1. Introduction
Climate change is one of today’s most important global environmental issues, which directly affects all natural ecosystems and socio-economic systems [
1]. One of the main strategies to mitigate climate change is based on the soil functioning as a carbon (C) sink [
2]. Consequently, the number of studies on soil organic carbon (SOC) changes in different ecosystems is increasing [
3,
4,
5,
6,
7].
Nowadays, forests are covering about 30% of the terrestrial landscape in the world [
8], and they are the main pool for terrestrial C stock, where the main C storage occurs in the forest biomass and soils [
9]. In the forest ecosystem, C storage occurs in the tree aerial biomass—stems, branches and foliage—ground vegetation [
10], tree-root biomass, forest floor, and soil [
11]. In the forest ecosystem, the soil C stock is important given that it is in dynamic equilibrium with the vegetation [
12]. Furthermore, in terrestrial environments, soil organic matter (SOM) is the largest organic C pool; the quantity and quality of organic C in soils can be affected by vegetation through influencing the inputs and outputs of SOM [
13]. Through differences in litter quality, plant species have an influence on the storage and dynamics of C in soils, as reported in several previous studies [
14].
Thus, forest management is accepted as a tool for removing atmospheric CO
2 [
15]. For this reason, some authors [
16] have studied whether unmanaged forests have higher contributions to climate change mitigation than managed forests because of their potential of C sequestration in living biomass and soil, which is the role of the old-growth forests as global C sinks as pointed out by some studies [
17]. Some publications also show how the management of the forest soils for agricultural purposes changes the soil C equilibrium [
18]. Thus, given that changes in land use and fire events cause the loss of C, which is emitted to the atmosphere [
19], it is important to know the C stock in different pools and if this C is allocated into labile or stable soil C components.
The storage of C in soils is mainly in the form of SOM [
20], for which humic substances comprise a major proportion [
21]. Furthermore, given the different chemical properties of humic substances, several studies have focused on developing processes for the separation of their fractions [
22]. On the basis of the different solubility of acids in the components of the organic matter, labile and recalcitrant pools, which have different turnover times, can be separated [
23]. Humic substances have a high C content and long turnover times; that is, a long time is needed to form them. These substances can be fulvic acid, humic acid, and humin. Fulvic acids are the most abundant humic substances in the soils, given their solubility [
24], and this property is important in order to know more about their environmental behaviour [
25]. Humin represents the most recalcitrant fraction of organic matter.
For this reason, estimating, quantitatively, the influence of tree species on C stored in forest soils is one of the issues that is receiving a significant amount of attention, but how to optimise C sequestration and how vegetation affects the stability of SOC are less understood. However, recently, some investigations have been conducted in order to analyse the influence of tree species on the vertical distribution of the labile and recalcitrant C [
13]; some of these studies were conducted on Mediterranean forests [
26] and in semi-arid environments [
27].
In order to mitigate global warming, C sequestration has been promoted by implementing effective measures for the conservation of old forest ecosystems and new plantations [
28]. A particular case is for arid and semi-arid ecosystems, for which the impacts of climate and atmospheric CO
2 changes are still unclear [
29]. For these reasons,
thurifera forests, which are usually old and semi-arid ecosystems, can play an important role in C sequestration.
J. thurifera, a tree of the Cupressaceae family, is found mainly in Morocco, Algeria, Spain, and certain regions of France and Italy. The forests of these trees are important from different points of view: ecological, socioeconomic, floristic, and cultural perspectives [
30]. Given this interest, several studies have been published on this millenarium tree [
31,
32,
33,
34,
35,
36], the most recent works by García-Morote et al. [
37] and Lahouel et al. [
38]. However, semi-arid Mediterranean soils are still not studied enough, in particular,
Juniperus woodlands [
37].
In this work, the influence of J. thurifera on the storage and stabilisation of organic C in soil was analysed. For this purpose, the variability of the organic C concentration in soil samples was studied considering the effect of several factors, such as the sex of the tree, the diameter of the tree trunk, and the canopy effect. For the analysis of the stability of the organic matter, the distribution of the recalcitrant organic matter fractions in the soil profile was obtained. Finally, a better quantitative and qualitative knowledge of the SOC in a thuriferous juniper semi-arid forest has been obtained, which should be useful for evaluating the role of this forest soil in C sequestration.
