0623-B1

Vegetation structure of Quercus dominated forests in Slovenia, Europe

Lado KUTNAR ¹

Abstract

In Slovenia, Europe, on 225 research plots dominated by common oak (Quercus robur) and by sessile oak (Quercus petraea) the vegetation structure has been analysed. Plots of Q. robur are located in five forest complexes, and plots of Q. petraea in four complexes. In the tree layer beside the dominant two oak species we found 27 other species, among them Carpinus betulus, Picea abies, Quercus cerris and Fagus sylvatica have significant share of growing stock. Based on the undergrowth vegetation (shrub and herb layer, terricolous mosses) the Detrended Correspondence Analysis (DCA) separated clearly between plots with dominant Q. robur and those with Q. petraea. The undergrowth vegetation proved to be a valuable indicator of the site conditions and of forest management in the past as well. Based on ordination, lowland common oak forests of long standing co-natural management (Hraščica and Krakovski gozd complexes) are separated from the common oak forests where spruce was favoured by the forest management (Cigonca and Dobrava complexes) and from the man-made common oak stands on primary sites of Q. petraea (Polom). DCA clearly differentiated the sessile oak forests in warmer climate of Sub-Mediterranean region (Panovec), and in warmer sites of Pre-Pannonian region (Pi_ece) from other sessile oak forests (Bojanci, Bukovnica). The main gradients of vegetation structure and of diversity of species as well main ecological gradients in different oak forests were obtained by ordination technique.

1. Introduction

Forests are a dominant part of the environment in Slovenia. They cover more than half of the country (approx. 56 %). Oaks species (Quercus spp.) grow on about 55% of the total forest area in Slovenia. As a main or admixed tree species five autochtonous oak species (Quercus robur, Q. petraea, Q. cerris, Q. pubescens, Q. ilex) represent around 8% of the total growing stock. The oak forests grow mostly in the lowland areas and their hilly margins, where the percentage of forest cover is low, and human population density is high. In such regions the ecological and generally beneficial function of forest are very important. Due to human impact, which results from their vicinity to settlements, oak forests are most altered forest communities. In addition strong sources of air pollutants or changes of groundwater level and other causes diminish the stability and resistance if oak forests. Especially, the floodplain forest ecosystems throughout Europe have always been under heavy anthropogenic impact (Klimo & Hager 2001).

Although many studies of different oak forests in Slovenia were done (e.g. Accetto 1974, Puncer & Zupančič 1981, Dakskobler 1997, Zupančič 1997, Čater et al. 2001) they did not focus much on the structure of vegetation.

Thus, the aim of this study was to determine the main gradients of vegetation structure of the oak forests in Slovenia.

2. Material and methods

2.1 Study area and research plots

Altogether 9 oak forest complexes was selected in different parts of Slovenia. Five forest complexes of common oak (Quercus robur L.) are located mostly in eastern part of Slovenia (Fig. 1): I. Polom, II. Krakovski gozd, III. Dobrava, IV. Cigonca, V. Hraščica.

Four forest complexes of sessile oak (Quercus petraea (Matt.) Liebl.) are more dispersed (Fig. 1): 1. Panovec, 2. Bojanci, 3. Pišece, 4. Bukovnica.

Figure 1: Location of the oak forest research complexes in Slovenia

Most of the Q. robur complexes are located in the floodplains of rivers on the margins of Pannonian basin (Smolej 1995). On the Q. robur complexes deep hydomorphic soil are prevailing (Kalan 1995), which have developed under the influence of either water logging above less permeable soil layers (pseudogley soils) or of high groundwater-table (gley soils - amphigleys and hypogleys). The exception is Polom complex which differs from the other essentially. It lies in the rolly permeable limy hill country near Kočevje, where some surface waterflow could appears occasionally but there is no direct groundwater influence. The selected Q. petraea complexes are placed on more or less permeable hilly terrain with very different types of forest soils out of direct groundwater influences.

