Keywords
How to Cite
Abstract
Introduction
Forest condition monitoring in Ukraine was initiated in the late 1990s as a part of the URIFFM research activities by methods harmonized with the UN-ECE ICP Forest I level methodology. Since 2000, forest condition monitoring has been conducted at the national level. It is an integral part of the state environmental monitoring system of Ukraine and is also ensured by Ukraine’s international obligations. According to the decision of the National Security and Defense Council of Ukraine dated 25.04.2013, it is necessary to optimize the environmental monitoring networks.
The aim of the study was to conduct a comparative analysis of the monitoring results for the full and sparse grids and to assess the level of representativeness of the sparse grid for the forest condition monitoring in Ukraine.
Materials and Methods
The data source was Ukrainian forest monitoring database with plots’ coordinates and survey results for years 2013 and 2015.
The existing forest monitoring grid (5 ? 5 km, full grid) covers all administrative regions of Ukraine. By means of Q-GIS a sparse grid (15 ? 15 km) was designed as a subsample of the full grid in Ukraine.
We compared the distribution of monitoring plots by forest condition types and tree species generally for Ukraine, as well as by natural zones by means of the ?2 method for both grids, and with the data of the forest fund database of Ukraine as of 2011. The ?2 method evaluated the coverage of areas, the tree species representation and the distribution of sample trees by standard defoliation classes (0 class (healthy trees, 0–10% defoliation), 1 class (11–25%), 2 class (26–60%), 3 class (61–99%), 4 class (100%)) for both grids.
Results
The I level forest monitoring grid covers all natural zones in Ukraine in proportion to the forested area. The full grid consists of 1,457 plots; the sparse grid includes 383 of them. Both grids cover 20 types of forest site conditions, the distribution of tree species does not significantly differ. At the same time, the total number of the monitoring plots, and therefore, actual expenditures, are reduced by more than 50%.
Both grids are representative of the Forest Steppe and Polissia (Forests) zones but are not representative for other natural zones (Steppe, Carpathian, and Crimea).
At the full grid 33,773 sample trees of 58 tree species are estimated annually. As for the sparse grid, the number of trees will be reduced to 8,890 (37 tree species) there. The 15 most represented species make 96.6% of all sample trees. The representation of tree species in the sparse grid is not significantly different from that in the full one.
Comparisons of tree species distribution by standard defoliation classes and age groups by the ?2 method showed that the sparse grid does not significantly differ from the full one for these groups (except for the beech tree). The reporting, which will be provided on the basis of monitoring data on the sparse grid, will accurately reflect the general condition of forests in Ukraine.
We found that the average defoliation values of most tree species and groups in the sparse grid were not significantly different from the full one (p = 0.05) (except Qurcus robur and Q. petrea, Fagus sylvatica, Abies alba and Robinia pseudoacacia). As the average defoliation is important in studying the dynamics, when implementing the sparse grid, the analysis will be performed for the plots which it includes only, and, accordingly, the differences (between the full and sparse grids) will not affect trends.
Conclusions
Designed sparse I level forest monitoring grid (15 ? 15 km) as a subsample of the full grid does not significantly differ from the latter in terms of its coverage of the natural zones, forest condition types and tree species composition. Observations at the sparse grid enable estimating the average defoliation rate, as well as standard reports on the distribution of sample trees by defoliation classes at the national level. The sparse grid usage allows reducing the actual cost of forest monitoring by more than 50%. The proposed sparse grid of monitoring plots can be used to optimize the density of forest monitoring grid currently applied in Ukraine. However, when taking a final decision on optimization, experts should bear in mind that the forest monitoring data in sparse grid is not representative for the Steppe, Carpathians and Mountain Crimea zones, and is totally representative for Forest-Steppe and Polissia. Therefore, when optimizing forest monitoring in the Steppe, Carpathians and Mountain Crimea natural zones it is recommended to keep the already existed density of monitoring plots.
2 Figs., 9 Tables, 23 Refs.