2. Materials and Methods
2.1. Site Description
The study area, the “Enebral forest” (
Figure 1), is located at Sierra de Cabrejas in Soria Province (central Spain), and is predominantly covered by
J. thurifera, partially mixed with
Juniperus communis L.,
Quercus ilex subsp.
ballota (Desf). Samp., pines (
Pinus sylvestris L. and
Pinus pinaster Aiton), and with scarce vegetation of
Lavandula latifolia, Medicus, and
Thymus zygis L. The main characteristics of the study site are shown in
Table 1. The site has a continental climate, with dry summers (June to September) and cold winters. The soils of this forest have been classified as Lithic Eutrudepts and Lithic Xerorthents according to the United States Department of Agriculture (USDA) [
39] classification. The forest has a density of 209 trees/ha, and the average distance between trees is 12 m. A forest inventory was performed in circular plots of 5000 m
2, for which all trees with a diameter at breast height (DBH) of greater than 7 cm were counted.
2.2. Soil Sampling
In order to take soil samples, a total number of 40 trees were selected according to their sex and their trunk diameter. According to the diameter (d) of the trunk, four groups of trees were taken into account, described as (A) trees with trunk diameters of between 7 and 10 cm, (B) trees with trunk diameters of between 10 and 20 cm, (C) trees with trunk diameters of between 20 and 30 cm, and (D) trees with trunk diameters of greater than 30 cm. An equal number of trees of either sex in each group were chosen: 10 trees in each group and, in total, 20 male and 20 female trees. The sampling was performed in summer-time.
Four soil samples were taken of up to 30 cm in depth, when it was possible, around each tree: two close to the trunk, and two far from the trunk, taken in both directions, north (N) and south (S). The sampling points close to the trunk and under the influence of the tree canopy and the trunk were denoted by N-I and S-I, and the others far from the trunk and without the influence of the canopy or the tree were denoted by N-II and S-II. Finally, a total of 160 soil samples (40 trees × 4 samples per tree) were collected, as cores of 3 kg for each sample.
Additionally, and in order to characterise SOM, an analysis of the humus fractions was performed. For this purpose, two kinds of soil profile were selected. One of these is denoted by Profile 1, which consists of the sampling profiles of those soils located close to the largest tree trunks, and the other is Profile 2, which corresponds to the profiles in the sampling points located far from the smallest tree trunks. In each profile, soil samples were taken at different depths, given the characteristics of the soils in each case. The ranges of the depths for Profile 1 were 0–10 cm, 10–20 cm, and 20–30 cm, and the ranges for Profile 2 were 0–7 cm, 7–12 cm, and 12–15 cm.
2.3. Soil Preparation and Analyses
All of the soil samples were dried (35–40 °C) and sieved through a 2 mm sieve. The fraction greater than 2 mm corresponded to the gravel content, which was determined gravimetrically and is shown in
Table 2 as the percentage relative to the total soil sample. The fraction of soil smaller than 2 mm was used for determining the percentage of granulometric fractions and selected soil properties.
Silt, sand, and clay percentages were determined by a discontinuous sedimentation process based on the densimeter method. The pH was measured potentiometrically using a solution of soil/water (1:2.5
w/
v). The total organic carbon (TOC) was determined following the method of Walkley and Black [
40]. The percentage of CO
3−2 content was measured according to Nelson [
41], using a calcimeter. Total nitrogen was determined with a Heraeus V-Analyser, model Macro N, (Labexchange, Burladingen, Germany).
Additionally, two parameters,
TOCint and
TOCext, associated to the positions close and far to the trunk, respectively, were calculated as follows:
and
where
TOCN(I) and
TOCS(I) were the TOC in the north and the south in the proximity of the trunk, and
TOCN(II) and
TOCS(II) were the TOC referring to the north and south far from the trunk.
2.4. Soil Organic Carbon Fractions
Humic substances can be operationally classified into three separate fractions defined in terms of their solubility properties [
42]. Laboratory procedures attempt to divide the fractions into the subfractions of humic acids, fulvic acids and humins. Humic acids are extrated from soils with alkaline solutions and turn into insoluble precipitates after acidification. Fulvic acids are yellowish in colour and are soluble in both acid and alkaline solutions. Humins are insoluble in acid or alkaline solutions.
For the analysis of the humus, 50 g of soil of each sample, previously sieved, were used for the isolation of their fractions.