Due to ecological regions in Slovenia (Kutnar et al. 2002), all complexes of Q. robur except Polom are located in different sub-regions of the Pre-Pannonian region. The Polom complex lies in Pre-Dinaric region. The three of four Q. petraea complexes are located in different parts of the Pre-Pannonian region. The exception is Panovec complex which is situated in Sub-Mediterranean region.

Polom complex was described as a potential site of the sessile oak-beech forest Querco-Fagetum (Smole & Kutnar 1994, Smole 1995). The vegetation of Dobrava, Cigonca and Hraščica complexes were described as forests of common oak and hornbeam Querco roboris-Carpinetum. The Krakovski gozd complex belongs to the Pseudostellario-Quercetum roboris association (Accetto 1974).

The Panovec complex belongs to Sub-Mediterranean oak forests Carici umbrosae-Quercetum petraeae var. geogr. Sesleria autumnalis. Forest of the Bojanci complex was described as a Epimedio-Carpinetum association. The Pi_ece complex lies in region of potential beech forest Hacquetio-Fagetum var. geogr. Ruscus hypoglossum, and Bukovnica in region of potential hornbeam forest Pruno padi-Carpinetum betuli.

Altogether 225 research plots were selected. In each oak complex (Fig. 1) 25 square (20×20 metres) plots were chosen.

2. 2 Field sampling

The trees with the breast-diameter exceed 10 centimetres were taken into account (Azarov 1995). The undergrowth vegetation of the research plots was surveyed according to the standard Central European method (Braun-Blanquet 1964). The cover of species in shrub, herb and moss layer (only terricoulus mosses) were estimated. The sources of the nomenclature were: Martinčič et al. (1999) for vascular plants; Corley et al. (1981) for mosses.

2.3 Data analysis

On all research plots, based on phytosociological relevés of the undergrowth vegetation following parameters were calculated: species richness (S), sum of cover estimations of all species per plot, mean species cover, Shannon [H' = - _ (p_i log (p_i))] and Simpson [D = 1 - _ p_i ² ] diversity indexes, and Evenness [E = H'/ln (S)]; p_i - share of plant species of total,

Based on field measurement the growing stock of oak trees was analyzed by Azarov (1995). We calculated also number of trees per plot and number of different tree species. Due to Azarov (1995) we estimated the total growing stock, growing stock of the most common tree species: Quercus robur, Quercus petraea, Carpinus betulus, Fagus sylvatica, Picea abies, Acer campestre, Quercus cerris, Alnus glutinosa and Tilia cordata;

The main structure gradients of vegetation were extracted by the detrended correspondence analysis (DCA) (Hill & Gauch 1980). As a criteria for ordination of the research plots floristic composition was used. The DCA was carried out with the PC-ORD program (McCune & Mefford 1999). We calculated the Spearman correlation coefficients between the DCA axes (plot scores) and: a) tree layer parameters; b) undergrowth vegetation parameters;

3. Results and discussion

On 225 research plots, altogether 29 tree and shrub species with the breast-diameter over 10 centimetres were found. The most frequent tree species are following (in parentheses are numbers of research plots with their presence): Quercus robur (124 plots), Quercus petraea (100), Carpinus betulus (90), Fagus sylvatica (56), Picea abies (50), Acer campestre (42), Quercus cerris (18), Alnus glutinosa (16), Tilia cordata (16), and Pinus nigra (12). Other less frequent tree and shrub species with the breast-diameter over 10 centimetres are following: Pyrus pyraster, Sorbus torminalis, Prunus avium, Crataegus monogyna, Pinus sylvestris, Ostrya carpinifolia, Fraxinus ornus, Larix decidua, Populus tremula, Acer pseudoplatanus, Abies alba, Sorbus aria, Fraxinus excelsior, Betula pendula, Acer platanoides, Ulmus glabra, Ulmus laevis, Corylus avellana and Ilex aquifolium. The total number of trees and shrubs with the breast-diameter over 10 centimetres were 4003. In average 17.8 tree (shrub) individuals per 20×20 metres plots were found. The number of tree (shrub) individuals per plot ranges between 4 and 35. The estimation of total growing stock per plot varies from 5.1 to 48.8 m ³ .

