The Short-Term Effects of Heavy Thinning on Selected Soil Carbon Pools and Microbial Activity in a Young Aleppo Pine Forest

Aleppo pine (Pinus halepensis Mill.) is the most common pine species in the western Mediterranean basin [1], and is well adapted to climatic constraints, particularly water scarcity and high seasonal temperature variation [2]. In fact, during the 20th century, it was the most widely used species in the reforestation of semi-arid areas, with a total area of approximately 2 million hectares in Spain (i.e., 11.37% of the forest area) [3]. Despite the good climatic adaptation of this species, the reforestation of Aleppo pine has been controversial [4]. The high temperatures and low humidity typical of the semi-arid climate, together with the presence of sometimes dense fuel stands, make this species highly vulnerable to fire [5]. A further handicap is that post-fire regeneration of this species is often hyper-dense at a young age, which usually leads to competition between trees for light and soil resources [3,6,7].

Thinning is a widely used forest management technique that reduces stand density and thus competition for natural resources [8]. Recently, the effects of thinning on tree and shrub growth, biodiversity, carbon stocks, hydrological processes, soil physicochemical properties, soil microbial biomass and community structure, soil enzyme activities, and soil microclimate have been the subject of several comprehensive reviews and meta-analyses [9,10,11,12,13,14,15]. Among the results of these studies are that thinning can promote individual tree growth and alleviate drought stress by increasing the soil water availability to the remaining trees and by developing more extensive individual root systems over time, particularly in high-density stands that tend to have less developed root systems [16,17,18,19,20,21]. In fact, good thinning management can increase shrub and herb diversity [10,15,22]. These positive effects could be explained, at least in part, by the increase in soil temperature and soil moisture following moderate or heavy thinning [9,15,23]. However, some studies have also reported negative aspects of thinning. For example, the thinning effect of drought mitigation tends to decrease over time [24]. Yang et al. [25] found, in a meta-analysis, that forest thinning increased soil CO₂ and N₂O emissions and decreased CH₄ uptake under certain climate conditions. In addition, the effects of thinning may vary depending on its intensity and the time elapsed since the treatment. Comparing the effects of high-density thinning with other thinning treatments or unthinned stands allows researchers and forest managers to assess the impacts of different thinning intensities on various aspects of forest ecology. Zhang et al. [14] suggested a moderate thinning treatment (amount of tree removal: 30%–60% intensity) for soil nutrient conservation benefits. However, for the specific case of P. halepensis in water-limited habitats, Calev et al. [17] indicate that high-intensity thinning treatments (≥50% basal area removal) are effective for treating excessively dense mature (>40 years) stands of P. halepensis under drought stress.

Forests play a key role in the carbon cycle, and represent the largest terrestrial carbon reservoir on Earth, with about 30%–70% of organic C stored below ground, making soil the largest terrestrial pool [26,27]. Soil organic carbon (SOC) is complex in terms of its composition and physical structure and, together with its fractions, is the subject of active research [9,28,29,30,31]. Forest thinning has a major impact on SOC pools, given the changes that occur in the soil microenvironment, organic matter inputs and microbial metabolism [13]. However, the heterogeneity of the thinning intensity, recovery stage and microclimatic conditions increases the uncertainty about the effects of thinning on the SOC and other carbon pools. On the positive side, changes in the SOC after thinning may be related to increased space and light for understorey growth, increased activity of soil microorganisms or residual tree roots, or decomposition of debris left on the forest floor or incorporated into the soil [9,11,15,32]. In fact, increases in SOC following thinning have been reported in several case studies. For example, Ma et al. [29] found that SOC increased in the short term after moderate thinning of Larix principis-rupprechtii plantations, suggesting that this could be due to changes in the labile carbon pool, as well as improved environmental conditions for microorganisms to decompose organic residues. Gong et al. [33] observed that moderate thinning increased SOC stocks more than other thinning intensities with significant differences after five years of recovery. However, some authors reported that soil organic carbon did not change after thinning [11,14,34]. Controversially, Yang et al. [35] stated that heavy thinning reduced SOC nine years after treatment, whereas low and moderate thinning did not change it.

