|
l
I

.7"?-

ll; \lllllllflll l 111ml ‘llllll

This is to certify that the
thesis entitled

DIFFERENTIAL EFFECTS OF GROWTH AND LOSS PROCESSES IN
CONTROLLING NATURAL PHYTOPLANKTON POPULATIONS

presented by

WILLIAM GRADY CRUMPTON

has been accepted towards fulfillment
of the requirements for

 

Ph. D. degree in him..—

Zué Jp/

Major professor
Robert G. Wetzel

Date__2_4_J_ulLJ.28_O___

0-7 639

 

JUL181999
0928 00

 

 

OVERDUE FINES:

25¢ per day per item

RETURNING LIBRARY MATERIALS:

Place in book return to remove
charge from circulation records

 

 

 

 

DIFFERENTIAL EFFECTS OF GROWTH AND LOSS PROCESSES IN

CONTROLLING NATURAL PHYTOPLANKTON POPULATIONS

by

WILLIAM GRADY CRUMPTON

A DISSERTATION

Submitted to
Michigan State University
In partial fulfillmentISI the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Botany and Plant Pathology

1980

 

ABSTRACT

DIFFERENTIAL EFFECTS OF GROWTH AND LOSS PROCESSES IN
CONTROLLING NATURAL PHYTOPLANKTON POPULATIONS
by

William Grady Crumpton

An investigation was made of factors controlling algal succession
in a small, oligotrOphic lake during summer stratification. Weekly
measurements were made of growth rate, sedimentation rate and
population density for each of the dominant phytoplankton species.
Weekly diel measurements were made of zooplankton grazing rates using

14C labeled algae. Cyclotella michiganiana was the dominant alga

 

through the end of June at Which time Cyclotella comensis began to
increase, becoming the dominant by August. In August, high grazing
pressure caused the rapid declines of both E: michiganiana and £3

comensis which were followed by an increase of Sphaerocystis

 

Schroeteri. The combined effect of greater growth rates and lower loss

 

rates of C. comensis resulted in its dominance over of C. michiganiana.

In contrast, the g. comensis - S: Schroeteri succession clearly

 

resulted from differential mortality alone. It is likely that the
importance of losses due to sedimentation and/or grazing is large in
many lakes and that interspecific competition may be less important in

actually controlling seasonal succession.

ACKNOWLEDGEMENTS

I am deeply grateful to Dr. Robert G. Wetzel for his continued
advice, encouragement, and support throughout this investigation and my
graduate career. I am also very grateful to the remaining members of
my graduate committee, Drs. Donald J. Hall, George H. Lauff and Earl E.
Werner. Each of them has contributed greatly and differently to the
development of this work and especially to my own development.

I have benefited considerably from interactions with fellow
graduate students and researchers including Art Stewart, Dave Francko,
Kay Gross, Dave Hart, Jim Grace, Jim Gilliam, Joyce Dickerman, Gary
Mittelbach, Gary King and Dick Smith. A special thanks goes to Dr.
Clyde Goulden for open discussions and encouragement. Dr. Michael Klug
generously and without complaint allowed the use of his laboratory for
which I remain grateful. I would also like to acknowledge the
technical help of Jay Sonnad and graphical assistance of Anita Johnson.
A very special debt is owed to Dr. Gerald A. Moshiri for his unreserved
support and encouragement and to my parents for fostering an early
self-dependence.

I would most of all like to thank Beverly who has grown with me
these past eight years and whom I love dearly.

This work was supported by the U. S. Department of Energy
(EY-76-S-02-1599, COO-lS99-l78) and the National Science Foundation
(BMS-75-20322).

ii

TABLE OF CONTENTS

LIST OF TABLES .................................................... iv

LIST OF FIGURES ................................................... v

INTRODUCTION ...................................................... 1

DESIGN AND METHODOLOGY ............................................ 2

Estimates of phytoplankton population size and r ........ 3

Estimates of u and p .................................... 4

Estimates of ZOOpIankton biomass and g .................. 4

Estimates of s and f .................................... 7

RESULTS AND DISCUSSION ............................................ 8

LITERATURE CITED .................................................. 21
APPENDIX A. DIALYSIS CHAMBERS FOR THE MEASUREMENT OF

PHYTOPLANKTON GROWTH RATE EN SITU .................... 24

Literature Cited ..................................... 34

APPENDIX B. A METHOD FOR THE PREPARATION OF PERMANENT,
QUANTITATIVE MOUNTS OF PHYTOPLANKTON FOR
CRITICAL MICROSCOPY .................................. 37
Literature Cited ..................................... 48

APPENDIX C. NOVOLACS: SYNTHETIC RESINS SUITABLE FOR MOUNTING
BIOLOGICAL MATERIALS ................................. 49

Literature Cited ..................................... 52

iii

LIST OF TABLES

Table Page
1 Comparison of the net concentration and transfer
technique with Haney's (1973) in situ technique ...... 6
APPENDIX B
1 Mean number of organisms ml'1 + S.D. of four
replicates ........................................... 46

iv

Figure

1A

1B

10

1D

LIST OF FIGURES

Population curves for Cyclotella michiganiana

 

 

 

 

( ---------- ), Cyclotella comensis (-————————4),

and Sphaerocystis SchroeterI ( ----- ). Error

bars denote 95% confidence limits ......................
Transparency (Secchi depth) in meters ..................

Grazing rates as Z of epilimnion filtered hr'l.
The first of each pair of bars represents daytime
rates; the second represents nighttime rates.
Each bar presents mean of five replicates

(average C.V. = 22%) ...................................

Daytime ( -------- ) and nighttime 0————————)
epilimnetic ZOOplankton biomass. Each point
represents mean of three replicates (average
C.V. = 10%). Dominant zooplankton were
GOpepods (>902) from.May through July and

Daphnia spp. (>902) from August through October .......
Population curve (lower figure, ------- ) and rate

budget (upper figure) for Cyclotella michiganiana
showing u (----9, g (hatched area), ?'(-———————J,

 

3 (open area between hatching and'?), and r ( ------ )....

Population curve (lower figure,——————) and rate
budget (upper figure) for Cyclotella comensis
showing u (--------0, g (hatched area), r'
(-—————-J, 3 (open area between hatching and T),

 

and r ( --------- ) .....................................

Population curve (lower figure, - - -) and rate
budget (upper figure) for Sphaerocystis Schroeteri
showing ll (-———-), ? (-——-—-———), 3 (open area

 

between b and F), and r( -------- ) .....................

Population curves (lower figure) and simplified
rate budgets (upper figure) for Cyclotella
michiganiana ( -------- ), Cyclotella comensis
(-———-———;) and Sphaerocystis Schroeteri (- - - -).
For each rate budget, the upper line represents

 

 

 

 

u and the lower line represents r .....................

V

Page

9
9

9

9

l3

..14

..15

..16

Figure Page

APPENDIX A
1 Dialysis chamber ....................................... 27

2 Predicted (o) and observed (0) densities for
Cyclotella michiganiana. Error bars represent
53% coandence limits on phytoplankton counts.
Predicted densities have additional variance
due to errors in estimates of g (mean C.V. ca.
20%) and 3 (mean C.V. ca. 15%) ......................... 32

 

3 Predicted (o) and observed (0) population
densities for Cyclotella comensis. Error bars
denote 95% confidence limits on phytoplankton
counts. Predicted densities have additional
variance due to errors in estimates of g (mean

 

C.V. ca. 20%) and 5 (mean C.V. ca. 15%) ................ 33
APPENDIX B
1 Settling chambers showing positions for (A) normal
filling and settling, (B) filling for nonturbulent
rinse and (C) dumping overlying water .................. 4O
2 Empty vapor chamber .................................... 41

vi

INTRODUCTION

The dynamics of algal succession are of fundamental importance in
the ecology of freshwater lakes and, accordingly, much effort has gone
toward delineating factors which contribute to the rise and decline of
phytoplankton populations. The results of such study now make it
apparent that while losses due to grazing, sinking, and other processes
may at times be quite substantial, their impact will vary for different
populations and conditions. In addition, losses may have less apparent
effects upon the size of a rapidly reproducing population than upon that
of a stationary one. In order to assess the actual impact of various
processes, it is necessary to conduct studies in which population
increases due to growth and decreases due to grazing, sinking, and other
processes are determined simultaneously.

