Australian thrips of the Haplothrips lineage (Insecta: Thysanoptera)

Keys are presented to the 11 genera and 50 species, including 21 new species, of Thysanoptera in Australia that are related to the worldwide genus Haplothrips Amyot and Serville. These taxa belong to what, in recent literature, has been called the “Haplothrips‐lineage”, that is, one of the three major radiations among the 2700 species and 350 genera of Thysanoptera Phlaeothripinae. The group is redefined, and the available tribal name Haplothripini shown to be appropriate. The character states on which the definition is based are discussed, and a list given of the 34 genera worldwide that can be included. The Australian species in these genera exhibit a diversity of biologies. Three genera involve species that invade galls induced by other thrips: Androthrips monsterae (Moulton) from New Guinea is newly recorded from Australia; Mesothrips jordani Zimmermann from South‐East Asia is newly recorded from Australia, with two new synonyms; the Asia‐Pacific genus, Euoplothrips Hood, includes two species in northern Australia. Three haplothripine genera are associated with grasses: one widespread genus, Apterygothrips Priesner, considered polyphyletic, includes only one Australian species; the only known species of Dyothrips Kudo is Oriental but extends into tropical Australia; Podothrips Hood, a circum‐tropical genus of thrips predatory on grass‐living coccoids, has 11 Australian species, six newly described. One grass‐associated genus, Bamboosiella Ananthakrishnan, is excluded from the Haplothripini. An Oriental genus of leaf‐ and flower‐living species, Dolichothrips Karny, includes one species in northern Australia, D. reuteri (Karny); Membrothrips Bhatti in which this species has been placed is considered a synonym. Karnyothrips Watson includes two species introduced to Australia, both predators of coccoids. Priesneria Bagnall includes three species from Australia, of which one fungus‐feeding species is newly described. Xylaplothrips Priesner is a widespread but ill‐defined, polyphyletic genus that currently includes three little‐known Australian species. Haplothrips is the main focus of this study, and character state variation among the Australian species is discussed. In total 24 species of Haplothrips from Australia are recognised, 14 being new species. Also included are the following, one new generic synonym, four new species synonyms, and one new combination. The 250 species worldwide in the genus are usually associated with the flowers of Asteraceae and Poaceae. In Australia, several of the species are specific to flowers, particularly of Poaceae and Cyperaceae but not Asteraceae. Almost half of these Australian species are presumed to be predatory on other small arthropods, and two have unusual host associations—with salt marsh Chenopodiaceae, and with the sori of Dicksonia tree ferns.


Introduction
The classification of the 5800 species currently recognised worldwide in the insect order Thysanoptera is the subject of considerable disagreement, and is not based on sound phylogenetic principles . Two sub-orders are recognised, and molecular evidence increasingly supports a sister-group relationship between these Terry et al. 2005). The first sub-order, Terebrantia, includes eight families with a total of about 2400 species. A recent ''reorganisation'' by Bhatti (2006), recognising 28 families and six superfamilies within an ''Order Terebrantia'', is nonphylogenetic, remaining within the phenetic tradition followed by that author in earlier publications (see Grimaldi and Engel 2005). In contrast the second sub-order, Tubulifera, includes a single family with about 3400 species (Mound 2007). This family, the Phlaeothripidae, comprises two subfamilies, and the morpho-systematics of the smaller of these, the Idolothripinae with about 700 species, has been examined extensively (Mound and Palmer 1983). In contrast, the systematic relationships amongst the species of Phlaeothripinae remain unclear, with the only proposals for formal supra-generic groupings being essentially phenetic (Priesner 1961;Bhatti 1994Bhatti , 1998 through emphasising the significance of unusual autapomorphies. In examining the Thysanoptera fauna of such a geographically large and ecologically diverse area as Australia, there are thus two serious impediments: very large numbers of undescribed species exist, and there is no satisfactory classification within which to deal with them. This paper is part of a continuing effort (Crespi et al. 2004) to overcome these problems, and is focused on the species within one sub-group of the Phlaeothripinae.
Despite an increase in the past 30 years in the number of recognised Australian species of the insect order Thysanoptera from 287 to over 700 (Mound and Minaei 2006), the current total probably represents less than 50% of the real fauna. In the two most speciesrich groups, the Terebrantia family Thripidae, and the Tubulifera subfamily Phlaeothripinae, there are many undescribed taxa represented in the Australian National Insect Collection, Canberra, and field surveys for these small insects have been essentially non-targeted over most parts of this continent. The primary objective of this paper is to give some account of the species diversity within the genus Haplothrips, this being one of the two thrips genera most frequently encountered by entomologists in Australia, and thus a group about which questions arise frequently. The second genus of particular general interest, Thrips Linnaeus, was fully examined recently (Mound and Masumoto 2005). A secondary objective of this paper is to reconsider one part of the weak suprageneric classification within the Phlaeothripinae, targeting those genera that are considered to be related to Haplothrips Amyot and Serville, and focusing particularly on genera currently known from Australia. Although far from a phylogenetic analysis, this is the first attempt in 50 years to consider the systematic relationships and biological diversity of this substantial group of insects.

Systematic problems in the Phlaeothripinae
The Phlaeothripidae is the only family recognised in the suborder Tubulifera (Mound and Morris 2007). The monophyly of this family is beyond question, and of the two recognised subfamilies, the Idolothripinae also seems likely to prove essentially monophyletic (Terry et al. 2005). However, the Phlaeothripinae is probably paraphyletic with respect to the Idolothripinae, and for the 2700 species in 360 genera currently recognised in the subfamily there is no satisfactory suprageneric classification. Priesner (1961) proposed 10 tribal and 18 subtribal names within this group, but his system was essentially phenetic, with many of his subgroups based on particularly unusual autapomorphies that are considered now to have little phylogenetic significance. Similarly, Bhatti (1994Bhatti ( , 1998) recognised 10 small families in this group, each comprising between one and 30 species, but with over 3000 species in the single-family Phlaeothripidae. An alternative approach to the Phlaeothripinae (Mound and Marullo 1996) built on original work by Stannard (1957) and recognised three informal lineages-the ''Haplothrips-lineage'', the ''Liothrips-lineage'', and the ''Phlaeothrips-lineage''. These were weakly defined on a set of character states, both morphological and biological, and the present paper concerns the members of the first and smallest of the three. Presented here is the first attempt to discuss the character states on which membership of the lineage might be based, and to list the genera that might be considered members of this lineage.
In one of 10 proposed tribes of Phlaeothripinae, the Haplothripini, Priesner listed 34 generic names, although four of these are now placed as synonyms of Haplothrips and two of Podothrips Hood. Moreover, a few genera that he listed seem only distantly related to Haplothrips. The Australian genus Asemothrips Hood has rather broad maxillary stylets and is currently placed in the phlaeothripine tribe Docessissophothripini (Mound and Palmer 1983). Similarly, as discussed below, Bamboosiella Ananthakrishnan (5Antillothrips Stannard) is best considered a member of the ''Phlaeothrips-lineage'', and this also applies to two Neotropical genera, Epomisothrips Hood and Zaliothrips Hood, listed by Priesner (Table I); the former with rather broad maxillary stylets, and the latter with the terminal antennal segments fused (Mound and Marullo 1996). More difficult to assess are the relationships of the South African genus Talitha Faure, and also the rather similar Agrothrips Jacot-Guillarmod. Both of these involve species that are divergent from the Haplothrips group in the form of the male genitalia, and have a sternal glandular area in males. Unfortunately, neither of these genera is clearly defined, and the relationships of the species remain equivocal. With these changes, the Haplothripini of Priesner represents the ''Haplothrips-lineage'' of Mound and Marullo, and this tribe is redefined below.

Radiation within the Haplothripini
Of the 34 genera listed here as members of this tribe (Table II) four have not been reexamined and are included because Priesner (1961) listed them as Haplothripini. As interpreted here, this group is largely from the Old World. Only three New World radiations are involved: the small genus Bagnalliella Karny in which the species are all associated with the leaves of Yucca, and two genera of predatory species, Leptothrips Hood and Karnyothrips Watson. A fourth New World genus, Goniothrips Hood, comprises a single aberrant Haplothrips species with the third antennal segment strongly asymmetric and Table I. Genera excluded from the Haplothrips lineage.

Antillothrips Stannard5Bamboosiella
Asemothrips Hood, 1919 Australia Bamboosiella Ananthakrishnan, 1957 Oriental Epomisothrips Hood, 1954 Neotropical Zaliothrips Hood, 1938 Neotropical bearing a single small sensorium. This species breeds in grass flowers in several southern States of the USA (Mound and Marullo 1996). Moreover, of the 250 recognised species of Haplothrips less than 20 are considered endemic to the New World, all of these being Nearctic (Mound and Marullo 1996) apart from three species of the subgenus Trybomiella Bagnall from south of the Tropic of Capricorn (Mound and Zapater 2003). Systematic studies on the Thysanoptera have come largely from the northern hemisphere, thus traditionally the genus Haplothrips has been seen as the typical element in the Haplothripini. However, 70% of Haplothrips species are Holarctic, none is from the Neotropics, and most species are associated with the flowers of Asteraceae and Poaceae. This suggests that much of the radiation in the genus may have occurred relatively recently, after the separation of Africa from South America. Most species of Haplothrips genus have the third antennal segment bearing only one or two sensoria, and these states, as discussed below, are assumed to be derived by reduction from a plesiotypic condition involving three sensoria on the third antennal segment. This plesiotypic condition occurs in the members
of six genera in the tribe. Four of these, Androthrips Karny, Euoplothrips Hood, Mesothrips Zimmermann, and Dolichothrips Karny, are endemic to the Asian/Pacific area. Species in the first three of these live in galls induced by other thrips species, whereas species of Dolichothrips live in flowers and flower buds. A fifth genus, Xylaplothrips Priesner, constitutes a difficult systematic problem because, although many species have the plesiotypic antennal sensorium arrangement and are found invading the galls of other thrips species, the type species of the genus is probably not congeneric with these kleptoparasitic species, as discussed further below. The sixth genus, Neoheegeria Schmutz, is now considered to comprise a small group of southern Palaearctic species that are often associated with the flowers of Lamiaceae such as Stachys (Minaei et al. 2007); other species that were treated in this genus by earlier authors (Priesner, 1964) are now placed in Haplothrips. An undescribed species known to be the pollinator of Macaranga trees in South-East Asia (Moog et al. 2002), and referred to as ''Neoheegeria sp.'', is more appropriately considered a species of Dolichothrips. Also similar to Dolichothrips is the monobasic New Guinea genus Tetragonothrips Moulton, despite the presence of unusually stout tergal marginal setae. The Haplothripini thus may have originated from a group of tropical leaf-feeding species, including kleptoparasitic species in galls, and radiated subsequently into the northern hemisphere as flower-feeders and as predators.

