Thysanoptera in Australia

Thysanoptera diversity

In 2012, the Australian Biological Resources Study web site listed from Australia a total of 826 species of Thysanoptera (ABRS 2012). About 500 of these are members of the sub-order Tubulifera, but the present work is targeted at the other sub-order, the Terebrantia, that includes almost all the thrips species that are known as crop pests.

Thrips are commonly thought of as pestiferous and essentially flower-living. However, this is a very incomplete view of the biological diversity within this Order. Almost half of the known species of thrips feed only on fungi, either in leaf litter or on dead branches (Morse & Hoddle 2006; Tree & Walter 2012), and a considerable number feed only on green leaf tissues on which some species induce galls (Mound 2008; Tree & Mound 2009). A few thrips are predators of other arthropods (Mound & Teulon 1995; Mound, 2004a, 2011a; Mound & Reynaud 2005), and in South America two species are ectoparasitic on various Homoptera (Cavalleri, 2010). Worldwide, 6100 species are recognised, with another 1600 species-group names in synonymy (Gaston & Mound 1993; Mound 2012a), but the thrips fauna of tropical areas remains poorly known. The total of 826 species recorded from Australia probably represents little more than 60% of the total thrips fauna of this continent. The fauna of Tasmania and also Western Australia remains largely unexplored, and the northern tropical fauna and its relationships to the fauna of Indonesia and the Oriental region is particularly poorly studied (Mound & Tree 2011).

Taxonomic studies on Australian thrips

Between the years 1750 and 1995, the number of thrips species that were described during each 50 year period, and that were recorded, at least subsequently, from Australia, was 2, 5, 14, 380 and 172. Most of the earliest names are of European insects subsequently introduced to Australia. Only one native species was named before 1900, the Giant Thrips, Idolothrips spectrum, apparently from specimens collected by Charles Darwin. By 1915 less than 20 species were known from Australia, but in the following ten years more than 100 species were named. In the four years 1926–1929 over 160 additional species were described, although many of these are now placed in synonymy. The final wave of descriptive activity from the first half of the century was published posthumously in 1968 from Dudley Moulton of California (1878–1951); of the 20 new species included, eight are now placed in synonymy, and a further eight allocated to different genera.

Up until the 1960s, studies on thrips in Australia were focused sharply onto one or other of two separate disciplines, with little interdisciplinary communication. One focus produced a series of outstanding ecological studies on the Plague Thrips, Thrips imaginis (see Andrewartha 1934; Evans 1935). The other produced descriptive taxonomic papers that were largely devoid of biological observations, and were often superficial even by the descriptive standards of the period. A.A. Girault was particularly notorious amongst the descriptive taxonomists. This author made available 139 species-group names for Australian thrips in a series of privately published, un-illustrated, notes that were as remarkable for their outrageously expressed opinions on a wide range of topics as they were for their lack of descriptive detail (Gordh et al. 1979). His slide preparations are exceptionally poor, often with specimens of different species under one cover-glass, and with a surprising number of these cover-glasses broken; also the labels are poorly written and sometimes misleading. He appears to have recognised species largely from colour, rather than from structural details; for example, the specimens he labelled as Thrips lacteicorpus involve pale individuals of three species, Thrips australis, T. imaginis and T. tabaci. He was employed at one time to study the biology and control of banana thrips, but his descriptive taxonomy was based largely on single specimens with no biological information. Consequently, 77 of the 139 species he described are now placed in synonymy. Thrips workers in the Northern Hemisphere similarly produced a high rate of synonymy when describing Australian species. Intra-specific variation in colour and structure was not recognised. The number of Australian species-group names erected by each of the major authors, together with the percentage of those names currently placed as synonyms, is: Girault 139 (55%); Bagnall 63 (33%); Karny 40 (42%); Hood 47 (17%); and Moulton 39 (33%).

Faunistic studies on thrips within Australia

Within Australia, taxonomic studies on thrips failed to develop, partly due to the lack of adequate reference collections, but also due to the absence of any useful introduction to the subject. Kelly accumulated many notes about thrips, and after his death these were compiled into a book (Kelly & Mayne 1934). Unfortunately, this included many errors, as did a list of thrips associated with fruit trees in the region of Perth (Newman 1935). In contrast, the keys to common Australian thrips by Hattie Vevers Steele (1935) (Mrs H.G. Andrewartha) proved particularly useful. Her work recognised that certain biological characteristics of thrips (e.g. vagility, variable host-range, intra-specific variation) create problems for species identification. Unfortunately, biological insights from the ecological studies by her husband did not penetrate further into thrips taxonomy for many years.

