Table 1.
Model rankings from the maximum-likelihood analysis of the evolutionary relationship between drought adaptation and mycorrhizal strategy for each dataset version analyzed.
| Evolution mode | Hidden rates | Model structure | k rates | Dataset v1 | Dataset v2 | Dataset v3 | Dataset v4 | Dataset v5 | Dataset v6 | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| AICc | AICcWt | AICc | AICcWt | AICc | AICcWt | AICc | AICcWt | AICc | AICcWt | AICc | AICcWt | ||||
| Independent | 1 | ER | 2 | 2846.1 | <0.01 | 2846.1 | <0.01 | 2808.3 | <0.01 | 2808.3 | <0.01 | 1308.3 | <0.01 | 1308.2 | <0.01 |
| Independent | 1 | SYM | 4 | 2823.2 | <0.01 | 2823.2 | <0.01 | 2811.0 | <0.01 | 2785.4 | <0.01 | 1288.4 | <0.01 | 1288.3 | <0.01 |
| Independent | 1 | ARD | 8 | 2817.1 | <0.01 | 2817.1 | <0.01 | 2772.2 | <0.01 | 2772.2 | <0.01 | 1285.1 | <0.01 | 1285.1 | <0.01 |
| Independent | 2 | ER | 8 | 2684.4 | <0.01 | 2684.3 | <0.01 | 2619.5 | <0.01 | 2619.6 | <0.01 | 1234.6 | <0.01 | 1234.6 | <0.01 |
| Independent | 2 | SYM | 12 | 2651.7 | <0.01 | 2651.7 | <0.01 | 2586.9 | <0.01 | 2586.9 | <0.01 | 1217.9 | 0.01 | 1217.9 | 0.01 |
| Independent | 2 | ARD | 20 | 2625.9 | 0.04 | 2624.9 | 0.01 | 2589.3 | <0.01 | 2568.3 | <0.01 | 1215.1 | 0.03 | 1215.1 | 0.02 |
| Independent | 3 | ER | 18 | 2682.5 | <0.01 | 2682.4 | <0.01 | 2618.9 | <0.01 | 2618.9 | <0.01 | 1236.0 | <0.01 | 1234.8 | <0.01 |
| Independent | 3 | SYM | 24 | 2649.3 | <0.01 | 2655.2 | <0.01 | 2585.8 | <0.01 | 2585.8 | <0.01 | 1224.4 | <0.01 | 1225.6 | <0.01 |
| Independent | 3 | ARD | 36 | 2630.2 | <0.01 | 2630.3 | <0.01 | 2562.8 | <0.01 | 2564.9 | <0.01 | 1233.7 | <0.01 | 1233.8 | <0.01 |
| Dependent | 1 | ER | 1 | 4038.5 | <0.01 | 4038.4 | <0.01 | 3833.5 | <0.01 | 3833.5 | <0.01 | 1845.2 | <0.01 | 1845.2 | <0.01 |
| Dependent | 1 | SYM | 9 | 2764.0 | <0.01 | 2764.0 | <0.01 | 2753.0 | <0.01 | 2753.0 | <0.01 | 1258.8 | <0.01 | 1258.8 | <0.01 |
| Dependent | 1 | ARD | 18 | 2706.5 | <0.01 | 2706.5 | <0.01 | 2675.8 | <0.01 | 2696.7 | <0.01 | 1254.9 | <0.01 | 1254.9 | <0.01 |
| Dependent | 2 | ER | 4 | 3894.8 | <0.01 | 3894.8 | <0.01 | 3701.0 | <0.01 | 3700.9 | <0.01 | 1783.7 | <0.01 | 1783.6 | <0.01 |
| Dependent | 2 | SYM | 20 | 2659.4 | <0.01 | 2659.4 | <0.01 | 2596.0 | <0.01 | 2596.0 | <0.01 | 1207.8 | 0.97 | 1207.8 | 0.97 |
| Dependent | 2 | ARD | 38 | 2619.4 | 0.96 | 2616.3 | 0.99 | 2544.0 | 1.00 | 2553.6 | 1.00 | 1227.2 | <0.01 | 1224.3 | <0.01 |
| Dependent | 3 | ER | 9 | 3878.8 | <0.01 | 3878.8 | <0.01 | 3688.0 | <0.01 | 3688.0 | <0.01 | 1778.6 | <0.01 | 1778.5 | <0.01 |
| Dependent | 3 | SYM | 33 | 2636.7 | <0.01 | 2643.8 | <0.01 | 2588.2 | <0.01 | 2592.6 | <0.01 | 1222.1 | <0.01 | 1222.2 | <0.01 |
| Dependent | 3 | ARD | 60 | 2632.3 | <0.01 | 2635.5 | <0.01 | 2561.7 | <0.01 | 2580.9 | <0.01 | 1259.9 | <0.01 | 1266.0 | <0.01 |
The k rate is the number of evolutionary transition rate parameters independently estimated by a given model. The AICc is the sample size-corrected Akaike information criterion, while the AICcWt is the weighted AICc that provides the relative likelihood of each model per dataset version. The text in bold corresponds to the best-fitted model for each dataset version according to the AICc. The assembly of datasets v1 to v6 is summarized in Table 3 and described in the Methods.
ER all rates are equal, SYM rates between any two states do not differ, ARD all rates differ.