The Inﬂuence of Vegetation Succession on Bearing Capacity of Forest Roads Made of Unbound Aggregates

: The aim of the research was to verify a common opinion concerning a positive inﬂuence of plants on the bearing capacity and durability of forest roads made of unbound aggregates. The surface bearing capacity is deﬁned as the ability to transfer tra ﬃ c loads without any excessive deformations which would hinder regular use of the surface and shorten its durability. It is a signiﬁcant functional feature of any road. The article analyzed the inﬂuence of road surface plant succession on its bearing parameters. The research was conducted on sections of experimental road constructed using macadam technology and reinforced partly with a biaxial geogrid. Measurements were taken with a lightweight Zorn ZFG 3000 GPS type deﬂectometer with a 300 mm pressure plate radius and 10 kg drop weight which allowed to measure dynamic deformation modulus (E vd ) and s / v parameter regarded as an indicator of compaction accuracy of the studied layer. E vd values and s / v parameters, which were obtained by measuring the road pavement covered in vegetation and after having it mechanically removed (mowed), were submitted to the analysis; next, they were compared with the results of an analysis done on areas naturally deprived of the plant cover and located in the immediate vicinity of the measuring points. The conducted research has indicated unfavorable inﬂuence of vegetation succession on the bearing parameters of the analyzed sections. The greatest drop in the mean E vd value was 39%, and s / v parameter deteriorated as much as 9%. Hence, a regular mowing of the road surface (including the maneuvering, storage and passing areas) should be taken as standard and mandatory procedures of forest road maintenance.


Introduction
Currently in Poland, in the forest road construction process, the roads are usually built from natural aggregate, whereas recycled materials reinforced with geosynthetics are used less often [1].
In the past, plant materials were used in road construction, especially on wetlands, such as for example, wooden poles, fascine or brushwood were used to lay the base course [2][3][4]. In Poland, a study on such a technology is conducted aimed at use of wooden rollers poles) or willow mat to reinforce the road courses constructed in high nature-value wetlands [5,6]. Among the technologies that use plant materials applied in the other parts of the world positive results are achieved by reinforcing the ground of the road course with coconut fibres i.e., [7][8][9][10][11]. Moreover, at times, during fieldwork it is possible to observe, makeshift methods of improving the bearing capacity of natural dirt road by using bark or woodchips or other methods of dirt road, and even airfield reinforcing such as sodding [12].
Roadside vegetation performs a very important protective function against soil erosion [13,14]. It protects the cut slopes and fill slopes, prevents fine material outwashing and sediment . Figure 2. The bearing capacity testing with lightweight deflectometer on the experimental forest road section no. 2 in spring 2020, in the place where vegetation had been previously removed-"E" symbol in Table 2 (Photo: S. M. Grajewski).
The vegetation covering the experimental road sections was at most 0.5 m high. In the process of data analyzing, the type of road pavement (on the section no. 1 and section no. 2) and the preparation manner of the measurement sites were taken into account as well (Table 2). Forests 2020, 11, x FOR PEER REVIEW 3 of 13 conducted in the direct vicinity, where the area was naturally devoid of vegetation cover (reference points). . Figure 2. The bearing capacity testing with lightweight deflectometer on the experimental forest road section no. 2 in spring 2020, in the place where vegetation had been previously removed-"E" symbol in Table 2 (Photo: S. M. Grajewski).
The vegetation covering the experimental road sections was at most 0.5 m high. In the process of data analyzing, the type of road pavement (on the section no. 1 and section no. 2) and the preparation manner of the measurement sites were taken into account as well (Table 2). The bearing capacity was measured with a lightweight ZFG 3000 GPS type deflectometer (Zorn Instruments GmbH & Co. KG, Hansestadt Stendal, Germany) with a 300 mm radius pressure plate and a 10 kg drop weight. The instrument was used to measure dynamic deformation modulus (E vd ) and s/v parameter (ratio of mean load settlement (deflection) after three impact LFWD tests to mean settlement velocity (deflection rate)), which is an indicator of compaction accuracy of the studied layer [38]. The E vd values and s/v parameter obtained during the tests on the road pavement covered with vegetation and after its mechanical removal were later compared with the results of the tests conducted in the direct vicinity, where the area was naturally devoid of vegetation cover (reference points). The vegetation covering the experimental road sections was at most 0.5 m high. In the process of data analyzing, the type of road pavement (on the section no. 1 and section no. 2) and the preparation manner of the measurement sites were taken into account as well ( Table 2). The results of bearing capacity measurements taken in the places deprived of vegetation, located in the direct vicinity of spots from set "A" sites (reference point)

