The Impact of Road Investments on the Forest Environment—Case Study: The Impact of Asphalt Roads on the Health Condition and Growth of Trees

Abstract

The study was conducted within the framework of the R&D project entitled “Environmental aspects of reconstruction and construction of forest roads using selected asphalt binders”. The aim of this study was to determine the impact of an asphalt road segment located in the Przymuszewo Forest District, the Regional Directorate of the State Forests in Toruń, Poland (53°57′24.14″ N, 17°34′38.30″ E), on growth increments and health condition of trees in a 60-year-old pine stand. The first stage of the investigation required laboratory analyses and field observations to assess the impact of mineral filler and asphalt mixtures on roadside habitats. Chemical analyses were performed on samples of wood and soil collected from transects located at three distances from the asphalt road, as well as dendrological and dendrometric analyses in circular sample plots of 0.02 ha in analogous transects parallel to the selected asphalt pavement. Analyses of soils and wood showed no negative environmental impact associated with the road. Contents of all the assayed elements in wood were lower compared to their contents in soil, which indicates their markedly reduced phytoextraction. Statistical analyses showed that the asphalt road has an effect on the adjacent forest mass, but that this effect is statistically insignificant with respect to the mean annual increments diameter at breast height and the defoliation of trees growing at various distances from the road.

1. Introduction

In Poland, forests cover almost 30% area of the country and are administered by the State Forests National Forest Holding (PGL LP). Forest roads in Poland serve economic, fire control and social functions [1,2]. Some forest roads are accessible for traffic as public roads. The State Forests districts also execute joint road investments with local government units. These roads are crucial for local communities, reducing transport exclusion of inhabitants in non-urban areas. Such road investments, when properly consulted and executed, are highly appreciated by local communities, particularly in densely forested areas with poorly developed public infrastructure.

Obviously, only the most important segments of forest roads and roads administered by local governments need to be constructed as hard surface roads [3]. The construction or reconstruction of forest roads and joint investment projects need to be preceded by public consultations. During the consultation process involving both local residents, local government bodies, agencies and organizations promoting environmental protection, it is essential to present credible arguments concerning the ecological, economic and social aspects of proposed technologies. Forest roads accessible for public traffic have to be safe and durable. Pavements made from unbound aggregates do not meet these requirements. Asphalt is a commonly used binder for road construction around the world. The success of this material is due to the popularization of the production process, the ease of use in construction and the possibility of recycling. For this reason, it is advisable to search for new road construction technologies, including those based on asphalt binders [4]. It also needs to be stated that some asphalt roads, after several decades of service life, require repairs or reconstruction, e.g., the application of another layer, preferably from inert asphalt chemically neutral for the forest environment [5].

In view of the above, at the stage of formal legal preparations for the investment project, it is crucial to consider the environmental aspect [6]. Many public roads, including motorways and expressways, run through forested areas taken over from the State Forests National Forest Holding (PGL LP) on the power of a special legal act [7]. Knowledge of the asphalt and the environment interactions is essential for all stakeholders, already at the stage of formal legal preparations for their investments, and social consultations. Vehicle traffic on forest and countryside roads is unique in character and highly seasonal. Design speed for forest roads is 30 km/h. For this reason, it is advisable to develop low-cost technologies using asphalt binders. Identification of the potential impact of asphalt on the environment, including forests, is also important when constructing public roads.

Numerous studies show that roads may have a diverse impact on the forest environment [8], particularly in areas of considerable nature value [3]. Analyses were conducted on the impact of forest roads on vegetation and soil conditions [9,10], stand structure and tree growth [11,12]. A significant impact was found to result from the fragmentation of forest complexes [13]. The environmental impact of forest roads may be evident at the stage of the construction of forest roads [14] and during their service life [15]. Excessive amounts of some trace elements may destabilize the homeostasis of the soil environment and weaken the condition, and thus the abundance, of forest stands. The harmfulness of trace elements depends not only on the concentration, but also on their forms of occurrence [16]. Particularly large emissions of trace elements can occur in the vicinity of roads with very heavy traffic. The deposition of chemical pollutants may result more from exhaust emissions than road construction [17,18].

