Elena Litchman
Population and community responses of phytoplankton
to ?uctuating light
Received: 24 September 1997 /Accepted: 8 July 1998
Abstract Light is a major resource in aquatic ecosys-
tems and has a complex pattern of spatio-temporal
variability, yet the e?ects of dynamic light regimes on
communities of phytoplankton are largely unexplored. I
examined whether and how fluctuating light supply
a?ects the structure and dynamics of phytoplankton
communities. The e?ect of light fluctuations was tested
at two average irradiances: low, 25 lmol quan-
ta m
)2
s
)1
and high, 100 lmol quanta m
)2
s
)1
in 2- and
18-species communities of freshwater phytoplankton.
Species diversity, and abundances of individual species
and higher taxa, depended significantly on both the
absolute level and the degree of variability in light sup-
ply, while total density, total biomass, and species rich-
ness responded only to light level. In the two-species
assemblage, fluctuations increased diversity at both low
and high average irradiances and in the multispecies
community fluctuations increased diversity at high
irradiance but decreased diversity at low average irra-
diance. Species richness was higher under low average
irradiance and was not a?ected by the presence or
absence of fluctuations. Diatom abundance was in-
creased by fluctuations, especially at low average irra-
diance, where they became the dominant group, while
cyanobacteria and green algae dominated low constant
light and all high light treatments. Within each taxo-
nomic group, however, there was no uniform pattern in
species responses to light fluctuations: both the magni-
tude and direction of response were species-specific. The
temporal regime of light supply had a significant e?ect
on the growth rates of individual species grown in
monocultures. Species responses to the regime of light
supply in monocultures qualitatively agreed with their
abundances in the community experiments. The results
indicate that the temporal regime of light supply may
influence structure of phytoplankton communities by
di?erentially a?ecting growth rates and mediating spe-
cies competition.
Key words Temporal heterogeneity ? Light fluctuations ?
Phytoplankton ? Community structure ? Diversity
Introduction
Light is a major factor influencing phytoplankton and
has a complex pattern of spatial and temporal variabil-
ity. Temporal frequencies of light variation range from
fast fluctuations caused by surface waves (on the order
of 10 Hz) to seasonal changes of irradiance (10
)7
Hz).
The light levels experienced by phytoplankton cells can
vary from the complete darkness in the aphotic zone to
irradiances greater than 1500 lmol quanta m
)2
s
)1
at
the surface. Phytoplankton are sensitive not only to the
absolute average light level but also to the pattern of
light supply, i.e., constant versus fluctuating light (e.g.,
Ferris and Christian 1991; Prezelin et al. 1991; Reynolds
1994). As phytoplankton experience a dynamic versus
constant light supply, changes may occur in the bio-
chemical composition (Gibson and Foy 1988; Kroon
et al. 1992; Ibelings et al. 1994) and optical properties of
cells (Stramski et al. 1993), and in the rates of photo-
synthesis (Harris and Piccinin 1977; Marra 1978;
Grobbelaar et al. 1992; Kroon et al. 1992) and respira-
tion (Beardall et al. 1994).
The majority of studies on the e?ects of fluctuating
light supply have concentrated on the biochemical and
physiological levels. However, light fluctuations may
a?ect higher levels of biological organization, up to the
population and community levels. Physiological changes
due to light fluctuations potentially can be translated
into changes in long-term growth rates. Several studies
E. Litchman
1
Department of Ecology, Evolution and Behavior,
University of Minnesota,
1987 Upper Buford Circle,
St. Paul, MN 55108, USA
Present address:
1
Smithsonian Environmental Research Center,
P.O. Box 28, Edgewater, MD 21037, USA
e-mail: litchman@serc.si.edu, Fax: +1-301-2617954
Oecologia (1998) 117:247?257 ? Springer-Verlag 1998
have demonstrated di?erential growth-rate responses
of phytoplankton to fluctuating light (Que
?
guiner and
Legendre 1986; Gibson and Foy 1988; Stramski et al.
1993; Ibelings et al. 1994). If these growth-rate responses
are species-specific (cf. Ibelings et al. 1994), then light
fluctuations are likely to a?ect the community level as
well, i.e., cause changes in dominance patterns, species
dynamics and diversity (Tilman et al. 1982).
These potential community-level e?ects of light fluc-
tuations remain largely untested. Brzezinski and Nelson
(1988) showed that fluctuating light (light:dark cycle)
facilitated the coexistence of two species of planktonic
diatoms competing for ammonium, probably by tem-
porally separating nutrient uptake by the two species.
Similarly, van Gemerden (1974) showed that two strains
of sulfur bacteria were able to coexist under fluctuating
light but not under constant light. The generality of
these results is unknown, as well as how di?erences in
variability of light supply would influence more complex
multispecies phytoplankton communities such as those
found in nature.
A question of particular interest for community
structure is whether fluctuating light supply can promote
coexistence and increase species diversity. The study of
Brzezinski and Nelson (1988) provides an example of
coexistence of two diatoms facilitated by light fluctua-
tions at non-limiting levels. For limiting irradiances,
theory predicts competitive exclusion under a constant
light supply (Tilman 1982; Huisman and Weissing 1994).
Fluctuating supply of a limiting resource may, however,
increase the number of coexisting species as was shown
theoretically by Levins (1979) and Armstrong and
McGehee (1980). This was confirmed experimentally for
nitrogen- and phosphorus-limited phytoplankton
(Turpin and Harrison 1979; Sommer 1984, 1985; Grover
1991a). However, it is unknown whether a fluctuating
light supply of either limiting or non-limiting level can
promote diversity of multispecies phytoplankton
communities.
The goal of this study was to investigate whether the
structure and dynamics of multispecies communities are
sensitive to the degree of variability in light supply. In
particular, I sought to determine whether the taxonomic
composition, dominance patterns, dynamics of individ-
ual species, and total biomass of the community depend
on the temporal regime of light supply and whether light
fluctuations can enhance diversity. These questions were
explored using a laboratory multispecies assemblage of
freshwater phytoplankton. In addition, because the
community and species responses to environmental
fluctuations may depend on species richness (Tilman
1996), I performed the same experiment with a com-
munity consisting of only two species, selected based on
their high abundance in the multispecies experiment.
