Toxorhynchites (Lynchiella) haemorrhoidalis (Fabricius)

Harbach, Ralph E. & Wilkerson, Richard C., 2023, The insupportable validity of mosquito subspecies (Diptera: Culicidae) and their exclusion from culicid classification, Zootaxa 5303 (1), pp. 1-184 : 118-120

Toxorhynchites (Lynchiella) haemorrhoidalis (Fabricius)

subspecies haemorrhoidalis ( Fabricius, 1787) —original combination: Culex haemorrhoidalis . Distribution: Argentina, Bolivia, Brazil, Colombia, Ecuador, French Guiana, Guyana, Paraguay, Peru, Suriname, Trinidad and Tobago, Venezuela (updated from Knight & Stone 1977).

subspecies separatus ( Lynch Arribálzaga, 1891b) —original combination: Megarhina [sic] separata (subspecific status by Lane 1951). Distribution: Argentina, Bolivia, Brazil, Paraguay ( Wilkerson et al. 2021, Nicaragua deleted).

subspecies superbus ( Dyar & Knab, 1906a)—original combination: Megarhinus superbus (subspecific status by Lane 1951). Distribution: Belize, Colombia, Costa Rica, Cuba, Ecuador, French Guiana, Guatemala, Honduras, Mexico, Nicaragua, Panama, Suriname, Trinidad and Tobago, Venezuela ( Wilkerson et al. 2021).

The taxonomic history of Tx. haemorrhoidalis involves four nominal species: Culex haemorrhoidalis Fabricius, 1787 , Megarhina separata Lynch Arribálzaga, 1891b , Megarhinus lynchi Dyar & Knab, 1906a and Megarhinus superbus Dyar & Knab, 1906a . Megarhina separata was considered a synonym of haemorrhoidalis as early as Lutz & Neiva (1913). Dyar (1928) followed earlier workers in recognizing Megarhinus lynchi as a valid species; it was regarded as a questionable synonym of haemorrhoidalis by Edwards (1932a) and formally synonymized with separatus by Lane (1939). Five years later, Lane (1944) treated haemorrhoidalis , separatus and superbus as separate species; however, he distinguished the last two based only on distribution: superbus in Central America and separatus in Argentina. It seems Lane (1951), without explanation, was unable to further support the specific rank of separatus and superbus and reduced them to subspecies of haemorrhoidalis . Lane (1953), in accordance with his earlier interpretation, distinguished the two subspecies based only on distribution, and stated in a note that “We have placed T. superbus and T. separatus as subspecies of T. haemorrhoidalis . Such a course taken by us is strengthened by the fact that the zoogeographical distribution of the three forms is quite distinct.” This is not clear, but we believe Lane intended to say that the distributions of each of the three forms are distinct from one another. Except for Vargas (1953), who either disagreed with or was unaware of Lane (1951, 1953) and recognized superbus as a species in Central America, separatus and superbus have continued to be recognized as subspecies to this day.

Dyar & Knab (1906a), in a discussion explaining why they gave the new name lynchi to the species in Argentina previously identified as haemorrhoidalis by Lynch Arribálzaga (1891b), Theobald (1901a), Giles (1902) and Blanchard (1905), stated that “Great confusion has been caused by basing the diagnosis on the tarsal markings without reference to sex. We find that when the sexual differences are considered the tarsal markings are a useful guide in the diagnosis of the species and are a much more constant character than has been supposed.” It is noteworthy that Dyar & Knab treated separatus , also originally described from Argentina, as a synonym of haemorrhoidalis , and described superbus as a new species based on specimens from Trinidad and Mexico. Their concept of superbus also included the identification of haemorrhoidalis by Williston (1900) based on specimens from Cuba, French Guiana and Mexico, as well as the identification of Megarhinus violaceus by Dyar & Knab (1906a) and Coquillett (1906) based on specimens collected in Central America. Dyar (1928) subsequently separated lynchi and superbus from haemorrhoidalis based on the presence or absence of a basal pale band on hindtarsomere 2, present in haemorrhoidalis and absent in the other two. He distinguished lynchi and superbus based on environmental location: superbus “From the northern edge of the tropics” and lynchi “From the southern edge of the tropics”. These distinctions were restated, using slightly different terminology, by Lane (1944, 1953), an indication that he either accepted the observations of Dyar without further study or he was unable to find additional characters to distinguish the three nominal forms.

