Appendix E. Results from subsampling studies from the entire data set.
This appendix includes tables showing results from subsampling studies from the entire data set, as well as statistical test for the comparisons illustrated in Fig. 4 in the text.
TABLE E1: Number of 5000 randomly drawn subsets of the studies from the full data set for which the upper 95% confidence limit for the specified log response ratio effect size was less than zero (<0) or the lower 95% confidence limit was greater than zero (>0), under three criteria for non-independence of data from the same study (see text).
Effect |
Different agents and independent controls |
Different agents |
Different controls |
|||
<0 |
>0 |
<0 |
>0 |
<0 |
>0 |
|
Individual effect of enemies |
5000 |
0 |
5000 |
0 |
5000 |
0 |
Overall effect of enemies |
5000 |
0 |
5000 |
0 |
5000 |
0 |
Individual effect of mutualists |
0 |
5000 |
0 |
5000 |
0 |
5000 |
Overall effect of mutualists |
0 |
5000 |
0 |
5000 |
0 |
5000 |
Enemyenemy interaction effect |
0 |
0 |
0 |
0 |
0 |
0 |
Mutualistmutualist interaction effect |
0 |
0 |
0 |
0 |
0 |
449 |
Enemymutualist interaction effect |
0 |
4109 |
0 |
4208 |
0 |
4888 |
TABLE E2: Results of homogeneity tests for comparisons illustrated in Fig. 4. All tests have 1 df.
Effect |
Field vs. greenhouse: |
Herbaceous vs. woody plants: |
Crop plants vs. non-crops |
|||
Qb |
P |
Qb |
P |
Qb |
P |
|
Herbivore individual effects |
0.268 |
0.605 |
1.045 |
0.307 |
4.530 |
0.033 |
Pathogen individual effects |
7.313 |
0.007 |
7.844 |
0.005 |
2.526 |
0.112 |
Mutualist individual effects |
1.171 |
0.279 |
0.272 |
0.602 |
0.572 |
0.449 |
Enemyenemy interactions |
0.839 |
0.360 |
0.429 |
0.513 |
0.856 |
0.355 |
Enemymutualist interactions |
0.078 |
0.780 |
0.008 |
0.929 |
0.175 |
0.676 |