Photoperiodism and thermoperiodism trigger endogenous rhythmical fluctuations in hormonal processes in cambium and the formation of various wood components. Growth in diameter in trees consists in the formation (deposition) of annual growth rings. Diameter increment depends on climatic conditions (temperature, precipitation) and the availability of nutrients in a given year. It is subjected to considerable fluctuations and its causes are divided into short- and long-term factors. Short-term fluctuations in the increment of the diameter at breast height are connected primarily with weather conditions changing from year to year, mainly temperature and precipitation. Long-term fluctuations are complex in character, influenced by a complicated system of factors. In the opinion of Bruchwald and Dmyterko [19], the main cause of long-term changes in diameter growth is related with the competition between individual specimens within the stand. In this respect, allelopathic interactions also need to be considered. The other causes of long-term fluctuations in growth increment may be classified into three groups: biotic, abiotic and anthropogenic. Biotic factors include mainly insects and fungi. Abiotic factors causing a reduced increment in diameter as well as intensive thinning of the stand are connected with the lowering of the groundwater levels and prolonged waterlogging of forest soils after flooding. Anthropogenic factors lead to changes primarily in habitat elements which, in turn, directly affect fluctuations in tree growth processes. Moreover, long-term effects of industrial emissions or exhaust emissions cause adverse habitat changes, themselves leading to reduced growth rate and mass die-back of trees in the stand. External factors have a permanent impact on tree health.

The aim of this study was to determine the impact of an asphalt road segment located in the Przymuszewo Forest District on the growth increment and health of trees in a pine stand. The investigated road was modernized in 2006 using a mineral–asphalt recapping layer in accordance with the PN-S-96025:2000 standard.

In connection with the implementation of the objectives set in the first stage of the research, measurements and field observations were carried out, and the obtained empirical material was the subject of laboratory analyses on the impact of mineral fillers and mineral–asphalt mixtures on roadside habitats.

The following research hypotheses were proposed:

• Asphalt binders used to reconstruct the forest road have a significant impact on tree growth increments in the pine stand and the tree health;
• The adverse impact of the asphalt pavement is correlated with the distance from the road.

2. Materials and Methods

Investigations were conducted within within the framework of the research and development project entitled “Environmental aspects of reconstruction and construction of forest roads using selected asphalt binders”. The Poznań University of Life Sciences is the contractor for LOTOS Asphalt Sp. z o. o. with its registered seat in Gdańsk.

The first stage in the study included required laboratory analyses and field observations to assess the potential impact on the roadside environment of mineral–asphalt mixtures type AC 11S 50/70 and AC 11S 50/70 WMA as well as SMA 16 JENA 50/70 and SMA 16JENA 50/70 WMA based on 50/70 and 50/70 WMA asphalts (Figure 1).

The detailed scope of performed works included:

• Laboratory analyses concerning the rate of leaching of heavy metals and hydrocarbons from provided samples of composites. Samples were subjected to leaching with simulated rainwater and distilled water. Moreover, total contents of trace elements essential for the environment were analyzed;
• Determination of the effect of the technical condition of the pavement on the intensity of the leaching process. Chemical analyses were performed on samples pre-crushed in a compressive strength testing machine (simulating the pavement ageing process);
• Chemical analyses of wood and soil samples collected from transects located at three distances from the analyzed asphalt road;
• Field observations comprising dendrological and dendrometric analyses of selected pine stands in transects parallel to the selected asphalt pavement.

This paper presents investigations specified in points 3 and 4.

2.1. Analyses of Soil Composition (Auxiliary Analyses)

Soil samples were taken in the center of each of the 9 circular plots. Thus, for zone I they were 16 m from the edge of the asphalt road, zone II, 41 m and zone III, 66 m. The distance between the centers of the circular surfaces and the sampling site in a zone was 25 m. Analyses were conducted following extraction of a dried soil sample using the hot plate aqua regia digestion method in accordance with the PN-ISO 11466:2002 standard. Contents of selected elements (As, Ba, Cd, Cr, Cu, Mo, Ni, Pb, Sb, Se and Zn) were determined using inductively coupled plasma/optical emission spectrometry (ICP-OES) based on the PN-EN ISO 11885:2009 standard, while in the case of mercury, atomic absorption spectrometry was applied following the PN-ISO 16772:2009 standard. Moreover, the mineral oil/petroleum hydrocarbon index was determined according to PN-ISO 16703:2011, while total contents of polycyclic aromatic hydrocarbons were assayed based on the PN-ISO 13877:2004 standard.

