Looks like specific practices, rather than the organic/conventional dichotomy, determine outcomes for soil health (measured in terms of soil biota).

... which is that if you forego a harvested crop to grow green manure, you're lowering average yields in a way that increases the land footprint of farming. To offset this, you would have to grow crops elsewhere, and that might lead to N emissions.

I myself am using GYGA (paired with FAO data) to try to understand if recent signs of yield stagnation in cropping systems around the world are due to yields hitting a ceiling, or to inadequate management. Stay tuned!

GYGA now covers a substantial share of crop production for important staple crops like wheat, rice, soy, and corn.

This problem is getting smaller and smaller, however, as the Global Yield Gap Atlas (www.yieldgap.org) adds to its database of potential yields and yield gaps.

Why don't we just use bottom-up methods? The reason is that their main strength--validation with local data--is also their main weakness: it is hard to scale up.

Earlier papers, such as this one by van Ittersum et al (www.sciencedirect.com/science/arti...), have also shown that results from top-down models can substantially over- or under-estimate potential yields at local to regional levels.

This latest study, titled "Statistical approaches are inadequate for accurate estimation of yield potential and gaps at regional level," shows that top-down methods are unreliable assessed against better-validated bottom-up methods.

If, on the other hand, these top-down methods can't distinguish between rainfed and irrigated cropping systems, they might, wrongly, infer that rainfed systems could reach the yields of irrigated systems. This leads to an overestimation of potential yields.

There are many problems with top-down methods, however. For example, even the best farmers in a region can be well below the real yield potential given soils, climate, and the best available agricultural technology. This would lead to underestimating potential yields.

The advantage of top-down methods is that they are easy to scale: in the simplest case, with some climate data and yield statistics from FAO, you can create a map of potential yields across the world.

Top-down methods typically take potential yields to be some percentile of current, actual yields within each climate zone. Bottom-up methods use field data combined with crop growth models and weather data to estimate potential yields.

Knowing the potential yields of crops is important, because it tells us something about how far we can get in meeting food demand without technological breakthroughs. There are two ways of estimating potential yields: top-down and bottom-up.

Total production goes down, but the marginal land yields so little in relation to inputs that it doesn't generate a net profit. Also shows the advantage of precision ag technologies that can pinpoint yields with very high spatial resolution.

Here's the Cafaro paper: overpopulation-project.com/wp-content/u...

My take here (with @jennbernstein.bsky.social): linusblomqvist.substack.com/p/beyond-the...

It's true, it was just never a big share. I'm not sure I'll publish the paper - issues like small sample size that could prevent publication in a good journal. So its findings should be taken as suggestive. There is anecdotal evidence (quotes, newspaper articles) that substitutes played a big role.

Worth noting that the countries that didn't have state-run or heavily state-subsidized whaling all basically stopped whaling long before any international regulations came into place.