A New General Approach to Quantify Nitrogen Fixation Exemplified for

A New General Approach to Quantify Nitrogen Fixation Exemplified for
36
The Open Marine Biology Journal, 2009, 3, 36-48
Open Access
A New General Approach to Quantify Nitrogen Fixation Exemplified for
the Baltic Proper
Lars Håkanson*, Julia K. Hytteborn and Andreas C. Bryhn
Department of Earth Sciences, Uppsala University, Villav. 16, SE-752 36 Uppsala, Sweden
Abstract: This work uses empirical data from the HELCOM database and a new empirically-based model to predict the
concentration of cyanobacteria in the Baltic Proper. The aim has been to estimate nitrogen fixation. The inherent variabilities/patchiness in the variables regulating nitrogen fixation are great. This means that different approaches may provide
complementary information so that several relatively uncertain estimates may together provide less uncertainty in the estimate for nitrogen fixation in a given system. We show that there is marked variability in nitrogen fixation among different years (a factor of 20 between the year 2001 with the smallest value and 2005 with the highest value of about 900 kt/yr
of N-fixation). The mean value for the period from 1997 to 2005 was 190 kt/yr. TN/TP based on median monthly data has
been higher than the Redfield ratio of 7.2 since 1994. 6.5% of all individual data (n = 3001) from the surface-water layer
(44 m) in the Baltic Proper for samples with temperatures higher than 15°C (when risks of getting cyanobacteria blooms
are favoured) have TN/TP lower than 7.2. The mean TN/TP is 20 for surface-water sites with temperatures higher than
15°C, indicating that the average trophic conditions in the Baltic Proper are likely more limited by phosphorus than nitrogen. Nitrogen fixation is an important contributor to the nitrogen concentration and we give overall budgets for nitrogen
and phosphorus in the Baltic Proper, including nutrient data from land uplift, which is the most important contributor for
nutrients and often neglected in discussions about sources of nutrients to the Baltic Sea.
Keywords: Nitrogen fixation, cyanobacteria, phosphorus, nitrogen, salinity, temperature, marine systems, empirical model,
Baltic Proper.
INTRODUCTION
Quantifying nitrogen fixation is essential for estimating
the nitrogen balance in marine waters and understanding the
causes of marine eutrophication. Since nitrogen fixation is
triggered by high temperatures, low non-gaseous nitrogen
concentrations and high phosphorus concentrations in the
water, this process may determine which nutrient regulates
primary production in the long run, and thus which nutrient
should be abated in order to remediate marine eutrophication
[1, 2].
Fig. (1) illustrates the main processes regulating external
fluxes (atmospheric input and river inflow) of nutrients (nitrogen and phosphorus) to a given aquatic system, internal
fluxes (sedimentation, resuspension, diffusion, denitrification and burial) including the very important relationship
between the amount of the nutrient in dissolved (bioavailable) and particulate form (the only part that can settle out
due to gravity) forms. The focus of this work is on nitrogen
fixation.
Many papers discuss nitrogen fixation in the Baltic Sea
[3-7]. Table 1 gives two empirical models for lakes as a
background for this work. The first is the well-known
OECD-model [8] predicting mean summer chlorophyll-a
concentrations from total phosphorus concentrations (TP).
This model yielded a coefficient of determination (r2) of 0.77
*Address correspondence to this author at the Department of Earth Sciences,
Uppsala University, Villav. 16, SE-752 36 Uppsala, Sweden;
E-mail: [email protected]
1874-4508/09
when tested for lakes from a wide trophic level domain
(characteristic TP-concenterations from 2.5 g/l, i.e.,
oligotrophic, to 100 g/l, i.e., hypertrophic). The second
empirical model in Table 1 concerns the main topic of this
work, a predictive model for cyanobacteria based on readily
accessible x-variables, such as the TP-concentration in this
model (from [9]), which gave an r2 of 0.71 when tested for
29 systems covering a very wide limnological domain (TP
from 8 to 1300 g/l). Smith’s empirical model for cyanobacteria applies to lakes.
The aim of this paper is to use a new empirically-based
model from [10] to predict cyanobacteria and nitrogen fixation in the Baltic Proper. This model will be outlined in the
next section.
Table 2 provides background for this work. This table
summarizes transport processes to, within and from the Baltic Proper. The value (bolded) for nitrogen fixation is 130
000 tons per year. This value indicates an order-ofmagnitude figure for fixation in the Baltic Proper. The value
is, however, very uncertain. References [11] and [12] gave
values of 100 kt/yr, [13] gave 130 kt/yr for the Baltic Proper
and the Gulf of Finland, [14] estimated values from 18.2 to
186 kt/yr for the Baltic Proper and the Mecklenburg Bay and
[4] noted a ranged of 30 to 260 kt/yr for the period 1992 to
1997. Yearly estimates of nitrogen fixation from the Baltic
Proper are given in Table 3 (according to [5]). These values
were calculated as increases in TN minus atmospheric deposition plus particulate N sedimentation at the Landsort Deep,
and extrapolated to other basins according to local increases
in TN. The summer increase in TN strongly coincided with
blooms of N-fixing cyanobacteria in 1997 and 1998 and
2009 Bentham Open
Quantify Nitrogen Fixation Exemplified for the Baltic Proper
Nitrogen
fixation by many
but not all
cyanobacteria
The Open Marine Biology Journal, 2009, Volume 3
Atmospheric wet and
dry deposition
(mainly N)
Biouptake and
growth
??
Phytoplankton
N & P;
16:1 (atoms) or
7.2:1 (g);
turnover time, 2-4
days
37
River and open
water input to
coastal area
??
Dissolved
N&P
PF ??
Decomposition
by bacterioplankton
Particulate
N&P
Sedimentation
Internal loading
(resuspension and diffusion
of P)
Sediment
N&P
??
Denitrification (N)
Burial
Land uplift
Fig. (1). Overview of important transport processes and mechanisms related to the concept of “limiting” nutrient. PF is the particulate fraction.
these values are about 60% of those calculated by [6]. [7]
estimated the N-fixation rate in the Baltic Proper to about a
factor of 2 times the values in Table 3 (434 - 792 kt/yr),
while earlier estimates have generally been much lower and
have not included night measurements, the complete growing season, or the important contribution from phytoplankton
smaller than 10m.
Hence, all approaches to quantify the annual fluxes of
nutrients in Fig. (1) are more or less uncertain. To quantify
nitrogen fixation should be one of the most difficult and uncertain because the cyanobacteria show such very high coefficients of variation and this will be discussed in a following
section.
Table 1.
Regressions Illustrating the Key Role of Total Phosphorus (TP) in Predicting Chlorophyll-a (Chl,
Summer Values) and the Biomass of Cyanobacteria
(CB). r2 = Coefficient of Determination, n = Number
of Lakes Used in the Regression
Eq.
