Journal of Plankton Research Vol.18 no.6 pp.1009-1032.1996
Regulation of bacterioplankton production and biomass in an
oligotrophic cleanvater lake—the importance of the
phytoplankton community
Katarina Vrede
Institute of Limnology, Uppsala University, Norbyvagen 20, S-752 36 Uppsala,
Sweden
Abstract. Estimations of bacterioplankton production and biomass were carried out in enclosure
experiments during two consecutive years (1989 and 1990) in oligotrophic clearwater Lake Njupfatet.
The lake was limed in November 1989, and the experiments were carried out both in 1989 (unlimed)
and in 1990 (limed). Bags (—3001) were manipulated with inorganic phosphorus and nitrogen, organic
carbon, and metazoan zooplankton abundance. Both years, bacterial production was stimulated by
inorganic nutrients alone and in combination with organic carbon. However, the increase in bacterial
production when inorganic nutrients were added alone was much stronger in 1990thanin 1989. In 1989.
bacterial production increased strongly only when inorganic nutrients and organic carbon were added
together. The phytoplankton community was dominated by the cyanobacterium Merismopedia tenuissima during 1989, and the phytoplankton biomass increased only slightly when receiving inorganic
nutrients. In 1990, when the lake had been limed, M.tenuissima had completely disappeared and the
phytoplankton community, dominated by Chrysophyceae and Chlorophyceae, responded strongly to
additions of inorganic nutrients. The increased phytoplankton productivity in 1990 may have resulted
in increased release of organic carbon, and this in turn that the carbon limitation of bacterioplankton
production decreased from 1989 to 1990. Zooplankton had a positive effect on bacterioplankton production in 1989, but no effect in 1990. The loss of bacterial biomass approximated 60% of the bacterial
production in 1989, while in 1990 it almost equalled the bacterioplankton production.
Introduction
Both bacterioplankton and phytoplankton utilize inorganic nutrients for their
growth and heterotrophic bacteria require organic carbon as well. Bacteria have,
at least at low concentrations, a more efficient uptake of dissolved inorganic phosphorus than phytoplankton (Currie and Kalff, 1984a; Jansson, 1993) and bacterioplankton have been reported to dominate the uptake of both ammonium
(Wheeler and Kirchman, 1986) and inorganic phosphorus (Currie and Kalff,
1984b) in aquatic systems. These findings indicate that bacterioplankton cover
their nitrogen (N) and phosphorus (P) requirements better than algae and that
organic carbon (C) may be the primary limiting nutrient for bacterioplankton.
However, the C:P and C:N ratios of bacterioplankton vary (Goldman etal., 1987;
Tezuka, 1990) and bacteria at least occasionally are P subsaturated (Vadstein and
Jensen, 1988). Also laboratory experiments have indicated that bacteria can be P
limited (Vadstein and Olsen, 1989). These observations have led to a lessrigidview
of the regulation of bacterioplankton growth, and today P is often considered to be
the most important factor limiting bacterioplankton growth in lakes and brackish
water (Morris and Lewis, 1992; Zweifel et al., 1993).
Bacterioplankton and phytoplankton biomass and production are correlated
frequently (e.g. Bird and Kalff, 1984; Cole etal., 1988), and extracellular release of
C from phytoplankton is an important carbon source for bacterioplankton (Munster and Chr6st, 1990). However, this close correlation between bacterioplankton
© Oxford University Press
1009
K.Vrede
and phytoplankton may be due not only to bacterial dependence on phytoplankton carbon; it may also be a result of direct limitation by bacterio- and phytoplankton of one or several nutrients. In eutrophic Calder Lake, only bacterioplankton
responded in their growth when inorganic P was added alone, whereas both bacterioplankton and phytoplankton responded when inorganic P and N were added
together (Le et al., 1994). However, the latter bacterial response was not an
indirect effect of stimulated phytoplankton growth and increased release of
organic substances that the bacteria could utilize. Rather, it was direct and independent of phytoplankton growth. Also direct limitation of bacterial growth by
inorganic nutrients has been reported from mesotrophic Lake Dillon, where bacterioplankton growth was limited by P alone or in combination with N or C (Morris
and Lewis, 1992), and from the Baltic Sea where bacteria were limited primarily by
P (Zweifel et al., 1993).
The importance of extracellular release of C from phytoplankton varies widely,
but covers an average <50% of the bacterial carbon demand (Baines and Pace,
1991). Phytoplankton species composition can explain some of the variation, as the
quantity and quality of extracellularly released organic compounds can vary
between phytoplankton species (Hellebust, 1965; Sundh, 1992). It has also been
shown that the amount of extracellularly released organic compounds from phytoplankton increases with increasing primary production, but levels off at rates of
high primary production in lakes (Baines and Pace, 1991). In humic lakes, organic
carbon derived from phytoplankton may be less important than in clearwater
lakes, as bacteria can utilize humic substances (Tranvik and Hofle, 1987; Sundh
and Bell, 1992). Other potential carbon sources for bacterioplankton are organic
complexes released from protozoans (Nagata and Kirchmann, 1992) and sloppy
feeding, defecation, and excretion by zooplankton (Lampert, 1978; Jumars et al.,
1989).
Losses of bacterioplankton biomass may be due to several factors, e.g. grazing,
sinking, or cell lysis induced by viruses or parasitic bacteria. Planktonic bacteria
often have high growth rates, but only a small fraction of the carbon produced is
found in the form of new bacterial biomass. Several studies have shown that grazing alone sometimes equals bacterial production (e.g. Sanders etai, 1989; Chrzanowski and Simek, 1993). Whereas the phytoplankton community includes many
organisms which by size, digestibility or toxicity may be well protected against
grazing (i.e. large filamentous or colony forming species), the bacterioplankton
community is grazed effectively by a wide range of organisms. Bacterivory is found
among heterotrophic and mixotrophic flagellates (Pringsheim, 1963), ciliates
(Sanders et al., 1989), and metazoan zooplankton (Glide, 1989; Arndt, 1993).
To study regulation of bacterial production and biomass in oligotrophic systems,
enclosure experiments were carried out in the oligotrophic clearwater Lake Njupfatet (Bell et al., 1993; Blomqvist et al., 1993,1995) during two consecutive years.
Bags were manipulated with inorganic nutrients (nitrogen and phosphorus),
organic carbon (glucose), and metazoan zooplankton abundance.
1010
Regulation of bacterioplanfcton production and biomass
Method
The study site
Lake Njupfatet is a dimictic, oligotrophic, clearwater lake located in central Sweden at 62° Ol'N, 16°23' E. The lake is small (0.12 km2), with a total drainage area of
only 0.62 km2. The lake was naturally slightly acidic until it was limed in November
1989. In total, 16 000 kg offinelyground limestone (calcite) were spread over the
lake by helicopter. Another 4000 kg of calcite were spread over the water outflows
(into the lake). The addition of lime corresponded to 0.17 kg nr2 (Blomqvist et al.,
1995). The experiments were carried out within a project designed to study the
effect of liming combined with nutrient additions on the pelagic ecosystem. In the
present investigation, four experiments were carried out: two in August 1989 when
the lake was acid and two in August 1990 when the lake was limed. The lake was
sampled every other month November-April and twice a month May-October
from November 1988 to October 1990. Composite samples representing 'epilimnion' (0-6 m) and 'metalimnion + hypolimnion' (6-12 m) were taken at five stations in the central part of the lake. Water from different 2-m layers (0-2,2-4,4-6,
6-8,8-10 and 10-12 m) was taken at the stations, corresponding to the proportion
by volume between these layers in the lake. Samples were taken for analyses of
water chemistry, bacterioplankton abundance and biomass, species composition,
abundance and biomass of phytoplankton, mixotrophic and heterotrophic flagellates, ciliates, heliozoans and metazoan zooplankton. For a more detailed description of water chemistry and the pelagic ecosystem of the lake, see Blomqvist et al.
