Table 3 Optimal discriminate analysis models for R. aurora breeding abundance addressing the seven variables examined. Presentation is identical to Table 2
| Â | Training analysisa | LOO analysisb | ||||||
|---|---|---|---|---|---|---|---|---|
Variable | Model | Number | Accuracy (%) | p< | ESS | Accuracy (%) | p< | ESS |
AREA | – | – | – | 0.1082 | 43.8 | – | – | – |
BULL | – | – | – | 0.5757 | 20.8 | – | – | – |
CON | If CON ≤ 96.26 predict ≤ 1 egg mass | 8 | 87.5 | 0.0001 | 87.5 | 75.0 | 0.0005 | 70.0 |
|  | If CON > 96.26 predict > 1 egg mass | 20 | 100.0 |  |  | 95.0 |  |  |
EMG | If EMG ≤ 2526 predict ≤ 1 egg mass | 8 | 100.0 | 0.0448 | 55.0 | – | 0.2184 | 25.0 |
|  | If EMG > 2526 predict > 1 egg mass | 20 | 55.0 |  |  |  |  |  |
FISH | If FISH = 0 predict ≤ 4 egg masses | 12 | 91.7 | 0.0390 | 41.7 | 91.7 | 0.0241 | 41.7 |
|  | If FISH = 1 predict > 4 egg masses | 16 | 50.0 |  |  | 50.0 |  |  |
FRP | If FRP ≤ 58.6% predict ≤ 40 egg masses | 19 | 84.2 | 0.0012 | 73.1 | 84.2 | 0.0005 | 73.1 |
|  | If FRP > 58.6% predict > 40 egg masses | 21 | 88.9 |  |  |  |  |  |
PCF | If PCF ≤ 56.0% predict ≤ 69 egg masses | 21 | 90.00 | 0.0006 | 81.0 | 76.2 | 0.0070 | 61.9 |
|  | If PCF > 56.0% predict > 69 egg masses | 7 | 100.00 |  |  | 85.7 |  |  |