Range for TP
r2
n
Units
Reference
Chl
0.28·TP0.96
2.5-100
0.77
77
μg/l
[8]
CB
43·TP0.98
8-1300
0.71
29
μg/l
[9]
It is important to identify all major sources of nutrient
inputs not just to the Baltic Proper but to all polluted water
systems and to quantify all major fluxes because this will
determine the expectation one would have on different, often
costly, measures to reduce nutrient emissions and improve
ecosystem conditions. If the total inflow of nitrogen to the
Baltic Proper from countries/processes is 1 800 000 tons per
year (Table 2), and if Swedish anthropogenic emissions are
just 2.6% of all the nitrogen transport to the Baltic Proper,
this will determine the expectations that one would have on
such reductions, and also provide a possibility to compare
the cost-effectiveness of alternative approaches to reduce
nitrogen fluxes to the system.
The average composition of algae is given by the Redfield ratio 7.2, if the calculation is done in g (see [15] and
[16]) and 16:1, if the calculation is based on the number of
atoms.
Another critical uncertainty in Fig. (1) concerns the
equilibria between nutrients in dissolved and particulate
phases, the time scales of these interactions, and what is actually meant by “limiting” nutrient. At short time scales
(seconds to days), it is evident that the causal agent regulating/limiting biouptake and primary production is the concentration of the nutrient in bioavailable forms [17]. This also
implies that the bioavailable forms are quickly regenerated
so their supply is poorly described by their concentrations at
any given sampling site [18]. This also explains why there
are no practically useful predictive models for chlorophyll
based on dissolved forms (DIN, DIP, phosphate, nitrate,
etc.), but rather on TN or TP (see Table 1 or [19, 20]). At
longer time scales (weeks to years), and for all practical purposes in water management, one must recognize the difference between what is causally the “limiting” agent in primary production at short time scales, and what is the form of
the nutrient limiting predictions of chlorophyll or cyanobacteria at the ecosystem scale for longer periods of time. This
paper uses an empirical model for total cyanobacteria, which
concerns the latter aspects, the ecosystem scale (the entire
Baltic Proper) using monthly data.
The problem to understand and predict TN-concentrations
in marine systems is accentuated by the fact that there are no
(to the best of our knowledge) practically useful models to
quantify the particulate N-fraction in saltwater systems (but
such approaches are available for phosphorus in lakes and
brackish systems, see [21]. In mass-balance modelling, it is
imperative to have a reliable algorithm for particulate nitrogen, since the particulate fraction (PF) is the only fraction that
by definition can settle out due to gravity.
38 The Open Marine Biology Journal, 2009, Volume 3
Table 2.
Håkanson et al.
Background Budget on the Transport of Nitrogen and Phosphorus to and from the Baltic Proper (t/yr); from [48]. The
Data from the Swedish Environmental Protection Agency (SNV [49]) Concern Mean Values for the period 1982 and 1989;
Data from HELCOM [43] is for 2000
Total-N
SNV
HELCOM
Sweden
44 300
Baltic states
Total-P
H&L*
SNV
HELCOM
46 636
1780
1219
72 600
145 697
1890
5408
Finland
-
35 981
-
1874
Russia
-
90 229
-
5863
Poland
109 900
191 521
19 100
12 698
Germany
20 000
20 602
2750
512
Denmark
51 000
27 664
7860
1193
Sum inflow from countries:
297 800
558 046
33 380
H&L*
A. From countries/rivers
500 000
Order-of-magnitude value:
28 767
30 000
B. From processes and water inflow from adjacent basins
Precipitation
289 900
192 400
Nitrogen fixation
130 000
-
3420
-
-
-
Land uplift
480 000
160 000
Inflow from Kattegat
120 000
14 000
Inflow from Bothnian Sea
340 000
14 000
Sum from processes:
1 261 000 – 1 359 000
191 420
Order-of-magnitude value:
1 300 000
190 000
Total inflow (A+B):
1 800 000
220 000
To the Bothnian Sea
340 000
24 000
To Kattegat
260 000
18 000
600 000
40 000
C. Water outflows to adjacent basins
Total outflow:
D. Other terms
Burial in sediments
Denitrification
(3·180 000)* = 540 000
(220 000 – 40 000) = 180 000
(1 800 000 - 600 000 – 540 000) = 660 000
* The nitrogen concentration is set to be 3 times higher than the phosphorus concentration in these sediments. H&L = [48]. The role of land uplift and the task of finding the most
cost-efficient remedial strategy for the Baltic Sea are discussed in more detail in [55].
From previous modelling work [22], one can conclude
that it is also very difficult to quantify denitrification (Fig.
1). Denitrification depends on sediment red-ox conditions,
i.e., on sedimentation of degradable organic matter and the
oxygen concentration in the deep-water zone, but also on the
frequency of resuspension events, on the presence of mucusbinding bacteria, and on zoobenthos and bioturbation. Given
this complexity, it is easy to understand why empirically
well-tested algorithms to quantify denitrification on a
monthly basis do not exist to the best of our knowledge. The
atmospheric wet and dry deposition of nitrogen may (as indicated in Fig. 1) be very large (in the same order as the
tributary inflow) and patchy [23], which means that for large
coastal areas and smaller systems distant from measurement
stations, the uncertainty in the magnitude of atmospheric wet
and dry N-deposition is also generally very large.
Table 3.
Yearly N-Fixation (kt/yr) in the Baltic Proper Based
on Data from [5]
Year
Minimum
Maximum
1994
290
350
1995
360
420
1996
180
250
1997
340
430
1998
180
330
Quantify Nitrogen Fixation Exemplified for the Baltic Proper
The Open Marine Biology Journal, 2009, Volume 3
TN/TP 29 (by weight) and are much less abundant at
higher ratios, while nitrogen-fixing cyanobacteria tend to
dominate at TN/TP 22 [28]. [9] found TP to be a better
predictor of the biomass of cyanobacteria (CB) than TN and
TN/TP.
The question about “limiting” nutrient is central in
aquatic ecology and has been treated in numerous papers
[19, 20, 23-26]. The salinity is of paramount importance to
the predictions of chlorophyll and cyanobacteria. The salinity influences the aggregation of suspended particles [21], of
particular interest in understanding variations in water clarity, which in turn regulate the depth of the photic zone and
hence primary production. The saltier the water, the greater
the flocculation of suspended particles. An increased salinity
together with N-limitation and increased light seem to control factors of N-fixing cyanobacteria in estuaries, according
to [27].
THE MODEL
This approach does not concern cyanobacteria produced
in the benthic zone. In the derivation of the model [10], data
from many databases were used, and Table 4 gives a compilation of information used as well as data used for the Baltic
Proper (from the HELCOM data base). Fig. (2A) gives the
(log-log) regression between chlorophyll-a concentrations
and total cyanobacteria (CB = median values for the growing
season) using lake data. One can note a highly significant
and mechanistically understandable strong positive co-
The abundance of cyanobacteria compared to other algal
groups is closely related to the TN/TP-ratio. Cyanobacteria
have been found to dominate lake primary production at
Table 4.
