Changes in horizontal structure of natural beech forests on an altitudinal gradient in the Sudetesby Zdeněk Vacek, Stanislav Vacek, Lukáš Bílek, Jiří Remeš, Igor Štefančík



Forestry / Plant Science


2015, vol. 73, 33–45

Zdeněk Vacek, Stanislav Vacek, Lukáš Bílek, Jiří Remeš,

Igor Štefančík

Changes in horizontal structure of natural beech forests on an altitudinal gradient in the Sudetes

Received: 13 December 2013; Accepted: 25 September 2014

Abstract: The article describes horizontal structure of the tree layer, natural regeneration, snags and crown projections of natural beech stands on three permanent research plots in the wide altitudinal range in the

Krkonoše Mts (Czech Republic). The spatial structure was classified from 1980 to 2010 and subsequently the prediction of spontaneous development with an outlook for 30 years (to 2040) was done by growth simulator. Hopkins-Skellam index, Pielou-Mountford index, Clark-Evans index and Ripley’s K-function were calculated. Further, the vertical structure and total diversity index was evaluated. The horizontal structure of individuals in the tree layer had not changed significantly during the monitored years. Tree spatial pattern of the lowest altitude lying herb-rich beech forest was mostly regular to random, in acidophilous mountain beech forest predominantly random and in fragments of beech groups around the timberline aggregated. Juvenile growth on all investigated plots was distributed aggregated and snags randomly. The horizontal structure of crown projection centroids had always higher values toward the regularity than tree layer and was random to regular. The result of principal component analysis also confirmed that spatial pattern was dependent on the altitude, but also on the number of trees.

Additional key words: European beech (Fagus sylvatica L.), spatial pattern, structural indices, forest dynamics, development prediction

Addresses: Z. Vacek, S. Vacek, L. Bilek, J. Remeš, I. Štefančík, Czech University of Life Sciences, Prague,

Faculty of Forestry and Wood Sciences, Kamýcká 129, 165 21 Prague, Czech Republic, e-mail:

I. Štefančík, National Forest Centre, T.G. Masaryka 22, 460 01 Zvolén, Slovak Republic


Modifications of spatial patterns of important forest attributes such as living mature trees and their crowns, deadwood and natural regeneration along altitudinal gradient may result from various environmental conditions such as climate, edaphic conditions, disturbance regime and human impact.

Surprisingly these aspects have been studied along extensive altitudinal gradients worldwide (Barrera et al. 2000; Motta et al. 2006; Holeksa et. al. 2007;

Girardin et al. 2014), but less information has been gathered from temperate zone with special focus on beech dominated forests.

Due to its ecological plasticity and broad ecological amplitude European beech (Fagus sylvatica L.) occurs 34 Zdeněk Vacek et al. over a wide range of mesic soils, with pH ranging from 3.5 to over 7.0, and humus form mull to mor with the exception of pseudogleys, or soils with reducing conditions within 20 cm from the soil surface (Le Tacon 1981; Otto 1994). In central Europe beech dominates the major and central part of the moisture and nutrient range of forests and is absent only where rain is insufficient, or where the soil is too dry (Ellenberg 1996). In the conditions of the Czech

Republic, beech occurred originally in the submontane, montane and subalpine zones from 300 m a.s.l. to 1300 m a.s.l. (Neuhäuslová et al. 1998), but was on most sites replaced by Norway spruce (Picea abies /L./ Karst) as economically more interesting species.

In the present Czech Republic, the representation of beech has been reduced from more than 40% of the natural representation to less than 8% today (Ministry of Agriculture of the Czech Republic 2013) and most beech forest have been modified in their tree species composition and structure.

In Central European conditions extensive remnants of old-growth beech forests remain to a larger extent in the Carpathians, but in the absence of strong human impact valuable examples of natural or near-natural beech stands are also known from the Central European middle-mountains including also the mountain range of the Sudetes (Jeník 1998).

Man-made forest stands mostly have lower volumes of dead wood (standing and fallen) – (Christensen et al. 2005), simplified DBH and age structure and regular distribution of trees, while natural forest stands that originate from natural regeneration (from seeds, vegetative sprouts or by layering) usually have an aggregated or randomly irregular initial distribution (Vacek et al. 2010a, 2010b) and generally higher structural heterogeneity (Rademacher et al. 2001;

Rozas 2006). In the course of stand development this type of distribution changes toward moderately regular distribution in favourable environmental conditions (Korpeľ 1995; Wolf 2005), while in less favourable conditions more irregular or aggregated structures are expected (Vacek et al. 2010b).

Commonly, structural indices and functions are used to study the structure of forest stands. In numerous studies on horizontal structure of forest stands, distribution indices based on a distance of trees to the nearest neighbour have frequently been used for a long time. Probably the best-known aggregation index R (Clark and Evans 1954) compares the actual distance of a tree to its nearest neighbour with a distance fulfilling the condition of purely random stand structure given by the Poisson probability distribution. Frequently are also used distribution indices based on a distance between a randomly selected point and actual positions of trees. The first index of this type was proposed by Hopkins and

Skellam (1954); it is based on the principle that the population has a random distribution in case that the distribution of distances from any point to its nearest neighbour coincides with the distribution of distances from a randomly selected tree to its nearest neighbour. The same principle was applied e.g. by

Mountford (1961).

In the seventies of the 20th century the first distribution functions were proposed (Geyer 1999) with the objective to express the horizontal structure in a continuous way. Their advantage is that they document the intensity of particular types of distribution to various distances (Pretzsch 2009). The frequently used K-function (Ripley 1977) shows the mean value of the number of individuals situated at a distance smaller than or equal to r from a randomly chosen individual.