Convergence des modèles aquatiques de photosynthèse pour la planète/ Convergent aquatic photosynthesis modelling for our planet

Séminaire organisé par Olivier Bernard et Filipa Lopes du 14 au 19 septembre 2020

 

Participants

Bruno Assis Pessis, Sakina-Dorothee Ayata, Melika Baklouti, Caroline Baroukh, Olivier Bernard (organisateur), Hubert Bonnefond, Gael Bougaran,Benoît Chahuat, David Demory, Walid Djema, Damien Eveillard, Andrea Fanesi, Eric Fouilland, Thomas Lacour, Filipa Lopes (organisatrice), Francis Mairet, Carlos Martinez Von Dossow, Bastien Polizzi, Antoine Regimbeau, Magali Ribot, Antoine Sciandra.

Résumé 

Le phytoplancton joue un rôle clé dans notre écosystème, non seulement il est à l’origine de 50 à 85 % de notre oxygène et il consomme un quart du CO₂ anthropique, mais le phytoplancton est aussi un point central pour capter l’énergie solaire, grâce à la photosynthèse. L’énergie stockée dans les molécules de carbone organique soutient en outre tout l’écosystème aquatique. Une meilleure quantification des flux de carbone capturés par le phytoplancton dans un monde en réchauffement, grâce à la modélisation mathématique, est donc de la plus haute importance pour prédire plus précisément le climat futur de notre planète. En outre, domestiquer la capacité du phytoplancton à récolter l’énergie solaire et à la stocker sous forme de glucides ou de lipides, qui peuvent être transformés en biocarburant, est également une formidable opportunité de réduire notre empreinte environnementale. Les modèles mathématiques représentant la croissance et la photosynthèse partagent de nombreuses caractéristiques communes pour la prévision des flux de carbone océan-atmosphère ou pour l’évaluation de la production industrielle de biocarburant à base de phytoplancton. L’objectif de l’atelier était d’abord de partager et de consolider différentes visions de la modélisation de la photosynthèse aquatique, avec différentes communautés scientifiques. L’atelier a été suivi par 19 chercheurs, dont 6 chercheurs débutants et 13 chercheurs confirmés. Il s’avère que différentes approches combinées sont nécessaires pour représenter la grande diversité des échelles de temps et des facteurs limitants (selon le domaine : océan/procédés de production). L’atelier a révélé combien il est fructueux d’examiner, de confronter et de tirer le meilleur parti des développements réalisés dans différents domaines scientifiques : océanographie, modélisation des plantes et biotechnologie. Le développement de modèles phytoplanctoniques cohérents est une question importante pour mieux évaluer les flux de CO2, estimer l’évolution de la biodiversité et donc pour prévoir les effets des dérèglements climatiques. Mais c’est aussi un formidable moyen d’optimiser le processus de production, soit en améliorant sa conception, soit en sous-tendant la surveillance et le contrôle en ligne, ce qui rapproche le procédé de son potentiel théorique. Nous pensons qu’en unissant nos efforts, nous pouvons proposer un ensemble cohérent et réaliste de modèles de photosynthèse qui intègre les aspects clés de la photosynthèse du phytoplancton.

 

Summary

Phytoplankton plays a key role in our ecosystem, not only because it produced from 50 to 85% of our oxygen, or because it consumes 1/4^th of the anthropic CO₂: Phytoplankton is a central point to capture solar energy, through photosynthesis. The energy stored in organic carbon molecules further supports the full aquatic ecosystem. Better quantifying the carbon fluxes captured by phytoplankton in a changing world, through mathematical modelling, is thus of utmost importance to more accurately predict the future climate of our planet. Moreover, taming phytoplankton ability to harvest solar energy and store it under the form of carbohydrates or lipids, which can be turned into biofuel, is also a formidable opportunity to reduce our environmental footprint. The mathematical models representing growth and photosynthesis share many common features for ocean-atmosphere carbon flux predictions or for assessing industrial production of phytoplankton biofuel. The objective of the workshop was first to share and consolidate different visions of aquatic photosynthesis modelling, with different scientific communities. The workshop was attended by 19 researchers, including 6 junior and 13 senior researchers, respectively. It turns out that different combined approaches are required to represent the vast diversity of time scales and limiting factors (depending on the domain: ocean/production process). The workshop revealed how fruitful it can be to examine, confront and take the best out of the developments carried out in different scientific fields: oceanography, plant modelling and biotechnology. Developing consistent phytoplanktonic models is an important issue to better assess the CO₂ fluxes, estimate the evolution of biodiversity and therefore to predict effects of climate change. But it is also a tremendous way to optimize the production process, either by improving its design, or by supporting on-line monitoring and control leading the process closer to its theoretical potential. We believe that, by unifying our efforts, we can propose a coherent and realistic bundle of photosynthesis models that incorporates key aspects of phytoplankton photosynthesis.

