Nonlinear growth curves of sorghum roots under compost/traditional fertilization

Plant roots are a major pool of total carbon in the planet, and their dynamics are directly relevant to greenhouse gas balance. Composted wastes are increasingly used in agriculture for environmental and economic reasons, but their role as a substitute for traditional fertilizers needs to be tested on all plant components. Compost application was compared to traditional fertilization for its effect on roots of Sorghum bicolor Moench x S. sudanense (Piper) Stapf. in a three-year experiment (2007-2009) in Southern Italy. Plant roots were monitored through sequential images taken with a digital microscope from 4 transparent acrylic access tubes per treatment, buried in the soil at 45° angle up to the soil depth of 60 cm from the surface. Roots on each 166 mm x 124.5 mm image were analysed and attributed to three categories: white, dark and gone. Total roots length, and surface area and average roots diameter were measured on each frame for the three root categories. Roots measurements distributions were assumed to follow an exponential dispersion model, namely the Tweedie distribution, in order to account for zero-inflation in the relative data. Generalized Additive Models (GAM’s) were used to evaluate the nonlinear longitudinal variation of roots measurements along “days after sowing”, together with the possible effects of fertilization treatment, depth and year and eventually adjusting for the confounding effect of the atmospheric temperature. GAM’s parameters were estimated by penalized likelihood maximization with smoothing degree given by generalized cross validation minimization. Residuals and likelihood ratio tests were used for model validation and comparison