Equifinality, sloppiness, and emergent structures of mechanistic soil biogeochemical models

Marschmann, Gianna; Pagel, Holger; Kügler, Philipp; Streck, Thilo

doi:10.1016/j.envsoft.2019.104518

Cited by 38 publications

(37 citation statements)

References 90 publications

(116 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Alternatively, one could allow for independent variations of several traits as a function of nutrient availability (Sistla et al, 2014, where modes (i), (ii), and (iii) are implemented), but at the cost of additional parameters to estimate via data fitting. Since even the simplest microbial models can be overparameterized and exhibit equifinality issues (Marschmann et al, 2019), we preferred to implement one mechanism at a time and compare model fitting across our four variants. In fact, even with our minimal calibration approach-and in contrast with our intuitive Berg and McClaugherty (1989), green from Osono andTakeda (2004, 2005), gray from Osono (2017), and blue from Hirobe et al (2004); the 1:1 line is depicted as a solid gray line.…”

Section: Discussion Putative Resource Use Modes Emerging From Earlier Modelsmentioning

confidence: 99%

Modeling Microbial Adaptations to Nutrient Limitation During Litter Decomposition

Manzoni

Chakrawal

Spohn

et al. 2021

Front. For. Glob. Change

View full text Add to dashboard Cite

Microbial decomposers face large stoichiometric imbalances when feeding on nutrient-poor plant residues. To meet the challenges of nutrient limitation, microorganisms might: (i) allocate less carbon (C) to growth vs. respiration or excretion (i.e., flexible C-use efficiency, CUE), (ii) produce extracellular enzymes to target compounds that supply the most limiting element, (iii) modify their cellular composition according to the external nutrient availability, and (iv) preferentially retain nutrients at senescence. These four resource use modes can have different consequences on the litter C and nitrogen (N) dynamics–modes that selectively remove C from the system can reduce C storage in soil, whereas modes that delay C mineralization and increase internal N recycling could promote storage of C and N. Since we do not know which modes are dominant in litter decomposers, we cannot predict the fate of C and N released from plant residues, in particular under conditions of microbial nutrient limitation. To address this question, we developed a process-based model of litter decomposition in which these four resource use modes were implemented. We then parameterized the model using ∼80 litter decomposition datasets spanning a broad range of litter qualities. The calibrated model variants were able to capture most of the variability in litter C, N, and lignin fractions during decomposition regardless of which modes were included. This suggests that different modes can lead to similar litter decomposition trajectories (thanks to the multiple alternative resource acquisition pathways), and that identification of dominant modes is not possible using “standard” litter decomposition data (an equifinality problem). Our results thus point to the need of exploring microbial adaptations to nutrient limitation with empirical estimates of microbial traits and to develop models flexible enough to consider a range of hypothesized microbial responses.

show abstract

Section: Discussion Putative Resource Use Modes Emerging From Earlier Modelsmentioning

confidence: 99%

Modeling Microbial Adaptations to Nutrient Limitation During Litter Decomposition

Manzoni

Chakrawal

Spohn

et al. 2021

Front. For. Glob. Change

View full text Add to dashboard Cite

show abstract

“…HydroLorica is a reduced-complexity model, which means that it simulates the most important processes affecting soil and landscapes using simplified process descriptions. Reducing model complexity promotes critical evaluation of essential processes, reduces calculation time and prevents extensive data requirements and over-120 parameterization (Hunter et al, 2007;Kirkby, 2018;Marschmann et al, 2019;Snowden et al, 2017;Temme et al, 2011).…”

Section: Model 115mentioning

confidence: 99%

“…Only when landscape are stable, flat and free of trees, changes in soil properties are not influenced by changes in terrain. In such settings, a 1D soil profile evolution model would suffice to simulate soil development in different landscape positions (Finke, 2012;Minasny et al, 2015). When rates of geomorphic processes far exceed those of 530 pedogenic processes, for example in tillage-dominated systems, a landscape evolution model would suffice (e.g.…”

