234 



Fishery Bulletin 100(2) 



1- 



10 20 30 40 50 60- 



July 



3n 



1- 



Mil 



□_ 



10 20 30 40 50 60- 



October 



n n 



10 20 30 40 50 60- 



August 



Ha 



^ r-l _ 



4 

 2 

 



10 20 30 40 50 60- 



Total 



nnn 



10 20 30 40 50 60- 10 20 30 40 50 60- 



Catch (number of fish) 



Figure 2 



Distributions of the catch rate of ayu by month for 104 sampled season-permit 

 anglers in the Nakagawa River. 



modes of the number of fishing days per season-permit an- 

 gler were five for all months. The variation in the niunber 

 of fishing days among anglers was also large (Fig. 3). 



Monthly plots of the total number of fish caught versus 

 the number of fishing days showed linear relationships 

 (Fig. 4). The variation in Figure 4 indicates differences in 

 the skill of the anglers. The monthly number of anglers 

 decreased over the fishing season. Figure 5 shows the 

 monthly changes in the total number of fishing days, the 

 total number of fish caught, and the catch rate for the 104 

 sampled anglers. The decline in number offish caught was 

 largely due to the decrease in fishing days. The change 

 in catch rate indicated a decline in the abundance of the 

 stock. 



The mean body weight of ayu was greatest in June (Fig. 

 6) and was affected by a method of fishing for ayu called 

 "Tomo-zuri" angling, which takes advantage of the attack- 

 ing behavior of ayu when another fish enters its territory. 

 Anglers attach a "call" fish (a live ayu) above a treble hook 

 that snares the territorial wild fish, as it attacks the "call" 



fish. Because larger individuals establish territories ear- 

 lier than smaller ayu, fish caught in June were predomi- 

 nantly the larger individuals. 



Reflecting the monthly trend in the number of fishing 

 days, 89'7,_ of the total annual catches of season-permit an- 

 glers and 98'>; of those of day-permit anglers were taken 

 from June to August (Table 2). The catch by day-permit 

 anglers was substantially smaller than anticipated, esti- 

 mated at about 29^ of season-permit anglers' catch in both 

 numbers and total weight. CVs ranged from 7% to 12% 

 in June and July for all parameters; however, they were 

 higher in August and September, ranging from 10% to 

 209; . In October, CVs exceeded 43% for total catch in num- 

 ber and weight. The decreasing precision of the monthly 

 catch rate estimates was caused by the decrease in anglers 

 («,,) (Figs. 4 and 5). _ 



The CVs of annual estimates of M and Mj, by method 1 

 were about T'r. but that of R'"' was about 20':i (Table 2i 

 because we evaluated the covariance terms for the number 

 of catches and fishing days between months; those were 



