Anthophora has a worldwide distribution and, containing about 450 species, is one of the most common genera within Apidae. Some species of Anthophora have been managed for pollination services (Stubbs & Drummond 1999), and Anthophora abrupta Say (Hymenoptera: Apidae) has been suggested as a potential species that could prove manageable for pollination of crops in the southeastern United States (Graham et al. 2015). Despite this potential, only a few types of nesting site structures have been documented. Other than vertical riverbanks (Frison 1922), only man-made structures, such as adobe walls (Norden & Scarborough 1982; Norden 1984), bags of colloidal clay (Graham et al. 2015), and clay banks underneath a porch (Rau 1929), have been documented as nesting sites for A. abrupta.

Anthophora abrupta is a commonly occurring species in the eastern United States (Mitchell 1962). It is a gregariously nesting bee species that has been documented nesting in clay-containing substrates (Norden 1984; Graham et al. 2015), often near riverbanks (Frison 1922). Anthophora abrupta is often known as the “turret-building bee” because the bees construct turret-like projections that cover the opening of their nest entrance (Rau 1929). The exact function of the turrets is unknown, but the structures may protect the nests from parasites or rain, provide a social purpose for aggregation, or possess a thermoregulatory function (Walsh & Riley 1868; North & Lillywhite 1980; Norden 1984). Although A. abrupta females are gregarious nesters, each individual constructs a separate tunnel that leads to a single nest (Norden 1984). To construct tunnels, female bees use water to soften the compacted and hardened clay (Rau 1929). Following construction, females waterproof their individual tunnel and nest with a glandular secretion from their Dufour's gland (Norden et al. 1980).

Tree falls due to weather events (e.g., hurricanes, tornadoes, high winds) are ecologically important forest disturbances that can create numerous novel niches (Schaetzl et al. 1989a; Greenberg & McNab 1998). Trees that fall often expose their root plate (or root wad), an intact mass of soil and root that, depending upon soil type, can contain large quantities of clay. Although root plates can persist in place for decades (Small 1997), the structures eventually form a relict tip-up mound where the vertical root plate once existed, along with a resulting depressed pit from which the previously intact root system was displaced (Schaetzl et al. 1989b). It has been well established that downed wood and snags within forested habitats contribute to biological richness for many invertebrates, vertebrates, and plants (Harmon et al. 1986; Berg et al. 1994). Here we provide evidence of vertical root plates in southeastern forests being utilized as A. abrupta nesting sites.

Seven A. abupta aggregation nests were found in root plates at 5 sites in the southeastern United States (Fig. 1). The first site was at the Bluebonnet Swamp Nature Center, East Baton Parish, Louisiana (30.371058°N, 91.106173°W), where 2 root plates were found containing A. abrupta nests. The first A. abrupta nest at this site was discovered in 2009 in a root plate (labeled BS1) from a hardwood tree that had fallen in 2008 due to Hurricane Gustav. This nest, situated approximately 50 m from the edge of a swamp, remained active for 4 yr, but the number of bees observed became less each year as the root plate turned into a low-profile tip-up mound. In 2012, a second nest was discovered in a root plate (BS2) approximately 25 m away from the first root plate at the Bluebonnet Swamp Nature Center. The second site was located in the Green River Game Lands in North Carolina (35.262792°N, 82.293511°W), where 2 root plates from fallen hardwood trees were found to contain an A. abrupta nest. The first root plate (GR1) with an A. abrupta nest was discovered in 2012 (Fig. 1) and was still active as of 2016. In 2016, a second root plate (GR2) containing an A. abrupta nest was located approximately 100 m from GR1. The third site was discovered in 2016 near Cedar Bluff, Alabama, along Mud Creek (34.180230°N, 85.481536°W) (Fig. 1), where an A. abrupta nest was found in an extremely large root plate (MC) of an oak tree (Quercus sp.; Fagaceae) that was situated approximately 10 m from a creek (Fig. 1). The fourth site was located along Salem Lake in North Carolina (36.104083°N, 80.166550°W), where a small, old root plate (SL) contained some visible turrets in 2016. One additional root plate (FB) was discovered at a fifth site in early 2017 within Fontainebleau State Park, St. Tammany Parish, Louisiana (30.332611°N, 90.015353°W). This root plate was heavily colonized with numerous intact turrets.

Particle sizes of the soils were determined by pipette-extraction methods (Robinson 1922). Table 1 shows the percentages of sand, silt, and clay within each of the 7 root plates that contained A. abrupta aggregations. Soil was sampled from all sites in 2016 except Fontainebleau State Park, which was sampled in 2017. Root plates BS1 and BS2 no longer had evidence of active nests when soil was sampled, whereas all other root plates had evidence of recent nests (visible, intact turrets).

Fig. 1.

Root plate examples from Green River Game Lands (North Carolina), Fontainebleau State Park (Louisiana), and Mud Creek (Alabama). Map depicts physiographic provinces and relative locations of root plate sites. Upper right picture shows close-up of Anthophora abrupta turrets from BS root plate. BS = Bluebonnet Swamp Nature Center (Louisiana), FB = Fontainebleau State Park (Louisiana), GR = Green River Game Lands (North Carolina), MC = Mud Creek (Alabama), SL = Salem Lake (North Carolina).

We located 7 root plates containing A. abrupta aggregations within 4 very distinct physiographic regions (Coastal Plain, Ridge and Valley, Piedmont, and Blue Ridge physiographic provinces). Previously described nesting sites of A. abrupta include vertical river banks, clay banks, and colloidal clay blocks (Frison 1922; Rau 1929; Graham et al. 2015). The primary nesting site of A. abrupta in forested habitats is unknown. Because vertical root plates are a common feature of forested habitats, these structures provide highquality A. abrupta nesting sites in areas that have scarce availability of exposed clay or other compacted soil and, thus, may serve as the primary nesting material in forests for these bees. Although vertical riverbanks (Frison 1922) are the only documented natural nesting site of A. abrupta, with the advent of widespread dam building, this habitat may now be rare making root plates desirable structures for A. abrupta nests.

Table 1.

Results from grain size analysis of each root plate that contained Anthophora abrupta nests. Listed are the percentages of sand, silt, and clay and the identified soil type.

Cane (1991) showed that Anthophora (A. abrupta included) bees preferred soil with a relatively higher percentage of clay compared with soil types preferred by many other ground nesting bees. This preference of relatively higher clay content for nesting structure may give them the ability to monopolize the use of root plates that usually contain high percentages of clay. In our nested root plate samples, we found a range of clay concentrations (13–33%), but all A. abrupta nests occurred in a type of loam, consistent with the observation of Cane (1991).

Graham et al. (2015) showed that it is possible to expand nesting of A. abrupta by placing suitable nesting material nearby active nest sites. Other Anthophora species have been used as commercial pollinators in Germany (Thalmann & Dorn 1990) and Japan (Maeta et al. 1990) and have similar ecologies as A. abrupta. Therefore, A. abrupta displays features that may make it a suitable pollinator species for potential management (Graham et al. 2015). The use of root plates for nesting sites in multiple habitat types also provides potential for nests to be easily located and augmented with artificial nesting sites placed nearby root plates.

We thank Chase Kimmel for GIS assistance and Allyn Irvin for providing help with photographs. We also thank Niky Hughes, Andrew Vigueira, and Belinda Vigueira for sharing the location of the root plates. This research was funded through a United States Department of Agriculture-Joint Fire Science Program (project No. FA-FON0013-0001).