U.S. ENVIRONMENTAL PROTECTION AGENCY
NATIONAL EUTROPHICATION SURVEY
WORKING PAPER SERIES
COMPONENTS CONTRIBUTING TO LIGHT
EXTINCTION IN NATURAL WATERS:
METHOD FOR ISOLATION
WORKING PAPER NO, 369
PACIFIC NORTHWEST ENVIRONMENTAL RESEARCH LABORATORY
An Associate Laboratory of the
NATIONAL ENVIRONMENTAL RESEARCH CENTER - CORVALLIS, OREGON
and
NATIONAL ENVIRONMENTAL RESEARCH CENTER - LAS VEGAS, NEVADA
•&GPO 697-O32
-------
COMPONENTS CONTRIBUTING TO LIGHT
EXTINCTION IN NATURAL WATERS:
METHOD FOR ISOLATION
WORKING PAPER NO, 369
NATIONAL EUTROPHICATION SURVEY
OFFICE OF RESEARCH AND DEVELOPMENT
U,S, ENVIRONMENTAL PROTECTION AGENCY
MftYl976
-------
COMPONENTS CONTRIBUTING TO LIGHT EXTINCTION IN
NATURAL WATERS: METHOD FOR ISOLATION
by
Jacob Verduin, Llewellyn R. Williams, and Victor W. Lambou
Water and Land Quality Branch, Monitoring Operations Division
Environmental Monitoring and Support Laboratory
Las Vegas, Nevada 89114
Working Paper No. 369
NATIONAL EUTROPHICATION SURVEY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
May 1976
-------
COMPONENTS CONTRIBUTING TO LIGHT EXTINCTION IN
NATURAL WATERS: METHOD FOR ISOLATION
The Lambert-Beer's Law relating light extinction to depth
(d) and concentration of light-quenching substances (c)
I = I0
can be expanded to the form
I = I0 e-
-------
total extinction value is about 0.7 m~l. Consequently, ksS + kp
+ ka = 0.7 m~l. The extinction due to suspensoids must be about
0.24 m~l, and the extinction due to water alone is 0.03 m~*. There-
fore, the value of kp must be (0.7 - 0.27) = 0.43 nr*. Apparently,
the brown stain is responsible for about 60% of the extinction of
this lake.
Furthermore, S can be divided into three components:
(1) Scn], the suspensoids associated with chlorophyll,
(2) S0, the suspensoids associated with non-chlorophyll-
related organic matter, and
(3) S-j, the inorganic suspensoids.
If one obtains dry weight of suspensoids, their carbon content,
and their chlorophyll content along with light extinction data, these
three components can be separated. A regression of carbon on chloro-
phyll will serve to establish a carbon:chlorophyll ratio (the slope
of the regression line), and the point at which it intersects the
ordinate will indicate the amount of carbon not associated with chloro-
phyll. However, the ratio of dry weightrcarbon in these suspensoids
must be known. The research of Antia et al. (1963) can be used to
derive a dry weight:carbon ratio of about 4 for diatom communities, and,
/
if S0 is derived primarily from such communities, it is perhaps reasonable
to apply the same conversion factor to that component.
Lorenzen (1968) has published carbon:chlorophyll regressions,
for southeast Pacific communities, which can be used to demonstrate
this separation. He studied an area of upwelling by deploying a
-------
drogue and following it for 5 days. Initially, the euphotic zone
(to 1% of surface light) extended to 33 m (k = 0.14 m"1), and the
chlorophyll concentration averaged 1.7 mg m~3 in the euphotic zone.
During the 5-day study, the euphotic zone shoaled to about 14 m
(k = 0.33 m ), and the average chlorophyll concentration increased
to 11 mg m~3. The carbon:chlorophyll regression had an average slope
of 40 mg carbon per milligram chlorophyll, and it intercepted the
ordinate at a value of 98 mg carbon per cubic meter. Consequently,
a phytoplankton dry weight on day 1 of about
Schl = 1.7 x 40 x 4 x ID'3 = 0.272 gm m"3
and a non-phytoplankton organic dry weight of about
S0 = 98 x 4 x 10~3 = 0.392 gm nT3
can be estimated. And so, the extinction due to phytoplankton would
be estimated as 0.272 x 0.12 = 0.033 m'1, and that attributable to
non-phytoplankton organic would be 0.392 x 0.12 = 0.047. If
ka = 0.03 m"1 is added, these three components account for an extinc-
tion of 0.11 m . The observed extinction on day 1 was 0.14 m~l.
