Soil Responses to Winch-Assisted Thinning Harvester Traffic on Steep Slopes in South Korea

Abstract

Background: Winch-assisted harvesting is an alternative to traditional cable yarding on steep slopes, offering improved operational efficiency and fewer limitations. Knowledge on the effects of winch-assisted harvesting on soil disturbance are limited. This study aimed to assess the effects of winch-assisted and conventional tracked harvester operations on soil compaction and machine slippage in a clear-cut stand with sandy loam soil. Methods: We evaluated changes in soil physical properties, in depth and extent, along machine operating corridors with and without winch-assist across slope gradients ranging from 30% to 52% and up to three machine passes. Results: The relative increase in bulk density differed between treatments. In the non-assisted corridors, the bulk density increased by 18%, 12%, and 11% at depths of 0–10, 10–20, and 20–30 cm, respectively; the winch-assisted corridors showed smaller increases of 12%, 5%, and 3% at the corresponding depths. The winch-assisted plots did not show a significant reduction in rut depth compared with the non-assisted plots, a result likely influenced by site-specific dry soil conditions. Conclusions: These results highlight the potential of winch-assisted systems to reduce horizontal soil disturbance, though their effectiveness in limiting rutting remains variable under dry conditions.

1. Introduction

In South Korea (hereafter referred to as Korea), forests cover approximately 6.3 million hectares, accounting for 64% of the total land area, with timber production forests making up 35% of the total forest area [1]. Approximately 77% of forested areas are situated on steep terrains with slopes exceeding 40%, highlighting the challenging topography of Korea’s forests [1,2]. Logging operations on steep slopes generally pose greater challenges than those on more moderate terrain [3,4,5].

Although cable logging is a viable option for steep terrains, it is often not preferred owing to its high operating costs and the significant time required for cable rigging and planning [6,7]. Consequently, timber harvesting in Korea has primarily relied on manual felling using chainsaws since the 1990s, followed by extraction using specialized methods, such as small shovels, winching, or cable logging [8]. Among these, small-shovel logging has become the preferred method due to its relatively low cost and operational simplicity compared with cable logging [9]. However, this method has been linked to frequent worker accidents and concerns regarding soil disturbance and environmental degradation [10,11]. Therefore, safer, more efficient, and environmentally responsible logging practices suitable for steep terrain in Korea remain urgently needed.

Ground-based harvesting was developed to improve the productivity of felling and extraction, as well as to enhance safety [3,12,13]. Despite these advantages, the adoption and performance of these systems are inherently constrained by terrain-related factors, such as slope gradient, soil strength, and surface roughness [14,15]. Although early guidelines suggested that wheeled skidders and crawler tractors could operate on slopes of up to 50% and 60%, respectively, later field experience recommended lowering these limits to approximately 30% and 40%, respectively, to reduce soil erosion and improve machine stability [3,16]. In response to these limitations, modern technologies, such as winch-assist systems, have been developed to support machine operations on steeper terrains by improving traction and stability.

Winch-assist technology, which employs a tensioned wire rope anchored upslope to improve traction and gradeability on steep terrain, has attracted global attention for its performance and safety advantages [17,18]. Various winch-assisted harvesting systems are utilized across Europe, New Zealand, and the Pacific Northwest region of the United States [19,20]. Winch-assisted systems have recently been introduced for thinning and clear-cutting operations in Korea. However, their widespread adoption remains limited due to the high capital requirements. Moreover, the potential adverse effects of these systems, particularly soil compaction and displacement, have not yet been thoroughly assessed relative to conventional non-winch-assisted harvesting systems [18,21]. Although empirical studies remain limited, existing research suggests that winch-assisted systems can reduce soil disturbance by improving machine stability and minimizing track slippage on steep terrain. Theoretical studies have highlighted their potential to enhance both traction and gradeability [17]. Similarly, Visser and Stampfer [3] reported that decreased slippage in winch-assisted operations may result in less soil damage than in untethered systems. These findings provide a scientific basis for hypothesizing the soil protection benefits of winch-assisted technologies. However, field-based experimental evidence remains warranted to validate these assumptions [18,22].

Therefore, in this study, we aimed to compare the effects of soil compaction and machine slippage on winch-assisted and conventional tracked harvester operations in a clear-cut stand with sandy loam soil conditions. Specifically, we analyzed the depth and extent of changes in soil physical properties across machine operating corridors with and without winch-assist, accounting for the slope gradient and the number of machine passes. To isolate the effects of machine movement alone, the harvester was operated without felling, enabling the assessment of soil impact from uphill and downhill travel within the clear-cut area. Based on previous observations and operational expectations, we hypothesized that winch-assisted harvester traffic on steep sandy loam slopes would (1) reduce the magnitude and vertical extent of soil compaction and (2) decrease crawler slip compared with conventional untethered traffic.

