the species composition of catch unknown ex ante. Such an assumption at

the firm level is most probably poor at best. Florida, for example, has

reasonably distinct grouper and red snapper fisheries (Cato and

Prochaska, 1977). The main use of the estimated price equations,

however, is directed towards analyzing the fishery in aggregate. Thus,

the validity of the use of the estimated price equations in applying

economic optimization criteria to the estimated catch equations depends

on the degree to which the fishery in aggregate behaves as if non-

selective fishing occurs.

 The estimated price equations are presented in Tables 6 and 7.

Table 6 contains the single equation ordinary least squares parameter

estimates of the price equations which were specified in equations (67)

and (68). It may be recalled that substantial contemporaneous correla-

tion in the disturbances of these equations was anticipated. This

a priori expectation implied that a two stage Aitken's estimator was

appropriate for estimating the price equations. The parameter estimates

resulting from this estimator are presented in Table 7.

 A priori, all of the estimated parameters corresponding to the

catch variables were hypothesized to be negative. Examination of the

parameter estimates in Table 6 demonstrates that the single equation

ordinary least squares estimator resulted in several wrong signs.

However, all of the time trend variables had the expected (positive)

sign. In contrast, estimation using the Aitken's two stage estimator

resulted in only one parameter estimate of the wrong sign. As with the

ordinary least square estimates, all of the estimated time trend parame-

ters take on the expected sign.