(1995). In addition to enclosure manipulations, carried out in August each year,
the conditions in the lake at the beginning of the experiments are presented in this
paper.
Experimental set-up
The experimental set-up was the same in 1990 as in 1989. Manipulations with metazoan zooplankton (Z), inorganic nitrogen and phosphorus (NP), and organic carbon (C) were carried out in transparent plastic bags fixed on a wood frame. The
enclosures had a surface area of ~0.15 m2 open to the air and a depth of 2.1 m. They
were closed at the bottom and did not touch the sediment. The enclosures were
filled with ~300 1 of lake water taken at 1 m depth. The water had been gently
filtered through a 100 ^.m net to remove larger zooplankton.
Inorganic nitrogen (KNO3) and inorganic phosphorus (KH2PO4) were added to
the enclosures to produce a gradient withfivedifferent concentrations (Figure 1).
The N:P ratio of added nutrients was 10:1 by weight. The concentrations ranged
from one to approximately four times the lake concentration on a total phosphorus
basis and from one to two times the lake concentration on a total nitrogen basis.
The nutrient gradient was set up in duplicate. Zooplankton collected with a 300
jtm net were added to one of the two gradients to approximate the natural lake
concentration; no zooplankton were added to the other gradient. Enclosures 1-10
(Figure 1) belong to this experiment.
The effect of C was studied by adding glucose to two other enclosures. One
received 500 jig glucose-C 1~' only (enclosure 11; Figure 1), while the second
1011
K-Vrede
increasing N and P concentrations
KH 2 PO 4 added(ugP litre 1 )
1
KNO., added (|ig N litre )
5
50
10
15
25
100
150
250
/ooplankton present
(lake concentration)
zooplankton absent
zooplankton absent
glucose added (500 ng C litre ')
Fig. 1. Experimental set-up of the gradient of inorganic nutrients and organic carbon additions in Lake
Njupfatet in 1989 and 1990. Each square corresponds to one enclosure. Enclosure numbers 1-10 are
referred to as the gradient of inorganic nutrients and numbers 6,8,11 and 12 are the experiment with
organic carbon additions.
(enclosure 12) received 500 \ig glucose-C I"1 in addition to inorganic nitrogen
(KN03) and inorganic phosphorus (KH2PO«) in the same concentration as in
enclosure number 8 in the gradient described above. No zooplankton were added
to these enclosures. By comparing the enclosures (number 11 and 12) with number
8 in the nutrient gradient (lacking zooplankton) and the enclosure that had
received neither inorganic nutrients nor zooplankton (number 6), the effect of C
alone and in combination with NP could be studied.
The experiments were run for 3 weeks and the enclosures were sampled every
third day for analyses of water chemistry, bacterioplankton abundance, biomass
and production, species composition, abundance and biomass of phytoplankton,
mixotrophic and heterotrophicflagellates,ciliates, heliozoans and metazoan zooplankton. However, metazoan zooplankton were not determined in the enclosures
receiving organic carbon in 1989, and ciliates, heliozoans and metazoan zooplankton in the corresponding enclosures in 1990. Samples were normally taken
between 9 and 11 a.m. However, thefirstsampling was carried out between 11a.m.
and 1 p.m., starting one hour after the manipulations. Before sampling, the water
in each enclosure was mixed thoroughly. Six litres of water were sampled from
each enclosure with a 1.5 m tube sampler on each sampling occasion. Three litres of
the water were filtered through a 40 jim net for analyses of metazoan zooplankton,
the remainder being used for the other analyses. Ciliates and heliozoans are
treated as a single group in this paper. Potential bacterial grazers were grouped as
heterotrophic flagellates, mixotrophic flagellates, ciliates + heliozoans, rotifers
and cladocerans. Potential mixotrophic flagellates were grouped according to
Pringsheim (1963).
1012
Regulation of bacterioplankton production and biomass
Biological and chemical analyses
Biomass of bacterioplankton was determined by counting in an epifluorescence
microscope (1250x magnification) after staining with acridine orange (Bell et al.,
1983). The samples were preserved with formaldehyde (final concentration 4%)
immediately after sampling. To convert bacterial cell volumes to carbon content a
conversion factor of 0.28 pg C u,nr3 was used (Simon and Azam, 1989). The cell
volumes of 64 samples from 1990 (different treatments and sampling occasions)
were determined. One hundred cells from each sample were measured in the
microscope, giving an average cell volume of 0.06 ± 0.005 (95% CL) jim3, which
was used for all samples. Phytoplankton, heterotrophic and mixotrophic flagellates, ciliates, heliozoans and metazoan zooplankton were preserved with acidified
Lugol's solution immediately after sampling and later analysed in an inverted
microscope after sedimentation. Phytoplankton, heterotrophic and mixotrophic
flagellates, ciliates and heliozoans were counted in 100 and 400x magnification.
Biovolumes were estimated using geometrical formulas and converted to biomass,
assuming a density of 1 g cm-3. For common phytoplankton species at least ten cells
were measured and, for rare species, one 'typical' cell was measured. For common
ciliates and heliozoans species at least twenty were measured and for rare species,
at least one. Metazoan zooplankton were counted in 63 and lOOx magnification.
Biomasses of crustacean zooplankton were determined using data on body lengthdry weight relationships from other Swedish lakes (H.Olofsson, unpublished
data). All animals found, up to a number of 100, were measured. Biomasses of
rotifers were calculated using conversion factors from other Swedish lakes
(H.Olofsson, unpublished data). Biomasses were converted to carbon using the
following conversion factors: cyanophytes = 22% of wet weight (Ahlgren, 1983);
chlorophytes = 16% of wet weight; other phytoplankton = 11% of wet weight
(Lundgren, 1978 and references therein); ciliates and heliozoans = 19% of wet
weight (Putt and Stoecker, 1989); crustacean zooplankton = 48% of dry weight
(Andersen and Hessen, 1991); rotifers as for crustacean zooplankton.
Bacterioplankton production was determined by measuring incorporation of
[3H]thymidine into macromolecules according to Bell et al. (1993). Triplicate or
duplicate 10-15 ml water samples and one killed blank (2% formaldehyde final
concentration) were incubated in glass scintillation vials for 1 h in situ with 25-30
nM [3H]thymidine (40-60 Ci mmol'; Amersham). The incubation was stopped by
adding formaldehyde to afinalconcentration of 2%. The samples were incubated
immediately after sampling. Later macromolecules were extracted with ice-cold
trichloroacetic acid (TCA) (5% final concentration) and the samples were filtered
onto 0.45 )im pore-size cellulose acetate membranes. The samples were washed
five times with ice-cold 5% TCA and five times with ice-cold 80% ethanol. The
samples were assayed with an LKB model 1217 Rackbeta liquid scintillation counter. To convert moles of thymidine incorporated into DNA to cells produced, a
conversion factor of 2 x 1018 cells produced per mole thymidine incorporated was
used (Bell, 1990).