Unfortunately, such comprehensive investigations are rare and most
studies address community level processes, which reveal very little
regarding the pOpulations involved. If phytoplankton ecologists are to
realistically address algal succession, it is imperative they initially
investigate population level processes. Primary production is obviously
dependent upon species characteristics as are sedimentation and other
transport processes. Indeed, grazing, even by filter feeders, is a much
more discriminatory process than is often realized (Porter 1977; Infante
1978; Starkweather and Gilbert 1978). Many other processes seem to
display a high degree of specificity as to the phytoplankton affected
(e.g. death due to parasitism or physiological extremes). If
successional changes are to be understood, the differential effects of
the above processes upon individual phytoplankton populations must be

considered.

 

The need for such consideration has not escaped previous
investigators, and attempts have been made to address the dynamics of
individual phytoplankton populations. The work of Lund (1949, 1950,
1954) is a well-known example of excellent investigation along these
lines. More recently, Knoechel and Kalff (1975, 1978) have utilized
autoradiography in an effort to determine the relative importance of
growth rates and loss rates in algal succession. However, neither these
nor other investigations have combined actual measurements of growth
rates and loss rates to delineate phytoplankton population dynamics.

Such basic information from field systems is essential to an
understanding of the structure and function of natural communities.
Recent laboratory and theoretical investigations greatly increased our
understanding of certain major processes (Dugdale 1967; Greeney, et al.
1973; DrOOp 1974; O'Brien 1974; Tilman 1977). However, an understanding
of these individual processes provides limited insight unless we know

the frequency, intensity, and duration of their operation in nature.

DESIGN AND METHODOLOGY
The objective of this study was to construct a detailed population
growth equation for each of several major species of phytoplankton in
Lawrence Lake, a small (4.96 ha., zm = 12.6 m, E'= 5.9 m), oligotrophic
lake in southwestern Michigan (for a more detailed description, see
Wetzel ££.El: 1972). The components of the equation may be

conceptualized as:

?'= u - g - s - p - f (1)
where ?'= predicted increase
u = growth (by cell division)

 

 

g = mortality due to grazing

s = loss due to sedimentation

p = mortality due to parasitism or physiological
extremes

f = flushing losses

are expressed as instantaneous rates for each population. The
logistical barriers to such an investigation are of course enormous and
a primary reason for the lack of previous study. This investigation was
limited to a consideration of epilimnetic populations during summer
stratification in 1979 and specifically to the diatom-green algal
succession Which is a prevalent feature of this lake. Temperature
profiles were obtained twice weekly in order to track the epilimnetic-
metalimnetic boundary. Simultaneous estimates of growth rates and loss
rates were obtained for each phytoplankton population studied. Actual
rates of increase (r) were obtained independently based on population
changes. With these data, it is possible to assess the relative

importance of these processes in phytoplankton p0pulation dynamics.

Estimates of phytoplankton population size and r:

 

Replicate phytoplankton samples were collected weekly from a
central station using an integrating tube sampler and were preserved
with 0.4% Lugol's solution. The samples were processed and the
phytoplankton counted using the technique described in Appendix B.
Using a combination of field and strip counts, between 100 and 1000
cells of each species were counted for each sample and 95% confidence
limits calculated as in Lund, Kipling, and Le Cren (1958).
Instantaneous daily rates of increase (r) between successive sampling

periods were calculated from

 

N7 = Noer7 (2)

where No and N7 represent population estimates for successive, weekly

samples.

 

Estimates of u and p:

In this study, I used dialysis chambers (Appendix A) suspended in
§i£2_to estimate growth rates of natural phytoplankton populations at
weekly intervals. The use of these chambers was critically evaluated
experimentally and in Lawrence Lake. Exchange rates of dissolved
material was very rapid and estimates of growth rates were not
statistically significantly different from those of natural populations
under the conditions of this study (Appendix A).

Triplicate dialysis chambers were filled with lake water from which
the dominant grazers had been removed by filtration through 153-um mesh
Nitex® netting. An aliquot of this water was preserved with 0.4%
Lugol's solution for estimates of initial cell numbers (CO below). The
chambers were suspended at the depth of mean epilimnetic light intensity
using a surface-tethered buoy to provide agitation of the chambers via
wave action. After four days of incubation, the chambers were collected
and the incubated samples were preserved with 0.4% Lugol's solution.
Phytoplankton counts were made as described above on initial (Co) and

final (C4) samples. Instantaneous growth rates (u) were calculated from
C4 = C08114 (3)

For each incubation, physiological death (p) was estimated based on
direct microsc0pic evaluations of changes in numbers of dead cells.

Cellular and subcellular integrity were the criteria used to distinguish

 

living from dead cells. Cells were considered viable unless there was
visible damage to their cell walls or plastids.

Based on the work of Haney and Hall (1975), two sampling periods
were chosen for measuring ZOOplankton grazing rates. ,Daytime (2-3 hours
past midday) and nighttime (2-3 hours past sunset) grazing rates (5
replicates each) were measured at weekly intervals using 14C-labeled,

log-phase cultures of Chlorella pyrenoidosa of a mean cell diamter of

 

5 pm. Zooplankton were collected from the epilimnion using vertical
tows with a calibrated net (153-um mesh size) and immediately
transferred to a feeding suspension of unfiltered lake water with ca.

103 cells ml"1 of labeled 9, pyrenoidosa (1-2 dpm cell-l). The effect

 

of this collection and transfer was evaluated by comparison with a
modification of Haney's (1973) technique. No significant difference was
found between collection rates estimated by the two methods (Table 1).
After a six-minute feeding interval, the zooplankton were collected on
153-um mesh netting, narcoticized with carbonated water, transferred to
liquid scintillation vials and preserved with 95% ethanol. These entire
samples were later processed using tissue solubilizers (Packard Soluene
350) and their radioactivities determined with liquid scintillation
radioassay. Counting efficiencies were obtained from a quench curve
constructed using internal standards. For controls, zooplankton were
killed with ethanol prior to the feeding incubation and then processed
as were the other samples. Based on the volume of water from which the
zooplankton were collected and on their feeding activity, grazing rates
were calculated as % volume cleared hr‘l. Daytime and nighttime hourly
rates were assumed constant for daytime and nighttime periods, and were

used to calculate instantaneous rates of mortality due to grazing (g).

Table 1. Comparison of the net concentration and transfer technique

with Haney's (1973) in situ technique.

Relative collection rate (: S.D.; n = 3)

 

 

 

 

Species Net concentration Haney chamber
Daphnia pulex 4.2 + 0.4 4.4 :_0.4
2.3 1 0.2 2.0 :_0.2
Daphnia galeata mendotae 1.6 i 0.2 1.5 :_0.2
Calanoid c0pepods 0.34 + 0.04 0.37 :_0.05

 

For biomass estimates, triplicate zooplankton samples were
collected as described above, narcoticized with carbonated water and
preserved with 95% ethanol. Total zOOplankton volume (ul zooplankton
1'1) was measured for each sample using a graduated column. ZOOplankton
were not counted, but representative samples were examined and the major

forms identified.

Estimates of s and f:

 

Current methodology does not allow an in §i£u_measure of species-
specific sinking rates. Therefore, in this study, I estimated sinking
losses from the epilimnion during a specified interval using
sedimentation traps placed just below the epilimnetic~metalimnetic
boundary. The placement of the traps was adjusted twice weekly to the
gradually changing epilimnetic-metalimnetic boundary.

The traps used in this study were cylinders with a heightzwidth
ratio of 23 (43 cm long x 1.9 cm dia.). The critical work of Hargrave
and Burns (1979) indicates that traps of this design collect with
95-100% efficiency in water moving less than 5 cm sec'l. Thus, When
suspended in metalimnetic layers, as in this study, such traps should
provide reasonable estimates of sedimentation.

