Does biodiversity of estuarine phytoplankton depend on hydrology?

Phytoplankton growth in estuaries is controlled by factors such as flushing, salinity tolerance, light, nutrients and grazing. Here, we show that biodiversity of estuarine phytoplankton is related to flushing, and illustrate this for some European estuaries.The implications for the definition of reference conditions for quality elements in estuaries of different types are examined, leading to the conclusion that constraints on the number of estuarine and coastal types that may be defined for management purposes require that quality classes take into account natural variability within types, in order to be ecologically meaningful. We develop a screening model to predict the growth rate required for a phytoplankton species to be present under different flushing conditions and apply it to estuaries in the EU and US to show how changes in physical forcing may alter biodiversity. Additional results are presented on the consequences for eutrophication, showing that changes in residence time may interact with species-specific nutrient uptake rates to cause shifts in species composition, potentially leading to effects such as harmful algal blooms.We discuss applications for integrated coastal zone management, and propose an approach to normalization of estuarine phytoplankton composition as regards species numbers.     

A reevaluation of the cultural eutrophication of Lake Okeechobee using multiproxy sediment records.

Lake Okeechobee, the hydrological lynchpin of the Everglades ecosystem, is the subject of an ambitious, multiagency restoration effort aimed at reducing phosphorus inputs and resulting algal blooms and impaired water clarity. This restoration is predicated on returning the lake to something closer to its predisturbance condition, but that goal has been challenged on the premise that the lake has always been eutrophic. The resolution of this debate and the appropriateness of the nutrient reduction goals thus depend on obtaining a reliable sediment record of past limnological conditions--the aim of this study. Because of the potential for severe sediment mixing from tropical storms, this investigation used multiple dating tools to examine the integrity of the sediment record and then analyzed proxies for nutrient enrichment, phytoplankton composition, and paleoproductivity. Sediment profiles for atmospheric pollutants, fertilizer contaminants, and radiocesium from three widely spaced cores showed good preservation of stratigraphic detail and coherence with the 210Pb chronologies. These results demonstrated that sediment stratigraphy is largely intact and retains a reliable record of limnological change. Geochemical proxies provide strong evidence of increased nutrient loading beginning ca. 1950. Concentrations of sediment P double, and N:P and C:N ratios drop, while those for N isotopes (delta15N) increase. At the same time, tracers of phosphate fertilizers (uranium, vanadium, and arsenic) rise. These changes are synchronous among cores and constitute a robust, internally consistent record of increasing water-column P. Biotic responses are manifested in higher concentrations and in changing composition of fossil algal pigments, including (1) large increases in the concentrations of chemically robust carotenoids, (2) corresponding decreases in the ratios of pigments from diatoms to chlorophyte and cyanobacterial algae, and (3) increases in UVR-photo-protective compounds indicating greater prevalence of surface algal blooms. This study provides strong evidence that Lake Okeechobee has experienced accelerated eutrophication linked with post-1950s land use changes in its watershed, a conclusion consistent with the nutrient reduction goals of the Lake Okeechobee Protection Program. The results contradict recent claims that the lake's trophic state has not changed over time, as well as the assertion that sediments of large shallow lakes cannot support a reliable chronology of past events.     

A vertical-compressed three-dimensional ecological model in Lake Taihu, China

A three-dimensional ecological model on the basis of the analyses of environmental characteristics is set up for Lake Taihu, one of the largest shallow lakes in China. The hydrodynamic processes, nutrient cycling, chemical processes and biological processes are integrated in the model. Model state variables include: water current, surface displacement, nutrients of nitrogen and phosphorus, as well as their different forms such as ammonia nitrogen, nitrate nitrogen, phosphate phosphorus, etc., biomasses of macroplankton, phytoplankton, zooplankton and fish, and also the nutrient levels of macroplankton and phytoplankton. A nutrient budget and sediment transformation are also coupled in the model. The data from January 17, 1997 to January 18, 1998 are use to calibrate the model. The model results have shown good agreement with the observations. It implies that the model could be used for the lake environmental management and research for examining the processes and determining the water quality. The reasons of deviations between the modelled results and the observed values are also discussed. There are six factors that explain the deviations of the modelled results from the observed values and they can be grouped into two sets. One set of problems is associated with the standard deviation introduced by sampling and analyses. The second set of problems can be solved by introduction of processes lacking in the present model (resuspension, phytoplankton transportation mode under the wind with low speed, shifts in species composition and varied size of phytoplankton and zooplankton). The latter two processes should be included in the model at a later stage by integration of a structurally dynamic approach into the three-dimensional model.     

