Tony Branham, the district conservationist with the Natural Resources Conservation Service, has done his homework.
To show soil absorption properties of no-till vs. conventional till he took samples of Genesee soil, a bottom-land soil that tends to be well drained. Fortunately for him, two farmers owned part of the same piece of land he sampled. One of them believed in tillage every year, the other no-tilled every year for the past several years, flood plain or not.
Branham took his samples, then went into the woods and took a sample of undisturbed soil of the same soil type. As far as he knows, it had never been farmed or turned over.

Branham did this experiment for a group of farmers from south-central Indiana, all the way from Franklin to Bedford and many points in between. He placed the chunks of all three types of soil into three separate columns of water. They were suspended in small wire baskets.
The tilled soil was the first to break apart. Particles began crumbling off and fell to the bottom. The no-till soil lost some soil particles, but not many. The undisturbed spoil remained intact.
No-till acted most like the undisturbed soil, Branham notes. The reason for the breakdown of the conventional soil was that it had no way to soak up water. There were very few pores or worm channels to let water in. Instead, the water pressured it to dissolve and fall to the bottom as sediment. Soil particles are the number one sediment found in streams and lakes.
The no-till soil and the undisturbed soil absorbed water. In each case there were natural channels and pores that allowed the water to infiltrate and stay within the soil mass.