A revision of Lachnodius Maskell (Hemiptera, Coccomorpha, Eriococcidae)

Abstract Lachnodius Maskell is a genus of three named species that are part of an Australian radiation of felt scale insects that induce galls on Eucalyptus and Corymbia (Myrtaceae). A female’s gall usually consists of an open-top pit in swollen plant tissue. Depending on the species, galls can occur on a host’s leaves, buds, stems, or trunk. Here, we redescribe the named species: L.eucalypti (Maskell), L.hirsutus (Froggatt) and L.lectularius (Maskell), and describe seven new species: L.brimblecombei Beardsley, Gullan & Hardy, sp. n., L.froggatti Beardsley, Gullan & Hardy, sp. n., L.maculosus Beardsley, Gullan & Hardy, sp. n., L.melliodorae Beardsley, Gullan & Hardy, sp. n., L.newi Beardsley, Gullan & Hardy, sp. n., L.parathrix Beardsley, Gullan & Hardy, sp. n., L.sealakeensis Gullan & Hardy, sp. n. Descriptions are based primarily on adult females, but for some species short diagnoses of nymphal stages also are provided. The taxonomic history of Lachnodius is reviewed, with notes on their biology and ecology. A key to species based on the morphology of adult females is provided, and lectotypes are designated for Dactylopiuseucalypti Maskell and Lachnodiuslectularius Maskell.


Introduction
In Australia, species of Eucalyptus and the closely related genus Corymbia are host to many species of gall-inducing felt scale insects (Gullan et al. 2005; García Morales et al. 2016). Most belong to one of two major radiations: the genus Apiomorpha Rübsaamen and a distantly related clade of Myrtaceae-feeding species (Cook and Gullan 2004). Lachnodius Maskell is one of at least eight genera belonging to the latter clade. Our aim here is to revise Lachnodius, with re-descriptions of the three currently recognized species and descriptions of seven new species. Maskell (1896) erected the genus Lachnodius for Dactylopius eucalypti Maskell, which he had described in 1892, and two other species, L. hirtus Maskell and L. lectularius Maskell, which he described as new. Fernald (1903) designated L. eucalypti as the type species. Beardsley (1982) synonymized the monotypic genus Pseudopsylla Froggatt with Lachnodius after study of the type specimens of the type species P. hirsutus Froggatt. This brought the number of described species of Lachnodius to four. Hardy et al. (2011) then transferred L. hirtus to their genus Lobimargo Hardy & Gullan, and the tally of Lachnodius species went back to three. Adult females of Lachnodius can be distinguished from other genera of felt scales found on Eucalyptus based on the morphological features in the keys of Hardy and Gullan (2007) or Hardy et al. (2011).

Taxonomic history and phylogenetic relationships of Lachnodius
In Maskell's brief definition of Lachnodius, he did not speculate on how it was related to other scale insects. Fernald (1903) placed Lachnodius in her Dactylopiinae, which included the presently recognized families Asterolecaniidae, Eriococcidae, Kermesidae and Pseudococcidae. Froggatt (1921) and Morrison and Morrison (1922) followed the classification of Fernald. The Morrisons also proposed a close relationship between Lachnodius and Sphaerococcopsis Cockerell. Ferris (1955) was puzzled by Lachnodius; he considered erecting an entirely new family for it, before opting to place it awkwardly in Pseudococcidae. Incidentally, the form that Ferris illustrated under the name Lachnodius eucalypti is certainly not that species, but may be L. lectularius. Hoy (1963) assigned both Lachnodius and Sphaerococcopsis to the Eriococcidae, a family to which he applied broad limits. Beardsley was of the opinion that Lachnodius and Sphaerococcopsis could not be placed easily into either Eriococcidae or Pseudococcidae (Beardsley 1972(Beardsley , 1974. He agreed with Ferris, that these genera constituted a previously unrecognized familylevel taxon, but one that was more closely related to the Eriococcidae than the Pseudococcidae. Koteja (1974) followed suite, and held Lachnodius to be a distinct family-level taxon, tentatively placed in his asterolecaniid group of families, on the basis of comparative studies of the labium, salivary pump and clypeolabral shield of adult females. In sum, the phylogenetic relationships of Lachnodius and Sphaerococcopsis were an enigma.
In the first scale insect phylogeny inferred from DNA sequence data, Cook et al. (2002) found support for a monophyletic group comprised of an unidentified Lachnodius species, Tanyscelis mammularis (Froggatt) and Ascelis praemollis Schrader (both of the latter being members of Eriococcidae in its current form); all three species induce galls on myrtaceous hosts. Then, in a more comprehensive estimate of the phylogeny of eriococcids, Cook and Gullan (2004) found these same three taxa inside a clade of Myrtaceae-feeding species that formed part of a larger clade of species from the Southern Hemisphere. They also found that the Eriococcidae is not monophyletic, as per previous suggestions based on morphological studies (Cox and Williams 1987;Hodgson 2002). The classification of scale insects has yet to be reconciled with this finding, but the most likely resolution will entail the recognition of the Myrtaceae-feeding (MF) clade of Cook and Gullan (2004) as a formal family-level taxon. This group would include many other mostly gall-inducing genera in addition to Lachnodius and Sphaerococopsis. Thus, it seems that Ferris, Beardsley, and Koteja were correct: Lachnodius and Sphaerococcopsis are not a natural fit in any of the existing scale insect families.

