sign in sign up
<<
Home , Acanthococcus (bug), Aclerdidae, Adelges, Asterolecanium, Beesoniidae, Coccus (insect)

MOLECULAR PHYLOGENETICS AND EVOLUTION Molecular Phylogenetics and Evolution 25 (2002) 43–52 www.academicpress.com

A preliminary phylogeny of the scale (: : Coccoidea) based on nuclear small-subunit ribosomal DNA

Lyn G.Cook, * Penny J.Gullan, 1 and Holly E.Trueman 2

School of Botany and Zoology, The Australian National University, Canberra, ACT 0200, Australia

Received 9 July 2001

Abstract

Scale insects (Hemiptera: Sternorrhyncha: Coccoidea) are a speciose and morphologically specialized group of plant-feeding bugs in which evolutionary relationships and thus higher classification are controversial.Sequences derived from nuclear small-subunit ribosomal DNA were used to generate a preliminary molecular phylogeny for the Coccoidea based on 39 representing 14 putative families.Monophyly of the archaeococcoids (comprising , sensu lato, and Phenacoleachia) was equivocal, whereas monophyly of the neococcoids was supported., represented by Puto yuccae, was found to be outside the remainder of the neococcoid clade.These data are consistent with a single origin (in the ancestor of the neococcoid clade) of a chromosome system involving paternal genome elimination in males.Pseudococcidae () appear to be sister to the rest of the neococcoids and there are indications that (soft scales) and (lac scales) are sister taxa.The (felt scales) was not recovered as a monophyletic group and the eriococcid Eriococcus sensu lato was polyphyletic. Ó 2002 Elsevier Science (USA).All rights reserved.

1. Introduction comprise the Margarodidae sensu lato (as in Morrison, 1928), the Ortheziidae (Koteja, 1974a, 1996), the The scale insects (Hemiptera: Sternorrhyncha: Coc- Carayonemidae (Kozaar and Konczne Benedicty, 2000), coidea) are obligatory plant parasites which are found in and sometimes the (Danzig, 1980; all terrestrial zoogeographical regions except Antarctica. Koteja, 1996).However, most of the morphological There are estimated to be more than 7700 species of characters that define the archaeococcoids (abdominal coccoids assigned variously to 20 or more families (Ben- spiracles, compound eyes in males, and an XX–X0 Dov et al., 2001; Gullan and Kosztarab, 1997; Koteja, chromosome system) are plesiomorphies that occur 1974a).Often the families are divided into two major, more widely in the Hemiptera and thus paraphyly of informal groups which sometimes are treated as super- archaeococcoids is likely (e.g., Foldi, 1997). families, the archaeococcoids or archaeococcids The neococcoids, which comprise all of the other ( ¼ Orthezioidea) and the neococcoids or neococcids families and most of the species of scale insects, are ( ¼ Coccoidea sensu stricto) (Borchsenius, 1950; Danzig, treated currently as a monophyletic group.It has been 1980; Kosztarab and Kozaar, 1988; Kosztarab, 1996; defined by synapomorphies such as a chromosome sys- Koteja, 1974a, 1996; Miller, 1984).The archaeococcoids tem involving paternal genome elimination (PGE) (Danzig, 1980; Nur, 1980), needle-like apical setae on * Corresponding author.Fax: +61-2-61255573. the labium (Koteja, 1974b, 1996), shared structural and E-mail address: [email protected] (L.G. Cook). developmental features of the ovaries (Szklarzewicz, 1 Present address: Department of Entomology, University of Cali- 1998), and the absence of abdominal spiracles (Morri- fornia, Davis, CA 95616-8584, USA. 2 son, 1928).Among the neococcoids sensu lato, only Puto Present address: Department of Biology, Imperial College of Science, Technology, and Medicine, Imperial College Rd., London, (XX–X0), sometimes considered part of the Pseudo- UK. coccidae (e.g., Danzig, 1980; Kosztarab, 1996; Miller

