Estimating Root Production: Comparison of 11 Methods in Shortgrass Steppe and Review of Biases

Estimating root production has been difficult due to multiple potential biases associated with both old and new methods. This shortgrass steppe site is the only place we are aware of that can compare most methods including sequential coring, ingrowth cores, and ingrowth donuts, C-14 pulse-isotope dilution, C-14 pulse-isotope turnover, rhizotron windows, and minirhizotron, and indirect methods including nitrogen budget, carbon flux, simulation carbon flow model, and regression model. We used the studies at this site, other comparisons, a summary of potential directional biases, and different ways of calculating estimates in a logical, comparative approach of evaluating methods. Much of the literature for root production is based on sequential biomass coring, a method resulting in erroneous estimates. Root ingrowth estimates of production are generally conservative compared to minirhizotron and isotope turnover methods. The size of the ingrowth area may be the most important determinant of the underestimation. Estimates based on pulse-isotope dilution are also erroneous due to non-uniform labeling of tissues. Uniform labeling is not an assumption of the pulse-isotope turnover method, and this method has the least severe potential biases. Root production estimates from pulse-isotope turnover were lower than those using minirhizotron when the most common method of calculation was used. This agrees with literature concerning bomb C-14 continuous-isotope labeling comparisons with minirhizotron, although some potential biases between isotope methods are different. However, good agreement between pulse-isotope turnover and minirhizotron were obtained when minirhizotron estimates were calculated from regression of decomposition versus production to equilibrium and when pulse-isotope turnover estimates were calculated from two-phase life-span regressions. This minirhizotron method bypasses biases associated with the artificial surface similar to root-cohort methods that may be practical only in mesic systems, and takes into account both short- and long-lived roots and corrects for soil-isotope contamination that the continuous-isotope labeling bomb C-14 method is not able to account for. Comparisons of these direct methods are also made with four indirect methods.