2.2. Analyses of Wood Composition

Analyses were performed following extraction of a dried wood sample using the hot plate aqua regia digestion method according to the PN-EN 13657:2006 standard. Contents of selected elements (As, Ba, Cd, Cr, Cu, Mo, Ni, Pb, Sb, Se and Zn) were determined by applying inductively coupled plasma/optical emission spectrometry (ICP-OES) following PN-EN ISO 11885:2009, while mercury was assayed by atomic absorption spectrometry according to PN-EN ISO 12846:2012+Ap1:2016-07 and the PB/I/11/D:10.04.2020 testing procedure developed by the laboratory performing the analyses.

2.3. Dendrological and Dendrometric Analyses

For the analyses of the impact of the asphalt road on annual growth ring increment and crown defoliation, it was decided to select a pure Scots pine stand aged approx. 60 years. Such a criterion was met by the stand located in the Przymuszewo Forest District (53°57′24.14″ N, 17°34′38.30″ E) Figure 2. The basic forest inventory data for this stand are as follows: pine (in the Polish forest classification So10, i.e., 100% pine) aged 58 years, subcompartment area 6.37 ha, forest habitat type (Polish classification of TSL)—fresh coniferous forest (Bśw), moderate crown closure, stocking 1011 trees per 1 ha, average diameter at breast height 19 cm, mean tree height 18 m, site index 27.15 and growing stock 265 m³/ha. Three zones were established in the stand and, in each of them, three circular sample plots were marked. The distance between the centers of the circular sample plots within individual zones and between the zones was 25 m. The circular sample plots were 0.02 ha in size, which followed the recommendations binding in the Large-scale Forest Condition Inventory in Poland.

Centers of the three circular sample plots located in the first zone (closest to the asphalt road) were situated at a distance of 16 m from the road edge. This distance was related, among other things, to the presence of a fire protection zone (free from trees) running along the roads. Inclusion of this zone into the circular sample plots would prevent an objective comparison of analyzed values of the features characterizing individual trees and the entire stand. Centers of the circular sample plots in the second zone were located at a distance of 25 m from those in the first zone (41 m from the road edge) while, in the third zone, it was by 25 m from those in the second zone (66 m from the road edge). The distance of the centers of the circular sample plots within individual zones was also 25 m. The center of each circular sample zone was stabilized and their boundaries were established. In the plots, trees were numbered and permanently marked at a height of 130 cm. Diameter at breast height (DBH) was measured twice (using a DigiTech Professional II caliper by Haglof) for each living tree. The first DBH measurement was performed so that the caliper arm faced the center of the plot, while the second measurement was perpendicular to it. Moreover, the height of each tree was measured using an electronic Vertex IV hypsometer.

The degree of defoliation in pines was determined using the bioindicator method. The degree of crown damage in Scots pines was assessed in two directions using binoculars according to three criteria: defoliation estimated accurate to 5%, the number of needle generations (in the apical part of the crown) and crown opening (

Table 1. Degrees of crown damage of Scots pine (Pinus sylvestris L.) [20].

Degree of Damage	Defoliation [%]	Number of Needle Age-Groups *	Tree Crown **
0—no damage	0–10	3	complete
1—weak	11–25	2–2.5	slightly attenuated
2a—moderate	26–40	1.5–2	moderately attenuated
2b—medium	41–60	1–1.5	strongly attenuated
3—strong	>60	1	very strongly attenuated
4—complete	100	-	-

* Assessment in the upper (illuminated) part of the crown, ** Assessment of the entire crown.

, Figure 3) [20]. As a result of the conducted inventories, the mean defoliation rate and the percentage share of trees in individual classes of tree damage were determined.

After all diameters at breast height and heights of the trees were measured in each circular sample plot, the dimensions of three sample trees were calculated for each circular plot. The mean sample trees were selected following the Urich method (division of trees measured in a circular sample plot into three diameter size classes with identical numbers of trees in a class). Thus, all trees growing at each circular sample plot were represented by three trees (a tree of a small (M), medium (S) and large (D) diameter at breast height). From each sample tree (with nine trees in each zone) two core samples were collected using an increment borer. A total of 54 samples were taken (6 for each plot and thus 18 for each zone). One core sample was collected in the northerly direction (N) and the other—in the easterly direction (E). Annual DBH increments were measured using the Regent Instruments WinDENDRO Regular STD4800 measuring system accurate to 0.001 mm (Sainte Foy, QC, Canada). Annual increments in the northerly (N) and easterly (E) directions determined on each sample tree were used to calculate the mean annual increment in diameter for the trees.

