3. Results and Discussion . 



The results of the regression analysis of the seven growth periods 

 are given graphically in Figure 19. The arithmetic instantaneous growth 

 rate (Kg) for each of the periods is given in Table 8 and Figure 20. 



The seven regression lines are straight and represent a linear- 

 regression analysis. Growth in most biological systems is generally 

 found to be an exponential function best fitted by a curved line. Since 

 it was not apparent by observation of scatter diagrams whether the results 

 were best filled by a linear or exponential function, both linear and 

 curvilinear regressions on all data were run and compared with the 

 correlation coefficients. All the correlation coefficients from 

 rectilinear and curvilinear were so close that it appeared to make no 

 difference. Therefore, the linear analyses were chosen for presentation. 



In many growth studies, the growth rate is uniform except among the 

 youngest and oldest clams where the rate is accelerating or decelerating. 

 When the data are plotted, a typical sigmoid growth curve is produced. 

 Since growth measurements of clams below 12 millimeters and above 125 

 millimeters in length were lacking in regions where rapid growth rate 

 changes occurred, the data fit a linear regression. Typical curved 

 growth lines were reported for Pismo clams by Weymouth (1923) and Coe and 

 Fitch (1950) , but examination of the graphs indicated that the growth was 

 essentially linear between 10 millimeters and 110 millimeters, the range 

 of most length measurements. 



The seven regression lines were all significantly different from 

 and showed a negative slope, indicating a decrease in growth increment 

 with an increase in size (= age) . 



To determine if the growth rates for all clams differed at different 

 times of the year, two separate analyses were run. First, the slope (b^.^) 

 of each regression line against the slope of the other regression lines 

 of that year was tested to see if there was a significant difference. 

 Two tests were used, F and t (Sokal and Rohlf, 1969). Results of 

 the tests indicated a significant difference (P = less than 0.05) between 

 the following paired intervals: 30 June to 9 August 1972 and 11 August 

 to 28 September 1972; 30 June to 9 August 1972 and 26 October to 20 November 

 1972; 11 August to 28 September 1972 and 26 October to 20 November 1972; 

 28 September to 26 October 1972 and 26 October to 20 November 1972; 

 8 May to 3 June 1973 and 3 June to 3 July 1973; and 10 April to 8 May 1973 

 and 3 June to 3 July 1973. The growth rate of the Pismo clam in Monterey 

 Bay, as illustrated in the analysis of the mean daily growth rate. Kg, 

 was not constant throughout the year. The mean daily growth rate varied 

 considerably over the course of the year (Table 8 and Fig. 20). The 

 highest rates occurred in the summer and early fall, and the lowest in 

 late fall and spring. The meager existing data suggest that the growth 

 rate is low in the winter. Coe and Fitch (1950) found that the growth 

 rates were lowest from November through February and highest in the summer. 



47 



