Effect of Nitrogen Reduction in Shallow Lakes

Most inland waters in Germany did not meet the good ecological status requirements stipulated in the EU Water Framework Directive for 2015. It was previously assumed that phosphorus is the primary determinant of water quality. In the BMBF funded project NITROLIMIT (www.nitrolimit.de) it has been proved that nitrogen is also a crucial control variable for many surface waters and its reduction ecologically meaningful. This has led to demands for procedures to ensure the reduction of nitrogen input.

Scientists from different disciplines and seven scientific institutions worked together on interrelated research modules to clarify whether nitrogen reduction is ecologically meaningful and economically feasible. Modelling was part of research module 2 that aimed to fill knowledge gaps about N- and P-turnover processes.

Scientific Questions

The "toy-model" found at these pages aims to demonstrate some of the main scientific questions:

  1. Which nitrogen sources fuel the development of phytoplankton in summer:
    • external nitrogen load entering the lake by its inflow?
    • atmospheric N2, fixed by specialized blue-green algae (N2 fixers)?
    • remineralized nutrients, originating from the organic matter at the sediment surface, especially from the dying algae of the spring bloom?
  2. Will it be likely, that N2 fixing cyanobacteria will become dominant?
  3. How is nitrogen and phosphorus turnover coupled?

Please note: this model is an extremely simplified representation of a natural lake. It aims to support understanding of main processes, and especially their dynamics and complexity. However, the model presented here is not able to make ultimate predictions.

Picture of Lake Langer See

Langer See (Brandenburg, Germany), one of the research objects of the project.

Simplified Lake Model with 3 Phytoplankton Groups

Model Overview

The model describes nutrient turnover, growth of phytoplankton and zooplankton, and its sedimentation and remineralization in a shallow lake. Main ideas were borrowed from a simplified version of the BELAMO lake model (Reichert, Mieleitner and Schuwirth 2016), while we added some more emphasis on processes in shallow lakes.The present version contains the following state variables.

Three functional groups of phytoplankton and one zooplankton group:

  • ALG_F1: blue-green algae that are able to fix atmospheric nitrogen N2 (diazotrophic cyanobacteria)  (gDM/m3)
  • ALG_F2: blue-green algae that are not able to fix atmospheric nitrogen N2 (non-fixing cyanobacteria) (gDM/m3)
  • ALG_F3: other phytoplankton groups (gDM/m3)
  • ZOO: zooplankton (gDM/m3)
Nutrients and oxygen:
  • HPO4: dissolved phosphorus (gP/m3)
  • NH4: ammonium-nitrogen (gN/m3)
  • NO3: nitrate-nitrogen (gN/m3)
  • O2: dissolved oxygen (gO/m3)

Sedimentation and remineralization were of special interest, so we distinguish different fractions of particulate organic matter. The model considers a degradable and a non-degradable (inert) fraction and keeps track from which living compartment (algae group or zooplankton) this material originates:

  • POMD_ALG_F1: particulate organic matter degradable (gDM/m3)
  • POMI_ALG_F1: particulate organic matter inert (gDM/m3)
  • POMD_ALG_F2: particulate organic matter degradable (gDM/m3)
  • POMI_ALG_F2: particulate organic matter inert (gDM/m3)
  • POMD_ALG_F3: particulate organic matter degradable (gDM/m3)
  • POMI_ALG_F3: particulate organic matter inert (gDM/m3)
  • POMD_ZOO: particulate organic matter degradable (gDM/m3)
  • POMI_ZOO: particulate organic matter inert (gDM/mm3)
  • SPOMD: sedimented particulate organic matter degradable (gDM/m2)
  • SPOMI: sedimented particulate organic matter inert (gDM/m2)

External forcing variables include water inflow, import of nutrients, organisms and organic matter; temperature, solar radiation (light), and light absorption by ice covering during the winter.

Figure 1 shows a simplified schematic representation of the model, and Figure 2 the so-called stoichiometry matrix. This matrix indicates which variables influence each other and in which direction.



Schematic Representation of the Model

image/svg+xml Nutrients Nutrients Algae3 Algae3 Nutrients->Algae3 Algae2 Algae2 Nutrients->Algae2 Algae1 Algae1 Nutrients->Algae1 Zoo Zoo Algae3->Zoo POM POM Algae3->POM Algae2->Zoo Algae2->POM Algae1->Zoo Algae1->POM Zoo->POM Outflow Outflow Zoo->Outflow remineralization SPOM SPOM sedimentation POM->Outflow Inflow Inflow Inflow->Nutrients N2 N2->Algae1 Ice Ice Ice->Algae3 Ice->Algae2 Ice->Algae1 Light Light Light->Ice Temp Temp Temp->Algae3 Temp->Algae2 Temp->Algae1 Temp->Zoo remineralization N2-fixation

Fig 1: Simplified representation of the most important model compartments and their connections.




