only twice in samples from Charlotte Harbor. In 

 Gulf waters adjacent to Tampa Bay, Charlotte 

 Harbor, and Pine Island Sound, G. breve was col- 

 lected every month. Regression analysis was used 

 to explore the relation between the mean monthly 

 concentrations of G. breve and the weighted average 

 input by the tributaries of Tampa Bay and Char- 

 lotte Harbor-Pine Island Sound. Weighted month- 

 ly inputs of iron, copper, phosphorus, and 

 chlorophyll "a" for the tributaries were calculated 



by the formula : 



X% ^X^ih) + X,i {11%) + Xui {111%) .... etc. 

 where X is the average value of the variable, /, //, 

 ... at the station or river, and 1%, !I%, . • ., the 

 percentage of the total monthly river discharge 

 for a particular river. 



Regression analysis was chosen because its 

 requirements are not as strict as those for corre- 

 lation, in which a normally distributed population 

 is prerequisite. Two relationships were tested — 

 polynomial and linear multivariate regression. 



In the polynomial regression, each variable 

 (total phosphorus, copper, iron, and chlorophyll 

 "a") was taken as the independent variable, and 

 G. breve as the dependent one. The linear, quad- 

 ratic, and cubic regression relationships were 

 calculated, and each tested for significance by 

 analysis of variance. Relationships were insignif- 

 icant at the 95-percent level. At the 80-percent 

 level, the linear regression relationship between 

 G. breve and iron was significant off both Tampa 

 Bay and Charlotte Harbor. One reason for the 

 weakness of this relationship may be that the 

 independent variables were not measured at the 

 places where G. breve was collected. 



SUMMARY 



A hydrological survey of HUIsborough, Alafia, 

 Little Manatee, Manatee, Myakka, Peace, and 

 Caloosahatchee Rivers, west Florida, was made 

 from January 1963 to February 1964. Monthly 

 changes in temperature, salinity, chlorophyll "a," 

 dissolved oxygen, inorganic phosphate-phosphorus, 

 total phosphorus, copper, and iron were deter- 

 mined. Data on precipitation and river discharges 

 and general geomorphological features of the area 

 are also given. 



The water temperature data for all rivers varied 

 from 12.8° C. (January 1965) to 32.4° C. (June 



and July 1964). Temperatures were low during 

 December, January, and February and variably 

 high from April through September. The vertical 

 differences of temperature were insignificant. 



Oligohaline (0.5 p.p.t. to 5 p.p.t.), mesohaline 

 (5 p.p.t. to 18 p.p.t.), and polyhaline (18 p.p.t. 

 to 30 p.p.t.) salinities were encountered at all 

 but two fresh-water stations. Salinity was reduced 

 in months of high rainfall and high river discharge. 

 The levels of chlorophyll "a" varied from 1.3 

 to 245.5 mg. per cubic meter and did not show 

 the expected seasonal cycle of spring and autumn 

 maximums and summer and winter minimums. 

 The highest concentrations were in the Peace 

 River at the fresh-water station. 



No serious depletion of dissolved oxygen was 

 indicated. In 80 percent of the observations, 

 oxygen saturation exceeded 50 percent. Maximum 

 values came in winter and minimums in summer. 

 Some supersaturation was detected in every 

 season. 



Concentrations of inorganic phosphate-phos- 

 phorus and total phosphorus were high in all 

 rivers; levels were exceptionally high in the Peace 

 and Alafia Rivers which flow du-ectly through an 

 area rich in phosphatic formations. Inorganic 

 phosphate-phosphorus was responsible for most 

 of the phosphorus. 



Copper values varied from 0.00 to 0.23 jugat. 

 per liter and were higher than those in the adjacent 

 Gulf of Mexico. They were markedly below the 

 lethal dose for Gymnodinium breve. 



The concentrations of iron were several times 

 higher in the rivers than in the adjacent Gulf of 

 Mexico waters. Values were highest at the fresh- 

 water stations in the Myakka and Peace Rivers 

 and lowest in the Hillsborough River. Temporal 

 changes in iron lacked a distinct trend. 



Variation in rate of flow has a marked effect 

 on the total quantities of the materials in the 

 rivers. 



The polynomial regression between total 

 phosphorus, copper, iron, chlorophyll "a" as the 

 independent variable and abundance of G. breve 

 as the dependent variable revealed no relation- 

 ships significant at the 95-percent level; iron and 

 G. breve were dbectly related at the 80-percent 

 level of significance in coastal waters off Tampa 

 Bay and Charlotte Harbor. 



FLORIDA'S WEST COAST TRIBUTARIES 



475 