Undescribed species diversity of Myrtaceae-feeding clade
The MF clade is species rich and divided into subradiations, each of which is largely restricted to a subclade of Myrtaceae (Cook and Gullan 2004;Gullan et al. 2005). The species diversity of radiations on Leptospermum and Melaleuca is almost entirely undescribed ( LG Cook pers. comm.). More progress has been made in documenting the species that feed on Eucalyptus and Corymbia. In fact, over the last decade we (e.g., Gullan 2007, 2010;Hardy et al. 2011;Semple et al. 2015) have approached complete coverage of the known diversity (which, of course, says nothing about the unknown diversity). Here, we make another step in that direction by describing seven new species and redescribing the three already named species of Lachnodius.

Materials and methods
Adult females and immature specimens from recent collections and from dry museum material were slide-mounted in Canada balsam, mainly using a method similar to that described in Gullan (1984b). The morphological terms mainly follow Williams (1985), Miller and McKenzie (1967) and Hardy and Gullan (2007). The adult females of a few species have tiny dorsal sclerotic pits or depressions that are referred to herein as urns or varioles, depending on their shape. Measurements were made using an ocular micrometer attached to a compound microscope. All are given as a range and based on maximum dimensions (e.g., the body width of a slide mounted specimen was measured across the widest transverse section, the location of which varies among specimens, and leg segment lengths were measured along the longest axis). Tarsal length excludes the claw. Spiracle length includes the muscle plate (apodeme). Setal lengths exclude the setal base. All illustrations of the insects were prepared by NBH and photographs of the live insects and galls were taken by PJG. requirement. We have registered each of the new names published in this paper with the Official Registry of Zoological Nomenclature (ZooBank) and cite the Life Science Identifiers (LSIDs) after the heading for each new name. Each LSID is a globally unique identifier for the nomenclatural act of naming a new taxon.
JWB is the coauthor of six new names for Lachnodius because he recognized these species in his unpublished work. We provide a short synopsis of his work in the Discussion. A large portion of this study was based upon JWB's collection, which is housed in the BPBM. The BPBM has allowed the holotype of any new Australian species from the JWB collection to be deposited in the ANIC (in correspondence of PJG in 1996).

Notes on parasitoids and the effects of parasitization
Slide-mounted adult female and second-instar specimens of Lachnodius frequently show evidence of attack by internal parasitoids. We have noted the chorion of parasitoid eggs and developing parasitoid larvae, sometimes evident only by their mandibles. In the field, JWB occasionally found obviously parasitized adult females of both L. eucalypti and L. lectularius within their galls on host trees. As the parasitoids matured, the parasitized host became a hard, brown husk from which the adult wasps eventually emerged. Several parasitoids can develop in one host, with as many as 20 adults emerging from a single adult female of L. lectularius. JWB determined that the wasps were a kind of Encyrtidae, possibly species of Metaphycus or related genera. There are no previously published records of parasitoids attacking Lachnodius.
It appears that parasitization affects the development of structures in the host's integument, in particular the macrotubular ducts, making identification of parasitized specimens potentially problematic. Compared to unparasitized individuals, parasitized female specimens identified as L. lectularius often have fewer or smaller macrotubular ducts. Normally, the dorsal macrotubular ducts are numerous and large (ca. 6-8 µm rim diameter). In some parasitized specimens the ducts are few, whereas in others they are abundant but small (ca. 2-3 µm in diameter and without well-defined rims).  (Figs 1, 2). Galls consist of a pit in swollen plant tissue with insect's dorsum either exposed or partially concealed. Females remain in their gall after their imaginal molt, and then at maturity, depending on species, either desert their gall and move elsewhere for oviposition, or remain in their gall for reproduction. Males, in the few species for which they are known, induce galls as first-instar nymphs but then, near the end of their second instar, vacate the gall and move to another site to form cocoons in which they complete their development. Diagnosis of adult females of the genus Lachnodius. Body outline circular to oval. In most species eyes on margin (on venter in L. froggatti, and absent in L. sealakeensis). Antennae six to seven-segmented. Pair of broad, blister-like frontal lobes between antennae; a series of elongate setae along posterior margin of lobes. Tentorial box usually with anterior aliform extensions. Labium either one-segmented, or composed of two fused segments; proximal segment indicated by a pair of setae on ventral surface; distal segment with one pair of ventral seta, one pair of fleshy apical setae, and one pair of dorsal setae. Legs well developed. Anus ventral, with sclerotic rim having fewer than ten setae (except in L. hirsutus), base of each seta surrounded by ring of minute pores. Anal lobes absent.

Taxonomy
Dorsum. Setae short to minute, ≤ 10 µm long (except up to 25 µm long on L. hirsutus). Microtubular ducts and one or two size classes of macrotubular ducts present; larger macrotubular ducts sometimes with one seta touching rim of dermal orifice; duct shaft of uniform width or constricted near vestibule; macrotubular ducts with vestibule weakly sclerotic and compressed, i.e., not cup-shaped. Derm membranous, sometimes with enlarged microtrichia, sometimes with concave sclerotic granules. Multilocular pores absent. Dorsum delimited by a marginal with fringe of setae, differentiated from other body setae, with shape flagellate, conical or sagittate; marginal fringe either complete around margin, or with break between thorax and abdomen, or with break between thorax and abdomen + break between meso-and metathorax.
