Appendix D. Results of additional GLM analyses.
TABLE D1. Results of general linear models that explain the evenness of substrate dwelling and seston feeding zooplankton and of phytoplankton in relation to the plant treatment (categorical variable) and to the initial primary productivity gradient (continuous variable).
|
Explanatory variables‡ |
df |
SS |
MS |
F |
P |
Substrate dwellers |
Intercept |
1 |
2.32 |
2.32 |
99 |
< 0.0001 |
PLANT |
2 |
0.47 |
0.24 |
10 |
< 0.001 |
|
CHLAin |
1 |
0.01 |
0.01 |
0.4 |
0.514 |
|
Error |
53 |
1.24 |
0.02 |
|||
Seston feeders |
Intercept |
1 |
3.24 |
3.24 |
165 |
< 0.0001 |
PLANT |
2 |
0.25 |
0.13 |
6 |
0.003 |
|
CHLAin |
1 |
0.16 |
0.16 |
8.35 |
0.006 |
|
PLANT × CHLAin |
2 |
0.30 |
0.15 |
7.54 |
0.001 |
|
Error |
53 |
1.04 |
0.02 |
|||
Phytoplankton |
Intercept |
1 |
4.51 |
4.51 |
223 |
< 0.0001 |
PLANT |
2 |
0.09 |
0.05 |
2.17 |
0.124 |
|
CHLAin |
1 |
0.45 |
0.45 |
22.4 |
< 0.0001 |
|
Error |
54 |
1.09 |
1.09 |
|||
Substrate dwellers§ |
Intercept |
1 |
5.21 |
5.21 |
222 |
< 0.001 |
PLANT |
2 |
0.47 |
0.24 |
10 |
< 0.001 |
|
PERIPH |
1 |
0.01 |
0.01 |
0.4 |
0.556 |
|
Error |
53 |
1.24 |
0.02 |
Note: Nonsignificant interactions between the predictor variables were omitted from the GLM models. For more information on the statistical analyses see Materials and Methods.
‡ PLANT is the plant treatment (no macrophytes, artificial macrophytes, real macrophytes; CHLAin is the initial chlorophyll a concentration at the moment of zooplankton inoculation and represents the initial primary productivity gradient; PERIPH: periphyton dry weight on container walls.
§ Evenness of substrate dwelling zooplankton analyzed with periphyton dry weight as continuous predictor variable.
TABLE D2. Results of general linear models that explain the taxon richness of substrate dwelling and seston feeding zooplankton and of phytoplankton in relation to the plant treatment (categorical variable), the imposed initial primary productivity gradient (CHLAin) and the realized productivity gradient at the end of the experiment (CHLAfin).
Organism group |
df |
SS |
MS |
F |
P |
|
Substrate dwellers |
Intercept |
1 |
259 |
259 |
182 |
< 0.001 |
PLANT |
2 |
120 |
60 |
42 |
< 0.001 |
|
CHLAin |
1 |
9.9 |
9.9 |
6.99 |
0.011 |
|
CHLAfin |
1 |
4.2 |
4.2 |
2.92 |
0.093 |
|
Error |
54 |
77 |
1.4 |
|||
Seston feeders |
Intercept |
1 |
340 |
340 |
357 |
< 0.001 |
PLANT |
2 |
5.1 |
2.6 |
2.7 |
0.078 |
|
CHLAin |
1 |
8.9 |
8.9 |
9.3 |
0.004 |
|
CHLAfin |
1 |
7.1 |
7.1 |
7.4 |
0.009 |
|
Error |
54 |
52 |
0.95 |
|||
Phytoplankton |
Intercept |
1 |
722 |
722 |
125 |
< 0.001 |
PLANT |
2 |
198 |
99 |
17 |
< 0.001 |
|
CHLAin |
1 |
5.9 |
5.9 |
1 |
0.316 |
|
CHLAfin |
1 |
40 |
40 |
6.9 |
0.011 |
|
Error |
53 |
307 |
5.8 |
Note: Nonsignificant interactions between the predictor variables were omitted from the GLM models. For more information on the statistical analyses see Materials and Methods.
TABLE D3. Results of general linear models that explain the evenness of substrate dwelling and seston feeding zooplankton and of phytoplankton in relation to the plant treatment (categorical variable) the imposed initial primary productivity gradient (CHLAin) and the realized productivity gradient at the end of the experiment (CHLAfin).
Organism group |
df |
SS |
MS |
F |
P |
|
Substrate dwellers |
Intercept |
1 |
1.63 |
1.63 |
89.8 |
< 0.001 |
PLANT |
2 |
0.67 |
0.33 |
18.4 |
< 0.001 |
|
CHLAin |
1 |
0.10 |
0.10 |
5.72 |
0.020 |
|
CHLAfin |
1 |
0.299 |
0.299 |
16.5 |
< 0.001 |
|
Error |
52 |
0.943 |
0.018 |
|||
Seston feeders |
Intercept |
1 |
3.219 |
3.219 |
178.26 |
< 0.001 |
PLANT |
2 |
0.232 |
0.116 |
6.421 |
0.003 |
|
CHLAin |
1 |
0.063 |
0.063 |
3.479 |
0.068 |
|
PLANT × CHLAin |
2 |
0.240 |
0.120 |
6.649 |
0.003 |
|
CHLAfin |
1 |
0.104 |
0.104 |
5.777 |
0.019 |
|
Error |
52 |
0.939 |
0.018 |
|||
Phytoplankton |
Intercept |
1 |
4.26 |
4.26 |
232 |
< 0.001 |
PLANT |
2 |
0.06 |
0.03 |
1.63 |
0.205 |
|
CHLAin |
1 |
0.21 |
0.21 |
11.6 |
0.001 |
|
CHLAfin |
1 |
0.116 |
0.116 |
6.34 |
0.015 |
|
Error |
53 |
0.973 |
0.018 |
Note: Nonsignificant interactions between the predictor variables were omitted from the GLM models. For more information on the statistical analyses see Materials and Methods.
TABLE D4. Results of two alternative general linear models that explain the taxon richness of seston feeding zooplankton in relation to the proportional density of Daphnia (D. galeata and D. magna) and the plant treatment, with and without incorporation of the initial primary productivity gradient (CHLAin).
Explanatory variables‡ |
df |
SS |
MS |
F |
P |
|
CHLAin not in model |
||||||
Intercept |
1 |
768.95 |
768.95 |
799.87 |
< 0.001 |
|
PLANT |
2 |
3.42 |
1.71 |
1.78 |
0.178 |
|
DAPHNIA% |
1 |
6.94 |
6.94 |
7.22 |
0.010 |
|
Error |
55 |
52.87 |
0.96 |
|||
CHLAin in model |
||||||
Intercept |
1 |
194.41 |
194.41 |
252.75 |
< 0.001 |
|
PLANT |
2 |
3.50 |
1.75 |
2.28 |
0.112 |
|
CHLAin |
1 |
11.34 |
11.34 |
14.74 |
< 0.001 |
|
DAPHNIA% |
1 |
0.50 |
0.50 |
0.64 |
0.426 |
|
Error |
54 |
41.53 |
0.77 |
Notes: For more information on the statistical analyses see Materials and Methods. Species richness in this analysis refers to the total number of seston feeding zooplankton species except Daphnia and was obtained through rarefaction with cut off levels of 90 individuals.
‡ PLANT is the plant treatment (no macrophytes, artificial macrophytes, real macrophytes); DAPHNIA% is the relative density (%) of Daphnia in the community of seston feeding zooplankton; CHLAin is the initial chlorophyll a concentration at the moment of zooplankton inoculation and represents the initial primary productivity gradient.