Table 4. Model results for Red River Delta, Wave 1 (December ‘03 –February ‘04).
GLMM | BRT | |||||
---|---|---|---|---|---|---|
Coefficient | p | s.e. | Relative Influence (%) | s.e. | Rank | |
Intercept | 108.702 | 0.001 | 33.225 | n/a | ||
Urbanicity | 0.91 | 0.04 | 13 | |||
Urbanicity: rural | 0 | |||||
Urbanicity: peri-urban | 0.009 | 0.986 | 0.518 | |||
Urbanicity: urban | 0.179 | 0.383 | 0.205 | |||
Land-use diversity (Gini-Simpson index) | 0.959 | 0.338 | 1.000 | 4.26 | 0.04 | 8 |
Chicken density* | 0.970 | 0.012 | 0.385 | 8.28 | 0.06 | 2 |
Duck-rice area density | -6.879 | 0.129 | 4.525 | 6.07 | 0.04 | 5 |
Chicken flock size diversity (Gini-Simpson Index) | 2.424 | 0.073 | 1.352 | 10.62 | 0.05 | 1 |
Duck & goose flock size diversity (Gini-Simpson Index) | 0.051 | 0.920 | 0.511 | 5.61 | 0.05 | 6 |
Percentage land under rice* | -1.643 | 0.362 | 1.802 | 5.03 | 0.06 | 7 |
Percentage land under aquaculture* | -0.974 | 0.799 | 3.832 | 6.30 | 0.04 | 3 |
Annual precipitation* | -32.626 | 0.003 | 11.114 | 6.21 | 0.09 | 4 |
Compound Topographical Index* | -7.193 | 0.355 | 0.119 | 3.36 | 0.04 | 10 |
Shortest distance to nearest national highway* | -0.079 | 0.082 | 0.045 | 1.59 | 0.06 | 12 |
Shortest distance to nearest provincial highway* | -0.027 | 0.472 | 0.038 | 0.49 | 0.09 | 14 |
Shortest distance to nearest town* | 0.173 | 0.034 | 0.082 | 2.27 | 0.05 | 11 |
Shortest distance to nearest lake* | -0.041 | 0.730 | 0.119 | 3.57 | 0.09 | 9 |
Autoregressive term | n/a | 35.43 | 0.04 | |||
AUC-ROC | 0.802 | Trg = 0.827, Eval = 0.737 |
* Transform of the type log10(1+x) was used
s.e. = standard error, Rank = rank of relative influence excluding the rank of the autoregressive term, AUC-ROC = Area Under the Curve of the Receiver Operating Characteristic, Trg = Training, Eval = Evaluation