Table 2.

Statistical tests of intraspecific relationships between phenology and abundance change

Dependent variable	Independent variable	Data subset	Interacting covariates	n	AIC	Marginal R2	Effect size (s.e.)	Χ2 (P)
Change in abundance	Change in emergence date	Full data set	Voltinism	1677	−5255.4	0.028	–	9.18 (0.002)
			Voltinism∗ Class	1677	−5261.4	0.032	–	0.17 (0.679)
		Univoltine species	–	1038	−3182.7	0.000	−0.00 (0.00)	0.02 (0.886)
			Class	1038	−3179.4	0.002	–	0.81 (0.368)
		Univoltine habitat specialists	–	99	−254.9	0.007	0.01 (0.02)	0.59 (0.441)
		Univoltine wider-countryside generalists	–	939	−2933.2	0.0001	−0.00 (0.00)	0.13 (0.723)
		Multivoltine species	–	639	−2029.7	0.074	0.01 (0.00)	57.50 (<0.001)
			Class	639	−2086.9	0.084	–	0.006 (0.938)
		Multivoltine habitat specialists	–	14	−32.46	0.002	0.01 (0.04)	0.06 (0.800)
		Multivoltine wider-countryside generalists	–	625	−1995.7	0.078	0.01 (0.00)	58.63 (<0.001)
		Multivoltine species (increasing abundance)	–	295	−903.9	0.068	0.01 (0.00)	21.46 (<0.001)
		Multivoltine species (declining abundance)	–	344	−1162.7	0.091	0.01 (0.00)	33.80 (<0.001)

An overall model was constructed in each case with species as a random effect, and with voltinism as an interacting covariate if indicated. Significance of each model was tested using a likelihood ratio test, and tests that had statistical significance (P < 0.05) are indicated in bold