39
Compilation of Main References (A), Data and Statistics (B) for the General, Comprehensive Data Base. Note that All
Data Used in this Data Base Represent Median Surface Water Values for the Growing Season, and (C) Data from HELCOM (http://www.ices.dk/ocean/asp/helcom/helcom.asp?Mode=1) Used in this Work for Water Temperature, Salinity,
Total-N and Total-P from the Entire Baltic Proper Between 1990 and 2005 and Chlorophyll-a Data from 1974 to 2005
A. Area
Country
Salinity Range
Main References
Chesapeake Bay
U.S.A.
<2-35
http://www.chesapeakebay.net/data/wq_query1.cfm?db=CBP_WQDB
Italian coast
Italy
15-40
http://dataservice.eea.europa.eu/dataservice/metadetails.asp?id=836
Crimean lakes
Ukraine
25-275
[51]
Baltic Sea
Sweden
2.4-5.5
Magnus Karlsson, ÅF, pers. comm.
Bothnian Bay
Sweden
0.3-1.5
http://dataservice.eea.europa.eu/dataservice/metadetails.asp?id=836
West coast
Sweden, Norway
6.4-27
http://dataservice.eea.europa.eu/dataservice/metadetails.asp?id=836
Swedish lakes
Sweden
0
http://info1.ma.slu.se/db.html
Baltic coast
Sweden, Finland
5-7.5
[20, 52]
Lakes, shelf, marine
Several
0-36
[25, 50]
Ringkobing Fjord
Denmark
3.9-14.5
[53, 54]
B.
Total-N (g/l)
Total-P (g/l)
Chlorophyll (g/l)
Salinity psu
Mean (MV)
533
40.5
7.9
15.9
Median (M50)
461
32.1
5.2
12.5
Minimum
1
<1
0.12
0
Maximum
2110
1090
60.3
275
Standard deviation (SD)
320
61.4
9.27
14.4
Coeff. of variation among systems (CV)
0.60
1.52
1.17
0.91
Number of data (n)
658
582
709
714
C.
Temp (°C)
Salinity
Total-N (g/l)
Total-P (g/l)
Chl-a (g/l)
Min
-0.71
1.29
7.0
0.3
0.1
Max
24.02
24.89
2493
780.6
71.6
Mean (MV)
6.44
8.41
298.1
44.83
2.11
Median (M50)
4.96
7.40
291.4
25.71
1.70
Standard dev. (SD)
4.27
2.33
78.39
41.93
2.23
Coeff. of var. (CV)
0.66
0.28
0.26
0.94
1.06
Number of data (n)
43125
43502
41690
44152
14384
40 The Open Marine Biology Journal, 2009, Volume 3
Håkanson et al.
y = 0.83x + 0.35; r2 = 0.86; n = 76 lakes
>0]
y = 5.85x - 4.01; r2 = 0.76; n = 86 [if
[if CB
BG>0]
18
5
A.
16
B.
4
14
12
BG^0.25
CB^0.25
log(1+CB)
log(1+BG)
3
2
1
10
8
6
4
0
2
-1
-1
0
1
2
3
4
5
6
0
.5
1
1.5
log(Chl)
5
SWT=15
n=74
2
2.5
3
3.5
log(TP)
C.
log(1+BG)
log(1+CB)
4
3
2
BG=100
CB=100
1
0
8
10
12
14
16
18
20
22
SWT
Fig. (2). (A) Regression between log(1+CB) and log(Chl) based on data from 76 lakes. The figure also gives the coefficients of determination (r2) and the regression lines. Cyanobacteria (CB) in g ww/l (ww = wet weight), chlorophyll-a concentrations (Chl) in g/l. (B) Regression between CB (transformed into CB0.25; CB in g ww/l) and log(TP) (TP in g/l) using median values for the growing season from 86
systems with CB-values higher than zero. (C) Scatter plot between cyanobacteria (transformed into log(1+CB); CB in g ww/l) versus surface water temperatures (SWT in °C) based on data from 74 systems. From [10].
variation between these two measures of primary production
in lake water. This relationship might have looked different
had it been based on daily, weekly or monthly values and it
is probable that this co-variation also exists for many marine
systems. The following calculations first show maximum
values for nitrogen fixation.
In a following section, we will also discuss the fraction of
N-fixing species of the total biomass of cyanobacteria in the
Baltic Proper.
CB Versus TP
The basic regression is given in Fig. (2B) (r2 = 0.76; n =
86). This regression includes data from more systems than
the equation given by [9] and it also yields a higher r2-value
(0.76 as compared to 0.71).
CB Versus Temperature
In the literature, temperatures between 15 and 17 ºC have
been reported as the minimum for cyanobacteria blooms in
Quantify Nitrogen Fixation Exemplified for the Baltic Proper
The Open Marine Biology Journal, 2009, Volume 3
41
Model for cyanobacteria
CB in μg ww per l
TP
TN to TP
YTNTP
TN
Salinity
SWT
Ysal
YSWT
CB = ((5.85·log(TP)-4.01)4)·Y TNTP·YSWT ·Ysal
CB in μg ww/l
Salinity in psu
SWT= Surface water temperature in °C
Total-N (TN) in μg/l
Total-P (TP) in μg/l
YTNTP = if TN/TP < 15 then (1-3·(TN/TP/15-1)) else 1
YSWT = if SWT 15 then (0.86+0.63·((SWT/15)^1.5-1)) else (1+1·((SWT/15)^3-1))
Ysal = if salinity <10 then (2.1+1.1·((salinity/10)^2-1)) else (2.1-115·((salinity/10)^0.01-1))
Model domain: 4 < TP < 1300; 165 < TN < 6830; 0 < salinity < 40; 8 < SWT < 25
Fig. (3). Outline of the model to predict median summer values of cyanobacteria in lakes and coastal areas (from [10]).
freshwater systems and in the Baltic Sea [7, 29, 30]. Laboratory experiments on cyanobacteria also support this conclusion [31-33]. There are also reports that cyanobacteria have a
requirement of temperatures of about 20-21 ºC to form
blooms. Those reports are from a freshwater lake in Canada
[34], the North Pacific Ocean [35] and an estuary in Australia [36]. The optimal growth temperatures in laboratory experiments are usually around 25 ºC for many species [31, 33,
37], but these experiments often use species from temperate
areas. With higher temperatures, growth rate usually starts to
decrease. In field data from the Baltic Sea [3], this decrease
in growth rate is not shown because there are few occasions
with temperatures higher than 20 ºC.
