Soil carbon and nitrogen changes following root-plowing of rangeland

The effects of root-plowing on soil organic carbon and nitrogen were invest igated by comparing paired undisturbed native rangeland with root-plowed si tes in the southern Great Plains. Time since root-plowing ranged from 4 to 22 years. We hypothesized that following root-plowing (1) soil carbon would initially drop hut recover to the level of untreated range within a 5-10 y ear period, and (2) the permanent removal of mesquite trees, which enhance ecosystem carbon and nitrogen and provide shade that lowers soil temperatur e, would result in a slow decline in soil carbon and nitrogen in this ecosy stem. There were not significant differences due to treatment for either soil car bon mass (g m(-2)) (P=0.81) or nitrogen mass (P=0.62). There were significa nt differences in soil carbon mass (P=0.0014) with respect to elapsed time since plowing. The upper soil layer (0-100mm) had higher carbon levels (P=0 .0001) than the deeper soil layer (100-200mm)(1422 +/- 210 g m(-2) vs. 1111 +/- 206 g m(-2)). Differences in soil nitrogen were similar to those of so il carbon. There were significant differences in nitrogen among years-since -root-plowing observations (P=0.003) and the upper soil layer had higher ni trogen levels than the deeper soil layer (138 +/- 18 g m(-2) vs. 107 +/- 18 g m(-2)(P=0.0001). When the data were analyzed using paired native site values as a covariate to account for site differences, the sites that had been root-plowed 4 year s previously had higher soil carbon (P<0.08) and nitrogen (P<0.09) than the sites root-plowed 11, 16, and 22 years previously. These results are the o pposite of what was hypothesized. This is probably due to root-plowing bein g a nonrecurring treatment that did not invert the soil or remove the peren nial grass cover. The slight increase in carbon measured 4 years after root -plowing was possibly caused by the large amount of dead tree roots in the soil after plowing. This would immediately increase the total amount of dea d plant material entering the decomposing pool, elevating carbon levels tem porarily before they returned slowly to previous levels. There was no trend of decreasing soil carbon or nitrogen over the 22 year period covered. It does not appear that removal of mesquite trees changes soil carbon or nitro gen levels in this ecosystem relative to native rangeland with mesquite tre es.