Based on the floristic composition, nine oak forest complexes are well separated in the plot ordination space (Fig. 2, 3, 4). DCA analysis clearly differentiates between the Quercus robur dominated plots and the Q. petraea dominated plots. In DCA1 vs. DCA2 ordination (Fig. 2 and 3), plots of Quercus robur are below the diagonal of graph (start from point DCA1=0, DCA2=0), and plots of Q. petraea are above this diagonal.

As a result of similarity of floristic composition (Fig. 2 and 3), in 3-D ordination space the Dobrava plots and Cigonca plots are grouped close together. On these plots are swamped, gleyed soils resulting from impact of high groundwater-table (Kalan 1995). The Hraščica and Krakovski gozd plots are quite similar in floristic composition (Fig. 2). However, in 3-D ordination space (Fig. 3), the significant differences between these plots are shown (Krakovski gozd plots obtained higher DCA3 scores than Hraščica plots). It is mainly a result of different soils conditions of these plots (Kalan 1995). On Hraščica plots eutric brown soils were found, and on Krakovski gozd plots the gleyed swampy soils. As a result of intensive ground-water influence on Krakovski gozd, Cigonca and Dobrava plots, the similarity in Figure 3 is shown.

Due to difference in site and climate conditions, the third DCA axis segregates the Bojanci plots from the Bukovnica plots. The Panovec plots and the Pi_ece plots obtain low scores along the DCA1 axis (Fig. 2 and 3). Resulting from pretty similar local climate conditions, in ordination space they lie close together. Although they origin from different part, their floristic composition shows on quite similar site condition. Due to relative humid and cold soils on flysch of Panovec plots, as well due to moderate north exposition of plots, the effect of warm Mediterranean climate is not so pronounced. On the other hand, the Pi_ece plots are situated on very warm southern slopes in the Pre-Pannonian region.

In the Polom region, the Q. robur acorns were important nutrition source of the domestic pigs in the past (Smolej 1995). Thus, in this region the forest stands dominated by the Q. robur are man-made. The oak trees is growing under conditions, where groundwater has no direct influence. Potentially, the Polom region is site of mixed forest of Q. petraea and Fagus sylvatica. Consequently, the undergrowth vegetation of Polom plots is more similar to that on plots overgrown with the Q. petraea forests than with the Q. robur forests. Therefore, the Polom plots lie quite close to Q. petraea plots in ordination space (Fig. 2 and 3).

Vectors of Q. robur and of Q. petraea show increasing of growing stock of these two tree species (Fig. 2 and 3). They confirm the differentiation between Q. robur and Q. petraea dominated plots based only on floristic composition of undergrowth layers. It is proved also with calculation of Spearman rank correlation (Table 1).

Besides similar floristic composition, Cigonca and Dobrava plots have significant share of Picea abies in tree layer (Fig. 2 and 3). In specific conditions of these plots, the Norway spruce was favoured by the forest management in the past. The significant positive correlation was found between the first DCA axis and the growing-stock of Picea abies (Table 1). Apart these tree species, in DCA analysis (Fig. 3) only vector of the Fagus sylvatica growing stock is shown. The growing stock of beech were found to increase towards the Bukovnica plots. Among nine common tree species we have analysed (Table 1) the weak negative correlation between DCA2 axis and Carpinus betulus growing stock, and between DCA2 and Acer campestre growing stock is shown too. The number of different tree species slightly increasing towards plots in lower parts of the DCA ordination space (Fig. 2, Table 1).

Figure 2. DCA analysis of research plots and vectors of tree layer parameters (Axis1 vs. Axis2)

Legende. Q. robur complexes: PO-Polom (I), KG-Krakovski gozd (II), DO-Dobrava (III), CI-Cigonca (IV), HR-Hraščica (V); Q. petraea complexes: PA-Panovec (1), BO-Bojanci (2), PI-Pišece (3), BU-Bukovnica (4).

Figure 3. DCA analysis of research plots and vectors of tree layer parameters (Axis1 vs. Axis3)

In the undergrowth vegetation layers, the total number of 254 species were recorded. In average 31 plant species per plot were found. The number of species per plot ranges between 4 and 70.

