•  
  •  
 

Abstract

Riparian areas are an important resource for many wildlife species, especially in the arid Southwest. Understanding species occurrence in areas dominated by nonnative vegetation is important to determine whether management should be implemented. Saltcedar (Tamarix spp.) is one of the most prevalent and invasive nonnative woody species in riparian areas in the southwestern United States, and it can alter vegetation structure. Little is known about how meso (medium) and large carnivores and herbivores use or avoid stands of saltcedar. We investigated mammal activity in 3 riparian forest types along the San Pedro River in southeastern Arizona: nonnative saltcedar, native mesquite (Prosopis spp.) woodland, and native cottonwood (Populus fremontii) and willow (Salix gooddingii) forest. We determined mammal use, diversity, and occupancy across riparian forest types. We sampled mammals along approximately 25 river kilometers between July 2017 and October 2018, using 18 trail cameras (6 per forest type) spaced 1 km apart. We summarized environmental variables around the cameras and reduced the variables to 4 components using a principal component analysis. We observed 14 meso and large mammal species, including bobcat (Lynx rufus), coyote (Canis latrans), and white-nosed coati (Nasua narica) over 7692 trap nights. Occupancy of some species may have been influenced by the environmental components, but models showed high standard errors, making it difficult to draw firm conclusions. Our results do not indicate that nonnative saltcedar was actively avoided. There were no discernible differences in detections of species or taxonomic groups among forest types; however, the highest species diversity was in mesquite woodlands. We did not find evidence of competitively dominant predators (coyotes and mountain lions [Puma concolor]) altering the temporal activity patterns and forest use of subordinate species (gray fox [Urocyon cinereoargenteus] and bobcat), as each pair used the same riparian space at similar times.


Las áreas ribereñas son un recurso importante para muchas especies silvestres, especialmente en el árido suroeste de Estados Unidos. Entender los fenómenos asociados a la presencia de estas especies en áreas dominadas por vegetación exótica resulta muy importante para determinar si se debe implementar algún tipo de manejo. El cedro salado (Tamarix spp.) es una de las especies leñosas exótica más prevalente e invasiva en las áreas ribereñas del suroeste de Estados Unidos y puede alterar la estructura de la vegetación. Se conoce poco acerca de cómo los carnívoros y herbívoros, tanto medianos como grandes, usan (o evitan) los rodales de cedro salado. Aquí investigamos los patrones de actividad de mamíferos en tres tipos de bosques ribereños a lo largo del río San Pedro en el sureste de Arizona: cedro salado exótico, bosque nativo de mezquite (Prosopis spp.) y bosque nativo de álamo (Populus fremontii) y sauce (Salix gooddingii). Determinamos el uso, la diversidad y la presencia de mamíferos en todos estos tipos de bosques ribereños. Muestreamos mamíferos a lo largo de aproximadamente 25 kilómetros del río entre los meses de julio de 2017 y octubre de 2018, utilizando 18 cámaras de fototrampeo (seis por cada tipo de bosque) espaciadas por 1 km de distancia. Resumimos las variables ambientales alrededor de las cámaras y las reducimos a cuatro componentes mediante un análisis de componentes principales. Observamos 14 especies de mamíferos medianos y grandes, incluyendo el lince rojo (Lynx rufus), el coyote (Canis latrans) y el coatí de nariz blanca (Nasua narica) durante 7692 noches de fototrampeo. Si bien la presencia de algunas especies puede haber sido influenciada por los componentes ambientales, los modelos mostraron altos errores estándar dificultando la obtención de conclusiones firmes. Sin embargo, nuestros resultados no indican que los mamíferos eviten el cedro salado exótico. Tampoco hubo diferencias discernibles en la detección de especies o grupos taxonómicos particulares entre los diferentes tipos de bosques; aunque, la mayor diversidad de especies se encontraba claramente en los bosques de mezquite. No encontramos apoyo para hipótesis asociadas a la existencia de depredadores competitivamente dominantes (coyotes y pumas [Puma concolor]) que pudieran alterar los patrones de actividad temporales y el uso forestal de especies subordinadas (e.g., zorro gris [Urocyon cinereoargenteus] y lince rojo) ya que cada pareja mostró un uso contemporáneo del mismo bosque ribereño.

81.1.7 Supplementary Material 1.pdf (22 kB)
Mean and standard errors for habitat variables measured at each camera site (n = 18) in 3 riparian forest types along the lower San Pedro River, Arizona, USA.

81.1.7 Supplementary Material 2.pdf (22 kB)
Component matrix showing the habitat variables organized by their variable loadings on 3 principal components (PC1–PC4) derived from a rotated principal component analysis (PCA) using a Varimax rotation.

81.1.7 Supplementary Material 3.pdf (21 kB)
Independent detections of each species for each season (SP – spring, SU – summer, Fall, Winter) in each of 3 riparian forest types along the lower San Pedro River, Arizona, USA.

81.1.7 Supplementary Material 4.pdf (17 kB)
Average number of independent detections of mammals per 100 trap nights from 18 cameras deployed in 3 riparian forest types along the lower San Pedro River, Arizona, USA.

81.1.7 Supplementary Material 5.pdf (61 kB)
Mean and standard error of independent detections per 100 trap nights on San Pedro River, Arizona, USA. Plots include data from July 2017 to October 2018 for the families Canidae (Canis latrans and Urocyon cinereoargenteus) and Felidae (Lynx rufus and Puma concolor).

81.1.7 Supplementary Material 6.pdf (18 kB)
Real estimates for occupancy (psi) and detection probability (p) and 95% confidence intervals for each species during the summer season along the lower San Pedro River, Arizona, USA.

81.1.7 Supplementary Material 7.pdf (32 kB)
Models of occupancy (psi) for the summer season in 3 riparian forest types along the lower San Pedro River, Arizona, USA.

81.1.7 Supplementary Material 8.pdf (18 kB)
Two-species occupancy models depicting occupancy (psi) and detection probability (r) for mountain lion (dominant species) and bobcat (subordinate species).

81.1.7 Supplementary Material 9.pdf (18 kB)
Two-species occupancy models depicting occupancy (psi) and detection probability (r) for coyote (dominant species) and gray fox (subordinate species).

81.1.7 Supplementary Material 10.pdf (289 kB)
Top row: Statistical power against overall effect size (amount of variance between null and top model) with an 80% chance of detecting the effect if it is true (Type II error). Bottom row: Statistical power over potential combinations of number of sites and replicate quantities.

Share

COinS