Nymphomyia holoptica Courtney 1994

Nymphomyia holoptica

Palaeodipteron sp.—Dudgeon, 1989:193.

DIAGNOSIS.—Larva: median tooth of postmentum projecting anteriorly to level of adjacent lateral teeth and with one pair of lateral serrations; lateral teeth without lateral serrations, notch between lateral teeth I and II as deep as notches separating other teeth. Pupa: rostral hooks divergent, abdominal sternites without hooks, tergites II–VIII with spine rows. Adult: ommatidia of compound eyes (three pair) contiguous ventrally, abdominal segment VIII set with setiform dorsolateral sensilla; male with gonocoxites enlarged and extending anteroventrally to beyond margin of segment VIII, gonostylus bifurcate, with two similarly shaped, curved lobes; female with narrow, elongate lobe on posterolateral margin of tergite VIII. and posteroventrally directed cerci.

DESCRIPTION.—Larva: Postmentum (Figure 31) not trilobed anteriorly; median tooth projecting anteriorly to level of adjacent lateral teeth and with one pair of lateral serrations; lateral teeth without lateral serrations; notch between lateral teeth I and II as deep as notches separating other lateral teeth.

Pupa: Rostral hooks elongate, divergent; antennal sheath elbowed. Microsculpture: abdominal sternites without hooks; tergites II-VII with posterior spine rows. Anal hooks short, with small dorsal serrations. Segment IX sexually dimorphic: male with single enlarged, posteroventral lobe; female without lobe, segment somewhat shortened (Figures 32, 34).

Adult : Rostrum apically truncate (Figures 29, 30). Cranium glabrous anterodorsally, from apex of rostrum to near anterior margins of eyes; microtrichia absent posteroventrally, Ommatidia of compound eyes contiguous ventrally, contact between three pairs of facets; microtrichia between dorsal facets only. Terminal sensilla of antenna apical. Empodium of foretarsus short, not extending beyond apex of tarsal claws; claws with prominent basal tooth (Figure 39).

Terminalia (Figures 35, 36): Stenites V and VI and tergite VIII simple, without elongate processes; tergite VIII with large dorsolateral patch of setiform sensilla. Membrane between tergites VIII and IX without setiform sensilla. Gonocoxites fused broadly to cerci, elongate, projecting anteriorly to just beyond posterior margin of sternite VIII, invested with microtrichia, curved medially and tapered toward apex, with setiform sensilla along inner margin. Gonostylus bifurcate, of two glabrous, similarly shaped, curved lobes; anterior lobe relatively more tapered in apical half than posterior lobe, not overlapping posterior lobe from opposite side; anterior lobe relatively more curved, acute apically (Figure 35). Aedeagus short, thick, broadest at base; when not exserted, extending anteriorly to abdominal segment VII.

Adult : Similar to male. Terminalia (Figure 33): Tergite VIII with large dorsolateral patch of setiform sensilla, glabrous anterolateral tubercle (vestigial spiracle?), and elongate, ventrally projected lobe. Sternite VIII membranous, microtrichia generally sparse. Cerci (?) broad, directed posteroventrally.

TYPE MATERIAL.—Holotype : “Loc: Tai Po Kau Forest stream, Hong Kong, China [New Territories, southwestern shore of Tolo Harbour, ~140 m] Date:—1981, coll: David Dudgeon.” Specimen in Canada balsam on slide [CNC], Allotype : same data as holotype; in Canada balsam on slide [CNC]. Paratypes: “HONG KONG, New Territories, southwestern shore of Tolo Harbour, Tai Po Kau Forest stream, ~140 m, 22 Jul 1983 [coll. David Dudgeon]” [2 larvae, instar IV; 2 pupae , ]. Paratypes in Canada balsam on slides [USNM].

DISTRIBUTION (Figure 96).—Known only from Hong Kong.

ETYMOLOGY.—Species name from the Greek holos, for entire, and optikos, sight, in reference to the extensively fused compound eyes of the adult.