The primary task was to calculate statistical characteristics for annual DBH increments for trees in individual zones and to detect significant changes between values of annual diameter increments for trees growing in the three analyzed zones within the last 21 years. For this purpose, the analysis of variance ANOVA was applied (using the Statistica 13.3 software package) taking into consideration the basic assumptions of the ANOVA tests, such as normal distribution of variables in each population (group) and homogeneity of variance in all the populations (groups). It was also assumed that groups of trees of different heights have no significant differences in annual growth and are analyzed as a uniform sample. If the first of the assumptions was not met, then the non-parametric Kruskal–Wallis test and the median test were applied. If the assumption of variance homogeneity was not met, means were evaluated using the Welch test. Due to the small sample sizes, the normality of distribution for annual DBH increments was assessed using the Kolmogorov–Smirnov test with the Lilliefors test and the Shapiro–Wilk test. In this study, the significance level α was assumed at p = 0.05.

3. Results

3.1. Analysis of Soil Composition

Analyses were conducted on soil samples collected from three transects at a distance of 16, 41 and 66 m, respectively, from the asphalt road. The median of concentrations was adopted as the representative value. Results of the analyses are presented in

Table 2. The content (mg/kg⁻¹) of the analyzed elements/groups of chemical compounds in the soil depending on the distance from the road and the requirements for group III (soil in forests) according to the Regulation of the Minister of the Environment of 1 September 2016, on the method of assessing soil contamination (Journal of Laws No. U. of 2016, item 1395) [21].

Chemical Element/ Group of Compounds	Distance from Road			Group III
	16 m	41 m	66 m
As	<5.00	<5.00	<5.00	50.00
Ba	6.40	9.40	8.00	1000.00
Cd	<0.05	<0.05	<0.05	10.00
Cr	3.10	2.60	3.20	500.00
Cu	82.00	1.00	<0.40	300.00
Hg	<0.05	<0.05	<0.05	10.00
Mo	<0.40	<0.40	<0.40	100.00
Ni	1.30	1.40	1.40	300.00
Pb	10.00	17.00	17.00	500.00
Sb	<5.00	<5.00	<5.00	-
Se	<5.00	<5.00	<5.00	-
Zn	6.50	8.50	8.00	1000.00
PAH *	<0.10	0.17	0.24	1.00
Mineral oil	78.00	120.00	92.00	300.00

* Polycyclic aromatic hydrocarbons.

.

Recorded results are compared directly to the requirements for soil quality in forests (group III) specified in the Resolution of the Minister of the Environment of 1 September, 2016, on the method to assess soil surface contamination (the Journal of Laws (Dziennik Ustaw), 2016, item 1395) [21]. In terms of all the investigated parameters, the tested samples (not only for the values of medians, but also for all individual results) met the above-mentioned requirements. Moreover, no uniform trend was found for changes in the concentrations of the analyzed elements or compounds correlated with the distance from the road.

3.2. Analysis of Wood Composition

Analyses were performed for samples of wood collected—similarly as in the case of soil analysis—in three transects at the distance of 16, 41 and 66 m, respectively, from the road. Sampling sites for wood samples were identical with sites for soil sample collection. The median of concentrations was adopted as the representative value. Analyses of wood were limited to analyses of the elemental composition. Results are given in

Table 3. The content (mg/kg) of the analyzed elements in wood depending on the distance from the road.

Chemical Element	Distance from Road
	16 m	41 m	66 m
As	<5.00	<5.00	<5.00
Ba	4.50	1.60	1.30
Cd	<0.05	<0.05	<0.05
Cr	3.90	<0.30	<0.30
Cu	0.79	<0.40	<0.40
Hg	0.14	<0.05	<0.05
Mo	<0.40	0.42	<0.40
Ni	2.20	<0.40	<0.40
Pb	<1.00	<1.00	<1.00
Sb	<5.00	<5.00	<5.00
Se	<5.00	<5.00	<5.00
Zn	12.00	5.90	7.20

.

