ABSTRACT 



Each year pound nets fished in the ocean off the coasts of New York and New 

 Jersey catch large quantities of shad. The majority of these fish are believed to 

 be native to the Hudson and Connecticut Rivers and therefore, these catches should 

 be considered in any management plan for the two rivers. To establish a manage- 

 ment plan which would include the ocean fisheries, estimates of the racial composi- 

 tion of this catch must be made. 



In this paper an analysis of some meristic counts for shad is presented to support 

 the racial theory. Samples were examined and it was found that the meristic 

 counts used could be considered representative of the populations. An analysis 

 of variance of the characters gave evidence for the existence of races. 



A discriminant function is presented whereby a mixed sample of Hudson and 

 Connecticut River shad can be separated. Meristic data collected from Hudson 

 River shad in 1939 and Connecticut River shad in 1945 are used to construct the 

 discriminant function. The mean value of this function for the Hudson River, 

 1939, is 74.103 and for the Connecticut River, 1945, is 70.940. 



The discriminant function obtained will correctly classify approximately 81 

 percent of a mixed sample of Hudson and Connecticut River shad. Meristic data 

 collected from the Hudson River in 1940 were substituted into this discriminant 

 function and out of 105 fish, 16 were incorrectly classified; this is in good agreement 

 with the theoretical 19 percent misclassification. The number of misclassifications 

 can be considerably reduced if the individuals falling close to the mid-point between 

 the two populations are not classified. By refusing to classify about one-half of 

 the sample, the number of wrong classifications is reduced to 3.7 percent. Several 

 methods of estimating the population composition of a mixed sample of shad are 

 presented. 



