ECOPHYSIOLOGY OF THE TOXIC CYANOBACTERIUM MICROCYSTIS AERUGINOSA
ACROSS NUTRIENT GRADIENTS
By
Geoffrey Horst

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Fisheries and Wildlife – Doctor of Philosophy
Ecology, Evolutionary Biology and Behavior – Dual Degree
2013

ABSTRACT
ECOPHYSIOLOGY OF THE TOXIC CYANOBACTERIUM MICROCYSTIS AERUGINOSA
ACROSS NUTRIENT GRADIENTS
By
Geoffrey Horst
Harmful algal blooms (HABs) are a significant and growing problem worldwide in both
freshwater and marine ecosystems. In many freshwater systems, cyanbacteria blooms can cause
significant risks to public health due to the production of potent toxins. An important objective
in managing harmful phytoplankton blooms is determining what factors influence the toxicity of
bloom-forming species. In this dissertation, I combined lab and field experiments along with
field measurements in Lake Erie to determine whether the production of microcystin (a potent
liver toxin) by the cyanobacterium Microcystis is strongly affected by nutrient limitation. The
field and lab experiments demonstrated that nitrogen-limited conditions led to significantly
reduced microcystin quota relative to phosphorus-limited or nitrogen and phosphorus saturated
conditions. Results an extensive seasonal survey in the western basin of Lake Erie revealed a
similar influence of nitrogen availability on microcystin quota, indicating that nitrogen limitation
appears to have a robust influence on the toxin quota of Microcystis. These results are consistent
with stoichiometric considerations in that production of nitrogen-rich secondary metabolites,
such as microcystins, should be reduced disproportionately under nitrogen limitation. The
results of this work may be used to inform public decision making about managing nitrogen and
phosphorus inputs into freshwater lakes.

ACKNOWLEDGMENTS

I would like to thank several individuals who helped me during my graduate school career at
Michigan State University. First, I would like to thank my wife, Katie Horst, who patiently
supported my efforts to pursue a PhD and even helped with some of the technical research work
at the Kellogg Biological Station. Second, I would like to thank my committee members: Drs.
Pat Soranno, Scott Peacor, Steve Hamilton, Elena Litchman and especially Orlando (Ace)
Sarnelle. Finally, I would like to acknowledge the help of several individuals who helped me
with my field research, including: Joel Berry, Joe Northrup, Reni Kaul, Jeff White, Juli Dyble,
Dave Weed, Nina Consolatti, Sherry Martin, Paul Horst and Jennifer Horst.

iii

TABLE OF CONTENTS

LIST OF FIGURES

v

CHAPTER 1
Nitrogen availability increases the toxicity of the harmful cyanobacterium Microcystis
aeruginosa
INTRODUCTION
METHODS
Gull Lake enclosure experiment
Laboratory experiment
Lake Erie surveys
mcyB gene copy density
Statistics
RESULTS
Gull Lake enclosure experiment
Laboratory experiment
Lake Erie Surveys
DISCUSSION
FIGURES
REFERENCES

1
1
5
5
9
11
12
13
14
14
15
16
17
21
27

CHAPTER 2
Phosphorus limits Microcystis aeruginosa abundance in eutrophic western Lake Erie
INTRODUCTION
METHODS
RESULTS
DISCUSSION
FIGURES
REFERENCES

33
33
37
40
44
49
56

iv

LIST OF FIGURES

Figure 1.1 Total phosphorus, chlorophyll a (Chl a), soluble reactive phosphorus
(SRP), bulk alkaline phosphatase activity (APA) and nitrate-nitrogen
(NO3-) concentrations in the Gull Lake enclosure experiment (all
units in ug/L). Values represent means ± SE (n = 10 enclosures per
treatment).

21

Figure 1.2 (A) Microcystis biomass (dry weight) and (B) microcystin toxin quota
(ug toxin per mg Microcystis dry biomass) across nutrient treatments
in the Gull Lake enclosure experiment on August. 6, 2007 (Day 31).
Values represent means ± SE (n = 10). Bars sharing the same letter
(A, B) were not significantly different following Bonferonni adjusted
post-hoc analysis at p = 0.05.

22

Figure 1.3 Microcystin per mcyB gene (pg/106 genes) across nutrient treatments
in the Gull Lake enclosure experiment on August 6, 2007 (Day 31).
Values represent means ± SE (n = 10). Bars sharing the same letter
(A, B, C) were not significantly different following Bonferonni
adjusted post-hoc analysis at p = 0.05.

23

Figure 1.4 Microcystin toxin quota (ug toxin per mg Microcystis dry biomass)
across nutrient treatments in the monoclonal lab experiment. Values
represent means ± SE (n = 8 for N-limited, n = 4 for P-limited and
WC media). Bars sharing the same letter (A, B) were not
significantly different following Bonferonni adjusted post-hoc
analysis at p = 0.05.

24

Figure 1.5 (A) Microcystis dry biomass; and (B) microcystin toxin quota, nitrate
(NO3-) and total phosphorus (TP) concentrations throughout the
summer of 2008 in the western basin of Lake Erie. (C) Measured
microcystin levels (solid line) and theoretical toxin levels based on
using a fixed toxin quota of 3.1 ug/mg (average quota from 22 July).
Values are mean ± SE averaged across sites within the basin on each
date.

25

Figure 1.6 Microcystin toxin quota (ug toxin per mg Microcystis dry biomass)
as a function of NO3- (ug N/l) in: (A) the enclosure experiment; (B)
the lab experiment with Microcystis strain; and (C) mean values on a
given date from sites in the western basin of Lake Erie (points from
left to right represent 22-Sep, 3-Sep, 18-Aug, 4-Aug, and 22-Jul.

26

Figure 2.1 Map of the western basin of Lake Erie and locations of the survey
sampling stations. Map derived from Chaffin et al. (2011).

49

v

Figure 2.2 Taxon-specific alkaline phosphatase activity using enzyme-labeled
fluorescence (ELF) and epi-fluorescence microscopy. Chlorophyll a
in cells fluoresces orange-red, while expression of alkaline
phosphatase activity is represented by bright-green precipitates
formed by the alkaline phosphatase enzyme. Panel A and B illustrate
colonies of Pediastrum and Microcystis aeruginosa that are not
expressing ELF, while panels C and D illustrate colonies expressing
ELF. For interpretation of the references to color in this and all other
figures, the reader is referred to the electronic version of this
dissertation.

50

Figure 2.3 (Top) Microcystis aeruginosa (ug dry biomass/L) and chlorophyll a
(ug/L); (middle) total phosphorus (ugP/L) and nitrate (ugN/L); and
(bottom) water temperature (oC) and Secchi disk depth (m)
throughout the summer of 2008 in the western basin of Lake Erie.
Values are means ± SE averaged across sites within the basin on each
date.

51

Figure 2.4 Total phosphorus (ug P/L) and Microcystis aeruginosa biomass (ug
dry weight/L) in relation to distance from the mouth of the Maumee
River in the western basin of Lake Erie. Trend lines are plotted for
significant correlations (p <0.05).

52

Figure 2.5 Microcystis aeruginosa biomass (ug dry weight/L) as a function of
total phosphorus (ugP/L) during the sampling dates leading up to
peak Microcystis abundance (Solid circles: 22 Jul, 4 Aug, 18Aug and
3 Sep) and after the bloom (Open circles: 22 Sep). Trend line
represents the linear regression of log10 transformed biomass and
log10 transformed phosphorus concentrations for the sampling period
leading up the bloom (22 Jul – 3 Sep).

53

Figure 2.6 Microcystis aeruginosa biomass (ug dry weight/L) as a function of
bulk alkaline phosphatase activity (APA) (nM/ug chlorophyll
a/L/min) during the sampling dates leading up to peak Microcystis
abundance (Solid circles: 22 Jul, 4 Aug, 18Aug and 3 Sep) and after
the bloom (Open circles: 22 Sep). Trend line represents a linear
regression of log10 transformed biomass and log10 transformed APA
for the sampling period leading up to the bloom (22 Jul – 3 Sep).

54

Figure 2.7 The fraction of Microcystis aeruginosa colonies expressing taxonspecific enzyme labeled fluorescence (ELF) of alkaline phosphatase
as a function of bulk alkaline phosphatase activity (APA) (nM/ug
chlorophyll a/L/min) for the entire seston community during on 4
Aug, 18 Aug, 3 Sep and 22 Sep. Spearman rank correlation
coefficient, rs = 0.337, t = 2.06, df = 33, p = 0.023.

55

vi

CHAPTER 1

Nitrogen availability increases the toxicity of the harmful cyanobacterium
Microcystis aeruginosa
INTRODUCTION
Harmful phytoplankton blooms (HABs) are a major environmental concern in both
marine and freshwaters throughout much of the globe (Huisman et al. 2005) and the impacts of
HABs on both ecosystems and human health (Chorus & Bartram 1999; Carmichael 2001; Smith
& Schindler 2009) are well documented. One of the most critical aspects of HAB ecology is the
production of toxins by HAB species during blooms. In many cases, the impact of HABs on
ecosystem and human health is directly related to the concentration of phytoplankton-produced
toxins in the water. As one noteworthy example, concentrations of microcystin, a potent liver
toxin produced by Microcystis aeruginosa (hereafter Microcystis), can exceed the World Health
Organization guideline for drinking water during the summer months in Lake Erie which is the
drinking-water source for four million people (Fuller et al. 2002).
The concentration of toxins in the water is a function of both the biomass of the toxin
producing species and the toxin quota (i.e. the amount of toxin per unit biomass). Toxin quota is
a highly variable trait within a species. For example, among genotypes within some species,
-1

toxin quota can range from essentially zero to over 5 µg toxin mg dry weight (Lukač 1993;
Sivonen & Jones 1999). The implication of a highly variable toxin quota is that blooms with
very high biomass and low toxin quota may actually pose a smaller concern (with respect to
toxin concentrations in the water) than a bloom with less biomass but higher toxin quota.
Therefore, understanding what factors contribute to variation in toxin quota is fundamental to
understanding the impacts of HABs.
1

Microcystis is a cosmopolitan cyanobacterium that develops toxin producing blooms in
both temperate and tropical freshwaters. Within populations of Microcystis, toxin quota can vary
by either: a) changes in the relative abundance of more toxic vs. less toxic genotypes (i.e.
intraspecific competition); or b) due to changes in toxin production induced by biological and/or
physical factors (i.e. phenotypic plasticity). For example, Briand et al. (2009) showed dramatic
spatio-temporal changes in the relative abundance of toxin producing genotypes of Microcystis
in a large reservoir over the course of the growing season. Kardinaal et al. (2007) demonstrated
in lab experiments that non-toxic strains of Microcystis outcompeted toxic strains under light
limited conditions and light limitation induced by self-shading might explain a commonly
observed temporal trend in some lakes for blooms to become less toxic (toxin per unit biomass)
over time.
Alternatively, many laboratory studies indicate that microcystin quota is a plastic trait
and can be a function of growth rate (Watanabe et al. 1989; Orr & Jones 1998; Oh et al. 2000),
nutrient concentration (Oh et al. 2001; Downing et al. 2005; Van de Waal et al. 2009a),
temperature (Westhuizen 1985), light levels (Utkilen and Gjølme 1992; Wiedner et al. 2003) and
grazers (Jang et al. 2003). With respect to Microcystis, several lab-scale studies have shown a
strong positive correlation between nitrogen concentration (or N:P ratio) and toxin quota (Orr &
Jones 1998; Long et al. 2001). However, these data are confounded by the fact that toxin quota
was also positively correlated with growth rate as a result of increased nitrogen availability.
More recent studies have parsed out the effects of growth rate and confirmed that toxin quota is
positively affected by increased nitrogen levels (or N:P ratio) independent of growth rate (Lee et
al. 2000; Oh et al. 2000; Downing et al. 2005; Van de Waal etal 2009). Jahnichen et al. (2011)

