This reminds me of a research project I did in high school. I was using a handful of Rhizobia species and a number of nutrients, attempting to find some non-nitrogen and non-root nodule substance which could promote Rhizobia growth in soil, partially to help reduce reliance on nitrogen heavy fertilizers. Ultimately the research didn't pan out, with no statistical different in any of the samples.

Turns out the solution was something like a locally applied CRISPR to force more of the roots to become nodulated (word?) and form the Rhizobia symbiosis.