Skip to main content
. 2016 Sep 7;6:32569. doi: 10.1038/srep32569

Table 3. The result of multinomial logistic regression for factors important for the presence of treefrog species within rice-paddy complexes.

Factor H. japonica
both species
B ± SE Wald df, P B ± SE Wald df, P
Complex 1
 Bush −0.14 ± 0.005 7.588 1, 0.006 −0.027 ± 0.012 5.392 1, 0.020
 Road 0.027 ± 0.024 1.232 1, 0.267 0.013 ± 0.031 0.186 1. 0.666
Complex 2
 Bush 0.001 ± 0.002 0.069 1, 0.793 0.005 ± 0.003 2.305 1, 0.129
 Road 0.019 ± 0.010 3.709 1, 0.054 0.007 ± 0.017 0.168 1, 0.682
 Forest −0.012 ± 0.002 23.824 1, <0.001 −0.010 ± 0.005 3.747 1, 0.053
Complex 4
 Bush −0.011 ± 0.006 3.360 1, 0.067 −0.022 ± 0.007 10.346 1, 0.001
 Road 0.062 ± 0.027 5.159 1, 0.023 0.055 ± 0.029 3.752 1, 0.053
 Forest −0.029 ± 0.011 7.571 1, 0.006 −0.032 ± 0.011 8.965 1, 0.003
Complex 5
 Bush 0.005 ± 0.003 2.978 1, 0.084 0.000 ± 0.004 0.002 1, 0.961
 Road 0.043 ± 0.010 18.549 1, <0.001 0.023 ± 0.011 4.242 1, 0.039
 Forest 0.001 ± 0.001 1.110 1, 0.292 0.000 ± 0.001 0.017 1, 0.897

There are four possible combinations for the occurrences of the two treefrog species: H. suweonensis only, H. japonica only, both species, and neither Hylid species. Because the sample sizes for H. suweonensis-only populations were small, multinomial logistic regressions were conducted for H. japonica only and for both species. “Forest” could not be calculated for complex 1. Significant values are in bold.