An integrated modeling approach to predict trophic state changes in a large Brazilian reservoir

Uma abordagem de modelagem integrada para prever mudanças no estado trófico em um grande reservatório brasileiro

Carolina Cerqueira Barbosa, Maria do Carmo Calijuri, Phelipe da Silva Anjinho e André Cordeiro Alves dos Santos 

ABSTRACT

Process-based ecological models have been used to study freshwater ecosystems and water quality on a broad scale. However, it is also of pivotal importance to incorporate watershed dynamics and nutrient releases in the downstream freshwaters. Integrated modeling approaches have been used to understand the combined effect of climate warming and land use and land cover (LULC) changes in lake ecosystems.

Moreover, most basin-scale water quality models require many datasets and parameters to perform reliable simulations which contributes to reduce studies in poorly monitored basins, most of them located in the Global South. In this study, we developed a coupled hydrological-biogeochemical-ecological modeling framework forced by two regionalized climate models and three LULC change scenarios to forecast trophic state changes in a subtropical multipurpose reservoir for the decade 2050-2060. The projections indicated an average air temperature increase between 2°C and 3°C and a downward trend of the average rainfall and longwave radiation for the 2050s in comparison to the last decade.

We found a pattern of 28% increase in total phosphorus (TP) and total chlorophyll-a (TChla) concentrations in the reservoir compared with the historical baseline. The climate warming projections along the 2059 projected LULC and basin’s increased economic development scenarios have predicted trophic state index (TSI) shifts between mesotrophic and eutrophic conditions (53.3<TSI<57.7).

On the other hand, one of the climate projections along the reduced deforestation scenario indicated a trend towards oligotrophication between 2054 and 2056, however higher phosphorus availability (60µg.Unknown node type: fontUnknown node type: fontUnknown node type: font<TP<100µg.Unknown node type: fontUnknown node type: fontUnknown node type: font) and phytoplankton biomass (50µg.Unknown node type: fontUnknown node type: fontUnknown node type: font<TChla<97µg.Unknown node type: fontUnknown node type: fontUnknown node type: font) would be expected for the entire decade compared to recent years. The proposed coupled modeling framework demonstrated the potential of open-source tools in water quality management studies, especially for poorly monitored basins, based on climate change trends and human pressure.

Keywords: Climate change, Land Use and Land Cover changes, GLM-AED2, Water quality

RESUMO

Modelos ecológicos baseados em processos têm sido utilizados para estudar ecossistemas de água doce e a qualidade da água em larga escala. No entanto, também é de fundamental importância incorporar a dinâmica das bacias hidrográficas e a liberação de nutrientes nas águas doces a jusante. Abordagens de modelagem integrada têm sido utilizadas para compreender o efeito combinado do aquecimento climático e das mudanças no uso e cobertura da terra (LULC) em ecossistemas lacustres.

Além disso, a maioria dos modelos de qualidade da água em escala de bacia requer muitos conjuntos de dados e parâmetros para realizar simulações confiáveis, o que contribui para reduzir os estudos em bacias mal monitoradas, a maioria delas localizadas no Sul Global. Neste estudo, desenvolvemos uma estrutura de modelagem hidrológica-biogeoquímica-ecológica acoplada, forçada por dois modelos climáticos regionalizados e três cenários de mudança de LULC para prever mudanças no estado trófico em um reservatório multiuso subtropical para a década de 2050-2060.

As projeções indicaram um aumento médio da temperatura do ar entre 2°C e 3°C e uma tendência de queda da precipitação média e da radiação de onda longa para a década de 2050 em comparação com a última década. Encontramos um padrão de aumento de 28% nas concentrações de fósforo total (PT) e clorofila-a total (TChla) no reservatório em comparação com a linha de base histórica. As projeções de aquecimento climático ao longo dos cenários de desenvolvimento econômico ampliado da bacia e do LULC projetados para 2059 previram mudanças no índice de estado trófico (IST) entre condições mesotróficas e eutróficas (53,3 < IST < 57,7).

Por outro lado, uma das projeções climáticas ao longo do cenário de desmatamento reduzido indicou uma tendência à oligotrofização entre 2054 e 2056, porém com maior disponibilidade de fósforo (60 µg). Unknown node type: fontUnknown node type: fontUnknown node type: font<TP<100µg. Unknown node type: fontUnknown node type: fontUnknown node type: font) e biomassa fitoplanctônica (50µg. Unknown node type: fontUnknown node type: fontUnknown node type: font<TChla<97µg. Unknown node type: fontUnknown node type: fontUnknown node type: font) seria esperado para toda a década, em comparação com os anos recentes.

