Eutrophic Lake Kasumigaura Research Articles

Seasonal variability and regulation of bacterial production in a shallow eutrophic lake

ABSTRACTThe seasonal variation of bacterial production (BP) in a shallow, eutrophic Lake Kasumigaura was clarified from 2012 to 2016. During the studied period, BP fluctuated from 1.9 to 138 μg C L−1 d−1. There were no significant correlations between BP and bacterial abundance in any season, suggesting a strong top‐down regulation on BP throughout the year. On the other hand, BP was also related to bottom‐up regulation factors such as water temperature, phosphorus, and primary production (PP) annually. During winter, BP was positively correlated with chlorophyll a concentration, suggesting that autochthonous substrates were relatively important for BP. Moreover, BP was positively correlated with heterotrophic nanoflagellates, ciliates, and copepods, suggesting higher availability of substrates for BP. In summer, although there was no significant correlation between BP and PP, rainfall amount showed significant negative correlations with both BP and PP, suggesting depressed PP from relatively lower solar irradiance coupled with unfavorable weather conditions that decreased the substrate supply for bacteria. These results suggest that temporal variation of BP was regulated not by allochthonous, but by autochthonous substrates during both the highest (summer) and lowest (winter) productive seasons, even in a shallow, eutrophic lake. PP in autumn was approximately half that of spring due to lower solar irradiance, although water temperatures during both seasons were similar and nutrient concentrations during autumn were higher. On the other hand, BP in autumn was comparable with that in spring, and the bacterial carbon demand (= BP + bacterial respiration; 1.12 ± 0.79 g C m−2 d−1) was comparable to PP (1.16 ± 0.53 g C m−2 d−1), suggesting the relative importance of higher allochthonous substrates relative to other seasons.

Dynamics of particulate phosphorus in a shallow eutrophic lake

We tested the hypothesis that in shallow, eutrophic Lake Kasumigaura, the concentration of particulate phosphorus (PP) is controlled by biogenic P (P in living or dead phytoplankton and bacterial cells), rather than by resuspension of inorganic P in sediment. Increases in wind velocity and turbidity were associated with bottom shear stress exceeding the critical value for the lake (τc=0.15Nm−2); this increased turbidity was due to sediment resuspension. However, concentrations of PP; HCl-extractable, reactive P in PP (P-rP); and HCl-extractable, non-reactive P in PP (P-nrP) were not correlated with wind velocity (PP vs. wind velocity: r=0.40, p>0.05). Rather, the P-nrP concentration accounted for approximately 79% of PP, and the concentrations of PP, P-rP, and P-nrP were correlated with the particulate organic carbon (POC) concentration (POC vs. PP: r=0.90, p<0.01; POC vs. P-rP: r=0.82, p<0.01; POC vs. P-nrP: r=0.86, p<0.01). In our 31P nuclear magnetic resonance spectroscopy results, mononucleotides accounted for the largest proportion among the detected P compound classes. In addition, concentrations of mononucleotides, orthophosphate, and pyrophosphate were significantly higher in samples with high POC concentrations, whereas the DNA-P concentration was not. These results suggest that biogenic P affects PP concentrations more strongly than does sediment resuspension, and the production of biogenic P creates a pool of mononucleotides, a class of easily degradable P, even in shallow, eutrophic Lake Kasumigaura.