Because community responses should depend on the
di?erential responses of individual species, I also ex-
amined the e?ects of the same light regimes that were
used in community experiments on growth rates of the
two species in monocultures. This study thus examined
the e?ect of temporal variability in light supply at three
levels of complexity: in monocultures, and in two- and
multispecies communities.
To isolate the e?ect of the degree of variability of
light supply from the e?ect of the total irradiance, I
contrasted light regimes of the same total irradiance
dose, with and without fluctuations. For simplicity,
I used only one type of fluctuation (square wave),
although in nature many types of temporal fluctuations
are superimposed on each other. Underwater irradiance
varies not only in intensity, but also in spectral com-
position, and there is experimental evidence that spectral
fluctuations can be important (e.g., Anderson 1993).
However, a consideration of their possible e?ects on
phytoplankton was beyond the scope of this paper.
I used fluctuations of two periods, 1 h and 8 h, so
that the time scales of experimental light variation were
similar to time scales of vertical mixing in the epilimnion
of a typical lake (Imberger 1985). The chosen fluctuation
periods are also comparable to two important photo-
physiological time scales of phytoplankton (less than
2 h, associated with electron transport and greater than
5 h, associated with pigment turnover) (Neale and
Marra 1985) and thus are likely to a?ect physiology and
possibly species dynamics. The e?ects of fluctuations
were investigated at two di?erent levels of irradiance,
one within the range of growth-limiting irradiances, and
the other closer to saturating levels, because theoretical
and experimental studies suggest that the e?ect of light
fluctuations of the same frequency might depend on the
average level of irradiance (Thornley 1974; Dromgoole
1988; Stramski et al. 1993). The amplitude of fluctua-
tions was ?70?80% of the average irradiance for both
irradiance levels. The frequency and amplitude of ex-
perimental light fluctuations are comparable to those of
fluctuations resulting from phytoplankton movements in
the thermocline due to internal wave mixing (Denman
and Gargett 1983).
Materials and methods
Multispecies community experiment
To investigate the e?ect of light fluctuations at the community
level, a laboratory community of 18 species of freshwater phyto-
plankton from three classes (green algae, diatoms and cyanobac-
teria) (Table 1) was grown at six di?erent light regimes. At each
of the two average levels of irradiance, 25 and 100 lmol quan-
ta m
)2
s
)1
, there were three treatments with the same total irradi-
ance: constant irradiance and square-wave high-low light
fluctuations with periods of either 1 h or 8 h. This design allowed
separation of the e?ect of fluctuations per se from the e?ects of the
total irradiance dose. There was no dark period in any of the
treatments. The periods of low and high irradiance were of equal
duration. At low average irradiance (25 lmol quanta m
)2
s
)1
),
light varied between 15 and 35 lmol quanta m
)2
s
)1
, growth-lim-
iting and subsaturating irradiances as inferred from growth-irra-
diance curves of selected species (E. Litchman, unpublished work).
At high average irradiance (100 lmol quanta m
)2
s
)1
), light fluc-
tuated between 65 and 135 lmol quanta m
)2
s
)1
, subsaturating
and saturating irradiances, respectively. Fluctuations in irradiance
248
were achieved by periodically turning on and o? additional light
sources. Light was provided by Philips ??cool white?? fluorescent
tubes. Irradiance was measured with a spherical quantum sensor
(LI-COR 185 B).
The algae were grown in WC freshwater medium (Guillard
1975) in 1 l Erlenmeyer flasks (300 ml culture volume) at 20?C, a
temperature suitable for most freshwater species (Reynolds 1987).
Each treatment had three replicates. Species were inoculated to
have the same order of magnitude biovolume contribution to
community. However, due to cell clumping in the inocula of some
species, initial biomasses of species di?ered more than expected
(Table 1). The initial cell or filament densities, due to cell size dif-
ferences, also varied significantly among species (see Results). Once
per day, 90 ml of the culture was replaced by fresh sterilized me-
dium (dilution rate of 0.3 day
)1
). Cultures were swirled several
times daily. Optical densities of cultures were measured at 750 nm
with a spectrophotometer (Shimadzu UV160, 1 cm path length) to
estimate light attenuation in the cultures. Phosphorus (PO
4
-P) and
nitrogen (NO
3
-N and NH
4
-N) concentrations were measured ac-
cording to Strickland and Parsons (1972) after 2 and 3 weeks from
the start to check for the nutrient-replete conditions. The experi-
ment was run for 3 weeks; samples were taken once per week,
preserved with Lugol?s iodine and counted using the inverted mi-
croscope technique (Lund et al. 1958). For abundant species (with
proportions >0.1 of the total cell counts) at least 400 cells or
filaments were counted, and for rare species the entire counting
chamber was screened. Biovolumes of each species were calculated
from cell measurements of at least 20 individuals using appropriate
geometric formulae (Wetzel and Likens 1991). Shannon diversity
indices were determined for both cell densities and biomasses
(expressed as biovolumes) (Sommer 1995).
Statistical analyses
To test the e?ects of average light level and degree of variability in
light supply on density of each species in the community, I used
two-way ANOVA on log-transformed densities with ??fluctuation
regime?? and ??average light level?? as main factors. If the e?ect of
fluctuaton regime was significant, within each of the two average
light levels cell densities were compared among di?erent fluctuation
regimes separately for each species using one-way ANOVA. The
Student-Neuman-Keuls test was used for multiple pairwise com-
parisons. A similar analysis was performed on biovolumes of major
taxonomic groups and on diversity indices.
Two-species community experiment
Because the sensitivity of species in a community as well as of
aggregate community characteristics to fluctuations might depend
on species richness (e.g., Tilman 1996), I tested the e?ects of the
same light regimes in a simple community consisting of two species
that were dominant in the multispecies community experiment,
Phormidium luridum and Nitzschia sp. These two species were
grown under the same six light regimes that were used in the
multispecies community experiment. The initial species densities
were also similar to their initial densities in the multispecies ex-
periment (c. 10
3
ml
)1
). Species densities and their relative abun-
dances in each treatment were determined after 2 and 3 weeks from
the start. Counts were done using a Sedgewick-Rafter chamber,
and at least six random transects were counted for each sample
(Wetzel and Likens 1991). The e?ects of average irradiance and
fluctuation regime on species abundances and diversity were tested
as in the multispecies community.