Biological diversity in the Haplothripini
This tribe apparently comprises about 34 genera worldwide, Haplothrips being the largest with almost 250 species, and 14 of the genera being monotypic. Most Haplothrips species feed and breed in flowers; very few live on leaves without feeding in flowers, although a considerable number are predators on other arthropods. Moreover, the predatory habit is typical of species in several of the smaller genera related to Haplothrips, and some of these are gall-invading species in the palaeotropics. Despite this, most of the gall-invading species in this tribe appear to be kleptoparasitic phytophages rather than predators. The preponderance of flower-living and predatory species in the Haplothripini can be contrasted with the situation in the two larger lineages of Phlaeothripinae: the ''Phlaeothrips-lineage'', in which species are mainly fungus-feeding, and the ''Liothripslineage'' in which species are mainly leaf-feeding (Mound and Marullo 1996).
Within Haplothrips, about 70% of the described species are Holarctic, where they breed particularly in the flowers of Asteraceae and Poaceae. A few breed in the flowers of several other unrelated plant families, but remarkably few feed only on leaves (Mound and Zapater 2003). Predatory species of Haplothrips in the Holarctic are typically members of the Haplothrips minutus species-group, and commonly feed on mites on the branches of trees (Mound et al. 1976;Palmer and Mound 1990). But the predatory habit also occurs amongst various unrelated species within the genus in other parts of the world, including Australia (Bailey and Caon 1986) and Japan (Kakimoto et al. 2006). Moreover, within the Haplothripini there are predatory species in at least four smaller genera. Podothrips comprises about 30 species, mainly from the Old World tropics, and these are considered to be predators of Coccoidea living on Poaceae and Bambusaceae. Leptothrips comprises about 30 species from the New World that are predators of mites and Lepidoptera eggs. Members of the two genera Karnyothrips and Xylaplothrips are usually considered to be generalist predators on other small arthropods, although there is a lack of direct evidence for most of the species (Palmer and Mound 1990;Mound and Marullo 1996). In contrast, the Asian genus Dolichothrips is a group of phytophagous species commonly associated with young buds, but with one undescribed species known to be a pollinator (Moog et al. 2002). Most of the species in the genera mentioned above are essentially monomorphic, with a body size rarely exceeding 2 mm. In contrast, the body size in species of Euoplothrips and Mesothrips is considerably larger, and these species are structurally polymorphic both within and between sexes. These species live in leaf galls, probably as kleptoparasites of other thrips species rather than as primary gall-inducers (Marullo 2001;Mound and Morris 2005).

Biological diversity amongst Australian Haplothripini
In the northern hemisphere Haplothrips species are particularly associated with the flowers of Asteraceae and Poaceae, and in Britain the genus comprises 10% of the total Thysanoptera fauna. In contrast, the composition of the Australian insect fauna is very different (Austin et al. 2004). Not only do the species of Haplothrips comprise scarcely 3% of the known thrips fauna of Australia, but they include no species specific to any of the highly diverse flora of Asteraceae on this continent, and very few that are specific to the equally large family Poaceae. The black plague thrips, H. froggatti, is abundant in grass flowers across much of Australia, and H. bellisi is associated with grasses near Darwin. H. anceps occurs in grass flowers but is particularly common in the inflorescences of certain Cyperaceae, and this family also provides the host plants for H. gahniae and H. driesseni, and possibly also for H. avius. Two native species are associated with a range of flowers, H. victoriensis in the south and H. haideeae in the north, and three apparently introduced species are also found in flowers-H. leucanthemi in the south, H. gowdeyi in the north, and H. robustus across much of the continent. Two species breed in the flowers of Amaranthaceae, H. gomphrena and H. varius, and this is possibly also true for H. ordi despite the type series having been taken from Sapindaceae. Two species have particularly unusual host associations: H. dicksoniae is associated with the sori of a tree fern, and H. salicorniae with saltmarsh Chenopodiaceae. Two species are predatory on other small arthropods, H. bituberculatus and H. collyerae, and this is probably also true of H. howei, H. lyndi, and H. ficii. Also possibly predatory is H. acaciae that is widespread across the continent, living amongst Acacia phyllodes that have been webbed by spiders. The host associations of these Australian Haplothrips species are thus very different from the host associations found in the northern hemisphere.
The other Australian members of the Haplothripini exhibit a diversity of biologies. The species of Euoplothrips and Mesothrips appear to be leaf-feeding kleptoparasites in the galls of other thrips, particularly on Ficus, whereas Androthrips species are probably predatory on gall thrips. Dolichothrips species seem to be associated with flower buds. The species in Apterygothrips and Dyothrips live in grasses and are probably phytophagous, whereas Podothrips species all live in grasses but appear to be predators on coccoids. The two recorded species of Karnyothrips, both introduced, are also scale insect predators.

Character states in the Haplothripini
Members of the Haplothripini share most of the following character states: antennal segment III with one, two or three sensoria, IV with four (rarely fewer) sensoria; antennal segment VIII distinct but broadly-based; head with maxillary bridge present; compound eyes not reduced; pronotum with notopleural sutures complete and five pairs of major setae developed; prosternal basantra well developed; metathoracic sternopleural sutures not developed; macropterous, fore wing constricted medially; pelta triangular; male with tergite IX setae S2 short and stout, sternite VIII without glandular area, pseudovirga slender.

Antennal segment III sensoria
In this tribe, the plesiomorphic condition of the third antennal segment presumably involved three sensoria, two on the outer apical margin and one on the inner. However, this condition is retained in only a few members of the tribe, including Androthrips, Dolichothrips, Euoplothrips, Mesothrips, and Neoheegeria, together with some of the species currently placed in Xylaplothrips. Most haplothripine species bear only two sensoria on this segment, one on the outer and one on the inner apical margins, but within Haplothrips, and also within Podothrips, a considerable number of apparently unrelated species have lost the inner sensorium and thus bear only one of these structures. Of particular interest is a new species described below, H. fici, in which most of the available specimens have one sensorium on the third segment, but a few specimens from the same series have two equally large sensoria. Recent examination of one paratype of the African genus Senegathrips failed to confirm the presence of more than two sensoria on the third antennal segment, despite the original description reporting four sensoria. Observations on the number of antennal sensoria require specimens to be slide mounted to a high standard, and often require the use of a 6100 oil-immersion lens.

Antennal segment IV sensoria
The plesiomorphic condition of the fourth antennal segment presumably involved four sensoria, two on the outer apical margin and two on the inner, and this condition is retained in most species of Haplothripini. However, the number has been reduced to three or to two in several unrelated groups of species, including species of Apterygothrips, Podothrips, and Haplothrips. In South Africa, H. stoffbergi Faure was described as having only three sensoria on this segment, and H. callani Faure as having only two. In Australia, H. bituberculatus usually has four sensoria on segment IV but often has only three, and the left and right antennae of single individuals commonly have a different number of sensoria. H. collyerae, described originally from New Zealand in Apterygothrips, has two major sensoria together with a third small one on the fourth segment, and two new species are described below with only two sensoria on the fourth segment, including the continent-wide H. acaciae. Most of the Haplothrips species mentioned in this paragraph have two sensoria on the third antennal segment; the exceptions are H. stoffbergi and a new Australian species, H. dicksoniae, both of which have one sensorium on the third segment and three on the fourth (also H. fici in which the third segment has either one or two sensoria). Reduction in the number of sensoria on both of these antennal segments is therefore assumed to have occurred independently more than once within the Haplothripini, and this must be taken into account when considering relationships.

Antennal segment shape
The members of Haplothripini have eight distinct antennal segments, but the eighth is more or less broadly based in almost all of the species (Figures 16-28). In contrast, the seventh segment is usually pedicellate, but is broadly based in two grass-living species in Australia. The third segment varies in shape in association with the number of sensoria that it bears; it tends to be more symmetrical when two sensoria are present than when three or only one are present. The base of the third segment is swollen into a ring-shaped ridge in the species placed in Priesneria ( Figure 5), but this structure is probably not a good indicator of relationships, because two species of Podothrips are now known with a similar swelling (Figure 28), and a ring at the base of this segment occurs in one or more species within distantly related genera, including Jacotia (see Mound 1995) and Carientothrips (see Mound 1974).

Maxillary bridge
The anterior ends of the maxillary guides are linked by a maxillary bridge in the head of all members of the Haplothripini, and the length of this bridge, that is, the distance between the maxillary stylets, differs between species. However, a maxillary bridge occurs also in a few unrelated taxa in the ''Phlaeothrips-lineage'', its presence thus being either plesiomorphic or homoplastic. The form of the bridge, and the extent to which the maxillary stylets are retracted into the head, are characters that must be used with caution, because they are often distorted during microslide preparation.

Compound eyes
Reduction in size of the compound eyes, sometimes to only a few ommatidia, is widespread amongst members of the ''Phlaeothrips-lineage'', but it is rare in the ''Liothrips-lineage'' and the Haplothripini.

Prothoracic notopleural sutures
These sutures (5epimeral sutures) delimit a pair of sclerites at the posterolateral angles of the pronotum. They are complete, extending to the posterior margin of the pronotum, in almost all the members of the Haplothripini, one exception in Australia being Dyothrips pallescens ( Figure 2).

Pronotal setae
The pronotum of Phlaeothripidae plesiotypically bears five pairs of major setae: anteromarginals (am), anteroangulars (aa), midlaterals (ml), epimerals (ep), and posteroangulars (pa). Many species in the Haplothripini have all five pairs of these setae well developed, but reduction of one or more pairs is not uncommon, and Haplothrips timori is described below with only the epimeral setae developed. The form of the apices of the major setae is often considered in the descriptions of these thrips, being variously described as ''pointed'' or ''blunt'', but the interpretation of such differences often depends on the magnification and type of microscope illumination used.

Prosternal basantra
The presence of well-developed basantra (5praepectus) appears to be an autapomorphy for the Haplothripini. However, these structures also occur in some members of the   ''Phlaeothrips-lineage'', and the weak sclerotisation of the basantra in Bamboosiella species suggests that this genus is probably not related to Haplothrips.

Fore leg armature and polymorphism
The fore tarsus of many species in this lineage bears a triangular tooth on the inner margin, this being particularly large in larger males. In the species of Karnyothrips this tooth arises closer to the apex of the fore tarsus and is directed forwards rather than at right angles to the tarsal axis ( Figure 6). A similar tarsal tooth ( Figure 41) occurs in several species of Australian Haplothrips. This ''fore tarsal tooth'' appears to be homologous with the ''inner hamus'', a prolongation of the inner apex of the fore tarsus. The different shapes and positions of the tooth presumably result from differing developmental growth rates of the tissues around the inner apex of the tarsus. This is particularly evident in species such as H. gomphrenae, in which the male has the tarsal tooth lateral but the female has a small distal tooth. The fore tibia in most Podothrips species bears a sub-apical tubercle on the inner margin ( Figure 64). In Euoplothrips species the fore tibia bears a series of tubercles on the inner margin, and the fore femur bears a long stout tubercle medially, particularly in large individuals ( Figure 7). Androthrips species have a particularly characteristic swollen fore femur, with a series of minute tubercles distal to a pointed sub-basal spur ( Figure 4). Enlarged fore legs occur in large males of various Haplothrips species, but polymorphism and enlarged fore legs are much less common in the Haplothripini than in species of the other two lineages, suggesting that male/male competition is of lesser significance among these species.

Mesopresternum
This sclerite is usually transverse and boat-shaped in the members of Haplothripini (Figures 35, 63). However, it is often eroded medially and thus reduced to a pair of lateral triangles on the meso-eusternal anterior margin ( Figure 15).

Metathoracic sternopleural sutures
These sutures, extending posteriorly from the mid-coxal cavities, are usually present in Phlaeothripidae, but are generally not developed, or only faintly indicated, in members of the Haplothripini. They are well developed in species of Talitha, and also occur in one species of Neoheegeria (Minaei et al. 2007).

