Productivity Analysis of Modifi ed Farm Tractors in Forwarding Industrial Wood Products

Public demands for forest products and services have been increasing gradually in many countries all over the world. Among the forest products, mostly round woods are produced in Turkey due to high economic value. On the other hand, there is a signifi cant increase in demand for industrial woods in the domestic market in recent years. The reason behind this increase is the need of domestic companies in the wood-based panel industry for a large amount of wood raw material. . Farm tractors are the most widely used forest machinery in the extraction of industrial wood in Turkey. Farm tractors can be effectively used in mechanized harvesting operations after receiving some modifi cations and additional attachments. The modifi ed farm tractors are affordable machinery for small logging contractors since their initial purchasing costs and hourly operating costs are quite low comparing with the forest harvesting equipment (i.e. skidder, forwarder, cable yarder). In this study, productivity of a modifi ed farm tractor was analyzed during forwarding of industrial wood products. The study was implemented in Pinus nigra stand located in (?) the city of Afyonkarahisar in Turkey. In the study, the effects of forwarding distance on productivity were evaluated using statistical analysis. The results indicated that the most time-consuming work stage was unloading the truck and piling the wood products at the landing area, followed by ( in terms of time-consumption) collecting woods and loading the tractor at the stump. The average productivity of the forwarding operation was 6.06 stere/h (stere is the amount of wood pile with the height, width, and depth of one meter). It was also found that the productivity of the forwarding was signifi cantly affected by the forwarding distance.

Introduction. Due to the high economic value, round-wood harvesting takes a leading position among the forest products in Turkey. On the other hand, there has been a signifi cant increase in the demand for industrial wood in the domestic market in recent years. One of the reasons for this increase is that domestic fi rms need a large amount of wood raw materials in the wood-based panel industry (Bilici, Akay, & Abbas, 2019). Factories processing industrial wood raw materials have to satisfy their increasing raw material needs timely and regularly for the continuity of their production. Most of the wood raw material demand (60%) in the domestic market is satisfi ed by the General Directorate of Forestry (GDF) (Türk, 2016). Under conditions of increasing competitiveness in Turkey's forest products industry, in order to respond to the demand for raw materials of the forest products sector and, on the other hand, to meet the need for community forest products, forests are to be managed and function on the principles of sustainability.
According to the OGM 2017 annual report (GDF, 2017), it is stated that annual wood consumption in Turkey is around 31 million m 3 , approximately 21 million m 3 of it is produced by GDF, and the rest is covered by private sector and imports. According to GDF's production plans for 2018 (URL-1), it was reported that the target stumpage sale was 25.7 million m 3 in which 20 million m 3 was industrial wood and 5.7 million m 3 was fi rewood. The production amount for 2018, where the realization rate is 99%, was 24.85 million m 3 . Approximately 19 million m 3 of this amount was industrial wood and the rest was fi rewood. The largest amount of wood products was fi ber-chip board and followed by the log production (Tab. 1). This also corresponds to 75% of the industrial wood produced.
The forested areas in Turkey generally have a steep terrain, which increases the costs of traditional logging method based on human and animal power, compared to machine-based logging operations (Eroğlu, Sarıyıldız, Küçük, & Sancal, 2016). The fragmented structure of the forest lands and operating in small harvesting units are also the key factors that increase the cost of timber extraction. In addition, traditional logging methods are limited in terms of amount of extracted wood per unit area. In today's social and economic conditions, various mechanized logging equipment is used in forestry activities in Turkey in order to increase the productivity while protecting the forest ecosystem and to improve the socio-economic life of the forestry workers who mostly live in forest villages. The forest operations in which mechanization is most commonly used are tree felling and the extraction of forest products from the felling sites (Erdaş Acar, & Eker, 2014). However, the utilization rate of mechanized logging methods in Turkey is still quite low, while the target is to increase this rate to around 15%. Farm tractors are used in many countries in the concept of small-scale forest harvesting (Yu, Gallagher, Mitchell, & O'Neal, 2016). In Turkey, the mechanized equipment that is most frequently used in timber extraction is the modifi ed farm tractors as well (Gülci & Akay, 2017). Proto et al. (2018) reported that the modifi ed farm tractors are also widely used in European countries such as Italy, Croatia, etc.