Venter. Sometimes larger than dorsum. Setae flagellate, in transverse rows across each abdominal segment, scattered along submargin, in clusters anterior to each coxa. Microtubular ducts usually absent (L. eucalypti with scattered microtubular ducts on head); macrotubular ducts similar to those on dorsum. Quinquelocular pores dense around vulva, clusters around each spiracle, scattered along submargin and across each body segment.
Venter. Larger than dorsum. Ventral setae 40-180 µm long; elongate setae medial of each coxa 150-225 µm long; longest setae on head 205-350 µm long. Macrotubular ducts similar to those on dorsum; in transverse band across each abdominal segment, scattered throughout submargin, medial of meso-and metacoxa. Quinquelocular pores 5 µm in diameter, found wherever setae occur, in transverse band across each segment, in a dense band along submargin, dense on posterior abdominal segments and around each spiracle.
Notes. The slide-mounted adult female of L. brimblecombei is most similar to that of L. lectularius. Each has a marginal fringe of close-set setae, and the dorsum densely beset with macrotubular ducts of a single type, none of which have a seta touching the dermal orifice. In life the two are easy to distinguish. The adult female of L. brimble-combei induces a deep stem or bud gall with considerable swelling of the surrounding tissue that covers a portion of the female's dorsum (Fig. 1a). If the gall occurs on the stem, it causes the stem to bend (Fig. 1a), often sharply. The adult female of L. lectularius also induces a gall on the stem or bud of the host, but the gall does not cover any portion of the female's dorsum (Fig. 2a, b), and if on a stem does not make it crooked. Slide-mounted specimens of L. brimblecombei can be distinguished from those of L. lectularius by having (1) a dense marginal band of quinquelocular pores on the venter (absent in L. lectularius) and (2) the marginal setae longer than the anal ring setae (marginal setae shorter than anal ring setae in L. lectularius).
The habitat of the Victorian specimens of L. brimblecombei, which develop in galls formed on flower buds, is different from that of the type specimens from Queensland (with galls as in Fig. 2a). But we found no significant morphological differences among specimens from the different states, except for slightly smaller fringe setae and possibly fewer quinquelocular pores in the Victorian specimens. In case further study reveals that the Victoria specimens are different, we have restricted the type series to specimens collected by AR Brimblecombe in Queensland and three adult females from New South Wales, one of which is a DNA voucher.
At Wilson's Promontory in Victoria, galling caused by L. brimblecombei on E. baxteri reduces bud survival and flowering (Andersen 1989). Monitoring of tagged eucalypt shoots showed that, although less than 12% of buds were galled by L. brimblecombei (misidentified as Opisthoscelis sp.), the presence of galled buds often caused the abscission of nearby non-galled buds and galling on flowering stems often resulted in the loss of entire inflorescences, but these losses may be exacerbated by low water availability. On heavily galled shoots, the proportion of flowers producing mature fruit was correlated negatively with degree of galling, indicating that there was no compensatory increase in the success of the non-galled buds. Thus L. brimblecombei may decrease the fitness of its host, especially during periods of low rainfall.
A live adult female from Wild Cattle Creek State Forest in New South Wales was covered in white powdery wax and had a marginal fringe of white wax filaments ca. 0.2 mm long. Evidence of attack by parasitoid wasps was seen in several of the specimens studied. Two females from Redland Bay, Queensland, contained mandibles of parasitoid larvae, and the specimen from Mittagong, NSW also was parasitized.
We also examined one large (ca. 10 mm long) adult female that may be a developmentally abnormal specimen of L. brimblecombei or it might be a new species. It was collected from a stem pit on E. fasciculosa at Belair in South Australia (ANIC). It differs from typical adult females of L. brimblecombei in having reduced and distorted legs and antennae, many more dorsal fringe setae and in lacking the narrow marginal band of disc pores.
Etymology. Pioneering Australian coccidologist AR Brimblecombe recognized this species and used the manuscript name 'Lachnodius geniculatus' to refer to it in his dissertation [citation of this name here is NOT intended to be for nomenclatural purposes; the name is not valid]. This species is named in Brimblecombe's honor. The species epithet is a noun in the genitive singular. Diagnosis. Loose marginal fringe with minute sagittate setae; microtubular ducts on venter of head; macrotubular ducts with distal attenuation.
First-instar nymph (n = 14 from Bundoora, Victoria). This instar was redescribed and figured well by Morrison and Morrison (1922, figure 14) and only some additional information is provided here. Newly hatched individuals ca. 380-400 µm long; feeding first-instar nymphs removed from leaf galls 550-600 µm long, broadly oval in outline, with venter expanded, balloon-like, to fill gall cavity, dorsum flat. Slide-mounted specimens with medial to submedial dorsal derm bearing small sclerotic spots, mostly 1-2 µm in greatest dimension; marginal setae mostly falcate (incorrectly described as 'flabellate' by Morrison & Morrison) except posterior three pairs lanceolate but often with apex jagged or notched, each marginal seta 15-30 µm long. Pair of elongate caudal setae ca. 65 µm long.