Wing development and shape
The fore wings of members of the Haplothripini are more or less constricted medially, and this appears to constitute another autapomorphy of the group. Moreover, wing reduction is not common among the members of this tribe. Micropterous and apterous species are rare, and such species are usually segregated to the genus Apterygothrips; the phylogenetic significance of this is clearly limited. Two new species are described below in the genus Haplothrips of which one is known only from apterous adults of both sexes, and the other has short-winged males and fully winged females. Moreover, Apterygothrips collyerae, in which wings are present or absent, is here transferred to Haplothrips. Similarly difficult to assess is the absence of duplicated wing cilia from the posterior margin near the fore wing apex. Species of Haplothrips that lack duplicated cilia are usually placed in a subgenus, Haplothrips (Trybomiella), but this group also appears to be polyphyletic, being based on this single ''loss apomorphy''.

Pelta
The first abdominal tergite is reduced to a median pelta in Phlaeothripidae, the shape of which often seems to be a good indicator of relationships (Figures 11, 13). Within the Haplothripini the pelta is usually triangular, whereas in members of the ''Phlaeothripslineage'' the posterior margin of the pelta is broader, thus producing a bell-shaped structure as is found in members of the genus Bamboosiella.

Tergal chaetotaxy
The abdominal tergites of Phlaeothripidae typically bear two pairs of sigmoid wingretaining setae, although these setae are reduced or even absent in wingless individuals such as some members of Apterygothrips. Within the Haplothripini, species of Euoplothrips and Mesothrips have at least one extra pair of small sigmoid setae anterolateral to the main two pairs, and this is also true in Asian species of Dolichothrips. Most Haplothrips species have a series of small straight discal setae on tergites II-IV lateral to the wing-retaining setae. However, new species are described below that lack these setae. The anal setae are usually no more that 1.5 times as long as the tube, although in Karnyothrips species they are more than 2.0 times longer than the tube.

Male tergite nine
Phlaeothripidae typically have three pairs of major setae on the ninth abdominal tergite, and males in the Phlaeothripinae usually have the second pair, S2, much shorter and stouter than S1 and S3. This condition is constant among the members of the Haplothripini, although it is variable in the ''Liothrips-lineage'' sometimes even within species (Del-Claro and Mound 1996).

Male sternal glandular areas
In contrast to males in the Idolothripinae, the males of Phlaeothripinae commonly have a glandular area on the eighth abdominal sternite. Despite this, within the Haplothripini a glandular area is developed rarely on this sternite, being found only in some New World species that are placed in Karnyothrips, as well as in species of Talitha and Agrothrips.

Male pseudovirga
The shape of the apex of the aedeagus has been used as a species discriminant in the genus Haplothrips. However, microslide preparation of male genitalia is difficult, and no author ever states how many such slides have been prepared and studied. There is thus no statistical basis for evaluating the possibility of variation in these structures. In the studies reported here, only one Australian species, H. froggatti, has been noted to have a distinctive narrow apex to the pseudovirga, sometimes with a pair of small apical swellings; this has been confirmed with many slides but is very difficult to photograph. In contrast, despite the preparation of many slides, no differences have been found between the pseudovirgal apices of the Australian species H. anceps, H. driesseni, and H. victoriensis, these being essentially similar to the genitalia of the European species H. leucanthemi (Figures 29-32).

Australian taxa excluded from Haplothripini
Contrary to Pitkin (1973Pitkin ( , 1976) the following taxa are considered here to be unrelated to Haplothrips.
Described from a single inadequately slide-mounted specimen (Figure 1), this is one of several species described by Girault that cannot at present be recognised satisfactorily (Mound 1996). Superficially it is similar to Haplothrips lyndi described below, but it has a much longer basal stem to the fourth antennal segment and is probably only distantly related. The third antennal segment bears only one sensorium, the fore wings have no duplicated cilia and are possibly not constricted medially, and the maxillary bridge, prosternal sclerites, and pelta are not visible. There is thus no reason to consider the species a member of Haplothrips, as decided provisionally by Pitkin (1973), although placement in the North American genus Bagnalliella is equally unjustified. This genus was treated by Pitkin (1973Pitkin ( , 1976Pitkin ( , 1977, under the synonymic name Antillothrips, as being related to Haplothrips. However, it seems more likely to be a member of the ''Phlaeothrips-lineage'' (Okajima 2006), because the expanded base of the pelta is unlike that of any member of the Haplothripini, the maxillary bridge is not developed, and the prosternal basantra are very weakly developed or entirely absent. In total, 27 species are listed in the genus, mostly from South-East Asia (Okajima 1995), with one species widespread around the world in association with various Poaceae, and two known from Australia.   The species in this genus are considered ''gall thrips'' but they are probably kleptoparasites or predators rather than gall inducers. A total of 12 species have been named, but apart from a key to five species from India (Ananthakrishnan and Sen 1980), most of these cannot be identified without re-examining original material. Of these named species, 11 are from the Oriental Region: four were described from India or Sri Lanka, four from Indonesia between Java and Papua, and one each from China, Taiwan, and the Philippines. The 12th species is based on a single male specimen from Cameroon that is presumed to have been lost (Pitkin and Mound 1973).
Described from two females and one male taken on the leaves of Monstera deliciosa at Koitaki, New Guinea, this species is here recorded from Australia for the first time. It has been taken from rolled leaf galls on five unrelated plants in the Northern Territory: Alyxia spicata (Apocynaceae), Antidesma ghesembilla (Euphorbiaceae) and Smilax australis (Smilacaceae) at Darwin, Timonius timor (Rubiaceae) at Kakadu, and Choriceras tricorne (Euphorbiaceae) on the Cobourg Peninsula. The galls on these plants were induced by different undescribed species of the Teuchothrips complex. As indicated by Ritchie (1974), this species is similar to A. flavitibia Moulton, described on one female from northern India. It is also similar to A. kurosawai Priesner from the Philippines, although Reyes (1994) incorrectly stated that that species has only two sensoria on the third antennal segment whereas it actually has three (pers. comm. Richard zur Strassen, in litt. 2006). These three names possibly represent a single widespread species that has brown femora but yellow tibiae.

Recognition
Varying considerably in body size; body and femora brown, tibiae and tarsi yellow; antennal segments III-VI largely yellow; fore wings pale. Antennae eight-segmented, III with three sensoria, IV with four. Postocular setae longer than distance between eyes, weakly capitate. Pronotal am setae minute, four pairs of major setae long and capitate. Fore femur inner margin in both sexes with sub-basal spur and an irregular row of small tubercles (Figure 4), this spur and also fore tarsal tooth largest in large individuals. Prosternal basantra present; mesopresternum almost complete medially. Metanotum with several small setae on anterior half. Fore wings constricted medially, with 5-10 duplicated cilia. Tergites II-VI with two pairs of sigmoid setae; tergite IX setae S1 weakly capitate and about 0.8 as long as tube. Male tergite IX S2 short and stout.
This genus was last redefined by zur Strassen (1966) to include eight species, but has been used subsequently to include any species resembling Haplothrips in general appearance but either lacking wings or with the wing-retaining setae reduced. Currently there are 41 species worldwide that are listed in this genus, but these are unlikely to represent a single lineage. Wing reduction and wing loss have clearly occurred more than once within the Haplothripini, and not always in closely related species, and Apterygothrips is presumably a polyphyletic assemblage. Furthermore, in this paper one fully apterous species is described in Haplothrips, a second new species is described in which the females are macropterous but the males micropterous, and the wing polymorphic species, Apterygothrips collyerae, is here transferred to Haplothrips.
Among the species currently placed in Apterygothrips there appears to be one speciesgroup of grass-living thrips, and the only known Australian member of the genus, A. australis, belongs to this group. A second Australian species of wingless Haplothripini has been studied from Kununurra in the north of Western Australia. This species was living on dead twigs, and apparently represents an undescribed apterous species that is related to the ''Apterygothrips'' species from dead wood in India (Pitkin 1976). One of these Indian species, with two sensoria on the fourth antennal segment and the pterothorax exceptionally apteriform, was designated type species of Tamilthrips Bhatti (1995), but the phylogenetic significance of such a genus based essentially on extreme wing loss remains open to question.

Apterygothrips australis Pitkin
Apterygothrips australis Pitkin 1973, p 328. This is a common species in grasses that is widespread across southern Australia, having been taken near Brisbane, Sydney, Canberra, Adelaide, and Perth. Pitkin (1973) refers to two macropterous females out of a total of 71 specimens examined, but pointed out that some micropterous individuals have the tergal wing-retaining setae more or less developed.

Recognition
Usually micropterous, wing lobe less than 0.25 of thorax width; colour light brown, antennal segment III and all tarsi yellow; antennal segment III with one sensorium, IV with two; head with maxillary stylets not retracted to postocular setae, about one-third of head width apart; postocular setae capitate, also all five pronotal setae, one sub-basal wing seta, and setae S1 and S2 on tergite IX; fore tarsus with small tooth. The genus Membrothrips was erected for a single species, and the Australian species discussed below was added subsequently. The genus was differentiated from the South-East Asian genus Dolichothrips because of the absence of additional pairs of sigmoid setae lateral to the two pairs of tergal wing-retaining setae, and also because the pseudovirga of the male genitalia is not slender. The first of these character states appears to be related to body size, and the second fails to take into account the variation in male genitalia amongst Haplothripini that are not associated with flowers. M. indicus has not been re-examined for this paper, but M. reuteri is essentially similar in structure and colour to Dolichothrips macarangae Moulton that lives on Macaranga tanarius in Taiwan. The differences between them in tergal chaetotaxy appear to be related to the differences in their body size. All of the D. reuteri specimens collected from Macaranga in Australia are much smaller than those collected from this plant in Taiwan. These smaller specimens all lack the additional tergal sigmoid setae that are present in the Taiwan specimens. Moreover, the species that is known to pollinate Macaranga species in Asia (Moog et al. 2002), and was referred to as ''Neoheegeria sp.'', varies considerably in size, and the larger individuals have additional sigmoid setae on the tergites but not the smaller specimens from the same series. The genus Dolichothrips is closely related to the genus Mesothrips, although the species of one live in flowers and buds, whereas the species of the second live in galls.

Recognition
Dark brown species with little sexual dimorphism; antennae eight-segmented, III with three sensoria, IV with four; head longer than wide, postocular setae acute, mouth cone pointed and extending between fore coxae ( Figure 3); fore tarsus with tooth minute in female, larger in male; fore wing constricted medially, with duplicated cilia; basantra well developed, mesopresternum forming two triangles; metanotum weakly reticulate; tergites usually with additional pairs of wing-retaining setae lateral to the two major pairs; tergite IX setae long and acute.

Dolichothrips reuteri (Karny)
Liothrips reuteri Karny 1920, p 40. Reported to be based originally on a single female collected at ''Cedar Creek'' Queensland (deposited in the Museum at Stockholm), a second female with identical data is present in the collections of the Senckenberg Museum, Frankfurt. This dark brown species with bright yellow tibiae has been collected more recently in considerable numbers at Paluma and Cape Tribulation in northern Queensland, and also around Darwin in Northern Territory. At these localities it was found in the apical buds of both Macaranga tanarius and Hibiscus tiliaceus, two plants that grow commonly along the shore. None of the available specimens has additional sigmoid setae on the abdominal tergites.
Although considered a genus by Bhatti (1995), the phylogenetic position of the single species involved possibly lies within the genus Haplothrips. It differs from the species in that genus in having the prothoracic notopleural sutures incomplete ( Figure 2). The statement by Bhatti (1995) that it also differs in having a complete mesopresternum is incorrect, because this character state occurs in H. aculeatus, the type species of Haplothrips, as well as several other members of the genus. Although stated by Kudo to have a long mouth cone, the apparent length of this structure depends largely on slide-mounting procedures, because the mouth cone is directed ventrally in life and can rotate posteriorly under coverslip pressure. Despite the lack of duplicated wing cilia, the relationships of Dyothrips probably do not involve Trybomiella, but the genus is retained here until such time as molecular data become available.
Described originally from Queensland, and subsequently from China, the synonyms of this species are listed by Pitkin (1973). It is now recorded from Australia, Bangladesh, China, Fiji, India, Pakistan, Thailand, Taiwan, and Japan (Okajima 2006), and also New Caledonia (Bournier and Mound 2000). A grass-living species, specimens have been seen in Australia from coastal regions between Brisbane and Cairns, and also around Darwin.

