Table 3. Regression analyses between kelp cover and physical variables at the transects in Port Stephens (PS), Batemans (Ba) and Tasman Peninsula (Ta) on East Australia.^# .

PS	Variable	SS(trace)	pseudo-F 1,30	P	R2
Marginal tests	Depth	1360.10	10.77	<0.01	0.26
	Sand	293.53	1.81	0.11	0.06
	CDOM	49.23	0.29	0.60	0.01
	Chlorophyll	799.91	5.52	0.02	0.16
	Salinity	623.53	4.13	0.04	0.12
	Temperature	1816.20	16.35	<0.01	0.35
Sequential test	All combined				0.77
Best		BIC	R2
	All except CDOM	137.95	0.76
Ba	Variable	SS(trace)	pseudo-F 1,20	P	R2
Marginal tests	Depth	22.18	2.64	0.13	0.12
	Sand	12.80	1.44	0.25	0.07
	CDOM	44.89	6.18	0.02	0.24
	Temperature	0.09	0.01	0.92	0.00
Sequential test	All combined				0.39
Best		BIC	R2
	CDOM, Temperature, Chlorophyll	46.16	0.38
Ta	Variable	SS(trace)	pseudo-F 1,53	P	R2
Marginal tests	Depth	14561.00	77.01	<0.01	0.59
	Sand	9675.90	34.40	<0.01	0.39
	CDOM	2019.90	4.75	0.03	0.08
	Chlorophyll	10185.00	37.50	<0.01	0.41
	Salinity	11696.00	48.10	<0.01	0.48
	Temperature	5375.80	14.84	<0.01	0.22
Sequential test	All combined				0.74
Best		BIC	R2
	Depth, Temperature, Chlorophyll	279.64	0.73

^# The proportion of the variability explained by all variables combined was obtained in sequential tests using R² selection criterion and Forward selection procedure. The overall best solution was obtained using the Bayesian Information Criterion (BIC), using all possible combinations of variables. P-values were calculated using 9,999 permutations. Ba: salinity and chlorophyll were removed from the model as they were strongly correlated with depth and temperature, respectively (|r|>0.95; see Table A in S1 Appendix).