Each week, triplicate sedimentation traps were suspended just below
the epilimnetic-metalimnetic boundary and harvested the following week.
The traps were initially filled with filtered (0.5-um) lake water, and a
layer of Lugol-NaCl solution was placed in the bottom of each trap to
preserve sedimenting organisms. The material collected was assumed to
represent cumulative loss from the epilimnion per surface area of trap
opening. Instantaneous rates of loss due to sedimentation (s) were

calculated from

 

8

s = £7 sNtdt (4)

where Nt = Noert = population size at each instant and S = total number
of cells sedimented volume”1 during the 7-day trapping period.

Integrating and solving,
s = (rS)/(N7-No) (5)

Base flow of Lawrence Lake inlet streams is very low compared to
epilimnion volume and flushing losses could only be significant during
storm events. Flushing losses (f) were calculated assuming that all
precipitation impinging on the drainage basin went through the
epilimnion. At no time did flushing losses calculated in this manner
reach more than a few per cent and they are thus assumed to be

insignificant.

RESULTS AND DISCUSSION

From spring through midsummer, Cyclotella michiganiana and g.

 

comensis strongly dominated the phytoplankton in terms of numbers,
biomass and production. In August there was a precipitous decline of

Cyclotella followed by an increase of Sphaerocystis Schroeteri (Figure

 

 

1A). These three species comprised >90% of the total phytoplankton from
mid-June through September, and the following analyses are restricted to
their dynamics.

The decline of Cyclotella and subsequent increase of Sphaerocystis

 

 

is strikingly mirrored by changes in Secchi depth water transparency

(Figure 1B). Data shown in Figure 1C suggests that grazing pressure was

 

a major factor in the Cyclotella-Sphaerocystis succession. Grazing

rates were relatively low through June, reached a peak in August during

Figure 1A.

Figure 1B.

Figure 1C.

Figure 1D.

 

Population curves for Cyclotella michiganiana ( -------- ),
Cyclotella comensis (————————), and Sphaerocystis
Schroeteri (- - - -). Error bars denote 95% confidence
limits.

Transparency (Secchi depth) in meters.

Grazing rates as % of epilimnion filtered hr-l. The first
of each pair of bars represents daytime rates; the second
represents nighttime rates. Each bar presents mean of

five replicates (average C.V. = 22%).

Daytime ( -------- ) and nighttime C—————-—-) epilimnetic
zooplankton biomass. Each point represents mean of three
replicates (average C.V. = 10%). Dominant zooplankton were
c0pepods (>90%) from May through July and Daphnia spp.

(>90%) from August through October.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

SEP ' OCT

 

 

 

'JUL'AUG

'MAY'JUN

 

 

 

PM L n d
1W1 - lullu
i L. L Illllllll
_T Ll. in.“
in
H41 . Id
Ill,
0 IH| J]
1 l J
L. 1..
l I.
4
IN I
i 4
"Us . ,
Ill
1 I
J
L. A B In
_ _ . _ F _ _ r _
O 2 4 6 8 w 3 2 .l 0
.ME Ihmwo

—| ~50: away—.5
w<¢340> thU¢wa

C

— p _

 

 

D

b b _ _

 

4 3 2

n09 x Edi 9.de

8 6 4 2

.Lo. : 205.2388 .3

‘ MAY ' JUN ‘ JUL 1 AUG ' sap ‘ ocr

 

O 11

the decline of Cyclotella and remained high for the duration of the

 

study. These seasonal patterns in grazing were related to quantitative
changes in the ZOOplankton community (Figure 1D). In addition, the
seasonal increase in grazing pressure in early August (Figure 1C)
corresponded to a shift from a copepod community to one dominated by

Daphnia pulex and 2. galeata mendotae.

 

 

It is obvious that there was a striking diel periodicity in grazing
pressure, with consistently higher rates at night (Figure 1C). It is
also obvious that this diel pattern was related to increases in
zooplankton biomass at night, presumably due to migration from deeper
strata (Figure 1D). However, the well established nocturnal increase in
feeding activity of individual zooplankton (Haney and Hall 1975) must
also have contributed to the higher grazing pressure at night and it is
not possible to separate the relative contribution of that effect.

Estimates of b, s, and r are avilable for each population.

However, grazing rates are based on collection of Chlorella and are thus

not independent estimates for each population. Cyclotella_michiganiana

 

and E. comensis are small centric diatoms whose diameters ranged from
8-16 pm and from 4~10 pm respectively in this study. Gut analyses

showed that both Cyclotella species were readily ingested by Lawrence

 

Lake zooplankton. While some selectivity is likely, the methods
employed do not allow separate estimates of grazing mortality for these
two pOpulations. It is therefore necessary for the current study to

assume both Cyclotella were grazed at the measured rates. In contrast,

 

Sphaerocystis Schroeteri is a large colonial green alga, and colonies

 

commonly exceeded 40 um in diameter. Gut analyses Showed no ingestion

by Lawrence Lake zooplankton (including Daphnia pulex and 23 galeatae

 

12

mendotae) during the period of this study. Sphaerocystis is certainly

 

ingested by large Daphnia (Porter 1976, 1977) and gut analyses of
Daphnia collected from Lawrence Lake in October, 1979 showed substantial
ingestion at that time. However, since ingestion was not found during

the period studied here, g is set equal to zero for Sphaerocystis.

 

In the successional pattern shown in Figure 1A, C. michiganiana was

 

the dominant alga through the end of June at which time 9. comensis

began to increase. During July, 9: michiganiana declined slightly while

 

£3 comensis continued to increase becoming the clear dominant by August.
With the increased grazing pressure in August, there was a rapid decline

of both E: michiganiana and E. comensis followed by an increase of S:

 

Schroeteri.

 

Returning to the conceptual framework established earlier (Eq. 1),
it is now possible to critically evaluate the roles of growth, grazing
and sedimentation in the dynamics of these phytoplankton populations.
Physiological death was never sufficiently significant to measure for
any of these populations and, as stated above, flushing losses were

negligible. The population growth equations can thus be simplified to
F‘= u - g - s (6)

for C. michiganiana and C. comensis, and to

 

T—r-‘p-S (7)

for S3 Schroeteri.

 

These rate equations and corresponding population curves are
plotted for each population in Figures 2, 3 and 4. Growth rates of £3

michiganiana (Figure 2) declined initially, but showed no clear trend

 

after late June. However, sedimentation losses were quite large and

l3

 

 

 

 

 

0.3
0.2
0.1 V'
3
0 a3
5.
q -0.l “I
2 h
x 2 '- "0.2
"I
'2" +
m 1- :§"+.o .3
j ,: $3.
LU .0: 00‘.
u ,5
.00.... ‘90,
‘ JUN ' JUL ' AUG
Figure 2. Population curve (lower figure, ----) and rate budget

(upper figure) for Cyclotella michiganiana showing

 

 

u G--9, g (hatched area), ?'( ), 3 (Open area

between hatching and T), and r(----).

l4

 

J 0.3
0.2
0.1

 

-0.l .
-O.2
-O.3
-0.4

CD
r,r,p,g, or s

 

s: 4'
9
._" 3-
L.
2
Q 2'
LU
L)

‘—

 

 

 

 

JUN ' JUL lAUGj SEP

Figure 3. Population curve (lower figure, ) and rate budget

 

(upper figure) for Cyclotella comensis showing u G—-—-),

 

g (hatched area), F-( ), 3 (open area between hatching

and T), and r (----).

 

15

 

 

 

 

- - 0.2 :
\/\ 0
— '0. - 0.] a:
0nuuuo.... [h
\\\\‘-../" L:
_-—.—-—.—.—.—._. o
m
I
O
r-
F
I I
_l *I
2 /,
m 1
_l 1 — / _
._J I
"J 1!
‘tx’
AUG' SEP '
Figure 4. Population curve (lower figure, - - - -) and rate budget

(upper figure) for Sphaerocystis Schroeteri showing

 

 

u (---), I'— ( ), 3 (open area between D and F), and

r <---->.