2. Materials and Methods

2.1. Study Area

This study was conducted in a clear-cut site where forest residues were collected, located in Yuchon-ri, Gandong-myeon, Hwacheon-gun, Gangwon-do, Korea (GPS coordinates: 127°47′52.0″ E, 38°24′52.5″ N; Figure 1). The research area covered approximately 5.3 ha and previously consisted of Pinus koraiensis plantations, which accounted for 90.2% of the area prior to harvesting. The mean diameter at breast height (DBH) and height were 21.9 cm and 10.0 m, respectively. The mean basal area was 24.4 m²/ha, and the total stand biomass (roundwood and residues) was 184.7 green tons (gt)/ha at an average moisture content of 50%, estimated from stem volume, forest biomass, and forest yield tables. The harvest unit was located on a relatively steep slope exceeding 47.7%. According to the soil particle size and texture classes, the soils were sandy loams.

The conventional cut-to-length (CTL) harvesting system at the site was implemented as a semi-mechanized operation, with two chainsaw operators conducted felling and processing at the stump, while small shovels and excavator-mounted equipment (5.0 t, 0.2 m³ bucket) performed extraction tasks. Small shovels sorted and carried merchantable 2–4 m logs together with logging residues (irregular circular forms; Figure 2) from the stump to the roadside. This small-shovel-based extraction method is a regionally specific practice widely used in mountainous forests of South Korea but is uncommon in many other regions. Further information about the operational characteristics and performance of small-shovel extraction systems can be found in previous studies [10]. Extraction activities were performed in the downhill direction; the logs and slash piles were subsequently transported to the landing by the tracked carrier.

2.2. Experimental Design and Equipment

Subsequently, experimental sites were selected, and research was conducted in areas with minimal disturbances from small shovel traffic and spur road construction. Previous studies have shown that trafficking by small shovels significantly alters soil physical properties, with bulk density increasing (by 4%–18%) and porosity decreasing (by 10%–24%) compared to undisturbed or shallowly disturbed areas [10]. For the experiment, six plots were established within the study site to maintain comparable operating conditions, such as slope gradient, soil type, and soil moisture content. This ensured that the observed differences could be attributed to the presence or absence of winch assistance rather than to environmental variability.

A field experiment was designed, as shown in Figure 3, to assess the differences in soil disturbance and machine slippage when operating on mountainous slopes under winch-assisted and non-assisted conditions. Each experimental plot consisted of twos corridors: one in which the machine traveled the slope with winch assistance, and the other without winch assistance. Three round-trip passes of machine were conducted in each corridor, with the machine traveling uphill and returning downhill along the same route. Soil sampling and rut-depth measurements were performed after each pass.

The winch-assist system employed in this study was a KDH-40 (Konrad Forsttechnik GmbH, Preitenegg, Austria), a purpose-built forestry machine weighing approximately 11 t equipped with 450 mm track width, designed for mechanized harvesting on steep terrain. It is equipped with a tethered winch mechanism that provides supplementary traction by anchoring it to an upslope tree or stump (Figure 4). Notably, the system allows for self-anchoring, enabling the operator to secure the cable to suitable anchor points (e.g., standing trees) without external assistance. The winch provides a pulling force of 30 kN and operates with an 80 m 11 mm swaged wire rope. This configuration enhances machine stability, reduces slippage, and enables safer operation on slopes beyond the limits of conventional ground-based methods [17].

2.3. Data Collection

To compare the extent of soil disturbance caused by machine traffic on the slopes, changes in soil bulk density and rut depth were measured. Soil samples were collected after each machine passage to determine the bulk density.

In each corridor, five transects were established at 5 m intervals. Samples were extracted from two survey zones along each transect: (i) a control area where machine tracks did not pass, and (ii) a wheel-track zone where tracks passed (Figure 5). Soil samples from the wheel-track zone were consistently collected from the left track across all corridors.