Notes. The adult female of L. eucalypti could be confused most easily with that of L. froggatti sp. n. Each induces pit galls on leaves and may be covered by waxy secretions. The adult female of L. eucalypti differs from that of L. froggatti by having (1) a marginal fringe of alternating sagittate and conical setae (marginal setae of L. froggatti hair-like to capitate); (2) eyes on margin (eyes on venter of L. froggatti); (3) no dorsal macrotubular ducts with setae touching rim of dermal orifice (dorsum of L. froggatti having some macrotubular ducts with a seta touching dermal orifice); and (4) microtubular ducts on ventral surface of head (absent in L. froggatti). Also, in life the secretions covering an adult female of L. froggatti are woolly, in contrast to the clumpy, powdery secretions that cover an adult female of L. eucalypti. Populations of L. eucalypti are known from all eight Australian states and territories. Specimens of L. eucalypti have been collected most commonly from E. camaldulensis, which is the most widely distributed species of Eucalyptus in Australia (Brooker 2002), but they also have been taken from a number of additional species of Eucalyptus in three sections (Adnataria, Exsertaria, and Maidenaria) of the subgenus Symphyomyrtus. Two populations of adult females probably both from E. camaldulensis (Windjana Gorge in northern Western Australia and near Alice Springs in the Northern Territory) have the sagittate setae of the marginal fringe of more uniform length and larger (15-20 µm long) compared with populations from the eastern and southern states in which the sagittate setae vary in size from 6-18 (mostly < 15) µm long on individual specimens. Due to this difference, we have excluded the females collected in the Northern Territory and Western Australia from the description above. Freshly collected specimens suitable for DNA sequencing might allow a decision on the species status of this morphological variation.
Life history data for L. eucalypti were obtained by JWB from a population that infested mature trees of E. camaldulensis on the campus of La Trobe University, Bundoora, Victoria, during the spring, summer, and fall of 1971-72. Beginning on 29 September 1971, adult females of L. eucalypti were collected while ovipositing on the bark of trunks and major branches of host trees. Oviposition was intermittent between then and mid-February 1972. Individual females appeared to complete oviposition within a short period of two or three days. The eggs were pink and laid in a single layer that formed a long, flat ribbon, 4-6 eggs wide, the top and sides of which were enclosed by a waxy secretion (Fig. 1c). Individual ovisacs were sometimes more than 5 cm long, straight or curved, and contained on the order of several hundred eggs (although no counts were made). A shrivelled, moribund female was often found at the end of an ovisac.
In the laboratory, eggs hatched 7-10 days after deposition. On host trees, the newly-eclosed first-instar nymphs migrated from the oviposition sites to the foliage, where they settled on the upper surfaces of young leaves. Feeding by each nymph resulted in a shallow pit gall on the leaf surface, which enclosed the nymph and grew along with it. The dorsal surface of settled first-instar and second-instar nymphs was nearly flat, smooth, and shiny, without evident waxy secretions. The ventral part of the nymph's body filled the cavity of the pit gall, while the dorsal margin overlapped and sealed the edge of the gall cavity.
In second-instar females the legs are poorly developed and apparently non-functional. Male nymphs, which can be distinguished from females in the second instar by the presence of fully developed legs, developed in leaf galls similar to those of females. Second-instar males eventually abandoned their galls and migrated to the bark of trunks and branches of host trees where they formed ovoid cocoons in protected situations. In the laboratory, males formed cocoons under paper lining the bottom of the petri dishes in which they were held. Cocoons were formed of whitish filaments, which issued from the dorsal tubular ducts.
Females remained in their galls after molting to the third (adult) instar, and continued to feed for an undetermined period, until fully developed. They then abandoned their galls and migrated to the bark to oviposit. When and where mating took place was not determined. At La Trobe University, the population of L. eucalypti did not appear to reproduce synchronously. Although ovipositing females were observed only during the spring and summer months (September to February), individuals of all stages were found on the trees during late January.  described this species based on adult females, pupal and adult males, and first-instar nymphs, collected from a tree referred to as E. amygdalina. The following year, Maskell (1893) indicated that his type material of Dactylopius eucalypti was from South Australia, and that the specimens were collected under bark. It appears that Maskell received the type material from the South Australian collections accumulated by Frazer S Crawford of Adelaide, an economic entomologist with an interest in Coccoidea. However, the identification of the host tree as E. amygdalina is problematic if the insects came from South Australia, because this eucalypt is endemic to Tasmania. Specimens of a second collection, which Maskell received from WW Froggatt in Sydney, were in pit galls in the leaves of E. robusta. This difference in the site of collection on the host trees apparently gave Maskell the impression that the species developed both in leaf galls and under bark, and presumably he was unaware that adult females migrate from leaf galls to bark prior to oviposition.
The Maskell collection in the NZAC contains six slides of L. eucalypti, four of which we consider to be type material. The four slides with type specimens are labeled "Dactylopius eucalypti" with the word "Dactylopius" crossed out and "Lachnodius" written above it. These labels also have the locality as "Australia" and the date as "1886." The slides bear (1) an adult female, (2) an adult male, (3) three first-instar nymphs, and (4) adult male parts (part of the thorax, two antennae, and two legs). The other two Maskell slides of this species in the NZAC contain (1) an adult female and (2) eight first-instar nymphs and bear later collection data (1893 and 1894) and therefore could not have been part of the material on which Maskell based his description. Beardsley had intended to designate the 1886 slide bearing the adult female as the lectotype of Dactylopius eucalypti Maskell, and labelled it as such in 1972 but this action was not published until now.