 

Figure 5.

16

 

- 0.3
- 002
-‘ 00]

 
    
 

 

i

.'

C)
rorp

'o

C

I

O
‘o
’0

_ -002
- -O.3
l l l I _O.4

 

£5

00

N

CELLS ML“ x10-3

 

 

 

 

Population curves (lower figure) and simplified rate budgets
(upper figure) for Cyclotella michiganiana (--"-),

Cyclotella comensis (

 

) and Sphaerocystis Schroeteri
(- - - -). For each rate budget, the upper line represents

H and the lower line represents r.

l7

effectively offset growth, so that T'was near zero from late June
through July. In August, additional losses due to grazing reduced F'to
less than zero. Growth rates of E: comensis (Figure 3) declined through
most of its growth period but stabilized and then increased as grazing
pressure (and presumably nutrient regeneration) increased in August.
Sedimentation losses were occasionally large but generally insufficient
to offset growth. Consequently, E'remained positive until August when
high grazing losses reduced F’of C: comensis to well below zero. Growth

rates of S3 Schroeteri (Figure 4) showed no clear trends, but

 

sedimentation losses (the only significant loss suffered) never offset
growth so that"? was always positive during the study period.

Given the close correspondence between'? and r in Figures 2, 3, and
4, plots of u and r may be used to examine the role of growth and loss
processes in the seasonal succession of these phytOplankton. The

difference between‘u and r represents net loss (d), such that
r=u"d (8)

These rates and corresponding population curves are plotted in
Figure 5. Although C. comensis had consistently higher growth rates

than C. michiganiana, it is obvious that their succession was not

 

simply due to differential growth. 9: comensis was smaller than 9.

michiganiana and, as might be expected, suffered lower sedimentation

 

losses. It is the combined effect of higher growth rates and lower loss
rates of C. comensis that led to the observed pattern. In contrast, the

C. comensis-S. Schroeteri succession clearly resulted from differential

 

mortality.

These results have interesting implications regarding the relative

 

l8

importance of growth and loss processes in phytoplankton pOpulation
dynamics and thus community structure. In particular, these results
contradict the argument that mortality due to sedimentation and grazing
is generally insignificant and that competition is the major force
controlling phytoplankton p0pulations. For example, the idea that the
epilimnion is a continuously, vigorously mixed layer is widespread in
aquatic ecology. As a result, losses due to sedimentation are generally
thought to be insignificant for the smaller phytoplankton. However, in
Lawrence Lake, temperature profiles indicated that periods of
epilimnetic stagnation were quite common, particularly in midsummer.
During these times, sedimentation losses were quite large and a major
force affecting community structure. Most likely, this loss is true for
a majority of lakes since, particularly in temperate, continental
regions, midsummer includes a number bf rather long periods of very low
winds and very high heat inputs. As a result, periods of epilimnetic
stagnation are probably accompanied by substantial losses of
phytoplankton by sedimentation.

Another widespread idea is that mortality due to grazing is
relatively unimportant in phytoplankton community structure. It is
generally conceded that if grazing pressure were sufficiently high, it
could alter community structure, but it is argued that grazing pressure
is generally low. However, in Lawrence Lake, it is obvious that grazing
mortality was a major factor affecting phytoplankton population dynamics
and resultant community structure during the midsummer period
investigated. While it is not reasonable to generalize on the
importance of grazing from this single detailed study, similar results

of grazing effectiveness exist in the literature at this time. Studies

 

19

which conclude that grazing is insignificant rarely involve actual
measurements of grazing rates. In contrast, when grazing rates have
been measured directly, the results have consistently indicated grazing
to be of significance at certain times of the year.

Competition for nutrients or other resource limitations have long
been suggested as important factors in the seasonal succession of
phytoplankton, although the argument has recently been formalized
(Kilham and Tilman 1979). Field demonstration of the importance of
competition in phytoplankton pOpulation dynamics is sorely lacking.
Based on growth rate data of this study, it is possible to speculate on

the importance of competition in the Cyclotella michiganiana - C.

 

comenis succession in Lawrence Lake.

In order for resource competition to affect succession, growth
rates must be affected. Although the effects of competition might be
delayed because of internal nutrient stores, unless these effects
finally appear as reduced growth rates, no operative role can be argued
for competition. An analysis of the growth rate curve of 93 comensis
suggests possible intraspecific competition. AS the population
increased, growth rates steadily declined. However, with the onset of
high grazing pressures and presumably high rates of nutrient
regeneration in August, growth rates first stabilized and then increased
to near maximum. However, an analysis of growth rate curves (Figure 5)

shows no depression of Cyclotella michiganiana growth rates by E:

 

comensis. Although interspecific competition may have been involved,

its effects did not appear as modifications in growth rate before loss

 

processes eliminated C. michiganiana from the community. The

implications of these results seem obvious; namely it can not be

20

assumed, and should not be expected, that such highly dynamic systems as
lakes will remain stable for sufficiently long time periods for
competitive displacement to occur. This lack of stability is especially
true when the effects of competition are delayed due to internal
nutrient stores which are commonplace in planktonic algae.

Clearly, despite nmch effort the field of phytOplankton pOpulation
dynamics and succession is little understood. However, future
developments will profit most from carefully executed investigations of
natural populations in concert with detailed mechanistic studies in the

laboratory and appropriate theoretical advances.

21

L ITE RATURE CITED

Droop, M. R. 1974. The nutrient status of algal cells in continuous
culture. J. Mar. Biol. Assoc. U. K. 54:825-855.

Dugdale, R. C. 1967. Nutrient limitation in the sea: Dynamics,
identification and significance. Limnol. Oceanogr. 12:685-695.

Greeney, W. J., D. A. Bella, and H. C. Curl. 1973. A theoretical
approach to interspecific competition in phytOplankton
communities. Amer. Natur. 107:405-425.

Haney, J. F. 1973. An in situ examination of the grazing activities

 

of natural zooplankton communities. Arch. Hydrobiol. 72:87-132.

Haney, J. F. and D. J. Hall. 1975. Diel vertical migration and
filter-feeding activities of Daphnia. Arch. Hydrobiol.
75:413-441.

Hargrave, B. T. and N. M. Burns. 1979. Assessment of sediment trap
collection efficiency. Limnol. Oceanogr. 24:124-135.

Infante, A. 1978. Natural food of herbivorous ZOOplankton of Lake
Valencia (Venezuela). Arch. Hydrobiol. 82:347-358.

Kilham, P. and D. Tilman. 1979. The importance of resource
competition and nutrient gradients for phytoplankton ecology.
Arch. Hydrobiol. 13:110-119.

Knoechel, R. and J. Kalff. 1975. Algal sedimentation: The case of a
diatom-blue-green succession. Verh. Internat. Verein. Limnol.

19:745-754.

22

Knoechel, R. and J. Kalff. 1978. An in situ study of the
productivity and population dynamics of five freshwater
planktonic diatom species. Limnol. Oceanogr. 23:195-218;

Lund, W. G. 1949. Studies on Asterionella formosa Hass. I. The

 

origin and nature of the cells producing seasonal maxima. J.
Ecol. 37:389-409.

Lund, W. G. 1950. Studies of Asterionella formosa Hass.

 

II. Nutrient depletion and the spring maximum. J. Ecol.
38:1-35.

Lund, W. G. 1954. The seasonal cycle of the plankton diatom Melosira
italica subsp. subartica 0. Mull. J. Ecol. 42:151-179.

Lund, W. G., C. Kipling, and E. D. LeCren. 1958. The inverted
microscope method of estimating algal numbers and the statistical
basis of estimations by counting. Hydrobiologia 11:143-170.

O'Brien, W. J. 1974. The dynamics of nutrient limitation of
phytoplankton algae: A model reconsidered. Ecology 55:135-141.

Porter, K. G. 1976. Enhancement of algal growth and productivity by
grazing zooplankton. Science 192:1332-1334.

