Table 3.
Clade | sqrt (Altitude) | ln (Precipitation) | ln (Emberger Index) | Min Temp | Max Temp |
---|---|---|---|---|---|
error = 0 | |||||
1 | 0.29978 | 0.60730 | 0.75952 | 0.12975 | 0.15825 |
2 | 0.16626 | 0.90511 | 0.90077 | 0.17785 | 0.17675 |
3 | 0.26579 | 0.52435 | 0.44646 | 0.76296 | 0.11206 |
4 | 0.52548 | 0.56360 | 0.76298 | 0.14681 | 0.30656 |
5 | 0.11212 | 0.49581 | 0.42547 | 0.98678 | 0.14469 |
error = estimated squared standard error | |||||
1 | 0.31380 | 0.63645 | 0.80282 | 0.12743 | 0.00213 |
2 | 0.33394 | 0.89657 | 0.89803 | 0.11836 | 0.00327 |
3 | 0.62014 | 0.48266 | 0.40712 | 0.97213 | 0.00208 |
4 | 0.69243 | 0.58770 | 0.80586 | 0.13598 | 0.00518 |
5 | 0.17833 | 0.46473 | 0.39064 | 0.99029 | 0.00262 |
For each variable, 33 models were evaluated: 32 O-U models comprising all 2^5 possible models allowing or disallowing shifts in equilibrium value at the base of the 5 focus clades; and the Brownian motion (constant variance) model. Values in each cell are Bayes information criterion (BIC) weights summed over all models allowing a change at the node indicated. This estimates the posterior probability that there was a shift in environmental conditions at the base of the clade estimated integrating over all models evaluated. Measurement error for the lower portion of the table was estimated as indicated in the text.