The stress gradient hypothesis (SGH) predicts that facilitation and competition vary inversely along stress gradients with facilitation more frequent and stronger at high stress. However the relationship depends on the type of stress, life history strategies, and location of interacting species relative to their ecological optima. At the community level, facilitation is predicted to increase stability at intermediate stress levels but decrease it at high stress levels and may reduce invasibility by increasing the abundance or diversity of resident species, which then provide resistance to invasion. We tested the SGH across 75 study sites located along complex stress gradients created by overlapping non-resource (heat, cattle herbivory) and a resource (water) based stress gradients. We examined pair-wise interactions between the foundational shrub Artemisia tridentata wyomingensis (benefactor) and eight focal herbaceous species (beneficiaries) that co-occur, but differ in their ecological optima along the stress gradients. The focal species included six native and two non-native species representing a range of competitive responses and stress tolerances. We inferred interaction outcomes from spatial patterns of association by comparing focal species cover beneath Artemisia canopies and in adjacent interspaces. We assessed whether findings translated into meaningful effects on community stability.
Results/Conclusions
We identified three groups of study sites that differed in combined stress (MRPP, A = 0.24, p < 0.0001) and spatial patterns of association at the community level (A= .19, p < .0001). Facilitation was strongest and most frequent at highest stress levels. Contrasting ecological optimum among the natives and non-natives led to strikingly different patterns: competition was strongest for natives at lowest stress, but for non-natives, competition was strongest at highest stress. At intermediate stress, facilitation was observed for two native species; however, only one native species was uniquely associated with the under-shrub community, there was a positive relationship between the under-shrub and site-level communities (Mantel statistic (R) = 0.71, p < .001), and the community was dominated by native species. At high stress, facilitation was observed for all native species with four native species uniquely associated with the under-shrub community; there was no relationship between the under-shrub and site-level communities (R =.06, p = .52); the community was dominated by non-native species. Artemisia facilitation enhanced stability at intermediate stress by providing a refuge for native species, which resisted invasion. At high stress, facilitation was destabilizing when natives became “obligate” beneficiaries and dependent on facilitation for their persistence.