Limnol. Oceanogr.

, 25(S), 1980,855-864
@ 1980, by the American Society of Limnology and Oceanography, Inc.
In situ measurement of the settling velocity of organic
carbon particles and 10 species of phytoplankton
N. M. Burns and F. Rosa
National Water Research Institute, Canada Centre for Inland Waters,
P.O. Box 5050, Burlington, Ontario L7R 4A6
Abstract
An in situ method of simultaneously measuring the settling velocity of three size ranges of
particulate organic carbon (POC) particles and 10 species of phytoplankton is described.
The measurements took 2 h and were done at four different periods during the diurnal cycle.
The POC fluxes of particles in the size ranges l-10 p, lo-64 p, and ~64 p were about equal
for each size range and showed very little diurnal variation. In contrast, the phytoplankton
showed considerable response to the varying light intensities of a day. The flagellates mi-
grated downward at sunset; a species of diatom settled more slowly during daylight than at
night; a blue-green species showed maximum buoyance at midnight and some green species
changed from downward sinking on a bright day to upward movement on a dull day.
The settling velocity of phytoplankton
was considered as an important variable
in one of the earliest quantitative phyto-
plankton system models (Riley et al.
1949). Since then, numerous estimates of
settling velocities of phytoplankton and
particulate carbon have been used in dif-
ferent models; however, relatively few in
situ measurements have been made of
the settling velocities of different phyto-
plankton species (Eppley et al. 1968;
Reynolds 1975, 1976; Happey-Wood
1976; Kamykowski and Zentara 1977) or
particulate organic carbon (POC) (Burns
and Pashley 1974; Spencer et al. 1978;
Linnegren 1979). In situ measurements
are essential because the settling velocity
of phytoplankton can vary with season,
nutrient concentrations, age of the pop-
ulation, time of day, and relative bright-
ness of the day. More such measurements
are needed if we are to reach an under-
standing of the factors controlling the set-
tling of living and detrital organic parti-
cles.
We here compare the POC settling ve-
locities and fluxes in three size ranges:
l-10 p, lo-64 p, and ~64 p, as defined
by the mesh sizes of the screens used.
We also compare the settling velocities
of phytoplankton with those of POC par-
ticles of similar size. Finally, we estimate
the contribution of phytoplankton carbon
to the settling flux of POC from the epi-
limnion. The effect of the brightness and
time of day on the settling velocity of 10
different phytoplankton species is de-
scribed.
We thank L. Janus for counting the
phytoplankton and P. J. Wade for remov-
ing the zooplankton from POC samples.
We also thank J. Leslie for his help.
Methods
The settling experiments were carried
out at Lake St. George, a small, eutrophic
lake (area = 5.8 ha, maximum depth = 15
m), about 40 km north of Toronto, Ontar-
io. Mooring systems were installed in the
middle of the lake so that cables ran from
submersible floats through pulleys at-
tached to anchor weights and then to
winches mounted ashore. The floats
could be submerged by winching the ca-
bles in, thereby providing taut-wire
moorings. The floats were allowed to sur-
face and the settling chambers retrieved
by releasing the winched cables.
During the experiment the epilimnion
was 4 m thick and isothermal at about
20C and the thermocline extended from
4 to 11 m. We sampled 10 liters of water
from the 3-m depth, Axed this water in
darkened containers, and then poured it
into two settling chambers which were
placed on the mooring cables. The cables
were winched in so that the floats were
just submerged and the l-m-long settling
chambers sat between 2.5- and 3.5-m
depth.
856
Burns and Rosa
SECTION A-A
I
- -REMOVABLE BAFFLES
,-PLATE RELEASE
DRAIN HotES
SPRINGS
SEPARATOR PLATE
OPEN POSITION
Fig. 1. Cross-sectional diagram of in situ settling
chamber.
The settling chamber used (Fig. 1) is
a refinement of an earlier design (Burns
and Pashley 1974). The chambers and
baffles were made from clear acrylic plas-
tic. The chambers were submerged and
left for 2 h; a messenger from the sub-
merged float then released the plate from
its position beside the chamber and seg-
mented the chamber into a 4.O-liter top
section and a 0.35-liter bottom section.
