Islands of Fertility: A Byproduct of Grazing?

Abstract

Listen

Translate article

Empirical work in arid shrubland systems has documented a distinct spatial pattern of soil nutrient distribution, with higher concentrations of nutrients under shrub canopies compared to bare ground interspaces between shrubs. This “fertile island” pattern is considered characteristic of arid ecosystems. However, recent work at a desertified shrubland site reported no island of fertility pattern for nitrogen inside a long-term grazing exclosure despite a strong fertile island pattern outside the exclosure, leading to a hypothesis that the pattern may be associated with grazing. In this paper we examine this hypothesis in two ways. First, we collected soil nitrogen data inside and outside two long-term livestock exclosures. Second, we examined published work on the fertile island pattern for nitrogen, phosphorus and carbon, and its association with livestock grazing. At the field sites, there was a strong fertile island pattern for soil nitrogen in the grazed area outside each exclosure but no fertile island pattern in the ungrazed area inside the fence. From the literature, we identified 49 studies, which represent 99 cases that examine differences in soil nutrients under and between shrubs (that is, tests of the fertile island pattern). In 92 cases, fertile islands were present for at least one of the nutrients. All cases where livestock grazing was present exhibited fertile island patterns for nitrogen and most did so for carbon and phosphorus. In contrast, fertile islands were less consistent where livestock grazing was absent; all nine cases that had been ungrazed for more than 30 years did not exhibit fertile islands for nitrogen. Our empirical data and literature review suggest that the fertile island pattern may be an artifact of grazing rather than an inherent property of arid shrublands. Further, we suggest that the grazing history of a site needs to be explicitly considered when documenting spatial patterns of soil nutrients in arid systems. This has significant implications for our understanding of soil nutrient dynamics in arid systems, the management of rangelands, and the potential for reversal of desertification.