Sampling in each survey zone involved removing the organic layer (O horizon) and extracting soil cores of 100 mL at depths of 5, 15, and 25 cm, corresponding to the 0–10, 10–20, and 20–30 cm soil layers, respectively. Samples were oven-dried in the laboratory in a forced-convection oven at 105 °C for 15 or more hours, until the mass loss after 1 h of additional drying was less than 0.1%, to determine dry weight [23]. Bulk density (g$·$ cm⁻³) was calculated by dividing the dry weight by the sample volume of 100 cm³ (Equation (1)). The bulk density increase rate (%) in the wheel was calculated relative to the bulk density of the control zone for each transect (Equation (2)).

$$\mathrm{B}\mathrm{u}\mathrm{l}\mathrm{k} \mathrm{d}\mathrm{e}\mathrm{n}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{y} (\mathrm{g}·{\mathrm{c}\mathrm{m}}^3)={\displaystyle \frac{\mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t} \mathrm{o}\mathrm{f} \mathrm{d}\mathrm{r}\mathrm{y} \mathrm{s}\mathrm{o}\mathrm{i}\mathrm{l}\left(\mathrm{g}\right)}{100 \left({\mathrm{c}\mathrm{m}}^3\right)}}$$

(1)

$$\mathrm{B}\mathrm{u}\mathrm{l}\mathrm{k} \mathrm{d}\mathrm{e}\mathrm{n}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{y} \mathrm{i}\mathrm{n}\mathrm{c}\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{s}\mathrm{e} \mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e}\left(\mathrm{\%}\right)={\displaystyle \frac{{\mathrm{B}\mathrm{u}\mathrm{l}\mathrm{k} \mathrm{d}\mathrm{e}\mathrm{n}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{y}}_{\mathrm{W}\mathrm{T},\mathrm{M}}(\mathrm{g}·{\mathrm{c}\mathrm{m}}^{-3})-{\mathrm{B}\mathrm{u}\mathrm{l}\mathrm{k} \mathrm{d}\mathrm{e}\mathrm{n}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{y}}_{\mathrm{R}}(\mathrm{g}·{\mathrm{c}\mathrm{m}}^{-3})}{{\mathrm{B}\mathrm{u}\mathrm{l}\mathrm{k} \mathrm{d}\mathrm{e}\mathrm{n}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{y}}_{\mathrm{R}}(\mathrm{g}·{\mathrm{c}\mathrm{m}}^{-3})}}\times 100$$

(2)

The rut depth was measured at both left- and right-hand wheel tracks along each transect. The mean of the two measurements was used for the analysis. Rut depth was defined as the vertical distance from the original ground surface to the center of the track depression, excluding the height of the soil displaced upward, and was measured using a leveling rod and folding ruler ([24]; Figure 5).

To quantify crawler slip, we measured the crawler linear velocity (Vc) and ground-referenced travel velocity (Vt). Two incremental rotary encoders (E58H12-360-3-N-5, Autonics Co., Ltd., Busan, South Korea; Figure 6) were mounted on the left and right travel-motor shafts to estimate Vc (m·s⁻¹), and a multi-band GNSS receiver (ZED-F9P, u-blox Holding AG, Thalwil, Switzerland; Figure 7) was installed above the machine center to obtain Vt (m·s⁻¹).

The sampling frequency was set at 1 Hz. Time stamps from the encoders and GNSS were synchronized, and records with encoder dropouts or GNSS loss of fix were removed before analysis. The crawler slip ratio (CS) was computed as:

$$\mathrm{C}\mathrm{r}\mathrm{a}\mathrm{w}\mathrm{l}\mathrm{e}\mathrm{r} \mathrm{s}\mathrm{l}\mathrm{i}\mathrm{p} \mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}, \mathrm{C}\mathrm{S} (\mathrm{\%})=1-{\displaystyle \frac{\mathrm{V}\mathrm{t}\left(\mathrm{m}·{\mathrm{s}}^{-1}\right)}{\mathrm{V}\mathrm{c} (\mathrm{m}·{\mathrm{s}}^{-1})}}\times 100$$

(3)

2.4. Statistical Analysis

All statistical analyses were performed in R software (version 4.3.1) with a significance level of α = 0.05. Data normality was assessed using normal quantile plots and the Shapiro–Wilk test. To examine the effects of both winch-assisted and conventional (non-assisted) machine traffic, two-sample t-tests, assuming unequal variances, were conducted at each soil depth to compare bulk density and displacement between trafficked and non-trafficked areas. Additionally, a one-way ANOVA was used to evaluate differences in bulk density across different numbers of machine passes; however, no significant differences were observed; therefore, post hoc analyses were not performed. Crawler slip (CS, %) was analyzed using ANCOVA with Assistance (winch-assisted vs. non-assisted) as a fixed factor and local slope gradient (%) as a covariate; models were fitted separately for uphill and downhill segments. The model assumptions were checked, and no material violations were detected.

