Carbon content of forest floor and mineral soil in Mediterranean Pinus spp. and Oak stands in acid soils in northern Spain

Forest Systems 25(2), e065, 13 pages (2016) eISSN: 2171-9845 http://dx.doi.org/10.5424/fs/2016252-09149 Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria (INIA) RESEARCH ARTICLE OPEN ACCESS Carbon content of forest floor and mineral soil in Mediterranean Pinus spp. and Oak stands in acid soils in Northern Spain Celia Herrero*1,2, María Belén Turrión1,3, Valentín Pando1,4, Felipe Bravo1,5 1Sustainable Forest Management Research Institute. University of Valladolid-INIA. E.T.S. Ingenierías Agrarias. University of Valladolid. Avda. Madrid 44, 34071 Palencia, Spain. 2ECM Ingenieria Ambiental, S.L. C/ Curtidores 17. 34004 Palencia, Spain. 3Departamento de Ciencias Agroforestales. E.T.S. Ingenierías Agrarias. Avda. Madrid 57, 34004 Palencia, Spain. 4Departamento de Estadística e Investigación Operativa. E.T.S. Ingenierías Agrarias. Avda. Madrid 57, 34004 Palencia, Spain. 5Departamento de Producción Vegetal y Recursos Forestales. E.T.S. Ingenierías Agrarias. Avda. Madrid 44, 34004 Palencia, Spain Abstract Aim of study: The aim of the study was to determine the baseline carbon stock in forest floor and mineral soils in pine and oak stands in acid soils in Northern Spain. Area of study: The study area is situated in northern Spain (42° N, 4° W) on “Paramos y Valles” region of Palencia. Material and methods: An extensive monitoring composed of 48 plots (31 in pine and 17 in oak stands) was carried out. Litter layers and mineral soil samples, at depths of 0-30 cm and 30-60 cm, were taken in each plot. An intensive monitoring was also performed by sampling 12 of these 48 plots selected taken in account species forest composition and their stand development stage. Microbial biomass C (CMB), C mineralization (CRB), and soil organic C balance at stand level were determined in surface soil samples of intensive monitoring. Main results: No differences in soil C content were detected in the two forest ecosystems up to 60 cm depth (53.0±25.8 Mg C ha-1 in Pinus spp. plantations and 60.3±43.8 Mg C ha-1 in oak stands). However, differences in total C (CT), CMB and CRB were found in the upper 10 cm of the soils depending on the stand development stage in each species forest composition (Pinus nigra, Pinus pinaster, Pinus sylvestris and Quercus pyrenaica). Plots with high development stage exhibited significant lower metabolic quotient (qCO2), so, meant more efficient utilization of C by the microbial community. The C content in the forest floor was higher in pine stands (13.7±0.9 Mg C ha-1) than in oak stands (5.4±0.7 Mg C ha-1). A greater turnover time was found in pine ecosystems vs. oak stands. In contrast, forest floor H layer was nonexistent in oak stands. Research highlights: Results about litterfall, forest floor and mineral soil dynamics in this paper can be used strategically to reach environmental goals in new afforestation programs and sustainable forest management approaches. Keywords: C stocks; pine; Quercus pyrenaica; litter; metabolic quotient (qCO2). Citation: Herrero, C., Turrión, B., Pando, V., Bravo, F. (2016). Carbon content of forest floor and mineral soil in Mediterranean Pinus spp. and Oak stands in acid soils in Northern Spain. Forest Systems, Volume 25, Issue 2, e065. http://dx.doi.org/10.5424/ fs/2016252-09149. Received: 16 Dec 2016. Accepted: 23 May 2016 Copyright © 2016 INIA. This is an open access article distributed under the terms of the Creative Commons Attribution-Non Commercial (by-nc) Spain 3.0 Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Funding: This study has been made possible through research under the FORSEE project - INTERREG III B, financed by the European Union, and through the University of Valladolid grant program and by the Ministry of Economy and Competitiveness of the Spanish Government project Ref: AGL2011-29701-C02-02. Competing interests: The authors have declared that no competing interests exist. Correspondence should be addressed to Celia Herrero de Aza: [email protected] countries must quantify the size, spatial distribution and changes to their soil organic carbon (SOC) stocks. Understanding the mechanisms and factors of SOC dynamics in forest soils is important for identifying and enhancing natural sinks for C sequestration to mitigate climate change effects (Lal, 2005). Carbon storage in forest soils is influenced by soil properties (Zou et al., 2005), climate (Turrión et al., 2009) and topography (Vande Walle et al., 2001). In addition, forest species composition (Oostra et al., 2006), Introduction Soil organic carbon (SOC) is the largest terrestrial pool of carbon and, at least, three times larger than the pool of atmospheric CO2 (Jobbágy & Jackson, 2000). All signatory countries in the United Nations Framework Convention on