1055-7903/02/$ - see front matter Ó 2002 Elsevier Science (USA).All rights reserved. PII: S1055-7903(02)00248-8 44 L.G. Cook et al. / Molecular Phylogenetics and Evolution 25 (2002) 43–52 and Miller, 1993), the eriococcid genus Lachnodius for inclusion as outgroups.The main aims of the (2N–2N), and the (2N–2N) do not ex- present study were to: (1) provide a hypothesis of the hibit chromosome systems with PGE (Nur, 1980).In relationships among the higher taxa of the Coccoidea, PGE systems, both male and female zygotes are diploid (2) test the monophyly of the Eriococcidae and Erio- and there are no sex chromosomes.Instead, the pater- sensu lato, and (3) provide new data to assist nally inherited haploid set of chromosomes becomes with interpreting coccoid evolution and reassessing heterochromatic, or is lost, in males during early em- higher classification. bryogenesis. Relationships among coccoid families are largely unknown or not supported well by morphological data 2. Materials and methods (Foldi, 1997; Gullan and Kosztarab, 1997; Kosztarab, 1996).Individual families are often well characterized by 2.1. Specimens and DNA extraction autapomorphies but there is controversy concerning the monophyly of some of the traditionally recognized The coccoid taxa used, their current taxonomic families such as the Margarodidae sensu lato (see Foldi, classification (following Ben-Dov et al., 2001), and col- 1997; Gullan and Sjaarda, 2001; Miller, 1984) and the lection localities are shown in Table 1.We obtained Eriococcidae (Cox and Williams, 1987; Miller and sequences from 39 scale species belonging to 14 Gimpel, 2000), which are defined mostly by symplesio- families, including all major families and several species- morphies.In the past, Eriococcidae has subsumed sev- poor families whose relationships have been problem- eral other currently recognized families or has been atical.The monophyly of each of the three largest broken into a number of less inclusive taxa (for review families, Coccidae, , and Pseudococcidae, see Miller and Gimpel, 2000).Within the Eriococcidae, has not been questioned and thus only a few species of even the definition of the type genus, Eriococcus, has each of these groups were included. oscillated from a restricted concept (e.g., Borchsenius, The cuticles of adult females were slide-mounted to 1950; Miller and Gimpel, 1996) to a broad one (e.g., allow identification and to provide voucher specimens. Miller and Gimpel, 2000; Williams, 1985). Vouchers are deposited in the Australian National In- A robust phylogeny of the Coccoidea is required to sect Collection, CSIRO Entomology, Canberra and in stabilize classification and to test hypotheses about scale the Bohart Museum of Entomology at the University of insect biology, including biogeography, and evolution, California, Davis.DNA was extracted from fresh or in particular the evolution of chromosome systems and ethanol-preserved specimens using the salting-out host–parasite relationships, the origins of gall induction, method of Sunnucks and Hales (1996).A chloroform and the correlates of radiation events.To date, few wash was performed prior to precipitation with ethanol studies have attempted to reconstruct the phylogeny of if solids or excessive pigments were present. the whole group other than through intuitive phylo- grams (Boratynski and Davies, 1971; Herrick and Seger, 2.2. PCR and sequencing 1999; Kosztarab, 1996; Kosztarab and Kozaar, 1988; Nur, 1980) or preliminary cladistic analysis of mor- The 50 region of the nuclear SSU rDNA was ampli- phological data. fied using the primers 2880 (50-CTGGTTGATCCTGCC Cladistic studies either have had limited taxonomic AGTAG-30) (Tautz et al., 1988) and B- (50-CCGCGGC scope (e.g., Miller, 1984; Miller and Hodgson, 1997) or TGCTGGCACCAGA-30) (von Dohlen and Moran, have not examined node support in a critical manner 1995).PCRs (25 ll) contained 5 pmol of each primer, (Foldi, 1997).This lack of a reliable phylogenetic esti- 10 mM Tris–Cl (pH 8.3), 3 mM MgCl2,50mMKC1, mate is due primarily to the very reduced and highly 0.2 mM each dNTP, 1 unit of Taq-polymerase (Fisher modified morphologies of adult female scale insects, Taq Fl; Biotech), and 2 ll of template.Amplification which leads to there being few phylogenetically infor- was carried out in a Corbett Research FTS-960 thermal mative features that can be scored for taxa across all sequencer using a 45-s denaturation at 94 °C and a 30-s recognized families. extension at 72 °C and using a 5 °C stepdown program This is the first phylogenetic hypothesis for Coc- for annealing with the first cycle at 65 °C and the final 35 coidea based on molecular data.Because of the ap- cycles at 45 °C.PCR products were precipitated and parent age of some of the older lineages within the purified using ammonium acetate:ethanol (1:10) fol- Coccoidea and the likelihood of substitutional satura- lowed by a 70% ethanol wash.Target products from tion in faster-evolving genes, the nuclear small subunit PCR which contained minor bands were excised from a ribosomal RNA gene (SSU rRNA) was considered 1% agarose gel following electrophoresis and purified most appropriate.SSU rRNA data were available for using the BresaClean DNA purification kit (GeneWorks other sternorrhynchans from previous studies (Camp- Cat.No.BT-3000) following the manufacturer’s bell et al., 1994, 1995; von Dohlen and Moran, 1995) instructions. L.G. Cook et al. / Molecular Phylogenetics and Evolution 25 (2002) 43–52 45

Table 1 Coccoidea used in this study Species Current family Source Tissue Length of GenBank classification SSU rDNAa Accession No. Araucaricoccus queenslandicus Brimblecomb Margarodidae Australia wf 563 Callipappus sp.Margarodidae Australia o,e 606 purchasi Maskell Margarodidae Australia e 593 Phenacoleachia zealandica (Maskell) Phenacoleachiidae New Zealand wf 601 australiensis Kozaar & Konczne Ortheziidae Australia wf 564 Benedicty urticae (Linnaeus) Ortheziidae Hungary wf 590 Puto yuccae (Coquillet) Putoidae USA wf 643 neobrevipes Beardsley Pseudococcidae von Dohlen and 598 U20429 Moran (1995) sp.Pseudococcidae Africa wf 596 callitris Williams Pseudococcidae Australia wf 598 Paralecaniini sp.Coccidae Australia wf 601 Austrolecanium sassafras Gullan & Hodgson Coccidae Australia wf 545 corni (Bouchee) Coccidae Australiab wf 599 Myzolecanium sp.Coccidae Australia wf 598 Austrotachardia angulata (Froggatt) Kerriidae Australia wf 624 Austrotachardia sp.Kerriidae Australia wf 624 Allokermes sp.1 USA wf 605 Allokermes sp.2 Kermesidae USA wf 605 aurantii Maskell Diaspididae Campbell et al. 569 U06475 (1994) Diaspididae sp.Diaspididae Australia wf 573 Maskellia globosa Fuller Diaspididae Australia wf 571 minus Lindinger USA wf 574 Bambusaspis bambusae (Boisduval) Asterolecaniidae Philippines wf 581 Frenchia casuarinae Maskell Asterolecaniidae Australia wf 594 Ascelis praemollis Schrader Eriococcidae Australia o, e 590 Cylindrococcus spiniferus Maskell Eriococcidae Australia o, e 584 Eriococcus aceris (Signoret) Eriococcidae Hungary wf 588 Eriococcus buxi (Boyer de Fonscolombe) Eriococcidae France wf 592 Eriococcus coccineus Cockerell Eriococcidae USA wf 584 Eriococcus eucalypti Maskell Eriococcidae Australia wf 572 Eriococcus sp.‘‘Hakea’’ Eriococcidae Australia wf 588 Eriococcus spurius (Modeer) Eriococcidae USAb wf 586 Lachnodius sp.Eriococcidae Australia o 584 Madarococcus viridulus Hoy Eriococcidae New Zealand wf 566 Opisthoscelis mammularis Froggatt Eriococcidae Australia wf 578 sp.Beesoniidae Malaysia wm 580 confusus (Cockerell) Dactylopiidae von Dohlen and 608 U20402 Moran (1995) Dactylopius austrinus De Lotto Dactylopiidae Australiab e 608 Stictococcus sjostedti Cockerell Stictococcidae Africa wf 606 Note.wf, whole female; o, ovary; e, embryos or eggs; wm, whole male. a Length (bp) of the 50 region of ssu rDNA amplified with primers 2880 and B-. b Introduced species, not native to collection locality.