The modifi ed farm tractors are used in different ways as a skidder, loader, and forwarder (Öztürk & Akay, 2007). Cable systems can be also applied in both uphill and downhill directions by farm tractors (Marenče & Krč, 2016;Lee et al., 2017). Transporting forest products from stump to landing is a diffi cult and dangerous job. For the farm tractors to be used in these hard works, especially motor power, tail shaft and hydraulic equipment are of great importance. The tractor must be structurally and powerfully suited to forestry works. For instance, there should be a correct ratio between the tractor's engine power and weight. Besides, the weight has a positive effect in terms of increasing the pulling force of the tractor and decreasing wheelslip.
Since farm tractors have proportionally more engine power than their weight, wheelslip increases in heavy jobs such as skidding and winching in the forest. In this case, it can cause work accidents and damage to the machines. In order to eliminate the drawbacks caused by the decrease in the weight of the tractor, additional weights are attached to various parts of the tractors. When extracting industrial woods, carrier saddles are attached in the front and at the back of the tractor to carry wood products, which also help to balance the weight. Besides, additional measures can be taken to empower traction such as reducing tire infl ation pressures and using an open profi le tire or attaching a chain to the rear tires (Öztürk & Akay, 2007). Due to the drums attached to the back of the farm tractors, they can be used as a skidder. The rotation of these drums, which depend on the steel cable rolling capacity of a certain length, is carried out with the power taken from the tail shaft of the tractor.
Tractors can get the logs out of the forest by moving in the forest along skid roads, and they can pull the logs from the slopes under the road to the side of the road with their drums by standing on the roadside landing ( Fig. 1). In areas where the slope of the land is less than 30%, the products can be removed from the felling site by skidding them on the skid roads with the farm tractor ( Figure 1). Since the farm tractor's mobility is limited in mountainous areas where the slope is more than 30%, the tractor stands in a suitable place on the forest roadside and removes the products from the stump by cable winching method (Türk, 2011). When the transported products are industrial wood in the unit of stere (i.e. height, width and depth is a meter of wood), removal operations are carried out by means of a saddle attached to the tractor (Fig. 2). In this way, wood is loaded on the saddles mounted on the front and rear part of a tractor and carried to the landing (Türk, 2016).
The productivity of the equipment and methods used in the logging operations are generally determined depending on the operation time. The most commonly used work measurement technique in calculating the operation time is the time study method. Within the scope of the time study, the work is divided into subsections or stages, and the standard time is determined. In order to apply the time study, the work needs to be subdivided into stages. Basic tools used in time studies are chronometers and time study forms. Besides, distance measuring instruments, clinometer, and metal compass are also used in time study. Time studies are applied in three different ways; continuous time measurement, repeated time measurement and work sampling (Gülci, 2014). In the continuous time measurement technique, the work is constantly monitored and the value read from the chronometer is recorded at the end of the work stages. In the repeated time measurement technique, the chronometer is read at the beginning of the stages, and it is reset at the end and restarted. In the work sampling, the work is observed with equal time intervals and the current work stages are recorded. In this study, the productivity of a modifi ed farm tractor during the extraction of stere wood products was analyzed. The study was carried out in a Black pine stand in the city of Afyonkarahisar in Turkey. In the study, the effects of the skidding distance on productivity were also evaluated using statistical analysis.
Material and Methods. Study Area. Afyonkarahisar Forest Enterprise Chief (FEC), chosen as the study area, is located within the boundaries of Afyonkarahisar Forest Enterprise Directorate (FED) in Eskişehir Forest Regional Directorate (Fig. 3). Forest resources of the Afyonkarahisar FED are given in Tab. 2. Afyonkarahisar FED is about 172,000 hectares, of which 43,000 hectares are covered with forests. The dominant tree species in the section 428 where the extraction study was carried out was Black pine. The average ground slope in the fi eld was determined as 30%.