Section: 223mentioning

confidence: 99%

“…Soil evolution models simulate a range of physical, chemical and biotic processes that affect the properties of soils through space and time (Minasny et al, 2015;Vereecken et al, 2016). Such models have been developed for a range of scales, varying from 1D soil profiles to 3D soil landscapes (Finke, 2012;Minasny et al, 2015;Temme and Vanwalleghem, 2016). One-dimensional soil profile models generally provide a high level of detail and process coverage, but they lack the simulation of essential feedbacks and interactions that can occur between soils on a landscape scale (van der Meij 80 et al, 2018).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Modelling soil and landscape evolution – the effect of rainfall and land use change on soil and landscape patterns

Meij¹,

Temme²,

Wallinga³

et al. 2019

Preprint

View full text Add to dashboard Cite

Abstract. Humans have substantially altered soil and landscape patterns and properties due to agricultural use, with severe impacts on biodiversity, carbon sequestration and food security. These impacts are difficult to quantify, because we lack data on long-term changes in soils in natural and agricultural settings and available simulation methods are not suitable to reliably predict future development of soils under projected changes in climate and land management. To help overcome these challenges, we developed the HydroLorica soil-landscape evolution model, that simulates soil development by explicitly modelling the spatial water balance as driver of soil and landscape forming processes. We simulated 14500 years of soil – formation under natural conditions for three scenarios of different rainfall inputs. For each scenario we added a 500-year period of intensive agricultural land use, where we introduced tillage erosion and changed vegetation type. Our results show substantial differences between natural soil patterns under different rainfall input. With higher rainfall, soil patterns become more heterogeneous due to increased tree throw and water erosion. Agricultural patterns differ substantially from the natural patterns, with higher variation of soil properties over larger distances and larger correlations with terrain position. In the natural system, rainfall is the dominant factor influencing soil variation, while for agricultural soil patterns landform explains most of the variation simulated. The cultivation of soils thus changed the dominant factors and processes influencing soil formation, and thereby also increased predictability of soil patterns. Our study highlights the potential of soil-landscape evolution modelling for simulating past and future developments of soil and landscape patterns. Our results confirm that humans have become the dominant soil forming factor in agricultural landscapes.

show abstract

“…Models are evaluated based on their ability to replicate measured soil respiration (both from incubation and field measurements). To reduce any biases with model fitting or model equifinality (Christiansen, 2018;Marschmann et al, 2019) we evaluate a range of parameter estimation approaches and data types.…”

Section: Introductionmentioning

confidence: 99%

Evaluating an exponential respiration model to alternative models for soil respiration components in a Canadian wildfire chronosequence (FireResp, v1.0)

Zobitz

Aaltonen

Zhou

et al. 2021

Preprint

View full text Add to dashboard Cite

Abstract. Forest fires modify soil organic carbon and suppress soil respiration for many decades since the initial disturbance. The associated changes in soil autotrophic and heterotrophic respiration from the time of the forest fire however, is less well characterized. We analyzed models of soil autotrophic and heterotrophic respiration with a novel dataset across a fire chronosequence in the Yukon and Northwest Territories of Canada. The dataset consisted of soil incubation experiments and field measurements of soil respiration and soil carbon stocks. The models ranged from a Q10 (exponential) model of respiration to models of heterotrophic respiration using Michaelis-Menten kinetics parameterized with soil microbe carbon. For model evaluation we applied model selection metrics (Akaike Information Criterion) and compared predicted patterns in soil respiration components across the chronosequence. Parameters estimated with data from the 5 cm soil depth had better model-data comparisons than parameters estimated with data from the 10 cm soil depth. The model-data fit was improved by including parameters estimated from soil incubation experiments. Models that incorporated microbial carbon with Michaelis-Menten kinetics reproduced patterns in autotrophic and heterotrophic soil respiration components across the chronosequence. Autotrophic respiration was associated with aboveground tree biomass at more recently burned sites, but this association was less robust at older sites in the chronosequence. Our results provide support for more structured soil respiration models than standard Q10 exponential models.

show abstract

Equifinality, sloppiness, and emergent structures of mechanistic soil biogeochemical models

Cited by 38 publications

References 90 publications

Modeling Microbial Adaptations to Nutrient Limitation During Litter Decomposition

Modeling Microbial Adaptations to Nutrient Limitation During Litter Decomposition

Modelling soil and landscape evolution – the effect of rainfall and land use change on soil and landscape patterns

Evaluating an exponential respiration model to alternative models for soil respiration components in a Canadian wildfire chronosequence (FireResp, v1.0)

Contact Info

Product

Resources

About