Climate Change have to implement a national system for reporting carbon stock changes in the agriculture, forests and other land uses sector (Schulp et al., 2008). As part of such a system, Celia Herrero, María Belén Turrión, Valentín Pando, Felipe Bravo Forest Systems August 2016 • Volume 25 • Issue 2 • e065 2 Valles” Region in Northern Spain were considered. On the other hand, in those studies three types of forests were considered (conifer, broadleaf and evergreen broadleaf forests), however, the species composition was not taken account. Finally, litter layers were not included in the mentioned analyses. Knowing the actual C stocks and the dynamics on different forests ecosystems in a region combined with the information on spatial distribution of these ecosystems can be used to improve insight in spatial distribution of SOC stocks under forest land use systems and to estimate the potential capacity of C sequestration of forest soils. Differences in SOC stocks between tree species could give an indication of the effects of future management changes, which can have a strong impact on SOC sequestration. In general, soils reverted to natural ecosystems and under climax vegetation have more soil organic carbon contents than those under managed ecosystems (Lal et al., 1995). Our hypothesis was that the oak stands would show more soil C content because they constituted the climax vegetation in this region. Additionally, tree species are expected to differ in mitigation potential (Schulp et al., 2008). This knowledge may help the manager’s choice of in case of afforestation /reforestation with the aim of enhancing soil C stock (Lal, 2005). The objectives of this paper were: A) to quantify the C content of forest floor and mineral soil under four Mediterranean pine species (Pinus pinaster Ait., Pinus sylvestris L., Pinus nigra Arn.) and under oak ecosystems (Quercus pyrenaica Willd.) in acid soils in Northern Spain. B) to determine the microbiological factors most directly related to the C fixation in the studied soils. C) to estimate the balance of the soil carbon stock, considering litterfall and organic matter decomposition. Material and methods Site description The study area is situated in northern Spain (42° N, 4° W) on “Paramos y Valles” region of Palencia (Figure 1). Geomorfologically is a wide platform, commonly denominated “acidic plateau”, with mean slope of 3% dismembered in a series of tentacles by the erosive action of the current fluvial net. Altitude ranges from 800 to 1000 m asl, while climate can be classified as Humid Temperate Mediterranean. Mean annual temperature is 10.3ºC and mean annual rainfall is 630 mm (mean summer precipitation equal to 107 mm). The soils can be classified mainly as Cambisols and Regosols (JCyL, 2013). stand development stage (Turner & Lamber, 2008), land use (Oliver et al., 2004), litter production and decomposition (Kavvadias et al., 2001), disturbances and silviculture or forest management (Oostra et al., 2006; Jandl et al., 2007) play an important role in SOC balance (Lal, 2005). On the other hand, forest floor is an essential component in the relations among soil and vegetation in the wooded ecosystem and litterfall is a principal pathway for the return of nutrients to the soil (Smith et al., 2015). It has been shown that the amount of litter accumulated in the forest floor is influenced by the nutrient capital of the stand as well as the decomposition rate of the litterfall. As the soil C pool is mostly determined by the balance between C input by litterfall and rhizodeposition and the release of C during decomposition; this part of the C cycling must also be taken into account. Inventories of SOC provide suitable data at different details levels depending on the research objectives and resources. So, while assessing SOC at large scale enables us to define carbon levels considering all the spatial variability, estimating SOC stock at stand level is also necessary to obtain relevant information about the parameters affecting C storage and to adopt appropriate soil management practices. In this sense, studying microbial and C mineralization fractions allowed us to assess the microbial soil activity and to monitor soil changes in a relatively short term (Powlson et al., 1987). Soil microbial biomass carbon (CMB) and the specific respiratory activity of soil microbial biomass (CRB) are sensitive to changes in the quantity and quality of soil organic matter and ecosystem stability (Insam, 1990). The metabolic quotient (qCO2) indicates the energy requirements of soil microorganisms (Anderson, 2003), the level of soil microbial stress (Wardle & Ghani, 1995) and the efficiency of soil microbial populations for substrate utilization. Along with the ratio between the soil microbial biomass C and the total C (CMB/CT), the metabolic quotient reflects organic matter