A new species of Andrena (Trachandrena) from the Southwestern United States (Hymenoptera, Andrenidae)

A new species of Andrena Fabricius, 1775, subgenus Trachandrena Robertson, 1902 is described and illustrated, A. hadfieldi sp. nov., from Arizona, United States. The new species, presently known only from the female holotype, was collected in a Malaise trap in 1994, and remained unstudied until recently. In addition, Trachandrena is compared to similar subgenera in North America to assist in recognizing new members.


Introduction
Andrena Fabricius, 1775 is one of the largest genera of bees, with 1,556 species (Ascher and Pickering 2020). Dubitzky et al. (2010) estimated that there are likely ca 2,000 species, suggesting there are many undescribed species, especially in Mesoamerica and in the dry regions of Central Asia. Though the genus is mainly Holarctic, it extends into Mesoamerica, parts of Africa and tropical Asia (Michener 2007).
The subgenus Trachandrena Robertson, 1902 is represented by 30 species globally (Gusenleitner and Schwarz 2002;Michener 2007), 24 of which occur in the Since LaBerge's (1973) revision of the subgenus Trachandrena in the Nearctic region no additional North American species have been recognized. Here a new species of Trachandrena is described from Arizona, United States. A diagnosis and full description of the female is provided. In addition, a partial key, modified from that of LaBerge (1973) is provided to allow females of the new species to be recognized from other species.