3. Results

3.1. Soil Characteristics of Machine Operating Corridors After Clear-Cut

Soil bulk density measured prior to machine traffic did not significantly differ between trials with and without the winch-assist, indicating similar baseline conditions for assessing soil physical properties (p > 0.05;

Table 1. Descriptive statistics for bulk density and total porosity of undisturbed areas in tethered and untethered corridors.

	Depth	Tethered Corridors	Untethered Corridors	p-Value
Bulk density	(cm)	(g·cm−3)	(g·cm−3)
	0–10	1.04 ± 0.15	1.03 ± 0.16	>0.05
	10–20	1.15 ± 0.14	1.17 ± 0.13	>0.05
	20–30	1.22 ± 0.12	1.21 ± 0.11	>0.05
Moisture	(cm)	(%)	(%)
	0–10	17.1 ± 6.9	17.8 ± 7.3	>0.05
	10–20	15.7 ± 2.9	15.9 ± 1.9	0.021
	20–30	15.1 ± 2.8	15.8 ± 2.5	>0.05

). Although a statistically significant difference in soil moisture content was observed at the 10–20 cm soil depth before machine intervention (p-value = 0.021), the values were generally comparable across the winch-assisted and non-assisted corridors. Therefore, the chosen trial sites likely exhibited similar soil physical and hydrological conditions prior to the machinery use operation.

3.2. Soil Disturbance

Post-harvester traffic data were compared with control data across both untethered and tethered corridors at all depths and sample sublocations (Figure 8). Under non-assisted trafficking, significant increases in bulk density were observed in track measurements at depths of 0–10, 10–20, and 20–30 cm between tracks (p < 0.01). Untethered tracks showed the least significant increase at depths of 20–30 cm, whereas in track measurements, the bulk density increased by 18% at 0–10 cm, 12% at 10–20 cm, and 11% at 20–30 cm. The bulk density also increased with increasing soil depth. In addition, increasing the number of machine passes elevated the bulk density. However, no significant differences were observed at depths of 0–10 and 20–30 cm (p > 0.05). At depths of 10–20 cm, a significant difference was detected between the different pass numbers (p = 0.041).

The winch-assisted corridors exhibited increases in bulk density at soil depths of 0–10, 10–20, and 20–30 cm between the tracks. At this site, the largest increase in bulk density was detected in the surface layer (12% at 0–10 cm, 5% at 10–20 cm, and 3% at 20–30 cm); however, these increases were not significant (p-value > 0.05). In addition, although bulk density tended to increase with the number of machine passes, no significant differences were observed at any soil depth (0–10, 10–20, and 20–30 cm; p > 0.05).

The vertical soil displacement resulted in an average rut depth increment per machine pass of 3.48 ± 2.22 cm. Soil displacement was observed in 40.0 and 26.7% of the nonassisted and winch-assisted corridors, respectively. After three machine passes, rut depth reached 9.99 ± 3.20 cm in the non-assisted section of the machine operating trail and 10.98 ± 2.22 cm in the winch-assisted section (

Table 2. Mean (±standard deviation) rut depth collected at the non-assisted and winch-assisted corridors.

	Non-Assisted Corridor	Winch-Assisted Corridor	p-Value
Machine passes 1	6.36 ± 2.72 cm	6.79 ± 2.39 cm	0.668
Machine passes 2	8.53 ± 3.64 cm	9.15 ± 1.93 cm	0.631
Machine passes 3	9.99 ± 3.20 cm	10.98 ± 2.22 cm	0.421

); however, these differences were not significant (p > 0.05). Therefore, deep ruts do not occur predominantly in the non-assisted plots.

3.3. Crawler Slippage

The average crawler slip during uphill travel without winch assistance was 25.9 ± 9.5%, whereas the winch-assisted condition showed a significantly lower value of 17.0 ± 4.5% (ANCOVA, p < 0.05). During downhill travel, the average crawler slip without the winch was 11.7 ± 4.3%, while the winch-assisted condition exhibited a higher value of 13.4 ± 1.1%, showing a significant difference (ANCOVA, p < 0.05). The relatively lower slip observed during downhill travel without winch assistance likely resulted from the overall sliding motion of the machine, which reduced the measured track–slip ratio. In contrast, the use of the traction winch increased the traction control and machine stability, thereby enhancing the safety of downhill operations (Figure 9). When the influence of slope was considered, uphill slip under non-assisted conditions showed a clear tendency to increase on steeper gradients, whereas winch assistance effectively constrained slip even on slopes of 40%–50%. Conversely, downhill travel did not exhibit a consistent monotonic relationship between slip and slope.

