Skip to main content
. 2011 Jul 12;6(7):e21115. doi: 10.1371/journal.pone.0021115

Table 1. Modified log-likelihood [−2log(L)] and parameter estimates for the different models.

Dispersal kernel parameters βh Preference parameters αh df −2.log(L)
4 habitats
βP βC βB βD αP αC αB αD
M0 4 1.62 0.54 0.43 0.018 0.008 4 12991
MG 4 2.01 2.11 1.14 0.71 0.54 0.40 0.036 0.014 7 12394
3 habitats (grouping Host plant+Crop residue)
βP+C βB βD αP+C αB αD
M0 3,(P+C) 1.63 0.95 0.01 0.04 3 12991
MG 3,(P+C) 2.04 1.09 0.74 0.91 0.02 0.07 5 12395
3 habitats (grouping Ditch+Bare soil)
βP βC βB+D αP αC αB+D
M0 3,(B+D) 1.62 0.55 0.43 0.02 3 12993
MG 3,(B+D) 1.97 2.14 1.08 0.56 0.41 0.03 5 12422
2 habitats (grouping Host plant+Crop residue and Ditch+Bare soil)
βP+C βB+D αP+C αB+D
M0 2 1.63 0.97 0.03 2 13014
MG 2 2.04 1.08 0.94 0.06 3 12445
1 single habitat (grouping Host plant+Crop residue+Ditch+Bare soil)
βP+C+B+D αP+C+B+D
MG 1 = MG 2 1.89 0.25 1 14769

Parameter subscripts: P, host plant; C, crop residue (litter-covered soil); D, ditch; B, bare soil. P+C means that host plant and crop residue habitats are pooled in a single category.

Habitat-dependent (resp. habitat-independent) models with k types of habitat are denoted MG k (resp. M0 k).