Notes. The adult female of L. froggatti is most similar to that of L. eucalypti. See notes for L. eucalypti for a comparison. Populations of L. froggatti have been sampled from New South Wales, Victoria, and South Australia. It is known to feed on hosts in the subgenera Eucalyptus (section Eucalyptus) and Symphyomytrus (sections Adnataria and Maidenaria). The live adult female is white to pale cream or yellow in life, and mature females produce copious dorsal glassy wax filaments and white powdery wax (Fig. 1d, e). The females have been found only on the leaves and the pit below the female's body may be up to 1.5 mm deep (Fig. 1e). The leaf area around the feeding insect is often depressed and discolored or necrotic, and the opposite surface of the leaf has a bulge; on very young foliage, the female causes leaf curling.
Froggatt's first accession notebook (Gullan 1984a) has an entry for the specimen that we have designated as holotype, as follows: "(303) Dactylopius eucalypti ?Large funnel leaf Penrith (No 1) (Berlese No 233)". The words "?Large funnel leaf " are written in different handwriting and inserted in the original entry. The mention of a Berlese number refers to part of this collection being sent to Berlese (presumably the Italian coccidologist Antonio Berlese) as a previous entry says "(Sent to Berlese No 230)". It seems that Froggatt confused L. froggatti with L. eucalypti, as shown by his identification of our holotype of L. froggatti (discussed above) as L. eucalypti, and also the following record. Two paratype females listed below have a Froggatt number of 27, which Froggatt's first accession notebook records as from Wallsend, which is one of the localities listed by Froggatt (1917Froggatt ( , 1921 for L. eucalypti. We have restricted the type series of L. froggatti to specimens from New South Wales. All specimens in the Froggatt collection are from this state.
Etymology. This species is named in honor of the collector of the type material, the late WW Froggatt, an Australian entomologist employed by the New South Wales Department of Agriculture during the early decades of the 20 th century. Froggatt was the first to seriously attempt a systematic treatment of the scale insect fauna of Australia. The species epithet is a noun in the genitive singular.
Notes. The adult female of L. hirsutus can be distinguished from all other species by the combination of 6-segmented antennae, extremely long marginal setae (350-450 µm long), and the scarcity of quinquelocular pores, which occur only near the spiracular openings. The anal ring of L. hirsutus is also unique among Lachnodius species; it has six or fewer ring setae present, with only a few minute pores near the base of each seta.
In his redescription of this species, Beardsley (1982) omitted the length of the fourth segment from the antennal formula. The correct segment lengths (µm), from the base to the apex, are: 150, 120, 400, 200, 130, and 50. Froggatt (1921: 6) stated "The female coccids produce solid woody galls on the branchlets of eucalypts with an irregular opening on the upper surface (Fig. 1f ). At female maturity, the gall of L. hirsutus probably splits open at the apex to reveal the female, because enclosed globular twig galls of nymphs have been collected in association with galls resembling those of L. hirsutus (Gullan et al. 2005). It is not clear whether the host genus of Froggatt's type material was Corymbia or Eucalyptus, since the original description simply says "an undetermined species of eucalyptus [sic]". The bloodwood eucalypts were not recognized as a genus (Corymbia) separate from Eucalyptus until more recently (Hill and Johnson 1995).
Notes. Adult females feed in a pit in a swollen stem or bud of the host eucalypt (Fig. 2a, b). The body color is variable; it is green with a red longitudinal stripe on the dorsum of younger females and fully orange or red to brown in older females. In life, females can lift up their abdomen and expose their venter. Each seta forming the marginal fringe surrounding the dorsum is covered in a glassy secretion. The life history of L. lectularius is similar to that of L. eucalypti. For details see Notes under L. eucalypti. One exception is that the galls of developing young females of L. lectularius are located on succulent young twigs and buds rather than on leaves. Mature females of L. lectularius were collected by JWB from Eucalyptus camaldulensis at La Trobe, and on other hosts and localities in Victoria during a relatively short period (February 14 to 20, 1972). This suggests that L. lectularius may reproduce with more synchrony than L. eucalypti. Eggs from females of L. lectularius held in the laboratory by JWB began to hatch approximately two weeks after oviposition.
In an unpublished manuscript, JWB treated as a separate species some of the larger specimens of what we consider to be L. lectularius. He noted that these specimens closely resemble the type material of L. lectularius and that the first-instar nymphs were identical, but pointed out several differences: specifically, the larger females have longer setae, more tubular ducts, a larger anus, more expanded tibial apices, and more translucent pores on the hind legs. Each of these traits appears to be correlated with body size across Lachnodius species. Therefore, we have opted to interpret this as part of the phenotypic variation found within L. lectularius. Maskell (1896) described the adult female, the second-instar female, and the firstinstar nymph of this species. Apparently, his description was based on material sent to him from Victoria by Mr C French. Type material of this species consists of specimens on 6 unstained slides prepared by Maskell, one adult female in the USNM and eleven adult females mounted by JWB from specimens from two boxes of dried material in NZAC. The original Maskell slides are labelled "Dactylopius lectularius" with "Dactylopius" crossed out and "Lachnodius" written above it, and "1895 -W. M. M." There are no locality or collector data on these slides. The dry material was labelled only "Dactylopius lectularius -Australia" but the boxes were lost (see explanation in Materials and Methods). Only one of the original Maskell slides contains an entire adult female. JWB labelled that specimen as the lectotype in 1972 but this action was not published until now (see below). Of the remaining slides (paralectotypes), one contains female mouthparts, one the posterior body and antenna of an adult female, one a second-instar female, and two contain first-instar nymphs. When JWB slide-mounted specimens from Maskell's dry material in 1972, he labelled the slides with the collection data from the original description (but with the wrong collector name), rather than what was written on the box.