Spring 2019
The results of bearing capacity measurements taken in the places covered with vegetation The results of bearing capacity measurements taken in the places deprived of vegetation, located in the direct vicinity of spots from set "C" sites (reference point) D  The results of bearing capacity measurements taken in the places with vegetation after its mechanical eradication (mowing to the level of the road pavement) Spring 2020 The results of bearing capacity measurements taken in the sites deprived of vegetation in the direct vicinity of spots from set "E" (reference point) Autumn 2020 Results of bearing capacity measurements taken in the places with vegetation after its mechanical eradication (mowing to the level of the road pavement) Autumn 2020 The results of bearing capacity measurements taken in the sites deprived of vegetation in the direct vicinity of spots from set "G" H 2 Autumn 2020 The results of bearing capacity measurements taken in the places with vegetation after its mechanical eradication (mowing to the level of the road pavement) Autumn 2020 The results of bearing capacity measurements taken in sites deprived of vegetation in the direct vicinity of spots from set "I" J *-data sets A, C, E, G and I were prepared to check the effect of vegetation on the bearing capacity of road surfaces and to compare to the B, D, F, H and J data sets, respectively.
The plant species which occurred in the study area were catalogued in detail, determining the species structure and surface coverage. The degree of coverage was determined by geodesic methods on randomly selected sections of the road surface. The species structure was recognized by analyzing the species within randomly selected square segments with sides of 100 cm. The plant inventory made it possible to select the bearing capacity test sites where vegetation coverage reached 80-90%.
For statistical evaluation of differences between groups of variables, the Tukey test for different N's, available in Statistica™ 13 (TIBCO ® Software Inc., Palo Alto, CA, USA), was used.

Results
During the spring 2019 and 2020 field work, in the study area, in total 10 vegetation taxa aged 1-3 years were catalogued ( Table 3). The vegetation coverage was about 80%. The species structure was by far dominated by a young generation of silver birch (Betula pendula Roth) and Scots pine (Pinus sylvestris L.), which covered the greatest part of the research area. In autumn 2020, a total of 34 plant species aged 1-4 years were identified, with 30 taxa on the road section no. 1, and 16 taxa on the section no. 2 ( Table 3). The degree of road surface coverage increased to about 90%.   Table 4 and Figure 3). The differences shown were confirmed statistically with the exception of C and D data sets relations ( Table 4). The presence of vegetation on the road pavement, regardless of the manner the measurement was taken, resulted in lowering the E vd value by 18% (in cases I-J), and sometimes even by 39% (in cases B-A). There is also a noticeable 13% (D-J), 17% (B-H) and 19% (F-J) decrease in the value of the dynamic modulus of deformation in the third testing season, comparing to the previous test results. The road pavement of section no. 1 (without the biaxial geogrid), despite displaying initially higher bearing capacity (B), responded with significantly lowered values of the dynamic deformation modulus values once the vegetation appeared (A). The development of vegetation succession on the road pavement also has a negative influence on the obtained values of s/v parameter (Table 4, Figure 4), but the observed differences in values are much lower and reach between 3% (F-E) and 10% (G-H). These relationships were not statistically confirmed, except for the G-H data sets relationship (Table 4).
It is worth emphasizing that the development of vegetation is accompanied by a significant increase of value variability for both E vd as well as s/v ratio which is, expressed with coefficient of variation z p ( Table 4, Figures 3 and 4).  Key: DSS-data set symbol, N-number of measurements; Evd-dynamic deformation modulus; x , xmin, xmax-mean, minimum and maximum values; SD-standard deviation; zp-coefficient of variation; s/v-ratio of mean plate settlement to settlement velocity, *-DSS marked with the same letters do not differ significantly-Tukey test for different N (α = 0.05).  Table 2). *-DSS marked with the same letters do not differ significantly-Tukey test for different N (α = 0.05).  Table 2). *-DSS marked with the same letters do not differ significantly-Tukey test for different N (α = 0.05).
Forests 2020, 11, x FOR PEER REVIEW 9 of 13  Table 2). *-DSS marked with the same letters do not differ significantly-Tukey test for different N (α = 0.05).

Discussion
Contrary to the popular opinion about the positive influence of vegetation on the forest road pavement, it is necessary to state that the occurrence of vegetation succession on the analyzed road  Table 2). *-DSS marked with the same letters do not differ significantly-Tukey test for different N (α = 0.05).