Contents of all the determined elements were lower than their levels in the soil, which indicates their markedly limited phytoextraction, and simultaneously the greatest efficiency of this process in plants growing in the nearest vicinity of the road (16 m).

3.3. Dendrological and Dendrometric Analyses

3.3.1. Annual Diameter Increments at Breast Height

Basic statistical characteristics such as arithmetic means and minimum and maximum values of annual DBH increments in the three analyzed zones in the years 2000–2021 are given in

Table 4. Basic statistical characteristics of the annual increments of DBH of trees in the three analyzed zones in 2000–2021.

Year	Zone I			Zone II			Zone III
	Average	Min.	Max.	Average	Min.	Max.	Average	Min.	Max.
	[mm]
2000	1.580	0.740	2.304	2.180	0.960	3.346	2.034	1.520	3.096
2001	1.726	0.454	2.833	2.134	1.214	3.705	2.264	1.460	2.668
2002	1.796	0.406	2.436	2.243	1.103	3.332	2.419	1.222	3.606
2003	1.435	0.531	2.178	1.798	0.988	2.652	1.769	0.749	2.958
2004	1.264	0.419	1.910	1.907	0.881	3.204	1.818	0.934	2.757
2005	0.822	0.334	1.392	1.097	0.214	2.073	1.118	0.532	1.647
2006	0.867	0.318	1.482	0.974	0.227	2.081	1.103	0.730	1.863
2007	1.172	0.272	1.920	1.125	0.199	2.148	1.553	1.019	2.178
2008	1.253	0.221	2.239	1.337	0.295	2.509	1.675	1.095	2.399
2009	1.285	0.621	2.094	1.588	0.304	3.272	1.694	1.087	2.743
2010	1.065	0.463	1.753	1.165	0.357	2.005	1.239	0.758	1.644
2011	0.997	0.631	1.816	1.014	0.355	1.962	0.984	0.488	1.564
2012	1.315	0.561	1.996	1.396	0.557	2.240	1.486	0.895	2.325
2013	1.227	0.625	1.871	1.416	0.549	2.324	1.426	0.719	2.243
2014	1.420	0.919	1.836	1.696	0.579	3.161	1.511	0.673	2.279
2015	0.957	0.437	1.206	1.196	0.541	2.159	0.966	0.425	1.329
2016	0.895	0.693	1.309	1.075	0.439	2.080	0.765	0.232	1.170
2017	0.709	0.400	0.946	0.905	0.341	1.956	0.672	0.148	1.186
2018	0.698	0.258	1.037	0.964	0.281	2.054	0.799	0.183	1.675
2019	0.840	0.346	1.381	1.111	0.294	1.906	1.003	0.241	1.963
2020	0.815	0.153	1.237	1.033	0.295	1.875	1.003	0.398	2.076
2021	0.706	0.285	1.145	0.886	0.298	1.578	1.071	0.470	2.072

. In the first zone of trees growing the closest to the asphalt road for the 21 analyzed years, 10 years were characterized by mean increments below 1 mm, while 11 years had increments exceeding this value. The greatest mean annual increment was recorded in 2002 (1.796 mm) and it was 2.57-fold greater than the smallest from 2018 (0.698 mm). Mean minimum increments in the analyzed years did not exceed 1 mm and ranged from 0.153 (2020) to 0.919 mm (2014). When analyzing the maximum mean values of annual increments for trees in this zone, it needs to be stressed that only in 2017 was this value (0.946 mm) below 1 mm. In the other years, this increment was greater than 1 mm, with the highest value recorded in 2001 (2.833 mm).

In zone II, the smallest mean DBH increment was recorded in 2021 (0.886 mm). The mean increment below 1 mm was found also in the case of three years. In the other 17 analyzed years the mean increments exceeded 1 mm, with the greatest value recorded in 2002 (2.243 mm). Mean minimum increments ranged from 0.199 mm (2007) to 1.214 mm (2001), while maximum increments ranged from 1.578 mm (2021) to 3.705 mm (2001).

