Social insects are highly successful invaders of novel habitats and, of these, wasps and yellowjackets are quite successful worldwide. Vespula germanica (German yellowjackets) have become a pernicious cosmopolitan problem – yet the degree of their urban association and domination remains unclear. We investigated the invasion and behavior of V. germanica in Western New York along a rural-urban gradient to understand their conspecific dominance and their foraging behavior around human activities. We placed yellowjacket baits in urban, suburban and rural sites through the summer season to assess their dominance, and we conducted behavior trials at picnic tables to observe their responses to human activity. Given that the invasive ecotype prefers to nest in human structures, we predicted we would find the greater V. germanica dominance near structures and in urban areas. Moreover, given that they learn to associate food with cues, we predicted that they would associate human presence/activity at picnic tables with food. We found that V. germanica dominated everywhere (70-95% of all Vespula wasps), but dominance was highest in suburban and urban areas, with a preference for foraging near human structures. Surprisingly, behavior trials revealed that the yellowjackets were less attracted to baits with humans nearby, suggesting an avoidance of human activity rather than learned affinity. These findings reinforce V. germanica's attraction to human structures and urban environments, and their high negative impact on native yellowjacket populations.

The study area fell within the Western New York (U.S.) geographic area, bounded on the northern and eastern edges by Lakes Ontario and Erie, the western edge by Rochester, NY and the southern edge by the Alleghany Plateau. The study area included urban (the greater metropolitan areas of Buffalo, Niagara Falls, and Rochester) suburban and rural habitats and encompassed about 12,000 km². Within that area, we acquired permission from municipal and private entities, and from that pool randomly we selected sites that were urban (dominated by impermeable surfaces; no single-family housing), suburban (permeable surfaces such as lawns and trees; dominated by single-family housing) and rural (dominated by permeable surfaces such as agricultural fields and forests; sporadic single-family housing). Every two weeks for 20 weeks (June-November 2022) we sampled 9 sites (3 of each: urban, suburban and rural). We captured no insects during the first baiting (June 13), so we resampled the same sites for the second baiting (June 27), resulting in n = 99 baiting events at 90 unique sites (n = 30 site in each land use category). Given that the urban-rural ‘gradient’ can be considered more conceptual than empirical with boundaries that can be hard to identify (Wandl et al. 2014; Warren et al. 2018), we treated land use as categorical (urban, suburban and rural) and selected different sites from within each category for each sampling. We also chose this approach because destructively sampling workers at the same location through the summer might confound results with time. By using land use as a treatment, however, we retained the ability to quantify seasonal effects. The minimum distance between sites was 1.5 km; the maximum was 115 km.

At each baiting location each week, we placed a Victor yellowjacket trap filled with 60% Italian water ice concentrate (diluted 1:1 with water) and 40% beer (Goldstein 1996). We chose the Victor traps because we were sampling in urban and suburban areas, and the commercial traps are recognizable to the public. The sugar and yeast attract the Vespula yellowjackets, and the alcohol kills them. Whereas other methods of trapping Vespula yellowjackets may be more effective overall (Landolt 1998; Haupt et al. 2019), given that we were leaving traps in public areas, and we were measuring relative and not absolute abundance, we chose this approach because it has been demonstrated effective in an urban setting (Goldstein 1996) and to minimize non-target animal interference and toxicity. We used a trap and bait (sugar and alcohol) combination normally used to manage Vespula populations that is not attractive to nontarget insects, including other wasps and bees, but there is little data comparing bait preferences amongst Vespula species – though there is some indication that V. germanica workers prefer protein baits more than some other Vespula species (Harris 1991; Harris and Oliver 1993; Pereira et al. 2013). Generally, the traps were placed on hard surfaces, like picnic tables, road curbs and porch steps. The traps were left out from morning to dusk, ~8 hours. We avoided placing traps during inclement weather, which sometimes lasted for several days.

During preliminary research, we noticed that Vespula yellowjackets swarmed around people eating outdoors, and a preliminary class project suggested that they were more likely to visit baits with students standing nearby than those left alone. Given that the preliminary project was conducted in September, we suspected that, if this behavior was present, it might be learned from previous summer foraging. Hence, we sought to test whether yellowjackets on campus learn to associate food with people eating outside at picnic tables. We randomly chose four separate picnic tables (from 17 potential picnic tables; all at least 100 m apart) on the SUNY Buffalo State University campus and placed a Victor trap on each for one hour once a week for nine weeks between August and October 2022. Because we were investigating the association between human activity and food, and not human food, we used the same baiting formula as the gradient study given that it would not be as familiar to the campus foraging Vespula yellowjackets. For one hour at ‘lunchtime’ (12-1 pm), a single person sat at two of the four picnic tables 0.5 m from the Victor trap (the other two tables were left alone as controls). The humans made moderate movements as if they were eating during the trial periods.

All collected wasps were identified to species using Akre (1981), Buck (2008), Kimsey and Carpenter (2012). Voucher specimens are stored at Buffalo State University.