 

Report

Phytoplankton play a key role in our ecosystem. It produced from 50 to 85% of our oxygen, and consumes one fourth of the anthropic CO₂. But more than that, phytoplankton plays a major role to capture solar energy, through photosynthesis. The energy stored in organic carbon molecules further supports the full aquatic ecosystem. The biotechnological potential of phytoplankton (defined by microalgae and cyanobacteria), has been re-discovered in the last decade (Chisti, 2007). Microalgae and cyanobacteria are cultivated mainly for human or animal nutrition, cosmetics and pharmaceutics (Williams & Laurens, 2010).Indeed, they are recognized for the extraordinary diversity of molecules they can produce: proteins, lipids, vitamins, antioxidants, pigments (Wijffels and Barbosa, 2010). In particular, microalgae turn out to be an emerging and promising solution for future bioenergy, with major economic developments on the horizon of a decade. Without using arable land, they offer the possibility of producing third generation biofuels from industrial CO₂ sources while extracting their nutrients from recycled water. Algal biofuel productivity could be 10 to 30 times higher than that  currently achieved with oleaginous terrestrial plants. Research and investment in the microalgae industry have been greatly stimulated in the last decades.

Oceans and culturing process are very different but they also share some common points that were discussed during the workshop. In particular, the optical depth is very different but the cells close to the enlighten surface receive most of the solar light. Consequently, they become photo-saturated, or even photo-inhibited, both causing reduced biomass yield. But also they shadow the other cells beneath, which can even experiment darkness.

Similar models are used to predict CO₂ fluxes at global scale and organic carbon accumulation for phytoplankton grown in photobioreactors or high-yield ponds (raceways) (Bernard, 2011). A closer look at the processes for photon harvesting and for transmitting the energy to the Calvin cycle for capturing CO₂ illustrates the nonlinear and dynamical aspects inherent to photosynthesis. As a result, it is currently difficult to use a common model that would at the same time predict the impact of mixing on the growth rate or determine how phytoplankton will respond to the increase in average ocean temperature, according to IPCC scenarios. In addition, phytoplankton growth spans different time scales from light reaction, photosynthesis metabolism to pigment acclimation. These time windows for coping with suboptimal light, nutrient or temperature conditions are generally not all integrated into the existing biological models either dedicated to process optimization or to predict phytoplankton production in the ocean. Accurate models able to robustly represent outdoor production of phytoplankton are therefore still missing. Some models have been proposed, but they are hardly validated. In the most advanced studies, successful datasets where predictions of the models significantly correlate with records, only cover periods of weeks, and for a very limited set of variables, mainly biomass. Developing consistent phytoplanktonic models is an important issue to better assess the CO₂ fluxes, estimate the evolution of biodiversity (Demory et al. 2018) or to predict the potential of the phytoplankton-based processes. It can also dramatically contribute to optimize the production process, either by improving its design, or by supporting on-line monitoring and control leading the process closer to its theoretical potential.

The objective of the workshop was to discuss these existing mathematical models predicting photosynthesis under realistic conditions, with the idea of paving the way towards future developments to enhance model accuracy. Several coexisting models were therefore explored and discussed, for different purposes, and different phytoplanktonic growth conditions This is what makes the modelling of phytoplankton complicate, the model has to be tailored to case and the objectives for which it has to be designed. The most crucial points for developing such modelling approach were discussed, and illustrated with the individual experiences of the researchers issued from different fields (oceanography, bioprocess, algae physiology, modelling). The most striking points, that were clearly highlighted confronting the viewpoints from several domains are summarized hereafter. We believe that, by unifying our efforts, we can propose a coherent and realistic bundle of photosynthesis models that incorporates key aspects of phytoplankton photosynthesis.