In zone III, located the farthest from the asphalt road, the smallest mean DBH increment was recorded in 2017 (0.672 mm) and it was over 3.5-fold smaller than that observed in 2002 (2.419 mm). Even greater discrepancies were found between the minimum mean annual increments, since the minimum value recorded in 2017 (0.148 mm) was over 10-fold smaller than the highest minimum value recorded in 2000 (1.560 mm). Mean maximum increments noted in this zone fall within the range from 1.170 mm (2016) to 3.606 mm (2002).

Generally, since 2000, the mean DBH increment in mean sample trees growing the closest to the asphalt road (zone I) was smaller than the increments recorded for trees from zones II and III (except for 2007, 2011, 2016 and 2017, when the mean was slightly higher than in zone III).

Figure 4 presents coefficients of variation for annual DBH increments of trees growing in the three zones distinguished by the distance from the asphalt road in the period from 2000 to 2021. The smallest variability in the annual DBH increments in zone I was recorded in 2014 (24.3%). In zone II, such a situation occurred in 2003 (31.6%), while in zone III, it was in 2001 (19.6%). The greatest coefficients of variation amounted to over 50% and in the individual zones they were as follows: I—53.4%, II—66.1% and III—66.7%. During the analyzed last 21 years, when comparing variability in individual zones, only in zone I was the greatest variation recorded in two years (2001 and 2002), while in zone III, it was recorder in three years (2003, 2018 and 2019). In the course of 17 years, the coefficients of variation for annual diameter increments were greater in zone II than in zones I and III.

Table 5. Distribution normality analysis, tests for homogeneity of variance and significance tests for differences in annual breast height increments (from 2000 to 2021) of trees in three zones of distance from the asphalt road.

Year	Analysis of the Normality of the Distribution Using the Shapiro–Wilk Test			Homogeneity of Variance Test		Significance of the Differences
				Levene’s Test	Brown–Forsythe Test	Test F	Kruskal–Wallis Test
	p [%]
	Zone I	Zone II	Zone III
2000	0.7632	0.2317	0.1562	0.1479	0.4479	0.1428
2001	0.9011	0.3439	0.0884	0.3116	0.3648	0.2427
2002	0.1645	0.7814	0.9607	0.6575	0.6858	0.1731
2003	0.8380	0.7806	0.9819	0.5533	0.6180	0.3492
2004	0.8049	0.9183	0.3560	0.2184	0.2479	0.0816
2005	0.2615	0.9696	0.9092	0.1788	0.2019	0.3082
2006	0.5756	0.5561	0.1440	0.2669	0.2988	0.6030
2007	0.5320	0.4788	0.6563	0.1138	0.2274	0.2345
2008	0.9405	0.1773	0.0254				0.4148
2009	0.7840	0.3849	0.3543	0.0079	0.0674		0.4259
2010	0.7542	0.5105	0.2917	0.0737	0.1011	0.7120
2011	0.0808	0.3999	0.5310	0.1310	0.2775	0.9899
2012	0.6252	0.1914	0.1514	0.4555	0.4762	0.8263
2013	0.8990	0.2200	0.7126	0.0756	0.2679	0.6707
2014	0.2723	0.6334	0.9824	0.0294	0.0727		0.7423
2015	0.1153	0.7308	0.2746	0.1820	0.1725	0.3239
2016	0.0467	0.5949	0.2136				0.5758
2017	0.6734	0.3326	0.6412	0.0217	0.0732		0.7581
2018	0.5700	0.5844	0.4468	0.0777	0.1057	0.5030
2019	0.9817	0.5894	0.7896	0.1248	0.2402	0.5113
2020	0.4986	0.6042	0.2720	0.2153	0.4423	0.5813
2021	0.3001	0.4299	0.3173	0.0864	0.0828	0.2233

presents the results of the Shapiro–Wilk test (in view of their high power compared to the other tests). Normal distribution was not found for the annual increments in diameter of trees growing in zone III (located the farthest from the road) in 2008 and in zone I (the closest to the road) in 2016 (the computer calculated p value is lower than the assumed significance level α = 0.05). Another important assumption of the ANOVA test is the homogeneity of variance in all the groups (annual DBH increments in the three analyzed zones based on the distance from the road). Levene’s test (Table 5) indicates that the condition of variance homogeneity is not met for the annual DBH increments for trees growing in the three zones in 2009, 2014 and 2017 (the calculated p is lower than the assumed significance level α = 0.05). Thus, since the assumptions of the ANOVA tests are not met in the case of five years (2008, 2009, 2014, 2016 and 2017), the non-parametric Kruskal–Wallis test was applied. For the other 17 years, parametric tests were performed. Results of both tests are also given in Table 5. Computer calculated coefficients of significance p (calculated using both a parametric and non-parametric test) in each year have values greater than the assumed level of significance α = 0.05. For this reason, a null hypothesis was adopted assuming that the average annual diameter increments of trees growing in the three zones (differing in their distance from the road) are identical (i.e., they do not differ statistically significantly). This shows that the asphalt road has an effect on the adjacent forest mass, but this effect is statistically insignificant on the mean annual DBH increments of trees growing at different distances from this road.

