DIGITALNA ARHIVA ŠUMARSKOG LISTA
prilagođeno pretraživanje po punom tekstu




ŠUMARSKI LIST 9-10/2018 str. 52     <-- 52 -->        PDF

Descriptive data on the sold and harvested timber was refined from an initial database that accounted for more than 800,000 m3, over bark (o.b.). Refinement was necessary due to the data inconsistency of some of the harvesting contracts. Specific parameters coming from the timber selling acts such as the felling type - FT, forecasted harvesting methods - HM, cut-block area - TA (ha), sold volume per cut-block - LS (m3 over and under bark on main tree species), removal intensity - RI (m3 × ha-1), tree size - TS (m3 × tree-1) and pruning condition - PC (% of the tree height), terrain slope - S (°), extraction distance - ED (m), initial tendering price - ITP (RON × m-3), final tendering price (selling price) - FTP (RON × m-3) and the difference in price paid by the contractors PD (RON × m-3) were coupled with data coming from the contractors side such as the effectively used harvesting systems (HS) for a number of 1192 harvesting contracts (644,055.00 m3 o.b.) spanning the 2012-2014 period. Harvesting systems were described based on the equipment used and they were categorized in (1) motor manual tree felling and processing - tractor skidding (MMF-SKID), (2) motor manual tree felling and processing - cable yarding (MMF-CY) and (3) motor manual tree felling and processing - animal powered logging (MMF-AL). Such data was provided by the local contractors based on the request of the RFA Baia Mare and it was paired with the contract data into a Microsoft Excel sheet. The types of silvicultural systems under the analysis were clear felling (CF) which in the Romanian mountains is usually carried out in the case of spruce stands on areas less than 3 hectares, thinning (THI) which in Romania is less intensive in terms of removals, salvage felling (SAL) which is usually implemented to extract the wood affected by various factors such as drought, windthrow or mass infestations on relatively large areas, sanitary felling (SAN) which aims to remove locally affected wood and selective felling (SEL) which is the regular system to extract the wood under the continuous cover forestry practice in Romania assuming that such forests are old enough - reached their rotation - to carry on the main harvests. In such a case, a selective felling is done to remove 10 to 33% of the growing stock per decade, providing the needed space for natural regeneration. Given the fact that the forest managers categorize the wood extracted to create the space needed for skid roads or other harvesting and transportation infrastructure as salvage felling, in this study we used the same approach.
Data was analyzed statistically using the Microsoft Excel software. The authors chose this software package as it enables several functionalities of data refining, calculation and statistical analysis. The later was carried out sequentially in order to get the descriptive statistics of the studied conditions, to attempt to model the variation of tendering prices as a function of the harvesting conditions and to analyze in detail such prices. A first step was that of treating the outliers which was described above. Inconsistencies were found as all of the initial data was manually filled into the database. Then the descriptive statistics of the harvesting conditions and selling prices were developed, followed by an attempt to model the tendering prices as a function of harvesting conditions. To this end, a stepwise backward regression procedure using a confidence threshold set at α = 0.05 was used to test the significance of predictor variables and of the developed models (p ≤ 0.05) following a correlation analysis (not shown in this study) which assumed a threshold of R = 0.75 to treat the multicollinearity of the independent variables as described in Sabo & Poršinsky (2005). Exclusion of a given variable within a pair of highly correlated variables was made based on logical assumptions on which of them would be more suitable for the regression analysis. All of the procedures used were those specific to general statistics techniques as described, for instance in Zar (2010).
Finally, a detailed analysis of the difference in price was carried out assuming that stacks with the difference between auction price and selling price under the auction step were negotiated and negotiated stacks are less demanded by the harvesting companies. For simplicity reasons, when assessing the prices per species, stacks having more than 60% volume of one species were considered for that species or group of species. This was the reason for considering only two group of species or species: beech and conifers. The analysis was done in Romanian currency (RON). The average annual exchange rates for the three considered years were RON/EUR 4.4560 in 2012, 4.4190 in 2013 and 4.4446 in 2014 as published by the Romanian Central Bank.
3. Research results
Rezultati istraživanja
3.1. Descriptive statistics of the harvesting conditions and tendering prices – Deskriptivna analiza uvjeta pridobivanja drva i natječajnih cijena
Table 1 shows the summary statistics of operational variables and tendering prices. Cut-block area varied widely between 0.1 and 287 hectares as being specific to some Romanian wood selling practices. The minimum value is specific to those extractions aiming to create space within the forest for the skidding roads. Such extraction types were categorized in this study as salvage cuts (SAL). The maximum value was that specific to sanitary extractions (SAN) where the Romanian praxis is that of grouping several compartments or (and) compartment parts, possibly to cope with very low removal intensities. This study covered the harvesting data coming from more than 22,000 hectares. The mean cut-block area was of about 19 hectares. Accordingly, the volume sold per cut-block varied between 3 and 2,868 m3 o.b., averaging 540.3 m3 o.b. (513.4 m3 under bark,