Concerning the type material of L. lectularius, Maskell (1896: 401) only stated that "Mr. French has sent me a number of specimens and says, 'It does great damage to young trees at Mooroopna, Goulburn River, Victoria'." Therefore, we assume that all of his material was from this one source. Specimens in the dry material are mostly parasitized mummies, and JWB only obtained two satisfactory slide preparations. Both of these adult females show evidence of having been parasitized, containing parasitoid mandibles, encyrtiform eggshells, and small sclerotized first-stage parasitoid larvae. Maskell (1896: 401) recorded the habitat from which the type material was derived as "In Australia, on Eucalyptus rostrata." Eucalyptus rostrata is a junior synonym of E. camaldulensis, the "river red gum," a common species throughout southeastern Australia (Chippendale 1988).
The adult female specimens of L. lectularius in the Maskell collection do not agree in all details with his published description and figures. We consider that the discrepancies are errors in Maskell's interpretation. Morrison and Morrison (1922) noted that Maskell's descriptions often were inaccurate. Here we point out the mismatches between his description and specimens. In his 1896 description, Maskell stated that the adult female has an "Epidermis bearing many short fine hairs, and near the cephalic and abdominal extremities are two curved series of stronger spiny hairs, about sixty in each." His figure of the female abdomen (Maskell 1896: Plate XXI, fig. 16) shows a series of spine-like setae in the area behind the anal ring. In the drawing these are thicker and more conical in form than the fringe setae, which are depicted (Plate XXI, fig. 17) as being nearly digitiform. By contrast, the Maskell specimens do not have conical or spiniform setae posterior to the anal area, although many of the setae appear to have been broken off and look somewhat like stiff bristles. On the other hand, in fresh preparations, the ventral setae in this region are quite elongate. Thus, we think that he simply confused body surfaces. Maskell also refers to a pair of "strong short conical spines" set close together between the antennae in some specimens, but not in all. In this position in the lectotype, we found a pair of parasitoid mandibles, which he must have mistaken for spines. Likewise, he mistook several pairs of parasitoid mandibles for spines in his description of the second-instar female. Maskell counted 24 of these structures, which is consistent with JWB's observation that the encyrtids that attack Lachnodius species can be highly gregarious.
In his notes, JWB recorded having studied two specimens that were not seen by PJG or NBH: Queensland: two adult females: Eucalyptus sp., gall no. 9, Acacia Ridge, Brisbane, 10 Jan1968, EC Dahms (these probably are housed in the Queensland Museum in Brisbane).  Coll. No. 453" (USNM). Note that JWB made an error in writing the collector as "W.W. Froggatt", as the original specimens were collected by C. French. Also, the dry material that JWB mounted did not bear the collection data that he put on his slide labels, but was added by JWB based on the data cited in Froggatt's original description. Diagnosis. Dorsum with numerous sclerotic invaginations; marginal fringe of conical setae; some dorsal macrotubular ducts with seta touching rim; anal ring invaginated.
Venter. Ventral setae 22-60 µm long; elongate setae medial of each coxa decreasing in size caudad: ca. 100 µm long near fore coxa, ca. 55 µm long near hind coxa; longest setae on head 120-140 µm long. Macrotubular ducts of two types: (i) larger ducts with shaft subtending vestibule constricted, each ca. 22 µm long, with rim of dermal orifice ca. 6 µm wide, found along posterior margin and in transverse band across abdominal segment IV; (ii) smaller ducts with uniform shaft diameter ca. 15 µm long, with rim of dermal orifice ca. 4 µm wide, along margin anterior of larger ducts, in transverse rows across abdominal segments, amongst clusters of setae medial of each coxa. Quinquelocular pores of two distinct size-classes: (i) larger pores 5-6 µm in diameter, found on posterior abdominal segments; and (ii) smaller pores 3-4 µm in diameter, near spiracles and along margin.
Etymology. The species name is taken from the Latin noun macula meaning spot, referring to the shallow, sclerotic pits on the dorsal body surface, combined with the Latin suffix -osus to give the meaning abundance of spots or spotted. The species epithet is a Latin masculine adjective.
Notes. Adult females of L. maculosus are most similar those of L. melliodorae and L. parathrix. See notes under L. melliodorae for a comparison. Adult females of L. maculosus can be distinguished by having (i) two size classes of macrotubular duct on both the dorsal and ventral body surfaces (L. melliodorae and L. parathrix have only one size class per body surface); and (ii) numerous minute sclerotic invaginations on the dorsum, each with interior margin sinusoidal (L. parathrix without minute sclerotic invaginations, L. melliodorae with minute sclerotic invaginations urn-shaped, interior margin convex).
Notes. Adult females of L. melliodorae are most similar those of L. parathrix and L. maculosus. These three species share (i) a marginal fringe composed of close-set setae inter- Young adult females collected near Benalla, Victoria, by PJG in 1996 and1997 were pale yellow in life with dorsomedial longitudinal stripe of red-wine color; the anterior spiracular furrow was visible as a pale line on each side of the body. The dorsum was naked (no secretion), but each seta in the marginal fringe was covered in a glassy secretion. The second-instar female had a salmon-colored dorsum.