input and availability in the soils, efficiency of conversion to microbial C, C losses from soil, C stabilization by the mineral fractions and maintenance requirements of the soil microbial community (von Lutzow et al., 2002; Turrión et al., 2012). Knowing the soil cycling and the actual capacity of C sequestration of the forest ecosystems would be necessary for calculating the maximum amount of C that is potentially able to return from the trees to the soil. In Spain, there have been three attempts to estimate SOC stocks for the whole country, conducted by Rodríguez-Murillo (2001), Chiti et al. (2012) and Doblas-Miranda et al. (2013). Although these studies presented an assessment of SOC stocks in forests, shrublands and grasslands of Peninsular Spain based on field measurements in a high amount of soil profiles, in general, no data from forest soils of “Paramos y Forest Systems August 2016 • Volume 25 • Issue 2 • e065 3 Carbon in acid soils in Mediterranean forest ecosystems in Spain Extensive monitoring A total of 48 plots (circular plot of radii = 15 m) were selected on the systematic grid from the National Forest Inventory with an intensity of one plot per 2 km, taking the species composition into account (Forsee, 2005). In the region, stands established at the same time, with homogeneous site conditions and similar land use history were sampled. There were 31 plots established on Pinus spp. plantations and 17 on oak stands. The main characteristics of the plots are shown in Table 1. Mineral soil samples were taken in each plot at depths of 0-30 cm and 30-60 cm, as Forsee project established in its protocol (Forsee, 2005). Each plot was divided in four parts, and from 6 to 8 individual randomized samples were obtained for each part. The forest floor accumulated on the ground above the mineral soil was sampled at each point. This was carried out by using a 900 cm2 square sampling frame and collecting all organic material without distinguishing Forests cover 61,570.5 ha (33.0% of total area). The main natural forest types are extensive stands of pyrenean oak (Quercus pyrenaica Willd.), holm oak (Quercus ilex L.) and portuguese oak (Quercus faginea Lam.). As a result of an extensive pine plantation program carried out mostly during the 1960’s in non-arable lands, Pinus stands cover 41.5% of the total forest area. These are young and middle-aged plantations (around 30-60 years old) of Pinus sylvestris L. (23%), Pinus nigra Arn. (21%) and Pinus pinaster Ait. (5%). The plantations are mainly a mixture of the three species, although there are also pure stands. Isolate gaps can be found inside these plantations where mosaics of heather (Erica spp.) and rock roses (Cistus spp.) are present. Sampling procedures Two different soil sampling procedures, extensive and intensive monitoring (Figure 1), were carried out in order to attain the objectives. Forest area Extensive monitoring plots Intensive monitoring plots Figure 1. Location of the study area and the sampled plots of both inventories (extensive and intensive monitoring) at the “Paramos y Valles” region, north of Palencia (Spain). Celia Herrero, María Belén Turrión, Valentín Pando, Felipe Bravo Forest Systems August 2016 • Volume 25 • Issue 2 • e065 4 determined by the dominant tree height and other stand characteristics like stem vigor and quality, tree density and the intensity of the silvicultural treatments applied. In each species composition category, there were three stages of stand development: high, medium and low. High stage was characterized by the best conditions in stand quality and tree size (higher dominant height and higher vigor and quality of the trees, because of a higher intensity of silvicultural treatments). In the opposite case, low stage, was the stage where the worst stand and tree quality characteristics were found. Medium stage is the intermediate level between high and low. So, one plot by each species (n=4) and stand development status (n=3) were considered in this monitoring (n=12, Table 2). A soil profile was opened in each intensive monitoring plot. Soil horizons were described and soils were classified (WRB, 2006). In each intensive monitoring plot, floor layers and 0-10 cm mineral topsoil were sampled. Five sampling points were established on each plot, located in the center of the plot and at a distance of 5 m to North, South, East and West directions. In these points the different organic horizons of the forest floor (undecomposed litter fraction, L; fragmented fraction, F; and humified fraction, H) were differentiated and collected on an area basis, using a 30 cm x 30 cm wooden frame. The samples were mixed, homogenized and composited to form one sample per plot and type of horizon. Litter traps and decomposition litter bags were installed. Litter traps were installed to assess annual litterfall dynamics. Only