WATER POLLUTION CONTROL RESEARCH SERIES 16010 EQA 10/71
Dissolved and Particulate
Organic Carbon
in Some Colorado Waters
^
i ,
,!ilT,tt ĞM'
H\Vl\ll 1
,ffiiMi';nMfnaa.
ğ':.?/ğĞğ/
\
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Series describes the
results and progress in the control and abatement of pollution
in our Nation's waters. They provide a central source of
information on the research, development and demonstration
activities in the Environmental Protection Agency, through
inhouse research and grants and contracts with Federal, State,
and local agencies, research institutions, and industrial
organizations.
Inquiries pertaining to Water Pollution Control Research
Reports should be directed to the Chief, Publications Branch
(Water), Research Information Division, R&M, Environmental
Protection Agency, Washington, D.C. 20^-60.
-------
DISSOLVED AND PARTICULATE ORGANIC CARBON
IN SOME COLORADO WATERS
by
Edward B. Reed
Colorado State University
Department of Zoology
Fort Collins, Colorado 80521
for the
Office of Research and. Monitoring
ENVIRONMENTAL PROTECTION AGENCY
Project #16010 EQA
October 1971
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20102 - Price $1.00
-------
EPA Review Notice
This report has been reviewed by the Environmental Protection
Agency and approved for publication. Approval does not
signify that the contents necessarily reflect the views and
policies of the Environmental Protection Agency nor does
mention of trade names or commercial products constitute
endorsement or recommendation for use.
ii
-------
ABSTRACT
Organic carbon in a variety of lakes, reservoirs and streams was measured
by acid-persulfate digestion and infrared absorption.
Organic carbon was designated as net seston carbon, filter seston carbon
or dissolved based on filtering techniques.
Concentrations of dissolved-carbon ranged over an order of magnitude;
i.e. from about 1 to 11 g/m ; filter seston amounts varied from about
0.2 to 1 g/m ; the net fraction of organic carbon ranged from less than
0.1 to over 0.2 g/m . In general bodies of water judged to be meso- to
eutrophic contained more organic carbon than those waters tending to-
ward oligotrophy.
Repeated sampling of selected bodies of water revealed that amounts of
organic carbon, either dissolved or particulate, fluctuated considerably
over short periods of time; however the dissolved fraction probably
varied less than particulate forms of carbon.
Samples of all carbon fractions were incubated at room temperature in
darkness for periods ranging up to 90 days. Concentrations of organic
carbon varied erratically and unpredictably during incubation.
Almost daily measurements of carbon in samples incubated over three week
periods also revealed erratic changes, with no clear reduction in total
organic carbon. Metabolic activities of heterotrophic bacteria (and
algae?) probably complicated interpretation of results.
This report was submitted in fulfillment of Project #16010 EQA under the
sponsorship of the Water Quality Office, Environmental Protection Agency.
iii
-------
CONTENTS
Section Page
I CONCLUSIONS 1
II RECOMMENDATIONS 3
III INTRODUCTION 5
IV METHODOLOGY AND PROCEDURE 7
Field Sampling
Laboratory Procedures
Methodological Modifications
Sensitivity of the Method
Accuracy of Method
Filter Seston Carbon
Incubation
Variability Amount Replicate Samples
V FIELD AND LABORATORY RESULTS, PRIOR TO 1 JANUARY 1971 25
Dissolved Organic Carbon (DOC)
Net Seston Carbon (NSC)
Relationship of NSC to Dry Weight of NS
VI FIELD AND LABORATORY RESULTS, AFTER 1 JANUARY 1971 55
Introduction
Bear Lake
Dixon Reservoir
Spring Creek
Glacier Creek
Incubation Studies
VII DISCUSSION 71
VIII ACKNOWLEDGMENTS 77
IX REFERENCES 79
-------
FIGURES
1. Mean dissolved organic carbon in water column, Station A,
Horsetooth Reservoir, on four dates and after incubation. 28
2. Dissolved organic carbon (left ordinate) and net seston
carbon, Dixon Reservoir. 29
3. Vertical distribution of dissolved organic carbon in Bear
Lake, 30 July and after various periods of incubation. 30
4. Vertical distribution of dissolved organic carbon in Cub
Lake, 30 July and after various periods of incubation. 31
5. Vertical distribution of dissolved organic carbon in Green
Lake, 11 August and after various periods of incubation. 32
6. Vertical distribution of dissolved organic carbon in Bear
Lake, 11 August and after various periods of incubation. 33
7. Mean net seston carbon in water column, Station A, Horse-
tooth Reservoir on two dates and after incubation. 34
8. Mean net seston carbon at 1 m depth, Dixon Reservoir,
five dates. Numbers above bars indicate number of replicate
samples. 35
9. Vertical distribution of dissolved organic carbon, Station A
Horsetooth Reservoir . 36
10. Vertical distribution of net seston carbon, Station A, Horse-
tooth Reservoir. Note change in scale for October. 37
11. Temporal and spatial distribution of dissolved organic
carbon in Spring Creek. 38
12. Vertical distribution of dissolved organic carbon in Bear
Lake. 63
13. Vertical distribution of net seston carbon in Bear Lake. 64
14. Vertical distribution of filter seston carbon in Bear Lake. 65
15. Dissolved organic carbon (triangles), filter seston carbon
(open circles) and net seston carbon in Dixon Reservoir. 66
16. Dissolved organic carbon, four stations, Spring Creek. 67
vi
-------
FIGURES (Continued)
Page
17. Dissolved organic carbon, four stations, Glacier Creek. 68
18. Changes in amounts of dissolved organic carbon (triangles),
filter seston carbon (open circles) and net seston carbon
from Dixon Reservoir during incubation (carboy experiment). 59
19. Changes in amounts of dissolved organic carbon (triangles),
filter seston carbon (open circles) and net seston carbon
from Dixon Reservoir during incubation (bottle experiment). 70
vn
-------
TABLES
No. Page
1. Effect of gas washing bottle in instrument train, values
as percentage absorption. 14
2. Detection of dissolved carbon in low concentrations, per-
centage absorption. 15
3. Analyses of Horsetooth Reservoir water with and without
addition of urea. 16
4. Dissolved organic carbon in Cub Lake water following dilu-
tion with distilled water. 17
5. Carbon determinations of Dixon Reservoir water. 18
6. Analyses of net seston incubated with and without addition
of nitrate, separate regression. 19
7. Analyses of net seston incubated with and without addition
of nitrate, combined regression. 20
8. Analyses of net seston incubated with and without addition
of nitrate, raw percentages of absorbance. 21
9. Analyses of Blue Lake DOC after 15 and 16 days of incuba-
tion in the laboratory. 22
10. Carbon content in replicate samples of net seston, Dixon
Reservoir, 29 September 1970. 23
11. Dissolved organic carbon (mg/liter) in Horsetooth Reser-
voir water (Station A), 15 July 1970 and after various
periods of incubation. 39
12. Dissolved organic carbon (mg/liter) in Horsetooth Reser-
voir water (Station A) 13 August 1970 and after various
periods of incubation. 40
13. Dissolved organic carbon (mg/liter) in Horsetooth Reser-
voir water (Station A) 9 September 1970 and after various
periods of incubation. 41
14. Dissolved organic carbon (mg/liter) in Horsetooth Reser-
voir water (Station A) 17 October 1970 and after various
periods of incubation. 42
viii
-------
TABLES (Continued)
No.
15. Dissolved organic carbon and net seston carbon in Horse-
tooth Reservoir water (Station C) 19 August 1970. 43
16. Dissolved organic carbon (mg/liter) in Bear Lake water
30 July and 23 September 1970 and after incubation. 44
17. Dissolved organic carbon (mg/liter) in Cub Lake water
30 July 1970 and after incubation. ' 45
18. Dissolved organic carbon (mg/liter) in Green Lake water
11 August and 3 September 1970 and after incubation. 46
19. Dissolved organic carbon (mg/liter) in Blue Lake water
11 August and 3 September 1970 and after incubation. 47
20. Net seston carbon (mg/liter) Horsetooth Reservoir
(Station A) two dates 1970. 48
21. Net seston carbon (mg/liter) Horsetooth Reservoir
(Station A) 9 September 1970 and after incubation. 49
22. Net seston carbon (mg/liter) Horsetooth Reservoir
(Station A) 17 October 1970 and after incubation. 50
23. Net seston carbon (mg/liter) in three lakes. 51
24. Dissolved organic carbon, net seston carbon and dry
weight of net seston, Bear Lake 23 September 1970. 52
25. Carbon content and dry weight of net seston, Dixon
Reservoir. All samples from 1 meter depth. 53
26. Carbon content and dry weight of net seston, Horse-
tooth Reservoir (Station A) 17 October 1970. 54
27. Physical characteristics of waters sampled. 61
28. Concentrations of dissolved (DOC), net seston (NSC)
and filter seston carbon (FSC) in Dixon Reservoir
water during incubation after heating to 80C. 62
IX
-------
SECTION I
CONCLUSIONS
Amounts of dissolved and particulate carbon in streams, lakes and reser-
voirs may fluctuate widely over short intervals of time. Dissolved
organic carbon tended to be more uniformly distributed with depth in
thermally stratified lakes than was the particulate fractions. The
total amount of organic carbon seemed to correspond to the trophic
status of the bodies of water investigated. That is, lakes judged to
be the most productive contained the largest instantaneous amounts of
organic carbon.
Under the conditions of this study, organic carbon did not decline uni-
formly or predictably in samples incubated in darkness at room tempera-
ture. The behavior of incubated samples could not be satisfactorily
related to the apparent trophic status of the bodies of water from
which samples were drawn.
-------
SECTION II
RECOMMENDATIONS
The temporal and spatial distribution of dissolved and particular or-
ganic carbon should continue to be investigated in a number of waters
covering a broad trophic spectrum. Inorganic carbon as well as organic
should be monitored in order to better elucidate the movements of car-
bon in the freshwater environment. Such information is essential to
evaluating the role of carbon in eutrophication.
Laboratory investigations of the mineralization of organic carbon during
incubation should be conducted. Inorganic carbon should be monitored
in order to stoichiometrically follow transfers within the total pool
of carbon. Particular attention should be paid to the role of bacteria
in incubated samples.
-------
SECTION III
INTRODUCTION
A large number of observations made in many areas and depths of the
world's oceans have clearly revealed the presence of a sizeable fraction
of detrital organic material that is refractory to bacterial breakdown.
The absolute quantity for any geographic area appears to be remarkably
constant and independent of local environmental conditions.
A parameter indicative of the degree of eutrophication of a body of
water and insensitive to local or seasonal conditions would be of con-
siderable value in detecting incipient pollution arising from unwanted
eutrophication as well as theoretical importance.
The objectives of the research were 1) to measure instantaneous amounts
of organic carbon, both particulate and dissolved, in a number of fresh-
waters ranging from unproductive alpine tarns to moderately productive
reservoirs on the plains; 2) to determine if a fraction of particulate
organic carbon refractory to bacterial decomposition could be demon-
strated and 3) to relate refractory or nonrefractory organic carbon to
trophic levels of the waters studied.
-------
SECTION IV
METHODOLOGY AND PROCEDURE
Organic carbon is, for the purposes of this investigation, divided into
three categories based on particle size: 1) net seston organic carbon
(NSC) - particles retained by plankton netting of 70 nm mesh; 2) filter
organic carbon (FSC) - particulate carbon passing the 70 nm net, but
retained on glass fiber filters, i.e. lower limit of about 1 - .5 nm;
3) dissolved organic carbon (DOC) - carbon in solution or particles
fine enough to pass through glass fiber filters.
