The discovery of a population of N. asteris on a patch of sawtooth sunflowers (Helianthus grosseserratus M. Martens) (Fig. 1E) occurred at Mutt Run Off-Leash Dog Park in Lawrence, KS on 22 September 2025 at approximately 17:45. Subsequent observations took place on 25 September 2025 at 18:45, 26 September 2025 at 19:00, and 28 September 2025. Count data for analysis was obtained on 26 September 2025 and is reported in Table 1. Other types of bees and insects at the site included non-parasitic eucerine bees, most likely in the genera Melissodes Latreille, 1829, Eucera Scopoli, 1770, or Svastra Holmberg, 1884, various members of Lepidoptera, soldier beetles in Chauliognathus Hentz, 1830, and the spotted cucumber beetle, Diabrotica undecimpunctata Mannerheim, 1843.

Figure 1. 

Photos of Helianthus grosseserratus and aggregatory Nomada asteris Swenk, 1913. A. Five males aggregating on a single flower; B. Two males on petals of flower; C. Male N. asteris resting on flower petal; D. Female on flower petal; E. Patch of H. grosseserratus where observations were made.

To determine statistical significance of the observational data (Table 1), the authors ran three statistical tests in R Core Team (2025), testing grouping or clustering at flower patches. A Chi-square goodness of fit test following a uniform distribution was performed using the chisq.test function. Then, a second Chi-square goodness of fit test following a Poisson distribution was performed. Lastly, the data was assessed using a Poisson model via the glm function.

Photos were taken in situ using an iPhone 16 and a Nikon D3100 Digital SLR camera with a Yongnuo 50 mm 1: 1.8 (YN50 mm F1.8N) full frame lens. It was not possible to collect specimens from the site without a collection permit from the Kansas Department of Fish and Wildlife. Museum specimen photos were taken using a Macropod Pro 3D photomicrography system from Macroscopic Solutions®, consisting of a Canon EOS 6D Mark II camera. Zerene StackerTM software package was used to condense images into a single, fully focused image, which were later combined into plates using Adobe Photoshop® CC. Scalebars were added at this time.

Identification of the N. asteris observed was done by comparing photos taken in-situ to museum specimens at the University of Kansas Snow Entomological Collection. Additionally, keys from Broemeling and Moalif (1988), Michener (2007), and Odanaka (2024) were used for visible traits. Defining characteristics included: from Broemeling and Moalif (1988): antennal scape globose, supraclypeal area, sides of face up to and around antennal sockets, labrum, clypeus, ring extending behind compound eyes almost to vertex, pronotal lobes, tegulae, scutellum, and metanotum, creamy-white, antennae, legs, ferruginous, remainder of body dark fuscous to black. Odanaka (2024) was used to differentiate between N. asteris and other, morphologically similar species, namely members of the N. vincta group. From this, a single character was used in tandem with those previously noted to determine that the observed species was not a member of the N. vincta group (lack of maculations on propodeal sides).

Identification was determined by comparison of the specimen to holotype photos on Discoverlife (Ascher and Pickering 2025), comparison to holotype description in Swenk (1913), and through the keys of Broemeling and Moalif (1988) and Michener (2007). Several key characteristics were considered during the identification of this specimen; from Swenk (1913): black or blackish small spot behind ocelli, a spot on each side of collar, the tegulae (in reference to the pronotal lobes), elevated portion of mesoscutellum and metanotum, yellowish white, antennae red, the first three joints brightly so, the following joints lightly suffused with dusky, tergite 2 with large suboval yellowish white lateral spots, 3 and 4 with similar but slightly smaller spots, venter clear red; from Broemeling and Moalif (1988): 3 submarginal cells in forewing, abdominal terga never with complete transverse yellow or cream-colored bands, posterior-medial triangle on scutum darkened to fuscous. The female specimen identified differs from the holotype specimen in tergite 1 with small yellowish white lateral spots. The authors have attributed this to intraspecific variation and have identified the specimen to N. asteris. A live female (Fig. 1D) was found at the observation site. Though partially obscured by the angle of photo, several key characteristics are visible that distinguish this species from other, morphologically similar species. These characteristics, taken from Broemeling and Moalif (1988), include the yellowish white spots on each side of the collar, the tegulae, elevated portion of mesoscutellum and metanotum (Fig. 1D), first three antennal segments bright red, the remaining dusky, and incomplete transverse yellow or cream-colored bands. Darkened markings on the thorax comparable to those described in Broemeling and Moalif (1988) are also visible.

