Image analysis systems have facilitated rapid measurements of fine root length (RL), diameter (RD), volume (RV), etc. The WinRHIZO system is unlike other image analysis systems in that it can detect, and make corrections for, areas of root overlap. It is designed to be capable of using both Regent’s non-statistical method (WinRHIZO method) and Tennant’s statistical method (line-intersect method), and can simultaneously output the root measurements by both methods when they are chosen at the same time. This study tested: (1), the efficacy of the overlap correction function in the WinRHIZO system; and (2), the consistency of fine root measurements between the WinRHIZO and Tennant methods with two sets of root measurement data from winter wheat ( Triticum aestivium L.). The results showed that there were significant differences in RL, RD and RV between small root samples with and without stumps. The impact of root stumps outweighed the overlap correction efficacy in WinRHIZO. The values from the Tennant method are significantly different from those using the WinRHIZO method, although both results are statistically closely correlated. This indicated how critical it was to use without-stump root samples when using image analysis systems to measure RL, RD, RV, etc., and to keep in mind that a significant difference in root measurements may be methodologically related when comparing the results of various experiments from these two methods. Our research results bear important implications for the study of root ecology.
Read full abstract