Table 2.
Linear regressions of the relation between mean air (independent variable) and mean water temperature (dependent variable) in one lowland (LL) and two highland sites (HL) in western Kenya.
| a | b | R2 | ANOVA | |
|---|---|---|---|---|
| Small-sized pool | ||||
| Kisian (LL) | .794 | 8.646 | .803 | F1,37 = 147, P < 0.001 |
| Lyanaginga (HL) | .911 | 6.209 | .487 | F1,29 = 26.6, P < 0.001 |
| Fort Ternan (HL) | .856 | 7.654 | .719 | F1,14 = 33.3, P < 0.001 |
| Medium-sized pool | ||||
| Kisian (LL) | .677 | 11.802 | .745 | F1,38 = 108, P < 0.001 |
| Lyanaginga (HL) | .846 | 7.745 | .407 | F1,29 = 19.2, P < 0.001 |
| Fort Ternan (HL) | .677 | 11.715 | .639 | F1,28 = 47.8, P < 0.001 |
| Large-sized pool | ||||
| Kisian (LL) | .700 | 11.400 | .743 | F1,39 = 110, P < 0.001 |
| Lyanaginga (HL) | .710 | 10.852 | .405 | F1,29 = 19.0, P < 0.001 |
| Fort Ternan (HL) | .503 | 15.395 | .483 | F1,29 = 26.1, P < 0.001 |
a slope; b intercept; LL lowland area; HL highland area
Empirical data and regression lines are plotted in Figure 2.