Department for Ecoinformatics, Biometrics and Forest Growth

Faculty of Forest Sciences and Forest Ecology
at the
University of Göttingen

German version of this page / Deutsche Version dieser Seite

Publications of the research group

Projects

Announcements

Winfried Kurth, mathematician and computer scientist

Research focus: Further development of formal languages and software tools for dynamical 3D plant models in the framework of ecosystem research (software GROGRA), methods for the evaluation and formalization of quantitative morphological data, computer graphics
Individual homepage: http://www.uni-forst.gwdg.de/~wkurth
e-mail: wk((at)) informatik.uni-goettingen.de

Former members of staff:

Helge Dzierzon, forest scientist

Research focus: Forest stand modelling, statistical data analysis, comparisons between tree models, tools for discretisation and analysis of forest trees and stands
Individual homepage: http://www.uni-forst.gwdg.de/~wkurth/memb_hd.html
e-mail: hdzierz((at)) gmail.com

Ulrike Singer, forest scientist

Research focus: 3D visualisation of forest stands and landscapes, object-oriented modelling, graphical interfaces
Individual homepage: http://www.uni-forst.gwdg.de/~wkurth/memb_us.html
e-mail: usinger((at)) uni-forst.gwdg.de

Michael Schulte, biologist

Research focus: Hydraulic architecture, gas exchange and stomatal control in trees, process-oriented modelling and model validation, application and further development of the software HYDRA
Individual homepage: http://www.uni-forst.gwdg.de/~wkurth/memb_ms.html
e-mail: mi-schu((at)) t-online.de

Gustavo Anzola, forest scientist

Research focus: Linkage of different structural plant models and process-oriented models (AMAP, HYDRA, HYDRO, GROGRA), inclusion of stress factors in 3D growth models
Individual homepage: http://www.uni-forst.gwdg.de/~wkurth/memb_ga.html
e-mail: ganzola((at)) web.de

The group is participant of the Forest Ecosystem Research Centre Göttingen.

Cooperation partners:

Programme Modélisation des Plantes (AMAP team) at the CIRAD (Montpellier, France)

LIGNUM team at the Finnish Forest Research Centre METLA (Finland)

Department of Growth and Yield at the Forest Research Station of Lower Saxony, Göttingen

Chair for Forest Growth Modelling, Technical University of Munich

Department of ecology and ecosystem research at the Albrecht-von-Haller-Institute, University of Göttingen, Prof. Dr. Christoph Leuschner

IPK Gatersleben, Dr. Gerhard Buck-Sorlin

Center for Environmental Systems Research, University of Kassel

Institute of Forest Ecology at the University of Brno (Czech Republic), Prof. Dr. Jan Cermák.

Further ecosystem research centres in Germany:

PZÖ Kiel

BITÖK Bayreuth

UFZ Leipzig

GSF Munich

***********************************************************************

Plant Architecture Information Systems (PAIS)

Model dokumentation database UFIS

Environment, Forestry, Global Change, Simulation etc. (by Michael Sonntag)

***********************************************************************

Artificial Life:

German AL workshops

Santa-Fe Institute

Jan T. Kim's link page with AL links

Plant modelling in the narrower sense:

Team of Przemyslaw Prusinkiewicz (Calgary)

Team of Peter Room (Queensland) (there are also further links, literature and software)

AMAP

The rule-based language of "stochastic sensitive growth grammars" as an extension of parametric L-systems (Prusinkiewicz & Lindenmayer 1990) was developed to describe algorithmically the changement of the morphology of forest trees in time, taking endogenous and exogenous factors into account, and to create systematically 3-dimensional simulations of tree crowns. At different tree species, mainly at spruce, morphological measurements were carried out to get a basis for the design and parameterization of such rule systems.

The software GROGRA (Growth Grammar Interpreter; Kurth 1994) creates time series of 3-dimensional crown structures from the rules; the basic elements of these structures (annual shoots) can additionally bear non-geometrical attributes. Furthermore, GROGRA contains several analysis tools and data interfaces.

The generated architectures serve as an "ecomorphological basis model" for different process-oriented simulation models. There is already realized a model of tree-internal water flow (HYDRA; Früh 1995), based on the artificial tree structures.

The project is associated with the Forest Ecosystem Research Centre, University of Göttingen. It is financed by the DFG (Deutsche Forschungsgemeinschaft, German association for the advancement of science) in the form of several subprojects.

Keywords: plant architecture, architectural plant model, 3D plant model, virtual plant, Lindenmayer system, L system, L-system, growth grammar, morphological model, plant morphological development, branching architecture, tree architecture, virtual tree, 3D tree model, structural plant model, functional-structural model, FSTM, sensitivity, sensitive L-system, structure-process-linkage, plant growth simulation, tree hydraulic architecture, tree water flow, GROGRA, HYDRA

MOTIVATION, AIMS OF RESEARCH:

The development of the central-European forests during the next decades will be influenced on the one hand by the increasingly fostered "close-to-nature" forms of silviculture, on the other hand by various unplanned anthropogenous impacts (nutrient depositions, groundwater depletion, possible climatic change). The classical yield tables, which were developed from long-term measurements and were tailored to conventional treatment practices, loose their reliability under these circumstances (Spellmann 1991). Thus, there is now at first a need for a causally oriented understanding of the interrelations of forest trees with their environment. The forest-damage research of the last two decades, too, arrived at gaps of knowledge in the foundations of its research area (e.g., Abschlußdokumentation... 1993).

