Natural Regeneration of Beech Forests in the Strict Protected Area of the Plitvice Lakes National Park

Background and Purpose: The study presents the results of an investigation of regeneration processes, growth, development and survival of young growth by field measurement and three-dimensional visualization of horizontal and vertical structure. The results are based on the ten-year investigation (1998-2009) on a permanent experimental plot in a mountain beech forest with dead nettle tree (Lamio orvale Fagetum sylvaticae Ht. 1938) in conditions of passive protection.


INTRODUCTION
Almost half of the continental territory of the Republic of Croatia (48%) is covered by forests.Of the total forest areas approximately 56% consists of beech dominated forests, beech forests with sporadic sessile flowered oak, and mixed beech-fir forests [1].In contrast to many countries, where the natural composition of the forest has been significantly changed by the activity of man, in Croatia a large part of the forests have retained its natural characteristics mainly due to the endeavours of the foresters and the nature of forest management based on the principles of sustainable development.This determined the stability and conservation of forest ecosystems and offered the possibility of establishing national parks during the middle and second half of the past century, in which the main, or one of the main, basic natural phenomena are the forests.The Plitvice Lakes National Park, one of eight national parks in Croatia, was established in 1949 for the protection of the hydrologic system of the lakes, forests and other ecosystems, and natural phenomena.In recognition of the great importance of the forest for the future of the Plitvice Lakes National Park, employees of the Croatian Forest Institute (formerly Forest Institute, Jastrebarsko) established four forest reserves with a total surface area of 1 347 ha: Medveđak (1976), Čorkova uvala-Čudinka (1977), Kik-Visibaba (1979) and Rječica-Javornik (1981).The basic objective of establishing forest reserves was to determine the basic (zero) condition of vegetation, determine structural relations and to monitor the further growth and development of forest ecosystems, particularly the condition and possibility of their natural regeneration as the basic factor for permanent forest ecosystem sustainability and survival.
This problem has been studied by many forest experts in Croatia.Seventy years ago prof.I. Horvat began the first systematic phytocoenological investigation in the Risnjak National Park.The start of forestry scientific research in virgin forests of Croatia can be attributed to the investigations of Čorkova uvala in the Plitvice Lakes National Park, which were commenced by academician Milan Anić in 1957.With the object of monitoring the development of forests in natural conditions in the area of the Plitvice Lakes National Park the investigation of Cestar et al. [2] should be mentioned, who, after carrying out typological investigations, showed that the method of performed management did not encourage the occurrence of young growth, particularly of beech.Hren [3] investigated the structure of the beech virgin forest "Ramino korito", and Prpić [4] investigated the characteristics of the beech-fir virgin forest "Čorkova uvala" in the Plitvice Lakes National Park.Klepac [5,6] advocated active protection of the forests in the Plitvice Lakes National Park, and in 1994 proposed ecological management of the forests with emphasis on the need to enable permanent natural regeneration of forests.In his investigation Novotny et al. [7] pointed to the growth and development of basic structural elements and elements of regeneration in the Plitvice Lakes National Park.
These investigations, with periodic measurements on permanent experimental plots, carried out by employees of the Croatian Forest Research Institute (basic structural elements, indicators of growth structures and development of tree crowns, including the number and quality of young growth), indicate that the possibility of satisfactory natural regeneration in national parks is questionable [8][9][10][11][12][13][14].The aforementioned investigations showed that, although nature is continuously active, we cannot be satisfied only with its activity.Long-term study has shown that passive protection clearly does not give the expected results.
This study aimed to investigate regeneration processes, growth, development and survival of young growth in a mountain beech forest with dead nettle tree (Lamio orvale -Fagetum sylvaticae Ht. 1938) in conditions of passive protection.For that purpose field measurements on a permanent experimental plot in 1998 and 2009 were conducted as well three-dimensional visualization of horizontal and vertical structure.

Study Area
The investigation was performed in the "Medveđak" Forest Reserve on a permanent experimental plot in a natural stand of mountain beech forest with dead nettle (Lamio orvale-Fagetum sylvaticae Ht.1938) 570 m above sea level (Figure 1).The reserve is situated within a larger forest complex of beech forests, in the north-eastern part of the "Plitvice Lakes" National Park.The reserve comprises three compartments with a total area of 156.3 ha.The highest point of the reserve is 875 m, and the lowest 580 m above sea level.The reserve is located on a geological base of limestone with three soil types.On the high positions and on the ridges is humus (black soil) on limestone (10%), on the steep slopes shallow brown soil on limestone (20%), on the less steep slopes moderately deep brown soil on limestone (40%), and in karst sinkholes is loessial soil or illimerised soil (30%) [9].Inclination ranges from 10 0 to 25 0 .In the south-eastern lowest part of the reserve are karst sinkholes, from which the terrain rises up towards the north-east up to the highest point, and again over the ridge descends towards the north and north-east.

Data Collection
In 1998 a permanent experimental plot was established, 1 ha in size (plot coordinates: N=44 0 53' 09"; E=15 0 38' 01") according to the method of Dubravac and Novotny [15] as a part of a network of permanent experimental plots established in the national parks of Croatia (Risnjak, Plitvice Lakes, Paklenica, Mljet, Brijuni).The plots were established with the aim of monitoring the dynamics of forest ecosystems in conditions of strict protection of nature.The age of the stand at the time of the establishment of the plot was 147 years.All trees with diameter at breast height (DBH) greater than 7.5 cm were marked and their basic characteristics measured (DBH, tree height and stem length).In the most homogenous part of the plot a sub-plot was set up, 60m x 60m in size, on which the spatial arrangement of the trees was recorded, their horizontal crown projections were mapped.Furthermore, elevation data of each tree were recorded, according to which a digital terrain model (DTM) was created (Figure 2).In three

RESULTS
Basic structural characteristics of the experimental plot are shown in Table 1.
Comparing the obtained data with growth and yield tables [17] which have volume on I. cite class of 646 m 3 •ha -1 , it may be seen that the obtained volume of the researched stand is considerably higher, mainly due to passive protection of the stand and absence of the management activities.More details on the stand structure elements and comparison between two measurements (1998 and 2009) may be found in the paper of Novotny et al. [7].

