coefficients of 0.93 and 0.89 were determined between I P and TSP, and IP and SP measurements, respectively A median value for both the IP/TSP and IP/SP ratios of 0.65 was computed, based on combined data from the five sites. (3) A very good linear relationship with a correlation coefficient of 0.93 was found to exist between hi-vols using glass-fiber filters and hi-vols using Whatman 41 cellulose filters. No site dependency was found to exist, but a TSP loading dependency was evident in the relationship. This dependency was probably due to the water sorbtion problem associated with cellulose filters but was small enough not to affect the quality of the relationship between the two measures. The results of this study indicate that the two principal disadvantages in the standard method for measuring TSP can be adequately compensated for by the use of dichotomous samplers and hi-vols using cellulose filter media. Acknowledgment
The authors thank Mrs. Catherine Cassidy for typing the manuscript, and Mr. Gary Lanphear and Mrs. Carol Clas for drafting the figures. We also express our thanks to Mr. James Hyde and M,r. Robert Forrester for their assistance during earlier phases of this study.
Literature Cited (1) “The Clean Air Act as Amended August 1977”;U.S. Government Printing Office, Nov 1977.
(2) Miller, F.J.; Gardner, D. E.; Graham, J. A,; Lee, R. E.; Wilson, W. E.: Bachmann, J. D. J . Air Pollut. Control Assoc. 1979,29,6. (3) Pate, J . B.; Taber, E. C. Am. Ind. Hyg. Assoc. J . 1962,145. (4) Levaggi, D. A,; Sandberg, J. S.; Feldstein, M.; Twiss, S. J . Air Pollut. Control Assoc. 1976,26,6. (5) Pace, T. G.; Meyer, E. L., Jr. Proc., Annu. Meet.-Air Pollut. Control Assoc. 1979,72,79-47.2. (6) Spengler, J. D.; Turner, W. A.; Dockery, D. W. Proc. Annu. Meet.-Air Pollut. Control Assoc. 1980,73,80-43.4. (7) Record, F. A.; Wiltsee, K. W., Jr.; Bradway, R. M. “Final Report on Program to Measure, Analyze and Evaluate TSP in Massachusetts”; GCA Corp., GCA/Technical Division: Bedford, MA, Sept 1979; Contract 5-538-01. ( 8 ) Neustader, H. E.; Sidik, S. M.; King, R. B.; Fordyce, J. S.;Burr, J. C. Atmos. Environ. 1975,9,1. (9) U S . Environmental Protection Agency “Quality Assurance Handbook for Air Pollution Measurement Systems”; EPA-BOO/ 4-77-027a; May 1977; Vol. 11. (10) Kolak, N. P.; Hyde, J.; Forrester, R. “Particulate Source Contributions in the Niagara Frontier”; EPA 902/4-79-006; Dec 1979. (11) Skogerboe, R. K.; Dick, D. L.; Lamothe, P. J. Atmos. Enuiron. 1977,11,3. (12) Harker, A. B.; Richards, L. W.; Clark, W. E. Atmos. Enuiron. 1977 11, 1. (13) Appel, B. It.;Wall. S. M.; lokiwa, Y.; Haik, M. Atmos. Enuiron. 1979,13,2. (14) Stevens, R. K.; Dzubay, T. G.; Russwurm, G.; Rickel, D. Atmos. Enuiron. 1978,12,1. (15) John, W.; Reischl, G. Atmos. Environ. 1978,12,10.
Received for review January 11,1980.Accepted November 10,1980. This work was supported in part by the U.S. Environmental Protection Agency under Contract No. 68-02-2880.