litterfall was collected. On the other hand, within each plot, three square 30 cm x among different layers. Both soil and forest floor samples were mixed, homogenized and composited to form one sample of each type per plot (forest floor, 0-30 cm mineral soil and 30-60 cm mineral soil). Samples were transported to the laboratory. Mineral soil samples were air-dried, sieved and % of coarse soil materials ([Ø > 2 mm]) was calculated. Intensive monitoring To obtain a better understanding of the factors most directly related to C content on forest floor and mineral soil, twelve of these 48 plots were selected for an intensive monitoring. These points were chosen taking into account two criteria: the forest species composition and stand development stage to consider the different stand situations. Species composition was described by dominant species: P. nigra, P. pinaster, P. sylvestris and Q. pyrenaica (target species basal area ratio higher or equal to 90%). The stand development stage was Table 1. Main characteristics of the extensive monitored plots. Forest stands n trees ha-1 BA (m2 ha-1) QMD (cm) Ho (m) Pinus spp.31 803±344 23.2±8.3 22.2±13.2 9.92±3.81 Oak stands 17 457±501 6.6±8.4 11.6± 8.1 5.12±3.87 Where BA is the stand Basal Area; QMD is quadratic mean diameter; Ho is dominant height (following Asmann definition); n: number of plots. NFI consider as minimum diameter 7.5 cm as DBH to be recorded. Table 2. Main characteristics of the intensive monitored plots (n =12) in the study area. Plot Forest stands Stand dev. Lith. Soil type trees ha-1 BA (m2 ha-1) QMD (cm) Ho (m) Age (years) 1P. nigra High 23 CMca 1367 40.8 19.49 14.4 36 2P. nigra Medium 27 CMdy 1833 40.0 16.67 11.1 24 3P. nigra Low 27 CMdy 689 21.0 19.68 7.5 33 4P. pinaster High 27 CMdy 478 38.6 32.09 15.2 44 5P. pinaster Medium 27 CMdy 689 36.7 26.05 13.1 28 6P. pinaster Low 33 RGdy 667 29.3 23.65 12.7 41 7P. sylvestris High 30 CMdy 733 34.0 24.29 15.9 32 8P. sylvestris Medium 27 CMdy 1100 37.4 20.81 13.2 35 9P. sylvestris Low 27 CMdy 678 31.0 24.14 6.5 41 10 Q. pyrenaica High 27 CMdy 322 13.6 23.21 12.1 88 11 Q. pyrenaica Medium 27 CMdy 778 24.0 19.82 9.8 51 12 Q. pyrenaica Low 30 CMdy 677 2.0 6.11 4.0 14 Where Stand dev. is the stand development stage; Lith is lithology type by geologic map (JCyL, 1995) [23: detritic and carbonate sediments (Garumnian facies), 27: conglomerates and microconglomerates, 30: River texture edges, sands, clays and silts and 33: eolic sands]; Soil type (WRB, 2006) [CMca: Calcaric Cambisol, CMdy: Dystric Cambisol, RGdy: Dystric Regosol]; BA is the stand Basal Area; QMD is quadratic mean diameter; Ho is dominant height (following Asmann definition); Age is the stand age. Forest Systems August 2016 • Volume 25 • Issue 2 • e065 5 Carbon in acid soils in Mediterranean forest ecosystems in Spain Microbiological soil analyses Microbiological soil analyses were carried out in topsoil samples from the intensive monitoring. Microbial biomass C (CMB) was determined by the fumigation-extraction procedure of Vance et al. (1987). Before extracting the microbial biomass, air dried soil (< 2 mm) was incubated at 60% field-capacity and room temperature for one week. Chloroform fumigation of the soils (24h) released microbial cytoplasm into the soil environment. After subsequently extracting the cell material with 0.5 M K2SO4 from fumigated and nonfumigated samples, total organic carbon was quantified as a reference for the microbial biomass. The extracts were analyzed for organic C by using SKALAR TOC/ TN automatic analyzer. The microbial biomass was calculated by biomass C = EC / kEC, where EC is the organic C extracted from fumigated soil minus that extracted from non-fumigated soil, and kEC is the extractable fraction of microbial biomass C after fumigation. A kEC value of 0.38, recommended by Joergensen et al. (1995), was used. Potential soil respiration (CRB) was determined in closed jars and under laboratory-controlled conditions following the Isermeyer method (Alef, 1995). Soil samples were wetted to 60% of water holding capacity and incubated in 1 L jars at 29ºC for 3 days. Evolved CO2 was trapped in 0.5 M NaOH and it was determined by means of back-titration of the remaining NaOH with HCl. The metabolic quotient (qCO2) represents the potential soil respiration per unit microbial biomass, and was calculated as the quotient between CRB and CMB. Microbial quotient (CMB/CT) represented the fraction of microbial biomass C with respect to the total soil organic C (Anderson & Domsch, 1993). All results were obtained by triplicate determinations and were expressed on the basis of oven-dry weight soil. Organic horizons, litter pool and topsoil analysis In all samples from intensive monitoring (organic horizons, litterfall from traps, litter bags and upper mineral soils), total C and total N were determined by dry combustion