The Influence of Slope Positions on the Recovery Response of Compacted Soil Properties and Enzyme Activity in an Oriental Beech Stand in the Hyrcanian Forests, Iran

Several studies emphasize the effects of slope position on divergences of soil properties in forest ecosystems, but limited data is available on the impact of slope position on recovery levels of soil, which were exposed to compaction due to machine traffic. This study examined the effects of slope position (i.e., S; summit, BS; backslope, and TS; toeslope) on recovery rate of soil properties and enzyme activity four years after ground-based harvesting operations were performed on machine operating trails, compared to the undisturbed areas (UND) in the Hyrcanian forests (north of Iran). Soil properties and enzyme activity of compacted soil in machine operating trails showed significant trend differences among the slope positions. A significantly lower soil bulk density, penetration resistance, soil moisture, aggregate stability, pH, sand, and C/N ratio were found in TS compared to the values recorded in the BS and S treatments. Conversely, total porosity, macroporosity, silt, clay, organic C, total N, available nutrients (i.e., P, K+, Ca2+, and Mg2+), fulvic and humic acid, earthworm density and dry mass as well as fine root biomass were higher in TS than in the BS and S treatments. Soil microbial respiration, MBC, NH4+, NO3−, N mineralization, and MBN were significantly higher in the UND areas followed by TS > BS > S treatment. The highest activity levels of enzymes (i.e., urease, acid phosphatase, arylsulfatase, invertase, and ß-N-acetylglucosaminidase) were detected in the UND areas, followed by TS > BS > S treatment. The highest recovery levels of all soil properties and enzyme activity were found in TS, followed by BS > S treatment. However, the full recovery of soil properties did not occur even after a 4-year period, compared to the UND areas. Our study results highlight the significance of the slope position in augmenting divergence in soil properties and enzyme activity after ground-based machine traffic.


Introduction
Catena slope positions in the forest ecosystems play a crucial role in creating micro-environments and soil heterogeneity [1], which result in particular ecological and biological processes of soil, and lead to miscellaneous nutrient cycling, composition/structure of the microbial population, as well as enzyme activity [2,3]. The hydrological processes and soil formation stages are regulated by The Hyrcanian deciduous temperate forests are known as unique ecosystems in the northern hemisphere because they encompass a diverse range of macro-and micro-topographies. Several studies highlight the effects of topography and slope position on divergences of soil properties in forest ecosystems, but no data is currently available on the impact of slope position on recovery levels of soils, which were exposed to machine-induced compaction through harvesting operations performed in forest stands. This study aimed to elucidate the effects of slope positions on recovery levels of compacted soil properties and enzyme activity within four years after machine operations performed on machine operating trails, compared to undisturbed or control (UND) areas. We tested the hypothesis that slope position may affect the recovery process of soil physical, chemical, and biological properties, as well as enzyme activity.

Site Description
The study site is an old-growth oriental beech stand in the Kheyrud forest of the Hyrcanian forests (40 • 46 N, 55 • 49 E). The investigated site has a slope ranging from 1-40% facing north and an elevation in the range of 1170-1220 m above sea level. The mean annual rainfall is 1420 mm with the highest precipitation occurring in October and the lowest in July [14]. The climate of the investigated area is humid cold with a one-month dry period in July. Mean annual temperature is 7.9 • C with the hottest and coldest months in July and January, respectively. The soils are Alfisols with clay loam texture (according to the USDA Soil Taxonomy) from limestone; belonging to the upper Jurassic and lower Cretaceous periods. Forest stands are composed of beech (Fagus orientalis Lipsky) and are accompanied with other species including hornbeam (Carpinus betulus L.), velvet maple (Acer velutinum), and Caucasian alder (Alnus subcordata C.A.M.). The main herbaceous species of the ground vegetation are Cyclamen coum Mill., Prunella vulgaris Huds. Rhynchocorys maxima Rieht., Galium odoratum L., Mercurialis perennis L., Oplismenus undulatifolius Ard., Euphorbia amygdaloides L., and Viola sieheana Becker. Semi-mechanized forest operations were performed by chainsaw in March 2014 where trees were felled, delimbed and bucked to size. The processed logs were then extracted by a Timberjack 450C wheeled skidder in August 2014. The four-wheel drive skidder had an empty weight of 10.3 metric tons (tire inflation pressure set to 220 kPa) and the average load volume was 2.9 cubic meters. All traffic was performed on machine operating trails of 3.5 m in width.

