response of phytoplankton to weather changes in esthwaite water · 2019-09-04 · degree project in...
TRANSCRIPT
Response of phytoplankton to weather changes in Esthwaite Wateras estimated by a submersible spectrofluorometer
Degree project in biology, 2007 Examensarbete i biologi 20 p, 2007Biology Education Centre and Department of Limnology, Uppsala University / Centre of Ecology and Hydrology, Lancaster, EnglandSupervisors: Kurt Pettersson and Stephen Maberly
Maria Wiik
1
Abstracts
The causes of fluctuating composition and dynamics of phytoplankton is a subject worthwhile studying to understand how the aquatic ecosystem works and how it would respond to climate change. Responses of phytoplankton composition and dynamics due to fluctuating environmental parameters affected by weather conditions were investigated in this study on a seasonal to a daily scale. Phytoplankton group concentration was measured using thesubmersible spectrofluorometer, bbe FluoroProbe. It has the ability to detect fluorescence signals from the phytoplankton groups; blue-green algae, green algae, cryptophytes and diatoms & dinoflagellates independently allowing continuous concentration measurements of each phytoplankton group. Traditional methods of monitoring phytoplankton composition and concentration tested and completed the results of the FluoroProbe.
Although deviation between spectrofluorometer and traditionally measured chl-a concentration and phytoplankton composition was noted, the methods indicated similar fluctuations in phytoplankton concentration and group fluctuations. A strong negative correlation between blue-green algae and solar radiation was noted and mainly caused by downward vertical migration at high solar radiation. A pulse of soluble reactive phosphorus coinciding with a heavy rainfall caused a shift in the composition to a dominance of green algae of high diversity in the surface waters. The green algae group was further positively correlated with the warm surface water temperature and high stability of the stratification. In the beginning of the green algae bloom, smaller species with high affinity for nutrients dominated and at the end of the bloom species with higher nutrient storage capacity were dominant. Cryptophytes showed a similar pattern as green algae increasing with high SRP and stability. The dominant cryptophyte Cryptomonas sp. showed daily upward migration due to its motility which was higher when solar radiation was low. With decrease in solar radiation stability decreased and caused a shift towards blue-green algae due to its ability of migration to overcome the vertical separation between nutrients and light and exploit both resources. Diatoms were positively correlated with wind and mixing and were abundant in the surface water at conditions similar to blue-green algae. Green algae showed a distinct daily decrease correlated to solar radiation due to non-photochemical quenching as a protection to over-excitation at the hours of highest solar radiation. In conclusion it was seen that phytoplankton community structure and size was highly affected by the fluctuating water column conditions correlated to changing weather.
2
1 Introduction ............................................................................................................................. 32 Background ............................................................................................................................. 4
2.1 Seasonal and diel response to meteorological and environmental variables ................... 43 Material and methods .............................................................................................................. 6
3.1 Study site .......................................................................................................................... 63.1.2 Phytoplankton dynamics in Esthwaite Water............................................................ 6
3.2 Sample collection ............................................................................................................. 73.3 Phytoplankton composition.............................................................................................. 73.4 Meteorological and thermal data...................................................................................... 8km from Esthwaite Water on Monday to Friday.................................................................... 83.5 Nutrient and light ............................................................................................................. 93.6 Chlorophyll a analysis...................................................................................................... 93.7 Data analysis .................................................................................................................... 93.8 Statistical analysis .......................................................................................................... 10
4 Results ................................................................................................................................... 114.1 Accuracy of FP measurements........................................................................................... 11
4.2 Seasonal patterns ............................................................................................................ 124.2.1. Weather and light conditions ................................................................................. 124.2.3 Phytoplankton dynamics ......................................................................................... 144.2.3 SRP and pH ............................................................................................................. 15
4.3 Correlations between abundance of phytoplankton groups and physical data............... 164.3.1 Green algae ............................................................................................................. 164.3.2 Blue-green algae ..................................................................................................... 164.3.3 Diatoms & Dinoflagellates...................................................................................... 164.3.4 Cryptophytes............................................................................................................ 174.3.5 Redundancy analysis (RDA).................................................................................... 17
4.4 Short-term patterns......................................................................................................... 194.4.1 Weather ................................................................................................................... 194.4.2 Diel variation in phytoplankton concentration ...................................................... 19
5 Discussion ............................................................................................................................. 215.1 Fluoroprobe measurements ............................................................................................ 215.2 Correlations between phytoplankton and environmental variables ............................... 215.3 Phytoplankton group dynamics and daily patterns ........................................................ 22
6 Conclusions ........................................................................................................................... 25Acknowledgements .................................................................................................................. 26References ................................................................................................................................ 27
3
1 Introduction
In lakes, especially small and shallow ones, weather variables play a central role in creating fluctuating conditions. Daily solar radiation, air temperature and wind are variables directly or indirectly affecting phytoplankton, their productivity and composition. Weather conditionsinfluence variables in the lake such as temperature, light condition, stratification and nutrient conditions. In turn this determines the success of different phytoplankton divisions and species, and therefore, the total primary production in the lake (Fig. 1). All phytoplankton groups require sufficient light for photosynthesis and sufficient nutrients to produce biomass (Reynolds 1984). The availability of these varies temporally and spatially and is strongly affected by water movements. Phytoplanktons are generally denser than water, so without water movements and circulation non-motile species would generally sink down in the water column to levels where light is not sufficient for photosynthesis (Huisman & Sommeijer2002). Water circulation is generated by weather conditions such as warming, cooling andwind stress (Talling 1981).
Figure 1. Effects of weather conditions on phytoplankton composition. (Gaedke et al. 1998)
In the future, the effects of presumed climate changes on the environment, including lake ecosystems, may become apparent. Better understanding of the pelagic food web as affectedby weather conditions is therefore of great importance for evaluating possible effects (Gaedke et al. 1998). As primary producers, phytoplanktons are important links in the food web of lake ecosystems. The short generation time of phytoplankton, in the order of days, allow quick responses to environmental changes. Short-term weather conditions can rapidly alter the performance of specific phytoplankton (Reynolds 2006). Knowledge of the influence of weather conditions on species’ composition and seasonal periodicity is therefore important both when generally studying lake ecosystems and in relation to climate change.
Previous studies in this area have mainly been based on manual algal counts from samples taken at a limited temporal resolution, resulting in inadequate information to draw significant conclusions. Recently, modern techniques have made high resolution phytoplankton measurements possible. In this study a relatively new instrument FluoroProbe (bbe Moldaenke), was used to perform automatic and continuous measurements of the phytoplankton groups blue-green algae (cyanobacteria), green algae (chlorophytes), cryptophytes (cryptophyceae) and diatoms (bacillariophyceae) & dinoflagellates(dinophyceae) in combination based on their independent fluorescence spectra. Although many species within each phytoplankton group share similar ecological requirements and behaviour there are exceptions in each group. It is obviously not sufficient to simply study
solar radiation
airtemperature& wind
water temperature stratification &
vertical mixing
light climate, primary production
process rateslight climatenutrient supply
Shifts in competition strength and taxonomical composition in phytoplankton
4
phytoplankton group changes in order to fully understand responses of environmental variables to phytoplankton succession. Regular but less frequent traditional sampling was used for phytoplankton species detection, to aid interpretation of FluoroProbe data. The meteorological and environmental variables incident solar radiation, wind speed, air and water temperature, rainfall, soluble reactive phosphorus, silica and pH were measured continuously, parallel to the phytoplankton measurements. Investigations of the most influential environmental variables to phytoplankton dynamics were performed on long-termand daily scales. Esthwaite Water was chosen as the study area because of its richness in phytoplankton and the weather induced fluctuating conditions in the lake. The lake has been scientifically studied since 1939 and historical data are therefore available for the site. It is of great scientific interest to study phytoplankton dynamics continuously and with high frequency to understand succession patterns. It is hoped that this study will unravel and quantify the importance of weather changes to the composition and dynamics of a phytoplankton community.
2 Background
2.1 Seasonal and diel response to meteorological and environmental variables
There is a strong relationship between phytoplankton dynamics and physical changes, and the interactions prevent equilibrium conditions (Sousa 1984). Environmental variables act on phytoplankton dynamics at short-term, long-term and seasonal scales. Meteorological events, hydrological processes and chemical gradients affect the vertical and horizontal distribution of phytoplankton at daily to weekly scales, whilst gradual changes in solar radiation, thermal stratification, nutrient replenishment or depletion act at long-term and seasonal scales (Salmaso 2003).
Diel and seasonal changes in the input of solar radiation cause the most significant changes in lakes, increasing the water temperature and the rate of phytoplankton photosynthesis (Talling 2004). Heating of the upper water layer causes thermal stratification, a vertical separation of the water column into layers more or less resistant to mixing with each other due to temperature and density differences. Solar radiation and wind strength determine the strength of thermal stratification and the depth of the mixed layer (Talling 1981). Non-motile phytoplankton depend on conditions in the upper mixed layer since the depth of the mixed layer affects light and nutrient conditions (Huisman & Weissing 1995). Light conditions also depend on the clarity and colour of the water, expressed by the euphotic depth. As a rule of thumb significant phytoplankton photosynthesis takes place down to the euphotic depth, defined as the depth at which photosynthetically active radiation (PAR, 400-700 nm) is 1% of that just below the water surface (Kirk 1994). Wind conditions determine the degree of turbulence in the mixed layer, affecting the distribution of phytoplankton with respect to environmental gradients and, in turn, their growth. High production and photosynthesis of phytoplankton increase pH by CO2 depletion. Elevated pH is most likely during stratified conditions with high phytoplankton biomass whilst strong wind and rainfall cause mixing and lead to decline in pH (Maberly 1996).