Discussion
Contrary to the popular opinion about the positive influence of vegetation on the forest road pavement, it is necessary to state that the occurrence of vegetation succession on the analyzed road constructed with unbound aggregate has a negative impact on its bearing capacity parameters.
While converting the dynamic deformation modulus (E vd ) into the values of secondary deformation modulus (E 2 ), which is simply done by doubling the E vd values [40,41], it should be noted that the observed significant decrease of bearing capacity parameters caused by development of vegetation can lead to failing to meet the minimum requirements in this scope i.e., E 2 ≥ 100 (140) MN·m −2 for road pavement made of unbound crushed stone aggregates [37,[42][43][44].
The values of dynamic deformation modulus, which were measured on the surface without the vegetation on both experimental sections, appeared to differ significantly ( Table 4). The reason behind it can be found in the reported problems related to a proper determination of the bearing capacity of the structure reinforced with geosynthetics [45][46][47]. Nevertheless, the results of the research of both sections in the areas with vegetation suggest that the presence of the geogrid in the road pavement limits the negative impact of the vegetation developing on the pavement (Table 4, Figures 3 and 4). Apparently, the effective wedging of the aggregate by this rigid geosynthetic [48,49] allows at least to slow down the loosening of the aggregate structure by the root systems [29,30] resulting in a decrease in bearing capacity parameters.
The recorded decrease in the value of dynamic modulus of deformation in the third season of testing in comparison to the previously obtained results can be explained by the autumn conditions of the measurements in 2020: the higher humidity of the tested system "pavement structure-road surface" had a negative impact on the obtained parameters of the bearing capacity of the road surface [41,[50][51][52][53]. However, the recorded decrease in bearing capacity of the road surface may also be affected by the continuous free growth of vegetation root systems, reaching its third growth season, which change the structure of aggregate layer. This is in line with observations made by [22], according to which the dry density of road surface layer decreases with the increase in vegetation cover, which results from successive development of the roots causing an increase in the porosity of surface material.
The observed deterioration of the bearing capacity of the pavement can also be associated with the successive accumulation of organic matter on the pavement after each vegetation season. The stages of plant succession in time are described by [14]. The authors state that immediately after commissioning, the road surface is free of organic pollutants. However, if the strongest seedlings survive after germination, they use remains of successively dying other plants. The assimilation apparatus of plants growing in the vicinity of the roadway also falls down on the pavement. The resulting plant cover shades and cools the surface, initiating at the same time the soil-forming process, including the decomposition of organic compounds and biochemical transformation involving acidic substances secreted by the roots of newly-grown plants [54]. The substances resulting from these processes, including humic acids, migrate deep into the sub-base, affecting the road structure. The organic matter in periods of high moisturization absorbs water, which freezes in winter. As a result, the organic substance penetrating the surface course of the road, alternately increases and decreases its volume, and thus loosens the course structure. The need to remove organic matter from forest roads was indicted by [35], among others. Removing surface organic matter can delay the succession process on roads used occasionally.
The root systems of plants themselves can also cause damage to the surface. Pioneer plant root architecture tends to penetrate the substrate both vertically and horizontally [28]. On the analyzed forest road surfaces, the dominant species was silver birch. The birch has root system of the intensive type, the small adventitious roots are grouped into bundles that penetrate the ground firmly ( Figure 5). Similarly, the roots of other pioneer species listed in the inventory results (see Table 3) strongly penetrate the road surface.
The root systems of plants themselves can also cause damage to the surface. Pioneer plant root architecture tends to penetrate the substrate both vertically and horizontally [28]. On the analyzed forest road surfaces, the dominant species was silver birch. The birch has root system of the intensive type, the small adventitious roots are grouped into bundles that penetrate the ground firmly ( Figure  5). Similarly, the roots of other pioneer species listed in the inventory results (see Table 3) strongly penetrate the road surface. Overground parts (shoots) and root systems are shown. The average weight of grey poplar was: 2.3 g, and silver birch 3.1; the weight of above-ground and underground parts were: 1.1 g and 1.2 g respectively for grey poplar; 2.0 g and 1.1 g for silver birch (Photo: A. Czerniak). Overground parts (shoots) and root systems are shown. The average weight of grey poplar was: 2.3 g, and silver birch 3.1; the weight of above-ground and underground parts were: 1.1 g and 1.2 g respectively for grey poplar; 2.0 g and 1.1 g for silver birch (Photo: A. Czerniak).

Conclusions
(1) The significant deterioration of bearing capacity parameters of the pavement caused by the developing vegetation succession on the surface indicates that systematic mowing of the surface (including the maneuvering, storage and passing areas) should be regarded as a standard and mandatory practice in order to preserve roads in forests. (2) Cutting off the vegetation close to the ground does not lead to the significant increase in bearing capacity parameters, which means that the owners/maintenance crew of the road should counteract any vegetation growth, and as a final measure eradicate it as soon as possible. (3) On the forest roads exploited less intensively, it is recommended to mechanically remove the organic matter that accumulates on the surface. This treatment will delay the growth of vegetation on the surface. Humus removal is easier and cheaper than mechanical removal of plants and safer when using herbicide compounds.