After the chambers were retrieved from
the mooring, water samples were drained
from the two sections of each and pre-
pared for analysis.
Duplicate 30-ml subsamples of water
from each sample were set aside for phy-
toplankton analysis. The rest was
screened through a 64-pm mesh screen
and then through a lo-pm mesh screen;
200 ml of this filtrate was then filtered
through a Whatman GF/C glass-fiber fil-
ter for POC analysis. The material re-
tained on the 64-pm screen was resus-
pended and preserved with Lugols
solution, the zooplankton manually re-
moved, and the samples were analyzed
for POC content. The material retained
on the lo-qn screen was washed directly
onto glass-fi her filters and analyzed for
POC content; examination of a random
selection of i,hese samples indicated very
few zooplancton in them.
Although each day was divided into
four periods (sunrise, 0600-0800; mid-
day, 1100-1300; sunset, 1700-1900; mid-
night, 2300-.OlOO), we could only carry
out two expf:riments per day. Over the 4
days, we were able to do two experi-
ments for each period.
The 10 dominant species were select-
ed from a riumber of samples from the
3.0-m depth. These species had cell or
colony sizes which would permit almost
all of them to pass through the 64-pm
mesh screerl, but be retained on the lo-
pm mesh; they were counted in all sam-
ples and formed the basis of comparison
of phytoplar kton with particulate organic
carbon in tire 10-64-p size range. The
carbon contc:nt of the different algae was
estimated from their size (Mullin et al.
1966).
Each 30-ml duplicate was settled out
and examined under an inverted micro-
scope. The number of cells counted var-
ied with the species: a minimum of 100
individuals of the rarer species were
counted and 2,000 or more individual
cells of the colonial species. A three-way
ANOVA of the phytoplankton data was
carried out and the mean-square error of
the complete data matrix per species
used to calculate the variance in the de-
termination of a concentration of that
species. This variance was then used in
calculating the standard deviation in the
computed settling velocities.
The calclllation of the settling velocity
and settlirrg flux of any variable is
straightforwfard if the concentration of the
variable in ,:he top and bottom sections of
the settlirq: chamber have been deter-
mined, The volume of water in the upper
and lower sections is measured and these
values, together with data on the concen-
tration difference between the upper and
lower sections, make it possible to cal-
culate the net settling velocity and net
Settling velocities of organic particles 857
settling flux of the particulate organic car-
bon or phytoplankton.
The initial concentration of POC in the
settling chamber, Ci, is given by
ci = (V,C, + V,C,)/(V, + V&
(1)
where Vt is volume of top section; Cl,
final concentration of POC in top section;
Vfi, volume of bottom section; and Cb,
final concentration of POC in bottom sec-
tion. The net transport from the top sec-
tion of the bottle to the lower section, f,
is given by
f= WCb - G> (2)
d( V&/V, + VJ (3)
where d = Cn - Cl. If F is the net settling
flux, then F = f/(cross-sectional area) x
(settling time). The net settling velocity,
S, is given by
S = FICi.
(4)
The relatively short settling time ensures
that the concentration of particles at the
level of the separator plate (Fig. 1) re-
mains close to the original concentration.
The only variable in the right-hand side
of Eq. 3 is d and thus the uncertainty of
the flux is determined by the uncertainty
in d.
Validation of settling chamber
techniques
The in situ settling chambers should
not alter the settling velocities of the phy-
toplankton or POC particles if this tech-
nique is to be valid. Since the settling
chambers are a new design, we hope to
demonstrate that the settling velocities
determined using them are similar to
those under natural conditions.