Field Sampling - Samples of water and included seston were obtained with
either a Kemmerer or Van Dorn sampler. Rental boats are available at
Horsetooth Reservoir near Fort Collins; an inflatable rubber raft was
packed to remote lakes. Vertical distribution of carbon was assessed
by taking samples at each 1-m interval from surface to near bottom.
In order to reduce variability, samples were integrated by combining
adjacent probes according to this scheme: in the uppermost 10-m stratum
immediately adjacent samples were combined, i.e. 1 and 2m; 3 and 4; 5
and 6 etc. Between depths of 11 and 25 m, three adjacent samples were
composited, i.e. 11, 12, and 13 m; 14, 15 and 16 m, etc. At depths
exceeding 25 m, compositing was done by 4 m intervals, i.e. 26, 27, 28,
and 29 m, etc.
Vertical profiles of temperature were usually obtained at each sampling
site.
Numerous investigations of sestonic and dissolved carbon have clearly
indicated that the dissolved fraction is substantially greater than the
particulate (i.e. Birge and Juday 1934). In order to secure sufficient
NSC for reliable analyses and simultaneously reduce the amount of water
to be transported from the field to the laboratory, NSC was strained
from the water at the time of sampling. Strainers were made by cutting
the bottoms from polyethylene bottles of 1000 and 500 ml capacities.
Holes large enough to remove most of the ends were drilled in the caps;
discs of 70 u netting were cut to fit snugly within the caps.
Two to 12 liters of water could be strained before clogging of the net
became a serious problem, thus yielding sufficient net seston for analyses.
From the water passing through the strainer, 250 to 1000 ml samples were
retained for FSC and DOC fractions and taken to the laboratory.
Laboratory Procedures - Samples were passed through a Millipore vacuum
filtering device fitted with Whatman GFC glass fiber filters. Known
volumes of water were filtered; amounts varied according to amount of
seston present. Filtering stopped when vacuum greater than 4 psi was
required to overcome clogging of filters. Water passing glass fiber
filters was saved for DOC determinations and some chemical analyses
-------
(i.e. hardness, calcium, NO,., PO,, and SiO_). Chemical analyses were
incidental to the main objectives of the investigation and will not be
discussed in the present report.
Menzel and Vaccaro (1964) developed a technique for determining organic
carbon. Their procedures with slight modification, to be discussed
later, were followed. In essence, organic carbon is digested with per-
sulfate and converted to gaseous carbon dioxide. The gas is passed
through a Beckman infrared analyzer where infrared absorption by the
CO is determined. Inorganic carbon present as CO- from the atmosphere
or as HCO is removed by acidification and stripping with nitrogen prior
to persulfate digestion. Digestion is accomplished in sealed glass
ampoules autoclaved for 40 min at 31 psi.
For DOC, 5 ml of filtered lake water was introduced directly into 10 ml
ampoules, heated and analyzed.
Net seston was divided into aliquots in the laboratory, one for imme-
diate analysis, the others to be analyzed after varying periods of in-
cubations. Aliquots of net seston were suspended in 5 ml of distilled
water. Analyses proceeded as for DOC.
It was planned to make aliquots of FSC as was done with NSC. Filters
as received from the manufacturer are contaminated with carbonaceous
compounds that influence analyses. Menzel and Vaccaro (1964) recommended
precombusting filters to remove carbon contamination. The only furnace
available at the outset of this study melted filters when set at its
lowest temperature. Wet oxidation using sulfuric acid and potassium
dichromate was tried, but proved unsatisfactory. Filters were assayed
for carbon in the hope that a "blank" or correction factor could be
obtained. Wide variation in results dashed this hope; therefore, no
determinations of FSC were made prior to 1 January 1971.
The infrared analyzer was calibrated by preparing solutions of dextrose
containing 2.5, 5.0, 10.0, 20.0, 40.0, 1000.0 and 200 mg C/liter.
Series were prepared by serial dilution of a concentrated stock solution.
New stock solution was made for each series of calibration standards.
Calibration series overlapped - i.e. each new series was checked against
the previous series.
Before unknown samples were analyzed, the machine was set to give 100%
scale deflection at a known carbon value i.e. for routine determination
of DOC 10 mg C/liter; heavy net seston sometimes required that the
machine be set at 40 or even 100 mg C/liter.
After setting, not less than three different standards were analyzed,
i.e. with the analyzer set on 10 mg C/liter, samples of 10, 5 and 2.5 mg
C/liter were used in addition to distilled water. From the calibration
run a least squares regression line relating carbon to percentage of
deflection on scale was computed. Unknowns were then analyzed and
-------
percentages of deflection converted to carbon values from the regression
line.
Methodological Modifications - Menzel and Vaccaro (1964) cautioned
that ampoules must be sealed in an oxyflame to guard against contami-
nation from combustion products. At first, ampoules were sealed in an
ordinary Bunsen burner flame since oxyequipment was not immediately
available. Combustion contamination was at once revealed by erratic
and unpredictable values using known concentrations of dissolved dex-
trose and distilled water. After oxyflame equipment became available,
no further trouble of this sort was experienced.
Menzel and Vaccaro (1964) recommended Mg (CIO,') as an agent to dry the
gas stream immediately prior to entering the infrared analyzer. We
substituted indicating Drierite for Mg (CIO,) with excellent results.
Drierite was changed as soon as it changed from blue to pink in the
upper one fourth of the drying tube.
Strickland and Parsons(1968) in their discussion of Menzel and Vaccaro's
method pointed to the importance of organic-free water. Following their
suggestions distilled water containing potassium persulfate and phos-
phoric acid was refluxed prior to redistilling in all glass apparatus.
Because Menzel and Vaccaro (1964) studied carbon dissolved in sea water,
they were concerned with chlorine contamination, resulting from per-
sulfate action on sea water. They passed the CO- - Nitrogen stream
through a washing bottle containing potassium dichromate dissolved in
sulfuric acid. Chlorine contamination seemed unlikely, at least in
most of the waters to be investigated; therefore, an experiment testing
the effect of removal of the gas washing bottle was performed. A least
squares regression was determined from water containing 0, 2.5, 5.0,
and 10.0 mg C/liter with the washing bottle in the instrument train;
four replicate samples of Dixon Reservoir water were then analyzed.
The wash bottle was removed and the entire process repeated. Least
squares regression equations were, with bottle in place, Y=7.42X + 28.30;
with bottle removed, Y=7.07X + 26.30. Apparently the wash bottle did
not enhance performance (Table 1) and thus has been removed from routine
use.
Sensitivity of Method - The ability of the method to detect low concen-
trations of dissolved carbon was evaluated by preparing replicate samples
containing 0, 1.0, 2.0 and 3.0 mg C/liter. Absolute amounts of dissolved
carbon in 5 ml ampoules were, 0, .005, .010 and .015 mg respectively.
Because 95% confidence intervals did not overlap (Table 2); we concluded
that the analyses could detect the differences in carbon concentrations.
A regression line was established with 10 mg C/liter standards producing
full scale deflection or 100% absorbance. Carbon values were then picked
from the regression line using percentages obtained from sensitivity
study (Table 2). Means and ranges were 0 mg C/liter, .35 mg, 0.-.75;
-------
1 mg C/liter, 1.40, 1,05-1.75; 2 mg C/liter, 3.0, 2.40-3.45 and 3 mg
C/liter, 5.5, 4.45-6.60.
Clearly the use of a regression line to convert percentages of absorp-
tion to carbon concentration introduces problems that are distinct from
those associated with accuracy and precision of detection of carbon in
unknown s amp1e s.
Accuracy of_ Method - It is possible to calibrate the analyzer against
weighed amounts of carbon, but does the method convert all organic car-
bon present to carbon dioxide? Insight to this problem was gained by
adding a weighed amount of urea to filtered water taken from Horsetooth
Reservoir. Eight replicates were analyzed; eight replicates of Horse-
tooth water lacking the urea spike were also analyzed (Table 3). The
analyzer was adjusted so that 10 mg C/liter, as dextrose, produced
full scale deflection. Mean carbon content of unspiked water was 3.20 mg
C/liter and that of the spiked samples was 6.35 mg C/liter. The spike
consisted of 15.9 mg urea in one liter of water, thus 15.9 v 12.0 _ ...
1 X 60.6 ~
mg C. The mean value of spiked samples, 6.35 minus 3.15, carbon added
as urea equals 3.20, the mean value for unspiked samples.
Undoubtedly, 100% recovery of added carbon is fortuitous, but low coeffi-
cients of variability are reassuring.
Sensitivity of analyses and method was evaluated by diluting lake water
with distilled water (Table 4). Water from Cub Lake, a small brown-
water montane lake, was filtered, diluted and analyzed with the machine
calibrated at 10 mg C/liter.
Non-overlap of confidence limits suggests that the method and machine
discriminated among 0, one-half and two-thirds dilutions. Percentage
of error (SE X 100/Y) increased with increasing dilution, namely 1.3,
2.1 and 4.9%.
The machine was recalibrated at 20 mg C/liter; filtered Cub Lake water,
full strength and 50% diluted, was re-analyzed. Two determinations of
undiluted water yielded identical results, 8.8 mg C/liter. Two analyses
of 50% dilution gave a mean of 5.2 mg C/liter with a 95% limit of 2.66-
7.74, or nearly 40% error.
Filter Seston Carbon - After unsuccessful attempts to either decontami-
nate filters or find a suitable correction factor, we tried analyzing
filtered and unfiltered lake water, hoping to estimate FSC by difference.
Three replicates of unfiltered Cub Lake water were compared with two
replicates of filtered water with the carbon analyzer calibrated at
20 mg C/liter. Identical values of 8.8 mg/liter were obtained from
filtered samples. Mean of unfiltered samples was 9.2 mg/liter, but the
95% limits were 8.3-10.1, thus suggesting insignificant difference.
Percentage of error for three unfiltered samples was 2.3%.
10
-------
Similar experiments were done using water from Dixon Reservoir with the
analyzer calibrated on 10 mg C/liter (Table 5).
Most likely these differences are not significant. The recorder scale
is probably not accurate to more than +_ one small division, i.e., 1%
of full deflection. Because the largest value is only 2 percentage
points greater than the smallest, machine error could account for dif-
ferences.
Rupture of algal cells during filtration could, however, cause abnor-
mally high values of DOC and simultaneously diminish FSC. Rupture of
fragile cells is a problem frequently encountered in C-14 uptake assays
of primary production rate (Vollenweider, 1969).
Experiments were repeated with Dixon Reservoir water and the analyzer
calibrated at 5 mg C/liter. Three replicates of unfiltered water each
gave 115% absorbancy while three filtered values were 115, 115 and 120%.
All values were determined by extrapolation, a dangerous procedure
since one cannot assume linearity when changing scales on the analyzer.
Cub Lake and Dixon Reservoir were chosen because previous studies (Reed
and Reed 1970 and unpublished) indicated substantial standing crops of
nannoseston in their waters. The method of difference is not satis-
factory for distinguishing FSC from DOC in our procedure.