Growth rate experiment
The e?ect of light fluctuations on growth rates of the two species
that were dominant in the multispecies community experiment,
Phormidium luridum and Nitzschia sp., was investigated in batch
culture experiments (Kilham 1978). The imposed fluctuations,
medium, temperature and light sources were as in the community
dynamics experiments. Monocultures of each species were grown in
1-l Erlenmeyer flasks (750 ml culture volume) with three replicates
of each treatment. The initial cell densities were c. 50 cells ml
)1
.
Experiments were run for 6 or 7 days, and samples were taken
every day or every 2 days. Growth rates were determined as slopes
of a least-squares linear regression of the ln-transformed cell con-
centrations versus time and compared among treatments using
two-way ANOVA. To test whether the observed di?erences in
Table 1. List of species used in the multispecies experiment (ATCC
American Type Culture Collection, LCCC the Loras College
Culture Collection, Iowa, U of MN University of Minnesota Plant
Biology Department, UTEX the Culture Collection of algae at the
University of Texas, Austin, Squaw L., WI isolated by the author)
Species Source Culture
number
Abbreviations
in figures
Initial biovolume,
10
5
lm
3
ml
)1
Bacillariophyceae
Asterionella formosa Hassal LCCC 384 Ast 0.6 ? 0.2
Cyclotella meneghiniana Kutzing LCCC 868 Cycl 4.3 ? 1
Fragilaria crotonensis Kitton LCCC 369 Frag 6 ? 0.6
Navicula sp. U of MN Nav 2 ? 0.4
Nitzschia sp. U of MN Nitz 12 ? 2
Cyanophyceae
Anabaena flos-aquae (Lyng.) Brebisson ATCC 22664 Anab 6 ? 0.9
Gloeocapsa sp. U of MN Gloeo 0.4 ? 0.3
Phormidium luridum var. olivace Boresch UTEX 426 Phorm 20 ? 3
Oscillatoria tenuis Ag. UTEX 1566 Osc 2 ? 0.4
Chlorophyceae
Chlamydomonas reinhardtii Dang. UTEX 89 Chlam 2 ? 0.5
Cosmarium botrytis Meneghini UTEX 301 Cos 0.02 ? 0.01
Mougeotia sp. UTEX LB 758 Moug 12 ? 2
Pandorina morum (Mu
?
ller) Bory UTEX 788 Pand 21 ? 3.6
Pediastrum boryanum (Turp.) Menegh. Squaw L.,WI Ped 0.6 ? 0.1
Scenedesmus quadricauda (Turp.) Bre? b. UTEX 614 Scen 14 ? 3
Selenastrum capricornutum Printz UTEX 1648 Sel 1 ? 0.2
Sphaerocystis schroeteri Chodat U of MN Sph 30 ? 4.7
Staurastrum gracile Ralfs UTEX LB 562 Staur 8 ? 1.6
249
growth rates are associated with di?erences in chlorophyll con-
centration (Richardson et al. 1983), chlorophyll a concentration of
monocultures and cell biovolumes were determined for each
treatment in the end of experiment (Wetzel and Likens 1991).
Chlorophyll a concentrations per cell and cell biovolumes were
compared among treatments using two-way ANOVA.
Results
In all three experiments (monocultures, two-species and
multispecies communities) both average light level and
its temporal variability a?ected various population and
community characteristics. These e?ects will be de-
scribed below for each experiment type. However, be-
cause the main goal was to test the e?ects of temporal
variability of light supply, the results will be presented to
highlight those e?ects (e.g., separate graphs for two
average light levels).
Multispecies community experiment
Taxonomic composition and total biovolume
The temporal regime of light supply had a significant
e?ect on the taxonomic composition of the community.
Biovolume contribution of green algae and diatoms, but
not cyanobacteria, depended on the fluctuation regime.
At low average irradiance after 2 weeks, green algae had
higher absolute biovolume in constant than in fluctuat-
ing light (two-way ANOVA, P < 0.05). In contrast,
diatoms at low average irradiance had significantly
higher biovolume under fluctuating light, becoming the
dominant group under those light regimes (Fig. 1). At
high average irradiance, diatoms were the least abund-
ant taxon in all three light regimes, and had their lowest
abundance under constant light (Fig. 1). Green algae
and cyanobacteria had similar biovolumes in the com-
munity in all three fluctuation regimes (Fig. 1). At high
average irradiance there were no statistically significant
di?erences in biovolumes among di?erent fluctuation
regimes for any of the taxonomic groups (Fig. 1). The
e?ect of the average light level was significant
(P < 0.05) for all taxa after both 2 and 3 weeks from
the start.
The temporal regime of light supply did not have a
significant e?ect on total biovolume. Total biovolume
after 2 weeks was, however, significantly higher
(P < 0.05) under high average irradiance than under
low average irradiance: on average it increased from
1.5 ? 10
7
lm
3
ml
)1
at the start of the experiment to
2 ? 10
7
lm
3
ml
)1
at low average irradiance and to
4.4 ? 10
7
lm
3
ml
)1
at high average irradiance. From 2
to 3 weeks, average total biovolume did not change
significantly at low average irradiance (P > 0.05) and
declined (P < 0.05) at high average irradiance to the
average of 2.9 ? 10
7
lm
3
ml
)1
, so that the di?erences
between high and low irradiance levels were no longer
significant.
Species abundances
Densities of all species depended significantly on the av-
erage irradiance level, as determined by two-way ANO-
VA. Some species were also sensitive to the degree of
variability in light supply and the interaction of these two
factors. At low average irradiance light fluctuation re-
gime significantly a?ected densities of seven species after
2 weeks and of five species after 3 weeks (Fig. 2a,b). At
high average irradiance, the e?ect of light fluctuations on
species densities was less pronounced: only one species
after 2 weeks and two species after 3 weeks were sensitive
to the degree of variability in light supply (Fig. 2c,d).
Species diversity and richness
Species diversity in the community depended signifi-
cantly on the degree of variability in light supply. At low
average irradiance, species diversity measured by Shan-
non index (calculated based on biomass) declined during
experiment and was significantly lower under fluctuating
light than under constant light (Fig. 3). At high average
irradiance diversity declined even more rapidly, but, in
contrast to low irradiance treatments, diversity was
significantly higher under fluctuating light (Fig. 3).