2

additionally reports the limiting effects of sulphur concentrations on microcystin quota as well as
an inverse relationship between quota and phosphorus concentrations.
Demonstrating statistically-significant effects of environmental factors in tightlycontrolled lab experiments with clonal monocultures is relatively straightforward given the high
precision possible in the laboratory. However, it remains to be demonstrated which of many
potentially-important factors has major effects on toxin quotas in natural populations, since the
latter are genetically diverse (Briand et al. 2009) and embedded within complex communities of
interacting species and genotypes.
Within natural phytoplankton communities where Microcystis is a dominant species,
there is evidence that microcystin concentrations can be correlated with nitrogen and/or
phosphorus concentrations (Graham et al. 2004; Rinta-Kanto et al. 2009; Orihel et al. 2012).
Furthermore, Gobler et al. (2007) demonstrated with nutrient amendment assays to lake water
samples that supplemental nitrogen late in the growth season led to increased toxin
concentrations.
Only a handful of studies have specifically focused on toxin quotas in natural systems.
Welker et al. (2003) documented a general pattern where the highest quotas occurred early in the
season but decreased with increasing Microcystis abundance later in the season, but provided
only relatively minor details about the nutrient environment. A survey of lakes in Quebec with
generally low toxin concentrations (< 1 µg/L microcystin) revealed a positive correlation
between total nitrogen and toxin quota (Graham et al. 2004). By contrast, a multi-year survey of
hyper-eutrophic lakes in Alberta showed that microcystin toxin quota was significantly
positively correlated with phosphorus concentrations but inversely correlated with nitrate (NO3)

3

and ammonium concentrations. Clearly there is a need for more work on the controls of
microcystin production by cyanobacteria in natural systems.
In this study, we examined the role of nutrient limitation in controlling the microcystin
quota of Microcystis by combining the results of a large-scale enclosure lake enclosure
experiment, a single-genotype lab experiment and an extensive field survey of the western basin
of Lake Erie. Results from all three approaches were congruent in terms of the positive effect of
nitrogen availability on microcystin quota. Furthermore, genetic analyses of the microcystin
gene (mcyB) from the enclosure experiment suggests that phenotypic plasticity (rather than
changes in gene copy density) was largely responsible for this effect. This observed positive
effect of nitrogen availability on the toxicity of Microcystis is particularly notable because the
high N:P ratios in most freshwater lakes indicate that excess available nitrogen is common
(Downing & McCauley 1992), particularly in landscapes with intensive row-crop agriculture
where anthropogenic nitrogen inputs considerably exceed biological demand (Arbuckle &
Downing 2001).

4

METHODS
Gull Lake enclosure experiment
2

Gull Lake is a large, hardwater lake (surface area = 8.2 km , mean depth = 12 m,
maximum depth = 31 m) in southwestern Michigan adjacent to Michigan State University’s
Kellogg Biological Station. The lake is oligotrophic with average summer total phosphorus (TP)
-1

-1

of 8 µg L and chlorophyll a of 4 µg L . Ammonium levels are typically very low (less than 15
-1

-1

µg L ) but NO3 levels are usually above 300 µg L due to significant external contributions
from the watershed (Bruesewitz et al. 2012), and phytoplankton growth is P-limited (Hamilton et
al. 2009). Sulfate is readily available and concentrations show little seasonal variability (mean,
-1

23 mg L : Hamilton unpublished data). Zebra mussels invaded the lake in 1994, and since then
Microcystis biomass levels have been significantly higher than pre-invasion (Sarnelle et al.
-1

2005). Microcystin concentrations in the mixed layer range up to 0.5 µg L , which is about 8
times higher than a lake of comparable total phosphorus without zebra mussels (Sarnelle et al.
2010). The lake is ideal for experimental investigations of microcystin quota since Microcystis
is the only significant microcystin producer and its low P and high nitrate-N concentrations allow
manipulation of N and P availability by simply adding additional P.
During the summer of 2007 we conducted an enclosure experiment in the lake to test for
effects of P addition and Dreissena polymorpha (zebra mussel) density on the abundance and
toxin-quota of Microcystis. The enclosures were deployed in the same manner as described in
Hamilton et al. (2009). In brief, enclosures were constructed from transparent polyethylene
tubing (2 m diameter,10 m long, total volume ~ 31,000 L) that was heat-sealed along the bottom
end and open at the top to allow exposure to the atmosphere. Thirty enclosures were suspended
5

from a floating dock that was anchored in ~15 m of water. Over the course of 2 days (27-28
June), the enclosures were filled with lake water pumped from 4 m depth (mixed layer = ask
Jeff) and passed through a 150 µm mesh net to remove crustacean zooplankton that might have
been killed by the pump. Natural densities of zooplankton were then re-established on 29 June
by pooling repeated zooplankton tows (1 m diameter, 150 µm mesh) from the lake epilimnion
(7-8 m depth) and allocating equal aliquots to the enclosures.
Enclosures were randomly assigned to three P treatments intended to simulate ambient
-1

oligotrophic (no P addition, TP ~8-10 µg L , referred to as “low P”), mesotrophic (TP amended
-1

-1

to ~15 µg L , “medium P”) and mildly eutrophic (TP amended to ~25 µg L , “high P”)
conditions. We added P alone to simulate a eutrophication gradient since the N:P ratio in lakes
typically decreases with increases in P (Downing et al. 2005) and since Gull Lake already has
high NO3. Over the course of 7 days (29 June – 5 July), we dripped a concentrated P solution
(as NaH2PO4) into the enclosures to reach the target TP levels. We subsequently maintained
target TP levels via weekly monitoring and addition of additional phosphate solution as needed.
Within each set of 10 enclosures for each P treatment, we also established a gradient of
-2

Dreissena ranging from 0 – 4 g m (dry tissue mass) on 5 July (day 0 of the experiment). We
found no effects of Dreissena density on microcystin quota (as we had found earlier: Knoll et al.
2008), total phytoplankton biomass, concentrations of available nutrients, or indices of
phytoplankton nutrient limitation, so we focus on the effects of phosphorus treatment in this
paper. Dreissena had a positive influence on Microcystis biomass in the enclosures, as reported
previously (Vanderploeg et al. 2002; Sarnelle et al. 2005; Knoll et al. 2008). Response variables
were averaged across the Dreissena densities for all analyses reported in this paper.
6

We sampled for phytoplankton and nutrients weekly in the enclosures by pooling
duplicate depth-integrated tube samples (2.5 cm inside diameter) from the surface to 7 m depth
(approximate depth of mixed layer). Sampled water was stored in cleaned cubitainers and
transported in coolers to a lakeside laboratory, where samples were initially processed within one
hour into four sub-samples: 1) raw water sample stored at 4 C; 2) filtered water sample passed
through an A/E glass fiber filter and stored at 4 C; 3) seston collected on A/E filters and
immediately frozen at -20 C; and 4) water preserved in Lugol’s solution for microscope
counting of the phytoplankton.
All nutrient analyses were performed within 48 hours of collection. Soluble reactive P
(SRP) and ammonium (NH4) were measured on filtered water samples using long-pathlength
spectrometry (Wetzel and Likens 2000). NO3 was measured on filtered water samples by
Dionex membrane-suppression ion chromatography. TP was measured on persulfate-digested
raw water samples and then analyzed spectrophotometrically as SRP. Chlorophyll was extracted
from frozen filters in 90% ethanol and measured fluorometrically (Welschmeyer 1994).
Particulate microcystins were extracted from frozen filters with 75% methanol and then diluted
1:4 with deionized water before analysis by Enzyme-Linked ImmunoSorbent Assay (ELISA)
using a commercial kit (Envirologix, Inc.). The detection limit for this protocol was ~0.02 µg L
1

.
Microcystis biomass was measured via inverted-microscope analysis on Lugol’s

preserved samples taken on the last day of the experiment (day 31). Sub-samples (50 ml) were
transferred into 25 mm-wide settling chambers with a cover-slip bottom. After a 5-day settling
7

-

period, the entire chamber surface was surveyed for Microcystis colonies and the surface area of
each colony was measured at 100X magnification from digitized photographs (SPOT Software V
4.0.9, Diagnostic Instruments). The cumulative colony surface area for a chamber was converted
-1

into Microcystis cell concentration (cells L ) based on the surface area to cell count relationship
established for Microcystis samples in Lake Erie (R2 = 0.95, n = 21) after confirming that cell
packing density within Gull Lake colonies was not significantly different than that within Lake
Erie colonies. Microcystis dry biomass concentration was then calculated by converting cell
concentration to biovolume (based on measurements of cell dimensions) and subsequently
3

converting biovolume to dry biomass assuming a dry weight (mg) to biovolume (mm ) ratio of
-1

0.42. Microcystin toxin quota (µg mg dry weight) was calculated by dividing the particulate
-1

-1

microcystin concentration (µg L ) by Microcystis dry biomass concentration (mg L ).
Bulk-alkaline phosphatase activity (APA), a measure of phosphorus limitation of the
planktonic community, was measured fluorometrically in 10 ml aliquots of freshly-collected raw
water samples from the enclosures (Pettersson 1980). Microcystis-specific alkaline phosphatase
activity was assessed using enzyme-labeled fluorescence (ELF) microscopy (Gonzalez-Gil and
Aguileral 1998). This microscopy-based method is analogous to bulk alkaline phosphatase
analysis, whereby phosphorus-limited organisms produce the enzyme alkaline phosphatase
which is then detected visually for cells of a particular species. ELF substrate (0.1 ml, 1:20 v:v in
buffer solution; E6601 from Invitrogen, USA) was added to separate 10 ml subsamples of
freshly-collected raw water samples from the enclosures. After 1 hr, the treated subsample was
filtered onto black 1 µm membrane filters and rinsed with phosphate-buffered saline to inhibit

8

further enzyme activity. The filters were then permanently mounted onto microscope slides and
stored at -20 C for up to three months before analysis.
Using a Nikon Eclipse E600 epifluorescence microscope, the mounted membrane filters
were visually scanned for Microcystis colonies and evidence of alkaline phosphate activity as
indicated by bright green precipitates on the surface of colonies. Each Microcystis colony (up to
200 per slide) was counted and marked as having either positive or negative enzyme activity,
with the percent of active colonies per sample serving as a relative index of the population’s
alkaline phosphatase activity.

