Soil Biology and Biochemistry 165 (2022) 108515 Contents lists available at ScienceDirect Soil Biology and Biochemistry journal homepage: www.elsevier.com/locate/soilbio Seasonal drought in Mediterranean soils mainly changes microbial C and N contents whereas chronic drought mainly impairs the capacity of microbes to retain P S. Marañón-Jiménez a, b, *, D. Asensio a, b, J. Sardans a, b, P. Zuccarini a, b, R. Ogaya a, b, S. Mattana c, J. Peñuelas a, b a CSIC, Global Ecology CREAF- CSIC-UAB, Cerdanyola del Valles, 08193, Barcelona, Catalonia, Spain b CREAF, Cerdanyola del Valles, 08193, Barcelona, Catalonia, Spain c Research Group Plants and Ecosystems (PLECO), Department of Biology, University of Antwerp, Wilrijk, B-2610, Belgium A R T I C L E I N F O A B S T R A C T Keywords: Intensification of droughts may aggravate the generally low capacity of Mediterranean soils to store C and nu­ Drought stress trients and induce soil C:N:P stoichiometric imbalances through its impact on soil microbial biomass and activity. Microbial nutrients Soil microbes may nonetheless have different responses to seasonal and chronic drought, but very few studies Microbial osmolytes investigate long-term drought periods under field conditions. This study compares the effects of seasonal drought Soil enzymes versus the impacts of 16 years of chronic experimental drought on microbial biomass and nutrients and assess the Soil nutrients Water limitation implications for soil nutrient availability and biogeochemical functioning in a Mediterranean forest. The chronic drought treatment reduced substantially and persistently microbial biomass C, N and particularly P, probably due to P-sparing community shifts or microbial adaptations. The smaller microbial N pool and lower minerali­ zation activity contributed to the accumulation of C- and N-rich organic compounds in the soil and to a lower availability of mineralized forms of N during the vegetation growing season. As a result, chronic drought con­ ditions may increase the risks of N losses from the plant-soil system in Mediterranean ecosystems. Microbial C:N ratios remained unaltered under chronic drought compared to control, likely associated with the equivalent accumulation of C- and N-rich osmolytes by microbial communities. In contrast, microbial biomass increased its C content relative to N content in response to seasonal drought, but also reduced considerably its N and P pool. Therefore, while microbial P was more sensitive to chronic water stress, microbial N and C were more closely coupled to the seasonal fluctuations of water availability. 1. Introduction litter decomposition (Santonja et al., 2017) and soil organic matter (SOM) mineralization rates (Marañón Jiménez et al., 2011), with the The frequency and intensity of droughts in the Mediterranean region consequent lower nutrient release in assimilable forms for plants. have increased since 1950, accompanied by more irregular and torren­ Drought impacts generally take longer to appear in Mediterranean tial rain (Hartmann et al., 2013; Vicente-Serrano et al., 2014). Intensi­ vegetation (Vicente-Serrano et al., 2013). In the long term, the direct fication of droughts and changed patterns of precipitation associated effect of water limitation together with the consequent lower nutrient with climate change can have large impacts on the frequently accesibility imposed by drought also reduce plant growth, nutrient up­ nutrient-poor Mediterranean soils (Yaalon, 1997). Summer droughts take and foliar nutrient concentrations (Kreuzwieser and Gessler 2010; decrease soil enzymatic activity (Sardans and Peñuelas, 2005; 2010), Sardans and Peñuelas 2007; Sardans et al. 2008a, 2020). Abbreviations: acp, acid phosphatase activity; akp, alkaline phosphatase activity; bgl, β-glucosidase activity; Cmicro, microbial C; CUE, carbon use efficiency; EOC, extractable organic carbon; ETN, extractable total nitrogen; Pt, extractable total phosphorus; Kmicro, microbial K; Mnmicro, microbial Mn; Momicro, microbial Mo; Nmicro, microbial N; Pi, bicarbonate-extractable inorganic P; Pmicro, microbial P; Po, bicarbonate-extractable organic P; prot, protease activity; SOM, soil organic matter; sresp, soil CO2 efflux; stemp, soil temperature; swclab, gravimetric soil water content measured in the laboratory; ure, urease activity. * Corresponding author. CSIC, Global Ecology CREAF- CSIC-UAB, Cerdanyola del Valles, 08193, Barcelona, Catalonia, Spain. E-mail addresses: s.maranon@creaf.uab.cat, s.maranon@creaf.uab.es (S. Marañón-Jiménez). https://doi.org/10.1016/j.soilbio.2021.108515 Received 2 May 2021; Received in revised form 3 December 2021; Accepted 4 December 2021 Available online 6 December 2021 0038-0717/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). S. Marañón-Jiménez et al. Soil Biology and Biochemistry 165 (2022) 108515 Seasonal and chronic drought, however, may have divergent effects this compartment to store nutrients, although the net effects of drought on soil biogeochemical functioning (Rousk et al., 2013; Hoover and on microbial biomass C are largely inconsistent in the literature (Deng Rogers, 