Table 1.
Overview of five major global meta-analyses from long-term observational data on individual wild species with diverse distributions in terrestrial (T), marine (M) and freshwater (F) systems
| Study | Number of species and functional groups | Species in given system (% of all) |
Plants in each system (% of each system) |
Species showing significant long-term change in phenologies, distributions, abundances or morphology (% of all) | Changes consistent with local or regional climate change (% of species that showed change) | ||||
|---|---|---|---|---|---|---|---|---|---|
| T | M | F | T | M | F | ||||
| Parmesan and Yohe 2003 | 1598 | 85 | 13 | <2 | 63 | 0 | 0 | 59 | 841 |
| Root et al. 2003 | 1468 | 94 | 5 | <1 | 49 | <1 | <1 | 40 | 821 |
| Root et al. 2005 | 130 | 100 | 0 | 0 | 65 | 0 | 0 | 100 | 921 |
| Rosenzweig et al. 2008 | 55 studies | 65 | 13 | 22 | 44 | 14 | 42 | – | 902 |
| Poloczanska et al. 2013 | 857 | 0 | 100 | 0 | 0 | 16 | 0 | 76 | 831 |
Each study includes data from multiple continents and oceans, and together there is representation from every continent and every major ocean. Not all studies provided all metrics; missing information is indicated with a dash. Percentages are approximate and estimated for the studies as whole-individual analyses within the studies may differ. The specific metrics of climate change analysed for associations with biological change vary somewhat across studies, but most use changes in local or regional temperatures (e.g. mean monthly temperature or mean annual temperature), with some using precipitation metrics (e.g. total annual rainfall). Individual species were analysed by Rosenzweig et al. (2008) but data on species not provided in publication—percentages shown are based on numbers of studies. 1P < 10−15; 2P < 0·001 (from binomial test against random expectation of 50-50 chance of change in either direction—either consistent or not consistent with local or regional climate change).