SEARCH

SEARCH BY CITATION

Keywords:

  • carbon sequestration;
  • land cover change;
  • Prosopis;
  • Santa Rita Experimental Range;
  • semidesert grassland;
  • shrub encroachment;
  • soil nitrogen;
  • soil organic carbon

Abstract

Recent trends of increasing woody vegetation in arid and semiarid ecosystems may contribute substantially to the North American C sink. There is considerable uncertainty, however, in the extent to which woody encroachment alters dryland soil organic carbon (SOC) and total nitrogen (TN) pools. To date, studies assessing SOC and TN response to woody plant proliferation have not explicitly assessed the variability caused by shrub age or size and subcanopy spatial gradients. These factors were quantified via spatially intensive soil sampling around Prosopis velutina shrubs in a semidesert grassland, using shrub size as a proxy for age. We found that bulk density increased with distance from the bole (P < 0.005) and decreased with increasing shrub size (P= 0.056), while both SOC and TN increased with shrub size and decreased with distance from the bole (P < 0.001 for both). Significant (and predictable) spatial variation in bulk density suggests that use of generic values would generate unreliable estimates of SOC and TN mass, and subcanopy SOC pools could be overestimated by nearly 30% if intercanopy bulk density values were applied to subcanopy sites. Predictive models based on field-documented spatial patterns were used to generate integrated estimates of under-shrub SOC and TN pools, and these were compared with results obtained by typical area-weighting protocols based on point samples obtained next to the bole or at a specified distance from the bole. Values obtained using traditional area-weighting approaches generally overestimated SOC pools relative to those obtained using the spatially integrated approach, the discrepancy increasing with increasing shrub size and proximity of the point sample to the bole. These discrepancies were observed at the individual plant scale and for landscapes populated by various shrub size classes. Results suggest that sampling aimed at quantifying shrub encroachment impacts on SOC and TN pools will require area-weighting algorithms that simultaneously account for shrub size (age) and subcanopy spatial patterns.