Ecological Archives E089-127-A2

D. Lawrence Venable, Arturo Flores-Martinez, Helene C. Muller-Landau, Greg Barron-Gafford, and Judith X. Becerra. 2008. Seed dispersal of desert annuals. Ecology 89:2218–2227.

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
(0.57–0.87)
0.45 1.95
SCBA
6.39B 0.2475 0.29
(0.23–0.36)
0.18 0.80
Treatments          
Both Removal
7.31A 0.2820 0.61
(0.47–0.78)
0.39 1.69
Open Removal 7.84A 0.2617

0.79
(0.62–1.01)

0.51 2.20
Shrub Removal 4.99B 0.2866 0.19
(0.15–0.25)
0.12 0.53
Slope          
Down-slope 7.04A 0.2564 0.53
(0.42–0.66)
0.34 1.47
Up-slope 6.39A 0.2530 0.38
(0.31–0.48)
0.24 1.07
Treatment  ×  Slope interaction          
Both  ×  Down-slope 7.52AB 0.4658 0.67
(0.45–1.01)
0.43 1.87
Both  ×  Up-slope 7.10AB 0.4168 0.55
(0.38–0.79)
0.35 1.52
Open  ×  Down-slope 9.01A 0.4007 1.42
(1.00–2.02)
0.90 3.94
Open  ×  Up-slope 6.68ABC 0.4339 0.44
(0.30–0.65)
0.28 1.23
Shrub  ×  Down-slope 4.59C 0.4627 0.16
(0.10–0.23)
0.10 0.43
Shrub  ×  Up-slope 5.40BC 0.4627 0.23
(0.16–0.35)
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.

[Back to E089-127]