The effects of soil properties and vegetation cover on the sedimentation of forest roads

This study was conducted to investigate the interaction of soil properties and vegetation cover on soil loss from forest road prism. Rainfall simulations were carried out on road surface, fillslope and cutslope. Runoff and sediment samples were collected every 4 min and then total soil loss was measured. Samples from top soil were randomly collected for analysis and grouping soil into A, B and C. Results showed that on cutslope, the highest soil loss was detected for soil group B, where sediment concentration in runoff was 21.83 g L -1 and vegetation cover was 0 to 30%. On fillslope, the highest soil loss was detected for soil group C, where sediment concentration in runoff was 18.07 g L -1 and vegetation cover was 10 to 40%. On road surface, the highest soil loss was detected for soil group A, where sediment in runoff was 8.99 g L -1 and vegetation cover was 2 to 5%.


INTRODUCTION
Globally, soil erosion is one of the most important environmental problems which threaten soil and water resources (Cerdà, 2007).An Iran, due to the arid climate and the human land use is affected by soil erosion and land degradation.Soil properties is one of the main parameters that affect runoff and soil loss processes.Water erosion of forest soil is naturally very low and can be neglected, but after clear cutting for road construction, the bare soil is exposed to rainfall and other erosive agents (Croke et al., 2001;Foltz et al., 2009).Road erosion is also found during the construction of the roads and the embankments and soil development only took place at a low rate.So, under this situation the inherent sensitivity of forest soil to erosion is appeared especially on steep slopes (Jordán and Martínez-Zavala, 2008).If a forest road is planned and constructed without considering region erodibility, the engineers would be faced to high cost to solve road sedimentation problem.Road is the main type of transportation system in Hyrcanian forests of Iran (Parsakhoo et al., 2009).Most of these roads traversed through hilly and mountainous areas.These mountainous roads experience numerous hazards such as water erosion and landslides which cause disruption, injuries and losses to life and economy  (Mohammadkhan et al., 2011).Kavian et al. (2010) reported that soil initial moisture, percentage of soil organic matter, bulk density and sand percent are most effective factors in runoff generation.Also, the results illustrated that percentage of soil organic matter, soil initial moisture and silt percent affect on soil erosion.Najafian et al. (2010) showed that both life form and vegetation cover significantly influenced the amount of runoff and sedimentation.Greatest runoff was found for forbs while grasses had lowest amount of runoff.Also bare ground and 100% cover showed the greatest and lowest amount of runoff, respectively.The concentration of sediment for shrubs was significantly greater than grasses and in 100% cover it was significantly lower than 50% and bare ground.Scientific understanding of factors influencing soil loss from forest road can effectively be used by managers in the design and sustainable management of road networks (Pappas et al., 2008;Olive and Marion 2009).Road prisms, including cutslopes, road surfaces, and fillslopes, can be important contributors of sediment to streams in forested watersheds (Bold et al., 2009;Morgan, 2005).
Interaction between soil properties and vegetation cover is one of the main factors responsible for soil loss (Fu et al., 2009).Monitoring of soil loss using rainfall simulator and runoff plots was cost-effective and provided valuable information about soil erosion caused by road construction.Thus, this study was conducted to (i) grouping soil properties for different parts of forest roads and (ii) investigate the effects of the interaction of some soil properties and vegetation cover on runoff and soil loss from road prism and forest lands located in Hyrcanian zone of Iran.