Fig. (2C) gives data on the relationship between CB
(log(1+CB)) and surface water temperatures (SWT in °C)
from 74 systems. One can note that all systems with CB exceeding 100 g/l (median values for the growing season)
have temperatures higher than 15 °C. Fig. (3) gives a compilation of the model. The dimensionless moderator for temperature influences on CB (YSWT) is given by:
If SWT 15 °C, then
YSWT = (15/17.5)·(1+0.63·((SWT/15)1.5-1)) else
YSWT = (1+1·((SWT/15)3-1))
(1)
This means that if SWT = 15 °C, YSWT = 0.86; if SWT =
17.5, YSWT = 1; if SWT = 25, YSWT = 1.48, etc. So, when the
temperature is 25 °C, the risks of getting high concentrations
of cyanobacteria a factor of 1.48 higher than at 17.5 °C, if all
else is constant.
CB Versus Salinity
In hypertrophic lakes, the biomass of cyanobacteria can
be very high with concentrations of about 100 mg/l [9]. [38,
39] found no data on N-fixing planktonic species in estuaries
and coastal seas, except for the Baltic Sea and the PeelHarvey estuary, Australia. Also results from [40] support this
general absence of N-fixing cyanobacteria in estuaries. There
are more than 10 nitrogen fixing cyanobacteria species in the
Baltic Proper [6]. The number of species and the nitrogen
fixation rates have been considerably revised upwards during
recent years [7]. A field study in the Baltic Sea [3] indicates
that in this brackish environment species of cyanobacteria
have, interestingly, the highest biomass at salinities of 7 – 8
psu and that the blooms in Kattegat and Belt Sea are more
frequent if the salinity is below 11.5 psu. Results from a
laboratory experiment with cyanobacteria from the Baltic
Sea support the highest growth rates at salinities between 5
and 10 psu [33]. According to [41], blooms of cyanobacteria
in marine environments may not be as common as in freshwater systems. In marine systems, there are just a few dominant genera. In a field study in the Pacific Ocean [35], there
was no correlation between the salinity and the abundance of
cyanobacteria and no cyanobacteria were found in the cooler,
less saline waters. In this model (Eq. 2), the influences of the
42 The Open Marine Biology Journal, 2009, Volume 3
Håkanson et al.
If the salinity < 10 psu, then
Ysal = (2.1+1.1·((salinity/10)2-1)), else
Ysal = (2.1-115·((salinity/10)0.01-1))
(2)
This means that at a salinity of 10 psu, Ysal is 2.1 and CB
a likely factors of 2.1 higher than in freshwater systems.
CB Versus TN/TP
Fig. (4A) shows a scatter plot between log(CB) and
TN/TP. High CB-values only appear in systems with relatively low TN/TP. The regression between log(CB) and
TN/TP attains a maximum value (r2 = 0.73;
log(CB)=0.142·TN/TP+5.47; n = 61) if only data from systems with TN/TP smaller than 40 are used. The upper curve
(circles) in Fig. (4B) gives the r2-values when only systems
with TN/TP smaller than 10, 15, 20,….. 100, respectively,
were used in the regressions. The lower curve in Fig. (4B)
gives similar results when only systems with higher TN/TP
were used. Note that there is no statistically significant
(p<0.01) relationship between log(CB) and TN/TP if TN/TP
is higher than 15.
lation of characteristic CV-values for many of the variables
discussed in this work for 58 lakes. One should note from
Table 5 that cyanobacteria should be expected to have significantly higher CVs than most other variables and are
therefore more difficult to predict with a high certainty. It is
also important to note that dissolved inorganic nutrient fractions have considerably higher CVs than TN and TP. Thus,
dissolved inorganic nutrient fractions are, in relation to TN
and TP, not only poorly correlated with chlorophyll and Secchi depth [44], but given their high inherent uncertainties, it
is much more costly to determine reliable mean/median values that can be used in ecosystem models, whose scientific
quality and usefulness is given by their predictive power.
y=-0.052·x+4.52
5
A.
4
3
log(CB)
salinity on cyanobacteria are motivated by empirical data
from many systems [10]:
1
The basic regression between CB and TP given by Eq. 1
is complemented with a dimensionless moderator related to
TN/TP, YTNTP, defined by:
0
If TN/TP < 15 then YTNTP = (1-3·((TN/TP)/15-1)) else Y TNTP
=1
(3)
-1
This means that for systems with TN/TP (based on median values for the growing season) higher than 15, one can
use the basic regression without any correction, but for systems with TN/TP < 15, Eq. 3 is used. If, e.g., TN/TP = 7.2,
then Y TNTP = 2.56, and the CB-value a factor of 2.56 higher
than the value suggested by the basic regression.
Measured N-fixation tends to follow a similar pattern as
the prevalence of cyanobacteria [6, 38, 39]. Analyses using
gene sequencing techniques have suggested that more organisms than we currently know may fix nitrogen in both lakes
and marine systems [42].
DATA AND DATA UNCERTAINTY
To estimate the N-fixation in the Baltic Proper, we have
used one main dataset, the HELCOM data compiled in Table
4C [43]. All calculations have used median monthly values
on TN, TP, salinity and temperature since these are the
obligatory driving variables in the model.
The uncertainty in the model variables may be expressed
by the coefficient of variation (CV = SD/MV; SD = standard
deviation; MV = the mean value) and Table 5 gives a compi-
Lake
Marine
-2
0
.8
.7
10
20
30
40
50
60
70
80
B.
90
100
r2, <
r2, >
.6
r2
The general model for cyanobacteria may give rather
uncertain predictions for systems with high TN/TP and low
temperatures. However, during such conditions, the Nfixation should be small. Predicting conditions with high CB
is evidently more important in calculations of N-fixation.
Due to the fact that all methods to estimate N-fixation in
entire systems at longer time scales (such as month or years)
are very uncertain, there is good reason to regard the approach presented here as complementary to other approaches.
2
.5
.4
.3
.2
.1
0
0
10
20
30
40
50
60
70
80
90
100
TN/TP
Fig. (4). (A) Scatter plot between log(CB) and TN/TP. The regression line is based on data from systems where TN/TP < 15. (B)
Compilation of r2-values between log(CB) and TN/TP using data
from systems when successively smaller (<) and higher (>) TN/TP
have been omitted. The maximum r2 is obtained for systems with
TN/TP < 40; very low r2-values are obtained if the regressions are
done for systems with TN/TP > 15 (from [10]).
The basic factors regulating CVs for chlorophyll and
cyanobacteria (i.e., nutrients, temperature, light, salinity,
predation, analytical uncertainties related to sampling)
should be similar in most aquatic systems. [45] showed that
for the River Danube, the CV for the cyanobacteria was, on
average, a factor of 1.3 higher than for chlorophyll.
This means that one should expect that the r2-value one
can hope to achieve would be lower for cyanobacteria than
Quantify Nitrogen Fixation Exemplified for the Baltic Proper
The Open Marine Biology Journal, 2009, Volume 3
43
for chlorophyll and this is also the information conveyed in
Table 1 (r2 = 0.71 and 0.77, respectively).
uncertainty in the calculated values is higher than the difference between these two years.
Table 5.
Fig. (5) illustrates the modelled monthly concentrations
of cyanobacteria from 1996 to 2005. One can see clear seasonal patterns and that the modelled values are very high for
2005, which also had a very high and much discussed algal
bloom.