On all 225 research plots Carpinus betulus is the most frequent plant species of undergrowth vegetation. It is present on 173 research plots. Other very frequent plant species are Acer campestre (129 plots), Prunus avium (128 plots), Crataegus monogyna (124 plots). In the undergrowth vegetation, the dominant tree species Quercus robur (113 plots) and Quercus petraea (94 plots) are frequent too. Other common ligneous plant species are Corylus avellana (109), Ligustrum vulgare (90) and Euonymus europaea (83). The herb species with high frequency are Anemone nemorosa (124), Athyrium filix-femina (116), Polygonatum multiflorum (111), Viola reichenbachiana (94), Pulmonaria officinalis (94), Carex brizoides (90), Galium sylvaticum (85), and Ajuga reptans (80). The most common moss species is Polytrichum formosum that occurs on 73 research plots.

The species richness were found to increase towards the Polom plots (Fig. 4). On Polom plots the mean number of species in the undergrowth layers is 62. In average, high number of species have been found on the Pi_ece plots (49 species/plot) and on the Krakovski gozd plots (45 species/plot) too. On Dobrava and on Cigonca plots we found very few species (9 and 12).

Due to high number of species and due to comparatively balance of their cover, the Polom, Pi_ece and Krakovski gozd plots have high Shannon (H') and Simpson (D) diversity index. With exception of the Polom plots, where Corylus avellana dominates in shrub layer, the evenness index (E) is also high on these plots.

The mean species cover increasing towards the Dobrava and Cigonca plots (Fig. 4, Table 1). Besides the low number of species, almost all of these plots are completely overgrown by Carex brizoides.

The sum of all species cover is high on Polom and Krakovski gozd plots. These are plots with high number of species in herb and in shrub layer.

Figure 4. DCA analysis of research plots and parameters of undergrowth vegetation (Axis1 vs. Axis2)

Legende. Q. robur complexes: PO-Polom (I), KG-Krakovski gozd (II), DO-Dobrava (III), CI-Cigonca (IV), HR-Hraščica (V); Q. petraea complexes: PA-Panovec (1), BO-Bojanci (2), PI-Pi_ece (3), BU-Bukovnica (4).

Table 1. Spearman rank correlations between DCA scores and 1) tree layer parameters, 2) undergrowth vegetation parameters.

4. Conclusion

Despite heavy anthropogenic impact, the variety of oak forests exist in Slovenia. Especially, the Quercus robur and Quercus petraea dominated forests create a wide range of different types with very divers vegetation structure.

By the ordination technique, diversity of the vegetation structure of different oak forests were obtained. The ordination has differentiated clearly between plots with dominant Q. robur and those with Q. petraea. Among them, due to the site conditions and the man influences, many different oak forests have been shown. By the ordination of oak forest based on the floristic composition, the main ecological gradients were extracted too. The undergrowth vegetation proved to be a valuable indicator of the site conditions and of forest management in the past as well.

Acknowledgements

Thanks are to colleagues M. Sc. Igor Smolej, Ivan Smole, Evgenij Azarov-Stjopa and Janko Kalan at the Slovenian Forestry Institute for all useful comments and all help. I wish to thank to all other collaborators for the field and technical assistance.

References

Accetto, M., 1974. Združbi gabra in evropske gomoljčnice (Pseudostellario-Carpinetum) ter doba in evropske gomoljčnice (Pseudostellario-Quercetum) v Krakovskem gozdu. Gozd. vestn., 32 (10): 357-369. In Slovenian.

Azarov, E., 1995. Qualitative and quantitative characteristics of oaks on permanent research plots. In: Smolej, I. and H., Hager (eds.). Oak decline in Slovenia: Endbericht über die Arbeiten 1995. Gozdarski in_titut Slovenije, Ljubljana, Institut für Waldökologie, Wien, pp. 26-34.

Braun-Blanquet, J., 1964: Pflanzensoziologie. Grundzüge der Vegetationskunde. Springer, Wien, New York. 865 pp.