Impacts of long-term increase in silicon concentration on diatom blooms in Lake Kasumigaura, Japan

In the eutrophic Lake Kasumigaura in Japan, a trend of dissolved Si (DSi) concentration was detected over the last three decades, probably caused by the DSi release enhanced by an increase in sediment resuspension for the same period (Arai et al., Limnol., 13, 81–95, 2012). The present study described the long-term trends of the magnitude and seasonality of diatom blooms in the lake during 1981–2010 using the database and assessed the influencing factors for the trends by the numerical simulation of DSi and diatoms. The box model was developed based on the lake budgets (inflow, outflow, release and sedimentation) and the simple diatom growth model depending on DSi, temperature and light condition. As results, database analysis detected a long-term trend of increasing diatom abundance and a shift of blooming season from spring and autumn to the winter–spring period. Si could be regarded as a main nutrient factor limiting diatom growth by analyzing N:P:Si ratios. Our model simulation relatively well-reproduced the increasing trend and the shift of seasonality of DSi and diatoms, even though peaks of diatom blooms were underestimated in some years. Among input variables, the concentration of resuspended sediments radically increased. The model simulation with the input variables or parameters changed suggested as follows: (1) the recent DSi release from resuspended sediments enhanced diatom abundance and (2) the degradation of light condition caused by resuspension affected the shift of blooming season. These findings implicate the significance of the interactions between sediments and water to phytoplankton blooms.

Bacterial Contribution to Dissolved Organic Matter in Eutrophic Lake Kasumigaura, Japan

Incubation experiments using filtered waters from Lake Kasumigaura were conducted to examine bacterial contribution to a dissolved organic carbon (DOC) pool. Bacterial abundance, bacterial production, concentrations of DOC, total dissolved amino acids (TDAA), and total dissolved neutral sugars (TDNS) were monitored during the experiments. Bacterial production during the first few days was very high (20 to 35 μg C liter(-1) day(-1)), accounting for 40 to 70% of primary production. The total bacterial production accounted for 34 to 55% of the DOC loss during the experiment, indicating high bacterial activities in Lake Kasumigaura. The DOC degradation was only 12 to 15%, whereas the degradation of TDAA and TDNS ranged from 30 to 50%, suggesting the preferential usage of TDAA and TDNS. The contribution of bacterially derived carbon to a DOC pool in Lake Kasumigaura was estimated using d-amino acids as bacterial biomarkers and accounted for 30 to 50% of the lake DOC. These values were much higher than those estimated for the open ocean (20 to 30%). The ratio of bacterially derived carbon to bulk carbon increased slightly with time, suggesting that the bacterially derived carbon is more resistant to microbial degradation than bulk carbon. This is the first study to estimate the bacterial contribution to a DOC pool in freshwater environments. These results indicate that bacteria play even more important roles in carbon cycles in freshwater environments than in open oceans and also suggests that recent increases in recalcitrant DOC in various lakes could be attributed to bacterially derived carbon. The potential differences in bacterial contributions to dissolved organic matter (DOM) between freshwater and marine environments are discussed.

Silicon budget of eutrophic Lake Kasumigaura, Japan

Determining the dynamics of silicon in lakes, one of the essential nutrients for diatoms, is valuable for understanding aquatic environmental problems. The dissolved silicon (DSi) and biogenic silicon (BSi) budgets in Lake Kasumigaura, a shallow eutrophic lake in Japan, during the last three decades were assessed based on the analysis of dated sediment cores and a water quality database. Sediment cores (100 cm long) were taken at the center of Lake Kasumigaura in 2005, 2007, and 2009 and at two other sites in 2007. BSi contents of the dated sediments were determined by wet alkaline digestion. The net sedimentation rates of BSi were defined as the difference between the DSi load from inflowing rivers and the DSi and BSi loads from the outflow of the lake, calculated using DSi concentrations and diatom abundances in the lake from 1980 to 2007 and DSi concentrations of the inflowing rivers during 1994, 2007, and 2009. The gross sedimentation rates of BSi were estimated by multiplying BSi concentrations in lake water by the diatom sinking rate reported by previous studies. Budgetary calculations based on the database showed that 60–70 % of DSi inputs from the inflowing rivers during the 27 years could ultimately be accumulated as diatom frustules in bottom sediments in the lake. The sediment analysis revealed that the amount of BSi accumulated in the lake from 1980 to 2007 was 2.0–2.6 × 1011 g, similar to the amount based on the database of 1.3–2.4 × 1011 g. Although the gross sedimentation rates of BSi likely increased, the net sedimentation rates of BSi decreased significantly from 6–10 × 109 g year−1 in the 1980s to 2–6 × 109 g year−1 in the 2000s, suggesting a fast recycling of BSi in recent years caused by an increase in sediment resuspension and regeneration. The sediment core information and the water quality database can be used for calculating the long-term silicon budgets in Lake Kasumigaura. An increase in the DSi release rates was identified, which is consistent with recent sediment resuspension. Comparing the sediment core information with the database suggests the long-term dissolution of sediment BSi; however, analysis of the BSi content in sediment cores representing a much longer time period is needed to confirm this.

