Effectiveness of water diversion and erosion control structures on skid trails following timber harvesting

Ecological Engineering 105 (2017) 370–378

Contents lists available at ScienceDirect

Ecological Engineering journal homepage: www.elsevier.com/locate/ecoleng

Effectiveness of water diversion and erosion control structures on skid trails following timber harvesting Ali Masumian a , Ramin Naghdi b,∗ , Eric K. Zenner c a b c

Department of Forestry, University Campus 2, University of Guilan, Rasht, Iran Department of Forestry, Faculty of Natural Resources, University of Guilan, P.O. Box 1144, Sowmeh Sara, Iran Department of Ecosystem Science and Management, Penn State University, University Park, PA, USA

a r t i c l e

i n f o

Article history: Received 29 November 2016 Received in revised form 10 April 2017 Accepted 12 May 2017 Keywords: Best management practices Runoff Soil loss Water bar

a b s t r a c t Sediment in forested watersheds is produced primarily from highly disturbed areas such as skid trails. Forestry best management practices (BMPs) have been developed to minimize erosion and water quality problems, but the efficacies of various BMP options such as water bars are not well documented. The aim of this study was to evaluate the effects of different distances (slope lengths) between water diversion structures (water bars) on runoff volume and soil loss on different skid trail gradients on two soils with different textures. The treatments were located in an Iranian temperate forest and included combinations of three levels of trail gradient (<10%, 10–20% and >20%), four different distances between water bars (25, 50, 75, and 100 m), and two soil textures (clay loam and silt loam). Results showed that runoff volume increased curvilinearly and soil loss linearly with distances between water bars regardless of the soil texture and trail gradient. The greater distances on trail gradients >20% resulted in the highest amounts of runoff and soil loss; shorter distances on trail gradients <10% resulted in the lowest runoff and soil loss amount for the two tested soil textures. On the clay loam soil, 50 and 75 m were the most effective distances between water bars for trail gradients >20 and <20%, respectively. On the silt loam soil, 25 m and 50 m were the most effective distances between water bars for trail gradients >20 and <20%, respectively. The results of our study confirm that slope angle is a primary factor in controlling surface runoff and soil loss on skid trails and that soil texture becomes increasingly important as slope gradients become steeper. Therefore, reducing skid trail slope during construction skid trails is recommended to decrease surface runoff and soil loss in forest operations. Further, BMPs should consider soil texture in addition to slope gradient when recommending spacing between water bars. © 2017 Elsevier B.V. All rights reserved.

1. Introduction In forest stands, compaction following forest operations that employ large, heavy, powerful machinery is one of the main causes of soil disturbance (Brais, 2001; Rohand et al., 2004; Najafi et al., 2009; Naghdi et al., 2015). Skidding operations typically alter the physical soil structure and hydrology by increasing soil bulk density and soil strength, breaking down aggregates, decreasing porosity and pores sizes within the soil, and decreasing aeration and infiltration capacity, which can ultimately increase the potential for water runoff and erosion (Najafi et al., 2009; Solgi et al., 2014; Naghdi et al., 2015, 2016). These changes alter the way air and water move

∗ Corresponding author. E-mail addresses: [email protected] (A. Masumian), [email protected], [email protected] (R. Naghdi), [email protected] (E.K. Zenner). http://dx.doi.org/10.1016/j.ecoleng.2017.05.017 0925-8574/© 2017 Elsevier B.V. All rights reserved.

through the soil as well as the ability of roots to grow in the soil (Richard et al., 2001). Although soil compaction can change many important soil physical properties, perhaps the most detrimental effect is a drastic reduction in hydraulic conductivity, which ultimately facilitates increased levels of soil erosion due to reduced infiltration, increased runoff and poor drainage (Solgi et al., 2014). The primary mechanism of increased soil erosion following compaction appears to be due to the noticeable increase in surface runoff (Greacen and Sands, 1980). Natural soil erosion rates in forested areas tend to be very low, but erosion on cut and fill slopes and the road surface following road construction and small changes in soil compaction following skidding may enhance adverse consequences on runoff and erosion (Ramos-Scharrón and MacDonald, 2005; Solgi et al., 2014). Traffic on forest roads and ground skidding systems on skid trails are major sources of sediment generation in forest settings (Jusoff and Majid, 1996; Hartano et al., 2003). For example, in a water-