NOTES
Free, Proteinaceous, and Humic-Bound Amino Acids in River Water Containing High Concentrations of Aquatic Humus+ Charles R. Lytle and Edward M. Perdue* Department of Chemistry, Portland State University, P.O. Box 751, Portland, Oregon 97207 -
______
Amino acids were determined by CLC analysis at 12 sampling sites along the Williamson River, Oregon, and its principal tributaries at monthly intervals for a period of 2 yr. The five most abundant amino acids maintained the relative order Gly > Asp > Ala > Ser N Glu for the majority of sampling sites and showed no seasonal change. A separation technique using XAD-7 macroreticular resin is described for fractionating total amino acids. At the sampling sites analyzed, 196%of the dissolved amino acids were associated with dissolved humic material. Correlation of discharge with total amino acid concentration suggests surface runoff as the primary source of amino acids in the river. --Introduction
Dissolved organic matter (DOM) in rive1 water is piimarily allochthonous in nature and is composed chiefly of humic substances, carbohydrates, and proteinaceous matter (1). Humic substances are stable polymers believed to be formed as byproducts of microbial degiadation of plants and can OCCUI in natural waters through leaching horn soils ( 2 ) In recent t A preliminary report of this work was piesented at the 33rd Northwest Regional Meeting of the American Chemical Society Seattle, WA, June 14,1978,
224
Environmental Science & Technology
years, studies have shown that dissolved organic matter can play an important role in the ability of natural waters to complex metals, thus affecting trace-metal transport and toxicity (3-7). Highly colored streams draining marshes or swamps contain unusually high concentrations of DOM and are thus often the most advantageous natural systems for studying the role of DOM in the transport of trace metals and other pollutants. The Williamson River in Klamath County, Oregon, not only possesses high concentrations of DOM but also provides a unique, “before-and-after” situation. The river begins as a clear spring and flows -25 mi through basaltic terrain and then drains into Klamath Marsh. After passing through the marsh, the river is dark brown in color and contains high concentrations of DOM. After joining Spring Creek and the Sprague River, the Williamson drains into Upper Klamath Lake, -35 mi from the marsh (8, 9). Since the Williamson provides -46% of the water and nutrients flowing into Upper Klamath Lake, the river is also a logical focus for one of the causes of Klamath Lake’s intense, seasonal bloom of the cyanobacterium, Aphatiizomenon flus-aquae (10, 11). This paper focuses uti the proteinaceous portion of DOM in the Williamson River. The abundance and the fractionation of the carbohydrate portion have been reported (12). The first part of this study presents a 2-yr, monthly survey of total 0013-936X/81/0915-0224$01.00/0 @ 1981 American
Chemical Society
amino acids in the river and its principal tributaries. The second part describes a fractionation study of the amino acids utilizing a macroreticular resin. Macroreticular resins have been used to isolate humic substances from seawater (13, 14) and freshwaters ( 1 5 ) , and their relative performances have been recently compared ( 1 6 ) .This study presents a novel use of the XAD-7 resin to achieve a separation of dissolved humic substances and dissolved proteinaceous material. This separation allowed for the identification and quantification of four distinct fractions of amino acids found in the river system.
Materials Reagents. Methanol and isopentyl alcohol (Baker reagent grade) were refluxed over magnesium turnings 1 h and then distilled, and the 65.0 and 127.0 "C fractions, respectively, were taken. Both alcohols were stored over activated molecular sieve pellets, type 13X (Matheson, Coleman, Bell), and the bottles stored in a desiccator. Acetyl chloride (Baker Instra-analyzed) was refluxed 1h with dimethylaniline, which had been previously purified by being passed through two alumina columns (M. Woelm, ICN Pharmaceuticals), and then distilled, and the 51.0 "C fraction was taken. Acetonitrile (Matheson, Coleman, Bell Spectro-quality), ethyl acetate (Baker Instra-analyzed), and heptafluorobutyric anhydride (Pierce, 1-mL ampules) were used without further purification. Amino acids were obtained from Sigma Chemical Co. Water was 18.5 MQ cm from a Barnstead Nanopure cartridge system with reverse-osmosis pretreatment. Glassware. All glassware was washed in detergent and water, rinsed thoroughly, soaked overnight in hot, alcoholic KOH, and then exhaustively rinsed with