Karnyothrips Watson

Karnyothrips Watson View in CoL

Karnyothrips Watson, 1923 View in CoL , 23. Replacement name for Karynia (sic!) Watson, 1922 (error for Karnyia ). Type-species: Karynia weigeli Watson, 1922 View in CoL (= Anthothrips flavipes Jones, 1912 View in CoL ), by monotypy.

Of the nearly 40 genera belonging to the tribe Haplothripini , only about 10 genera are distributed in the New World. These include Leptothrips Hood View in CoL and Bagnalliella Karny View in CoL , which are endemic to the New World, but most others are probably of Old World origin. Even the largest genus Haplothrips View in CoL , comprising more than 250 species, includes less than 20 endemic species in the New World ( Mound & Minaei 2007), and only nine species are distributed in the Neotropics ( Cavalleri et al. 2016). In contrast, before this study, 50 species of Karnyothrips View in CoL had been known, of which 32 species are recorded in the New World. However, this includes two widespread species, K. flavipes View in CoL and K. melaleucus View in CoL , which are considered to be of Old World origin. As a result of current study, the number of Karnyothrips species is increased to 76, of which 46 are distributed in the Old World. Moreover, considering that the presumed relatives of Karnyothrips View in CoL , Androthrips View in CoL , Mesandrothrips View in CoL (in part, see below), and Xylaplothrips View in CoL (in part, see below), are all of Old World origin, so the genus Karnyothrips View in CoL itself is very likely to have originated in the Old World. The existence of more species in the Old World, as shown here, seems to support this conjecture, but the degree of substantial surveys varies greatly from region to region, and it is not possible to make a definitive decision at this time. In addition, although the habits of Xylaplothrips species are uncertain, the species in the remaining three genera, Androthrips View in CoL , Karnyothrips View in CoL and Mesandrothrips View in CoL (in part), may all be predatory, and it seems to be that they have evolved from a predaceous ancestor. In contrast, although the genus Haplothrips View in CoL is somewhat similar to these genera in appearance, the species are basically phytophagous, and its origin may be different. There are a few predatory species among this huge genus, but they may have been derived secondarily from phytophagous congeners.

The generic definition described below is based on 43 species recognized here from Asia, but exclusive of the New World species. Therefore, this definition does not apply to the whole of the New World species.

Generic definition. Body slender, wings fully developed, but often reduced, rarely absent. Head longer than wide, surface usually smooth, with no transverse fine striae; cheeks without stout setae; a pair of postocular setae situated near cheeks. Eyes and ocelli normal, but smaller in micropterae and apterae. Antennae eight-segmented; segment VIII variable in form and length, conical to weakly pedicellate, often widely and closely joined to segment VII; segment III with one (0+1), two (1+1) or three (1+2) sense cones, segment IV with four (2+2 +1) or fewer major sense cones; campaniform sensillum on segment II situated near apex, at least before middle. Mouth-cone short and rounded; maxillary stylets long, retracted far into head capsule, at least reaching postocular setae, narrower than half of head width, often close together; maxillary bridge present. Prothoracic notopleural sutures usually complete, rarely incomplete; am setae reduced or vestigial, other four pairs of major setae developed. Basantra well developed, relatively far apart from ferna; prospinasternum usually large, rarely small; mesopresternum usually eroded at anterior margin, narrowly boat-shaped or divided into two or three small plates. Fore tarsus with a distinct tooth arising at apex of inner margin, directed forwards. Metathoracic sternopleural sutures absent; both furcal arms usually fused medially at least in macropterae, with or without short spinula. Forewing, if developed, weakly constricted medially, with or without duplicated cilia. Pelta usually trapezoidal. Abdominal tergites II (or III) to VII usually each with two pairs of wing-retaining setae at least in macropterae, but rarely with only one pair. Tergite IX posteromarginal setae S1 variable in length, pointed or expanded. Tube tapering, straight-sided, 0.4–0.7 times as long as head. Terminal setae variable in length, but usually much longer than tube. Male: usually not show extreme allometric growth; pore plate on sternite VIII absent; tergite IX S2 setae short.

Sense cone formula of Karnyothrips from Asia

We agree with Mound and Minaei (2007) that the plesiomorphic condition of the sense cone formula on antennal segments III and IV is three (1+2) on III and four (2+2) on IV. This plesiomorphic condition is retained in a few members of the Haplothripini , Androthrips , Dolichothrips , Mesothrips , Neoheegeria , and Mesandrothrips ( Mound & Minaei 2007; Mound & Tree 2019). Of these, Androthrips and Mesandrothrips may be more closely related to Karnyothrips than the remaining three genera. Previously, Karnyothrips has been discriminated from these related genera together with the largest genus Haplothrips in having different sense cone formula of the antennal segments IV, or of both segments III and IV. Androthrips and Mesandrothrips have plesiomorphic three and four sense cones on segments III and IV respectively, whereas Karnyothrips has two or fewer sense cones on segment III and three or fewer sense cones on segment IV, but K. flavipes and K. merrilli are exceptional with four sense cones on IV ( Mound & Marullo 1996). Haplothrips usually has two or fewer sense cones on the segment III and four on the segment IV, but rarely fewer. However, Okajima (2006) described nine Karnyothrips species from Japan, of which three species, K. acer , K. pacificus and K. robustus , have three and four sense cones on the segments III and IV respectively. Moreover, in this study, five additional species, K. capitatus sp. n., K. inpallens sp. n., K. pitkini sp. n. K. tenax sp. n. and K. vietnamensis sp. n., are described, which also have three (1+2) and four (2+2 +1) sense cones on these antennal segments. Table 1 shows the sense cone formula of every species of Karnyothrips recognized here from Asia, and species are arranged in order of the sense cone number from fewest to most. From this table, the following four facts are possibly revealed: 1) the interspecific variation in the number of sense cones is clearly continuous and exhibiting a morpho-cline; 2) there is no significant gap in the number of sense cones; 3) there are few, but important, intraspecific variations in the number of sense cones, e.g. K. quadriconus and K. variabilis have two or three sense cones on segment III, and K. flavipes has three or four on segment IV. 4) 21 species have four sense cones on segment IV, of which 13 species have two on segment III but 10 species have three sense cones. In particular, K. vietnamensis sp. n., which is morphologically very similar and possibly closely related to K. flavipes , has three and four sense cones on the segments III and IV. Moreover, Cavalleri et al. (2016) reported intraspecific variation in the sense cone number on these antennal segments in the genus Mirothrips , known only from South America and considered closely related to Karnyothrips . These species usually have two or three sense cones on segment III and three or four on IV. In particular, M. analis has two to four sense cones on segment IV. Judging from these facts, the difference in sense cone formula seems not to satisfactorily distinguish these Karnyothrips species at genus level. However, it is still useful as a character to classify at species level in the Haplothripini if the intraspecific variation is understood.