A. Masumian et al. / Ecological Engineering 105 (2017) 370–378

shed in the southeast U.S., road prisms, i.e., road surface, ditches, and banks accounted for the vast majority (80%) of sediment delivery to streams and rivers (Van Lear et al., 1995). Soil erosion often follows enhanced physical impacts of raindrops on soil surfaces that can cause surface runoff (Froehlich, 1995) and enhanced sedimentation is the consequence of increased soil erosion and mass movements after heavy storms or prolonged rainy periods (Ziegler and Giambelluca, 1997; Gucinski et al., 2001; LaMarche and Lettenmaier, 2001; Jordan-Lopez et al., 2009). Nonetheless, forest roads and skid trails are necessary to provide managers access to harvest areas (Akay et al., 2008). However, the standard to which roads are constructed differs and depends on the proposed end use, the amount of harvestable and marketable wood per unit area, and terrain conditions (Grace and Clinton, 2007). Forest roads are typically built to a higher standard than skid trails, including regular use of gravel for road surfaces, crowning, ditches and culverts, and water bars typically lacking in skid trails (Stringer et al., 1998; Grace and Clinton, 2007; NRCS, 2014). Skid trails on the other hand are typically cleared areas within the forest (NRCS, 2014) and are repeatedly used by machinery to carry harvested trees to the landing/main roads during the harvesting operation, leading to high frequencies of machinery traffic over the same trail (Zenner et al., 2007). Skid trails are typically located in areas where proper road construction operations cannot be performed; thus they have the potential to contribute large amounts of sediment to waterways until the trail is properly protected after termination of the skidding operations (Grace and Clinton, 2007). Roads of minimum or below standard as well as skid trails often accelerate soil erosion losses over time or lead to sudden mass failures that can introduce large quantities of sediment into the waterways and reduce water quality (Patric, 1976; Yoho, 1980; Swift, 1985; Binkley and Brown, 1993; Grace et al., 1998; Grace, 2002, 2005; Grace and Clinton, 2007). In fact, most of the erosion features in forests connected to stream channels may actually originate from skid trails (Litschert and MacDonald, 2009) due to altered subsurface hydrology and decreased hydraulic conductivity on compacted and less permeable surfaces that can result in erosion during rain events (Croke et al., 2001; Croke and Mockler, 2001; Jackson et al., 2005; Grace, 2005). The severity of adverse impacts of skid trails is further related to the slope gradient (Akbarimehr and Naghdi, 2012a,b), traffic volume (Akay et al., 2008; Solgi et al., 2014), vegetation cover (Cerdà, 2007; Lee et al., 2013), mechanical pressure (Battiato et al., 2013), road surfacing material (Akay et al., 2008), seasonality and rainfall intensity (Martínez-Zavala et al., 2008), soil texture (Croke et al., 2001), and the time since construction of the skid trail (Fu et al., 2010). Although soil erosion is affected by many factors, the soil type/texture, the gradient of the slope, and the presence/absence of ground vegetation seem to have pivotal roles in determining soil erodibility potential (Morgan, 1986). Soil texture determines the susceptibility of a soil to erosion in that erosion rates can differ among various soil types under the same conditions of rainfall intensities, slope gradients and amounts of vegetation cover (Hussein et al., 2007; Mohamadi and Kavian, 2015). The slope gradient plays an important role in soil loss (Jordan-Lopez et al., 2009), because runoff velocity can increase with the increasing slope gradients and lead to excessive soil erosion/loss (Koulouri and Giourga, 2007; Kateb et al., 2013). Indeed, sediment yields per unit area of machine operating trail strongly depend on the gradient of the trail (Akbarimehr and Naghdi, 2012a), while in contrast even the largest storm events do not generate any runoff and sediment on flat control plots (Solgi et al., 2014). To minimize runoff and soil erosion, Best management practices (BMPs) typically recommend implementation of soil erosion control practices designed to minimize the delivery of sediment and pollutants to natural drainage lines (Wallbrink and Croke, 2002).

371

While it is customary to design water diversion structures for forest roads, the importance of skid trails as sources of runoff and sediments has not been sufficiently considered (Martínez-Zavala et al., 2008), even though machine operating trails may have greater potential for erosion due to less elaborate water control structures (Wear et al., 2013). BMPs often recommend water bars as diversion structures that are quick and easy to construct, particularly in places where other best management practices such as ground cover by slash, vegetation, or similar treatments are deemed ineffective for controlling soil erosion and sediment discharge because of soil condition, slope gradient, slope length, and costs (Stringer et al., 1998; Wade et al., 2012). Water bars are intended to slow the speed of flowing water and divert flowing water from a road or retired skid trail to adjacent forest land (Miller, 2006) and can be effectively adopted in areas with large rainfall amounts (Wade et al., 2012). As a consequence, water bars are a very effective sediment control strategy capable of reducing runoff generation and limiting sediment yield and delivery to adjacent areas (Wallbrink and Croke, 2002). Although water bars seem to be a particularly beneficial instrument for erosion control on retired skid trails in mountainous forests that exhibit steep slopes and road gradients (Akbarimehr and Naghdi, 2012a,b), there is no consensus on recommended distances of water bars among different BMPs (Copstead et al., 2003). Further, recommended maximum distances between water bars do not depend on soil texture (NRCS, 2014), even though soil texture is a major determinant of soil erodibility potential (Morgan, 1986). To expand research of a recent study that determined the distances between water bars that minimized runoff and soil loss for a clay-loam soil (Akbarimehr and Naghdi, 2012b), this study aims to determine the distances between water bars that minimize runoff and soil loss for skid trails on two different soil textures and three different slope gradients. Thus, the specific aims of this study were to (i) quantify the amount of runoff and soil loss generated at different distances between water bars on clay loam and silt loam on three slope gradient classes, (ii) determine potential interactions among soil textures, slope gradient, and distances, (iii) identify the distances between water diversions on skid trails that minimize runoff and soil loss for soils of different textures and for different slope gradients, and (iv) discuss the relationship between soil texture and soil loss.

2. Material and methods 2.1. Study area This research was conducted between November 2015 and January 2016 in Shenrood forest, Guilan province, northern Iran (36◦ 13 N and 36◦ 15 N and 53◦ 10 E and 53◦ 15 E) (Fig. 1). The area is predominantly covered by stands of oriental beech (Fagus orientalis Lipsky) and common hornbeam (Carpinus betulus) with canopy cover of 0.81 (Site 1) and 0.79 (Site 2). The area is characterized by brown forest soils formed on unconsolidated limestone. Soils have a moderately deep profile and are classified as Eutric Cambisols (FAO/UNESCO, 1990) and Typic Eutrudepts (USDA Soil Taxonomy, 1998). Soil textures in the studied skid trails were determined based on particle size analysis using the Bouyoucos hydrometer method (Kalra and Maynard, 1991) and classified as a clay loam (Site 1) and silt loam (Site 2) soil (Table 1). The average depth of the soil to bedrock ranged between 60 cm (Site 1) and 70 cm (Site 2). The elevations of the two study sites were approximately 900–1100 m above sea level with a northerly aspect. The average annual rainfall recorded at the closest national weather station, located 20 km from the research area is 1130 mm, with a maximum mean monthly rainfall of 140 mm in October and a minimum rainfall of 25 mm in

372

A. Masumian et al. / Ecological Engineering 105 (2017) 370–378

Fig. 1. Study area.