Sequencing used ABI (Perkin–Elmer) BigDye termi- coccus buxi and two individuals of E. aceris to provide nator chemistry following the manufacturer’s instruc- confirmation of the SSU rDNA sequence for each of tions using an ABI Prism 377 automated DNA sequencer. these taxa because their placement is critical to eriococcid All DNA fragments were sequenced in both the forward . and reverse directions.Amplifications of genomic DNA from were run under different PCR con- 2.3. Sequence and phylogenetic analyses ditions (varying annealing temperature and magnesium concentrations) to determine whether different copies of Sequences were edited using Sequencher 3.0 (Gibbs SSU rDNA were present and preferentially amplified and Cockerill, 1995).Three sequences (Table under the different conditions.Additionally, DNA 1) and sequences from the following seven , rep- was extracted separately from three individuals of Erio- resenting the three aphidoid families (Blackman and 46 L.G. Cook et al. / Molecular Phylogenetics and Evolution 25 (2002) 43–52

Eastop, 2000), were obtained from GenBank (Benson et tests (Faith, 1991).Groupings assumed to be mono- al., 2000) for inclusion as outgroups: (Acyr- phyletic under previous hypotheses but not appearing in thosiphon pisum X62623, U27819; Melaphis rhois U27820; the unweighted MP tree were assessed by determining Mindarus kinseyi U27821; Rhopalosiphon padi U27825; the number of extra steps required for that clade to Schizaphis graminum U27826), ( sp. appear and tested using T-PTP.The a priori groups U20400), and ( notabilis tested were archaeococcoids comprising families Mar- U20398). garodidae, Ortheziidae, and Phenacoleachiidae; Puto A preliminary alignment was made using CLU- within, or as sister to, the Pseudococcidae (Danzig, STAL_X (Thompson et al., 1997) with further adjust- 1980; Miller and Miller, 1993); Puto within, or as sister ments made by eye with reference to two predicted to, neococcoids; Phenacoleachia as sister to Pseudococ- secondary structure models for SSU rRNA: the model cidae (Cox, 1984; Miller and Miller, 1993); Eriococcidae for A. pisum (Kwon et al., 1991) and the generalized sensu lato;andEriococcus sensu lato (e.g., Miller and eukaryotic SSU rRNA model of Van de Peer et al. Gimpel, 2000; Williams, 1985). (2000).Regions of conflict between the secondary structure model and each of the scale insect sequences 2.3.2. Neighbor-joining (NJ) were compared using minimum energy RNA folds ob- Models of varying complexity (base frequencies, sub- tained for each helix using Mfold (Mathews et al., 1999; stitutional classes, percentage of invariant sites (I), site- Zuker et al., 1999). Some highly variable regions, par- specific rates (C)) were compared using log likelihood ticularly the E10 region that is highly expanded in scale ratio tests (Goldman, 1993; Yang and Nielsen, 1998).The insects and some related taxa, were not able to be simplest model that did not differ significantly from more aligned unambiguously across all taxa even after con- complex models was chosen.Bases occurred in equal sideration of secondary structure and were excluded frequencies and the model approximated a Kimura two- from further analyses. parameter model (K2P þ C) (Kimura, 1980) with site- specific rates (C shape parameter ¼ 0.16). 2.3.1. Maximum-parsimony (MP) Heuristic searches were performed using PAUP* 2.3.3. Maximum-likelihood (ML) 4.0b4a (Swofford, 2000) and comprised 100 random ML was run only with a reduced taxon set because of addition sequence starting trees, TBR branch swapping, the computational difficulties involved with including all and no maxtrees restrictions.Several weighting schemes taxa.An exemplar taxon was chosen from each of the were used: (1) equal weights; (2) transversions weighted clades with strong support in MP and NJ analyses.To twice that of transitions; (3) stem sites weighted half that test whether the unexpected inferred relationship be- of single-stranded regions, based on an assumption that tween E. buxi, C. spiniferus, the beesoniid, and S. sjos- pairing sites covary and thus represent a single character tedti was the result of long-branch attraction, each of (e.g., Soltis et al., 1999); and (4) partition-dependent these taxa were included in ML analyses.ML was run weighting in which sites were assigned to one of four under three evolutionary models: (1) Jukes–Cantor categories for each of the two secondary structure (Jukes and Cantor, 1969), (2) K2P assuming equal rates models: double-stranded helix stem regions, single- for all sites, and (3) K2P þ C with parameters estimated stranded hairpin loops, unpaired single-stranded regions from the data. within a double-stranded helix, and other unpaired single-stranded regions.Variable sites within the paired regions of double-stranded helices were further catego- 3. Results rized as fully covarying if a change conserved a G–C or A–U pairing or as a neutral change if variation involved 3.1. Sequence variation a G–C and G–U pairing or a A–U and G–U pairing. The partition-dependent weighting scheme was run for The length of the edited fragment ranged from 545 bp the two secondary structure models with weighting as in Austrolecanium sassafras to 643 bp in Puto yuccae follows: fully covarying sites ¼ 1, neutral change ¼ 1, (Table 1).The aligned region used in analyses comprised hairpin loops ¼ 1, and all other sites ¼ 2. 507 bp of which 85 sites were parsimony informative. Rate heterogeneity tests were conducted with Phyltest There was no base compositional bias among taxa. (Kumar, 1996).Bootstrap tests (Felsenstein, 1985) were Only a single length product of SSU rDNA was am- conducted using 1000 replicates and the same settings as plified from I. purchasi under the different PCR condi- those in the original analyses but including only one tions.Identical sequences were obtained from different representative of taxa with identical sequences in the individuals of each of I. purchasi, E. buxi,andE. aceris. aligned regions.Internal nodes were assessed using No taxa of scale insects shared identical complete nu- Bremer support (decay index) (Bremer, 1988) and to- cleotide sequences although, in the aligned regions used pology-dependent permutation tail probability (T-PTP) in phylogenetic analyses, E. aceris was identical to L.G. Cook et al. / Molecular Phylogenetics and Evolution 25 (2002) 43–52 47

