Ecology is full of fleeting interactions that nevertheless can reveal the rules structuring communities and ecosystems. These cryptic interactions—subtle, ephemeral, and often context-dependent—are frequently overlooked (Janzen, 1974) yet can illustrate underappreciated links between species and guilds. Indeed, cross-guild, parallel interactions can reveal the fundamental mechanism driving trophic interactions that transcend spatial and temporal scales and taxa (Schmitz, 2005). Here, we document a cryptic and ephemeral food chain linking foxes, berries (an important food source for foxes in summer), scat, flies, and rove beetles—an interaction structure that differs markedly from classical producer–consumer–predator food chains. For nearly a decade, our research team has conducted fieldwork on Isle Royale National Park in Lake Superior, United States, to investigate the trophic and behavioral effects of gray wolves (Canis lupus) on meso-carnivores, namely, red foxes (Vulpes vulpes) and American martens (Martes americana; Rodriguez Curras et al., 2024). Across approximately 750 km of trails hiked each year, we collect fox and marten scats for genetic and dietary analysis (Lacin Alas et al., 2025) while subsequently trapping foxes to monitor their behavior, diet, and activity patterns. On September 10, 2021, we were fortunate to observe one of those “cryptic interactions that often escape the eye” (Janzen, 1974): prior to collecting a fox scat on the Feldtmann Trail near Cumberland Point (Figure 1A) we observed a gold-and-brown rove beetle (Ontholestes cingulatus)—also called “carrion beetles” (hereafter rove beetle)—perched on top of the scat with its brightly colored golden-yellow tail waving in the air as a small mass of hovering flies circled above. As a large black fly of an unidentified species (Family: Sarcophagidae) descended, the rove beetle leapt up approximately 10 cm, captured, killed, and began consuming the fly (Figure 1B). The following year, we set out to document rove beetles along the scat transects and en route to fox trapping locations. Out of a total of 148 red fox scats (all ≤4 days old, as transects were surveyed at 4-day intervals) encountered and collected in the field along approximately 700 km of trail, we (along with two field technicians) detected rove beetles on approximately 10% of the total fox scats (N = 15). On two occasions, we detected two individuals on a single scat, one usually larger (estimated ~50% larger) than the other. In addition to the documented predation event reported here, we observed similar hunting behavior on two other occasions, although these were not photographed. We therefore distinguish between a single documented predation event and repeated observations of beetle presence on fox scats. However, we did not observe rove beetles on the scats of any other species (wolf, marten, or moose) that we encountered during our transects, including >30 moose droppings (though their larvae have been previously observed hatching from moose droppings in Isle Royale; Egan Lacin Alas et al., 2025), we did not quantify fruit content in individual scats where beetles were observed. Thus, we present the proposed link between carnivore scats, fly attraction, and beetle predation as a hypothesis for future study. Following our observations, we reviewed all available publications in Web of Science, Scopus, and ScienceDirect using “Ontholestes cingulatus” as a keyword (N = 10), we found no documentations of this species utilizing terrestrial carnivore scats as a hunting ground, though they have been observed on river otter (Lontra canadensis) latrines in New Brunswick, Canada (Gallant et al., 2022). We expanded our search to include the family (Staphylinidae) AND “hunting” (yielding N = 169 records) and found records of other species (e.g., Leistotrophus versicolor) using two main tactics: ambushing flies at scat/carrion or on decomposing matter and prey-luring by releasing scented abdominal secretions (Forsyth Lacin Alas et al., 2025). Adult rove beetles are ambush predators that rely on the irregularly distributed resources of their prey, specifically, those that exploit ephemeral dung, carrion, or fungi (e.g., Drosophila spp.) as a hunting ground (Teskey, 1969; Worthen, 1989a). Our observation complements prior work on scat-associated predator communities (e.g., Forsyth Hanski Teskey, 1969; Worthen, 1989a, 1989b). In one experiment controlling the presence of rove beetles on a food source, face flies (Musca autumnalis) ceased oviposition or flew away if a rove beetle was detected (Teskey, 1969). Worthen (1989a), (1989b), showed that rove beetles structure fly communities via keystone predation, decreasing the natal recruitment of competitively dominant species and thereby releasing subordinate competitors from intraspecific competition. Within invertebrate food webs, rove beetles serve as voracious predators of ovipositing insects (e.g., Drosophila spp.), structuring communities and potentially providing feedback loops into vertebrate food webs, a hypothesis that warrants future testing. Vertebrate scats are an important resource for a variety of plant and invertebrate species, serving as a mode of dispersal and biogeochemical cycling as well as a nutritional resource and microhabitat, and—as we described above—can facilitate trophic interactions. For example, due to their large size, the seeds of the jicaro (Crescentia alata) and guanacaste (Enterolobium cyclocarpum) in Central and South America were likely dispersed by large mammals of approximately 500 kg (Janzen Schmitz, 2008) and large-scale studies featuring pumas (Puma concolor; Laundre et al., 2010), which have suggested that ambushing predators can have strong effects on prey densities and behaviors, thereby influencing community structure and ecosystem function (Preisser et al., 2007). Cryptic interactions, specifically those that link modular food webs, may influence interspecific interactions across guilds (McCann et al., 1998). Invertebrate communities, and specifically flies, have important direct and indirect impacts on vertebrate populations through competition for resources, as pests, pollination services, and disease/parasite transmission. Predators such as rove beetles, then, may regulate invertebrate prey in ways that could cascade to vertebrates—an ecological consequence we propose as a testable hypothesis rather than a demonstrated outcome of the present study. Our observation of rove beetles preying on flies at red fox scats reveals a cryptic connection between vertebrate and invertebrate food webs. Such cross-guild interactions highlight how ephemeral vertebrate resources can structure invertebrate communities and, further, exemplify the cascading and facultative nature of species interactions. Just as carrion shapes vertebrate predator–scavenger dynamics, fox scats act as microhabitats that attract, concentrate, and expose invertebrate prey to predation. This parallel—fatal attractions (sensu; Sivy et al., 2017)—underscores that the form and function of trophic interactions are remarkably conserved across taxonomic boundaries. Our findings highlight that vertebrate and invertebrate food webs are not silos but interwoven, dynamic components of food webs, and that attention to these subtle mechanisms can yield new insights into the structural connections that link otherwise disparate components of ecosystems. We would like to thank M. Romanski and L. Potvin for their steadfast support to our research program. We would also like to thank T. Peltier, M. E. Martin, J. Marchman, K. L. Black, M. R. Cooper, P. J. Manlick, M. Dudenhoeffer, B. Lacin Alas, J. Bonessi, E. Olsen, and S. Heirs. Research associated with this observation was supported by grants from the US National Park Service (P20AC00057, P22AC01601), the National Park Foundation, and the Shadle Fellowship from the American Society of Mammalogists. We thank John Pastor, Oswald J. Schmitz, and one anonymous reviewer for their insightful comments to our manuscript. Individual components of Figure 2 were generated using Adobe Firefly. The authors declare no conflicts of interest.
Curras et al. (Fri,) studied this question.