ANALYSIS OF THE QUALITY OF PRIMARY FOREST ACCESSIBILITY: THE CASE STUDY

Relative forest accessibility percent of accessible forest area by forest roads in comparison with total forest area is the most important indicator of the quality of primary forest accessibility. The accessible forest area by forest roads is determined by the bounded area around forest roads. Today, in the area with steep and variable slopes of terrain, a double targeted geometrical extraction distance of timber is used for the width of the bounded area around forest roads, and the forest road spacing is used in the area with a mild and uniform slope of the terrain. Both parameters depend on the targeted density of forest roads. Modern information technologies (IT) like geographical information systems (GIS) enable the quality spatial and statistical analysis of different kinds of data whose result is not accessible forest areas by current primary forest traffic infrastructure only, but also an insight into the spatial distribution of insufficient accessible areas into the forest area. The spatial distribution of these areas is significant for spatial distribution of the new routes of forest roads. The research is done in the area of Forest Management Unit (FMU) Prosara, for which the spatial analysis of a digital terrain model (DTM) determined the mostly hilly relief area. The average relative forest accessibility, based on double targeted geometrical extraction distance of timber, is 35% for the actual network of forest roads, and targeted forest road spacing is 60% for the upgraded network of forest


Abstract
Relative forest accessibility percent of accessible forest area by forest roads in comparison with total forest area is the most important indicator of the quality of primary forest accessibility.The accessible forest area by forest roads is determined by the bounded area around forest roads.Today, in the area with steep and variable slopes of terrain, a double targeted geometrical extraction distance of timber is used for the width of the bounded area around forest roads, and the forest road spacing is used in the area with a mild and uniform slope of the terrain.Both parameters depend on the targeted density of forest roads.Modern information technologies (IT) like geographical information systems (GIS) enable the quality spatial and statistical analysis of different kinds of data whose result is not accessible forest areas by current primary forest traffic infrastructure only, but also an insight into the spatial distribution of insufficient accessible areas into the forest area.The spatial distribution of these areas is significant for spatial distribution of the new routes of forest roads.The research is done in the area of Forest Management Unit (FMU) Prosara, for which the spatial analysis of a digital terrain model (DTM) determined the mostly hilly relief area.The average relative forest accessibility, based on double targeted geometrical extraction distance of timber, is 35% for the actual network of forest roads, and targeted forest road spacing is 60% for the upgraded network of forest roads.

Research area / Područje istraživanja
The FMU Prosara was selected as a case study which is managed by Forest Administration (FA) Gradiška (Figure 1).It is located in the north of B&H.The total area of the FMU Prosara is 3,812.49ha, the length of the forest road is 28 km, and the density of forest roads is 7.3 m/ha (IRPC, 2013).
The impact of forest roads on the value of forest accessibility and density of forest roads depends on their length, their spatial distribution and their suitability for timber extraction.Poršinsky et al. (2017) defined four basic and five spatial criteria for the determination of forest road density.The basic criteria defined the impact of a forest road on its density based on the features, limitations and suitability for timber extraction.The spatial criteria for determination of the impact of forest roads on their density are based on the location of forest road in the forest area.These criteria are used for the determination density of forest roads for the area of a FMU but not for the area of the compartment or sub-compartment.
The second indicator of forest accessibility is the average extraction distance that can be geometrical and real for a theoretic, actual and upgraded network of forest roads.It has an impact on the productivity of means of transport and the costs of transport.
The indicators of the forest accessibility quality are relative forest accessibility and the coefficient of efficiency of forest roads network.
Hence the need to define the aims which will contribute to a comprehensive analysis of quality primary forest accessibility:

Methods of research / Metodi istraživanja
Calculation of relative forest accessibility requires defining the targeted density of forest roads for a specific area.It can be calculated depending on the costs of transport of timber and cutting volume of timber.In addition, the targeted density of forest roads may depend on the intensity of forest management and terrain or relief characteristics of the observed area.
The base for the determination of the targeted density of forest roads is the terrain characteristics of the FMU Prosara since the costs of transport are very changeable at this period of time.Terrain characteristics can be obtained by spatial analysis of DTM.The result of this analysis is the map of relief regions which was obtained based on the classification of elevations according to Bertović (1999).
The targeted density of forest roads is the base for the calculation of the width of accessible forest area by forest roads.The average targeted geometrical extraction distance and targeted forest road spacing will be used as the width of accessible forest area by forest roads.Theaverage targeted geometrical extraction distance is the most often used for defining the width of accessible forest area by forest roads (Backmund, 1966;Hayati et al., 2012;Hentschel, 1996;Lepoglavec, 2014;Pentek et al., 2005;Petković, 2019;Petković & Potočnik, 2018;Petković et al., 2019;Sachs, 1968;Sokolović & Bajrić, 2013).
The average targeted extraction geometrical distance is obtained by the ratio of the average targeted real extraction distance and the average value of the extraction factor.The extraction factor is a factor of geometrical distance increasing due to elevation, slope, and the presence of obstacles at the ground.
They are then entered into ESRI ArcGIS 10 Geoprocessing tool Buffer for the definition of the bounded area around forest roads which represents accessible forest area by forest roads.This area goes over the border of the FMU and it is necessary to cut these parts of the area by using ESRI ArcGIS 10 Geoprocessing tool Clip at the end (Figure 2).The result is an accessible forest area by forest roads within the boundaries of the FMU.
This area is compared with the total area of the FMU and it is obtained a percent of the FMU area which is accessible by forest roads.It represents the relative forest accessibility O R .
The relative forest accessibility can be assessed as insufficient if the percentage of accessible forest area by forest roads is lower than 55% of the total forest area, poor if it is between 56 and 65% of the total forest area, hardly good if it ranges between 66 and 75% of the total forest area, very good if it is between 76 and 85% of the total forest area and excellent if it ranges over 85% of the total forest area (Pentek et al., 2005).
The second importance of the calculation of relative forest accessibility, besides the percentage of accessible forest area by forest roads, is an insight into the spatial distribution of inaccessible forest areas within the boundaries of the FMU.It is necessary to design new forest roads in these areas if we want to make them accessible.These routes for new forest roads are designed as the ground lines with a maximum grade of 8% and 10% for short lengths of ground lines in special terrain conditions at the strategic level of planning of forest roads on the digital map with contours.The purpose of designing ground lines for new forest roads is to achieve the targeted density of forest roads in the FMU and the analysis of relative forest accessibility for the upgraded network of forest roads.This analysis is based on the width of