Experimental Design
To study the effects of catena position on recovery values of compacted soil, four machine operating trails were selected four years after skidding operations were completed (2014) (Figure 1a-c). More specifically, three catena positions were selected in each machine operating trail including the summit (S), backslope (BS), and toeslope (TS) as well as the undisturbed or control areas (UND) as treatments. Sampling plots were established in different trail segments exposed to a high level of machine traffic (> 15 machine cycles; a machine cycle consisted of one unloaded and one loaded pass with the skidder). In each catena position, three plots were randomly established and a plot (with length of 20 m and width of 4 m) was randomly selected for soil sampling in August 2018. In each selected sample plot, five transects were set up perpendicular to the longitudinal axis of the trail with a spacing of 4 m between transects (Figure 1b,c). Three of the five transects were randomly selected for soil sampling [17]. In each sample plot (Figure 1c), nine soil samples (six samples in tracked locations and three samples from between tracks) were collected from the machine operating trails and three soil samples were collected in the undisturbed or control areas. The undisturbed or control areas (UND area) were established at a distance of 20 m from the machine operating trails in each plot (Figure 1b,c). In total, 144 soil samples (i.e., 4 machine operating trails × 3 slope positions × 12 samples in each plot (nine samples in the compacted area + three samples in the control area)) were collected and analyzed in August 2018.

Data Collection and Laboratory Analysis
A steel cylinder (length of 40 mm and diameter of 56 mm) was used to collect soil samples from the surface soil of 0-10 cm. Following the extraction, soil samples were placed in plastic bags, sealed, labeled, and transported to the lab for further analysis. Soil samples were weighted after collection and then oven-dried at 105 °C until reaching a constant mass to determine the water content and the soil bulk density (ds). The hydrometer method was used to determine the soil particle size distribution for particles smaller than 0.075 mm [30], and the larger particles by sieving through a series of sieves of varying apertures. The water desorption method was applied to determine the macroporosity [31]. The soil penetration resistance (PR) was measured using an analog hand-held soil penetrometer (Eijkelkamp 06.01.SA penetrometer with a 60° cone and a 1 m maximum measuring depth). The wet sieving procedure was used to determine the aggregate stability [32]. To assess the soil particle density (dp), the ASTM D854-00 2000 standard was applied and the formula as (Total porosity (TP) = [1 − (soil bulk density (ds)/ soil particle density (dp))] × 100) was used to determine total porosity (TP).
To analyze the biological properties, approx. 2 kg of soil was also collected from each sampling point, transported to the lab, and stored in plastic bags at 4 °C. Soil pH with a ratio of soil to water of 1:2.5 was measured using the Orion Ionalyzer (Model 901) pH meter. EC was measured by an Orion Ionalyzer EC meter with a ratio of soil to water of 1:2.5 solution. Soil organic C was determined by applying the Walkley-Black technique [33] and total N by using the Kjeldahl method [34]. The