Since the chambers are baffled, the de-
gree of turbulence in them is less than
that in the lake. Turbulence, however,
only affects particle settling velocities
when the particle Reynolds numbers are
~0.5 (Bloesch and Burns 1980; Hutch-
inson 1967) and even these larger parti-
cles are hardly affected by the degree of
turbulence prevalent in most lakes and
oceans. In our study, very few of the par-
ticles were large and the turbulence
levels in the small protected lake were
very low; thus the decrease of turbulence
in the settling chamber would be unlike-
ly to affect the settling velocity of the par-
ticles. The reason for minimizing the tur-
bulence in the settling chambers was to
prevent the resuspension of material that
had settled onto the bottom of the cham-
bers. Turbulence can be induced in un-
baffled chambers by their rotation on the
mooring line or by jerking as the cham-
bers are being set or retrieved. Also, set-
tled material can be resuspended by con-
vection currents caused by a temperature
difference between the water in the
chamber and the surrounding lake water
when the chambers are immersed.
We did tests to determine whether the
system of baffles was adequate to prevent
stirring of the water in the bottom seg-
ments of the chambers. KMnO, crystals
put on the bottom of the chamber colored
the water strongly for a distance of about
5 mm from each crystal; the chambers
were then placed in an environment 5C
colder, and later 5C warmer, than the
water inside. The baffling prevented con-
vection currents from reaching the floor
of the bottom section and did not disturb
the water colored by the crystals. The
chambers with KMnO, crystals in the
bottom section were then hoisted on a
simulated mooring system in a workshop
and rotated on the mooring line: again,
the colored water remained undisturbed.
Finally, after closing the dividing plate,
we simulated a violent retrieval of the
system from the lake and found that this
did mix the colored water throughout the
bottom segment but none of this colored
water mixed into either the box on the
side of the chamber or the top segment
of the chamber. A violent retrieval would
thus not invalidate the experiment pro-
viding that the plate segmented the cham-
ber before retrieval, which is the normal
experimental procedure.
We have compared the settling veloc-
ities determined for some species by our
settling chamber technique with veloci-
ties determined by other investigators.
The average settling velocity for Fragi-
lariu crotonensis measured during our
study was 0.27 + 0.13 m * d-l (see Fig. 3d).
858 Burns and Rosa
04 a0
PHYTOPLANKTON CARBON lo-64~.
Fig. 2. Settling velocities and fluxes of phyto-
plankton and detrital organic carbon in 10-64-p
range.
This is as close as could be expected to
a value of 0.43 5 0.22 m-d-l determined
for the same species by Reynolds (1976)
from observations of its disappearance
from the epilimnion of a small English
lake. In an in situ experiment, we found
Asterionellu formosu to settle with a
mean velocity of 0.40 + 0.28 m * d-l. Lab-
oratory measured settling velocities for
this species (Smayda 1974; Titman and
Kilham 1976) averaged 0.31 m * d-l, with
observed values ranging from 0.09 m * d-l
for cells in growth phase to 0.63 for cells
in stationary phase. Phytoplankton set-
tling velocities cannot be compared
strictly because even for the same
species they can be quite variable. How-
ever, the above comparisons do illustrate
that -the in situ settling chambers mea-
sure flux and settling velocity of the nat-
ural organic particles at least as well as
other techniclues now in use.
Our settling chamber technique mea-
sures the intrinsic settling velocity of par-
ticles, not their loss rate from an epilim-
nion. The loss rate can only be obtained
if both the c.ownward settling velocities
and the upward transport of particles by
vertical water movements are known.
The upward transport of particles and the
suspending action of Langmuir circula-
tions, however, seldom reach the full
depth of an epilimnion (Scott et al. 1969)
and thus caL se relatively little resuspen-
sion of particles positioned just above the
thermocline, For this reason the down-
ward settling; flux, measured with settling
chambers jltst above the thermocline,
will approximate the loss rate of particles
from the ep! limnion (Bloesch and Burns
1980).
