| title: | The composition and dynamics of Estonian forest communities under anthropogenic disturbance |
|---|---|
| reg no: | ETF5478 |
| project type: | Estonian Science Foundation research grant |
| subject: |
1.10-1.16. Bio-Geo Sciences |
| status: | accepted |
| institution: | TU Faculty of Biology and Geography |
| head of project: | Jaan Liira |
| duration: | 01.01.2003 - 31.12.2006 |
| description: | In order to describe structural dynamics in forest communities under human activity, firstly one has to define and describe 'core-types' for reference, and secondly, to develop discrimination methodology free of researcher subjectivity to differentiate disturbed/young communities from 'core-types'. The most popular forest site-type system by E. Lõhmus is half-subjective and the lack of original empirical data makes it unsuitable for statistical analyses. The evaluation methodology of forest natural authenticity should be based on empirical sample data, the analysis should be as universal as possible, independent of assumptions about distribution, easy to use and interpret. One of the most promising techniques is Artificial Neural Network Analysis, successfully implemented to discriminate natural and ruderal grasslands, and to evaluate natural quality of grasslands in Denmark. This type of analysis should be applied for forest communities to detect structural and compositional changes in forest communities, particularly considering increasing public demand for automated decision algorithms within the frame of inventory works. The field works will be carried out in four years in co-operation with other workgroups (e.g. J. Paal - 3-4 site-types) to cover most of the variability of Estonian forests (old-grown, young and disturbed). As the result (i) the revised and updated forest site-type classification-network, based on empirical field data, it will be flexible system for updates, useful reference database for new observations or communities, and non-subjective tool to evaluate natural quality of observed forests. (ii) The sample database will be used to generate regional species pool for forests in certain environmental conditions, and (iii) to describe ecological niche distributions of forest species. The results of the study (iv) about the general trends in dynamics of structure and composition of forests, and about indicator species are significant to evaluate natural quality of young and mature forest communities, and to improve forest protection policies. |
| project group | ||||
|---|---|---|---|---|
| no | name | institution | position | |
| 1. | Jaan Liira | TU Faculty of Biology and Geography | researcher | |