Table 2. Multi- and uni-variate analyses of life-history variation, p-values associated with the effects of latitude and population structure.
Latitude | Population structure | ||||
Trait | df | F | % Explained variance | F | % Explained variance |
Multivariate analysis | |||||
(excluding secondary dormancy) | 1,109 | 16.06*** | 11% | 12.51** | 9% |
(including secondary dormancy) | 1,75 | 22.01*** | 20% | 7.79** | 7% |
Univariate analysis | |||||
Primary dormancy | 2,107 | 4.56* | 5% | 6.99** | 9% |
Secondary dormancy | 2,75 | 0.32 | n.s. | 2.66• | 4% |
Vegetative growth rate | 2,107 | 1.29 | n.s. | 2.16 | n.s. |
Flowering time | 2,108 | 7.95*** | 11% | 0.16 | n.s. |
Interaction between the two gradients was not significant (not shown). Because of missing values in secondary dormancy, the multivariate analysis was conducted with and without secondary dormancy. P-values were established with F test based on permutations (see methods). n.s.: not significant, • p<0.1, * p<0.05, ** p<0.01, ***p<0.001. The percentage of variance explained by latitude or population structure gradients is given. df: degrees of freedom for numerator and denominator, respectively.