Figure 2 describes the procedure proposed by Duchaufour and Jacquin [
42] that was followed. The first step involved a physical separation of the soil with a mix of bromophenol and ethanol (density of 1.6 g/mL), for which the floating particle fraction was denoted by the soil light fraction (SLF) and the rest was denoted by the soil heavy fraction (SHF), which was used for the isolation of the humus fractions following a chemical process. The chemical process began with the separation of the soil soluble fraction (SSF) using deionised water at 20 °C. After this, the insoluble fraction was dissolved in a solution of Na
4P
2O
7 (0.1 N) in Na
2SO
4 (5%) fixed at pH 7.0 and was centrifuged at 5000 rpm. The extract was taken as the total humic fraction (THF-1). For one side, the liquid was acidified down to pH 2.0 and centrifugated at 5000 rpm, to obtain a yellowish supernatant solution containing the fulvic acids fraction (FA1) and a precipitate of humic acids (HA1). Residue 1 was extracted with a mixture of Na
4P
2O
7 (0.1 N) and NaOH (0.1 N) regulated at pH 12.0; from this extraction, a new THF (THF-2) was isolated and separated from the THF. The total humic extract was acidified with HCl down to pH 2.0, and was then centrifuged. Again, a supernatant liquid containing the fulvic acids fraction (FA2) and a fraction corresponding to humic acids (HA2) was obtained. The organic C in the fractions was measured using a total organic carbon analyser (Leco CHN-2000, St. Joseph, MI, USA).
In order to estimate the grade of stabilisation considering different organic fractions, some parameters can be defined and calculated. The ratio between the total humic acid (
HA) and total fulvic acid (
FA), denoted by (
HA/
FA), is the index of polymerisation, which allows one to study the grade of stabilisation of the organic matter [
43]. Additionally, the degree of humification (
DH) is used in this work as a tool for comparing the composition of organic matter in different horizons. The
DH parameter is defined in different ways in different studies, but in the present work the following expression [
43] was used:
2.5. Statistical Analysis
The statistical method used was an analysis of variance (ANOVA), to determine the existence or absence of significant differences within two and four groups of experimental results, which, in this work, were defined by several properties. The analysis of the influence of several vegetation factors (sex, orientation, diameter of the trunk, and proximity to the trunk) on physical and chemical characteristics, as well as organic C, was performed through one-way ANOVA. The main effects were considered to be significant at the p < 0.05 level. The mean separation was conducted using the Fisher’s least significant difference (LSD) method. The data reported in the tables are the mean values and the precision, expressed as the standard deviation. The software SPSS 15.0 (SPSS Inc., Chicago, IL, USA) was used for the statistical analyses.
4. Discussion
From the analysis that has been carried out, we can conclude that there is a clear influence of the tree on the accumulation of C in soil. The thickness of the trunk and the protection of the canopy are factors to take into account in order to determine the variability of the concentration and composition of the humus, but neither the sex of the tree nor the north–south orientation has an influence. Quantitatively, higher contents of C exist under the canopy of the tree, as well as in the proximity of larger trees. The litter on the soil is the main reason for this, given that it increased with the size of the trunk, was non-existent far from the trunk and out from under the canopy, and was also almost negligible in the proximity of the smallest trees. The litter contributes fresh organic matter but with a protection that improves the humification processes in the soil. Thus, older trees have a higher organic C concentration.
The analysis of the vertical distribution of the TOC shows that the upper layer has a higher concentration and that it decreases with depth. The gradients of decreasing TOC along Profile 1 and Profile 2 were 0.04% and 0.05% TOC/cm, respectively, which means that there is an influence of the tree in the depth distribution of the organic C. The analysis of the C/N along the profiles shows that the mineralisation process is slower in soils under the protection of the canopy, and, by the analysis of the litter, is slower than in soils without the influence of the tree. Thus, both analyses suggest the importance of the vegetation in the soil processes. Soils under the influence of the trees, close to the trunk and under the canopy, and particularly for trees with a larger trunk, present higher concentrations of organic C, and its variation with depth is less than for soils far from the influence of trees. The value of the TOC from soil samples taken close to group D trees (TOC
int = 3.45%) was found to be similar to that reported for a similar forest by González and Candás [
36], which was 3.760%.
Apart from the mineralisation process, the other organic matter process is humification, and this takes place in a different way at different depths. The vertical distribution of the C fractions shows the results of this humification process. Qualitatively, the organic matter in the samples taken from sites close to the trunk and under the canopy had a higher stability, given that the humin fraction was higher in these sites, at any depth. The humin fraction is associated to recalcitrant C, given that its compounds have a slow degradation and a high resistance to biodegradability, and, thus, it contributes to long-term C storage in soils [
13]. Other parameters, such as the total content of fulvic and humic acids, support the hypothesis of the influence of the vegetation on the stabilisation of the organic matter in the soils.
