Does vascular plant encroachment affect parameter importance for modelling carbon dioxide fluxes in temperate raised bogs?

Undisturbed raised bogs are characterised by permanent water saturation which prevents decomposition of peat, limits the spread of vascular plants and makes the ecosystem a peat moss dominated sink of carbon. Lower water levels e.g. due to climate change or drainage endanger ecosystem functions and lead to an altered vegetation composition. In particular, encroaching vascular plant species are a growing thread to natural or restored peatlands in Central Europe. While previous work often focussed on how greenhouse gas emissions or ecosystem functions of raised bogs respond to changing environmental conditions, the impact of encroaching vascular plants is only sparsely covered. As process-based SVAT models are able to simulate different vegetation compositions and their respective water and carbon fluxes, they are an optimal tool to answer this relevant question. However, this requires the relevant processes of a vegetation shift impacting the system’s water and carbon relations to be correctly implemented.

Models capable of simulating bryophyte processes are sparse. We use the process-based SVAT model 'pyAPES' including a bryophyte layer, developed and tested for boreal peatlands, at a bog site in northern Germany. The overall objective of this study is to identify whether pyAPES is able to simulate carbon fluxes of a temperate raised bog under a changing vegetation composition. We addressed four research questions: (i) Does model parametrization need to be changed when using pyAPES for temperate conditions? (ii) How do these changes affect modelled gross primary production (GPP) and ecosystem respiration (R_eco)? (iii) Which photosynthetic and respiratory parameters are most crucial for model performance? (iv) Does the order of crucial parameters depend on vegetation composition and on environmental conditions?

We answer these questions by calibrating pyAPES to measured soil temperature (T_s), water table depth (WTD), seasonal dynamics of vascular plant leaf area index (LAI) and GPP and R_eco fluxes. Morris sensitivity analysis (MSA) is conducted with the calibrated model to investigate parameter impacts on modelled GPP and R_eco.

Preliminary results show that pyAPES performs well for a temperate raised bog after adaptation of the model parameters. Most important parameters for calibrating pyAPES were parameters of the unimodal Van-Genuchten-Mualem water retention model for both moss and peat and Farquhar parameters, which are sparse in literature.

MSA is conducted for GPP and R_eco using annual sums. Boundaries for MSA are set to ± 20% around initial parametrisation in order to derive a standardized rank of parameter importance as well as to observed boundaries from literature to cover the whole range of possible site conditions. Subsequent analysis will give evidence whether meaningful parameter inference with respect to ecosystem carbon fluxes is possible under different site conditions.

Further, we investigate the impact of changing vascular plant LAI and moss biomass due to encroachment on ecosystem carbon fluxes by applying full factorial parameter combinations inferring possible shifts in model sensitivities. In a last step, we investigate intraannual shifts of sensitivities in half-hourly resolution to assess the impact of dynamic environmental conditions like WTD and moss surface temperature.