Table 1
Formula for modified “SM-organic” nutrient medium
| Ingredients | Quantity |
|---|---|
| Bottled water to makea | 1 L |
| CaHPO4 | 99 mgb |
| KH2PO4 | 75 mg |
| MgSO4.7H2O | 75 mg |
| Soluvit®c | 10 mL |
| Amina-Fed | 0.92 mLe |
| Kinetin | 5 mg |
| Saccharose (sucrose) | 10 g |
| Activated charcoal | 1 g |
| Danish agar | 12 g |
| Pineapple juice | 25 mL |
| 0.1 M H2SO4 for pH adjustment | 15–20 drops |
| Potato (Solanum tuber tuberosum) | 1 cm3 per culture vessel |
Table 2
Combinations of “dummy” or indicator variables that uniquely identify each scarification method for the purpose of logistic regression analysis. The NaClO method is specified as reference, in which case X2 = X3 = 0
| Variable | NaClO | CENTRIFUGE | PRESOAK |
|---|---|---|---|
| X2 | 0 | 1 | 0 |
| X3 | 0 | 0 | 1 |

Figure 1
Effect of time and three disinfection-plus-scarification methods (NaClO, CENTRIFUGE, and PRESOAK) on the proportion of germinated vessels of Anacamptis laxiflora 189 d after sowing. Markers are measured points (n = 16) and lines are fitted models.

Figure 2
Germinated seeds of Anacamptis laxiflora 189 d after sowing.
Table 3
Logistic regression results and parameter estimates for log odds-ratios from Eq. [2]. The regressors are duration of soaking in NaClO and scarification method. The NaClO method was used as the reference
| Model | -LL | DF | χ2 | p > χ2 |
|---|---|---|---|---|
| Difference | 142.21 | 3 | 284.42 | <.001 |
| Full | 145.08 | |||
| Reduced | 287.29 | |||
| R2 | 0.50 | |||
| N | 432 |
| Term | N | Estimate | SE | z | p > |z| | OR | 95% C.I. |
|---|---|---|---|---|---|---|---|
| Intercept (β0) | 144 | -3.8866 | 0.447 | -8.694 | <.001 | 0.021 | [0.008 – 0.047] |
| Duration (β1) | 432 | 0.1023 | 0.009 | 10.790 | <.001 | 1.108 | [1.089 – 1.130] |
| Centrifuge (β2) | 144 | 0.1969 | 0.363 | 0.543 | 0.587 | 1.218 | [0.598 – 2.492] |
| Presoak (β3) | 144 | 0.9947 | 0.373 | 2.664 | 0.008 | 2.704 | [1.315 – 5.709] |

Figure 3
Effect of three disinfection-plus-scarification methods on the proportion of germinated vessels at 41 versus 189 d. At 189 d, the parameter estimates for the odds-ratios were not statistically significant for CENTRIFUGE (p>|t| = 0.720) and PRESOAK (p>|t| = 0.068) compared to the reference method, whereas at 41 d the parameter estimate was statistically significant for CENTRIFUGE (p>|t| < 0.001). Centrifugation seems to severely delay germination, although given enough time the proportion of germinated vessels catches up.

Figure 4
Inverse prediction: The curves predict disinfection/scarification time for a given germination rate, according to fitted model.