tap and then Barnstead water. Linear polyethylene bottles (Nalgene) were washed as above, soaked in 1 M HNO3 overnight, and then rinsed as above. Chromatography.Dowex 50W-X4,200/400 mesh (Biorad) was prepared by the method of Kaiser et al. (17). Elution regimes for the columns used in desalting hydrolysates were determined by the liquid-scintillation monitoring of [3H]Gly and [3H]Glu (New England Nuclear) test elutions. Column recoveries were determined by GLC analysis of known amino acid standard mixtures before and after desalting. Macroreticular resin XAD-7 (Rohm and Haas) was prepared as recommended by the manufacturer (18).Separation efficiency was determined by the liquid-scintillation monitoring of [14C]algal protein and [14C]algal protein hydrolysate (ICN Pharmaceuticals) test elutions. At pH 2, 83% of the algal protein and 88%of the algal protein hydrolysate were nonretained by the resin. The efficiency of removal of humic substances was determined to be -99% by monitoring absorbance at 420 nm, pH 10. Gas-liquid chromatography columns were 12-ft, 2-mm i.d. glass (Supelco) packed with 3% SE-30,100/120 Gaschrom-Q (Applied Science Laboratories) using the method of Leibrand and Dunham (19). Analyses were performed on a Hewlett-Packard gas chromatograph, Model 5750, equipped with a flame ionization detector. The operating parameters were the following. Column temperatures: 90 "C (initial) and 250 "C (final). Programming: 5-min initial isothermal hold, 2 "Clmin to 140 "C, 4 "C/min to 250 "C. Injector and detector temperatures: 275 "C. Carrier (Nz) flow: 30 mL/min. Sample volume: 1pL. Peak areas were automatically calculated by a Hewlett-Packard reporting integrator, Model 3380A. Methods Sampling. From September, 1977, through September, 1979, samples were taken monthly from 1 2 sampling sites along the Williamson River and its tributaries. The approximate location of these sites is shown in Figure 1.Abnormally low snow pack in the Cascade Mountain Range prior to and during this study lowered discharge into Klamath Marsh to
i WR
lob
SPRING CREEK\
SPRAGUE
u
RIVER
UPPER
10 krn
Figure 1. Location of sample sites.
the extent that no flow occurred out of the marsh through sampling site WR-50 during the three fall seasons observed. Flow did occur at site WR-56 and throughout the rest of the river because of numerous springs along the river between WR-50 and WR-56. No samples were taken a t Big Springs (BS-10) during the winter months because of inaccessability due to snow. Stream samples for the monthly survey were collected in 65-mL linear polyethylene (LPE) bottles and preserved in 0.02% sodium azide and stored a t 5 "C until derivitization. Samples for the fractionation study were taken in January, 1980, at WR-21, WR-32, and WR-50 in 4-L LPE bottles and preserved in 0.02% sodium azide. Before storage a t 5 "C, the samples were deaerated for 15 min with prepurified Nz. Assay for Humic Carbon. A sample of humic substances was isolated from the Williamson River by using the method of Perdue (ZO),and its carbon content analyzed. Various known weights were dissolved in distilled water, buffered to pH 10, and read at 420 nm. A linear-regression analysis gave humic carbon as a function of absorbance at 420 nm (21). Derivitization for GLC Analysis. After the volumes were carefully measured, the water samples for the routine monthly survey were acidified to pH -2 with 6 M HC1 and evaporated under a stream of dry, prepurified Nz at 50 "C to a final volume of -2 mL. The samples were then quantitatively transferred to 3-mL Reacti-vials (Pierce Chemical Co.), fitted with Teflon-lined screw caps, and evaporated to -1.0 mL. An equal volume of 12 M HC1 was added, and samples were hydrolyzed, under Nz, for 22 h at 110 "C. The hydrolysates were evaporated under a stream of dry, prepurified N2 at 50 "C and taken up in 1 mL of 0.1 M HCI. They were then desalted on 0.5-mL columns of Dowex 50W-X4, 200/400 mesh. Five thousand nanograms (50 yL of a 100 ng/yL stock solution) of y-amiVolume 15, Number 2, February 1981 225
!-
BULK SAMPLE
1200mL
0.45 micron
IOOOmL
HYDROLYZE
I
DESALT DERlVlTlZE (DISSOLVED)
XAD-7 RESIN
1 MARSH 1
-
POST-MARSH SITES
+
/LAKE
DESALT DERlVlTlZE (TOTAL)
100 mL
I
PRE-MARSH SITES 4
Flgure 4. Two-year averages for total amino acids and humic carbon. HYDROLYZE DESALT DERlVlTlZE (HUMIC)
Table 1. Weighted Average Concentrations of Amino Acids in the Williamson River System over 2 yr amlno acid
112 SAMPLE
112 SAMPLE
HYDROLYZE DESALT DERlVlTlZE (FREE+ PROTEIN)
DERlVlTlZE (FREE)
Figure 2. Fractionation scheme, 0
m
m
N
TIME (rnln) w
D
s
S
B
DFigure 3. Typical chromatogram: total amino acids, KL-10, Sept 1978.
