Past Warm Episodes did not Cause Extinctions
Willis, K.J., Bennett, K.D., Bhagwat, S.A. and Birks, H.J.B. 2010. Perspective: 4 deg C and beyond: what did this mean for biodiversity in the past? Systematics and Biodiversity 8: 3-9.
The first period they examined was the Eocene Climatic Optimum (53-51 million years ago), during which time the atmosphere's CO2 concentration exceeded 1200 ppm and tropical temperatures were 5-10°C warmer than modern values. Yet far from causing extinctions of the tropical flora (where the data are best), the four researchers report that "all the evidence from low-latitude records indicates that, at least in the plant fossil record, this was one of the most biodiverse intervals of time in the Neotropics." They also note that "ancestors of many of our modern tropical and temperate plants evolved ...when global temperatures and CO2 were much higher than present...indicating that they have much wider ecological tolerances than are predicted based on present-day climates alone."
The second period they examined consisted of two rapid-change climatic events in the Holocene -- one at 14,700 years ago and one at 11,600 years ago -- during which times temperatures increased in the mid- to high-latitudes of the Northern Hemisphere by up to 10°C over periods of less than 60 years. During these events, there is evidence from many sites for rapid plant responses to rapid warming. And the authors note that "at no site yet studied, anywhere in the world, is there evidence in the fossil record for large-scale climate-driven extinction during these intervals of rapid warming." On the other hand, they report that extinctions did occur due to the cold temperatures of the glacial epoch, when subtropical species in southern Europe were driven out of their comfort zone.
The study of Willis et al. also makes use of recent historical data, as in the case of the 3°C rise in temperature at Yosemite Park over the past 100 years. In comparing surveys of mammal fauna conducted near the beginning and end of this period, they detected some changes, but no local extinctions. Thus, they determined that for all of the periods they studied, with either very warm temperatures or very rapid warming, there were no detectable extinctions.