Table 2. Results of GLMM testing of the influence of honey bee abundance on wild bee and small bee richness in 2020 vs. 2013.

Data were pooled across plots at each site and sampling period in each year. Covariates included greenspace type (park, community garden, cemetery), sampling period, and the proportion of herbaceous vegetation and impervious surface within a 500 m buffer around each site. Herbaceous vegetation was the amount of green vegetation greater than 3m in height within the buffer zone was used as a proxy for floral resource availability within the foraging range of wild bees. As sampling efforts differed between years, bee richness is the rarified species richness values, standardized to equal sample coverage between years. Predictors for each model were scaled and centered on zero. β is the coefficients for each predictor, CI is the 95% confidence intervals and p is the significance. Significant predictors for either bee richness or abundance are in bold.

	Wild bee richness			Small bee richness
Predictors	β	CI	p	β	CI	p
(Intercept)	3.61	2.76–4.46	<0.001	2.67	1.81–3.52	<0.001
Honey bee abundance × Year	−0.20	−0.37–0.03	0.023	−0.18	−0.35–0.00	0.047
Honey bee abundance	0.01	−0.13–0.15	0.894	0.02	−0.12–0.16	0.741
Year	−0.14	−0.30–0.02	0.087	0.00	−0.16–0.17	0.961
Herbaceous veg. (%)	−0.06	−0.45–0.33	0.751	−0.03	−0.42–0.35	0.863
Impervious surface (%)	−0.44	−0.96–0.08	0.097	−0.34	−0.86–0.18	0.202
Greenspace type (garden)	0.00	−0.17–0.17	0.973	0.16	−0.02–0.34	0.074
Greenspace type (park)	−0.26	−0.49–0.03	0.027	−0.00	−0.23–0.23	0.973
Sampling Period (2)	0.05	−0.15–0.26	0.600	0.05	−0.14–0.25	0.581
Sampling period (3)	0.01	−0.19–0.22	0.893	−0.06	−0.26–0.14	0.547
Sampling period (4)	−0.21	−0.42—0.00	0.049	−0.39	−0.61–0.18	<0.001
Sampling period (5)	−0.25	−0.45—0.04	0.020	−0.40	−0.62–0.19	<0.001