nobutanoic acid was added as an internal standard to the 2 M NH3 eluants. The NH3 fractions were evaporated under a stream of dry, prepurified Nz at 50 "C. Derivitization to the isoamyl heptafluorobutyrates was accomplished by the method of Zanetta and Vincendon (22) with the modification of using acetyl chloride instead of gaseous HC1 to prepare the acid alcohols (23). Fractionation Study. The steps in the fractionation scheme are outlined in Figure 2. Each of the three river samples was analyzed in triplicate; 100-mL portions of each sample were analyzed directly for total amino acids; 1100-mL portions were filtered through prewashed, 0.45-pm filters (Millipore), and 100 mL of this filtrate was analyzed for dissolved amino acids. Because of irreproducibility of results on samples eluted off the filters (presumably due to microbial growth on the filters and/or inability to quantitatively wash amino acids off the filters), particulate amino acids were calculated as the difference between total and dissolved amino acids. The remaining 1000 mL of filtered sample was acidified to pH 1.9 with 12 M HCl and passed through a 15-mL column Environmental Science & Technology
mol
YO
11.8 18.7 4.7
6.1 9.5
4.7 2.3 3.5
amino acid
mol %
methionine phenylalanine aspartic acid lysine tyrosine glutamic acid arginine histidine
3.6
5.2 13.6
3.0 0.3 9.6 1.2
2.2
LI
W
226
alanine glycine valine threonine serine leucine isoleucine proline
of XAD-7. The humic substances were eluted with 40-50 mL of 0.1 M NaOH. A portion of the 0.1 M NaOH eluant was analyzed for humic amino acids. The 1000 mL of filtered sample that passed through the XAD-7 column was rotary evaporated to a final volume of -40 mL, desalted on a 1.5-mL Dowex 50W-X4,200/400 column, and split into two equal portions. One portion was analyzed as previously described for free plus protein amino acids. The second portion was derivitized without a preceding hydrolysis step to give free amino acids only. Protein amino acids were calculated to be the difference between free-plus-protein and free amino acids. Standards. Standard mixtures of 17 amino acids were prepared, spiked with y-aminobutanoic acid (GABA) as an internal standard, and run in the range of 4-80 ng of each amino acid. For plots of nanograms of amino acid vs. area of amino acidlarea of 50 ng of GABA, correlation coefficients ranged from 0.977 to 0.999, with an average of 0.992. The hydrolysis, desalting, and derivitization procedure was checked by analyzing a known protein, raccoon a-hemoglobin. A plot of found residues vs. expected residues yielded a slope = 0.99, intercept = 0.1, and r = 0.970. Analytical blanks gave no detectable amino acids.
Results Monthly Survey. As shown in Figure 3, 15 major amino acids were separated. Asparagine and aspartic acid yield a single peak, as do glutamine and glutamic acid. On a few occasions when total amino acids were above 10 pM, other amino acids gave separate but very small peaks. These included 0-Ala, Hyp, Orn, and Tyr. Trp is lost in acid hydrolysis. The results for each sampling site, averaged over the entire 2-yr study, are summarized in Figure 4. The input from the marsh is clearly seen, as are the diluting effects of the many riverbed springs between WR-50 and WR-56 and of Spring Creek. At WR-50, the value shown was averaged only over those months that water was flowing out of the marsh. The range was 1.5 to 15.9 pM. The 2-yr averages for humic carbon are also shown in Figure 4 and are seen to closely follow the amino acid averages. KL-10
.
9 0 1 . .
. .
,
.
,
.
,
. .
,
.
.