It seems that current knowledge on photosynthesis becomes less consensual as soon as light intensity saturates the photosynthetic capability, triggering photoinhibition and photoprotection mechanisms. At the same time reactive oxygen species have a strong impact on cell metabolism and a combined set of acclimation processes to reduce the photon harvesting capability and to absorb the excess of energy is triggered. These mechanisms are more difficult to represent mathematically and the rationale behind (in terms of what the cell is optimising) is often not straightforward. Further developments in this domain must be carried out.

Photoacclimation, defined as pigment modification in response to light, is rarely accounted in the models while it can result to an error by a factor two in the growth rate estimation. When considering the cell physiology, it seems that photoacclimation is not optimal as a growth strategy, but clearly protects the cell by reducing photodamages. In the ocean as in a photobioreactor, the light spectrum is changing with depth and is  modulated by absorption and scattering. Red and blue lights are rapidly absorbed by concentrated biomasses. Modelling the role of light colour on photosynthesis does not seem to be too challenging and there are currently representations of growth rate as a function of light quality. Coupled with a simple spectral Lambert-Beer model, it can be an efficient option to represent photosynthesis with depth. However, including the regulation mechanisms in such models is more difficult. Indeed, it has been shown that several photoreceptors sense particular wavelength and trigger regulation mechanisms. Depending on the species, red and/or blue light are necessary especially for the cell to complete its full cycle. Other photoreceptors are involved in the photoacclimation process and can modify some of the pigments in the cell.

The question of modelling the growth rate in fluctuating light conditions is debatable. Recent models seem to be able to accurately represent the fluorescence under different light signals (Bernardi et al. 2017), including very shaky pattern. Growth rate has been proved to be explained by the Han model (Han, 2002) for periodic light signals. The key photosynthetic processes with time scales up to an hour (photoproduction, photoinhibition and photoregulation) are well described. When the light signal is not periodic, and results from the computation of light reconstructed with a Lagrangian approach, the prediction capability of the models is hindered more complicated phenomena are triggered. It is likely that the processes for dissipating energy interact with photosynthesis and mask the photosynthetic response. Much work remains to be carried out to understand these aspects.

Paradoxically, the effect of temperature on phytoplankton has not been studied as widely as light. In particular, the acclimation mechanisms when temperature are shifted are rarely described and have not been included in models. It results that there is a marked uncertainty when models are used to predict fluctuating temperatures in industrial processes or scenario of temperature increase to predict the impact of global warming. On top of acclimation, the temperature adaptation mechanisms triggered at long time scale must also be accounted for. It seems evident that experimental work must clarify the ability of different species to adapt and determine the temperature conditions for which they can survive. Acclimation and adaptation when temperature is fluctuating must also be more attentively considered. These questions become of outmost importance for the biotechnological processes as in the natural environment. In particular, we must rapidly understand how phytoplankton can genomically evolve in such complex and changing environment. This is the cornerstone of a prediction of the evolution of the marine ecosystem in a warming ocean. In particular, quantifying the way pigments are responding to temperature changes is crucial. These aspects have been deeply investigated for the response to light but stay poorly described with temperature.

Algal biofilms are emerging for the production of proteins, wastewater treatment and are a promising solution for green chemistry, including biofuel. Modelling algal biofilm growth is still challenging and the available studies are presently rare. The structure of the biofilm plays a key role in its activity, and it is triggered by the limiting nutrients or by the excess of oxygen. Water can paradoxically limit photosynthesis, as pointed out by some models. The developments in photosynthesis modelling for planktonic cultures must benefit to biofilm, with some specificities due to the light and chemical gradients which take place in the biofilm.