We attempted a direct check on the set-
tling chamber technique by comparing
downward Cluxes measured by settling
chambers with those measured by a sed-
iment trap. This is not an easy compari-
son because the settling chambers mea-
sure an almost instantaneous flux with a
2-h settling period, while sediment traps
measure a flux averaged over an exposure
period of a few days or more. In this test,
we set two settling chambers at 8 m and
two more at 25 m four times during a 24-h
period and simultaneously exposed a
sediment tr,ip for 24 h midway between
the other dt:pths and within the thermo-
cline at 15 n. Even when placed in tur-
bulent water, sediment traps of good de-
sign will measure the downward flux of
particles in the calm water just above the
bottom of the trap (Hargrave and Burns
1979; Bloe:ch and Burns 1980); for this
Table 1. Average particulate organic carbon concentrations (mg C * m-), fluxes (mg C * m- * d-l), and
settling velocities (m.d-1) in three different size ranges, Lake St. Gt orge, 5-9 September 1977.
L-10 /.I, 10-61/L MP
Concentration 1,107 + 66 198 + 3.0 136 & 8.8
Fluxes 271 -+ 36 306 k 41 312 k 46
Settling velocities 0.24 k 0.03 1.54 ?I 0.22 2.32 k 0.16
2 avz flux = 889 + 71
Settling velocities of organic particles 859
(a)
(d)
UPWARD
BEHAVIOR
5
-SLlNklSE MID&Y SUNSET M~DF~~GHT
041
it
0211PWAm
(b)
It
01
UPWARD
(e)
NORMAL DAY
SUNRISE MlDbAVcr MloFilGHT
0.5
02 DOWNWARD
05
1 I---.
SUNRISE MIDDAY SUdSET MIDNIGHT
a. Cryptomonas erosa AVCJ SETTLING VELOCITY q 0.31 + 032m*DAY
b. Cryptomonas marsonll
Avg SETTLING VELOCITY q 0.32f 032m-DAY-
c. Rhodomonas mlnuta Avg SETTLING VELOCITY q 0.07 f 0.2lm*DAY-
d Fragilaria crotonensis Avg SETTLING VELOCITY = 0.27f 013m*DAY-
e Gomphosphaeria lacustns Avg SETTLING VELOCITY q 0.11 f 0.05m.DAY-
f. Anabaena spiroides Avg UPWARD VELOCITY = 0.10 k Ollm - DAY-
g. Selenastrum minutum
Avg SETTLING VELOCITY = 0.15 f 013m *DAY4
h. Closterium parvulum
Avg SETTLING VELOCITY = 0.18 f Ollm-DAY-
i Scenedesmus acutlformls Avg SETTLING VELOCITY q 0.10 f O.O6m*DAY-
1 Lagerhaemla quadnseta Avg SETTLING VEUXITY q 0.08 f O.llm-DAY-
Fig. 3. Settling velocities of different phytoplankton species as a function of time and relative bright-
ness of a day. Numbers refer to dates in September 1977 when experiments were done.
860
Burns and Rosa
Table 2. Light intensities (lux x 103) measured at midday at Lake: St. George, 6-8 September 1977.
0.5 1.0
Depth (m)
2.0 3.0 4.0 Weather conditions
6 Sep
7 Sep
8 Sep
5,9 Sep
3.90 2.76 1.59 1.04 0.60 Overcast, occasional dull sun
2.15 1.55 0.94 0.58 0.31 T-hick cloud cover all day
- 11.09 6.44 3.34 1.93 E:arly morning mist then
continued bright sun
No midday values measured
reason, the flux measured in the calm
water inside settling chambers should be
similar to that measured in the calm
water inside sediment traps in the same
water column. The average fluxes mea-
sured by the settling chambers in our ex-
periment were 0.40 and 0.38 g
POC * mm2 * d-l at the 8- and 25-m depths.
The flux measured by the sedimentation
trap was essentially the same as that mea-
sured by the settling chambers, 0.40 g
POC. rnm2. d-l.
Red ts
The results of the organic particle set-
tling investigation are summarized in Ta-
ble 1. The values shown are averages of
all those determined except for the
>64-p fraction, which does not include
any midnight values. There were large
numbers of zooplankton in the bottom
segment samples taken at midnight, and
although nearly all were manually re-
moved, the few remaining were enough
to contaminate the samples and invali-
date these values.
The small particles in the 1-10-p range
represented more than five times the car-
bon of the particles in the other two size
ranges. Thus, although the small parti-
cles settle slowly, the large concentration
of small particles resulted in their carbon
.
flux being almost equal to the flux ob-
served for the other two size ranges.