Time Study. The productivity of the mechanized logging equipment and systems used in harvesting operations are determined depending on the operation time. In this study, the repeated time measurement method, which is one of the most common time study methods in forestry, was used in order to analyze the productivity of the timber extraction carried out with the New Holland 55-56 S farm tractor (Gülci, 2014). Technical features of the farm tractor used in the study are given in Tab. 3.   The work measured in timber extraction is divided into sub-sections and work stages. In this study, a digital chronometer was used to measure the operation time in four work stages; 1) the time for the unloaded tractor to reach the point where the wood products are located, 2) the time to collect and load the products at the stump, 3) the time for the loaded tractor to reach the landing area, 4) the time of unloading tractor and stacking the wood products at the landing (Figure 4). The fi rst stage starts with the tractor being unloaded on the ramp and ends when it stops for loading the products at the harvesting site. At the second stage, the forest workers collect the products at the stump and load them on the tractor. The third stage starts with the movement of the tractor from the harvesting area and ends when the tractor reaches the landing area. At the last stage, the tractor is unloaded on the landing and the wood products are stacked. The lost time and delay time was not evaluated in the study. (1) (2) (4) Figure 4. The work stages of the tractor skidding: 1) the unloaded tractor to reach the point where the wood products are located; 2) collect and load the products at the stump; 3) the time of the loaded tractor to reach the landing area; 4) the time of unloading tractor and stacking the wood products at the landing.
Time study forms were used to record time measurements for the extraction method evaluated in the study. In addition to the time measurements, the skidding distances measured by a steel tape and the quantity of products (stere) stacked on the ramp were recorded on the forms. A total of 45 turns were recorded during timber extraction. The amount of product per turn in the study was 5.5 stere in 40 turns, 5 stere in four cycles, and 4 stere in one cycle. Hourly productivity (P) in stere per hour for each turn is calculated as follows: (1) v = Timber volume per turn (stere) t = Total time per turn (min) Statistical Analysis. Within the scope of hourly productivity and time measurement studies, to examine the total cycle time and the factors affecting effi ciency, fi rstly the correlation and then regression analyzes were conducted. As the skidding (forwarding) distance varied in the study, only the skidding distance was evaluated as an independent variable since the amount of product transported remained constant in 89% (5.5 stere) of the cycles. The relationship between the skidding distance variable and the cycle time and productivity were examined with the one-way ANOVA analysis at 0.05 signifi cance level.
Normality test based on the Skewness and Kurtosis coeffi cients was performed to examine the normality of the variables. In order to investigate the effect of skidding distance on cycle time and productivity, skidding distance was evaluated in three classes (short: < 200 m, medium: 200-300 m, long: > 300 m). The investigation of the relationships between the variables was carried out with the Pearson Correlation Test and the determination of mathematical models for the independent variables using the Linear Regression Analysis. SPSS 16.0 software was used to conduct statistical analysis in the study.
Results and Discussion. The basic statistical values (average, maximum, minimum and standard deviation) obtained for the productivity and the skidding distance factor are given in Tab. 4. It was found that the average skidding distance was 250 m while the average productivity of transporting stere wood with a saddle attached to the tractor was determined as 6.06 stere/hour. In a similar study where the average skidding distance was 320 m, the hourly productivity was reported as 7.7 m 3 /h in the skidding operation with a farm tractor (Öztürk, 2010). The Skewness and Kurtosis coeffi cients were examined to check if the variables showed normal distribution. The closer these values are to zero, the closer the data is to the normal distribution. In addition, when the skewness and kurtosis coeffi cients are divided by their standard deviation values, the data is normally distributed if the result is in the range of -2 to 2 (George & Mallery, 2003). The results revealed that variables show normal distribution (Tab. 5).  The basic statistical values (mean, maximum, minimum and standard deviation) of the time measurements of the work stages in the tractor skidding study are given in Tab. 6. According to the results, the average total time for skidding with the tractor was determined as 59.12 minutes. When the work stages are evaluated, it was found that the most time-consuming stage in a skidding cycle was unloading tractor and stacking the wood products at the landing (35.97%), followed by (in terms of time consumption) collecting and loading the products at the stump (33.14%).
