LandS: Vegetation modeling based on Ellenberg's ecological indicator values

Abstract

We present LandS, a new version of the Gras Model. The Gras Model was designed to simulate grassland development at local scales based on Ecological Indicator Values (EIVs) for different grassland management practices. In LandS, we complemented the existing set of EIVs with a second set representing several environmental factors: light, moisture, temperature, soil pH and nitrogen, also known as Ellenberg's EIVs. These new EIVs make the model more versatile and applicable to a wide range of sites across Central Europe. For example, it can be used on sites with dry or moist, acidic or calcareous soils in grassland or forest environments. We have also improved the implementation of the model by introducing version control and moving species and site-specific variables to data input files, so that species sets can be easily swapped for application in new study sites. We demonstrate the use and behavior of the model in two simulation experiments exploring interactions mediated by Ellenberg's EIVs, using input files to apply the model to different landscapes. We also provide detailed guidance on species selection and calibration, and discuss model limitations.•LandS is an improved version of the GraS Model for simulating vegetation development at the local scale.•It includes Ellenberg-like indicator values for environmental variables for inverse prediction of species occurrence and composition.•The model is now flexible enough to be used for study sites throughout Central Europe, using data input files for species initialization.

Keywords: LandS Model (Landscape Succession Model); Modeling environmental conditions; Plant sociology; Semi-natural grassland; Species composition; Vegetation development.

Graphical abstract

Fig. 1

General model concept.

Fig. 2

Processes and scheduling. Process names correspond to those of the submodels described in detail in Section ODD7.

Fig. 3

An example decision tree for deriving vegetation types from species composition, the name of species is referring to their simulated cover, first published by Siehoff et al. .

Fig. 4

Screenshots showing parts of the output GUI of two different simulation runs: at the start of the calibration process (left) and after calibration (right) with the dynamic equilibrium of the examined community.

Fig. 5

Control functions for site conditions (A) and land use (B) at different species EIVs. (A) Control functions are calculated here for regular species EIVs with a slope of 1, and for indifferent species EIVs with a reduced slope of 0.3 and a reduced optimum of 0.6. (B) There are no indifferent species for the land use.

Fig. 6

Patchy test landscape. Different patch types: (0) surrounding open landscape; (1) dry, acidic and nutrient-poor grassland; (2) dry, calcareous and nutrient-poor grassland; (3) wet floodplain grassland; (4) acidic and nutrient-poor moorlands; (5) alkaline/calcareous and nutrient-poor moorlands; (6) woodland edge; (7) nonspecific woodland; (8) acidic and nutrient-poor woodland; (9) alkaline and nutrient-poor woodland; (10) alkaline and nutrient-rich woodland; for further information on patch types with site EIVs see Table 13.

Fig. 7

Average species cover in each patch type (consisting of 4 cells) over time. Different patch types: (0) surrounding open landscape; (1) dry, acidic and nutrient-poor grassland; (2) dry, calcareous and nutrient-poor grassland; (3) wet floodplain grassland; (4) acidic and nutrient-poor moorlands; (5) alkaline/calcareous and nutrient-poor moorlands; (6) woodland edge; (7) nonspecific woodland; (8) acidic and nutrient-poor woodland; (9) alkaline and nutrient-poor woodland; (10) alkaline and nutrient-rich woodland; for further information on patch types with site EIVs see Table 13 and for species characteristics see Table 14.

Fig. 8

Average species cover in the test plot. Calibrated equilibrium states of the grassland community in the fertilized (A) and unfertilized form (B) and the simulated transition from fertilized to unfertilized after the initial calibration (C).

Fig. 9

Cover development of Arrhenatherum elatius in each of the 100 cells of the test plot for the fertilized meadow after the final calibration.