Detecting change in belowground carbon stocks: Statistical feasibility and future opportunities from loblolly pine forests

The substantial magnitude of belowground carbon (C) stocks suggests that even minor proportional changes in soil C concentrations could significantly affect global carbon budgets. However, measuring stock changes with sufficient statistical confidence proves difficult due to soil C heterogeneity. In the southeastern US, planted loblolly pine forests have extensive potential to manifest background C fluxes through monitoring networks and to modify C flows under intentional management regimes. To understand feasibility of estimating belowground C stocks, we evaluated current knowledge of and ability to detect change in belowground C subpools (i.e., coarse woody material, O horizon, mineral soil, detritus, fine roots) in planted loblolly pine stands by compiling statistical distributions from existing databases. We performed power analyses to quantify minimum detectable change in each subpool given current sample sizes, as well as minimum sample sizes required to detect desired magnitudes of change with adequate statistical confidence. The mineral soil subpool could report the smallest magnitude of change (12.9 %, power = 0.8, α = 0.05), and smaller changes could be detectable at individual depth increments. No subpool had enough measures to detect change < 10 % across the region (power = 0.8, α < 0.2). At minimum, tens of thousands more samples (>2200 % sampling increase, power = 0.8, α ≤ 0.2) would be needed to detect the smaller stock changes (i.e., 2 %) across the species’ native range in which current monitoring and market frameworks are interested. Measurement gaps highlight pools for prioritized monitoring. Lack of statistical power associated with all pools necessitates greater data collection, reconsideration of statistical conventions, or examination of alternative approaches to conclusive change detection.