Comparison of ordinal logistic regression models for survival of C. multigeniculata joint cuttings prior to initiating watering treatment (6 July–13 Aug) and comparison of failure-time models for survival of joint cuttings during watering treatment (13 Aug–20 Nov).
| Survival prior to watering treatmentz | |||
|---|---|---|---|
| Model | AICc | ΔAIC | wi |
| Intercept | 112.9053 | 45.8534 | 0.0000 |
| CI | 67.0518 | 0.0000 | 0.8963 |
| CI Soil|Water | 71.3648 | 4.3130 | 0.1037 |
| Survival during watering treatmenty | |||
|---|---|---|---|
| Model | AICc | ΔAIC | wi |
| Intercept | 57.316 | 9.678 | 0.0040 |
| Soil | 47.638 | 0.000 | 0.5101 |
| Soil Water | 49.191 | 1.553 | 0.2347 |
| Soil Water Vol | 51.355 | 3.717 | 0.0795 |
| Soil Water CI | 51.571 | 3.933 | 0.0714 |
| Soil|Water | 51.894 | 4.256 | 0.0607 |
Notes: Models are ranked by ΔAICc and only those models with some level of support (ΔAICc < 7) are presented, along with the intercept model for comparison. Akaike weights (wi) are included for comparing the relative importance of the initial joint cutting and treatment attributes: Joint cutting condition index (CI), joint cutting volume (Vol), soil mix (Soil), and watering treatment (Water). Maternal plant did not occur in any of the models with ΔAICc < 7.
↵z Analysis prior to watering treatment implementation (1 mo) included all 80 C. multigeniculata joint cuttings collected.
↵y Analysis during watering treatment included the 39 C. multigeniculata joint cuttings that remained alive after 1 mo, minus 3 joint cuttings that never rooted during the experiment although the shoots remained green and succulent (N = 36).