Biogenic Phosphorus Compounds in Sediment and Suspended Particles in a Shallow Eutrophic Lake: A 31P-Nuclear Magnetic Resonance (31P NMR) Study

Differences in biogenic phosphorus (P) compounds between sediment and suspended particles in aquatic environments are important for understanding the mechanisms of internal P loading, but these differences are still unknown. We used solution-state (31)P-nuclear magnetic resonance spectroscopy ((31)P NMR) with NaOH-ethylenediaminetetraacetic extraction to detect the multiple P compounds in suspended particles and sediment in the eutrophic Lake Kasumigaura, including orthophosphate monoesters, orthophosphate diesters, pyrophosphate, and polyphosphate. We tested the hypothesis that there is a significant difference between these groups in suspended particles and sediment. Biogenic P other than orthophosphate was found in significantly higher proportions in suspended particles (74.3% of total P) than in sediment (25.6%). Orthophosphate monoesters were comparatively more abundant in suspended particles, as indicated by the ratio of orthophosphate diesters to monoesters (average, 0.31 for suspended particles; 1.05 for sediment). The compounds identified as orthophosphate monoesters by (31)P NMR spectroscopy originated mainly from phospholipids (α-glycerophosphate and β-glycerophosphate) and ribonucleic acid (RNA-P), whereas the orthophosphate diesters included mostly DNA (DNA-P). These results suggest that the dynamics of orthophosphate diesters, the production of DNA-P, or the degradation of phospholipids, play an important role in P cycling in Lake Kasumigaura.

Modeling mitigation strategies for toxic cyanobacteria blooms in shallow and eutrophic Lake Kasumigaura, Japan

This paper explores mitigation scenarios for toxic cyanobacteria blooms in Lake Kasumigaura is located about 60–90 km northeast of Tokyo, in the southeast of Ibaraki Prefecture, Japan. Dominant species of cyanobacteria (Microcystis aeruginosa, Microcystis viridis and Microcystis ichthyoblabe) produce highly potent toxins as (Microcystin-(Leucine + Arginine, MC-LR), Microcystin-(Arginine + Arginine, MC-RR) and Microcystin-(Tyrosine + Arginine, MC-YR) in the Lake. Toxin production is generally the result of two major factors-natural processes and human interferences. Both factors have an extreme influence on the generation of cyanobacteria toxins within lake ecosystems. To address these factors, we propose two concepts for mitigation. The first concept is intended for examining the natural process of toxin production behavior within the lake and the second concept is used for evaluating inflow of wastes and nutrients from human activities that form toxins. Our research aims to combine both strategies to mitigate impacts of toxins, by examining trends of cyanobacteria nutrient sources, buoyancy regulations that influence bloom formation, and the environmental conditions that spur blooms. This study proposes a simple IMPACT (Integrating Mitigation Policies for Aquatic Cyanobacteria Toxin) model for diminution strategies of harmful algal blooms and their toxins. A mixed-methods approach is employed, nested within the Environmental Systems Analysis (ESA) tools e.g. scenario analysis and stakeholder analysis. The quality of the lake is assessed through a combination of observation and field study analysis. The findings suggest that successful mitigation of cyanobacteria toxins is highly dependent on multi-functional, multi-stakeholder involvement, and relevant intergovernmental policy. Without integrating approaches among different stakeholders, diverse socioeconomic activists, local-national policymakers and effective policy measures, prevention of cyanobacteria toxin production within lakes becomes extremely complex and difficult. The proposed IMPACT model could be a decision framework for identifying suitable policies that mitigate cyanobacteria impacts.