Recognition
Macropterous and distinctively bicoloured; body light brown with metathorax and first abdominal segment yellow, antennal segments III-VI yellow, also all tarsi and apices of tibiae; wing pale. Head longer than wide, antennal segment III with one sensorium, IV with four; postocular setae capitate, also all five pronotal setae and tergite IX S1 and S2; fore tarsal tooth not present in either sex; prothoracic notopleural sutures incomplete.
Six species are known in this genus, two from Australia, three from Pacific islands, and one from India, although the identity and affinities of the latter species remain in doubt. Marullo (2001) provided a key to these species, and indicated that they appear to be kleptoparasites within leaf galls induced by other thrips species. In general Euoplothrips are associated with Gynaikothrips species in Ficus rolled leaves, although E. bagnalli occurs in leaves galled by various thrips species on several unrelated plants.

Recognition
Dark brown species exhibiting size-correlated polymorphism in fore leg armature; antennae eight-segmented, III and IV each with three sensoria; fore femur of both sexes with median spur on inner margin (Figure 7), fore tibia with two small tubercles on inner margin, fore tarsus with stout tooth; prosternal basantra small; fore wings constricted medially; tergites II-VI with additional sigmoid setae lateral to the major wing-retaining setae. Described from northern Queensland, this species has been found along the east coast as far south as Sydney, and was described under a synonymic name from Papua New Guinea (Marullo 2001). It has been taken from thrips galls on the following unrelated plants: Alyxia spicata (Apocynaceae), Citriobatus pauciflorus (Pittosporaceae), Ficus macrophylla (Moraceae), Smilax australis (Smilacaceae), and Tetrastigma nitens (Vitaceae). The smallest male has the tubercles scarcely developed on the fore femur and fore tibia.

Euoplothrips platypodae Marullo
Euoplothrips platypodae Marullo 2001, p 97. This species is closely similar to E. bagnalli, but is known only from Ficus platypoda in the northern part of Western Australia, where it was found living in galls induced by an undescribed species of Gynaikothrips.
The subgenus Trybomiella has been used to include 18 species of Haplothrips, each of which lacks fore wing duplicated cilia, but the variation amongst these species suggests that the subgenus is probably polyphyletic. Despite this, some of the species from Australia, South Africa, and southern South America (Mound and Zapater 2003) are sufficiently similar to each other to suggest that they might constitute a lineage distinct from the rest of the genus. In the absence of any molecular data bearing on this morphologically homogeneous group, the genus Haplothrips is interpreted broadly. Molecular data are needed to test relationships, both within this large genus, and with such satellite genera as Apterygothrips, Karnyothrips, and Xylaplothrips. The genus now includes rather more than 260 described species worldwide.

Recognition
Usually macropterous, rarely with wings reduced or absent. Body colour usually brown, antennal segment III commonly paler, fore wings usually pale except for basal area. Head usually a little longer than wide, maxillary stylets deeply retracted, rarely less than onethird of head width apart, maxillary bridge well developed. Vertex not strongly sculptured, postocular setae usually capitate and extending beyond posterior margin of eyes, rarely short and acute. Antennal segment III with one or two sensoria (rarely none), segment IV usually with four sensoria but sometimes only two or three; antennal segment VIII usually short and broad at base. Pronotum usually with five pairs of capitate setae, but sometimes one or more of these no larger than discal setae; epimeral sutures complete; prosternum with four sclerites, paired basantra and paired ferna, also a median spinasternum; mesopresternum either complete, or eroded to paired lateral triangles. Metanotum weakly sculptured, median pair of setae usually arising on anterior half of sclerite. Fore tarsus usually with tooth present laterally or near inner apex; tooth sometimes large in large males; mid and hind tarsi with ventrolateral hamus present. Fore wing constricted medially, duplicated cilia usually present; three pairs of sub-basal setae. Male with no glandular area on sternite VIII; tergite IX setae S2 short and stout; pseudovirga of aedeagus slender.
Female macroptera. Body colour light brown to brown, mid and hind tarsi brownish yellow, fore tarsi and basal half of antennal segment III yellow; major setae weakly shaded, fore wing shaded at base, then pale but variably shaded medially and distally. Head as broad as long, maxillary stylets wide apart, retracted to postocular setae, maxillary bridge complete ( Figure 9); postocular setae capitate, extending beyond posterior margin of compound eyes. Antennal segments III-IV each with two sensoria; segment VIII distinctly constricted at base ( Figure 16). Pronotum with five pairs of long, capitate setae. Metanotum reticulate medially, median setae arise on anterior half of sclerite. Mesopresternum transversely boatshaped but slender medially. Fore tarsal tooth small, acute, directed forwards. Fore wing with about five duplicated cilia, sub-basal setae capitate. Tergite II ( Figure 11) with wingretaining setae weak, III-VII with anterior pair weaker than posterior pair; posteromarginal setae S1 long and capitate on tergites II-VII, but shorter on VIII; tergite IX setae S1 bluntly pointed, S2 finely acute.

Comments
On four occasions colonies of this species have been taken from webbed phyllodes of Acacia species, and at three of these sites the webbing was certainly produced by spiders not by Lepidoptera (for the fourth sample the field notes remain equivocal). Despite this association, it is not known if the species is predatory or phytophagous. It is one of only two known species in Haplothrips with two sensoria on both the third and fourth antennal segments.
No consistent differences have been found between specimens, including type specimens, identified as H. anceps and as H. angustus, despite the statement by the original author that the latter species has longer antennal segments. Pitkin (1973) suggested that there was a difference in the length/width ratio of the third antennal segment (1.3-1.8 in anceps, but 1.8-2.2 in angustus), but this does not appear to be a satisfactory discriminant because of the variation within individual field samples. Moreover, it does not correlate satisfactorily with any other structural differences, such as the form of setal apices, and the male genitalia of the two forms are identical. Although treated here as one species, specimens from northern Australia commonly have the third antennal segment rather shorter (anceps-form) than specimens from southern Australia (angustus-form), but some genetic differences are to be expected between populations spread over an area from Adelaide to Cape York. In structure this species is closely similar to H. victoriensis, and the male pseudovirga of these species cannot be distinguished satisfactorily. This group of common thrips requires more detailed study, with particular attention to the host plants on which breeding occurs. In southern Australia H. anceps appears to be associated with flowering sedges such as Eleocharis (Cyperaceae), but samples are often taken from various Poaceae. The host records from northern Australia are less clear.

Recognition
Body and legs dark brown, mid and hind tarsi brownish yellow, fore tarsi and apex of fore tibiae yellow, at least basal half of antennal segment III yellow, IV light brown; pronotal major setae weakly shaded; fore wings pale except at base. Head with maxillary stylets about one-third of head width apart ( Figure 10). Antennal segment III with one sensorium, IV with four. Pronotum with five pairs of capitate setae, am setae short; mesopresternum transverse and boat-shaped. Fore tarsal tooth in female minute but variably large in male. Metanotum weakly reticulate medially, median setal pair arising on anterior half; fore wing sub-basal setae S1 and S2 capitate, S3 capitate or pointed; about eight duplicated cilia present. Pelta triangular ( Figure 12); tergite IX setae S1 and S2 blunt. Male pseudovirga with apex spoon-shaped ( Figure 29). Haplothrips angusi sp. nov.

Measurements of one female (in
Female macroptera. Body colour light brown, legs extensively yellow; antennal segment III yellow, IV-VIII brown; major setae pale; fore wings weakly shaded. Head longer than wide, maxillary stylets slightly less than one-third of head width apart, retracted just anterior to postocular setae, maxillary bridge faint; postocular setae long and capitate. Antennal segment III conical with one small sensorium, IV with two sensoria; segment VIII broad at base (Figure 18). Pronotum with five pairs of long, capitate setae; basantra unusually small; mesopresternum eroded and very slender medially (Figure 15). Mesonotal lateral setae capitate. Metanotum with no sculpture medially, median setae arise on anterior half of sclerite. Fore tarsal tooth minute, scarcely visible. Fore wing with five duplicated cilia, subbasal setae capitate. Pelta relatively large with rounded lateral margins ( Figure 13). Tergites II-VII with posteromarginal setae S1 long and capitate; tergite IX setae S1 long and acute; tube short, anal setae long.

Comments
The single specimen on which this species is based is presumably teneral, and more mature individuals can be expected to have the body and basal antennal segments darker. However, the species has an unusual group of character states, including the antennal sensorium formula, the minute fore tarsal tooth, and the reduced prosternal sclerites (Figure 15), and would be very difficult to place to genus using the concepts of previous authors.

Haplothrips avius sp. nov.
Female macroptera. Body colour brown, all tarsi and apices of tibiae yellow, antennal segment III yellow, major setae pale, fore wing shaded medially and at base. Head slightly longer than wide, maxillary stylets one-third of head width apart, retracted anterior to postocular setae, maxillary bridge complete medially ( Figure 14); postocular setae capitate, extending just beyond posterior margin of compound eyes. Antennal segment III with two sensoria, IV with four sensoria; segment VIII relatively elongate (Figure 19). Pronotum with five pairs of capitate setae. Mesonotal lateral setae capitate. Metanotal median setae arise on anterior half of sclerite, with no small discal setae on anterior half.

Comments
Despite being collected at sites more than 2000 km apart no significant differences can be seen between the specimens from South Australia and the one from northern Queensland. This is a much smaller species than H. driesseni described below, with the maxillary stylets further apart, but it is probably yet another species that is associated with the inflorescences of particular Cyperaceae. Despite the lack of fore wing duplicated cilia, neither of these species is closely similar in other character states to the typical members of subgenus Trybomiella, such as H. robustus or even H. timori.

Haplothrips bellisi sp. nov.
Female macroptera. Body colour light brown to brown, all tarsi and apices of fore tibiae yellow, antennal segment III brownish yellow, IV light brown, major setae and fore wing pale. Head as broad as long, maxillary stylets wide apart, almost V-shaped, not retracted to postocular setae, maxillary bridge incomplete medially; postocular setae capitate, extending beyond posterior margin of compound eyes. Antennal segments III-IV relatively short with sensoria well-developed, one on III and two on IV; segment VII broad at base. Pronotal am setae no larger than discal setae (Figure 33), four pairs of major setae capitate. Mesonotal lateral setae minute. Metanotal median setae arise on posterior half of sclerite, with 4-10 small discal setae on anterior half. Mesopresternum usually fully divided into two triangles. Fore femur with external apical margin slightly thickened; fore tarsal tooth small, acute, arising at inner apex of tarsus. Fore wing with two to five duplicated cilia, sub-basal setae S1 and S2 capitate, S3 longer and acute. Tergites II-III with only one or two discal setae lateral to wing-retaining setae; posteromarginal setae S1 broadly capitate on tergites II-VII, but weakly capitate on VIII-IX; S2 on IX blunt.