Determined through comparison to descriptions in Long (1959), Long (1961), Heiser et al. (1969), and Gudžinskas and Petrulaitis (2014), as well as to photos from Kansas State University Libraries: Kansas Wildflower & Grasses (2026).

As noted in the Methods section, the initial observation of N. asteris was on 22 September 2025 at approximately 17:45. At the time of the initial observation, the authors could identify the bees to Nomada but were unaware of the species. Individuals were somewhat active but did not fly between flowers unless disturbed. This was taken as an indicator that the males were beginning to roost, though the authors did not record data at this time due to the unplanned nature of this observation. On 25 September 2025 at 18:45, the authors recorded 38 roosting males present within the patch (Fig. 1E).

To better estimate the number of roosting males and investigate the possibility of movement to another nearby sunflower patch, on 26 September 2025 at 19:00 the authors recorded the number of N. asteris present in the main patch (Fig. 1E) and walked throughout the rest of the approximately 45-acre field where the sample patch was located to assess the number of N. asteris at other sunflower patches in the same area. The authors considered a patch as any grouping of H. grosseserratus ramets at least 4 feet in diameter and at least 10 feet away from another grouping. 25 males and one female N. asteris were recorded in the main sample patch (Fig. 1E). After visually assessing other isolated patches of sunflowers, the authors counted only 15 other N. asteris, all scattered throughout the other patches. There were no N. asteris found on other flowering plant species in the field. Data obtained is within Table 1. The authors returned to the sample patch on subsequent dates (28 September 2025 through 4 October 2025) and found one, then zero N. asteris, respectively. While the H. grosseserratus were blooming for several weeks after the initial finding, there were no further observations of N. asteris— it is likely that the authors’ discovery occurred near the end of their typical adult season, making this chance observation a truly rare find.

The significant results of this study suggest that the observations on aggregatory behavior of male N. asteris are non-normal, indicating overdispersion relative to the number of available H. grosseserratus patches. While overdispersion, specifically clustering, is considered common in ecological systems (Barron 1992; Richards 2008; Harrison 2014), the authors cannot determine the cause without additional data collection. Herein, the authors speculate on the possible drivers of aggregatory behavior and include considerations for future studies.

One possible explanation may be rooted in thermoregulatory techniques. Males across all patches were observed to have three distinct roosting behaviors: individuals on leaf margins, individuals on petal margins, and groups of individuals on petal margins (Fig. 1A–D). This is unique from other observations in the literature on male roosting behavior in that nearly all individuals observed roost along petal margins instead of along grass stems or in the flower’s center (Rau and Rau 1916; Alcock 1998; Vereecken and McNeil 2010). The outward preference for the margins of petals over other, previously documented relative locations could be reliant on the position of H. grosseserratus flowers relative to the sun. Numerous studies have addressed the eastward orientation of sunflowers in Helianthus, including Atamian et al. (2016) and van der Kooi (2016). Additional studies have examined the thermal properties of flowers themselves, particularly members of Asteraceae, finding that both the central disks and petals absorb, retain, and generate heat (Sapir et al. 2006; Dietrich and Körner 2014; van der Kooi et al. 2019). This explanation is furthered by the consideration of “shelter aggregations”, as noted in Allee (1927), in which bees aggregate because of small amounts of available shelter material. Given the small body size of N. asteris, it is likely that this ectothermic species would be highly reliant upon both solar radiation and would have retained floral heat for thermoregulation overnight and before flying. An alternative explanation that runs parallel to this floral thermal dependence is that the grouping of multiple N. asteris on single petals may reduce risk of freezing, promoting more ideal physiological conditions for clustered individuals. This has only been documented in bees honeybee-sized and larger (Xylocopa Latreille, Ostwald et al. 2022), but we cannot truly dismiss this possibility without additional research into this behavior in smaller bee taxa.