using the automated C/N analyzer (CHN-2000 LECO). Statistical analyses A Linear Mixed Model (LMM) with a betweensubjects factor (forest ecosystem, with two levels Pinus 30 cm litter traps placed randomly under the canopy, to take into account spatial variability at the plot level. The traps were established in May 2008 and were collected every two months throughout the entire year (5 times). The sampling ended in May 2009. The total trap capacity was extracted each time, weighted and dried until constant weight in the laboratory. On the other hand, to estimate the decomposition litter rate, foliar litter bags were installed in the each plot. A total of 144 litter bags were prepared using polyvinyl screen mesh (0.5 mm in the bottom and 2 mm in the top) of approximately 25 cm × 25 cm. In each plot, air-dried leaf litter (7 g), obtained previously from the upper soil in each species composition plot, was put into the bags. Three 3 litter bags were extracted from each plot every 2 months after the start of the decomposition study (the same 5 times). Contents of litter traps, decomposition litter bags and organic horizon samples were oven-dried at 70ºC, weighed in the laboratory and an aliquot was ground by a mill to plant samples. Analyses procedures Physical and chemical soil properties Some physical and chemical soil properties (bulk density, texture, pH, cation exchangeable capacity, Ca, Mg, K and Na exchangeable cation concentrations, sum of bases total C, total N and C/N) were determined in mineral soil samples of extensive monitoring. Bulk density was measured through the core method (Blake & Hartge, 1986) in the field with volumetric steel rings and soil dry weight. Texture was determined by Bouyoucos Hydrometer Method (MAPA, 1994), pH (soil:solution ratio of 1:2.5) was measured using a pH-meter. Cation exchange capacity (CEC) was determined by the Bascomb method (Bascomb, 1964). Exchangeable cations were extracted with 0.1 M BaCl2, Ca and Mg were determined by atomic absorption spectrophotometry and K and Na by emission spectrophotometry (MAPA, 1994). The sum of the bases (SB) was calculated as sum of Ca, Mg, K and Na exchangeable. Soil organic matter was analyzed by the Walkey and Black method. Soil total C concentration was determined multiplying easily oxidable carbon concentration with a factor of 1.3 (MAPA, 1994). Total N was analysed by Kjeldahl method (Jones et al., 1991). The C/N relationship was calculated as the ratio between soil C and total N. Accumulated SOC in each depth was calculated taking into account C concentration, bulk density, their thickness and the percentage of gravels. Celia Herrero, María Belén Turrión, Valentín Pando, Felipe Bravo Forest Systems August 2016 • Volume 25 • Issue 2 • e065 6 Results General soil parameters In Table 3 some physical and chemicals properties of the studied soils by depths (0-30 cm) and (30-60 cm) and forest ecosystems (Pinus spp. and oaks) are shown. So, in oak forest ecosystems, significant higher values of clay content were found in the second depth [30-60 cm] than in the upper depth. Also, between depths, in both forest ecosystems (Pinus spp. and oak stands), significant higher values were found in 0-30 cm respect to 30-60 cm in exchangeable K, CT and N concentrations. However, in oak stands, smaller values were found in CEC and C/N in the upper depth in comparison to 30-60 cm depth. On the other hand, between forest ecosystems, in the upper depth, while significant smaller values were found in pH, CEC, exchangeable K and Mg, SB and N concentration in Pinus spp. than in oak stands, a significant higher value of C/N was found at this depth in Pinus spp. In contrast, at 30-60 cm of depth, all the significant values were smaller in Pinus spp. than in oak stands (clay content, pH, CEC, exchangeable K and Mg, and SB). The interaction depth*forest ecosystems was significant only in Ca concentration. In this cation, in the second depth, significant higher values were found in oak than in Pinus spp. forests, while in the upper depth the values were similar. Finally, the soil C content did not present significant differences by depth and by forest ecosystems, either by the interaction between depth and forest ecosystems. Soil C accumulation was 53.0 (±25.8) Mg C ha-1 in Pinus spp. plantations and 60.3 (±43.7) Mg C ha-1 in oak stands (Table 3). spp and oak) and a within-subjects factor (depth, with two levels, 0-30cm and 30-60cm) and their interaction, was applied to the soil parameters in the extensive inventory. When the effect studied turned out to be significant, differences among levels were evaluated using the Tukey test. In the intensive inventory, a LMM with two between-subjects factors and their interaction was fitted to assess the differences in the C fractions of the topsoil and forest floor layers. The factors considered in the