Simplifying the protocol for the quantification of generalized soil fertility gradients in grassland community ecology

Abstract

Aims

A new method, based on the latent variable measurement model, was recently developed to quantify generalised soil fertility (F_G) in grassland ecosystems; i.e. the productive capacity of a soil at a multispecific scale when non-soil variables are held constant. To predict F_G values, this method uses the relative growth rates (RGR) of Festuca rubra, Trifolium pratense, Triticum aestivum and Arabidopsis thaliana grown in intact soil.

We investigated three simplifications of this method; namely, (i) the use of sieved soil instead of intact cores, (ii) the use of a single measure of biomass production instead of RGR and (iii), a combination of these two modifications.

Methods

Using 26 grassland soils from southern Quebec (Canada), F_G values were predicted for each method and compared using Pearson correlation coefficients. We also evaluated the performances of these different F_G values, as well as measures of NO₃⁻, P and K in predicting aerial net primary production (NPP) of the vegetation in these sites in a common non-soil environment.

Result

The four methods had comparable and very consistent F_G values and, although they were not numerically equivalent, they were equivalent in predicting natural NPP of plant communities, and these predictions were better than the direct measures of NO₃⁻, K and P flux rates as measured by Plant Root Simulator probes.

Conclusions

Therefore, any of these methods could be used to assess the generalized fertility.