Incubation - One of the objectives of the study was to follow the release
of carbon during incubation of samples in the laboratory. At first,
breakdown of DOX was followed in 300 ml BOD bottles, later 60 ml reagent
bottles of filtered lake water were stored in dark boxes at room tempera-
ture for periods lasting to 90 days. Net seston was incubated in dis-
tilled water. One possible objection is that bacteria may have become
nutrient limited thus producing unnatural rates of breakdown of organic
matter. Therefore, replicate samples of net seston obtained from Horse-
tooth Reservoir were divided into three lots, one for immediate carbon
analysis, two to be incubated. One lot of incubated samples was treated
with Ca(NO_)9 at 5 mg/liter. Because this experiment illustrates a
number of problems encountered in incubation studies, it will be de-
scribed in detail.
Several liters of reservoir water were obtained and thoroughly mixed in
a large container in the laboratory. One-liter subsamples were drawn,
with continuous stirring. Each subsample was strained through 70 u
plankton netting. The catch was carefully rinsed into 10 ml ampoules
with 5 ml of distilled water and treated in the routine manner.
The infrared analyzer was calibrated at 20 mg C/liter (0.1 mg C produced
full scale deflection). A regression line was determined using standards
of 20, 10, 5, and 2.5 mg C/liter and distilled water (0 carbon). Six
replicates were then analyzed.
11
-------
Twelve more replicates were drawn and strained in the same manner; to
six, Ca(NO ) was added. Each ampoule was covered with a cap of alumi-
num foil. Samples were incubated in a dark box at room temperature,
22C+ 3C for nine days. They were then analyzed for carbon content
after calibrating the analyzer with 20 mg standards. A new regression
line was determined.
Thus, this experiment involved not only the possible effects of the
nitrate on incubation but a number of other sources of variation:
1) subsampling, 2) possible loss of seston in straining, 3) contamina-
tion of distilled water used in black, 4) standards and 5) inherent in
machine.
Before discussing results of this particular experiment, it should be
emphasized that it was done rather early in our experiences with the
analyzer and procedure. Later the analyzer was allowed to run continuously,
thus eliminating warm-up problems which may have influenced earlier re-
sults. Reproducibility of standards received constant attention and,
as far as can be determined, was well within other sources of error.
Despite repeated efforts including double glass distillation we were
unable to achieve zero absorbance from distilled water blanks. Exami-
nation of a graph of absorbance vs. carbon content given by Menzel and
Vaccaro (1964) revealed that their regression line did not pass through
the origin but gave a positive intercept on the ordinate, indicating
absorbance from supposedly carbon free water.
Distilled water that would yield zero absorbance was not achieved, al-
though following the procedure of Strickland and Parson (1968) "carbon-
free" water that regularly reads less than 10% absorbance, usually less
than 5% was obtained. Thus, regression equations were influenced by
different values for supposedly carbon free water. The regression
equation for the control series of samples, those analyzed immediately
upon collection, was Y = 4.40 X + 13.64. Mean carbon content was .033
mg C per liter of reservoir water strained (Table 6).
After nine days, the incubated samples were analyzed; new regression
equation was Y = 4.10 X + 20.50. Mean of enriched samples was .030
mg C/liter and that of unenriched samples was .037 (Table 6).
These data indicated no significant difference among control and incu-
bated samples at the 95% level. Confidence levels of control and unen-
riched incubation samples did not overlap at 70% level.
A new regression equation was derived using blanks and standards of the
two separate regressions and percentages of absorbance reconverted to
carbon (Table 7). Results were not noticeably improved. Means retained
the same relative rankings, but amounts were nearly halved. Results
were analyzed using raw absorbance percentages; thus eliminating re-
gression equations (Table 8).
12
-------
The mean of unenriched incubated samples apparently differed significantly
at 70% level from the means of the other samples. Omitting conversion
of absorbance to carbon slightly reduced percentages of error, but did
not alleviate the seeming problem of incubated means exceeding the con-
trol. Perhaps nine days incubation was too short a period for signifi-
cant differences to develop; however, the difficulties of obtaining
exactly the same regression equations at different times are illustrated.
Ways in which incubated samples could gain significant amounts of carbon
during incubation will be considered in the discussion of results of
other incubation experiments.
Difficulties in securing large amounts of seston and in storing and
handling samples in the laboratory discouraged multiple sampling of
seston. Therefore, incubation of NSC samples was not run in duplicate;
that is, only one sample represented any given depth over a particular
time interval. In contrast, rather large volumes of water for DOC
determinations could be readily obtained and stored.
Presumably DOC would change little in 24 hours after having been incu-
bated for 15 days, thus offering the possibility of checking reproduci-
bility of machine operation and laboratory procedures.
Samples of Blue Lake water representing various depths were analyzed
after 15 days of incubation. On the next day a second series was run.
The machine was recalibrated for the second trial and two separate
regression lines determined for converting absorbancy to mg C/liter.
There was no significant difference in means of the two trials (Table 9).
High variability indicated that DOC was not evenly distributed in the
water column. From a methodological point of view, it is gratifying
that the samples retained their rankings relative to each other.
Variability Among Replicate Samples - The variability in carbon content
among replicate samples of net seston was evaluated by drawing a quantity
of water from Dixon Reservoir and transporting it to the laboratory where,
with continuous stirring, five aliquots were taken for carbon analysis.
The analyzer was calibrated so that .50 mg C (100 mg C/liter) produced
full-scale deflection; regression equation was Y = .84 X + 22.08.
Results are summarized in Table 10.
Percentage of error in NSC determination seems high, but perhaps not
excessively so when possible sources of error are considered; for exam-
ple, in drawing subsamples, loss in straining and washing seston into
ampoules and problems inherent in large particles, eight large Daphnia
cannot be evenly distributed among five ampoules.
13
-------
Table 1. Effect of gas washing bottle in instrument train, values as
percentage absorption
Bottle
n
Y (%)
S2
s
SE
95% limits
error
CV
in place
4
74.75
21.43
4.63
2.31
67.40 - 82.10
3%
6%
Bottle removed
4
71.62
2.21
1.49
.74
71.38 - 71.86
1%
2%
Abbreviations used in this and subsequent tables:
n, number of observations; YJ mean of observations
o
S , variance; s, standard deviation
SE, standard error s//n~; 95% limits, fiducial limits
P=.95; error, 100 SE/y; CV, 100 s/y
14
-------
Table 2. Detection of dissolved carbon in low concentrations,
percentage absorption
Concentration, mg. C/liter 0123
n 7 8 10 10
Y (%) 27.36 34.81 '46.95 65.80
S2 7.727 8.924 36.914 125.678
s 2.780 2.987 6.076 11.210
SE 1.051 1.056 1.921 3.545
95% limits 24.79- 32.31- 42.60- 57.78-
29.93 37.31 51.30 73.82
error 7, 4345
CV 10 10 13 19
15
-------
Table 3. Analyses of Horsetooth Reservoir water with and without
addition of urea
Spiked Unspiked
n
Y absorbance 71.56 47.00
S2 12.60 10.00
s 3.55 3.16
SE 1.26 1.12
error % 2 2
CV 57
95% Confidence limit, 68.59 - 74.53 44.36 - 49.64
absorbancy
95% limits, 5.95 - 6.70 2.85 - 3.50
carbon mg/liter
16
-------
Table 4. Dissolved organic carbon in Cub Lake water following
dilution with distilled water
83 1/3
n 233 1
0
2
7
percentage
50
3
.95 4.22
.02 .03
.14 .16
.10 .09
dilution
66 2/3
3
2.63
.05
.23
.13
Y mg C/liter 7.95 4.22 2.63 1.25
S2
s
SE
95% limits 6.68-9.22 3.82-4.62 2.07-3.19
error % 125
CV % 2 4 9
17
-------
Table 5. Carbon determinations of Dixon Reservoir water
unfiltered
filtered
percentage
absorbance
77.0
76.5
78.5
mean
carbon
mg/liter
7.05
7.00
7.20
7.08
percentage
absorbance
78.5
78.0
78.5
carbon
mg/liter
7.20
7.15
7.20
7.18
18
-------
Table 6. Analyses of net seston incubated with and without addition
of nitrate, separate regression
n
Y (mg C/liter)
S2
s
SE
95% limits
error ?<,
CV
Initial
6
.0331
.0000383
.00619
.00277
.0260-.0402
8
19
Enriched
6
.0302
.0000297
.00544
.00244
.0239-.0365
8
18
Unenriched
6
.0368
.0000543
.00736
.0033
.0283-.0452
9
20
19
-------
Table 7. Analyses of net seston incubated with and without addition
of nitrate, combined regression
Y mg C/liter
S2
s
SE
957o limits
error %
CV 7o
Initial
6
.0151
.0000104
.00322
.00144
.0114-.0188
10
21
Enriched
6
.0162
.0000049
.00221
.00099
.0137-.0187
6
14
Unenriched
6
.0199
.0000122
.00349
.00156
.0159-.0239
8
18
20
-------
Table 8. Analyses of net seston incubated with and without addition
of nitrate, raw percentages of absorbance
Initial Enriched Unenriched
n 666
Y (%) 42.75 44.58 50.66
S2 29.67 12.84 36.57
s 5.48 3.72 6.05
SE 2.44 1.66 2.70
95% limits 36.49-49.01 40.30-48.86 43.71-57.61
error % 6 4 5
CV % 13 8 12
21
-------
Table 9. Analyses of Blue Lake DOC after 15 and 16 days of incubation
in the laboratory
depth
1-2 m
3-4
5-6
7-8
n
Y (mg C)
S2
s
SE
95% limits
error
CV
day 15
1.90 mg C/liter
1.30
1.10
3.30
4
1.90
.987
.993
.573
.32-3.48
30
52
day 16
2 . 15 mg
1.55 mg
1.25 mg
3.25 mg
4
2.05
.780
.883
.510
.64-3.45
25
43
C/liter
C/liter
C/liter
C/liter
22
-------
Table 10. Carbon content in replicate samples of net seston, Dixon
Reservoir, 29 September 1970
n 5
Y mg/liter .352
S2 .0303
s .1741
SE .0778
95% limits .1360-.4880
error % 22
CV % 49
23
-------
SECTION V
FIELD AND LABORATORY STUDIES PRIOR TO 1 JANUARY 1971
Dissolved Organic Carbon - The results of this investigation contain two
sorts of data, namely, observations on the instantaneous concentrations
of organic carbon at various depths in several different kinds of water
bodies, and secondly, on changes in organic carbon concentrations after
varying periods of incubation. Since incubation studies must include
instantaneous values at moment of sampling, it is convenient to present
results of incubation studies first. T always refers to date of sampling,
T T , T and T refer to 15, 30, 60 and 90 days after T . In the
tables to follow, values are usually the average of two, sometimes three
determinations or, occasionally only one determination.
Mean value of dissolved organic carbon in water column at Station A on
Horsetooth Reservoir, a large mesotrophic body of water at an altitude
of 1656 m, ranged from 2.48 mg/liter to 6.29; from 67 to 82% remained
after 90 days incubation (Tables 11-14). Mean carbon fell in first 15
days of incubation in two cases and in two others rose to a value
higher at T than at T (Figure 1).
A series of samples was obtained over the deepest water in Bear Lake,
a montane lake, on 30 July 1970 for incubation (Table 16). After 90
days, the mean carbon content was reduced from 5.30 mg/liter to 2.47,
or a reduction of 53%. A rapid loss of carbon occurred during the first
15 days of incubation; little change took place in the next 15 days,
but carbon increased to 82% of the original value at the end of 60 days.
During the last 30 days, carbon content declined rapidly (Figure 3).