A estrutura de modelagem acoplada proposta demonstrou o potencial de ferramentas de código aberto em estudos de gestão da qualidade da água, especialmente para bacias mal monitoradas, com base nas tendências das mudanças climáticas e na pressão humana.

Introduction

The deterioration of water quality in freshwater ecosystems worldwide has been driven by several factors, mainly due to water level changes and inflow nutrient loadings (Gilboa et al., 2014). It is important to identify and quantify the drivers and pressures related to water pollution (Teurlincx et al., 2019) taken in account the main ongoing problems (Downing, 2014), e.g., eutrophication (Liu et al., 2021), agriculture impacts (Lopes et al., 2020), overexploitation (Simonovic and Arunkumar, 2016), water withdrawal (Feldbauer et al., 2020) and climate change (O’Reilly et al., 2015; Woolway et al., 2020).

Climate change impacts on lakes and reservoirs have been extensively studied in the last years (Ladwig et al., 2018; Magee and Wu, 2017; Jeppesen et al., 2017; Sahoo et al., 2016). The increase in water temperatures, modification of stratification patterns and thermal regime are some of the consequences already observed (Paerl and Huisman, 2008, Moe et al., 2016). Changes in rainfall and climate patterns have been generating immediate responses in the biogeochemical and ecological dynamics of watersheds and lakes (Adrian et al., 2009).

Disruption of biogeochemical cycles may impact the loading of dissolved organic matter and nutrients in the lake and affect the magnitude, variability and balance of in-lake metabolic processes (Brighenti et al., 2018) and increase eutrophication (Sinha et al., 2017). Thus, it has also driven the development, proliferation and maintenance of cyanobacteria blooms (Wells et al., 2015). Land use and land cover (LULC) changes have also driven eutrophication and have been investigated in order to improve water quality. On the other hand, even with the expected intensification of urbanization rate for the next decades, restoration efforts have shown to have effectiveness in the water quality control of some catchments (Fu et al., 2021).

The inference whether climate change or LULC changes are the most important drivers in water quality deterioration is not so straightforward. In this regard, integrated modeling approaches are important to advance our comprehension of how aquatic biogeochemical dynamics respond to meteorological conditions (Curtareli et al, 2015), human pressures (Munar et al 2018, 2019, Liu et al., 2015) and changing climate (Zhang et al, 2019). Integrating geography information system (GIS) technologies also have been giving insightful qualitative answers for water resource management studies (Liu et al., 2018). However, less than 5% of the modeling studies have tested coupled models to represent climate-catchment-lakes processes and only 6% have been conducted in Central/South America (Soares and Calijuri, 2021).

In the present study, we aimed to develop a coupled modeling framework based on open-source deterministic models and GIS tools to predict trophic state changes in a subtropical reservoir for the 2050s. We investigated how downscaled climate projections based on two regional climate models (RCM) along with three LULC change scenarios can facilitates our understanding of the combined effect of climate warming and landscape dynamics on the productivity of an aquatic ecosystem.

Following this, we draw conclusions on the performance and potentialities of the models and also the implications of trends in water level and water quality for reservoir operation to prevent eutrophication and not impair the drinking water supply. The proposed integrated modeling approach facilitates our current comprehension of water quality management to incorporate climate change mitigation strategies to prevent the water crisis and control water quality and catchment zoning to protect threatened landscapes.

Section snippets

Study site

Itupararanga reservoir is a large lake built in 1914 in the Southeast of Brazil in the Alto Sorocaba basin to support multiple uses, mainly hydropower generation and drinking water supply for almost 1 million people. The reservoir surface area is 29.9 km² and a water depth range of 14.5-23 m (Barbosa et al., 2021).

The climate is of the Cwa type, characterized by winter precipitation of less than one-tenth of the amount in mid-summer and air temperatures above 22°C during the summer, according

Models performance

The SMAP model was used to simulate daily flows upstream from the Itupararanga reservoir. The values of the calibrated parameters are shown in Table S4 in the Supplementary material. The hydrological model reasonably represented seasonality differences and captured flow peaks from the daily observed rainfall and evaporation data (Fig. 4). The PBIAS metric was 19% and r=0.52 in the calibration period and PBIAS=4% and r=0.69 in the validation one.