Observation of sense cones requires great care. Sometimes we needed to use a ×100 oil-immersion objective lens. For this study, the authors examined more than 1,500 specimens, on all of which sense cones on both left and right antennal segments III and IV were observed. This was not so easy, as sometimes some sense cones were small and difficult to observe, and could not be observed well in a few dozen specimens.

Systematic position of Karnyothrip s

In conclusion, we speculate that the closest relatives of Karnyothrips are limited to a part of each of Mesandrothrips and Xylaplothrips . Some species closely related to both M. inquilinus and X. fuliginosus (these species-groups are referred to in this paper as Mesandrothrips (in part) and Xylaplothrips (in part) respectively) are very closely related to Karnyothrips , but the remaining species now classified into these two genera are not included. In addition, a monobasic genus Glenothrips Priesner , which was erected for Cryptothrips biuncinatus Karny from Java, may also be very closely related to these genera ( Priesner 1921). It can be distinguished from Mesandrothrips (in part) only by having a fore tibial inner sub-apical tooth, and these two genera may represent a single genus ( Dang et al. 2014). These thrips share some important structures in considering systematic relationships, such as the following: fore tarsus with a tooth, that arises on inner apex of tarsus and is directed forwards (this structure may be homologous with ‘inner hamus’, see Mound & Minaei 2007, p. 2929); maxillary stylets long and relatively close together, at least reaching postocular setae, separated by less than half of head width, with rather distinct maxillary bridge; prothorax with four pairs of major setae developed, anteromarginal setae reduced, short and acute; mesopresternum reduced medially, if transverse then relatively narrow at middle, not complete. Moreover, these four genera cannot be distinguished satisfactorily by any single character, and they can barely be distinguished by a combination of several character states. For example, the differences in the distance between prosternal basantra and ferna and in the length of the terminal tube setae are sometimes useful to discriminate Karnyothrips from both Mesandrothrips (in part) and Xylaplothrips (in part), but there are several species that have intermediate character states in both structures. The basantra and ferna of Karnyothrips are usually more widely separated from each other, but the differences are not clear in some Karnyothrips species, such as K. acutus , K. semiflavus and K. yoshi (cf. Figs 43 View FIGURES 41–47 , 190 View FIGURES 188–193 & 241 View FIGURES 240–246 ). Similarly, the terminal tube setae are usually considerably longer than the tube in Karnyothrips , but are almost as long as the tube in K. yoshi , or only slightly longer in K. mucidus . Besides, the genus Androthrips Karny may also be closely related to these thrips, despite the substantial difference in appearance. It has most structures, other than the armature on the fore femur and tibia, mostly shared with Karnyothrips and Mesandrothrips (in part). In particular, one Karnyothrips species, K. tenax sp. n. from Southeast Asia, has somewhat similar armature on the fore leg, that is, the femoral sub-basal indistinct inner hump (not distinct tubercle) ( Fig. 217 View FIGURES 212–220 ) and the tibial sub-apical inner tubercle or scale ( Fig. 218 View FIGURES 212–220 ). In consequence, it seems to be somewhat intermediate between these two genera. However, it has the prosternal basantra and ferna widely separated ( Fig. 215 View FIGURES 212–220 ), a condition commonly found in Karnyothrips but quite dissimilar to Androthrips , and the terminal tube setae more than twice as long as the tube. In addition, an undescribed grass-living Androthrips species from East Asia between Thailand and the Ryukyu Islands, Japan, shows considerable intraspecific variation in body size, and small individuals have the fore femur and tibia scarcely armed with indistinct tubercle or scale; they cannot be distinguished satisfactorily from Karnyothrips species. Furthermore, although Xylaplothrips (in part) species are usually found on dead branches, their feeding habits are currently unknown. Species of Androthrip s, Mesandrothrips (in part), and Glenothrips uncinatus , are undoubtedly predatory and usually found in leaf galls or on green leaves together with certain phytophagous phlaeothripine species. Similarly, most Karnyothrips species are considered to be predatory (see introduction). Therefore, there is a possibility that these five genera, Androthrips , Glenothrips , Karnyothrips , Mesandrothrips (in part) and Xylaplothrips (in part), could be treated as a single genus. Nevertheless, at present, both Mesandrothrips and Xylaplothrips are undoubtedly polyphyletic assemblages.