Comments
This is one of two new species described here, in addition to the widespread H. froggatti, in which the anterior area of the metanotum bears several minor setae. H. bellisi also resembles H. froggatti in having antennal segment VII broad at the base, not pedicellate. However, the external apical margin of the fore femora is slightly recurved, particularly in males, much as in species of the Ethiopian genus Chiraplothrips. Similar fore femora are also known in the unrelated species of the Thripidae genera, Chirothrips Haliday and Arorathrips Bhatti as well as in the Aeolothripidae species, Desmothrips chirus Mound and Marullo (1998), and it is probably significant that all of these thrips live in grass flowers.

Haplothrips bituberculatus (Girault)
Podothrips bituberculatus Girault 1927, p 2. This species is widespread but never abundant across the whole continent of Australia. Judging from collecting records it is probably predatory on other small arthropods, because only a few specimens are ever taken at one time, and these are found on dead twigs, in galls, and on leaves. In structure it is variable both within and between sites. The number of sensoria on the fourth antennal segment is either three or four, even on the same individual, although most specimens studied from southern Australia have four, and most specimens studied from northern Australia have three. In males, the pronotal anteromarginal setae are often shorter, and the fourth antennal segment longer, than in females. The fore wings are distinctly shaded medially, although specimens from Kununurra in the north of Western Australia have the wings paler.

Recognition
Body and femora dark brown, tarsi yellow, hind tibiae yellow at base and apex but variably brown medially; antennal segment III brownish yellow, remaining segments brown; pronotal major setae light brown; fore wings shaded medially. Head slightly longer than wide; maxillary stylets almost V-shaped, maxillary bridge about one-third of head width; antennal segment III with two sensoria, IV with three or four. Pronotum with five pairs of capitate setae, am setae slender; mesopresternum transverse. Fore tarsal tooth in female small but variably large in male. Metanotum very weakly reticulate medially, median setal pair arising on anterior half; fore wing sub-basal setae capitate; five to seven duplicated cilia present. Tergite IX setae S1 and S2 pointed but not acute.
Large males with two small, seta-bearing tubercles on inner margin of fore tibiae, and fore femora and fore tarsal tooth enlarged; small males with slender fore legs and no tibial tubercles; pronotal anteromarginal setae sometimes small, even no larger than discal setae. This species was described as widespread in the North and South Islands of New Zealand, where it was recognised as a predator on the eggs of the mite species, Panonychus ulmi. The original description also referred to two specimens collected near Hobart, Tasmania, and recent observations by John Ireson (Tasmanian Institute of Agricultural Research) indicate that the larvae of this thrips are predatory on the larvae of a thrips species (Sericothrips staphylinus) that is useful in the biological control of gorse (Ulex europea). A single specimen has also been studied from north-western Tasmania, near Trowutta.

Measurements of one female (in
As discussed above, the genus Apterygothrips is a poorly defined assemblage of species that lack wings, or have the tergal wing-retaining setae reduced or absent, and there is no clear evidence that it is anything other than polyphyletic. Considering the variation in wing form and development, also the number of antennal sensoria, among the Australian species of the genus Haplothrips it seems most appropriate to place collyerae in this genus.

Recognition
Macropterous or micropterous. Body colour brown, mid and hind tibiae yellow at base but shaded brown distally and along external margin; all tarsi yellow; antennal segment III yellow, remaining segments brown ( Figure 20); major setae pale, fore wing shaded except in basal quarter. Head longer than wide, maxillary stylets about one-third of head width apart, retracted to postocular setae, maxillary bridge complete ( Figure 34); postocular setae blunt. Antennal segments III-IV with sensoria slender two on III and two +1 on IV. Pronotal am setae no larger than discal setae, aa and ml setae also short. Mesonotum and metanotum without sculpture medially. Mesopresternum transverse (Figure 35). Fore tarsal tooth small, acute, arising at inner apex of tarsus. Fore wing without duplicated cilia, sub-basal setae blunt; fore wing of micropterae shorter than width of thorax. Tergite II with wing-retaining setae minute (Figure 36), III-VII each with two pairs; tergites II-VII posteromarginal setae S1 weakly capitate to blunt; tergite IX setae acute.
Female macroptera. Body colour brown to light brown, all tarsi and apices of tibiae yellow, hind tibiae more extensively yellow except at base; antennal segment III yellow, IV-V light brown, remaining segments dark ( Figure 21); major setae light brown, fore wing weakly shaded medially. Head slightly longer than wide ( Figure 44); maxillary stylets at least onethird of head width apart, retracted to postocular setae, maxillary bridge complete; postocular setae capitate, extending beyond posterior margin of compound eyes. Antennal segment III slender (Figure 21), with one sensorium, IV with two or three sensoria. Pronotal am setae no larger than discal setae, four pairs of major setae capitate. Mesonotal lateral setae capitate. Metanotal median setae arise on anterior half of sclerite. Mesopresternum transverse. Fore tarsal tooth small, directed forwards at inner apex of tarsus. Fore wing with three to four duplicated cilia, sub-basal setae capitate. Tergites II-III with almost no sculpture and only one or two discal setae lateral to wing-retaining setae ( Figure 37); posteromarginal setae S1 long and capitate on tergites II-VII, but short and capitate on VIII; tergite IX S1 long and blunt, S2 acute; wing-retaining setae on tergite II small, III-VI with anterior pair smaller than posterior pair. Male microptera. Similar to female, but fore wing slightly shorter than width of pterothorax; fore tarsal tooth very small; tergite IX setae S1 weakly capitate, S2 short and stout; apex of pseudovirga broad.

Material examined
Holotype R macroptera: Australian Capital Territory, Tidbinbilla, from Dicksonia fronds with sori, 25 April 1999 (LAM 3685). Paratypes: 3R, 5" taken with holotype; same locality and plant species, 1R, 3" with larvae and pupa, 16 January 1999; 3R, 2", 31 July 1999; 7R, 5" with larvae, 30 September 2001. New South Wales, 35 km SE of Queanbeyan, Tallaganda Forest, 15R, 15" with larvae from Dicksonia fronds with sori, 24 June 2006. This relatively small species is unusual within the genus, both in biology and structure. It has been found, together with its bright red larvae and pupae, among the sori on mature fronds of the tree fern, Dicksonia antarctica. Moreover, it has been found only where the tree ferns were growing in particularly moist situations, almost within the splash zone of a stream. The species has not been taken from the similar-looking Cyathea tree ferns where these grow in the same area. The first abdominal tergite, the pelta, is unusual in shape, with the anterior margin usually transverse (Figure 37), and all of the available males are micropterous with a rather broad apex to the aedeagus. The shaded fore wings resemble those of the widespread species, H. bituberculatus, but the hind tibiae are brown at the base but yellow medially.
Female macroptera. Body colour brown to dark brown, tarsi and base of antennal segment III yellowish brown; fore wing shaded medially. Head longer than wide ( Figure 45); maxillary stylets less than one-fifth of head width apart, retracted anterior to postocular setae, maxillary bridge complete; postocular setae capitate, extending to posterior margin of compound eyes. Antennal segment III with two slender sensoria, IV with four sensoria. Pronotal am setae no larger than discal setae, four pairs of major setae capitate but aa and ml small. Mesonotal lateral setae capitate. Metanotum without sculpture medially, median setae arise medially on sclerite. Mesopresternum transverse. Fore tarsal tooth small, at inner apex of tarsus. Fore wing with no duplicated cilia, sub-basal setae capitate. Tergites II-III with neither sculpture nor discal setae lateral to wing-retaining setae ( Figure 38); posteromarginal setae S1 long and capitate on tergites II-VII; tergite IX S1 capitate, S2 weakly capitate.
Male macroptera. Similar to female in colour and structure; tergite IX setae S2 short and stout; apex of pseudovirga narrowly spoon-shaped.

Comments
The host plant of this species remains unknown, although it is likely to be a montane species of Juncaceae or Cyperaceae. The head is relatively long and the stylets relatively close together, and it is one of the few Haplothrips species with a distinctly shaded fore wing.
Female macroptera. Body colour brown, tarsi and tibiae yellow, femora yellow washed with brown on external margins; antennal segments III-V largely yellow, VI brownish yellow, VII-VIII brown ( Figure 22); major setae and fore wing pale. Head longer than wide, maxillary stylets more than one-third of head width apart, maxillary bridge complete; postocular setae long and capitate, arising close to posterior margin of compound eyes. Antennal segment III with two sensoria or with only one (inner sensorium reduced or absent), IV with three sensoria; segment VIII broad at base. Pronotum with five pairs of long broadly capitate setae. Mesonotal lateral setae capitate. Metanotum with longitudinal lines of sculpture on posterior half, median setae arise on anterior half of sclerite. Mesopresternum boat-shaped. Fore tarsal tooth well developed. Fore wing with three to five duplicated cilia, sub-basal setae capitate. Tergite II with sculpture but no discal setae lateral to wing-retaining setae ( Figure 39); tergites II-VIII with posteromarginal setae S1 long and capitate; tergite IX setae S1 and S2 long and capitate ( Figure 40).
Measurements of holotype female (in mm). Body length 1500. Head, length 160; median width 120; postocular setae 16. Pronotum, length 110; width 220; major setae am 20, aa 14, ml 22, epim 40, pa 36. Fore wing length 520; sub-basal setae 28, 30, 48. Tergite IX setae S1 50, S2 60, S3 70. Tube length 100; basal width 50; anal setae 110. Antennal segments 36,38,38,36,24. Male macroptera. Similar to female; fore femora and fore tarsal tooth sometimes large; tergite IX setae S2 short and pointed. This species is remarkable, both for its antennal sensorium formula and for the capitate setae on the ninth abdominal segment. The variation in number of sensoria on the third antennal segment is particularly curious, because such variation is usually accompanied by variation in size of the affected sensoria. But in the available specimens of this species the sensorium on the inner apex of the third segment is either well developed and equal in size to the outer sensorium, or entirely absent. All of the specimens from Lake Argyle have only one sensorium present on both antennae, and this is also true of one of the females from Kununurra, but the other two female paratypes and one male paratype have two large sensoria on the third segment of both antennae. At two localities this thrips was taken from leaves in association with Ascirtothrips arafura Mound, on whose larvae it was possibly feeding. The larvae of H. fici are predominantly yellow, although the pupae have a little red pigment.

Haplothrips froggatti Hood
Haplothrips froggatti Hood 1918, p 130. This species has been referred to as ''the black plague thrips'', because it sometimes occurs in such vast numbers across Central Australia. It is distinguished from all but two other members of the genus in Australia by the presence of extra setae on the anterior half of the metanotum (Figure 42), also by the unusually short and dark third antennal segment ( Figure 23). It is a grass-living species and, although widespread across Australia, it is associated primarily with the arid areas. No specimens have been studied from Tasmania, Victoria, southern South Australia, or southern Western Australia. Similarly no specimens have been seen from the moist areas of north-eastern Queensland nor from Darwin, although it is common around Kununurra in the north of Western Australia.