Laboratory experiment
To clarify the mechanistic effect of N versus P limitation on the microcystin quota of
Microcystis in Gull Lake, we conducted a laboratory experiment in which a Microcystis strain
from the lake was exposed to varying nitrogen versus phosphorus availability. We used a toxinproducing strain of Microcystis (E1-6) that was isolated from Gull Lake in 2007. In contrast to
most culture-collection strains, our strain had retained the typical colony-forming phenotype
when the laboratory experiment was conducted.
Before the experiment, Microcystis strain E1-6 was grown in four 2-L culture vessels
containing sterilized half-strength WC-S media (Stemberger 1981), with an initial N:P ratio of
20:1 (250 µM N as KNO3, 12.5 µM P as K2HPO4). These culture vessels were grown under
constant light (PAR of 100 µmol m s ) with no supplemental aeration at 19 C. After two
-2 -1

weeks the culture vessels were pooled together and 250 ml of this seed inoculum was distributed
to each of 16 500-ml sterilized flasks. The flasks were randomly assigned to one of three

9

nutrient treatments where the relative concentrations of N and P were modified from the half
strength WC media recipe: 1) P-limited (250 µM N: 2.5 µM P [100:1]); 2) N-limited (31.25 µM
N: 12.5 µM P [2.5:1]); and 3) half-strength WC media. The treatment flasks were filled to 450
ml with their respective media and set under the same fluorescent lighting on a 12hr:12hr
light:dark cycle at 19 C.
After a six-day initial growth period, we initiated a semi-continuous dilution regime of
-1

0.1 d to maintain desired nutrient availabilities in the treatments. Dilutions were achieved by
removing (after thorough mixing) 90 ml (i.e. 20%) of culture every two days and aseptically
replacing with the respective media for each treatment. To ensure that inorganic carbon
availability was not limiting, we included two sets of four flasks for the N-limited treatment with
the additional set receiving aeration (to alleviate potential C limitation) starting on day 34. On
day 50 (from the start of the dilution regime), we ended the experiment and sampled the flasks
for nutrients, microcystin and Microcystis. In the N-limited treatment, there was no significant
difference in biomass or microcystin levels between flasks that were aerated and not, therefore
the results for all eight flasks were pooled.
Chlorophyll a, microcystin toxin and SRP were analyzed in the same fashion as the
-

enclosure experiment. Concentrations of NO3 were analyzed with 2

nd

derivative UV

spectroscopy (Crumpton et al. 1992) on a Perkin Elmer Lambda 20 UV/Vis spectrophotometer.
Microcystis biomass was estimated from Lugol’s preserved samples following a NaOH digestion
(Reynolds and Jaworski 1978) to break the colonies into individual cells for easier counting.

10

Lake Erie surveys
We conducted an extensive seasonal survey of the western basin of Lake Erie, the most
eutrophic of the Laurentian Great Lakes, from July through September of 2008. Nutrient
concentrations in the basin are heavily influenced by loading from the Maumee River (TP
-1

typically above 200 µg L ) that enters from the southwest. The Maumee contributes more than
half of the phosphorus in the southern and western portions of the basin where the Microcystis
blooms are a recurring water quality problem (Schwab et al. 2009; Chaffin et al. 2011;
Bridgeman et al. 2012). Ten sampling stations, including one in the mouth of the Maumee
River, were selected to provide a broad representation across the basin, with stations spaced
approximately 6-10 km apart. Sampling trips were conducted at intervals of approximately two
weeks, starting on 9 May and continuing through 22 September in order to capture seasonal
changes in the physiology of Microcystis before, during and after the annual bloom.
On each sampling trip, water samples from each station were collected at 0.5 and 1 m
depth with a van Dorn bottle and pooled into clean 10 L cubitainers. The western basin of Lake
Erie is considered to be completely mixed during the summer due to the generally shallow
depths and large fetch for wind mixing. Temperature profiles confirmed that the lake was
generally isothermal (i.e. < 1 C difference from top to bottom) and therefore subsurface
samples were assumed to be to be representative of the water column. Secchi disk depth was
recorded at each station and temperature, light, pH, conductivity, in situ phycocyanin and
dissolved oxygen profiles were taken from at least four stations on each sampling trip.
Water samples were kept at ~ 4 C immediately following collection and during transport
back to the Michigan State University campus in East Lansing, MI. Samples for chlorophyll a,

11

microcystin and nutrients were processed within 12 hr of collection and analyzed in the same
-

manner (with NO3 via spectrophotometry) as described for the enclosure and lab experiments.
Microcystis biomass was measured via microscopic analysis of Lugol’s preserved samples as
described for the enclosure experiment. Microcystin quota was then calculated as particulate
toxin divided by Microcystis biomass. Since toxin quota is highly sensitive to the biomass
estimate in the denominator, we only included samples for which Microcystis was relatively
-1

common (biomass concentrations > 1 mg L dry weight) in our statistical analyses. This
criterion excluded five, before-bloom, samples from 22 July.

mcyB gene copy density
We processed Microcystis samples collected from the Gull Lake experimental enclosures
to assess whether treatment effects on microcystin quota were associated with effects on the
frequencies of the mcyB gene, which is required for microcystin production (Kurmayer et al.
2002). Protocols for quantifying mcyB gene copies are described in detail by Dyble et al.
(2008).
Briefly, water samples from the enclosure were filtered on 0.8 µm membrane filters to
collect Microcystis. The filters were stored at -80 oC until all the samples could be processed.
Cells were then lysed using DNAzol and DNA was extracted using standard phenol-chloroform
procedures. The mcyB gene was amplified for quantification in 20 µL volumes using an
Applied Biosystems® 7500 Fast Real-Time PCR system. Amplification used fast conditions
with an initial denaturing step of 95 C for 10 minutes followed by 40 cycles of 3 seconds at 95

C and 30 seconds at 60 C.

The standard curve for absolute quantification was derived from

12

Hudson Bay strain E4-13, but relative differences are sufficient for our objectives (Kurmayer and
Kutzenberger 2003). The critical threshold values for the samples were compared to the
standard curve panning five orders of magnitude (10^2 – 10^6 genes/filer) using Applied
Biosystems SDS software.

Statistics
All statistical analyses were performed using Systat 9 software. In some cases, values
were log transformed to correct for homogeneity of variances. Comparisons among treatments
in the experiments were tested following the Bonferonni correction for multiple comparisons.

13

RESULTS
Gull Lake enclosure experiment
We were successful in establishing and maintaining TP near target levels in the
-1

-1

-1

enclosures (i.e. low ~ 8-10 µg L , medium ~ 15-20 µg L and high ~ 25 µg L ) (Fig. 1.1).
-1

Soluble reactive phosphorus (SRP) was less than 2 µg L for the low and medium P treatments,
-1

while concentrations were marginally higher and reached 4 µg L in the high P treatment by day
31. Phytoplankton biomass (measured as chlorophyll a) increased markedly in response to the P
addition treatments (Fig. 1), as expected given that Gull Lake is strongly P limited (Hamilton et
al. 2009). After the initial strong response to P addition, chlorophyll declined in the P-addition
treatments, and this decline was most pronounced in the high-P treatment (Fig. 1.1).
-1

At the beginning of the experiment, NO3 was ~300 µg N L and remained relatively
unchanged throughout the experiment in the low-P treatment (i.e. no phosphorus addition, Fig.
1.1). NO3 concentrations in P-amended treatments decreased and by day 24 had fallen to
-1

undetectable levels in the high P treatment and to less than 150 µg L in the medium-P
-1

treatment. Ammonium levels remained less than 20 µg L in all treatments throughout the
experiment.
Bulk alkaline phosphatase activity for the planktonic community was inversely related to
ambient phosphorus throughout the experiment, with enzyme activity in the low phosphorus
treatment typically 3-4 times higher than in the medium and high phosphorus treatments (Fig.
1.1). At the taxon-specific level, the relative alkaline phosphatase activity for Microcystis
14

measured with enzyme-labeled fluorescence microscopy was also significantly higher with 34%
of colonies expressing activity in the low P treatment vs. 8% in the high P treatment by day 31
(ANOVA F2,27 4.16, p = 0.027). These data indicate that phosphorus addition alleviated P
limitation in the phytoplankton generally, and in Microcystis specifically.
Microcystis biomass on day 31 was affected by both Dreissena (ANOVA F1,24 7.99, p<
0.01) and P addition (Fig. 1.2a; ANOVA F2,24 4.22, p = 0.027) but with no significant
interaction. The response of Microcystis biomass to P addition was similar to that observed for
chlorophyll on day 31, with highest abundance in the medium-P treatment (Figs. 1.1, 1.2). More
importantly, the microcystin quota of Microcystis (µg of toxin per mg of Microcystis biomass)
was negatively affected by P addition by the end of the enclosure experiment (Fig. 1.2; ANOVA
F2,27 3.73, p = 0.037), and thus varied positively with available NO3 concentration (Figs. 1.1,
1.2)
Density of the mcyB gene was highest in the enclosures with the lowest available N (high
P treatment), indicating no selection for strains with low copy numbers at lower NO3.
Microcystin per mycB gene copy was highest at high NO3 (low P treatment) and lowest at low
NO3 (high P treatment, Fig. 1.3).

Laboratory experiment
Microcystin quota of the Gull Lake strain of Microcystis in the laboratory experiment was
significantly affected by nutrient treatments. (Fig. 1.4; ANOVA F3,12 7.87, p = 0.004). Toxin
quota was lowest in the N limited treatment and highest in the P limited treatment. Microcystis
15

grown in media replete with N and P (WC media) had the same quota as the high N/low P
treatment. These results indicate that N availability, not P availability, was the primary nutrientmediated factor influencing toxin quota.

Lake Erie Surveys
During the Lake Erie survey in 2008, Microcystis was the only significant microcystin
producing taxon present. Anabaena was present at some sites but always at levels less than 1%
of the total phytoplankton biomass. Averaging across the western basin, Microcystis biomass
-1

increased by over two orders of magnitude during August, reaching a peak of 3500 µg L on 3
September before declining in late September (Fig. 1.5). Basin-wide microcystin quota declined
more or less continuously, and by more than two fold, from July to September. The decline in
quota coincided with a ten-fold decline in NO3 from late July to late September (Fig. 1.5). Total
-1

-1

phosphorus increased from about 33 µg L to 82 µg L during this same period (Fig. 1.5).
Thus, results from the western basin of Lake Erie are congruent with results from the field and
laboratory experiments in showing a negative relationship between microcystin quota and NO3
availability.

16

DISCUSSION
In this paper we have presented multiple lines of evidence supportive of a major role for
N availability in mediating the toxin quota Microcystis, one of the most important
cyanobacterium species responsible for harmful algal blooms in freshwaters. Results from a field
experiment in large enclosures showed a strong association between reduced NO3 concentration
(generated by our P additions) and reduced microcystin quota (Figs. 1.2, 1.6). This relationship
was shown to be consistent with observations from a seasonal survey of the western basin of
Lake Erie showing concomitant declines in NO3 and microcystin quota (Figs. 1.5, 1.6). Genetic
evidence from the enclosure experiment suggested that decreases in microcystin quota were not
driven by decreases in copy density of the microcystin-producing gene, mcyB, but rather more
likely the result of phenotypic changes presumably mediated by N availability (Fig. 1.3). Taken
together, these observations suggest a strong influence of N availability on cellular production of
microcystin in nature, but do not demonstrate that low N is the cause of low quota.
To demonstrate cause and effect and phenotypic plasticity in Microcystis, we followed up
field results with a laboratory experiment in which we exposed a single genotype of Microcystis
from Gull Lake to strong N and P limited conditions. This experiment demonstrated that cells
grown under N limitation had about a 2.5 fold lower toxin quota compared to N replete
conditions, whereas P limitation had no detectable effect on toxin quota (Fig. 1.4). Thus, we
conclude that N availability is a likely a strong causative driver of microcystin quota in natural
populations of Microcystis, and that this effect may largely be mediated by phenotypic responses
by individuals cells..
We established a qualitatively consistent pattern of increased microcystin quota with
2

increased available N across a spectrum of study systems (less than 1 liter to over 100 km ).
17

Quantitatively, we compared across these study scales by regressing toxin quota on available
NO3. As expected, there was a significant positive linear relationship between quota and NO3
for each of the study systems (Fig. 1.6). Despite large differences among study systems, the
range of microcystin quota was fairly constrained across the three systems (~1 - 3 µg toxin per
mg Microcystis dry biomass).
The positive effect of N availability on microcystin quota that we observed is largely
consistent with previous laboratory research (Lee et al. 2000; Oh et al. 2000; Downing et al.
2005; Van de Waal et al 2009). More recently, Janichen et al. (2011) found that microcystin
quota decreased with increasing P (as we observed in the enclosure experiment). Our laboratory
experiment showed, however, that the P "effect" in our enclosures was more likely directly
driven by N availability, not P availability. In lakes, increasing P concentrations typically shift
systems towards N limitation (Downing and McCauley 1992), which may have driven the
decrease in quota observed by Janichen et al. (2011).
A number of studies have shown that the proportion of the Microcystis population
containing toxin producing genes can vary in space and time (Briand et al. 2009, Otten et al.
2012), including in the western basin of Lake Erie (Rinta-Kanto et al. 2009). Briand et al. (2012)
suggest that the toxicity or toxin quota of a bloom can be explained by variation in the proportion
of toxic vs. non-toxic strains of the Microcystis population. If selection for strains of Microcystis
more capable of toxin production was causing the increase in toxin quota at higher N levels, then
we would have expected to observe a higher relative mcyB gene frequency (i.e. mycB genes per
Microcystis biomass). In fact, the relative mcyB gene frequency was actually higher in the
enclosures with the lowest available N (i.e. the high P treatment).