It was not until the mid 1960s that the concept of inter- and intra-population variation, together with an appreciation that ecology and taxonomy are inter-related aspects of an organisms’ evolutionary biology, was applied to Australian thrips. Taxonomic studies on the species associated with galls on Acacia trees (Mound 1971a) led to a particularly intensive series of biological studies on this ecological system. These culminated in a major volume that used Acacia thrips as a ‘model clade’ to examine the evolution of ecological and behavioural diversity (Crespi et al. 2004), and studies on this clade have continued (Bono & Crespi 2006, 2008; McLeish et al. 2006). Taxonomic studies have focused either on particular ecosystems, such as leaf litter and dead wood (Mound 1974, 2007; Mound & Minaei 2006), or the diversity on particular plant groups, such as the genus Geijera (Mound 1971b), and the genus Casuarina (Mound et al., 1998), or the Poaceae (Mound 2011). Other studies have focused on particular genera (Odontothripiella, Pitkin 1972; Anaphothrips, Mound & Masumoto 2009; Scirtothrips, Hoddle & Mound 2003; Thrips, Mound & Masumoto 2005), or higher taxa (Haplothripini, Mound & Minaei 2007; Urothripini, Mound 1972; Dendrothripinae, Mound 1999; Sericothripinae, Mound & Tree 2009).

Resource utilisation by thrips

As a result of extensive field work over the past 50 years, various patterns of resource utilisation by thrips have become clear, including flowers, leaves, grasses, fungi and predation on other arthropods.

Thrips and flowers

In much of the southern half of Australia the flower-habitat is dominated by the plague thrips, Thrips imaginis, although north of Brisbane this species is replaced by the closely related Thrips safrus. In the eastern rainforests the most common flower thrips is Thrips setipennis, but in the flowers of Eucalyptus across much of the continent Thrips australis predominates. Cranothrips, Odontothripiella and Thrips have each developed a range of host-specific species, although some species in each of these genera are oligophagous. Similarly, a few of the species of Pseudanaphothrips seem to be host-specific, but the biological significance of some of the named species in this genus remains in doubt due to lack of field work on host specificity and structural variation (Mound & Palmer 1981). Cycadothrips includes three described species that breed only in the male cones of Macrozamia cycads in eastern, western and central Australia, and the genus Anaphothrips and its relatives involve at least 50 species that live in flowers from a wide range of plant families (Mound & Masumoto 2009). Similarly, the flowers of native Poaceae support an interesting diversity of specialist thrips taxa (Mound 2011).

Kirk (1997) summarised the available information on pollination by thrips species, and indicated that the importance of these insects as pollinators had probably been overlooked. Since then other studies have emphasised the significance of thrips, including Williams et al. (2001) for a rainforest tree, Wilkiea huegeliana, in eastern Australia, and Mound & Terry (2001) and Terry et al. (2004) for Macrozamia cycads. The close association of particular thrips species with particular flowers, coupled with the mobility of these insects, suggests that pollination by thrips is more general than has been proven experimentally. For example, Thrips knoxi is abundant in the flowers of various Lomandra species, Thrips wellsae is abundant in the flowers of certain Epacridaceae in the south-eastern mountains (Mound & Masumoto 2005), and Oxythrips australopalmae is abundant in the flowers of two palm trees in northern Queensland, Normanbyia and Arch­ontophoenix (Mound & Tree 2011). However, precise demonstrations of pollination by Thysanoptera, such as that by Thrips antiaropsidis of Antiaropsis decipiens (Moraceae) in New Guinea, remain unusual (Zerega et al. 2004).

Thrips and leaves

In the moist tropics and eastern coastal regions of Australia, many species of Phlaeothripidae are known to feed on green leaves or apical buds (Mound 2008). Moreover, there are many such species in the dry areas, particularly on the foliage of Casuarina and Acacia species (Mound 1970; Crespi et al. 2004). Leaf-feeding Phlaeothripidae are usually host-specific, and thus members of this family are rarely considered pests. In contrast, some leaf-feeding species of Thripidae are polyphagous, with many introduced species being pests of non-native cultivated plants. Particularly noteworthy is the absence from the leaves of Eucalyptus species of any Thysanoptera, the solitary exception being a polyphagous panchaetothripine species, Australothrips bicolor. Members of the Thripidae subfamily Panchaetothripinae are all leaf-feeders, but these are primarily tropical in distribution, with few species occurring in southern Australia (Mound 2009). Members of another thripid subfamily, the Dendrothripinae, are also leaf-feeding, and there are several endemic species and genera (Mound 1999). The fourth thripid subfamily, the Sericothripinae, is also well represented in Australia, with both native and introduced species feeding primarily on leaves (Mound & Tree 2009), with one significant in the biological control of the noxious weed, Ulex europeaus (Ireson et al, 2008).