Recognition
Body and legs dark brown, fore tarsi and base of antennal segments III and IV lighter. Maxillary stylets more than one-third of head width apart; antennal segment III short with one sensorium, IV with four. Pronotum with five pairs of capitate setae; mesopresternum transverse. Fore tarsal tooth absent in female, usually small in male. Metanotum with several pairs of minor setae on anterior half; fore wing with sub-basal setae all capitate, 7-13 duplicated cilia present. Tergite IX S1 long and weakly capitate, S2 acute. Male pseudovirga with apex slender and bearing distinctive apical lobes.
Measurements of one female (in mm) (Queensland, Birdsville, April 1998 Female macroptera. Body colour brown, tarsi yellowish brown, antennal segment III mainly light brown but with base yellow, remaining segments brown; major setae pale, fore wing shaded medially. Head as broad as long, maxillary stylets about one-third of head width apart, maxillary bridge complete; postocular setae acute, arising close to posterior margin of compound eyes. Antennal segments III and IV each with two sensoria; segment VIII broad at base, VII with pedicel relatively broad, VI with apex broadly truncate ( Figure 41). Pronotum with setae aa, am, and ml scarcely longer than discal setae, epim and pa setae bluntly pointed. Mesonotal lateral setae small and pointed. Metanotum very weakly reticulate medially, median setae arise on anterior half of sclerite. Mesopresternum transversely boat-shaped but extending forward medially toward spinasternum. Fore tarsal tooth small, acute and slightly hooked, arising at anterior margin of tarsus. Fore wing with about five duplicated cilia, sub-basal setae S1 and S2 blunt, S3 acute. Tergites II-III with no discal setae lateral to wing-retaining setae ( Figure 43); tergites II-V with posteromarginal setae S1 long and capitate, VI-VII with these setae acute; tergite IX setae long and finely acute.

Comments
Taken only from the inflorescences, not the leaves, of Gahnia, this species was considered in the field to be the common H. anceps that occurs widely on Cyperaceae. Subsequent examination indicated that it is one of only two Haplothrips species with two sensoria on both the third and fourth antennal segments. It is an unusual species in that only two pairs of pronotal setae are larger than the discal setae, and the pedicel of antennal segment VII is relatively broad.
Female macroptera. Body colour brown with red internal pigment, fore tarsi and apices of fore tibiae paler, also antennal segment III ( Figure 24); major setae pale, fore wings pale except at base. Head as wide as long, maxillary stylets slightly less than one-third of head width apart, retracted to compound eyes, maxillary bridge complete; postocular setae small but capitate, extending just beyond posterior margin of compound eyes. Antennal segment III with two sensoria, IV with four sensoria. Pronotum with four pairs of weakly capitate setae, ml setae no larger than discals, am setae scarcely twice as long as discals. Mesonotal lateral setae capitate. Metanotum reticulate medially, median setae arise on anterior half of sclerite. Mesopresternum eroded to two small lateral triangles. Fore tarsal tooth very small, arising at inner apex of tarsus. Fore wing with no duplicated cilia, sub-basal setae capitate. Pelta triangular; tergites II-IV with many lines of sculpture and several discal setae lateral to wing-retaining setae; posteromarginal setae S1 capitate on tergites II-VIII; tergite IX setae S1 and S2 bluntly rounded.

Comments
This species and H. varius are both associated with the flowers of Amaranthaceae in the arid zone, but the setal lengths are strikingly different in the available samples, as indicated in the key. It is smaller than H. robustus, with the major setae much smaller and the male genitalia distinctive. The larvae are bright red.
Although described from Barbados, this species is probably African in origin (Mound and Marullo 1996) but is now widespread around the tropics and sub-tropics in various flowers. In Australia it is not commonly collected, and males have not been found here. It seems to be associated with areas of domestic gardens, having been collected around Sydney and Brisbane, as well as at Port Douglas, Kununurra, and Melville Island. It is fairly distinctive in appearance, with the fourth antennal rather large and yellow, and the setae on the ninth tergite longer than the tube and acute.

Recognition
Body colour brown to dark brown, fore tarsi and apices of fore tibiae yellow, antennal segments III-V yellow (Figure 25), VI yellow with apex shaded; major pronotal setae brown, fore wing pale. Head slightly longer than wide, maxillary stylets one-third of head width apart, retracted to postocular setae, maxillary bridge complete; postocular setae weakly capitate. Antennal segment III with two sensoria, IV with four sensoria. Pronotum with five pairs of capitate setae. Metanotum weakly reticulate medially, median setae arise on anterior half of sclerite. Mesopresternum broadly boat-shaped but slender medially. Fore tarsal tooth minute in female (larger in male). Fore wing with about seven duplicated cilia, sub-basal setae capitate. Tergite IX setae S1 and S2 finely acute, slightly longer than tube. Female macroptera. Body colour light brown to brown; all tarsi yellow, also apical fifth of hind tibiae, apical third of mid-tibiae, and most of the fore tibiae; antennal segments III-VI yellow with varying shades of brown apically; major pronotal setae light brown, fore wing pale. Head almost as broad as long, maxillary stylets at least one-third of head width apart, retracted almost to postocular setae, maxillary bridge complete; postocular setae weakly capitate. Antennal segment III with two sensoria, IV with four sensoria. Pronotum with five pairs of capitate setae. Mesonotal lateral setae capitate. Metanotal median setae arise medially on sclerite, usually with two to four small discal setae on anterior half. Mesopresternum broadly boat-shaped. Fore tarsal tooth absent. Fore wing with about 10 duplicated cilia, sub-basal setae S1 and S2 capitate, S3 longer and bluntly pointed. Tergite II with no discal setae lateral to wing-retaining setae; posteromarginal setae S1 long and capitate on tergites II-VI, but blunt on VII-VIII; tergite IX S1 bluntly pointed, S2 finely acute.

Comments
The sharply yellow apex of the middle and hind tibiae appears to be diagnostic for this species, but the available specimens from the southern part of the range, in New South Wales, have slightly darker antennal segments and the fore wings slightly shaded medially, in contrast to the specimens from tropical Australia. Moreover, the additional setae on the anterior half of the metanotum have not been seen on any of the paratypes from Kununurra in Western Australia. It is possible that this is the species that was referred to as predatory on Thrips palmi on eggplant at Darwin by Young and Zhang (2001), as the only remaining voucher specimen from those observations is listed above as a male paratype. The name of this species recognises the work of Haidee Brown in identifying so many thrips species in the Darwin area.
Female aptera. Body colour brown with extensive red pigment internally, fore tarsi and antennal segment III yellowish; major setae brown. Head longer than wide (Figure 48), ocelli absent, maxillary stylets more than one-third of head width apart, maxillary bridge complete; postocular setae finely acute. Antennal segment III with two sensoria, IV with four sensoria; segment VII strongly pedicillate, VIII broad at base. Pronotum with setae aa, am, and ml scarcely longer than discal setae, epim and pa setae bluntly pointed. Mesonotal lateral setae minute. Metanotum transverse with no sculpture medially. Mesopresternum slender medially or eroded to small paired lateral triangles. Fore tarsal tooth long and acute. Fore wing lobe very small. Pelta D-shaped ( Figure 50); tergal wing-retaining setae small and straight; tergites II-IX with posteromarginal setae S1 long and finely acute.

Comments
The biology of this species and its bright red larvae is not known, but all of the specimens were collected in association with dead twigs and branches, and they were probably predatory on mites rather than phytophagous. This is the first fully apterous species to be placed in the genus Haplothrips. Despite the absence of wings, it is not considered to be related to any Apterygothrips species, particularly because of the body size polymorphism in males.

Much of the European literature refers to this species under two separate names:
Haplothrips niger from red clover flowers (Trifolium), and H. leucanthemi from ox-eye daisy flowers (Chrysanthemum leucanthemum). However, the measurements given as discriminants by Priesner (1964), and also by Schliephake and Klimt (1979, p 254), overlap considerably without distinguishing two entities. All of the specimens identified as H. niger in Britain (Morison 1949) were subsequently re-identified as H. leucanthemi (Mound et al. 1976) because of the great variation in body size and setal lengths amongst specimens taken together in England in the same flowers. Mound and Walker (1986) pointed out that although H. niger is abundant in New Zealand in red clover flowers no males had been found. These authors therefore suggested that H. niger is a parthenogenetic form of H. leucanthemi that is specific to clover flowers. In Australia, only a few specimens of this species, all females, have been studied, collected near Sydney, Wagga, Melbourne, and Adelaide. The pseudovirga of males from C. leucanthemum flowers in England have the apex weakly swollen (Figure 31).

Recognition
Body colour brown to dark brown, fore tarsi and base of antennal segment III yellow; fore wing with base extensively shaded. Head slightly longer than wide, maxillary stylets onethird of head width apart, retracted to postocular setae, maxillary bridge complete; postocular setae acute, usually not reaching posterior margin of compound eyes (Figure 51). Antennal segment III with two sensoria, IV with four sensoria. Pronotal setae small and acute, am and ml setae not longer than discals. Metanotum weakly reticulate medially, median setae arise on anterior half of sclerite. Mesopresternum eroded medially into two triangular sclerites. Fore tarsal tooth minute in female (large in large male). Fore wing with 7-12 duplicated cilia, sub-basal setae acute or blunt. Tergite IX setae S1 bluntly pointed, much shorter than tube, S2 acute.
Female macroptera. Body colour light brown to brown, tarsi yellow, tibiae yellow washed with brown on external margins, fore femora yellow distally, antennal segment III and base of IV yellow, remaining segments brown; major setae and fore wing pale. Head longer than wide (Figure 52), maxillary stylets about one-third of head width apart, retracted to postocular setae, maxillary bridge complete; postocular setae capitate, extending beyond posterior margin of compound eyes. Antennal segment III with two sensoria, IV with three sensoria; segment VIII broad at base. Pronotum with five pairs of long capitate setae. Mesonotal lateral setae capitate. Metanotum weakly reticulate medially, median setae arise on anterior half of sclerite. Mesopresternum transverse. Fore tarsal tooth small. Fore wing with one or no duplicated cilia, sub-basal setae capitate. Tergite II with no discal setae lateral to wing-retaining setae; tergites II-VII with posteromarginal setae S1 long and capitate; tergite IX setae S1 long and slender but weakly capitate, S2 blunt.

Comments
The shape of the head and antennal segment III (Figure 52), also the long capitate pronotal setae, resemble ''Bagnalliella nigricoxa Girault'' that was referred to the genus Haplothrips by Pitkin (1973). However, the single specimen on which that species is based has only one sensorium on the third antennal segment, and a long basal stem to the fourth segment ( Figure 1). In contrast, H. lyndi is probably related to the widespread Haplothrips bituberculatus.
Female macroptera. Body and legs brown, fore tarsi and apices of fore tibiae yellow; antennal segment III largely yellow, IV pale brownish yellow, V-VIII brown; pronotal major setae and fore wing pale. Head longer than wide, maxillary stylets retracted to compound eyes, one-quarter of head width apart, maxillary bridge complete ( Figure 53); postocular setae capitate, extending beyond posterior margin of compound eyes. Antennal segment III with two sensoria, IV with four sensoria; segment VIII broad at base. Pronotum with four pairs of capitate setae, ml setae no longer than discals. Mesonotal lateral setae capitate. Metanotum reticulate medially, median setae arise on anterior half of sclerite. Mesopresternum eroded to two small lateral triangles. Fore tarsal tooth very small, at inner apex of tarsus. Fore wing with no duplicated cilia, sub-basal setae capitate, their bases arranged in a triangle. Pelta triangular ( Figure 54); tergites II-IV lateral to wing-retaining setae with sculpture and one or more discal setae; tergites II-VIII with posteromarginal setae S1 long and capitate; tergite IX setae S1 and S2 long and capitate.

Comments
This is a typical member of the Haplothrips (Trybomiella) group, with short stout antennal segments. It is smaller than the two common species, H. robustus and H. varius, with distinctively paler antennae, and the pseudovirga of the male has an unusual, small, terminal lobe.

