FISHERY BULLETIN: VOL. 70, NO. 4 



silicate by molybdate complexing (Mullin and 

 Riley, 1955), and phosphate by molybdate com- 

 plexing (Murphy and Riley, 1962) followed the 

 procedures outlined by Strickland and Parsons 

 (1968). Salinity was determined by conductiv- 

 ity measurements on an Autolab^ salinometer/ 



Ammonium was determined by the phenolhyo- 

 chlorite method (Solorzano, 1969) and urea by 

 the urease method (McCarthy, 1970), but both 

 of these methods deserve additional comment. 

 Solorzano states that for ammonium analyses 

 the absorbance of samples is stable for at least 

 24 hr when a 0.5% sodium nitroprusside reagent 

 is used, whereas, at higher concentrations a high 

 and unstable blank which increases with time 

 can result. Difficulties such as those expected 

 at higher sodium nitroprusside concentrations 

 were, however, occasionally experienced when 

 using the recommended concentration. Efforts 

 made to determine the source of this problem 

 were complicated by its erratic occurrence, e.g., 

 differences from day to day using the same re- 

 agents. No solution to the problem was found. 

 It is, however, of little consequence when ab- 

 sorbances are read within a short interval of 

 time, e.g., 15 min, since the optical density in- 

 creased uniformly in blanks, samples, and stan- 

 dards. When analyzing as many as 50 samples 

 at once, the reagent additions were staggered 

 in time so that the absorbance of a sample was 

 always read between 1.00 and 1.25 hr later. 



A problem was encountered with the Worth- 

 ington URC crude urease which was recommend- 

 ed for the urease method. Several batches of 

 this product were purchased over a period of 

 a year, and following the outlined purification 

 procedure, the preparations were almost identi- 

 cal in activity and blank. Shortly after the 

 urease-urea method was published (McCarthy, 

 1970), two batches of the same product were 

 received which yielded both less activity and 

 higher blanks than the preceding preparations. 

 Owing to these difficulties, Sigma Type III urease 

 was used subsequently. The resultant blank is 

 higher (0.030-0.050 OD units/10-cm cell) than 



previously described and approximately 3 mg 

 rather than 0.5 mg of the enzyme preparation 

 must be added to each sample. A 0.1-0.2 ml 

 Biopipette'^ was found to be accurate and rapid 

 enough to permit direct addition of the concen- 

 trated enzyme preparation without dilution. In 

 addition, it was found that by centrifuging the 

 final urease preparation approximately 1,000 X 

 g for 20 min and discarding the pellet, turbidity 

 of the preparation could be reduced. For labor- 

 atory analyses of urea McCarthy (1970) recom- 

 mended the use of aluminum foil coverings for 

 the reaction vessels, but it should be noted that 

 if the sample contacts the foil an erroneously 

 high ammonium value will result. If foil cov- 

 erings are used, care should be taken when mov- 

 ing the flasks, and they should be discarded after 

 the addition of the oxidizing solution. After the 

 addition of the last reagent, color will not result 

 from ammonium added to the sample whereas 

 contact with the aluminum foil will still inter- 

 fere with the results. 



The nutrient data were shown to depart from 

 normality in distribution by plotting them on 

 normal probability paper, so parametric statis- 

 tics could not be applied. The Mann-Whitney 

 U test (Tate and Clelland, 1957) was used for 

 comparing medians from different stations or 

 different periods. This test compares the central 

 tendency of two distributions and does not as- 

 sume similarity in variance. The Tukey-Siegel 

 modification of the Mann-Whitney U test was 

 used to compare variability between sets of data, 

 the median regression procedure (Tate and Clel- 

 land, 1957) was used to calculate the regression 

 lines, and the tern coefficient test (Kendall, 1955) 

 was used to determine correlations. The a for 

 significant differences was taken at the 0.05 level 

 except when there was multiple testing, in which 

 cases 0.05/w was used where n — the number 

 of times a test was run with interrelated data. 



The data for dissolved oxygen, chlorophyll a, 

 phaeophytin a, silicate, phosphate, phytoplank- 

 ton species and numbers, temperature, and sa- 

 linity will be reported elsewhere (Kamykowski) .' 



^ Reference to trade names in the publication does not 

 imply endorsement of commercial products by the Na- 

 tional Marine Fisheries Service. 



* Kamykowski, D. Some physical and chemical as- 

 pects of the phytoplankton ecology off La JoUa, Cali- 

 fornia. Manuscript in preparation. 



1264 