Samples were obtained from Cub Lake, a brown-water montane lake, 30 July
1970 (Table 17). After 90 days of incubation mean DOC content had de-
creased 49% (Figure 4).
Two alpine lakes (Tables 18 and 19) were sampled on 11 August 1970. In
general, mean DOC decreased in the first 15 days of incubation; composite
samples from three depths in highly oligotrophic Green Lake exhibited
similar trends in loss of DOC (Figure 5). Samples from five depths in
Blue Lake behaved dissimilarly and, after 90 days, mean loss of DOC was
only 6% (Figure 6).
Observed extreme values of DOC at Station A, Horsetooth Reservoir ranged
from 1.50 to 8.20 mg/liter. Early in the summer maximum values were
found in the upper 2 meters (Figure 9). In general, DOC increased in
the water column as summer progressed to autumn and on 17 October ex-
hibited a bimodal distribution. DOC apparently collected in the hypolim-
nion (Figure 9) and in the epilimnion.
Details of the vertical distribution of DOC in the upper 14 meters of
Horsetooth Reservoir at Station C on one occasion are shown in Table 15.
The epilimnion extended down to 9 m and metalimnion ranged from 10 to 14 m.
25
-------
Spatial differences in DOC in Horsetooth Reservoir are suggested by the
data of Tables 12 and 15. Stations C and A are about 1 km apart, but
Station C is in the flow of the inlet; whereas Station A is not in the
direct flow through the reservoir. The stations were sampled about one
week apart. Little difference was observed in the upper 10 m, but
from 11-14, Station C contained about 1 mg C/liter less than Station A.
Apparently the amount of DOC in one body of water can vary considerably
over short periods of time, Dixon Reservoir, a small reservoir near
Horsetooth Reservoir, was sampled on five occasions between 7 August
and 29 September. All samples were drawn from a depth of 1 meter.
DOC ranged from 3.98 to 7.35 mg/liter (Figure 2).
Elsewhere DOC ranged from about 1.20 to 2.50 mg/liter in Green Lake; 1.42
to 4.40 in Blue Lake; 3.90 to 5.98 in Bear Lake and was about 8 mg/liter
in Cub Lake.
Net Seston Carbon - Because NSC comes from rather large particles which
are often unequally distributed among subsamples, results are expected
to be more variable than in DOC, which, in theory at least, is equally
distributed within one water sample. Losses of samples and other labora-
tory malfunctions further reduced the number of useable results arising
from incubation studies of net seston.
Net seston samples were obtained at Station 2, Horsetooth Reservoir on
15 July and 13 August 1970 (Table 20). Initial values ranged from
.0303 mg C/liter (9-10 m, 13 August) to .0020 mg C/liter (35-37 m, same
date). Mean NSC value on 15 July was .0117 mg/liter; this increased to
.0130 mg/liter on 13 August. After 60 days of incubation, eight samples
contained a mean of .022 mg C/liter or 188% initial value.
The series of 9 September (Table 21) and 17 October (Table 22) are more
complete. Extreme initial values ranged from .157 mg C/liter (1-2 m,
17 October) to a low of .0082 (38-41 m, 9 September ). The mean values
for NSC during incubation ranged from 94 to 162% of initial value for
the 9 September series (Figure 7). Initial mean value for the 17
October series was much greater, but fell to 55% of the original value
after 90 days.
Net seston from Dixon Reservoir was not incubated in the first half of
this investigation; the reservoir was, however, repeatedly sampled to
gain insight into changes in NSC. Between 15 August and 15 October
1970, the amount of NSC varied by an order of magnitude, the largest
value being 12 times the smallest (Figure 8).
Within 30 days, the mean NSC of Bear Lake samples collected 30 July was
reduced 91% (Table 23). A series of NSC samples drawn from Blue Lake
on 3 September averaged slightly less in carbon and lost 93% after 60
days incubation. Carbon in net seston taken from Green Lake 3 September
averaged slightly over one-half as much as Blue Lake and ultimately lost
about the same percentage after 60 days of incubation (Table 23).
26
-------
At Station A in Horsetooth Reservoir, NSC varied from .002 to .157 ing/liter.
Depth and time trends (Figure 10) are not easy to summarize; NSC did
increase through the summer and usually the maximum was found at some
depth shallower than 10 m. Bear and Blue lakes usually showed about .04
mg NSC/liter and Green Lake about half of that amount (Table 23). NSC
values in Dixon Reservoir ranged over an order of magnitude in two months
time, 0.029 to 0.352 mg/liter (Figure 8).
The relationship of NSC to dry weight of net seston was examined on
several occasions. Vertical profile of NSC vs dry weight was examined
at Station A, Horsetooth Reservoir, 17 October; carbon was 10% of dry
weight of seston (Table 26).
NSC vs dry weights of seston was examined on two different occasions in
Dixon Reservoir; during a bloom of Aphanizomenon, mean weight of carbon
was 16% of mean dry weight (Table 25). On 12 October when seston dry
weight was nearly double that on 29 September, carbon was less than 1%
of dry weight. These percentages are low for carbon; Kerr, Paris and
Brockway (1970:) reported that carbon accounts 50-77% of dry weight
of algae and about half of the dry weight of bacteria. Reed and Reed
(1970) showed that inorganic materials compose a large part of the ses-
ton in Dixon Reservoir.
Samples were taken from Bear Lake 23 September and analyzed for NSC, dry
weight of NS and DOC by depth strata. On the average NSC was only 4.4%
of the seston dry weight (Table 24) and DOC was 344 times greater than
NSC.
Observations of DOC in Streams - DOC was measured in Spring Creek, a
plains stream near Fort Collins, on four occasions. Mean values at eight
stations along the 10 km length of the stream ranged from 1.22 to 13.8
mg/liter (Figure 11). On 7 August the Cache la Poudre River was sampled
at seven locations ranging from 10 km upstream from Fort Collins to 6 km
downstream from the city limits. Eleven determinations varied between
3.35 and 4.55 mg DOC/liter. Effluent from a tertiary sewage treatment
plant contained 7.48 mg DOC/liter, where it entered the Poudre River.
27
-------
'0 '15 '30 '60 '90
Figure 1. Mean dissolved organic carbon in water column,
Station A, Horsetooth Reservoir on four dates
and after incubation.
28
-------
7-
5-
o:
t 4
oğ
E
2-
I -
15 30 15 30
August September October
Figure 2. Dissolved organic carbon (left ordinate)
and net seston carbon, Dixon reservoir.
29
-------
a:
UJ
oğ
E
'0 '15 '30 '60 190
Figure 3. Vertical distribution of dissolved organic carbon
in Bear Lake, 30 July and after various periods
of incubation.
30
-------
u>
'0
Figure 4.
'15 '30 '60 '90
Vertical distribution of dissolved organic carbon in Cub Lake,
30 July and after various periods of incubation.
-------
u;
oğ
E
T0 TI5 T30 T60 T90
Figure 5. Vertical distribution of dissolved organic carbon in
Green Lake, 11 August and after various periods of
incubation.
32
-------
UJ
Figure 6. Vertical distribution of dissolved organic carbon in Bear
Lake, 11 August and after various periods of incubation.
-------
.08
.07
.06-
.05-
.04-
LJ
.03^
.02H
.01-
17 OCTOBER
9 SEPTEMBER
-r
-r
'15
'30
'60
'90
Figure /, Mean net seston carbon in water column, Station A,
Horsetooth Reservoir on two dates and after
incubation.
34
-------
.4-
.3-
.1-
rfl
6
n
30 30
August September October
Figure 8. Mean net seston carbon at 1 m depth,
Dixon Reservoir, five dates. Numbers
above bars indicate number of replicate
samples.
35
-------
Oct 17
OJ
1 -?
-4
5-6
7-P
9-10
IP ' "2
1C 1 w>
14 !£
It)
17-19
20-22
23-25
O O . O Q
3O ^^
34-37
38-41
42-45
I i
Ju y
r*~
T
r "
1
L
r
i
i
__
.J
15 .---"
. - - j '
-^J ~~~
i
r _ -
.I _---"'
i . . "
[
1 - "~~
r j
1 . - - " '
^^^
1
,.-- ^
h - - "
____^ "
__^- '
^-- ""~~
_^~~^- ""
Aug 13
-
_
i
i
\
.. -
\
r""
j
"
j --
-'
I
i
^ ---
-
-
-
""""
. -
.^
___
...
- "
- "
___
"' "
- "
- '
Sapj 9
-
_._.-.^
'
_-
_
^
r
r""
L
,
.-
I
[
- '
__--"
- "
^ -
^_ - "
.-
.._ ~~
-
_.--
_ .- J
^_-
"
~^
^^-~-~~~
_-.~- "^
4mg/lf*f
*
-^" "
-
. '
^,^~~"
-~ "
--"
__. - -"^"
~~~~
"
1
J
Figure 9. Vertical distribution of dissolved organic carbon, Station A, Horsetooth Reservoir.
-------
Oct 17
Figure 10. Vertical distribution of net seston carbon, Station A, Horsetooth
Reservoir. Note change in scale for October.
-------
I 7 VIII
la
Ib
7 VIII
J25VIU
17IX
23 X I
7VIH
I 7 VIII
25 VIII
I 17 IX
25 VIII
1
3
S
Figure 11. Temporal and spatial distribution of dissolved organic
carbon in Spring Creek,
38
-------
Table 11. Dissolved organic carbon (mg/liter) in Horsetooth Reservoir
water (Station A) 15 July 1970 and after various periods of incubation.