Data Collection and Laboratory Analysis
A steel cylinder (length of 40 mm and diameter of 56 mm) was used to collect soil samples from the surface soil of 0-10 cm. Following the extraction, soil samples were placed in plastic bags, sealed, labeled, and transported to the lab for further analysis. Soil samples were weighted after collection and then oven-dried at 105 • C until reaching a constant mass to determine the water content and the soil bulk density (ds). The hydrometer method was used to determine the soil particle size distribution for particles smaller than 0.075 mm [30], and the larger particles by sieving through a series of sieves of varying apertures. The water desorption method was applied to determine the macroporosity [31]. The soil penetration resistance (PR) was measured using an analog hand-held soil penetrometer (Eijkelkamp 06.01.SA penetrometer with a 60 • cone and a 1 m maximum measuring depth). The wet sieving procedure was used to determine the aggregate stability [32]. To assess the soil particle density (dp), the ASTM D854-00 2000 standard was applied and the formula as (Total porosity (TP) = [1 − (soil bulk density (ds)/ soil particle density (dp))] × 100) was used to determine total porosity (TP).
To analyze the biological properties, approx. 2 kg of soil was also collected from each sampling point, transported to the lab, and stored in plastic bags at 4 • C. Soil pH with a ratio of soil to water of 1:2.5 was measured using the Orion Ionalyzer (Model 901) pH meter. EC was measured by an Orion Ionalyzer EC meter with a ratio of soil to water of 1:2.5 solution. Soil organic C was determined by applying the Walkley-Black technique [33] and total N by using the Kjeldahl method [34]. The available phosphorous (P) was determined using the Olsen method with a spectrophotometer, and available potassium (K), calcium (Ca), and magnesium (Mg) (by ammonium acetate extraction at pH 9) by applying an atomic absorption spectrophotometer [35]. The earthworm sampling and counting was manually done at the surface soil with area of 25 × 25 cm and 0-10 cm depth after removing the litter layer. After collection, the earthworms were washed, weighed, and oven-dried at 60 • C for 24 h to determine the earthworm dry mass [17]. To measure fine root biomass, fine roots (<2 mm in diameter) were extracted from each sample and dried at 70 • C to a constant mass [35]. By measuring the CO 2 evolved in a 3-day incubation experiment at 25 • C, soil microbial respiration was measured [35]. The chloroform fumigation-extraction method was also used to determine the microbial biomass carbon (MBC) and nitrogen (MBN) in the soils [34]. Soil NH 4 + and NO 3 − were extracted with 2 M KCl solution (with a ratio of soil to solution of 1:5) and determined using the colorimetric techniques [25]. N mineralization was determined by aerobic incubation of the soils [36]. Urease activity (EC 3.5.1.5) was analyzed using 200 µmol urea as substrate, incubated for 2 h at 37 • C. The acid phosphatase activity was determined in a MUB buffer (pH 6.5), incubated for 1 h at 37 • C [25]. A p-Nitrophenyl sulphate was used for incubation for 1 h at 37 • C to analyze the arylsulphatase activity (EC 3.

Statistical Analyses
A factorial experiment with a complete block design was randomly assigned to the treatments (TS, BS, S, and UND). Generalized linear modeling (GLM) was used to relate the recovery of soil properties and enzyme activity to treatment. To compare soil properties among different slope positions, one-way analysis of variance (ANOVA) was used. The normality and homogeneity of variance were verified with the Kolmogorov-Smirnov and the Levene tests (α = 0.05). The post hoc test was used to verify the statistically significant differences between the treatments by the Tukey test at P ≤ 0.05. The relationships between soil physical, chemical, and biochemical as well as biological properties with soil enzyme activity in four treatments were determined using the Pearson correlation. All statistical tests were performed using the SPSS software package (release 17.0; SPSS, Chicago, IL, USA).

Soil Physical, Chemical, and Biological Properties
Results showed that all the soil physical, chemical, and biological properties tested (with the exception of EC) were influenced by the treatments (p < 0.001). Because the soil properties originating from the different UND areas were similar without any statistical differences, these data were grouped and reported as mean values. Average soil bulk density and penetration resistance were lower in the UND areas and TS position as compared to the other positions (Table 1). More specifically, average soil bulk density was 33% higher in the S position as compared to the UND areas whereas penetration resistance was 80% higher in the S position as compared to the UND areas. Total porosity and macroporosity, soil moisture, aggregate stability, and silt were significantly higher in the UND areas and TS position than the amounts recorded at the BS and S positions. Additionally, soil moisture and aggregate stability were 50% and 40% lower in the S position as compared to the UND areas, respectively. The highest sand content was measured in the S and BS positions. The highest recovery values of all soil physical properties tested were detected in the TS position, followed by values at the BS and S positions, compared to the UND areas.
Concerning chemical properties, there were no significant differences in EC among slope positions and UND areas. Soil pH was significantly higher in S positions than in BS, TS and the UND areas. Also, soil C/N ratios were significantly lower in S positions than in BS, TS and the UND areas. Soil organic C, N, available nutrients (i.e., P, K, Ca, and Mg), fulvic acid, and humic acid were also highest in the UND areas followed by TS and lowest in the BS and S positions. Likewise, average soil pH and C/N ratio were 24% and 44% higher in the S position as compared to the UND areas, whereas SOC and N were 3.7 and 5.1 times lower in the S position as compared to the UND areas, respectively.
Significantly higher values of earthworm density and dry mass were measured in the UND areas followed by TS > BS > S treatment. Fine root biomass showed no significant difference between the UND areas and TS. Furthermore, earthworm density was 6.4 times lower in the S position as compared to the UND areas, whereas fine root biomass was 45% lower in the S position than the UND areas.