E. spurius and E.sp.ex Hakea, Opisthoscelis mammularis as sister group to the gall-inducing taxa Opisthoscelis, was identical to Lachnodius sp., and Allokermes sp.1 and Lachnodius, and Ascelis.The Dactylopiidae fell within Allokermes sp.2 had identical sequences.Expansion in sensu Miller and Gimpel (1996) (repre- the E10 accounted for most of the extra length in the sented in the analysis by E. aceris, E. coccineus, E. sp.ex P. yuccae sequence. Hakea, and E. spurius) in all analyses. There was up to 5.5% divergence ðK2P þ CÞ among One of the most strongly supported relationships coccoid taxa currently recognized as margarodids, among the scale insects was the clade comprising the equivalent to that between each of the margarodids and beesoniid, S. sjostedti (Stictococcidae), and the two the neococcoid taxa.There was 4% divergence within the eriococcid taxa E. buxi and Cylindrococcus spiniferus Coccidae, up to 3.5% within several eriococcid clades and (hereafter called the BSE clade).There was strong the Asterolecaniidae, 2% between the two ortheziids, bootstrap, Bremer (Fig.1), and character support (13 1.5% within Pseudococcidae, and only 0.4% divergence inferred substitutions) for this relationship.The sister among the diaspidids sampled.There is about 2% di- relationship between S. sjostedti and the beesoniid was vergence between the phylloxerid and the aphidids and not as well supported by the character changes along the only 0.4% variation within the Aphididae included. branch as all appeared homoplasious with respect to at least two other scale insect lineages.In addition, the 3.2. Phylogenetic relationships SSU rDNA of S. sjostedti and the beesoniid appeared to be evolving significantly faster than other taxa Trees recovered by MP and NJ methods, using differ- (P 6 0:01Þ.Therefore, MP analyses were run with and ent weighting schemes, were mainly congruent (Fig.1). without these two taxa.The tree topology was identical The archaeococcoids did not form a monophyletic group among other taxa and the same six MP trees were found in any analysis but their monophyly could not be rejected in both searches.Searches conducted without E. buxi (T-PTP reverse ¼ 0.998). The archaeococcoid polytomy and C. spiniferus failed to identify any alternative in the MP strict consensus (Fig.1) is the result of differing placement of S. sjostedti and the beesoniid.The BSE resolutions among the most parsimonius reconstructions. clade was recovered also in ML analyses despite the A monophyletic Margarodidae (Araucaricoccus, I. pur- tendency for long branches to be repulsed under this chasi,andCallipappus sp.) was recovered only in one of model (Felsenstein, 1978; Siddall and Whiting, 1999). the six unweighted MP trees and not in weighted MP or distance-based trees.The two ortheziids formed a clade in all methods. Phenacoleachia appeared as sister to the rest 4. Discussion of the scale insects in the ML analyses (Fig.2) and three of the MP trees and as sister to different groups of marg- 4.1. Chromosome systems arodid taxa in the other three MP trees. The placement of Putoidae was ambiguous. P. yuccae The scale insects display a very diverse range of sex did not fall among the other neococcoid taxa in any of determination mechanisms and chromosome behaviors, the analyses.It was placed as sister to the neococcoids or collectively called chromosome systems (Nur, 1980). variously as sister to groups of archaeococcoid taxa, These include XX–X0 sex determination; hermaphrod- including the ortheziids. itism; PGE; arrhenotokous, thelytokous, and deutero- The neococcoids (excluding Puto) were recovered as a tokous parthenogenesis; and a system (2N–2N) in which monophyletic group by all methods.Pseudococcidae there are no apparent chromosomal differences between appeared as sister to the other neococcoids under all the sexes. models and weighting schemes used.There was support It is likely that the XX–X0 system is the ancestral for a sister relationship between Coccidae and Kerrii- type for the scale insects (Nur, 1980) because this system dae.A monophyletic Eriococcidae was not recovered is present in some of the putatively oldest scale insect under any method or model and an a priori hypothesis lineages and occurs in aphids (Blackman, 1987) and of monophyly of the eriococcids was rejected (T-PTP other sternorrhynchans (White, 1973).PGE is thought reverse ¼ 0.018). The relative positions of the Diaspidi- to have evolved from an XX–X0 system (Nur, 1980), dae and Asterolecaniidae differed between MP and NJ perhaps via maternal control of sex and the suppression methods but both fell among the eriococcid taxa in all of recombination in males by X-chromosome meiotic analyses. drive (i.e., a type of 2N–2N system) (Haig, 1993). There The monophyly of the eriococcid genus Eriococcus has been some doubt as to whether there have been any sensu lato was rejected (T-PTP reverse ¼ 0.001). Taxa reversions from a PGE system to an XX–X0 system currently assigned to Eriococcus fell into three clades (Herrick and Seger, 1999; Nur, 1980), primarily because that each included taxa currently assigned to other Puto (XX–X0) has been thought to be closely related to genera.There was a sister relationship between Erio- Pseudococcidae (Miller and Miller, 1993), a group with coccus eucalypti and Madarococcus, and both appeared a PGE system (Nur, 1980). 48 L.G. Cook et al. / Molecular Phylogenetics and Evolution 25 (2002) 43–52

