Bembidium erasum LeConte, 1859: 83

Maddison, David R. & Sproul, John S., 2020, Species delimitation, classical taxonomy and genome skimming: a review of the ground beetle genus Lionepha (Coleoptera: Carabidae), Zoological Journal of the Linnean Society 189, pp. 1313-1358 : 1340-1344

Bembidium erasum LeConte, 1859: 83 . Lectotype

♀, designated by Erwin & Kavanaugh (1981: 59), in MCZ (type # 5490), examined, including genitalia and DNA. Type locality: ‘Oregon’ (original citation), restricted to Fort Klamath, Klamath County, Oregon, USA, by Erwin & Kavanaugh (1981: 59). Although Fort Klamath is a reasonable type locality for Erwin & Kavanaugh’s concept of Lionepha erasa (i.e. the species here called Lionepha probata ), the species to which the types of Bembidium erasum belong is not known from the southern half of Oregon. In order to move the type locality within the range of the current species, we here change the type locality to Marys Peak, Oregon (44.5104°N 123.5593°W); on this mountain, Lionepha erasa is the only known species of Lionepha with isodiametric microsculpture. GenBank accession numbers for DNA sequences of the lectotype are MN401767 , MN401952 , MN402005 , MN402129 and MN402322 ; accession number of sequence reads in NCBI’s Sequence Read Archive is SRR5230423.

Bembidion lindrothellus Erwin & Kavanaugh, 1981: 61 . Holotype ♂ in MCZ (type # 32549), examined, including genitalia and DNA. Type locality: Haines Highway Mile 31.5, Little Boulder Creek , Alaska, USA. New synonymy. GenBank accession numbers for DNA sequences of the lectotype are MN401768, MN401784, MN401905, MN401953, MN402006, MN402130, MN402255 and MN402323 ; accession numbers of sequence reads in NCBI’s Sequence Read Archive are SRR5230400 and SRR5230401.

Bembidion lummi Erwin & Kavanaugh, 1981: 62 . Holotype ♀, in CAS (type # 13652), examined. Type locality: Friday Harbour, San Juan Island, San Juan County, Washington, USA. Synonymy with L. chintimini established by Kanda et al. (2015).

Bembidion chintimini Erwin & Kavanaugh, 1981: 63 . Holotype ♀, in CNC (type # 16452), examined. Type locality: Marys Peak , 1220 m, Benton County, Oregon, USA. New synonymy.

Nomenclatural notes: Among the non-type specimens we examined, all small Lionepha with isodiametric or effaced microsculpture from Oregon, Washington, Idaho, British Columbia (BC) and Alaska (the area containing the type localities of the four names listed above) belong to two species: a common, widespread species known from southern BC south to California, east to Montana and Colorado, but not known from Alaska or the Oregon Coast Range; and a rarer species known from the Oregon Coast Range, the Cascades of central and northern Oregon , north through Washington, BC, to near Anchorage , Alaska .

The realization that the four primary types of the abovelisted names belong to only one of these species (the latter, rarer species) took several years, and was delayed by the fact that three of the primary types are females, one of which is teneral (the lectotype of Bembidium erasum ), and the fourth type (holotype of Bembidion lindrothellus ) is a teneral male with unpigmented, flattened genitalia.

Investigation of the rarer species, the one here called Lionepha erasa , began in 2010. Dissection of the first recognized males from Marys Peak, Oregon (type locality of Bembidion chintimini ) revealed an aedeagus indistinguishable from those from San Juan Island, Washington (type locality of Bembidion lummi ). The female holotype of B. chintimini is wingless and has slightly rounded shoulders. However, the Marys Peak population is wing-dimorphic, and winged individuals are in body form no different from the type series of Bembidion lummi . The elytral microsculpture of the holotype of B. chintimini is perfectly isodiametric (against Erwin & Kavanaugh, 1981), thus matching that of B. lummi . Other characters mentioned by Erwin & Kavanaugh as distinguishing the two populations are not consistent with available specimens. The lack of evident morphological differences, combined with effectively identical DNA sequences in specimens from Oregon, British Columbia and Alaska suggested that the Marys Peak populations are the same species as populations further north, and for this reason, Bembidion chintimini and B. lummi were synonymized by Maddison in Kanda et al. (2015).

This left in question the specimens considered to be Bembidion lindrothellus by Erwin & Kavanaugh, which are at first glance similar to the Marys Peak and other populations of ‘ Bembidion chintimini ’. Specimens classified as Bembidion lindrothellus are reported to be paler, but all specimens mentioned in Erwin & Kavanaugh (1981) are teneral. The unsclerotized aedeagus of the holotype of Bembidion lindrothellus made comparison of internal sac sclerites difficult. However, the internal sac membrane that rests in the left-most position has a species-specific microsculpture in Lionepha , and the microsculpture scales of the holotype of Bembidion lindrothellus from Alaska match those of Marys Peak specimens. A non-teneral male was also collected by Lindroth at the type locality of B. lindrothellus , but was not included in the type series, perhaps as the specimen was housed in Lindroth’s collection in Lund, Sweden.This specimen is presumably the one whose genitalia Lindroth figured as Bembidion brumale (1963: fig. 127f). We have examined that specimen, and it is indistinguishable from specimens of ‘ Bembidion chintimini ’ from Alaska, British Columbia, Washington and Oregon, including details of the internal sac. Most critically, DNA sequences of the holotype of Bembidion lindrothellus are identical in eight studied genes to those of other specimens from throughout the range ( Figs 5–7). It is thus evident that the holotypes of Bembidion chintimini , B. lindrothellus and B. lummi belong to a single species.

