Two recent papers that highlight this issue in an interesting way came out recently in Nature and the Proceedings of the Royal Society B. The first, Smith, Elliot and Lyons' "Methane Emissions from Extinct Megafauna,"details an estimate of the net reduction in methane emissions that followed the (widely hypothesized as human-induced) extinction of large mammals in the Americas around 10,000BCE. They find that a large (they give between 12.5 and 100%) portion of the reduction in methane that we observe around that time period (a short period of global cooling known as the Younger Dryas) could be ascribed to the extinction of megafauna during this time period and the concomitant loss of their digestion-related methane emissions. This is important climatologically because methane is a very potent greenhouse gas, despite its relatively short residency time in the atmosphere, and thus big swings in it can be a major driver of short-term climate change.
The second paper, Boyce and Lee's "An exceptional role for flowering plant physiology in the expansion of tropical rainforests and biodiversity," estimates the effect that angiosperm (i.e., flowering) vegetation has on the climate, specifically on precipitation. Angiosperms have dramatically higher transpiration (evaporative loss through the plant, functionally the same as "plant sweat") rates than non-angiosperm plants and thus can contribute quite a bit to local precipitation, especially when heavily concentrated in areas like the Amazon rain forest. Boyce and Lee find that changing the flowering vegetative biomass in the Amazon from angiosperm to entirely non-angiosperm in a climate model would result in a huge decrease in precipitation, implying that such species are a major driver of the Amazon's extraordinarily high precipitation (which in turn drives a variety of other effects in the climate, as well as affecting rainforest biodiversity).
Now, obviously both of these are new papers and the extent to which the hypotheses they propose will stand up to scrutiny is still unknown. My personal interpretation is that the megafauna methane emissions paper, while intellectually pleasing, probably details a relatively minor part of the climate system while the angiosperm paper is pointing out a very interesting and fundamental part of biosphere-climate interactions. But both papers are nonetheless great examples of how complex the climate system is, and how even now we're still discovering fundamental drivers of climate variability. It's this complexity that should give us pause when we think about our current greenhouse gas emissions: if such a relatively minor (at least climatologically speaking) event as the extinction of some of the larger mammal species could potentially have led to or at least contributed to the Younger Dryas, then forcing the climate in so major a fashion as we currently are should definitely give us pause.
*Note: Image copyrighted L. Cunningham 2006.