18

An alternative explanation is that toxin quota is a function of phenotypic plasticity caused
by a change in the abiotic or biotic environment. In this case, the toxin produced per mycB gene
was over 8 fold higher in the low-P (high NO3) vs. high-P (low NO3) enclosures (Fig. 1.3). This
evidence suggests a high plasticity in toxin production physiology mediated by N availability
and is consistent with the stoichiometric theory that production of nitrogen-rich secondary
metabolites, such as microcystins, should be reduced disproportionately under N limitation
(Sterner & Elser 2002; Van de Waal et al. 2009). Furthermore, Van de Waal et al. (2009) also
found that the most nitrogen-rich variants of microcystin were disproportionately reduced under
N limiting conditions.
Microcystis biomass levels and the microcystin quota both contribute to the overall toxin
concentrations in the water at a given location and time. Low biomass levels combined with
high toxin quotas may result in the same toxin concentrations as a large bloom with lower toxin
quotas. In our survey of the western basin of Lake Erie, microcystin concentrations were
2

strongly correlated with the dynamics of Microcystis biomass (r > 0.9) despite declining
microcystin quota. This indicates that biomass was the major driver of microcystin levels (Fig.
1.5). However, the role of seasonally-variable microcystin quota (and thus presumably N
availability as a driver) is also important, as illustrated by a simple hypothetical example.
During peak Microcystis biomass in early September, microcystin levels in the western basin
-1

peaked at ~3 µg L , or three times the WHO limit for drinking water (Fig. 1.5) despite a
-1

relatively low microcystin quota for Microcystis (~1.5 µg mg ). If the quota had remained at
-1

levels observed in July (~3 µg mg ), perhaps as a result of higher N availability, microcystin
-1

levels might have reached over 6 µg L , or six times the WHO limit. Of course, we do not yet
19

understand how an increase in N availability in the basin would affect the ecological success of
Microcystis relative to non-microcystin containing taxa, so further research on the competitive
dynamics of Microcystis under varying N:P ratios is needed.
The results of this work have potential implications for the monitoring and management
of toxic Microcystis blooms. Current nutrient control strategies in freshwaters focus almost
exclusively on P, supported by decades of research demonstrating the tight coupling between P
and eutrophication in lakes (Schindler 1974; Schindler et al. 2008). However, Conley et al.
(2009) argue that N and P control is needed for many lakes since P flux with the sediments can
support large blooms of non nitrogen-fixing cyanobacteria, like Microcystis, when nitrogen is
readily available. We present the first conclusive experimental evidence for the role of N
availability in determining toxin quota in a natural lake environment and that evidence is
supported by lab culture experiments, genetic analyses and surveys in eutrophic Lake Erie. The
widespread N pollution of watersheds and recipient freshwaters (Carpenter et al. 1998) could
therefore be leading to the increased toxicity of harmful algal blooms, even under conditions of P
limitation. Thus, the control of N as well as P may be necessary to ameliorate harmful
phytoplankton blooms that produce toxins with ecological and public health impacts.

20

FIGURES

Figure 1.1 Total phosphorus, chlorophyll a (Chl a), soluble reactive phosphorus (SRP), bulk
alkaline phosphatase activity (APA) and nitrate-nitrogen (NO3) concentrations in the Gull Lake
enclosure experiment (all units in ug/L). Values represent means ± SE (n = 10 enclosures per
treatment).

21

Figure 1.2 (A) Microcystis biomass (dry weight) and (B) microcystin toxin quota (ug toxin per
mg Microcystis dry biomass) across nutrient treatments in the Gull Lake enclosure experiment
on August. 6, 2007 (Day 31). Values represent means ± SE (n = 10). Bars sharing the same
letter (A, B) were not significantly different following Bonferonni adjusted post-hoc analysis at p
= 0.05.
22

6

Figure 1.3 Microcystin per mcyB gene (pg/10 genes) across nutrient treatments in the Gull
Lake enclosure experiment on August 6, 2007 (Day 31). Values represent means ± SE (n = 10).
Bars sharing the same letter (A, B, C) were not significantly different following Bonferonni
adjusted post-hoc analysis at p = 0.05.

23

Figure 1.4 Microcystin toxin quota (ug toxin per mg Microcystis dry biomass) across nutrient
treatments in the monoclonal lab experiment. Values represent means ± SE (n = 8 for Nlimited, n = 4 for P-limited and WC media). Bars sharing the same letter (A, B) were not
significantly different following Bonferonni adjusted post-hoc analysis at p = 0.05.

24

Figure 1.5 (A) Microcystis dry biomass; and (B) microcystin toxin quota, nitrate (NO3) and total
phosphorus (TP) concentrations throughout the summer of 2008 in the western basin of Lake
Erie. (C) Measured microcystin levels (solid line) and theoretical toxin levels based on using a
fixed toxin quota of 3.1 ug/mg (average quota from 22 July). Values are mean ± SE averaged
across sites within the basin on each date.
25

Figure 1.6 Microcystin toxin quota (ug toxin per mg Microcystis dry biomass) as a function of
NO3 (ug N/l) in: (A) the enclosure experiment; (B) the lab experiment with Microcystis strain;
and (C) mean values on a given date from sites in the western basin of Lake Erie (points from
left to right represent 22-Sep, 3-Sep, 18-Aug, 4-Aug, and 22-Jul.
26

REFERENCES

27

REFERENCES

Arbuckle, K., and J. Downing. 2001. The influence of watershed land use on lake N: P in a
predominantly agricultural landscape. Limnology and Oceanography 970–975.
Briand, E., M. Bormans, C. Quiblier, M.-J. Salençon, and J.-F. Humbert. 2012. Evidence of the
cost of the production of microcystins by Microcystis aeruginosa under differing light and
nitrate environmental conditions. PloS one 7: e29981.
Briand, E., N. Escoffier, C. Straub, M. Sabart, C. Quiblier, and J.-F. Humbert. 2009.
Spatiotemporal changes in the genetic diversity of a bloom-forming Microcystis aeruginosa
(cyanobacteria) population. The ISME journal 3: 419–29.
Bridgeman, T. B., J. D. Chaffin, D. D. Kane, J. D. Conroy, S. E. Panek, and P. M. Armenio.
2012. From River to Lake: Phosphorus partitioning and algal community compositional
changes in Western Lake Erie. Journal of Great Lakes Research 38: 90–97.
Bruesewitz, D. a., J. L. Tank, and S. K. Hamilton. 2012. Incorporating spatial variation of
nitrification and denitrification rates into whole-lake nitrogen dynamics. Journal of
Geophysical Research 117: 1–12.
Carmichael, W. W. 2001. Health effects of toxin-producing cyanobacteria: “The CyanoHABs”.
Human and Ecological Risk Assessment 7: 1393–1407.
Carpenter, S., N. Caraco, D. Correll, R. Howarth, A. Sharpley, and V. Smith. 1998. Nonpoint
pollution of surface waters with phosphorus and nitrogen. Ecological Applications 8: 559–
568.
Chaffin, J. D., T. B. Bridgeman, S. a. Heckathorn, and S. Mishra. 2011. Assessment of
Microcystis growth rate potential and nutrient status across a trophic gradient in western
Lake Erie. Journal of Great Lakes Research 37: 92–100.
Chorus, I., and J. Bartram. 1999. Toxic cyanobacteria in water: a guide to their public health
consequences, monitoring and management. 416.
Conley, D. J., H. W. Paerl, R. W. Howarth, D. F. Boesch, S. P. Seitzinger, K. E. Havens, C.
Lancelot, and G. E. Likens. 2009. Controlling eutrophication: nitrogen and phosphorus.
Science 323: 1014–1015.
Crumpton, W. G., T. M. Isenhart, and P. D. Mitchell. 1992. Nitrate and organic N analyses with
second-derivative spectroscopy. Limnology and Oceanography 37: 907–913.
Downing, J. A., and E. McCauley. 1992. The nitrogen:phosphorus relationship in lakes.
Limnology and Oceanography 37: 936–945.
28

Downing, T. G., C. Meyer, M. M. Gehringer, and M. van de Venter. 2005. Microcystin content
of Microcystis aeruginosa is modulated by nitrogen uptake rate relative to specific growth
rate or carbon fixation rate. Environmental toxicology 20: 257–62.
Dyble, J., G. L. Fahnenstiel, R. W. Litaker, D. F. Millie, and P. A. Tester. 2008. Microcystin
concentrations and genetic diversity of Microcystis in the lower Great Lakes.
Environmental Toxicology 23: 507–516.
Fuller, K., H. Shear, and J. Wittig. 2002. The Great Lakes: an environment atlas and resource
book., US EPA, Government Canada, Toronto, Canada.
Gonzalez-Gil, S., and A. Aguileral. 1998. Detection and quantification of alkaline phosphatase in
single cells of phosphorus-starved marine phytoplankton.
Graham, J. L., J. R. Jones, S. B. Jones, J. A. Downing, and T. E. Clevenger. 2004. Environmental
factors influencing microcystin distribution and concentration in the Midwestern United
States. Water Research 38: 4395–4404.
Hamilton, S. K., D. a. Bruesewitz, G. P. Horst, D. B. Weed, and O. Sarnelle. 2009. Biogenic
calcite-phosphorus precipitation as a negative feedback to lake eutrophication. Canadian
Journal of Fisheries and Aquatic Sciences 66: 343–350.
Huisman, J., H. C. P. Matthijs, and P. M. Visser. 2005. Harmful cyanobacteria, Vol 3. J.
Huisman, H.C.P. Matthijs, and P. Visser [eds.]. Springer.
Jahnichen, S., B. Long, and T. Petzoldt. 2011. Microcystin production by Microcystis
aeruginosa : Direct regulation by multiple environmental factors. Harmful Algae 12: 95–
104.
Jang, M.-H., K. Ha, G.-J. Joo, and N. Takamura. 2003. Toxin production of cyanobacteria is
increased by exposure to zooplankton. Freshwater Biology 48: 1540–1550.
Kardinaal, W. E. a, L. Tonk, I. Janse, S. Hol, P. Slot, J. Huisman, and P. M. Visser. 2007.
Competition for light between toxic and nontoxic strains of the harmful cyanobacterium
Microcystis. Applied and environmental microbiology 73: 2939–46.
Knoll, L. B., O. Sarnelle, S. K. Hamilton, C. E. H. Scheele, A. E. Wilson, J. B. Rose, and M. R.
Morgan. 2008. Invasive zebra mussels (Dreissena polymorpha) increase cyanobacterial
toxin concentrations in low-nutrient lakes. Canadian Journal of Fisheries and Aquatic
Sciences 65: 448–455.
Kurmayer, R., E. Dittmann, J. Fastner, and I. Chorus. 2002. Diversity of microcystin genes
within a population of the toxic cyanobacterium Microcystis spp. in Lake Wannsee (Berlin,
Germany). Microbial ecology 43: 107–18.