Losses of bacterioplankton biomass were calculated from bacterial biomass and
production. Change in bacterial biomass per day and bacterial production per day
1013
K-Vrede
were calculated between all sampling occasions for each treatment. The daily
median change in bacterial biomass and bacterial production was calculated from
these values. The change in bacterial biomass per day was subtracted from the
value of bacterial production per day in each treatment. This gave a value of daily
median loss of bacterial biomass. By multiplying this value by 21 (the number of
days that the experiment was run), the total loss of bacterial biomass could be
estimated. The gross bacterial biomass for the whole period of 21 days was calculated from accumulated bacterial production added to bacterial biomass on the
starting day. It was assumed that bacterial production was constant throughout the
day, which is a simplification. However, it gives an idea of the size of the losses of
bacterial biomass.
Methods for analyses of total phosphorus (TP), soluble reactive phosphorus
(SRP), total nitrogen (TN), nitrate (NO,), ammonium (NH4), alkalinity and pH
are described in Blomqvist et al. (1995).
Statistical analyses
The effects of the manipulations on bacterial biomass, bacterial production and
losses of bacterial biomass were analysed with multiple linear regression analyses
at the 5% significance level. The effects of potential bacterial grazers were also
tested. Variables with P > 0.05 were excluded from the regression models. The
regressions were carried out on average values of each treatment. Data from all
samplings except the starting day were used to calculated the average value. There
were no major differences, irrespective of whether total phosphorus or total nitrogen was used in the models, as their concentrations remained substantially constant throughout the experiment. Therefore only results from the models with
total phosphorus are presented. Concentrations of SRP and nitrate were not used
in the models as they were affected by uptake during the experiments. The differences between starting values within and between years were tested with Student's
/-test.
Results
Concentrations of inorganic nutrients and zooplankton
The lake and the experiments. The manipulations with NP and Z succeeded well.
The nutrient manipulations resulted in even gradients of both TP and TN (Figure
2), and the concentrations did not change appreciably during the experiment. The
natural concentrations of TP, TN, SRP, NO3 and NH< in the lake (Table I) were
similar to the lowest concentrations in the enclosures. In both years, the additions
of zooplankton resulted in concentrations similar to the lacustrine concentrations.
The enclosures that had received C had concentrations of inorganic nutrients similar to those in the corresponding enclosures in the gradients of NP, both years. The
results are therefore presented together below.
In 1989, the added phosphate was taken up the first day of the experiment and
was never detectable (<1 u-g P M). The NO3 concentrations declined, but were
measurable in all enclosures that had received NO, during the 21 days the experiments were run (Figure 3). The NH4 concentrations were low in most enclosures
1014
Regulation of baderioplanklon production and biomass
start day
£30
zooplankton absent
•
zooplankton present
r~| zooplankton absent
organic carbon added
611 27
Lake
3 812 4 9
510
[611 2 7 3 812 4 9
Enclosure
510
1 611 27
1 611 27
Lake
3812 49
S10
3812 49
510
Enclosure
? tOt.h p h T h ° r a ", d KOtaI n i t r O g e n i n t h e e n c l o s u r e s during the experiments and in
j p a t e t on the starting days of the experiments. The numbers refer to the enclosures in Figure 1
Error bars show standard deviations (N = 8).
from the beginning to the end of the experiments. Enclosures without zooplankton
had starting values of 0.02 ± 0.012 (average ± standard deviation) pig zooplankton
C 1-' and enclosures to which zooplankton were added, 48 ± 16.5 ^g C I1. The zooplankton biomass in the lake when the experiment started was 44 u.g C I"1. The
added zooplankton community was dominated by calanoid copepods. Eudiaptomus gracilis and Heterocope appendiculata constituted 80% of total zooplankton
biomass and the remaining 20% of the cladoceran, Diaphanosoma brachyurum.
Polyarthra vulgaris dominated in enclosures not receiving zooplankton. Average
values of zooplankton biomass in the nutrient gradient experiments are presented
in Figure 4. The starting values of bacterial production in the enclosures
that had received zooplankton were significantly higher than in the ones without
(paired r-test P = 0.023) in 1989. Paniculate organic matter could have been added
together with the zooplankton addition and immediately stimulated bacterial
activity, but this seems unlikely considering the coarse mesh-size of the net that
was used when collecting the zooplankton. There was no difference on a bacterial
biomass basis on the starting day between the enclosures that had received zooplankton and those that had not.
In 1990 the concentrations of SRP were measurable at times but usually near the
detection Umit (Figure 5). The highest concentrations were found on the starting
day and on day four in the two enclosures that had received the greatest addition
of nutrients. By day ten the SRP concentrations in these enclosures were as low
as in the other enclosures. The NO3 concentrations were measurable early in
1015
K.Vrede
Table I. Water chemistry in Lake Njupfatet in 1989 (before liming) and 1990 (after liming). Average
values and range are from composite samples and August values from epilimnion when the experiments started
Parameter
1989
Average
Tot-P(jigP|-')
SRP((igPl-')
Tot-N(u.gNI-')
NO, (pig N I-1)
NH 4 (u.gN|-')
pH
Alkalinity
(mequiv. I"1)
9
1
242
11
18
5.6
0 03
1990
Range
6-13
0-3
171-301
3-34
0-50
53-6.0
0.01-0.05
August
Average
Range
August
8
<1
222
<10
5
5.7
7
1
260
25
19
7.2
4-9
0-2
206-349
10-78
6-46
6.6-7.6
6
1
251
12
9
7.5
0.02
0.28
0.19-0.34
0.31
the experiments in all enclosures that had received NO3, but undetectable
(<10 ng N I"1) or near detection limit towards the end of the experiments in all
enclosures. As in 1989, NH4 concentrations were low in all enclosures throughout
the course of the experiments. Enclosures without zooplankton had starting values
of 0.25 ± 0.053 (xg zooplankton C I*1 and enclosures to which zooplankton were
added, 47 ± 6.6 n-g C I"1. The zooplankton biomass in the lake at the beginning of
the experiment was 67 u>g C I"1. The calanoid copepods, E.gracilis and H.appendiculata, made up 60% of the added zooplankton biomass. Two rotifers, Conochilus
unicornis and C.hippocrepis constituted 34% and a cyclopoid copepod, Cyclops
scutifers, 6%. In enclosures with no zooplankton addition, rotifers, and especially
P.vulgaris dominated. The relatively high biomass in the enclosures with high
nutrient concentrations and no zooplankton added was due to P.vulgaris (Figure
4). There was no difference in bacterial production or biomass on the starting day
between enclosures that had received zooplankton and those that had not.
Bacterial biomass and production
The lake. The bacterioplankton biomass in the epilimnion varied during the season
in the lake and was highest during the winter (Figure 6). When the phytoplankton
biomass increased in the spring, the bacterioplankton biomass decreased. In
August, when the experiments were performed, the bacterioplankton biomass was
lower in 1989 than in 1990.
The gradient of inorganic nutrients in 1989. Bacterial biomass and bacterial production increased with increasing nutrient concentrations, in both years (Figure 7).