model were the species forest composition with four levels: Pinus pinaster, Pinus sylvestris, Pinus nigra, and Quercus pyrenaica, and the stand development stage with three levels: high, medium and low. So, the total effects were 12. When the effect studied turned out to be significant, differences among levels were evaluated using the Tukey test. All the statistical analyses were performed using Proc MIXED from SAS 9.1 (SAS Institute Inc., 2010). A descriptive study was carried out to ascertain the annual fluxes of inputs and decomposition of litter in forest ecosystems. Differences in the parameters litterfall amount, C concentration and C/N relationship were evaluated among forest ecosystems and sampling time through Tukey’s procedure by Proc GLM from SAS 9.1 (SAS Institute Inc., 2010). Decomposition litter rate constants (k, in year-1) were calculated from the different sampling time data by using Equation 1 [Eq. 1] X=X0e−kt [1] where Xo is the initial weight of needles in the litter bags, X is the weight of needles after time t in years and k is the decomposition rate constant (Olson, 1963). Table 3. Physical and Chemicals properties of the soils by depths and forest ecosystems. Depth (cm) Forest ecosystem Bulk density (g cm-3) Clay (%) pH CEC (cmolc kg-1) Exchangeable cations (cmolc kg-1)SB (cmolc kg-1) C (g kg-1) C (Mg ha-1) N (g kg-1)C/N K Ca Mg Na 0-30 Pinus spp. 1.10 aA 9.1 aA 5.4 aA 12.00 aA 0.35 aA 2.62 aA 0.52 aA 0.04 aA 3.52 aA 21.0 aA 30.0 aA 0.9 aA 23.6 aA Oak stands 1.09 aA 12.8 aA 5.8 aB 13.92 aB 0.61 aB 4.02 aA 0.73 aB 0.04 aA 6.68 aB 21.1 aA 38.1 aA 1.2 aB 17.7 bA 30-60 Pinus spp. 1.25 aA 11.1 aA 5.4 aA 10.63 aA 0.25 bA 1.89 aA 0.33 bA 0.04 aA 2.50 aA 14.0 bA 22.9 aA 0.6 bA 25.0 aA Oak stands 1.22 aA 18.8 bB 5.9 aB 14.14 bB 0.43 bB 5.99 aB 0.75 aB 0.04 aA 6.15 aB 13.1 bA 22.2 aA 0.7 bA 18.3 aA FACTOR: Depth ns ** ns ns ** ns ns ns ns *** ns *** ns Forest ecosystems ns *** *** *** *** *** *** ns *** ns ns ** * Depth* Forest ecosystems ns ns ns ns ns ** ns ns ns ns ns ns ns Note: SB: Sum of bases; CEC: cation exchangeable capacity, Different lower letters showed significant differences in each forest ecosystems between depths at 95% Tukey’s test. Upper letters showed significant differences in each depth between forest ecosystems forest at 95% Tukey’s test. Significant levels: ***: (p<0.01); **: (p<0.05); *: (p<0.1) ; ns: no significant. Forest Systems August 2016 • Volume 25 • Issue 2 • e065 7 Carbon in acid soils in Mediterranean forest ecosystems in Spain higherst values were found in the highest stand development stage. For the CRB/CT ratio significantly higher values were found in the high than in the medium and low stand development in Pinus nigra plots. For the other forest species the significant lowest CRB/CT values were found in the plots with high stand development stage. For the qCO2, no significant differences were found in Pinus nigra stands. However, in the other forests, low stand development plots showed significant higher qCO2 values than high stand development stages. Forest floor Carbon The average amount of forest floor was 16.5 Mg ha-1, showing average values of 26.9 (±7.4) Mg ha-1 in pine ecosystems and 11.8 (±6.5) Mg ha-1 in oak stands. The carbon content in this pool was significantly higher in pine [13.7 (±3.8) Mg C ha-1] than in oak stands [5.4 (±3.0) Mg C ha-1]. The amount and the carbon content in forest floor varied among species composition, stand development stage and the interaction between the two factors (Table 5). Analyzing the different forest floor layers Soil microbial activity and C mineralization Microbial biomass C represented between 0.56-1.37 % of total C (mean 0.94 ± 0.30 %). Significant differences were found (Table 4) in the interaction between species composition and stand development stage in the CT, the ratio CMB/CT , CRB/CT and the metabolic quotient (qCO2). So, these variables, in the studied soils, were different in the different species composition depending on the stand development stage considered. In Pinus nigra stands, the significant lowest CT value was found in the high stand development stage plot. In contrast, significant higher CT values were found in the high development stage stands than in the others for Pinus sylvestris and Quercus pyrenaica forests. For Pinus pinaster significant higher CT values were found in low development stages in comparison to the other two stages. For the ratio CMB/CT, different patterns were found. So, no significant differences were found in Pinus nigra stands, and significant different values were found in each stand development stage in Pinus pinaster stands. In Pinus sylvestris, a significant higher value was found in lower stand development stage than in the others, however in Quercus pyrenaica, the significant Table 4. Values of CT content, CMB/CT ratio, CRB and metabolic quotient (qCO2) in the upper 10 cm soil by species forest composition and stand development stage Species Forest Composition Stand Development CT (gC 100g-1) CMB/CT (gCMB kgCT-1) CRB/CT (gCRB kgCT-1) qCO2 (gCRB gCMB-1) Pinus nigra High 1.32 b 8.42 a 9.15 a 1.11 a Medium 2.27 a 6.86 a 5.06 c 1.05 a Low 2.55 a 7.33 a 7.21 b 1.05 a Pinus pinaster High 2.15 b 11.21 b 4.72 c 0.43 c Medium 2.56 b 13.52 a 7.48 a 0.55 b Low 3.20 a 8.62 c 6.29 b 0.73 a Pinus sylvestris High 4.08 a 5.57 b 1.90 c 0.37 c Medium 3.42 b 5.21 b 4.70 b 1.17 a Low 3.39 b 12.93 a 9.19 a 0.71 b Quercus pyrenaica High 4.37 a 13.66 a 4.29 c 0.32 b Medium 2.92 b 11.27 b 10.44 a 0.93 a Low 2.72 b 8.74 b 9.44 b 1.22 a FACTOR Species Forest Composition *** *** *** *** Stand Development ns *** *** *** Species Forest Composition * Stand Development *** *** *** *** Where CT is the total organic C content; CMB is the soil microbial biomass C and CRB is the biomass respiration C; qCO2 is the metabolic quotient (CRB/CMB). Different letters means significant differences among stand development stages within each species forest composition. Significance levels: ***: (p<0.001). ns: not significant. Celia Herrero, María Belén Turrión, Valentín Pando, Felipe Bravo Forest Systems August 2016 • Volume 25 • Issue 2 • e065 8 nual litterfall carbon was 8.9 (±1.8) and 2.9 (±0.7) Mg C ha-1 year-1 in Pinus spp. and oak stands, respectively. Foliage peaked in autumn (Figure 2), where more than 50% of the total litter was fallen in that season in the four species compositions. Litterfall showed different values of C/N along the sampling period in the different forest stands (Figure 3). Smaller values were found in oak stands along the year. The litter bag experiment showed that degradation of dry organic matter decreased along the year in the different forest stands (Figure 4). The models (p<0.0001) allowed us to determine the value of the decomposition rate constant k (year-1). The k value was different between Pinus spp. and oak stands. They were 0.18 and 0.46 years-1 for the dry organic matter in Pinus spp. and oak, respectively, and 0.16 and 0.32 years-1 for the C content in Pinus spp. and oak stands, respectively. Discussion In this paper, C stock in forest floor and mineral soils in pine and oak stands in acid soils in Northern Spain has been determined. The study area showed significant (L, F and H), the mass was substantially greater in the L, H layers than in F layer in most of the plots (Table 5). The C content in H layer showed significantly smaller values in oak stands. In most of the plots of the four species compositions, the carbon concentration ranged in the following order: L>F>H. Within each forest composition, the stand development stage showed different C amounts and contents of L and F layers without a clear tendency, except for H layer. While a higher C was found in L layer in the highest stand development stage for P. nigra and P. sylvestris, the opposite behavior was found for P. pinaster and Q. pyrenaica. However, in the different stand development stages of Pinus spp. species compositions, similar values were found in H layer. The values of the relation C/N in pine ecosystems (L>98; F>50 and H>38) contrasted to those obtained in oak stands (>46 in L and F layers). Litterfall and litter decomposition Figure 2 shows the temporal evolution of C in litterfall at the four forest ecosystems. The average an- Table 5. Values of C concentration, Carbon content and C/N ratio of the different litter layers (L, F and H) by species forest composition and stand development. Species Forest Composition Stand Development L F H CLCFCH (C/N)L(C/N)F(C/N)H (Mg ha-1) (g kg-1) (Mg ha-1) (g kg-1) (Mg ha-1) (g kg-1) (Mg ha-1) P. nigra High 8.7 a 5.4 a 7.0 a 497 a 4.3 a 464 a 2.4 ac 325 a 2.4 a 139.1 a 84.6 a 53.6 a Medium 3.9 bc 4.1 a 9.8 a 497 a 1.9 b 425 a 1.7 c 327 a 3.4 a 113.9 b 54.9 b 45.5 a Low 4.9 b 4.9 a 11.0 a 546 b 2.6 a 544 b 2.5 a 425 b 6.5 a 98.4 b 80.0 a 53.1 a P. pinaster High 4.2 a 4.5 a 13.8a 494 a 2.1 a 438 a 1.9 a 223 a 3.6 a 131.3 a 63.8 a 46.5 a Medium 7.9 c 4.4 a 9.6 a 497 bc 3.9 b 463 ab 2.0 a 270 ab 2.5 a 100.7 bc 65.7 a 47.0 a Low 5.9 b 3.9 a 7.2 a 455 ac 2.7 a 473 b 1.8 a 316 b 2.4 a 129.1 ac 71.4 a 46.5 a P. sylvestris High 8.3 a 5.4 a 16.7 a 503 a 4.2 a 466 a 2.5 a 311 a 5.5 ac 98.3 a 60.5 a 47.7 a Medium 4.4 b 2.8 b 15.8 a 496 b 2.2 b 470 a 1.3 b 397 c 6.3 c 102.9 a 50 b 38.0 b Low 8.2 a 4.2 a 13.4a 497 b 4.1 a 475 a 2.0 a 273 b 3.6 a 128.6 b 74.2 c 51.2 a Q. pyrenaica High 2.6 a 4.5 abc 0.0 473 a 1.2 a 393 a 1.9 a - 0.0 63.2 a 46 a - Medium 4.1 b 2.3 ac 0.0 436 b 1.7 bc 339 a 0.7 c - 0.0 48.8 b 46.4 a - Low 4.3 b 4.4 b 0.0 439 b 1.9 b 319 a 1.5 a - 0.0 51.3 b 46.3 a - FACTOR Species Forest Composition *** ns * *** *** *** ** *** ** *** *** *** Stand Development ns *** ns ** ** ns *** ** ns *** *** *** Species Forest Composition * Stand