Shannon diversity index calculated from cell densities
exhibited similar trends. The number of species present
(species richness) in the samples was not significantly
di?erent (compared by one-way ANOVA) between the
fluctuating and constant light treatments at low or at
high average irradiance.
Fig. 1a,b. Biovolumes of three taxa in the multispecies experiment at
the start and after 3 weeks under di?erent light regimes: a low average
irradiance (25 lmol quanta m
)2
s
)1
); b high average irradiance
(100 lmol quanta m
)2
s
)1
)
250
Species richness declined over the course of the ex-
periment. At low average irradiance after 2 weeks,
densities of 7 species had declined from initial densities
in all three treatments (Fig. 2a), and after 3 weeks 9
species had declined or completely disappeared in one or
more treatments (Fig. 2b). At high average irradiance
species loss occurred faster and was more pronounced
than at low average irradiance: after 2 weeks at high
average irradiance, 10 species had completely disap-
peared from samples (Fig. 2c), and after 3 weeks, 11
species were absent from the samples and 3 more species
were absent in some treatments (Fig. 2d). Species di-
versity and richness were significantly higher (P < 0.05)
in the low light treatments than in high light treatments.
Two-species community experiment
The two-species assemblage was also sensitive to the
temporal regime of light supply. After 3 weeks, the
density of Nitzschia was significantly higher under fluc-
tuating than in constant light at both low and high
average irradiances, and its abundance was the highest
under low fluctuating light (Table 2). After 2 weeks
Fig. 2a?d Cell densities of in-
dividual species (mean ? 1 SE)
in the community in the course
of experiment in low average
irradiance treatments,
25 lmol quanta m
)2
s
)1
: a
after 2 weeks, b after 3 weeks;
and in high average irradiance
treatments, 100 lmol quan-
ta m
)2
s
)1
: c after 2 weeks and
d after 3 weeks. The initial
densities of each species are also
shown. Species abbreviations
are given in Table 1. Means
with no common letters above
the bars are significantly
di?erent (P < 0.05)
251
Phormidium had significantly lower density in the 8 h
fluctuating light treatment at low average irradiance; at
high irradiance, however, no significant di?erences in
filament density were found (Table 2). Nitzschia?s rela-
tive abundance also responded positively to light fluc-
tuations (Table 2). At low average irradiance, after
3 weeks Nitzschia and Phormidium had approximately
equal densities in the fluctuating light treatments. At
high average irradiance Phormidium was the dominant
species in all treatments, but the density of Nitzschia was
still higher under fluctuating light than under constant
light (Table 2). After 3 weeks, the Shannon diversity
index, which is a measure of species evenness in case of
constant number of species, was significantly higher in
the fluctuating light treatments at both low and high
average irradiances and the period of fluctuations did
not have a significant e?ect (Table 2). The average light
level also had a significant e?ect on community com-
position: after 2 weeks, densities of both Nitzschia and
Phormidium, as well as total biovolume were signifi-
cantly higher under high average irradiance, as tested by
two-way ANOVA.
Growth rate experiment
Growth rates of Nitzschia and Phormidium were higher
at high average irradiance; they also depended on tem-
poral variability in light supply (Fig. 4). The responses
to light fluctuations were species-specific and depended
on the average irradiance level (Fig. 4). In the constant
light treatment, growth rate of Phormidium was signifi-
cantly higher than the growth rate of Nitzschia at both
low and high average irradiances (P? 0.0004 and
P? 0.02 respectively). In the fluctuating light treat-
ments, growth rates of the two species were not signifi-
cantly di?erent (P > 0.05). Interspecific similarity in
Fig. 3. Species diversity in the multispecies community experiment.
Shannon diversity index (mean ? 1 SE) was calculated based on
species biovolumes. Within each light level treatments di?ering in
fluctuation regime were compared by one-way ANOVA. Significant
di?erences between treatments are indicated by di?erent letters above
the bars
Table 2. Results of the two-species experiment. Means of three
replicates ? SE are given (low light 25 lmol quanta m
)2
s
)1
average irradiance, high light 100 lmol quanta m
)2
s
)1
average
irradiance, CL constant light treatment, FL1 fluctuating light
treatment of 1 h period, FL8 fluctuating light treatment of 8 h
period). Values within each light level were compared across
treatments with one-way ANOVA; proportions were arcsine
square-root transformed before the analysis
At the start CL FL1 FL8 P
Nitzschia density, 2.5 ? 1 Low light 2 weeks 6.4 ? 0.7 8.5 ? 2 5.5 ? 1 NS
10
4
cells ml
)1
3 weeks 19 ? 0.8 43 ? 9 58 ? 2 0.008
High light 2 weeks 39 ? 15 26 ? 3 37 ? 2 NS
3 weeks 7.4 ? 2 12 ? 0.4 15 ? 2 0.05
Phormidium density, 2.6 ? 1 Low light 2 weeks 9.7 ? 1 10 ? 1.2 4.4 ? 0.7 0.003
10
4
filaments ml
)1
3 weeks 62 ? 1 43 ? 9 47 ? 3 NS
High light 2 weeks 51 ? 3 47 ? 3 50 ? 4 NS
3 weeks 56 ? 8 44 ? 9 49 ? 8 NS
Proportion of Nitzschia 0.49 ? 0.01 Low light 2 weeks 0.4 ? 0.03 0.46 ? 0.04 0.56 ? 0.01 0.023
(density-based) 3 weeks 0.24 ? 0.01 0.5 ? 0.03 0.55 ? 0.02 <0.0001
High light 2 weeks 0.44 ? 0.01 0.36 ? 0.02 0.42 ? 0.02 0.05
3 weeks 0.11 ? 0.01 0.21 ? 0.01 0.24 ? 0.01 <0.001
Total biovolume, 1.3 ? 0.1 Low light 2 weeks 4.3 ? 0.5 4.8 ? 0.4 2.5 ? 0.4 0.02
10
7
lm
3
ml
)1
3 weeks 24 ? 5 20 ? 4.7 23 ? 1.4 NS
High light 2 weeks 23 ? 3 22 ? 1 26 ? 1.5 NS
3 weeks 20 ? 3 16 ? 3 18 ? 3 NS
Shannon diversity index 0.69 ? 0.02 Low light 2 weeks 0.67 ? 0.1 0.69 ? 0.01 0.69 ? 0.07 NS
(density-based) 3 weeks 0.55 ? 0.05 0.69 ? 0.06 0.68 ? 0.06 0.03
High light 2 weeks 0.69 ? 0.03 0.65 ? 0.03 0.68 ? 0.03 NS
3 weeks 0.36 ? 0.01 0.52 ? 0.04 0.56 ? 0.06 0.03
Shannon diversity index 0.74 ? 0.03 Low light 2 weeks 0.60 ? 0.05 0.44 ? 0.03 0.68 ? 0.08 0.04
(biovolume-based) 3 weeks 0.44 ? 0.04 0.66 ? 0.06 0.68 ? 0.05 0.01
High light 2 weeks 0.63 ? 0.06 0.57 ? 0.08 0.62 ? 0.07 NS
3 weeks 0.26 ? 0.01 0.41 ? 0.05 0.45 ? 0.04 0.02
NS P>0.05
252
growth rates was the greatest under low average irradi-
ance.