In 1989, there was a positive response of bacterial production to TP in combination
with a positive response to total zooplankton biomass (TZ) (Table II). Bacterial
biomass was positively correlated to TP alone, but there was no correlation to TZ
either alone or in combination with TP. The increase in bacterial biomass with
increasing TP levelled off at the highest nutrient concentrations (Figure 7). In
1989, bacterial production was significantly negatively correlated to mixotrophic
flagellates and total ciliates + heliozoans biomass in combination with a positive
response to TP. This correlation was probably not due to a direct effect. Instead it
1016
6 —*— 11
7 — ° — 12
--••- 8
9
10
zooplankton absent
Fig. 3. Nitrate and ammonium concentrations in enclosures in Lake Njupfatet in 1989.
3
Enclosures:
zooplankton present
zooplankton absent
organic carbon udded
a
I
Io
•u
s
•a
a
2,
a
f
lCVrede
zooplankton present
zooplankton absent
•
D
•
•
•
1989
Rotiftra
CUdocera
Coprpoda
Ciliophora + HrliozokU
Heterotrophic flagellates
17
27
1
Total phosphorus (jig P T )
Fig. 4. Average metazoan zooplankton biomass and the relative distribution of Rotifera, Cladocera and
Copepoda, and the biomass of Ciliophora + Heliozoida and heterotrophic flagellates versus average
total phosphorus in the nutrient gradient in 1989 and 1990 in Lake Njupfatet (N = 7, starting values
excluded).
can be a result of zooplankton grazing. The biomasses of mixotrophic flagellates
and ciliates + heliozoans were lower in enclosures where larger zooplankton were
present (Figures 4 and 8). This gave a negative correlation between bacterial production, and mixotrophic flagellates and ciliates + heliozoans respectively, as
bacterial production responded positively to TZ. Also in 1989, neither bacterial
biomass nor production was correlated to the total biomass of phytoplankton,
heterotrophicflagellates,rotifers or cladocerans. Nor was bacterial biomass correlated to the biomass of mixotrophicflagellatesor ciliates + heliozoans either alone
or in combination with TP. The losses of bacterioplankton biomass were not correlated to TP, TZ, or any of the potential grazers.
The gradient of inorganic nutrients in 1990. In 1990 there was a significant positive
relationship between TP concentration and bacterial biomass (Table II). The positive correlation between TP concentration and bacterial production was weaker
than in 1989, but still significant. In 1990, the increase in bacterial production
levelled off at the highest TP concentrations (Figure 7). Bacterial biomass and
production were not correlated to TZ in 1990. The losses of bacterioplankton
biomass were positively correlated to TP in combination with rotifer biomass
(Table II), but they were not correlated to TZ.
Bacterial biomass, production and losses of bacterioplankton biomass were all
positively correlated to total phytoplankton, mixotrophicflagellatesand ciliates +
heliozoans alone. However, there were probably autocorrelations as the parameters were correlated with TP concentrations (Figures 4 and 8), just as were
1018
zooplankton present
zooplankton absent
10
20
zooplankton absent
organic carbon udded
Fig. 5. Nitrate, ammonium and soluble reactive phosphorus (SRP) concentrations in enclosures in Lake Njupfatet in 1990. Symbols as in Figure 3.
,-, 300
1990
a.
o;
5"
g
a
Io
•o
o
s
i
I
o'
ao
ICVrede
Experiments 1989
100 - r
_•—
80
-•- Baclerioplanklon
— Total phytoplankton
•
Cyanophyceae
-"
Experiments 1990
60
E
40
--
o
5
20
N D J F M A M J
1989
J A S O N
DJ
FMAMJ
JA
S
1990
Fig. 6. Bacterioplankton and total phytoplankton biomass in epilimnion in Lake Njupfatet, November
1988 to September 1990. The starts of the experiments are indicated. Each dot represents one sampling
occasion.
bacterial biomass, production and losses of bacterioplankton biomass. Neither
heterotrophic flagellates, rotifer or cladoceran zooplankton biomass were significantly correlated to bacterial biomass or production in 1990, nor was loss of bacterioplankton biomass to heterotrophic flagellates or cladoceran zooplankton
biomass.
Additions of organic carbon in 1989 and 1990. In 1989, bacterial production did not
respond to additions of C alone (Figure 9). However, when C and inorganic nutrients were added together, there was a steep continuous increase in bacterial production. In 1990 when the lake was limed, C when added alone stimulated bacterial
production slightly, but the strong response in bacterial production occurred when
NP was added alone or in combination with C.
Comparisons between the years. There is a striking difference between the years
regarding the responses of bacterial production and biomass to the treatments.
The same addition of NP, alone or in combination with C, resulted in greater maximum bacterial production and biomass in 1990 than in 1989 (Figures 9 and 10). The
same pattern is seen with average bacterial biomass and production in the NP
gradient. Average values of both bacterial biomass and production were much
higher in 1990 than in 1989 (Figure 7). However, the initial values of bacterial
biomass and production in the enclosures were also higher in 1990 (27 ± 5 p.g C I"1
and 0.033 ± 0.014 u,g C 1-' h"1 in 1989 versus 32 ± 6 u.g C I"1 and 0.065 ±0.028 p-g
C 1-' fr1 in 1990). The differences between the years were statistically significant
(paired r-test, bacterial biomass: P - 0.0104, bacterial production: P = 0.0394).
When the averages of the ten starting values of bacterial biomass and production in
the NP gradients are compared with the maximum value for the entire experiment,
bacterial biomass increased at most 2.5-fold in both 1989 and 1990 and bacterial
production at most 8.5-fold in 1989 and 9.5-fold in 1990.
1020
Regulation of bacf erioplanklon production and biomass
70
o
IS
50
40-
p
u
pa
3020y = 24 + 0.70TP
100
10
20
30
1
Total phosphorus (jig P I" )
B
3
1-
1990
0.4-
y = 0.033 + 0.014TP
03-
Jc "-1
I ai
1989
y = - 0.011 + 0.005TP + 0.001TZ
ao
10
20
30
Total phosphorus (jig P 1 )
Fig. 7. Results from manipulations of inorganic nutrients and zooplankton in enclosures in Lake Njupfatet in 1989 and 1990. The experiments were run for 21 days. (A) Regressions of average bacterioplankton biomass on average total phosphorus in enclosures in the gradient of inorganic nutrients
(N = 7, starting day values excluded). (B) Regressions of average bacterioplankton production on average total phosphorus in enclosures in the gradient of inorganic nutrients (JV = 7, starting day values
excluded). The regression line for 1989 is not drawn, as the regression is dependent on both TP and TZ.
(•) 1989 zooplankton present, (•) 1989 zooplankton removed, (a) 1990 zooplankton present, (o) 1990
zooplankton removed.
Grazers
The increase in bacterial production did not cause a corresponding increase in
bacterial biomass. The losses of bacterial biomass were greater in 1990 than in 1989
(Figure 11). During the 21 days the experiments were run, 3-47 jtg C I"1 of produced bacterial carbon were not recovered as new bacterial biomass in the NP gradient in 1989. This corresponds to 57 ± 38% of the bacterial production. In 1990 a
larger proportion of the bacterial carbon produced was lost (18-172 u.g C I"1)- The
bacterial production was also greater in this experiment, where losses equalled
production (93 ± 16% of bacterial production). In enclosures enriched with glucose the losses were 15-50 \ig C I'1 in 1989, and 32-174 jtg C I 1 in 1990, i.e. similar
to losses in enclosures without added glucose.
