TABLE 1.
Model set for response variablesa
| Variables in the model | Kb for models B and C | AIC for model: |
K for models A, D, and E | AIC for model: |
|||
|---|---|---|---|---|---|---|---|
| A | B | C | D | E | |||
| Null model | 2 | 1,148 | 798 | 14,745 | 1 | 15,127 | 2,526 |
| Source | 3 | 1,148 | 772 | 14,726 | 2 | 13,894 | 2,381 |
| Bacteria | 3 | 1,124 | 782 | 14,742 | 2 | 11,685 | 2,013 |
| Worm | 3 | 1,150 | 790 | 14,745 | 2 | 9,478 | 2,377 |
| Source + Bacteria | 4 | 1,124 | 749 | 14,722 | 3 | 10,906 | 1,712 |
| Source × Bacteria | 5 | 1,124 | 743 | 14,724 | 4 | 9,876 | 1,664 |
| Source × Bacteria + Worm | 6 | 1,126 | 737 | 14,724 | 5 | 8,261 | 1,511 |
| Source × Bacteria + Source × Worm | 7 | 1,126 | 737 | 14,725 | 6 | 8,196 | 1,512 |
| Source × Bacteria + Source × Worm + Bacteria × Worm | 8 | 1,128 | 738 | 14,727 | 7 | 7,862 | 1,469 |
| Source × Bacteria × Worm | 9 | 1,130 | 738 | 14,725 | 8 | 6,530 | 1,440 |
Response variables: A, fertility of the mother; B, number of colonized worms; C, bacterial load in the progeny; D, number of bacteria in plate; E, bacterial load in the mother. The best model for each data set is shown in boldface.
K, number of parameters. For models B and C, plate is included as a random effect (see Materials and Methods).