Coefficients of Variation (CV) for Twelve Water
Variables Based on Data from 58 Swedish Lakes
(from [10]). The Values Describe Conditions During
the Growing Season and are Typically Based on 66
Measurements and in No Case Less than 6
Variable
Median
Min
Max
Cyanobacteria
1.76
0.64
3.91
Ammonium
0.74
0.42
3.91
Nitrate + nitrite
0.73
0.09
2.78
Total algal biomass
0.68
0.34
1.83
Phosphate
0.58
0.24
1.14
Chlorophyll
0.43
0.25
1.29
Total-P
0.37
0.23
0.78
Total-N
0.24
0.04
0.72
Secchi depth
0.21
0.11
0.51
Temperature
0.14
0.10
0.37
Calcium
0.11
0.03
0.83
pH
0.04
0.01
0.09
RESULTS
Calculating Maximum Nitrogen Fixation in the Baltic
Proper
Table 6 exemplifies the calculation routine using data for
2004. The table gives the median monthly values for surfacewater temperatures (i.e., for samples taken above the theoretical wave base of 44 m in the Baltic Proper; see [46]),
salinity, TN, TP and TN/TP, as well as the dimensionless
moderators for salinity (Ysal), surface-water temperature
(YSWT) and TN/TP (YTNTP). The calculated monthly concentrations of total cyanobacteria (CB in g/l) are given as well
as the calculated maximum potential N-fixation. The table
also shows the dimensional adjustments used in the calculation. From Table 6, one can note:
•
•
The maximum potential N-fixation (in 2004) in the
Baltic Proper was about 255 kt, with highest values in
July, August and September.
This corresponds very well to the overall values given
in Table 3.
The calculated lowest and highest values for the
maximum potential annual N-fixation in the Baltic
Proper for the period 1997 to 2005 are given in Table
7. One can note:
•
The variation among the years is very high: the
smallest value is 45 kt, the highest 908 kt.
•
The average annual value is 191 kt.
Both [5] and [6] measured higher rates in 1997 than in
1998, but our results show higher rates in 1998 (58 and 71
kt, respectively). However, the order-of-magnitude is about
the same in our measurements for these two years, and the
Using the model, one can also clarify the factors contributing to the calculated values. Fig. (6) illustrates predictions
if the four x-variables (i.e., the monthly median values of
temperature, salinity, TN and TP) were reduced by 25%. A
reduction in TN would increase the predicted biomass of
total cyanobacteria considerably, while reductions in salinity,
temperature and TP would lower the predicted values; the
clearest response would be from reductions in temperature.
According to our results, the high CB-values in 2005 may be
attributed to relatively high TP-concentrations and high temperatures in the summer and fall.
TN/TP and Cyanobacteria
If TN/TP is lower than 7.2, phytoplankton species that
can take up dissolved nitrogen of atmospheric origin will be
favoured. Empirical data in Fig. (4) show that there is a
threshold limit for the ratio not at 7.2 but rather at 15. Fig.
(7) gives a scatter plot of all available data on TN/TP (n =
24048) from the surface-water layer in the Baltic Proper
from 1990 to 2005; Fig. (8) shows variations in median
monthly TN/TP in relation to the Redfield ratio of 7.2 and
the threshold ratio of 15. From these two figures, one can
note that there are no major changes in the general temporal
trend. There is also a very large scatter in the data and clear
seasonal patterns.
Table 8 gives a compilation of TN/TP statistics for the
Baltic Proper, divided into four categories: (1) all data from
the surface-water layer, (2) all data for situations with temperatures higher than 15 °C, (3) all data with Temp > 15 °C
and from water depths < 20 m, and (4) all data with Temp >
15 °C, water depths < 20 m and TP > 10 g/l. This table also
gives information on the percentage of the data with TN/TP
smaller than 7.2 and smaller than 15. The main conclusion
from this table is that less than 7% of the values are smaller
than 7.2 and that between 30 and 50% of the TN/TP-ratios
are smaller than 15. Fig. (9) gives a frequency distribution
for the data at sites with temperatures higher than 15°C.
The key information in Figs. (7-9) and Table 8 is that the
conditions in the surface-water layer of the Baltic Proper
often favour cyanobacteria. However, in more than 70% of
the situations when the water temperature is higher than 15
°C and the risks of blooms of cyanobacteria are highest, the
system is not N-limited but P-limited (Fig. 9). To minimize
the overall primary phytoplankton production in the Baltic
Proper (eutrophication), one should mainly reduce the discharges of phosphorus to the systems. To achieve a goal of
less eutrophication in the Baltic Proper, N-reductions would
be more or less meaningless, which is well in line with results presented by [1]. In 30% of the situations, when TN/TP
is lower than 15, reductions in N would favour nitrogen fixing cyanobacteria, which should be avoided.
Fig. (10) gives a regression between median monthly
concentrations of chlorophyll-a and modelled concentrations
44 The Open Marine Biology Journal, 2009, Volume 3
Table 6.
Håkanson et al.
Calculation of Monthly and Annual Maximum Potential Nitrogen Fixation in the Baltic Proper Using Data from 2004
Month
Temp
Salinity
TN
TP
CB
N-Fix
2004
°C
psu
g/l
g/l
g ww/l
t/Month
1
3.8
7.1
276
22.9
12.0
1.56
0.02
1.59
10
2363
2
2.4
7.2
266
25.7
10.4
1.57
0.00
1.93
4
937
3
2.0
7.1
275
29.4
9.3
1.55
0.00
2.13
4
870
4
3.5
6.9
261
25.4
10.3
1.52
0.01
1.95
12
2917
5
6.2
7.1
254
21.1
12.0
1.55
0.07
1.59
34
8171
6
10.0
7.1
259
18.9
13.7
1.55
0.29
1.26
82
19748
7
12.5
8
16.5
7.1
265
20.0
13.3
1.55
0.58
1.35
202
48451
7.0
270
19.5
13.9
1.54
0.96
1.23
284
68315
9
13.4
7.0
269
19.2
14.0
1.54
0.71
1.20
197
47328
10
9.7
7.0
265
20.8
12.8
1.53
0.27
1.44
110
26359
11
8.5
6.9
261
19.8
13.1
1.53
0.18
1.37
62
14899
12
6.0
7.1
265
25.7
10.3
1.56
0.06
1.94
61
14619
TN/TP
Ysal
YSWT
YTNTP
Max. annual N-fixation (t)
254977
CB·8.7·1012·10-9·0.2·(0.42/4.19)·0.147·(1/3.2)·30.
CB·V (m3)·(g to t)·(ww to dw)·(dw to C)·(C to N)·(1/d)·(d to month).
of total cyanobacteria. The co-variation is significant
(p = 0.0008), but the scatter around the regression line is
considerable (r2 = 0.11). Two reasons for this low r2-value is
that the range in the monthly chlorophyll data is relatively
small in the Baltic Proper (from 0.2 to 4 g/l), and that the
CV for the empirical chlorophyll data is high.