Sample
depth, m
0
1-2
3-4
5-6
7-8
9-10
11-13
14-16
17-19
20-22
23-25
26-29
30-33
34-37
38-41
42-45
T
io
-
5.28
1.75
1.90
3.35
2.65
2.18
2.85
1.50
2.40
1.80
2.25
-
2.20
-
2.10
T15 T30
-
3.92 4.70
4.55 3.55
2.25 3.55
3.52
1.62 3.58
3.62
3.85
3.70 3.25
4.20 3.25
3.80 3.30
2.60 3.20
4.50 3.55
3.70 4.00
1.08 2.85
3.85
T60
-
3.65
3.25
2.40
2.55
2.35
2.55
2.50
2.50
2.20
2.55
2.25
2.20
2.55
2.35
2.55
T90
-
1.80
1.85
2.05
1.80
2.05
1.65
2.45
1.85
2.10
1.90
1.65
1.25
1.40
1.85
1.15
Y 2.48 3.26 3.57 2.56 1.79
100 Ti/To 100 131 144 103 72
39
-------
Table 12. Dissolved organic carbon (mg/liter) in Horsetooth Reservoir
water (Station A) 13 August 1970 and after various periods of incubation
Sample
depth, m
0
1-2
3-4
5-6
7-8
9-10
11-13
14-16
17-19
20-22
23-25
26-29
30-33
34-37
38-41
42-45
T T-i
"o 1
4.10
8.20
5.30
4.25
3.55
3.35
3.32
3.05
3.15
3.50
3.10
5.38
3.48
3.50
3.52
3.52
5 T30
4.05
5.82
6.60
6.05
6.95
4.52
3.67
4.22
4.30
2.88
3.07
3.27
3.32
3.50
3.18
3.80
T60
1.30
2.70
3.75
2.75
3.25
2.75
3.30
3.50
2.05
1.85
2.45
1.85
1.60
1.85
1.95
1.15
T90
2.25
2.20
4.40
4.25
3.00
3.30
3.25
4.35
2.55
2.30
3.30
2.65
2.25
3.15
2.35
2.02
4.02 4.32 2.38 2.97
100 1ħ/T0 100 107 59 74
40
-------
Table 13. Dissolved organic carbon (mg/liter) in Horsetooth Reservoir
water (Station A) 9 September 1970 and after various periods of incuba-
tion
Sample
depth, m
0
1-2
3-4
5-6
7-8
9-10
11-13
14-16
17-19
20-22
23-25
26-29
30-33
34-37
38-41
41-45
To
3.85
3.95
3.75
3.95
3.55
3.80
3.65
-
3.85
2.90
2.80
2.80
2.90
3.40
3.20
3.85
T15
2.15
2.50
2.45
2.40
2.25
2.30
2.15
2.05
2.05
2.05
1.95
2.00
1.85
1.70
1.95
2.15
T30 T60
3.00
2.40
2.25
2.65
2.60
2.40
2.50
3.00
2.95
2.85
2.50
2.30
2.55
2.90
2.40
3.00
T90
3.60
1.95
2.05
2.35
2.10
2.10
2.10
1.85
1.70
2.00
2.50
2.70
2.25
2.90
2.65
-
3.48 2.12 2.64 2.32
100 Tħ/T0 100 62 76 67
41
-------
Table 14. Dissolved organic carbon (mg/liter) in Horsetooth Reservoir
water (Station A) 17 October 1970 and after various periods of incuba-
tion
Sample
depth, m
0
1-2
3-4
5-6
7-8
9-10
11-13
14-16
17-19
20-22
23-25
26-29
30-33
34-37
To
5.85
5.95
7.50
7.62
6.65
5.42
6.38
5.50
4.82
4.52
6.22
6.55
6.68
8.35
T13 ?3
4.05
3.50
4.15
5.10
4.40
4.90
3.90
4.25
2.20
3.00
2.80
2.85
2.95
3.00
0 T60 T90*
-
5.25
-
6.87
4.60
5.75
5.60
5.15
5.75
3.62
3.90
5.50
3.82
5.95
6.29 3.65 5.15
100 Ti/T0 100 58 82
* average of three determinations/depth
42
-------
Table 15. Dissolved organic carbon and net seston carbon in Horsetooth
Reservoir water (Station C) 19 August 1970
Sample
depth, m
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
DOC
mg/liter
2.5
3.6
3.6
3.3
3.3
3.6
3.4
3.4
3.2
3.4
4.0
2.8
2.0
2.2
2.4
NSC
mg/liter
-
.043
.034
.041
.035
-
.041
.044
.043
.042
.044
.043
.029
.031
.025
.038
43
-------
Table 16. Dissolved organic carbon (mg/liter) in Bear Lake water
30 July and 23 September 1970 and after incubation
30 JULY
Sample
depth, m
1-2
3-4
5-6
7-8
9-10
Y
100 Ti/T0
1-2
3-4
5-6
7-8
To
5.43
4.97
4.85
5.42
5.98
5.30
100
5.00
4.60
3.90
4.25
Tl5 T30 T60
3.85 3.35 4.55
3.72 3.02 3.95
3.62 3.10 4.00
3.72 3.15 4.25
3.60 3.45 4.95
3.70 3.21 4.34
70 61 82
23 SEPTEMBER
3.90
3.70 Samples lost
3.15
3.75
T90
2.55
2.15
2.30
2.50
2.85
2.47
47
4.44 3.62
100 Ti/T0 100 82
44
-------
Table 17. Dissolved organic carbon (mg/liter) in Cub Lake water
30 July 1970 and after incubation
Sample
depth, m
1-2
3-4
T
8.70
8.20
T15
6.25
6.08
T30
6.50
6.30
T&0
7.05
7.25
T90
4.25
4.40
Y 8.45 6.16 6.40 7.15 4.32
100 T-/T 100 73 76 85 51
45
-------
Table 18. Dissolved organic carbon (ing/liter) in Green Lake water
11 August and 3 September 1970 and after incubation
11 AUGUST
Sample
depth, m
0
1-2
3-4
5
TO
1.30
2.50
2.30
1.85
T15
-
1.15
0.30
0.55
T30
-
2.42
0.98
1.45
T60
-
2.30
0.95
1.70
T90
-
1.70
0.0
1.00
2.22 0.67 1.62 1.65 0.90
100 T-/T 100 30 73 74 40
3 SEPTEMBER
1-2 2.32
3-4 1.84 Samples lost
5-6 1.20
Y 1,80
46
-------
Table 19. Dissolved organic carbon (mg/liter) in Blue Lake water
11 August and 3 September 1970 and after incubation
2.80
11 AUGUST
Sample
depth, m
0
1-2
3-4
5-6
7
7.5
Y
100 Ti/T0
1-2
3-4
5-6
7-8
To T15 T30
4.60
4.40 4.35 3.85
3.25 1.90 3.12
3.30 1.40 2.70
2.85 0.90 2.12
2.65 3.20 3.60
3.29 2.35 3.08
100 71 94
3 SEPTEMBER
3.60
3.32 Samples lost
1.42
2.87
T60 T90
-
4.40 5.00
3.95 3.70
2.95 2.55
1.90 2.50
2.20 1.75
3.08 3.10
94 94
47
-------
Table 20. Net seston carbon (mg/liter) Horsetooth Reservoir
(Station A) two dates 1970
Sample
depth, m
1-2
3-4
5-6
7-8
9-10
11-13
14-16
17-19
20-22
23-25
26-29
30-33
34-37
38-41
42-45
Station A
15 July
TO T60
.0176 .033
.0125
.0158
.039
.0225
.0125 .018
.0159 .021
.0093
.0149
.0060 .018
.0073
.0047 .015
.0087
.012
.0049 .016
13 AugUi
To
.0165
.0167
.0171
.0237
.0303
.0233
.0154
.0178
.0064
.0048
.0039
.0039
.0020
.0069
.0069
.0117
.022
.0130
100 Ti/Tc
100
188
48
-------
Table 21. Net seston carbon (mg/liter) Horsetooth Reservoir (Station A)
9 September 1970 and after incubation
Sample
depth, m
1-2
3-4
5-6
7-8
9-10
11-13
14-16
17-19
20-22
23-25
26-29
30-33
34-37
38-41
To T15
.0270
.0216
.0154
.0120
.0272
.0297
.0165
.0140
.0120
.0137
.0090
.0097
.0130
.0082
T30
.0208
.0212
.0207
.0129
.0185
.0244
.0174
.0134
.0132
.0118
.0065
.0064
.0105
.0032
T60
.034
.034
.026
.029
.024
.025
.030
.044
.025
.023
.015
.002
.013
.041
T90
.019
.019
.032
-
.014
.017
.042
.019
.014
-
.001
.002
.008
.011
.0164
.0144
.026
.015
100
100
88
162
94
49
-------
Table 22. Net seston carbon (mg/liter) Horsetooth Reservoir (Station A)
17 October 1970 and after incubation
Sample
depth, m
1-2
3-4
5-6
7-8
9-10
11-13
14-16
17-19
20-22
23-25
26-29
30-33
34-37
TO
.157
.091
-
.141
.118
.068
.059
.062
.055
.062
.035
.029
-
*
.096
.067
.074
.062
.045
.033
.029
.040
.040
.034
.016
.015
_
JO T60 T90*
.0731
.0491
.0633
.0729
.0446
.0436
.0375
.0380
.0402
.0367
.0200
.0219
.0263
.080
.046
.0436
100
100
58
55
50
-------
Table 23. Net seston carbon (mg/liter) in three lakes
BEAR LAKE - 30 JULY 1970
Sample
depth, m
1-2
3-4
5-6
7-8
9-10
Y
100 T-^/T
1-2
3-4
5-6
7-8
1-2
3-4
5-6
7
Y
100 T./T0
TO
.0764
.0461
.0456
.0643
.0415
.0578
100
.0066
.0167
.0098
.0186
.045
.045
.048
.059
.049
100
T15 T30
.0050
.0059
.0046
.0052
.0045
.0050
9
BEAR LAKE - 23 SEPTEMBER
-
BLUE LAKE - 3 SEPTEMBER
.046
.051
.048
T60
-
-
.0039
.0025
.0039
.0031
.0034
7
1-2
3-4
5-6
GREEN LAKE - 3 SEPTEMBER
.0261 .021
.0228 .011
.0393 .051
100
.0294
100
.028
.0023
.0023
.0023
9
L90
51
-------
Table 24. Dissolved organic carbon, net seston carbon and dry weight
of net seston, Bear Lake 23 September 1970
Sample
depth, m
1-2
3-4
5-6
7-8
DOC*
mg/liter
5.00
4.60
3.90
4.25
NSC**
mg/liter
.0066
.0167
.0098
.0186
dry weight**
mg/liter
.26
.41
.12
.40
lOOc/
dry wt
2.5
4.1
8.2
4.6
* mean of two determinations/depth
** " " three " "
52
-------
Table 25. Carbon content and dry weight of net seston, Dixon
Reservoir. All samples from 1 meter depth
n
SE
95% limits
error
CV
29 SEPTEMBER
dry weight
mg/liter
6
2.23
.3143
.561
.229
1.64-2.82
10
25
.3520 mg/liter
.0303
.174
.078
.1360-.4880
22
50
* .0293 mg carbon
12 OCTOBER
s
SE
95% limits
error
CV
4.05
.1914
.437
.1545
3.69-4.41
4
10
96.33 % absorbance*
1.26
1.123
.4586
95.15-97.51
.5
1
53
-------
Table 26. Carbon content and dry weight of net seston, Horsetooth
Reservoir (Station A) 17 October 1970
Sample
depth, m
1-2
3-4
5-6
7-8
9-10
11-13
14-16
17-19
20-22
23-25
26-29
30-33
34-37
dry weight
mg/liter
1.20
1.10
1.20
1.60
1.20
0.80
0.80
0.67
0.53
0.60
0.35
0.05
0.30
carbon
mg/liter
.157
.091
-
.141
.118
.068
.059
.062
.055
.062
.035
.029
0.80 .080
54
-------
SECTION VI
FIELD AND LABORATORY RESULTS AFTER 1 JANUARY 1971
Introduction - In view of fluctuations in carbon concentrations both in
natural waters and in samples incubated in the laboratory, it was deemed
desirable to narrow the sampling scope and to increase frequency of
sampling. Field work was concentrated on, but not confined to, two
small bodies of water, Bear Lake and Dixon Reservoir (Table 27) and
on two small streams.
Bear Lake was sampled at approximately weekly intervals. Water was
taken from each meter of depth from the ice surface to maximum depth.
Net seston was strained from 500 mis of water and variable but known
volumes, usually 100 to 300 mis., of strained water were filtered for
filter seston carbon. DOC was determined from 5 ml samples.
Dixon Reservoir was sampled at intervals ranging from 1 to 7 days.
All samples were taken at a depth of 1 m.;net seston, filter and dis-
solved carbon values were determined.
Two small streams, one, Spring Creek, had already received some sampling
prior to 1 January. Glacier Creek rises in tarns and snowfields at
high altitudes in Rocky Mountain National Park and in addition to other
sources received the small outflow from Bear Lake. Glacier Creek was
sampled at four stations, ranging in altitude from 9220 feet down to 8000.
Laboratory procedures remained as before except that a filter fraction
was added. This was made possible by acquisition of an electric oven
that could be controlled so that glass fiber filters could be preheated
just below the melting point. Filters were preheated for 2 hours.