SOC is usually divided into two main fractions: active C and stable C. Among the active SOC fractions are microbial biomass carbon (MBC) and water-soluble organic carbon (WSOC), which are readily mineralized and may reflect management practices, making them an indicator for assessing the quality of SOC pools [36]. WSOC is considered to be the most mobile and reactive organic carbon fraction and the most important carbon source for soil microorganisms [37]. Ma et al. [29] found that WSOC was mainly derived from partially decomposed plant litter, reflecting short-term nutrient storage and acting as a substrate for soil microbial activity. Thus, WSOC content may also vary with stand thinning. However, Chen et al. [38] did not observe differences in WSOC among different thinning treatments (low intensity thinning, high intensity thinning and control) seven years after thinning. It has been reported that the effect of thinning on MBC may also be influenced by thinning intensity, with some studies reporting that high-intensity thinning was the only variable that increased MBC [38,39]. Zhou et al. [11] found that thinning changed the microbial community structure, but not the total microbial biomass, suggesting that microorganisms adapt to thinning by changing the microbial community structure rather than by changing the microbial biomass. Recently, Zhang et al. [15] reported that thinning had positive effects on SOC, dissolved organic carbon and MBC, especially in the late stage (>6 years). In general, soil microbial respiration (BSR) can provide an estimate of soil microbial activity. It has been reported that, after thinning, soil respiration increases [14,40], decreases [41], or no change is observed [34]. Yang et al. [42] reported that these inconsistent results may be related to the fact that heterotrophic respiration and autotrophic respiration respond independently to thinning. Zhang et al. [9] reported that light and moderate thinning increased soil heterotrophic respiration in the early recovery phase (≤2 years after thinning) while heavy thinning had no significant effect. Microorganisms secrete soil enzymes (e.g., β-glucosidase, AP or UA) that promote C and N assimilation by plants. Variable results have also been reported for enzyme activity after thinning. For example, Zhou et al. [43] found that thinning inhibited or had no effect on C-degrading enzymes, but stimulated N- and P-degrading enzymes. In related research, Lull et al. [44] explain the relationship between enzyme activity, climate, and soil properties, and note that results may vary with thinning intensity and soil texture. In line with this, Zeng et al. [45] found that the effects of thinning on the activity of extracellular soil enzymes vary with time during the recovery of the forest after treatment. Therefore, the effect of thinning on carbon pools and microbiological activity can vary depending on different aspects such as climate, forest ecosystem (i.e., dominant species, stand age, etc.), thinning intensity, time elapsed since treatment, and soil properties, among others [9,15].

A very dense pine forest, regenerated after a fire in 1992, was heavily thinned in October 2012. The research of this study is part of a global project to evaluate the effects of adaptive forest management on growth dynamics, water fluxes and soil variables in an Aleppo pine regeneration forest [12,23,34,46,49]. A variety of soil physical (GWC), chemical (SOC and WSOC), and biological (microbial and enzymatic) parameters were assessed at different sampling dates during a five-year period after thinning. In this context, the main objective of this study was to determine the effects of thinning applied (i.e., severe thinning of young pine growing in a high density population) at soil level in the short term. Coarse woody debris was retained in situ to prevent soil erosion and improve soil organic matter. Intense thinning can lead to a rapid decline in aboveground carbon stocks, and belowground carbon pools may exhibit variable responses, and this can significantly influence soil carbon pools and microbial enzyme activity with more pronounced short-term effects. In a first approach, three blocks each of treated and control plots were planned for the study. However, after analysing the samples collected, it was decided that a better approach would be to divide the study areas into zones according to soil properties. The differences in soil properties are related to the tomography presented by del Campo et al. [49], since a rock mass present in the upper zone of both experimental plots limits soil depth, which means that zI has less soil depth and higher stoniness compared to zII (Table S1). A significant difference was the higher level of SOC in zI, although forest cover and forest management were similar in both zones (zI and zII). This could be due to accumulation of stable organic matter in this shallow zI. This could explain that the decomposition of organic matter by microorganisms and its translocation to the deep soil is limited, and therefore in zI there was a higher organic carbon content. Note that the boundary between these two zones is not well defined (Figure 1c), and that the heterogeneity of the terrain makes it impossible to establish a clear boundary between them.

SOC has shown to have positive effects on the mechanical properties of the soil, improving its strength, bulk density and porosity, which favours infiltration, drainage and its storage capacity [59,60]. However, this is not the case at the forest floor level, which may be understandable given its higher water holding capacity and greater exposure to climatic variations [61,62]. For the period studied, no differences were found in the GWC of forest floor and mineral soil between the treated and control plots in both study zones. On average, zI had significantly higher GWC_ms (about 1.4-fold) than zII, which is justified by the higher organic matter content observed in zI (Figure 4, Table S1). The low precipitation in the first year after thinning could explain partly the small differences in soil moisture in the treated plot compared to the control. The lower GWC content until July 2014 is explained by the lack of rainfall. In fact, during the early recovery period of 1–3 years after thinning there was a drought episode that slowed down and reduced soil changes. The results for GWC in mineral soils were also confirmed in laboratory column tests to determine WHC (Table S2).