Thus, it appears that the extinction due to inorganic particles plus
pigments dissolved in ocean water must be small, accounting for a
remaining extinction of only about 0.03 m"1.
The maximum chlorophyll concentration observed during the 5-day
study {11 mg m~3) represents phytoplankton dry weight amounting to
Schl = 11 x 40 x 4 x 10-3 = 1.76 gm m~3.
-------
The non-phytoplankton component estimate is as before (0.392 gm m~3)
because the carbon:chlorophyll regression spanned the 5-day study.
The extinction due to phytoplankton at this time would be
1.76 x 0.12 = 0.21 m"1, and the two components account for an
extinction of 0.21 + 0.047 - 0.257 nr1. Adding ka (0.03), we have
accounted for an extinction of 0.287 rrrl. The observed extinction
was 0.33 m~l, yielding a remainder of 0.043m~l attributable to the
combined influence of inorganic particles and dissolved pigments.
Figure lisa graph of extinction versus chlorophyll for
Lorenzen's drogue stations. (Lorenzen provided the data for this
graph in a personal communication; the data also appear in the
report of the Anton tfruun Cruise 15, 1966.) Lorenzen also
sampled several stations outside the upwelled region. Their
carbonrchlorophyll regression had a slope of 70, and the intercept
was at 93 mg m~3. He did not report extinction values for these
stations, but the non-phytoplankton carbon agrees closely with that
in the upwelled area, suggesting that this component was not
influenced importantly by the 5-day phytoplankton pulse.
Riley (1956) published a curvilinear regression equation
relating chlorophyll to extinction in the marine environment. It
has the form
k = 0.04 + 0.0088 C + 0.054 C2/3
where C = chlorophyll concentration in milligrams per cubic meter.
-------
0.4
0.3
0.2
0.1
8 10 12
Chi a. (mg m~3)
14 16 18 20
Figure 1. Graph of extinction versus average chlorophyll
concentrations, in the euphotic zone, for Lorenzen's drogue
stations. The carbon chlorophyll ratio averaged 40.5 and the
regression line intercepted the ordinate at 97.6 mg dhlorophylI
per cubic meter.
-------
Riley's equation indicates that the extinction not correlated
with chlorophyll is small (0.04 m~l), most of it representing ka
(0.03 m~l). It also indicates that the extinction per milligram
of chlorophyll is variable, having a high value at low chlorophyll
concentrations (0.06 nr mg'1, when chlorophyll is at 1 mg m~3) and
lower values at high chlorophyll concentrations (0.02 m^ mg~l, when
chlorophyll is 100 mg m~3). Such variation can be accounted for on
the basis of a changing carbon:chlorophyll ratio. For example, if
the ratio were 120 [a typical carbon:chlorophyll ratio for waters
with low chlorophyll concentrations (Lorenzen 1968)] then the
dry weight:chlorophyll ratio in a diatom community would be about
120 x 4 x 10"3 = 0.480 gm dry weight per milligram chlorophyll.
And, the extinction associated with chlorophyll would be 0.48 x
0.12 = 0.058 m^ tng-1, agreeing well with Riley's equation at low
chlorophyll concentrations, and with Small and Curl's (1968)
regression slopes in communities whose chlorophyll concentrations
were in the 0 to 3 mg nr3 range (see also Tyler 1975).
However, if the carbon:chlorophyll ratio is 30 [Lorenzen (1968)
and Antia et al. (1969) both found low carbon:chlorophyll ratios for
waters containing high chlorophyll concentrations], the dry weight
per milligram of chlorophyll would be 0.120 gm, and the extinction
Q.01M ? ,
associated with chlorophyll would be 0.14 m mg'1, in essential
agreement with Riley's equation at high chlorophyll concentrations.
-------
The preceding discussion has been confined to diatom communities
in which the dry weight:carbon ratio is assumed to be 4. However,
many freshwater communities are dominated by chlorophycean and
cyanophycean phytoplankton which do not have siliceous cell walls.