However, there is an older name. The type series of Bembidium erasum consists of four females. These specimens have traditionally been considered to belong to the common, widespread species here called Lionepha probata . Females of these two isodiametrically microsculptured species are difficult to tell apart, especially those with less-extreme prothoracic proportions (neither wide nor narrow). Although there are distinctions in the lobe of the female bursa of fully sclerotized individuals, interpretation of tenerals is more tenuous. Specimens in the type series of Bembidium erasum are all teneral, with prothoraces of moderate width, and thus there is no clear morphological evidence to place them to species. The type series was provided by George Suckley ( LeConte, 1859), presumably captured during his travels as naturalist for the governor of Washington Territory during 1853–57 ( Cooper & Suckley, 1859). The type series is from ‘Oregon’, which at the time encompassed the current area of Oregon, the southern half of what is now Idaho and some parts of Wyoming and Montana ( Barry, 1932). Suckley’s travels in Oregon included areas within the range of both species ( Cooper & Suckley, 1859), and thus geography provides no clues about species membership. However, DNA data from the lectotype (and two of the paralectotypes; Sproul & Maddison, 2017) makes it clear that these specimens belong to the current species ( Figs 5–7; Supporting Information, Fig. S1). Thus, the valid name of this species is Lionepha erasa , with Bembidion chintimini , B. lindrothellus and B. lummi as junior synonyms.

Diagnosis: A small, convex species ( Fig. 1A) with isodiametric elytral microsculpture ( Fig. 15A, B; effaced in most males) and pale tibiae contrasting against the darker body. Prothorax moderately wide [PW/EL 0.467 –0.478, average 0.473 (N = 6),> 0.47 in most specimens]. Aedeagus with straight ventral surface, with apex not expanded ( Fig. 11A, B). Internal sac of aedeagus with distinct but small nub; sclerite CH 1 as in Figure 14A. Female bursa with relatively small dorsal microtrichial patch ( Fig. 9B).

Most similar to L. australerasa , but also difficult to distinguish externally from L. probata . Compared to both L. australerasa and L. probata , L. erasa is paler, with the tibiae distinctly paler than the body; L. erasa also tends to have a more convex body than in those species. In contrast to L. australerasa , sculpticells of the elytra are more perfectly isodiametric, although in a few specimens they are slightly transverse; the lateral margins of the pronotum of L. erasa are more rounded, with the sides distinctly converging posteriorly, and the hind angle always quite obtuse; the striae of elytra are slightly weaker, with smaller, less distinct punctures, often with only the three inner striae being clearly evident. Compared to L. probata , L. erasa has a narrower prothorax.

Additional characteristics: Body length 3.47– 4.41 mm, although rarely above 4.2 mm. Antenna piceous. Legs with femora rufopiceous, and tibiae pale rufous or dark testaceous, except in specimens from the Cascades, whose tibiae can be rufopiceous. Hind wings dimorphic, with some individuals fully winged (e.g. DNA2615) and others brachypterous with narrow and sloped shoulders (e.g. DNA2616).

Geographic variation: Some individuals in populations from inland British Columbia have slightly transverse microsculpture (e.g. 15 km E New Denver, Zincton Summit , BC; MZLU) and are larger than typical for L. erasa . Specimens from the Cascades are darker, with darker tibiae, than specimens from the Oregon Coast Range .

The only notable variation in DNA sequences observed is in 28S: the two Alaskan specimens (DNA 4059 and the holotype of Bembidion lindrothellus ), as well as the lectotype of Bembidion erasum , are missing four bases (‘ATTA’) present in other specimens; this missing section occurs just after base 613 in the sequences of the holotype of Bembidion lindrothellus in GenBank (accession MN402323 ).

Note: This is the species referred to as Bembidion brumale in Lindroth (1963), but the type of Bembidion brumale is a member of Lionepha casta .

Distribution: Known from the Oregon Coast Range from Prairie Peak northward, as well as in the Cascades from Lost Prairie Campground, Oregon, northwards to Mt. St. Helens, at low elevation on the San Juan Islands and Vancouver Island, eastward to the western slopes of the Rocky Mountains and north to east of Anchorage, Alaska ( Fig. 20); not known from south of 44.2°N. Found from 15 to 1500 m elevation. Collected in all months of the year except January, with lower frequency in midsummer; most commonly collected in the fall and early spring. In the Oregon Coast Range, where these beetles are found in high-elevation grasslands (e.g. on Marys Peak and Mt. Hebo), they are only active above ground once the rains start in August or September, until the rains cease in late spring. In June and July they are not in evidence, even at night.

water, a specimen has also been found along the shores of a stream (DNA4144 from Mount Hood, Oregon).