Lake phosphorus dynamics and climate warming: A mechanistic model approach

Model results in this work indicate that lakes may respond very differently to climate change depending on their physical character. A physical lake model and a mechanistic phosphorus model are combined with two temperature scenarios generated by a regional climate model (RCM) in three sites in central Sweden—Lake Erken and two basins of Lake Mälaren (Galten and Ekoln). In the phosphorus model water mixing, mineralization, diffusion and biouptake are temperature dependent. In the simulations, Lake Erken is much more sensitive to climate warming than the two basins of Lake Mälaren, and the reason is shown to be the much longer water residence time in Lake Erken (7 years), stressing the importance of internal lake processes. In Galten and Ekoln the water residence times are less than 1 year, and the effects of water temperature changes are small. In Lake Erken the concentration of epilimnetic-dissolved phosphorus is almost doubled in spring and autumn in the warmest climate scenario. Since the lake is mostly phosphorus limited, this means that the potential for phytoplankton production is almost doubled. The implication would be that in Lake Erken, and in other eutrophic lakes with long water residence times, eutrophication problems may become serious in the future, and that managers may need to take action today in order to maintain good water quality in these lakes.     

Implementation of ecological modeling as an effective management and investigation tool: Lake Kinneret as a case study

The need for scientifically based management of lakes, as key water resources, requires the establishment of quantitative relationships between in-lake processes responsible for water quality (WQ) and the intensity of major management measures (MM, e.g. nutrient loading). In this paper, we estimate the impact of potential changes in nutrient loading on the Lake Kinneret ecosystem. Following validation of the model against a comprehensive dataset, we applied an approach that goes beyond scenario testing by linking the lake ecosystem model DYRESM–CAEDYM with a set of ecosystem variables included in a pre-assessed system of water quality indices. The emergent properties of the ecosystem predicted from the model simulations were also compared with lake data as a form of indirect validation of the model. Model output, in good agreement with lake data, indicated differential effects of nitrogen and phosphorus nutrient loading on concentrations, and major in-lake fluxes, of TN and TP, and dynamics and algal community structure. Both model output and lake data indicated a strong relationship between nitrogen loading and in-lake TN values. This relationship is not apparent for phosphorus and only a weak relationship exists between phosphorus loading and in-lake TP. The modeling results, expressed in terms of water quality, allowed establishment of critical/threshold values for the nutrient loads. Implementation of the ecological modeling supplemented with the quantified set of WQ indices allowed us to take a step towards establishment of the association between permissible ranges for water quality and major management measures, i.e. towards sustainable management.     

Spatial and temporal hydrodynamic and water quality modeling analysis of a large reservoir on the South Carolina (USA) coastal plain