Fig.1.Strict consensus of the six topologies obtained from unweighted maximum-parsimony analyses (length ¼ 229, CI ¼ 0.515, RI ¼ 0.775). Thick branches are those which are present also in all best trees under all other models (maximum-parsimony using partition-dependent weighting schemes based on the two secondary structure models, a transition:transversion weighting or a simple stem:loop weighting; neighbor-joining using a Kimura two-parameter model with gamma shape parameter ¼ 0.16). Bootstrap values (1000 replications) above 50 are shown above nodes, Bremer values are shown below nodes.All nodes have significant T-PTP support ðP 6 0:01Þ.Current family placements of the scale insect taxa are indicated on the right.The arrowed node represents the neococcoids (with Puto excluded).A single origin of paternal genome elimination (PGE) is inferred, with two reversions to a 2N–2N chromosome system indicated as crosses.The nomenclature for species of Eriococcidae follows Miller and Gimpel (2000) in using Eriococcus sensu lato, rather than recognizing Acanthococcus as a separate genus.

The phylogenetic estimates based on SSU rRNA are chromosome systems, are clearly embedded within the consistent with a single origin of PGE in scale insects (at PGE clade, supporting the hypothesis that Lachnodius the base of the neococcoid clade) and no reversions to the has undergone a reversal of PGE to a 2N–2N chromo- plesiomorphic XX–X0 sex determination system because some system (Haig, 1993; Herrick and Seger, 1999; Nur, Puto appears to have diverged prior to the origin of PGE. 1980).Thus, there appear to have been two independent Lachnodius and Stictococcus, both taxa with 2N–2N reversals from PGE to a 2N–2N system because Sticto- L.G. Cook et al. / Molecular Phylogenetics and Evolution 25 (2002) 43–52 49

Fig.2.Single topology obtained using maximum-likelihood.Thicker branches are those obtained under all ML models used (Jukes–Cantor, Kimura two parameter with transition:transversion ratio estimated from the data and assuming equal rates per site, and K2P with both transition:trans- version ratio and gamma estimated from the data).Current family placements of the scale insect taxa are indicated on the right.A single origin of paternal genome elimination (PGE) is inferred, with two reversions to a 2N–2N chromosome system indicated as crosses. coccus and Lachnodius do not cluster together and each is rodidae and are thus in accord with cladistic analysis of most closely related to taxa retaining a PGE system. morphological data in which relationships among the margarodid higher taxa are unresolved (Gullan and 4.2. Taxonomic implications Sjaarda, 2001).However, the extent of molecular dif- ferentiation among the margarodid lineages, relative At higher taxonomic levels, these SSU rDNA data to that among recognized neococcoid families, favors are largely congruent with those derived from morpho- Koteja’s (1974a, 1996, 2000a,b, 2001) suggestion that logical studies. the currently recognized margarodid subfamilies war- rant family status. 4.2.1. Archaeococcoids The enigmatic Phenacoleachia zealandica does not These molecular data are uninformative with regard appear to be allied with the Pseudococcidae as suggested to the monophyly of the archaeococcoids and Marga- by Cox (1984) and Miller and Miller (1993).The dis- 50 L.G. Cook et al. / Molecular Phylogenetics and Evolution 25 (2002) 43–52 tinctness of P. zealandica indicates that there is no rea- (Ben-Dov et al., 2001; Miller and Gimpel, 2000), may son to change the family status of Phenacoleachiidae.It need to be subdivided.There is very strong support for may be sister to the rest of the scale insects or belong the type species of Eriococcus, E. buxi, forming a clade among the archaeococcoid lineages, as suggested by with C. spiniferus, an Australian gall-inducing eriococ- Danzig (1980) and Koteja (1974a).An origin of the cid, and species from two other families (Beesoniidae Phenacoleachiidae lineage early in scale insect evolution and Stictococcidae) (BSE clade).It has been suggested is consistent with the occurrence in Phenacoleachia of previously that beesoniids may be affiliated with the plesiomorphic features (Gullan and Cook, 2002), such eriococcids (e.g., Takagi, 1992) based on the similarity as abdominal spiracles and a labial apical organ, which of tubular ducts, and both Cylindrococcus and some occur also in other archaeococcoids but not neococc- beesoniids induce bract galls on their host plant.It has oids. not been suggested previously that the four taxa of the The phylogenetic placement of Puto remains unre- BSE clade form a natural group to the exclusion of other solved.There is no support for a close relationship be- taxa traditionally assigned to the Eriococcidae. tween P. yuccae and the pseudococcid taxa sampled, The close relationship between Cylindrococcus and E. consistent with findings that endocytobiosis in Puto is buxi to the exclusion of other species of Eriococcus very different from that of pseudococcids (Buchner, suggests that Eriococcus sensu lato (as treated by Miller 1965; Tremblay, 1989). and Gimpel (2000) and Williams (1985)) is polyphyletic. This has taxonomic and systematic implications because 4.2.2. Neococcoids E. buxi is the designated type species of Eriococcus The apparent sister relationship of the Pseudococci- (Melville, 1982) and of the Eriococcidae. dae to the rest of the neococcoids (excluding Putoidae) The other species currently assigned to Eriococcus that has been suggested previously based on either intuitive are included in this study (E. aceris, E. coccineus, E. assessment (e.g., Danzig, 1980; Herrick and Seger, 1999; eucalypti, and E. spurius) fall into two distinct clades Kosztarab, 1996; Kosztarab and Kozaar, 1988) or cla- (referred to as the El and E2 clades in Fig.1). E. eucalypti distic analysis (Miller, 1984) of morphology.Putoids belongs to the E2 clade and thus does not appear to be and pseudococcids have similar adult female morphol- closely related to E. buxi, even though both species pos- ogy (Miller and Miller, 1993) and therefore the posi- sess enlarged ducts (see Cook and Gullan, 2001).The tioning of Puto outside the main neococcoid clade other eriococcid clade (El) comprises taxa assigned to suggests that this morphology may be plesiomorphic for Acanthococcus by Miller and Gimpel (1996) but currently the neococcoids, as suggested by Koteja (1990, 1996). placed in Eriococcus (namely E. aceris, E. coccineus, and There is support for a close relationship between E. spurius), an undescribed Australian species (E. sp.ex Kerriidae ( ¼ Tachardiidae) and Coccidae, consistent Hakea), and the insects, genus Dactylopius,of with the similarity of their salivary pumps (Koteja, the monotypic family Dactylopiidae. 1976), crawler and adult male morphology (Miller, Serious nomenclatural implications for the Eriococ- 1991), and putative homologies in the anal areas of both cidae stem from the results presented here.These mo- groups (Lit and Gullan, 2001).However, the resolution lecular data, however, should be regarded as preliminary of the sister relationships of the Kerriidae and the Coc- and the current concept of the Eriococcidae should be cidae require data from the and Micrococci- retained until more taxa are sampled, independent data dae, small families suggested to have close affinities with are derived from other genes, and morphological sup- the Kerriidae (Foldi, 1997; Miller and Hodgson, 1997). port for the clades recognized by the SSU rDNA data can be assessed.In addition, the region of SSU rRNA 4.2.3. Eriococcidae used does not provide sufficient resolution of the rela- Eriococcids primarily are defined by the lack of tionships among the archaeococcoids and among some synapomorphies of other families, which led Cox and of the neococcoid families despite resolution of both Williams (1987) to suggest that Eriococcidae may be deeper and more derived nodes.We suggest that DNA paraphyletic.In the present study, Eriococcidae is ren- sequences from faster-evolving regions, a more extensive dered paraphyletic by the inclusion of the Asterolecon- sampling of major families, and the inclusion of ex- iidoe, Beesoniidae, Dactylopiidae, Diaspididae, and emplars from presently unrepresented minor families are Stictococcidae.These molecular data do not provide required to help improve the current understanding of support for the inclusion of either the Diaspididae or the scale insect phylogeny. Asterolecaniidae among taxa currently recognized as belonging to the Eriococcidae.However, there is sup- port for the inclusion of the Beesoniidae, Dactylopiidae, Acknowledgments and Stictococcidae among the eriococcid taxa.The SSU rDNA data strongly suggest that the large genus Erio- We thank Peter Cranston and Rob DeSalle for coccus, currently comprising about 350 described species helpful comments on earlier drafts of the manuscript. L.G. Cook et al. / Molecular Phylogenetics and Evolution 25 (2002) 43–52 51