Description of the study area
Lat Talar forest within the watershed number of 71 in Hyrcanian zone of Iran was selected as the study area.The research site has a total area of 3801 ha.The mean annual temperature of the study area is 15°C.Minimum altitude is about 300 m and the maximum 1650 m.The mean slope was 50%.The location of the research is in latitude 53° 9′ 40″ to 53° 13 ′55″ E and longitude 36° 12′ 55″ to 36° 15 ′ 45″ N. The region has very moist to mid moist and cold climate with a mean annual precipitation of 800 mm.The bedrock is typically marl, marl lime and limestone with a soil texture of loam and clay loam.Forest stands are dominated by Fagus orientalis Lipsky, Carpinus betulus and herbaceous species of Carex silvatica, Buxus hyrcanus, Berachypodium silvaticum, Ruscus hyrcanus, Phyllitis scolopendrium, Rubus hyrcanus L. and Polypodium auidinum.Forest roads in the studied area are used by truck, motorcycle, Landrover and nissan.The traffic density is 5 vehicles per day.Philonotis marchica (Hedw.)Brid.and Rubus hyrcanus L. are dominant species on cutslopes and fillslopes of road (Figure 1).

Data collection
Totally, 38 rainfall simulations were randomly done all over the area Caspian Sea for road surface (10 replications), fillslope (13 replications) and cutslope (15 replications).A portable single nozzle rainfall simulator simulated rain with drop size of 3 mm for duration of 20 min at intensity of 32.4 mm h -1 (Figure 2).Water with a mean temperature of 23°C was rained using a Schlick r86510 nozzle mounted 2 m above the ground onto a squared area of 0.48 m 2 that is bordered by a steel structure on cut slope and road surface.Runoff and sediment samples were collected every 4 min by gauge and then total soil loss was measured.The clear water was separated from deposited sediment.Sediment was air-dried for a week in temperature of 30°C.Besides, it was oven-dried at 105°C for at least 2 h and then weighted using digital balance.25 samples of the top soil (0 to 20 cm deep) were randomly collected from cutslope, fillslope and road surface.Soil texture was determined by the Bouyoucos hydrometer method.Lime percentage (T.N.V or CaCO3) was measured using the NaOH titration method.Soil organic carbon was determined using the Walkley-Black technique.Soil bulk density was measured using a relation of soil dry weight to volume of sampling cylinder (484 cm 3 ) (Koc et al., 2008).Runoff coefficient is calculated as Equation 1 (Figure 2).

 
Where, RC is runoff coefficient in %, Rh is runoff height in mm and Ph is rainfall height in mm.

Statistical analysis
In grouping process of plots, each table of soil data sets was analyzed using principal component analysis (PCA) in PC-ORD software.Then, data were statistically analyzed using GLM procedure in SAS program.Student Newman Keuls (SNK) multiple comparison test at probability level of 5% was used to compare means among groups and diagram designed by Excel software.

Effects of vegetation cover and soil groups on soil loss from cutslope
Figure 3 shows the PCA correlation circle for the soil variables on cutslope.Soil properties on cutslope were classified into three groups based on PCA.The purpose of this analysis is to define groups of items based on their similarities.The properties of each group have been recorded in Table 1.
The runoff and soil loss in soil groups B and C was significantly more than that of soil group A (Table 2).A soil with a high percentage of silt and clay particles has a greater erodibility than a sandy soil under the same  conditions.Although all soils are potentially susceptible to water erosion, silts, silt loams and loams are most at risk because there is no adhesion forces among silt particles, (Ziegler et al., 2001).The low-vegetated road cutslope increase their runoff coefficient.The soil loss was lower than 90 g m −2 h −1 on  the high vegetated road cutslope and greater than 200 g m −2 h −1 on the low-vegetated ones in study area (Table 3).Vegetation provides a protective layer or buffer between the atmosphere and the soil by means of its canopy, roots, and litter components (Mohammad and Adam, 2010).Many studies have emphasized the importance of vegetation cover on soil loss.On cutslope, the highest soil loss was detected for soil group B, where sediment concentration in runoff was 21.83 g L -1 and vegetation cover was 0 to 30% (Table 4).There was no slope class of 30 to 45% in soil group B.

