engleski

Responses of forest herbs to available understory light measured with hemispherical photographs in silver fir - beech forest in Croatia

Introduction Light conditions are very important determinant of forest understory vegetation pattern. To further examine that relation we investigated response of forest herbs to available light conditions in the forest of silver fir (Abies alba) and beech (Fagus sylvatica) which are most frequent altimontane forest on Croatian karst, using the hemispherical photographs for measuring available light. Our special interest was focused on the possibility of estimation light condition parameters (which are hardly gatherable at field) on the large areas using the remotely sensed data, as potential support to the spatial prediction of phytodiversity and particular plant species occurrences (Jelaska et al. 2003). All in all 151 circular plots with 25 meters diameter were surveyed, alongside with the ten meters diameter circular plot, positioned inside bigger ones. Data about the floristic assemblage were gathered in four equal sectors of the smaller circle, divided by cardinal points of the compass, and overall for large plots. Hemispherical photographs of forest canopies were taken in the centre of each plot during the 2002 growing season and analysed with Gap Light Analyzer software (Frazer et al., 1999). Using the sky-region grid compound of 36 azimuth and 9 zenith regions taking into account polar projection distortion of the lens two canopy characteristics were calculated for each photograph: Leaf Area Index (LAI) and Canopy Openness. Canopy characteristics were calculated for the whole view of sky, and for portions of the sky that corresponds to the four sectors of the smaller plot. For portion of sampled area satellite Landsat ETM+ image was available which enable comparison of canopy features with Normalized Difference Vegetative Index (NDVI). Relations between mentioned variables (floristic, derived from hemispherical photographs and derived from satellite image) were investigated by univariate correlations, canonical correspondence analyses (CCA ; ter Braak and Smilauer, 2002), classification trees (CT) and logistic regression. Results and discussion Statistically significant negative correlation was detected between canopy openness and NDVI for best-contrasted photographs, while LAI was positively correlated with NDVI for all analysed photographs. LAI was negatively correlated with number of present plant species for best-contrasted photographs. Using the CCA, herbaceous plant species that are most influenced by the light conditions were detected. According to CT analysis that included altitude, slope, northness (cosine of aspect) and LAI as predictor variables, LAI was important predictor variable for estimating the number of present plant species, as well as for occurrence of particular plant species. Besides CT, logistic regression models were developed and tested for their accuracy on independent data for selected species as well. Conclusions These results lead to the future modelling of silver fir - beech forest phytodiversity and spatial distribution of particular plant species as a function of environmental estimators (geomorphologic variables based on digital terrain model and light conditions based on remotely sensed NDVI). Acknowledgements This work was supported by OIKON Ltd. - Institute for Applied Ecology. We wish to thank stuff of Forest Managements from Delnice, Senj, Ogulin and Gospić for their hospitality during our field work. References Frazer, G.W., Canham, C.D. and Lertzman, K.P., Gap Light Analyzer (GLA), Version 2.0. Simon Fraser University, Burnaby, British Columbia, and the Institute of Ecosystem Studies, Millbrook, New York, 36 pp, 1999. Jelaska, S.D., Antonić, O., Nikolić, T., Hršak, V., Plazibat, M. and Križan, J., Estimating plant species occurrence in MTB/64 quadrants as a function of DEM-based variables - a case study for Medvednica Nature Park, Croatia. Ecological Modeling. 170, 333-343, 2003. Ter Braak, C.J.F. and Smilauer, P., CANOCO Reference manual and CanoDraw for Windows User's guide: Software for Canonical Community Ordination (version 4.5). Microcomputer Power (Ithaca, NY, USA), 500 pp, 2002.

canopy openness; digital terrain model; LAI; NDVI; phytodiversity

nije evidentirano