Hindgut haven turned sour: Habitat change as the driver of mammoth and horse declines in North America prior to human settlement?

Some of the arguments for climate mediated megafauna extinctions I find most interesting are made by Dale Guthrie, a palaeobiologist from the University of Alaska.

Guthrie believes megafauna in Alaska and the Yukon were hard hit by climate change long before people arrived. This view stems from a better understanding of ecosystem communities. In the past, ecologists believed that Pleistocene habitats and communities were the same we see today, only in different locations. Guthrie was one of the first to suggest that this may not be the case (Levy, 2011).

Pleistocene habitats were not simply modern ones in different locations: communities then were very different, with assemblages of organisms no longer found anywhere near each other.
.
Pleistocene Alaska was cold and dry; known as "mammoth steppe" (Levy, 2011). Few trees grew and grassland dominated. This habitat faded away at the end of the Ice Age, along with many of its inhabitants, including the woolly mammoth and rhino.

Bison and moose, however, prevailed (Guthrie, 2006). Why was that? Bison and moose are ruminants who ferment their food in a large forestomach called the rumen. Plants build their cell walls from cellulose; a substance only bacteria have evolved to digest. The rumen harbours cellulose munching bacteria which breaks down the plant matter before digestion. Plants also produce toxins designed to put off browsers. Fermentation occurring in the rumen enables these herbivores to cope with these toxins (Levy, 2011).  

However all this processing means that digestion is slow. Horses and elephants have another strategy: they also ferment their food, not in a rumen but in a pouch called a cecum, behind the small intestine. The cecum is further along the digestive tract and can therefore absorb proteins that would otherwise be broken down by microbes in a ruminant. So horses can make do on tough, old grasses - a diet which wouldn't support a ruminant (Levy, 2011).

The cold, arid plains of Ice Age Alaska had short growing seasons and lots of grass: perfect conditions for hindgut (cecum) fermenters like the horse and mammoth. But as the climate warmed, trees and shrubs colonised the landscape, beginning at the banks of the new meltwater streams (Levy, 2011). This vegetation was armed with toxins only a ruminant could digest. Guthrie believes these habitat shifts, occurring 14-15,000 years ago, drove the decline of horse and mammoth, long before the Clovis people arrived (Guthrie, 2006).

This idea is corroborated by the fact that the region's most recent horse fossils date from this window of time. During this same period, bison, elk and moose (ruminants) populations exploded and expanded north. Mammoths became less common at this time before finally vanishing 13,000 years ago (Levy, 2011). Guthrie sees the deterioration of the hindgut haven that was Ice Age Alaska as the driving force of megafauna decline, with humans possibly later dealing a final blow. From this he suggests that the Blitzkrieg, keystone removal and palaeodisease models for megafauna extinctions can be rejected, at least for the Alaska-Yukon area (Guthrie, 2006).