1021
K.Vrede
Table II. Summary of regressions from the enclosures with the gradient of inorganic nutrients in Lake
Njupfatet in 1989 and 1990
r2
P-values
Regression Intercept
TP
TZ
0.0365
1989 Models
BB = 24.091 +0.698TP
BP = -O.011 + 0.005TP
+ 0.001 TZ
0.571
0.951
0.0115
<0.0001
00004
ns
0.0115
<0.0001
1990 Models
BB = 32.014 + 0.742TP
BP = 0.033+ 0.014TP
B B L = - 0 . 0 1 8 + 0.008TP
+ 0.012 ROT
0.748
0.642
0.801
0.0012
0.0054
0.0035
<0.0001
ns
ns
0.0012
0.0054
0.0223
ROT
0.0218
Only significant regressions are shown BB = bacterial biomass, BP = bacterial production.
BBL = bacterial biomass losses, TP = total phosphorus, TZ = total zooplankton biomass, ROT = rotifer biomass, ns = not significant. All the biomass and production values are expressed in carbon units.
All potential grazers increased in biomass in 1990 vis-a-vis 1989. Potential mixotrophicflagellateshad the largest biomass among the grazers in all enclosures, the
second largest group being ciliates + heliozoans both years (Figures 4 and 8). However, ciliates + heliozoans is a very rough classification, as it includes all ciliates and
heliozoans that were found, and therefore several species feed on other organisms
than bacteria. The largest groups in the bags were Urotricha sp. 1 (~15 \i.m) +
Balanion sp. (could not be separated), Stichotricha sp., and heliozoans in 1989; and
Urotricha sp. 1 (—15 |xm) + Balanion sp. and Strombidium sp. in 1990. Also mixotrophicflagellatesis a mixed group as it includes all potential mixotrophic flagellates. On biomass basis, larger Ochromonadaceae spp. (5-10 u.m), Dinobryon
sociale var. americanum, Gymnodinium cf. lacustre, and Peridinium inconspicuum
were the most important groups among potential mixotrophic flagellates in 1989
and Ochromonadaceae spp. (5-10 u.m), Cryptomonas cf. marssonii, Gymnodinium cf. lacustre, and Rhodomonas lacustre in 1990. These species and groups constituted at least 10% each of the total mixotrophic biomass. It is noteworthy that
heterotrophicflagellates,which have bacterioplankton as their main food source,
in several bags increased their biomass by a factor of two in 1990, compared with
1989 (Figure 4). Heterotrophicflagellateswere Monosigales sp. and Polytoma granuliferum in 1989, and Katablepharis ovalis was the only species in 1990. The most
important bacterial grazer among rotifers, on biomass basis, was Polyarthra vulgaris in 1989. In 1990 P. vulgaris and Kellicottia longispina dominated in enclosures
not receiving zooplankton, and Conochilus unicomus and C.hippocrepis in bags
receiving zooplankton. Diaphanosoma brachyurum was the only cladoceran.
Phytoplankton biomass
In August 1989, the phytoplankton community was dominated by the cyanobacterium Merismopedia tenuissima. The cyanobacterium constituted 81% of the
1022
Regulation of bacterioptankton production and biomass
zooplankton absent
zooplankton present
0 ('>anoph\it-jf
I
H Mixolrnphk- flagrllati
H < >lhiT ph> tttplanktun
i
24
Total phosphorus (ug P I"')
Fig. 8. Average phytoplankton biomass, and the relative importance of potential mixotrophic flagellates and cyanobacteria. versus average total phosphorus in the nutrient gradient in Lake Njupfatet in
1989 and 1990 (/V = 7, starting values excluded).
phytoplankton biomass (on carbon basis) in the lake when the experiments were
initiated. Chrysophytes, dinoflagellates and green algae constituted 8,7 and 4% of
total phytoplankton biomass, respectively. In 1990 M.tenuissima had completely
disappeared from the lake and was not replaced by any other phytoplankton
species, which resulted in a lower phytoplankton biomass in 1990 than in 1989
(Figure 6). At the onset of the experiment in 1990, chrysophytes and green algae
dominated and constituted 45 and 37% of biomass, respectively. The biomass of
Dinophyceae was 15% of total phytoplankton biomass, Cryptophyceae was 5%
and no cyanophytes were present at all. The most important species on biomass
basis (over all enclosures and time) were Ochromonadaceae spp. (5-10 u.m), Cryptomonas cf. marssonii, Gymnodinium cf. lacustre, and Mallomonas cf. tonsurata
(>10% each of total phytoplankton biomass) in 1990. The total phytoplankton
biomass failed to respond significantly to increasing nutrient concentrations in
1989, as the dominant M.tenuissima responded weakly to the NP manipulations
(Figure 8). Merismopedia tenuissima constituted 71% of total phytoplankton biomass in the experiments and all other phytoplankton had biomasses <10% of total
(over all enclosures and time). In 1990 there was a significant positive response of
phytoplankton biomass to increasing concentrations of NP (r2 = 0.96, P < 0.0005)
(Figure 6).
1023
s
o
•
0
20
40 •
60 •
80
imi
0
20 -
40 -
60 •
1990
—A—
--•80 •
•
o—
100
Day
10
10
no uJJilinns (6)
NP(S)
C(ll)
C + NPU2)
\
V *
X
20
20
1989
0,4
zooplankton absent
Day
10
20
Fig. 9. Bacterial biomass and production in the experiments with organic carbon additions in Lake Njupfatet in 1989 and 1990.
Error bars show standard deviations, and are sometimes too small to be seen behind the symbols. Numbers in brackets refer to
enclosure numbers in Figure 1. Additions: N = nitrogen, P = phosphorus, C = carbon. Note the different scales on the y-axes.
Si
u
nO
C
rial
i
Regulation of bacteriopbrakton production and biomass
zooplankton present
zooplankton absent
00
1989
80
—*—
NoaJdiuonsih,
- - • -
Iflpy P|-I(X|
1989
— A — ItygPI -'|9i
60 40
20
0
•
10
20
20
icterial pi-oduct ion
0.3
aa
0.0
1.0
0.8
0.6
0.4
0,2
0.0
20
10
20
Dav
Fig. 10. Bacterial biomass and production in the nutrient gradient in Lake Njupfatet in 1989 and 1990.
Error bars show standard deviations, and are sometimes too small to be seen behind the symbols. Both
nitrogen and phosphorus were added to the enclosures, however only the phosphorus additions are
shown in the legends. Numbers in brackets refer to enclosure numbers in Figure 1. Note the different
scales on the y-axes.
1025
K-Vrede
1990
1989
" T
10 T zooplankton present
zooplankton present
8-
• Bacterial production
• Losses of bacterial biomass
W-r
12
20
10-|
zooplankton absent
">, 8 - _ organic carbon added
29
14
17
26
zooplankton absent
organic carbon added
•? 6 5
4-
3 20-
wmzzLM&A.
9
16
Total phosphorus (jig P litre'1)
15
Total phosphorus (ug P litre 1 )
Fig. 11. Average bacterial production and rate of biomass losses versus average total phosphorus (N =
8) in enclosure in Lake Njupfatet in 1989 and 1990.
Discussion
The main experimental design of this study was a gradient of inorganic nutrients.
