Every now and then it occurs to me that I should document some of the hypotheses in my mind that might make interesting research questions at some indefinite point in the future. Here’s one:

If the historical reference community for a given site was a patchy grassland, with alternating dense perennial grass patches and sparsely vegetated patches (usually these are more or less linear or arcuate and perpendicular to the slope), grazing that is at a level that reduces perennial grass cover while still leaving perennial grasses dominant will also increase the size of both the grass patches and the sparsely vegetated patches (the latter, presumably, increasing more).

And another hypothesis, closely related to the above:

In an area with two ecological sites / topographic units, both having perennial grassland as the historical reference community but one typified by much denser grassland than the other (an example would be a draw surrounded by loamy plains), reductions in perennial grass cover caused by grazing will be greatest immediately adjacent to the more densely vegetated ecological site. This causes a long-term pattern in which what would have been a more or less bell-curve-shaped distribution of grass density in a cross section of the draw and its surroundings into a trimodal distribution, with a tall central peak and two small but broad peaks at some distance (presumably set by factors like terrain complexity, climate / water availability, forage availability relative to stocking rate, and so forth) from the draw. The tall central peak would be the next to diminish and eventually disappear. At one extreme, in a broad valley surrounded by low rocky hills, this can create a situation in which grass cover is inversely proportional to soil moisture; grasses remain on the rocky slopes where runoff is high and water-holding capacity relatively low while being eliminated from the more finely textured soils and shallower slopes near the valley bottom, and from the valley bottom itself.

As a result of this, utilization of grasses by livestock may be inversely proportional to grass cover, even in sites relatively near water with limited overall forage availability, if slopes adjacent to a drainage are steep or merely rocky. At the extreme of that phenomenon, you may have slight or no utilization of perennial grasses under heavy grazing pressure. This is the landform equivalent of seeing perennial grasses only within shrubs and can produce a map of forage availability and forage utilization that, read naively, would indicate minimal grazing pressure and relatively abundant forage.