Soil Microbial Properties and Enzyme Activity
Soil microbial properties and enzyme activity significantly differed among treatments and the UND areas (Table 2; p < 0.001). Soil microbial respiration, MBC, N mineralization, and MBN were significantly higher in the UND areas followed by TS > BS > S treatment (Table 2). However, NH 4 + and NO 3 − did not show any significant differences between the UND areas and TS, but were significantly higher in the UND areas and TS than in the BS and S treatments. Moreover, soil microbial respiration, MBC, and MBN were 55, 77, and 65% lower in the S position as compared to the UND areas, respectively. The highest activity levels of urease, acid phosphatase, arylsulfatase, invertase, and ß-N-acetylglucosaminidase were found in the UND areas, followed by TS > BS > S treatment. Significant positive correlations (p < 0.05) were reported among enzyme activities, soil moisture content, SOC, N, available P and K, fulvic acid, humic acid, earthworm dry mass, soil microbial respiration, fine root biomass, MBC, NH 4 + , NO 3 − , N mineralization, and MBN (Table 3). Soil bulk density, pH, and C/N ratio had significant negative correlations (p < 0.05) with enzyme activities.

Soil Physical, Chemical, and Biological Properties
The current study demonstrated that soil physical, chemical, and biological properties were influenced by topography and more specifically slope positions within the area trafficked by ground-based machines. Particularly, soil moisture and organic matter content were regulated by topography and slope positions, which is in line with [4,7,9]. Because water is directly involved in chemical and biological activities of decomposition and degradation as well as in physical degradation, changes of soil moisture directly affected the evolution of soil profiles and biological activities of the soil [8,38]. Likewise, previous studies reported that topography and slope positions can introduce heterogeneity in soil properties [1,6,8]. The highest recovery levels of soil physical properties including soil bulk density, total porosity, macroporosity, penetration resistance, and aggregate stability were observed under the TS treatment. However, the recovery values of soil physical properties in the TS treatment were still lower than the values found in the UND areas within a four-year period following machine traffic. These results are consistent with findings from Moorman et al. [39], stating that the particulate and dissolved materials displaced across the slope position from the summit to the toeslope resulted in soil detachment and erosion in the summit and backslope, and sediment deposition in the toeslope. Similarly, our results revealed that the recovery values of soil physical properties and soil moisture were significantly higher in the TS position than in the S and BS positions.
Our results highlight the importance of slope positions on the recovery levels of compacted soil properties over a 4-year period after traffic. Following harvesting operations, soil bulk density and penetration resistance remarkably increased in the machine operating trails in the tested slope positions (i.e., TS, BS, and S), compared to the UND areas. Consistent with our findings, several studies reported that soil physical properties (e.g., bulk density and penetration resistance) were significantly increased after machine traffic [10,29,40]. Four years after skidding operations, the recovery levels of soil bulk density and penetration resistance were higher in the TS than in the BS and S treatments, compared to the UND areas. Furthermore, by transporting the eroded materials and organic matter as well as litters from S to the TS position on bare mineral soil, water storage capacity augmented, which in turn suppressed the runoff and soil loss [9,17,41]. Consequently, SOC and soil moisture were enhanced and augmented soil fauna, which in turn resulted in reclamation of soil aggregates [42,43]. In contrast, soil aggregates in the S and BS positions were destructed after ground-based machine traffic and were left directly exposed to raindrop impacts, which led to increased runoff and detached soil particles. The effects from the mechanical stress caused during harvesting operations impaired air permeability and air-filled pore connectivity, which resulted in suppressing the activities of soil fauna as reported by Horn et al. [12], Cambi et al. [44], and Flores Fernández et al. [28].