Mesandrothrips contains about 20 species ( Mound & Tree 2019), but the majority of these species may not be congeneric with M. inquilinus , the type-species of the genus. M. pictipes from India and M. dubius from Japan, have the maxillary stylets rather wider apart from each other, almost V-shaped. This wider maxillary bridge situated low in head and lying around the basal collar of the head (= postoccipital ridge), the well developed prothoracic anteromarginal setae, the mesopresternum not reduced medially and the fore tarsal tooth situated on inner margin, suggests these species may not be so closely related to Karnyothrips . These species could possibly be transferred from Mesandrothrips to some different genus (or genera). Under such a condition, M. inquilinus and its close relatives could not be transferred immediately to Karnyothrips , because inquilinus is the type-species of the genus Mesandrothrips .

The situation is exactly the same for Xylaplothrips that currently contains 11 species. About five of these species, including X. fuliginosus , the type-species of the genus, are closely related to Karnyothrips , but the rest of the species may not be closely related. Currently, Mesandrothrips and Xylaplothrips are distinguished by the different sense cone number on antennal segments III and IV ( Mound & Tree 2019), but some Karnyothrips species apparently bridge this morphological gap between these two genera. In any case, in order to avoid further confusion, no nomenclatural changes should be made until the generic positions have been determined of species that are not closely related to Karnyothrips , such as M. pictipes and M. dubius . Incidentally, Androthrips is the oldest generic name among these related five genera.

Cavalleri et al. (2013 & 2016) indicate that the genus Mirothrips Cavalleri et al. is closely related to Karnyothrips . Currently, it involves four predaceous species described from South America, of which M. albiter Cavalleri et al. and M. vespicola (De Santis) , prey on the eggs of social wasps.According to the original description ( Cavalleri et al. 2013), the genus may be distinguished from Karnyothrips and its relatives in having the following character states: antennal segment VIII long and slender; maxillary stylets not deeply retracted into the head and more than half of head width apart; one or two pairs of long setae anterolateral to the ferna; mesopresternum not reduced medially; metanotum with two pairs of setae; fore tarsal tooth small or absent; male abdominal sternite VIII with pore plate. The long prosternal setae in Mirothrips are especially peculiar amongst Phlaeothripidae , and a male sternal pore plate is absent in Asian Karnyothrips and its relatives. However, somewhat intermediate structures are often found, as follows: a few Asian Karnyothrips species have antennal segment VIII long and slightly constricted at base (cf. Figs 126 View FIGURES 122–127 & 219 View FIGURES 212–220 ); several Androthrips and Mesandrothrips (in part) species have the maxillary stylets relatively shorter and wider apart; some Androthrips species have the metanotum with an anterior pair (or two anterior pairs, or more) of setae in addition to the median setal pair; a few Androthrips species have the mesopresternum not distinctly reduced medially. Considering these, it can be inferred that Mirothrips is possibly a distinct genus that has a close relationship with Karnyothrips and its close relatives, but evolved independently in South America. However, some species of Karnyothrips in South America are known to have male sternal pore plate, therefore the relationship between them and Mirothrips needs to be examined.

Several Asian Karnyothrips species, such as K. antennalis Okajima , K. cyathomorphus Wang et al. , K. inflatus Okajima , K. oppositus sp. n. and K. triconus sp. n., have antennal segment III with a curious sub-basal ring-like swelling (cf. Figs 169 View FIGURES 164–169 & 226 View FIGURES 221–226 ). However, a similar structure is often found in other widely scattered genera not only in the Phlaeothripinae but also in the Idolothripinae ( Mound & Minaei 2007) . Especially noteworthy is the genus Priesneria Bagnall described from Australia ( Bagnall 1926). It is classified into the Haplothripini and currently includes six species, P. akestra Mound & Wells , P. kellyana Bagnall , P. longistylosa Pitkin and P. peronis Mound & Minaei from Australia, P. doliicornis (Bianchi) from Hawaii and P. insolitus (Ananthakrishnan) from India, but not all these species may be congeneric with the type-species, kellyana . According to the original descriptions, at least three species, doliicornis , kellyana and peronis , are very similar to Karnyothrips species in general appearance. These species are especially similar to K. cyathomorphus from China and K. inflatus from Japan in having antennal segment VII elongate and closely joined to segment VIII (cf. Fig. 168 View FIGURES 164–169 ) in addition to the sub-basal swelling of the segment III. Because of this, these species may appear to be distinguishable from Karnyothrips , but these antennal structures could well be included in the range of interspecific variation within the genus Karnyothrips as discussed below. Moreover, they cannot be distinguished satisfactorily from Karnyothrips by most characteristics other than the antennal structures. Therefore, the generic position of two Australian species, kellyana and peronis , needs to be considered in detail, especially to make a comparison with the generic definition of Karnyothrips given above. The remaining species, doliicornis , has already been re-transferred from Priesneria to Karnyothrips by Okajima (2006, p. 389–390), but this nomenclatural change has usually been overlooked. Furthermore, according to the original description of doliicornis , it is very similar to infratus, and it is even possible that they are of the same species. The genus Talitha Faure , described based on three apterous species from South Africa ( Faure 1958), also has similar antennal structure, and may also be related to Karnyothrips . However, Talitha may can be distinguished by having the campaniform sensillum on antennal segment II situated near the base or middle of the segment like that of genera classified into the Plectrothripini, metathorax with well developed sternopleural sutures, and only two pairs of prothoracic setae, posteroangulars and epimerals, are developed. Moreover, two of the three species, T. fusca (type of the genus) and T. glandifera , have a well developed pore plate on abdominal sternite VIII of males.

