Associations between a feral equid and the Sonoran Desert ecosystem Asociaciones Entre un Equino Salvaje y el Ecosistema del Desierto Sonorense

The effect of non-native herbivores on ecosystems and diversity has become a global concern in conservation. Management challenges associated with non-native free-roaming equids have existed for decades in a wide range of ecosystems yet have been difficult to resolve. Although much of the challenge is associated with non-biological considerations, empirical ecological research is crucial for guiding sound management decisions. We conducted a field study on the associations between feral burros (Equus asinus) and elements of the Sonoran Desert ecosystem in Arizona, USA, during 2017–2019. We identified areas with and without established burro herds, and collected data on vegetation, ungulate sign, small mammals, birds, and herpetofauna at multiple, randomly selected grids within these areas, while accounting for vegetation community and distance to water. We predicted that burros would be associated with differences in vegetation metrics such as lower ground cover, smaller perennial plant size, and lower plant density, foliage density, recruitment, and species richness among perennial native plants susceptible to burro foraging or trampling. We further predicted that these differences would be accompanied by lower density or relative abundance and lower species richness of small mammals, birds, and herpetofauna. Finally, because burro distribution has been documented to be associated with water in this arid landscape, we predicted that effects would be most pronounced near water. The results of our study did not consistently support our predictions, perhaps because of small sample sizes or, in several cases, inherent complexities associated with seasonal burro habitat use and plant phenology patterns. However, our study documented that the presence of this feral equid is associated with a number of key differences that may be ecologically important and have the potential to alter community structure in this sensitive arid ecosystem. In areas with established burro herds, we documented lower ground cover, plant density, foliage density, or smaller plant size in several species, and changes were often influenced by distance from water. For example, density of Engelmann's prickly pear cactus (Opuntia engelmannii) was 94% lower and Anderson wolfberry (Lycium andersonii) plants were 49% smaller in areas with established burro herds. In areas with burros, we also recorded lower density of white bursage (Ambrosia dumosa) in areas distant from water. Of notable concern was that our metric of recruitment indicated 63% lower recruitment in saguaro cactus (Carnegiea gigantea) and that foliage densities of yellow paloverde (Parkinsonia microphylla) and desert ironwood (Olneya tesota) were lower in areas with established burro herds. Data on some plant species did not support our predictions. For example, white bursage and Anderson wolfberry plants were found at similar densities in areas with and without established burros near water, but they occurred at lower densities far from water in areas with established burros. Our data revealed that in 4 of 7 small mammal species evaluated (Bailey's pocket mouse [Chaetodipus baileyi], desert pocket mouse [C. penicillatus], deer mice [Peromyscus spp.], and Merriam's kangaroo rat [Dipodomys merriami]), density was associated with an interaction between burros and distance to water, with lower densities close to water in burro areas. Contrary to predictions, 3 of these species (Bailey's pocket mouse, desert pocket mouse, and deer mice) exhibited higher densities in burro areas than in non-burro areas at grids farther from water. Density of a fifth species (Arizona woodrat [Neotoma devia]) was 68% lower in burro areas than in non-burro areas, and the densities of 2 species were not associated with burros. Across species, we did not find consistent patterns in our analysis of bird group density, with some species exhibiting a negative effect associated with burros and others exhibiting a positive effect. When we categorized birds by hypothesized nesting and foraging vulnerabilities (low, medium, high), vulnerability levels did not predict the effect of burros. However, all categories exhibited a negative burro effect distant from water but not close to water, contrary to our expectations. Relative abundance of common side-blotched lizards (Uta stansburiana) was 26% lower in areas with established burros, but data on other herpetofauna species did not support our predictions, with some species exhibiting higher relative abundance in areas with established burros. Our data did not reveal an association between burros and bird, small mammal, or herpetofauna species richness, but species richness of native perennial plants was higher in burro areas close to water. We recommend that future bird studies focus on riparian birds and nest success, and possibly evaluate potential effects in relation to other aspects of bird ecology such as feeding guilds or nesting ecology, and that future herpetofauna studies use survey methods that can better account for detection. Although some results did not support our predictions, our study documented negative associations between burros and a number of native plant species, and density in some small mammal species. These associations are important and of concern in and of themselves because changes in long-lived keystone plant species and in small mammal densities indicate that the long-term sustainability of portions of this ecosystem may be affected, and it is likely that these changes can have additional indirect effects on plants and wildlife in this ecosystem. Field data on ungulate sign (fecal groups and tracks) suggested that the associations detected in our study were related to burros and not cattle (Bos taurus) or native ungulates such as mule deer (Odocoileus hemionus) or bighorn sheep (Ovis canadensis). Our results indicate that the presence of established burro herds was associated with changes, primarily in the plant community that is critical for ecosystem function, and we suggest that current management of this feral equid may not be adequate for maintaining the long-term viability of this arid and fragile ecosystem.