Checklist of bees (Apoidea) from a private conservation property in west-central Montana

Abstract Background Here we present preliminary results from the first three years of a long-term bee survey conducted at a 3,840-ha private conservation property in the northern Sapphire Mountains and Bitterroot River Valley, and a pilot study at an associated 80-ha property in the Swan River Valley, Missoula County, Montana, USA. The survey includes hand-net, bowl-trap, and blue-vane trap collections. The resulting checklist comprises 229 bee species and morphospecies within 5 families, 38 genera and 91 subgenera. Of the total species in the list, 34 of them represent first state records Montana. This survey expands the number of bee species recorded in Montana to 366. Included in these species is Megachile (Eutricharaea) apicalis Spinola, showing a range expansion for this introduced bee. New information We present new distributional records for 34 bee species, including Megachile (Eutricharaea) apicalis Spinola, an introduced bee that was discovered to be resident in North America in 1984 in Santa Barbara County, California. This species has since expanded its range in the across the west, but had not been previously recorded in Montana.


Introduction
Wild bees play a vital role as pollinators in both agricultural and natural systems, and may be important to the success of habitat restoration projects (Menz et al. 2011, Williams 2011. Work has been done to document bee faunas of some western regions of North America, including Colorado (Scott et al. 2011), the Columbia Basin (Tepedino and Griswold 1995), north-central Washington (Wilson et al. 2010), and some US national parks and monuments (e.g. Messinger and Griswold 2002); but there are few large-scale bee surveys from Montana.
Reported here is the bee species list from the first three years of a long-term bee monitoring study we initiated in 2013 at MPG Ranch, a privately-owned conservation property in the northern Bitterroot Valley and Sapphire Mountains, and a pilot study from MPG North, an associated property in the Swan Valley. Both properties are in Missoula County of west-central Montana. A purpose of this study is to inventory the bee fauna on the properties.

Study Site
The 3,840 hectare MPG Ranch is a privately-owned conservation property in west-central Montana, in the Bitterroot River Valley and Sapphire Mountains of Missoula County (Fig.  1). Elevations range from 975m at the valley floor along the Bitterroot River, to 1860m on Mount Baldy, its highest point. The region has a mild, semi-arid climate with cool, moist winters and warm, dry summers. Mean precipitation ranges from 300mm at the valley floor to 350mm on mountain summits; most precipitation occurs in the winter as snow. Mean temperatures for the nearby town of Stevensville range from -3.6°C in December to 20.0°C in July.
For over a century prior to 2009, this property had been managed for livestock and agricultural crop production. During that time, most of the lower elevation grasslands were replaced with irrigated crops or introduced forage grasses. These areas are now out of production and undergoing restoration treatments to return them to a more natural state. Livestock has also been removed from the property as part of the restoration effort, and native ungulates (elk, mule deer, and whitetail deer) currently comprise the majority of vertebrate grazers. Habitats on the property include riparian bottomlands, dry open forests, mid-elevation sagebrush steppe and grasslands, montane grasslands, and montane mixed-conifer forest.
The MPG North, an associated 81 hectare conservation property separated from MPG Ranch by over 100 kilometers, is located in the Swan River Valley, also in Missoula County ( Fig. 1), at an elevation of 1200m. The climate in the Swan River Valley is cooler and wetter than the Bitterroot River Valley. At the nearby town of Seeley Lake, mean temperatures range from -6.1°C in December to 17.5°C in July, and mean annual preciptiation is 530mm. MPG North's habitat consists of mixed-conifer forest and clearings, and some small wetlands.

