breve abundance were poured from the col- 

 lection container into 2-liter Erlenmeyer 

 flasks previously rinsed with a small 

 amount of the sample water. During periods 

 of high air temperature, wet newspaper was 

 placed around the flasks to maintain a 

 stable temperature. Pieces of polyethylene 

 sheeting were placed over the mouth and 

 secured to the flask neck with rubber bands 

 to prevent spillage or contamination. 



Water samples for salinity analysis 

 were collected at all stations in 4-ounce 

 Duraglass prescription bottles. Pyrex 

 glass bottles of 250 ml. capacity with 

 glass stoppers were used to collect water 

 samples for copper analysis. Total phos- 

 phorus, inorganic phosphate, nitrate- 

 nitrite, and carbohydrate and protein 

 equivalents were determined from quick- 

 frozen samples collected in 200 by 25 mm. 

 pyrex glass culture tubes, fitted with 

 polyethylene or polyethylene-lined screw 

 caps. All containers used for sampling 

 were chemically cleaned prior to use, and, 

 after collection of the water sample, were 

 sealed with plastic electrical tape for 

 later analysis. 



Enumeration of Gymnodinium breve 



Water samples for G. breve analysis 

 were usually examined within 24 hours after 

 collection. Most of the samples were con- 

 centrated (the organisms given time to move 

 to the surface of the sample) under 40-watt 

 fluorescent lights, but available lighting 

 was also used when a return to the field 

 station was not feasible. Satisfactory 

 concentration of the organism was obtained 

 with lights from 8 to 18 inches above the 

 flasks. The period of concentration at 

 room temperature under ordinary conditions 

 was ideally 16-18 hours, preferably over- 

 night, and a minimum concentration period 

 of 4 hours was observed. 



Examination of sample aliquots for 

 the living organisms was made with a ster- 

 eoscopic, wide-field, inclined binocular 

 microscope equipped with 18X wide-field 

 eyepieces. The technique was as follows: 



1. After the samples had been con- 

 centrated, three to six 1-milliliter ali- 

 quots were pipetted from the area just 

 beneath the surface of the water sample and 

 each deposited in the depression of a 3- 

 depression micro-plate. If no G. breve 



were found in these aliquots, the water 

 sample was assumed to contain no G. breve 

 and the count was recorded as none. 



For general detection and for count- 

 ing of both concentrated and mixed samples, 

 magnifications of 54X and 144X were used. 



2. If G. breve were observed in the 

 initial concentrated aliquots, the flask 

 was inverted several times to redistribute 

 the organisms throughout the sample of 

 water. Then, depending on the numbers of 

 G. breve per milliliter (1-99; 100-1000; or 

 more than 1000) in the examination of con- 

 centrated samples, 1.0 ml., 0.1 ml., or 

 0.01 ml. aliquot was pipetted from within 

 the flask to each of 3 depressions in a 

 micro-plate. The number of G. breve ob- 

 served per milliliter or fraction was then 

 recorded for each of the 3 aliquots. If 

 the count did not agree in 2 out of 3 of 

 these aliquots, an additional 3 aliquots 

 were drawn from the flask and examined. 

 The numbers of G. breve for all aliquots 

 were averaged and shown as the number per 

 milliliter of water. In cases where G. 

 breve were found in the surface layer of a 

 concentrated sample, usually in very low 

 numbers, but not in the mixed, or when no 

 counts were made of a mixed (or agitated) 

 sample, the symbol P is used to indicate 

 that the organism was present. 



Counts were of the living organism; 

 no preserved or fixed specimens were used. 

 The counting technique assumes that after 

 several inversions of the flask G. breve 

 are randomly distributed. Until counts of 

 such agitated samples were instigated in 

 1955, however, most of the concentrated 

 samples were merely given an adjective 

 rating for abundance. 



Chemical analyses 



Salinities were determined by the 

 Mohr-Knudsen method described by Knudsen 

 (1901), Nitrate-nitrite determinations 

 were made by the method of Zwicker and 

 Robinson (1944) as modified by Marvin 

 (1955b). 



The method of Robinson and Thompson 

 (1948) was used to determine amounts of 

 inorganic phosphates and the method of 

 Harvey (1948) was used in total phosphorus 

 determinations. Copper determinations 

 were by the analytical method of Hoste, 