Table 1 Soil particle size distribution with corresponding soil textures for both research sites at a depth of 1–10 cm. The range of particle size was <0.002, 0.002–0.05 and 0.05–2 mm for clay, silt, and sand, respectively. Sample site

Site 1 Site 2

Soil particle size distributions (g 100 g−1 )

Soil texture

Sand

Silt

Clay

25 32

36 52

39 16

August (Naghdi et al., 2015). The mean annual temperature is 16 ◦ C, with lowest temperatures in February based on historic weather data of the previous 50 years. At the time of skidding, weather conditions were dry and warm with an average gravimetric soil moisture content of 23% (Site 1) and 20% (Site 2). To our knowledge and based on a pre-harvest survey, the site did not show any signs of previous timber harvesting or that the soil had been driven on before the experiment. At each of the two study sites, a combination of group selection and single tree selection silvicultural treatments were applied (Naghdi et al., 2016). In Hyrcanian forests, harvesting and silviculture operations are most commonly applied in the autumn and winter and the extraction of logs is usually completed in the spring and summer (Naghdi et al., 2016). Harvesting operations were performed by hand-felling and processing, followed by transportation of logs from the forest stand to the roadside by a pneumatic tyre “Timberjack 450C” cable skidder. After skidding, ten randomly placed sample plots (1 m2 area each) along the skid trails revealed that no forest floor or litter material remained on the skid trails (i.e., bare mineral soil), which were not treated with slash or other treatments. Further, no obvious surface features such as rutting or flow channels associated with surface runoff were detected on the skid trails.

2.2. Experimental design This study evaluated the amount of runoff and soil loss using four various distances between water diversion structures (water bars)

Clay loam Silt loam

on both sites (i.e., soil textures) to determine the distance between water bars that would provide the most effective erosion control on skid trails. Distances between water bars were 25 m, 50 m, 75 m, and 100 m. Because the longitudinal slope gradient can strongly affect water movement on skid trails, a skid trail was laid out at each site that covered a range of various slope gradients. Three longitudinal trail gradient classes were considered: <10% (actual range 3–7%), 10–20% (range 13–16%), and >20% (range 23–28%). After mapping the entire skid trail longitudinal profile in 10 m sections, the plots were established on the skid trails in form that the average slope of including sections within each plot were inside the mentioned ranges. For example, a plot was assumed in the first class once the average slope of all including sections was under 10%. Further, only those sections of the skid trail without lateral gradient were eligible to be chosen for this study. This was ensured by placing six transects perpendicular to the skid trail in potential plot areas and measuring the lateral gradient of each transect. Only plots without lateral gradient were retained for further study. Although this study was done as part of a commercial harvesting operation under real-life conditions, the harvesting operation was halted after the total traffic frequency of the loaded skidder in each study site reached 20 passes. This was done to ensure comparability along the skid trail profile and between sites. A total of 72 runoff plots were installed that included 24 combinations of two levels of soil texture (T), three levels of trail gradient (G), and four levels of distance (D) between water bars (2 (T) × 3 (G) × 4 (D) × 3 replicates). Each runoff plot was 4 m wide, with a minimum buffer zone of 10 m between plots. Runoff plots were

A. Masumian et al. / Ecological Engineering 105 (2017) 370–378

373

Fig. 2. Layout of the sample plots.

surrounded by wooden boards that were 30 cm tall and inserted 10 cm deep into the soil to control surface water movement from the inside to the outside of the plot area and vice versa. A ditch was constructed on the lower side of each plot so that all surface water runoff from inside the area could be collected in a tank with a capacity of 250 l (Fig. 2). A collecting trough made of a metal sheet that was covered with plastic or sheet metal to prevent direct entry of rainfall was positioned at the downslope end of each plot. Data were collected from a total of eight rainfall events; the time between these rainfall events was between one to two weeks. After each rainfall event and transferring the runoff material to the lab for further processing, tanks were emptied and cleaned. The volume of the surface runoff was determined by measuring the height of the water in the collecting tanks. Samples were taken to the laboratory, where after thoroughly mixing the water runoff and bringing all the sediments into suspension, a subsample of 1 l of collected runoff material was extracted, filtered under a vacuum through Whatman no. 1 filter paper, ovendried at 105 ◦ C for a day, and then weighed (i.e., the drying and weighing method; Sadeghi et al., 2007). The total amount of soil loss (g) per plot was computed by multiplying the dried weight of sediments per 1 l sample by the volumes (liters) collected in the tank of each plot. Finally, the total amount of runoff and soil loss per plot were divided by the total plot area and analyzed as runoff (mm) and soil loss (g) per m2 . Runoff volume and sediment loss from each plot were computed for each rainfall event.

2.3. Statistical analysis Mixed model (repeated measures) analysis was used to assess the significance of observed differences in average runoff and soil loss in runoff plots from the eight runoff events as a function

of distance between water bars on different trail gradients and soil textures and their interaction effects at a significance level of ˛ ≤ 0.05. Tukey’s HSD test was used to compare the amount of runoff and soil loss among the three trail gradients and the four different distances between water bars (main effects of categorical/class variables). Mixed model analysis was chosen because repeated observations (i.e., several rainfall events) occurred over time in the same plots. In this analysis, soil texture, distances classes, and slope gradient were considered as fixed effects and plots as random effects. Repeated measures analysis was also used to quantify the functional relationship between runoff amounts and soil loss and distances between water bars and slope gradients (both treated as continuous variables) for both soil textures and between amounts of soil loss and runoff. All statistical calculations were performed using SAS version 9.3.

3. Results Following skidding operations, average surface runoff volume, which ranged between 0.15 and 0.84 mm m−2 , was significantly affected by the main effects (all P ≤ 0.001), all two-way interactions (all P ≤ 0.040), and the three-way interaction of soil texture × trail gradient × distance (P ≤ 0.001). Similarly, average soil loss, which ranged from 0.9 to 25.9 g m−2 , was significantly affected by the main effects (all P ≤ 0.001), all two-way interactions (all P ≤ 0.003), but not the three-way interaction of soil texture × trail gradient × distance (P = 0.431) (Table 2). On both soil textures and all trail gradients, average amounts of runoff decreased with increasing distances between water bars and increased with increasing trail gradients at all distances between water bars (Table 3). In contrast, and regardless of soil texture, average amounts of soil loss increased with increasing distances

374

A. Masumian et al. / Ecological Engineering 105 (2017) 370–378

Table 2 P values based on analysis of variance of the effects of soil texture, skid trail gradient, and distance between water bars on runoff volume (l) and soil loss (g m−2 ). Source of variable

d.f.