Although individual treatments were not replicated, it is evident from Figures 2,3
and 5 that the treatments with and without zooplankton for a given nutrient
addition are similar, and that the nutrient additions gave gradients as intended.
Biological parameters (Figures 4,8 and 10) were also consistent within the experiments, which strengthen the results and the choice of this experimental design.
One problem with a gradient set-up, which is sampled several times, is that consecutive measurements from one enclosure are not independent. Therefore, I have
chosen to use the average values (for the whole period) for the regressions. However, in this way some information is lost, since the temporal dynamics are
excluded from the analyses.
There was a marked difference in bacterioplankton response to the treatments
in 1989 vis-a-vis 1990. In both years, NP additions had a significant positive effect
on bacterioplankton production. However, the increase in bacterial production in
1026
Regulation of bacterioplankton production and biomass
1989 was very small compared with the strong response to the NP addition in 1990.
Bacterioplankton growth in Lake Njupfatet seems to have been co-limited by inorganic nutrients and C in 1989, as the increase in bacterial production in 1989 was
strong only when NP and C were added together. In a clearwater lake, organic
carbon released from phytoplankton could be of great importance for bacterioplankton, as other sources, such as C of allochthonous origin, are small. It could
also be expected that NP additions would stimulate phytoplankton growth and
thereby increase the release of extracellular organic compounds from phytoplankton, as the release increases with increasing primary production (Baines and Pace,
1991). However, the phytoplankton biomass increased only slightly with
increasing NP concentrations in 1989 and the release of extracellular organic compounds from phytoplankton could therefore have been small. If bacterioplankton
were dependent on organic C released from phytoplankton, the small increase in
phytoplankton biomass could have resulted in carbon limitation of bacterioplankton and thus explain the weak response in bacterial production when NP was added alone.
In 1990, NP stimulated bacterioplankton production when added alone. There
are various pathways by which this stimulation could have occurred,
(i) Bacterioplankton were limited by inorganic nutrients only, and the stimulation
of production was directly attributable to increased NP concentrations,
(ii) Phytoplankton increased their growth when NP was added and thereby
increased their release of extracellular organic compounds. Bacterioplankton
were limited by organic carbon only and the stimulation of growth was indirect and
mediated by phytoplankton.
(iii) The stimulation of bacterioplankton production was both direct and indirect
(mediated by phytoplankton) as bacteria were co-limited by NP and C.
Bacterial production and phytoplankton biomass were correlated in 1990 and both
bacterioplankton production and phytoplankton biomass showed a strong positive response to the manipulations of NP. However, the slight increase in bacterial
production when C was added alone indicates that there was no strong limitation
of bacterial growth by C alone in 1990. This in turn suggests that the increase in
bacterial production when NP was added alone must have been a direct response,
alone or in combination with stimulation by phytoplankton (due to increased
release of extracellular organic compounds). It is difficult to separate a direct effect
of the manipulations on bacterioplankton from an indirect effect mediated by phytoplankton, in experiments like this where the entire plankton community is used.
The enclosures were sampled every third day, which may have been too
infrequently to distinguish the responses.
The major difference in phytoplankton species composition between the two
years may explain the different responses to the manipulations. The cyanobacterium M.tenuissima which completely dominated in August 1989, disappeared in
1990 due to liming of the lake (Blomqvist et al., 1993,1995). The weak response by
phytoplankton to the nutrient additions in 1989 was due to the lack of response by
this dominant species. This low response could have resulted in a scarcity of readily
utilizable carbon for bacteria. In 1990 the phytoplankton biomass was smaller on
1027
K.Vrede
the starting day than on the starting day in 1989, as no other phytoplankton species
had replaced the cyanobacterium in the lake (Figure 6 and Blomqvist et al., 1995).
Compared with the starting values, nutrient additions increased the average phytoplankton biomass up to 5-fold more in 1990 than in 1989. The increased phytoplankton growth may have resulted in increased extracellular release of organic
compounds that could be utilized by bacteria.
As pH and alkalinity increased in 1990 compared to 1989, total available inorganic carbon most likely also increased. Thus, there was more inorganic carbon
and inorganic nutrients available for phytoplankton in 1990 than in 1989, and it
may be assumed that the low response of phytoplankton in 1989 could have been
due to limitation by inorganic carbon. However, in fertilized Lake Mountaintop,
with a pH of 4.0-4.5, chlorophyll a values exceeded 100 mg nr3, and maximum
phytoplankton biomass was 14 mg ww 1-' (Yan and Lafrance, 1984). Recalculated
to carbon units (using a low carbon conversion factor of 0.11% of wet weight, see
Method) this corresponds to a phytoplankton biomass of 1540 u.g C I"1, which is
much higher than the maximum biomass (453 (xg CH) in Lake Njupfatet. The
primary production could have been limited by inorganic carbon at times in Lake
Mountaintop, but it did not prevent the development of a large standing stock of
phytoplankton. Therefore, it is unreasonable to believe that the difference in response to nutrients by phytoplankton between the years was due to limitation by
inorganic carbon.
Merismopedia tenuissima seems to be the clue to the different responses to the
manipulations in the two years, and this is supported by the development of bacterioplankton biomass in the lake. There was a significant increase in bacterioplankton biomass in 1990 compared to 1989 during the open water periods
(Blomqvist et al., 1995). The bacterioplankton biomass decreased in the spring
when the phytoplankton biomass increased and was smallest in late summer, both
in 1989 and in 1990. However, the biomass decreased more in August 1989 when
M.tenuissima dominated the phytoplankton community, than in 1990, which indicates that the cyanobacterium had a negative impact on the heterotrophic bacteria
in the lake.
As the bacterioplankton response to combined NP and C additions was very
small in 1989 compared with 1990, it is possible that M.tenuissima had other negative effects on bacterioplankton. This cyanobacterium species may have been able
to successfully compete for phosphorus with both bacterioplankton and other phytoplankton species in 1989 and store phosphorus in the cells, thereby suppressing
their growth. As M.tenuissima cannot use nitrate as a nitrogen source (Blomqvist
et al., 1989) it did not respond to the manipulations, even though it had sufficient
phosphorus. However, it has been shown that other phytoplankton species in the
lake (as well as the cyanobacterium) had phosphorus stored in their cells (Blomqvist et al., 1993). It is therefore conceivable that bacterioplankton are able to take
up P, as their P uptake is more efficient than that of phytoplankton (Currie and
Kalff, 1984a). Another explanation for the weak response by bacterioplankton in
1989 is that M.tenuissima may have some antibiotic effect on bacterioplankton.
The weak response to the additions of NP in combination with C in 1989 could also
have been due to the lower pH in 1989, or to limitation by one or several other
1028
Regulation of bacterioplankton production and biomass
nutrients. However, experiments carried out in June 1989 before the cyanobacterium dominated the phytoplankton community showed that at least pH is not the
reason for the lower response in August 1989 (K.Vrede, unpublished data), as
bacterioplankton responded vigorously to the nutrient manipulations in the June
experiments.
Autotrophic picoplankton can account for a large fraction of the carbon fixation
in oligotrophic lakes (Stockner and Porter, 1988). However, in Lake Njupfatet
autotrophic picoplankton are most likely of little importance, since very few cells
were recorded in the quantification of phytoplankton. Those which were found
were either M.tenuissima (Cyanophyceae) or Dictyosphaerium cf. botrytella
(Chlorophyceae), and were cells (1-2 u.m) that had been broken off from colonies.