The type series is restricted to specimens collected at Lower Plenty in Victoria, where JWB made several collections of all instars of this species.
Dorsum. Derm beset with weakly sclerotic spots, each 2-5 µm in greatest dimension. Dorsal setae ca. 5 µm long; two longitudinal rows on each side of body, medial row with two setae on prothorax, and one seta on each segment from mesothorax to abdominal segment VII, submedial row with three setae on prothorax and one seta on each segment from mesothorax to abdominal segment I. Microtubular ducts each ca. 5 µm long, with rim of dermal orifice subelliptical, ca. 2 µm wide and 3 µm long, two longitudinal rows on each side of body, medial row with one duct on each segment from head to metathorax + abdominal segment VIII, and submarginal row with two ducts on prothorax, one duct on each thoracic segment and abdominal segments I and V (Note: it was an arbitrary decision to include the ducts on each side of head and abdominal segment VIII in the "medial" row). Dorsum delimited by fringe of 31-34 setae on each side of body (excluding caudal setae), each seta 9-15 µm long and 7-9 µm width at base above socket, deltate, with anterolateral margin sinusoidal and posteromedial margin straight; weakly sclerotic cuticle surrounding each setal socket, these sclerotic areas coalescing around the three most posterior fringe setae and the caudal seta, forming broad caudal sclerotization.
Notes. The first-instar nymphs of L. melliodorae are most similar to those of L. froggatti and L. lectularius, which also have (i) dorsal sclerotic spots (poorly developed in L. froggatti); (ii) stout marginal setae subtended by patches of sclerotic cuticle; and (iii) broad caudal sclerotizations (but in L. lectularius, only the caudal seta and 1 enlarged fringe seta are part of the sclerotization). The first-instar nymphs of L. eucalypti lack both the sclerotized area surrounding the socket of each marginal seta and the broad caudal sclerotizations, but have dorsal sclerotic spots. The nymphs of all four species have a similar arrangement of ducts, pores and setae. The first-instar nymphs of L. melliodorae can be distinguished from those of L. eucalypti, L. froggatti and L. lectularius by the distinctive shape of the marginal setae: deltate, with anterolateral margin sinusoidal and posteromedial margin straight, and base broad (marginal setae of L. eucalypti mostly falcate, with setal base constricted; of L. froggatti falcate but shorter than those of L. eucalypti; of L. lectularius more elongate, conical, with both margins straight).
Etymology. The species name refers to the name of the host from which the type material was collected. The species epithet is in the genitive singular.
Etymology. This species is dedicated to Dr TR New, of the former Department of Zoology (now Ecology, Environment and Evolution), La Trobe University, who accompanied JWB during many collecting trips made in Victoria during 1971-72, and who guided JWB to the spot where this species was discovered. The species epithet is a noun in the genitive singular.
Notes. The holotype is the only specimen known for this species, but it is distinctive. The specimen is relatively small and probably not fully expanded. However, the modest size of the antennae and legs, in comparison with other twig gall-inhabiting species (e.g., L. lectularius) suggests that even fully expanded adults of L. newi would not measure much more than 4 mm long. The adult female of L. newi would be confused most easily with those of L. melliodorae, as both species have strongly recurved marginal setae. The adult female of L. newi can be distinguished from those of L. melliodorae by lacking urn-shaped sclerotic structures on the dorsum (present in L. melliodorae). In their place are heavily sclerotized microtrichia. The adult female of L. newi can be further differentiated from those of L. melliodorae by having no dorsal macrotubular ducts with a seta touching the rim of the dermal orifice (some present in L. melliodorae) and by lacking long setae at the middle of the posterolateral edge of antennal segment III.
Etymology. The species name is based on the Greek words para, meaning near or beside, and thrix, meaning hair, referring to the close-set setae forming the marginal fringe. It is a noun in apposition. Notes. The description of the adult female is based on specimens from both Victoria and New South Wales because they agree in all diagnostic features. The adult female from near Narooma has longer antennae and leg segments and appears to have no translucent pores on the hind coxae, compared with the two Victoria females, but this variation may be due to differences in nutrition or developmental temperatures. Adult females of L. parathrix are most similar those of L. melliodorae and L. maculosus, but can be easily distinguished by lacking the peculiar dorsal urns and varioles present in those species. The two known host species of L. parathrix belong to Eucalyptus series Radiatae and are characterized by the juvenile leaves having numerous oil glands (Brooker 2000 Diagnosis. Dorsum with dermal orifice of each microtubular duct surrounded by sclerosis; marginal fringe of truncate setae; dorsal setae capitate; anal ring invaginated. Description. Adult female (n = 16). Body outline oval; length 0.84-1.45 mm (1.18 mm for holotype), greatest width 0.70−1.28 mm (0.90 mm for holotype). Eyes not apparent. Antennae seven-segmented; length 230-355 µm; with two hair-like setae on segment I, one hair-like seta on segment II, one hair-like seta on segment III, two hair-like seta on segment IV, one fleshy seta on segment V, two hair-like setae + one fleshy seta on segment VI and six hair-like setae + three fleshy setae on segment V. Tentorial box with anterior extension of the dorsal arms, 138-190 µm long, 123-155 µm wide. Labium 60-75 µm long, 60-100 µm wide. Spiracles 45-68 µm long, 25-35 µm wide across atrium. Legs increasing in size caudad; fore legs: trochanter + femur 165-250 µm, tibia 105-175, tarsus 80-125 µm; mid legs: trochanter + femur 165−263 µm, tibia 105-165 µm, tarsus 90-125 µm; hind legs: trochanter + femur 190-275 µm, tibia 110-175, tarsus 85-125 µm; claw 28-38 µm; fore coxa with six setae, mid and hind coxae each with five setae, trochanter with four setae, femur with 6-8 setae, tibia with 6-8 setae, tarsus with 7-9 setae; tarsal digitules 45-60 µm long, claw digitules 25-40 µm long; translucent pores on all segments of hind leg. Anal ring invaginated, cuticle surrounding ring sclerotic, 38-75 µm wide, with 10-12 setae; ring setae 40-73 µm long. Pair of elongate caudal setae absent.