Two-dimensional, 31-segment, 61-channel hydrodynamic and water quality models of Lake Marion (surface area 330.7 km²; volume 1548.3×10⁶ m³) were developed using the WASP5 modeling system. Field data from 1985 to 1990 were used to parameterize the models. Phytoplankton kinetic rates and constants were obtained from a related in situ study; others from modeling literature. The hydrodynamic model was calibrated to estimates of daily lake volume; the water quality model was calibrated for ammonia, nitrate, ortho-phosphate, dissolved oxygen, chlorophyll-a, biochemical oxygen demand, organic nitrogen, and organic phosphorus. Water quality calibration suggested the model characterized phytoplankton and nutrient dynamics quite well. The model was validated (Kolmogorov–Smirnov two-sample goodness-of-fit test at P<0.05) by reparameterizing the nutrient loading functions using an independent set of field data. The models identified several factors that may contribute to the spatial variability previously reported from other research in the reservoir, despite the superficial absence of complex structure. Sensitivity analysis of the phytoplankton kinetic rates suggest that study site-specific estimates were important for obtaining model fit to field data. Sediment sources of ammonia (10–60 mg m⁻² day⁻¹) and phosphate (1–6 mg m⁻² day⁻¹) were important to achieve model calibration, especially during periods of high temperatures and low dissolved oxygen. This sediment flux accounted for 78% (nitrogen) and 50% (phosphorus) of the annual load. Spatial and temporal variability in the lake, reflected in the calibrated and validated models, suggest that ecological factors that influence phytoplankton productivity and nutrient dynamics are different in various parts of the lake. The WASP5 model as implemented here does not fully accommodate the ecological variability in Lake Marion due to model constraints on the specification of rate constants. This level of spatial detail may not be appropriate for an operational reservoir model, but as a research tool the models are both versatile and useful.     

Water management strategies against toxic Microcystis blooms in the Dutch delta.

To prevent flooding of the Dutch delta, former estuaries have been impounded by the building of dams and sluices. Some of these water bodies, however, experience major ecological problems. One of the problem areas is the former Volkerak estuary that was turned into a freshwater lake in 1987. From the early 1990s onward, toxic Microcystis blooms dominate the phytoplankton of the lake every summer. Two management strategies have been suggested to suppress these harmful algal blooms: flushing the lake with fresh water or reintroducing saline water into the lake. This study aims at an advance assessment of these strategies through the development of a mechanistic model of the population dynamics of Microcystis. To calibrate the model, we monitored the benthic and pelagic Microcystis populations in the lake during two years. Field samples of Microcystis were incubated in the laboratory to estimate growth and mortality rates as functions of light, temperature, and salinity. Recruitment and sedimentation rates were measured in the lake, using traps, to quantify benthic-pelagic coupling of the Microcystis populations. The model predicts that flushing with fresh water will suppress Microcystis blooms when the current flushing rate is sufficiently increased. Furthermore, the inlet of saline water will suppress Microcystis blooms for salinities exceeding 14 g/L. Both management options are technically feasible. Our study illustrates that quantitative ecological knowledge can be a helpful tool guiding large-scale water management.     

Impact of climate factors on cyanobacterial dynamics and their interactions with water quality in South Taihu Lake,China

Cyanobacterial bloom events in South Taihu Lake cause serious water quality problems and disturb aesthetic view of lake’s environment. In this study, correlations between cyanobacterial blooms and hydro-meteorological factors, including water quality, temperature and precipitation were investigated. Results demonstrated that South Taihu Lake was heavily affected by cyanobacteria and the proliferation of cyanobacteria due to variations in hydro-meteorological factors and water quality conditions. Water quality parameters, including COD, NH₃-N, TN and TP improved significantly since 2008 even at an elevated cyanobacterial bloom situation. Correlation analyses have shown that the development of cyanobacterial density and chlorophyll a concentration was sensitive to a wider temperature variation. The optimum temperature for cyanobacteria was 20°C, while extremely low and high temperatures were found to suppress their growth. Moreover, unusual rainfall patterns were measured during the study period (2003–2009), which showed an adverse impact on cyanobacterial development. Findings from this study suggested that seasonal lake’s water quality monitoring; suitable treatment of cyanobacterial blooms and strict policy implementation can solve the water quality issues in highly eutrophic lakes like Taihu.     