Soil texture Gradient Distance Soil texture × Gradient Soil texture × Distance Gradient × Distance Soil texture × Gradient × Distance *

1 2 3 2 3 6 6

P values* Runoff

Soil loss

≤0.001 ≤0.001 ≤0.001 0.013 0.037 ≤0.001 ≤0.001

≤0.001 ≤0.001 ≤0.001 0.002 ≤0.001 0.003 0.431

P values less than 0.05 are given in bold.

between water bars on all trail gradients and increased with increasing trail gradients at all distances between water bars (Table 4). The longest distance on trail gradients >20% resulted in the greatest, and the shortest distance on trail gradients <10% resulted in the lowest amounts of soil loss on both soil textures, with intermediate amounts of soil loss at intermediate distances. The effect of distance between water bars on the amounts of runoff and soil loss varied with trail gradient and depended on the soil texture. On the clay loam soil, the amounts of runoff were significantly greater at distances of 25 m than all other distances, which did not significantly differ among each other. The smallest amounts of runoff were observed at distances between water bars of 75 m on slopes up to 20% (0.15–0.16 mm m−2 ) and at distances of 50 m for slopes >20% (0.37 mm m−2 ). On the silt loam soil, the amounts of runoff were also greater at distances of 25 m than all other distances. Amounts of runoff did not differ significantly between distances of 75 m and 100 m at any slope gradient. The smallest amounts of runoff were observed at distances between water bars of 50 m on slopes up to 20% (0.27–0.30 mm m−2 ) and at distances of 75 and 100 m for slopes >20% (0.54 mm m−2 ). Averaged across both textures and all three slope gradients, runoff was 0.54 mm m−2 at 25 m, 0.34 mm m−2 at 50 m, 0.33 mm m−2 at 75 m, and 0.36 mm m−2 at 100 m distances.

On the clay loam soil, the effect of distances between water bars on the amounts of soil loss were smaller at distances of 25 m, but did not differ significantly between distances of 25 and 75 m on trail gradients <20% or between distances of 25 and 50 m on trail gradients >20%. The amounts of soil loss were significantly greater at a distance of 75 compared to 50 m on slope gradients >20% as well as between distances of 100 m and shorter distances on all slope gradients. The smallest amounts of soil loss were observed at distances of 25 m on all slopes (0.9–3.9 g m−2 ). On the silt loam soil, the amounts of soil loss were also lowest at distances of 25 m, did not differ significantly between distances of 25 and 50 m on trail gradients <20%, but did significantly increase between distances of 50 and 75 m and between 75 and 100 m. On trail gradients >20%, the amounts soil loss differed significantly among all distance classes. The smallest amounts of soil loss were observed at distances of 25 m on all slopes (1.3–5.7 g m−2 ). Averaged across both textures and all three slope gradients, soil losses were 2.5 g m−2 at 25 m, 3.6 g m−2 at 50 m, 7.8 g m−2 at 75 m, and 12.7 g m−2 at 100 m distances. Averaged across all distances and slope gradients, the amounts of runoff were significantly greater on silt loam (0.47 mm m−2 ) than clay loam (0.31 mm m−2 ). This was also the case when compared by distances and/or slope gradients (Table 3). Similarly, averaged across all distances and slope gradients, the amounts of soil loss were significantly greater on silt loam (9.4 g m−2 ) than clay loam (3.9 g m−2 ). This was also the case when compared by distances and/or slope gradients (Table 4). Differences in the amounts of soil loss on the silt loam versus the clay loam texture increased strongly with the distance between water bars and, averaged across soil gradients, was nearly four times greater on silt loams compared to clay loams at distances between water bars of 100 m. Runoff and soil loss increased very little from slope gradients <10% to 10–20% and did not differ significantly on both soil textures and at any distance between water bars. Runoff and soil loss were, however, significantly greater at slope gradients >20% than slope gradients <20%, on both soil textures and at any distance between water bars. The increase in slope gradient from <20% to >20% resulted in greater relative changes in runoff and soil loss on

Table 3 Means (±std) of runoff volume (mm m−2 ) on different soil textures, trail gradients, and distances between water bars. Gradient (%)

Soil texture Clay loam

Silt loam

Distance between water bars (m)

0–10 10–20 >20

Distance between water bars (m)

25

50

75

100

25

50

75

100

0.36aB ± 0.04 0.39aB ± 0.05 0.66bB ± 0.07

0.20aA ± 0.03 0.21aA ± 0.04 0.37bA ± 0.05

0.15aA ± 0.02 0.16aA ± 0.02 0.39bA ± 0.04

0.19aA ± 0.03 0.21aA ± 0.02 0.43bA ± 0.03

0.48aC ± 0.04 0.49aC ± 0.05 0.84bD ± 0.08

0.27aA ± 0.03 0.30aA ± 0.05 0.67bB ± 0.06

0.34aA ± 0.03 0.38aB ± 0.04 0.54bA ± 0.04

0.39aAB ± 0.03 0.38aB ± 0.03 0.54bA ± 0.04

Note: Different letters within each treatment show significant differences (P < 0.05). Capital case letters refer to the comparisons among the four distance classes at different trail gradients for each soil texture (row). Lower case letters refer to the comparison among the three trail gradient categories in each distance class and soil texture class separately (column).

Table 4 Means (±std) of soil loss (g m−2 ) on different soil textures, trail gradients, and distances between water bars. Gradient (%)

Soil texture Clay loam

Silt loam

Distance between water bars (m)

0–10 10–20 >20

Distance between water bars (m)

25

50

75

100

25

50

75

100

0.9aA ± 0.1 1.0aA ± 0.2 3.9bA ± 0.4

1.0aA ± 0.1 1.3aA ± 0.1 4.2bA ± 0.3

1.4aA ± 0.2 1.4aA ± 0.2 8.3bB ± 1.0

4.6aB ± 0.5 4.9aB ± 0.7 13.5bC ± 1.4

1.3aA ± 0.1 1.7aA ± 0.2 5.7bA ± 0.7

1.8aA ± 0.2 2.1aA ± 0.2 10.7bB ± 1.3

7.4aB ± 1.2 8.0aB ± 0.9 19.4bC ± 2.2

13.2aC ± 1.6 12.9aC ± 1.2 25.9bD ± 2.2

Note: Different letters within each treatment show significant differences (P < 0.05). Capital case letters refer to the comparisons among the four distance classes at different trail gradients for each soil texture (row). Lower case letters refer to the comparison among the three trail gradient categories in each distance class and soil texture class separately (column).

