|
Model
|
Fixed effects
|
Random effect
|
Random slope
|
Response
|
SE
|
AIC
|
ΔAIC
|
χ
2
|
DF
|
p
|
|---|
|
Home range area
(km
2
)
|
Month
|
PTT
|
Month|PTT
|
−215.7
|
36.8
|
1092
|
2
|
7.68
|
1
|
0.006
|
|
Month
|
PTT
|
Month|PTT
|
−214.9
|
38.3
|
1091
|
1
|
11.28
|
2
|
0.004
|
|
+ DTM
| | |
2.7
|
1.4
| | | | | |
|
DoY
|
PTT
|
DoY|PTT
|
−5.8
|
1.3
|
1094
|
4
|
2.27
|
1
|
0.132
|
|
DoY
|
PTT
| |
−6.5
|
1.2
| | | | | |
|
+ DTM
| |
DoY|PTT
|
3.4
|
1.3
|
1090
|
0
|
8.08
|
2
|
0.018
|
|
Centroid pairwise distance (km)
|
Month
|
PTT
|
Month|PTT
|
−21.3
|
4.7
|
719
|
2
|
5.39
|
1
|
0.020
|
|
Month
|
PTT
| |
−21.454
|
5.451
| | | | | |
|
+ DTM
| |
Month|PTT
|
0.285
|
0.145
|
717
|
0
|
9.72
|
2
|
0.008
|
|
DoY
|
PTT
|
DoY|PTT
|
−0.6
|
0.1352
|
722
|
5
|
5.00
|
1
|
0.025
|
- P-value and χ2 values were obtained by analysis of variance tests of each of the full models against a null model. Null models had Distance To Mainland (DTM) as the only fixed effect, to determine if month or Day of Year (DoY) was a significant predictor variable. DoY + DTM centroid pairwise distance model failed to converge and has been omitted. The Platform Transmitting Terminal (PTT) is unique to each whale, and was used as a random effect with by-whale random slope to capture the variation in movements between individual whales. Response is the unit change in the dependent variable (home range area in km2 or centroid pairwise distance in km) per unit increase in the fixed effect variable(s). For example, for each unit increase in month, the home range area decreased by 215.7 km2 with a standard error of ±36.8 km2. The selected model is bolded. The most parsimonious model within two Akaike information criterion (AIC) units of the lowest AIC was selected. SE = Standard Error; DF = Degrees of Freedom; p = p-value.