Four years of litter production and its associated movement on compacted soil, from upper to lower slope positions, are largely responsible to rehabilitate the organic matter content and nutrients cycling in the TS, compared to the S and BS treatments. Hence, the greater decomposition rate of organic material provided a higher release and propagation of nutrients in the TS, compared to the S and BS positions. Significant negative correlation (p < 0.05) was observed among SOC content and bulk density. By increasing the organic matter content through a thicker litter layer, the soil bulk density decreased in the TS positions. In turn, the litter layer absorbs rainwater, governing soil temperature and moisture, intercepting throughfall, regulating temperature fluctuations from/to soil surface, and decreasing soil particle detachment. All of these benefits can accelerate the natural rehabilitation of soil bulk density [14,17]. Clay and silt particles were also detached from upper slope positions (i.e., S and BS treatments) and deposited in the lower slope position (i.e., TS), which increased water storage capacity. Results of this study demonstrated that soil particle size distribution was significantly different among the slope positions and those from the UND areas. Because of this relocation of fine particles, the highest sand content was observed in the S treatments, while the highest contents of silt and clay particles were found in the TS treatment. Reasons for this have been described by Marques et al. [45], stating that the S treatment was exposed more to the erosion processes than the TS and BS treatments, which concentrated the sand particles following removal of finer particles.
Our results indicated that soil chemical properties were considerably modified after machine traffic compared to the UND areas, a finding supported by Cambi et al. [44] and Jourgholami et al. [17]. Similar to our findings, Jourgholami et al. [17] reported that soil pH significantly increased after ground-based skidding operations. Specifically, Karaca et al. [1] reported that soil pH and EC values increased from the backslope to the terrace, however, they found no significant differences in available nutrients among the topographic positions (i.e. backslope, footslope, and terrace). Zhu et al. [7] and Lozano-García and Parras-Alcantára [3] also reported that SOC was significantly higher at the toeslope than at the shoulder. The lower C/N ratio of soil was found in the TS followed by S > BS > UND areas. Previous studies reported that the slope position can introduce a microclimate, which in turn regulated soil moisture and solar radiation and resulted in spatial variability of SOC, N, and C/N ratio [3,9,46]. The higher humic and fulvic acids were observed in the TS treatment due to a higher soil moisture availability, which resulted in improved soil physical and chemical properties. Similar to our findings, Fu et al. [47] validated that soil physical and chemical properties levels were higher in the foot slopes than in the middle and lower slopes. The higher soil physical and chemical properties in the TS than S and BS treatments enhanced the earthworm activity and abundances as reported by Jourgholami et al. [13]. Fine root biomass was found to be higher in the TS followed by BS and S treatments. This is likely caused by the increase of soil bulk density and penetration resistance in the S and BS treatments, which decreased the elongation and propagation of plant roots [10,44].