In such communities, the dry weight:carbon ratio is approximately 2
(Strickland et al. 1969), and the extinction associated with chlorophyll
would be 0.024 m? mg~l when the carbonrchlorophyll ratio is 100, and
0.007 m^ mg~l when this ratio is 30. Megard (1974) has observed a
carbon:chlorophyll ratio of approximately 100 in a Minnesota lake,
although a community having low chlorophyll levels per cubic meter
exhibited a ratio of 347. He estimated an extinction associated with
chlorophyll of 0.03 m2 mg~l in Shagawa Lake near Ely, Minnesota.
Figure 2 was prepared from data (Reid et al. 1975) representing
three lakes in Canada's experimental lake area. It shows the slope
of the extinction:chlorophyll regression in each lake representing
a value of about 0.011 m2 per milligram of chlorophyll. Each lake,
however, had a large component of extinction not correlated with
chlorophyll (1.3 nT1 in lake 304, 1.1 rrT1 in lake 227, and 0.75 nT1
in lake 226). Because dry weight of suspensoids and carbon:chlorophyll
ratios were not presented in the report (Reid et al. 1975) from which
Figure 1 was derived, we do not know how much of the extinction was
caused by S0, S-j, or by pigments dissolved in the water. The rela-
tively low value of extinction per milligram of chlorophyll may reflect
dominance of chlorophycean and cyanophycean genera, or it may reflect
-------
n
D
n
n.
20 40 60 80 100
Chi a^ (mg m~3)
120
140
Figure 2. Graph of extinction (k) versus chlorophyll a_ in three
Canadian lakes.
Key
• Lake 304
D Lake 227
• Lake 226
-------
a low carbon:chlorophyll ratio. Schindler (1971) observed declining
carbon:chlorophyll ratios after fertilization of lake 227. But,
there is'a third possibility; it is well known that the 'cyanophycean
genera, Anabaena, Microcyetis, and Aphanizomenon form colonies so
large that they are visible macroscopically as tiny globules in the
water. It seems likely that such large particles must exhibit a
ks value of less than 0.12 nr gnT*, and the extinction per unit
chlorophyll associated with such large particles would be propor-
tionately reduced. '
In Table 1, the average values suggested here for estimating
the various components of extinction are presented. We believe the
averages are fairly reliable. Individual samples will show large
,s
deviations, just as one finds a considerable scatter around regression
lines, but recognition of the fact that the extinction associated with
a unit of chlorophyll is a function of the carbon:chlorophyll ratio
will improve the estimation of other components. Studies in which
underwater radiometer data, suspensoid dry weight data, and carbon:
chlorophyll regressions are obtained simultaneously will serve to
refine the averages appearing in Table 1.
-------
10
Table 1. Average values facilitating the computation of various
components contributing to light extinction.
Component
Average Value
Schl (diatom-derived,
chlorophyll -related
organic matter)
(non-diatqm-derived,
chl orophyl 1 -rel ated
organic matter)
SQ (diatom-derived,
non-chlorophyl 1 -related
organic matter)
S0 (non-diatom-derived,
non-chl orophyl 1 -rel ated
organic matter)
Si (inorganic matter)
0.12 m2 gnr1
0.03 m'1
k - (ksS + ka) m-1
Carbon/chlorophyll x 4 x 10"3* gm m~3
1 v
Carbon/chlorophyll x 2 x 10"3 gm m~3
Carbon x 4 x 10"3 gm m~3
Carbon x 2 x 10"3 gm m~3
S - (Schl + S0) gm m-3
*Carbon and chlorophyll are usually expressed as milligrams per cubic
meter (mg nr3), but S is expressed in grams per cubic meter (gm m-3),
-------
11
LITERATURE CITED
Antia, N. J., C. D. McAllister, T. R. Parsons, K. Stephens, and
J. D. H. Strickland. 1963. Further measurements of primary
production using a large-volume plastic sphere. Limnol. Oceanogr.
8:166-183.
Ilmavirta, V. 1974. Diel periodicity in the phytoplankton community
of the oligotrophic lake Paa"ja'rvi, southern Finland I. Phytoplankton
primary production and related factors. Annales Botanici Fennici
11:136-177.