Haplothrips robustus Bagnall
Haplothrips robustus Bagnall 1918, p 209. Haplothrips (Trybomiella) robustus Bagnall;Pitkin 1973, p 333. Recorded only from Australia, and taken widely between Tasmania and Torres Strait Islands, this species is known only from females. Moreover, it has been collected from a wide range of flowers, including many that are not native to this continent. These observations suggest that H. robustus may have been introduced to Australia from some other part of the world, such as eastern Africa. The females are closely similar to the females of H. clarisetis Priesner, an Afrotropical species that is known between South Africa, Egypt, and Iran. In that species, the fore wing sub-basal setae S3 are at least 1.5 times as long as the other sub-basal setae, and setae S1 on the ninth abdominal tergite are acute. In contrast, in H. robustus, the fore wing sub-basal setae S3 are much shorter, scarcely 1.3 times as long as the nearest seta, and setae S1 on the ninth tergite have blunt apices.

Recognition
Macropterous female, body and legs dark brown, fore tarsi, apices of fore tibiae, and sometimes antennal segment III lighter; pronotal major setae and fore wings pale except at base. Maxillary stylets about one-quarter of head width apart; antennal segment III with two sensoria, IV with four sensoria. Pronotum with four pairs of capitate setae, ml setae no larger than discals; mesopresternum eroded medially into two lateral triangles. Fore tarsal tooth very small. Mesonotal lateral setae capitate. Metanotum reticulate medially, median setal pair arising on anterior half. Fore wing exceptionally broad distally, with no duplicated cilia; sub-basal setae capitate with bases forming a triangle. Pelta broadly triangular; tergites II-IV lateral to wing-retaining setae with lines of sculpture and several small setae; tergites II-VIII with setae S1 capitate; tergite IX setae S1 and S2 bluntly rounded.

Haplothrips salicorniae Mound and Walker
Haplothrips (Trybomiella) salicorniae Mound and Walker 1986, p 54. Described from New Zealand on the saltmarsh plants known as marsh samphire or glasswort, identified at that time as a species of Salicornia, this thrips has been found widely across Australia. It is associated with the flowers of the native plants in the genera Halosarcia or Sarcocornia (Chenopodiaceae) that grow both near the coast and on inland salt flats. Males have not been seen.

Recognition
Macropterous. Body and legs brown, fore tarsi and base of antennal segment III paler; major setae and fore wings pale. Head longer than wide, maxillary stylets close together medially (Figure 55), maxillary bridge complete; postocular setae variable, usually and capitate and extending beyond posterior margin of compound eyes but sometimes short and pointed. Antennal segment III with two sensoria, IV with four sensoria. Pronotal am, aa, and ml setae variable, pa and epim setae capitate. Mesonotal lateral setae capitate. Metanotum weakly reticulate medially, median setae arise on anterior half of sclerite. Mesopresternum eroded to two lateral triangles. Fore tarsal tooth very small at inner apex of tarsus. Fore wing with no duplicated cilia, sub-basal setae capitate with bases in a triangle. Tergites II-IV with sculpture and discal setae lateral to wing-retaining setae; tergites II-VIII with posteromarginal setae S1 weakly capitate to blunt; tergite IX setae S1 and S2 blunt.
Male macroptera. Body and legs brown, fore tarsi and apices of fore tibiae yellow, antennal segment III yellowish brown on external margin at base ( Figure 26); major setae and fore wing pale. Head longer than wide, maxillary stylets about one-third of head width apart, maxillary bridge complete ( Figure 56); postocular setae no larger than minor setae on vertex. Antennal segment III with two sensoria, IV with four sensoria; segment VIII broad at base. Pronotal major setae no larger than discal setae, except capitate epimeral setae. Mesonotal lateral setae small. Metanotum reticulate medially, median setae arise on anterior half of sclerite. Mesopresternum eroded to two lateral triangles. Fore tarsal tooth small on inner margin of tarsus. Fore wing with no duplicated cilia; sub-basal setae capitate, arising in a straight line. Tergite II with dentate sculpture but no discal setae lateral to wing-retaining setae ( Figure 58); tergites II-VII with posteromarginal setae S1 strongly capitate; tergite VIII S1 short and pointed; tergite IX setae S1 and S3 capitate, S2 short and stout with apex blunt.

Comments
The single male specimen on which this species is described is unique amongst the described Haplothrips (Trybomiella) species in having only the epimeral setae on the pronotum well developed. Moreover, the presence of two pairs of capitate setae on the ninth abdominal tergite has not previously been reported for any male in this group. This is probably an Indonesian species, and a single female specimen collected at Dili, Timor Leste, appears to be conspecific, although the maxillary stylets are only 15% of the head width apart.

Haplothrips varius Hood
Haplothrips varius Hood 1918, p 128. Haplothrips (Trybomiella) varius Hood;Pitkin 1973, p 333. With its dark antennae and broad fore wings lacking duplicated cilia, this is a typical member of the subgenus Trybomiella. It breeds in the flowers of Ptilotus species, an Australian genus of Amaranthaceae, and has been collected widely across the arid zone, from Canberra in the south-east to Karratha and the Fitzroy River in the north of Western Australia.

Recognition
Macropterous, body and legs dark brown, fore tarsi and antennal segment III lighter; pronotal major setae and fore wings pale. Maxillary stylets about one-quarter of head width apart; antennal segment III with two sensoria, IV with four sensoria. Pronotum with five pairs of capitate setae; mesopresternum eroded medially into two lateral triangles. Fore tarsal tooth in female very small, in male well developed. Mesonotal lateral setae capitate. Metanotum weakly sculptured medially, median setal pair arising on anterior half. Fore wing exceptionally broad distally, with no duplicated cilia; sub-basal setae capitate with bases forming a triangle. Pelta broadly triangular; tergites II-IV lateral to wing-retaining setae with lines of sculpture and several small setae; tergites II-VII with setae S1 weakly capitate to blunt; tergite IX setae S1 blunt, S2 pointed. Male pseudovirga with apex transverse.

Haplothrips victoriensis Bagnall
Haplothrips victoriensis Bagnall 1918, p 208. Haplothrips jarvisi Kelly in Kelly andMayne 1934, p 46. syn. n. This is the most commonly encountered member of the genus in Australia, from south of the Tropic of Capricorn to Tasmania. No valid records from Northern Territory have been found, and in Queensland the most northerly valid record is from Nambour. In contrast, one female has been studied from Kununurra in northern Western Australia. This thrips lives in the flowers of many plants, but despite its abundance it seems to be of limited economic significance. In Tasmania it has caused contamination problems by entering the fruits of raspberries (pers. comm. Margaret Williams), but there is evidence that it can be a useful predator of mites (Bailey and Caon 1986). Both in general body structure and in the structure of the male genitalia this species is very similar to H. anceps, the colour character states given in the key above being the only recognised differences in adults, although the colour of the larvae is strikingly different. However, H. victoriensis is associated with flowers, whereas in southern Australia H. anceps is associated with grasses and sedges. The published description of H. jarvisi is totally inadequate, and in the absence of any specimens this name is placed as a synonym of H. victoriensis that is abundant at the type locality.

Recognition
Macropterous, body and legs dark brown, fore tarsi and base of antennal segments III and IV lighter ( Figure 27); pronotal major setae dark brown; fore wing with margins distinctly shaded. Maxillary stylets about one-third of head width apart ( Figure 59); antennal segment III about 1.8 times as long as wide, with one sensorium, IV with four. Pronotum with five pairs of capitate setae; mesopresternum transverse and boat-shaped. Fore tarsal tooth in female minute or absent, in male usually small and arising in basal half of tarsus. Metanotum weakly sculptured, median setal pair arising on anterior half; fore wing subbasal setae S1 and S2 capitate, S3 bluntly pointed; about 12 duplicated cilia present. Tergite IX setae long and acute. Male pseudovirga with apex spoon-shaped ( Figure 32).
Measurements of one female (in mm) (Tasmania, November 1999 This poorly defined genus currently includes 45 species, of which 30 are from North, Central or South America (Mound and Marullo 1996), with one from Ghana, three from India, two from the Philippines, and nine from Japan (Okajima 2006). Two species are widespread around the world, including Australia, and both of them have many synonyms. These two species are known to be predatory on scale insects (Palmer and Mound 1990), and other members of the genus possibly have a similar biology. Karnyothrips is currently considered polyphyletic, and there is no satisfactory definition of the genus. The included species are very similar to Haplothrips species, but usually have the fore tarsal tooth projecting forwards from the anterior margin of the tarsus (Figure 6), the prosternal basantra usually slightly longer than wide, the pronotal anteromarginal setae no longer than the discal setae, and the anal setae twice as long as the tube. In both species considered here, antennal segment III bears two small sensoria and segment IV bears three or four. This is the type species of the genus, because K. weigeli Watson is one of 11 junior synonyms. The maxillary stylets are deeply retracted into the head but are widely separated. The antennae are brown, except for segment III that is yellowish, and the mid and hind tibiae are often yellow. In Australia, specimens have been studied from northern Queensland and Lord Howe Island, in association with grass tussocks.
Although described from a greenhouse in Denmark, this species has been seen from many tropical localities around the world (Mound and Marullo 1996). In Australia it has been studied from Lord Howe Island, northern Queensland, and Darwin in association with grasses. It is a striking species, with the head and thorax brown, also abdominal segments IX and X, but with the rest of the abdomen and the mid and hind legs yellow. The maxillary stylets are closer together than in any other described species in the genus.
A total of 44 species is described in this genus, all from South-East Asia, particularly Indonesia, India, and Vietnam, but with two from Australia and one from Guam. In the absence of revisionary studies the significance of many of these species must remain in doubt, because one species discussed below is now recognised as highly variable. Moreover, although they are all ''gall thrips'' there appears to be little evidence that any of them induce galls. These thrips are commonly found in association with other thrips species, and they are probably phytophagous kleptoparasites rather than gall inducers. This suggestion is supported by the fact that both sexes vary greatly in body size, as in the related genus Euoplothrips. Although apparently differing in biology, Mesothrips species are closely related to those of Dolichothrips, as discussed above. Two species have been described from Australia, but these are here considered synonymous with the type species of the genus.