The carbon contained in the different
phytoplankton cells was estimated from
the cell volume (Mullin et al. 1966). The
total phytoplankton carbon concentration
and flux in .:he 10-64-p range was then
estimated by summing the values for
each param zter for each species. The
data in Fig. 2 clearly show that the phy-
toplankton carbon flux was negligible
compared with the detrital carbon flux
and that the average settling velocity of
phytoplankton was <lo% that of organic
detrital particles of similar size. The av-
erage phytoplankton settling velocity
was low because some organisms were
moving upward, partially counteracting
the effect OF others moving downward.
The increas 3 in the settling velocity of
the detrital organic carbon at sunset is
worth noting.
The experiments were done from mid-
night, 5 September, to sunrise, 9 Septem-
ber. The three intervening days showed
considerable difference in terms of rela-
tive brightness (Table 2). Some of the
phytoplanktlm species studied showed
an alteration in settling velocities accord-
ing to the relative brightness of a partic-
ular day (Eig. 3). All of the species
showed an alteration of settling velocities
Table 3. Nutrient concentrations observed at Lake St. George on 8 September 1977 (units-pgeliter -).
0.5 m 1.5 m 2.5 m 3.5 m 4.5 m 5.5 m 6.5 m 7.5 m 8.5 m 9.5 m
19.7% 19.7-z 19.7% 19.7% 19.6% 19.3% 17.5% 14.OC 9.4% 7.0% ,:E
0
N03- + NO,- 5 9 7 6 8 18 18 34 5 <5 <5
NH,+ 66 77 21 22 27 22 53 340 830 990 1,220
SRP 2.1 1.8 2.1 2.2 1.7 2.0 1.9 2.1 2.3 2.4 2.3
X filtered P 13.0 6.8 6.5 6.5 5.8 8.2 7.5 11.0 10.0 9.3 9.0
Settling velocities of organic particles 861
.
with time of day. While examining these
data, one should keep in mind that the
figures represent the response of the total
population of a species. Since senescent
or recently dead cells (counted as being
alive) will show little diurnal pattern or
response to light, they will settle down-
ward at a relatively constant rate; this
will diminish the estimated magnitude of
the upward or downward movement of a
group of live, migrating cells.
The scale of Fig. 3a and b is twice that
of Fig. 3c-j; thus, the flagellates Crypto-
monas erosa and Cryptomonas marsonii
can have greater settling velocities than
the other species shown. Although C.
erosa had a volume of about 2,000 p3 per
cell and C. marsonii about 400 t,cn, they
showed almost identical settling veloci-
ties and patterns of movement. Both
species remained almost neutrally buoy-
ant from before midnight to after midday,
but sank or migrated downward at a rel-
atively high rate just before darkness.
Rhodomonas minuta v. nannoplankti-
ca, also a flagellate, had a pattern fairly
similar to that of the other two crypto-
monads, but showed considerable re-
sponse to the relative brightness of the
day. It continued its early morning move-
ment upward until midday on the dull
day, 7 September, but began moving
downward after sunrise of the normal
day, 6 September.
The only diatom, Fragilaria crotonen-
sis, showed (Fig. 3d) a response different
from that of the flagellates, with the low-
est settling velocity at sunset at the end
of a bright day.
The two blue-green species in the
study reacted differently from each other
in their dirunal pattern. Gomphosphaeria
lacustris (Fig. 3e) did not show any dra-
matic changes in settling velocity, al-
though these colonial cells do appear to
have settled a little faster at the end of
the photosynthetic period, at sunset at
the end of the brightest day, 8 Septem-
ber. Anabaena spiroides (Fig. 3f) was up-
wardly buoyant except in the morning
when it was probably neutrally buoyant.
Its buovancv increased from sunrise to
midday and then again from sunset to
midnight.
The last species under consideration
(Fig, 3g-j) are green algae, which often
form a large part of the phytoplankton
community in Lake St. George toward
the end of the stratified period. The most
interesting feature of this group was the
tremendous difference in response of
those species to very dull conditions.