RESULTS / REZULTATI
The two relief regions were obtained by a spatial analysis of the DTM of the FMU Prosara according to Bertović's (1999) classification of elevation.
The share of the lowland region in the FMU is 37.5%, and the hilly 62.5% in the FMU.The FMU Prosara is a mostly hilly relief area (Figure 3).
The total length of forest roads, which were recorded by GPS, is 40,130 m in the FMU, and 31,007 m or 77% of the total length of forest roads make forest area accessible.Accordingly the actual forest accessibility is 8.13 m/ha (Table 1).The length of forest roads that make the lowland part of the FMU accessible is 6,813 m, while the hilly part of the FMU is 24,194 m.
accessible forest area by forest roads which is equal to the double average targeted geomet-rical extraction distance (Lepoglavec, 2014) and the targeted forest road spacing.
The targeted density of forest roads for the lowland relief region is 12 m/ha, while for the hilly region is 18 m/ha (Narodne Novine, 2015).Based on the share of relief regions in the FMU Prosara area, the average value of targeted density is 15.75 m/ha (Table 2).2014) and their share in the total area of the FMU.
The targeted forest road spacing is 635 m based on the average value of the targeted density of forest roads.
The accessible forest area by forest roads is 1,153 ha for double targeted geometrical extraction distance (466.87 m) and 1,527 ha for targeted forest road spacing (Figure 4).The relative forest accessibility ranges between 30 and 40% and it is considered insufficient (Pentek, 2005).
It is necessary to increase the actual density of forest roads by 8 m/ha to achieve a minimum density of forest roads for optimal forest management regarding the relief characteristics of the FMU.This means that it is necessary to design 30,500 m of new forest roads.
24,572 m (Figure 4) of new forest roads or 80.6% of the needed total length is designed, 97.5% of which makes the forest accessible by forest roads from two sides (Table 4).
The total length of the upgraded network of forest roads is 55,579 m based on the length of actual (31,007 m) and proposed (24,572 m) forest roads that make forest accessible.Based on it, the achieved density of forest roads is 14.6 m/ha.
The accessible forest area by forest roads for double average targeted geometrical extraction distance (466.86 m) is 2,029.56ha, while for targeted forest roads spacing (635 m)  The importance of relative forest accessibility is the percentage of accessible forest area by forests roads at first, but the most important is the percentage of the total forest area which remains outside the influence of forest roads.Those are areas in the forest area that are insufficiently accessible and it is necessary to make them accessible by new forest roads.The base for the calculation of the relative forest accessibility is a forest area accessible by forest roads which depends on its width, and the most often indicator of the width of accessible forest area is double average targeted geometrical extraction distance and the targeted forest road spacing is used for the width.
There are not many differences between the results of the analysis of relative forest accessibility based on the width of accessible forest area by forest roads depending on double targeted geometrical extraction distance and targeted forest road spacing, but it should be applied one or another parameter depending on relief conditions.The most important task in the case of using double-targeted geometrical extraction distance for the width of accessible forest area by forest roads, in the area with steep and variable slopes of the terrain is the determination of the extraction factor.The real extraction distance depends on it, and the costs of transport depend on the real extraction distance.These comprehensive spatial and statistical analyses can be done by using GIS software.
These require quality data that can be obtained from forest management plans, field research, DTM, and other sources of data.

Table 2 .
Weighted targeted density of forest roads / Tabela 2. Ponderisana ciljana gustina šumskih puteva The targeted average real extraction distance would be 317.46m, and the targeted average geometrical extraction distance would be 233.43m since the value of weighted extraction factor is 1.36 (Table 3) for the FMU Prosara.It was obtained by the multiplication of the extraction factors for lowland 1.20 and hilly 1.45 relief regions (Lepoglavec,

Table 4 .
Proposed forest roads / Tabela 4. Predloženi šumski putevi The targeted geometrical extraction distance and targeted forest road spacing depend on the targeted density of forest roads and it is necessary to define it.The targeted density of forest roads in this case should be 15.75 m/ha based on relief conditions.Petković & Potočnik (2018) and Petković (2019) determined that the targeted density of forest roads should be 16.5 m/ha based on the costs of transport or 17.47 m/ha according to the data obtained by IRPC & Šuma plan d.o.o.(2019).Differences between these results range from 0.7 to 1.8 m/ha.The density of forest roads for the upgraded forest road network is 14.6 m/ha and it is 1.15 m/ha lower than targeted.The density of the forest roads for the upgraded network is around 10 m/ha in lowland and 17.3 m/ha in hilly terrain conditions.It means that the targeted density of forest roads is almost achieved in hilly terrain conditions.