A. Masumian et al. / Ecological Engineering 105 (2017) 370–378

375

Fig. 3. Predicted runoff volume (y, mm m−2 ) as a function of different distances between water bars (D, m) and slope gradient (S, %) and their interactions on clay loam (A) and silt loam (B) based on repeated measures analysis. Regression equations for clay loam: y = 0.239 − 0.017 × S + 0.0006 × S2 − 0.0038 × D + 0.00003 × D2 + 0.0001 × S × D and for silt loam: y = 0.109 − 0.012 × S + 0.00066 × S2 − 0.00066 × D + 0.00003 × D2 .

Fig. 4. Predicted soil loss (y, g m−2 ) as a function of different distances between water bars (D, m) and slope gradient (S, %) and their interactions on clay loam (A) and silt loam (B) based on repeated measures analysis. Regression equations for clay loam: y = 8.035 − 0.771 × S + 0.026 × S2 − 0.184 × D + 0.002 × D2 + 0.005 × S × D and for silt loam: y = 2.090 − 1.288 × S + 0.047 × S2 + 0.120 × D + 0.006 × S × D.

the clay loam compared to the silt loam after distances between water bars exceeded 50 m. The functional relationships between runoff volume (Fig. 3) and soil loss (Fig. 4) with distance between water bars and slope gradient reveal different dynamics between runoff and soil loss. Runoff volumes clearly increase with increasing slope gradients on both soil textures, with minimal runoff at distances of 50 m and 75 m and greater values at 25 m and 100 m. Soil loss, particularly on silty clay, increases very strongly linearly with distance between water bars and with increasing slope gradient. Finally, the relationship between soil loss and runoff volume in both soil textures was strongly positive (P ≤ 0.001); the slopes of all curves increased with increasing distance between water bars (P ≤ 0.001; Fig. 5). Regardless of distance between water bars, the slopes were significantly steeper on silt loam than clay loam textured soils (significant three-way interactions of runoff × texture × distance; P ≤ 0.001). At all distance classes, runoff was continuously larger on silt loam than the clay loam soil.

with Solgi et al. (2014). Nonetheless, differences in the response dynamics in the amounts of runoff and soil loss between clay loam and silt loam soils support our central premise that effective distances between water bars that minimize soil loss would differ by soil texture. The influence of soil texture as an important factor that can affect runoff volumes and amounts of soil loss (i.e., soil erodibility) in this study is consistent with previous findings (Ekwue and Harrilal, 2010). While fine-textured soils are more easily transported by water than coarse-textured soils, soil cohesion determines how easily the soil is detached from a surface. Cohesive soils with fine textures, such as clays, are often very stable and do not erode easily; rather, it is the fine-textured soils with low cohesion (i.e., silt loams) that are the most erodible (Hussein et al., 2007). On the other hand, coarse-textured soils are only erodible when water velocities are high (Erpul and Canga, 1999). While it is well known that uncompacted silt soils are often the most erodible soils because the texture is fine and cohesion is relatively low (Greacen and Sands, 1980), it appears that this result can be extended to compacted silt loams as well. In this study, soil loss increased with increasing sand and decreasing clay contents. In silt loam soil, the lower percentage of clay content (18.1%, Table 1) likely decreased soil strength and the cohesiveness of soil particles, while the larger particle sizes led to lower soil cohesiveness and more loosely detached sand particles that are likely responsible for the greater soil loss compared to the clay loam soil (Compton, 2003). Our surface runoff and soil loss results confirm findings from previous studies that documented a considerable impact of increas-

4. Discussion Rainfall in mountainous landscapes is a primary factor of sediment detachment and movement, particularly on steep slopes (Akbarimehr and Naghdi, 2012b), made possible by surface runoff that provides the kinetic energy for overland flow to dislodge soil particles and cause soil loss (Abu Hammad et al., 2006). Consequently, the significant positive correlation between runoff and soil loss on skid trails observed in this study was expected and is line

376

A. Masumian et al. / Ecological Engineering 105 (2017) 370–378 30

Clay Loam 25 m Clay Loam 50 m Clay Loam 75 m Clay Loam 100 m Silt Loam 25 m Silt Loam 50 m Silt Loam 75 m Silt Loam 100 m

-2 Soil loss (g m )

25 20 15 10 5 0 0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

-2 Runoff (mm m )

Fig. 5. Predicted relationship between soil loss (y, g m−2 ) and runoff (x, mm m−2 ) for clay loam (CL, black) and silt loam (SL, gray) based on linear regression analysis. Regression functions differ by soil texture and distances between water bars. Regression equation for: CL25: y = −0.526 + 5.583 × runoff; CL50: y = −0.266 + 9.796 × runoff; CL75: y = −3.479 + 30.291 × runoff; CL100: y = −3.666 + 40.584 × runoff; SL25: y = −5.62 + 13.871 × runoff; SL50: y = −5.361 + 24.345 × runoff; SL75: y = −8.573 + 48.827 × runoff; SL100: y = −8.761 + 60.517 × runoff; R2 = 0.948.