Soil Microbial Properties and Enzyme Activity
Physical and environmental factors such as leaching, amount of organic matter, clay transfer, and soil moisture are affected by the position of slope [3,7]. In line with the current study, many studies reported a positive correlation between soil microbial biomass and activities with SOC [9]. Accordingly, the greater amounts of available nutrients detected in the TS than those recorded in the S and BS can be attributed to the larger litter input and flux from the upper slopes. Hence, the high litter decomposition rates due to the higher microbial activity can influence organic matter processes and nutrient cycling, which is in agreement with the results of Fu et al. [47] and Zhu et al. [7]. Therefore, soil microbial biomass and activity were more pronounced in the TS as compared to the S and BS treatments. In contrast, the higher soil temperature and the lower moisture content and litter input resulted in a less appropriate environment, thus declining the soil microbial communities in the S and BS treatments, compared to the TS treatment. Consistent with our findings, Baldrian and Stursova [22] and Yang et al. [25] reported that the spatial variation of soil moisture due to the different slope position can be affected by microbial biomass and activity. Our results showed that higher levels of SMR, MBC, NH 4 + , NO 3 − , N mineralization, and MBN were reported in the TS treatment as compared to the S and BS treatments. Therefore, the machine operating trails located at the TS position were attributed to the greater availability of nutrients for soil microbial communities. The TS position can supply favorable ambient conditions and high soil moisture as well as suitable quality of substrate [9,48]. In contrast, the S and BS positions showed prominently lower values of SOC and N microbial properties (e.g., MBC, NH 4 + , NO 3 − , N mineralization, and MBN) than those observed in the TS, which indicates an unfavorable ambient condition.
In accordance with findings from Kang et al. [46], Liu et al. [48] and Fazlollahi Mohammadi et al. [9], our results accentuated the significance of topography features, and in particular slope position, in augmenting divergence in enzyme activity. Enzyme activity in soils is dependent on the microbial communities (e.g., soil microorganisms) and soil physical, chemical, and biological properties, which are commonly regulated by tree species [22]. The highest enzyme activity (i.e., urease, acid phosphatase, arylsulfatase, invertase, and ß-N-acetylglucosaminidase) measured in the TS treatment can be largely explained by the higher soil moisture in the TS than in the BS and S treatments. This increased enzyme activity indicated a more favorable environment for soil microbial activities and biomass as well as abundance. Similarly, Fazlollahi Mohammadi et al. [9] reported that the higher soil moisture and deposition of organic matter increased the enzyme activity more significantly in the TS treatment than in the S and BS treatments. Fazlollahi Mohammadi et al. [9] also stated that the reduced quantity of substrate in the S and BS positions was the most important factor for suppressing enzyme activity.
The result of the current study revealed that the important drivers influencing enzyme activity were the organic matter and soil moisture. The lower slope positions are associated with the higher soil moisture [49] and the greater organic matter deposition [48], which lead to higher SOC and N in the substrate for soil microorganism activities [9].
Results of Pearson's correlation in this study demonstrated that the activity levels of urease, acid phosphatase, arylsulfatase, invertase, and ß-N-acetylglucosaminidase showed a significant positive correlation (p < 0.05) with soil moisture content, SOC and N microbial biomass, NH 4 + , and NO 3 − , which was compatible with previous studies [9,48,49]. However, soil physical properties (i.e., bulk density), pH, and C/N ratio were negatively correlated with enzyme activity. Previous studies revealed that the primary control for enzyme activity is soil pH, and that small changes in pH can have a significant effect on enzyme activity [50,51]. In addition, slope positions had significant influence on the soil physical, chemical, and biological properties as well as SOC and N microbial biomass, which are the important drivers that contributed to enzyme activity, findings that are also supported by Jourgholami et al. [13]. Our hypothesis that slope positions have a significant influence on the physical, chemical, and biological soil properties, as well as enzyme activity of compacted soil in previously trafficked machine operating trails is supported by our data.
The results of the current study could be applied in several regions with similar stand conditions (large-diameter deciduous trees, sloped terrain, close-to-nature forest management) since the slope position is a concept that is well accepted as a driver of soil physio-chemical and hydrological processes and their associated impacts on creating diverse soil properties. Hence, understanding the soil restoration dynamics on the different catena positions of a slope is valuable information that can be used to efficiently plan the off-road traffic of ground-based forest machines.

Conclusions
In the present study, the effects of slope positions (e.g., summit (S), backslope (BS), and toeslope (TS)) on the recovery levels of compacted soil properties and enzyme activity were tested on machine operating trails four years after they had been trafficked by ground-based forest machines, compared to the UND areas. Results revealed that the highest recovery of soil physical, chemical, and biological properties was found in the TS rather than in the S and BS treatments, but the levels were still lower than those recorded in the UND areas, even 4 years after ground-based skidding operations. Our results confirmed that soil microbial and enzyme activity differed significantly among slope positions and the UND areas. The microbial and enzyme activities of compacted soil were higher in the TS than in the S and BS treatments. However, the full recovery of these properties as compared to those measured at the UND areas did not occur over a 4-year period in the study area.