Table 5. MANOVA results for forest age structure variables.

		Climate effect		Fire effect£		Harvest effect§		Fire×harvest
Simulation years	Seral stage (%)	Variation explained (%)	t	Variation explained (%)	t	Variation explained (%)	t	Variation explained (%)	R2
150 (2140)	Establishment	0.4**	12.2	40.8**	77.9	58.6**	94.1	0.1**	1.00
	Early-stage	7.4**	15.2	5.3**	12.2	85.5**	39.6	0.6**	0.99
	Mid-stage	0.7**	22.1	97.3**	−155.9	0.08	19.7	1.8**	1.00
	Late-stage	0.002	1.5	77.0**	−239.6	22.2**	−116.4	0.8**	1.00
	Old-growth	0	0.77	5.1**	−106.2	94.8**	−419	0.09**	1.00
300 (2290)	Establishment	0.2**	−4.3	8.6**	12	89.4**	57	1.4**	1.00
	Early-stage	2.0**	10.9	6.6**	−14.1	90.8**	51.6	0	0.99
	Mid-stage	28.2**	9.6	52.2**	−7.6	7.3**	−1.9	1.5	0.88
	Late-stage	5.6**	14.9	15.5**	−6.5	71.9**	−26.7	6.1**	0.99
	Old-growth	0.2**	16.4	0.09**	8.4	99.2**	−220.3	0.4**	1.00

The t values test the hypothesis that the response between levels of main effects are equal, and significant (α = 0.01) differences are indicated in boldface. All three main effects were significant in both years. Only the fire × harvest interaction was always significant and was included in the model. Significant interactions are indicated by asterisks.

^**P<0.01.

Positive t value means that response variable increases as climate warms.

^£Positive t value means that response variable increases as fires increases.

^§Positive t value means that response variable increases when harvest is added.