We are extremely grateful for the specimens that were Phylogenetic Tests of Evolutionary Scenarios.Mem.Mus.Natl. provided by our colleagues Christian Burban, Imre His.Nat.173, 203–230. Foldi, Ray Gill, Hans-Peter Heckroth, Rosa Henderson, Gibbs, R.A., Cockeril, M., 1995. Sequencher. Gene Codes Corp., Ann Arbor. Chris Hodgson, Jan Koteja, Michael Kosztarab, Ferenc Goldman, N., 1993. Simple diagnostic statistical tests of models for Kozaar, Ireneo Lit, Chris Malumphy, Zvi Mendel, Dug DNA substitution.J.Mol.Evol.37, 650–661. Miller, Elizabeth Miller, and Mark Self.A permit to Gullan, P.J., Cook, L.G., 2002. Phenacoleachia Cockerell (Hemiptera: collect specimens of Phenacoleachia was provided by the Coccoidea: Phenacoleachiidae) re-visited.Bolletlino di Zoologia Department of Conservation, New Zealand, with the agraria e Bachicoltura, Ser.II 33 (3), 163–173. Gullan, P.J., Kosztarab, M., 1997. Adaptations in scale insects. Annu. assistance of Chris Green, and John Dugdale advised on Rev.Entomol.42, 23–50. collecting localities.Funding for the eriococccid re- Gullan, P.J., Sjaarda, A.W., 2001. Trans-Tasman Platycoelostoma search was provided by the Australian Biological Re- Morrison (Hemiptera: Coccoidea: Margarodidae) on endemic sources Study. Cupressaceae, and the phylogenetic history of margarodids.Syst. Entomol.26, 257–278. Haig, D., 1993. The evolution of unusual chromosomal systems in coccids: extraordinary sex ratios revisited.J.Evol.Biol.6, 69–77. References Herrick, G., Seger, J., 1999. Imprinting and paternal genome elimi- nation in insects.In: Ohlsson, R.(Ed.),Genomic Imprinting: An Ben-Dov, Y., Miller, D.R., Gibson, G.A.P., 2001. ScaleNet. http:// Interdisciplinary Approach.Springer, Berlin, pp.41–71. www.sel.barc.usda.gov/scalenet/scalenet.htm. Jukes, T.H., Cantor, C.R., 1969. Evolution of protein molecules. In: Benson, D.A., Karsch-Mizrachi, I., Lipman, D.J., Ostell, J., Rapp, Munro, H.N. (Ed.), Mammalian Protein Metabolism. Academic B.A., Wheeler, D.L., 2000. GenBank. Nucleic Acids Res. 28, 15–18. Press, New York, pp.21–132. Blackman, R.L., 1987. Reproduction, cytogenetics and development. Kimura, M., 1980. A simple method for estimating evolutionary rate In: Minks, A.K., Harrewijn, P. (Eds.), Aphids: Their Biology, of base substitutions through comparative studies of nucleotide Natural Enemies and Control.Elsevier, Amsterdam, pp.163–195. sequences.J.Mol.Evol.16, 111–120. Blackman, R.L., Eastop, V.F., 2000. Aphids on the World’s Crops: An Kosztarab, M., 1996. Scale Insects of Northeastern North America: Identification and Information Guide, second ed.Wiley, Chiches- Identification, Biology, and Distribution, Virginia Mus.Nat.Hist., ter. Martinsville.Special Publication No.3. Boratynski, K., Davies, R.G., 1971. The taxonomic value of male Kosztarab, M., Kozaar, F., 1988. Scale Insects of Central Europe. Junk, Coccoidea () with an evaluation of some numerical Dordrecht. techniques.Biol.J.Linn.Soc.3, 57–102. Koteja, J., 1974a. On the phylogeny and classification of the scale Borchsenius, N.S., 1950. Mealybugs and Scale Insects of the USSR insects (Homoptera, Coccinea) (discussion based on the mor- (Coccoidea).Akad.Nauk SSSR, Moscow, in Russian. phology of the mouthparts).Acta Zool.Cracoviensia 19, 267– Bremer, K., 1988. The limits of amino acid sequence data in 326. angiosperm phylogenetic reconstruction.Evolution 42, 795–803. Koteja, J., 1974b. Comparative studies on the labium in the Coccinea Buchner, P., 1965. Endosymbiosis