In contrast, Brettum (1989), in his survey of 165 Norwegian lakes using the same
technique, found 'u^algae' (—1-2 u>m diameter) to be quantitatively important in
oligotrophic lakes with pH 5-7. Furthermore, fertilization of oligotrophic Lake
Hymenjaure resulted in mass development of the green alga Choricystis coccoides
(0.8-1.5 x 0.5-1.0 (xm), which was also quantified using the same method (Holmgren, 1984). Thus, if autotrophic picoplankton were of importance in Lake Njupfatet, these organisms would have to be even smaller than those recorded by
Brettum and HolmgTen, as they otherwise would have been recorded during the
phytoplankton counts.
Zooplankton had a significant positive effect on bacterioplankton production in
1989. By excretion, defaecation and sloppy feeding, zooplankton can provide bacterioplankton with inorganic nutrients and C (Lampert, 1978; Jumars et ai, 1989)
and thereby stimulate bacterial production. If bacterioplankton were co-limited
by organic and inorganic nutrients in 1989, it can be assumed that zooplankton
increased the bacterial production by generating inorganic and/or organic nutrients. This is remarkable in a way, as the analysis of phytoplankton data from the
enclosures shows that the predominant colonial M.tenuissima was not grazed by
zooplankton (Blomqvist, 1996). Since there was no growth response by the cyanobacteria and since it was not grazed, the release of organic compounds through
sloppy feeding ought to have been lower in 1989, than in 1990 when 'edible' phytoplankton increased their biomass (Blomqvist etal., 1993). However, even though
the amount of organic compounds released due to zooplankton activity was less in
1989 than in 1990, it might have been more important to the bacteria when the
overall availability of organic compounds utilizable by bacteria was lower.
The bacterial biomass losses were lower in 1989 than in 1990, both on carbon
basis and as percentage of bacterial production. The estimations of losses of bacterial biomass and total bacterial production have weaknesses, as it was assumed
that bacterial production was constant during the day, and the calculations are
based on one value every third day. In spite of this, the result, that the losses of
bacterial biomass can be of the same order of magnitude as production, is consistent with other studies (Pace etal., 1990; Chrzanowski and Simek, 1993). It is difficult to find a good correlation between potential bacterial grazers and bacterial
biomass, which may be expected with the rather rough classification of grazers. In a
study like this, more specific methods for estimating grazing must often be used to
gain an insight into the fate of bacterioplankton biomass. However, it is inconceiv1029
K.Vrede
able that mixotrophic flagellates (the dominant bacterial grazer on a biomass
basis) were exclusively heterotrophic in 1990 when they had good access to inorganic nutrients, at least at the beginning of the experiments.
To summarize, this study in Lake Njupfatet indicates that the presence of a
cyanobacterium may profoundly influence bacterioplankton growth. Also, the
results show that bacterioplankton can be limited by both NP and C. Bacterioplankton production was closely co-limited by NP and C in 1989, whereas in 1990,
bacterioplankton production was limited primarily by NP alone, or by NP and C.
The strength of the response to the manipulations of NP and C differed between
1989 and 1990 and this co-occurred with a change in the phytoplankton community. When a cyanobacterium {M.tenuissima) dominated the phytoplankton
community, heterotrophic bacterioplankton responded only weakly to the
manipulations and the largest increase in bacterioplankton production was when
NP and C were added together. In 1990, the cyanobacterium was absent, and bacterioplankton exhibited a steeper increase in production, when NP was added both
alone and in combination with C. This increase may have been due to a direct
stimulation of the bacterioplankton production, or partly indirectly through phytoplankton. Phytoplankton biomass responded weakly to the nutrient manipulations in 1989 when the cyanobacterium dominated the phytoplankton
community, and the release of extracellular organic compounds may therefore
have been small. In 1990 the increase in phytoplankton biomass was pronounced
and the bacterioplankton response could have been mediated through phytoplankton production of bacterial substrates. Both the stronger responses in bacterial production to the manipulation of nutrients and the larger losses of bacterial
biomass in 1990 indicate that the microbial food web was more efficient when the
cyanobacterium was absent. Seasonal dynamics in bacterial limitation has been
demonstrated earlier (Morris and Lewis, 1992; Wang et al., 1992). However, to my
knowledge, this is the first study where limitation of bacterioplankton has been
correlated to changes in phytoplankton species composition.
Acknowledgements
I am indebted to the late Russell T.Bell (deceased in 1993) for supervision and
ideas. I thank Peter Blomqvist, Hans Olofsson and Ulrika Stensdotter-Blomberg
for allowing me to use their data on phytoplankton, metazoan zooplankton and
protozoan zooplankton, respectively. I acknowledge Raul Figueroa for field assistance and for running the chemistry analyses, Kjell Hellstrom for counting bacteria,
Jan and Ulla Domarhed for allowing me and my co-workers to use their barn as a
field station and Peter Blomqvist, Lars Tranvik, Tobias Vrede and two anonymous
referees for comments on the manuscript. This work was supported by a grant
from the Swedish Environmental Protection Board (SNV) to P.Blomqvist and
grants from the Swedish Natural Science Research Council (NFR) and SNV to
R.T.Bell.
References
Ahlgren.G. (1983) Comparison of methods for estimation of phytoplankton carbon. Arch Hydrobiol.,
98,489-508.
1030
Regulation or bactertoplankton production and biomass
Andersen.T. and Hessen.D.O. (1991) Carbon, nitrogen, and phosphorus content of freshwater zooplankton. LimnoL Oceanogr., 36, 807-814.
Arndt.H. (1993) Rotifers as predators on components of the microbial web (bacteria, heterotrophic
flagellates, ciliates)—a review. Hydrobiologia, 225/226.231-346.
Baines.S.B. and Pace.M.L. (1991) The production of dissolved organic matter by phytoplankton and its
importance to bacteria: Patterns across marine and freshwater systems. Limnol. Oceanogr., 36.
1078-1090.
Bell.R.T. (1990) An explanation for the variability in the conversion factor deriving bacterial cell production from incorporation of [-'Hjthymidine. Limnol. Oceanogr.. 35.910-915.
Bell.R.T.. Ahlgren.G.M. and Ahlgren.I. (1983) Estimating bacterioplankton production by measuring
PHJthymidine incorporation in a eutrophic Swedish lake. Appl. Environ. Microbiol., 45.1709-1721.
Bell,R.T., Vrede.K., Stensdotter.U. and Blomqvist.P. (1993) Stimulation of the microbial food web in
an oligotrophic, slightly acidified lake. Limnol. Oceanogr., 38,1532-1538.
Bird.D.F. and KalffJ. (1984) Empirical relationships between bacterial abundance and chlorophyll
concentration in fresh and marine waters. Can. J. Fish. Aqual. Sci., 41, 1015-1023.
Blomqvist.P. (1996) In-situ experiments on late summer phytoplankton responses to combinations of
altered nutrient availability and grazing pressure in oligotrophic Lake Njupfatet. central Sweden.
Arch. Hydrobiol. (in press).
Blomqvist.P., Bell.R.T., Olofsson.H., Stensdotter.U. and Vrede.K. (1993) Pelagic ecosystem responses
to nutrient additions in acidified and limed lakes in Sweden. Ambio, 22.283-289.