3.3.2. The Degree of Tree Defoliation

Basic descriptive statistics, i.e., the minimum and maximum values, arithmetic means, standard deviations and coefficients of variation for defoliation rates of trees in the three analyzed transects, are presented in Figure 5.

In the three transects, the defoliation rate of pines ranged from 5 to 50%, apart from one outlier in the case of transect II (75%). Mean defoliation rate decreased slightly from zone I to zone III. It ranged from 25.36% (III) to 28.08% (I) and, to a limited extent, exceeded the defoliation threat threshold, which for pine is 25%. Comparable standard deviations were recorded: from 7.5 to 9.9%, while the coefficients of variation ranged from 28.9 to 37.1%. Extreme values of variation would be further reduced if the outlier defoliation rate of 75% from zone II was disregarded.

The analysis of variance ANOVA was conducted to investigate the significance of differences between mean defoliation rate of trees growing at different distances from the asphalt road (in three transects). The null hypothesis assumed that the average defoliation of trees growing on three transects (differing in their distance from the road) are identical (i.e., do not differ statistically significantly). One of the basic assumptions for ANOVA tests is the normality of distribution for variables in each population (group).

Table 6. The results of Kruskal–Wallis test.

Dependent Defoliation [%]	ANOVA of Kruskal–Wallis Rank; Defoliation, Independent (Grouping) Variable: T1 Kruskal–Wallis Test: H (2, N = 182) = 5.452065 p = 0.0655
	Code	N Important	Sum Rank	Avarge Rank
TRANSECT I	1	73	7452.00	102.08
TRANSECT II	2	53	4592.00	86.64
TRANSECT III	3	56	4609.00	82.30

presents results of the Kruskal–Wallis test.

The results of the Kruskal–Wallis test were statistically significant, for p = 0.0655 at the assumed significance level of α = 0.05.

The null hypothesis assumed that the average defoliation of trees growing on three transects (at different distances from the road) are identical (i.e., do not differ statistically significantly). This indicates that the asphalt road did not have a significant effect on defoliation of trees growing at different distances from this road.

Conclusions from the analysis of descriptive statistics are confirmed by the interpretation of the share of trees in individual damage classes (Figure 6). The highest share of pines was observed in damage class 1 (defoliation 11–25%): from 45 (zone I) to 61% (III); followed by 2a (26–40%): from 30 (III) to 45% (I). In the other damage classes, the recorded frequencies were below 10%.

4. Discussion

The effect of forest roads on the health and growth increments of trees as well as stand structure is an important problem for forest protection and management. Watkins et al. [9] investigated the distribution of shrub plants and their relationships with unpaved forest roads in the northern landscape of deciduous forest in the Chequamegon National Forest, Wisconsin (USA). Results indicate that roads alter conditions within forests and thus the species composition and numbers of plants; however, the effects are limited in terms of the depth of penetration in managed forests. Picchio et al. [11] investigated the effect of forest roads on communities of trees and stand structure in three Italian and Iranian temperate forests. As the results showed, edges (20 m) of the forest road networks constitute a fine mosaic composed of various trees, combined with a moderate presence of dead trees. A similar taxonomic composition of the forest community was observed in three areas, from road edges to the forest interior. The first major difference concerned the number of species with lesser shade tolerance. Another principal difference was related with higher indices of biodiversity for trees growing along roads. The main similarities concerned the structure of living and dead trees. Studies on the effect of the distance of forest roads on the biodiversity of the shrub layer in young and mature oak stands in a French lowland forest were conducted by Avon et al. [22]. All vascular species and bryophytes were collected at five distances from roadsides up to 100 m to the adjacent stand. Species composition was analyzed depending on the distance from roads and the age of stands. The species composition varied greatly between roadside habitats and those within forests. The response of plants to the distance from roads depended also on the age of the stand. Even if the depth at which the effect of forest roads measured in lowland stands was evident was narrow, the construction of a new forest road has a significant impact on the dynamics of plant populations.

