Using radius frequency distribution functions as a metric for quantifying root systems

Abstract

Root radius frequency distributions have been measured to quantify the effect of plant type, environment and methodology on root systems, however, to date the results of such studies have not been synthesised. We propose that cumulative frequency distribution functions can be used as a metric to describe root systems because (1) statistical properties of the frequency distribution can be determined, (2) the parameters for these can be used as a means of comparison, and (3) the analytical expressions can be easily incorporated into models that are dependent upon root geometry. We collated a database of 96 root radii frequency distributions and botanical and methodology traits for each distribution. To determine if there was a frequency distribution function that was best suited to root radii measurements we fitted the exponential, Rayleigh, normal, log-normal, logistic and Weibull cumulative distribution functions to each distribution in our database. We found that the log-normal function provided the best fit to these distributions and that none of the distribution functions was better or worse suited to particular shapes. We derived analytical expressions for root surface and volume and found that they are a valid, and simpler method for incorporating root architectural traits into analytical models. We also found that growth habit and growth media had a significant effect on mean root radius.