Table 1.
House sparrow mass is predicted by predator abundance. (GLM of factors predicting mass variation, produced by reverse stepwise regression. Dependent variable is mass, N=7580 individual house sparrows, adjusted R2=0.14.)
| source of variation | sum of squares | d.f. | F | sig. |
|---|---|---|---|---|
| size (wing length) | 1816 | 1 | 499.7 | <0.001 |
| time of day | 459 | 1 | 126.2 | <0.001 |
| sparrowhawk abundancea | 434 | 1 | 119.3 | <0.001 |
| latitude | 413 | 1 | 113.6 | <0.001 |
| sex | 99 | 1 | 27.1 | <0.001 |
| month | 599 | 11 | 15.0 | <0.001 |
| longitude | 44 | 1 | 12.0 | <0.001 |
| day length | 13 | 1 | 3.7 | 0.056 |
| sex×month | 281 | 11 | 7.0 | <0.001 |
| model | ||||
| explained | 4611 | 29 | 43.7 | <0.001 |
| residual | 27 447 | 7550 | ||
| total | 32 058 | 7579 |
a
Parameter estimate for hawk abundance =−1.74±0.159.