The geologic diversity of landforms in the Southwest complicates efforts to evaluate impacts of land uses such as livestock grazing. We examined a research study that evaluated relationships between trout biomass and stream habitat in the White Mountains of east-central Arizona. That study interpreted results of stepwise regressions and a nonparametric test of ''grazed and ungrazed meadow reaches'' as evidence that livestock grazing was the most important factor to consider in the recovery of the Apache trout (Oncorhynchus apache Miller). That study had assumed that geologic variation was insignificant in the study area. However, lithologic and topographic differences between the felsic slopes of Mount Baldy and adjacent mafic plateaus influence many attributes of trout habitat. We tested the robustness of the earlier study by using its dataset and its method of stepwise regression, but with the addition of a variable representing geologic variation. The results suggested that geology was a highly significant predictor of trout biomass (P , 0.0001), whereas bank damage by ungulates was not a useful predictor of residual variation in trout biomass after accounting for geology (r 2 ¼ 0.015, P ¼ 0.290). However, the associations between natural variation and land use impacts in this spatial dataset confound attempts to make inferences concerning effects of livestock grazing upon trout. Despite fundamental problems in the analysis, the results of the earlier study were repeatedly cited in scientific literature and debates about grazing management. To fairly decipher relationships between ecological production and livestock grazing in diverse landscapes requires temporal studies with reliable methodologies and proper controls for landscape variation. Ignoring geologic variation has the potential to mislead conservation policies by inappropriately implicating land use, by undervaluing inherently favorable habitats, and by inflating expectations for inherently less favorable habitats.
ResumenLa diversidad geoló gica de los paisajes del Sudoeste Americano complica los esfuerzos para evaluar los impactos de los usos del suelo, como el apacentamiento del ganado. Examinamos una investigació n que evaluó las relaciones entre la biomasa de truchas y su há bitat ribereñ o en las White Mountains, de la region este-central de Arizona. Ese estudio interpretó resultados de regresió n gradual y una prueba no paramétrica de ''praderas apacentadas y no apacentadas,'' como evidencia de que el apacentamiento del ganado era el factor más importante a considerar en la recuperació n de la trucha ''Apache'' (Oncorhynchus apache Miller). Ese estudio había asumido que la variació n geoló gica era insignificante en el á rea del estudio. Sin embargo, las diferencias litoló gicas y topográ ficas entre las pendientes félsicas del Mount Baldy y las mesetas má ficas adyacentes influencian muchos atributos del há bitat de la trucha. Probamos la robustez del estudio anterior usando sus datos y su método de regresió n gradual, pero con la ad...