Minimising water loss is an esssential survival strategy for small insects in dry areas. Many thrips species achieve this either by inducing galls, or else by invading galls induced by other insects. On some phyllodinous Acacia trees one suite of thrips species induces pouch galls, and some of these have a caste of non-breeding wingless individuals that function as soldiers to repel intruding insects, particularly a suite of kleptoparasitic thrips (Crespi et al. 2004). Species of another suite on Acacia trees have the remarkable ability of glueing or sewing together two or more phyllodes to produce a domicile within which to breed (Bono & Crespi 2008). On several Casuarina species there are Phlaeothripidae that induce large woody galls; these are commonly invaded by one or more kleptoparasitic thrips species, and in Western Australia one of these kleptoparasitic thrips is noteworthy for dimorphism in the male genitalia (Mound et al. 1998). One native species, Cyrilthrips cecidis, is particularly interesting as it one of the few members of the Thripidae worldwide that is known to induce leaf galls (Tree & Mound 2009).

Thrips and grasses

Plants of the family Poaceae, and to a much lesser extent Cyperaceae, support a particularly wide diversity of Thysanoptera. Among the Aeolothripidae, Desmothrips reedi and Gelothrips cinctus are predators on small arthropods at the base of grasses. Among the Phlaethripidae, species of Bamboosiella and Podothrips are also considered predatory, some Haplothrips species are phytophagous and breed in the inflorescences, and several Idolothripinae breed at the base of grasses where they feed on fungal spores. But it is amongst the Thripinae that Poaceae have been particularly adopted as hosts (Mound, 2011). Most Thripinae that breed on Poaceae feed on leaves, with only a few breeding in the florets. Several species associated with pasture grasses in the South of the continent are clearly European immigrants, whereas in the North many species are shared with Southeast Asia. However, there are several endemic radiations of Thripinae on native Poaceae, and a number of endemic genera are now recognised.

Thrips and fungi

Fungus-feeding thrips in Australia are all members of the Tubulifera family Phlaeothripidae. Fungi are available to thrips on the surface of dead twigs and branches, and also in leaf-litter on the ground (Tree & Mound 2012). Both of these habitats are exploited in Australia, although the second probably includes the larger number of species. Thrips on dead bark are exposed to desiccation, and also to a wide range of predators including birds, lizards, ants and spiders. As a result, these thrips species exist as widely dispersed populations, and are rarely collected (Mound, 2008; Goldarazena, 2010). Leaf-litter in Australia, even shallow Eucalyptus litter, often contains a wide range of thrips taxa. The completely wingless species in litter sometimes exist as a mosaic of slightly different looking forms, presumably due to limited gene flow between localised demes; their taxonomy remains particularly doubtful (Mound 1972). The members of one subfamily, Idolothripinae, are presumed to feed only on fungal spores (Mound 1974, 2007; Tree et al. 2010). A recent study in eastern Australia showed that thrips abundance was higher in dry sclerophyll leaf litter than rainforest leaf litter, and that Eucalyptus leaf litter appears to support a richer thrips fauna than Acacia leaf litter (Tree & Walter 2012).

Thrips as predators

A few species of thrips, in all three major families, act as predators, and the range of prey includes thrips larvae, mites and their eggs, scale-insects and whitefly nymphs. These thrips are generally facultative predators, taking prey as well as feeding on plant tissues (Mound & Teulon 1995), as in various species of Aeolothripidae and also the pest species, Thrips tabaci and Frankliniella occidentalis. However, several groups of thrips are obligate predators, such as Scolothrips species that feed on mites (Mound et al. 2010; Mound, 2011), Aleurodothrips fasciapennis that feeds on scale insects, and some species of Karnyothrips that also attack scale insects (Palmer & Mound 1991). Haplothrips bituberculatus, a widespread and common species apparently feeding on mites on dead branches, is a native predator, as is Scolothrips ochoa that feeds on Raoiella mites on Eucalyptus leaves (Mound et al., 2010).

Adventive taxa among Australian thrips

Of the species of thrips recorded from Australia, at least 60 are introduced from other countries. These include common grass-living Thripidae of Europe, Aptinothrips, Chirothrips and Limothrips species, also leaf-feeding tropical species such as those in the genera Chaetanaphothrips, Heliothrips, Hercinothrips, Parthenothrips, Selenothrips and some Scirtothrips and Thrips. Most of these probably have been established in Australia for many years, but more recent introductions are the Western Flower Thrips, Frankliniella occidentalis, and the Lily-bulb Thrips, Liothrips vaneeckei. More difficult to evaluate are Asian species that have been recorded in recent years, such as the Melon Thrips, Thrips palmi, and the Rice Thrips, Stenchaetothrips biformis. Given the continuing lack of knowledge of the thrips fauna of northern Australia, it is not possible to know to what extent the northern coastal areas of this continent are part of the natural distribution of Asian species. Recent studies have demonstrated the presence in northern Australia of several species from across Asia (Mound & Tree, 2011), and some species are now well established, including Gynaikothrips ficorum, the leaf-rolling thrips of cultivated Ficus microcarpa trees (Mound et al. 1996; Tree & Walter 2009), the Oriental Lily-flower Thrips Taeniothrips eucharii (Mound & Tree, 2008), and several Panchaetothripinae including Helionothrips errans that has been found at Perth and at Newcastle on orchid leaves (Mound, 2009).

