Think about that.  If a study did not indicate that "at least one species shows change over time" or that there was "a strong association between the species trait and site-specific temperature," the study was ignored.  Talk about stacking the deck in favor of one's hypothesis!  If a study showed that a species' population was stable over time or did not show a strong association between one of its traits and changing temperature patterns - such as we would predict for the heat-limited boundary of a species' range, which consequence would tend to refute the CO2-induced global warming extinction hypothesis -- it was dropped from further consideration.

So just how extensive was this stacking of the deck?  Of the thousands of articles Root et al. examined, they selected a mere one hundred and forty-three for detailed scrutiny.  Does this massive filtering of the data mean there could be hundreds upon hundreds -- if not thousands -- of studies that run counter to their hypothesis?  There is a strong possibility that it does, especially in light of what we are about to learn about the studies they did use.

From among the 143 "qualifying" articles that survived their unique filtering process, Root et al. created two categories of studies.  Tier 1 studies, of which there were 85, were "those demonstrating a statistically significant trend for at least one species examined."  Tier 2 studies, of which there were 58, were "those in which statistical significance was not shown by the study's authors," which gives yet another indication of the paucity of pertinent data they employed in reaching their sweeping global conclusions.

Within these two tiers of papers, Root et al. evaluated changes in four species traits: (1) species densities at given locations and species ranges, (2) the timing of significant species life-cycle events, (3) species morphology, and (4) species genetic frequencies.  However, only the first of these traits provide data that are appropriate for evaluating the hypothesis that CO2-induced global warming will force plants and animals to migrate to cooler regions of the planet.

The impact of this last observation is such that of the 85 Tier 1 articles evaluated by Root et al. - again by their own admission - only eleven contain information on species range shifts; while of the 58 Tier 2 papers they evaluated, a mere thirteen deal with the subject.  Hence, out of the thousands of articles they originally examined, only two dozen contain data that might be appropriate for evaluating the CO2-induced global warming extinction hypothesis; and over half of them lay no claim to possessing any statistical significance, which truly speaks volumes about how underwhelming is the case their paper makes for Root's contention that "we're sitting at the edge of a mass extinction."  Nevertheless, to complete our analysis, we next evaluate each of the 24 potentially pertinent papers in the order in which Root et al. list them in their Supplemental Appendices 1 and 2, dealing first with their Tier 1 studies and then their Tier 2 studies.