Recognition
Dark brown species exhibiting size-correlated polymorphism in fore tarsal tooth development; antennae eight-segmented, III with three sensoria, IV with four sensoria; head sharply constricted to basal neck ( Figure 8) with a few small stout setae on cheeks, maxillary stylets retracted about half-way to postocular setae; prosternal basantra small, mesopresternum divided; fore tarsal tooth large in large individuals of both sexes; fore wing constricted medially; tergites II-V with additional sigmoid setae lateral to the major wingretaining setae.
The character states given for M. bianchii by Ananthakrishnan do not distinguish this from the species described by Hood, and neither of these can be distinguished satisfactorily from the widespread Oriental species M. jordani Zimmermann. In describing the species, Hood (1918) stated that M. australiae was perhaps merely a race of M. jordani, and in view of the wide range of variation within samples from any single locality there is no good evidence that more than one species is represented by these three names. Moreover, considering the variation that has been observed in samples both from Australia and from Kuala Lumpur, some of the other nominal species recognised in this genus by Ananthakrishnan (1976) are likely to be synonymous.
Prof. Shuji Okajima of Tokyo kindly provided the following information concerning variation in length (in mm) of the pronotal posteroangular setae amongst samples of M. jordani that he has collected in South-East Asia: Bali (17 females) 40-85; Singapore (two females) 65-80; Taiwan (about 50 females) 80-130; India (one female) 60; Japan Ryukyus (one large female) 130-140. The species is known from India, West Malaysia, Indonesia, Taiwan, China, and southern Japan (Ryukyu Islands) (Okajima 2006). In Australia, it has been taken widely in the coastal rainforests of Queensland, including the Botanic Gardens at Mt Cootha, Brisbane, at Noosa Head, Gordonvale, and Atherton. Although living in leaf-roll galls of Ficus species, the wide range of structural variation in both sexes suggests that this thrips is a phytophagous kleptoparasite in the galls of Gynaikothrips species, rather than a gall-inducer.
The members of this worldwide genus all appear to be coccid predators living on Poaceae. An account of 18 species in this genus was given by Ritchie (1974), including a list of four generic synonyms, although subsequently four further species have been described from South-East Asia and two from New Zealand. The Australian fauna of this genus now comprises 10 species plus the Asian species, P. lucasseni. Nine of the 10 species constitute a single lineage, the P. australis group, in that they share the following character states: fore tibial inner margin with sub-apical setal-bearing tubercle but no apical tooth; antennal segments III and IV each with two sensoria; fore wing with no duplicated cilia; tergite VII with wing-retaining setae small and straight. The 10th species is divergent, in that it is apterous, and antennal segment III lacks sensoria and has a pronounced sub-basal ring, but the fore tibial sub-apical tubercle is similar to the other members of the P. australis group. As is typical of Podothrips, in all of the Australian species the prosternal basantra are longer than wide, and the pronotal anteromarginal setae are minute. The range of body colour within the genus is remarkable, from largely clear yellow to uniformly brown, but with most Australian species strikingly bicoloured. As in Karnyothrips melaleucus, the anal setae of most species are particularly long.
Key to Podothrips species from Australia 1. Abdominal tergite I comprising median pelta and two lateral sclerites; antennal segment III with one sensorium; fore tibia inner margin with apical tooth, basal to which is a prominent setal-bearing tubercle ( Figure 64); fore femur inner margin with pronounced hump or at least rugose with several minute tubercles . . . . . . . lucasseni -Abdominal tergite I comprising median pelta only; antennal segment III with two sensoria; fore tibia inner margin without an apical tooth, sub-apical setal-bearing tubercle present but sometimes small; fore femur inner margin smooth . . . 2 2. Tergite IX setae S1 (both sexes) and S2 (females)   Female macroptera. Bicoloured, head, thorax, and abdominal segments IV-X brown, I light brown, II yellow, III yellow with brown median area; antennal segments I and most of II brown, remaining segments yellow but VIII a little darker; tibiae and tarsi yellow, fore femora yellowish brown, mid and hind femora brown with base sometimes paler. Structure typical of group; major setae all capitate except S2 on abdominal segment VII, also S1 and S2 on IX; antennal segment III relatively short with sensoria small; fore wing sub-basal seta S3 weakly capitate. The brown body and yellow legs of this species (Figure 66) constitute a uniform pattern that has been found in samples taken at the base of grasses at many sites across the dry inland areas of Australia, between the Spencer Gulf in South Australia, Rockhampton in Queensland, and the Hamersley Range of Western Australia. It is one of only three species in which the setae on tergite IX are not acute, and is also one of only three species treated here that have the metathoracic sternopleural sutures developed.
Female aptera. Bicoloured, body and legs yellow except abdominal segments VIII-X and antennal segments V-VIII ( Figure 72). Head with no ocelli, compound eyes reduced to less than 20 facets; postocular setae long and capitate; vertex without sculpture; antennal segment III short with pronounced sub-basal ring but no sensoria (Figure 28), IV with two sensoria. Fore tibial inner margin with sub-apical papilla small and bearing a seta; fore tarsal tooth not large. Meso and metanota without sculpture, wing axillary sclerites not developed; mesopresternum eroded into several pieces; metathoracic sternopleural sutures well developed. Pelta broad, without sculpture; tergites with wing-retaining setae no larger than median discal setae; tergite VIII with S1 and S2 capitate, IX with S1 and S2 acute.

Comments
Ananthakrishnan (1967) proposed a sub-genus Saucrothrips for a single species from India with antennal segment III very similar to barrowi. However, these two are not likely to be closely related, because the Indian species, S. scitulus, has the typical tooth-like tubercle at the inner apex of the fore tibiae that is found in many Podothrips species from around the world. In contrast, all members of the P. australis group discussed here lack this tubercle, and P. barrowi is probably an Australian endemic derived from within this group. It is the only species discussed here that has the metathoracic sternopleural sutures developed, apart from P. australis and P. lucasseni. Unlike so many of the species in this genus, the apical antennal segments are brown not yellow.

Podothrips lucasseni (Krü ger)
Phloeothrips lucasseni Krü ger 1890, p 105. This is a widespread Asian species that is sometimes associated with sugar cane or even rice. It is recorded by Ritchie (1974) from Hawaii to Bangladesh, with both hawaiiensis Moulton and oryzae Priesner placed as synonyms. It differs from all of the Australian species in this genus in the character states indicated in the key (Figure 64), and also in the uniform brown body and femora contrasting with the yellow tibiae. Two small females have been studied from Darwin.

Podothrips moundi Ritchie
Podothrips moundi Ritchie 1974, p 274. Described from Port Moresby, New Guinea, this species is here recorded from Australia based on specimens taken at several sites near Darwin and Kakadu in Northern Territory, and also from 15 km east of Broome in Western Australia. The postocular and pronotal posteroangular setae are capitate in the type specimens, but vary in the Australian specimens, both pairs of setae sometimes being acute. The legs are generally yellow (Figure 69), but the hind femora are sometimes shaded brown.

Podothrips orion Ritchie
Podothrips orion Ritchie 1974, p 277. This is the only member of the Australian lineage within Podothrips that is uniformly brown with brown legs. Despite the colour, it is similar in structure to the other species, with the major setae all capitate except for those on tergite IX, and the fore tibia with a sub-apical setalbearing tubercle. It is known only from the dry areas of South Australia near Port Augusta.

Podothrips regina sp. nov.
Female macroptera. Bicoloured, head, antennal segment I, and thorax dark brown, also all femora and abdominal segments VI-X; abdominal segments I-IV yellow, V yellow shaded with brown; tibiae, tarsi, and antennal segments II-VIII yellow. Structure typical of group, all major setae capitate except S3 on tergites VII and IX; three capitate sub-basal wing setae; tergite VII with anterior pair of wing-retaining setae small and sigmoid but posterior pair small and straight.

Comments
Very similar to P. australis in structure, apart from the fore wing sub-basal setae and lack of metathoracic sternopleural sutures, this species is readily distinguished by the colour of the body and legs. The two males excluded as paratypes have the anterior abdominal segments more extensively shaded brown.
Female macroptera. Bicoloured, head and abdominal segments VI-X brown, pronotum yellow with anterior margin lightly shaded, pterothorax and abdominal segments I-V yellow ( Figure 70); antennal segment I brown, rest yellow. Structure typical of group, except postocular, pronotal posteroangular, and tergite VIII S1 setae all acute, remaining major setae capitate; tergite IX major setae long and acute, tergite VII with S3 setae long and acute; tergites III-V each with two pairs of sigmoid wing-retaining setae of which each anterior pair is small; tergites II, VI, and VII with these setae short and straight. Fore wing with one short and one minute sub-basal setae.
Measurements of holotype female (in mm

Comments
Similar in colour pattern to P. websteri, but with the abdomen slightly less extensively yellow, this species also differs in having several pairs of major setae acute not capitate.
Female microptera. Bicoloured, head, antennal segment I, and base of II brown, also abdominal segments VIII-X, rest of body and legs yellow ( Figure 71). Structure typical of group, except antennal segment III conical without lateral swelling; fore tarsal tooth not large; fore tibia with sub-apical setal-bearing tubercle very small; fore wing shorter than width of thorax; mesopresternum fully developed; tergites II-VII with anterior pair of wingretaining setae very small and straight, also posterior pair on II and VII; major setae all capitate except S3 on tergite IX. Male microptera. Similar to female but smaller. Tergite IX setae S2 stout but 0.8 as long as setae S1.

Measurements
Female macroptera. Similar to micropterae. Fore wing pale without duplicated cilia; subbasal setae scarcely 15 mm; anterior pair of tergal setae small and straight.

Material examined
Holotype

Comments
One female paratype collected with the holotype has the tube unusually short and presumably aberrant. Although some of the character states in the description are associated with micropterae, the body colour of this species is unique.
Podothrips xanthopus Hood Podothrips xanthopus Hood 1919, p 82. This is a widespread species in the coastal regions of Queensland and New South Wales north of Sydney, but has also been taken in the central parts of Northern Territory near Tennant Creek. Throughout this range, P. xanthopus is uniform in structure and colour pattern ( Figure 68). In contrast, the new species P. anomalus varies in colour pattern, judging from specimens taken widely across the tropical parts of northern Australia. Although treated by Pitkin (1973) as a genus with two species from Australia and one each from India and New Caledonia, the phylogenetic position of these taxa is far from clear. They all share a curious ring-like swelling near the base of the third antennal segment ( Figure 5), but a similar structure occurs in various species within unrelated genera, as discussed above under character states of the Haplothripini. The type species of the genus has the two terminal antennal segments broadly joined, forming a single outline, and there is only one sensorium on the third segment and two on the fourth; these character states also occur in the closely related new species described below. The pseudovirga at the apex of the aedeagus in males of these two species is slender but apparently rather shorter than is described for any Haplothrips species. In contrast, the Australian species, P. longistylosa, has the eighth antennal segment long and slender, unlike that of any other member of the Haplothripini, and has two sensoria on the third segment and four on the fourth. This species seems unlikely to be closely related to the first two species. The two non-Australian species have two sensoria on the third and fourth antennal segments, but have not been re-examined for this study. A macropterous female and an apterous male from Mt Tamborine, Queensland, considered by Pitkin (1973) to represent this genus, possibly belong to the South African genus Talitha Faure, otherwise unknown in Australia, because the male has a glandular area on sternite VIII and the setae on the ninth tergite are longer than the tube. Described from a single apterous female taken at Melbourne, a second female has been studied that was collected from dead twigs at Canberra in 1994. Moreover, two males that apparently represent this species were collected on dead twigs of Acacia 140 km south of Broome in the north of Western Australia. These males have all of the major dorsal and lateral setae capitate, except that setae S2 on tergite IX are short and pointed, and S3 finely acute.

Priesneria longistylosa Pitkin
Priesneria longistylosa Pitkin 1973, p 327. Known only from three macropterous females taken from leaf litter at Trangie, New South Wales, the antennae of this species are unique in the Phlaeothripidae, with the eighth segment long and slender but the third segment with a sub-basal ring-like swelling.
Female aptera. Body, antennae, and legs light brown, mid and hind tarsi yellow with apices of these tibiae variably yellow. Head longer than wide, without ocelli; maxillary stylets retracted to postocular setae and about one-third of head width apart, maxillary bridge complete. Antennae with segment VIII closely joined to VII; segment III with one sensorium, IV with two sensoria; segment III with distinctive sub-basal swelling ( Figure 5). Pronotum with four pairs of major, acute, setae, am setae not much larger than discal setae. Fore tarsus with very small tooth at inner apex. Metanotum with no sculpture medially. Mesopresternum fully eroded medially; metathoracic sternopleural sutures long. Pelta with rounded margins; wing-retaining setae short and straight, major setae all acute except S2 on tergite VIII.