Porter, K. G. 1977. The plant-animal interface in freshwater
ecosystems. Amer. Sci. 65:159-170.

Starkweather, P. L. and J. J. Gilbert. 1978. Feeding in the rotifer

Brachionus calyciflorus. IV. Selective feeding on tracer

 

particles as a factor in trOphic ecology and in situ technique.

 

Verh. Internat. Verein. Limnol. 20:2389-2394.
Tilman, D. 1977. Resource competition between planktonic algae: An

experimental and theoretical approach. Ecology 58:338-348.

In" '

 

23

Wetzel, R. G., P. H. Rich, M. C. Miller, and H. L. Allen. 1972.
Metabolism of dissolved and particulate detrital carbon in a

temperate hard-water lake. Mem. Ist. Ital. Idrobiol. 29 Suppl.:

185-243.

APPENDIX A

DIALYSIS CHAMBERS FOR THE MEASUREMENT OF

PHYTOPLANKTON GROWTH RATE IN SITU

ABSTRACT
A dialysis/cage culture technique is described and its use for

estimating growth rates of natural phytoplankton pOpulations in situ is

 

critically evaluated. Samples from which grazers had been eliminated

were incubated in situ in Plexiglas chambers with Nuclepore

 

polycarbonate membrane side walls. Growth rates of phytoplankton in
these chambers were corrected for measured grazing and sedimentation
losses and compared with actual rates of increase of natural
populations. No Significant differences were found between the two

estimates.

24

25

A major obstacle to the study of phytoplankton population dynamics
has been the lack of suitable methodology. Species-level productivity
measurements have been a particular problem, and as a result,
phytoplankton have commonly been considered a single, homogeneous entity
in terms of productivity. Recent advances using track autoradiography
held promise of extending the 14C-uptake technique to the consideration
of species-level productivity (Knoechel and Kalff 1976). Unfortunately,
the fixatives required for autoradiography cause large and variable
losses of 140 (Silver and Davoll 1978, Brock 1978, Crumpton and Wetzel
unpublished data). In addition, the 14C technique suffers from such
inherent problems as bottle effects, the uncertain relationship between
14C-uptake and particulate carbon fixation, and difficulties in
estimating nighttime respiration and excretion losses.

Given this bleak outlook for meaningful estimates of 14C-uptake, a
more suitable and more direct measure of growth rate might be obtained
from changes in population size. Estimates of growth rates based on
changes in population size have been attempted only rarely for
phytoplankton due to the uncertainties introduced by grazing, sinking,
and other loss processes. In order to calculate growth rates from
population changes, these losses must either be eliminated or measured.

One very promising approach which has been poorly exploited is the
use of dialysis or cage cultures suspended in 31523 With this approach,
grazing and transport losses can be eliminated and mortality from other
processes estimated from changes in numbers of dead cells. The growth
rate of each population can thus be estimated based on changes in the
number of living and dead cells over a specified time interval. Despite

a long history in the study of micro-organisms (reviewed by Schultz and

 

26

Gerhardt 1969), dialysis/cage culture has seldom been applied to the
study of aquatic algae and only recently extended to the study of mixed
natural populations. This area has been recently reviewed by Jensen,
Rystad and Skoglund (1972), Sakshaug and Jensen (1978) and Bide, Jensen
and Melsom (1979), although the works of Braune (1966a & b, 1970a & b,
1972, and 1975), Pierre (1969), Owens SE 31. (1977) and Laake (1978)
were apparently overlooked. While previous studies have provided
considerable information on algal growth in dialysis/cage culture, none
has critically evaluated the use of cage cultures for estimating growth
rates of natural populations. In this paper, I describe an improved
dialysis/cage culture chamber and critically evaluate its use for
estimating growth rates of natural phytoplankton populations in_§i£23
Dialysis/cage culture chambers (Figure l) were constructed from
cast Plexiglas tubing (7.6cm OD, 0.64cm wall thickness). In case of
slight irregularities in the Plexiglas stock, the components of each
chamber were cut as adjoining sections. Holes were drilled through as
shown and fitted with stainless steel or nylon machine screws. A
combined sampling-port-suspension-point was made by tapping a 0.64cm
hole through the central section of the Chamber and fitting it with a
nylon machine screw. These Chambers are similar in design to those
described by MacFeters and Stuart (1972). However the MacFeters and
Stuart chambers were designed for use with axenic cultures of bacteria,
and although they have been used in phytoplankton study (Owens ££_alx
1977; G. A. McFeters personal communication), they are not ideally
suited for that purpose. For applications where sterile conditions are

not required, the chambers described above are more suitable. These

 

 

27

 

Figure 1. Dialysis Chamber

 

28

chambers are easier to fill and empty and their thinner walls cause less
light attenuation and spectral modification.

For assembly (Figure 1), the inside face of each retainer ring is
wetted and pressed to a 76mm dia., 1.0-um pore size Nuclepore
polycarbonate membrane. The membrane will adhere to the ring and these
units are then carefully positioned on each side of the central section
(Figure 1). A dissecting needle or thin wire is then run through each
screw hole, puncturing the membranes to allow the screws to be inserted
smoothly. The screws are then inserted and tightened with moderate
pressure.

Filling the chambers is a simple matter of adding sample through
the port provided and sealing the chamber with the nylon screw. Flow
through 1.0-um and smaller pore size membranes is relatively slow and
handling time is sufficient for most purposes. Chambers may be
transported or held for longer periods if immersed in water.

Diffusion rates into the chambers were measured using 14C-glucose.
Triplicate chambers were suspended in a 10-1 tank of continuously mixed
water containing 100-uCi 14C-glucose. Duplicate 100-ul samples were
taken from each chamber at appropriate intervals and their activities
determined. Transport followed first order kinetics with a half time of
ca. 1 hour.

The chambers were tested for leakage of algae as follows.
Triplicate chambers were filled with dense cultures of Chlorella

pyrenoidosa (mean cell diameter ca. 5-um) and placed in separate, sealed

 

flasks of media. These flasks were shaken vigorously for ca. 48 hours
and the external media examined for algal cells using the method of

Utermbhl (1958). No cells were found in any instance.

 

29

Field studies were conducted in Lawrence Lake, Michigan to evaluate
the use of the dialysis/cage culture chambers for estimating growth
rates of natural phytoplankton populations. POpulatiOn growth in the
chambers was corrected for losses due to grazing and sedimentation and
compared with actual changes of epilimnetic populations. Growth rates,
grazing losses (g), and sedimentation losses (3) were measured as part
of a larger study extending from May through October 1979 (main body
this thesis). Grazing and sedimentation losses were lowest for

Cyclotella michiganiana during June and for Cyclotella comensis during

 

 

July. These periods were chosen to evaluate the use of dialysis/cage
culture chambers for estimating growth rates for these populations since
the effects of errors in grazing and sedimentation measurements would be
minimal at these times. However, the test must still be considered
conservative since some errors are derived from these other
measurements.

Replicate phytoplankton samples were collected twice weekly from a
central station using an integrating tube sampler and were preserved
with 0.4% Lugol's solution. The samples were processed and the
phytoplankton counted using the technique described in Appendix B.

Using a combination of field and strip counts, between 100 and 1000
cells of each species were coUnted for each sample and 95% confidence
limits calculated as in Lund, Kipling, and Le Cren (1958).

At weekly intervals, six replicate dialysis chambers were filled
with lake water from which the dominant grazers had been removed by
filtration through 153-um mesh Nitex® netting. An aliquot of this water
was preserved with 0.4% Lugol's solution for estimates of initial cell

numbers (Co below). The chambers were suspended at the depth of mean

30

epilimnetic light intensity using a surface-tethered buoy to provide
agitation of the chambers via wave action. On day four and again on day
seven, three chambers were collected and the incubated samples were
preserved with 0.4% Lugol's solution. Phytoplankton counts were made as
described above on initial (Co) and incubated (Ct) samples.