Estimating the critical phosphorus loading of shallow lakes with the ecosystem model PCLake: Sensitivity, calibration and uncertainty

There is a vast body of knowledge that eutrophication of lakes may cause algal blooms. Among lakes, shallow lakes are peculiar systems in that they typically can be in one of two contrasting (equilibrium) states that are self-stabilizing: a ‘clear’ state with submerged macrophytes or a ‘turbid’ state dominated by phytoplankton. Eutrophication may cause a switch from the clear to the turbid state, if the P loading exceeds a critical value. The ecological processes governing this switch are covered by the ecosystem model PCLake, a dynamic model of nutrient cycling and the biota in shallow lakes. Here we present an extensive analysis of the model, using a three-step procedure. (1) A sensitivity analysis revealed the key parameters for the model output. (2) These parameters were calibrated on the combined data on total phosphorus, chlorophyll-a, macrophytes cover and Secchi depth in over 40 lakes. This was done by a Bayesian procedure, giving a weight to each parameter setting based on its likelihood. (3) These weights were used for an uncertainty analysis, applied to the switchpoints (critical phosphorus loading levels) calculated by the model. The model was most sensitive to changes in water depth, P and N loading, retention time and lake size as external input factors, and to zooplankton growth rate, settling rates and maximum growth rates of phytoplankton and macrophytes as process parameters. The results for the ‘best run’ showed an acceptable agreement between model and data and classified nearly all lakes to which the model was applied correctly as either ‘clear’ (macrophyte-dominated) or ‘turbid’ (phytoplankton-dominated). The critical loading levels for a standard lake showed about a factor two uncertainty due to the variation in the posterior parameter distribution. This study calculates in one coherent analysis uncertainties in critical phosphorus loading, a parameter that is of great importance to water quality managers.     

洞庭湖浮游植物增长的限制性营养元素研究

近20年水质监测资料表明,洞庭湖水体富营养化日趋严重。洞庭湖水体主要污染物为氮和磷,而营养盐赋存形态及其含量对浮游植物生长的影响在洞庭湖尚未见报道。2011年9月至2012年8月对洞庭湖浮游植物生物量及主要营养盐赋存形态与含量进行监测,同时利用藻类增长的生物学（NEB）评价方法对限制浮游植物增长的营养盐进行了研究,并分析了浮游植物生物量与各营养元素之间的相关性。结果表明：洞庭湖主要污染物总氮（TN）和总磷（TP）的年平均值分别为1.90 mg·L-1和0.093 mg·L-1,溶解态无机氮（DIN）平均占ρ（TN）比例为87%,溶解态总磷（DTP）平均占ρ（TP）比例为70%。洞庭湖水体中,DIN是TN的主要贡献者,且不同形态DIN的贡献大小依次为ρ（NO3--N）〉ρ（NH4＋-N）〉ρ（NO2--N）;磷形态组成中,TP主要以溶解反应性磷（SRP）存在。春季洞庭湖水体中ρ（TN）、ρ（TP）较高,这一结果可能源于春季面源污染。洞庭湖水体中ρ（Chla）与氮显著正相关,与磷显著负相关。NEB 实验结果表明氮对洞庭湖浮游植物生长有明显的促进作用,其幅度随氮浓度的增加而加强,而磷对浮游植物的生长影响不大,有时出现抑制作用,硝态氮与磷之间不存在交互作用。因此,氮可能是洞庭湖浮游植物增长的主要限制性营养因子,这一研究暗示在洞庭湖富营养化控制过程中应特别注重氮的控制。     

The development of a multi-species algal ecodynamic model for urban surface water systems and its application

An ecodynamic model that can simulate four phytoplankton species has been developed to deal with the unique characteristics of urban river systems which has manmade river profile, flow controlled by gates, severe eutrophication status, and fragile aquatic ecosystem. The ecodynamic model was developed referencing two typical models: the water quality simulation model WASP and ecological model CAEDYM. The model can simulate 11 state variables: dissolved oxygen, carbonaceous biochemical oxygen demand, ammonia nitrogen, nitrate nitrogen, organic nitrogen, inorganic phosphorus, organic phosphorus and four phytoplankton species with zooplankton as a boundary condition. The ecodynamic model was applied to Sihai section of the Beijing urban river system, where serious algal blooms broke out in recent years. The dominant phytoplankton species are Cyanophyta, Chlorophyta, Bacillariophyta, and Cryptophyta. Site-specific data on geometry, meteorology, pollution sources, and existing ecosystem parameters were collected and used for model calibration and verification The model results mimic observed trends of water quality and phytoplankton species succession and can be used for forecasting algal blooms as well as assessment of river management measures.     

