Non-invasive Methods to Assess Co-occurrence of Mammalian Carnivores

During 2003–2005, I surveyed 82 sites on the North and South rims of Grand Canyon, Coconino County, Arizona, to test effectiveness of four non-invasive techniques for detecting carnivores and to assess patterns of co-occurrence among pairs of species. Techniques were not equally effective for detecting carnivores. Searches of transects for feces, tracks, and other evidence yielded the greatest number of detections; remotely triggered cameras and track plates had the greatest probabilities of detecting common species and also produced detections of smaller and rarer carnivores; and hair traps generally were ineffective. Even after accounting for variation in probabilities of use of habitats by species due to characteristics of sites, two pairs of carnivores had limited co-occurrence. Coyotes (Canis latrans) and bobcats (Lynx rufus) did not co-occur at sites on the North Rim, and coyotes and gray foxes (Urocyon cinereoargenteus) co-occurred less than one-half as frequently as expected in the South Rim. RESUMEN—Entre 2003 y 2005, muestreé 82 sitios en los bordes norte y sur del Gran Canyon, condado de Coconino, Arizona, para probar la eficacia de cuatro técnicas no-invasivas para detectar carnı́voros mamı́feros, y evaluar los patrones de coocurrencia entre pares de especies. Las técnicas no fueron igualmente efectivas para detectar carnı́voros. Búsquedas de transectos para excrementos, huellas y otra evidencia produjeron el número mayor de detecciones; cámaras de disparador-remoto y superficies preparadas para registrar huellas tuvieron la mayor probabilidad de detección de especies comunes y también produjeron detecciones de carnı́voros pequeños y raros; y trampas de pelo generalmente fueron inefectivas. Aún después de tener en cuenta la variación en la probabilidad del uso de hábitat de especies debido a caracterı́sticas del sitio, dos pares de carnı́voros tuvieron coocurrencias limitadas. Coyotes (Canis latrans) y gatos montés (Lynx rufus) no coocurrieron en sitios en el borde norte, y coyotes y zorros grises (Urocyon cinereoargenteus) coocurrieron con frecuencia menos de la mitad de lo esperado en el borde sur. Co-occurrence of species is important to consider when surveying for and modeling distributions of mammalian carnivores. The majority of models used to predict distributions of species are developed solely using characteristics of habitats and for one species at a time (Scott et al., 2002). However, biotic interactions in communities may influence results of field surveys and predictive models should consider the likelihood that detecting one species depends on presence of another (MacKenzie et al., 2006). Prior research on diets (Fedriani et al., 2000; Neale and Sacks, 2001a; Carvalho and Gomes, 2004) and movements of carnivores (Fedriani et al., 1999; Neale and Sacks, 2001b; Constible et al., 2006) indicates limited coexistence and partitioning of resources among species, especially at fine spatial scales (Scognamillo et al., 2003). Researchers have employed several approaches to examine co-occurrence of carnivores, including analysis of use of space by radiocollared individuals (Neale and Sacks, 2001b), comparison of probability derived from predictive-occurrence models (Alexander et al., 2006), and null-model analysis of presence-absence matrices (Gotelli, 2000). Null models are used to generate random patterns of composition and distribution of species that would be expected in absence of interactions among species. These randomized distributions can then be compared with empirical data to assess how likely the observed patterns THE SOUTHWESTERN NATURALIST 56(2):231–240 JUNE 2011