168 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



Realizing that the pumping rate is a resultant of 

 all the factors that influence the oyster, we accept 

 it as the hest available measure of physiological 

 activity. The method used in measuring the 

 pumping rate is described in the Appendix, 

 page 183. 



COLLECTING WATER SAMPLES 



Sea water was pumped from the bay and sup- 

 plied to the laboratory at Pensacola, Fla., as a 

 continuous flow without storage, because it was 

 desired to follow the changes in the bay water 

 as they occurred. 



During the course of the work, it became 

 apparent that water samples were required at 

 frequent intervals, owing to the wide and frequent 

 variations in salinity and carbohydrate concentra- 

 tion noted in our preliminary tests. In September 

 1948 we began taking samples every 2 hours 

 throughout the day and night, and in November 

 1949, the interval was shortened to 1 hour. The 

 samples were collected by means of a special 

 device described in the Appendix (fig. A-2). 



CHARACTERISTICS OF 

 CARBOHYDRATES 



STABILITY 



Since the effects of storage on these organic 

 compounds, which we have defined as carbohy- 

 drates, were not known, it was necessary to 

 determine then - stability while in storage so as to 

 establish the validity of values from samples 

 which had been collected in the automatic sampler. 

 We assumed that no significant change would 

 occur during the intervals between analyses, and 

 tested this assumption by collecting a series of 

 triplicate samples from the laboratory sea-water 

 supply (table 1). In each series, one was analyzed 

 immediately, one was held overnight at room 

 temperature, and one was held overnight in the 

 refrigerator. The mean difference between the 

 samples analyzed immediately and those held in 

 the refrigerator was plus 0.007 milligram/liter, 

 and for those held at room temperature was plus 

 0.094 milligram/liter. Since even the larger of 

 these differences was within the range of error of 

 our readings, we accepted this method of sampling 

 as adequate, despite the delays inherent in making 

 the determinations. 



Table 1. — Effect of standing time on carbohydrate con- 

 centration (mg./l.) of sea water at Pensacola, Fla., 1949 



EFFECTS OF FILTERING AND CENTRIFUGING 



Since the carbohydrate content of diatoms or 

 other organisms present in varying numbers as 

 particulate matter might cause erratic results in 

 the carbohydrate analysis, we tested the magni- 

 tude of this factor, by making a series of tests on 

 filtered and on centrifuged sea water, both 

 separately and in combination. The results of 

 these tests are shown below. 



Carbohydrate concentrations of three samples 

 before and after certrifuging ' showed the following 

 changes : 



Before After Loss 



Sample 1 10.7 10.5 -0.2 



Sample 2 13.9 13.7 -0.2 



Sample 3 13.9 13.9 



Carbohydrate concentration of four samples 

 before and after filtering 2 showed the following 

 changes : 



Before After Loss 



Sample 1 11.5 11.3 -0.2 



Sample 2 11.2 11.2 0.0 



Sample 3 13.9 13.6 -0.3 



Sample 4 14.1 13.9 -0.2 



From these results it became apparent that 

 each method was effective in removing particulate 

 matter. A sample containing 10.8 mg. of carbo- 

 hydrates to the liter was then centrifuged and 

 found to contain 10.5 mg. to the liter. An aliquot 

 of that same sample, when filtered, contained an 



i Centrifuged for 10 minutes at force 336. See Appendix, p. 182. 

 » Zeitz filter, porosity permitting 1.5 liters per hour at 40 millimeters 

 vacuum. 



