Estimating total body heat dissipation in air and water from skin surface heat flux telemetry in Weddell seals

Table 2 Summary LME model outputs predicting TBHD in wet and dry conditions

Model	State	k	ΔAIC	L-ratio
TBHD ~ HF_head + HF_neck + HF_axilla + HF_flank	Wet	4	0	9,654,101
TBHD ~ HF_flank	Wet	1	5,885,893	3,768,202
TBHD ~ HF_axilla	Wet	1	7,591,208	2,062,887
TBHD ~ HF_neck	Wet	1	8,518,031	1,136,064
TBHD ~ HF_head	Wet	1	9,411,749	242,346.1
Null model	Wet	0	9,654,093	–
TBHD ~ HF_head + HF_neck + HF_axilla + HF_flank	Dry	4	0	6,922,081
TBHD ~ HF_axilla	Dry	1	4,132,645	2,789,429
TBHD ~ HF_flank	Dry	1	5,942,993	979,081.3
TBHD ~ HF_neck	Dry	1	6,338,992	583,082.3
TBHD ~ HF_head	Dry	1	6,625,552	296,522.6
Null model	Dry	0	6,922,072	–

Models are fitted separately for the ‘wet’ and ‘dry’ subsets of TBHD data. AIC for single-variable and null models are presented relative to the full model (4 sensor inputs). Log-likelihood ratios (L-ratios) are relative to the Null models and indicate improved fit in all cases (p < 0.0001)

ISSN: 2050-3385