4. Discussion

The application of winch assistance in ground-based harvesting systems has been shown to alter the interaction between forestry machinery and the underlying soil, potentially reducing soil disturbances on steep terrains [4,22]. The findings of this study indicated that winch-assisted operations provide a horizontal benefit by reducing track slippage, thereby mitigating soil disturbances in the direction parallel to the ground surface. Across all treatments, the magnitude of soil compaction decreased with depth, indicating a diminishing influence of surface loading on deeper layers. However, in the vertical direction, the rut depth did not significantly differ between the winched and conventional corridors. These contrasting results suggest that the effects of winching assistance may be directionally dependent, with the benefits being more pronounced in the horizontal plane. This pattern is consistent with the mechanics of machine–soil interaction, as slip is strongly governed by traction demand, which increases when the gravitational downslope component becomes larger on steeper gradients [3].

This directional pattern is consistent with the results of previous studies. Winch-assisted systems are thought to improve machine stability and reduce soil disturbances [17]. Green et al. [21] attributed reduced soil compaction to decreased track wander under cable tension. Similarly, Allman et al. [25] observed smaller increases in bulk density in winch-assisted operations (up to 17.39%) than in unassisted operations (up to 22.35%). Garren et al. [4] reported a 16% average increase in rut bulk density on slopes of 27–38°. In this study, cable-assisted operations resulted in a smaller increase in penetration resistance, suggesting a reduced shear load on the soil surface owing to improved traction and reduced slippage. Taken together, these results support the potential of winch-assisted systems to mitigate mechanical stress and soil compaction during steep-slope harvesting.

While previous studies highlighted the importance of traffic frequency, especially during the initial passes [26,27], our regression results showed no significant effect of the number of passes. A possible explanation is inconsistent machine travel paths. Field observations indicated lateral deviations between passes [21], which may have contributed to a broader zone of influence. Similar degrees of track wandering were observed in both winched and non-winched operations, suggesting that tethered systems may not consistently limit lateral deviations under all conditions.

Interestingly, the rut depth measurements in this study did not support the commonly reported trend of reduced soil impact with winch assistance. The rut depths were relatively shallow across treatments and comparable to or lower than those previously reported [28,29]. Notably, Garren et al. [4] observed rut depths ranging from 0 to 9.5 cm in winched-forwarder trials. In contrast, our study showed a slightly greater rut depth in the winch-assisted plots. The effectiveness of traction-assist systems in reducing rutting can vary widely depending on site-specific factors such as soil texture and soil moisture, and increases in traction only translate into reduced rutting under favorable soil conditions [17]. Poltorak et al. [30] and Rollerson [31] emphasized the role of soil moisture in rut formation, with wetter conditions leading to deeper ruts. Given that our experiment was conducted during a dry summer, overall trafficability was high, potentially minimizing the benefits of winch-assisted reduction in vertical soil disturbance.

These findings should be interpreted considering several contextual factors. First, the observed responses reflected machine travel alone, without the added effects of active felling or forwarding. Second, the dry-site conditions likely reduced the extent of compaction and rutting, limiting insight into the effectiveness of winch-assisted operations under wetter, more vulnerable soil conditions. Third, although lateral deviations in machine travel were observed, they were not quantified, thereby preventing a full spatial analysis of the track wander. Finally, the study was confined to a single forest site with a moderate slope range (30%–52% gradients), and machine operation parameters (e.g., winch tension, travel speed, and operator behavior) were not standardized. These factors might have introduced uncontrolled variability, limiting the generalizability of the results.

To strengthen these insights, future studies should examine a wider range of operational scenarios, including varying soil moisture conditions, and multiple site types such as steeper slopes, finer-textured soils, and different forest stand compositions. Standardizing the machine parameters and quantifying the lateral machine movements can improve the applicability and generalizability of the results.

5. Conclusions

Our study compared the site impacts of harvester traffic with and without cable assistance. Although limited to a single case study, the results demonstrate that:

(1)

Tethered operations reduce the spatial extent of machine-induced compaction, particularly in the surface layer;

(2)

Crawler slip during both uphill and downhill travel was significantly lower under winch assistance;

(3)

Rut occurrence was less frequent in the winch-assisted corridors, although rut depth did not differ significantly between treatments.

These findings indicate that traction-assisted technology has strong potential for mitigating soil disturbance under unfavorable field conditions.