Appendix B. ANOVAs, main effects, and fits to raw data.
TABLE B1. A split-split-plot ANOVA was conducted with SAS proc MIXED on the natural log of the Student t dispersal parameter, u, (A) testing for differences between reproductive inhibition treatments (Shrub, Open, and Both removals) slopes (up-slope vs. down-slope) and species (Pectocarya recurvata vs. Shismus barbatus). (B) with slope coded as a continuous variable. Satterthwaite's approximate degrees of freedom were calculated for the appropriate error terms.
A) Slope – Fixed Effect
Effect Num df Den df F P Treatment (Removal) 2 7.81 29.44 0.0002 Slope 1 7.51 2.64 0.1451 Treat × slope 2 7.50 5.35 0.0360 Species 1 16.30 21.72 0.0003 Treatment × species 2 16.20 0.98 0.3950 Slope × species 1 16.40 0.00 0.9713 Treatment × slope × species 2 16.40 0.10 0.9062 B) Slope – Continuous Effect
Effect Num df Den df F P Treatment (Removal) 2 7.88 29.30 0.0002 Slope 1 25.0 10.75 0.0031 Treat × slope 1 25.0 6.41 0.0180 Species 1 25.9 21.72 0.0001 Treatment × species 2 25.9 0.96 0.3952 Slope × species 1 25.0 0.10 0.7591 Treatment × slope × species 1 25.0 0.04 0.8434
TABLE B2. Log-transformed dispersal parameter, ln(u) of the one parameter Student t kernel (Clark et al. 1999), and its standard error calculated from least square means, and corresponding mean (with 90% confidence intervals), median, and 95th percentile dispersal distances for significant main effects and interaction effects from a split-split-plot ANOVA conducted with SAS proc MIXED. Effects with same superscript letter did not differ by Tukey-Kramer means comparisons at P < 0.05.
Dispersal distance (m) Species ln(u) Std error mean median 95% PERE
7.60A 0.2491 0.70
(0.570.87)0.45 1.95 SCBA
6.39B 0.2475 0.29
(0.230.36)0.18 0.80 Treatments Both Removal
7.31A 0.2820 0.61
(0.470.78)0.39 1.69 Open Removal 7.84A 0.2617 0.79
(0.621.01)0.51 2.20 Shrub Removal 4.99B 0.2866 0.19
(0.150.25)0.12 0.53 Slope Down-slope 7.04A 0.2564 0.53
(0.420.66)0.34 1.47 Up-slope 6.39A 0.2530 0.38
(0.310.48)0.24 1.07 Treatment × Slope interaction Both × Down-slope 7.52AB 0.4658 0.67
(0.451.01)0.43 1.87 Both × Up-slope 7.10AB 0.4168 0.55
(0.380.79)0.35 1.52 Open × Down-slope 9.01A 0.4007 1.42
(1.002.02)0.90 3.94 Open × Up-slope 6.68ABC 0.4339 0.44
(0.300.65)0.28 1.23 Shrub × Down-slope 4.59C 0.4627 0.16
(0.100.23)0.10 0.43 Shrub × Up-slope 5.40BC 0.4627 0.23
(0.160.35)0.15 0.65
Within Open Down-slope, P. recurvata Open into Shrub Up-slope, S. barbatus ![]()
![]()
Shrub into Open Down-slope, P. recurvata Shrub into Open Down-slope, S. barbatus ![]()
![]()
FIG. B1. The number of seedlings at different distances from the boundaries of suppressed areas, with the fitted population-level dispersal curves (assuming a one parameter Student t dispersal kernel and Poission error) for several sample transects.