Soil carbon and nitrogen in relation to shrub size and death in a semi-arid grassland

Nutrient accumulation as fertile islands beneath invasive trees and shrubs in grasslands may provide opportunities for carbon sequestration. In a southwestern USA grassland, our objectives were to describe 1) the accumulation beneath Prosopis velutina (velvet mesquite) and isolated grass plants, and 2) the loss of accumulated nutrients 40 y after P. velutina death. We compared organic carbon (OC), total nitrogen (TN) and δ13C in soil organic matter among large living, large dead, and small living P. velutina and open grassland, and between grass plants and bare ground. Soil samples were collected at 0–5, 5–10, 16.8–23.2, and 36.8–43.2 cm depths, and separated into five size/density fractions: particulate organic matter (Macro- and Micro-POM), mineral associated organic matter (Micro-MAOM), Silt, and Clay. We expected that differences in OC, TN, and δ13C among soil fractions would suggest mechanisms and rates of accumulation with P. velutina persistence and loss following P. velutina death. Soil OC and TN accumulation was ~ 80–750% greater for large P. velutina than open grassland for whole soil and 4 of 5 fractions at 0–5 cm depth, but only 50–250% greater at 5–10 cm depth for whole soil and 3 of 5 fractions. Total OC and TN accumulation at 0–10 cm depth was 6.12 kg C m− 2 and 0.55 kg N m− 2, respectively. Accumulation did not occur in whole soil or any fraction at 16.8–23.2 cm and 36.8–43.2 cm depths, or in the Micro-MAOM fraction at any depth. Accumulation under small P. velutina was less than large plants, and not significantly different from open grassland. Beneath isolated grass plants, accumulation of TN occurred at 0–5 cm depth, and OC accumulation at 0–5 and 5–10 cm depths in whole soils only, but change in δ13C did not accompany accumulations. Forty years after death of large P. velutina, 67–106% of accumulated OC and TN were lost from whole soil and soil fractions at 0–5 cm depth. At 5–10 cm depth, loss (78–93%) was only detected in whole soils. Greater accumulation of OC and TN in the POM than the Silt and Clay fractions is consistent with the large physical size of recent organic matter inputs from P. velutina, but no differences in loss rates among fractions following P. velutina death suggests density dependent rates of organic matter consumption. Declines in δ13C accompanied OC accumulation and increases occurred during loss. A 20–30 y mean residence time (MRT) for whole soil and the Clay fraction over those 40 y is suggested by changes in OC and δ13C, but MRT based on changes in OC and δ13C differ for Macro-POM (1–3 y versus 20–25 y, respectively) and Silt (> 50 y versus 20–30 y, respectively). Activities that remove P. velutina should focus on small plants and protect large plants thereby maintaining their carbon sequestration potential.