Using a modelling approach, we could assess how thinning intensity and frequency affect the provision of various ecosystem services under two contrasted climatic scenarios. This kind of analysis can be useful in adapting forest management and planning to climate change, where the expected provision of ecosystem services can be integrated into the decision-making process. Overall, we observed that a thinning regime with light intensities and long frequencies provided an overall higher production of mushrooms, carbon, and timber, and ensured more dead trees to serve as habitat for species. However, a thinning regime with heavy and frequent harvests favoured a higher diameter growth and a higher provision of blue water, although all the other ecosystem services were at lower rates.
4.1. Effect of thinning regimes on forest dynamics
One of the traditional objectives of thinning is to change forest structure, so that the productive potential of the stand is concentrated on a smaller number of trees, but with a larger diameter, and therefore a higher market value. Our model correctly reproduced the positive effect of thinning intensity on diameter growth, as proven in many previous observational studies (e.g., del Río et al., 2008; Marchi et al., 2018; Montero et al., 2001; Ricardo Ruiz-Peinado et al., 2017; Varmola & Salminen, 2004). On the other hand, the effect of thinning intensity on basal area or volume is not so clear. Ameztegui et al. (2017) and del Río et al. (2008) showed that low-intensity thinnings did not decrease significantly the final basal area in comparison with a non-intervention plot. Yet, excessively heavy thinnings may lead to a loss in forest stock when the growth of the remaining trees cannot compensate for the loss of the extracted ones, as we observed here. As for the frequency, it is well known that the effects of thinnings are transient and disappear around 20 years after the intervention (Ameztegui et al., 2017). Our results show that the final productivity of the stands only matches that of undisturbed plots when frequencies are long. Low final densities can be convenient when the objective is the production of good quality, large-size wood (Montero et al., 2001). However, it should be noted that in our study, the thinning intensity was kept constant during the whole rotation, while Montero et al. (2001) recommend reducing intensity with stand age, particularly in the last thinning.
4.2. Effect of thinning regimes on ecosystem services provision
Among all the factors that influence mushroom yield, structure is the only one that can be modified, through silvicultural actions. Indeed, forest management can promote mushroom production (Bonet et al., 2012; Palahí et al., 2009). de-Miguel et al. (2014) have shown that basal area has a significant influence on mushroom production, while other studies showed greater dependence on forest age or dominant height (Martínez-Peña et al., 2012; Tomao et al., 2017). In our study, the production of mushrooms decreased when a high percentage of the basal area was removed, and the highest amount of mushrooms was produced when a non-intervention regime was simulated. Contrary to our results, de-Miguel et al. (2014) found that maintaining the current management of Catalan forests (removing 25% of annual forest growth) would lead to a decrease in the production of 5%, while in the absence of management, it dropped by 11% (Bonet et al., 2012). Although a decrease in canopy cover may mean more water available in the soil, an excessive decrease may foster evaporation rates, negatively affecting fungal production (Bonet et al., 2012). We also observed differences in mycological productivity between the two stands studied, agreeing with previous studies which have observed higher productions in Pinus sylvestris stands than in Pinus nigra (de-Miguel et al., 2014; Morán-Ordóñez et al., 2020)
Total timber production decreased with thinning frequency and, above all, with the intensity of the interventions, although the differences between intensities were reduced for longer frequencies, suggesting an interaction between both variables. In general, the highest total timber volume was obtained for weak intensities (15%). Pinus sylvestris showed greater adaptation to strong thinnings, as shown by the fact that moderately intense regimes resulted equally or more productive than control plots, while for Pinus nigra moderate thinnings always produced lower wood volume than unthinned ones. The effects of thinning intensity on wood production are widely known (e.g., del Río et al., 2008, 2017; Zanchi et al., 2014), while the effect of frequency has not been as well studied. Varmola & Salminen (2004) showed the importance of early thinning in optimizing timber production, especially of large dimensions. Although thinnings decrease the biomass at the stand level, they increase diameter growth (Cabon et al., 2018), thus increasing the economic value of the forest (Baumgras et al.,1989). Despite total yields can be similar, the larger tree size obtained in the more intensive treatments can lead to higher economic returns, since larger trees usually imply more profitable destinations and lower unit extraction costs (Montero et al., 2001). On the other hand, faster growth can lead to lower wood density, which could have repercussions on wood quality (Jaakkola et al., 2005; Peltola et al., 2007; Russo et al., 2019).
Many studies have shown that one of the direct effects of forest management is the reduction of the on-site total carbon stock compared to a non-intervention regime (e.g., Duncker et al., 2012; Ruiz-Peinado et al., 2013, 2016; Zanchi et al., 2014). When the off-site carbon is considered, the carbon stocked by managed forests is however usually higher (Bravo-Oviedo et al., 2015). Ruiz-Peinado et al. (2016) found that the total carbon stock – including in-situ carbon and ex-situ carbon from removed trees – was not significantly different between an unthinned plot and a plot with light thinning regime. Yet in our study, when a thinning regime with light intensity and long frequencies was simulated, the level of in-situ carbon at the end of the simulated period reached the level of a non-intervention forest, while the total carbon stock was highest in the thinning regimes with light intensity. The differences between treatments would have been greater if we had considered the ex-situ carbon fixed by the wood products resulting from thinnings.