Another explanation may be reduced individual predator risk. Alcock (1998) believed this to be the driver for male aggregations, noting that this may provide protection via dilution effect— a strategy that greatly lowers the probability that an individual will be eaten by a predator. This is a behavior readily seen in other organisms, including other insects (Foster and Treherne 1981; Wrona and Dixon 1991; Mooring and Hart 1992). It may be that the unusual positional behavior of N. asteris is another predator avoidance technique, driven by relative positions of primary arthropod predators, such as crab spiders, on flowers (Morse 1981; Heiling et al. 2005, 2006). Several studies have shown that bees can see well-camouflaged predators and will actively avoid flowers where crab spiders are waiting (Heiling and Herberstein 2004; Heiling et al. 2005; Huey and Nieh 2017). Schmalhofer (1999) notes that the comfortable temperature range that summertime crab spiders can inhabit tends towards warmer temperatures, suggesting that spiders adapted to the high temperatures during the day may be less successful hunters at night due to their own thermal limitations. Following this logic, it is possible that the crab spiders are also reliant upon the heat retained by the flowers of H. grosseserratus overnight. Thus, perhaps the N. asteris are roosting on the edges of flower petals to avoid crab spiders that may be positioned around the warm central disk of the flowers, relying on the thermal limits of the predators during cooler hours of the day and night, and roosting in groups to increase likelihood of survival should the crab spider still be capable of hunting.

Multi-year usage dependent on proximity to females may also drive aggregatory behavior in N. asteris. It has been documented that the males of some solitary bee species will return to the same site for nocturnal roosting both within and across seasons, and different generations being involved across seasons (Wcislo 2003). It is thought that the overnight location where males reside is determined by proximity to females, either where they will emerge or where they will forage, to increase chance of mating (Pinheiro et al. 2017). Males are highly competitive when it comes to mating opportunities (Velthuis and Gerling 1980; Paxton 2005; Alcock 2013). Additionally, multi-year nesting aggregations are common in non-parasitic, solitary bees because of preferential nesting conditions (Michener et al. 1958; Danforth et al. 2019; Tsiolis et al. 2022). Foraging cleptoparasitic female bees are likely to stay close to these aggregations (Danforth et al. 2019; Litman 2019; Moens et al. 2024). Given that H. grosseserratus is a perennial plant, it is likely that the patch observed here will be used again next year by male N. asteris. If this is the case, it may allude to the location of females and thus the host species, as well as the foraging preferences of the two species. Nomada are known parasites of several genera, most commonly Andrena Fabricius, 1775 (Cardinal et al. 2010; Litman et al. 2013; Danforth et al. 2019). Given the time of year and observations made by the authors, it is entirely plausible that N. asteris utilizes either a member of Andrena (Callandrena) Cockerell, 1898, or one of the eucerine species viewed at and around the observation site (Michener 2007; Danforth et al. 2019). While members of Callandrena are common in Kansas throughout the summer and fall, none were seen at the time of observation. Future investigations should be conducted with the goal of determining if this observation was a simple one-off and, if not, if N. asteris males exhibit a multi-year roost location preference. Additionally, future studies should aim to decipher diel behavior of both sexes, as well as locate the host species.

A tentative female specimen of N. asteris was reported on the civilian science platform iNaturalist in Norman, Oklahoma in late September, 2024 (Wingert 2024). Visible morphological characteristics, such as the darkened thoracic markings highlighted in Broemeling and Moalif (1988), might suggest that this specimen is N. asteris, but without additional images showing more of the key features confirmation of the observation is impossible. The specimen was flagged as a potential N. asteris by John S. Ascher, the noted taxonomist behind DiscoverLife, but the species identification remains unresolved. However, it would not be out of the question for this to be a representative of the species given the location where the specimen was observed. The authors find it extremely likely that N. asteris is a prairie-associated species and is possibly distributed throughout the plains regions of Kansas, Nebraska, and Oklahoma. Future bee diversity and sampling surveys are needed to ascertain the true distributional range of this elusive bee.