FISHERY BULLETIN: VOL. 74, NO. 2 



Apaiach.col. 



Figure l. — The Apalachicola Bay system with permanent sam- 

 pling stations for long-term studies concerning fluctuations 

 of populations of epibenthic fishes and invertebrates. 



bottom with a 1-liter Kemmerer bottle. Tempera- 

 ture was measured with a stick thermometer 

 and/or a YSP dissolved oxygen meter. Salinity 

 was determined with a temperature-compensated 

 refractometer periodically calibrated with stan- 

 dard seawater. Color was measured with a (Hach) 

 American Public Health Association platinum- 

 cobalt standard test while turbidity was deter- 

 mined with a Hach model 2100A turbidimeter. 

 Light penetration readings were taken with a 

 standard Secchi disk. River flow data were pro- 

 vided by the U.S. Army Corps of Engineers 

 (Mobile, Ala.) while local climatological informa- 

 tion was provided by the Environmental Data 

 Service, NOAA, U.S. Department of Commerce. 



Biological collections were made with 5-m (16- 

 foot) otter trawls (%-inch mesh wing and body; 

 %-inch mesh liner). Repetitive, 2-min trawl tows 

 were taken at each station at speeds of 2-3 knots. 

 Seven subsamples were taken at stations 1, 2, 4, 

 5, and 6 while two samples were taken at stations 

 lA, IB, IC, 3, and 5A. All organisms were pre- 

 served in 10% Formalin, sorted and identified to 

 species, measured and/or counted (standard 

 length for fishes; total length for shrimps; 

 carapace width for blue crab, Callinectes sa- 

 pidus). Stations 1 and 4 were also sampled at 

 night, approximately 1-2 h after sunset for the 

 first 2 yr of the study. 



All statistical analysis was carried out using an 

 interactive computer program designed for the 

 study of extensive data collections. The extent of 



interstation community similarity was tested 

 using the C\ index of overlap (Morisita 1959; 

 Horn 1966). This index determines the probabil- 

 ity that two randomly drawn samples from popu- 

 lations X and Y will be the same species relative 

 to the probability that two individuals of the 

 same species will be drawn from population X or 

 Y alone. 



K = 





1 = 1 





i = l 



yr 



Yi 



^Yellow Springs Instrument Co. Reference to trade names 

 does not imply endorsement by the National Marine Fisheries 

 Service, NOAA. 



where S = number of species 



X, and J, = number of individuals of the 

 ith species in populations 

 XandF 

 X and Y = total number of individuals in 



the two communities 

 k^ and X^ = measures of diversity (Simp- 

 son 1949) as modified for 

 sampling with replacement 

 (Horn 1966). 



Values for this index range from (no species in 

 common) to 1. A hierarchical (stepwise) multiple 

 regression analysis was carried out using 

 monthly population size as the dependent vari- 

 able. Various physicochemical and biological 

 parameters (temperature, salinity, chlorophyll a, 

 turbidity, color, Secchi disk depth, total depth, 

 local rainfall, wind speed and direction, tidal 

 stage, river flow, and dissolved oxygen) were used 

 as the independent variables. All such functions 

 were tested in the same month of collection and 

 with a 1-mo lag in the physicochemical parame- 

 ters. Due to the relatively high number of inde- 

 pendent variables, the stepwise regression was 

 used whereby one variable at a time was sys- 

 tematically introduced into the equation, and, at 

 each step, the variable added was the one giving 

 the greatest increase in the multiple correlation 

 coefficient. While not necessarily giving the 

 "best" equation, this method is computationally 

 feasible, and frequently gives results comparable 

 to methods that would determine all possible re- 

 gressions. Since the salinity, color, and turbidity 



312 