of with Plant Microor- (Homoptera).Zeszyty Nauk.Akad.Rolniczej Krak oow Rozprawy ganisms.Wiley, New York. 27, 1–162. Campbell, B.C., Steffen-Campbell, J.D., Gill, R.J., 1994. Evolutionary Koteja, J., 1976. The salivary pump in the taxonomy of the Coccinea origin of whiteflies (Hemiptera: Sternorrhyncha: Aleyrodidae) (Homoptera).Acta Zool.Cracoviensia 21, 263–290. inferred from 18S rDNA sequences.Insect Mol.Biol.3, 73–89. Koteja, J., 1990. Paleontology. In: Rosen, D. (Ed.), World Crop Pests, Campbell, B.C., Steffen-Campbell, J.D., Sorensen, J.T., Gill, R.J., Vol.4A, Armored Scale Insects: Their Biology, Natural Enemies 1995.Paraphyly of Homoptera and inferred and Control.Elsevier, Amsterdam, pp.149–163. from 18S rDNA nucleotide sequences.Syst.Entomol.20, 175–194. Koteja, J., 1996. Scale insects (Homoptera: Coccinea) a day after. In: Cook, L.G., Gullan, P.J., 2001. Are the enlarged ducts of Eriococcus Schaefer, C.W. (Ed.), Studies on Hemipteran Phylogeny. Proceed- (Hemiptera: Coccoidea: Eriococcidae) plesiomorphic? Entomolog- ings of Thomas Say Publications in Entomolgy.Entomol.Soc.Am. ica 33, 59–66. Lanham.MD, pp.65–88. Cox, J.M., 1984. Relationships of the Phenacoleachiidae (Homoptera: Koteja, J., 2000a. Advances in the study of fossil coccids (Hemiptera: Coccoidea).In: Verhandlungen des Zehnten Internationalen Sym- Coccinea).Polski Pismo Entomol.69, 187–218. posiums uuber€ Entomofaunistik Mitteleuropas (SIEEC X), 15–20 Koteja, J., 2000b. Scale insects (Homoptera, Coccinea) from Upper Aug.1983, Budapest, pp.339–341. New Jersey amber.In: Grimaldi, D.(Ed.),Studies on Cox, J.M., Williams, D.J., 1987. Do the Eriococcidae form a Fossils in Amber, with Particular Reference to the Cretaceous New monophyletic group? Boll.Lab.Entomol.Agrar.Filippo Silvestri Jersey.Backhuys, Leiden, pp.147–229. Portici 43 (Suppl.), 13–17. Koteja, J., 2001. Essays on coccids (Hemiptera: Coccinea): paleontol- Danzig, E.M., 1980. Coccids of the Far Eastern USSR (Homoptera, ogy without fossils? Prace Muz.Ziemi 46, 41–53. Coccinea) with Phylogenetic Analysis of Coccids in the World Kozaar, F., Konczne Benedicty, Zs., 2000. Carayonemidae of the Fauna.Nauka, Leningrad (in Russian), also available in a 1986 Neotropical Region with descriptions of new genera and species English translation: Oxonian Press Pvt.Ltd, New Delhi. (Homoptera: Coccoidea).Folia Entomol.Hungarica 61, 71–82. Faith, D.P., 1991. Cladistic permutation tests for monophyly and Kumar, 1996.PHYLTEST: a program for testing phylogenetic nonmonophyly.Syst.Zool.40, 366–375. hypotheses, Version 2.0. Felsenstein, J., 1978. Cases in which parsimony and compatibility Kwon, O.Y., Ogino, K., Ishikawa, H., 1991. The longest 18S methods will be positively misleading.Syst.Biol.27, 401–410. ribosomal RNA ever known: nucleotide sequence and presumed Felsenstein, J., 1985. Confidence limits on phylogenies: an approach secondary structure of the 18S rRNA of the pea , Acyrtho- using the bootstrap.Evolution 39, 783–791. siphon pisum.Eur.J.Biochem.202, 827–833. Foldi, I., 1997. Defense strategies in scale insects: phylogenetic Lit Jr., I.L., Gullan, P.J., 2001. Comparative morphology of the anal inference and evolutionary scenarios (Hemiptera, Coccoidea).In: tubercle and associated structures of some lac insects (Hemiptera, Grandcolas, P.(Ed.),The Origin of Biodiversity in Insects: Coccoidea, Kerriidae).Entomologica 33, 119–126. 52 L.G. Cook et al. / Molecular Phylogenetics and Evolution 25 (2002) 43–52