Speciation of Aluminum in Lake Kasumigaura, Japan

Speciation of aluminum (Al), including inorganic monomeric Al (Ali, the sum of aquo, hydroxy, and inorganically complexed forms), organic monomeric Al (Alo, the organically complexed form), and colloidal mineral Al (Alc, the fine particulate form that passes through a 0.4-μm pore size membrane filter), was investigated in circumneutral (pH 7–9) eutrophic Lake Kasumigaura, Japan. The mean concentrations calculated from the result of monthly observations during the period from January to December 2003 were compared among ten stations. Ali decreased downstream from the mouth of the influent rivers (0.07–0.10 μM) to the mouth of the effluent river (0.03 μM). Alc, which also decreased downstream (from 0.11 to 0.06 μM), was the major fraction (>50%) of total dissolved Al at all stations. Alo remained around 0.02 μM at all stations. In Lake Kasumigaura, a high concentration of Si (average 0.23 mM) seems to be involved in the removal of Ali and formation of Alc. Also, a high concentration of Ca (ave. 0.43 mM) seems to be acting as a competitive cation for complexation with a high concentration of dissolved organic carbon (ave. 0.26 mM), and inhibit the formation of Alo.

富栄養化した霞ケ浦の堆積物に養殖の痕跡を残すリンの形態と分布

富栄養化した霞ケ浦の堆積物に養殖の痕跡を残すリンの形態と分布

Missing role of groundwater in water and nutrient cycles in the shallow eutrophic lake Kasumigaura, Japan

AbstractA process‐based model was developed, NICE‐LAKE (NIES (National Institute for Environmental Studies) Integrated Catchment‐based Ecohydrology), which includes interactions between surface water, canopy, unsaturated water, aquifer, lake and rivers, and used it to model the shallow eutrophic Lake Kasumigaura in Japan. By estimating the spatial distribution of the hydrological cycle, the model shows that groundwater withdrawal greatly affects groundwater distribution and seepage and indirectly influences lake water level. The simulated seepage agrees excellently with the budget‐derived value calculated from the observed groundwater level, lake level and isotope analyses. The model showed that groundwater seepage and groundwater concentrations are important contributors to the eutrophication of Lake Kasumigaura, an important contribution not recognized in previous studies of the lake. Groundwater entering the lake from the north side is contaminated with high concentrations of nitrate and ammonia from intensive pig and cattle raising and cultivated fields. The simulation showed that this high nitrogen load plays an important role in the eutrophication of the lake (the nitrogen load in inflowing groundwater is 30% of river inflow and 4 times that from wastewater treatment plants) in spite of government policies to prevent overland flow of nutrients into the lake. Our results show that NICE‐LAKE is a powerful tool for forecasting how the water quality of the lake will be affected by the (illegal) disposal of animal excreta in the surrounding open fields. Copyright © 2007 John Wiley &amp; Sons, Ltd.

Effect of iron complexation with dissolved organic matter on the growth of cyanobacteria in a eutrophic lake

Dissolved organic matter (DOM) in lake water can have a harmful effect on algal growth through either direct toxicity or iron limitation induced by its complexation with iron. We assessed the effect of iron complexation with DOM on algal growth in the eutrophic Lake Kasumigaura, Japan, using 2 species of cyanobacteria, Microcystis aeruginosa and Planktothrix agardhii, by combining the algal growth potential (AGP) test and pretreatment of DOM decomposition by UV irradiation. We also determined which nutrient limited the growth of the 2 species in the lake water using the AGP test: iron limited the growth of the 2 species, as did nitrogen and phosphorus in Lake Kasumigaura. Moreover, the growth of the 2 species was inhibited by iron complexation with DOM. This is consis- tent with our results regarding iron speciation in Lake Kasumigaura by cathodic stripping voltam- metry: most (>99.9%) of the dissolved iron was present as organic species in the lake water. Further- more, data from our AGP test suggest that iron requirement or iron availability between the 2 species are different. This difference in growth characteristics between algal species would be an important determinant of the dominance of specific algal species.

