Table 2. Results of mixed effects models testing effects of predator exclusion and vertical stratum.

Models tested the fixed effects of predator exclusion (cage vs. control) and vertical stratum (UN, understory; LC, lower canopy; MC, middle canopy; UC, upper canopy) on the density of various groups of arthropods surveyed from June–September 2007 on 20 sugar maple trees (Acer saccharum Marsh.), while considering the random effects of tree identification (not shown); comparisons between vertical strata are presented as stepwise contrasts. Bold text indicates significant p-values.

Response group	Factor	DF	F	p	Comparisons
All arthropods	Treatment	1, 131	174.98	<0.0001	Cage > Control****
	Stratum	3, 131	172.95	<0.0001	LC < UN****
	Trt*Strat	3, 131	1.98	0.1196	MC < LC **
					UC < MC****
Spiders	Treatment	1, 131	123.90	<0.0001	Cage > Control ****
	Stratum	3, 131	106.55	<0.0001	LC < UN ****
	Trt*Strat	3, 131	2.49	0.0635	MC < LC**
					UC < MC**
Beetles	Treatment	1, 131	47.55	<0.0001	Cage > Control****
	Stratum	3, 131	43.16	<0.0001	LC < UN****
	Trt*Strat	3, 131	1.55	0.2044	MC < LC
					UC < MC****
True bugs	Treatment	1, 131	11.76	0.0008	Cage > Control***
	Stratum	3, 131	5.44	0.0015	LC < UN*
	Trt*Strat	3, 131	0.24	0.8681	MC < LC
					UC < MC
Caterpillars	Treatment	1, 131	7.98	0.0055	Cage > Control**
	Stratum	3, 131	11.04	<0.0001	LC > UN
	Trt*Strat	3, 131	0.53	0.6638	MC < LC
					UC < MC**

Notes.

Comparisons between strata were conducted using contrasts with backwards difference coding, significance of comparisons (t₁₃₄) designated by

^****p < 0.0001.

^***p < 0.001.

^**p < 0.01.

^*p < 0.05.