Nutrients are crucial for the production of biomass. The two most crucial nutrient supplies for phytoplankton are nitrogen and phosphorus and in addition for diatoms, silica is essential. The supply of phosphorus is normally much less than nitrogen (nitrate and ammonia) in lake ecosystems and most lakes are phosphorus limited, although many upland lakes may be nitrogen co-limited (Maberly et al. 2002). Rainfall is the most important meteorological
5
variable that can enhance the total phosphorus levels significantly in the lake through run-off from the drainage area (Reynolds 2006). Partial breakdown of stratification can increase theconcentration of phosphorus in the mixed layer. A specific requirement for silica is seen in diatoms, as it is an important component of their cell walls. Utilisation of silica by diatoms can reduce the concentration in the mixed layer, causing competition or deficiency. A concentration below 0.05 mg Si /L is the lower limit for many diatom species and is often the cause of the abrupt spring bloom decline (Lund 1964).
Changes in environmental variables produce constant fluctuations in the lake environment making the lake ecosystem highly dynamic with respect to phytoplankton species composition. In the recent study by Madgwick et al. (2006) on Esthwaite Water it was found that phytoplankton community change is not always instantaneously correlated to environmental variables but depends upon antecedent conditions. Increased correlations with phytoplankton community variables were shown for all environmental variables including stability, solar radiation and soluble reactive phosphorus (SRP) when considering antecedent conditions of 6-7 days rather than instantaneous environmental measurements. A problem arising when estimating the quantitative importance of individual physical variables to phytoplankton dynamics is their colinearity, as well as their relationships with other biological and chemical ”predictors”. The relevance of high resolution studies is therefore high when attempting explanations of dynamical patterns (Madgwick et al. 2006).
To understand long-term changes in phytoplankton groups and species, it is necessary to study their responses to short-term environmental variability, in order to identify their ecological requirements and behaviour. The abundance and distribution of phytoplankton on a daily basis fluctuates mainly due to external factors such as light and water currents induced by wind and thermal stratification (Frempong 1981). The ability for active or passive movements by phytoplankton is also important. These are properties of great advantage, allowing a species to overcome the vertical separation between light and nutrients. Active movements are undertaken by species with one or more flagella, like dinoflagellates some green algae and cryptophyte species. Passive movements are undertaken mainly by blue-green algae species with gas-vacuoles, which permit the regulation of buoyancy and position in thewater column (Reynolds et al. 1987). Light in excess can cause photoinhibition as a result of damage to the photosynthetic apparatus by destructive photooxidation (Falkowski et al.1994). Exposure to high irradiance therefore reduces buoyancy by collapse or decreased production of gas vacuoles. At low irradiance increased production of gas-vacuoles tends to be induced (Dinsdale & Walsby 1972) Blue-green algae further tend to decrease buoyancy when major nutrients such as phosphorus, nitrogen and carbon dioxide are limited (Klemer 1991). Endogenous rhythms also affect the distribution and activity of phytoplankton on a daily basis (Eppley et al. 1968). The diel variation of non-motile species tends to be determined mainly by wind-induced water movements. Suppression of diel variation in non-motile phytoplankton tends to appear with wind induced water currents causing random distribution (Frempong 1981).
The bbe FluoroProbe has been used in published studies by Gregor et al. (2005) and See et al. (2005) monitoring the phytoplankton community parallell to traditional methods analysing Chl-a and taxonomy. Both studies evaluated that the FluoroProbe is a good and effective toolin observing the phytoplankton community structure, especially in tandem with other methods.
6
3 Material and methods
3.1 Study site
The study was performed on Esthwaite Water (5421’N, 30’W), a small and shallow lake in the English Lake District. Esthwaite Water is one of the most eutrophic and productive lakesin the area which is related to its position in a shallow valley with relatively rich alluvial soil (Heaney et al. 1986). Average winter concentrations of SRP and total phosphorus for 2004-2005 were 14 μg/L and 55 μg/L respectively. The high biomass and diversity of phytoplankton in Esthwaite Water makes it an appropriate site for studying short-term changes in a phytoplankton community. Esthwaite Water is an elongated lake with a length of 2.5 km and an area of 1 km2. The catchment area is 17 km2 and the average water retention time is about 90 days, maximum depth is 15.5 m and mean depth 6.4 m. The short retention time results in significant effects of rain flushing on seasonal dynamics making Esthwaite Water a good lake to study short term changes (George et al, 2004). Annual precipitation in the Lake District is high (1500-4000 mm) increasing the incidence of flushing events (Reynolds & Lund, 1988).
Figure 2.Depth contour map of Esthwaite Water Position of bbe FP, Meteorological monitoring and water sampling; Position of long-term monitoring site, pH, temperature profiles
3.1.2 Phytoplankton dynamics in Esthwaite Water
The phytoplankton composition in Esthwaite Water is diverse with a periodically high biomass and chlorophyll a concentrations above 100 g/L: indicating the high productivity of the lake. The seasonal biomass maxima commonly occur as a diatom bloom in spring and a blue-green algae bloom in late summer (Talling 1999). The spring maximum is dominated by diatom species that are sensitive to SiO2 depletion and stratification but tolerant of light deficiency. After the spring bloom an abundance of single celled green algae, preferring shallow mixed layers, and the flagellated Cryptomonas and Peridinium, tolerating low light, is likely. The late summer maxima often consists of the filamentous and buoyant blue-green algae Anabaena flos-aquae and Aphanizomenon flos-aquae sensitive to mixing and poor light, and sometimes by species of the often deeply distributed dinoflagellate Ceratium (Reynolds2006, Talling & Heaney 1984). Additional maxima in spring and early summer have occurred, but are less frequent. The upper stratified layer holds most of the phytoplankton biomass during summer although deep maxima can develop through vertical migration (Talling, 1999). Buoyant blue-green algae such as Anabaena flos-aquae and Aanabaena
7
solitaria occur deeper down in the water column at high solar radiation. Motile flagellates such as Ceratium hirundinella and Cryptomonas sp. are also present and respond to changing light levels by migrating towards the surface water with increasing irradiance during daytime (Frempong 1981).
3.2 Sample collection
The study was carried out from 11th of May until 15th of August 2006 to cover phytoplankton dynamics during late spring and summer. Phytoplankton fluorescence was continuously recorded at a fixed buoy situated in the northern basin at a depth of 13 m (Fig. 2). Once or twice a week an integrated water sample from 0-5 m depth was collected for chlorophyll a(Chl-a) analysis. Meteorological data were recorded at the same location. Water temperature and pH were recorded from an adjacent buoy (Fig. 2). Results from fortnightly sampling of Esthwaite Water, performed by staff at the Centre of Ecology and Hydrology in Lancaster, were used for nutrients, light attenuation and algal counts. Daily rainfall data was collected by the Freshwater Biological Association at Windermere.
3.3 Phytoplankton composition
An automatic submersible spectrofluorometer, bbe FluoroProbe (FP), was deployed from the centre of the buoy on Esthwaite Water and submersed to a measuring depth of 1.1 m. Fluorescence measurements were collected continuously every 30 minutes with a measuring duration of 5 seconds. Data were recorded continuously from the 11th of May until 15th of August. The data were retrieved twice a week until 23rd of June at the same time as instrument servicing. From 23rd of June the same procedure was performed once a week. Instrument services consisted of removing the protective shell from the FP and cleaning the screens of the light emitting diodes to prevent false measurements caused by contamination.
The principle of algal group specific determination with the bbe FP is the fact that algae of the same division contain a similar quantity and quality of photosynthetic pigments. All oxygenic photosynthesis is based on the pigment chl-a, found in every phytoplankton species and the fluorescence emission from chl-a is measured by the FP. However, specific accessory pigments can additionally be found in the different groups (Table 1) that pass energy to chl-a. The five different divisions of algae; green algae, blue-green algae, dinoflagellates and diatoms in combination and cryptophytes can be distinguished by the FP because the different accessory pigments result in differential absorption of given wavelengths (Fig. 3). The relative fluorescence intensity from diatoms, dinoflagellates and chrysophytes coincide with each other (Fig. 3) and cannot be distinguished individually. The FP uses LEDs (light emitting diodes) of five selected wavelengths appropriate to excite the different accessory pigments differentially (Beutler et al. 2002). The LED light pulses are reemitted by the phytoplankton as chl-a fluorescence and detected and further processed. During the measurements the spectra of the sample are stored in the instrument and in the downloading stage the data are converted to give a chl-a concentration for each division. The FP was initially calibrated with known concentrations of algal cultures representative of each division (http://www.bbe-moldaenke.de). Since all phytoplankton generally contain a similar quantity of chl-a, fluorescence can be used as an estimator of biomass (Falkowski & Kiefer 1985). In vivo fluorescence per chl-a unit can vary with changes in species composition, physiological state and particularly with exposure to strong sunlight which should be taken in consideration (Kromkamp & Forster 2003, Gregor et al. 2005).