Materials and methods
For consistency with species treatments published elsewhere, the description generally follows the format used by LaBerge (1973). Other terminology and measurement methods follow that of Michener (2007); body length was measured as the sum of the length from the antennal base to the posterior propodeal surface and the length of the metasoma in lateral view. The following abbreviations are used: F = flagellomere, numbered from base to apex; pd = puncture diameter; T = metasomal tergum, and S = metasomal sternum, both numbered from the base to apex.
Photomicrography was undertaken with a Canon EOS 5D Mark II digital camera with an MP-E 65 mm 1:2.8 1-5× macro lens. Measurements were made with an ocular micrometer on a Nikon SMZ1000 stereomicroscope.
Diagnosis. Trachandrena, particularly the females, are relatively easy to recognize among most other subgenera of Andrena in the Nearctic region based on the combina-tion of the coarsely rugose metapostnotum (i.e., propodeal triangle) (Fig. 1) and the generally coarse body sculpturing (being less strongly sculptured in most other subgenera), the structure of the metasomal terga, especially T2, which have broad apical marginal zones that usually extend ½ the median length of the tergum (Fig. 2B) or more ( Fig. 2A, C, D) (other subgenera in the Nearctic have narrower marginal zones), and the characteristic structure of the facial fovea, which is typically much narrower in the lower half ( Fig. 3) (more parallel-sided for entire length in other Nearctic subgenera). As indicated above, in the Nearctic region Trachandrena is most similar to Plastandrena and Scrapteropsis. In addition to both of these subgenera having terga with much narrower marginal zones (Fig. 4), both sexes of Plastandrena usually have weakly to strongly curved inner hind tibial spurs (Fig. 5A) whereas these are straight in Trachandrena (Fig. 5B), and the facial fovea of females of both subgenera are typically broad throughout. Males of Scrapteropsis are difficult to distinguish from Trachandrena (Viereck 1924), though each has unique genital capsules (LaBerge 1971(LaBerge , 1973, the marginal zone of T2 is longer in Trachandrena, the antenna usually being slightly longer. Male Trachandrena also have S6 usually flat, not with a reflexed apical margin or with apicolateral teeth as in some Plastandrena (i.e., A. crataegi;LaBerge 1969;Michener 2007). In addition, most Plastandrena in North America, excluding A. crataegi, have the clypeus yellow or otherwise maculated (LaBerge 1969), not black as in Trachandrena.  Diagnosis. The female of Andrena hadfieldi is unique among Trachandrena in the Nearctic region in having very wide (i.e., at least 2/3 of the median tergal length) marginal zones of T2-T4 which are shiny and impunctate (Fig. 2D), a feature shared only with A. cleodora (Viereck) (Fig. 2C). Andrena hadfieldi is smaller than A. cleodora (9 mm, versus 10-13 mm body length in A. cleodora), and differs from A. cleodora in hav- ing the terga black ( Fig. 2D) instead of black with strong metallic bluish reflections (Fig. 2C), and in having entirely pale pubescence, including the scopa (Figs 6, 7); the hair on the metasoma (Fig. 2C) and scopa of A. cleodora are black, and the pubescence of the dorsum of thorax is yellowish to red (subspecies cleodora; widespread in western North America) or entirely black (subspecies melanodora Cockerell; known from southern California). The structure of the pubescence on the dorsum of the thorax also differs between these two species, being long, very thin, and weakly plumose in A. hadfieldi (Fig. 6), but shorter and densely plumose, almost scale-like, in A. cleodora (Fig. 8). The process of the labrum in A. hadfieldi is more than 3× as wide basally as long medially (Fig. 9A); in A. cleodora the labral process is larger, subtriangular, with the base 2.5× as wide as the medial length (Fig. 9B). The body surface sculpture of A. hadfieldi is much finer than for A. cleodora; as examples, the face of A. hadfieldi is generally more finely and sparsely punctate, with shiny interspaces > two pd on the lower paraocular area (Fig.  10A), while in A. cleodora the lower paraocular area is more coarsely and closely punctate (interspaces < pd) (Fig. 10B); the surface of the propodeal corbicula is smooth with a few short rugae in A. hadfieldi (Fig. 11A), while coarsely rugose in A. cleodora (Fig. 11B).
Structure. Labrum with process trapezoidal, more than three times as wide at base as long medially, apical edge entire (Fig. 9A). Clypeus with coarse, close round to irregular shaped punctures, becoming finer apically, interspaces shiny and linear, less than 0.5 pd, without obvious median impunctate line but with a small shiny subapical boss extending for less than 1/5th median length of clypeus (Fig. 10A). Supraclypeal area with distinct round punctures separated by 0.5 pd, surface rather shiny (Fig. 10A). Mandible short, extending beyond middle of labrum by about ¼ its length in repose. Malar space extremely short (Fig. 10A). Lower paraocular area shiny with small punctures separated by > two pd (Fig. 10A). Face above antennal socket with rugulae extending to ocelli, without obvious punctures. Facial fovea long, extending from middle level of lateral ocellus to basal edge of clypeus; lower portion narrow, from below level of antennal socket about 1/3 as wide as upper portion, outer edge slightly incurved from inner margin of compound eye just above level of antennal socket, this area smooth, shiny and impunctate (Fig. 10A). Compound eye just over three times as long as broad in frontal view, inner margin converging slightly toward mandibles. Genal area in profile about as broad as compound eye, surface shiny with minute punctures separated by 2 pd, posterior half dull with reticulate shagreening, without apparent punctures except near base of mandible. Vertexal area above lateral ocellus subequal to one ocellar diameter, dulled by crowded punctures and dense reticulate shagreening. Antennal scape length equal to combined length of F1-F3; F1 about 1.5 times as long as broad at apex, and 1.5 times longer than F2; F2-F5 quadrate, F6-F9 about 1.2 times longer than broad, F10 more elongate, about 1.5 times longer than broad.
Pronotum somewhat shiny, with distinct punctures dorsally, separated by about two pd, laterally mostly impunctate with surface somewhat dull. Mesoscutum with large, round deep punctures, between parapsidal lines and posteromedially separated mostly by ½ to one pd, anteriorly and laterally separated by less than ½ pd, becoming somewhat rugosopunctate along anterior edge (Fig. 7), surface dull, reticularly shagreened. Scutellum similarly punctured though punctures slightly sparser anteriorly, and surface mostly shiny. Metanotum dull and tessellate laterally, becoming somewhat shiny and punctate medially. Metapostnotum with rather shallow but distinct rugae, these somewhat irregular (Fig. 12); dorsolateral and posterior surfaces of propodeum moderately coarsely rugosopunctate, tessellate, dull; propodeal corbicular surface moderately shiny, tessellate, with a few short rugae (Fig. 11A). Mesepisternum coarsely rugose and somewhat shiny. Metepisternum surface smooth, with a slight shine. Fore femur with base round in outline. Posterior hind tibial spur straight. Tarsal claws with a small subbasal tooth. T1 shiny and largely impunctate, with basal area (= disc) punctures obscure, shallow, sparse, separated by ≥ 5pd; marginal zone impunctate and shiny, occupying about 3/5 th of median length of tergum, a few sparse minute punctures visible at extreme lateral edge, surface smooth (Fig. 2D). T2 with marginal zone clearly longer than basal area (about 3/4 th medial with); basal area shiny with punctures separated by one pd, a   narrow dull impunctate area adjacent to gradulus; marginal zone smooth, shiny and impunctate (Fig. 2D). T3-T4 similar to T2 but marginal zone of T4 shorter, about 3/5th of median length of tergum (Fig. 2D). T5 with basal area broader than marginal zone, basal area tessellate, with distinct punctures separated by 2-3 pd, becoming closer adjacent to marginal zone (Fig. 2D); marginal zone dull, largely obscured by prepygidial fringe (Fig. 2D). Pygidial plate U-shaped with rounded apex about ½ as wide as base, with sharply pointed, internal, raised triangular area on median surface. S2-S5 with surface somewhat shiny, punctures uniformly dense, separated by 1 pd; S2 with a medial U-shaped depression.
Male unknown. Distribution. United States, southern Arizona. Etymology. It is a privilege to name this new species after Canadian astronaut Col. Chris Hadfield for his many achievements as a pilot, astronaut, author, lecturer, and science educator.
Partial key to species of Trachandrena (modified from LaBerge 1973)

Discussion
Nothing is known about the biology of this species as it is known only from the holotype female which was collected in a Malaise trap. However, like other Trachandrena in the Nearctic region, this is a vernal species that probably visits spring flowering trees and shrubs (LaBerge 1973). It is hoped that by publishing a full description of the new species and a partial key to the Trachandrena in North America that includes this species, that additional specimens, including the male, will be discovered in other entomology collections or through survey work. The American Southwest is one of the most diverse regions for bees globally (Michener 1979(Michener , 2007, so it is likely that many other bee species await recognition.