Dynamic, mechanistic, molecular‐level modelling of cyanobacteria: Anabaena and nitrogen interaction

Abstract

Phytoplankton (eutrophication, biogeochemical) models are important tools for ecosystem research and management, but they generally have not been updated to include modern biology. Here, we present a dynamic, mechanistic, molecular-level (i.e. gene, transcript, protein, metabolite) model of Anabaena - nitrogen interaction. The model was developed using the pattern-oriented approach to model definition and parameterization of complex agent-based models. It simulates individual filaments, each with individual cells, each with genes that are expressed to yield transcripts and proteins. Cells metabolize various forms of N, grow and divide, and differentiate heterocysts when fixed N is depleted. The model is informed by observations from 269 laboratory experiments from 55 papers published from 1942 to 2014. Within this database, we identified 331 emerging patterns, and, excluding inconsistencies in observations, the model reproduces 94% of them. To explore a practical application, we used the model to simulate nutrient reduction scenarios for a hypothetical lake. For a 50% N only loading reduction, the model predicts that N fixation increases, but this fixed N does not compensate for the loading reduction, and the chlorophyll a concentration decreases substantially (by 33%). When N is reduced along with P, the model predicts an additional 8% reduction (compared to P only).

The response of nitrogen fixing cyanobacteria to a reduction in nitrogen loading

“…Crucially, we expect that while it may be possible for Nostocales to fix sufficient N for their own requirements, they will not in general fully offset large reductions in N input to a lake and therefore total N available for production will fall. Hellweger et al () came to similar conclusions, and their model based results could be validated using a similar but longer duration experimental setup to ours. Importantly, we do not advocate relaxing control of P, rather that dual control of N and P may have benefits over controlling P alone.…”

“…There have also been experimental studies in the Neuse River Estuary and in lake Taihu in which reduction of N loading did not promote, or only marginally promoted, an increase in Nostocales abundance (Paerl et al, ; Piehler, Dyble, Moisander, Pinckney, & Paerl, ). In a dynamic, mechanistic, molecular‐level model of N 2 ‐fixing cyanobacteria in a hypothetical lake, reduction of N loading by 50% was only partly compensated for by fixation and led to a 33% reduction in chlorophyll a (Hellweger et al, ).…”

From Genes to Ecosystems in Microbiology: Modeling Approaches and the Importance of Individuality

“…A model of Anabaena —nitrogen interactions was developed by Hellweger et al ( 2016b ). This model simulates the uptake of various forms of nitrogen and early intracellular assimilation pathways.…”

The effectiveness of cyanobacteria nitrogen fixation: Review of bench top and pilot scale nitrogen removal studies and implications for nitrogen removal programs

“…Indeed, limnologists and managers consider P management to be of paramount importance. However, not all P management programs have successfully mitigated cyanobacterial blooms, at least in the short term (Schindler et al 2016), and several authors have questioned the predominance of the P management paradigm, arguing that dual nitrogen (N) and P controls (i.e., supplementing P controls with N controls) would be more effective at preventing cyanobacterial blooms than P controls alone (Lewis and Wurtsbaugh 2008;Paerl 2009;Scott and McCarthy 2010;Paerl et al 2011;Hellweger et al 2016). In many cases, this means adding N controls to existing point source P control programs.…”

“…The most valid proof of concept test in this case is direct N removal from experimental systems because it most closely resembles N control programs at a watershed scale. Inferences from relatively indirect evidence not employing N removal, such as descriptive correlations of field observations, N addition bioassays, computational modeling, etc., on the hypothesized effectiveness of N controls(Lewis and Wurtsbaugh 2008;Paerl 2009;Scott and McCarthy 2010;Paerl et al 2011;Hellweger et al 2016) can be useful in the early stages of developing an idea, but the risk of misinterpreting indirect evidence is much larger than the risk of misinterpreting direct tests of N removal. Therefore, direct N removal tests should be seen as critical elements of the decision making process because they lower uncertainty.To what extent do N 2 fixing cyanobacteria adapt to loss of N?…”