Apterygothrips View in CoL , with almost 40 species, is also related to Karnyothrips View in CoL , with a similar fore tarsal tooth. But this genus is another polyphyletic assemblage, with only several species from the Mediterranean and African regions that are possibly congeneric with the type-species, A. haloxyli Priesner. At View in CoL least haloxyli View in CoL differs from Karnyothrips View in CoL in having the pronotal anteromarginal setae well developed and the compound eyes coarsely facetted ( Priesner 1933; zur Strassen 1966). This species has not been examined for the present studies, but several species that appear to be closely related to haloxyli View in CoL have been examined: A. canarius (Priesner) View in CoL , A. hispanicus (Bagnall) View in CoL , A. longiceps View in CoL zur Strassen and A. priesneri View in CoL zur Strassen. These species are not always apterous, and have slight differences in the endoskeleton of the metathorax from Karnyothrips View in CoL . In Apterygothrips View in CoL , both furcal arms tend to widely separated from each other without spinula in both macropterae and apterae, while in Karnyothrips View in CoL , they are usually connected medially with short spinula at least in macropterae. Moreover, these species may be more similar to Haplothrips View in CoL rather than Karnyothrips View in CoL in general structures. In any case, the majority of the species listed under Apterygothrips View in CoL are probably not congeneric with the type-species. Many of them are probably included in this genus only by the invalid reasons that they resemble apterous Haplothrips species but have fewer sense cones on antennal segment IV and a forwardly directed fore tarsal tooth. A comprehensive re-assessment of their generic positions is required. Among them, according to the original descriptions, several species seem to be somewhat similar to Karnyothrips View in CoL . Two Apterygothrips species originally described under the genus Xylaplothrips View in CoL from India, A. jogensis View in CoL and A. rubiginosus View in CoL , are very similar to Karnyothrips View in CoL ( Ananthakrishnan & Jagadish 1967 & 1970). However, both these species have a small but important characteristic that is not found in Karnyothrips View in CoL , and re-examination of the original specimens is needed. According to Pitkin (1976), jogensis View in CoL has no sense cone on antennal segment III, a condition not mentioned in the original description, and rubiginosus View in CoL has the prothoracic midlateral setae reduced.

Podothrips Hood View in CoL and its two relatives, Okajimathrips Bhatti View in CoL and Praepodothrips Priesner & Seshadri View in CoL , are apparently somewhat similar to Karnyothrips View in CoL in having the body proportions slender and the fore tarsal tooth directed forwards like an inner hamus. Moreover, they are predators of scale insects and inhabit bamboo and/or grass, like certain Karnyothrips species mentioned above. However, these three genera may not be so closely related to Karnyothrips View in CoL , even if their origin is from the same predaceous ancestor. They usually have the sternopleural sutures well developed and both furcal arms separated from one another with no spinula in the metathorax even in macropterae, whereas Karnyothrips View in CoL have no sternopleural suture and both furcal arms are usually fused medially with a short spinula, at least in macropterae. But there is an exception to this that seven of nine species of australis - group of the genus Podothrips View in CoL from Australia have no metathoracic sternopleural sutures ( Mound & Minaei 2007) and are apparently somewhat similar to Karnyothrips species. However, their remaining structures are typical of Podothrips View in CoL . This is probably the result of independent evolution among the genus during the long isolation of this landmass, and it may not indicate relatedness of these two genera.

When we see Karnyothrips species alive in the open field, they can be almost correctly distinguished from some genera that are very similar in appearance, such as Haplothrips . This is probably due to differences in their body shape and behavior or movements. In Karnyothrips species the body shape is elongate and more slender ( Figs 11–40 View FIGURES 6–20 View FIGURES 21–30 View FIGURES 31–40 ) in comparison with Haplothrips species ( Figs 6–10 View FIGURES 6–20 ). Occasionally, Karnyothrips species may have considerably distended abdomen when alive ( Fig. 2 View FIGURES 1–5 ). Moreover, the head is relatively small and the legs tend to be shorter in Karnyothrips . The slower movements of Karnyothrips than those of Haplothrips are mainly due to their relatively shorter legs compared to their elongate bodies. However, taxonomists usually observe immovable specimens mounted on glass slides, and usually compare certain parts of structures rather than the whole body.

Structural variations among Asian Karnyothrips species

Each of the following structures can be important for the discrimination of species.

1) Coloration: About half the species from Asia are bicolored brown and yellow, and the color patterns vary between the species. The most common color pattern is found in K. brevipilosus , K. melaleucus and K. triconus , the abdomen is largely yellow, sometimes with a brown median marking on each tergite, in contrast to the head, thorax and posterior end of abdomen including tube that are brown to dark brown (cf. Figs 3 View FIGURES 1–5 , 15 View FIGURES 6–20 , 26, 27 View FIGURES 21–30 & 37 View FIGURES 31–40 ). K. flavescens , K. insignis , K. macrommatus , K. maculatus and K. submaculatus have similar color pattern, but the metathorax is largely yellowish and anterior end of abdomen shaded with brown (cf. Figs 17 View FIGURES 6–20 , 24, 25 View FIGURES 21–30 and 36 View FIGURES 31–40 ). K. inpallens also has similar color pattern, but prothorax is yellowish ( Fig. 21 View FIGURES 21–30 ). Moreover, K. semiflavus has the head and thorax largely yellow but the abdomen largely dark brown ( Fig. 33 View FIGURES 31–40 ), and K. pallucidus has the body entirely yellow (Fig. 158). The color of antennal segments is often useful for discriminating species. Two sympatric species, K. flavipes and K. flavicornis , are very similar but easily distinguished by the color of antennal segment III (see key couplet 26).

