Climate change tracking using a century of phenology data in Colorado

Background/Question/Methods

 Determining the biological relevance of climate change trends generally requires data that spans decades in order to account for year-to-year variability. Such data are especially important for understanding effects of climate change on rare or endangered species. Herbarium specimens have been used to document changes in phenology over periods as long as 100 years by considering collection date as a proxy for flowering time, with a previous study in Massachusetts observing evidence of an earlier spring by a week. Given that documented climate changes have been more dramatic in the interior of the U.S., we asked 1) to what degree are Colorado species showing earlier flowering times? Are spring ephemeral species there, which are more likely to be cued by temperature, more likely to show a change over time?  2) Are rare plants more likely to show sensitivity to climate change? 3) Can climate data be used to predict collection date, if that measure really does reflect phenology?  To address these questions, we amassed label data from over 600 flowering herbarium specimens of Colorado plants collected from 1870 to 2009 from herbaria across the country. Digital label data were verified and flowering status recorded with visual inspection of physical specimens. 

Results/Conclusions

We found Colorado species to be flowering approximately 20 days earlier now versus a century ago.  A significant negative correlation was found between collection date and year for both a group of six Colorado spring ephemeral species and for 12 Federally Listed rare/endangered Colorado species.  The former were chosen because of the abundance of collection data found at a single herbarium, and the latter used label data from herbaria across the country.  As a control for changes in collector behavior, all collection dates of all specimens in one Colorado herbarium, including non-flowering specimens, were also regressed against year and found no pattern. Climate data from the PRISM Climate Group's Data Explorer, explained more than 70% of variability in collection date, based on both temperature and precipitation of the collection year. These findings suggest that both spring blooming and rare species are highly sensitive to the climate changes observed in the center of the country over the past century, with changes more than twice as large as those observed on the East Coast.