Here is another article about people studying the ecological benefits of gophers:
The Nature Conservancy in Washington State is studying Mazama pocket gophers by catching them and inserting electronic tags under the critters' skin (similar to the chips used for pets). This will allow them to learn more about the range of individual gophers.
O.J. Reichman and Eric W. Seabloom have also published an article on the pocket gopher in Ecology and Evolution. (January, 2002) in which they too noted the soil-moving effects of the gopher. They observed that the gopher burrows accelerated erosion on shallow slopes and slowed erosion on steep slopes. I have not noticed increased erosion on my gently sloping land. In fact, the gopher borrows encourage the water to penetrate the soil, rather than running off. The run-off causes erosion. When the water soaks into the soil instead of running off, none of the soil is washed away.
Another article Lisa F. Cantor and Thomas G. Whitham in Ecology discusses the gopher's role in preventing aspen invasion of mountain meadows.
Grasses, incidentally, tend to benefit from root pruning by gophers (I have read this and observed it first-hand), so gopher activity would encourage prairie landscapes in regions that do not provide ideal growing conditions for trees.
This article by Jake Weltzin, Steve Archer, and Rod K. Heitschmidt looks really interesting: "Small Mammal Regulation of Vegetation Structure in a Temperate Savannah," Ecology: Vol. 78, No. 3, pp. 751-763. This article focuses on the prairie dog rather than gophers. Since prairie dogs don't live here, I don't have any first-hand knowledge of them, but they apparently eat roots and seeds and live in burrows, so their ecological niche would be similar to that of the gophers.
I think I may have to buy a few of these articles so I can read the entire articles rather than just the abstracts. Here is an excerpt from the abstract of the Weltzin, Archer, and Heitschmidt article:
These data illustrate how transitions from grassland to woodland vegetation can be mediated by a rodent herbivore. They further demonstrate how purposeful or inadvertent removal of native herbivores can have unforeseen effects on plant species composition and landscape physiognomy. Investigations of environmental constraints on vegetation distribution and abundance should take into account the historical role of herbivores in shaping the present system. Inconsistencies among historic accounts of woody plant distribution and abundance in semiarid western North America may be resolved by considering population dynamics of prairie dogs. Widespread eradication of this formerly abundant rodent has eliminated a significant constraint to woody plant establishment on many semiarid grassland and savanna landscapes and has thereby facilitated transitions to shrubland and woodland states. Past land management designed to remove one perceived impediment to livestock production appears to have contributed significantly to development of another management problem that is now a major detriment to sustainable livestock production.