For both Nitzschia and Phormidium, cell biovolumes
did not di?er significantly among treatments either
within or between the two average irradiances. Chloro-
phyll a concentrations per cell were not significantly
di?erent among treatments within one average irradi-
ance level. The average cellular chlorophyll a concen-
trations were, however, significantly higher under low
average irradiance than under high average irradiance:
1.6 ? 10
)3
lg Chl a cell
)1
versus 0.6 ? 10
)3
lg Chl a/
cell (P? 0.002) for Phormidium and 1.5 ? 10
)3
lg
Chl a cell
)1
versus 1.1 ? 10
)3
lg Chl a cell
)1
(P? 0.01)
for Nitzschia.
Discussion
Multispecies community experiment
Variability in light supply (constant versus fluctuating
light) had a significant e?ect on the structure and dy-
namics of a laboratory multispecies community of
phytoplankton. The abundance of the three taxonomic
groups (diatoms, cyanobacteria, and green algae) in the
community di?ered among treatments with the same
average irradiance but di?erent fluctuation regimes.
Diatoms had higher density and biovolume under fluc-
tuating light, reaching their maximum abundance under
low fluctuating light, where they were the dominant
group (Fig. 1). Diatom blooms in lakes often occur
when light conditions are highly variable and average
irradiances are low (Reynolds 1994), e.g., during spring
and fall turnover when cells circulate through the whole
water column. Better performance of diatoms (higher
densities and proportions in communities) under mixing
conditions is usually attributed to the positive e?ect of
turbulence keeping heavy-celled diatoms suspended in
the water column (Reynolds 1984; Ki?rboe 1993). In
addition, as the experiments described above suggest,
fluctuating light itself may stimulate growth of diatoms.
This could be the result of an adaptation to fluctuating
light associated with turbulent conditions which are
favorable for diatoms (Ki?rboe 1993). In enclosure ex-
periments on artificial mixing, Reynolds et al. (1984) had
found that diatoms were dominant under mixing con-
ditions, when the light climate was highly variable, while
green algae, including Sphaerocystis, grew better during
quiescent periods. In the present study, some green al-
gae, including Sphaerocystis, also had higher abundance
under constant than fluctuating light of low average
level.
Within each group, however, species responses to a
particular light regime were not uniform: both ampli-
tude and direction of the responses were species-specific.
Among diatoms, Nitzschia and Fragilaria had higher
densities under fluctuating light, while densities of As-
terionella, Cyclotella, and Navicula either did not di?er
among light regimes or were lower under fluctuating
light (Fig. 2b). Among cyanobacteria, Phormidium and
Anabaena at low average irradiance had higher densities
under constant light, but Gloeocapsa and Oscillatoria
were more abundant under fluctuating light. Interest-
ingly, Oscillatoria is known to respond positively to
turbulence (Reynolds 1987; Klemer and Barko 1991).
Moreover, Gibson (1985) found that growth rate of
Oscillatoria agardtii was slightly inhibited by continuous
versus intermittent light, which is consistent with lower
abundance of Oscillatoria under constant light observed
in this study. Among green algae, at low irradiance
Sphaerocystis and Pandorina had higher densities under
constant regime, while densities of Pediastrum, Scene-
desmus and Selenastrum did not di?er among constant
and fluctuating light treatments. Therefore, species-spe-
cific di?erences should be taken into consideration when
determining the e?ects of the light supply dynamics on
species even within a taxonomic group.
In contrast to species abundances, total biomass
measured as biovolume depended only on the average
irradiance, and not on the dynamics of light supply. The
lower sensitivity of total biomass to environmental
variation, compared to individual species sensitivity,
seems to be due to di?erential responses of species
compensating each other and thus reducing the net e?ect
on community biomass. This may be a general property
of communities (May 1974; Tilman 1996). Abundances
of some species were still changing between 2 and
3 weeks (Fig. 2), so that at the end of experiment com-
munity composition possibly was not at equilibrium, at
least in the low-light treatments. The goal of this study,
however, was not to determine the steady-state com-
munity composition, but rather to examine whether and
how a multispecies community would respond to dif-
ferences in light regime over a relatively short time. As
these experiments demonstrate, community composition
is sensitive to the degree of temporal variability in light
Fig. 4. Growth rates (mean ? 1 SE) of Nitzschia and Phormidium
under di?erent light regimes. For each species growth rates within
each average irradiance were compared by one-way ANOVA.
Significant di?erences between treatments are indicated by di?erent
letters above the bars
253
supply: di?erences were detected as soon as 2 weeks
from the start.
Species richness and diversity declined during the
experiment in both high and low light treatments. Spe-
cies that declined or disappeared quickly from all
treatments (e.g., Cosmarium, Mougeotia, and Staura-
strum) had low initial densities and possibly low growth
rates (large-celled species tend to have lower growth
rates; Reynolds 1984) and could have been washed out.
However, species loss was more pronounced in the high
average irradiance treatments, where the growth rates of
species were higher. This suggests that some species
might have been competitively excluded and that com-
petitive exclusion occurred faster under high average
irradiance. Some theoretical models also suggest that the
rates of competitive exclusion may be greater when
growth rates of competing species are high (e.g., Huston
1994).