Crown structure
From the layout of horizontal projections of crowns in the first measurement in 1998, it was determined that the crown cover of ground amounted to 96%.Mean area of a horizontal crown projection amounted to 53.67 m 2 ranging from 5.94 to 158.36 m 2 .Crown projection area of trees exhibits a nonlinear relationship with the DBH, in the form of the power function (Figure 4).
In the left lower quadrant of the experimental plot an advanced regeneration was found in the opening of the stand canopy, which had resulted from the die-back of one dominant beech tree from the upper canopy layer prior to the establishment of the plot.From the DBH of the dead tree, the area of the horizontal crown projection was estimated to be 145 m 2 .The opening created in the canopy layer stimulated the occurrence of the advanced regeneration.In order to estimate the growth dynamics of the crown structure development between the two measurements, two stand profiles were set up on a part of the plot, 20 m x 60 m in size (Figure 5).In the first measurement the total area of the horizontal crown projections of 27 trees in the profile amounted to 1,492.24m 2 , with the average size of projection area for one tree at 55.27 m 2 .Prior to the second measurement dieback occurred in 3 trees of the total area of the horizontal crown projection of 69.27 m 2 .However, an increase occurred in the mean

Process of natural regeneration
As has already been stated that the process of natural regeneration was mainly influenced by the opening of the canopy due to the die-back of one dominant beech tree and by development of advanced regeneration (Figure 5).The area of the advanced regeneration in 1998 amounted to 124.7 m 2 which increased fourfold in the second measurement, to 512.2 m 2 .By measuring the numbers of the regeneration and shrubs on three strips (Figure 2) on total area of 360 m 2 in 1998 and 2009, an increase in the overall number of plants from 40.8 to 56.2 plants per hectare, was determined (Table 2).The greatest changes occurred in the share of beech seedlings, the number of which increased fourfold from 3,556 plants per hectare in 1998 to 12,694 plants per hectare in the second measuring.In the percentage share common beech increased from 8.7% to 22.6% of all plants of young growth and shrubs, and took a dominant role among the young crop of trees.With regard to the height structure of the young beech trees, their increased number is most visible in height class up to 30 cm (Figure 6).The majority of the young beech plants is of poor quality and it is questionable how they will continue their further development in the conditions of passive protection.
Today, information on stand structure (spatial arrangement of trees, tree measurements, and particularly crown measurements) may be visually presented by means of one of numerous computer programmes.On the basis of a digital model of tree-crown projections, measured values in the field (tree heights, stem length, length and width of crowns) and standard bases, a threedimensional photo-realistic digital model of a stand was produced.In the production of the model the spatial arrangement of trees and phenotype of crown forms were taken into account [18].In this investigation stand structure is visualized in the programme packet EnVision (USDA Forest Service, USA) which is shown in Figure 7.

DISCUSSION
Although the observation was carried out in a relatively short period of time (10 years), the methods and results of this investigation present new contributions to understanding crown structure dynamics and the process of natural regeneration of pure beech stands in conditions of passive protection.Investigations so far [e.g.[19][20][21][22][23][24], also confirmed in this study, indicate that natural regeneration should be carried out on initiated regenerative cores, by opening the canopy in small areas and groups.Similar to this research, results of the research conducted by Rugani et al. [25] showed that gaps smaller than 500 m 2 are the dominant driving force of stand development.Therefore, we think that opening canopy in small areas present the best form of regeneration for nature, especially in protected forest ecosystem.In order to attain scientifically based plans for deciding on the future of forest stands of protected areas, particularly in conditions of climatic changes today, it is necessary to constantly improve and broaden methods of monitoring by means of a network of permanent experimental plots in all protected forest areas.The possibilities offered by modern computerised models for high quality presentation of obtained knowledge should not be ignored.Furthermore, modern remote sensing methods and data (e.g.airborne images, satellite images, LiDAR) may be used for characterising forest dynamics on large-scale areas.For instances, Hobi [26] founded that high resolution satellite images (WorlView-2) have a large potential for forest canopy modelling (for characterising forests' gap dynamics).
By acknowledging existing laws and provisions in areas which are not under a regime of strict protection, it is necessary to assist natural processes which are already in progress in the forest ecosystems.The aforementioned activities are the fundamental method for optimization of priority functions of the forests with specific assignment.Such work should include two important principles: 1) high-production forest is biologically the most stable forest, which offers the greatest generally useful functions and 2) professional silvicultural activities are at the same time the condition and only way in which we can, on the one hand utilize productive functions of the forest, and on the other hand ensure the protection and natural regeneration in these forests.
Knowledge of these problems clearly indicate that increased engagement of the forestry profession is necessary, particularly in those protected areas whose basic phenomena, and most recognized characteristics, are forest ecosystems.In such cases forest interventions should be adapted to the preservation and natural characteristics of the forest ecosystems on particular sites.This is also confirmed by other researches conducted in the 'Medveđak' forest reserve [20,24,27].Therefore, so called 'closeto-nature-forestry' silvicultural approach [19,23,28], with existed but reduced human activities, which is accepted in most of the Central European countries may be adapted in forest managing of Croatian Natural Parks, as well.

FIGURE 5 .
FIGURE 5. Scheme of the development of initiated advanced regeneration between two measurements