New records of Orussus minutus Middlekauff, 1983 (Hymenoptera: Orussidae) represent a significant western range expansion

Abstract Background Orussus minutus is an uncommonly collected parasitoid sawfly known from the eastern United States. New information We report specimens Orussus minutus Middlekauff, 1983, from Arkansas, Iowa, Minnesota, and Manitoba, which represent new state and province records and significantly expand the known range of the species west from previous records; provide collection information for unpublished specimens housed in the United States National Museum collection, which includes new state records for West Virginia and Michigan; and report two specimens housed in the Biological Museum at Lund University that represent new state records for Connecticut.


Introduction
Orussidae have long interested entomologists because of their parasitoid larvae, which are unique among non-apocritan Hymenoptera, phylogenetically important position between basal Hymenoptera ("Symphyta") and Apocrita, and because they are rarely collected (Middlekauff 1983, Pesarini and Turrisi 2003, Vilhelmsen 2003. Middlekauff (1983) provided an excellent review of the literature concerning the feeding biology and hosts of orussid larvae. Briefly summarized, a number of authors reported orussid larvae develop in wood (Harrington 1887a, Konow 1902, Gaulle 1906 and associate with beetle and sawfly larvae (Wachtl 1882, Rudow 1909. Harrington (1887b) first hypothesized that orussid larvae may be parasitoids, though he considered it more likely they fed on wood. Rohwer (1912) and Burke (1918) provided convincing evidence that orussids are parasitoids as they reported Orussus larvae pupating in old cerambycid larval galleries and attacking buprestid larvae. Subsequent authors investigated oviposition behavior and larval feeding; they found that adult female orussids deposit eggs into frassfilled galleries of and directly onto larvae of wood-boring Coleoptera and Hymenoptera and that larval orussids feed upon those larvae (Cooper 1953, Rawlings 1957, Powell and Turner 1975. Currently, Orussidae are known or suspected to parasitize Buprestidae, Cerambycidae, Siricidae, and Xiphydriidae (Table 1). relationships within Orussidae, the most robust phylogenetic analysis was produced by Vilhelmsen (2003). His analysis recovered most genera as monophyletic, though Vilhelmsen abandoned the use of subfamilies and tribes, as "[e]nforcing a strictly cladistics classification at these levels would require recognition of many redundant taxa without enhancing the information content".
Orussidae are uncommonly collected and rare in collections. For example, despite a cumulative 25,000 trapping hours (314 separate 1-2 week collection events) using Malaise traps over the last five years by the authors around Arkansas, no additional specimens beyond the three reported herein were captured with this trapping method and David Smith (USDA, SEL), who has had success collecting orussids in Malaise traps (e.g., Smith 2006, Smith 2008, Barrows and Smith 2014, has only collected 33 specimens of O. minutus in 35 years of collecting with an average of 15 Malaise traps set per year (David R. Smith, pers. comm. 18 August 2015). Additionally, new species continue to be described, even in heavily collected areas such as California (e.g., Vilhelmsen 2005, Blank et al. 2010, Vilhelmsen et al. 2014. Several species are known only from one or a few localities and specimens and the known ranges of many species continue to expand as new specimens are collected (Ahnlund and Ronquist 2001, Vilhelmsen and Smith 2002, Pesarini and Turrisi 2003, Pesarini and Turrisi 2006, Choi and Suh 2011. Orussus is represented five species in North America north of Mexico: O. occidentalis (Cresson, 1879) has been reported from Southern British Columbia east to Ontario, south in the western United States to southern California, Nevada, and New Mexico; O. thoracicus (Ashmead, 1898) has been reported from Colorado, Washington, Oregon, and California; O. sayii (Westwood, 1835) has been reported from Ontario south to Louisiana, west to Indiana; O. terminalis (Newman, 1838) has been reported from New England and Ontario west to Iowa and Illinois, south to Maryland; and O. minutus (Middlekauff, 1983) has been reported from New York to Georgia west to Illinois (Middlekauff 1983, Vilhelmsen 2003, Blank et al. 2010, Vilhelmsen et al. 2013).

Materials and methods
Two orussids (1 male, 1 female) were collected along the Buffalo National River in the lower collector of an aerial SLAM (sea-land-air-Malaise) trap (MegaView Science Co., Ltd., Taichung, Taiwan) and a black multifunnel trap (ChemTica International, S.A., Heredia, Costa Rica); a third specimen (1 female) was collected via aerial netting in the Kessler Mountain Reserve. Both localities are mixed secondary deciduous forest dominated by oak and hickory that were logged approximately 80-100 years ago. Specimens were identified to species using published keys (Middlekauff 1983, Vilhelmsen et al. 2014) and have been deposited in the University of Arkansas Arthropod Museum.
Stereomicrographs of the Arkansas specimens were taken with a Cannon EOS 40D camera (Tokyo, Japan) attached using a Diagnostic Instruments DD20NLT 2.0X camera mount (Sterling Heights, Michigan, USA) to a Nikon SMZ1500 stereomicroscope (Tokyo, Japan). The microgrpahs were processed and final plates arranged in Adobe Illustrator (San Jose, California, USA).
David R. Smith kindly provided label information for specimens housed in the United States National Museum; previously unpublished specimens are reported herein. Additional unpublished specimens were found by searching the databased collection of Lund University Biological Museum (Lund University 2015), BugGuide (Hatfield 2008, Alexander 2011, Liberta 2014, Zhang 2014, and Flickr (King 2014).

Distribution
New York south to Georgia west to Manitoba, Iowa, and Arkansas.

Discussion
The Arkansas specimens and those shared as photographs on Bugguide and Flickr significantly expand the known range of O. minutus westward (Fig. 3). Morphological determination of the Arkansas specimens was confirmed by genetic data and the species is easily identified due to its small size and distinct markings, so it is highly unlikely the photographed specimens are not O. minutus.
Many of the USNM specimens were collected by David R. Smith during 35 years of Malaise trapping specifically for sawflies. However, most recently collected specimens, especially those from Pennsylvania, were found as non-target species during various exotic species monitoring programs that utalized Lindgren multifunnel traps (David Smith, pers. comm., 28 Aug. 2015). The abundance of these specimens emphasize the utility of examining, or at least collecting and sending to the appropriate specialist, non-target species in mass trapping surveys, such as was suggested by Skvarla and Holland (2011). Precise figures for the number of traps and amount of effort that was involved in the Pennsylvania surveys is unavailable, so we are unable to compare the efficiency of Malaise trapping compared to Lindgren funnel trapping; however, the number of O. minutus that were collected in Lindgren funnel traps suggests that it may be a useful tool for collecting Orussus. Given the current records, O. minutus is likely present throughout most of Eastern North America. The concentration of specimens from northern Virginia and Pennsylvania reflect collecting effort and specimen recoginition rather than true abundance and further collecting in the southeastern United States and Canada should produce additional specimens from those areas.
Finally, records found through Bugguide and Flickr join a growing list of discoveries made via citizen science and social media websites (e.g., Otto and Hill 2011, Winterton et al. 2012, Gonella et al. 2015 and help underscore the importance of such resources in descriptive biology and natural history.