Lorenzen, C. J. 1968. Carbon:chlorophyll relationships in an
upwelling area. Limnol. Oceanogr. 13:202-204.
Megard, R. 0., and P. P. Smith. 1974. Mechanisms that regulate
growth rates of phytoplankton in Shagawa Lake, Minnesota. Limnol.
Oceanogr. 19:279-296.
Reid, R. A., D. W. Schindler, and R. V. Schmidt. 1975. Light
measurements in the experimental lakes area, 1969-73. Technical
Report #559. Dept. of Environment, Freshwater Institute, Winnipeg,
Manitoba.
Riley, G. A. 1956. Oceanography of Long Island Sound, 1952-1954.
2. Physical oceanography. Bull. Bingh. Oceanogr. Collect. 15:15-46.
Schindler, D. W. 1971. Carbon, nitrogen, and phosphorus and the
eutrophication of freshwater lakes. J. Phycol. 7:321-329.
Small, L. F., and R. C. Curl. 1968. The relative contribution of
particulate chlorophyll and river tripton to the extinction of
light off the coast of Oregon. Limnol. Oceanogr. 13:84-91.
Strickland, J. D. H., 0. Holm-Hansen, R. W. Eppley, and R. J. Linn.
1969. The use of a deep tank in plankton ecology. I. Studies
of the growth and composition of phytoplankton crops at low
nutrient levels. Limnol. Oceanogr. 14:23-34.
Tyler, J. e. 1975. The in situ quantum efficiency of natural
phytoplankton populations. Limnol. Oceanogr. 20:976-980.
Verduin, J. 1954. Phytoplankton and turbidity in western Lake
Erie. Ecology 35:550-561.
-------
12
Verduin, J. 1964. Principles of primary productivity: photosynthesis
under completely natural conditions. NATO Advanced Study Institute.
Pages 221-238 in D. F. Jackson (ed.), Algae and man. Plenum Press,
New York.
Whitney, L. V. 1938. Transmission of solar energy and the scattering
produced by suspensoids in lake waters. Trans. Wise. Acad. 31:201-221,
-------
13
APPENDIX
WORKING DETAILS
When underwater light meter data are obtained, it is most
convenient to plot the data immediately on a semi-log graph, as
shown by the figure. If the data lie on a single, reasonably
straight line, one can assume that the suspensoid concentration
is essentially homogeneous vertically, and the depth associated
with 1% (DO.01^ °^ surface light (the euphotic zone) can be
identified simply by locating the point on the graph where the
line has traversed two log cycles.
However, if the plot indicates two or more lines, then one
must assume that waters of differing turbidity are superposed.
The slope of each line (DQ.QI) should be determined separately,
and it is advisable to collect samples from the approximate center
of each vertical stratum. The strata will usually have different
suspensoid dry weights, and in some cases they may differ in
dissolved pigments also.
The extinction coefficient can be derived from each DQ QI
value as is shown on the graph, k_ for the upper layer is
4.6/11 = 0.4 nrl, and k^ for the lower layer is 4.6/6.3 = 0.73 nT1.
When suspensoids are collected by centrifuge, or filter, the
importance of resuspending any suspensoids that may have settled
out in the sample bottle must be emphasized. The millipore filters
-------
14
must be dried at the same oven temperature as that to which the
filter-plus-suspensoids will be subjected before its tare weight is
determined.
The pitfalls of carbon and chlorophyll analysis are well
documented in the literature, but the precaution concerning resus-
pending settled materials before filtering samples for those deter-
minations is also worth emphasizing. Moreover, it is an interesting
fact that numerous sets of carbon and chlorophyll analyses are in
existence, but no dry-weight-of-suspensoid data, measured alonq with
the carbon and chlorophyll data, are available to relate to the
observed extinction coefficients.
-------
15
IT)
l/J
Q.
6
4
2 -
1
8
6
4 .
2 .
1
" 8
en
o 6
4 -
1
8
6
4
DEPTH (meters)
Graph of light intensity versus depth showing broken-line plot
indicating strata of differing turbidity. The single arrows
indicate depth associated with 1% of surface light for each
stratum, the double arrow indicates depth associated with 1%
of surface light in the total water column.
------- |