Effects of vegetation cover and soil groups on soil loss from fillslope
Figure 4 and Table 5 show the PCA correlation circle for the soil variables and the properties of each group on fillslope, respectively.Minimum runoff coefficient and sediment concentration as well as soil loss was observed in soil group B (P<0.05), this might be due to the high content of CaCO 3 and sand (Table 6).Soil texture determines the rate at which water drains through a saturated soil; water moves freely through sandy soils than it does through clayey soils (Clinton and Vose, 2003).In humid Mediterranean mountainous area it was detected that the sand content and clay content of the soil have significant correlations with runoff rates on the three parts of the roads, affecting in a negative or positive manner respectively.The organic matter content contributed to reduce the runoff rate on the road surface and on the fillslope.Soil loss was negatively correlated to the sand content and organic matter content, but positively correlated to clay content (Jordán-López et al., 2009).Shixiong et al. (2006) indicated that vegetation cover could significantly reduce sediment yield from unpaved roads.The above-ground components of the vegetation, such as leaves and stems, partially absorb the energy of the erosive agents of water and wind, so that less is directed at the soil, whilst the below-ground components, comprising the rooting system, contribute to the mechanical strength of the soil (Morgan and Rickson, 1995).In current study runoff and soil loss from fillslope plots with higher percentage of vegetation cover was lower than that of plots with lower vegetation cover, but this difference was not significant which could be due to the effects of soil properties, different plant forms or different amount of plant residues (Table 7).On fillslope, the highest soil loss was detected for soil group C, where sediment concentration in runoff was 18.07 g L -1 and vegetation cover was 10 to 40% (Table 8).

Effects of vegetation cover and soil groups on soil loss from road surface
Figure 5 and Table 9 show the PCA correlation circle for the soil variables and the properties of each group on road surface, respectively.Minimum runoff coefficient and sediment concentration as well as soil loss was observed in soil group C (P<0.05), this might be due to high content of CaCO 3 , organic matter and sand (Table 10).Calcium carbonate of soil is an effective factor on conjunction of clay particles (Feiznia et al., 2005).There was no significant difference between soil loss from plots with vegetation cover of 2-5 and 5-8% (Table 11).

Soil groups
Bulk density (g cm    On road surface, the highest soil loss was detected for soil group A, where sediment concentration in runoff was 8.99 g L -1 and vegetation cover was 2 to 5% (Table 12).

Conclusions
Soil properties and vegetation cover are two effective factors on sediment yield from different parts of forest road.In this study it was concluded that the runoff and soil loss was more in soils which had high content of silt and clay.The maximum sedimentation and soil loss from cutslope was observed in soil group B due to high moisture and low bulk density.In fillslope, minimum sedimentation and soil loss was detected in soil group B, because high content of sand and CaCO 3 in this group.This finding was also observed for road surface and forest ground in soil group C and A, respectively.It is possible to repair cut and fillslopes with revegetation methods.Moreover, soil properties can be improved

Figure 1 .
Figure 1.Geographical position of the study area.

Figure 2 .
Figure 2. Rainfall simulator and runoff collection by the gauge.

Table 1 .
Characteristics of the soil groups in cutslope.

Table 2 .
Effect of soil groups on runoff and sedimentation from cutslope.

Table 3 .
Effect of vegetation cover on runoff and sedimentation from cutslope.

Table 4 .
Interaction of soil and vegetation cover on runoff and sedimentation from cutslope.

Table 5 .
Characteristics of the soil groups in fillslope.

Table 6 .
Effect of soil groups on runoff and sedimentation from fillslope.

Table 7 .
Effect of vegetation cover on runoff and sedimentation from fillslope.

Table 8 .
Interaction of soil and vegetation cover on runoff and sedimentation from fillslope.
Figure 5. PCA analysis of soil on road surface.

Table 9 .
Characteristics of the soil groups in road surface.

Table 10 .
Effect of soil groups on runoff and sedimentation from road surface.

Table 11 .
Effect of vegetation cover on runoff and sedimentation of road surface.In a row, means with the same letter are not significantly different based on Student-Newman-Keuls test, Alpha=0.05.

Table 12 .
Interaction of soil and vegetation cover on runoff and sedimentation from road.
through mulching, hydro-seeding and etc.