Fig. 4 Relationship between bee visits and floral resource (‘FR’) predictors from the best models (in Table 1) measured across four time periods (T1–T4) in four spatial scales (transect, 250 m, 500 m, 750 m) over one season, represented by the corresponding hypotheses: H2b, H3a, or H3b. Lines represent the negative binomial relationship between an FR predictor and bee visits, with a 95% confidence interval. FR were calculated with all unknown areas in a given radius assigned the median volume of FR measured across all known resource-providing land types during a given time period.
Appendix
Table A1 Land-types used to quantify landscape structure and floral resources around sampling locations. The range in proportion of each land type across all sampling periods is represented as the percent of total area within a 750-m radius around each sampling location. Areas with water, exposed or barren land, and urban or developed land were not digitized, and exact proportions were not calculated for these land types. The unknown crop, hedgerow, and potentially resource-providing crop land types were not assessed for floral resources, and a range is presented for values assigned in models run, from the minimum to the median floral resource volume from all resource-providing land types in a given spatial scale and sampling period.