29

Kurmayer, R., and T. Kutzenberger. 2003. Application of Real-Time PCR for Quantification of
Microcystin Genotypes in a Population of the Toxic Cyanobacterium Application of RealTime PCR for Quantification of Microcystin Genotypes in a Population of the Toxic
Cyanobacterium Microcystis sp . Applied and Environmental Microbiology 69: 6723–6730.
Lee, S. J., M.-H. Jang, H.-S. Kim, B.-D. Yoon, and H.-M. Oh. 2000. Variation of microcystin
content of Microcystis aeruginosa relative to medium N:P ratio and growth stage. Journal of
Applied Microbiology 89: 323–329.
Long, B. M., G. J. Jones, and P. T. Orr. 2001. Cellular microcystin content in N-limited
Microcystis aeruginosa can be predicted from growth rate. Applied and Environmental
Microbiology 67: 278–283.
Lukač, M. 1993. Influence of trace metals on growth and toxin production of< i> Microcystis
aeruginosa</i>. Toxicon 31: 293–305.
Oh, H., S. Lee, J. Kim, and H. Kim. 2001. Seasonal variation and indirect monitoring of
microcystin concentrations in Daechung Reservoir, Korea. Applied and Environmental
Microbiology 67: 1484–1489.
Oh, H.-M., S. J. Lee, M.-H. Jang, and B.-D. Yoon. 2000. Microcystin Production by Microcystis
aeruginosa in a Phosphorus-Limited Chemostat. Applied and Environmental Microbiology
66: 176–179.
Orihel, D. M., D. F. Bird, M. Brylinsky, H. Chen, D. B. Donald, D. Y. Huang, A. Giani, D.
Kinniburgh, H. Kling, B. G. Kotak, P. R. Leavitt, C. C. Nielsen, S. Reedyk, R. C. Rooney,
S. B. Watson, R. W. Zurawell, and R. D. Vinebrooke. 2012. High microcystin
concentrations occur only at low nitrogen-to-phosphorus ratios in nutrient-rich Canadian
lakes. 1462: 1457–1462.
Orr, P., and G. Jones. 1998. Relationship between microcystin production and cell division rates
in nitrogen-limited Microcystis aeruginosa cultures. Limnology and Oceanography
Otten, T. G., H. Xu, B. Qin, G. Zhu, and H. W. Paerl. 2012. Spatiotemporal patterns and
ecophysiology of toxigenic microcystis blooms in Lake Taihu, China: implications for
water quality management. Environmental science & technology 46: 3480–8.
Pettersson, K. 1980. Alkaline phosphatase activity and algal surplus phosphorus as phosphorusdeficiency indicators in Lake Erken. Archiv fur Hydrobiologie 89: 54–87.
Reynolds, C. S., and G. H. M. Jaworski. 1978. Enumeration of natural Microcystis populations.
British Phycological Journal 13: 269–277.
Rinta-Kanto, J. M., E. A. Konopko, J. M. DeBruyn, R. A. Bourbonniere, G. L. Boyer, and S. W.
Wilhelm. 2009. Lake Erie Microcystis: relationships between microcystin production,

30

dynamics of genotypes and environmental parameters in a large lake. Harmful Algae 8:
665–673.
Sarnelle, O., J. Morrison, R. Kaul, G. Horst, H. Wandell, and R. Bednarz. 2010. Citizen
monitoring: testing hypotheses about the interactive influences of eutrophication and mussel
invasion on a cyanobacterial toxin in lakes. Water Research 44: 141–150.
Sarnelle, O., A. E. Wilson, S. K. Hamilton, L. B. Knoll, and D. E. Raikow. 2005. Complex
interactions between the zebra mussel, Dreissena polymorpha, and the harmful
phytoplankter, Microcystis aeruginosa. Limnology and Oceanography 50: 896–904.
Schindler, D. 1974. Eutrophication and recovery in experimental lakes: implications for lake
management. Science 184: 897–899.
Schindler, D. W., R. E. Hecky, D. L. Findlay, M. P. Stainton, B. R. Parker, M. J. Paterson, K. G.
Beaty, M. Lyng, and S. E. M. Kasian. 2008. Eutrophication of lakes cannot be controlled by
reducing nitrogen input: results of a 37-year whole-ecosystem experiment. Proceedings of
the National Academy of Sciences of the United States of America 105: 11254–8.
Schwab, D. J., D. Beletsky, J. DePinto, and D. M. Dolan. 2009. A hydrodynamic approach to
modeling phosphorus distribution in Lake Erie. Journal of Great Lakes Research 35: 50–60.
Sivonen, K., and G. J. Jones. 1999. Cyanobacterial toxins, p. 41–111. In I. Chorus and J. Bartram
[eds.], Toxic cyanobacteria in Water: A guide to Their Public Health Consequences,
Monitoring, and Management. E & FN Spoon, London.
Smith, V. H., and D. W. Schindler. 2009. Eutrophication science: where do we go from here?
Trends in ecology & evolution 24: 201–7.
Stemberger, R. S. 1981. A general approach to the culture of planktonic rotifers. Canadian
Journal of Fisheries and Aquatic Sciences 38: 721–724.
Sterner, R. W., and J. J. Elser. 2002. Ecological Stoichiometry, Princeton University.
Utkilen, H., and N. Gjølme. 1992. Toxin Production by Microcystis aeruginosa as a Function of
Light in Continuous Cultures and Its Ecological Significance. Applied and Environmental
Microbiology 58: 1321–1325.
Vanderploeg, H. A., T. F. Nalepa, D. J. Jude, E. L. Mills, K. T. Holeck, J. R. Liebig, I. A.
Grigorovich, and H. Ojaveer. 2002. Dispersal and emerging ecological impacts of PontoCaspian species in the Laurentian Great Lakes. Canadian Journal of Fisheries and Aquatic
Sciences 59: 1209–1228.
Van de Waal, D. B., J. M. H. Verspagen, M. Lürling, E. Van Donk, P. M. Visser, and J.
Huisman. 2009a. The ecological stoichiometry of toxins produced by harmful

31

cyanobacteria: an experimental test of the carbon-nutrient balance hypothesis. Ecology
letters 12: 1326–35.
Van de Waal, D. B., J. M. H. Verspagen, M. Lürling, E. Van Donk, P. M. Visser, and J.
Huisman. 2009b. The ecological stoichiometry of toxins produced by harmful
cyanobacteria: an experimental test of the carbon-nutrient balance hypothesis. Ecology
letters 12: 1326–35.
Watanabe, M. F., K.-I. Harada, K. Matsuura, M. Watanabe, and M. Suzuki. 1989. Heptapeptide
toxin production during the batch culture of two Microcystis species (Cyanobacteria).
Journal of Applied Phycology 1: 161–165.
Welker, M., H. von Dohren, H. Tauscher, C. E. W. Steinberg, and M. Erhard. 2003. Toxic
Microcystis in shallow lake Muggelsee (Germany) - dynamics, distribution, diversity.
Archiv fur Hydrobiologie 157: 227–248.
Welschmeyer, N. A. 1994. Fluorometric analysis of chlorophyll a in the presence of chlorophyll
b and phaeopigments. Limnology and Oceanography 39: 1985–1992.
Westhuizen, A. 1985. Effect of temperature and light on the toxicity and growth of the bluegreen alga Microcystis aeruginosa (UV-006). Planta 163: 55–59.
Wetzel, R. G., and G. E. Likens. 2000. Limnological Analyses, 3nd ed. Springer-Verlag.
Wiedner, C., P. M. Visser, J. Fastner, J. S. Metcalf, G. A. Codd, and L. R. Mur. 2003. Effects of
light on the microcystin content of Microcystis strain PCC 7806. Applied and
Environmental Microbiology 69: 1475–1481.

32

CHAPTER 2

Phosphorus limits Microcystis aeruginosa abundance in
eutrophic western Lake Erie
INTRODUCTION
Harmful algal blooms (HABs) are a significant and growing problem worldwide in both
freshwater and marine ecosystems (Chorus and Bartram 1999; Huisman et al. 2005; Paerl and
Huisman 2008). In many freshwater systems, cyanbacteria blooms can cause significant risks to
public health due to the production of potent toxins (Carmichael 1994, 2001; Chorus and
Bartram 1999). The most cited causes for HABs relate to excessive nutrient loading, especially
phosphorus (Carpenter et al. 1998; Paerl and Huisman 2008). In general, as phosphorus levels
increase, phytoplankton biomass increases (Schindler 1974) and in particular, the proportion of
cyanobacteria in the plankton community increases (Smith 1986; Watson et al. 1997; Downing et
al. 2001). For this reason, substantial efforts and funding have been put in place to reduce
phosphorus loading to freshwater ecosystems.
Lake Erie, in the Laurentian Great Lakes, historically experienced severe HAB events
leading up to the 1970’s, dominated by nitrogen-fixing cyanobacteria such as Anabaena and
Aphanizomenon (Matisoff and Ciborowski 2005). Point source phosphorus control measures
implemented by the Great Lakes Water Quality Agreement in 1972 dramatically reduced
phosphorus loading (Richards et al. 2002). Subsequently, phytoplankton abundance in Lake Erie
began to decrease through the mid 1990’s, especially with respect to cyanobacteria abundance
(DePinto et al. 1986; Makarewicz and Bertram 1991; Makarewicz 1993; Fahnenstiel et al. 1998).
However, since the mid 1990’s, phytoplankton abundance, especially the non-nitrogen fixing
cyanobacterium, Microcystis aeruginosa, has increased significantly despite nutrient control
measures (Vanderploeg et al. 2001; Conroy et al. 2005; Ouellette et al. 2006). Extensive
33

Microcystis blooms visible with satellite imagery, are now an annual occurrence in the western
basin of the lake (Vincent et al. 2004; Millie et al. 2009; Chaffin et al. 2011a; Bridgeman et al.
2012a) and microcystin (a potent liver toxin) levels routinely exceed the World Health
-1

Organization’s limit of 1 ug L at sites throughout the western basin (Brittain et al. 2000; Dyble
et al. 2008; Rinta-Kanto et al. 2009). Given that over 5 million people depend on western basin
of Lake Erie as a source of drinking water, transportation, commercial fisheries and recreation
(Fuller 2002), it’s critical to understand the limiting factors controlling Microcystis abundance in
order to mitigate its impact within the ecosystem.
The western basin of Lake Erie is functionally different from the central and eastern
being relatively shallow (average depth < 15 m), having a short hydraulic retention time and is
heavily influenced by the Maumee River, which contributes only 10% of the total water flow but
over 35% of the phosphorus loading to the basin (Schwab et al. 2009). As a result, there is
typically a steep nutrient gradient extending outwards from the mouth of the Maumee River at
the southwest corner of the basin with total phosphorus spanning a range from hypereutrophic to
nearly oligotrophic conditions within a distance of less than 40 km (Moorhead et al. 2008). It is
well established from field and lab studies that Microcystis grows faster with increasing
phosphorus levels (Gobler et al. 2007; Jiang et al. 2008; Davis et al. 2009) and therefore it is not
surprising that the highest abundance of Microcystis in the lake occurs closer to the mouth of the
Maumee River where total phosphorus is highest (Dyble et al. 2008; Chaffin et al. 2011a).
Temporally, Microcystis blooms usually occur from late August through September (RintaKanto et al. 2009; Chaffin et al. 2011a; Bridgeman et al. 2012a).
Numerous studies have documented that Microcystis is prevalent in the western basin and
that its abundance appears to be associated with the discharge plume of the Maumee River (e.g.
34