Experiments using blank, preheated filters revealed that they caused
absorbency when treated and processed through the infrared analyzer;
however, dispersion among replicates was small. Thus, blank filters
were run with each batch of filters containing seston. An average
percentage of absorbency for blanks was figured and this amount sub-
tracted from the percentage of absorbency of each seston filter prior
to conversion to carbon value.
Three incubation experiments were conducted. In the first, a 10-gallon
polyethylene carboy was filled with water and included seston from
Dixon Reservoir. In the laboratory, the carboy was covered by a large
cardboard box to exclude light. The carboy was sampled for NSC, FSC,
and DOC over a three-week period. Adequate stirring of the carboy prior
to sub-sampling created problems, thus another incubation series was run.
This time a number of 1-liter polyethylene bottles were filled with
Dixon Reservoir water and seston. These were stored in the dark in the
laboratory and sampled over a three-week period. Each bottle could be
55
-------
thoroughly mixed prior to sub-sampling and the contents of each were used
completely. Despite our precautions, it is probable that the bottles re-
ceived slightly unequal amounts of seston during filling at the Reservoir.
Because the presence of live large zooplankters could have affected the
previous incubation experiments, a new one was initiated on 29 April.
Using a Kemmerer water bottle, 30 1 of water was obtained from a depth
of 1 m in Dixon Reservoir. In the laboratory, the container filled with
lake water was submerged in a water bath at 80C and held there for 30
minutes after the lake water had reached 80C. This was done to kill crus-
taceans and other large zooplankters.
After heating, the lake water was thoroughly mixed and poured into a
number of one liter polyethylene bottles and allowed to cool to room
temperature; approximately 2-3 mis of raw unheated lake water were added
to provide bacterial seeding.
The bottles were tightly capped and stored in a dark box at room tempera-
ture.
On nine occasions between 29 April and 24 May, two bottles were removed
from the dark box and their contents analyzed for dissolved, filter
and net fractions of organic carbon. Each bottle was handled separately,
thus giving duplicate samples.
Bear Lake - DOC ranged from about 2.5 to 10 mg/liter over nine sampling
dates (Figure 12). Most values were between 3.5 and 5.5 mg C/liter,
however. Values for 23 January are so much greater than on most other
dates that error is suggested. DOC tended to be concentrated in the
deeper waters. No overall increase or decrease with time is immediately
evident in Figure 12.
NSC in Bear Lake did not exhibit clear-cut trends in changeover time
nor were there consistent patterns of depth distribution (Figure 13),
although the maximum value for any sampling tended not to be found in
the deepest water. Noticeable fractions of the net seston were the
copepod Diaptomus nudus and the cladoceran Daphnia rosa. These animals
are able to adjust their positions in the water column in response to
a number of changing environmental factors including light intensity.
Vertical migrations would tend to confound distributional patterns in
net seston.
In contrast to NSC, FSC tended to accumulate in deep water, just above
the lake bottom (Figure 14). Observations of previous years (Reed 1970
and unpublished) have revealed the accumulation of autochthonous detritus
in the deep water of Bear Lake during winter stagnation and summer strati-
fication. Ratios of NSC : FSC : DOC were approximately .02 : .20 : 1.0;
i.e. DOC was 50 times greater than NSC, but only 5 times FSC.
56
-------
During the period of these observations, Bear Lake was covered by 65-80 cm
of ice; snow depths ranged from 0 to 75 cm.
Dixon Reservoir - Between 9 February and 24 March, DOC varied from 5.82
to 10.70 mg/liter and averaged 6.71. FSC averaged .4933 mg/liter and
ranged from .2900 to .9438 mg. The range and mean for NSC were .0487
to .2250 and .1187 mg/liter. Mean DOC value was 13 times that of FSC
and 60 times greater than mean NSC; FSC was about 5 times larger than
NSC. All fractions fluctuated greatly and without clear patterns
(Figure 15). Without direct microscopic examination, one cannot say
whether an increase in particulate fraction was the result of synthetic
or destructive processes; i.e. a pulse of small plankters could swell
FSC values as well as detrital particles resulting from die-off and
decay of a plankter population. Moreover, the relationships among
NSC, FSC and DOC are more likely to result from some past condition
rather than to be the result of conditions obtaining at the moment of
sampling. On two occasions, decrease in DOC was contemporary with
increase in particulate carbon. Forsberg (1967) attributed simultaneous
decrease in dissolved carbon with increased chlorophyll to uptake by
phytoplankters.
Spring Creek - DOC in Spring Creek ranged from about 2 to 8 mg/liter
(Figure 16). Spring Creek winds along the edge of Fort Collins and
receives agricultural and suburban drainage, in addition to roadside
ditches. Correlation of changes in DOC with respect to precipitation
has not yet been examined; perhaps peaks may have resulted from inputs
of allochthonous carbon. In any event, the sampling stations did not
retain relative rankings, although stations 4 and 6, those furthermost
downstream, often had greatest carbon values.
Glacier Creek - DOC in Glacier Creek also fluctuated between stations
and over time (Figure 17), although not as widely as in Spring Creek.
On 29 March, stations 3 and 4 were obviously being influenced by local
snowmelt and runoff. At higher elevations, the snow was not yet melting.
Glacier Creek in winter is smaller than Spring Creek and certainly seems
more thoroughly isolated from its surroundings by ice and show than does
Spring Creek.
Incubation Studies - Results of the first incubation run, Dixon carboy,
are shown in Figure 18. DOC increased as FSC decreased and then declined
irregularly to about the initial value. After an increase following
about a week of incubation, FSC remained below initial value, although
fluctuating. NSC exhibited an enormous increase after a week of incu-
bation; this value is so extreme relative to others that the possibility
of error is very difficult to dismiss. Live Daphnia and copepods were
observed in the carboy at the close of the experiment, thus the one
high value of NSC could have resulted from a chance concentration of
large zooplankters in the subsamples. There is no clear indication
that amount of organic carbon in the carboy decreased appreciably during
incubation. Possible destruction of carbon may have been offset by
57
-------
heterotrophic C0_ fixation and it is conceivable that sufficient light
reached the carboy to permit some photosynthesis, although it would
surely have been carried on at a very low light level.
The last incubation series, Dixon bottle, (Figure 19) is no less
bewildering than Dixon carboy. Again, there is no indication that there
was less total carbon at the conclusion of incubation than at the start;
in fact, NSC and FSC appear to have increased. In this experiment,
photosynthesis can be ruled out; the bottles were stored in complete
darkness. Live zooplankters were, however, still present in the final
bottle analyzed.
A number of unknowns come to mind: carbon fixation by heterotrophic
bacteria and facultative heterotrophic algae, "surface effect" which
would be largely relative to the volume of each liter bottle. If bacteria
grew profusely on the inner surfaces of the bottles, they could be dis-
lodged in mixing intended to suspend net seston and thus perhaps appear
in FSC. Dynamics of zooplankton metabolism probably also affected
different carbon fractions in unknown ways.
The results of the third incubation experiment are summarized in Table 28.
Prior to heating on 29 April, Dixon Reservoir water contained a total of
nearly 7 mg organic carbon/liter, of which 6.19 was DOC, .042 NSC, and .758
FSC.
Concentrations of fractions exhibited a decrease between 29 April and
3 May (Table 28). All increased in the next two days and all decreased
between 17 May and 19 May, otherwise, there were no common trends of
increase or decrease. Between 14 and 17 May, DOC and NSC increased
substantially, but FSC decreased slightly.
After the period of initial decrease and recovery, i.e. 10 May through 24,
DOC fluctuated little, about a mean value of 5.25 mg/1. NSC was more
variable, about a mean concentration of .036 mg/1, but no clear trends
are evident. Certainly the data do not suggest a progressive loss of
carbon from large particles. FSC varied about a mean value of .609 mg/1
and, as was the case with NSC, did not exhibit a clear trend toward diminu-
tion as incubation proceeded.
The data are few and possible uneven distribution of particulate carbon
among the liter bottles and problems of analysis of filter seston
add to the difficulties of drawing firm conclusions from this experiment.
If the bottles can be regarded as closed systems, i.e. the caps were
tight enough to permit no passage of gaseous carbon dioxide and no ex-
changes of carbon between bottle walls and contents occurred. The organic
carbon behaved in a very conservative fashion. The initial amount of
organic carbon was 6.99 mg/1., assuming the final value was 5.94; then
about 1 mg of organic carbon was lost per liter of water. The rate of
loss was about 0.04 mg/1 day.
58
-------
Considering only the first post-heating sample, about 2 mg organic carbon
were lost per liter of reservoir water; most of which was recovered by
the time of the second post-heating sampling. If the period between
heating and 5 May is regarded as a time of readjustment and stabilization,
one may compare the total organic carbon present on that date (6.66) with
that at the conclusion of the experiment (5.94) to arrive at a loss of
0.7 mg/1 and .037 mg/1, day.
The presence of large, live zooplankters seems not to have confounded
this experiment as it might have in earlier incubation attempts. Warming
to 80C was lethal to cladocerans and copepods in the water. Possibly
eggs, nauplii or other immature forms could have been introduced to some
bottles at the time raw, unheated Dixon Reservoir water was added; no
live cladocerans or copepods were observed, however. Live zooplankters
are not regarded as having influenced the outcome of this experiment in
any way.
Possible roles of viable small zooplankters are not so easy to assess.
Conceivably some rotifers or protozoans might have survived the heating
and certainly could have been introduced unnoticed with the raw water.
Bacteria are, of course, the most significant unknown.
A major assumption of this procedure was that bacteria would either sur-
vive the heating or be introduced into the bottles in the unheated water.
Should either failed to have happened and the bottles resulted in a
series of sterile environments, changes in carbon could be regarded as
fluctuations due to error in measuring a constant amount of carbon - a
possible but highly unlikely situation.
Regarding each bottle as a closed system, it may conceptually be dia-
grammed as having two carbon components, one inorganic and one organic.
carbon dioxide dissolved
bicarbonate
carbonate particulate
Inorganic Organic
carbon carbon
Within each box there may be exchanges between forms and states, e.g.
particulate carbon may become dissolved, or bicarbonate give up CO , etc.
There are, of course, exchanges between boxes. In the case of this ex-
periment, there appears to have been a net loss from the organic carbon
component to presumably the inorganic box. In retrospect it is clear that
pH, CO and alkalinity should have been monitored to ascertain if the in-
organic fraction did increase.
59
-------
Loss from the organic box could not be clearly ascribed to either
dissolved or particulate forms. Possibly as bacteria reduced the
carcasses of zooplankters and phytoplankters and bits of debris con-
stituting the particulate matter, their (bacteria) biomass increases
tended to offset decreases of the substrate. Both bacteria and
detritus could have been added to dissolved carbon. CO. produced in
bacterial respiration could be one route from organic to inorganic box.
Bacterial fixations of carbon dioxide is one possible route from inorganic
to organic box. Further speculation about a dynamic system without
further data seems fruitless. But within the conditions of this experi-
ment, particulate organic fractions did not exhibit drastic decreases.
