Table 2.
Shoot K+/Na+ ratio | Root K+/Na+ ratio | Root proline | |||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Categorical Moderator | Qm | n | phetero | df | I2 | Qm | n | phetero | df | I2 | Qm | n | phetero | df | I2 |
A | |||||||||||||||
% NaCl in salt stress*** | 11.4 | 249 | 0.003 | 2 | 82.4 | 6.7 | 95 | 0.009 | 2 | 85.1 | |||||
Stress severity | 6.8 | 158 | 0.009 | 1 | 85.2 | 1.0 | 56 | 0.326 | 1 | 0.0 | 0.0 | 90 | 0.913 | 1 | 0.0 |
AM taxa** | 43.7 | 242 | <0.001 | 6 | 86.3 | 9.0 | 88 | 0.061 | 4 | 55.7 | 28.7 | 111 | <0.001 | 3 | 89.5 |
Host water status*** | <0.01 | 23 | 0.952 | 1 | 0.0 | 0.0 | 23 | 0.952 | 1 | 0.0 | |||||
Host type | 23.7 | 249 | <0.001 | 2 | 91.6 | 1.1 | 95 | 0.591 | 2 | 0.0 | 4.9 | 123 | 0.086 | 2 | 59.2 |
Woodiness | 0.2 | 249 | 0.697 | 1 | 0.0 | 3.1 | 95 | 0.077 | 1 | 68.0 | 3.5 | 123 | 0.060 | 1 | 71.7 |
Shoot size | 21.8 | 248 | <0.001 | 2 | 90.8 | 4.7 | 74 | 0.095 | 2 | 57.5 | 0.9 | 116 | 0.636 | 2 | 0.0 |
Regression Moderator | n | intercept | slope | phetero | I2 | n | intercept | slope | phetero | I2 | n | intercept | slope | phetero | I2 |
B | |||||||||||||||
Time of exposure | 234 | 0.691 | −0.004 | 0.070 | 14.0 | 88 | 0.964 | −0.010 | 0.002 | 18.6 | 123 | −0.196 | 0.005 | 0.015 | 15.8 |
Root colonization | 225 | 0.222 | 0.006 | 0.009 | 17.3 | 85 | 0.113 | 0.005 | 0.172 | 25.2 | 118 | 0.153 | −0.002 | 0.413 | 16.0 |
Qm = between study heterogeneity; n = sample size; phetero = probability that all observed (total) heterogeneity is due to sampling error (expected variation); df = degrees of freedom, levels within a moderator; I2 = percentage of total heterogeneity due to true differences in summary effects among levels of a moderator. The levels of of each categorical moderator with their summary effect sizes, CIs, number of studies and significance values are given in Figures 2–4.
Total amino acid concentration of shoots also showed significant heterogeneity but there were not enough studies within moderator levels to perform moderator analysis for this summary effect.
A single species was included as a level in the AM taxa moderator if at least 8 studies from >1 paper reported data for it. Those species with data from <8 studies or just 1 paper were grouped into “other species.”
% NaCl in salt stress and Host water status are blank for root proline because of insufficient articles providing these data.