富营养化水体中藻类生长限制因素的确定及其应用

在富化营养化的巢湖，围隔实验结果表明磷和其它营养元素不是水体藻类的生长限制因素、藻类生长的正磷酸盐阈值浓度为０．０１９ｍｇ／ｌ，它低于巢湖实际浓度。湖水的矿物性浊度很高，净生产力在一米水深以下呈现负值，数据表明学强在大数时间是藻类生长的限制因素。在巢湖治理过程中，需大幅降低流域内的磷负荷，使湖水平均溶解态总磷浓度从目前的０．０４９ｍｌ／ｌ降到０．０１９ｍｇ／ｌ以下。因此巢湖在治理和恢复是一个缓慢的     

Analysis of the trade-off between economic growth and the reduction of nitrogen and phosphorus emissions in the Poyang Lake Watershed, China

Lake eutrophication leading to water pollution is a major global concern. In recent years, rapid economic growth and the increase in the intensity of resource exploitation in China have caused the influx of nitrogen and phosphorus into lakes. This in turn has led to more severe lake eutrophication, more frequent outbreaks of algal blooms, and the degradation of lake ecosystems. An effective plan balancing economic growth with the reduction of nitrogen and phosphorus emissions is greatly needed. The design and implementation of such a plan requires the collection and analysis of pertinent data. In this paper, we use the environmental computable general equilibrium (ECGE) model to identify the most effective way to balance economic growth with the reduction of nitrogen and phosphorus emissions. For the multiregional analysis, we use social accounting matrices (SAMs) and a provincial trade matrix based on the assumptions of the gravity model. We consider the Poyang Lake Watershed as a case study to illustrate the utility of the model. Based on present conditions in the Poyang Lake Watershed, restricting nitrogen and phosphorus emissions from sectors with the highest emissions is more effective for balancing economic growth and the reduction of nitrogen and phosphorus emissions than restricting nitrogen and phosphorus emissions from all sectors.     

Exploitation and destabilization of a warm, freshwater ecosystem through engineered hydrological change

Exploitation of freshwater resources is having catastrophic effects on the ecological dynamics, stability, and quality of those water resources on a global scale, especially in arid and semiarid regions. Lake Kinneret, Israel (the Biblical Sea of Galilee), the only major natural freshwater lake in the Middle East, has been transformed functionally into a reservoir over the course of approximately 70 years of hydrological alterations aimed mostly at producing electrical power and increasing domestic and agricultural water supply. Historical changes in lake chemistry and biology were reconstructed using analysis of sedimentary nutrient content, stable and radioisotope composition, biochemical and morphological fossils from algae, remains of aquatic invertebrates, and chemical indices of past light regimes. Together, these paleolimnological analyses of the lake's bottom sediments revealed that this transformation has been accompanied by acceleration in the rate of eutrophication, as indicated by increased accumulation rates of phosphorus, nitrogen, organic matter, phytoplankton and bacterial pigments, and remains of phytoplankton and zooplankton. Substantial increases in these indices of eutrophication coincide with periods of increased water-level fluctuations and drainage of a major upstream wetland in the early to middle 20th century and suggest that management of the lake for increased water supply has degraded water quality to the point that ecosystem stability and sustainability are threatened. Such destabilization may be a model for eutrophication of freshwater lakes in other arid regions of the world in which management emphasizes water quantity over quality.     