Endemicity in the Australian thrips fauna

The faunal relationships of Australian Thysanoptera are considered in Austin et al. (2004) and Mound (2004b), and a large number of genera and species are clearly endemic. Some distributions are particularly interesting, such as Dorythrips in the Melanthripidae that includes two species in Western Australia and one in Chile, and Cranothrips that includes one species from South Africa and a series of species across Australia (Pereyra & Mound 2009). Similarly, Jacotia in the Phlaeothripidae has one species in South Africa and several in Australia (Mound 1995). Phlaeothripidae dependent on the Australian plant genera Acacia, Casuarina and Geijera are clearly endemic (Crespi et al. 2004; Mound et al. 1998; Mound 1971b). In contrast, fungus feeding species in the genera Bactrothrips, Carientothrips, Malesiathrips, Mecynothrips, and Nesothrips show close relationships to the faunas of Southeast Asia, Papua New Guinea and the Pacific islands. Moreover, a few taxa in Southern Australia show relationships to New Zealand (Mound 2006). Of the 23 genera of Aeolothripidae worldwide, five are known only from Australia. In the Thripidae, the genera Odontothripiella and Pseudanaphothrips represent endemic Australian radiations, and this is also true of some genera related to Anaphothrips (Mound & Masumoto 2009), and the genera Aliceathrips and Masamithrips that are associated with Poaceae (Mound 2011). In the semi-arid parts of Australia, the fauna of Phlaeothripidae includes several endemic radiations, including those on Geijera, Casuarina and Acacia

Diagnosis and classification of the Thysanoptera

Thrips differ from other insects in the asymmetry of their mouth parts. The right mandible does not develop beyond the embryo, although the left mandible forms an elongate, solid stylet in larvae and adults. The maxillary stylets (laciniae) are also asymmetric. These are linked together longitudinally by a tongue and groove system to produce a feeding tube with a sub-apical aperture and one central channel that also functions to inject salivary secretions. On each leg, adult and larval thrips have an eversible, adhesive, pretarsal arolium, from which the old name Physapoda (bladder-foot) is derived. The modern ordinal name, Thysanoptera (fringe-wings), is derived from the long cilia that border the slender wings of adults, although in many species adults are wingless.

The life history of thrips is unlike that of other insects. Following two actively feeding larval stages, there are two (three among Phlaeothripidae) non-feeding pupal stages before the adult is produced. These pupal stages are found at ground level in most flower-living species, but normally occur on the food substrate in leaf- and fungus-feeding species. The Order Thysanoptera is considered part of the Hemipteroid complex, the Paraneoptera. But the life history distinguishes the group, and it is best regarded as an exopterygote that has independently adopted holometaboly.

Worldwide, nine extant families of Thysanoptera are currently recognised (also five families of fossil thrips: Mound, 2011), and these are grouped into two suborders, the Terebrantia and the Tubulifera (Mound et al. 1980; Mound & Morris 2007). An alternative phenetic classification (Bhatti 1994, 2006) recognised two Orders and many small families, but this involves problems. Recognition of Xaniothripidae for a single Australian genus Xaniothrips implies that the included species comprise a lineage that separated from all other Phlaeothripidae before the radiation of all the other phlaeothripid genera. The sub-order Tubulifera is therefore considered to include a single family, Phlaeothripidae, and this comprises two subfamilies, the Phlaeothripinae with 2820 described species in 375 genera worldwide, and Idolothripinae with 715 species in 80 genera. In Australia, there is no general introduction to the Phlaeothripidae, but revisions of many genera and more extensive groups are available, and currently about 500 species in 120 genera are listed from this continent. Publications on major segments of the Australian fauna of Phlaeothripidae include: Crespi et al. 2004; Mound & Minaei 2007; Mound 2008; Mound et al. 1998; Mound 1972). An introduction to the Idolothripinae of Australia (Mound, 1974) included 70 species in 23 genera, and a key to world genera of this subfamily is web-available (Mound 2012).

Five Terebrantia families occur in Australia, each of which occurs in other parts of the world. In contrast, each of the three families not represented in Australia has a restricted distribution. The Heterothripidae includes about 70 species in four genera, all from the New World; the Stenurothripidae (=Adiheterothripidae) includes three small genera – one of four species from date palms in western Asia and the Mediterranean, but the other two are monobasic genera from California; the Fauriellidae includes five species in four genera with a disjunct distribution between South Africa, California and southern Europe. The phylogenetic significance and relationships of these three families remains in doubt.

Oz thrips taxa