,
,
SbMPLlNG MONTH l 9 / 7 7 - 9 / 7 9 1
Figure 5. Discharge, humic carbon, and total amino acids at Sprague
River over 2 yr. FRACTION UNFILTERED
WR21 0 7 8 L O 03
PARTICULATE” 19- of u n f i l t ~ r i d l
FILTERED
SAMPLING SITES WR32
120A0.14
141V.l
116%
0.46*0.05
1.00*0.08
WR50 2.38kOll
1
2.234007
FREE
“found b y d l f f a r m m
Figure 6. Results of
fractionation study (concentrationsin pmol/L.)
shows anomalous behavior that is most probably due to the high concentration of algae present in Klamath Lake during the summer, fall, and early winter months (>30 000 cells/mL of Aphanizomenon flos-aquae). When total amino acids is plotted against humic carbon, significant r 2 correlations are found. As an example, for the December 1978 field trip, the least-squares line gave an r 2 of 0.976, disregarding the anomalous KL-10 point. This evidence suggested that most of the amino acids were associated with humic substances. The mole percentages of individual amino acids, averaged over all of the sampling sites for 2 yr, are given in Table I. In general, the five most abundant amino acids were Gly > Asp > Ala > Ser E Glu. This order showed no significant variation through the river system throughout the 2 yr. This relative order has been found in other river systems by other workers (24,25). Generally, total amino acid concentrations were several times higher in the winter and spring months than in the summer and fall months. This seasonal pattern was also observed for humic carbon and discharge. The positive correlation among these parameters is exemplified in Figure 5 , which shows the data from the Sprague River sampling site (SR-65). On the basis of these results, it appears likely that the principal source of humic carbon and amino acids is surface runoff, which flushes these components from water-saturated soils during periods of high discharge. Similar variations of total organic carbon with discharge have been previously noted (26,27). In view of the biological lability of free amino acids, proteins, etc., it seems likely that those amino acids which are mobilized from soils are already associated with humic substances. Thus the overall seasonal variability of amino acid concentrations in this river system is probably best explained
by a predominant discharge-related pattern on which relatively minor biological perturbations are superimposed. Fractionation Study. To test the hypothesis that the amino acids were associated with humic substances, we chose three sampling sites: WR-21, well before the marsh where the Williamson is a clear, dilute stream; WR-32, directly in the marsh; and WR-50, immediately after the marsh. The results of the fractionations are shown in Figure 6. In all three cases, the humic-associated amino acids were 196% of the total dissolved amino acids. The amounts of free amino acids were below detection limit, implying that the remaining dissolved amino acids were either proteinaceous or possibly humicassociated amino acids that had bled through the XAD-7 column. The high percentage of particulate amino acids at WR-21 can be accounted for by the comparatively high concentration of diatoms at this site. Cell counts average -8000 cells/mL (mostly Navicula ssp.), while WR-32 and WR-50 average -1700 cells/mL (mostly Fragilaria ssp.). At all three sampling sites, humic-associated amino acid carbon accounted for -1% humic carbon. This is in agreement with other reports (25).