We have also investigated if it would be possible to test
these model predictions of cyanobacteria against empirical
data on cyanobacteria. The basic idea was to use information
accessible from SMHI (see SMHI website), but that turned
out to be very difficult for several reasons. The satellite images only give qualitative (not quantitative) information related to non specified algal blooms in the uppermost water
layer at defined hours and the results depend very much on
the cloudiness. Our model predictions concern the inventory
of total cyanobacteria in the entire surface-water layer (0 to
44 m) in the entire Baltic Proper on a monthly basis, and to
transform the information conveyed by the satellite images
has been beyond the possibilities of this work.
Table 7.
Compilation of Calculated Annual Maximum Potential Nitrogen Fixation in the Baltic Proper Between
1997 and 2005. Note that the Results from 2005 are
Based on Interpolated Data for the Missing Months
Year
N-Fix (t/yr, max.)
1997
58303
1998
70960
1999
60443
2000
116916
2001
45181
2002
131898
2003
73642
2004
254977
2005
908087
The Fraction of N-Fixing Cyanobacteria
Min
45181
All calculations so far are based on total concentrations
of cyanobacteria (CBtot) using the model in Fig. (3). The
fraction of N-fixing cyanobacteria (CBfix) is not identical to,
but lower than, CBtot. We have estimated the ratio CBfix/CBtot
using data from the Baltic Proper. The data comes from two
monitoring stations, one outside the Askö research station
(B1) and one at the Landsort deep (BY31). We have accessed the data at, http://www2.ecology.su.se/dbbm/JS.html.
Max
908087
Mean
191156
Median
73642
Cyanobacteria from the order Nostocales can fix nitrogen
in heterocysts. Nitrogen fixation is rather energy consuming,
with each heterocyst requires 12 to 20 photosynthetic cells to
provide the necessary energy supply [47].
SD
276448
CV
1.45
n
9
Recent studies indicate that other primary producers
(non-heterocystous cyanobacteria) smaller than 10 μm may
contribute considerably to nitrogen of the total N-fixation in
Quantify Nitrogen Fixation Exemplified for the Baltic Proper
The Open Marine Biology Journal, 2009, Volume 3
the Baltic Proper may be fixation [6]; measurements indicated that up to 43% done by these organisms, while not
contributing much to total biomass.
Table 8.
45
values since N-fixation from “small organisms” is not included in this ratio.
25% reduction
in:
2000
Compilation of Statistical Information TN/TP in the
Surface-Water Layer of the Baltic Proper Using
Data from 1990 to 2005
1800
Surface-water layer
Baltic Proper
TN
Min.
Temp
>15 °C
Depth
<20 m
TP
>10 g/l
1.05
1.74
1.74
1.74
Max.
171
76.9
76.9
59.4
Mean
16.5
20.1
20.4
18.7
Median
15.2
19.5
19.8
18.8
SD
8.1
9.2
9.2
7.4
CV
0.49
0.46
0.45
0.40
n
24048
3001
2880
2577
% < 7.2
7.0
6.5
6.0
6.6
% < 15
49.1
28.8
27.5
30.7
Cyanobacteria (μg ww/l)
1600
All
1400
Actual
1200
Sal
1000
800
600
TP
400
200
CB (μg ww/l)
1000
Surface-water layer
Baltic Proper
0
1996
40
60
Month (1 = Jan. 1996)
80
100
120
2005
Fig. (6). Simulations to illustrate how 25% reductions in TN, TP,
water temperature and salinity (Sal) would likely influence the concentrations of cyanobacteria in the surface-water layer of the Baltic
Proper using data from 1996 to 2005. The curve marked “Actual” is
based on empirical data from the given period.
100
10
[10] calculated CBfix/CBtot using lake data from [9], deriving a value of 0.33, which is considerably lower than
these results for the Baltic Proper.
1
.1
Temp
20
20
40
60
Month (1 = Jan. 1996)
80
100
120
2005
Fig. (5). Modelled concentrations of total cyanobacteria (g ww/l;
ww = wet weight) in the surface-water layer of the Baltic Proper
1996 to 2005.
The mean and median values when the water temperature
is higher than 15°C (the threshold temperature for cyanobacteria) are between 0.84 and 0.97, respectively. The actual Nfixation may, however, be close to the modelled maximum
15
TN/TP
Because heterocyst-forming species of cyanobacteria do
not always fix nitrogen and because small non-heterocystous
cyanobacteria may contribute to N-fixation, it is complicated
to find a reliable value of CBfix/CBtot. Further complicating
matters include the high inherent CV for cyanobacteria and
the fact that reliable data for the entire surface-water layer
for longer periods of time from the Baltic Proper (or from
most systems) are scarce. In this estimation of CBfix/CBtot,
only the heterocyst-forming species have been considered.
An average ratio of 0.85 was calculated from all 580 values
(see Table 9).
7.2
Month (1 = jan. 1990)
Fig. (7). TN/TP from all (n=24048) surface-water samples in the
Baltic Proper from 1990 to 2005.
46 The Open Marine Biology Journal, 2009, Volume 3
Håkanson et al.
30
species contribute, on average, less than 1% of the biomass
of cyanobacteria.
25
Min: 1.74
Max: 76.9
MV: 20.2
M50: 19.5
SD: 9.2
CV: 0.46
n: 3001
%<7.2: 6.5
%<15: 28.8
Threshold, 15
15
10
Redfield, 7.2
5
0
0
20
40
60
80
100
120
140
160
Number of data
TN/TP
20
180
Month (1 = Jan. 1990; 193 = Jan. 2006)
Fig. (8). Variations in median monthly TN/TP in the surface-water
layer of the Baltic Proper from 1990 to 2005 in relation to the Redfield ratio of 7.2 and the threshold ratio of 15 (see text).
The mean compositions of cyanobacteria species or genus at the two monitoring stations in the Baltic Proper in
situations with water temperatures higher than 15°C are
given in Fig. (11). This diagram is included here so that
these results from the Baltic Proper may be translated to
other systems where different cyanobacteria with other
CBfix/CBtot-ratio may prevail. The two nitrogen fixing species Aphanizomenon sp. and Nodularia spumigena are dominant, with 80% of the biomass. Then follow Cyandictyon
spp., Pseudanabaena limnetica, Anabaena spp. and Anabaena lemmermannii with a few percent each. The remaining
Table 9.
TN/TP
Fig. (9). Frequency distribution for TN/TP for all samples with
temperatures higher than 15°C during 1990 to 2005 in the surfacewater layer of the Baltic Proper. 6.5% of the data have TN/TP <
7.2; 28.8 % of the data ratios < 15.