Development *** ** ns *** *** ** ** *** ns *** *** *** Where L is the amount of undecomposed litter fraction; F is the amount of fragmented fraction and H is the amount of humified fraction; CL, CF and CH are carbon concentration (g kg-1) and amount (Mg ha-1) in L, F and H fractions, respectively; (C/N)L is the C/N relationship on the L layer; (C/N) F is the C/N relationship on the F layer; (C/N)H is the C/N relationship on the H layer. Significance levels: *: (p<0.05). **: (p<0.01). ***: (p<0.001). ns: not significant. Forest Systems August 2016 • Volume 25 • Issue 2 • e065 9 Carbon in acid soils in Mediterranean forest ecosystems in Spain of CEC found. This fact, join the significant smaller values of C/N in oak stands reflected the higher organic matter quality in this type of forest ecosystems in comparison to Pinus spp. stands. On the other hand, our results showed that similar soil C stock exists underground, while different forest ecosystems grow aboveground. Our hypothesis was that the oak stands would show more soil C content because they constituted the climax vegetation. However, the oak stand degradation causes that this ecosystem accumulates C in a manner similar to middle-aged plantations that are growing and are being managed for wood production. Selvicultural prescriptions to improve the vigor of oak stands would be necessary to increase the C fixation in this type of forest. In addition, the C fixed allowed us to assess the repercussions of pine plantations on soil C in this area. Plantations were established in non-productive lands. In the study area, Mulas et al. (2015) assessed C budgets under different land use systems. Their results showed that C concentration in surface horizon (0-10 cm) was 16.9 g kg-1 under shrubs and 19.0 g kg-1 under crops. The comparison of C in the different land uses reflects the importance of forest ecosystems in the region. Afforestation programs have to consider this fact. The lack of litter inputs from the previous ecosystem is progressively compensated by litter inputs from the newlyestablished forest plantation. After several years, a new soil C equilibrium is eventually reached, when carbon outflows from decomposition are balanced by carbon inflows from litter production (Jandl et al., 2007). Amounts of SOC accumulation found in this study agree with previous works. Rodriguez-Murillo (2001) estimated the soil C under different types of land use and soil in peninsular Spain obtaining under conifers mean values of 75 Mg C ha-1 and under broadleaves forests mean values of 93.6 Mg C ha-1. Turrión et al. (2009) found that soils under Q. pyrenaica forests in Western Spain (Sierra de Gata Mountains) had a high capacity to accumulate SOC (between 33 and 185 Mg C ha-1), while higher values were found in chestnut coppice forest (195 Mg C ha-1) by Gallardo & González (2004). In this paper, two different inventories have been carried out. The first one, the extensive inventory allowed us to consider soil spatial variability and to know the importance of the soil C in the ecosystem. Bernoux et al. (2002) emphasized the importance of large-scale studies to refine global estimations obtained by the aggregation of local estimates. Previous studies carried out on the tree biomass in this region (Herrero & Bravo, 2012) showed that Pinus spp. tree biomass fixed an average static value of 42.6 Mg C ha-1, while oak stands fixed 5.2 Mg C ha-1. Total C fixed in the different pools higher clay contents at 30-60 cm depth in oak stands due to the presence of a cambic horizon at this depth. In contrast, in Pinus spp., clay contents didn´t show significant differences between depths probably due to mixing of soil horizons when the afforestation was performed. The higher clay contents found in the second depth in oak stands could explain the higher values Figure 2. C content (g kg-1) in litterfall picked up by sampling time. Note: d.w. is dry weight. Aug is August; Oct is October; Dec is December; Feb is February. C [g kg-1] of d.w. Sampling time Pinus pinaster Quercus pyrenaica Pinus nigro Pinus sylvestris Aug 08 Oct 08 Dec 08 Feb 09 May 09 2.0 1.6 1.2 0.8 0.4 0.0 Figure 3. Relation C/N of litterfall picked up in the traps by sampling time in the different species compositions. Note: Aug is August; Oct is October; Dec is December; Feb is February. C/N 50 40 30 20 10 0 Sampling time Aug 08 Oct 08 Dec 08 Feb 09 May 09 Quercus pyrenaica Pinus nigro Pinus pinaster Pinus sylvestris Sampling time Pinus pinaster Quercus pyrenaicaPinus nigro Pinus sylvestris May 08 Aug 08 Oct 08 Dec 08 Feb 09 May 09 1.0 0.9 0.8 0.7 Figure 4. Evolution of the ratio X X0 by sampling time by species compositions in litter bag experiment. Note: Aug is August; Oct is October; Dec is December; Feb is February. X is the weight of needles after time; X0 is the initial weight of needles in the litter bags. X X0