Table 1. Periodic regression results; N = 29 in all cases. Top catch is the catch of the most experienced anglers.

Dependent variable	Best predictor	Number of fisha or hoursb	F	P c	R2	Max. relative effectd (%)	Best two predictors	P	R2	Max. relative effect (%)
Muskies Inc. catch data
Daytime catche	cosine2θ	228,172	37.4	<0.0001e	0.58	4.8	cosine2θ, −sineθ	<0.0001	0.59	4.8
Nighttime catchf	cosineθ	7708	18.6	0.0002	0.41	14.2	cosine2θ, cosineθ	<0.0001	0.74	28.2
Ontario catch	cosine2θ	73,421	81.5	<0.0001	0.75	10.1	cosine2θ, −cosineθ	<0.0001	0.77	11.6
Minnesota catch	cosine2θ	28,227	7.0	0.01	0.21	4.0	cosine2θ, cosineθ	0.03	0.24	5.5
Wisconsin catch	cosine2θ	70,371	1.5	0.23	0.05	Ndg	cosine2θ, cosineθ	0.37	0.07	Nd
Ohio catch	−sine2θ	13,124	4.6	0.16	0.15	Nd	−sine2θ, −cosineθ	0.07	0.18	ND
Lake of the Woods catch	cosine2θ	22,259	31.4	<0.0001	0.54	10.4	cosine2θ, cosineθ	<0.0001	0.63	14.9
Mille Lacs catch	cosine2θ	1262	3.8	0.06	0.12	Nd	cosine2θ, sine2θ	0.08	0.17	Nd
Lake Vermilion catch	cosine2θ	1771	25.6	<0.0001	0.49	19.1	cosine2θ, −sine2θ	0.0001	0.50	19.7
Lake St. Clair catch	cosine2θ	17,058	13.8	0.0009	0.34	14.0	cosine2θ, −cosineθ	<0.0001	0.59	26.4
High expertise catchh	cosine2θ	50,171	21.7	<0.0001	0.46	5.3	cosine2θ, −sine2θ	0.0001	0.50	5.6
Low expertise catch	cosine2θ	50,802	21.8	<0.0001	0.45	5.0	cosine2θ, −sineθ	0.0002	0.48	5.1
Department of Natural Resources creel data
Wisconsin effort	−sineθ	91,095	1.4	0.25	0.05	Nd	cosineθ, −sineθ	0.32	0.08	Nd
Mille Lacs effort	cosine2θ	44,856	1.2	0.29	0.04	Nd	cosine2θ, sineθ	0.55	0.04	Nd
Lake Vermilion effort	cosine2θ	17,113	12.4	0.002	0.31	23.8	cosine2θ, cosineθ	0.001	0.40	36.0
Wisconsin CPUE	−sine2θ	1445	0.48	0.49	0.02	Nd	−sine2θ, sineθ	0.64	0.03	Nd
Mille Lacs CPUE	cosine2θ	358	2.2	0.15	0.07	Nd	cosine2θ, −cosineθ	0.17	0.13	Nd
Lake Vermilion CPUE	sine2θ	334	3.2	0.08	0.11	Nd	cosine2θ, cosineθ	0.12	0.15	Nd

^aCatch and CPUE models.

^bEffort models.

^cThe familywise significance rate (α) for 8 geographic group F-tests using the Muskies Inc. catch data is 0.006; the familywise significance rate for 6 tests using the creel data is 0.009.

^dThe percent increase in predicted number of muskellunge caught or predicted hours fished on the peak days relative to catch or fishing effort on random days. For catch, the maximum relative effect was calculated using back-transformed predicted catch.

^eAll catch models except “Night” excluded nighttime records and records from November – May.

^fIncludes all night records except records from November – May.

^gNd  =  not determined (regression F not significant).

^hExcludes nighttime records and records from November – May.