滇池水体磷的时空变化与藻类生长的关系

水体磷的时空变化与藻类生长的关系对研究水体富营养化具有十分重要的作用。采用GPS定位,对滇池海埂、斗南、罗家村、新街、昆阳等5个代表性位点监测断面水体总磷、可溶性磷及叶绿素a含量进行了为期1年（2003年5月至2004年5月）的动态研究,并在滇池海埂位点进行了日变化试验,全面分析了滇池不同区域、不同层次、不同时期水体总磷和可溶性磷的年变化、日变化及水体氮/磷比对藻类生长的影响。结果表明,滇池水体磷与藻类生长呈现显著的年变化和日变化特征,显示了滇池全湖水体总磷与叶绿素a周年变化呈显著正相关,水体可溶性磷与叶绿素a呈正相关趋势;海埂位点水体总磷与叶绿素a日变化呈显著正相关,水体可溶性磷与叶绿素a日变化呈显著的负相关,水体氮磷比与叶绿素a呈显著正相关。表明水体磷负荷对藻类生长影响呈现显著的水体区域性和水层差异性和季节性,藻类生长主要吸收水体中的可溶性磷,暗示了滇池水体磷是藻类生长的主要限制因子之一。     

Development of Lake Ladoga ecosystem models: modeling of the phytoplankton succession in the eutrophication process. I

New models of Lake Ladoga ecosystem and the results of modeling are presented. In the first part the model of phytoplankton succession in the process of anthropogenic eutrophication of the lake is considered under the evolution of the phosphorus loading. The still continued anthropogenic eutrophication of the lake started in 1962 when the phosphorus load began to increase. Since 1962 during the evolution of the lake’s state from oligotrophic to developed mezotrophic one, the structure of phytoplankton community dominating species was significantly changed as well as its total productivity. The system state in the model is described by 14 parameters: nine phytoplankton complexes, zooplankton, dissolved organic matter, detritus, dissolved mineral phosphorus and dissolved oxygen. The number of parameters of this model is noticeably larger than that of previous models created by the authors. The relative dynamics of phytoplankton complexes in the lake’s ecosystem evolution was simulated by the new model. It is shown that the modeling results are adequately corresponding to the observation data. The results of phytoplankton structure modeling allow to estimate the impact of phytoplankton on the water quality as well as give the prediction of the lake’s ecosystem evolution with the changes of the phosphorus loading.     

水动力条件对藻类影响的研究进展

水动力过程是影响水体富营养化状态和水华爆发的重要因素,水动力因素对藻类影响的研究对于富营养化水体藻类控制具有重要意义。归纳分析近年来关于流速、流态对藻类生长和种类变化的研究报道;就水动力条件对藻类的影响及其作用机理等详细地进行了文献综述。水动力条件对藻类生长的影响分为流速和流态两个方面,不论是单一藻种还是混合藻类,低流速、小扰动有利于藻类的生长和聚集,流速增大则导致Chla浓度先递增后递减,不同藻类的临界流速并不相同;藻类生长随着湍流程度的增加而逐渐受到抑制,抑制作用与水流流态（层流、过渡流、湍流）无明显相关关系,水体流态的变化造成水流剪应力的变化,藻类种类的差异导致其对水流剪应力的响应变化。水动力条件变化引起的藻类种群结构变化,主要表现为水体混合加剧导致优势种群的转换。水动力条件对藻类影响的作用原理主要是引起了光强的改变、细胞长度的变化、营养盐运送及捕食行为变化等。综观当前的研究成果,水动力能否真正阻止藻类细胞的生长或聚集,影响藻类生长或种类变化的扰动的最低水平以及水动力对藻类影响的作用机理是这一领域未来研究的重点所在。     

Assesment of the response of a shallow macrotidal estuary to changes in hydrological and wastewater inputs through numerical modelling