Representing the metabolism of bacteria or yeast has been extensively studied, validated and used in the models over the past decades. Metabolic and genomic knowledge for microalgae is much more recent, partial and limited to a few species (Baroukh et al. 2015). Beyond the reconstruction of the metabolism of these phototrophic microorganisms from the knowledge of their genomes, identifying their ability to adapt to the strong temporal variability of their environment remains a limit to the reconstruction of metabolic networks. An associated challenge is that the cells are rarely in a steady state. Traditional approaches relying on the balanced growth assumptions and metabolic equilibrium are thus inadequate for phytoplankton. Bridging the gap between detailed metabolic models with hundreds of parameters and macroscopic models based on limited information and reduced calibration is still challenging. There is also a trend to consider the metabolism of the holobionte embedding the metabolic networks of different actors of the ecosystem. This natural biodiversity, with assemblages of phytoplankton species, bacteria, viruses, etc. whose overall effect on productivity (in the ocean as in industrial systems) is misidentified (Cardinale et al., 2011). The shift in complexity is tremendous, but at the same time it provides a global picture and offers explanations to symbiotic behaviours and overyielding resulting from the association of different organisms. Evaluating and representing the positive or negative influence of the natural flora on productivity and overall photosynthetic efficiency is a delicate topic for which the metabolic and genomic tools developed in the framework of single-species populations must be extended. Accounting for the interactions between microorganisms is crucial for both oceanography (improving predictions by better accounting for symbioses in the ecosystem) or for industrial purpose (with the development of micro-permaculture). An ecosystem vision based on functions in the ecosystem more than organism-centred is probably more flexible and more adapted. The new generation of photosynthetic models must now target the whole microbial community and better represent the interactions between the key actors of the ecosystem.

An important challenge of photosynthesis modelling is to find biological models that can be integrated into a 3D physical scheme (in photobioreactors, biofilms as in the ocean) able to represent in a relevant way the different time scales involved. The modelling needs are different in oceanography and bioprocess, but in both cases, laboratory experiments have to be extrapolated to a much larger scale (process scale, i.e. hectares to ocean scale, ie billion km²) and account for the microbial communities growing together with phytoplankton. Many modelling works underestimated the difficulty to extrapolate from lab-scale records to large-scale outdoor productivity under fluctuating climate conditions. It is a difficult issue, since primary production (or large-scale industrial production) are complex processes resulting from a nonlinear combination of different factors with intricate impacts.

Calibrating these nonlinear models of high dimension, for non-linear complex dynamical systems accounting for multiple possible factors (light, temperature, nutrients) is very challenging. Conventional methods are often inadequate, and we must develop new approaches dedicated to these forced nonlinear systems (Bernard, 2011). Experiences from different fields must be shared and mutualised. It would probably be illusory to identify, from experimental data, the whole set of parameters of a mechanistic model coupling hydrodynamics and metabolism. Similarly, to optimize the functioning of these systems, we must first develop dedicated methods to reduce these models to coarser grains models which are more tractable analytically.

 