8
Table 1. Accessory photosynthetic antenna pigments of the phytoplankton groups. (Van den Hoek el al. 1995), + pigment occur in some species.
chlo
roph
yll b
chlo
roph
yll c
fuco
xant
hin
peri
din
phyc
ocya
nin
allo
phyc
o-cy
anin
phyc
oery
thri
n
phyc
obil
isom
es
zeax
anth
in
echi
neoo
n
cant
haxa
nthi
n
myx
oxan
toph
yll
osci
llax
anth
in
Blue-green algae X X X X X X X X XGreen algae XCryptophytes X X XDiatoms X XDinoflagellates X + XCrysophytes X X
In addition, the number of specific phytoplankton species was counted on a fortnightly basis with an inverted microscope after preservation and sedimentation in Lugol’s iodine solution (Lund et al. 1958). Cell, filament and colony counts were converted to biovolumes v(µm3/cell) according to Madgwick et al. (2006) and Laurence Carvalho (unpublished, CEH Edinburgh).
Figure 3: Relative fluorescence intensities for phytoplankton divisions measured by the FP: bluegreen algae (cyanophyceae), green algae (chlorophyceae), cryptophytes (cryptophyceae) and the integrated group of diatoms (bacillariphyceae) and dinoflagellates (dinophyceae) relatively normalised by the intensity of the LED. Deviation for species within the groups are marked as dots and triangles.
3.4 Meteorological and thermal data
Wind speed was continuously recorded at an interval of two minutes with a Vector A100L2-WR anemometer with an accuracy of 2%. Incident solar radiation was measured at two minute intervals with a Didcots instrument DRN-305 radiometer with an accuracy of approximately 10 % and an albedo of 0.07 according to Nunez et al. (1972). Depth profiles of temperature were measured in situ with a twelve metre PRT (Platinum Resistance Thermometer) thermistor chain recording lake temperature at length intervals of 1 metre, every two minutes with an accuracy of 0.1°C (Rouen et al. 2000). Air temperature was measured at two minute intervals from a buoy on the lake using a Skye Instruments SKH2012
9
probe, comprising a PRT with a maximum error of 0.05C. Rainfall was measured by the Freshwater Biological Association at Windermere Ferry House about 4 km from Esthwaite Water on Monday to Friday.
3.5 Nutrient and light
Concentrations of SRP (Soluble Reactive Phosporus) and Silica (SiO2) were sampled fortnightly and determined according to Mackereth et al. (1978). Light (μmol photons m2/s) was measured fortnightly with a LI-250A light meter made by LI-COR Biosciences at each metre of the water column down to 9 m depth.
3.6 Chlorophyll a analysis
Chl-a concentration was measured twice a week from 10th of May until 23rd of June and thereafter once a week. These data were used to compare with the total chl-a concentration measured by the FluroProbe. Chl-a concentrations were measured by filtering 100-200 ml of each sample through a glass microfibre filter (Whatman GF/C) of 25 or 50 mm diameter. The pigments were extracted by boiling the filters in tubes containing 7.6 ml methanol (95%). The tubes were centrifuged for 10 minutes at 2500 rpm after which absorbance was measured at the wavelengths 665 and 750 nm. Chl-a concentration could then be calculated from the absorbance values according to Talling (1974).
3.7 Data analysis
Stability was determined according to the Schmidt method (Hutchinson, 1957). Schmidt stability (S) is defined as the amount of physical work (J/m2) required to mix the water column to a uniform temperature. Schmidt stability quantifies the stability of the stratification to wind disruption. The value depends upon the current temperature profile and lake size and morphometry. Morphometrical information for Esthwaite Water is presented in Ramsbottom (1976). Schmidt stability (S) was calculated using the formulas:
mz
zzg dzAzzA
gS
0
max0
)()(
S= Schmidt stability (J/m2), g = acceleration due to gravity (m/s2), A0 = surface area (m2), Az
= area at depth z (m2), zm = the maximum depth (m), z = the density of pure water at depth z(kg/m3), max = the maximum density of pure water (kg/m3). Zg = depth of centre of gravity(m). v=volume of the lake (m3);
The mixing depth was calculated from the thermistor data and is defined as the circulating upper layer of warm water with more or less uniform temperature (Wetzel 2001). Mixing depth in this study was determined as the depth where the temperature difference between surface water and deeper water was 1 °C.
mz
zg dzzAV
z0
1
10
Light conditions in the upper water column (1-5 m depth) could be estimated through determination of the euphotic depth, zeu. The value of zeu can be determined from the vertical light attenuation coefficient Kd by the simplified formula 4.6/Kd assuming that Kd is constant with depth (Kirk 1994). The vertical attenuation Kd coefficient for downward attenuation was calculated as the slope of the measured and log-transformed light values in the water column.
Relationships between phytoplankton dynamics and meteorological and thermal data wereinvestigated. Phytoplankton concentration, meteorological and thermal data were averaged over a 24 hour period. Meteorological and thermal factors were further studied for correlations at 24, 48, 72, 96, and 120 hours prior to the time of phytoplankton sampling. The purpose was to see if longer-term weather conditions were more strongly correlated with phytoplankton dynamics than short-term weather changes. Linear interpolation of SRP, SiO2
and water column light attenuation was performed to produce values at the same temporal resolution as the FP measurements.
Diel variations in phytoplankton group abundance caused by external factors in the environment including, active or passive migration and other endogenous rhythms was also investigated. Daily patterns influenced by solar radiation and wind, the most important variables at the daily scale, were investigated.
When comparing total chl-a measured by the FP and traditional analysis, average night concentrations between midnight and 4 am were used. This was to eliminate underestimation of chl-a through non-photochemical quenching during the light hours. Excess light can cause destruction of the photosynthetic apparatus by over-excitation. Non-photochemical quenching will cause lower quantities of fluorescence signals due to relaxation of the photochemical centre (Long et al. 1994).
3.8 Statistical analysis
Multiple linear regression was used to determine the effects of environmental factors on phytoplankton dynamics, and their significance. Multiple regression allows the effects of multiple independent variables to be simultaneously tested and modelled. For each phytoplankton group, the partial effects of each environmental variable were evaluated when controlling for all other variables. Initially starting with all factors, one after one they wereeliminated until the model only included statistically significant factors. Non-significant terms (P>0.05) were excluded using a stepwise procedure. The method created a model for the dynamics of each phytoplankton group including only statistically significant relationships with environmental variables. Multiple regression was performed using MINITAB 14. The adjusted coefficient of determination (R2
adj) was used to describe to what degree the variance in phytoplankton abundance was explained by the model. Variance inflation factors for model terms were <5, indicating that there was no problem with colinearity between environmental variables. The models were tested for normality, homogeneity and independence using MINITAB.
These univariate analyses were complemented with a multivariate analysis of the phytoplankton community data. Canonical and direct gradient analysis was performed to relate the phytoplankton groups directly to measured environmental factors. In this case linear relationships were assumed and RDA (Redundancy analysis) was used (Rao 1964, Ter Braak 1986). RDA can be described as a series of multiple regressions followed by a principal component analysis. It is an important tool in interpretation of species-environment
11
interactions. The analysis was carried out using CANOCO 4.5 for Windows. For the analysis 1000 permutations were performed. A biplot of the result was created, to visualise the species-environment interactions, showing ordination assemblages along the first two ordination axes.
Pearson correlations for relationships studies and the student’s paired t-tests with the significance level P<0.05 were performed in this study.
4 Results
4.1 Accuracy of FP measurements
Measurements of total chl-a concentration were significantly different (t-test, t = 6.0, n = 23 P<0.001) from the FP and the analysed chl-a. The FP measured 31 % of the concentration value from the traditional chl-a analysis according to a linear regression of the two methods (Fig. 4). Although the results were related the FP seems to underestimate the chl-aconcentrations. However it should be noted that the FP conducts measurements at 1.1 m depth compared to the integrated water samples from 0-5 m depth for analysed chl-a.
y = 0.3123x + 4.3738
R2 = 0.3092
0
10
20
30
40
0 10 20 30 40 50
Fluoroprobe+Analysed
Linear(Fluoroprobe+Analysed)
Figure 4. Linear relationship between Chl-a concentration measured by FP and traditionally analysed based on sampling 1 or 2 times each week in Esthwaite Water between May 10th and August 8th 2006.
The percentage of the total concentration for each group measured by the FP (Fig. 5b) compared to the percentage based on biovolume estimations (Fig. 5a) show similar patterns with blue-green algae as the overall dominant phytoplankton group. The blue-green algae decreased drastically once during the period according to the FP which is not expressed inbiovolume estimations. The decrease in blue-green algae detected by the FP in the middle of July is not detected until August 1st. The green algae and cryptophytes have higher proportion overall in FP measurements but a similar pattern, with increasing proportion in late July.Although significant differences were noted between the methods, the dynamic patterns weresimilar which indicates that the FP functioned reasonably well in this study.
Analysed Chl-a (g/L)
FP
Ch
l-a
(g/
L)
12
0%
20%
40%
60%
80%
100%
16-M
ay
30-M
ay
13-Ju
n
27-Ju
n
11-Ju
l
25-Ju
l
0%
20%
40%
60%
80%
100%
16-M
ay
30-M
ay
13-Jun
27-Jun
11-Jul
25-J
ul
Figure 5. Percentage of phytoplankton groups in Esthwaite Water based on a) phytoplankton counts andbiovolumes between May 16th and July 25th 2006 in Esthwaite Water (weekly samples). b) FP measurements on the same dates as a.
Servicing of the FP was revealed to be of great importance. The concentration of the phytoplankton group diatoms & dinoflagellates was significantly different before and after cleaning of the LED screens (Fig. 6) Discounting results taken in May when concentrations of diatoms & dinoflagellates were negligible, there was a statistically significant decrease in the concentration of these algae after servicing (t-test, t = 2.8, n = 14, P=0.016). The average decrease was 18 % of the precedent concentration.
