The Nine-banded Armadillo (Dasypus novemcinctus) is a widespread burrowing species with an expanding geographic range across the southeastern and midwestern United States. Armadillos dig numerous, large burrows within their home ranges and these burrows are likely used by a diverse suite of wildlife species as has been reported for other burrowing ecosystem engineers such as Gopher Tortoises (Gopherus polyphemus), Desert Tortoises (Gopherus agassizi), and Black-tailed Prairie Dogs (Cynomys ludovicianus). We used motion-triggered game cameras at 35 armadillo burrows in 4 ecoregions of Arkansas and documented 19 species of mammals, 4 species of reptile, 1 species of amphibian, and 40 species of bird interacting with burrows. Bobcat (Lynx rufus), Coyote (Canis latrans), Eastern Cottontail (Sylvilagus floridanus), Gray Fox (Urocyon cinereoargenteus), Gray Squirrel (Sciurus carolinensis), Northern Raccoon (Procyon lotor), Virginia Opossum (Didelphis virginiana), and unidentified rodents (mice and rats) were documented using burrows in all four ecoregions. We documented wildlife hunting, seeking shelter, rearing young in, and taking over and modifying armadillo burrows. The rate of use was highest in the Mississippi Alluvial Valley, a landscape dominated by agriculture, where natural refugia may be limited and rodents are abundant. Armadillo burrows are clearly visited and used by numerous wildlife species to fulfill various life stage requirements and this list will likely expand if more attention is devoted to understanding the role of armadillos burrows. Armadillos are important ecosystem engineers, and their ecological role warrants more investigation and attention as opposed to only being viewed and managed as agricultural and garden pests.

Beginning in March 2020, we located burrows by walking the sites and opportunistically encountering burrows. We selected burrows where we felt cameras would not attract attention from hikers or recreators and where the field of view of the burrow was relatively unobstructed by vegetation. At selected burrows we set motion-triggered wildlife cameras. We used several types of cameras for this study (Reconyx Microfire, Bushnell HD Aggressor, Bushnell Core, and Browning Strike Force HD). All cameras were set to take a burst of 3 photos each time they were triggered and to have a reset period between 0 and 8 seconds (depending on the available options of the camera model) before triggering again. Cameras were placed on nearby trees or tripods placed between 2 and 3 m from the burrow entrance. Cameras were set approximately 50-75 cm off the ground and angled towards the burrow entrance to ensure we captured all wildlife interacting with the burrows. The amount of time that individual cameras were left in place varied from 1 month to slightly over a year due to various factors including camera failures, flooding, and making accommodations to landowners. All cameras were removed from the field by June 2021.

Data was processed using Timelapse 2.0 to review all photographs and assign species ID to each wildlife trigger. We combined all photographs taken within a 5 min period as a single detection to reduce the likelihood of double-counting individuals. For each wildlife detection we identified the species present, the number of individuals visible, and we recorded which part of the burrow an animal interacted with including passing by (no interaction with burrow), apron (animal interacting with the mounded sand, bare ground, or leaf litter piled in front of the burrow), entrance (animals sniffing, foraging, or inspecting the opening to the burrow), or interior (animals that moved beyond the entrance of a burrow to enter, exit, or investigate the tunnel of the burrow). We also extracted the date and time of all interactions with burrows. While we attempted to identify each trigger to the species level, for most rodents we categorized them simply as “mice” or “rats”. For all other vertebrates, we excluded detections where we were unable to identify the species.

For additional details about this dataset, please use the following ReadMe.txt file.