Table 2. Model results for shrub density in the South Field in 2016.
| Source (fixed factors)a | Shrub density (plants m-2)b | ||
|---|---|---|---|
| df c | F | P-valued | |
| (i) 3-factor model: | |||
| Irrigation | 1,2.084 | 0.23 | 0.67 |
| Seeding strategy | 3,24 | 10.34 | <0.001 |
| Shrub origin | 1,28 | 35.58 | <0.001 |
| Irrigation x Strategy | 3,24 | 1.93 | 0.15 |
| Irrigation x Shrub origin | 1,28 | 0.18 | 0.68 |
| Strategy x Shrub origin | 3,28 | 10.80 | <0.001 |
| Irrigation x Strategy x Shrub origin | 3,28 | 0.27 | 0.85 |
| (ii) 4-factor model | |||
| Irrigation | 1,2 | 0.02 | 0.91 |
| Seeding strategy | 1,12 | 29.88 | <0.001 |
| Shrub origin | 1,16 | 15.49 | 0.001 |
| Grass origin | 1,12 | 4.00 | 0.07 |
| Irrigation x Strategy | 1,12 | 1.89 | 0.19 |
| Irrigation x Shrub origin | 1,16 | 0.34 | 0.57 |
| Irrigation x Grass origin | 1,12 | 5.26 | 0.04 |
| Strategy x Shrub origin | 1,16 | 14.43 | 0.002 |
| Strategy x Grass origin | 1,12 | 0.32 | 0.58 |
| Grass origin x Shrub origin | 1,16 | 0.36 | 0.56 |
| Irrigation x Strategy x Shrub origin | 1,16 | 0.33 | 0.57 |
| Irrigation x Strategy x Grass origin | 1,12 | 2.24 | 0.16 |
| Irrigation x Shrub origin x Grass origin | 1,16 | 3.23 | 0.09 |
| Strategy x Shrub origin x Grass origin | 1,16 | 0.07 | 0.80 |
| Irrigation x Strategy x Shrub origin x Grass origin | 1,16 | 0.08 | 0.78 |
Model results for shrub density in the South Field in 2016, shown for the (i) 3-factor model (Model P<0.001, Adj. R2 = 0.84, N = 72) and (ii) 4-factor model (Model P = 0.002, Adj. R2 = 0.73, N = 48). Too few shrubs were encountered in the North Field for analysis. See Methods for details.
a Fixed factors (and levels) included in the 3-factor model: Irrigation treatment (Spring, Fall+Spring), seeding strategy (I, II, III, IV), and shrub origin (local, distant). The 4-factor model added grass origin (wild-collected, commercial) and included only seeding strategy levels I and II (i.e., the treatments where grasses were seeded). Random effects of block, sub-block (nested within block), and plot (nested within sub-block, block) were also included in models but significance of these effects are not shown.
b Shrub density was log transformed prior to analysis.
c Degrees of freedom (df), indicated as numerator, denominator df.
d Bolded P-values indicate significant (P<0.05) model effects.