0
1
2
3
4
5
6
7
01-Ju
n
05-Ju
n
08-Ju
n
12-Ju
n
15-Ju
n
19-Ju
n
23-Ju
n
29-Ju
n
04-Ju
l
12-Ju
l
19-Ju
l
25-Ju
l
01-A
ug
04-A
ug
Date
chl-
a (u
g/L
)
before
after
Figure 6. Diatom & dinoflagellate Chl-a concentrations in Esthwaite Water June 1st –August 4th 2006, measured before and after instrument service
4.2 Seasonal patterns
4.2.1. Weather and light conditions
Water column temperature during the period is presented in Figure 7 and rain, solar radiation, stability, wind speed and air temperature are presented in Figure 8a and 8b. The total precipitation during the period May 11th-August 15th was 281 mm. During the measured period, May had the most precipitation. There were long periods without rain with high irradiance and air temperatures which resulted in a very shallow stratification and strong stability (Figure 8a). Highest stability and shallowest mixing depth was seen in July when
Blue-green algae
Cryptophytes
Green algae
Diatoms & Dinoflagellates
(b)(a)
Date Date
chl-
a (u
g/L
)
13
solar radiation, and air temperature showed highest mean. There was one short breakdown in stratification in July caused by a day of 30 mm rain. Between July 11th and August 1st there was no rainfall.
Figure 7). Isotherms (°C) measured in the water column of Esthwaite water May 10th - August15th 2006.
Periods with cool, windy weather and rainfall caused weakened stratification with deepening and cooling of the mixed layer. The most extensive mixing periods were at the end of May to the beginning of June to middle of August (Fig. 7). Lowest stability was calculated between May 20th and June 1st after an extended period of rain and high winds (Figs 8a, b). The wind strength varied during the study period with July being the least windy month.
0
100
200
300
May-06 Jun-06 Jul-06 Aug-06
0
10
20
30
40
Schmidt stability Solar radiation Rainfall
0
5
10
15
20
25
May-06 Jun-06 Jul-06 Aug-06
0
1
2
3
4
5
6
Air temperature Wind speed
Figure 8. Daily mean values of weather variables in Esthwaite Water May 11th-August 15th a) schmidt stability, solar radiation and rainfall, b) air temperature and wind speed.
Light conditions in the upper water column described by euphotic depth (zeu) varied during the period and decreased further into the summer months with the increase of phytoplankton biomass. Zeu was shallower than the mixing depth for short periods creating an environment that could cause light limitation to non-motile phytoplankton species (Fig. 9). These periods coincided with mixing periods.
Dep
th (
m)
a) b)
Date Date
So
lar
rad
iati
on
(W
/m2 )
Sch
mid
t st
abil
ity(
J/m
2 )
Rai
nfa
ll (m
m)
Air
tem
p. (
°C)
Win
d s
pe
ed (
m/s
)
150 180 210
-10
-5
89101112131415161718192021222324
JulyJune August
Date
14
0
2
4
6
8
May-06 Jun-06 Jul-06 Aug-06
Time
zeu mixing depth
Figure 9. Mixing depth and interpolated zeu in Esthwaite Water May 11th-August 15th 2006
4.2.3 Phytoplankton dynamics
The phytoplankton dynamics measured by the FP are illustrated in Figure 10. By the start of the measuring period the yearly diatom spring bloom had declined and was substituted by a dominant blue-green algae population. After a rapid decline of the blue-green algae population coinciding with a strong mixing event at the end of May the phytoplankton group diversity increased and the groups were all more or less abundant showing a variable dynamic pattern. The missing data over 4 days in end of June was caused by a power cut. The maximum total chl-a concentration was reached at the end of July until the beginning of August when all groups coexisted and reached their highest concentration during the period. Blue-green algae were the dominant group throughout most of the period except at the end ofJuly/beginning of August when green algae dominated. At the beginning of July there was a decline in total chl-a concentration mainly caused by a decrease in the dominant blue-green algae. At the end of July there was a decline of blue-green algae independent of the other groups lasting for two weeks. Total concentration of chl-a remained high due to a significant increase in green algae and cryptophytes. The concentration of cryptophytes was generally lower compared to the other groups. There were however two distinct peaks in cryptophyte abundance. A small peak was seen after the first decline in total phytoplankton concentration and a similar, but larger peak in end of July coinciding with a dominance of green algae.
Biovolume estimations from the 0-5 m zone were performed until August 1st around the end of the summer bloom. Altogether 86 taxa were identified and the highest diversity was seen in late July during the period of maximum phytoplankton biomass. The blue-green algae mainly consisted of species with gas vacuoles. At the start of the study period, the population was dominated by Anabaena circinalis/flos-aquae which after a couple of weeks was succeededby Aphanizomenon flos-aquae which remained dominant during the whole period. Between the two population declines in July there was a peak where the colonial non-buoyant species Aphanothece clathrata was briefly abundant. The cryptophyte group was strongly dominated by the flagellate Cryptomonas spp. throughout the period together with a high number of the small sized flagellated Rhodomonas sp which increased simultaneously with Cryptomonas spp. The green algae group was the most species rich group with altogether 40 different taxa. Colonies of Coenochloris fottii were plentiful and dominated the green algae biomass untilthe start of the bloom in July when species diversity increased. In June an abundance of the species Chlamydomonas sp., Chlorella sp. and later Cosmarium depressum were abundant. In
m
Date
15
late July, with the rapid increase in green algae abundance, a higher proportion of single cell species dominated particularly Chlorella sp and also colonial species like Micractinium sp.,Pauschulzia pseudovolvox and Chlamydocapsa s.. By the end of the bloom with a decline in stability species diversity decreased and the population was dominated by Cosmarium bioculatum and Paulschulzia tenera. The diatoms & dinoflagellates had very low concentrations in May after the spring bloom collapse. From the beginning of June, the population was increasing in size with growing numbers of the diatoms Asterionella formosa, Fragilaria crotonensis and Dinobryon divergens. The crysophyte Dinobryon was only dominant at one point in middle of June where after F. crotonensis completely dominated the population.
0
10
20
30
May-06 Jun-06 Jul-06 Aug-06
Green algae
Bluegreen algae
Diatoms & Dinoflagellates
Cryptophytes
Total concentration
Figure 10. Phytoplankton composition in Esthwaite Water based on daily averages of FP measurements, May 11th-August 15th 2006. The gap of data at end of May was caused by a power cut.
4.2.3 SRP and pH
The dynamics of SRP, pH and SiO2 are illustrated in Figure 11. Fortnightly measurements of SRP concentrations revealed a strong increase in July. Phytoplankton biomass increased shortly after and pH reached its’ daily maximum of pH 9.8. The concentration of SiO2 had a high peak after the diminishing of the diatom spring bloom in the beginning of May but decreased once more through diatom uptake. Asterionella formosa and Fragilaria crotonensiswere the most abundant diatoms and are both sensitive to silica depletion and stability. Lower concentrations than the threshold value were detected in one measurement in July. Fragilariacrotonensis was still abundant during this time indicating that diatoms were never severelysilica depleted during the period.
Date
chl-
a (u
g/L
)
16
0
2
4
6
8
10
12
May-06 Jun-06 Jul-06 Aug-06
0
0.2
0.4
0.6
0.8
SiO
2 (
mg
/l)
pH SRP SiO2
Figure 11) pH in Esthwaite water and interpolated concentrations of SRP and SiO2 from fortnightly integrated samples between from 1-5 m depth, May 11th-August 15th 2006.
4.3 Correlations between abundance of phytoplankton groups and physical data
The variables that showed significant relationships with phytoplankton group abundance in multiple regression and Redundancy Analysis (RDA) were solar radiation, wind speed, rainfall, SRP, stability and mixing depth, air temperature, pH, zeu and SiO2. The strength of the relationships, and their statistical significance, varied when phytoplankton were related to environmental data from 1-5 days prior to sampling.
4.3.1 Green algae
The strength of thermal stratification, quantified by the Schmidt stability, revealed the strongest relationship with the abundance of green algae (Table 2). A negative relationship was found with zeu and a marginally significant (P<0.10) positive relationship with SRP concentration. These variables explained 90 % of the variation in the abundance of green algae. When multiple regression was performed with environmental variables averaged for 2 to 5 days prior to the time of phytoplankton sampling, the same explanatory variables were statistically significant predictors, with increasing importance of stability and elimination of SRP (R2
adj = 0.86). Pearson correlation between green algae and stability was r =0.71, P<0.001.
4.3.2 Blue-green algae
During the study period blue-green algae displayed a strong negative relationship with solar radiation on a 24 h short-term basis (Table 2). On a long term basis (2 to 5 days) pH was included as explanatory variable with increasing positive correlation.
4.3.3 Diatoms & Dinoflagellates
Analysis of diatoms and dinoflagellates were performed on the results from the beginning of June. The concentrations before this were negligible. Diatoms & dinoflagellates showed a negative relationship with solar radiation and zeu (Table 2). The importance and significance of solar radiation was retained on a long-term basis (2-5 days), when SRP and rainfall were also included in the final statistical model. An increasing abundance of diatoms and dinoflagellates was related to increasing SRP concentrations and rainfall. The fact that this group consists of three functional phytoplankton groups with widely different characteristics can make the result difficult to interpret. Since diatoms are the main dominant group in this
Date
SiO
2 (m
g/L
)
pH
, SR
P(
g/L
)
17
study the characteristics of the group diatoms & dinoflagellates is mainly determined by the requirements of diatoms.
4.3.4 Cryptophytes
SRP was the most important variable affecting the abundance of cryptophytes (Table 1). On along-term basis (3-5 days) rainfall had a significant negative effect on the abundance of cryptophytes.
