Skip to main content
. 2014 Mar 19;9(3):e91726. doi: 10.1371/journal.pone.0091726

Table 4. AIC results for linear mixed-effects models where mean logit-transformed path straightness is a function of ecoregion, sea surface temperature (SST), logarithm of surface chlorophyll a concentration (chla), logarithm of SST gradient (SSTg), logarithm of surface chla gradient (chlag), and logarithm of bathymetry (bathy).

Model p1 AIC ΔAIC2 ω3
null 4 7258.06 390.51 0
ecoregion 9 6927.63 60.08 0
ecoregion + SSTa 10 6927.26 59.71 0
ecoregion × SSTb 11 6921.62 54.07 0
ecoregion + chla1 10 6908.31 40.76 0
ecoregion × chlab 15 6908.97 41.42 0
ecoregion + SSTga 10 6928.73 61.18 0
ecoregion × SSTgb 11 6929.90 62.35 0
ecoregion + chlaga 10 6923.94 56.39 0
ecoregion × chlagb 15 6922.50 54.95 0
ecoregion + bathymetrya 10 6871.17 3.62 0.07
ecoregion × bathymetryb 11 6873.15 5.60 0.03
ecoregion + SSTa + bathymetrya 11 6871.52 3.97 0.06
ecoregion × SSTb + bathymetrya * 12 6867.55 0.00 0.42
ecoregion + chlaa + bathymetrya 11 6871.76 4.21 0.05
ecoregion × chlab + bathymetrya 16 6868.89 1.34 0.21
ecoregion + SSTga + bathymetrya 11 6873.13 5.58 0.03
ecoregion × SSTgb + bathymetrya 12 6874.93 7.38 0.01
ecoregion + chlaga + bathymetrya 11 6871.61 4.06 0.05
ecoregion × chlagb + bathymetrya 16 6870.84 3.29 0.08
1

p: number of parameters in the model.

2

ΔAIC: difference in AIC value between best fitting model and other model.

3

ω: Akaike weight.

a

Slopes held constant in regions of interest.

b

Slopes and intercepts allowed to vary in regions of interest.

*Best fitting model.