Corley, M. F. V., A. C. Crundwell, R. Düll, M. O. Hill and A. J. E. Smith, 1981. Mosses of Europe and the Azores: an annotated list of species, with synonyms from the recent literature. J. Bryol. 11: 609-689.

Čater, M., L. Kutnar and M. Accetto, 2001. Slovenian lowland and floodplain forests. In: Klimo, E. and H. Hager (eds.). The floodplain forests in Europe: current situation and perspectives. European Forest Institute Research Report, 10. Leiden; Boston; Köln: Brill, pp. 233-248.

Dakskobler, I., 1997. Phytosociological characteristics of holm oak Quercus ilex L. stands on Mount Sabotin and above the source of the Lijak river (Western Slovenia). Acta biol. slov., 41 (2-3): 19-42. In Slovenian with English summary.

Hill, M. O. and H. G. Gauch, 1980. Detrended correspondence analysis, an improved ordination technique. Vegetatio 42: 47-58.

Kalan, J., 1995. Basic soil analyses.- In: Smolej, I. and H. Hager (eds.). Oak decline in Slovenia: Endbericht über die Arbeiten 1995. Gozdarski in_titut Slovenije, Ljubljana, Institut für Waldökologie, Wien, pp. 43-46.

Klimo, E. and H. Hager. 2001. Executive Summary. In: Klimo, E. & Hager, H. (eds.) The Floodplain Forests in Europe: Current Situation and Perspectives. European Forest Institute Research Report 10, Brill, Leiden, Boston, Köln, pp. VI-XI

Kutnar, L., M. Zupančič, D. Robič, N. Zupančič, S. Žitnik, T. Kralj, I. Tavčar, M. Dolinar, C. Zrnec, and H. Kraigher, 2002. The delimitation of the regions of provenance of forest tree species in Slovenia based on ecological regions. Zb. gozd. lesar., 67: 73-117. In Slovenian with English summary.

Martinčič, A., T. Wraber, N. Jogan, V. Ravnik, A. Podobnik, B. Turk, and B. Vreš, B. 1999. Mala flora Slovenije, Ključ za določevanje praprotnic in semenk. Tehniška založba Slovenije, Ljubljana, 845 pp. In Slovenian.

McCune, B. and M. J. Mefford, 1999. PC-ORD: Multivariate Analysis of Ecological Data, Version 4.0. MjM Software Design,Glenden Beach, Oregon. 237 pp.

Puncer, I. and M. Zupančič, 1981 Novi združbi gradna v Sloveniji (Melampyro vulgati-Quercetum petraeae ass. nova s. lat.). Scopolia, 2: 1-47. In Slovenian.

Smole, I., 1995. Vegetations- und Standortsverhältnisse der Ständigen Versuchs-flächen in den Eichenwäldern Sloweniens. In. Smolej, I. and H. Hager (eds.). Oak decline in Slovenia: Endbericht über die Arbeiten 1995. Gozdarski in_titut Slovenije, Ljubljana, Institut für Waldökologie, Wien, pp. 47- 59.

Smole, I. and L. Kutnar, 1994. Vegetacijske in rastiščne razmere na trajnih raziskovalnih ploskvah hrasta v Sloveniji, III. del: povzetek I. in II. dela naloge). Gozdarski inštitut Slovenije, Ljubljana, 50 pp. In Slovenian.

Smolej, I., 1995. Permanent research plots. In. Smolej, I. and H. Hager (eds.). Oak decline in Slovenia: Endbericht über die Arbeiten 1995. Gozdarski inštitut Slovenije, Ljubljana, Institut für Waldökologie, Wien, pp. 11-14.

Zupančič, M., 1997. (Sub)mediterranean floral element in the forest vegetation of the Submediterranean floric area of Slovenia. Razpr.- Slov. akad. znan. umet., Razr. naravosl. vede, 38: 257-298. In Slovenian with English summary.

¹ Slovenian Forestry Institute, Department of Forest Ecology
Večna pot 2, 1000 Ljubljana, SLOVENIA, Europe
Tel: (+386) 01 200 78 00/ Fax: (+386) 01 2573 589
E-mail: [email protected]
Internet site: http://www.gozdis.si/