Vincent et al. 2004), however, survey studies that capture the spatial and temporal variation in
bloom abundance and environmental factors that may affect Microcystis are rare. A survey by
Chaffin et al. (2011) captured the temporal resolution of the 2008 Microcystis bloom at five sites
that spanned much of the trophic gradient extending from the Maumee River and found higher
Microcystis abundance at sites just outside the river mouth where phosphorus was higher than
offshore sites. Additionally, Rinta-Kanto et al. (2009) surveyed between three and twelve sites
in the western basin two to three times per year for three years and demonstrated a significant
correlation between total phosphorus and microcystin and mcyD gene copies. Clearly, more
extensive sampling is required in order to fully-characterize the seasonal bloom and determine to
what extent phosphorus may be correlated with Microcystis abundance.
Since other environmental variables (e.g. water clarity) may co-vary with phosphorus
along the trophic gradient extending from the Maumee River, it is important to identify
additional independent measures of phosphorus-dependent growth of Microcystis before
assigning a more causal link between phosphorus loading and the Microcystis blooms. One
option is to measure phosphorus limitation in the phytoplankton community and specifically in
Microcystis since it is established that phosphorus limited phytoplankton have lower growth rates
(Litchman et al. 2003; Shen and Song 2007). A number of studies have independently
confirmed that phosphorus can be limiting to at least some members of the phytoplankton
community in the western basin using measures of biomass C:P ratios (Guildford and Hecky
2005; Rattan and Taylor 2012), nutrient addition assays (Wilhelm et al. 2003), and alkaline
phosphatase activity (Guildford and Hecky 2005). More specifically, Chaffin et al. (2011)
isolated Microcystis cells from the community and found that cells were typically phosphorus
limited on a C:P ratio basis across most sampling sites in the basin and that the protein content of

35

cells at the most distant sites from the Maumee River were lower than cells collected closer to
the river, perhaps as a result of higher ambient phosphorus levels.
The purpose of the present study was to: 1) capture spatio-temporal patterns of
Microcystis abundance in the western basin at a short enough frequency and large enough spatial
scale in order to fully characterize the extent of the annual Microcystis bloom and identify
potentially important environmental factors associated with the bloom; and 2) determine whether
the apparent association between Microcystis abundance and phosphorus is congruent with
indices of phosphorus limitation and thus strengthening the causal link between phosphorus
loading to the basin and Microcystis prevalence.

36

METHODS
We conducted an extensive seasonal survey of the western basin of Lake Erie, the most
eutrophic of the Laurentian Great Lakes, from May through September of 2008. Nutrient
concentrations in the basin are heavily influenced by loading from the Maumee River (TP
-1

typically above 200 µg L ) that enters from the southwest. The Maumee contributes more than
half of the phosphorus in the southern and western portions of the basin where the Microcystis
blooms are a recurring water quality problem (Bridgeman et al., 2012; Chaffin et al. 2011;
Schwab et al. 2009). Nine sampling stations were selected to provide a broad representation
across the basin, with stations spaced approximately 6-10 km apart (Fig. 2.1). Sampling trips
were conducted at intervals of approximately two weeks, starting on 9 May and continuing
through 22 September in order to capture seasonal changes in the physiology of Microcystis
before, during and after the annual bloom.
On each sampling trip, water samples from each station were collected at 0.5 and 1 m
depth with a van Dorn bottle and pooled into clean 10 L cubitainers. The western basin of Lake
Erie is considered to be completely mixed during the summer due to the generally shallow
depths and large fetch for wind mixing. Temperature profiles confirmed that the lake was
generally isothermal (i.e. < 1 C difference from top to bottom) and therefore subsurface
samples were assumed to be to be representative of the water column. Secchi disk depth was
recorded at each station and temperature, light, pH, conductivity, in situ phycocyanin and
dissolved oxygen profiles were taken from at least four stations on each sampling trip.
The cubitainers were kept at ~ 4 C following collection and during transport back to the
Michigan State University campus in East Lansing, MI, where the samples were processed into

37

four sub-samples: 1) raw water sample stored at 4 C; 2) filtered water sample passed through an
A/E glass fiber filter and stored at 4 C; 3) seston collected on A/E filters and immediately
frozen at -20 C; and 4) water preserved in Lugol’s solution for microscope counting of the
phytoplankton.
All nutrient analyses were performed within 72 hours of collection. Soluble reactive P
(SRP) was measured on filtered water samples using long-pathlength spectrometry (Wetzel and
-

Likens 2000) and concentrations of NO3 were analyzed with 2

nd

derivative UV spectroscopy

(Crumpton et al. 1992) on a Perkin Elmer Lambda 20 UV/Vis spectrophotometer. Total
phosphorus (TP) was measured on persulfate-digested raw water samples and then analyzed
spectrophotometrically as SRP. Chlorophyll was extracted from frozen filters in 90% ethanol
and measured fluorometrically (Welschmeyer 1994).
Microcystis biomass was measured via inverted-microscope analysis on Lugol’s
preserved samples. Sub-samples (50 ml) were transferred into 25 mm-wide settling chambers
with a cover-slip bottom. After a 5-day settling period, the entire chamber surface was surveyed
for Microcystis colonies and the surface area of each colony was measured at 100X
magnification from digitized photographs (SPOT Software V 4.0.9, Diagnostic Instruments).
The cumulative colony surface area for a chamber was converted into Microcystis cell
-1

concentration (cells L ) based on the surface area to cell count relationship established for
2

Microcystis samples in Lake Erie (R = 0.95, n = 21). Microcystis dry biomass concentration
was then calculated by converting cell concentration to biovolume (based on measurements of

38

cell dimensions) and subsequently converting biovolume to dry biomass assuming a dry weight
3

(mg) to biovolume (mm ) ratio of 0.42.
Bulk-alkaline phosphatase activity (APA), a measure of phosphorus limitation of the
planktonic community, was measured fluorometrically in 10 ml aliquots of freshly-collected raw
water samples from the enclosures (Pettersson 1980). Microcystis-specific alkaline phosphatase
activity was assessed using enzyme-labeled fluorescence (ELF) microscopy (Gonzalez-Gil and
Aguileral 1998). This microscopy-based method is analogous to bulk alkaline phosphatase
analysis, whereby phosphorus-limited organisms produce the enzyme alkaline phosphatase
which is then detected visually for cells of a particular species. ELF substrate (0.1 ml, 1:20 v:v in
buffer solution; E6601 from Invitrogen, USA) was added to separate 10 ml subsamples of
freshly-collected raw water samples from the enclosures. After 1 hr, the treated subsample was
filtered onto black 1 µm membrane filters and rinsed with phosphate-buffered saline to inhibit
further enzyme activity. The filters were then permanently mounted onto microscope slides and
stored at -20 C for up to three months before analysis.
Using a Nikon Eclipse E600 epifluorescence microscope, the mounted membrane filters
were visually scanned for Microcystis colonies and evidence of alkaline phosphate activity as
indicated by bright green precipitates on the surface of colonies (Fig. 2.2). Each Microcystis
colony (up to 200 per slide) was counted and marked as having either positive or negative
enzyme activity, with the percent of active colonies per sample serving as a relative index of the
population’s alkaline phosphatase activity.

39

RESULTS
Early season survey sampling trips on 9 May, 23 May and 16 June spanned a period of
rapidly increasing water temperatures from 12 to over 21 C (data not shown). Microcystis was
-1

not observed via microscopy (detection limit of 1 ug L ) until 22 July (Fig. 2.3a) and then only
-1

increased from about 26 to 44 ug L between 22 July and 3 August when averaged across all
sites within the western basin. Between 3 August and 3 September, Microcystis demonstrated
-1

strong exponential growth, increasing by nearly 100 fold to over 3500 ug L . Following the
-1

peak abundance in early September, Microcystis decreased to less than 750 ug L by 22
September. Averaged across sites, chlorophyll a followed the same temporal growth pattern as
-1

Microcystis, but concentrations only increased by about 10 fold from 5 ug L in July to a peak
-1

-1

of 57 ug L on 3 September before declining to about 30 ug L on 22 September.
Phosphorus concentrations in the western basin of Lake Erie are typically very high
compared to the central and eastern basins of the lake due to the strong influence of high
phosphorus loading from the Maumee River. Temporally, total phosphorus exhibited two phases
during the sampling season of 2008, with an average concentration among all stations in the
-1

basin between 30 and 40 ug-P L during June, July and early August (Fig. 2.3b). Between 16
August and 3 September, total phosphorus more than doubled to over 85 g/L and remained above
-1

90 ug L on 22 September. Nitrate, which is the dominant source of inorganic nitrogen in the
basin, declined progressively over the course of the sampling period (Fig. 2.3b). Concentrations

40

-1

-1

of nitrate averaged around 800 ug-N L through mid-July, then dropped to 200 ug-N L by
-1

mid-August and reaching as low as 33 ug-N L by 22 September.
Water temperatures in the western basin during the growing season rose quickly to over
20 C by mid-June and continued to climb to over 25 C by mid-July (Fig. 2.3c). Temperatures
remained between 23 and 25 C through 3 September and then dropped to 20 C by 22
September. Water clarity was generally inversely proportional to chlorophyll a concentration
and Microcystis abundance, with Secchi Disk depths averaging 2.5 to 3.5 meters among all sites
in June, July and early August leading up to the bloom (Fig. 2.3c). By 3 September, the Secchi
Disk depth had decreased to about 1 meter and remained low through 22 September.
Due to the strong influence of phosphorus inputs into the western basin of Lake Erie by
the Maumee River (Schwab et al. 2009), we expected, and in fact found, a strong total
phosphorus (TP) gradient extending outwards from the mouth of the Maumee River. On each
sampling date there was a significant negative correlation between total phosphorus and distance
from the mouth of the river (Fig. 2.4). As depicted in the temporal plots in Fig. 2.3, there as a
dramatic increase in overall TP concentrations in the basin between 18 Aug and 3 Sep but Fig.
2.4, illustrates that this increase was also characterized by a dramatic increase in the steepness in
the TP gradient, not simply an overall increase at all sites. For example, the sites closest to the
-1

mouth of the river typically averaged close to 50 ugP L through 18 Aug, yet reached 150 to 200
-1

ugP L by 3 Sep and afterwards. The most offshore sites also increased but less dramatically
-1

-1

from about 20 to 30 ugP L through mid-August up to 25-50 ugP L in September.

41

Microcystis biomass did not show a significant correlation with distance from the
Maumee River during the early sampling dates through mid-August while concentrations were
generally low (Fig. 2.4). However, as Microcystis reached peak abundance on 3 September,
there was a strong and significant negative correlation between biomass concentration and
distance from the Maumee River, paralleling the concentrations in total phosphorus and spatial
pattern continued through 22 September.
Given the apparent co-variation in total phosphorus and Microcystis leading up to the
peak bloom in early September, we regressed Microcystis against TP for all the sampling sites
sampled during the growth period from 22 July to peak abundance on 3 September (Fig. 2.5).
TP was, as expected, significantly correlated with Microcystis abundance over the growth period,
2

explaining nearly half of the variation (R = 0.48, p <0.0001), despite a multitude of other
2

potential abiotic and biotic variables that vary among sites within the >500 km sampling area
over the span of six weeks. On 22 September, following the peak of the bloom, Microcystis was
still positively correlated with TP, but at a lower slope than during the growth period (Fig. 2.5).
The strong correlation between TP and Microcystis abundance (Fig. 2.5) suggests that
phosphorus was potentially limiting the growth of Microcystis and perhaps other phytoplankton
species during the growth period in the western basin. In order to assess phosphorus limitation
among sites and sampling dates, we measured bulk alkaline phosphatase activity on the entire
plankton community. Despite this broad measure of phosphorus limitation in the entire
community, Microcystis abundance demonstrated a very tight negative correlation with alkaline
phosphatase activity (Fig. 2.5) during the growth period (22 July to 3 September) indicating that
Microcystis abundance increased as phosphorus limitation decreased (i.e. lower alkaline
phosphatase activity). Compared to total phosphorus, alkaline phosphatase activity for the entire
42

plankton community explained more than 70% of the variation in Microcystis abundance during
2

the growth period (R = 0.71, p < 0.0001). By 22 September, alkaline phosphatase activity was
very low for all the sites, yet Microcystis abundance had already started to decline, suggesting
that some other factor besides phosphorus limitation was likely limiting Microcystis growth after
the bloom (Fig. 2.6).
In order to determine whether Microcystis specifically was experiencing phosphorus
limitation, we employed a taxon-specific alkaline phosphatase activity using enzyme labeled
fluorescence (ELF) and epi-fluorescence microscopy. Among sites and dates during the survey,
Microcystis colonies displayed a range of ELF activity ranging from about 30% of colonies
expressing ELF-based alkaline phosphatase activity to 0% of colonies expressing ELF (Fig. 2.7),
indicating lower levels of alkaline phosphatase activity by Microcystis. For the period from 4
August till 22 September, the % of Microcystis colonies expressing ELF-based alkaline
phosphatase activity was correlated with overall bulk alkaline phosphatase activity (APA) for the
entire plankton community (Spearman rank correlation coefficient, rs = 0.34, t = 2.06, df = 33, p
= 0.023), indicating that phosphorus limitation for Microcystis was congruent with overall bulk
alkaline phosphatase activity for the entire plankton community.