Communications 

• Introduction à la photosynthèse (A. Sciandra, CNRS, Laboratoire d’Océanographie de Villefranche sur Mer)
• Introduction aux mécanismes de photoacclimatation (A. Sciandra, CNRS, LOV)
• Modeling the Links between Fluorescence and Growth (B. Chachuat, Imperial College London)
• Comment représenter les limitations multiples ? (M. Baklouti, MIO, Univ. Aix-Marseille, O. Bernard, Biocore, Inria, G. Bougaran, PBA, IFREMER, A. Sciandra, CNRS, LOV)
• Modélisation du Quenching Non Photochimique (NPQ) (F. Mairet, Ifremer, Laboratoire PBA, Nantes)
• La photosynthèse dans les modèles métaboliques (C. Baroukh, IRHS, INRAE)
• Modeling photoacclimation in microalgae and cyanobacteria (O. Bernard, Biocore, Inria)
• La colimitation N-P sous forte dépendance ? (G. Bougaran, PBA, IFREMER)
• Effet de la couleur de la lumière (F. Mairet, Ifremer PBA, C. Martinez, Biocore, Inria, O. Bernard, Biocore, Inria)
• Les processus de dissipation de l’énergie chez les microalgues (T. Lacour, PBA, IFREMER)
• Comparaison des outils de mesure de la production primaire : une question d’écophysiologie (T. Lacour, Laboratoire PBA, IFREMER)
• Fast photosynthesis scales. Is everything understood? (O. Bernard, Biocore, Inria, Chachuat, Imperial College London)
• Impact of temperature on phytoplankton growth rate (O. Bernard, Biocore, Inria, D. Demory, Georgia Institute of Technology, Atlanta)
• Light limitation in high density microalgae cultures (C. Martinez, Biocore, Inria, F. Mairet, PBA, IFREMER)
• On multiple limitation Multiplication vs minimum (O. Bernard, Biocore, Inria)
• Représentation de la production primaire dans les modèles biogéochimiques de l’océan (M. Baklouti , MIO, Univ Aix-Marseille)
• Photosensing ? (O.Bernard, Biocore, Inria, F. Mairet, PBA, IFREMER)
• Microalgae-Bacterial interactions (E. Fouilland, CNRS, Marbec)
• Interaction phytoplankton-predator ( D. Demory, Georgia Institute of Technology, Atlanta)
• Microalgae-Microalgae Interactions (E. Fouilland, CNRS, Marbec; W. Djema, Biocore, Inria)
• Microalgal-based biofilm cultivation systems (H. Bonnefond, Inalve)
• La photosynthèse dans les biofilms: de la biologie au modèle (F. Lopes, LGPM, CentraleSupélec, A. Fanesi, LGPM, CentraleSupélec)
• Modélisation de la croissance des biofilms photosynthétiques (B. Polizzi, Univ. Besançon, M. Ribot, Univ. Angers)
• Meta-organisms, and functional diversity (D. Eveillard, COMBI, Univ. Nantes, A. Regimbeau COMBI, Univ. Nantes)

 

Bibliography

Baroukh, C., Muñoz-Tamayo, R., Steyer, J. P., & Bernard, O. (2015). A state of the art of metabolic networks of unicellular microalgae and cyanobacteria for biofuel production. Metabolic engineering, 30, 49-60.

Baroukh, C., Muñoz-Tamayo, R., Steyer, J. P., & Bernard, O. (2014). DRUM: a new framework for metabolic modeling under non-balanced growth. Application to the carbon metabolism of unicellular microalgae. PloS one, 9(8), e104499.

Bernard,  O.,“Hurdles and challenges for modelling and control of microalgae for CO2 mitigation and biofuel production,” Journal of Process Control, vol. 21, no. 10, pp. 1378–1389, 2011.

Bernard, O., Mairet, F., & Chachuat, B. (2015). Modelling of microalgae culture systems with applications to control and optimization. In Microalgae Biotechnology (pp. 59-87). Springer, Cham.

Bernardi, A., Nikolaou, A., Meneghesso, A., Chachuat, B., Morosinotto, T. and Bezzo, F., 2017. Semi-empirical modeling of microalgae photosynthesis in different acclimation states–Application to N. gaditana. Journal of biotechnology, 259, pp.63-72.

Chisti, Y. (2007). Biodiesel from microalgae. Biotechnology Advances, 25:294– 306.

Demory, D., Baudoux, A. C., Monier, A., Simon, N., Six, C., Ge, P., … Bernard, O., Rabouille, S. (2018). Picoeukaryotes of the Micromonas genus: sentinels of a warming ocean. The ISME journal, 1.

Droop, M. R. (1983). 25 years of algal growth kinetics, a personal view. Botanica marina, 16:99–112.

Falkowski, P. G., & Raven, J. A. (2013). Aquatic photosynthesis. Princeton University Press.

Mairet, F., O. Bernard, P. Masci, T. Lacour, and A. Sciandra, “Modelling neutral lipid production by the microalga Isochrysis affinis galbana under nitrogen limitation,” Biores. Technol., vol. 102, pp. 142–149, 2010.

Tebbani, S., Filali, R., Lopes, F., Dumur, D., & Pareau, D. (2014). CO₂ Biofixation by Microalgae: Modeling, estimation and control. John Wiley & Sons.

Wijffels, R. H. and Barbosa, M. J. (2010). An outlook on microalgal biofuels. Science, 329(5993):796–799.

Williams PJB, Laurens LML. Microalgae as biodiesel & biomass feedstocks: Review& analysis of the biochemistry, energetics & economics. Energy and Environ Sci. 2010;3:554–590

 

---

---