Stoichiometry Matrix

stoichmat.html
Process HPO42- ALG_F1 ALG_F2 ALG_F3 ZOO O2 NH4+ NO3- POMDALG_F1 POMIALG_F1 POMDALG_F2 POMIALG_F2 POMDALG_F3 POMIALG_F3 POMDZOO POMIZOO SPOMD SPOMI
g_ALG_F1_NH4 Δ


Δ










g_ALG_F2_NH4
Δ

Δ










g_ALG_F3_NH4

Δ
Δ










g_ALG_F1_NO3 Δ


Δ










g_ALG_F2_NO3
Δ

Δ










g_ALG_F3_NO3

Δ
Δ










g_ALG_F1_N2 Δ


Δ











r_ALG_F1 Δ


Δ










r_ALG_F2 Δ


Δ










r_ALG_F3 Δ


Δ










d_ALG_F1






Δ Δ







d_ALG_F2








Δ Δ





d_ALG_F3










Δ Δ



g_ZOO_ALG_F1


Δ


Δ Δ







g_ZOO_ALG_F2


Δ




Δ Δ





g_ZOO_ALG_F3


Δ






Δ Δ



r_ZOO Δ


Δ










d_ZOO












Δ Δ

nitri




Δ









miner_POMD_ALG_F1 Δ



Δ









miner_POMD_ALG_F2 Δ



Δ









miner_POMD_ALG_F3 Δ



Δ









miner_SPOMD Δ



Δ









sed_POMD_ALG_F1














Δ
sed_POMD_ALG_F2














Δ
sed_POMD_ALG_F3














Δ
sed_POMI_ALG_F1















Δ
sed_POMI_ALG_F2















Δ
sed_POMI_ALG_F3















Δ
ex_O2




Δ











inflow Δ
Δ Δ





miner_POMD_ZOO Δ



Δ









sed_POMD_ZOO














Δ
sed_POMI_ZOO















Δ
sed_ALG_F1














Δ Δ
sed_ALG_F2














Δ Δ
sed_ALG_F3














Δ Δ
denitri_1
















zin



Δ












Fig2: Schematic representation of the model stoichiometry. Triangle up: positive dependency, triangle down: inverse dependency.

Reset to defaults



              

Acknowledgements

This is a derived work from the NITROLIMIT project, funded by the German Ministry of Education and Research (BMBF) under grant no. 033W015EN. The model is based on a teaching version of the BELAMO lake model (cf. publications of Reichert, Omlin, Mieleitner, Dietzel and Schuwirth below), with some own modifications from our group: Hannes Feldbauer, David Kneis, Thomas Petzoldt, Yue Zhao. Please consult the papers and web pages of our colleagues from EAWAG and ETH Zurich for more background information about the model.

Data were kindly provided by BTU Cottbus-Senftenberg, Department of Freshwater Conservation. Many thanks to Jacqueline Rücker, Björn Grüneberg, Andrew Dolman, Claudia Wiedner and Brigitte Nixdorf.

Note: The current version of model is intended for demonstration purposes only, not for predictions. It exemplifies a specific state of the discussion process and we are sill continuing to make the process descriptions more realistic. Our main contribution to modelling and process understanding in Nitrolimit part 2 was a 1D model of the sediment water interface, that needs more computer power. Its description and results can be found in written form in the final project report and in scientific publications, listed at the NITROLIMIT project page.

Please contact us in case of comments and questions.

References

Dietzel, A., Reichert, P. (2012) Calibration of computationally demanding and structurally uncertain models with an application to a lake water quality model Environmental Modelling and Software 38, 129-146.
Kneis, D. (2016) rodeo: A Code Generator for ODE-Based Models. R package version 0.6. https://github.com/dkneis/rodeo
Mieleitner, J., Reichert, P. (2008) Modelling functional groups of phytoplankton in three lakes of different trophic state. Ecological Modelling 211, 279-291.
Nixdorf, B., Grüneberg, B. & Rücker, J. (2016) Bilanzierung der saisonalen Stickstoffein- und -austräge sowie deren Umsetzungen in einem eutrophen Flachsee. Erweiterte Zusammenfassungen der Jahrestagung 2015 in Essen. Hardegsen: 69-76.
Omlin, M.; Reichert, P., Forster, R. (2001) Biogeochemical model of Lake Zürich: model equations and results. Ecological Modelling 141, 77-103
Omlin, M.; Brun, R., Reichert, P. (2001) Biogeochemical Model of Lake Zürich: Sensitivity, Identifiability and Uncertainty Analysis. Ecological Modelling 141, 105-123
Reichert, P., Schuwirth, N. (2010) A generic framework for deriving process stoichiometry in environmental models. Environmental Modelling and Software 25, 1241-1251
Reichert, P., Mieleitner, J. and Schuwirth, N. (2016) Modelling Aquatic Ecosystems. Course 701-0426-00 ETH Zürich. http://www.eawag.ch/en/department/siam/teaching/modelling-aquatic-ecosystems/
Rücker, J., Knie, M., Voss, M., Martienssen, M., Grüneberg, B., Kolzau, S., Nixdorf, B. (2016) Abschätzung des Stickstoffeintrages durch planktische Cyanobakterien (Nostocales). Erweiterte Zusammenfassungen der Jahrestagung 2015 in Essen. Hardegsen: 170-177.
Soetaert, K.; Petzoldt, T., Setzer, R. W. (2010) Solving Differential Equations in R: Package deSolve. Journal of Statistical Software 33(9), 1-25. http://dx.doi.org/10.18637/jss.v033.i09

Links

  • R: A free software environment for statistical computing and graphics
  • Shiny by Rstudio: Web application framework
  • Package deSolve: General-purpose differential equation solvers for R
  • Package rodeo: Code generator for ODE models in R



Figure scaling

Expert options

Parameters

Output data