CONCLUDING COMMENTS
The empirically-based approach to predict total nitrogen
fixation used in this work (from [10]) is general and could be
used for other aquatic systems in a wide salinity range. It is
The Ratio Between the Volume of Nitrogen Fixation Cyanobacteria to the Total Volume of Cyanobacteria from Two
Monitoring Stations in the Baltic Proper, Askö (Data from 1983 to 2005) and the Landsort Deep (Data from 1990 to
2005). MV = Mean Value for All Data; n = Number of Data. Temperature Data (T) from Askö from 1992 to 2004 and
Landsort Deep from 1990 to 2004 at 0 to 20 m Depth
Askö
Month
Landsort Deep
T(ºC)
n
CBfix/CBtot
T(ºC)
n
CBfix/CBtot
1
1.6
10
0.50
3.0
11
0.79
2
0.8
10
0.68
1.9
8
0.83
3
1.2
11
0.93
1.7
23
0.88
4
2.7
16
0.87
2.7
38
0.90
5
6.5
21
0.94
5.7
28
0.93
6
10.9
38
0.96
9.7
32
0.97
7
13.6
53
0.96
13.5
34
0.87
8
15.3
52
0.87
15.5
34
0.82
9
12.8
37
0.79
13.5
25
0.85
10
9.3
29
0.73
10.3
21
0.69
11
6.4
13
0.91
7.2
14
0.51
12
4.0
10
0.69
5.4
11
0.74
Median for T>15ºC
17.2
30
0.97
16.8
34
0.91
Mean for T>15ºC
17.2
30
0.92
16.6
34
0.84
Mean for all data
7.6
301
0.86
7.8
279
0.84
Quantify Nitrogen Fixation Exemplified for the Baltic Proper
The Open Marine Biology Journal, 2009, Volume 3
meant to give seasonal or monthly data for the surface-water
layer. When tested using empirical data for the Baltic Proper,
there was reasonably good overall correspondence between
the estimates using this model and other approaches. For the
future, it would be important to have better data on the fraction of nitrogen fixing cyanobacteria not just in the Baltic
Proper but more generally, so that a general algorithm could
be derived for this important ecosystem characteristic.
[2]
y = 0.47x + 1.54; r 2 = 0.11; n = 99; p = 0.0008
[6]
[3]
[4]
[5]
4
[7]
3.5
log(CB) [modeled]
[8]
3
[9]
[10]
2.5
[11]
2
1.5
[12]
1
[13]
[14]
.5
[15]
0
-1
1
[16]
Fig. (10). The relationship between empirical median monthly concentrations of chlorophyll-a (g/l) and modelled concentrations of
cyanobacteria (g ww/l) using data from 1997 to 2005 for the surface-water layer of the Baltic Proper.
[17]
-.8
-.6
-.4
-.2
0
.2
.4
.6
.8
log(Chl) [empirical]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
Fig. (11). Average composition of Cyanobacteria at water temperatures higher than 15 ºC at Askö (station B1) using data from 30 days
sampled between 1992 to 2004 and at the Landsort deep, (station
BY31) using data from 34 days collected between 1990 to 2004.
[26]
[27]
[28]
REFERENCES
[1]
Savchuk O, Wulff F. Modelling regional and large-scale response
of Baltic Sea ecosystems to nutrient load reductions. Hydrobiologia
1999; 393: 35-43.
[29]
47
Tyrrell T. The relative influences of nitrogen and phosphorus on
oceanic primary production. Nature 1999; 400: 525-31.
Wasmund N. Occurrence of cyanobacteria blooms in the Baltic Sea
in relation to environmental conditions. Int Revue ges Hydrobiol
1997; 82: 169-84.
Rahm L, Jönsson A, Wulff F. Nitrogen fixation in the Baltic
proper: an empirical study. J Mar Syst 2000; 25: 239-48.
Larsson U, Hajdu S, Walve J, Elmgren R. Baltic Sea nitrogen fixation estimated from the summer increase in the upper mixed layer.
Limnol Oceanogr 2001; 46: 811-20.
Wasmund N, Voss M, ad Lochte K. Evidence of nitrogen fixation
by non-heterocystous cyanobacteria in the Baltic Sea and recalculation of a budget of nitrogen fixation. Mar Ecol Prog Ser
2001; 214: 1-14.
Wasmund N, Nausch G, Schneider B, Nagel K, Voss M. Comparison of nitrogen fixation rates determined with different methods: a
study in the Baltic Proper. Mar Ecol Prog Ser 2005; 297: 23-31.
OECD. Eutrophication of waters. Monitoring, assessment and
control. OECD, Paris 1982; p. 154.
Smith VH. Predictive models for the biomass of blue-green algae
in lakes. Water Resour Bull 1985; 21: 433-9.
Håkanson L, Bryhn AC, Hytteborn JA. On the issue of limiting
nutrient and predictions of bluegreen algae in aquatic systems. Sci
Total Environ 2007; 379: 89-108.
Melvasalo T, Niemi Å, Niemitö L, Rinne I. On the importance of
nitrogen fixation in the Baltic Sea ecosystem. Symposium on Ecological Investigations of the Baltic Sea Environment, Riga 16-19
March 1983; pp. 176-89.
Leppänen JM, Niemi Å, Rinne I. Nitrogen fixation of cyanobacteria (blue-green algae) and nitrogen cycle of the Baltic Sea. Symposis 1988; 6: 181-94.
Rönner U. Nitrogen transformations in the Baltic proper: denitrification counteracts eutrophication, Ambio 1985; 14: 134-8.
Hübel H, Hübel M. Blaualgen-Wasserblüten in der Ostsee: Ursachen-Ausmass-Folgen. Dtsch Hydrogr Z (Suppl) 1995; 2:151-8.
Redfield AC, Ketchum BH, Richards FA. In: Hill N, Ed. The influence of organisms on the composition of sea-water. The Sea 2. Interscience, New York 1963; pp. 26-77.
Hassett RP, Cardinale B, Stabler LB, Elser JJ. Ecological
stoichiometry of N and P in pelagic ecosystems: comparison of
lakes and oceans with emphasis on the zooplankton-phytoplankton
interaction. Limnol Oceanogr 1997; 42: 648-62.
Vidal M, Duarte CM, Agusti S. Dissolved organic nitrogen and
phosphorus pools and fluxes in the central atlantic ocean. Limnol
Oceanogr 1999; 44: 106-15.
Dodds WK. Misuse of inorganic N and soluble reactive P concentrations to indicate nutrient status of surface waters. JNABS 2003;
22: 171-81.
Dillon PJ, Rigler FH. The phosphorus-chlorophyll relationship in
lakes. Limnol Oceanogr 1974; 19: 767-73.
Meeuwig JJ, Kauppila P, Pitkänen H. Predicting coastal eutrophication in the Baltic: a limnological approach. Can J Fish Aquat Sci
2000; 57: 844-55.
Håkanson L. Suspended particulate matter in lakes, rivers and
marine systems. The Blackburn Press, New Jersey 2006; p. 331.
Floderus S. The effect of sediment resuspension on nitrogen cycling in the Kattegat-variability in organic matter transport. UNGI
Report 71, Dr. thesis, Uppsala University, Sweden 1989.