Trihalomethane formation potential of dissolved organic matter in a shallow eutrophic lake

Dissolved organic matter (DOM) in water samples from the shallow eutrophic Lake Kasumigaura, the second largest lake in Japan, was fractionated and characterized by using resin adsorbents into 5 classes: aquatic humic substances (AHS), hydrophobic neutrals, hydrophilic acids (HiA), bases (BaS) and hydrophilic neutrals (HiN). Subsequently, the trihalomethane formation potential (THMFP), ultraviolet absorbance to dissolved organic carbon (UV:DOC) ratio, and molecular size distribution of the DOM, AHS and hydrophilic fractions (HiF) (HiF=HiA+BaS+HiN) were examined. The THMFP of HiF, normalized on a DOC basis, was found to be comparable to that of AHS (0.176 μmol THM mg C −1 vs. 0.195 μmol THM mg C −1, respectively). The importance of HiF over AHS as a THM precursor became more pronounced when THMFP was evaluated in terms of concentration. In this case, the THMFP of HiF was much greater than that of AHS (0.374 μmol THM l −1 vs. 0.229 μmol THM l −1, respectively). Molecular size distributions all exhibited a narrow size range and relatively low molecular weights. The weight-averaged molecular weights of DOM, AHS and HiF were 780, 957 and 606 g mol −1, respectively.

Fractionation and characterization of dissolved organic matter in a shallow eutrophic lake, its inflowing rivers, and other organic matter sources

Dissolved organic matter (DOM) in water from eutrophic Lake Kasumigaura, its inflowing rivers, and several other DOM sources in the lake catchment area was fractionated using resin adsorbents into five classes: aquatic humic substances (AHS), hydrophobic neutrals (HoN), hydrophilic acids (HiA), bases (BaS), and hydrophilic neutrals (HiN). The DOM-fraction distribution pattern and the ultraviolet absorbance to dissolved organic carbon ratio (UV/DOC ratio) were found remarkably effective for evaluating the characteristics of DOM in water. DOM-fraction distribution patterns were significantly different depending on the origin of the sample. AHS and HiA were found to be the dominant fractions in DOM in all samples studied. HiA prevailed over AHS in the lake water, whereas AHS were slightly more abundant than HiA in the river waters. AHS were in the great majority in forest streams and plowed-field percolates. HiA abounded in paddy-field outflow, domestic sewage, and sewage-treatment-plant effluent. Only domestic sewage contained a significant amount of HoN. The UV/DOC ratio also varied depending on the origin of the sample; the ratios in total DOM, AHS, and HiA were greater in river waters than in the lake water. The greatest ratio of AHS was found in paddy-field outflow and the lowest in domestic sewage. The UV/DOC ratios in the sewage-treatment-plant effluent were very similar to those in the lake water.

Causative microorganisms for musty odor occurrence in the eutrophic Lake Kasumigaura

The musty odor phenomenon in Lake Kasumigaura, Japan (a eutrophic lake that is used for various purposes, including water supply source) was examined. The causative microorganisms responsible for the production of the odor compounds, geosmin and/or 2-methylisoborneol (MIB), and the odor release mechanism were studied in vitro and in situ. The numbers of the filamentous P. tenue in the lake water column were closely correlated with the concentration of MIB, but not geosmin. The isolated monoxenic P. tenue in culture produced only MIB. Geosmin concentration was well correlated with the number of actinomycetes in the sediment. Forty isolates of actinomycetes from the sediment around the water supply intake tower produced both geosmin and MIB in culture. Furthermore, the average ratio of production of geosmin to MIB was 1.40–1.0. Actinomycetes in the sediment played an important role in geosmin production. We concluded that geosmin and MIB from actinomycetes in the sediment, and MIB from P. tenue in the water column were jointly responsible for the severe musty odor occurrence of the waters supply source at Lake Kasumigaura.