60
-------
Table 27. Physical characteristics of waters sampled
Water Body
Green Lake
Blue Lake
Bear Lake
Cub Lake
Horsetooth
Reservoir
Dixon
Reservoir
Glacier Creek
Spring Creek
Altitude Area Max. Depth
m ha m
3,520 1.4 7.0
3,400 1.4 8.3
2,888 4.8 11.0
2,629 3.5* 4.0
1,656 755.0 62.0
1,586 18.5 2.5
2,806-
2,562-
1,586-
1,500
Mean Depth
m
3.2
4.5
3.9
2.0*
24.0
1.8*
Remarks
Cirque lake
Rock basin lake
Moraine dam lake
Brown water lake
Fluctuates, data
are maxima
Fluctuates, data
are maxima
Mountain brook,
stony bed, rapid
flow
Plains brook, sandy
gravelly bed, slow
flow
1. Reed 1970
2. Nelson 1970
* estimated
-------
Table 28. Concentrations of dissolved (DOC), net seston (NSC) and
filter seston carbon (FSC) in Dixon Reservoir water during incubation
after heating to 80C. Values for 29 April were obtained prior to heat-
ing. All values are in mg C/liter and most are means of duplicate
samples
Sampling
Date
29 Apr
3 May
5
10
12
14
17
19
21
24
DOC
6.19
4.46
5.98
-
5.18
5.05
5.90
5.12
5.05
5.20
NSC
.042
.030
.043
.028
.027
.032
.052*
.038
.033
.039
FSC
.758
.524
.632
-
.520*
.652*
.640*
.532
.700*
_
*single determination
62
-------
Jan 23
o
i
2
3
4
5
6
7
8
9
10
Jan 30 Feb 17 Fib 23 Mar 9
Mar 13
Mar 20
Mar 27
0
1
2
3
4
5
6
7
8
10
10 mg
Figure 12. Vertical distribution of dissolved organic carbon in Bear Lake,
-------
Jan 9 Jan 16
Jan 23 Jan 30 Feb 17 Feb 23 Feb 27
^J
\
I
3
O
A
e;
P,
\J
~7
1
fl
Q
10
0
1
7
10
WAR 9
M,
AR 13
Ğ
MAR 20
fc
j
MAR
20
4
.lOmg
Figure 13. Vertical distribution of net seston carbon in Bear Lake,
-------
Jan. 30
Febl7 Feb 23
MAR 4
MAR 9
1
2
3
4
5
6
7
8
9
0
Ui
MAR 13
MAR 20
MAR 27
0
1
2
3
4
5
6
7
8
9
10
i
1
1
1
1
1
I.Omg
Figure 14. Vertical distribution of filter seston carbon in Bear Lake.
-------
.20
NSC
.10
9 11
17
February
24 26
2 4
10 16 18
March
24
.80
-.70
FSC
-.60
-.50
10.5
9.5
DOC
-8.5
h7.5
Figure 15. Dissolved organic carbon (triangles), filter seston carbon (open circles) and
net seston carbon in Dixon Reservoir.
-------
80
7.0
6.0
K O
\J. w
4.0
3.0
20
1.0
-
-
j
I
2
Fe
"
3
b
4
9
6
i
i
i
1
2
FĞ
3
b
4
II
-i
6
1
2
Fe
-
3
b
4
n
Ğ
. ,
i
2
Ft
b
2*
i
I
f
-t
M
ft
21
-^
ğ
-
Ml
ir
2
-i
i
2
Ml
3
ir
IH
4
4
e
i
I
2
Ma
3
r
4
1C
6
(
-
1
1
2
Ma
-
3
r
4
1C
6
p
1
1
2
ida
3
r
4
C
i
6
1
P-
1
ft
2
Ao
-
3
r
r-
4
2'
i i
6
I
Figure 16. Dissolved organic carbon, four stations, Spring Creek.
-------
oo
6.0
5.0
4.0
3.0
2.0
1.0
234
1234 1234
1234 1234 1234
234
Febl7 Feb 23 Feb 27 MAR 9 MAR 13 MAR 20 MAR 27
Figure 17. Dissolved organic carbon, four stations, Glacier Creek.
-------
.5
NSC
.4-
2-
.10 12
February
15 16 17
19
DOC
\-7
22
Figure 18. Changes in amounts of dissolved organic carbon (triangles), filter seston
carbon (open circles) and net seston carbon from Dixon Reservoir during
incubation (carboy experiment).
-150
.25
-I.OO
FSC
-.75
-5O
-------
.05-
22 24 26
February
17 19'
Figure 19. Changes in amounts of dissolved organic carbon (triangles), filter seston carbon
(open circles') and net seston carbon from Dixon Reservoir during incubation
(bottle experiment).
-------
SECTION VII
DISCUSSION
Strickland and Parsons(1968) pointed out that while some theoretical
doubt may exist, it is difficult to believe that the method of Menzel
and Vaccaro fails to fully oxidize all organic compounds dissolved in
sea water. Presumably the same is true for fresh water. Strickland
and Parsons stated that the method is capable of detecting organic carbon
in amounts as small as 0.09 mg/liter. As used in this study, the lower
limit of detection appears to be about 1 mg C/liter or .005 mg C/5 ml
ampoule, if the machine is calibrated so that 10 mg C/liter standard
produced full scale deflection. In the case of NSC, if the carbonaceous
material were strained from 12 liters of lake water, carbon was detected
at the level of about .004 mg/liter in nature.
Strickland and Parsons suggested calibrating the analyzer so that 2 mg
C/liter produced full scale deflection, thus giving a more sensitive
setting than 10 mg C/liter standards. Calibration with 2.5 and 5 mg
C/liter standards magnified noise and increased variability to unde-
sirable levels.
Usually, a number of replicate samples analyzed at one time had an error
(100 s / Y) of 10% or less. Thus, the errors associated with actual
analysis were no greater than those associated with drawing the samples
in nature.
One of the important pioneer studies of organic substances in fresh
water was that of Birge and Juday (1934). They distinguished between
particulate and dissolved forms of organic matter. Because they removed
the particulate fraction by centrifugation, their findings cannot be
compared with the net seston of this study. Their dissolved organic
fraction should be roughly comparable to ours. The Wisconsin limnologists
measured organic carbon in the surface waters of 529 lakes and found
values ranging from 1.15 to 28.5 mg/liter. This range easily embraces
the values reported here.
Using the method of Menzel and Vaccaro, Forsberg (1967) studied dis-
solved organic carbon in eight lakes of Uppland, Sweden. Forsberg
designated some lakes as Chara lakes and others as Potamogeton lakes.
In general, Chara lakes contained water showing from 10 to 55 mg color/
liter (platinum - cobalt scale); whereas Potamogeton lakes had from
65 to 85 mg color/liter. Carbon values ranged from 5 to 14 mg/liter in
Chara lakes and 11 to 14 mg/liter in Potamogeton lakes. The correla-
tion of carbon with color was r = 0.872. The dominant macrophyte
in Cub Lake is the pond lily, Nuphar polysepalum; the color of the water
is, however, probably similar to that of the lower range of Forsberg's
Chara lakes. Thus, the DOC values for Cub Lake, the lake in this survey
with the most heavily stained water, are similar to values for brown-
water Swedish lakes.
71
-------
Forsberg made successive determinations of DOC in three lakes over a
three-week period. Between 15 and 31 May, values of DOC in Ekoln varied
from about 11 to 14 mg/liter. Greatest difference between two consecu-
tive dates was 3 mg/liter. DOC values in Langsjon ranged from 5.2 to
6.8 mg/liter over the period 13 to 31 May. One mg/liter was the largest
variation between successive readings. Concentrations of DOC ranged
from 10.7 to 12.2 mg/liter in Trehorningen, a lake with much greater
humus staining than Cub Lake. No common patterns to variation in DOC
readings in the Swedish lakes were evident, although Forsberg noted
that DOC and chlorophyll exhibited opposite tendencies, i.e. when DOC
decreased, chlorophyll increased, supposedly indicating uptake of DOC
by phytoplankton.
In a somewhat similar vein, Goldman and Armstrong (1969) apparently
drew a number of subsamples at 24-hour intervals from lake and river
probes for assay of C-14 uptake. Considerable variation among subsamples
was found, but not discussed by these workers. Biologists have not
been quick to elucidate short term changes in biological parameters of
aquatic ecosystems, or for that matter, to document well their occurrences.
Concentrations of DOC in both Spring Creek and the Cache la Poudre
River were roughly similar to values reported by Weber and Moore (1967)
from the Little Miami River at Cincinnati. These investigators used
a CO. infrared analysis method that differed from the one used in this
study.~ DOC in the Little Miami ranged from 2.5 to 12.5 mg/liter and
averaged 6.4. Fluctuations of 5 mg/liter occurred between adjacent
weekly samplings. They concluded that DOC originated from several inde-
pendent sources and thus could not be correlated with single parameters
such as river discharge or standing crop of algae.
Some brief remarks regarding DOC and NSC in relation to other limnological
variables can be made. In the summer of 1968, mean standing crops of
seston in Green Lake were 3.51 g/m , Bear Lake 6.12 and Blue Lake 21.30
g/m (Reed 1970). Stimpfl (1966) reported mean annual seston crops of
1.9 mg dry weight/liter and net seston crops of 1.969/m for Horsetooth
Reservoir, thus the lakes had the same relative rankings in dry weight
of seston and DOC. In regard to NSC, Green, Blue and Bear lakes showed
the same ranking as for DOC; Horsetooth by its variation and by being
more frequently sampled could be placed on either end of the scale.
Vertical distribution of NSC and DOC in the lakes is difficult to sum-
marize briefly. In general, maxima of both DOC and NSC at any one
sampling tended to be found either in the upper layers or in the lower
layers, but not at mid depth. Ohle (1968) reported that in nutrient
rich lakes about 90% of the photosynthetic products of the epilimnion
are broken down in the metalimnion. He further noted that destruction
proceeds more slowly in the hypolimnion, but eventually increases at the
mud-water interface. Overbeck (1968) found positive correlation between
phytoplankton and bacteria maxima in lakes of east Holstein. He stated
that heterotrophic bacteria depended on the phytoplankton for organic
72
-------
substances, although bacteria and phytoplankton could exhibit different
distributions in the water column. Maxima might occur together or
bacterial maximum could occur in the hypolimnion simultaneously with a
bloom of phytoplankton in the epilimnion.
Fonden (1969) noted that bacterial maxima tended to occur in or near
the metalimnion in lake stations massively influenced by inflowing
rivers. Away from inflow, maximum numbers of bacteria tended to be
found in surface layers, or in any event, not in the hypolimnion. This
is, of course, in agreement with Ohle's observations. Thus, apparent
accumulation of carbon in deep water may not necessarily indicate the
presence of large bacterial populations.
Nothing is known of bacterial populations associated with either DOC
or NSC samples. It may be pointed out, however, that the simultaneous
occurrence of phytoplankton and bacteria maxima does not necessarily
imply a causal relationship. A bacterial maximum could be supported by
a previous phytoplankton bloom at the time that another phytoplankton
bloom developed. This could explain bacteria maximum in deep water
occurring simultaneously with a pulse of phytoplankton in the epilimnion.
Nelson (1970) found that very little primary production, measured by
C-14 uptake, occurred at depths greater than 6 m in Horsetooth Reservoir.
Determinations of both DOC and NSC tended to agree with Nelson's findings.
Support, or at least agreement, with Ohle's statement about retarded
rates of decomposition in the hypolimnion may also be found in the Horse-
tooth results in that particulate organic matter easily attacked by
bacteria could have been removed in the epilimnion, thus, accounting
for lower concentrations of NSC below the metalimnion. However, the
situation is undoubtedly more complex.
The nature of the particulate matter changed with depth in Horsetooth
(and probably in most of the other lakes as well). Net seston samples
taken below the metalimnion tended to be dominated by Daphnia and other
relatively large zooplankters; whereas, epilimnion samples contained a
preponderence of phytoplankters.
