Soil health responses to long-term grazing intensity gradients in two semiarid rangelands

Semiarid rangelands constitute nearly 30 % of the world’s grassland ecosystems and livestock grazing is the most widespread land use in these ecosystems. These semiarid rangelands provide a variety of ecosystem goods and services, many of which may depend on soil health. While advances have been made with indicators of soil health for croplands, similar efforts for rangelands are lacking. The North American Project to Evaluate Soil Health Measurements (NAPESHM) sampled soils to 15 cm from long-term (≥ 40 years) grazing treatments spanning a range of grazing intensity (ungrazed to heavy grazing) in two semiarid rangelands, shortgrass steppe (SGS) and northern mixed-grass prairie (NMP), in fall 2019. Soils were analyzed for chemical (permanganate oxidizable carbon [POXC] and soil organic carbon [SOC]), biological (mineralizable soil carbon [MinC], phospholipid fatty acids [PLFA], ACE protein, and β-glucosidase enzyme activity [BG]), and physical (saturated hydraulic conductivity [SHC], available water capacity [AWC], and aggregate stability) indicators of soil health. Light particulate organic carbon and mineral associated organic carbon fractions were also analyzed at the SGS. Additionally, annual net primary productivity and the relative production of warm-vs cool-season grasses were evaluated from 2010 to 2019. Soil health responses to grazing intensity treatments in these two semiarid rangeland ecosystems were generally inconsistent across and within chemical, biological, and physical indicators. For instance, POXC and SOC differed between the two rangeland ecosystems, but neither soil health response within a site was significantly affected by grazing intensity. MinC and saturated hydraulic conductivity consistently decreased as grazing intensity increased in both rangeland ecosystems, while all other biological and chemical indicators were either 1) solely influenced by rangeland ecosystem type, 2) the interaction between rangeland ecosystem and grazing intensity, or 3) unaffected by rangeland ecosystem or grazing intensity. At SGS, δ13 C values of both organic carbon fractions became less negative as grazing intensity increased, consistent with a greater proportion of warm-season perennial grasses and lower proportion of cool-season grasses. Our results suggest that generalizations about the effects of multi-decadal grazing intensity gradients in western Great Plains semiarid rangelands in North America on chemical, biological, and physical indicators of soil health remain elusive. • Soil carbon was affected by rangeland ecosystems (RE), not grazing intensity (GI). • Aboveground primary production (ANPP) decreased as GI increased at both RE. • Inconsistent responses to GI and RE within biological soil health indicators (SHI). • Inconsistent responses to GI and RE within physical SHI. • Only two of the nine SHI observed followed ANPP trends with GI across RE.