In his treatment of the Toxorhynchites (as Megarhinus ) of “ Brasil Meridional” (southern Brazil), Lane (1944) stated the following (translated from the Portuguese).

It is very interesting to note that, while this species [ haemorrhoidalis ] occurs in the Guianas and the Amazon Valley, the two related species are found, one in the North ( superbus ) and the other in the South ( separatus ). The distinguishing characteristics of these three species reside in the development of the abdominal tufts of the males and in the marking of the tarsi of the females.

We think it very likely that they represent a single species and that both superbus and separatus are just geographic forms. A definitive solution of this case is impossible for us due to lack of material.

The available descriptions of the leg markings of both sexes are confusing and lack explicit detail, but it appears that they are the same in all four nominal forms except for hindtarsomere 2 of females, which is pale basally in haemorrhoidalis and completely dark-scaled in lynchi , separatus and superbus . The lateral tufts on the posterior abdominal segments of males are said to be more strongly developed in haemorrhoidalis than they are in the other three nominal forms, but the degree of development has not been made explicit. Dyar & Knab (1906a) indicated that strictly red tufts are only present on abdominal segment VII in superbus but are present on segments VI and VII in haemorrhoidalis and lynchi . In contrast, Dyar (1928) stated that the male of haemorrhoidalis has “Abdominal red tufts on the last four segments”, and this is quoted verbatim by Lane (1953). In their identification keys, Dyar (1928) and Lane (1953) merely indicated that the “abdominal red tufts [are] well developed” in haemorrhoidalis and are “less developed” in lynchi ( separatus of Lane) and superbus . It is interesting to note that Dyar & Knab (1906a) distinguished the males of haemorrhoidalis and lynchi based on the length of “segments 3 and 4” [palpomeres 3 and 4] of the maxillary palpus—equal in length in haemorrhoidalis and 3 longer than 4 in lynchi . Considering what is now known about the development of the maxillary palpus of mosquitoes ( Harbach & Kitching 1998), “segments 3 and 4” are actually palpomeres 4 and 5. Oddly, this character was not mentioned in later works, and the descriptions of the maxillary palpi provided by Dyar (1928) and Lane (1953) are ambiguous. Dyar stated that haemorrhoidalis has “Palpi with the third joint long and pointed” (quoted verbatim by Lane), which surely must refer to the terminal palpomere, and Lane added that superbus has the “Last palpal segment long and acuminate” and separatus is “Similar to T. haemorrhoidalis superbus .”

As revealed by da Costa Lima (1931), Dyar (1928) failed to notice that Goeldi (1905) had described the egg, larva and pupa of separatus , and provided a color illustration of the adult male. It is interesting to note that hindtarsomere 2 is pale basally in the male illustrated by Goeldi, indicating that the species he described is not separatus , nor any of the other three nominal forms under discussion here, all of which have hindtarsomere 2 completely dark-scaled. Séguy (1950) provided a similar color illustration, which seems to correctly depict the male of haemorrhoidalis .

The larva and pupa of separatus were briefly described by Forattini & Lane (1952) based on a single larva captured in the Serra do Diabo region in the western area of São Paulo State in southern Brazil that was reared to an adult. The sex of the adult was not mentioned, but it is presumed to have been a female otherwise the specimen could not have been identified as separatus . The descriptions (in Portuguese) were repeated (in English) by Lane (1953). Dyar (1928) and Lane (1953) provided brief descriptions of the larva of separatus . Their descriptions lack comparable information except the former author reported that the siphon is “over four times as long as wide” whereas the latter author described the siphon as being “three and a half times basal width.” Lane (1953) also provided brief descriptions of the larva and pupa of the nominotypical form. He unintelligibly characterized the larval siphon as “slightly more than one time as broad as wide.” However, judging from his illustration of the terminal abdominal segments, he obviously meant to say the siphon is slightly longer than broad. Vargas (1953), in a key for the identification of larvae of species of Toxorhynchites (as Megarhinus ) known to occur in Venezuela, characterized the larval siphon of superbus as being twice as long as the saddle of segment X (transliterated from the Spanish). In comparison, Lane (1953) described the length of the siphon of superbus as being “two and a half times greatest width.” Although these authors expressed siphon length in different ways, the degree of actual difference seems to be greater than expected for individuals of the same species.