Etymology. The species name refers to the type locality, Sea Lake, Victoria. The name is an adjective with the suffix derived from the Latin -ensis, denoting place or locality.
Notes. Adult females of L. sealakeensis are most superficially similar to those of L. maculosus; both species occur under bark, have an invaginated anus, and relatively robust, subconical marginal setae. Adult females of L. sealakeensis can be readily distinguished from those of L. maculosus by (i) the lack of dorsal macrotubular ducts (two size-classes present in L. maculosus); (ii) dorsal microtubular ducts with sclerotic surrounds (these are unique among Sphaerococcopsis and Lachnodius spp.); (iii) capitate dorsal setae (lanceolate in L. maculosus); and (iv) the ventral surface of abdomen with macrotubular ducts with the shaft constricted distally (not constricted distally in L. maculosus). Capitate dorsal setae like those of L. sealakeensis are also present on adult female of Sphaerococcopsis platynotum Beardsley and S. umbilicus Beardsley. These differ from L. sealakeensis in having (i) 6-segmented antennae (7-segmented in L. sealakeensis); (ii) hind legs much larger than fore and mid legs (all legs subequal in L. sealakeensis); and (iii) venter much larger than dorsum (venter and dorsum subequal in L. sealakeensi).
This is the only species treated here that was not part of JWB's view of Lachnodius since the only known specimens were not collected until after his death. NBH and PJG share authorship of its name.
Material examined. Holotype: Victoria: adult female, on slide: ex pit under loose bark of Eucalyptus oleosa, ca. 6 km N of Sea Lake, intersection of Lake Tyrell Road and Calder Highway, 35.45S, 142.83E, NBH and PJG, 5 Feb 2005 (ANIC). Paratypes: Victoria: 17 adult females (all on separate slides), same data as holotype, one female is DNA voucher NH47 (ANIC except 2 slides in NMV; also 2 slides deposited in QDPC in 2009 but could not be located in 2018).

Discussion
This taxonomic work was begun in 1971 by the late Dr JW Beardsley, while he was a visiting Fulbright Research Scholar in the then Zoology Department, La Trobe University, Bundoora, Victoria. During that period, Beardsley encountered a number of undescribed taxa that he felt were new species of Lachnodius Maskell. Subsequently, he borrowed specimens of Lachnodius from Australian collections, especially those made by the late Ms HM Brookes (formerly of the Waite Agricultural Research Institute of the University of Adelaide) and PJG. He also visited the New Zealand Arthropod Collection in Auckland to examine Maskell's type material. The demands of other work, including ten years as the Chair of the Department of Entomology at the University of Hawaii, delayed progress on his Lachnodius revision. That delay ended with his retirement, and by 1993, he had completed written descriptions for several species. However, arthritis in his hands kept him from being able to illustrate them and, when Beardsley died suddenly on 5 February 2001 (Anwar 2001), his work on Lachnodius was still unpublished. At that point, PJG acquired Beardsley's notes and slides, with the intention of completing his work, an effort that NBH joined.
When PJG and NBH took over, they modified Beardsley's taxonomic concepts. Specifically, of the species he included in his unpublished revision of Lachnodius, they described one as a species of Opisthoscelis (Hardy and Gullan 2010), one as the sole member of the genus Heathcotia Hardy & Beardsley (Hardy et al. 2011), and six as species of Lobimargo Hardy & Gullan (Hardy et al. 2011). This paper treats what is left of Beardsley's concept of Lachnodius, and completes his revisionary work. Nevertheless, the monophyly of what we have left of Beardsley's concept of Lachnodius is uncertain. These species clearly are closely related to Sphaerococcopsis, as well as to Opisthoscelis Schrader and Tanyscelis Hardy & Gullan. It could be that some are more closely related to these other genera than to some of the other species of Lachnodius. More phylogenetic work is required to resolve these relationships.
to collect scale insects in Tasmania. The staff at the Melbourne Wildlife Sanctuary, La Trobe University, allowed us access the reserve in February 2005 to visit one of JWB's former main collection sites. Tim New, the late Ian Thornton and his colleagues (from the former Department of Zoology, La Trobe University, Bundoora, Victoria) provided laboratory space and other valuable assistance to JWB during 1971-72. The late Ms Helen M. Brookes (formerly at the Waite Agricultural Research Institute, University of Adelaide, South Australia) assisted JWB with his studies and provided important specimens. Financial support to JWB for research in Australia was provided by the Fulbright Exchange Program through the Australian American Educational Foundation. This work was supported by a National Science Foundation PEET grant DEB-0118718 to PJG, and an Australian Biological Resources Study Participatory Program research grants to Lyn Cook and PJG.