Wulff F, Rahm L, Larsson P, Eds. A systems analysis of the Baltic
Sea. Springer-Verlag, Berlin, Ecological Studies 2001; Vol. 148: p.
455.
Ryther JH, Dunstan WM. Nitrogen, phosphorus, and eutrophication
in the coastal marine environment. Science 1971; 171: 1008-13.
Guildford SJ, Hecky RE. Total nitrogen, total phosphorus, and
nutrient limitation in lakes and oceans: Is there a common relationship? Limnol Oceanogr 2000; 45: 1213-23.
Wetzel RG. Limnology. Academic Press, London 2001; p. 1006.
Moisander PH, McClinton E, Paerl HW. Salinity effects on growth,
photosynthetic parameters and nitrogenase activity in estuarine
planktonic cyanobacteria, Microbila. Ecology 2002; 43: 432-42.
Havens KE, James RT, East TL, Smith VH. N:P ratios, light limitation, and cyanobacterial dominance in a subtropical lake impacted
by non-point source nutrient pollution. Environ Pollut 2003; 122:
379-90.
Reynolds CS, Walsby AE. Water-blooms. Biol Rev 1975; 50: 43781.
48 The Open Marine Biology Journal, 2009, Volume 3
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
Håkanson et al.
Edler Å. Blue-green algal blooms and N:P ratio in the Baltic Sea.
Acta Bot Fennica 1979; 110: 57-61.
Konopka A, Brock TD. Effect of temperature on blue-green algae
(Cyanobacteria) in Lake Mendota. Appl Environ Microbiol 1978;
36: 572-6.
Lehtimäki J, Sivonen K, Luukainen R, Niemelä SI. The effects of
incubation time, temperature, light, salinity, and phosphorus on
growth and hepatotoxin production by Nodularia strains. Arch Hydrobiol 1994; 130: 269-82.
Lehtimäki J, Moisander P, Sivonen K, Kononen K. Growth, nitrogen fixation, and Nodularin production by two Baltic Sea Cyanbacteria. Appl Environ Microbiol 1997; 63: 1647-56.
McQueen DJ, Lean DRS. Influence of water temperature and nitrogen to phosphorus ratios on the dominance of blue-green algae
in Lake St. George, Ontario. Can J Fish Aquat 1987; 44: 598-604.
Marumo R, Asaoka O. Distribution of pelagic blue-green algae in
the Northen Pacific ocean. J Oceanogr Soc Jpn 1974; 30: 77-85.
Lukatelich RJ, McComb AJ. Nutrient levels and the development
of diatoms and blue-green algal blooms in a shallow Australian estuary. J Plankton Res 1986; 8: 597-618.
Robarts RD, Zohary T. Temperature effects on phytosynthetic
capacity, respiration, and growth rates of bloom-forming cyanobacteria. NZ J Mar Freshwater Res 1987; 21: 391-9.
Howarth RW, Marino R, Cole JJ. Nitrogen fixation in freshwater,
estuarine, and marine ecosystems 2. Biogeochemical controls. Limnol Oceanogr 1988; 33: 688-701.
Howarth RW, Marino R, Lane J, Cole JJ. Nitrogen fixation in
freshwater, estuarine, and marine ecosystems. 1. Rates and importance. Limnol Oceanogr 1988; 33: 669-87.
Marino R, Chan F, Howarth RW, Pace ML, Likens GE. Ecological
constraints on planktonic nitrogen fixation in saline estuaries. I.
Nutrient and trophic controls. Mar Ecol Prog Ser 2006; 309: 25-39.
Sellner KG. Physiology, ecology and toxic properties of marine
cyanobacteria blooms. Limnol Oceanogr 1997; 42: 1089-104.
Zehr JP, Jenkins BD, Short SM, Steward GF. Nitrogenase gene
diversity and microbial community structure: a cross-system comparison. Environ Microbiol 2003; 5: 539-54.
HELCOM. Baltic Sea Environment Proceedings No. 100 from the
HELCOM website 2000.
Received: October 17, 2008
[44]
[45]
[46]
[47]
[48]
[49]
[50]
[51]
[52]
[53]
[54]
[55]
Pienitz R, Smol JP, Lean DRS. Physical and chemical limnology of
59 lakes located between the southern yukon and the tuktoyaktuk
peninsula, Northwest Territories (Canada). Can J Fish Aquat Sci
1997; 54: 330-46.
Håkanson L, Malmaeus JM, Bodemar U, Gerhardt V. Coefficients
of variation for chlorophyll, green algae, diatoms, cryptophytes and
blue-greens in rivers as a basis for predictive modelling and aquatic
management. Ecol Modell 2003; 169: 179-96.
Håkanson L. Physical geography of the Baltic. The Baltic University. Session 1. Uppsala University 1991.
Vitousek PM, Cassman K, Cleveland C, et al. Towards an ecological understanding of biological nitrogen fixation. Biochemistry
2002; 57/58: 1-45.
Håkanson L, Lindgren D. On regime shifts and budgets for nutrients in the open Baltic Proper-evaluations based on extensive data
between 1974 and 2005. J Coast Res 2008; 24: 246-60.
SNV. Eutrophication of land, freshwater and the Sea (in Swedish,
Eutrofiering av mark, sötvatten och hav). Swedish Environmental
Protection Agency, Report 4134, Stockholm 1993; p. 199.
Håkanson L, Boulion V. The Lake Foodweb-modelling predation
and abiotic/biotic interactions. Backhuys Publishers, Leiden 2002;
p. 344.
Golubkov MS, Golubkov SM, Gubelit YI, Litvinchuk LF. Primary
production and composition of phytoplankton assemblages in hypersaline lakes of Crimea peninsula. Proc Zool Inst Russ Acad Sci
2006.
Wallin M, Håkanson L, Persson J. Load models for nutrients in
coastal areas, especially from fish farms (in Swedish with English
summary). Nordiska ministerrådet, Copenhagen 1992; 502: p. 207.
Petersen JK, Hansen JW, Laursen MB, Conley D. Regime shift in a
marine coastal ecosystem. Manuscript, National Environmental Research Institute and County of Ringkjøbing 2006.
Bryhn AC, Håkanson L, Eklund JM. Variabilities and uncertainties
in key coastal water variables as a basis for understanding changes
and obtaining predictice power in modelling. J Coast Res 2008; in
press.
Håkanson L, Bryhn AC. Eutrophication in the Baltic Sea-present
situation, nutrient transport processes, remedial strategies. Springer
Verlag, Berlin, Heidelberg 2008; p. 264.
Revised: December 2, 2008
Accepted: December 3, 2008
© Håkanson et al.; Licensee Bentham Open.
This is an open access article licensed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/bync/3.0/) which permits unrestricted, non-commercial use, distribution and reproduction in any medium, provided the work is properly cited.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Related manuals

Download PDF

advertisement