The aim of this study was to investigate the response to short-term changes in river freshwater discharges and in nutrients loadings (mainly from the treatment of urban wastewater), of the shallow macrotidal Urdaibai estuary (north of Spain), by using numerical tools. A two-dimensional hydrodynamic model and a water quality model were applied to the estuary, in order to better use it as a prediction tool in the study of the effects of variations in hydrodynamic conditions and in waste water inputs. The model was calibrated and verified using data measured under different hydrological conditions (spring and summer). A model calibration was carried out with field data measured during the summer, while the model validation was conducted for spring conditions. The calibration process allowed the model parameter definition, while the model validation permitted the verification of the calibrated parameters under different environmental conditions. The model results were in reasonable agreement with field measurements, in both the calibration and the validation phases. The model showed a significant decrease in phytoplankton concentration with river input increase. A study on the effects of nutrient input reduction from the Gernika Waste Water Treatment Plant (WWTP) was conducted. It showed a decrease in phytoplankton concentration with decreasing levels of nutrient discharge. This reduction was more pronounced in conjunction with the highest river discharge. In that case, a 50% decrease for the elimination of the WWTP discharge was observed.     

Phytoplankton carbon dynamics during a winter-spring diatom bloom in an enclosed marine ecosystem: primary production,biomass and loss rates

Phytoplankton production, standing crop, and loss processes (respiration, sedimentation, grazing by zooplankton, and excretion) were measured on a daily basis during the growth, dormancy and decline of a winter-spring diatom bloom in a large-scale (13 m³) marine mesocosm in 1987. Carbonspecific rates of production and biomass change were highly correlated whereas production and loss rates were unrelated over the experimental period when the significant changes in algal biomass characteristic of phytoplankton blooms were occurring. The observed decline in diatom growth rates was caused by nutrient limitation. Daily phytoplankton production rates calculated from the phytoplankton continuity equation were in excellent agreement with rates independently determined using standard ¹⁴C techniques. A carbon budget for the winter bloom indicated that 82.4% of the net daytime primary production was accounted for by measured loss processes, 1.3% was present as standing crop at the end of the experiment, and 16.3% was unexplained. Losses via sedimentation (44.8%) and nighttime phytoplankton respiration (24.1%) predominated, while losses due to zooplankton grazing (10.7%) and nighttime phytoplankton excretion (2.8%) were of lesser importance. A model simulating daily phytoplankton biomass was developed to demonstrate the relative importance of the individual loss processes.     

Source identification and prediction of nitrogen and phosphorus pollution of Lake Taihu by an ensemble machine learning technique

● A machine learning model was used to identify lake nutrient pollution sources. ● XGBoost model showed the best performance for lake water quality prediction. ● Model feature size was reduced by screening the key features with the MIC method. ● TN and TP concentrations of Lake Taihu are mainly affected by endogenous sources. ● Next-month lake TN and TP concentrations were predicted accurately. Effective control of lake eutrophication necessitates a full understanding of the complicated nitrogen and phosphorus pollution sources, for which mathematical modeling is commonly adopted. In contrast to the conventional knowledge-based models that usually perform poorly due to insufficient knowledge of pollutant geochemical cycling, we employed an ensemble machine learning (ML) model to identify the key nitrogen and phosphorus sources of lakes. Six ML models were developed based on 13 years of historical data of Lake Taihu’s water quality, environmental input, and meteorological conditions, among which the XGBoost model stood out as the best model for total nitrogen (TN) and total phosphorus (TP) prediction. The results suggest that the lake TN is mainly affected by the endogenous load and inflow river water quality, while the lake TP is predominantly from endogenous sources. The prediction of the lake TN and TP concentration changes in response to these key feature variations suggests that endogenous source control is a highly desirable option for lake eutrophication control. Finally, one-month-ahead prediction of lake TN and TP concentrations (R² of 0.85 and 0.95, respectively) was achieved based on this model with sliding time window lengths of 9 and 6 months, respectively. Our work demonstrates the great potential of using ensemble ML models for lake pollution source tracking and prediction, which may provide valuable references for early warning and rational control of lake eutrophication.