Comparing the proportion of the different fractions with respect to the TOC along both profiles, the effect of the vegetation on the humification process can be discussed. The most recalcitrant fraction (humin) had no variation with depth for the soils under the influence of the tree, but in the absence of this influence, this fraction increased with depth, as mineralisation processes are important in the upper level, but less so in the deeper layers in such unforested soils. The opposite occurred, in the soil profiles under study, for the SSF, which was negligible, but this fraction was the labile C composed of compounds, such as soluble sugars and other carbohydrates, which have been shown to play a dominant role in the evolution of CO
2 from soil as a result of preferential decomposition and rapid turnover [
51]. Other fractions of less-humificated C (SLF) were higher in the upper layers in both profiles, where the competition with mineralisation processes in unforested soils and the influence of the litter in the soils close to the vegetation have been taken into account. Additionally, the SLF decreased very sharply in Profile 2, and the absence of litter could explain this fact. Litter is not only a source of C, it also has a buffer effect against extreme temperatures, as well as a protective function against water and air erosion [
32].
A simple comparison with other forests can be performed.
Table 7 shows the results obtained in the present study with those of a previous work, for which the same methodology was followed [
52]. The data of the
thurifera forests correspond to average values obtained from Profile 1. All the values are given as the percentage of C with respect to 100 g of soil and also refer to the TOC (in parentheses) for each fraction. Analysing the table, it is clear than the organic C level is higher in oak forests, and
thurifera forests have a lower C content. Additionally,
Table 7 shows that the organic C in the different fractions is lower in the
thurifera forest than in oak and pine forests, given that the first soils are particularly poor in organic matter, as is usual in semi-arid soils.
An interesting point for analysis is the capacity of the different forests for the stabilisation of the organic C. For this purpose, the data referring to the TOC are very useful. These data show that the proportion of the humin (with respect to the TOC) in the
thurifera forest was considerably higher than for the other two forests (pine and oak): 64% versus 34% and 39%. This fact confirms that the soil in
J. thurifera forests presents a higher proportion of organic matter resistant to degradation and recalcitrant organic C than in oak and pine forests. Furthermore, the contents of the total fulvic and humic acids in the
J. thurifera forest were lower than those found in pine and oak forests (see
Table 7). However, if humic and fulvic acids are compared with humin, a proportion of humin 3.6 times higher than for acids is found in the
thurifera soils, versus the estimated 0.7 and 0.8 for pine and oak forests, respectively. Thus, taking into account that humin is organic matter that is more humificated than the fulvic and humic acids, the stabilisation of the organic matter in the oak and pine forests was lower than that for the
J. thurifera forest. Thus, the
thurifera forest could have a better capacity for the stabilisation of its low SOC than the pine and oak forests presented here. This fact shows that the SOM of
J. thurifera forest has a reduced biodegradability if it is compared with pine and oak forests, and this characteristic can be explained in terms of the vegetation and the type of microorganisms responsible for the humification process, as well as by the age of the forest and its history.
5. Conclusions
We can conclude that the studied
J. thurifera forest does not have a large amount of C if it is compared with the other forests mentioned in this work; however, its organic matter is more stable than those for the oak and pine forests of a previous work [
52]. The loss of C in the soil is due to degradation processes and it is increasing with global warming, as was pointed out in other works [
5]. The soil can be considered a large C sink of the forest ecosystem, according to one author [
53], but we conclude that this fact depends on the amount of organic C storage and its grade of resistance to biodegradability. The efficiency of a C sink is given by the balance between its capacity for C sequestration and for the conservation and the reduction of its natural loss. Consequently, in agreement with De Deyn et al. [
54], the efficiency of the C sequestration by the soil depends strongly on the age of the forest, as is the case for the semi-arid soil of the studied
J. thurifera forest. Thus, the
J. thurifera forest soils should be regarded as an active C sink, and indeed, an ideal C sink, given their organic matter stability.
The information obtained in the present work may be very useful for studies on optimising land management in Spanish mountain ecosystems, if an environmental perspective is adopted. The study reveals that, for ecosystems such as that presented here, vegetation destruction or inadequate substitution can change the equilibrium of the SOM and produce its irreversible degradation. This is a strong reason to support the protection of natural forests.