Table IV
Coefficients and fit statistics for linear mixed effects model of natural logarithm of number of sprouts per meter of pruned bole in 2002 and 2006 ln (N + 1) , two and six growing seasons after pruning respectively, as a function of redwood crown ratio remaining immediately after pruning in 1999–2000.
| Model component | Data | Parameter | Coefficient | s.e. | d.f. | t-value | Pr > | t | |
|---|---|---|---|---|---|---|---|
| Fixed effects | 2002 | Intercept (fi0) | 4.387495 | 0.390 | 10 | 11.24 | < 0.0001 |
| Slope (fi1) | –5.631559 | 0.726 | 10 | –7.76 | < 0.0001 | ||
| 2006 | Intercept (fi0) | 2.752960 | 0.357 | 8 | 7.71 | 0.0001 | |
| Slope (fi1) | –3.802147 | 0.738 | 8 | –5.15 | 0.0009 | ||
| No. groups | Std.dev. | Var. | |||||
| Random effects1 | 2002 | Study site | 7 | 0.345 | 0.119 | ||
| Treatment | 17 | 0.078 | 0.006 | ||||
| Clone | 130 | 0.359 | 0.129 | ||||
| Section | 157 | 0.644 | 0.415 | ||||
| Residual | – | 0.754 | 0.569 | ||||
| 2006 | Study site | 7 | 0.001 | 0.000 | |||
| Treatment | 17 | 0.270 | 0.073 | ||||
| Clone | 120 | 0.004 | 0.000 | ||||
| Section | 146 | 0.462 | 0.213 | ||||
| Residual | – | 0.844 | 0.712 | ||||
| Model | 2002 | n = 168; Bias = –5.75; RMSE = 22.6 | |||||
| 2006 | n = 155; Bias = –3.00; RMSE = 13.2 | ||||||
No. groups = number of data groups for each nested parameter e.g., 17 treatments nested within 7 study sites; s.e. = standard error for coefficient; d.f. = degrees of freedom; Std.dev. = standard deviation, a measure of variation attributed to random effect parameter; Var. = Std. dev squared; n= number of observations; Bias = mean of prediction errors (Eq. (1)), and RMSE = square root of the average squared prediction error (Eq. (2)) for number of sprouts per meter of stem predicted using equation (4).