Basing fertilizer recommendations on things other than soil tests?

Yesterday (Friday, Dec 7th), I attempted to host a webinar to discuss options and ideas about what to base fertilizer recommendations on other than soil tests. I had a terrible experience with GoToMeeting which I think is actually to blame on a stupid McAfee pre-install that I had not yet removed. Anyways, by the time I got it figured out I only a had four attendees sticking around. Thanks so much for your patience!

There are a number of things more influential on yield than soil tests (based on my research.) In the webinar I showed an excel graph of some data that we had gathered from a subset of the 350,000 acres we work with. Here is what we did... a yield query around ever sample site. Then, we plotted each soil test value and yield in a scatter graph and looked for regression lines. It looked pretty dismal. (insert actual image here!)

What this tells me is that although GPS soil testing is one of the most popular precision ag practices it is a far cry from making a significant impact on yield. For that reason I often advise taking some other things into consideration. Here are just a few:
1. What is your fertilizer budget?
2. What are your site specific yield goals?
3. What do you believe critical levels will be in the future?

With some dialogue a good advisor should be able to educate you about normal parameters and ideas that help you achieve your goals. Once that is in place your variable rate recommendations might have a greater influence derived from yields than actual tests. Or, perhaps your budget trumps your soil test?

When I think about the future, I get excited pondering software that makes all metrics of crop growth relative to each other. Lieberg Law of the minimum should trump any one particular attribute's relation to a perceived "standard" or "safe" range.

I'm also excited about ability to take a whole farm budget and distribute input products around like making a peanut butter sandwich... you spread in the divots first and then put a layer down across the whole piece.

We have gobs of data, now we just need software to automate the operating decisions that align with our goals. Happy farming!