Table 2. Summary of explanatory variable(s) of phytoplankton variation in Esthwaite Water between May 11th
and August 15th 2006. The results were obtained from multiple regression analysis and includes degrees of freedom(n), F-statistics (F) and adjusted proportion of explained variation (R2
adj.) The coefficient factor for each explanatory variable is given. Results of environmental variables from the preceding day to FP measurements were used in the analysis. Coefficients in parentheses are obtained from analysis with environmental variables from 5 preceding days.
Green algae(Chl-a, μg/L)
Blue-green algae(Chl-a, μg/L)
Diatoms & dinoflagellates(Chl-a, μg/L)
Cryptophytes(Chl-a, μg/L)
Variable 1 day 5 days 1 day 5 days 1 day 5 days 1 day 5 daysStability (J/m2)
0.006*** 0.007*** - - - - - -
SRP (μg/L)
0.037ns(0.07) - - 0.46*** 0.65*** 0.64***
zeu (m) -0.37*** -0.366***
- - -0.93*** - - -
Solar radiation(W/m2)
- - 0.027*** -0.04*** 0.007*** 0.012*** - -
Rainfall (mm)
- - - - - 0.15*** - 0.067***
pH - - - 2.07*** - - -Statisticsn 83 83 86 83 75 75 93 89F 244*** 262*** 66*** 95*** 34*** 48*** 153*** 96***R2
adj 0.90 0.86 0.43 0.70 0.47 0.65 0.62 0.68***
significant P<0.001, ns(0.07)
not significant with P=0.07
4.3.5 Redundancy analysis (RDA)
RDA, which is an extension of multiple regression, was performed to obtain complementary information about the relationship between phytoplankton species and environmental variables. Results after the 1st of June were used in the analysis to minimise erroneous correlations in the diatom and dinoflagellate group. In an RDA of the whole data series the correlations were different for diatoms and dinoflagellates but principally unchanged for the other groups. The environmental variables included explained 86.6 % of the total variation in the phytoplankton dynamics. The biplot in Figure 11 illustrates the variation along the first two of four ordination axes, with 84.6 % explanation of the total variation (F75=34.1, P=0.001). The biplot explains the relationship between each phytoplankton group and the environmental variables. Each axis is an assemblage of different environmental explanatory variables, illustrated by the angles between the axis and the vectors representing each variable. The length of an arrow is relative to the importance of the variable in theassemblage. Axis 1 is a function of SiO2 and zeu and to some degree wind speed and mixing
18
depth and axis 2 a function of Stability, air temperature and SRP. The angles between phytoplankton group arrows and environmental arrows indicate the correlations.
The phytoplankton groups cryptophytes and green algae were favoured in fairly similar conditions. Green algae were positively correlated especially with SRP but also pH, stability and air temperature. Cryptophytes were found at similar conditions but were more strongly correlated to stability and air temperature. Diatoms and dinoflagellates and blue-green algae were found in similar conditions and were most abundant in surface water when mixing depth, rain and wind speed were high and solar radiation was lower. All phytoplankton groups were negatively correlated with zeu.
There were significant correlations between environmental variables. Stability was strongly correlated with air temperature, which can be seen in Figure 12 (r =0.86, P=0.000) and with pH (r=0.79, P=0.000). Rain and wind speed were positively correlated (r= 0.32, P=0.005) and in turn negatively correlated with solar radiation. When SRP was high it positively affected the production of phytoplankton, mainly cryptophytes and green algae, which increased the pH but decreased zeu. SiO2 was correlated to zeu due to diatoms reaching the highest concentration when zeu was low followed by uptake of SiO2 and decreased concentration.
.
-1.0 1.0
-0.8
1.0
G
B
DD
C
pH
WS
S
Sol
SRP
Rain
zeu
air
SiO2
Mix
Figure 12. Biplot of RDA analysis of relations between phytoplankton groups and measured environmental factors based on measurements in Esthwaite Water, June 1st -August 15th , 2006. The result presented in the graph explains 84.6 % of the total species variation (F75:34.1, P=0.001). Blue arrows: Phytoplankton groups: B,Bluegreen algae; DD,Diatoms and Dinoflagellates; G,Green algae; C,Cryptophytes. Red arrows: Environmental factors: SRP; S, stability; air,air temperature; Sol, solar radiation, SiO2, Silica concentration; zeu, euphotic depth; WS, wind speed; Mix, mixing depth.
Axis 1
Axi
s 2
19
-1
-0.5
0
0.5
1
0 4 8 12 16 20
Cryptophytes
Green algae
Blue-greenalgae
Diatoms &Dinoflagellates
4.4 Short-term patterns
4.4.1 Weather
Wind speed varied greatly from day to day and was on average higher during daytime than night time. In July there were many completely still nights. The thermal structure and stability of the lake also varied on a diel scale associated with solar radiation, air temperature and wind speed (Fig. 13a). Stability was positively correlated with solar radiation (r = 0.31, P = 0.005) and started increasing by sunrise with highest values seen between 12.00 and 23.00 followed by a gradual drop during the night hours. Incident solar radiation varied on a daily basis depending on cloudiness which also affected the stability positively with less cloud and negatively on cloudy days.
4.4.2 Diel variation in phytoplankton concentration
Noticeable daily patterns could be seen among all phytoplankton groups which were more or less distinct through the period. The highest range between daily minima and maxima was positively correlated for each group with the maximum concentration for the individual group on the day with r = 0.82-1. Cryptophytes had high daily fluctuations with at least one measurement of 0 g/L causing the range to be equal to the maximal concentration and an r-value of 1.
The diel variation in cryptophyte concentration was on average very high with a distinct pattern. The low and depleted concentrations were primarily seen between 8 pm and 8 am and higher concentrations were found at daytime (Fig. 13b). Concentration of diatoms & dinoflagellates also increased gradually during the light hours. It is likely that the expressed diel variation pattern over the period is caused by diatoms which are dominant most of the period. Green algae and blue-green algae showed the opposite pattern with concentration decrease during daytime. The decrease for green algae was limited to early morning, around sunrise, between 4 am and 8 am.
0
1
2
3
4
5
0 4 8 12 16 20
0
100
200
300
400
500
Wind speed
Stability
Solar radiation
Figure 13. Average daily patterns in Esthwaite Water May 11th-August 15th 2006 a) wind speed (m/s), solar radiation (W/m2) and stability (J/m2). b) deviation from midnight concentrations for the phytoplankton groups, cryptophytes, diatoms & dinoflagellates, green algae and blue-green algae.
The regularity of the daily pattern was affected by environmental conditions. When analysing the effect of solar radiation and wind speed, days between July 10th and August 10th were
(a) (b)
chl-
α (
ug/
L)
20
-2
-1
0
1
0 4 8 12 16 20 24
-1
-0.5
0
0.5
1
1.5
2
0 4 8 12 16 20 24
chosen. This period showed fairly high concentrations of all phytoplankton groups with likelihood to show a significant daily pattern. Thermal stratification was established and stable during this period making it easier to detect the strong fluctuation of the variables wind and solar radiation and their effect on diel variation. Daily patterns for 5-10 days were averaged to determine the differences in daily patterns. High daily wind speed (>3m/s)parallel to low (<100 W/m2) or high (>200 W/m2) solar radiation in combination with low daily wind speeds (<1.5 m/s) were analysed. Days with strong wind resulted in less diel variation in blue-green algae than days with high or low solar radiation and little wind (Fig.14b). Days with high solar radiation showed a smooth and more distinct trend. The diel variation of cryptophytes and diatoms and dinoflagellates showed highest daytime increase with low solar radiation compared to days with high wind and high solar radiation (Figs 14c,d). The green algae diel variation did not show any significant difference at varying weather conditions (Fig. 14a). The extended rain period in May had a strong effect on the dominant blue-green algae. The range of maximum and minimum blue-green algae showed a positive correlation with rainfall (r = 0.33, P = 0.004) and revealed an altered pattern with increased concentrations during daytime.
-3
-2
-1
0
1
0 4 8 12 16 20 24
-2
-1
0
1
2
3
4
0 4 8 12 16 20 24
Figure 14. Daily average Chl-a deviation from midnight concentrations of a) green algae, b) blue-green algae, c) cryptophytes, d) diatoms & dinoflagellates during 5 daysin Esthwaite Water between July 10th and August 15th 2006 during weather conditions of high wind speed (>3 m/s), high solar radiation (>200 W/m2) and low wind speed (<1.5 m/s), low solar radiation (<100 W/m2) and low wind speed.
a) b)
c) d)
high wind speed
low solarradition, lowwind speed
high solarradiation, lowwind speed
Hour Hour
Hour Hour
chl-
a (u
g/L
)
chl-
a (u
g/L
)
chl-
a (u
g/L
)
chl-
a (u
g/L
)
21
5 Discussion
This study showed that seasonal and episodic weather changes affect the phytoplankton community to a greater or lesser extent. On a seasonal scale the phytoplankton population biomass and diversity increased during the summer months with higher solar radiation and air temperature. Episodic events like rainfall and strong winds which increased mixing caused disturbances in the population by disfavouring certain species and favouring others. Daily changes in solar radiation and wind speed affected the phytoplankton community on a daily scale.