There seems to be no clear evidence from this study that deep water
net seston in Horsetooth was more resistant to breakdown during incu-
bation than was that from shallower depths. There is a weak suggestion
that deep water net seston from Green and Blue lakes lost more carbon
during incubation than did samples from shallower water. Results of
incubation studies are inconclusive, but some trends are suggested.
From 67 to 82% of the initial DOC values remained in Horsetooth water
after incubation for 90 days. In one incubation series, 94% of the
initial value of DOC was found in Blue Lake water after 90 days. In
Bear, Green, and Cub Lakes, concentrations of DOC at the end of incu-
bation were about 50% less than the initial concentration.
Values of NSC were more variable, but NSC in Horsetooth was apparently
more resistant than that of any other body of water - 55 to 188% of
73
-------
mean initial values was found after incubation. Readings of less than
10% of initial NSC were obtained for Bear, Blue, and Green Lakes.
No direct observations were made of bacteria, hence any remarks con-
cerning them are completely speculative. However, one may assume that
the organic matter, both particulate and dissolved, taken from these
waters was complex and diverse in nature. Undoubtedly, a variety of
bacteria were brought into the laboratory. It is not unreasonable to
suppose that some were better suited to laboratory conditions than others.
Vaccaro (1969) found that no single kinetic uptake pattern characterized
the response of marine microplankton organisms in solutions enriched
with specific organic substances. In natural waters a broad spectrum
of organic substances is attacked by diverse assemblages of hetero-
trophic organisms. Vaccaro further pointed out that while results of
short term (3 hour) uptake studies of heterotrophs in enriched media
are unpredictable and erratic, irregular uptake patterns do tend to
stabilize over time. This may be explained in part by the development
of heterotrophic populations which are able to attack the specific
substrate in question. Once the heterotroph population has become
"adjusted" to the substrate (and in so doing changes from a mixed to
essentially pure culture?) uptake follows predictable kinetic patterns.
The gain in carbon by samples during incubation requires comment.
Because the gain was seen in many samples from different bodies of
water and because gain was noted in series incubated and analyzed at
different times, it cannot be dismissed simply as error. The following
tentative explanation is suggested. Samples were incubated in darkness
and covered to prevent entry of dust, but the aluminum caps over ampoules
and reagent bottles were not airtight, thus CO^ from the atmosphere could
enter. Heterotrophic bacteria were able to attack organic matter con-
tained in the samples for nutrients and at the same time fix carbon
from the atmosphere. Ampoules and reagent bottles provided a large
surface relative to the volume. The "surface effect" is well recognized
in C-14 uptake studies and in BOD determinations; in essence, the glass
surface provides a stable substrate upon which bacteria can attach and
multiply.
Ultimately, the bacteria exhaust some nutrient to a limiting level or
perhaps their metabolic products reach intolerable levels, but in any
event, the population cannot grow continuously in a limited environment.
A rapid decline ensues and the bacteria population either dies out or
fluctuates about some low level.
Probably some sorting out of species also occurs and any given sample
becomes dominated by one or a few species whose metabolic processes
are complimentary. If a sample is analyzed when a large bacterial popu-
lation is fixing atmospheric carbon, carbon value in excess of the ini-
tial content is possible.
Kuznetsov (1968) emphasized the importance of heterotrophic bacteria
in production dynamics of certain European lakes and reservoirs.
74
-------
Heterotrophic assimilation of carbonic acid accounted for 24% of total
organic matter produced in Lake Onega and in Rybinsk Reservoir produc-
tion of bacterial biomass exceeded photoplankton biomass production
in 1964. Rybinsk Reservoir apparently receives massive amounts of
allochthonous organic detritus. The detritus is broken down by bac-
teria, releasing carbon dioxide which is utilized by heterotrophic
bacteria and photoplankters.
The major thesis, that relative fractions of refractive and non-refractive
organic carbon, either particulate or in solution, in freshwaters might
offer clues as to trophic state remains neither strongly endorsed nor
denied by this investigation.
In the increased acuity of vision that comes with hindsight, the origi-
nal concepts of incubation studies appear naive, to say the least. The
results reported are not without value, although neat, clear-cut con-
clusions are not to be drawn. The incubation studies are, of course,
a variation on the classic BOD determinations, as seen from the view-
point of carbon rather than oxygen. Because of heterotrophic bacteria
(and algae?) the carbon story may be much more complicated than the
tale told by oxygen. Certainly, the results suggest that smooth decay
curves cannot describe the events occurring in the incubation studies.
In regard to natural waters, the findings indicate that organic carbon
concentrations, both dissolved and particulate, can and do fluctuate
markedly over short intervals of time. Flowing waters have long been
recognized to be more variable environments than lentic waters, thus, the
demonstration the DOC fluctuates with time in small streams can hardly
be considered astounding.
Few studies have followed changes of organic carbon for several weeks
in one body of water. The data thus have descriptive value, although
they cannot, at this time, be related satisfactorily to the body of
general limnological knowledge. One might suppose that carbon contained
in a population of plankters would eventually find its way from large
particulate chunks through fine particles to the dissolved state. Of
course, it is well recognized that healthy photoplankters release con-
siderable amounts of photosynthetically fixed carbon in their metabolic
processes. Inputs of carbon to the plankton community from littoral,
benthic and allochthonous sources offer opportunities for rapid changes
in organic carbon. Further complications arise from exchanges between
inorganic and organic fractions of carbon.
In view of its possible role as a limiting nutrient (Kerr, Paris and
Brockway 1970 ?) in some aquatic ecosystems, carbon deserves further
study. A number of papers presented at the annual meeting of the Ameri-
can Society of Limnology and Oceanography, University of Rhode Island
(August 1970) and at the Symposium on Nutrients and Eutrophication,
sponsored by the same society at W. K. Kellogg Biological Station (Febru-
ary 1971) referred to organic carbon. Reported values ranged over at
75
-------
least two orders of magnitude. Seemingly much work simply describing
instantaneous amounts of organic carbon remains to be done.
76
-------
SECTION VIII
ACKNOWLEDGMENTS
The support of the project by the Water Quality Office, Environmental
Protection Agency and the help provided by Dr. L. P. Seyb, the Grant
Project Officer, is acknowledged with sincere thanks.
77
-------
SECTION IX
REFERENCES
Ahl T. 1966. Chemical conditions in Osbysjon, Djursholm. Oikos,
17:33-61.
Birge, E. and C. Juday. 1934. Particulate and dissolved organic
matter in inland lakes. Ecol. Monogr. 4:440-474.
Duursma, E., 1960. Dissolved organic carbon, nitrogen, and phosphorus
in the sea. Neth. J. Sea Res. 1:1-141. (not seen, quoted in Weber
and Moore)
Fonden, R. 1969. Heterotrophic bacteria in Lake Malaren and Lake
Hjalmaren. Oikos 20:344-372.
Forsberg, C. 1967. Dissolved organic carbon in some lakes in Uppland,
Sweden. Oikos 18:210-216.
Goldman, C. and R. Armstrong. 1969. Primary productivity studies in
Lake Tahoe, California. Verh. Internat. Verein. Limnol. 17:49-71.
Kerr, P., D. Paris and D. Brockway. 1970(7). The interrelation of car-
bon and phosphorus in regulating heterotrophic and autotrophic popu-
lations and in aquatic ecosystems. Water Pollution Control Research
Series 16050 FGS 07/70.
Kuznetsov, S. 1968. Recent studies on the role of microorganisms in
the cycling of substances in lakes. Limnol. Oceanogr. 13:211-224.
Menzel, D. and R. Vaccaro. 1964. The measurement of dissolved organic
and particulate carbon in seawater. Limnol. Oceanogr. 9:138-142.
Nelson, W. 1970. Relation of Daphnia galeata mendotae population sta-
tistics to environmental variables in Horsetooth Reservoir, Colorado.
Ph.D. thesis, Colorado State University, 316 pp.
Ohle, W. 1968. Chemische and mikrobiologische Aspekte des biogenen
Stoffhaushaltes der Binnengewasser. Mitt. Internat. Verin. Limnol.
14:122-133.
Overbeck, J. 1968. Prinzipielles zum Vorkommen der Bakterien im See.
Mitt. Internat. Verein. Limnol. 14:134-144.
Pennak, R. 1945. Some aspects of the regional limnology of northern
Colorado. Univ. Colorado Studies. Series D 2:263-293.
Reed, D. and E. Reed. 1970. Estimates of seston crops by filtration
with glass fiber discs. J. Fish. Res. Board Canada 27:180-185.
79
-------
REFERENCES (Continued)
Reed, E. 1970. Summer seston crops in Colorado alpine and montane
lakes. Arch. Hydrobiol. 67:485-501.
Stimpfl, K. 1966. Changes in the quality of water impounded in Horse-
tooth Reservoir. M.S. thesis, Colorado State University, 66 pp.
Strickland, J. and T. Parsons. 1968. A practical handbook of seawater
analysis. Fish. Res. Board Canada Bull. 167.
Vaccaro, R. 1969. The response of natural microbial populations in sea-
water to organic enrichment. Limnol. Oceanogr. 14:726-735.
Vollenweider, R. 1969. A manual on methods for measuring primary pro-
duction in aquatic environments. Davis Co., Philadelphia. 213 pp.
Weber, D. and D. Moore. 1967- Phytoplankton, seston and dissolved or-
ganic carbon in the Little Miami River at Cincinnati, Ohio. Limnol.
Oceanogr. 12:311-318.
80
-------
1
Access/on Number
w
5
_ Subject Field & Group
05A,05C
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
Organization
Department of Zoology
Title
Dissolved and Particulate Organic Carbon in Some Colorado Waters
10
Authors)
Reed, Edward B.
16
Project Designation
EGA 16010
21
Note
22
Citation
23
Descriptors (Starred First)
*Lakes, carbon *Reservoirs, carbon //Nutrients, carbon
25
Identifiers (Starred First)
*Carbon, organic
27
Abstract
Organic carbon in a variety of lakes, reservoirs and streams was measured by
acid-persulfate digestion and infrared absorption.
Organic carbon was designated as net seston carbon, filter seston carbon or dissolved
based on filtering techniques.
Concentrations of dissolved carbon ranged over an order of magnitude; i.e. from about
1 to 11 g/m3; filter seston amounts varied from about p.2 to 1 g/m3; the net fraction of
organic carbon ranged from less than 0.1 to over 0.2 g/m3. In general bodies of water
judged to be meso- to eutrophic contained more organic carbon than those waters tending
toward oligotrophy.
Repeated sampling/ of selected bodies of water revealed that amounts of organic carbon,
either dissolved or particulate, fluctuated considerably over short periods of time; how-
ever the dissolved fraction probably varied less than particulate forms of carbon.
Samples of all carbon fractions were incubated at room temperature in darkness for
periods ranging up to 90 days. Concentrations of organic carbon varied erratically and
unpredictably during incubation.
Almost daily measurements of carbon in samples incubated over three week periods
also revealed erratic changes, with no clear reduction in total organic carbon. Meta-
bolic activities of heterotrophic bacteria (and algae?) probably complicated interpre-
tation of results.
Abstractor
Institution
WR:I02 (REV. JULY 1969)
WRSIC
SEND WITH COPY OF DOCUMENT, TO: WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON. D. C. 20240
OPO! 1 970389-930
------- |