Sampling for Soil Carbon Stock Assessment in Rocky Agricultural Soils.

Coring methods commonly employed in soil organic C (SOC) stock assessment may not accurately capture soil rock fragment (RF) content or soil bulk density (BD) in rocky agricultural soils, potentially biasing SOC stock estimates. Quantitative pits are considered less biased than coring methods, but are invasive and often cost-prohibitive. We compared fixed-depth and mass-based estimates of SOC stocks (0.3-m depth) for hammer, hydraulic push, and rotary coring methods relative to quantitative pits at four agricultural sites ranging in RF content from <0.01 to 0.24 m³ m^-3. Sampling costs were also compared. Coring methods significantly underestimated RF content at all rocky sites, but significant differences (p < 0.05) in SOC stocks between pits and corers were only found with the hammer method using the fixed-depth approach at the <0.01 m³ m^-3 RF site [pit 5.80 kg C m^-2, hammer 4.74 g C m^-2] and at the 0.14 m³ m^-3 RF site [pit 8.81 kg C m^-2, hammer 6.71 kg C m^-2]. The hammer corer also underestimated BD at all sites, as did the hydraulic push corer at RF 0.21. No significant differences in SOC stocks estimated using the mass-based approach were observed between pit and corers. Our results indicate that: 1) calculating SOC stocks on a mass basis can overcome biases in RF and BD estimates introduced by sampling equipment, and 2) a quantitative pit is the optimal sampling method for establishing reference soil masses, followed by rotary, then hydraulic push corers.