2) Antennal segments: The interspecific variation in the form of antennal segment VIII is especially important. It varies between species, but the variation shows continuous change without any gaps. In several species, e.g. K. macrommatus and K. submaculatus , segment VIII is slender and distinctly constricted at base and narrowly fused to segment VII (cf. Figs 126 View FIGURES 122–127 & 210 View FIGURES 206–211 ). In contrast, K. inflatus and K. oppositus have segment VIII conical and widely fused to VII, forming a single segment outline ( Fig. 168 View FIGURES 164–169 ). Moreover, these two species as well as K. cyathomorphus have segment VII somewhat elongate (cf. Fig. 168 View FIGURES 164–169 ), though several species do not have it elongate (cf. Figs 84 View FIGURES 80–85 & 126 View FIGURES 122–127 ). However, these are both extreme ends of the morpho-cline, and there are many intermediate conditions showing continuous changes. Several species, such as K. antennalis , K. inflatus , K. oppositus and K. triconus , have antennal segment III with a sub-basal ring-like swelling (cf. Figs 169 View FIGURES 164–169 & 226 View FIGURES 221–226 ). However, K. acutus and K. formosanus have a somewhat intermediate small sub-basal swelling ( Figs 46 View FIGURES 41–47 & 97 View FIGURES 92–97 ). Moreover, among acutus specimens those swellings show slight variation: that of the non-paratypic females from Taiwan are more distinct ( Fig. 47 View FIGURES 41–47 ) than that of the type series from Thailand ( Fig. 46 View FIGURES 41–47 ). Interestingly, a similar structure is found in certain species belonging to unrelated phlaeothripine genera. K. insignis is unusual with segment III very small, shorter than 0.7 times as long as segment IV (see Fig.133E View FIGURES 128–133 in Okajima 2006: 395).

3) Sense cone formula: This can be very important for identifying species, and is one of the very useful key characters. However, in some cases a sense cone is very small, not always completely recognizable or even invisible. Moreover, in rare cases there is variation among specimens of a species ( Table 1). For example, K. micrommatus usually has only one outer sense cone on antennal segment III, but often has an additional very small inner sense cone. Variation in sense cone formula is also found in K. flavipes on antennal segment IV, (1+2 +1) or (2+2 +1). Finally, K. quadriconus and K. variabilis usually have two sense cones on antennal segment III, but infrequently have three sense cones. As discussed above, this fact is quite important when considering the definition of the genus Karnyothrips .

4) Maxillary stylets: The type-species of the genus, K. flavipes , has the maxillary stylets rather widely separated from each other at the middle ( Figs 86 View FIGURES 86–91 ), although several species, such as K. lalae , K. melaleucus and K. submaculatus , have these stylets narrowly separated (cf. Figs 110 View FIGURES 110–115 , 134 View FIGURES 134–139 & 206 View FIGURES 206–211 ). However, there are some species with the stylet separation intermediate, and this variation shows continuous change without distinct gaps. Moreover, males tend to have stylets wider apart than females (cf. Figs 194 & 195 View FIGURES 194–199 ). Even in the type-species, flavipes , females have the stylets about one-fourth of head width apart ( Fig. 86 View FIGURES 86–91 ), whereas males have them about one-third of head width apart ( Fig. 87 View FIGURES 86–91 ). Moreover, males of K. melaleucus show intraspecific variation in the separation of the stylets ( Figs 135–137 View FIGURES 134–139 ).

5) Notopleural sutures: Although most species have the pronotal notopleural sutures complete, a few species, such as K. brevipilosus , K. melaleucus and K. oppositus , have those sutures incomplete. However, a few species are somewhat intermediate. K. acutus , for example, has this suture almost complete, but often incomplete.

6) Prosternum: The prosternal basantra are somewhat well developed, usually wider than long (cf. Figs 112 View FIGURES 110–115 & 184 View FIGURES 182–187 ), often almost as long as wide (cf. Figs 178 View FIGURES 176–181 & 223 View FIGURES 221–226 ), rarely a little longer than wide (cf. Figs 70 View FIGURES 68–73 & 118 View FIGURES 116–121 ). Moreover, they tend to be widely separated from the ferna (cf. Fig. 202 View FIGURES 200–205 ), and this may be somewhat important for defining the genus. The prospinasternum is variable in form and size depending on the species, being very small in K. longicaudus and K. tenax ( Figs 118 View FIGURES 116–121 & 215 View FIGURES 212–220 ), but very large in K. simpliceps ( Fig. 202 View FIGURES 200–205 ).

7) Mesopresternum: The mesopresternum is also variable, but morphologically stable within species. Several species, such as K. micrommatus , have the mesopresternum rather reduced to small lateral triangles (cf. Fig. 148 View FIGURES 146–151 ), though several species, such as K. flavipes and K. melaleucus , have this sclerite not strongly reduced medially, transverse, but not broad at middle (cf. Fig. 88 View FIGURES 86–91 ). This structure also shows continuous change.

8) Tergite IX S1 and S2 setae: The form and the relative length of the posteromarginal setae on abdominal tergite IX is usually useful for determining species. They are usually pointed or expanded, and shorter or longer than the tube.