Collection Methods
We captured bees from MPG Ranch using the protocol outlined in A Standardized Method for Monitoring Bee Populations: The Bee Inventory (BI) Plot (LeBuhn et al. 2003), modified by reducing the amount of time spent netting from 60 minutes to 30 minutes. We used 24 plots that had been chosen for a larger, multi-organism surveying effort occurring on the property. These 24 sampling plots were placed along an elevational gradient, from approximately 975m elevation to 1850m elevation. At each location we laid out two 50m transects, intersecting at 25m, creating an "X". We deployed 21, 3.25oz Solo brand plastic cups (Solo brand stock number PB6-0099) filled with soapy water at each sampling plot and placed them about 5m apart along the transects. Bowls were spray-painted fluorescent blue, fluorescent yellow, or left white. Seven bowls of each of the three colors were used at each site. With the exception of the first sampling round, we placed traps out at all sampling plots on the same day by 9am and collected the contents after 3pm, providing a snapshot of the bee species composition for that day. We seived bowl-trapped bees in the field and transferred them into 4oz Whirl-Pak bags with 70% isopropanol. We kept these samples refrigerated until shipment to the USDA-ARS Pollinating Insects Research Unit in Logan, UT, for identification. A total of 30 minutes of netting occurred at each site on each collection day in 2013 and 2015, when time and field staff availability permitted. Netting took place on bowl-trapping days or the following day, and was limited to a 100m radius of plot center. Samplers netted bees throughout the plot and attempted to sample from as many different floral resources as possible within the time frame. Netted bees were dispatched with ethyl acetate and kept frozen in labeled collection tubes until processed.
We sampled every 2-4 weeks throughout as much of the field season as possible to capture as much of the bee fauna as we could. We sampled each of the plots five times in 2013 (June 10-17, July 1, July 22, August 12, and September 9), seven times in 2014 (May 1, May 15, June 12, July 10, July 31, August 28, and September 17), and nine times in 2015 (April 21, May 4, May 18, June 8, June 29, July 20, August 10, September 1, and September 23).
Collecting events at MPG North were part of a pilot study and consisted of a small number of blue vane trap collections and some hand-netting in 2014 and 2015. Bee species caught at MPG North are included in this species checklist.
Species accumulation estimates were determined by using EstimateS (Colwell 2013).
Bee specimens were pinned and labeled with location information, date, collection method, and collector name. Except for two synoptic sets of voucher specimens that are kept at MPG Ranch, specimens are deposited in the U.S. National Pollinating Insect Collection at Logan, Utah.
Lasioglossum bees in the subgenus Dialictus were identified only to subgenus due to the difficulty in distinguishing between Dialictus species and the lack of comprehensive keys for the western United States. Species in the genus Sphecodes were also only identified to genus level due to the lack of available taxonomic literature for the area.

Range
Species ranges were determined using Global Biodiversity Information Facility (GBIF), DiscoverLife.org (Ascher and Pickering 2016), and U.S. National Pollinating Insects databases in order to determine how many species records were new to the state of Montana.

Analysis
Between 2013 and 2015 we collected a total of 64,747 bees representing 229 species and morphospecies across 38 genera in 5 bee families (Suppl. material 1). We had a total of 558 collection events (unique combinations of site and collection date). The number of species we collected is near the number of species we would expect to find in our study area based on species accumulation estimates, which predict the number of bee species in the area to be 255 (Fig. 2). Net collections yielded a total of 110 species, only three of which were species not also collected by pantraps. This number is lower than would be expected based on previous studies showing from 23% and 53% of the bee species collected by net to be unique when compared to pantraps (Grundel et al. 2011, Roulston et al. 2007, Wilson et al. 2008). This may be due to the reduced net collection time period (30 minutes instead of 60 minutes). In addition, many of the net collectors had no prior experience with insect collecting. Increased collection periods and improved collector training could raise the number of unique net collected species.

Discussion
Prior to our investigation, the number of known wild bee species in Montana was 337. Our study increases this number to a total of 366. This result may not be unexpected since Montana is largely rural and there have been few large scale bee inventories in the state.

Megachile apicalis
Megachile apicalis is an Old World species in the subgenus Eutricharaea. Although there are North American records for M. apicalis prior to 1932, no established populations were recorded until 1984, when resident populations were documented in Santa Barbara County, CA (Cooper 1984). Subsquent collections showed that the species had expanded its range to Oregon, Washington, and British Columbia, Canada, along with its most common floral host, Centaurea solstitialis (Barthell et al. 2001, McIver et al. 2009, Sheffield et al. 2011), a widespread, introduced weed that invades rangelands in the western United States. There is evidence for a positive link between the spread of non-native weeds and the presence of non-native bees (Hanley and Goulson 2003). Megachile apicalis has a Species accumulation curve generated for bee species sampled in 558 collection events between 2013 and 2015. Species estimates were generated using EstimateS. The blue line represents the mean species accumulation, and the light blue shaded area represents the upper and lower bound of 95% confidence interval for species estimate.
strong association with C. solstitialis and some researchers have suggested that these two Old World species, along with Apis mellifera, form an invasive mutualism (Barthell et al. 2001, McIver et al. 2009), in which each of the three species within the mutualism benefits from the presence of, and encourages the spread of the other species within the mutualism. Our collections are the first documentation of M. apicalis in Montana, indicating further range expansion of this adventive bee. Centaurea stoebe, a close relative of C. solstitialis, is a problematic weed species throughout MPG Ranch and Montana. Megachile apicalis captures increase in late summer/early autumn, when Centaurea stoebe bloom intensity peaks (MPG Ranch unpublished data), suggesting that M. apicalis may be preferentially using this Centaurea species in our area. Additionally, M. apicalis appears to be integrated throughout the MPG Ranch, since we have found M. apicalis at most of our sampling locations.