5.1 Fluoroprobe measurements
The measurements performed by the FP are at a shallow depth giving no information about the vertical distribution. This might be one factor contributing to the constantly lower Chl-aconcentration in the FP measurements compared to analysis of the 0-5 m depth profile. However, there are usually periods when the phytoplankton is more or less uniformlydistributed in the mixed layer. Lower Chl-a values measured by the FP compared with pigment analysis (HPLC) was also noted in Gregor et al. (2005), and was explained to high extent by the calibration method. In this study the calibration performed by bbe Moldaenke was applied. A different species composition than that of Esthwaite Water was likely used in the calibration and could be another reason for the discrepancy. According to biovolume estimates, blue-green algae decrease at the end of May and end of July was not as distinct as according to FP measurements. This indicates that descending vertical migration instead of population death may have been the cause of the decline. This would explain the negative correlation between solar radiation and blue-green algae. The FP measurements tend to favour non-motile species compared to migrating since they are more or less evenly distributed in the mixing layer. Although the FP and the biovolume estimates of the amount of green algae differed (Fig. 5), the proportion of the phytoplankton groups was similar and suggests that the FP is a valuable instrument of monitoring population dynamics. Estimations of phytoplankton biovolumes were rough since actual cell and colony size was not individually measured. This is an additional explanation for the differences in phytoplankton group proportion between the methods.
5.2 Correlations between phytoplankton and environmental variables
The strongest correlations between environmental variables and phytoplankton succession were with solar radiation, Schmidt stability and SRP. Extreme or extended conditions of rainfall and wind affected the phytoplankton in this study through secondary effects by disruption of the stability and deteriorating light conditions and occasionally nutrient levelsthat tend to cause shifts in phytoplankton composition and dynamics (Leitão et al. 2003, Vanni et al. 2006). Results from multiple regression and RDA showed analogous results forgreen algae and cryptophytes being favoured in similar conditions mainly during periods with high SRP and stability. Both blue-green algae and diatoms & dinoflagellates were prevailing in surface waters under conditions with rain, high wind and great mixing depth. Some factors did not express correlation in multiple regression compared with RDA results probably due to the elimination of interlinked factors. For example air temperature was strongly correlated with stability and wind speed with rainfall. Considering antecedent conditions up to 5 days,correlations were strengthened and variables included due to retaining effects of previous environmental conditions as suggested by Madgwick et al. (2006). RDA analysis showed
22
correlations between phytoplankton abundance and zeu, pH and SiO2 which likely werecaused indirectly through phytoplankton presence rather than actual community response. Positive correlation with pH can be explained by increased production at higher Chl-aconcentration (Maberly 1996). The negative correlation with zeu was likely caused by deteriorated light conditions at high biomass and particle concentration. The negative correlation with SiO2 was induced by utilisation of SiO2 by diatoms.
5.3 Phytoplankton group dynamics and daily patterns
The blue-green algae decrease at the beginning of July might be related to low SRP concentrations and migration to deeper more nutrient rich waters. The light conditions at the time were good. The euphotic depth was deeper than the mixing depth and the migration is likely since all requirements could be fulfilled at higher depths. Phosphorus-limitation seemsto induce blue-green algae migration to a higher degree than light limitation (Kromkamp & Walsby, 1999). Although the measured SRP concentrations are too infrequent to draw any definite conclusions, the measured concentration was low at the time of the decrease. The next decline occured close after in late July and during similar conditions but when SRP concentration was higher. With higher SRP concentrations diversity increased. A pulse of nutrient tends to greatly increase the diversity (Sommer 1985, Grover 1989) as well as induce a shift in species composition (Turpin & Harrison 1979, Sommer 1985, and Suttle et al.1987). With the SRP pulse the green algae and cryptophytes increased significantly, possibly inducing competition for light and nutrients. The newly appeared dominant blue-green algae species during the decline was the non buoyant Aphanothece clathrata that prefers nutrient-rich water columns whilst the buoyant species that have advantage at low nutrient levels (Reynolds 2006) may have migrated to greater depths or declined. Vertical niche separation is a solution to maintain diversity at limited conditions and often occur around the summer climax (Lindenschmidt & Chorus 1998) when nutrient levels are likely to be exhausted (Forsberg & Heyman 1984). Zeu at this time is deeper than the mixing depth which makes migration worthwhile in combination with less competition of nutrients at higher depth. Unfortunately algal counts were missing during this period and this distinct decline in blue-green algae being caused by migration and niche separation is only a hypothesis.
The migrating nature of the buoyant blue-green algae was principally noticed in the diel variation. Their buoyancy regulating property depends on irradiance and they tend to moveupwards at night and downwards during daytime, especially when light is high (Walsby, 1989). The pattern was most significant when solar radiation was high and wind speed low. This negative correlation with sunlight was seen both on a daily basis and in the long-term analysis. Most recent studies suggest that blue-green algae are abundant in summer time whenwater temperature and stability is high (Reynolds & Walsby 1975, Komarek 1985) due to their competitive ability to exploit light and nutrient gradients more effectively through vertical migration. However the conditions determining the success of blue-green algae arestill highly discussed. The results in this study showed high blue-green algae concentrationsthroughout most of the period with declines at periods with high long-term solar radiation andstability in surface water. However this study does not reveal the distribution at other depths than 1.1 m. Although the algal counts are limited the results suggests that the concentration is higher below 1.1 m depth. It seems like blue-green algae avoid the upper metre of the water column at high solar radiation with a remaining high concentration at greater depths.
The dominant peak of green algae was unexpected since the species in Esthwaite Water during summer maxima is usually dominated by buoyant blue-green algae or Ceratium sp.
23
according to Talling (1999) and Reynolds (2006). The green algae order Chlorococcales and blue-green algae are positively correlated to stability (Reynolds 1996) and are both more or less resistant to grazing by zooplankton by the presence of a thick cellulose cell wall, gelatinous envelope or colonial structure (Porter, 1973). The green algae bloom was highly correlated to the high stability in combination with high SRP conditions. A likely cause of the SRP enrichment was the high rainfall on the 10th of June. The limited frequency of SRP measurements makes it impossible to know at what day the rise occurred and whether it coincided with the rain. In some cases deepening of the mixing layer can cause entrainment of phosphorus to the upper water column and this may be one alternative or additional cause of the increase of SRP. In Esthwaite Water the entrainment of phosphorus from deeper water is of smaller importance than land runoff and the rainfall is a more likely cause for an increase of this nutrient (Miller, unpublished).
Former studies have shown that increased nutrient conditions often lead to dominance of green algae and blue-green algae (Scheffer et al. 1997). It has also been suggested that, in some circumstances, green algae can out compete blue-green algae due to their faster generation time (Jensen et al. 1994). Critical light intensity is an important factor involved. Phytoplankton species have different critical light intensities and at stable environmental conditions, species with the lowest critical light intensity will have an advantage (Tilman, 1982). In a study by Huisman et al. (1999) Chlorella vulgaris was found to be a better light competitor than Aphanizomenon flos-aquae through a suggested lower critical light intensity.At permanently high light, green algae prevailed and diatoms and bluegreen algae prevailed at permanent low light levels. These conclusions was supported by Flöder et al. (2002) and seen in the statistical analysis of this study. Fluctuating light conditions in the order of several days tends to sustain the phytoplankton diversity by preventing a condition of equilibrium and dominance by one species (Flöder & Burns 2005, Flöder et al. 2002).
At the start of the green algae bloom the former dominant colonial Coenochloris fottii, preferring clear waters and tolerant to low nutrients (Reynolds 2006), was substituted by a high biomass and diversity of green algae species and an increased proportion of single celled species. Small single celled species like Chlorella sp. were abundant at the beginning of the bloom due to their high requirements for nutrients (Reynolds 2006) and fast reproduction.Small sized species have a higher affinity for nutrient uptake (Laws 1975, Smith & Kalff 1983) and higher initial phosphorus uptake rates (Lehman & Sandgren 1982) than larger sized species. This makes small species more rapid in growth response to a phosphorous pulse. Spijkerman & Coesel (1998) showed that species with high initial phosphorus uptake rateshave a lower nutrient storage capacity than larger species with low initial uptake rate. This suggests that larger species with a high nutrient storage capacity would be successful when nutrient concentrations were high over a long period. A high nutrient pulse does not result in immediate cell division and all phytoplankton species have a lag period of varying length (Azad & Borchard 1970). Chlorella pyrenoidosa had a much shorter lag-phase than the larger green algae Cosmarium abbreviatum and support their short-term advantage to a nutrient pulse (Spijkerman & Coesel 1998). The diversity in green algae was higher in the beginning of the bloom but decreased by the end of the bloom when Cosmarium bioculatum and Paulschulzia tenera were completely dominant, possibly due to their higher storage capacityfor phosphorus. The green alga Cosmarium sp. is more tolerant of nutrient deficiency andprefer less stratified conditions than other green algae species (Reynolds 2006). At the end of the green algae bloom stability had started to decline in combination with decreased SRP, indicating a favourable condition for Cosmarium bioculatum. After the collapse of the green algae and cryptophyte bloom caused by a weather episode of low solar radiation, high wind
24
and rainfall decreasing the stratification remarkably, buoyant blue-green algae and diatoms were favoured once more.
Diatoms & dinoflagellates were generally favoured when solar radiation was low similar to blue-green algae. The overall dominant species Fragilaria crotonensis has no motility to migrate against or towards gradients of light or nutrient. The daily patterns of diatoms & dinoflagellates with increasing chl-a concentrations daytime can be correlated to higher wind speed during the daytime. With little water turbulence diatoms are at risk of sinking out of the mixed layer (Husiman & Sommeijer 2002) because they are denser than water. At night time when wind and turbulence is low, diatoms sink passively and the surface concentration is therefore likely to be less than daytime when the wind causes turbulence that intersperse the diatoms in the mixed layer. This positive correlation with wind and mixing was also seen on the long-term scale in the RDA analysis. Meroplanktonic diatoms, spending part of their life-time in the sediment, can be resuspended into the water column by wind and mixing in shallow lakes (Carrick et al. 1993). These events could have contributed to the higher concentrations of diatoms in Esthwaite Water during windy periods.