9) Tube: The comparative length of tube is sometimes useful for determining species, and shows continuous change. K. brevipilosus and K. oppositus have a relatively short tube, about 0.45 times as long as the head, whereas K. inflatus has the tube about 0.7 times as long as the head.

10) Terminal tube setae: The relative length of the terminal setae is usually long in this genus. K. yoshi has exceptionally short ones, almost as long as tube, whereas K. longicaudus has very long ones, more than 2.6 times as long as tube. Most of the remaining species have these setae intermediate, ranging from 1.3–2.5 times as long as tube, and the interspecific variation indicates continuous change.

Key to Karnyothrips species from Asia

[excluding K. expandosus Reyes View in CoL ; *: included from published descriptions]

1. Antennal segment III with one (0+1) sense cone............................................................. 2

- Antennal segment III with two (1+1) or three (1+2) sense cones............................................... 10

2. Antennal segment IV with two (1+1) major sense cones....................................................... 3

- Antennal segment IV with three (1+2) or (1+2 +1) major sense cone.............................................. 6

3. Tergite IX S1 setae expanded, shorter than tube............................................................. 4

- Tergite IX S1 setae pointed, almost as long as tube or longer................................................... 5

4. Body uniformly brown ( Fig. 22 View FIGURES 21–30 ); tube about 0.6 times as long as head................................... lalae sp. n.

- Body largely yellow (Fig. 158), but mesonotum and anterior portion of abdominal segment II very weakly shaded with brown; tube about 0.5 times as long as head, or a little longer................................... pellucidus View in CoL comb. n. (in part)

5. Postocular and prothoracic four major setae dilated; terminal tube setae a little longer than 2.0 times as long as tube..................................................................................... micrommatus sp. n. (in part)

- Postocular and prothoracic four major setae pointed or bluntly pointed, at least not distinctly dilated; terminal tube setae about 1.5 times as long as tube, at least much shorter than 2.0 times as long as tube................................ spinulus View in CoL

6. Body uniformly brown to dark brown; antennal segment III with sub-basal swelling; maxillary stylets about one-third of head width apart from each other..................................................................... antennalis View in CoL

- Body bicolored brown and yellow, or largely yellow; Antennal segment IV without sub-basal swelling; maxillary stylets narrower than one-fourth of head width apart from each other.................................................. 7

7. Body largely yellow (Fig. 158), with head largely yellow................................ pellucidus View in CoL comb. n. (in part)

- Body bicolored, with head brown........................................................................ 8

8. Tergite IX S1 setae pointed; prothorax yellow; if wings fully developed, fore wing without duplicated cilia......... alpha View in CoL *

- Tergite IX S1 setae dilated; prothorax brown; if wings fully developed, fore wing with duplicated cilia................. 9

9. Pterothorax brown; antennal segment IV with a small outer sense cone in addition to three major sense cones, (1+2 +1).................................................................................................. mucidus View in CoL *

- Pterothorax yellow ( Fig. 152 View FIGURES 152–157 ); antennal segment IV with three major sense cone, without small outer sense cone, (1+2)................................................................................................ nigriflavus View in CoL

10. Antennal segment III with two sense cones, (1+1).......................................................... 11

- Antennal segment III with three sense cones, (1+2)......................................................... 40

11. Antennal segment IV with two major sense cones, (1+1) or (1+1 +1)............................................. 12

- Antennal segment IV with three or four major sense cones, (1+2), (1+2 +1), (2+2) or (2+2 +1).......................... 22

12. Antennal segment IV with two major sense cones, without small outer sense cone, (1+1)............................ 13

- Antennal segment IV with a small outer sense cone in addition to two major sense cones, (1+1 +1)..................... 15

13. Antennal segment III with sub-basal swelling; antennal segments VII and VIII widely joined, formed a single segment outline; S1 setae on tergite IX dilated, shorter than tube; tube about 0.7 times as long as head........................... inflatus View in CoL

- Antennal segment III without sub-basal swelling; antennal segments VII and VIII well separated, narrowly joined; S1 setae on tergite IX pointed, longer than tube; tube shorter than 0.6 times as long as head................................... 14

14. Body uniformly brown ( Fig. 28 View FIGURES 21–30 ); head shorter than 1.4 times as long as wide ( Figs 146 & 147 View FIGURES 146–151 ); postocular setae longer than eyes, dilated; prothoracic ml setae well developed; tube about 0.6 times as long as head; terminal setae about 2.0 times as long as tube, or a little longer.......................................................... micrommatus sp. n. (in part)

- Head and thorax brownish yellow to yellowish brown, slightly paler than brown abdomen, but often thorax darker, at least anterior portion of head yellowish ( Figs 4, 5 View FIGURES 1–5 & 243 View FIGURES 240–246 ); head longer than 1.5 times as long as wide ( Fig. 240 View FIGURES 240–246 ); postocular setae shorter than eyes, pointed or bluntly pointed; prothoracic ml setae reduced; tube about half of head length; terminal setae almost as long as tube, or a little longer....................................................................... yoshi View in CoL

15. Tergite IX S1 setae dilated............................................................................. 16

- Tergite IX S1 setae pointed............................................................................ 18

16. Body uniformly brown to dark brown ( Fig. 29 View FIGURES 21–30 ); antennal segment III with sub-basal swelling ( Fig. 169 View FIGURES 164–169 ); antennal segments VII and VIII widely joined ( Fig. 168 View FIGURES 164–169 ), formed a single segment outline; prothoracic notopleural sutures usually incomplete............................................................................................ oppositus sp. n.