Mathews, D.H., Sabina, J., Zuker, M., Turner, D.H., 1999. Expanded Sunnucks, P., Hales, D.F., 1996. Numerous transposed sequences of sequence dependence of thermodynamic parameters improves mitochondrial cytochrome oxidase I–II in aphids of the genus prediction of RNA secondary structure.J.Mol.Biol.288, 911–940. (Hemiptera: Aphididae).Mol.Biol.Evol.13, 510–524. Melville, R.V., 1982. Opinion 1203. Eriococcidae Cockerell, 1899 Swofford, D.L., 2000. PAUP*. Phylogenetic Analysis Using Parsimony conserved: type species designated for Eriococcus Targioni- (* and Other Methods).Version 4.Sinauer, Sunderland, MA. Tozzetti, 1868 (Insecta: Homoptera).Bull.Zool.Nomenclat.39, Szklarzewicz, T., 1998. Structure of ovaries in scale insects. I. 95–98. Pseudococcidae, Kermesidae, Eriococcidae, and Cryptococcidae Miller, D.R., 1984. Phylogeny and classification of the Margarodidae (Insects, Hemiptera, Coccinea).Int.J.Insect Morphol.Embryol. and related groups (Homoptera: Coccoidea).In: Verhandlungen 27, 167–172. des Zehnten Internationalen Symposiums uuber€ Entomofaunistik Takagi, S., 1992. Mangalorea hopeae, a new beesoniid (Homoptera: Mitteleuropas (SIEEC X), 15–20 Aug.1983, Budapest, pp. Coccoidea) inducing galls on Hopea ponga in southern India. 321–324. Insecta Matsumurana 47, 10–32. Miller, D.R., 1991. Superfamily Coccoidea. In: Immature Insects, vol. Tautz, D., Hancock, J.M., Webb, D.A., Tautz, C., Dover, G.A., 1988. 2.Kendall/Hunt, Iowa, pp.90–111. Complete sequences of the rRNA genes of Drosophila melanogas- Miller, D.R., Gimpel, M.E., 1996. Nomenclatural changes in the ter.Mol.Biol.Evol.5, 366–376. Eriococcidae (Homoptera: Coccoidea).Proc.Entomol.Soc.Wash- Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins, ington 98, 597–606. D.G., 1997. The clustal_X windows interface—flexible strategies for Miller, D.R., Gimpel, M.E., 2000. A Systematic Catalogue of the multiple sequence alignment aided by quality analysis tools. Eriococcidae (Felt Scales) (Hemiptera: Coccoidea) of the World. Nucleic Acids Res.25, 4876–4882. Intercept, Andover, UK. Tremblay, E., 1989. Coccoidea endocytobiosis. In: Schwemmler, W., Miller, D.R., Hodgson, C.J., 1997. Phylogeny. In: Ben-Dov, Y., Gassner, G.(Eds.),Insect Endocytobiosis: Morphology, Physiology, Hodgson, C.J. (Eds.), Soft Scale Insects—Their Biology, Natural Genetics, Evolution.CRC Press, Boca Raton, FL, pp.145–173. Enemies and Control, vol.A.Elsevier, Amsterdam, pp.229–250. Van de Peer, Y., De Rijk, P., Wuyts, J., Winkelmans, T., De Wachter, Miller, D.R., Miller, G.R., 1993. A new species of Puto and a R., 2000. The European small subunit ribosomal RNA database. preliminary analysis of the phylogenetic position of the Puto group Nucleic Acids Res.28, 175–176. within the Coccoidea (Homoptera: Pseudococcidae).Jeffersoniana von Dohlen, C.D., Moran, N.A., 1995. Molecular phylogeny of the 4, 1–35. Homoptera: a paraphyletic taxon.J.Mol.Evol.41, 211–223. Morrison, H., 1928. A classification of the higher groups and genera of White, M.J.D., 1973. Cytology and Evolution, third ed. the coccid family Margarodidae.US Dept.Agric.Tech.Bull.52, 1– Cambridge University Press, London. 239. Williams, D.J., 1985. The British and some other European Eriococ- Nur, U., 1980. Evolution of unusual chromosome systems in scale cidae (Homoptera: Coccoidea).Bull.Br.Mus.Nat.His.(Ento- insects (Coccoidea: Homoptera). In: Blackman, R.L., Hewitt, mol.) 51, 347–393. G.M., Ashburner, M. (Eds.), Insect Cytogenetics. Blackwell Yang, Z.H., Nielsen, R., 1998. Synonymous and nonsynonymous Science, Oxford, pp.97–117. rate variation in nuclear genes of mammals.J.Mol.Evol.46, Siddall, M.E., Whiting, M.F., 1999. Long-branch attractions. Cladis- 409–418. tics 15, 9–24. Zuker, M., Mathews, D.H., Turner, D.H., 1999. Algorithms and Soltis, P.S., Soltis, D.E., Wolf, P.G., Nickrent, D.L., Chaw, S., thermodynamics for RNA secondary structure prediction: a Chapman, R.L., 1999. The phylogeny of land plants inferred from practical guide. In: Barciszewski, J., Clark, B.F.C. (Eds.), RNA 18S rDNA sequences: pushing the limits of rDNA signal? Mol. Biochemistry and Biotechnology.NATO ASI Series, Kluwer Biol.Evol.16, 1774–1784. Academic, Dordrecht, pp.11–43.