43

DISCUSSION
During the spring/summer of 2008, the western basin of Lake Erie experienced a
pronounced bloom of Microcystis during late August and early September (Fig. 2.3), which
resulted in microcystin toxin levels exceeding the World Health Organization drinking water
limit of 1 ug/L at several sites throughout the basin. The Microcystis population was not even
detectable via microscopy until mid-July when temperatures had warmed to 25 C, but then
grew exponentially through the first week September before dissipating. Noticeably, Microcystis
grew exponentially during a period of increasing total phosphorus levels (Fig. 2.3) and
Microcystis biomass was strongly correlated with total phosphorus leading up to the peak bloom
(Fig. 2.5). More pertinently, Microcystis was inversely correlated with alkaline phosphatase
activity (Fig. 2.6), a measure of phosphorus limitation in the phytoplankton community. In
particular, Microcystis colonies expressed taxon-specific alkaline phosphatase activity in relative
proportion to community-wide phosphorus limitation (Fig. 2.7). Together, these results indicate
that phosphorus strongly regulates Microcystis abundance in the western basin.
It has been previously demonstrated that there is a pronounced phosphorus gradient
extending from the mouth of the Maumee River (Moorhead et al. 2008), and this pattern was
exhibited consistently throughout our survey (Fig. 2.4). Notably, the relative slope of this
gradient increased dramatically between 18 August and 3 September, whereby total phosphorus
at sites closest to the river mouth increased by nearly four-fold. This relative change in the
gradient therefore explains the most of the apparent increase in phosphorus averaged over the
whole basin sampling area during the same period. Microcystis did not exhibit a strong gradient
until 3 September, perhaps because the slope of the phosphorus gradient in earlier dates was not
as pronounced despite being statistically significant.
44

Despite the complexities of a dynamic ecosystem with multiple abiotic and biotic
2

variables changing over the course of nine weeks and a 500 km survey area, there was still a
very strong positive correlation between total phosphorus concentrations and Microcystis
abundance, explaining nearly 50% of the variation (Fig. 2.5). Several studies have found similar
patterns across multiple lakes or within lakes (Oh et al. 2001; Otten et al. 2012), but we believe
this is the first study to quantitatively demonstrate a correlation between phosphorus and
Microcystis biomass during the course of developing bloom in Lake Erie. Additionally, this
pattern is consistent with both field and lab studies that have experimentally demonstrated that
increasing phosphorus increases Microcystis (Vézie et al. 2002; Strojsová et al. 2005; Gobler et
al. 2007; Davis et al. 2009).
In order to further evaluate the role of phosphorus in driving the dynamics of Microcystis
growth in the western basin, we quantified both the community-wide bulk alkaline phosphatase
activity and taxon-specific alkaline phosphatase activity for Microcystis. In comparison to total
phosphorus, bulk alkaline phosphatase activity was actually a better predictor of Microcystis
abundance, explaining over 70% of the variation (Fig. 2.6). It was surprising to find such a
strong relationship, but it has been well established that alkaline phosphatase activity is highly
dynamic in phytoplankton (Jansson and Petterson 1988) especially in cyanobacteria (Sakshaug
and Olsen 1986). Under phosphorus limitation, Microcystis, like other species should exhibit
lower growth rates (Litchman et al. 2003; Shen and Song 2007) and therefore its consistent that
Microcystis abundance should increase with decreasing levels of phosphorus limitation (i.e.
lower alkaline phosphatase activity).
Despite the strong relationship between bulk alkaline phosphatase activity and
Microcystis abundance, we also wanted to know whether we could specifically identify
45

phosphorus limitation in Microcystis without the laborious efforts of performing either nutrient
amendment assays (e.g. Gobler et al. 2007, Davis et al. 2009) or isolating Microcystis cells from
the community and analyzing C:P ratios. Fortunately, an enzyme-labeled fluorescence ELF
assay (Gonzalez-Gil and Aguileral 1998; Rengefors et al. 2003) allows for the rapid detection of
alkaline phosphatase activity via fluorescence microscopy (Fig. 2.2). Microcystis colonies did
express a modest range of ELF phosphatase activity that correlated positively with bulk alkaline
phosphatase activity (Fig. 2.7), demonstrating that Microcystis colonies were experiencing
phosphorus limiting conditions in conjunction with bulk alkaline phosphatase activity for the
entire plankton community. While the ELF method is relatively fast and simple to process, its
limitation may be that in many samples, there was no indication of taxon-specific ELF
phosphatase for Microcystis (i.e. 0% of colonies expressing), yet clearly there was some positive
level of bulk alkaline phosphatase activity.
The majority of the focus for this study has been on the factors controlling the growth of
Microcystis in the western basin of Lake Erie, however, we did collect samples during the period
of declining Microcystis biomass (3 Sep to 22 Sep) which may offer insight into factors that lead
to a bloom collapse. First, phosphorus levels remained high following the bloom (Figs. 2.3 and
2.4), alkaline phosphatase activity remained low, and no Microcystis colonies were expressing
ELF phosphatase activity, all of which indicates that phosphorus limitation was not apparent and
likely did not contribute to the population decline. One explanation for the decline may be
-1

nitrogen limitation as nitrate availability declined from over 800 ug L in early summer to only
-1

30 ug L by 22 Sep. This would have resulted in a nitrate-N:TP ratio of less than 1, which is
considered strongly nitrogen limited by many measures (Healey and Hendzel 1980). Second, the
water temperature dropped relatively rapidly from 24 to 20 oC between 3 September and 22
46

September and it is well established that Microcystis growth rates are strongly temperature
dependent in this range (Nalewajko and Murphy 2001; Davis et al. 2009). Third, water clarity
was dropping quickly with the concomitant rise in both chlorophyll a and Microcystis abundance
in late August and early September and it’s possible that light levels in the water column could
have limited Microcystis growth rates (Wiedner et al. 2003). It is plausible that a combination of
these environmental factors could have contributed to the rapid decline in Microcystis abundance
following the bloom.
Our study has provided multiple lines of evidence that Microcystis abundance is
regulated by phosphorus availability in the western basin of Lake Erie at least during the 2008
season. The implications of this work are that phosphorus loading into the western basin via the
Maumee River can have dramatic impacts on the scope of the Microcystis bloom in a given year.
Clearly, an overall increase or decrease in phosphorus loading would translate into some direct
impacts on Microcystis abundance, and unfortunately, the current trend indicates that phosphorus
loading from the Maumee has been increasing since the mid 1990’s (OEPA 2010). The timing
of the phosphorus loading within the growing season may also have unknown variable impacts.
For example, in 2008, there was a dramatic rise in the phosphorus gradient between 18 August
and 3 September (Fig. 2.4). The rapid increase in available phosphorus then contributed to the
rapid growth of phytoplankton, including Microcystis, within the basin (Figs. 3 and 4). If this
pulse of phosphorus had arrived later in September, then it’s possible that Microcystis would not
have been phosphorus limited at that point and would have been more limited by temperature or
nitrogen. Likewise, if the pulse had arrived earlier in the season when temperatures were lower,
then perhaps a different taxa besides Microcystis would have bloomed. Given that the Maumee
watershed is largely dominated by agriculture (Richards et al. 2002) there may be land

47

management practices that could reduce overall phosphorus loading or at least alter the timing of
the phosphorus loading into the river.

48

FIGURES

Figure 2.1 Map of the western basin of Lake Erie and locations of the survey sampling stations.
Map derived from Chaffin et al. (2011).

49

Figure 2.2 Taxon-specific alkaline phosphatase activity using enzyme-labeled fluorescence
(ELF) and epi-fluorescence microscopy. Chlorophyll a in cells fluoresces orange-red, while
expression of alkaline phosphatase activity is represented by bright-green precipitates formed by
the alkaline phosphatase enzyme. Panel A and B illustrate colonies of Pediastrum and
Microcystis aeruginosa that are not expressing ELF, while panels C and D illustrate colonies
expressing ELF. For interpretation of the references to color in this and all other figures, the
reader is referred to the electronic version of this dissertation.

50

Figure 2.3 (Top) Microcystis aeruginosa (ug dry biomass/L) and chlorophyll a (ug/L); (middle)
total phosphorus (ugP/L) and nitrate (ugN/L); and (bottom) water temperature (C) and Secchi
disk depth (m) throughout the summer of 2008 in the western basin of Lake Erie. Values are
means ± SE averaged across sites within the basin on each date.

51

Figure 2.4 Total phosphorus (ug P/L) and Microcystis aeruginosa biomass (ug dry weight/L) in
relation to distance from the mouth of the Maumee River in the western basin of Lake Erie.
Trend lines are plotted for significant correlations (p <0.05).

52

Figure 2.5 Microcystis aeruginosa biomass (ug dry weight/L) as a function of total phosphorus
(ugP/L) during the sampling dates leading up to peak Microcystis abundance (Solid circles: 22
Jul, 4 Aug, 18Aug and 3 Sep) and after the bloom (Open circles: 22 Sep). Trend line represents
the linear regression of log10 transformed biomass and log10 transformed phosphorus
concentrations for the sampling period leading up the bloom (22 Jul – 3 Sep).

53

Figure 2.6 Microcystis aeruginosa biomass (ug dry weight/L) as a function of bulk alkaline
phosphatase activity (APA) (nM/ug chlorophyll a/L/min) during the sampling dates leading up to
peak Microcystis abundance (Solid circles: 22 Jul, 4 Aug, 18Aug and 3 Sep) and after the bloom
(Open circles: 22 Sep). Trend line represents a linear regression of log10 transformed biomass
and log10 transformed APA for the sampling period leading up to the bloom (22 Jul – 3 Sep).

54

Figure 2.7 The fraction of Microcystis aeruginosa colonies expressing taxon-specific enzyme
labeled fluorescence (ELF) of alkaline phosphatase as a function of bulk alkaline phosphatase
activity (APA) (nM/ug chlorophyll a/L/min) for the entire seston community during on 4 Aug,
18 Aug, 3 Sep and 22 Sep. Spearman rank correlation coefficient, rs = 0.337, t = 2.06, df = 33, p
= 0.023.