- Body bicolored brown and yellow; antennal segment III without sub-basal swelling; antennal segments VII and VIII well separated, narrowly joined; prothoracic notopleural sutures complete........................................... 17

17. Head, prothorax and mesothorax largely yellowish ( Fig. 17 View FIGURES 6–20 ), partly shaded with brown; leges largely yellow. flavescens sp. n.

- Head, prothorax and mesothorax brown to dark brown ( Fig. 36 View FIGURES 31–40 ), fore femora largely brown, mid and hind femora largely pale brown;............................................................................... submaculatus sp. n.

18. Body uniformly brown to dark brown.................................................................... 19

- Body bicolored brown and yellow....................................................................... 20

19. Antennal segment III with small sub-basal swelling ( Figs 46 & 47 View FIGURES 41–47 )..................................... acutus sp. n.

- Antennal segment III without sub-basal swelling ( Fig. 61 View FIGURES 56–61 ).......................................... apoensis sp. n.

20. Head and thorax largely yellow ( Fig. 33 View FIGURES 31–40 ); prothoracic notopleural sutures complete; terminal tube setae about 1.5 times as long as tube................................................................................. semiflavus sp. n.

- Head and thorax brown to dark brown; prothoracic notopleural sutures incomplete; terminal tube setae about 2.0 times as long as tube............................................................................................. 21

21. Head very weakly sculptured with transverse striae at middle; maxillary stylets reaching postocular setae, 22–27µm apart from each other at middle in female ( Fig. 68 View FIGURES 68–73 ); mesopresternum very narrow at middle, often scarcely divided ( Fig. 70 View FIGURES 68–73 ); sub-basal wing setae S1 longer than S3; tergite IX S1 setae about 1.2 times as long as tube or shorter; antennal segments IV and V brownish yellow to yellowish brown: abdominal segment VIII largely brown ( Fig. 15 View FIGURES 6–20 )................ brevipilosus sp. n.

- Head almost smooth; maxillary stylets barely reaching eyes, 13–22µm apart ( Fig. 134 View FIGURES 134–139 ); mesopresternum broader at middle; sub-basal wing setae S3 longer than S1; tergite IX S1 setae 1.3–1.4 times as long as tube or longer; antennal segments IV and V yellow; abdominal segment VIII largely yellow in female ( Fig. 26 View FIGURES 21–30 ), but largely brown in male ( Fig. 27 View FIGURES 21–30 )....... melaleucus View in CoL

22. Antennal segment IV with three major sense cones, (1+2) or (1+2 +1)............................................ 23

- Antennal segment IV with four major sense cones, (2+2) or (2+2 +1)............................................ 27

23. Mesopresternum reduced medially, usually divided......................................................... 24

- Mesopresternum transverse, not divided.................................................................. 25

24. Body uniformly brown ( Fig. 35 View FIGURES 31–40 ); antennal segment IV without small outer sense cone, (1+2); antennal segment III without sub-basal swelling ( Fig. 205 View FIGURES 200–205 )................................................................ simpliceps sp. n.

- Body bicolored ( Figs 37–39 View FIGURES 31–40 ); antennal segment IV with a small outer sense cone, (1+2 +1); antennal segment III with sub-basal swelling ( Fig. 226 View FIGURES 221–226 )......................................................................... triconus sp. n.

25. Terminal tube setae much longer than 2.5 times as long as tube................................... longicaudus sp. n.

- Terminal tube setae about 2.0 times as long as tube.......................................................... 26

26. Antennal segment III entirely yellow.............................................................. flavicornis View in CoL

- Antennal segment III brownish at apical half, yellow at basal half ( Fig. 91 View FIGURES 86–91 )........................... flavipes View in CoL (in part)

27. Antennal segment IV with four major sense cones, without small outer sense cone, (2+2)........................... 28

- Antennal segment IV with two major sense cones, with a small outer sense cone, (2+2 +1)........................... 31

28. Antennal segment III with sub-basal swelling.................................................. cyathomorphus View in CoL *

- Antennal segment III without sub-basal swelling........................................................... 29

29. Body largely yellowish, but prothorax pale brown; antennal segment III extremely short, shorter than 0.7 times as long as segment IV, about 1.3 times as long as wide........................................................... insignis View in CoL

- Body uniformly brown or dark brown; antennal segment III normal............................................ 30

30. Head relatively short ( Fig. 62 View FIGURES 62–67 ), about 1.2 times as long as wide; tergite IX S1 setae almost as long as tube... breviceps View in CoL sp. n.

- Head long, longer than 1.5 times as long as wide ( Fig. 176 View FIGURES 176–181 ); tergite IX S1 setae shorter than tube. quadriconus sp. n. (in part)

31. Body uniformly brown to dark brown.................................................................... 32

- Body bicolored, at least intermediate abdominal segment paler than prothorax.................................... 35

32. Mesopresternum transverse ( Fig. 88 View FIGURES 86–91 ), not divided; antennal segment VIII not constricted at base ( Fig. 90 View FIGURES 86–91 )... flavipes View in CoL (in part)

- Mesopresternum reduced medially, divided into lateral two triangle plates (cf. Fig. 100 View FIGURES 98–103 ); antennal segment VIII weakly or scarcely constricted at base (cf. Fig. 102 View FIGURES 98–103 )................................................................. 33

33. Tergite IX S1 setae pointed, longer than tube..................................................... infectus sp. n.

- Tergite IX S1 setae expanded, shorter than tube............................................................ 34

34. Antennal segment III with small sub-basal swelling ( Fig. 97 View FIGURES 92–97 )..................................... formosanus sp. n.