Conclusion This study has shown that Klamath Marsh has a profound influence on the amount of amino acids and humic carbon in the Williamson River, Oregon. The 2-yr averages for total amino acids were -1 pM in the river before the marsh, -5 p M in the marsh itself, and -8 pM in the river directly after the marsh. The relative amounts of the most abundant amino acids remained constant throughout the river system in the order Gly > Asp > Ala > Ser N Glu and showed no significant seasonal changes. The concentrations of total amino acids and humic carbon showed significant correlations with each other and with discharge. A novel use of XAD-7 macroreticular resin made possible the separation of dissolved amino acids into humicassociated and non-humic-associated fractions. This fractionation study showed that 296% of the dissolved amino acids were associated with aquatic humus and that amino acid carbon comprised -1% of the humic carbon. It is proposed that the amino acids are associated with humic substances in soil and are carried into the stream by surface runoff. Acknowledgment We thank Dr. Bernice Brimhall for the sample of raccoon &-hemoglobin and James Sweet for the preparation and counting of algal samples. The assistance of Stanley Kunsman, United States Forest Service District Ranger, Chiloquin Ranger District, is gratefully acknowledged. Literature Cited (1) Reuter, J. H.; Perdue, E. M. Geochim Cosrnochirn. Acta 1977, 41, 325. (2) Stevenson, F. J.; Butler, J. H. A. in “Organic Geochemistry”;
Eglinton, G., Murphy, M. T. J., Eds.; Springer-Verlag: New York, 1969; Chapter 22. (3) Gachter, R.; Lum-She-Chan, K.; Chau, Y. K. Schweitz. Z. Hydrol. 1974,35, 252. (4) Andrew, R. W., Hodson, P. U., Konasewich, P. E., Eds. “Toxicity T o Biota of Metal Forms in Natural Water”; United States Environmental Protection Agency, Environmental Research Laboratory: Duluth, MN, 1976. (5) Sunda, W. G.; Lewis, J. M. Limnol. Oceanogr. 1978,23, 870. (6) Giesy, J. P.; Briese, L. A.; Leversee, G. J. Enuiron. Geol. (N.Y.) 1978,2, 257. ( 7 ) Saar, R. A.; Weber, J. H. Can. J . Chem. 1979,57, 1263. ( 8 ) Peterson, N. V.; McIntyre, J. R. “The Reconnaissance Geology and Mineral Resources of Eastern Klamath County and Western Lake County, Oregon”: State of Oregon Department of Geology and Mineral Industries: Portland, OR, 1970. (9) Leonard, A. R.; Harris. A. B. 1974, “Ground Water in Selected Areas in the Klamath Basin, Oregon”; United States Geological Survey Ground Water Report No. 21.
Volume 15, Number 2, February 1981 227
(10) Miller, W. E.; Tash, J. G. “Interim Report: Upper Klamath Lake Studies, Oregon”, Pub. WP-20-8; Pacific Northwest Water Laboratory: Corvallis, OR, 1967. (11) Gahler, A. R. “Field Studies on Sediment-WaterAlgal Nutrient Interchange Processes and Water Quality of Upper Klamath and Agency Lakes”, Working Paper No. 66. Pacific Northwest Water Laboratory: Corvallis, OR, 1969. (12) Sweet, M. S. Master’s Thesis, Portland State University, Portland, OR, 1979. (13) Mantoura, R. F. C.; Riley, J. P. Anal. Chim. Acta. 1975, 76, 97. (14) Stuermer, D. H.; Harvey, G. R. Mar. Chem. 1978,6, 55. (15) Weber, J. H.; Wilson, S. A. Water Res. 1975,9, 1079. (16) Aiken, G. R.; Thurman, E. M.; Malcolm, R. L.; Walton, H. L. Anal. Chem. 1979,51, 1977. (17) Kaiser, F. E.; Gehrke, C. W.; Zumwalt, R. W.; Kuo, K. C. J . Chromatogr. 1974,94, 113. (18) Rohm and Haas Co. “Summary Bulletin. Amberlite Polymeric Absorbants”, 1971. (19) Leibrand, R. J.; Dunham, C. L. Res./Dev. 1973,24, 32.
228
Environmental Science & Technology
(20) Perdue, E. M. ACS Symp. Ser. 1979, No. 93, Chapter 5. (21) Blunk, D., Portland State University, personal communication, 1979. (22) Zanetta, J. P.; Vincendon, G. J . Chromatogr. 1973,76, 91. (23) Felker. P.: Bandurski. R. S. Anal. Biochem. 1975.67. 245. (24) Peake,’E.;Baker, B. L.;Hodgson, G. W. Geochim.’Cosmochim. Acta. 1972,36, 867. 125) Beck. K. G.: Reuter. J. H.: Perdue. E. M. Geochim. Cosmochim. ‘ Acta. 1974,38, 341. ’ (26) Reuter. J. H.: Perdue. E. M. Geochim. Cosmochim. Acta. 1974. 38, 341. (27) Dahm, C. N.; Oregon State University, personnel communication, 1980.
Received for review June 16,1980. Accepted October 20,1980. This work was supported by a grant from the Office of Water Research and Technology, Washington, D.C., through the Water Resources Research Institute of Oregon State University.