Blomqvist.P., Bell.R.T., Olofsson.H., Stensdotter.U. and Vrede.K. (1995) Plankton and water chemistry in Lake Njupfatet before and after liming. Can. J. Fish. Aqual. Sci.. 52,551-565.
Blomqvist.P., Olsson.H., Olofsson.H. and Broberg.O. (1989) Enclosure experiments with low-dose
additions of phosphorus and nitrogen in the acidified Lake Njupfatet, central Sweden. Int. Rev. Ces.
Hydrobiol.,14,6U-63\.
Brettum.P. (1989) Alger som indikalor p& vannkvalitet i norske innsjper. Planteplankton. Norwegian
Institute for Water Research (Niva). Oslo. In Norwegian.
Chrzanowski.T.H. and Simek.K. (1993) Bacterial growth and losses due to bacterivory in a mesotrophic lake J Plankton Res., 15, 771-785.
ColeJJ., Findlay.S. and Pace.M.L. (1988) Bacterial production in fresh and saltwater ecosystems: a
cross-system overview. Mar. EcoL Prog. Ser., 43, 1-10.
Currie.D.J and KalffJ. (1984a) A comparison of the abilities of freshwater algae and bacteria to
acquire and retain phosphorus. LimnoL Oceanogr., 29,298-310.
Currie,D.J. and KalffJ. (1984b) The relative importance of bacterioplankton and phytoplankton in
phosphorus uptake in freshwater. Limnol. Oceanogr., 29, 311-321
GoldmanJ.C, Caron.D.A. and Dennett.M.R. (1987) Regulation of gross growth efficiency and
ammonium regeneration in bacteria by substrate C:N ratio. Limnol. Oceanogr., 32,1239-1252.
Gflde.H. (1989) The role of grazing on bacteria in plankton succession. In Sommer.U. (ed.), Plankton
Ecology: Succession in Plankton Communities. Springer-Verlag, pp. 337-364.
HellebustJ.A. (1965) Excretion of some organic compounds by marine phytoplankton. Limnol.
Oceanogr., 10,192-206.
Holmgren,S. (1984) Experimental lake fertilization in the Kuokkel area, northern Sweden; Phytoplankton biomass and its algal composition in natural and fertilized subarctic lakes. Int. Rev. Ges.
Hydrobiol.,69,m-8\1.
Jansson.M. (1993) Uptake, exchange, and excretion of orthophosphate in phosphate-starved Scenedesmus quadricauda and Pseudomonas K7. LimnoL Oceanogr., 38,1162-1178.
Jumars.P.A., Penry.D.L., BarossJ.A., Perry.MJ. and Frost,B. (1989) Closing the microbial loop: dissolved carbon pathway to heterotrophic bacteria from incomplete ingestion, digestion and absorption in animals. Deep-Sea Res., 36,483-495.
Lampert,W. (1978) Release of dissolved organic carbon by grazing zooplankton. Limnol. Oceanogr.,
23,831-834.
LeJ., WehrJ.D. and Campbell.L. (1994) Uncoupling of bacterioplankton and phytoplankton production in freshwater is affected by inorganic nutrient limitation. AppL Environ. Microbiol., 60,
2086-2093.
Lundgren.A. (1978) Experimental lake fertilization in the Kuokkel area, northern Sweden: Changes in
sestonic carbon and the role of phytoplankton. Verh. Internal. Ver. Limnol., 20,863-868.
Morris JD.P. and Lewis.W.M., Jr (1992) Nutrient limitation of bacterioplankton growth in Lake Dillon,
Colorado. LimnoL Oceanogr., 37,1179-1192.
M0nster,U. and Chrost.RJ. (1990) Origin, composition and microbial utilization of dissolved organic
matter. In Overbeck J. and Chrost, R J. (eds), Aquatic Microbial Ecology: Biochemical and Molecular Approaches. Springer, pp. 8-46.
1031
K-Vrede
Nagata,T. and Kirchman.D.L. (1992) Release of dissolved organic matter by heterotrophic protozoa:
implications for microbial food webs. Arch. Hydrobiol., 35,99-109.
Pace.M.L., McManus.G.B. and Findlay.S.E.G (1990) Planktonic community structure determines the
fate of bacterial production in a temperate lake. Limnol. Oceanogr., 35,795-808.
Pringsheim,E.G. (1963) Farblose Algen Ein Beiirag zur Evolutionsforschung. Gustav Fisher Verlag.
Putt.M. and Stoecker.D.K. (1989) An experimentally determined carbon:volume ratio for marine 'oligotrichous' ciliates from estuarine and coastal waters. LimnoL Oceanogr., 34, 1097-1103.
Sanders.R.W.. Porter.K.G., Bennett.SJ. and DeBiascA.E. (1989) Seasonal patterns of bacterivory
by flagellates, ciliates, rotifers, and cladocerans in a freshwater planktonic community. LimnoL
Oceanogr.. 34,673-687.
Simon.M. and Azam.F. (1989) Protein content and protein synthesis rates of planktonic marine bacteria. Mar. Ecol. Prog. Ser., 51, 201-213.
StocknerJ.G. and Porter.K.G. (1988) Microbial food webs in freshwater planktonic ecosystems. In
Carpenter.S.R. (ed.). Complex Interactions in Lake Communities. Springer, pp. 69-83.
Sundh.I. (1992) Biochemical composition of dissolved organic carbon released from natural communities of lake phytoplankton. Arch. Hydrobiol., 125, 347-369.
Sundh.I. and Bell.R.T. (1992) Extracellular dissolved organic carbon released from phytoplankton as a
source for heterotrophic bacteria in lakes of different humic content. Hydrobiologia, 229,93-106.
Tezuka.Y. (1990) Bacterial regeneration of ammonium and phosphate as affected by the carbon:nitrogemphosphorus ratio of organic substrates. Microb. Ecol., 19, 227-238.
Tranvik.LJ. and Hflfle.M.G. (1987) Bacterial growth in mixed cultures on dissolved organic carbon
from humic and clear waters. Appl. Environ. Microbioi, 53,482-488.
Vadstein.O. and Jensen.A. (1988) Growth and phosphorus status of limnetic phytoplankton and bacteria. Limnol. Oceanogr., 33,489-503.
Vadstein.O. and Olsen.Y. (1989) Chemical composition and phosphate uptake of limnetic bacterial
communities cultured in chemostats under phosphorus limitation. Limnol. Oceanogr., 34,939-946.
Wang,L., Miller.T.D. and PriscuJ.C. (1992) Bacterioplankton nutrient deficiency in a eutrophic lake.
Arch. Hydrobiol., 125,423-439.
Wheeler.P.A. and Kirchman.D.L. (1986) Utilization of inorganic and organic nitrogen by bacteria in
marine systems. LimnoL Oceanogr., 31, 989-1009.
Yan.N.D. and Lafrance.C. (1984) Responses of acidic and neutralized lakes near Sudbury, Ontario, to
nutrient enrichments. In NriaguJ. (ed.). Environmental Impact of Smelters, pp. 458-521.
Zweifel.U.L., Norrman3- and HagstrOm^A. (1993) Consumption of dissolved organic carbon by marine bacteria and demand for inorganic nutrients. Mar. Ecol Prog. Ser., 101, 23-32.
Received on April 20, 1995; accepted on February 2, 1996
1032