ing trail gradients on erosion (Koulouri and Giourga, 2007; Solgi et al., 2014). In this study, the increase in slope gradient >20% resulted in higher absolute amounts of runoff and soil loss with increasing distances between water bars. The reason for this drastic increase in runoff and soil loss is likely the much higher velocity of surface water at slope gradients >20%, which enhances the erosive power of the water (Ekwue and Harrilal, 2010). Greater flow velocities on steeper slope gradients measured in a silt loam soil has been shown to enhance the erosion capacity of soils (Zhang et al., 2002). In turn, greater flow velocity increases the detachment and transport of soil particles, as demonstrated in laboratory experiments (Fox and Bryan, 1999) and field investigations (Chaplot and Le Bissonnais, 2000). Greater flow velocity and greater volumes of runoff on the silt loam soil compared to clay soil in this study may also be the reason why soil loss amounts in the silt loam exceeded those in the clay soil following the same rainfall event. Consequently, the significant interaction between trail gradient and soil texture that resulted in greater runoff and soil loss on steeper slopes and on silt loam compared with clay loam soil is in line with Ekwue and Harrilal (2010) and calls for enhanced preventative measures under these conditions. One of these preventative measures is the construction of water bars. Results of this study clearly show that the distance between water bars strongly affected the amount of runoff and soil loss on skid trails. Interestingly, although the shortest distance between water bars (25 m) resulted in the lowest amounts of soil loss on both soil textures and at all slope gradients, runoff values were decoupled from soil loss at this distance and exhibited the greatest values per unit area (m2 ). Nonetheless, this is consistent with studies that document that the runoff per unit area is sensitive to plot scale (i.e., the catchment area as determined by the distance between water bars) and that an increase in slope length (i.e., distance between water bars) results in increases in runoff and soil loss (Chaplot and Le Bissonnais, 2003; Moreno-de las Heras et al., 2010). The substantial increase in runoff and soil loss with increasing distance between water bars observed on gradients >20% compared with gradients <20% on both soil textures may be explained by lower depression storage capacities and higher depression connectivity on the steeper slopes (Chaplot and Le Bissonnais, 2003). This result is further consistent with the finding that erosion (soil loss) is proportional to the product of road length (distance between water bars) and the square of the slope (E ∼ LS2 ; Luce and Black, 1999). It is challenging, however, to compare absolute values of runoff and soil loss to results from other studies due to differences in soils, slope gradients, road lengths and because surface erosion on skid trails depends strongly on rainfall amounts, intensities, and duration (Fu et al., 2010). In a previous study on a clay loam soil

with a slope gradient of 31# and distances between water bars of 25 m, 50 m, and 75 m, runoff values of 1.61 mm m−2 , 0.64 mm m−2 , and 0.55 mm m−2 , respectively, which is nearly twice the amount observed in this study for slopes <20%, translated into soil losses of 3.9 g m−2 , 5.0 g m−2 , and 17.2 g m−2 (Akbarimehr and Naghdi, 2012b). These losses are similar to values observed in this study up to distances of 50 m and nearly twice the losses that were observed in this study at 75 m distances. The significant interaction between soil texture and distance between water bars clearly shows that as the distance between water bars increased, the differences in soil loss between the silt loam and the clay loam soil widened and that for the same spacing between water bars (Fig. 5). Further, runoff amounts were generally greater in silty loam than clay loam soils and that the same amount of runoff leads to greater soil loss in silt loam than clay loam soils (Fig. 5). As a consequence, runoff control practices such as post-harvest rehabilitation of skid trails and forest roads and construction of water diversions are not only essential to reduce the quantity of runoff and to minimize the sediment movement (Croke et al., 2001; Grace and Clinton, 2007), but the frequency with which these measures are employed is of crucial importance if runoff and soil erosion are to be minimized, particularly on steep slopes and on silt loams. Despite lowest absolute values of soil loss at 25 m, distances up to 75 m between water bars may be sufficient (and may be more efficient and less costly) to maintain low levels of runoff and soil loss on trails with gradients <20% and distances up to 50 m on trail gradients >20% on clay loams. On silt loams, recommended distances are shorter: 50 m on trail gradients <20% and 25 m on trail gradients >20%. These distances are comparable to recommendations of 75 m for a trail on a 10% slope and 50 m for a trail on 31% slope on a clay loam soil (Akbarimehr and Naghdi, 2012b). Because minimizing runoff and soil erosion is important not only for soil and water conservation, but also for reducing nutrient discharge with runoff (Bjorneberg et al., 2000), water diversion techniques such as water bars are thus crucial elements for reducing loss of plant nutrients and maintaining the fertility of forest soils (Ekwue and Harrilal, 2010). Because rainfall amount, intensities, and duration were not measured in this study, more work is needed before these recommendations on distances between water bars can be applied in areas with different rainfall patterns.

5. Conclusions This study provides empirical evidence to support the conclusion that soil texture, trail gradient and distances between water bars can strongly affect runoff volumes and soil loss. Although

A. Masumian et al. / Ecological Engineering 105 (2017) 370–378

runoff and soil loss on skid trails cannot be avoided with the use of water bars, we conclude that maximum distances between water bars should depend on the soil texture and slope gradient, with shorter distances between water bars on more erodible soils such as silt loam soils and on steeper slopes. Even short distances of 25 m between water bars are only modestly effective in reducing runoff and soil loss, however, when slope gradients exceed 20%, presumably due to increased runoff volume and water velocity. Given that runoff and soil loss increased with increasing trail gradient and precipitously so at gradients >20%, the installation of physical water diversion structures (water bars) that is a commonly prescribed BMP and is generally considered the minimum level of BMP implementation in the Iran, may have to be supplemented with other skid trail retirement and rehabilitation techniques such as seeding of ground vegetation. We further recommend that BMPs consider soil texture when making recommendations on the spacing of water bars on slopes of soils with different textures.

