temporal trends in mussel populations among and within 

 the Reach 15 study sites. Density distributions combine 

 both mean density (no./m-^ and frequency distributions for 

 a species within each study site (i.e., % of population by 

 age or 5-mm size intervals). For example, the mean 

 density for A. plicata at Illiniwek (RM 492.4) was 

 10.34/m' and the percent of mussels within the 60-mm 

 shell length interval (55.01 to 60.00 mm) was 14.6%; 

 therefore, the calculated density of this size interval is 

 10.34/m- X 0.146 = 1.51/m-. Density distnbutions were 

 presented as histograms and in tabular format, the latter 

 allowing one to calculate the mean density of a specific 

 age group or size range by summing the mean densities of 

 all mussels within the desired group or range. 



Recruitment 



We evaluated recent recruitment for ten of the 

 more common mussel species we collected in Reach 15 

 during 1994-95 (Appendix F). The size criteria to define 

 a recent recruit was species specific and typically 

 represented mussels less than three years of age. For 

 most sfMscies, individuals less than 30-mm in length 

 constituted recent recruitment. However, the size was 

 reduced for small, short-lived species such as Truncilla 

 truncata (< 15 mm), Obliquaria reflexa (< 15 mm), and 

 T. donaciformis (<10 mm). Length-frequency and 

 density tables were used to determine the percentage (%) 

 and density (no./m'^ of recent recruits within the 

 population at each study site and for each year sampled 

 (i.e. Sylvan Slough 1983, 1985, 1987, and 1994-95) to 

 evaluate recruitment patterns over the past decade. 



Age and Growth 



The relationship between mussel age and growth 

 was evaluated using regression plots and regression 

 formulas. Therefore, it is crucial that the reader have a 

 basic understanding of these two techniques. We offer 

 the following brief explanations: 



R^ression plots are used to determine the degree of 

 relationship between the independent (X) and 

 dependent (Y) variables. Regression plots 

 attempt to fit a line to a series of data having 

 specific X,Y coordinates. The more closely the 

 data points fall along the line the better the 

 relationship. The proportion (or percentage) of 

 the total variation in Y that is explained or 

 accounted for by the fitted regression is termed 

 the coefficient of determination, r^, which may 



be thought of as a measure of the strength of the 

 relationship. 



Rq;ression formulas are mathematical equations which 

 describe the relationship between the X and Y 

 variables by evaluating the regression coefficient 

 or slope (b) and the y-intercept (a) of the best fit 

 regression line (2Lar 1984). Knowing the 

 parameter estimates of a and b for the regression 

 equation, one can calculate the value of Y 

 (dependent variable) at a stated value of X 

 (independent variable). The closer the r value 

 is to 1 the less variability there is in the data and 

 therefore the more reliable the estimate of Y. 



The species and number of individual mussels 

 used in growth analysis were limited to those which we 

 had aged or weighed in 1987 and 1994-95 (Table I). No 

 distinction was made regarding collection location (study 

 sites); rather, growth analysis was based on composite 

 mussel samples from all Reach 15 study sites. We used 

 a stepwise procedure (Zar 1984) in selecting the 

 regression formula which consistently provided the best fit 

 (i.e., highest r^) for mussel growth data. 



Age-size relationships were best described by S"*" 

 order polynomial regression formulas (y = a -I- b|X -I- 

 b-,x- + bjx'). Mean shell measurements of each of the 

 five commercial species (Appendix G, Part II) served as 

 the dependent variables and mussel age as the independent 

 variable in growth curves (regression plots). Regression 

 formulas were used to calculate shell size (i.e., length, 

 width, and height) at ages from 1 to 30 years. By 

 switching the variables we derived regression formulas for 

 each of the five commercial species to calculate age for a 

 given shell length or shell height. Formulas based on 

 shell length and age were used to calculate the age of all 

 mussels which had not been aged. 



Size-weight relationships were best described by 

 power regression formulas (y = ax''). Live and dry shell 

 weights of individual mussels served as the dependent 

 variables and shell length and height as the independent 

 variables in growth curves. Regression formulas were 

 used to calculate live weight and dry shell weight given 

 either shell length or height. 



Mussel age-frequency histograms were 

 constructed for five commercial species, A. plicata, M. 

 nerwsa, Q. quadrula, Q. metanexra, Q. pustulosa, and 

 two non commercial species, E. lineolata and O. reflexa. 

 These histograms represented all individuals regardless of 

 whether their ages had been determined from counting 

 growth bands (estimated) or calculated from 3"* order 



