COMPARISON OF TWO METHODS OF N-ETHYLCARBAZOLE 

 CARBOHYDRATE ANALYSIS » 



By Kenneth T. Marvin and Raphael R. Proctor Jr., (Chemists) 

 Bureau of Commercial Fisheries Biological Laboratory, Galveston, Tex. 



Two currently used N-ethylcarbazole (NEC) 

 methods for determining the amount of dissolved 

 substances in sea water which respond to the NEC 

 reagent and which will be referred to here as 

 "carbohydrates," are the original one described 

 by Erdman and Little 2 and the modification de- 

 scribed by Zein-Eldin and May (1958). It has 

 been assumed by many investigators that those 

 methods give comparable results when used to es- 

 timate the amount of carbohydrates in water. 

 Recently, however, data obtained through both 

 methods at the Bureau of Commercial Fisheries 

 Biological Laboratory in Galveston, Tex., sug- 

 gested that their corresponding estimates differ 

 significantly. This difference prompted us to de- 

 termine experimentally if the two methods yield 

 the same results when used to analyze the carbo- 

 hydrate content of aliquots from the same sample. 



The original method described by Erdman and 

 Little consists of adding the NEC reagent ( 1 g./l. 

 of 36 N H 2 S0 4 ) to the sample in one step, warming 

 the mixture in a water bath, and then comparing 

 color density with that of similarly treated arabi- 

 nose (D-(-) -Arabinose) standards. This proce- 

 dure will be called the "one-step" method. The 

 method described by Zein-Eldin and May is the 

 same as the above except that the NEC reagent is 

 added in two steps. We will therefore refer to this 

 procedure as the "two-step" method. Results of 

 both procedures are based on the density of color 



Note. — Approved for publication May 1. 1965. 



1 Contribution No. 195, Bureau of Commercial Fisheries Bio- 

 logical Laboratory, Galveston, Tex. 



2 Erdman, J. G. and A. B. Little. 1950. Analysis of marine 

 coastal and estuarine waters, 50 pp. Multiple fellowship of 

 Gulf Research and Development Co., Mellon Institute, Pitts- 

 burgh, Pa. (Unpublished.) 



development when NEC reagent reacts with car- 

 bohydrates. Units are expressed in mg. equival- 

 ents of arabinose. 



We compared the two methods by analyzing 

 carbohydrate determinations obtained routinely 

 from many samples with each procedure. The 

 one-step method yielded significantly higher 

 results (table 1). 



Table 1. — Concentration of carbohydrates in samples of 

 water collected from Galveston Island's East Lagoon during 

 the period February-August, 1963 



[Units are mg./l. of arabinose equivalents] 



Date, 1963 



Feb. 16. 

 Feb. 19.. 

 Mar. 1... 

 Mar. 12.. 

 Mar. 15.. 

 Mar. 19.. 

 Mar. 22.. 

 Mar. 26.. 

 Mar. 29.. 

 Apr. 2... 

 Apr. 9... 

 Apr. 23.. 

 May 8... 

 May 10.. 

 May 14. . 

 May 17.. 

 May 21.. 

 May 24.. 

 May 28.. 

 May 31.. 

 June 4.. 

 June 7. . 

 June 11. 



Method 



One- 

 step 



Mg./l. 

 4.15 

 4.00 

 3.80 

 4.55 

 4.35 

 3.65 

 3.50 

 3.20 

 2.55 

 3.11 

 5.14 

 3.17 

 4.06 

 3.43 

 3.91 

 3.94 

 3.28 

 2.31 

 3.03 

 3.60 

 4.13 

 3.33 

 4.58 



Two- 

 step 



Mg./l. 

 1.06 



.61 

 1.98 

 1.71 

 2.61 



.78 

 1.15 



.85 



.73 

 1.86 

 3.01 

 2.33 

 1.98 

 2.15 

 2.36 

 2.02 

 2.13 

 2 24 

 L71 

 2.25 

 2.70 

 1.99 

 2.63 



Date, 

 1963 



6-14 



6-18 



6-21 



6-25 



6-28 



7-2 



7-5 



7-9 



7-12 



7-16 



7-19 



7-23 



7-26 



7-30 



8-2 



8-7 



8-9 



8-13 



8-16 



8-20 



8-21 



8-22 



Method 



One- 

 step 



Mg./l. 

 3.80 

 4.71 

 3.92 

 3.66 

 4.64 

 5.69 

 5.85 

 7.21 

 5.07 

 4.28 

 5.36 

 4.28 

 6.50 

 4.71 

 5.07 

 5,07 

 7.85 

 4.57 

 5.00 

 4.07 

 2.55 

 3.60 



Two- 

 step 



Mg./l. 

 2.27 

 3.76 

 2.32 

 1.78 

 3.25 

 3.90 

 3.70 

 3.18 

 3.24 

 3.03 

 4.45 

 2.58 

 3.48 

 3.15 

 2.94 

 2.18 

 3.03 

 2.67 

 3.51 

 2.70 

 2.47 

 2.36 



Upon investigating the reason for the differences 

 noted between estimates secured by the two meth- 

 ods, we found that the color density resulting from 

 interacting NEC and carbohydrates in natural 

 waters varied only slightly with the method used. 

 However, when the NEC reagent was added to 

 standard or calibration samples that consisted of 



FISHERY BULLETIN: VOLUME 65, NO. 3, 



683 