At low average irradiance, light was limiting, as
indicated by higher total biomass under high average
irradiance than at low average irradiance. Contrary to
theoretical considerations (e.g., Levins 1979) and
experiments with nitrogen- or phosphorus-limited phy-
toplankton (Turpin and Harrison 1979; Sommer 1984),
fluctuations at this limiting irradiance did not increase
the number of coexisting species, and species diversity
was significantly lower under fluctuating low irradiance
(Fig. 3). At high average irradiance, however, diversity
was higher under fluctuating light (Fig. 3).
Several explanations for lack of a positive e?ect of
fluctuations on diversity and species richness at low ir-
radiance are possible. A positive e?ect of a fluctuating
light supply on diversity might have been noticeable in
these experiments after a longer time, as the community
approached equilibrium species composition. In the high
average light treatments, where growth rates and con-
sequently, community changes were faster than at low
irradiance, fluctuations increased diversity, which agrees
with this hypothesis.
An alternative explanation is that only a limited
range of frequencies, amplitudes and average irradiances
of light fluctuations can promote coexistence (E. Litch-
man and C. Klausmeier, unpublished work), and that
the chosen fluctuations were not in this range. In a
theoretical analysis, Grover (1991b) showed that under
nutrient competition the range of nutrient fluctuations
leading to coexistence may also be very narrow. In his
experiments, Sommer (1984, 1995) found a maximum
species diversity when phosphorus pulses or dilution
disturbances had a 1-day period, comparable to the
generation times of phytoplankton. Similarly, Kemp
and Mitsch (1979) showed in numerical simulations that
species coexistence occurred when the frequency of tur-
bulence had approximately a 1-day period. It is possible
that fluctuations in irradiance with a 1-day period (e.g.,
the daily light:dark cycle) could promote coexistence.
Brzezinski and Nelson (1988) found that such a light:-
dark cycle of non-limiting irradiance level facilitated
coexistence of species under nitrogen limitation. It re-
mains to be explored whether 1-day or longer periods of
fluctuation in irradiance can promote coexistence under
light-limited conditions.
A third explanation is that positive e?ects of fluctu-
ations on diversity are greater in a system with strong
competition. It is possible that in the low-light cultures,
despite light limitation, competition for light was not
severe, i.e., mutual shading was not high enough for
species to compete strongly for light. The optical den-
sities of the low-light cultures measured at 750 nm were
in the range of 0.08?0.12. Usually higher optical densi-
ties (>0.3?0.4) are required for light competition by
mutual shading (i.e., resource consumption) to be signi-
ficant (J. Huisman, personal communication). Nitrogen
and phosphorus concentrations were well above limiting
levels, c. 25 lmol ? l
)1
of PO
4
-P and 70 lmol ? l
)1
of
dissolved inorganic N (DIN; NO
?
3
-N+NH
?
4
-N) after
3 weeks, so that competition for these nutrients was also
unlikely.
In contrast to low light treatments, at high average
irradiance diversity was significantly higher under fluc-
tuating than under constant light (Fig. 3). Nitrogen and
phosphorus concentrations measured after 2 and 3 weeks
from the start of the experiment were above the limiting
levels (c. 20 lmol ? l
)1
of PO
4
-P and 60 lmol ? l
)1
of
DIN after 3 weeks), so that as in low-light treatments,
limitation by these nutrients was unlikely. However,
light attenuation by biomass was greater than in low-
light cultures (optical densities were 0.2?0.25 after
2 weeks), so that mutual shading was greater and species
might have competed for light. It is also possible that the
system was carbon-limited: the average pH in cultures
was around 9.5 (due to photosynthesis by a dense
community in a weakly bu?ered medium), so that
inorganic carbon was predominantly in the form of
carbonate and bicarbonate ions (Stumm and Morgan
1981), which are less readily taken up by phytoplankton
than is carbon dioxide (Maberly and Spence 1983). Such
conditions of considerable light attenuation by biomass
and low availability of inorganic carbon can be common
in lakes during bloom events. Cyanobacteria were the
most abundant group in all high-light treatments in
contrast to low-light treatments, where average pH was
below 9. These algae are known to be superior com-
petitors for inorganic carbon at high pH (Shapiro 1973;
Talling 1976). Diatoms, which generally are poor com-
petitors for inorganic carbon (Talling 1976), had low
abundance under high average irradiance. It is, there-
fore, possible that light fluctuations increased diversity
in a system limited by another resource, by a?ecting
competition for that resource. The importance of light
fluctuations in systems limited by CO
2
has been previ-
ously discussed by Gallegos et al. (1980). Under carbon
limitation, the positive e?ect of light fluctuations on
diversity is similar to the results of Brzezinski and Nel-
son (1988) for a simple two-species diatom community
where light fluctuations enabled coexistence of both
species competing for ammonium, while constant light
led to competitive exclusion. Further experiments are
254
needed to determine the exact mechanism of the diver-
sity enhancement by light fluctuations.
The two-species community experiment
The responses of species in the two-species community
were qualitatively similar to their responses in the mul-
tispecies community, e.g., higher density and relative
abundance of Nitzschia in the fluctuating light treat-
ments (Fig. 2, Table 2). Such similarity in the e?ects of
fluctuations on species in communities with di?erent
numbers of species indicates robustness of the observed
patterns. Species densities and especially relative abun-
dances were more sensitive than total biomass to light
fluctuations. Diversity (species evenness) was higher
under fluctuating light, at both high and low average
irradiances, because Nitzschia was more abundant under
fluctuating light. Low light treatments in this experiment
had greater light attenuation and possibly stronger
competition than the corresponding treatments in the
multispecies experiment, which could have led to higher
diversity.
Growth rate responses to fluctuating light
The growth rates of Nitzschia and Phormidium in
monocultures were sensitive to the regime of light sup-
ply. Growth-rate responses to light fluctuations were
species-specific and depended on the average irradiance
level and the frequency of fluctuations. Because growth
rate of phytoplankton directly depends on photosyn-
thesis, respiration and pigment turnover (Cullen 1990),
fluctuations a?ecting each of these processes may a?ect
growth. In the future, it would be of interest to clarify
further the link between physiology and population
dynamics by identifying the physiological processes that
led to the observed di?erences in growth rates.
The lack of significant di?erences in cellular chloro-
phyll a concentration among di?erent fluctuation
regimes indicates that processes other than chlorophyll
a turnover were responsible for di?erences in growth
rates. Both species showed a decrease in cellular chlo-
rophyll a concentration with increasing average irradi-
ance, which is in agreement with the general trend
observed for many microalgae (Richardson et al. 1983).