- Antennal segment III without sub-basal swelling ( Fig. 181 View FIGURES 176–181 )............................... quadriconus sp. n. (in part)

35. Head largely yellowish, with anterior portion brownish, but often largely brown with base yellowish.................. 36

- Head uniformly brown or dark brown.................................................................... 37

36. Prothorax brown ( Fig. 129 View FIGURES 128–133 )................................................................. maculatus sp. n.

- Prothorax yellow ( Fig. 141 View FIGURES 140–145 )................................................................ micans View in CoL comb. n.

37. Eyes largely developed ( Fig. 122 View FIGURES 122–127 ), about 0.4 times as long as head; metathorax largely yellowish at middle ( Fig. 24 View FIGURES 21–30 ); terminal tube setae about 1.4 times as long as tube.................................................. macrommatus sp. n.

- Eyes about one-third as long as head or smaller; metathorax brown; terminal tube setae about 2.0 times as long as tube, or longer............................................................................................. 38

38. Head about 1.6 times as long as wide ( Fig. 48 View FIGURES 48–55 ); maxillary stylets about one-sixth of head width apart from each other in female..................................................................................... affinis sp. n.

- Head about 1.4 times as long as wide (cf. Fig. 194 View FIGURES 194–199 ); maxillary stylets about one-fifth of head width apart from each other or wider in female...................................................................................... 39

39. Intermediate abdominal segments without median brown marking ( Fig. 34 View FIGURES 31–40 ); antennal segment II brown with apex yellowish; maxillary stylets wider than one-fourth of head width in female ( Fig. 194 View FIGURES 194–199 ); prospinasternum less than 40µm wide ( Fig. 196 View FIGURES 194–199 ); pronotal aa setae much shorter than postocular setae................................................. similis sp. n.

- Intermediate abdominal segments each with a median brown marking ( Fig. 40 View FIGURES 31–40 ); antennal segment II almost uniformly yellowish; maxillary stylets narrower than one-fourth of head width in female ( Fig. 227 View FIGURES 227–232 ): prospinasternum more than 50µm wide ( Fig. 229 View FIGURES 227–232 ); pronotal aa setae almost as long as postocular setae, or a little shorter.................... variabilis sp. n. (in part)

40. Fore femur with sub-basal inner hump; fore tibia with sub-apical inner tooth or scale; prosternal membrane between basantra and ferna well developed, with transverse dense striae................................................ tenax sp. n.

- Fore femur and tibia unarmed; prosternal membrane normal.................................................. 41

41. Antennal segment III without outer small sense cone, (2+2).............................. quadriconus sp. n. (in part)

- Antennal segment IV with an outer small sense cone, (2+2 +1).................................................. 42

42. Body uniformly brown to dark brown.................................................................... 43

- Body bicolored brown and yellow....................................................................... 48

43. Tergite IX S1 setae expanded, almost as long as tube or shorter, if pointed or bluntly pointed and longer than tube, antennal segments III and IV yellow and largely brown respectively................................................... 44

- Tergite IX S1 setae pointed, longer than tube, if antennal segment III yellow, segment IV also yellow.................. 46

44. Tube 0.5 times as long as head...................................................... quadriconus sp. n. (in part)

- Tube 0.6 times as long as head.......................................................................... 45

45. Base of antennal segment IV brownish, somewhat darker than segment III; antennal segment VIII conical, not constricted at base; maxillary stylets about one-third of head width apart from each other.................................. pacifics

- Base of antennal segment IV yellow, concolorous with segment III; antennal segment VIII slightly constricted at base; maxillary stylets about one-fourth of head width apart from each other...................................... robustus View in CoL (in part)

46. Antennal segment III brown, with basal one-third yellowish; each maxillary stylet more than two-fifths of head width apart; mesopresternum divided or very narrow at middle......................................................... acer View in CoL

- Antennal segment III yellow; each maxillary stylet less than one-third of head width apart; mesopresternum transverse, not divided............................................................................................ 47

47. Antennal segments IV and V brown, but IV with base yellowish; pelta trapezoid; antennal segment VIII weakly constricted at base................................................................................... robustus View in CoL (in part)

- Antennal segments IV and V yellow; pelta triangular; antennal segment VIII not constricted at bae..... vietonamensis sp. n.

48. Head largely brown, with posterior portion yellowish ( Figs 104 & 105 View FIGURES 104–109 ); prothorax largely yellowish, pterothorax pale brown; legs yellow ( Fig. 21 View FIGURES 21–30 )........................................................................ inpallens sp. n.

- Head and thorax brown or dark brown; at least fore femora largely brown....................................... 49

49. Intermediate abdominal segments each with a median large brown marking along antecostal suture ( Fig. 30 View FIGURES 21–30 ); prospinasternum small ( Fig. 171 View FIGURES 170–175 ); tube longer than 0.6 times as long as tube; terminal tube setae about 1.5 times as long as tube.. pitkini sp. n.

- Intermediate abdominal segments each with a median small marking along antecostal suture; prospinasternum large; tube shorter than 0.6 times as long as tube; terminal tube setae about 2.0 times as long as tube, or a little longer.............. 50

50. Eyes largely developed ( Fig. 74 View FIGURES 74–79 ), about 0.4 times as long as head; maxillary stylets about one-sixth of head width apart from each other................................................................................ capitatus sp. n.

- Eyes normally developed ( Fig. 227 View FIGURES 227–232 ), about one-third as long as head; maxillary stylets more than one-fifth of head width apart from each other.................................................................... variabilis sp. n. (in part)