Phylogenetics

INTERSPECIFIC RELATIONSHIPS.—Previous investigations have not explicitly discussed relationships within the Nymphomyiidae, yet phylogeny and classification have been implicit in recognition of three genera, two of which were monobasic. Ide (1965), by describing the Nearctic species Palaeodipteron walkeri as a new genus, initiated this trend toward monobasic genera. Criteria for placing this nymphomyiid in a genus other than Nymphomyia included “annulate” antennal flagellum (vs. “entire”), empodia “with stout spines” (vs. “brush-like”), and wing a “small scale” (vs. “large, triangular with long fringing setae”). I do not consider the former two differences particularly convincing, and even Ide correctly predicted that the wings of his specimens were in “a de-alated condition.” For these reasons and others (see below) I have synonymized Palaeodipteron under the type genus Nymphomyia. The terminalia of N. walkeri and N. alba, the only other species known at that time of Ide's description, are quite different and might justify ranking these taxa as separate genera; however, most terminalia differences reflect autapomorphies of individual species, which are of little use for delimiting higher taxa.

The Himalayan genus Felicitomyia was said to be “somewhat less specialized” than Nymphomyia and Palaeodipteron, but with certain attributes of both (Cutten and Kevan, 1970). Generic rank was based on adult Felicitomyia, which were purportedly “sexually dimorphic” in head structure, yet both known adults are males (i.e., the allotype “female” is actually a male (personal observation)). Consequently, stated differences between the females of Felicitomyia and other nymphomyiids are untenable. Cutten and Kevan described minor differences in the shape of the antenna, tibia, and other structures, but these characters are variable and insignificant. Cutten and Kevan also mentioned differences between the male terminalia of Felicitomyia and those of other nymphomyiids; e.g., abdominal segments V and VI without ventral appendages (present in Palaeodipteron), abdominal segment VII without paratergal projections (present in Nymphomyia). Again, one could argue that terminalia differences justify recognition of three genera, but as was true regarding Palaeodipteron, most differences reflect autapomorphies. In contrast to earlier analyses, the present study considers both morphology and phylogeny as a basis for formal classification (see below).

The present analysis evaluated 20 characters and eight taxa (Tables 2, 3; Figure 3), including a generalized outgroup based largely on the Culicomorpha, Tipulidae, and Ptychopteromorpha. Autapomorphies, including certain features of the male terminalia (e.g., appendages on abdominal segments V and VI in N. walkeri), provide no information about relationships and were therefore excluded from the analysis. Most multistate characters (9, 10, 12) were coded as additive, as outgroup comparison permitted logical arrangement in linear transformation series. Character 20, however, was coded as non-additive, because there was no logical basis for polarizing alternative character states in a transformation series; this was confounded further by lack of data on two species. Character analysis resulted in a single tree (Figure 3) with a length of 33 steps, and consistency and retention indices of 78 and 78, respectively; this hypothesis accepts homoplasy in six characters (3, 4, 5, 17, 18, 20).

