. Sudman, P.D., Villablanca, F.X., Spra- Simon, C Fratti, F., Beckenbach, A., Crespi, B„ Liu, dling, T.A., Demastes, J.W. and Nadler, S.A., H. and Flook, P., 1994. Evolution, weighting, and 1994. Disparate rates of molecular evolution in phylogentic utility of mitochondrial gene cospeciating hosts and parasites. Science 265: sequences and a compilation of conserved poly- 1087-1090. merase chain reaction primers. Annals of the Howarth. F.G. and Mull, W.P., 1992. Hawaiian Entomological Society of America 87: 651-701. insects and their kin. University of Hawaii Press: Strauss, S., 1994. Levels of herbivory and parasitism Honolulu. in host hybrid zones. Trends in Ecology and Evol- Kimura, M., 1980. A simple method for estimating ution 9: 209-214. evolutionary rate of base substitutions through Strong, D.R., Lawton, J.H. and Southwood. R., 1984. comparative studies of nucleotide sequences. Insects on plants. Blackwell: Oxford. Hawaii. Journal of Molecular Evolution 16: 111-120. Swezey, O.H., 1954. Forest entomology in Honolulu. Kumar, S., Tamura. K. and Nei, M., 1993. MEGA: Bishop Museum: Molecular Evolutionary Genetic Analysis. The Swofford, D.L., 1993. PAUP. Phylogenetic analysis- Pennsylvania State University: University Park, using parsimony. Sinauer: Sunderland, Mass. ofthe PA. Tamura. K. and Nei, M„ 1 993. Estimation num- in the control Labandeira. C.C. and Sepkoski, J.J.. Jr., 1993. Insect ber of nucleotide substitutions in humans and diversity in the fossil record. Science 261: 310- region of mitochondrial DNA Evolution 6: 315. chimpanzees. Molecular Biology and Miller, S.E. and Eldredge, L.G.. 1996. Numbers of 447-459. revolutionary process. Hawaiian species: supplement I. Bishop Museum Thompson, J.N., 1994. The Chicago. Occasional Papers 45: 8-17. The University of Chicago Press: Funk, V., 1995. Hawaiian biogeo- Mitter, C. and Brooks, D.R.. 1983. Phylogenetic Wagner, W.L. and hot spot archipelago. Smith- aspects of co-evolution. In: Futuyma, D.J. and graphy: evolution on a Press: Washington, D.C. Slatkin, M. (eds), Coevolution. Sinauer: Sunder- sonian Institution P.A. and Potts. B.M.. 1994. land, Mass. Whitham, T.G., Morrow, Plant hybrid zones as centers of biodiversity: the Mitter, C, Farrell, B. and Futuyma. D.J., 1991. Phy- of two endemic Tasmanian logenetic studies of insect-plant interactions: herbivore community Oecologia 97: 481-490. insights into the genesis of diversity. Trends in eucalypts. E.O., 1988. Biodiversity. National Academy Ecology and Evolution 6: 290-293. Wilson, Washington, D.C. Mitter, C. Farrell, B. and Wiegmann, B., 1988. The Press: ultimately biodiverse phylogenetic study of adaptive zones: has phyto- Wilson, E.O., 1996. Insects: the P. in: International Congress of phagy promoted insect diversification? A merican animals. XVI XX Proceedings: Firenze, Italy. Naturalist 132: 107-128. Entomology Zimmerman, E.C., 1948. Homoptera: Auchenorhycha. Moran, N., 1988. The evolution of host-plant alter- University of Hawaii Press: Honolulu. nation in aphids: evidence for specialization as a dead end. American Naturalist 132: 681-706.