The air2stream model was used to simulate the

Evaluation of model’s performance and reproducibility

The simulated daily upstream flows were considered satisfactory for calibration and in very good agreement for validation according with Moriasi et al., 2007. Likewise, the predictions of stream water temperature by Air2stream are in the range found in previous studies using the model (0.86°C<RMSE<1.52°C, Fenocchi et al., 2017; Toffolon and Piccolroaz, 2015). The simulated upstream TP and TN concentrations showed good fit criteria compared with the reference values from Moriasi et al. (2007).

Conclusion

The current study developed an integrated modeling approach forced by two regionalized climate models and LULC change scenarios to perform forecasting of trophic state in a subtropical reservoir for the 2050s and to compare the outputs with the historical observations. We have chosen open-source tools implemented in R and GIS to foster the reproducibility of the modeling approach. The contribution of uncertainties in the models’ processes was also taken into account in the modeling framework.

Software availability

Instructions to download and install the SMAP, Air2stream, GLM-AED2, and the QGIS MOLUSCE Plugin can be found at:

-SMAP: http://www.labsid.eng.br/software.aspx?id=1

-Air2stream: https://github.com/marcotoffolon/air2stream

– GLM-AED2: https://aed.see.uwa.edu.au/research/models/aed/download.html

-QGIS MOLUSCE Plugin: https://qgis.org/en/site/forusers/download.html and https://plugins.qgis.org/plugins/molusce/ (Last access: 2021-08-18)

The basin distributed load approach is a GIS implementation of

CRediT authorship contribution statement

Carolina Cerqueira Barbosa: Writing – original draft, Writing – review & editing, Conceptualization, Methodology, Software, Visualization. Maria do Carmo Calijuri: Supervision, Resources, Writing – review & editing, Writing – original draft. Phelipe da Silva Anjinho: Writing – review & editing, Methodology, Visualization. André Cordeiro Alves dos Santos: Writing – review & editing, Data curation.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work was supported by the doctoral scholarship provided by the Coordination for Improvement of Higher Education Personnel (CAPES). The authors declare no conflicts of interest. Two anonymous reviewers have provided valuable suggestions for improving this manuscript.

References (79)

