Ecological Archives E087-078-A6

Jarrod M. Thaxton and William J. Platt. 2006. Small-scale fuel variation alters fire intensity and shrub abundance in a pine savanna. Ecology 87:1331–1337.

Appendix F. ANCOVA table for stem density and ANOVA table for mortality.

TABLE F1. Results of mixed-model repeated-measures ANCOVA of stem density for all species pooled.

Source of variation
df
F
P
Fixed Effects:
 
 
 
Initial Density (Covariate)
1, 506
735.37
<0.001
Elevation
1, 10.8
1.48
0.25
Fuel
3, 496
15.51
<0.001
Elevation × Fuel
3, 496
0.64
0.59
Year
1, 509
3.77
0.05
Elevation × Year
1, 509
0.31
0.58
Fuel × Year
3, 509
1.5
0.21
Elevation × Fuel × Year
3, 509
1.02
0.38
 
 
 
 
Linear Contrasts (Fuel):
 
 
 
(Controls + Fuel Removal) vs. (Pine Needle Addition + Wood Addition)
1, 496
43.75
<0.001
Controls vs. Fuel Removal
1, 497
0.87
0.35
Pine Needle Addition vs. Wood Addition
1, 496
13.08
<0.001
 
 
 
 
Linear Contrasts (Fuel by Year):
 
 
 
[(Controls + Fuel Removal) vs.
(Pine Needle Addition + Wood Addition)] × (Post-fire 1 vs. Post-fire 2)
1, 509
0.11
0.74
[(Controls vs. Fuel Removal) × (Post-fire 1 vs. Post-fire 2)
1, 509
0.07
0.79
(Pine Needle Addition vs. Wood Addition) × (Post-fire 1 vs. Post-fire2)
1, 509
3.29
0.07

   Notes: Covariate = 1998 initial stem density. Error term for Elevation = Elevation × Transect (Ravine). Error term for Fuel and interactions = Subplot (Fuel × Elevation × Transect (Ravine)). Error term for Year and interactions = residual.

† Denominator degrees of freedom based on Satterthwaite approximation. Significance level is P < 0.05.

 

TABLE F2. Results of generalized linear mixed-model analysis of the probability of shrub mortality.

Source of variation
df
F
P
Fixed Effects:
 
 
 
Elevation
1, 9.18
7.72
0.02
Fuel
3, 509
5.68
<0.001
Elevation × Fuel
3, 509
0.21
0.89
Year
1, 523
5.97
0.015
Elevation × Year
1, 523
1.8
0.18
Fuel × Year
3, 520
1.98
0.11
Elevation × Fuel × Year
3, 520
0.98
0.4
 
 
 
 
Linear Contrasts (Fuel):
 
 
 
(Controls + Fuel Removal) vs. (Pine Needle Addition + Wood Addition)
1, 511
13.41
<0.001
Controls vs. Fuel Removal
1, 496
0.21
0.65
Pine Needle Addition vs. Wood Addition
1, 525
6.72
0.01
 
 
 
 
Linear Contrasts (Fuel by Year):
 
 
 
[(Controls + Fuel Removal) vs.
(Pine Needle Addition + Wood Addition)] × (Post-fire 1 vs. Post-fire 2)
1, 523
4.91
0.03
[(Controls vs. Fuel Removal) × (Post-fire 1 vs. Post-fire 2)
1, 508
0.9
0.34
(Pine Needle Addition vs. Wood Addition) × (Post-fire 1 vs. Post-fire2)
1, 534
0.14
0.7

   Notes: Species loss is defined as presence of stems of a species prior to fire, but absence of all stems of that species for two consecutive years post-fire. Error term for Elevation = Elevation × Transect (Ravine); Error term for Fuel and interactions = Subplot (Fuel × Elevation × Transect (Ravine)); Error term for Year and interactions = residual.

† Denominator degrees of freedom based on Satterthwaite approximation. Significance level is P < 0.05.



[Back to E087-078]