Aedes (Ochlerotatus) canadensis (Theobald)

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 : 28-30

Aedes (Ochlerotatus) canadensis (Theobald)

subspecies canadensis ( Theobald, 1901c) —original combination: Culex canadensis . Distribution: Canada, Dominican Republic, Mexico, United States (Alaska and continental) ( Wilkerson et al. 2021).

subspecies mathesoni Middlekauff, 1944 —original combination: Aedes (Ochlerotatus) mathesoni (subspecific status by Rings & Hill 1948). Distribution: Southeast United States (Alabama, Florida, Georgia, South Carolina) ( Rings & Hill 1948).

The nominotypical subspecies was described from specimens collected near Toronto, Canada. Theobald (1901c) stated: “A very distinct species, in which the legs are very characteristically marked, the last tarsal joint [tarsomere 5] of the hind legs being entirely dull white and the banding of the legs involving both sides of the joints. The dusky scaled abdomen, with the creamy-white basal lateral patches, is also very characteristic.” Additionally we note: Maxillary palpus with apical half of palpomere 2 pale; scutum dull purplish brown to deep chestnut brown, covered with varying patterns of curved golden-brown scales; wing all dark; and venter of abdomen covered with pale yellowish scales. In the Natural History Museum, London, Townsend (1990) found: “ Lectotype male, designated by Belkin, 1968b: 4 ‒ Canada: Ontario, near Toronto, Trout Creek. Paralectotypes (14) ‒ Canada: 1 male, 13 female.” “Belkin misread the type locality as ‘Front Creek’, and overlooked some of the paralectotypes.” Other pertinent descriptions include: Adult female, larva and male genitalia ( Carpenter & LaCasse 1955); larva and male genitalia ( Ross 1947); pupa ( Darsie 1951); and egg ( Craig 1956; Craig & Horsfall 1958; Ross & Horsfall 1965; Kalpage & Brust 1968; Horsfall et al. 1970). Notably, canadensis is cold adapted, emerging very early in the spring (reviewed by Horsfall 1972).

Notable characters of the larva include: Setae 1-A and 5–7-C multi-branched, 5–7-C not in a straight line, 6-C inserted far forward of 5-C; seta 1-M small and short, not reaching bases of prothoracic setae; comb with many scales in a patch, individual scales pointed and fringed with rather slender subequal spinules; siphon index 3.0–4.0, pecten with 13–24 evenly spaced spines on basal two-fifths of siphon, siphonal tuft (seta 1-S) usually 3–8-branched, inserted beyond pecten; saddle incomplete, extending about 0.67 down side of segment, ventral brush (seta 4-X) large, usually with 2 precratal setae; anal papillae tapered, about 1.5 times as long as saddle ( Carpenter & LaCasse 1955; Harrison et al. 2016).

Subspecies mathesoni was described, as a species, by Middlekauff (1944) from specimens collected in south-central Florida (Kissimmee) in late summer and autumn [early season occurrence not noted]. Middlekauff recognized mathesoni as quite similar to canadensis : “This species [ mathesoni ] is most closely related to Aedes canadensis (Theob.) , but that species [ canadensis ] lacks the nearly black scales of the mesonotum and the white areas on the legs are more extensive. The male genitalia of canadensis differ only slightly, the spines of the ninth tergites [terga] being less regularly placed and more abundant, and the tenth sternites [paraprocts] being somewhat curved and acutely [sic] apically.” “ Holotype, allotype and one paratype [are] in the collection of the United States National Museum.”

Rings & Hill (1946) described the larva of mathesoni using reared-associated specimens from Georgia (Camp Gordon) and from northeastern Florida (Camp Blanding). They found the larvae of canadensis and mathesoni “to be remarkably similar, the only apparent difference being in the degree of branching in the head hairs [setae] and antennal tufts [seta 1-A].” In general, mean branch counts for setae 5- and 6-C were lower in specimens from further north ( canadensis ) than southern specimens. They counted setal branches of 89 canadensis in places ranging from British Columbia to Florida and 33 mathesoni from Georgia and Florida: Seta 5-C, mean number of branches 6.77 (4–10) ( canadensis ), 9.8 (6–14) ( mathesoni ); seta 6-C 5.06 (range not given) ( canadensis ), 7.1 (4–9) ( mathesoni ). Rings & Hill compared specimens of mathesoni from Georgia and Florida and found the same tendency: “... individuals collected in the more southern portions of the geographical range showed relatively more branching than those from the north.” Seta 5-C, mean 8.9 ( Georgia), 10.7 (Florida); seta 6-C, 7.0 ( Georgia), 7.2 (Florida). We think these numbers are suggestive of clinal variation. Rings & Hill concluded: “...there is a very close phylogenetic relationship between the two forms and that these forms do not possess differences of specific rank.”