• T. Bucak et al.Modeling the effects of climatic and land use changes on phytoplankton and water quality of the largest Turkish freshwater lake: lake BeyşehirSci. Total Environ.(2018)
• D.G.F. Cunha et al.A trophic state index for tropical/subtropical reservoirs (TSI)Ecol. Eng.(2013)
• M.P. Curtarelli et al.Coupling remote sensing bio-optical and three-dimensional hydrodynamic modeling to study the phytoplankton dynamics in a tropical hydroelectric reservoirRemote Sens. Environ.(2015)
• D. Darko et al.Modeling the impacts of climate change on the thermal and oxygen dynamics of Lake VoltaJ. Great Lakes Res.(2019)
• K.J. Farrell et al.Ecosystem-scale nutrient cycling responses to increasing air temperatures vary with lake trophic stateEcol. Modell.(2020)
• A. Fenocchi et al.Relevance of inflows on the thermodynamic structure and on the modeling of a deep subalpine lake (Lake Maggiore, Northern Italy/Southern Switzerland)Limnologica(2017)
• M.M. Fernandes et al.Assessment of land use and land cover changes and valuation of carbon stocks in the Sergipe semiarid regionBrazil: 1992–2030. Land use policy(2020)
• Y. Gilboa et al.Defining limits to multiple and simultaneous anthropogenic stressors in a lake ecosystem – Lake Kinneret as a case studyEnviron. Model. Softw.(2014)
• Y. HuangUncertainty assessment in reservoir water quality modeling: implication for model improvementJ. Environ. Eng.(2015)
• P.J. JohnesEvaluation and management of the impact of land use change on the nitrogen and phosphorus load delivered to surface waters: the export coefficient modelling approachJ. Hydrol.(1996)
• L. Liu et al.Impacts of climate change and land use on the development of nutrient criteriaJ. Hydrol.(2018)
• N.J. Messina et al.Modeling response of water quality parameters to land-use and climate change in a temperate, mesotrophic lakeSci. Total Environ.(2020)
• S.J. Moe et al.Climate change, cyanobacteria blooms and ecological status of lakes: a Bayesian network approachEcol. Modell.(2016)
• M. Motew et al.Comparing the effects of climate and land use on surface water quality using future watershed scenariosSci. Total Environ.(2019)
• L.M.V. Soares et al.Deterministic modelling of freshwater lakes and reservoirs: current trends and recent progressEnviron. Model. Softw.(2021)
• M.L. Wells et al.Harmful algal blooms and climate change: Learning from the past and present to forecast the futureHarmful Algae(2015)
• C. Zhang et al.An ensemble modeling framework to study the effects of climate change on the trophic state of shallow reservoirsSci. Total Environ.(2019)
• S.C. Zipper et al.Continuous separation of land use and climate effects on the past and future water balanceJ. Hydrol.(2018)
• R. Adrian et al.Lakes as sentinels of climate changeLimnol. Oceanogr.(2009)
• P. Anjinho et al.Integrated empirical models to assess nutrient concentration in water resources: case study of a small basin in southeastern BrazilEnviron. Sci. Pollut. Res.(2021)
• C.C. Barbosa et al.Future projections of water level and thermal regime changes of a multipurpose subtropical reservoir (Sao Paulo, Brazil)Sci. Total Environ. 144741.(2021)
• F.G. Beghelli et al.Trophic state evolution over 15 years in a tropical reservoir with low nitrogen concentrations and cyanobacteria predominanceWater. Air. Soil Pollut.(2016)
• J. Brasil et al.Drought-induced water-level reduction favors cyanobacteria blooms in tropical shallow lakesHydrobiologia(2016)
• L.S. Brighenti et al.Importance of nutrients, organic matter and light availability on epilimnetic metabolic rates in a mesotrophic tropical lakeFreshwater Biol.(2018)
• Bruce O. BurnhamMarkov intertemporal land use simulation model. Southern journal of agricultural economicsSouthern Agric. Econ. Assoc.(1973)
• S.C. Chou et al.Assessment of climate change over south downscaling scenariosAm. J. Clim. Chang.(2014)
• L. Comte et al.Climate and land-use changes interact to drive long-term reorganization of riverine fish communities globallyProc. Natl. Acad. Sci.(2021)
• D.G.F. CunhaHeterogeneidade espacial e variabilidade temporal do Reservatório de Itupararanga: uma contribuição ao manejo sustentável dos recursos hídricos da Bacia do rio Sorocaba (SP). (Doctoral dissertation)(2012)
• D.G.F. Cunha et al.Uptake rates of ammonium and nitrate by phytoplankton communities in two eutrophic tropical reservoirsInt. Rev. Hydrobiol.(2017)
• M.R.A. da Costa et al.Effects of water level reduction on the dynamics of phytoplankton functional groups in tropical semi-arid shallow lakesHydrobiologia(2016)
• Downing, J.A., 2014. Limnology and oceanography: two estranged twins reuniting by global change 4, 215–232….
• FABH-SMT – Fundação Agência de Bacia Hidrográfica Sorocaba e Médio Tietê (2018). Relatório de Situação 2018. Available…
• J. Feldbauer et al.Managing climate change in drinking water reservoirs: potentials and limitations of dynamic withdrawal strategiesEnviron. Sci. Eur.(2020)
• Flato, G., Marotzke, J., Abiodun, B., Braconnot, P., Chou, S.C., Collins, W., Cox, P., Driouech, F., Emori, S., Eyring,…
• J. GarciaA qualidade da água, o processo de autodepuração e a proposição de cenários para gerenciamento sustentável do rio Sorocamirim (SP)(2013)
• H.V. Gupta et al.Status of auto- matic calibration for hydrologic models: Comparison with mul- tilevel expert calibrationJ. Hydrologic Eng.(1999)
• Hansen, N. 2016. The CMA evolution strategy: a tutorial. Available at https://arxiv.org/abs/1604.00772 (Last accessed…
• P.D.F. IslesThe misuse of ratios in ecological stoichiometryEcology(2020)
• S. Jachner et al.Statistical methods for the qualitative assessment of dynamic models with time delay (R package qualV)J. Stat. Softw.(2007)

There are more references available in the full text version of this article.