Table 3.
Effect of management interventions, landscape composition, time, and their interactions on predicted visitation rates per area flower cover
| Pollination 1 | Pollination 2 | OSR pollination | |
|---|---|---|---|
| Intercept | −.297 (.207) | .124 (.102) | −.349 (.193) |
| t = −1.438 | t = 1.224 | t = −1.813 | |
| FM | .054 (.015) | .061 (.013) | .097 (.019) |
| t = 3.564*** | t = 4.612*** | t = 5.187*** | |
| FS | .347 (.031) | −.408 (.027) | .381 (.038) |
| t = 11.233*** | t = −15.141*** | t = 10.013*** | |
| HET | −.116 (.133) | −.036 (.089) | −.056 (.143) |
| t = −0.870 | t = −0.406 | t = −0.393 | |
| OSR | −.267 (.056) | .742 (.051) | −.341 (.065) |
| t = −4.799*** | t = 14.593*** | t = −5.216*** | |
| Year | .0005 (.001) | −.001 (.001) | .0001 (.001) |
| t = 0.446 | t = −1.363 | t = 0.086 | |
| FM × OSR | .029 (.015) | .076 (.013) | .042 (.019) |
| t = 1.901 | t = 5.810*** | t = 2.230* | |
| FS × HET | .194 (.015) | .164 (.013) | .328 (.019) |
| t = 12.650*** | t = 12.290*** | t = 17.356*** | |
| FS × OSR | −.228 (.015) | −.562 (.013) | −.314 (.019) |
| t = −14.819*** | t = −41.967*** | t = −16.589*** | |
| FS × year | .010 (.001) | .008 (.001) | .012 (.002) |
| t = 6.466*** | t = 5.926*** | t = 6.414*** | |
| Observations | 2,800 | 2,800 | 2,800 |
| df | 2,653 (119) | 2,652 (119) | 2,653 (119) |
| AIC | 3,268.946 | 2,512.081 | 4,424.583 |
Results are from mixed‐effect models with identity of the land‐use map within the 7‐year sequence nested in landscape (site ID) as random effects. Given are coefficients, standard errors (in brackets), t‐values (t), and corresponding significance levels. Response variables, HET and OSR were z‐transformed prior to analysis. FM, field margins; FS, flower strips; HET, landscape heterogeneity; OSR, oilseed rape; df, degrees of freedom; AIC, Akaike information criterion. For heterogeneity, which does not change between years, degrees of freedom are given in brackets.
Significance levels: *p < .05; **p < .01; ***p < .001.