Soil carbon fractions are directly related to organic matter and microbial activity [63,64]. In a previous study, we found no direct improvement in SOC and WSOC content in a pine forest in the long term (eleven years after thinning) under a clay-loam soil texture by moderate thinning, but did under the same conditions under a low fertility sandy soil [44]. As explained above, the effects of thinning depend on multiple factors, and SOC and WSOC content can be improved or worsened depending on aspects such as the intensity of thinning, species, age, physicochemical properties of soils, microorganisms, etc. [33,34,65,66]. The significantly higher content (about 2- to 3-fold) of SOC_ms in zI compared to zII (Figure 4) is explained by the peculiarity of the terrain, which was not affected by the silvicultural treatments applied, nor were significant differences observed between treatment and control. In fact, the accumulation of organic matter in a smaller volume of soil justifies the higher zI values obtained. These results are in line with what was observed for the content of WSOC_ms, which was about 3.8-fold significantly higher in zI; on the contrary, WSOC_ff presented about 6.7-fold significantly higher values in zII. Higher WSOC values are usually correlated with greater soil moisture and higher SOC levels [44]. In addition, the autumn–winter of 2016 was wetter, with this being a possible explanation for the significantly higher WSOC_ms values observed in June 2017. No significant differences between treatment and control were observed for WSOC_ff. However, different studies report that heavy thinning, especially in the short term (i.e., the first 5 years), reduces litterfall and consequently the different organic carbon fractions [66,67]. The reduction in litterfall may be compensated by the organic debris left with the thinning treatment. As indicated by del Río et al. [68], heavy thinning results in a loss of volume yield, but the extent depends on location, site and stand age. No significant differences between T and C were found for MBC in zII, and those found in zI were not consistent. However, Kim et al. [69] observed higher MBC 7 years after an intermediate and a heavy thinning treatment, which was associated with the presence of higher amounts of residue in the soil. Comparing different plots, they reported difficulties in interpreting the relationship between the amount of thinning residues and the site-specific effect of thinning due to the high heterogeneity observed. MBC also confirmed previous results on labile carbon-fractions, showing significant differences of about 1.4-fold greater in zI than in zII (Figure 6). Soil organic carbon acts as a substrate and energy source for microbial biomass growth and activity, with greater differences observed for SOC in zI between the treated and control plots. In fact, soil microbial biomass is mainly found in organic matter and is essential for decomposition and formation of the soil carbon pool, which is used as an indicator of soil quality [70,71]. These results were correlated by Spearman (Figure 9), which showed a clear correlation in the SOC content (r = 0.85) when the data were examined globally between treatment and control, rather than when they were examined comparing zI and zII (r = −0.041). It should be noted that, although it was not possible to establish a clear relationship between the sampled data and climate based on the punctual measurements made, it can be observed that the main component of the abscissa axis in the PCAs (Figure 9) was mainly conditioned by soil and ambient temperatures at the negative end and by soil humidity at the positive end, indicating the close relationship between climate and microbial activity. Zhang et al. [14] explain the direct effects of increasing soil temperature and microbial activity by increasing thinning intensity. This is also explained by changes in microbial communities associated with climate change [72,73]. The ordinate axis of the PCA seems to be related to the treatment, with all the biological variables possibly influenced by the organic matter content, observing that zI seems to be more dependent on temperature than zII, which could be explained by the fact that it is in the high area of the mountain slope and receives greater solar radiation due to the lower slope and NW orientation. BSR is another quality indicator related to soil respiration and corresponds to the CO₂ released by microbial mineralisation of soil organic matter [34]. As a result, BSR measured 4 and 5 years after thinning did not show significant differences between the two zones or between the two treatments due to the high variability observed in the samples although, on average, zI had a higher respiration rate than zone II. In fact, BSR is directly related to organic matter content and influenced by the size and activity of the microbial biomass [74,75]. Mainly in zI it is observed that respiration is higher when the MBC is higher.

Two extracellular enzymes related to phosphorus and nitrogen cycles were selected in this study (i.e., AP and UA, respectively), which are also established as indicators of soil quality [76,77]. No significant differences were found between treatment and control for either enzyme activity, indicating no significant changes due to thinning. However, the spatial heterogeneity of both enzymes but mainly that of UA should be taken into account. Significantly, about 2.5-fold higher levels of AP were produced in zI than in zII (Figure 8), associated with higher SOC content. Slightly higher levels of UA were also observed in zI, but not significantly, and this enzymatic activity seemed to be independent of soil and treatment, perhaps because pine debris is a very poor nitrogen material. These results support the importance of the quality of organic matter in the microbial activity [78,79]. A high variability between sampling dates was observed indicating a high activity dependence with climate [44]. In fact, enzyme activity is extremely dependent on temperature, humidity, pH, substrate availability and other related soil chemical properties, as shown in several studies [80,81,82].