For many years, studies on the effect of road material and technologies on the forest environment have been conducted at the Department of Forest Engineering, the Poznań University of Life Sciences (PULS). To date, the analyses covered the impact of industrial waste used in road construction (slag, ashes, mining tailings) as well as cement binders supplemented with waste from power engineering facilities [15]. Most results indicated no chemical contamination from these materials used in road construction [23]. The commissioned R&D task was to indicate whether pavements constructed using asphalt binders also have no significant effect on the health and increments of trees, in this case pines.

The health condition was determined based on the defoliation of pines using the bioindicator method. The degree of crown damage in Scots pines was assessed using three criteria: defoliation estimated accurate to 5%, the number of needle generations (in the apical part of the crown) and crown opening [20]. Based on the conducted inventory, mean defoliation rates and the percentage shares of trees in individual damage classes were identified. The conducted ANOVA and the Kruskal–Wallis test positively verified the zero hypothesis, assuming that the mean annual diameter increments of trees growing in the three zones (different distances from roads) are identical (do not differ statistically significantly). This indicates that the asphalt road had no significant effect on mean annual DBH increments of pines growing at different distances from this road.

The course of change with age of the radial increment of trees is characterized by high values and their rapid decline in the initial (juvenile) period of the trees’ life, followed by lower values, which slowly decline to reach a course parallel to the age axis at the end of the trees’ life [24,25]. This trend is also observed for the radial increment of trees aged 58 years growing in the three analyzed zones of distance from the asphalt road for the last 21 years. For the analyzed 21 years, the average growth of trees in zone I (closest to the road) was once (2007) greater than in zone II and twice (2016 and 2017) greater than trees from zone III. However, the statistical analysis carried out indicates that there are no significant statistical differences between the increments recorded for each zone. In the case of strip roads, it is said that they affect both the entire stand and individual trees growing directly along the roads (edge trees). This is especially the case when strip roads are created, and significantly affects the reduction of production potential. However, this effect loses significance over time [26]. Marginal trees have an increased supply of light to their crowns and thus a larger living space. This results in increased growth in most cases [27,28,29,30,31]. However, this effect does not always occur [32,33]. The plots analyzed in this work were devoid of edge trees, and possible increased growth in them was not taken into account. Some of the authors of the works mentioned above point to the effect of the road on the species composition of plant communities. The analyzed stand was a solid pine stand without undergrowth and the asphalt road in this case had no statistically significant effect on a number of pine stand parameters. The limited leaching process of chemical pollutants from monolithic road structures is confirmed by other studies [34]. Even surfaces containing ashes are not susceptible to leaching processes [35]. A more important source of trace elements are dusts resulting from road use [36,37].

5. Conclusions

Within the R&D project entitled “Environmental aspects of reconstruction and construction of forest roads using selected asphalt binders”, in the first stage of the study, the required laboratory analyses and field observations were performed to assess the potential impact of mineral–asphalt mixtures on the roadside habitat. The study investigated the effect of the asphalt road segment located in the Przymuszewo Forest District on the increment and health of trees in a 60-year-old pine stand. Based on the analyses the research hypotheses were verified. Laboratory analyses of soils and wood collected from various distances from the roadway showed a lack of a negative environmental effects of roads. Contents of all detected elements in wood were lower than their contents in soil, which indicates their markedly limited phytoextraction, as well as that the highest efficiency of this process is found in plants growing in the nearest vicinity of the road. This was confirmed in the course of dendrological and dendrometric studies using statistical analyses. It was shown that the analyzed asphalt road had no significant effect on mean annual DBH increments and crown defoliation of pines growing at different distances from the road. This means that forest road segments with very high traffic intensity may be constructed using asphalt binders with no significant negative consequences for pine stands. In the context of the conducted studies, it is rational to search for new road technologies, including those based on asphalt binders. For this reason, in view of the importance of this problem in the natural and social context, these studies will be continued, and their scope and area of the study will be expanded.