Pearson Correlation Test was applied to reveal the relationship between hourly productivity and skidding distance with the tractor. Correlation test results are given in Tab. 7. The results showed that there was a signifi cant (p < 0.001) relationship between skidding distance and total cycle time and productivity at the 99% confi dence level. Accordingly, it was determined that as the skidding distance increases, the total cycle time increases and consequently the productivity decreases. In similar studies, it has been stated that the increase in tractor skidding distance increases the total cycle time which decreases the productivity (Behjou, Majnounian, Namiranian, & Dvorak, 2008;Kulak, Arkadiusz, & Grzegorz, 2017;Öztürk, Varsak, & Bilici, 2019). In the study, the distances were divided into three classes (short: ≤ 200 m, medium: 200-300 m, and long: > 300 m) to examine the effect of skidding distance on total cycle time and productivity in more detail. Then, using the One-Way ANOVA with 0.05 signifi cance level, the relationship between skidding distance classes and total cycle time and productivity was evaluated (Tab. 8).  When skidding distances were classifi ed, it was found that total cycle time increased from short skidding distance (48.32 min) to medium (61.32 min) and long skidding distance class (79.50 min). On the other hand, it was determined that the productivity value increased from long skidding distance class (3.96 stere/hour) to medium (5.65 stere/hour) and short skidding distance class (7.27 stere/hour).
Considering the correlation test results, the Linear Regression Analysis was used to reveal the relationship between independent variable (skidding distance) and dependent variable (total cycle time and productivity). According to the One-Way ANOVA results, it was found that the regression model related to the total cycle time and skidding distance was statistically signifi cant at 99% confi dence level (p < 0.001) (Tab. 9). The R 2 value obtained (0.819) showed that the regression model adequately explained the total cycle time for the farm tractor. The regression model with the dependent variable (y) representing the total cycle time and the skidding distance (x), which is the independent variable effective on the productivity, is shown in the following equation: y = -0,645 + 0,239x (2) According to the One-Way ANOVA results, it was found that the regression model related to total productivity and skidding distance was 99% confi dence level (p < 0.001 (Tab. 10). The R 2 value obtained (0.875) showed that the regression model adequately explained the productivity for the tractor. The regression model with the dependent variable (y) representing the productivity and the skidding distance (x), which is the independent variable effective on the productivity, is shown in the following equation:

Conclusions.
A signifi cant increase in industrial wood demand is observed on the global scale. In Turkey, the main reason for this increase is the high need for wood in the wood-based panel industry. Farm tractors that have been modifi ed in accordance with forestry activities can be used effectively in removing industrial wood product from the forest. Modifi ed farm tractors are more affordable machines in terms of initial purchase price and hourly operating costs compared to advanced harvesting machines (e.g. harvester, feller-buncher, forwarder, etc.). In this study, the productivity of a modifi ed farm tractor during the extraction of industrial wood products was analyzed by time analysis method. In addition, the effect of skidding distance on total cycle time and productivity was investigated. The study was carried out in the Black pine stand located in Afyonkarahisar Forest Enterprise Chief. In the study, the skidding distance was accepted as an independent variable and regression models were developed for total time and productivity estimates in the study of transporting products in saddles attached to the farm tractor. As a result of statistical analysis, it was determined that the most time-consuming working stage was the unloading and stacking of products on the landing. It was determined that the collecting the products and loading the products to the tractor was the second most time consuming stage. The results of the analysis showed that the productivity of wood extraction was signifi cantly affected by the skidding distance.