LARVA

1. median postmental tooth 0. one pair of lateral tines

1. two pairs of lateral tines

2. anterior margin postmentum 0. median tooth to level of lateral teeth

1. trilobed, median tooth recessed

PUPA

3. rostral sheath 0. elongate, slender

1. short, triangular

4. antennal sheath 0. straight

1. elbowed

5. abdominal sternites 0. without recurved hooks

1. with pair of recurved hooks

6. gonocoxite sheath 0. glabrous, without distinctive spine

1. with pair of sharp spines

7. gonostylus sheath 0. bilobed

1. enlarged, bilobed form lost

8. anal hook 0. large, elongate

1. short, triangular

ADULT

9. eye vestiture 0. without microtrichia between facets

1. microtrichia between dorsal facets

2. microtrichia between all facets

10. ventral eye bridge 0. absent

1. present, but facets not contiguous

2. contact between two pairs of facets

3. contact between three pairs of facets

11. tarsal claw 0. basal tooth present

1. basal tooth absent

12. paratergal projections 0. absent

1. one pair (ventral) present

2. two pairs (ventral and dorsal) present

13. abdominal segment VII 0. with dorsolateral patch of setiform sensilla

1. without dorsolateral patch of setiform sensilla

14. abdominal segment VII 0. setiform sensilla generally distributed on tergile

1. setiform sensilla clustered above ventral paratergal projection

15. gonocoxites and cerei 0. distinctly separable

1. indistinguishably fused

16. gonocoxites 0. extend posteriorly

1. extend anteriorly

17. gonostyli 0. simple

1. bifurcate

18. aedeagus (at rest) 0. short, extending (internally) anteriorly to segment VIII

1. elongate, extending anteriorly to segment VII

2. markedly elongate, extending to segment V

19. aedcagal sheath 0. tube-like, broad basally, gradually tapering to apex

1. markedly broad, sinuate laterally, truncate apically

20. abdominal segment VII 0. without ventrally directed, lateral processes or flaps

1. with narrow, lateral flap near posterior margin

2. with broad, lateral flap

Cladistic analysis of larval, pupal, and adult characters suggests that the Nymphomyiidae comprise two species groups with the following phylogenetic relationship (Figure 3): (1) alba group (alba + (levanidovae + rohdendorfi)) and (2) walkeri group ((walkeri + dolichopeza) + (brundini + holoptica)). This hypothesis is based largely on features of the adult cranium and terminalia and, to a lesser extent, on pupal and larval characters. A transformation series involving the adult eye (character 10, ventral eye bridge absent—bridge present, but facets not contiguous—contact between two pairs of facets—contact between three pairs of facets) is one of the most compelling characters. The latter two steps, as well as two other characters—(9) microtrichia between eye facets and (15) gonocoxites and cerci indistinguishably fused—support the monophyly of the walkeri group. Characters that help define the alba group include: (12) paratergal projections; (14) cluster of setiform sensilla above ventral paratergal projection; (19) aedeagal sheath broad, sinuate laterally, truncate apically. Other characters provide evidence for the monophyly of the Oriental species brundini + holoptica: (7) gonostylus sheath enlarged, not bilobed; (8) pupal anal hooks short, triangular; and (16) gonocoxites extend anteriorly. Although data suggest that the Nearctic fauna (walkeri and dolichopeza) is monophyletic, the basis for this hypothesis is weak, consisting of two reductive characters: (11) basal tooth of tarsal claw lost; (13) setiform sensilla on abdominal segment VII of male lost. Nymphomyia walkeri is clearly very distinctive, possessing more unique features than any other species, yet N. dolichopeza does not share this distinctness. Because differences between currently recognized genera are minor or reflect autapomorphies, their separate taxonomic status is untenable. I therefore synonymize Palaeodipteron and Felicitomyia with Nymphomyia.