This suggests that these species might acclimate to av-
erage irradiance level rather than to the maximum ir-
radiance they experience. Similarly, Ibelings et al. (1994)
found that the cyanobacterium Microcystis sp. re-
sponded to the total daily irradiance dose and not to the
maximum irradiance. The generality of such physiolog-
ical strategy across di?erent taxonomic groups needs
further investigation.
Di?erential responses of the growth rates of individ-
ual species to light fluctuations are likely to be the major
cause for observed di?erences at the community level
(cf. Tilman et al. 1982), assuming that loss rates were
similar for all species. There was good agreement be-
tween the response pattern of species in monoculture
and their abundance in both two- and multispecies
community: under low average irradiance, Phormidium?s
growth was inhibited by fluctuations (Fig. 4) and its
abundance in both communities was lower under fluc-
tuating light (Fig. 2, Table 2). When two species are
compared, there is also a good agreement between their
growth rates and relative dominance in communities: at
low average irradiance, under constant light the growth
rate of Nitzschia in monoculture was significantly lower
than the growth rate of Phormidium, while under fluc-
tuating light, growth rates of the two species were not
significantly di?erent (Fig. 4). This agrees well with
species densities in the two species experiment: Nitzschia
had lower than Phormidium density under low constant
light, and densities similar to the Phormidium?s under
low fluctuating light of both periods (Table 2). In the
multispecies community, this correspondence might not
hold because of the influence of other species (e.g.,
competitive interactions). Nitzschia had higher growth
rate in monoculture under high average irradiance but
its abundance in the multispecies experiment was higher
under low than under high average irradiance. A direct
comparison of the growth rate experiments and the
community dynamics experiments might also be com-
plicated by the fact that the growth rate experiments
were run as batch cultures, and the community experi-
ments were run as semicontinuous cultures; also the
duration of the experiments was di?erent. It is possible
that a long time-scale (>1 week) photoacclimation to a
particular light regime (Geider et al. 1996) led to changes
in growth rates which were not detected in the 1-week
growth-rate experiments, but could have been important
in the community dynamics experiments. These factors
may be responsible, for example, for the discrepancy
between relatively low growth rate of Nitzschia in batch
monocultures (c. 0.3 day
)1
) (Fig. 4) and its high density
in the community experiments with a dilution rate of
0.3 day
)1
(Fig. 2, Table 2).
In summary, the results of the community and
growth rate experiments demonstrate that light fluctua-
tions may a?ect multiple levels of phytoplankton orga-
nization, including the population and community
levels. Light fluctuations may change species abun-
dances and dominance patterns in both simple and
multispecies communities (e.g., diatom versus cyano-
bacterial dominance). Community changes may be
caused by di?erential e?ects of fluctuations on growth
rates of species and on their competitive interactions.
Light fluctuations may also increase species diversity.
Thus, variability in light may o?er one more explanation
for the ??paradox of the plankton?? (Hutchinson 1961),
the coexistence of many species of phytoplankton in a
seemingly homogeneous environment. The sensitivity of
phytoplankton communities to temporal regime of light
supply suggests that di?erent regimes of vertical mixing
(e.g., seasonal patterns) may cause shifts in phyto-
plankton communities, by altering not only the average
255
irradiance but light variability as well. Thus, better un-
derstanding of changes in phytoplankton communities
may require an explicit consideration of temporal het-
erogeneity in light supply.
Acknowledgements This research was in part supported by the
NSF Doctoral Dissertation Improvement Grant (DEB 94-23531),
the Sigma Xi Society and Dayton-Wilkie Fund of the Bell Museum
of Natural History awards. Comments by J. Huisman, C. Klaus-
meier, A. Klemer, H. MacIntyre, K. Schulz, J. Shapiro, R. Sterner,
D. Tilman and two anonymous reviewers are greatly appreciated.
References
Anderson TR (1993) A spectrally averaged model of light pene-
tration and photosynthesis. Limnol Oceanogr 38:1403?1419
Armstrong RA, McGehee R (1980) Competitive exclusion. Am
Nat 115:151?170
Beardall J, Burger-Wiersma T, Rijkeboer M, Sukenik A, Lemoalle
J, Dubinsky Z, Fontvielle D (1994) Studies on enhanced
post-illumination respiration in algae. J Plankton Res 16:
1401?1410
Brzezinski MA, Nelson DM (1988) Interactions between pulsed
nutrient supplies and a photocycle a?ect phytoplankton com-
petition for limiting nutrients in long-term culture. J Phycol
24:346?356
Cullen JJ (1990) On models of growth and photosynthesis in
phytoplankton. Deep-Sea Res 37:667?683
Denman KL, Gargett AE (1983) Time and space scales of vertical
mixing and advection of phytoplankton in the upper ocean.
Limnol Oceanogr 28:801?815
Dromgoole FI (1988) Light fluctuations and the photosynthesis of
marine algae. II. Photosynthetic response to frequency, phase
ratio and amplitude. Funct Ecol 2:211?219
Ferris JM, Christian R (1991) Aquatic primary production in re-
lation to microalgal responses to changing light: a review.