Closterium parvulum and Selenastrum
minutum reversed the normal downward
sinking trend to move upward during the
midday period of the very dull day, while
Scenedesmus acutiformis became neu-
trally buoyant at the end of the dull day.
Each species of green algae had a very
low average settling velocity.
Discussion
The data in Fig. 2 show that virtually
all the carbon flux out of the epilimnion
in the 1-64-p size range is the result not
of phytoplankton settling but of the set-
tling of organic detritus. While phyto-
plankton carbon made up about 25% of
the organic carbon in that size range, the
phytoplankton only accounted for about
1.7% of the downward flux of carbon as
a result of the low mean phytoplankton
settling velocities (Fig. 3).
The settling velocity and flux of detrital
organic carbon varied diurnally (Fig. 2),
with the maximum in both variables dur-
ing the sunset experiments. There was no
significant change in the concentration of
detritus at sunset; the increased flux was
entirely due to the increased settling ve-
locity of the particles. We at first had dif-
ficulty in finding a possible reason for the
diurnal variation in the settling velocity
of inanimate material. However, the data
on the zooplankton removed from the
samples showed that the sunset experi-
ments contained higher numbers of zoo-
plankton than any of the experiments
done at other times (the water for these
experiments was sampled 3 h before
darkness). Enright (1977) hypothesized
and confirmed (Enright and Honneger
1977) that zooplankton migrate upward
before darkness and begin grazing heavi-
862
Burns and Rosa
ly when first arriving in the zone of pho-
tosynthesis. The data in Fig. 2 would
tend to support these ideas in that, just
before darkness, an increased number of
actively grazing zooplankton would con-
vert phytoplankton and smaller detrital
particles into fecal pellets at a higher
rate. Since most fecal pellets settle faster
than 1.7 m . d-l (Ferrante and Parker
1977; Smayda 1971), the increased num-
ber of fecal pellets would increase the
average settling velocity of the organic
particles at sunset.
The different species of phytoplankton
showed very different responses to the
varying intensities of light during a day.
The most striking feature about the set-
tling behavior of C. erosa and C. mar-
sonii is the downward migration of these
organisms at sunset (Fig. 3a, b). The 2-h
settling period ran from 1700 to 1900
hours with the onset of darkness later, at
about 1945. These organisms were prob-
ably migrating downward in response to
the length of exposure to daylight rather
than to the gradual decrease in light in-
tensity before sunset. The pattern of de-
scent before darkness is similar to that
observed in test tank simulations of diur-
nal migration of marine dinoflagellates,
in which the organisms also migrated
downward an hour or two before the light
intensity changed abruptly when the
lights were switched off (Kamykowski
and Zentara 1977; Eppley et al. 1968).
Weiler and Karl (1979) found similar be-
havior in a laboratory population of Ce-
ratium furca and further observed that
this organism migrated upward shortly
before the lights were turned on. Some
of the flagellates in our investigation may
also have moved upward before sunrise;
we never did an experiment at this time
of day. The response of R. minuta (Fig.
3c) to high and low light intensities was
similar to that observed by Happey-
Wood (1976) who noted that R. minuta
moved upward under intermittent cloud
cover from before sunrise to midday and
moved downward as the skies cleared
during the afternoon and the light be-
came more intense. Although all three
flagellated species (Fig. 3a, b, c) are pho-
totactic, R. minuta is obviously far more
responsive t 3 relative light intensities
than the two Cryptomonas species.
Fragilaria :rotonensis (Fig. 3d) appar-
ently had its settling velocity diminished
by higher light intensity. There are no
other published studies on this relation-
ship for this organism but a maiine dia-
tom, Ditylum brightwelli, has been ob-
served to hnve its settling velocities
increased by photosynthesis (Anderson
and Sweeney, 1977).
The settling behavior of two other An-
abaena species, Anabaena flos-aquae 2
(Kanopka et ~1. 1978) and Anabaena cir-
cinalis (Reyrolds 1975), has been inves-
tigated. In both studies, the Anabaena
species were found to be buoyant at all
times, but their buoyancy was reduced
by photosynthesis. Our finding was sim-
ilar in that A. spiroides was observed to
be buoyant c.t nearly all times and most
so at midniglrt.
