Table 3.
Results from stepwise logistic regressions relating the presence and absence of each marker to the two first axes of the principal component analysis (PCA)1, the host-plant (HP), and the three interactions2.
| Markers | Axis1 | Axis2 | HP | Axis1 × HP | Axis2 × HP | Axis1 × Axis 2 |
|---|---|---|---|---|---|---|
| 11 |
0.19 (5.7 × 10-4) |
NS | NS | NS | NS | NS |
| 13 |
0.39 (8.22 × 10-6) |
-0.34 (5.3 × 10-4) |
0.53 (0.15) |
NS | NS | NS |
| 19 | NS |
-0.38 (4.2 × 10-4) |
0.56 (0.13) |
NS | NS | NS |
| 26 | 0.10 (0.03) |
-0.25 (5.4 × 10-4) |
0.98 (2.7 × 10-3) |
NS | NS | NS |
| 27 | 0.09 (7 × 10-2) |
NS |
-0.96 (1 × 10-4) |
NS | NS | NS |
| 29 |
-0.57 (4 × 10-4) |
2.40 (0.24) |
2.48 (0.43) |
0.37 (0.45) |
-2.74 (3 × 10-3) |
NS |
| 30 | -0.19 (5.4 × 10-3) |
1.37 (2.42 × 10-5) |
0.74 (0.59) |
NS | -1.13 (1 × 10-3) |
NS |
| 33 |
-0.26 (3.64 × 10-9) |
NS | 0.37 (0.13) |
NS | NS | NS |
| 38 |
-0.45 (4.31 × 10-21) |
0.90 (2.37 × 10-7) |
NS | NS | NS | NS |
| 39 |
0.16 (3.8 × 10-5) |
NS | -0.090 (0.45) |
0.27 (.09) |
NS | NS |
| 40 |
-0.18 (8 × 10-6) |
-0.26 (0.93) |
0.29 (0.19) |
NS | 0.43 (0.10) |
NS |
| 42 | -0.18 (0.37) |
-0.66 (6 × 10-5) |
-0.74 (0.02) |
0.34 (0.02) |
NS | NS |
| 52 |
0.33 (1 × 10-7) |
-1.6 (6 × 16-16) |
-0.22 (0.86) |
NS | 0.97 (1 × 10-3) |
NS |
| 62 | -0.18 (0.37) |
0.46 (4 × 10-3) |
0.34 (0.39) |
0.31 (3 × 10-4) |
NS | NS |
| 63 |
0.59 (1 × 10-22) |
-0.76 (0.89) |
-0.86 (0.16) |
-0.18 (0.20) |
0.80 (0.04) |
NS |
| 65 | -0.19 (0.37) |
-0.44 (8 × 10-5) |
0.45 (0.47) |
0.20 (0.09) |
NS | NS |
| 68 | NS | 0.93 (0.05) |
0.04 (0.52) |
NS |
-1.05 (4.3 × 10-5) |
NS |
1 Axis 1 and axis 2 are the two first PCA components (representing respectively 76 and 20% of total inertia) of the projection of altitude and nine climatic variables calculated from the latitude/longitude as yearly mean of monthly values for the period 1961 to 2001 (see Methods for more details).
2Statistically significant logistic regression coefficients are indicated in bold (p-values for each coefficient are given in parenthesis). NS: the correlation was not significant.