Spectroscopic and Foliar pH Model for Yield Prediction in a Symbiotic Corn Production

The agronomic management of symbiotic (S) inoculations, by means of bio-fertilizers (BF), is aimed at inducing modifications of the plant rhizosphere and thereafter of the phenotype and yield of the crop. It is here shown that the yield response of maize to a symbiotic treatment may be correlated to six easy-to-calculate indicator variables on the basis of the raw foliar pH, NIR-Spectroscopy of leaves, and the NIRS of hay litter-bags from soils. It has been confirmed, in a set of thirteen pairwise comparisons of Symbiotic (S) soil inoculated by BF vs. Control (non-inoculated soil; C), that the inoculation on average acidified the leaves by -3.7% pH units (P<0.0001). The responses in yield ranged from +25.2% to -9.2% (av.ge +3.5%; P = 0.03), but with average null responses in two centers and a significant response (+11%) in a third center. NIR-Tomoscopy scans (No. 574) were also performed on the leaves, and in addition, hay-litter-bags that had previously been buried in fields were dug up after two months, and 431 NIR- scans were acquired. The effect-size on the yield was expressed as the logarithm of the response ratio, i.e. the mean of the inoculated Symbiotic treatment divided by the mean of the non-inoculated Control for each pairwise comparison. A multiple regression model was developed to predict the symbiotic response to the treatment using six independent variables, including the squared litter-bag fingerprints, and an R2adj. level of 0.78 (P=0.01) was reached, with a standard error of ±4%. Validation in one external maize field, with a positive response to bio-fertilizers, demonstrates the juxtaposition of the estimated and accomplished yield. In a second experiment, with 40 pairwise comparisons, the two tested maize varieties did not respond to five types of bio-fertilizer, and the negative results were predicted at 84% (P 0.0012). The soil biota is a key factor for the application of appropriate microbial inoculants in the field, but the genotype/genotype interactions between the microbial strain (s) and the crop cultivar (s) require prior screening to obtain the desired results