References Abu Hammad, A.D., Borresen, T., Haugen, L.E., 2006. Effects of rain characteristics and terracing on runoff and erosion under the Mediterranean. Soil. Till. Res. 87, 39–47. Akay, A.E., Erdas, O., Reis, M., Yuksel, A., 2008. Estimating sediment yield from a forest road network by using a sediment prediction program and GIS techniques. Build. Environ. 43, 687–695. Akbarimehr, M., Naghdi, R., 2012a. Assessing the relationship of slope and runoff volume on skid trails (Case study: Nav 3 district). J. For. Sci. 58 (8), 357–362. Akbarimehr, M., Naghdi, R., 2012b. Determination of most appropriate distance between water diversions on skid trails in the mountainous forest, northern Iran. Catena 88, 68–72. Battiato, A., Diserens, E., Laloui, L., Sartori, L., 2013. A mechanistic approach to topsoil damage due to slip of tractor tyres. J. Agric. Sci. Appl. 2 (3), 160–168. Binkley, D., Brown, T.C., 1993. Forest practices as nonpoint sources of pollution in North America. Water Res. Bull. 29 (5), 729–740. Bjorneberg, D.L., Aase, J.K., Westermann, D.T., 2000. Controlling sprinkler irrigation runoff, erosion, and phosphorus loss with straw and polyacrylamide. Am. Soc. Agric. Biol. Eng. 43 (6), 1545–1551. Brais, S., 2001. Persistence of soil compaction and effects on seedling growth in northwestern Quebec. Soil Sci. Soc. Am. J. 65, 1263–1271. Cerdà, A., 2007. Soil water erosion on road embankments in eastern Spain. Sci. Total Environ. 378 (1–2), 151–155, http://dx.doi.org/10.1016/j.scitotenv.2007. 01.041. Chaplot, V., Le Bissonnais, Y., 2000. Field measurements of interrill erosion under different slopes and plot sizes. Earth Surf. Process. Landf. 25, 145–153. Chaplot, V., Le Bissonnais, Y., 2003. Runoff features for interrill erosion at different rainfall intensities, slope lengths, and gradients in an agricultural loessial hillslope. Soil Sci. Soc. Am. J. 67, 844–851. Compton, A., 2003. Runoff, Erosion and Sediment Control Best Management Practices for Yukon Placer Mining Operations. EDI Environmental Dynamics Inc, p. 29. Copstead, R.L., Johansen, D.K., Moll, J., 2003. Water/Road Interaction: Introduction to Surface Drains. Department of Agriculture Forest Service Technology and Development Program. 9877 1806—SDTDC, San Dimas, U.S, https://www.fs. fed.us/eng/pubs/html/wr p/98771806/98771806.htm. Croke, J., Mockler, S., 2001. Gully initiation and road-to-stream linkage in a forested catchment, southeastern Australia. Earth Surf. Process. Landf. 26, 205–217. Croke, J.C., Hairsine, P.B., Fogarty, P., 2001. Soil recovery from track construction and harvesting changes in surface infiltration, erosion, and delivery rates with time. For. Ecol. Manage. 143, 3–12. Ekwue, E.I., Harrilal, A., 2010. Effect of soil type, peat, slope, compaction effort and their interactions on infiltration, runoff and raindrop erosion of some Trinidadian soils. Biosyst. Eng. 105, 112–118. Erpul, G., Canga, M.R., 1999. Effect of subsequent simulated rainfalls on runoff and erosion. Turk. J. Agric. For. 23, 659–665. FAO/UNESCO, 1990. Soil map of the world, revised legend, World Soil Resources Report 60. FAO, Rome. Fox, D.M., Bryan, R.B., 1999. The relationship of soil loss by interrill erosion to slope gradient. Catena 38, 211–222. Froehlich, D.C., 1995. Peak outflow from breached embankment dam. J. Water Res. Plan. Manage. 121 (1), 90–97. Fu, B., Newham, L.T.H., Ramos-Scharron, C.E., 2010. A review of surface erosion and sediment delivery models for unsealed roads. Environ. Model. Softw. 25, 1–14. Grace, J.M., 2002. Control of sediment export from the forest road prism. Trans. ASAE 45 (4), 1127–1132. Grace, J.M., 2005. Forest operations and water quality in the South. Trans. ASAE 48, 871–880. Grace, J.M., Clinton, B.D., 2007. Protecting soil and water in forest road management. Trans. ASABE 50 (5), 1579–1584.

