a common set of survey stations. The inclusion of a single, rapidly re- 

 treating station in one period but not the other would drastically influ- 

 ence the difference in the means for the two periods. Recession data are 

 also less variable on the relatively high part of the beach face (see 

 Fig. 9). Using these two considerations, the average recession of the 

 176.92-meter contour was calculated for the 14 stations which included 

 that contour on surveys for all 3 years. The resulting average rate of 

 recession between 1971 and 1975 was not less, but 37 percent greater than 

 during the previous period. Thus, the mean recession rate measured be- 

 tween 1971 and 1975 increased even though the mean lake level had been 

 falling slowly during the last 2 years of this 4-year period. Rates of 

 recession calculated at the other principal elevations (see Fig. 9) also 

 increased for the 1971-75 period. The reason recession increased is 

 unknown; however, wave activity may have been more intense during this 

 period. Johnson and Hiipakka (1976) report that two unusually destruc- 

 tive storms in the 1972-73 storm season evidently removed 1.5 times as 

 much bluff material as had been eroded during the preceding 2.5 years at 

 the site of a temporary harbor near Bridgman, Michigan, 160 kilometers 

 south of the present study area. The cumulative effect of storm vari- 

 ability in the present study area is unknown, but since recession rates 

 did not decline there is no evidence that the beaches were approaching 

 equilibrium before 1975. 



The data are, therefore, consistent with the concept that the shore 

 lags several years behind in its response to the termination of a rapid 

 rise in water levels. 



The magnitude of the lag in terms of how much additional shore re- 

 cession actually occurred between the time when lake levels stabilized 

 and the time when profiles finally equilibrated, can not be calculated 

 directly because there was no survey during the year when levels first 

 stabilized. A good estimate, however, of the "latent recession" (i.e., 

 the response to the inherited stress which was not relieved by profile 

 adjustment until after the lake level had peaked) can be obtained by 

 assuming shore recession continued until the water level peaked in July 

 1973 at the same rate as existed between 1969 and 1971 (1.91 meters per 

 year). By subtracting the estimated recession before peak levels (1.91 

 meters per year X 25/12 years = 3.98 meters) from the known recession for 

 June 1971 to August 1975 (14.38 meters averaged from the same stations), 

 the remaining difference (10.6 meters) should be the recession which 

 occurred after the July 1973 peak in order to bring the profiles to the 

 near-equilibrium conditions interpreted for 1975. Given the uncertainties 

 involved, primarily the large variation in recession between stations and 

 the less than desirable timing of peak water levels between widely sep- 

 arated surveys, plus the general nature of the original prediction which 

 was based simply on early profile steepening, the extremely close agree- 

 ment between the 10.6 meters of calculated recession and the 11 meters 

 predicted must be largely ascribed to chance. The close agreement cer- 

 tainly supports the prediction procedure, but should not be taken as 

 indicative of the precision to be expected with this method. Additional 

 detailed, long-term studies of profiles, which are adjusting to new water 



34 