The positive effects of thinning on the provision of blue water are largely known and constitute the basis of the so-called eco-hydrological forest management (del Campo et al. 2017). According to our results, thinnings have the potential to revert the negative trend in blue water provision expected with climate change, but only when heavy and frequent thinnings are applied. More interestingly, the positive effect of heavy thinnings disappeared if more than 20 years elapse between thinnings, particularly for Pinus sylvestris under severe climate change, which matches previous studies on the duration of thinning effects (Ameztegui et al., 2017). This can be related to the higher adaptation of Pinus sylvestris to heavy thinnings that we mentioned above, but also to changes in the ratio between blue and green water when water becomes scarce (De Cáceres et al., 2021).
There is a lack of studies showing the effect of thinning on biodiversity in Europe and especially in the Mediterranean region. However, studies assessing the effect of forest management in general on biodiversity show that unmanaged forests provide better habitat for biodiversity (Torras & Saura, 2008), or that management practices that require less intensive interventions can provide good habitat for biodiversity (Duncker et al., 2012; Tomao et al., 2020). Accordingly, we found that low-intensity thinning regimes or control favoured a higher presence of standing dead trees, due to higher natural mortality by self-thinning. However, the lack of growth in the control plots led to the majority of standing dead trees being small in size, less favourable for hosting biodiversity (Blaser et al., 2013). Large-diameter trees are especially important for biodiversity as they host species such as bryophytes, lichens, fungi, and saproxylic beetles (Paillet et al., 2010). Our results also show that Pinus sylvestris stands produced more standing dead trees than Pinus nigra stands, as corresponds to the greater slope of their self-thinning line and already observed by Torras and Saura (2008) using data from the Third Spanish National Forest Inventory.
4.3. Trade-offs and synergies among multiple ecosystem services
When forest management is oriented towards multiple-use, not all ecosystem services can be equally maximized. In the regimes we have defined, mushroom, timber, and carbon production were positively correlated. Synergies between these services have also been shown by Morán-Ordóñez et al. (2020). Schwaiger et al. (2019) found that current carbon sequestration is negatively correlated with timber production, while total carbon sequestration is positively correlated with it. Although we found a weak correlation of biodiversity (expressed as large dead trees) with the other services, most of the studies have shown trade-offs between wood production and biodiversity (e.g., Duncker et al., 2012; Schwaiger et al., 2019; Sing et al., 2018), especially for high-intensity treatments. The provision of blue water was always negatively correlated with the other services. This result is especially relevant in water-oriented silviculture, as it makes it difficult to integrate blue water provision in multiple-use management decisions. Similar results were obtained by Ameztegui et al. (2017) and Morán-Ordóñez et al. (2020) for the Mediterranean region, while Duncker et al. (2012) found that forest management had a low influence on water quantity under continental conditions. Considering the trade-offs and synergies between ecosystem services, forest management at the landscape-scale should be designed considering the demands of different stakeholders in each location, or with a diversity of approaches to maximize the overall provision of services (Schwenk et al., 2012; Sing et al., 2018).
4.4. The impact of climate change and forest type on ecosystem services
The overall effect of climate change was to lower the growth of the trees, decline the production of mushrooms and timber, store less carbon, generate less blue water, and increase tree mortality. Previous studies in the area have also shown that climate change will reduce the growth of trees, especially Pinus sylvestris (Fernández-de-Uña et al., 2015; Gracia et al., 2002), which will automatically lead to carbon and volume losses (Alvarez et al., 2016). Concerning mushroom production and the availability of snags, our results show that climate change will produce a negative effect mainly in the second half of the rotation period, as the stand matures and approaches the end of the century. Overall, there were no large differences in service provision between RCP scenarios, and thinning regime proved much more important in determining forest dynamics and ecosystem services. However, some of the main impacts of climate change on forests will occur through changes in disturbance regimes, which were not considered here (Ruiz-Peinado et al., 2013).
We also found that Pinus nigra will be more affected by climate changes predicted by RCP 8.5 in comparison with RCP 4.5, and will support less production of ecosystem services. The same results have been obtained by Morán-Ordóñez et al. (2020), who have also shown that the forest type influences more the production of ecosystem services than the climate scenario. However, if climatic conditions get even harsher (RCP 8.5), Pinus sylvestris will also suffer, since it is known to have problems at the edge of its distribution (Benito Garzón et al., 2008; Fernández-de-Uña et al., 2015; Morán-Ordóñez et al., 2021).
4.5. Management implications
Pine forests in the region have multiple uses, such as wood and non-wood forest products, pasture production, fire prevention, hydrological protection, biodiversity conservation, recreation. However, management recommendations are still mostly based on the obtention of certain tree sizes that ensure good profitability for timber. In this manner, for an even-aged Pinus sylvestris forest on a high-quality site, where the goal is producing trees with diameters of ~ 50 cm, Piqué Nicolau et al. (2011) recommend a thinning regime from below with moderate to heavy intensities at every 10–20 years. For an even-aged Pinus nigra forest, from a high-quality site, where the goal is producing trees with diameters of ~ 40 cm, Beltrán Barba et al. (2012) recommend a thinning regime from below with moderate intensities every 10–20 years. In our study, we have shown that a thinning regime with heavy intensity every 10 or 15 years can lead to the diameter targets proposed by Piqué Nicolau et al. (2011) and Beltrán Barba et al. (2012), but this would automatically affect the provision of the other ecosystem services. Thus, we do agree with Montero et al. (2001) that an intervention regime that maintains a low density is justified only when the ultimate goal is large size timber and the stand is in a highly productive site. Other management recommendations by the same authors, which consider the protective function of these forests, advocate for less intense and frequent thinnings, which according to our results would be more compatible with a varied service provision. In fact, our study can complement these management recommendations by allowing us to quantify and integrate into the decision-making the provision of services under different climatic scenarios.