55

REFERENCES

56

REFERENCES

Bridgeman, T. B., J. D. Chaffin, D. D. Kane, J. D. Conroy, S. E. Panek, and P. M. Armenio.
2012a. From River to Lake : Phosphorus partitioning and algal community compositional
changes in Western Lake Erie. Journal of Great Lakes Research 38: 90–97.
Bridgeman, T. B., J. D. Chaffin, D. D. Kane, J. D. Conroy, S. E. Panek, and P. M. Armenio.
2012b. From River to Lake: Phosphorus partitioning and algal community compositional
changes in Western Lake Erie. Journal of Great Lakes Research 38: 90–97.
Brittain, S. M., J. Wang, L. Babcock-Jackson, W. W. Carmichael, K. L. Rinehart, and D. A.
Culver. 2000. Isolation and characterization of microcystins, cyclic heptapeptide
hepatotoxins from a Lake Erie strain of Microcystis aeruginosa. Journal of Great Lakes
Research 26: 241–249.
Carmichael, W. W. 1994. The toxins of cyanobacteria. Scientific American 270: 78–86.
Carmichael, W. W. 2001. Health effects of toxin-producing cyanobacteria: “The CyanoHABs”.
Human and Ecological Risk Assessment 7: 1393–1407.
Carpenter, S., N. Caraco, D. Correll, R. Howarth, A. Sharpley, and V. Smith. 1998. Nonpoint
pollution of surface waters with phosphorus and nitrogen. Ecological Applications 8: 559–
568.
Chaffin, J. D., T. B. Bridgeman, S. a. Heckathorn, and S. Mishra. 2011a. Assessment of
Microcystis growth rate potential and nutrient status across a trophic gradient in western
Lake Erie. Journal of Great Lakes Research 37: 92–100.
Chaffin, J. D., T. B. Bridgeman, S. a. Heckathorn, and S. Mishra. 2011b. Assessment of
Microcystis growth rate potential and nutrient status across a trophic gradient in western
Lake Erie. Journal of Great Lakes Research 37: 92–100.
Chorus, I., and J. Bartram. 1999. Toxic cyanobacteria in water: a guide to their public health
consequences, monitoring and management. 416.
Conroy, J. D., D. D. Kane, D. M. Dolan, W. J. Edwards, M. N. Charlton, and D. a. Culver. 2005.
Temporal Trends in Lake Erie Plankton Biomass: Roles of External Phosphorus Loading
and Dreissenid Mussels. Journal of Great Lakes Research 31: 89–110.
Crumpton, W. G., T. M. Isenhart, and P. D. Mitchell. 1992. Nitrate and organic N analyses with
second-derivative spectroscopy. Limnology and Oceanography 37: 907–913.

57

Davis, T. W., D. L. Berry, G. L. Boyer, and C. J. Gobler. 2009. The effects of temperature and
nutrients on the growth and dynamics of toxic and non-toxic strains of Microcystis during
cyanobacteria blooms. Harmful Algae 8: 715–725.
DePinto, J., T. Young, and L. McIlroy. 1986. Great Lakes water quality improvement.
Environmental Science & Technology 20: 752–759.
Downing, J. A., S. B. Watson, and E. McCauley. 2001. Predicting cyanobacteria dominance in
lakes. Canadian Journal of Fisheries and Aquatic Sciences 58: 1905–1908.
Dyble, J., G. L. Fahnenstiel, R. W. Litaker, D. F. Millie, and P. A. Tester. 2008. Microcystin
concentrations and genetic diversity of Microcystis in the lower Great Lakes.
Environmental Toxicology 23: 507–516.
Fahnenstiel, G. L., A. E. Krause, M. J. McCormick, H. J. Carrick, and C. L. Schelske. 1998. The
structure of the planktonic food-web in the St. Lawrence Great Lakes. Journal of Great
Lakes Research 24: 531–554.
Gobler, C., T. Davis, K. Coyne, and G. Boyer. 2007. Interactive influences of nutrient loading,
zooplankton grazing, and microcystin synthetase gene expression on cyanobacterial bloom
dynamics in a eutrophic New York lake. Harmful Algae 6: 119–133.
Gonzalez-Gil, S., and A. Aguileral. 1998. Detection and quantification of alkaline phosphatase in
single cells of phosphorus-starved marine phytoplankton.
Guildford, S., and R. Hecky. 2005. Phytoplankton nutrient status in Lake Erie in 1997. Journal of
Great Lakes … 31: 72–88.
Healey, F. P., and L. L. Hendzel. 1980. Physiological indicators of nutrient deficiency in lake
phytoplankton. Canadian Journal of Fisheries and Aquatic Sciences 37: 442–453.
Huisman, J., H. C. P. Matthijs, and P. M. Visser. 2005. Harmful cyanobacteria, Vol 3. J.
Huisman, H.C.P. Matthijs, and P. Visser [eds.]. Springer.
Jiang, Y., B. Ji, R. N. S. Wong, and M. H. Wong. 2008. Statistical study on the effects of
environmental factors on the growth and microcystins production of bloom-forming
cyanobacterium—Microcystis aeruginosa. Harmful Algae 7: 127–136.
Litchman, E., D. Steiner, and P. Bossard. 2003. Photosynthetic and growth responses of three
freshwater algae to phosphorus limitation and daylength. Freshwater Biology 48: 2141–
2148.
Makarewicz, J. 1993. Phytoplankton biomass and species composition in Lake Erie, 1970-1987.
Journal of Great Lakes Research 19: 258–274.

58

Makarewicz, J. C., and P. Bertram. 1991. Evidence for the restoration of the Lake Erie
ecosystem. Bioscience 41: 216–223.
Matisoff, G., and J. Ciborowski. 2005. Lake Erie trophic status collaborative study. Journal of
Great Lakes Research 31: 1–10.
Millie, D. F., G. L. Fahnenstiel, J. Dyble Bressie, R. J. Pigg, R. R. Rediske, D. M. Klarer, P. a.
Tester, and R. W. Litaker. 2009. Late-summer phytoplankton in western Lake Erie
(Laurentian Great Lakes): bloom distributions, toxicity, and environmental influences.
Aquatic Ecology 43: 915–934.
Moorhead, D., T. Bridgeman, and J. Morris. 2008. Changes in water quality of Maumee Bay
1928–2003. Checking the pulse of Lake Erie 1–32.
Nalewajko, C., and T. P. Murphy. 2001. Effects of temperature, and availability of nitrogen and
phosphorus on the abundance of Anabaena and Microcystis in Lake Biwa, Japan: an
experimental approach. Limnology 2: 45–48.
OEPA. 2010. Ohio Environmental Protection Agency (OEPA). Ohio Lake Erie Phosphorus Task
Force Final Report.
Oh, H., S. Lee, J. Kim, and H. Kim. 2001. Seasonal variation and indirect monitoring of
microcystin concentrations in Daechung Reservoir, Korea. Applied and Environmental
Microbiology 67: 1484–1489.
Otten, T. G., H. Xu, B. Qin, G. Zhu, and H. W. Paerl. 2012. Spatiotemporal patterns and
ecophysiology of toxigenic microcystis blooms in Lake Taihu, China: implications for
water quality management. Environmental science & technology 46: 3480–8.
Ouellette, A. J. a, S. M. Handy, and S. W. Wilhelm. 2006. Toxic Microcystis is widespread in
Lake Erie: PCR detection of toxin genes and molecular characterization of associated
cyanobacterial communities. Microbial ecology 51: 154–65.
Paerl, H. W., and J. Huisman. 2008. Blooms Like It Hot. Science 320: 57–58.
Pettersson, K. 1980. Alkaline phosphatase activity and algal surplus phosphorus as phosphorusdeficiency indicators in Lake Erken. Archiv fur Hydrobiologie 89: 54–87.
Rattan, K., and W. Taylor. 2012. Nutrient status of phytoplankton across a trophic gradient in
Lake Erie: evidence from new fluorescence methods. Canadian Journal of Fisheries and
Aquatic Sciences 111: 94–111.
Rengefors, K., K. C. Ruttenburg, C. L. Haupert, C. Taylor, B. L. Howes, and D. M. Anderson.
2003. Experimental investigation of taxon-specific response of alkaline phosphatase activity
in freshwater phytoplankton. Limnology and Oceanography 48: 1167–1175.

59

Richards, R. P., D. B. Baker, and H. Creek. 2002. Trends in Water Quality in LEASEQ Rivers
and Streams (Northwestern Ohio), 1975-1995. Journal of Environmental Quality 31: 90–96.
Rinta-Kanto, J. M., E. A. Konopko, J. M. DeBruyn, R. A. Bourbonniere, G. L. Boyer, and S. W.
Wilhelm. 2009. Lake Erie Microcystis: relationships between microcystin production,
dynamics of genotypes and environmental parameters in a large lake. Harmful Algae 8:
665–673.
Sakshaug, E., and Y. Olsen. 1986. Nutrient status of phytoplankton blooms in Norwegian waters
and algal strategies for nutrient competition. Canadian Journal of Fisheries and … 43: 389–
396.
Schindler, D. 1974. Eutrophication and recovery in experimental lakes: implications for lake
management. Science 184: 897–899.
Schindler, D. W. 1977. The evolution of phosphorus limitation in lakes. Science 195: 260–262.
Schwab, D. J., D. Beletsky, J. DePinto, and D. M. Dolan. 2009. A hydrodynamic approach to
modeling phosphorus distribution in Lake Erie. Journal of Great Lakes Research 35: 50–60.
Shen, H., and L. Song. 2007. Comparative studies on physiological responses to phosphorus in
two phenotypes of bloom-forming Microcystis. Hydrobiologia 592: 475–486.
Smith, V. H. 1986. Light and nutrient effects on the relative biomass of blue-green algae in lake
phytoplankton. Canadian Journal of Fisheries and Aquatic Sciences 43: 148–153.
Strojsová, A., J. Vrba, J. Nedoma, and K. Åœimek. 2005. Extracellular phosphatase activity of
freshwater phytoplankton exposed to different in situ phosphorus concentrations. Marine
and freshwater research 417–424.
Vanderploeg, H. A., J. R. Liebig, W. W. Carmichael, M. A. Agy, T. H. Johengen, G. L.
Fahnenstiel, and T. F. Nalepa. 2001. Zebra mussel (Dreissena polymorpha) selective
filtration promoted toxic Microcystis blooms in Saginaw Bay (Lake Huron) and Lake Erie.
Canadian Journal of Fisheries and Aquatic Sciences 58: 1208–1221.
Vincent, R. K., X. Qin, R. M. L. McKay, J. Miner, K. Czajkowski, J. Savino, and T. Bridgeman.
2004. Phycocyanin detection from LANDSAT TM data for mapping cyanobacterial blooms
in Lake Erie. Remote Sensing of Environment 89: 381–392.
Vézie, C., J. Rapala, J. Vaitomaa, J. Seitsonen, and K. Sivonen. 2002. Effect of nitrogen and
phosphorus on growth of toxic and nontoxic Microcystis strains and on intracellular
microcystin concentrations. Microbial ecology 43: 443–54.
Watson, S. B., E. McCauley, and J. A. Downing. 1997. Patterns in phytoplankton taxonomic
composition across temperate lakes of differing nutrient status. Limnology and
Oceanography 42: 487–495.
60

Welschmeyer, N. A. 1994. Fluorometric analysis of chlorophyll a in the presence of chlorophyll
b and phaeopigments. Limnology and Oceanography 39: 1985–1992.
Wetzel, R. G., and G. E. Likens. 2000. Limnological Analyses, 3nd ed. Springer-Verlag.
Wiedner, C., P. M. Visser, J. Fastner, J. S. Metcalf, G. A. Codd, and L. R. Mur. 2003. Effects of
light on the microcystin content of Microcystis strain PCC 7806. Applied and
Environmental Microbiology 69: 1475–1481.
Wilhelm, S., J. DeBruyn, and O. Gillor. 2003. Effect of phosphorus amendments on present day
plankton communities in pelagic Lake Erie. Aquatic Microbial Ecology 32: 275–285.

61