The subgenus Lynchiella Lahille, 1904 , to which haemorrhoidalis and 16 other species belong, is predominantly Neotropical, with an extension into eastern areas of the United States and southeastern Canada represented by Tx. rutilus (Coquillett, 1896) (see below). For the most part, the current taxonomy of Lynchiella dates back to Lane (1953) and Vargas (1953). Their studies were based almost entirely on adult mosquitoes, and the immature stages, as noted above, were described very superficially. With the exception of Tx. gerbergi Belkin, 1977 , Tx. guadeloupensis ( Dyar & Knab, 1906a) and Tx. portoricensis (von R̂der, 1885), the larva and pupa of which were fully described and illustrated by Belkin (1977), Augier et al. (2003) and Belkin et al. (1970), respectively, the complete larval and pupal chaetotaxy has not been studied for any other species of the subgenus. As noted by Belkin et al., Toxorhynchites “is amazingly similar in all [life] stages throughout its nearly worldwide distribution and great difficulty is experienced in identifying species. New World species have been diagnosed largely on the basis of light markings of the tarsi, which frequently differ in the 2 sexes and are not always reliable. The metallic coloration of the thoracic scales shows considerable differences among species but is subject to some variation and is difficult to describe accurately owing to marked changes in color depending on the angle of observation. Few specific differences have been noted in the male genitalia. To date no reliable characters have been found to separate any of the species of a group in the larval and pupal stages….”, which is obviously due to the fact that they have not been studied in comparative detail. It is worth noting that Belkin et al. provisionally applied the name superbus to the species in Cuba. Likewise, Belkin and his colleagues applied the name superbus to specimens reared from collections made in Colombia ( Heinemann & Belkin 1978c); Costa Rica ( Heinemann & Belkin 1977a); Guatemala, Honduras and Nicaragua ( Heinemann & Belkin 1977b); Mexico ( Heinemann & Belkin 1977c); Panama ( Heinemann & Belkin 1978a); Trinidad (the type locality of superbus ) ( Heinemann et al. 1980); and Venezuela ( Heinemann & Belkin 1978b; Navarro 1996). Those researchers identified haemorrhoidalis sensu stricto in collections made in northern Brazil and Ecuador (Heinemann & Belkin 1979); Colombia ( Heinemann & Belkin 1978c); and also French Guiana, Guyana and Surinam ( Heinemann & Belkin 1978b). It is certain that haemorrhoidalis sensu stricto and superbus are sympatric in at least Venezuela ( Navarro et al. 2007). It is interesting to note that the collections made in Colombia, French Guiana, Guyana and Surinam also included a species identified as sp. D, near superbus .

From the foregoing, it should be evident that the three subspecific forms may prove to be morphologically distinct, particularly in the larval stage; there is evidence that the distributions of haemorrhoidalis sensu stricto and superbus overlap in northern countries of South America; and available information indicates that separatus may be geographically separated from haemorrhoidalis sensu stricto and is restricted to areas southward of approximately latitude 20° south. In view of these indicators, we believe it is likely that haemorrhoidalis , separatus and superbus are separate species; thus, we here formally return separatus and superbus to their original specific status: Toxorhynchites (Lynchiella) separatus ( Lynch Arribálzaga, 1891b) and Toxorhynchites (Lynchiella) superbus ( Dyar & Knab, 1906a). We anticipate that molecular data will confirm these two forms and haemorrhoidalis are three separate species. Toxorhynchites separatus and Tx. superbus are currently listed as species in the Encyclopedia of Life. Based on available morphological data and having type localities in Argentina, Megarhinus lynchi Dyar & Knab, 1906a is retained as a synonym of Tx. separatus ( Lynch Arribálzaga, 1891b) .