Settling velocities of the remaining
five 0rganisinsX. lacustris, S. minu-
turn, C. parvullum, S. acutiformis, and
Lagerhaemk quadriseta (Fig. 3e, g, h, i,
j)-have not been reported in the litera-
ture, and we are unable to comment on
their hehavicr as we did not carry out any
physiologica studies.
We have lrere developed information
on the rate < at which phytoplankton
change their depth and how these rates
can change lvith time of day; it is of in-
terest to examine the other factors which
change with depth. The main variables
affecting pjrytoplankton, which also
change with depth, are light intensity,
temperature, and nutrient concentra-
tions. Light intensity (Table 2) decreased
considerably with depth and cloud cover;
Fig. 3 shows that five of the ten species
reacted to tl-e varying intensities on dif-
ferent days and that all the species react-
ed to the varying light intensities at dif-
ferent times of day.
We only studied the phytoplankton at
a single tem:?erature during the 4-day in-
vestigation and cannot comment on the
effect of temperature on settling velocity.
Nutrient concentrations were mea-
sured as a lunction of depth (Table 3).
Settling velocities of organic particles
863
The only soluble nutrient to show any
ganic carbon flux in the eutrophic lake
significant gradient through the thermo-
studied.
Cline was ammonium ion, which was
As a contrast to the pattern of constant
present in fairly high concentration in the
settling of detrital carbon, almost all of
epilimnion and was thus unlikely to be
the phytoplankton species investigated
limiting. Soluble reactive phosphorus
showed significant diurnal variation in
(SRP) showed no variation with depth their settling velocities.
and therefore would not affect the verti-
cal behavior of the plankton. References
We must thus conclude that light had
the greatest effect on the vertical distri-
bution of the phytoplankton. Happey-
wood (1976) came to the same conclu-
sion after a diurnal study of the vertical
distribution of phytoplankton. Reynolds
(1976) studied the vertical distribution of
several species over a summer and con-
cluded that nutrient availability did
cause changes in the species present,
but, for long periods when there were no
major changes in nutrient availability and
hence no changes in species composi-
tion, light was the controlling variable.
Conclusions
We have shown that settling chambers
can be used to measure the diurnal flux
of organic carbon out of the epilimnion.
The flux was almost equally divided be-
tween the three size ranges (l-10 k, lo-
64 p, and ~64 p). There was little varia-
tion in the carbon fluxes during the 24-h
diurnal period One exception was the
slight increase in both settling velocity
and flux of detrital organic carbon in thk
10-64-p size range shown during the
sunset period (Fig. 2). This may result
from an increased number of fast settling
fecal pellets produced by the presunset
upward migration of zooplankton.
Most phytoplankton were in the lo-
64-p size range and, although phyto-
plankton carbon comprised about 25% of
the particulate carbon in this size range,
phytoplankton carbon comprised only
1.7% of the total downward flux because
of the relatively low phytoplankton set-
tling velocities. In the other size ranges,
the phytoplankton carbon flux was a
much smaller fraction of the total carbon
flux. Thus, phytoplankton carbon fluxes
comprise ~1% of the total downward or-
ANDERSON, L. W., AND B. M. SWEENEY. 1977. Diel
changes in sedimentation characteristics of Di-
tylum hrightwelli: Changes in cellular lipid
and effects of respiratory inhibitors and ion-
transport modifiers. Limnol. Oceanogr. 22:
539-552.
BLOESCH, J., AND N. M. BUKNS. 1980. A critical
review of sedimentation trap technique.
Schweiz. Z. Hydrol. 41: in press.
BURNS, N. M., AND A. E. PASIILEY. 1974. In situ
measurement of the settling velocity profile of
particulate organic carbon in Lake Ontario. J.
Fish. Res. Bd. Can. 3 1: 291-297.
