F2 - High resolution carbon flux inference of phytoplankton-zooplankton interaction
Doctoral student: Marvin Mayerhofer 
Supervisors: Prof. Dr. Ferdi Hellweger , PD Dr. Sabine Hilt , Prof. Dr. Reinhard Hinkelmann 
Microbial systems are diverse and complex and by far not fully understood. Large aquatic microbial time series are available but how can these large datasets be used? In comparison to widely used empirical approaches, this study uses a high-resolution mechanistic inference method called FLUXNET to give insights into the complex realm of planktonic interactions. The focus lies on the phytoplankton-zooplankton interaction, especially the cyanobacteria-zooplankton interaction which still remains largely unexplored (Ger, et al., 2014). Broader and deeper knowledge on interaction behavior and adaption mechanisms of cyanobacteria is of emerging concern as the amount of unwanted blooms rise globally (Huisman 2018). Increased eutrophication, rising CO2 concentrations, and global warming are factors that enhance cyanobacteria blooms (Paerl & Huisman, 2009; Paerl & Otten, 2013) which then can have a negative effect on biodiversity, food webs, and water use (Paerl & Huisman, 2008) (Ullah, et al., 2018).
The study site for this research project is lake Müggelsee, a shallow eutrophic polymictic lake located south-east of Berlin (Germany)(Driescher, et al., 1993). Lake Müggelsee experienced fundamental ecological changes since 1990 in terms of nutrient loading, eutrophication and species invasion (Köhler, et al., 2005) (Gsell, et al., 2015; Shatwell & Köhler, 2018). In addition, large and continuous time series data exists for lake Müggelsee, which is therefore an interesting case for long term analysis.
Now the aim of this study can roughly be summarized in these two research questions:
- How does the cyanobacteria-zooplankton interaction shape the evolution and behavior of individual species in lake Müggelsee?
- What are significant patterns in the long-term phytoplankton-zooplankton interaction in lake Müggelsee?
- Figure 1: Time series data for selected species over the course of 24 years: Rhodomonas minuta/lacustris (RHO), Calanoid copepods (CCO), Cryptomonas (CRY), Daphnia juvenile (DJU).
- © Mayerhofer
- Figure 2: Schematic draw to illustrate the process of delumping. N=Nutrients, P=Phytoplankton, Z=Zoooplankton.
- © Mayerhofer
Methods & Results
FLUXNET is a mechanistically-constrained inference method for long-term data analysis that is applicable with a high taxonomic resolution. FLUXNET is calibrated to 24-years of time series data (1994-2017) of lake Müggelsee (Fig. 1) where the top 40 phytoplankton and zooplankton species are selected based on total biomass. The interaction of these species is then inferred by FLUXNET considering abiotic factors like nutrients, light, temperature and discharge. One of the main outputs is a carbon flux network of the individual interacting species.
FLUXNET follows a typical model development workflow and mimics natural speciation or diversification by de-lumping the components (Fig. 2). In each de-lump step, the parameters are optimized and after reaching a minimal error, one component is split into two, introducing a new component. To account for long term changes and factors that are not explicitly considered in this method, each year is set up individually. To get a glimpse of how this method works, fig. 3 shows the inferred carbon flux network for t= 124d during the spring bloom 1994 where all species are delumped.
For a more in-depth view on this method, FLUXNET was already applied to characterize phytoplankton-heterotrophic bacteria interactions via dissolved organic matter in a marine system (Mayerhofer, et al., 2021). Here, the resulting flux network shows a strong correlation between the abundance of bacteria species and their carbon flux during blooms, where copiotroph have a higher importance than oligotrophs. In addition, the fraction of exudates decreases during blooms and the functional similarity of phytoplankton (i.e. what they produce) decouples their association with heterotrophs.
- Figure 3: Top: Inferred carbon flux network for t=124d in 1994. Nodes are components and size indicates in/outflux (mmolC/L/d). Lines are fluxes and thickness is proportional to log flux (mmolC/L/d). Lines below a flux threshold of 0.01% are cut off to highlight most important fluxes. Bottom: Corresponding Chlorophyll-a content in µg/L. Black line represents time series data, green line model output.
- © Mayerhofer
Other UWI projects: F5, F6
Common topic group: Interfaces in urban freshwater systems 
- Driescher, E., Behrendt, H., Schellenberger, G. & Stellmacher, R., 1993. Lake Muggelsee and its Environment - Natural Conditions and Anthropogenic Impacts. Int. Revue ges. Hydrobiol., 78(3), pp. 327-343.
- Ger, K. A., Hansson, L.-A. & Lürling, M., 2014. Understanding cyanobacteria-zooplankton interactions in a more eutrophic world. Freshwater Biology, 59(9), pp. 1783-1798.
- Gsell, A. S., Özkundakcia, D., Héberta, M. & Adrian, R., 2015. Quantifying change in pelagic plankton network stabilityand topology based on empirical long-term data. Ecological Indicators, Band 65, pp. 76-88.
- Huisman, J. et al., 2018. Cyanobacterial blooms. Nature Reviews Microbiology, Band 16, pp. 471-483.
- Köhler, J. et al., 2005. Long‐term response of a shallow, moderately flushed lake to reduced external phosphorus and nitrogen loading. Freshwater Biology, 50(10), pp. 1639-1650.
- Mayerhofer, M. M. et al., 2021. Dynamic carbon flux network of a diverse marine microbial community. ISME Communications, 1(50).
- Paerl, H. & Huisman, J., 2008. Blooms like it hot. Science, Band 320, pp. 57-58.
- Paerl, H. & Otten, T., 2013. Harmful cyanobacterial blooms: causes, consequences, and controls.. Microbial ecology, 65(4), pp. 995-1010.
- Paerl, H. W. & Huisman, J., 2009. Climate Change: A Catalyst for Global Expansion of Harmful Cyanobacterial Blooms. EEnvironmental Microbiology, 1(1), pp. 27-37.
- Shatwell, T. & Köhler, J., 2018. Decreased nitrogen loading controls summer cyanobacterial blooms without promoting nitrogen-fixing taxa: Long-term response of a shallow lake. Limnology and Oceanography, 64(S1), pp. 166-178.
- Ullah, H., Nagelkerken, I., Goldenberg, S. & Fordham, D., 2018. Climate change could drive marine food web collapse through altered trophic flows and cyanobacterial proliferation. PLOS Biology, 16(1), p. e2003446.