The species dominating the green algae group throughout the study period were non-motile and so might be expected to show a similar daily pattern to diatoms. In contrast, however, a distinct decline in chl-a was seen from early morning until afternoon which did not change significantly at different weather conditions. This may result from the known physiology of green algae. The green alga Chlorella pyrenoidosa showed a decrease in active photosystem centres due to relaxation, non-photochemical quenching, with increasing irradiance towards a midday maximum (Kroon 1994). This behaviour is developed to minimize the risk of photoinhibition and to maintain a balanced growth under fluctuating irradiation. The xanthophyll pigment cycle, present in green algae, is important for photoadaptation. The presence of xanthophyll contributes to a higher capacity to adapt to high irradiance by inducing non-photochemical quenching and reducing fluorescence in proportion to chl-a content (Frank et al. 1994, Masojidek et al. 1999). In this study decreased fluorescence resulting in lower estimates of green algae chl-a concentration may have been induced by this behaviour. It suggests that green algae have the strongest photoadaptation compared to the other groups and also that the behaviour is so strong on a daily basis that non-photochemical quenching occurs even at low irradiance.
In contrast the daily pattern during windy days indicated less diel increase in chl-a than days with low winds such as in the study by Frempong (1981). The relationship between wind and turbulence is not always instantaneous. The power of the wind produces turbulence whichoften is prolonged despite the wind abating. This makes it difficult to relate wind with daily patterns of phytoplankton in situ. The limited days with high daytime winds during the investigated period was not enough to state whether high winds had an instantaneous effect on the daily patterns.
The high peak in the cryptophyte Cryptomonas sp. was triggered by the SRP increase which explains the strong positive correlation with SRP. The zeu was low at the end of July and possibly another reason for the increase since Cryptomonas sp. has the characteristics of tolerating low light levels through an effective light harvesting complex, and their ability to grow at low irradiance levels (Palmisano et al. 1985). The diel pattern of the dominant species Cryptomonas sp. showed a significant difference between day and night concentrations with surface concentrations increasing during daytime and decreasing during night time. Similar results were found in a study by Gervais (1997) with a well known migration behaviour of a
25
regular vertical ascending in daytime and descent in the water column at night. This behaviour is believed to improve the light supply in daytime and at night time increase nutrient supply and reduce the grazing pressure. Highest upward migration in this study was seen when solar radiation was less than 100 W/m2. With high solar radiation and high wind the migration to the surface water was less. On a day with high solar radiation, migration might not be as strong as on a less sunny day. Further, on days with very little sun when desired light intensity not can be achieved, migration tend to be delayed or not occur at all (Knapp et al. 2003) With high winds the ability to migrate was suppressed. This may be because rates of active swimming may be lower than water movement. It is likely that the first small peak at the beginning of June was due to migration behaviour, and small competitionfor light and nutrients due to the low total phytoplankton biomass at the time. The low zooplankton number and low grazing pressure on small species is another factor explaining the increase (Salmaso & Naselli-Flores 1999)
6 Conclusions
These results indicate that phytoplankton species composition and diversity of phytoplanktonwere highly affected by the changing weather resulting in fluctuating water column conditions. This suggests that changes in weather conditions create dynamic phytoplankton populations and prevent a state of equilibrium. It also reveals the complexity of the system with interlinking variables that complicates predictions of the outcome in population dynamics. Phytoplankton succession was mainly affected by solar radiation, stability of the water column and SRP concentration. The SRP pulse in July had remarkable effects on the summer phytoplankton population and promoted green-algae and cryptophytes. Blue-green algae and diatoms were promoted with less stability and SRP. Deviations between FP measurements, and traditional methods for determination of phytoplankton composition and chl-a concentration, were observed. Fixed at a constant depth the FP missed out migrating populations, avoiding the surface water at certain conditions. Measurements at further depths would be useful for a better interpretation of phytoplankton dynamics. Higher resolution of SRP concentration and additional measurements of nitrate and ammonium would be useful for further studies.
26
Acknowledgements
Special thanks to my English supervisor Stephen Maberly, Centre of Ecology and Hydrology, Lancaster for all the help with practical and theoretical issues in the process of this project. Many thanks to Helen Miller for help with fieldwork and expertise on Esthwaite Water and Mitzi de Ville for performing fortnightly algal counts and chemical analysis. I also want to thank Ian Jones for all help with data treatment, Stephen Thackerey for statistical advice, nutrient analysis and comments to the report, Alex Elliott for phytoplankton expertise and comments to the report and the Freshwater Biological Association. Thanks to my Swedish supervisor Kurt Pettersson for help with the report. Finally I would like to thank my fiancé Ben for his endless support.
27
References
Azad H.S. & Borchardt J.A. 1970. Variations in phosphorous uptake by algae. Environ. Sci. Technol. 4:737-747.
Beutler M., Wiltshire K.H., Meyer B., Moldaenke C., Lüring C., Meyerhöfer M., Hansen U-P. & Dau H. 2002. Photosynthesis research 72:39-53.
Carrick H.J., Alridge F.J. & Schelske C.L. 1993. Wind influences phytoplankton biomass in a shallow, productive lake. Limnol. Oceanogr. 38:1179-1192.
Dinsdale, M.T.& Walsby, A.E. 1972. The interrelations of cell turgor pressure, gas vacuolation and buoyancy in blue-green alga. Journal of experimental Botany 23: 561-570.
Eppley R.W. , Holm-Hanson O. & Strickland J.D.H. 1968. Some observations on the vertical migration of dinoflagellates. Journal of Phycology 4:333-340.
Falkowski P.G., Kiefer D.A. 1985. Chlorophyll a fluorescence fluorescence in phytoplankton: relationship to photosynthesis and biomass. Journal of plankton research. 7:715-731.
Falkowski P.G., Humphries.S. & Long S.P., 1994. Photoinhibition of photosynthesis in nature. Annu.Rev.Plant.Physiol.Plant.Mol.Biol. 45: 633-662.
Frank H.A., Cua A., Chynwat V., Young A., Gosztola D & Wasielewski M.R. 1994. Photophysics of the carotenoids associated with the xanthophylls cycle in photosynthesis. Photosynthesis research 41:389-395.
Frempong E. 1981. Diel variation in the abundance, vertical distribution, and species composition of phytoplankton in a eutrophic English lake. Journal of Ecology 69: 919-939.
Forsberg C. & Heyman U. 1984. Variation of water quality parameters in meso- to hypereutrophic lakes. Verh. Int. Verein. Limnol 22:730-3.
Flöder S., Urabe J. & Kawabata Z-I. 2002. The influence of fluctuating light intensitites on species composition and diversity of natural phytoplankton communities. Oceologia 133:395-401.
Flöder S. & Burns C.W. 2005. The influence of fluctuating light on diversity and species number of nutrient-limited phytoplankton. Journal of Phycology. 41: 950-956.
Gaedke U., Seifried A. & Kummerlin R. 1998. The influence of weather conditions on the seasonal plankton development in a deep and large lake (Lake Constance). Management of Lakes and Reservoirs during Global Climate Change. Editor: George D.G et al. Kluwer Academic Publishers. Dordrecht
Gervais F. 1997. Diel vertical migration of Cryptomonas and Chromatium in the deep chlorophyll maximum of a eutrophic lake. Journal of Plankton research. 19: 533-550.
28
George D.G., Maberly S.C., Hewitt D.P. 2004. The influence of the North Atlantic Oscillation on the physical, chemical and biological characteristics of four lakes in the English Lake District. Freshwater Biology. 49:760 -774.
Gregor J., Geriš R., Maršálek B., Heteša J. & Marvan P. 2005. In situ quantification of phytoplankton in reservoirs using a submersible spectrofluorometer. Hydrobiologia 548:141-151.
Grover J. P. 1989. Effects of Si:P supply ratio, supply variability, and selective grazing in the plankton: An experiment with a natural algal and protistan assemblage. Limnol. Oceanogr. 34: 349-367.
Heaney S.I., Smyly W.J.P., Talling J.F. 1986. Interactions of physical, chemical and biological processes in depth and time within a productive English lake during summer stratification. Internationale Revue der gesamten Hydrobiologie 71: 441-494.
Huisman J. & Weissing F.J. 1995. Competition for nutrients and light in a mixed water column: A theoretical analysis. Am. Nat. 146: 536-564.
Huisman J. & Sommeijer B. 2002. Maximal sustainable sinking velocity of phytoplankton. Marine Ecology Progress Series. 244:39-48.
Huisman J., Jonker R.R., Zonneveld C. & Weissing F.J. 1999. Competition for light between phytoplankton species: experimental tests of mechanistic theory. Ecology 80:211-222.
Hutchinson, G. E. 1957. A Treatise on Limnology, John Wiley & Sons, Inc, New York.
Jensen J.P., Jeppesen E., Olrik K. & Kristensen P. 1994. Impact of nutrients and physical factors on the shift from cyanobacterial to chlorophyte dominance in shallow Danish lakes. Canadian Journal of Fisheries and Aquatic Sciences, 51:1692–1699.
Kirk, J.T.O. 1994. Light and Photosynthesis in Aquatic Ecosystems. 2nd edition. University press, Cambridge.