Instantaneous growth rates (u) were calculated from

C
t = Coeut

A common problem with any incubation is failure to duplicate the
vertical movements and thus light regime of actual phytoplankton
populations. The importance of such failure is unsettled due to
conflicting reports (cf. Jewson and Wood 1975 with Harris and Piccinin
1977 and Marra 1978). However, the photoinhibitory effect often
observed in productivity studies is in many cases an artifact of
subjecting phytoplankton to extended periods of high light intensity
when they would otherwise have spent only a short time under such
conditions. Samples which are moved through the light-depth gradient

during incubation provide the most realistic estimate of in situ

 

productivity for epilimnetic populations. However, a point exists on
the light-depth profile where the productivity of a stationary sample
approximates that of the mean for the epilimnion, and in Lawrence Lake,
this point is near the depth of mean epilimnetic light intensity
(Wetzel, unpublished data). Despite numerous implicit assumptions,
incubating samples at this depth represents the most reasonable
compromise.

Growth rates (U) obtained using dialysis/cage culture chambers

were combined with estimates of instantaneous grazing (g) and

31

sedimentation losses (8) and used to predict population density

according to

NC = Noe(u-g-S)t.

These predicted population densities (NE) and actual population

densities (Nt) are compared for Cyclotella michiganiana in Figure 2 and

 

Cyclotella comensis in Figure 3. The correspondence of the two curves

 

is striking in each case. These results demonstrate that, under certain
conditions, dialysis/cage culture can indeed be used to estimate growth
rates of natural phytOplankton populations. The limitations to such use
are at present unclear, but at least for epilimnetic phytoplankton
populations the potential is great. The approach has been used in an
investigation of the differential effects of growth and loss processes
in a diatom-green algal succession in Lawrence Lake, Michigan (main body

this thesis).

 

32

 

IIIIIII]IITIII]

1200 - ++—
soo - 4+

400 — */ —

CELLS mI-l

 

 

 

O lllJIiilIIiiLil

14 21 28
JUNE

Figure 2. Predicted (o) and observed(o) densities for Cvclotella

 

michiganiana. Error bars represent 95% confidence limits

 

on phytoplankton counts. Predicted densities have
additional variance due to errors in estimates of

g (mean C.V. ca. 20%) and 3 (mean C.V. ca. 15%).

33

 

TIIIIIIIIIII]IIIT111111111111

2000 —
1500 - -

1000- U -

500 - i4 —

+/

CELLS ml-1

 

 

 

O.11lllllllllllllllllllllllIll

3O 5 1O 15 20 25
JUNE JULY

Figure 3. Predicted (o) and observed (0) population densities for

Cyclotella comensis. Error bars denote 95% confidence

 

limits on phytoplankton counts. Predicted densities have
additional variance due to errors in estimates of g (mean

C.V. ca. 20%) and 3 (mean C.V. ca. 15%).

34

Literature Cited
Braune, W. 1966a. Die Verwendung von Membranfilter-Kapseln zu
experimentellen Studien der WUchsleistung von Mikroorganismen
unmittelbar in Freiland-Gewassern. I. Bergrundung der Methode.
Limnologica 4f245-256.

Braune, W. 1966b. Experimentalle Ermittlung der Zellvermehrung von

 

Scenedesmus obliquus (Turp.) Kruger in Membranfilter-Kapseln
unmittelbar im Flie gewasser. Verh. Internat. Verein. Limnol.
163830-836.

Braune, W. 1970a. ‘In situ-Experimente zur Zellvermehrung von

Scenedesmus obliquus in FlieBgewassern. I. Jahreszyklus.

 

Limnologica 13371-376‘

Braune, W. 1970b. In situ-Experimente zur Zellvermehrung von

Scenedesmus obliquus in FlieBgewassern. II. Vertikalprofil.

 

Limnologica 15377-380°

Braune, W. 1972. Experimental investigations in situ into biomass

 

production of micro-alge and of natural algae biocoenoses in
flowing waters. Int. Revue ges. Hydrobiol. 215227-256.
Braune, W. 1975. Studien zur Algenbesiedlung der Saale im Raum Jena.
II. Vergleich der jahreszeitlichen Besiedlungsdynamik ober- und
unterhalb der Stadt. Limnologica 2f443-480.
Brock, T. D. 1978. Comment On "Loss of 140 activity after Chemical
fixation of phytoplankton" (M. W. Silver and P. J. Davoll).

Limnol. Oceanogr. 22}1083-1084.

35

Eide, I., A. Jensen, and S. Melsom. 1979. Application Of.éE“§i£E
cage cultures of phytoplankton for monitoring heavy metal
pollution in two Norwegian fjords. J. exp. mar. Biol. Ecol.
21f271-286.

Haney, J. F. 1973. An in;§é£2_method for the measurement of
ZOOplankton grazing rates. Limnol. Oceanogr. 16:970-977.

Hargrave, Barry T. and Burns, N. M. 1979. Assessment of sediment
trap collection efficiency. Limnol. Oceanogr. 245124-135.

Harris, G. P. and B. B. Piccinin. 1977. Photosynthesis by natural
phytoplankton populations. Arch. Hydrobiol. 825405-457.

Jensen, A., B. Rystad, and L. Skoglund. 1972. The use of dialysis
culture in phytoplankton studies. J. exp. mar. Biol. Ecol.
§f241-248'

Jewson, D. H. and R. B. Wood. 1975. Some effects on integral
photosynthesis of artificial circulation of phytoplankton through
light gradients. Verh. Internat. Verein. Limnol. 12}1037-1044.

Knoechel, R. and J. Kalff. 1976. Track autoradiography: A method
for the determination of phytoplankton species productivity.
Limnol. Oceanogr. 21f590-596.

Laake, M. 1978. Monitoring the effects of chemical and biological
waste water treatment in_§i£2_by dialysis cultures of freshwater
algae. Mitt. Internat. Verein. Limnol. 223453-472.

Lund, J. W. G., C. Kipling, and E. D. LeCren. 1958. The inverted
microscope method of estimating algal numbers and the statistical
basis of estimations by counting. Hydrobiologia 115143-170.

Marra, J. 1978. Phytoplankton photosynthetic response to vertical

movement in a mixed layer. Mar. Biol. 295203—208°

 

36

McFeters, G. A., and D. G. Stuart. 1972. Survival coliform bacteria
in natural waters: Field and laboratory studies with
membrane-filter chambers. Applied Microbiology: 343805-811.

Owens, 0. v. H., Dresler, P., Crawford, C. C., Tyler, M. A., and
Seliger, H. H. 1977. Phytoplankton cages for the measurement

in situ of the growth rates of mixed natural populations.

 

Chesapeake Science 18(4):325-333.

Pierre, J.-F. 1969. Etude experimentale du comportement in situ
d'une population diatomique maintenue en enceinte dialysante.
Hydrobiologia.33:364-368.

Sakshaug, E. and A. Jensen. 1978. The use of cage cultures in
studies of the biochemistry and ecology of marine phytoplankton.
Oceanogr. Mar. Biol. Ann. Rev. 16581-106.

Schultz, J. S. and P. Gerhardt. 1969. Dialysis culture of
micro-organisms: design, theory and results. Bact. Rev. 3351-47.

Silver, M. W. and P. J. Davoll. 1978. LoSs of 14C activity after
chemical fixation of phytoplankton: Error source for
autoradiography and other productivity measurements. Limnol.
Oceanogr..23:362-367.

UtermOhl, H. 1958. Zur Vervollkommnung der quantitativen

Phytoplankton-Methodik. Mitt. Int. Ver. Limnol. 2} 38p.

 

 

APPENDIX B

A METHOD FOR THE PREPARATION OF PERMANENT, QUANTITATIVE

MOUNTS OF PHYTOPLANKTON FOR CRITICAL MICROSCOPY

ABSTRACT
A method is described for the preparation of permanent, quantitative
mounts of phytoplankton for critical microscopy. Preserved
phytoplankton are settled onto cover glasses (or slides), dehydrated
using ethanol vapor substitution, and mounted in one of several ethanol-
soluble resins. The resulting slides are permanent, quantitative and of
unsurpassed optical quality. Resolution is excellent and, in most
cases, contrast so high as to eliminate the need for phase shift or

other contrast enhancement.

