these extremes. The samples in 195^ and 1957 indicated that the size 

 distribution for each reservoir remained approximately the same for both 

 years. The difference in mean lengths between the different locations 

 vas highly significant, indicating that a real difference exists in the 

 size composition of the catch. However, the data from each location, 

 when plotted, fit a normal distribution curve (figure k) . 



Population sampling . --After the mark and recapture study was 

 initiated on May 15, 1956^ in Wheeler Reservoir, 33 tags were recovered 

 during the remaining 3-nionth period in 1956 and 5^ tags during the 6-month 

 harvest season of 1957 (table 8) . The data from tag returns and the catch 

 of unmarked pigtoes by the commercial fishery were subjected to a chi- 

 square test to determine if the rate of tag recovery was random by time 

 periods in proportion to the total catch. The test showed that the differ- 

 ence in the ratio of tagged to vmtagged pigtoes caught for the various 

 months was not significant (X^r 8.17; P= 0.25). Based on the mortality 

 rate of the 100 tagged specimens used as controls (25 per cent in a con- 

 fined area), the total number of tagged survivors available to the fish- 

 ery was estimated at 750* 



Using the Petersen method, the pigtoe popiilation was estimated 

 at 29,55i<-,000, with upper and lower limits of 37^500,000 and 21,600,000 

 at the 95-percent level (table 9) • This estimate represents the escape- 

 ment population in the 7^^80>000 square yards of fishing area as of 

 August 31^ 1957^ after subtracting the commercial catch for the previous 

 9 -month period. 



The pigtoe population size by the area-density method, based on the k8 

 random SCUBA samples in Wheeler Reservoir, was estimated to be 20,566,000 

 with an upper limit of 28,199,000 and a lower limit of 12,9^+0,000 at the 

 95-percent level (table 10) . 



Since SCUBA sampling provided a direct method of measixring abun- 

 dance, the population estimate by this method was accepted as being more 

 representative of the true population than that arrived at by the Petersen 

 method. In addition, bottom sampling provided the only means of studying 

 the total population, composed of all species and available sizes. 



The best estimate of the size of the total population of all 

 species and sizes on the Wheeler beds in August 1957 vas 38,970>000 

 (table ll) . The population was composed of l8 species, 11 of which are 

 of commercial value. The pigtoe accounted for 52.7 percent of the total 

 population. The rest of the population was mostly of six other species: 

 pink wartyback (8.8 percent), white wartyback (7»8 percent), three-horn 

 {^.k percent), three-ridge (6.6 percent), elephantear (5.8 percent), and 

 butterfly (k.k percent). 



The population was not distributed at random on the bottom area 

 but was concentrated on specific beds. Mussels were found at only 39 of 

 the 48 stations sampled; density at popxilated stations varied ftom 1 to 

 21 mussels per square yard. The pigtoe occurred at 37 stations; density 

 varied from 1 to l6 per square yard. 



16 



