Smoke-Cued Emergence in Plant Species of Ponderosa Pine Forests: Contrasting Greenhouse and Field Results

Fire Ecology

Table 1 Summary of statistical results for three experiments examining seed, seed bank, and field responses to aqueous smoke in ponderosa pine forests, northern Arizona.

Effect	Response variables and statistics
Seed experiment^a	Emergence
	F	P > F	df
Species	45.3	<0.001	60
Smoke	43.6	<0.001	1
Species × smoke	3.4	<0.001	60
Seed bank experiment^b	Density			Richness
	t	P > t	df	t	P > t
Smoke	−2.6	0.034	8	−2.4	0.043
	Composition
	T	A	P > T
Smoke	−0.2	0.006	0.358
Field experiment^c	Plant cover			Richness/m²		Richness/9 m²
	F	P > F	df	F	P > F	F	P > F
Smoke	0.1	0.728	1	0.2	0.667	1.5	0.239
Plot [smoke]	1.7	0.153	16	1.2	0.391	2.0	0.096
Grazing	4.9	0.043	1	3.3	0.088	3.7	0.074
Smoke × grazing	0.6	0.461	1	0.5	0.489	3.7	0.074
Covariate	15.8	0.001	1	9.4	0.008	5.6	0.032
	Composition
	t	P > t	df
Smoke: grazed	−1.6	0.145	8
Smoke: ungrazed	−1.2	0.255	8

^a Two-factor analysis of variance with percentage of seeds emerging as the response variable.
^b Density and species richness of emergents analyzed using two-tailed paired t tests. Species composition (importance values) of emergents compared between treatments using blocked-multi-response permutation procedures (T = test statistic, A = chance-corrected within-group agreement, which provides a measure of within-group homogeneity; McCune and Grace 2002).
^c Univariate response variables analyzed as a split plot (aqueous smoke treatment as the whole-plot factor and grazing treatment as the subplot factor) with pre-treatment data as a covariate using analysis of variance. Species composition compared between control and aqueous smoke treated plots separately for grazed and ungrazed areas within plots by computing a Sørensen similarity (based on species importance values) between each pair of control and treated plots before and after treatment. A two-tailed paired t test evaluated the null hypothesis that the mean pre-treatment similarity between pairs of plots did not differ from the mean post-treatment similarity between pairs of plots.