Rings & Hill (1948) documented and illustrated the scutal patterns of canadensis and mathesoni and noted that there were many “intergrades”. They stated: “The data here presented are interpreted by the authors as evidence that mathesoni is a melanistic, geographical variation of A. canadensis .” They illustrated intergrades showing the extent of pale scaling on the hindtarsomeres and mapped the geographical extent of intergradation and showed a broad area of intergrades. Those north and south of this zone were easily identified as either canadensis or mathesoni , respectively. Darker coloration seen on the legs was explained as directly related to higher temperatures at decreasing latitudes. They found little or no differences (intergrades) in the larvae or male genitalia of the two forms. The notion that mathesoni was a subspecies of canadensis was introduced in this paper; however, their unattributed definition of subspecies allows for sympatry (which for us precludes subspecific status). They concluded: “The phylogenetic rank of subspecies is generally applied to a part of a species showing geographical variations which intergrade with a subspecies occupying different though usually adjacent and overlapping parts of the general range of the species.” They, however, continued to use the subspecies rank as a convention for pointing out observed variability.

Bickley (1981) furthered the observations of Rings & Hill (1946, 1948) in an analysis of adults and larvae of more than 1,400 specimens in the United States National Museum (Smithsonian Institution, National Museum of Natural History (USNM), Washington, D.C.). He concluded: “From the survey described here, involving specimens from 43 North American states and provinces, it may be concluded that the geographical range of Ae. canadensis mathesoni does not extend northward from southern Georgia and South Carolina. Only 1 character, the narrow pale rings on the hindtarsomeres, is available as a means of recognizing this rare subspecies, the validity of which is certainly complicated by the existence of intergrades.”

Harrison et al. (2016) also used the pale bands of the hindtarsomeres to separate canadensis canadensis from canadensis mathesoni , but they could not separate them in the larval stage. They stated (Note 14): “The two subspecies of Ae. canadensis are separated in the adult female key, but are not separated in the larval key. Those subspecies, Ae. canadensis canadensis and Ae. canadensis mathesoni represent a real taxonomic enigma that begs to be resolved with more study and the use of modern techniques.” An explanation for this enigma may lie in the abundant examples of mosquito morphological and physiological variation influenced by the environment, for example: Factors affecting larval setal branching ( Colless 1956; Mattingly 1975); altitude and latitude influencing diapause and the size of anal papillae in Wyeomyia smithii ( Coquillett, 1901) ( Bradshaw & Lounibos 1977) ; season and temperature affecting melanism in species of the genus Anopheles ( Harrison 1980) ; and the effect of temperature on morphological attributes of Anopheles merus (D̂nitz, 1902) ( Le Sueur & Sharp 1991).

Ongoing sampling of mosquitoes at several locations in North America is being carried out by the National Ecological Observatory Network (NEON). Their sampling is based on light trapping of adult females, which are then identified morphologically. A subsample is kept as morphological and DNA vouchers, with single legs used for sequencing of the barcoding region of the mtDNA COI gene. The NEON database (https://biorepo.neonscience. org/portal/collections/list.php?usethes=1&taxa=80835) has 571 records for canadensis canadensis (also written simply as canadensis ) and canadensis mathesoni . The database includes a limited number of sympatric records of nominotypical canadensis with canadensis mathesoni in the area of historical “intergrades,” i.e. Florida Disney Wilderness Preserve near Kissimmee and near Gainesville, Florida, which is relatively close to Camp Blanding. The Barcode of Life Data System (BOLD: boldsystems.org) includes a single entry for canadensis mathesoni (http://www.boldsystems.org/index.php/Taxbrowser_Taxonpage?taxon= Aedes %20canadensis) that corresponds to a specimen collected by NEON (specimen number MOSN4427-20), which allowed us to visualize, using tools available in BOLD, its genetic distance from more than 300 other nominal canadensis in the database. This specimen clustered very closely with all others at a distance of <1.0%, well within accepted COI barcode criteria for an individual species. This is a single data point, from a single gene, from unpublished data, but it suggests that further focused collecting and sequencing will probably reveal that canadensis and mathesoni are conspecific, at least using the COI gene.

We do not think there is any indication of genetic divergence, only easily explicable clinal variation. Plus, the two nominal taxa are apparently sympatric and, although weak, the single molecular indicator shows they are genetically very similar. We therefore think that subspecies mathesoni is a synonym of Aedes (Ochlerotatus) canadensis , which we formally assert here: mathesoni Middlekauff, 1944 , junior subjective synonym of Aedes (Ochlerotatus) canadensis ( Theobald, 1901c) . Future work may prove we have not interpreted the literature correctly, but until then mathesoni belongs as a synonym. An alternate hypothesis, that there is a northern species and a southern species with a zone of hybridization, is valid. We do not see signs of that, but it should not be totally discounted. The nominal subspecies mathesoni , which is listed as a species in the Encyclopedia of Life, must be removed from the list of valid species of Aedes .

There is one other synonym of Ae. canadensis , Culex nivitarsis Coquillett, 1904 , synonymized by Howard et al. (1917), who wrote: “Coquillett described Culex nivitarsis from two specimens and none have been since collected. Dr. C. S. Ludlow suggested to us that these specimens were only aberrations of A. canadensis , and we have adopted this view.”