ENRIGIIT, J. T. 1977. Diurnal vertical migration:
Adaptive significance and timing. Part 1. Selec-
tive advantage: A metabolic model. Limnol.
Oceanogr. 22 : 856-872.
-, AND H.-W. HONNEGER. 1977. Diurnal ver-
tical migration: Adaptive significance and tim-
ing. Part 2. Test of the model: Details of timing.
Limnol. Oceanogr. 22: 873-886.
EPPLEY, R. W., 0. HOLM-HANSEN, AND J. D.
STRICKLAND. 1968. Some observations on the
vertical migration of dinoflagellates. J. Phycol.
4: 333-340.
FERRANTE, J. G., AND J. I. PARKER. 1977. Trans-
port of diatom frustules by copepod fecal pel-
lets to the sediments of Lake Michigan. Lim-
nol. Oceanogr. 22: 92-98.
HAPPEY-WOOD, C. M. 1976. Vertical migration
patterns in phytoplankton of mixed species
composition. Br. Phycol. J. 11: 353-369.
HARGMVE, B. T., AND N. M. BURNS. 1979. As-
sessment of sediment trap collection efficiency.
Limnol. Oceanogr. 24: 1124-1136.
HUTCIIINSON, G. E. 1967. A treatise on limnology,
2. Wiley.
KAMYKOWSKI, D., AND S. J. ZENTAM. 1977. The
diurnal vertical migrgtion of motile phyto-
plankton through temperature gradients. Lim-
nol. Oceanogr. 22: 148-151.
KONOPKA, A., T. D. BHOCK, AND A. E. WALSDY.
1978. Buoyancy regulation by planktonic blue-
grcen algae in Lake Mendota, Wisconsin. Arch.
Hydrobiol. 83 : 524-537.
L;~NNEC,REN, C. 1979. Buoyance of natural popu-
lations of marine phytoplankton. Mar. Biol. 54:
l-10.
MULIJN, M. M., P. R. GOAN, AND R. W. EPPLEY.
1966. Relationship between carbon content,
cell volume and area in phytoplankton. Limnol.
Oceanogr. 11: 307-3 11,
REYNOLDS, C. S. 1975. Interrelations of photosyn-
864 Burns and Rosa
thetic behaviour and buoyance regulation in a
natural population of a blue-green alga. Fresh-
water Biol. 5: 323-338.
-. 1976. Sinking movements of phytoplankton
indicated by a simple trapping method. 1-A
Fragilnriu population. 2-Vertical activity
ranges in a stratified lake. Br. Phycol. J. 11:
279-291,293-303.
HILEY, G. A., H. STOMMEI,, AND D. F. BUMPUS.
1949. Quantitative ecology of the plankton of
the western North Atlantic. Bull. Bingham
Oceanogr. Collect. 12: 1-169.
SCOTT, J.T.,G.E. MEYER, R. STEWART,AND E.G.
WALTIIER. 1969. On the mechanism of Lang-
muir circulations and their role in epilimnion
mixing. Limnol. Oceanogr. 14: 493-503.
SMAYDA, T. J. 1971. Normal and accelerated sink-
ing of phytoplankton in the sea. Mar. Geol. 11:
105-122.
-. 1974. Scme experiments on the sinking
characteristic.; of two freshwater diatoms. Lim-
nol. Oceanog *. 19: 628-635.
SPENCER, D. W.! AND OTHERS. 1978. Chemical
fluxes from a sediment trap experiment in the
deep Sargass I Sea. J. Mar. Res. 36: 493-523.
TITMAN, D., ANI) P. KILHAM. 1976. Sinking in
freshwater plankton: Some ecological implica-
tions of cell nutrient status and physical mixing
processes. Li nnol. Oceanogr. 2 1: 409-4 17.
WEILER, C. S., AND D. M. KARL. 1979. Diel
changes in phased-dividing cultures of Ceru-
tiumfurcu (Ibinophyceae): Nucleotide triphos-
phates, aden {late energy charge, cell carbon
and patterns of vertical migration. J. Phycol.
15: 384-391.
Submitted: 20 November 1979
Accepted: 8 April 1980