Effects of iron limitation and aquatic humic substances on the growth ofMicrocystis aeruginosa

Aquatic humic substances (AHSs) from the eutrophic Lake Kasumigaura, Japan, significantly inhibited the growth of Microcystis aeruginosa at a concentration level similar to that found in the lake. AHSs were isolated and quantified using XAD-8 resin. The effects of these isolated AHSs on the growth of M. aeruginosa cultures were evaluated in a chemically defined medium. The production of siderophores, iron-specific complexing agents, of M. aeruginosa under iron limitation was also examined. Fulvic acid (the lower molecular size fraction of the AHSs) accounted for more than 85% of AHSs in Lake Kasumigaura. Under iron limitation, M. aeruginosa produced hydroxamate-type siderophores; however, they were not effective in enabling substantial recovery of growth. Microcystis aeruginosa growth was significantly inhibited when fulvic acid was 2 mg dry weight·L-1, similar to the concentration in the lake, even when siderophores were produced. Growth inhibition was probably due to a deficiency of iron caused by iron complexation with the fulvic acid. It was suggested that AHSs play an important role in formation and disappearance of M. aeruginosa blooms in eutrophic lakes.

Application of an in situ gradostat for a natural phytoplankton community in a eutrophic environment

The in situ gradostat ( Higashi et al., 1998) was applied to study the effect of oligotrophication on a eutrophic phytoplankton community. The in situ gradostat was launched in a eutrophic experimental pond (2360 ± 166 μg NO 3–N litre −1). Distilled water was pumped into the central (first) chamber and diffused radially outwards into the second and third chambers through the perforated walls, so as to achieve a nutrient gradient of every limnological trophic level from the first to the third chambers within 150 h. With this apparatus, the impact of a nitrate gradient on a natural phytoplankton community in eutrophic Lake Kasumigaura was determined. The mean growth rate of a eutrophic phytoplankton species Oscillatoria tenuis, which often dominates eutrophic environment such as this, was the lowest (−0.3 ± 0.5 day −1 at 209 ± 80 μg NO 3–N litre −1) in the oligotrophic chamber, and intermediate in the mesotrophic chamber (−0.3 ± 0.4 day −1 at 679 ± 130 μg NO 3–N litre −1) and higher in the eutrophic chamber (0.7 ± 0.9 day −1 at 2018 ± 212 μg NO 3–N litre −1). In contrast, the mean growth rate of a non-eutrophic phytoplankton species, Synedra ulna, was higher in the experimentally oligotrophic environment (0.4 ± 0.4 day −1) than that in the mesotrophic environment (−0.7 ± 0.5 day −1).

溶存フミン物質の藍藻Ｍｉｃｒｏｃｙｓｔｉｓ ａｅｒｕｇｉｎｏｓａの増殖に及ぼす影響

Aquatic humic substances (AHSs) from the eutrophic Lake Kasumigaura were isolated using XAD-8 resin. The effects of these isolated AHSs on the growth of Microcystis aeruginosa cultures were evaluated. M. aeruginosa was isolated from algal blooms in Lake Kasumigaura and grown in a chmically defined medium under iron limitation. The production of siderophores, iron-specific complexing agents, by M. aeruginosa was also examined. Under iron limitation, M. aeruginosa was confirmed to produce hydroxamate-type siderophores; however, they were not effective to enable substantial recovery of the growth rate. The fulvic acid (lower molecular-size fraction of AHSs) isolated from Lake Kasumigaura was found to significantly inhibit the growth of M. aeruginosa. The fulvic acid at 2 mg dry weight l-1, the same level as in the lake, significantely inhibited the growth of M. aeruginosa even though siderophores were produced. This inhibition was probably due to a deficiency of iron caused by iron complexation with the fulvic acid. These results suggest that AHSs, through their iron complexation reactions, may play an important role in formation of M. aeruginosa algal blooms.