ZOOGEOGRAPHY.—Structural and phylogenetic criteria help separate species of Nymphomyia into two groups, the northeastern Palaearctic alba group (alba + (rohdendorfi + levanidovae)) and the walkeri group, with the latter subdivided further into two, geographically distinct, lineages: (1) brundini + holoptica in the Oriental Region and (2) walkeri + dolichopeza in the Nearctic Region. While little can be said about zoogeographic relationships within these regions, the relationship between the Palaearctic, Oriental, and Nearctic faunas deserves comment. Of particular interest is the direction and timing of origin of the Nearctic nymphomyiid fauna. Numerous studies in cladistic biogeography have examined the origins of the North American insect fauna, particularly the relationships between eastern Nearctic groups and those from the western Nearctic or the Palaearctic (e.g., Allen, 1983; Noonan, 1986, 1988). Results of these studies have implications for the importance of Beringian vs. trans-Atlantic dispersal routes and, concomitantly, for the timing of faunal invasions. The Beringian land bridge existed periodically until recently, with its latest appearance during the Late Wisconsinan glaciation (ca. 18,000 years ago). Noonan (1988) posits that trans-Atlantic dispersal routes were severed approximately 20–35 million years ago, with more recent dates coinciding with the Thulean land bridge. Phylogenetic evidence suggests that the Nearctic nymphomyiid fauna is monophyletic and represents a single invasion of North America, yet the source of this invasion is unclear. The absence of nymphomyiids in either the western Nearctic or western Palaearctic is problematic. The sister-group of the Nearctic fauna is the Oriental clade brundini + holoptica (Figure 98), which could support the hypothesis of an ancient trans-Atlantic invasion; this requires that nymphomyiids later became extinct in the western Palaearctic. It is possible that the group has simply been overlooked in the western Nearctic or western Palaearctic and that appropriate collection methods will result in their discovery. However, because the Diptera fauna of Europe and western Asia is relatively well known, it is doubtful that nymphomyiids will be found in the western Palaearctic. Surveys of central and eastern Asian streams hold more promise. Further sampling in western North America, particularly the extreme northwest (e.g., Alaska), may lead to the discovery of nymphomyiids from the western Nearctic; however, I predict that nymphomyiids from this area will be more closely related to the alba group than to eastern Nearctic species. Nymphomyia rohdendorfi presently occurs in extreme northeastern Russia, and it is reasonable to predict that this or a related species could have invaded the arctic or subarctic regions of Alaska during the Pleistocene. The accepted phylogeny predicts that such an invasion would be more recent than and unrelated to that which established the eastern Nearctic fauna.

Although the range of nymphomyiids is fairly broad, including the Nearctic, Palaearctic, and Oriental regions, species distributions within these areas typically are restricted and disjunct. The Russian and North American faunas each contain one relatively widespread species (rohdendorfi and walkeri, respectively), but most nymphomyiids are known from one or a few localities. The present distribution of nymphomyiids, particularly in the Appalachian Mountains, can be interpreted as relictual; i.e, nymphomyiids were previously more widespread, and the disjunct pattern seen today reflects local extinction through much of their historical range. Whether or not this extinction was caused by natural (e.g., post-glacial warming of streams) or anthropogenic (e.g., water pollution) phenomena is unclear. Pesticide studies at Coweeta Hydrologic Laboratory have shown that nymphomyiids can recolonize impacted streams after just a few years (J.B. Wallace, 1990, pers. comm.). The mechanism of recolonization has not been established, but may depend on downstream drift of immature stages from tributary streams, or on colonization by adults from adjacent watersheds. Because of their presumedly short life and limited flight abilities, adult nymphomyiids are thought to be poor dispersers. These presumptions may be largely justified; however, demonstration that adults can survive in the laboratory for up to four days suggests that their life span may be greater than predicted. Furthermore, our ignorance about nymphomyiid flight behavior precludes definitive conclusions about adult dispersal capabilities. Wing morphology suggests that nymphomyiids are poor “active” fliers, but may be adept at “passive” (e.g., wind-assisted) flight. If so, aerial plankton samples in the appropriate habitats may provide adult nymphomyiids. Such collections would demonstrate that these flies have a greater dispersal potential than previously expected, and would provide an alternative to the above (vicariance-based) explanation for their disjunct distributions.

Clearly many discoveries about nymphomyiid phylogeny, biogeography, and ecology remain for future dipterists and aquatic biologists. Recent collections confirm that nymphomyiids are more widespread and diverse than was previously thought, and records from Hong Kong indicate that the group is not restricted to boreal and temperate zones. It is, therefore, probable that new records and species will be found in other areas, including parts of southeastern Asia and perhaps even mountain streams of the Neotropics. However, as is the case for many small, specialized insects, the key to future discoveries will be to adopt collection methods that effectively sample their unusual habitats.