Klemer A.R. 1991. Effects of nutritional status on cyanobacterial buoyancy, blooms, and dominance, with special reference to inorganic carbon. Can. J. Bot. 69:1133-1138.
Knapp C.W., deNoyelles F.Jr., Graham D.W. & Bergin S. 2003. Physical and chemical conditions surrounding the diurnal migration of Cryptomonas spp. (Crypophyceae) in a seasonally stratified Midwestern reservoir. J. Phycol. 39:855-861.
Komárek J. 1985. Do all cyanophytes have a cosmopolitan distribution? Survey of the freshwater cyanophyte flora of Cuba. Archiv für Hydrobiologie, Supplements – Algological Studies, 38/39, 359–386.
Kromkamp J. & Walsby A.E. 1999. A computer model of buoyancy and vertical migration in cyanobacteria. Journal of plankton research. 12:161-183.
29
Kromkamp J.C & Forster R.M. 2003. The use of variable measurements in aquatic ecosystems: differences between multiple and single turnover measuring protocols and suggested terminology. European Journal of Phycology. 38:103-112.
Kroon B.M.A. 1994. Variability of photosystem II quantum yield and related processes in Chlorella pyrenoidosa (Chlorophyta) acclimated to an oscillating light regime simulating a mixed photic zone. J. Phycol. 30: 841-852.
Laws E.A. 1975. The importance of respiration losses in controlling size-distribution of marine phytoplankton. Ecology 56:419-426.
Lehman J.T. & Sandgren C.D. 1982. Phosphorus dynamics of the prokaryotic nannoplankton of a Michigan Lake. Limnol. Oceanogr. 27:828-838.
Leitão M., Morata S.M., Rodriguez S. & Vergon J.P. 2003. The effect of perturbations on phytoplankton assemblages in a deep reservoir (Vouglans, France). Hydrobiologia 502:73-83.
Lindenschmidt K.E. & Chorus I. 1998. The effect of water column mixing on phytoplankton succession, diversity and similarity. Journal of Plankton Research 20:1927-1951.
Long S.P., Humphries S. & Falkowski P.G. 1994. Photoinhibition of Photosynthesis in Nature. Annual Review of Plant Physiology and Plant Molecular Biology. 35:633-662.
Lund J.W.G., Kipling C. & Le Cren E.D. 1958. The inverted microscope method of estimating algal numbers and the statistical basis of estimation by counting. Hydrobiologia, 11:143-170.
Lund J.W.G. 1964. Primary production and periodicity of phytoplankton Verhandlungen Int. Ver.Limnol. 15:37-56.
Maberly S.C. 1996. Diel, episodic and seasonal changes in pH and concentrations of inorganic carbon in a productive lake. Freshwater Biology. 35:579-598.
Maberly S.C., King L., Dent M.M., Jones R.I. & Gibson C.E. 2002 Nutrient limitation of phytoplankton and periphyton growth in upland lakes. Freshwater Biology 47: 2136-2152.
Mackereth F.J.H., Heron J. & Talling J.F. 1978. Water Analysis:some revised methods for limnologists. Scientific Publications of the Freshwater Biological Association No.36.
Madgwick G., Jones I.D., Thackerey S.J., Elliott J.A. & Miller H.J. 2006 Phytoplankton communities and antecedent conditions: high resolution sampling in Esthwaite water. Freshwater Biology 51:1798-1810.
Masojidek J., Torzillo G., Kobližek M., Kopecky J., Bernardini P., Sacchi A & Komenda J. 1999. Photoadaptation of two members of the Chlorophyta (Scenedesmus and Chlorella) in laboratory and outdoor cultures: changes in chlorophyll fluorescence quenching and the xanthophylls cycle. Planta 209:126-135.
Nunez, M., Davies, J. A. & Robinson, P. J. 1972. Surface albedo at a tower site in Lake Ontario. Boundary-Layer Meteorology: 3:77-86.
30
Palmisano A.C., SooHoo J.B., Sullivan C.W. 1985. Effects of four environmental variables on photosynthesis-irradiance relationships in Antarctic sea-ice microalgae. Marine Biology 94:299-306.
Porter K.G. 1973. Selective grazing and differential digestion of algae by zooplankton. Nature 244:179-180.
Ramsbottom, A. E. 1976. Depth Charts of the Cumbrian Lakes, FBA, Ambleside
Rao C.R. 1964. The use and interpretation of principal component analysis in applied research. Sankhya A. 26:329-358.
Reynolds C.S. & Walsby A.E.1975. Water blooms. Biol.Rev. 50:437-481
Reynolds C. 1984. The Ecology of Freshwater Phytoplankton. Cambridge University Press, Cambridge.
Reynolds C.S., Oliver R.I. & Walsby, A.E. 1987. Cyanobacterial dominance:the role of buoyancy regulation in dynamic lake environments. Zealand Journal of Marine and Freshwater research 21: 379-399.
Reynolds C.S & Lund, J.W.G. 1988. The phytoplankton of an enriched, soft-water lake subject to intermittent hydraulic flushing (Grasmere, English Lake District). Freshwater Biology, 47:1183-1188.
Reynolds C.S. 1996. The plant life of the pelagic. Verh. Ver. Limnol. 26:97-113.
Reynolds C.S. 2006. Ecology of phytoplankton. 1st edition. University Press, Cambridge.
Rouen, M. A., George, D. G. & Hewitt, D. P. 2000. Using an automatic monitoring station to assess the impact of episodic mixing on the seasonal succession of phytoplankton. Verhandlungen der Internationalen Verienigung für Theoretische und Angewandte Limnologie 27:2972-2976.
Salmaso N. & Naselli-Flores L. 1999. Studies on zooplankton of the deep subalpine Lake Grada. J. Limnol. 58:66-76.
Salmaso N. 2003. Life strategies dominance patterns and mechanisms promoting species coexistence in phytoplankton communities along complex environmental gradients. Hydrobiologia, 502: 13-36.
Scheffer M., Rinaldi S., Gragnani A., Mur L. & van Nes E.H. 1997. On the dominance of filamentous cyanobacteria in shallow, turbid lakes. Ecology 78:272–282.
See J.H., Campbell L., Richardson T.L., Pinckney J.L. & Rongjun Shen. 2005. Combining new technologies for determination of phytoplankton community structure in the northern gulf of Mexico. J. Phycol. 41:305-310.
31
Smith R.E.H. & Kalff J. 1982. Size-dependent phosporous uptake kinetics and cell quota in phytoplankton. Journal of phycology 18:275-284.
Sommer U. 1985. Nutrient competition between phytoplankton species in multispecies chemostat experiments. Arch. Hydrobiologie 96:399-416.
Sousa W.P. 1984. The role of disturbance in natural communitites. Ann. Rev. Ecol. Syst. 15:353-391.
Spijkerman E. & Coesel F.M. 1998. Different response mechanisms of two planctonic desmid species (chlorophyceae) to a single saturating addition of phosphate. Journal of phycology 34:438-445.
Suttle C.A., Stockner J.G. & Harrison P.J. 1987. Effects of nutrient pulses on community structure and cell size of a freshwater phytoplankton assemblage in culture. Can. J. Fish. Aquat. Sci. 44:1768-1774.
Talling, J. F. 1974 Photosynthetic pigments. General outline of spetrophotometric methods; specific procedures. In A Manual on Methods for Measuring Primary Productivity in Aquatic Environments (Ed, Vollenweider, R. A.) Blackwell Scientific Publications, Oxford, pp. 22-26.
Talling J.F. 1981. The development of attenuance depth profiling to follow the changing distribution of phytoplankton and other particulate material in a productive English lake. Archiv für Hydrobiologie 93:1-20.
Talling J.F., Heaney S.I. 1984. Evaluation of historical data on the nutrient status on Esthwaite water, Cumbria. Freshwater Biological Association. The Ferry House, Ambleside, Cumbria
Talling J.F. 1999. Some English Lakes as diverse and active ecosystems: a factual summary book and source book. Freshwater Biological Association, Ambleside. No. 6.
Talling J.F. 2004. Interrelation of diel and seasonal change, involving summer thermal stratification, weather variables and a mobile dinoflagellate in a productive English lake. Hydrobiologia 524: 215-227.
Ter Braak, C. J. F. 1986. Canonical Correspondence Analysis: a new eigenvector technique for multivariate direct gradient analysis. Ecology 67:1167-1179.
Tilman D. 1982. Resource competition and Community structure. Princeton University Press, Princeton.
Turpin D.H. & Harrison P.J. 1979. Limiting nutrient patchiness and its role in phytoplankton ecology. J.Exp. Mar. Biol. Ecol. 39:151-166.
Van den Hoek C., Mann D.G. & Jahns H.M. 1995. Algae : an introduction to phycology. 1st
edition. Cambridge University Press.
32
Vanni M.J., Andrews J.S., Renwick W.H., Gonzalez M.J. & Noble S.J. 2006. Nutrient and light limitation of reservoir phytoplankton in relation to storm-mediated pulses in stream discharge. Arch. Hydrobiol. 167:421-445.
Walsby A.E., Reynolds C.S., Oliver R.L. & Kromkamp J. 1989. The role of gas-vacuoles and carbohydrate content in the buoyancy and vertical distribution of Anabaena minutissima in lake Rotongaio, New Zeeland. Arch. Hydrobiol. Beih. Ergebn. Limnol. 32:1-25.
Wetzel R.G. 2001. Limnology, Lake and River Ecosystems. 3rd edition. Academic press. San Diego
Bbe Moldaenke http://www.bbe-moldaenke.de. 2006-07-15. 14:30