The state of theory formation in tree physiology and in silvicultural growth studies manifests itself - like in other scientific areas - in the models, which describe the relations in precise mathematical terms (and today find often their algorithmic realization in the form of computer simulations). Besides the purely empirical-deductive stand models, which are obviously insufficient for causal insights, there are currently two much-discussed directions of development in plant modelling:

(1) Process-oriented models which quantify the amount and dynamics of the pools of C, H₂O, N, and energy (e.g. McMurtrie & Wolf 1983, Mäkelä 1986, Bossel & Schäfer 1989, Sloboda & Pfreundt 1989, Bossel 1996).

(2) Structure-oriented models which reproduce the 3-dimensional tree and stand architecture on a morphological basis and thus make available a common structural matrix for light interception, flows of C, H₂O, N, and for mechanical processes (Reffye 1979, Bell 1986, Prusinkiewicz & Lindenmayer 1990, Kurth 1994).

However, to obtain progress in understanding tree growth and its control, it is necessary to consider the relations between branching structure and processes and hence to unify both approaches of development. It has to be recognized that the manipulation of the 3-dimensional structure of stands is the main way of control in forestry!

In fact, since about 10 years one can recognize in ecophysiological and ecosystemic modelling an orientation towards approaches which join the processes with the morphology of trees and tree stands. For different categories of processes, a gradual increase in difficulty results:

(1) Processes which occur - despite being determined by stand structure - outside the trees or at the boundary zone tree - environment: water flow in the soil, light interception. These are describable in terms of physics and hence to be modelled most easily (in principle).

(2) Processes in the interior of the trees which are still to a great extent describable in terms of physics. For the water flow in the tree, which belongs to this category, meanwhile the first whole-tree models were developed (Tyree 1988, Früh 1995); however, they do not yet reflect the close mutual dependence between the hydraulic systems of tree and soil.

(3) Processes in the trees which are mainly biologically determined, like carbon allocation and growth. Here, also, first modelling approaches which respect the relation to structure were developed (Ford & Ford 1990, Perttunen et al. 1996, Deleuze 1996), however, generally here only a very laborious progress in the theoretical synthesis of the various empirical material is to be expected, and this depends, after all, also on progress concerning the process categories (1) and (2).

Processes of the categories (2) and, even more, (3) are subject to biological control mechanisms which can only be understood on the level of the individual, i.e. of the organism (the tree). Thus, a purely stand-related consideration (even if structures are recognized) is no longer adequate here. Instead, for a causal understanding of processes of categories (2) and (3) and of their control, an individual-oriented research and modelling approach is well-suited, which finds applications also in other contexts of ecology (deAngelis & Gross 1992) and is supported on the side of computer science by object-oriented techniques (Breckling 1996). This approach is consistent with the above-mentioned structural plant models insofar, as - depending on the considered process and following the concept of modularity in botany (White 1979) - the tree itself can be dissolved into a hierarchical system of individual axes, growth units, internodes etc. (Rey et al. 1997), down to the meristems as the very entities where growth is to be localized and can be understood as the result of local influences as well as of biological rules in the hierarchical system "tree". Modular structural-functional models are thus capable to establish a bridge - built on firm theoretical knowledge - between process studies and the phenomenologically-oriented description level of botany (crown and shoot damage patterns, morphological vitality assessment: Roloff 1989).

This innovative approach in the area of ecophysiological theory formation and modelling was till today scarcely followed in Germany, and on an international level only by few workgroups. On December 12-13, 1996 the first "Workshop on Functional-Structural Tree Models" took place in Helsinki. The second FSTM Workshop (October 12-15, 1998, in Clermont Ferrand) attracted an even larger number of scientists from around the world. Here it became obvious that we have to do with a rapidly expanding, promising research area.

Despite of some recent progress, the models are still rather deficient concerning the integration of structure and processes, hence blocking the causal understanding:

• The models treat the water flows in the tree and in the soil still as isolated from each other. In the form of water uptake by the roots and in the dependence of stomatal conductivity from root water status we have, however, a close interdependence between the two nonlinear systems.
• In the models of light interception, there are certain discrepancies between approaches considering canopy layers, cubes, and plant organs, respectively. The corresponding problems of downscaling were made accessible by a secure theoretical analysis only recently (Knyazikhin et al. 1997).
• The coupling of several processes on the basis of structural information is realized only in very preliminary forms. This concerns especially the interrelation between water and carbon relations, the environmental control of tree-related processes, and the interaction of damaging and compensating mechanisms. Bassow et al. (1990) and Tyree & Alexander (1993) did already indicate this research gap.
• The use of process-related informations (intercepted light, water potential, mechanical stimuli) in the structure-generating growth models was only tested in very preliminary "test models" (Reffye et al. 1997, Kurth & Sloboda 1997). Progress is desirable here to organize the empirical knowledge and to help formulating new hypotheses to be tested in the experiment.
• There is a need for the development of flexible and powerful algorithms for the integration of structure and process.

Last not least, there is a lack of well-founded simplifications of the process models and growth models - simplifications which are based on a theoretically-founded knowledge of system behaviour and which can answer application-oriented questions: e.g. growth reactions to forest operations, relations between parameters of tree architecture and dynamics of water flow in the tree, light response curves of tree stands.

References:

Department for Ecoinformatics, Biometrics and Forest Growth
University of Göttingen
Büsgenweg 4, 37077 Göttingen
Germany

Phone +49-551-39-9715, -3464
Fax +49-551-39-3465

E-mail: wkurth((at)) informatik.uni-goettingen.de

Back to Ecoinformatics homepage

Back to Forestry Faculty homepage

To the homepage of the Chair for Graphics Systems at the University of Cottbus (where Winfried Kurth was active from 2001 until 2008)

To this website this disclaimer (in German) applies.