Aquat Sci 53:187?217
Gallegos CL, Hornberger GM, Kelly MG (1980) Photosynthesis-
light relationships of a mixed culture of phytoplankton in
fluctuating light. Limnol Oceanogr 25:1082?1092
Geider RJ, MacIntyre HL, Kana TM (1996) A dynamic model
of photoadaptation in phytoplankton. Limnol Oceanogr 41:
1?15
Gemerden H van (1974) Coexistence of organisms competing for
the same substrate: an example among the purple sulfur bac-
teria. Microb Ecol 1:104?119
Gibson CE (1985) Growth rate, maintenance energy and pigmen-
tation of planktonic Cyanophyta during one-hour light:dark
cycles. Br Phycol J 20:155?161
Gibson CE, Foy RH (1988) The significance of growth rate and
storage products for the ecology of Melosira italica ssp. sub-
arctica in Lough Neagh. In: Round FE (ed) Algae and the
aquatic environment. Biopress, Bristol, pp 88?106
Grobbelaar JU, Kroon BMA, Burger-Wiersma T, Mur LR (1992)
Influence of medium frequency light/dark cycles of equal du-
ration on the photosynthesis and respiration of Chlorella
pyrenoidosa. Hydrobiologia 238:53?62
Grover JP (1991a) Non-steady state dynamics of algal population
growth: experiments with two chlorophytes. J Phycol 27:70?79
Grover JP (1991b) Resource competition in a variable environ-
ment: phytoplankton growing according to the variable-inter-
nal-stores model. Am Nat 138:811?835
Guillard RRL (1975) Cultures of phytoplankton for feeding
marine invertebrates. In: Smith WL, Chanley MH (eds) Cul-
ture of marine invertebrate animals. Plenum, New York, pp
29?60
Harris GP, Piccinin BB (1977) Photosynthesis by natural phyto-
plankton populations. Arch Hydrobiol 80:405?456
Huisman J, Weissing FJ (1994) Light-limited growth and compe-
tition for light in well-mixed aquatic environments: an ele-
mentary model. Ecology 75:507?520
Huston MA (1994) Biological diversity: the coexistence of species
on changing landscapes. Cambridge University Press, Cam-
bridge
Hutchinson GE (1961) The paradox of the plankton. Am Nat
95:137?145
Ibelings BW, Kroon BMA, Mur LR (1994) Acclimation of pho-
tosystem II in a cyanobacterium and a eukaryotic green alga to
high and fluctuating photosynthetic photon flux densities,
simulating light regimes induced by mixing in lakes. New Phytol
128:407?424
Imberger J (1985) Thermal characteristics of standing waters: an
illustration of dynamic processes. Hydrobiologia 125:7?29
Kemp WM, Mitsch WJ (1979) Turbulence and phytoplankton di-
versity: a general model of the ??paradox of plankton??. Ecol
Model 7:201?222
Kilham SS (1978) Nutrient kinetics of freshwater planktonic algae
using batch and semicontinuous cultures. Mitt Int Ver Limnol
21:147?157
Ki?rboe T (1993) Turbulence, phytoplankton cell size, and the
structure of pelagic food webs. Adv Mar Biol 29:1?72
Klemer A, Barko J (1991) E?ects of mixing and silica enrichment on
phytoplankton seasonal succession. Hydrobiologia 210:171?181
Kroon BMA, Latasa M, Ibelings BW, Mur LR (1992) The e?ect
of dynamic light regimes on Chlorella. I. Pigments and cross
sections. Hydrobiologia 238:71?78
Levins R (1979) Coexistence in a variable environment. Am Nat
114:765?783
Lund JW, Kipling C, LeCren ED (1958) The inverted microscope
method of estimating algal numbers and statistical basis of
estimations by counting. Hydrobiologia 11:143?170
Maberly SC, Spence DHN (1983) Photosynthetic inorganic carbon
use by freshwater plants. J Ecol 71:705?724
Marra J (1978) Phytoplankton photosynthetic response to vertical
movement in mixed layer. Mar Biol 46:203?208
May RM (1974) Stability and complexity in model ecosystems, 2nd
edn. Princeton University Press, Princeton
Neale PJ, Marra J (1985) Short-term variation of Pmax under
natural irradiance conditions: a model and its implications. Mar
Ecol Prog Ser 26:113?124
Prezelin BB, Tilzer MM, Schofield O, Haese C (1991) The control
of the production process of phytoplankton by the physical
structure of the aquatic environment with special reference to
its optical properties. Aquat Sci 53:136?151
Que? guiner B, Legendre L (1986) Phytoplankton photosynthetic
adaptation to high frequency light fluctuations simulating those
induced by sea surface waves. Mar Biol 90:483?491
Reynolds CS (1984) The ecology of freshwater phytoplankton.
Cambridge University Press, Cambridge
Reynolds CS (1987) Cyanobacterial water-blooms. Adv Bot Res
13:67?143
Reynolds CS (1994) The role of fluid motion in the dynamics of
phytoplankton in lakes and rivers. In: Giller PS, Hildrew AG,
Ra?aelli DG (eds) Aquatic ecology: scale, pattern and process.
Blackwell, Oxford, pp 141?188
Reynolds CS, Wiseman SW, Clarke MJO (1984) Growth- and loss-
rate responses of phytoplankton to intermittent artificial mixing
and their potential application to the control of planktonic algal
biomass. J Appl Ecol 21:11?39
Richardson K, Beardall J, Raven JA (1983) Adaptation of uni-
cellular algae to irradiance: an analysis of strategies. New
Phytol 93:157?191
Shapiro J (1973) Blue-green algae: why they become dominant.
Science 179:382?384
Sommer U (1984) The paradox of the plankton: fluctuations of
phosphorus availability maintain diversity in flow-through
cultures. Limnol Oceanogr 29:633?636
Sommer U (1985) Comparison between steady and non-steady
state competition: experiments with natural phytoplankton.
Limnol Oceanogr 30:335?346
256
Sommer U (1995) An experimental test of the intermediate dis-
turbance hypothesis using cultures of marine phytoplankton.
Limnol Oceanogr 40:1271?1277
Stramski D, Rosenberg G, Legendre L (1993) Photosynthetic and
optical properties of the marine chlorophyte Dunaliella tertio-
lecta grown under fluctuating light caused by surface-wave fo-
cusing. Mar Biol 115:363?372
Strickland JD, Parsons T (1972) A practical manual of sea-water
analysis, 2nd edn (Publication 167). Fisheries Research Board
of Canada
Stumm W, Morgan JJ (1981) Aquatic chemistry, 2nd edn. Wiley,
New York
Talling JF (1976) The depletion of carbon dioxide from lake water
by phytoplankton. J Ecol 64:79?121
Thornley JHM (1974) Light fluctuations and photosynthesis. Ann
Bot 38:363?373
Tilman D (1982) Resource competition and community structure.
Princeton University
Tilman D (1996) Biodiversity: population versus ecosystem stabil-
ity. Ecology 77:350?363
Tilman D, Kilham SS, Kilham P (1982) Phytoplankton community
ecology: the role of limiting nutrients. Annu Rev Ecol Syst
13:349?372
Turpin DH, Harrison PJ (1979) Limiting nutrient patchiness and
its role in phytoplankton ecology. J Exp Mar Biol Ecol 39:151?
166
Wetzel RG, Likens GE (1991) Limnological analyses, 2nd edn.
Springer, Berlin Heidelberg New York
257