37

 

38

The inverted microsc0pe technique of Uterm6hl (1958) has allowed
quantitative analyses of even the most delicate u-plankton, and as a
result, has become the standard by which other techniques are judged.
Nonetheless, the Utermbhl technique does suffer a number of
shortcomings. Among the most crucial are the need for a suitably
equipped inverted microscope, the lack of a permanent record, and the
poor optical quality provided by long working distance condensers.

The obvious need for a better technique has not escaped previous
investigators, and several have attempted improvements. Sanford 35.31:
(1969) developed a settle-freeze method which has received some
application but probably lacks the accuracy necessary for general
acceptance. Sanford §£_313 (1969) reported unexplained discrepancies
between the settle-freeze method and that of Utermfihl (1958). Although
they do not discuss the possibilities, spillage during slide preparation
and morphological artifacts resulting from freezing are potential
sources of error.

A more promising approach was introduced by Coulon and Alexander
(1972) and later improved by Knoechel and Kalff (1976). This method is
less manipulative than that of Sanford 55 El. (1969) and contains fewer
potential sources of error. In addition to the well known advantages of
gentle sedimentation, the approach provides permanent and quantitative
mounts (Coulon and Alexander, 1972; Knoechel and Kalff, 1978; this
study). However, the water-soluble mounting media used by previous
investigators leave much to be desired. This problem was recognized by
Coulon and Alexander (1972), who sUggested the possibility of mounting
in alcohol- or xylene-soluble media. In addition, previous

investigators have settled phytOplankton directly onto slides despite

 

 

 

 

 

39

the considerable Optical advantages of settling them onto cover glasses.

In this paper, I describe the preparation of quantitative, resinous
mounts of phytoplankton for critical microscopy. Preserved
phytoplankton are settled directly onto cover glasses (or slides) using
the sliding-chambers develOped by Knoechel and Kalff (1976). The
supernatant is then removed and the algae are dehydrated using ethanol
vapor substitution. Once in ethanol, the algae can be mounted in any of
several ethanol-soluble resins. The resulting slides are permanent,
quantitative, and of excellent optical quality. While the entire
process takes several days, only 5-10 minutes actual work is required
per slide.

I acknowledge the helpful discussions of E. F. Stoermer, R.
Knoechel and A. J. Stewart, and the machining skills of W. Wetzel. J.
Kalff kindly provided sample chambers from which our design was taken.

The settling chambers (Figure 1) are based on the design of
Knoechel and Kalff (1976). With this design, it is possible to settle
phytoplankton directly onto cover glasses rather than slides, a
considerable advantage for critical microscopy. It is a simple matter
of inverting the bottom plate to use cover glasses rather than slides.
These chambers are easily constructed from Plexiglas® stock, or may be
obtained from Total Plastics Inc., 5271 Wynn Road, Kalamazoo, Michigan
49001. This firm now has complete specifications and will manufacture
chambers to order at a unit price of approximately $35.

The vapor chambers are essentially thin layer Chromotography
desiccators manufactured from DZ" (6.4mm) Plexiglas® (Figure 2). A final
cut across the face of the chamber provides a squared surface for

sealing the door. Studs are tapped into the face and the door secured

 

4O

 

 

 

 

 

 

 

 

Figure l. Settling chambers showing positions for (A) normal Filling
and settling, (B) filling for nonturbulent rinse and (C)

dumping overlying water.

 

 

 

 

41

 

Figure 2. Empty vapor chamber.

42

with knurled nuts. A layer of silicone grease or a ne0preme gasket
between the door and the chamber face provides an inert and airtight
seal. These chambers are available from Total Plastics Inc. at a unit
price of approximately $50.

Euparal, pleurax (Hanna 1949, von Stosch 1974), and novolacs
(Appendix C) have all proven to be useful mounting media. Euparal is a
proprietary mixture based on gum sandarac. It is miscible with 95%
ethanol, has a refractive index of 1.48, and may be obtained
commercially.

Pleurax is a synthetic resin first described by Hanna (1949). It
is miscible with 95% ethanol and has a refractive index of 1.75.

Pleurax is not available commercially but its synthesis (von Stosch
1974) is relatively simple. In this procedure, 110 parts phenol, 40
parts sulfur, and 0.1 part anhydrous sodium carbonate are refluxed at
180-190°C until all sulfur is reacted (5-9 hrs.). In my preparations
(as in Hanna 1949), excess phenol is driven off by further heating and
the resinous product dissolved in absolute ethanol.

Novolacs are a class of phenolic resins recently suggested as
mounting media for biological materials (Appendix C). They are miscible
with ethanol and have refractive indices greater than 1.6. I have
successfully used Union Carbide® BRPB 5215, a commercially available
novolac which is miscible with 60% ethanol and has a refractive index of
1.65. In my preparations, the stock resin is recrystalized from ethanol
to remove excess phenol, and the dried product is dissolved in absolute
ethanol.

Chambers are assembled as shown in Figure 1 using either slides or

#1 1/2 cover glasses (the bottom plate is inverted when using cover

 

 

43

glasses). For extremely critical work, it is best to use only
premeasured cover glasses of appropriate thickness (0.17mm for most
objectives). Each chamber is filled with preserved sample, and capped
with a cover glass (Figure 1A). Alternatively, for reasons discussed
below, fresh samples may be used and preservative added directly to the
chamber. If leakage is a problem, a layer of silicone grease may be
placed between the towers and their guide grooves. As with any settling
chamber it is imperative that a filling tube (UtermOhl 1958) or some
means be used to ensure the random distribution of settled organisms.
The filled chambers are placed in a moist atmosphere and the organisms
allowed to settle for at least 4 hr per cm of chamber height. A moist
atmosphere is necessary to prevent evaporative loss and possible
leakage. After settling is complete, the overlying water is removed by
sliding the chamber towers just past the edge of the chamber plates
(Figure 1C). I routinely follow this with a non-turbulent rinse of
deionized water for those samples where crystals might later form (high
concentrations of Lugol's, CaC03, or NaCl). For rinsing, each chamber
is positioned as shown in Figure 1B, filled with deionized water and
recapped. The towers are then very carefully slid back into position
over the settled material. After allowing time for diffusive mixing and
for settling of any resuspended organisms, the overlying water is again
removed as described above.

The settling chambers are next placed into vapor Chambers (Fig. 2)
and the towers carefully removed from the chamber plates. An open dish
of absolute ethanol (minimum of 15 ml per sample) is placed inside the
vapor chamber and the door immediately sealed. At least 72 hrs are

allowed for vapor substitution, after which the settling chambers are

 

44

transferred to shallow desiccators. In order to ensure that the
distribution of settled organisms is not affected, cover glasses are
positioned over each depression during the transfer. The cover glasses
are then slid away and the excess ethanol evaporated (takes 2-3 hours),
being careful not to dry out the organisms. Just as each sample
approaches dryness, the cover glass is slid back over the depression to
prevent further evaporation. In a dry atmosphere, the concentration of
ethanol on the organisms will not fall below the azeotrOpe of 95.6%,
assuming at least this concentration to start. Alternatively, settling
chambers may be left in the vapor chamber and the dish of ethanol
replaced with silica gel. The distribution of settled organisms is well
protected, but subsequent steps become more difficult and time
consuming. Once the excess ethanol has been removed, chambers are
disassembled and the organisms mounted in a suitable resin. Curing is
dictated by the nature of the sample and resin so that no set procedure
can be recommended.