Novel application of a back-propagation artificial neural network model formulated to predict algal bloom

This paper describes the novel application of an artificial neural network (ANN) model based on the back-propagation method formulated to predict algal bloom by simulating the future growth of five phytoplankton species and the chlorophyll a concentration in the second largest lake in Japan: eutrophic freshwater Lake Kasumigaura. Comparison of observed and calculated values showed that (i) seasonal variations in the biomass of Microcystis spp. were well-predicted with respect to the timing and magnitude of algal bloom, and (ii) the concentration of chlorophyll a, as an indicator of the total biomass of phytoplankton, was well predicted in general. The resultant correlations for the other species, however, showed that model learning was insufficient to effectively predict species biomass; thereby indicating that some unknown factors which are not represented by the set of water quality parameters used as model input data affect phytoplankton growth. A sensitivity analysis performed on input parameters showed that chlorophyll a concentration was mainly affected by PO4-P concentration, while cyanobacteria and diatom species were affected by NO3-N and NH4-N concentrations, respectively. These results indicate that the “algal bloom” ANN model achieved reasonable effectiveness with respect to learning the relationship between the selected water quality parameters and algal bloom.

Seasonal and longitudinal changes in copper and iron in surface water of shallow eutrophic Lake Kasumigaura, Japan

The concentrations of copper and iron in surface water of the eutrophic Lake Kasumigaura, the second largest lake in Japan, have been monitored at three sites monthly from April 1989 until March 1994. The metals were analyzed by a graphite furnace AAS after filtration with a 0.45-μm membrane filter. The concentrations of copper in the water were in the range 10 −9 to 5 × 10 −8 M and showed clear seasonal changes, being higher in summer and lower in winter. The concentrations of copper did not decrease on passage through the lake. The values of dissolved chemical oxygen demand (COD Mn) showed a similar behavior, an interaction of copper in the lake water with organic matter was estimated. The concentration of copper in the lake became sometimes higher than the value of EC 50 for Microcystis, however, the metal forms stable complex species with organic matter and the toxic effect may be reduced. The concentrations of iron, on the other hand, showed wide variation, but no obvious seasonal change. The concentrations of iron decreased very markedly during flow of water through the lake.

Novel application of a back-propagation artificial neural network model formulated to predict algal bloom

This paper describes the novel application of an artificial neural network (ANN) model based on the backpropagation method formulated to predict algal bloom by simulating the future growth of five phytoplankton species and the chlorophyll a concentration in the second largest lake in Japan: eutrophic freshwater Lake Kasumigaura. Comparison of observed and calculated values showed that (i) seasonal variations in the biomass of Microcystis spp. were well-predicted with respect to the timing and magnitude of algal bloom, and (ii) the concentration of chlorophyll a, as an indicator of the total biomass of phytoplankton, was well predicted in general. The resultant correlations for the other species, however, showed that model learning was insufficient to effectively predict species biomass; thereby indicating that some unknown factors which are not represented by the set of water quality parameters used as model input data affect phytoplankton growth. A sensitivity analysis performed on input parameters showed that chlorophyll a concentration was mainly affected by P04-P concentration, while cyanobacteria and diatom species were affected by N03-N and NH4-N concentrations, respectively. These results indicate that the “algal bloom” ANN model achieved reasonable effectiveness with respect to learning the relationship between the selected water quality parameters and algal bloom.