377

Grace, J.M., Rummer, B., Stokes, B.J., Wilhoit, J., 1998. Evaluation of erosion control techniques on forest roads. Trans. ASAE 41 (2), 383–391. Greacen, E.L., Sands, R., 1980. Compaction of forest soils – a review. Aust. J. Soil Res. 18, 163–189. Gucinski, H., Furniss, M., Ziemer, R., Brookes, M., 2001. Forest roads: a synthesis of scientific information. General Technical Report PNW-GTR-509. U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station, Portland, Oreg. Hartano, H., Prubhu, R., Widayat, A.S.E., Asdak, Ch., 2003. Factors affecting runoff and soil erosion: plot level soil loss monitoring for assessing sustainability of forest management. For. Ecol. Manage. 180, 361–374. Hussein, M.H., Kariem, T.H., Othman, A.K., 2007. Predicting soil erodibility in northern Iraq using natural runoff plot data. Soil Till. Res. 94, 220–228. Jackson, C.R., Martin, J.K., Leigh, D.S., West, L.T., 2005. A southeastern piedmont watershed budget: evidence for a multi-millennial agricultural legacy. J. Soil Water Conserv. 60, 298–310. Jordan-Lopez, A., Martinez-Zavala, L., Bellinfate, N., 2009. Impact of different parts of unpaved forest roads on runoff and sediment yield in a Mediterranean area. Sci. Total Environ. 407, 937–944. Jusoff, K., Majid, N.M., 1996. The impacts of skid trails on the physical properties of hill forest soils. Pertanika 9, 311–321. Kalra, Y.P., Maynard, D.G., 1991. Methods and Manual for Forest Soil and Plant Analysis. Re NOR-X-319. Forestry Canada, Northern Forestry Center, Alberta, Canada. Kateb, E.H., Zhang, H., Zhang, P., Mosandl, R., 2013. Soil erosion and surface runoff on different vegetation covers and slope gradients: a field experiment in Southern Shaanxi Province, China. Catena 105, 1–10. Koulouri, M., Giourga, C., 2007. Land abandonment and slope gradient as key factors of soil erosion in Mediterranean terraced lands. Catena 69, 274–281. LaMarche, J., Lettenmaier, D.P., 2001. Effects of forest roads on flood flows in the Deschutes River Basin, Washington. Earth Surf. Process. Landf. 26, 115–134. Lee, J.W., Park, C.M., Rhee, H., 2013. Revegetation of decomposed granite roadcuts in Korea: developing digger, evaluating cost effectiveness, and determining dimensions of drilling holes, revegetation species, and mulching treatment. Land Degrad. Dev. 24 (6), 591–604. Litschert, S.E., MacDonald, L.H., 2009. Frequency and characteristics of sediment delivery pathways from forest harvest units to streams. For. Ecol. Manage. 259, 143–150. Luce, C.H., Black, T.A., 1999. Sediment production from forest roads in western Oregon. Water Resour. Res. 35, 2561–2570. Martínez-Zavala, L., Jordán López, A., Bellinfante, N., 2008. Seasonal variability of runoff and soil loss on forest road backslopes under simulated rainfall. Catena 74, 73–79, http://dx.doi.org/10.1016/j.catena.2008.03.006. Miller, R., 2006. The Water Bar – Water Quality Update for Loggers and Other Forestry Professionals. Serving Northwest North Carolina, p. 2. Mohamadi, M.A., Kavian, A., 2015. Effects of rainfall patterns on runoff and soil erosion in field plots. Int. Soil Water Conserv. Res. 3, 273–281. Moreno-de las Heras, M., Nicolau, J.M., Merino-Martín, L., Wilcox, B.P., 2010. Plot-scale effects on runoff and erosion along a slope degradation gradient. Water Resour. Res. 46, 1–12. Morgan, R.P.C., 1986. Soil Erosion and Conservation. Longman, New York. Najafi, A., Solgi, A., Sadeghi, S.H., 2009. Soil disturbance following four wheel rubber skidder logging on the steep trail in the north mountainous forest of Iran. Soil Till. Res. 103, 165–169. Naghdi, R., Solgi, A., Zenner, E.K., 2015. Soil disturbance caused by different skidding methods in mountainous forests of Northern Iran. Int. J. For. Eng. 26, 212–224. Naghdi, R., Solgi, A., Eric, R., Labelle, Zenner, E.K., 2016. Responses of soil physical and chemical properties and maple (Acer velutinum Boiss) seedling growth to soil compaction following ground-based skidding. Eur. J. For. Res., http://dx. doi.org/10.1007/s10342-016-0986-3. Natural Resources Conservation Service (NRCS), 2014. Forest trails and landings, p. 6. Patric, J.H., 1976. Soil erosion in the eastern forest. J. Forestry 74 (10), 671–677. Ramos-Scharrón, C.E., MacDonald, L.H., 2005. Measurement and prediction of sediment production from unpaved roads, St John, US Virgin Islands. Earth Surf. Process. Landf. 30 (10), 1283–1304, http://dx.doi.org/10.1002/esp.1201. Richard, G., Cousin, I., Sillon, J.F., Bruand, A., Guérif, J., 2001. Effect of compaction on soil porosity: consequences on hydraulic properties. Eur. J. Soil Sci. 52, 49–58. Rohand, K., Al Kalb, A., Herbauts, J., Verbrugge, J.C., 2004. Changes in some mechanical properties of loamy soil under the influence of mechanized forest exploitation in a beech forest of central Belgium. J. Terramech. 40, 235–253. Sadeghi, S.H.R., Ghaderi Vangah, B., Safaeeian, N.A., 2007. Comparison between effects of open grazing and manual harvesting of cultivated summer rangelands of northern Iran on in filtration, runoff and sediment yield. Land Degrad. Dev. 18, 608–620. Solgi, A., Najafi, A., Sadeghi, S.H., 2014. Effects of traffic frequency and skid trail slope on surface runoff and sediment yield. Int. J. For. Eng. 25, 171–178. Stringer, J.W., Lowe, L., Smidt, M., 1998. Field guide to best management practices for timber harvesting in Kentucky, FOR-69, p. 73. Swift Jr., L.W., 1985. Forest road design to minimize erosion in the southern Appalachians. In: Blackman, B.G. (Ed.), Forest and Water Quality: A Mid-South Symposium. University of Arkansas, Little Rock, Ark, pp. 141–151. USDA, 1998. Soil Taxonomy: Keys to Soil Taxonomy, 8th ed. United States Department of Agriculture, Natural Resources Conservation Service, Washington, DC.

378

A. Masumian et al. / Ecological Engineering 105 (2017) 370–378

Van Lear, D.H., Taylor, G.B., Hansen, W.F., 1995. Sedimentation in the Chattooga River watershed. Tech. Paper No. 19. Clemson University, Department of Forest Resources, Clemson, SC. Wade, C.R., Bolding, M.C., Aust, W.M., Lakel, W.A., 2012. Comparison of five erosion control techniques for bladed skid trails in Virginia. South. J. Appl. For. 36, 191–197. Wallbrink, P.J., Croke, G., 2002. A combined rainfall simulator and tracer approach to assess the role of best management practices in minimising sediment redistribution and loss in forests after harvesting. For. Ecol. Manage. 170, 217–232. Wear, L.R., Aust, W.M., Bolding, M.C., Strahm, B.D., Dolloff, C.A., 2013. Effectiveness of best management practices for sediment reduction at operational forest stream crossings. For. Ecol. Manage. 289, 551–561.

Yoho, N.S., 1980. Forest management and sediment production in the South: A review. South. J. Appl. For. 4 (1), 27–36. Zenner, E.K., Fauskee, J.T., Berger, A.L., Puettmann, K.J., 2007. Impacts of skidder traffic intensity on soil disturbance, soil recovery, and aspen regeneration in north central Minnesota. North. J. Appl. For. 24, 177–183. Zhang, G.H., Liu, B.Y., Nearing, M.A., Huang, C.H., Zhang, K.L., 2002. Soil detachment by shallow flow. Trans. ASAE 45, 351–357. Ziegler, A.D., Giambelluca, T.W., 1997. Importance of rural roads as source areas for runoff in mountainous areas of northern Thailand. J. Hydrol. 196, 204–229.