Euparal and the novolac BRPB 5215 both give excellent preservation
of cellular and subcellular structures. In addition, contrast is
sufficient in either that flagella, plastids, pyrenoids, pectic
processes, and cell walls can be distinguished without resorting to
phase-shift or other contrast enhancement techniques. In BRPB 5215,
diatom frustules are also clearly visible, but in euparal, contrast is
insufficient without enhancement. Euparal has the additional
disadvantage that mounts remain fluid and thus relatively delicate while
BRPB 5215 mounts soon harden sufficiently to endure considerable
manipulation. Also, if for some reason mounts contain less than 95%

ethanol, BRPB 5215 mounts are still usable while euparal mounts become

45

milky and near worthless. Euparal does have the advantage that it has
been in use for many years and is known to be relatively permanent
wherease novolacs have only recently been introduced as mounting media.
However, novolacs are inert compounds and have proven extremely durable
as lacquers and varnishes. If care is taken to remove excess phenol
from the stock resin, BRPB 5215 mounts can be expected to last
indefinitely.

Pleurax gives excellent results with diatoms but with very little
else. In many cases, overall plastid structure is obscured, apparently
by increased prominence of substructures. This may be the result of
increased contrast of these substructures at higher refractive indices
or it may be that pleurax acts as a clearing agent. In addition, many
structures, including flagella and some cell walls and processes, lack
sufficient contrast in pleurax. In either case, structures needed for
identification of many algae are not readily visible in pleurax.

The accuracy of the sliding-chamber technique has been demonstrated
by previous investigators (Coulon and Alexander 1972, Knoechel and Kalff
1978). In order to test the accuracy of the procedures described here
and to further test the general approach, comparisons were made between
this technique and that of Utermbhl (1958). Samples were taken from two
small, hardwater lakes and preserved with 0.4% Lugol's solution. Four
replicate Utermbhl chambers and four replicate slides were prepared for
each sample. Using a combination of field and strip counts, at least
100 of each of the dominant species were counted for each replicate.

The results (Table 1) demonstrate no significant difference in accuracy
between the two techniques.

The most desirable technique is one Which makes it possible to

 

46

Table 1. Mean number of organisms ml"l + S.D. of four replicates.

 

 

 

__ Utermbhl Slides
Little Mill Lake - August 1979
Cyclotella comta 99 :_19 Q? + 11
Lawrence Lake - October 1978
Dinobryon divergens 747 :_53 787 + 58
- June 1979
Cyclotella michiganiana 1100 + 80 1160 + 50

 

- September 1979

RhodOmonas minuta 570 + 60 600 + 60

 

 

 

47

observe algae in near natural condition with not only intact cellular
and subcellular structures but also intact colonies and chains.

Problems often arise when samples have been preserved for any length of
time. Among other things, gelatinous linkages are weakened and the
vigorous agitation needed to homogenize stored samples can disrupt
colonies and especially loosely bound chains of diatoms. As a result,
valuable information is lost, leaving an incomplete and unrealistic
picture. Solitary cells are certainly not ecologically equivalent to
those in colonies or chains, differing with regard to nutrient dynamics,
sedimentation, grazing, and parasitism. Fortunately, it is relatively
simple to overcome this problem. Rather than preserving samples and
later subsampling, fresh samples are added directly to settling chambers
and then preserved or else freshly preserved samples are used. This may
also be done with UtermOhl chambers but their expense and impermanency
make such use impractical. In contrast, the technique described above
is ideally suited for such application and is a considerably more

practical approach to exacting phytoplankton analyses.

48

REFERENCES

Coulon, C., and V. Alexander. 1972. A sliding-chamber phytoplankton
settling technique for making permanent quantitative Slides with
applications in fluorescent microscopy and autoradiography.
Limnol. Oceanogr. 17:149-152.

Hanna, G. D. 1949. A synthetic resin which has unusual properties.
J. R. Microsc. Soc. 69:25-28.

Knoechel, R., and J. Kalff. 1976. Track autoradiography: A method
for the determination of phytOplankton species productivity.
Limnol. Oceanogr. 21:590-596. \

, and . 1978. An in_situ study of the

 

productivity and population dynamics of five freshwater
planktonic diatom species. Limnol. Oceanogr. 23:195-218.

Sanford, G. R., A. Sands, and C. R. Goldman. 1969. A settle-freeze
method for concentrating phytoplankton in quantitative studies.
Limnol. Oceanogr. 14:790-794.

von Stosch, H.-A. 1974. Pleurax: its synthesis and application to
the mounting and clearing for cell walls of diatoms,
dinoflagellates and other algae, and its use in a new method of
electively staining dinoflagellate armours. Arch. Protistenk.
116:132-141.

Utermbhl, H. 1958. Zur Vervollkommnung der quantitativen

Phytoplankton-Methodik. Mitt. Int. Ver. Limnol. 9.38p.

 

 

 

APPENDIX C

NOVOLACS: SYNTHETIC RESINS SUITABLE FOR MOUNTING BIOLOGICAL MATERIALS1

ABSTRACT
Novolacs are transparent, highly refractive (nD > 1.6), alcohol-
soluble resins. A commercially available novolac (Union Carbind® BRPB
5215, nDZO = 1.65) was used successfully in preparing mounts of
freshwater algae and microcrustacea. Additional applications for these

resins are apparent.

1 For further information including availability contact Union Carbide

Corporation, 270 Park Avenue, New York, N.Y. 10017.

49

 

50

Novolacs are transparent, highly refractive (nD > 1.6), phenolic
resins soluble in alcohol, acetone, esters, ethers, and to some extent
hydrocarbons (Roff & Scott, 1971). These polymers are permanently
fusible with melting points above 80°C. Since commercial production
began in 1910-1912 (as Bakelite), novolacs have been used as lacquers
and varnishes, and as bonding agents in filler-reinforced plastics.

Beyond their current applications, the solubility and optical
characteristics of novolacs suggested their potential utility in the
preparation of biological materials for light microscopy. To explore
this potential, I selected a commercially available resin (Union
Carbide® BRPB 5215)l soluble in 60% to 100% ethanol. Its refractive
index was determined by immersion in reference liquids (nD20 = 1.65).
The commercial product contains free phenol and other contaminants which
I remove by recrystalizing the novolac from ethanol with excess water.
The precipitate is collected by centrifugation, dried, and dissolved in
absolute ethanol.

The solubility and optical characteristics noted are useful in
preparing many kinds of biological materials for light microscopy. I
have had considerable success mounting fresh-water algae and
microcrustaceans in BRPB 5215. In my preparations, preserved organisms
are dehydrated using ethanol vapor substitution (Sanford £5.3l3 1969)
and mounted directly in a drop of resin-ethanol solution with a cover
glass. Curing is dictated by the nature of the sample, but if allowed
to sit at room temperature, mounts harden at their edges within a few
hours. Although the permanency of these mounts is unknown, novolacs
have proven to be durable lacquers and varnishes. If one is careful to

remove excess phenol from the stock resin, problems with deterioration

 

51

seem unlikely. All that can be said at present is that mounts of
diatoms prepared one year ago and mounts of green algae and cryptomonads
prepared six months ago have not deteriorated in any noticeable aspect.
Preservation of cellular and subcellular structures is excellent and
contrast is sufficient that flagella, plastids, pyrenoids, pectic
processes, frustules, and cell walls can be distinguished without
resorting to phase-shift or other contrast enhancement techniques. In
addition to excellent contrast and resolution, these mounts possess
useful fluorescent properties. Background in blue-light fluorescence is
acceptable and autofluorescence of many algae is preserved (E. F.
Stoermer, personal communications).

No commercial mounting media combines the solubility and optical
characteristics of BRPB 5215. Of those available, only euparal is
soluble in ethanol and unfortunately is both eutectic and has a much

lower refractive index (nD = 1.48) than BRPB 5215.

 

52

LITERATURE CITED
Roff, W. J. and Scott, J. R. 1971. Fibers, Films, Plastics and

Rubbers: A Handbook of Common Polymers, Butterworths, London.

 

688 pp.
Sanford, C. R., A. Sands, and C. R. Goldman. 1969. A settle-freeze
method for concentrating phytoplankton in quantitative studies.

Limnol. Oceanogr. 14:790-794.

 

L IIIIIII

"lfllifijgfujliigljillly1'111.1

5 0022