2016; Zuccarini et al., 2020). Short term seasonal drought can et al., 2021; Abbasi et al., 2020), and largely depend on the length and increase the accumulation of SOM and nutrients in forms unavailable to intensity of the drought period. plants (Sardans et al., 2008b; Peñuelas et al., 2018) by increasing the The role of microbial biomass in regulating nutrient availability in total amount of above- and belowground litter (Ogaya and Peñuelas, Mediterranean soils has nonetheless been poorly addressed. The most 2007; Martí-Roure et al., 2011; Padilla and Pugnaire 2007) together studies mainly focus on the short-term droughts, whereas there is a lack with reductions in potential soil enzymatic activities and microbial of comprehensive understanding on the long-term drought effects (Deng mineralization rates at low soil water contents (Sardans and Peñuelas, et al., 2021). Very few studies have identified the transient effects of 2010; Manzoni et al., 2012). Long term chronic drought reduces plant seasonal drought and the permanent impacts of chronic drought on nutritional status, growth and cover, decreasing the quantity and quality microbial stoichiometry and on the size of nutrient storage in the mi­ of plant litter inputs to the soil (Sardans et al., 2008a, 2020, Lloret and crobial biomass pool. We compared the effects of seasonal drought Granzow-de la Cerda, 2013) and soil nutrient availability (Peñuelas versus the impacts of 16 years of chronic experimental drought on the et al., 2018; Sardans et al., 2020). Lower soil protection and perme­ size of the nutrient pool stored in the microbial biomass compartment ability increase soil erosion (Li et al., 2011; Navarro-García et al., 2012) and assessed the implications for soil C, nutrient availability and soil and nutrient loss (Zaimes et al., 2012, Bloor and Bardgett, 2012), biogeochemical functioning in a Mediterranean forest. We call hereafter especially if drought is accompanied by more frequent and torrential microbial storage the size of the nutrient pool stored in the microbial rain. Chronic drought may therefore aggravate the generally low ca­ biomass compartment. We hypothesized that 1) both seasonal and pacity of Mediterranean soils to store C and nutrients (Beier et al., 2009) chronic drought would lead to reductions in microbial biomass and in and induce soil C:N:P stoichiometric imbalances (Sardans and Peñuelas, the nutrients stored in this compartment; 2) this reduction in microbial 2012; García-Angulo et al., 2020). biomass and nutrients caused by chronic drought is persistent across Soil microbial biomass plays a fundamental role in the stabilization seasons; 3) the diminished capacity of the microbial biomass community of soil C (Liang et al., 2017; Miltner et al., 2012) and in the regulation of (as a whole) to store nutrients under chronic drought conditions is fol­ nutrient availability in climates with marked seasonality (Bardgett et al., lowed by a lower nutrient availability for vegetation during its growing 2013), such as in Mediterranean ecosystems (Qiu et al., 2009). A large period. fraction of the available nutrients in Mediterranean soils is stored in microbial biomass (Díaz-Raviña et al., 1995; Aponte et al., 2010). This 2. Material and methods microbial nutrient pool oscillates throughout the year with seasonal changes in environmental factors, and these temporal dynamics may 2.1. Study site determine the amounts of nutrients released or immobilized in the soil and therefore their availability to plants (Díaz-Raviña et al., 1995). Soil The study was carried out in a natural holm oak (Quercus ilex) forest microbial communities adapt during dry periods by synthesizing in the Prades Mountains in Catalonia, northeastern Iberian Peninsula extracellular polysaccharides (Marchus et al., 2018) and storing simple (41◦ 21′ N, 1◦ 2′ E). The vegetation is characterized by a dense forest, with C- and N-rich osmolytes, such as amino compounds in bacteria and Q. ilex, Phillyrea latifolia and Arbutus unedo the main arboreal species. polyols in fungi, in high concentrations to combat osmotic stress The soil is a Dystric Cambisol with a depth of 35–90 cm (Sardans and (Schimel et al., 2007). These adaptations allow microbes to survive and Peñuelas 2010). The climate is typically Mediterranean, with hot and preserve part of their cellular and nutrient contents during drought dry summers (mean temperature of 20.5 ± 0.3 ◦ C, total precipitation of (Gutiérrez-Girón et al., 2014). The capacity of the soil microbial biomass 64.8 ± 7.0 mm, and mean soil moisture of 10.9 ± 0.7% vol.) and warm pool to store nutrients becomes particularly important during periods of winters (mean temperature of 4.8 ± 0.2 ◦ C, total precipitation of 110.6 drought, because the decrease in vegetation nutrient uptake and in­ ± 16.0 mm, and mean soil moisture of 20.8 ± 1.1% vol.). The springs crease in foliar senescence may lead to the accumulation of nutrients in (mean temperature of 10.8 ± 0.4 ◦ C, total precipitation of 237.4 ± 24.0 partially decomposed organic matter on the soil surface and to increase mm and mean soil moisture of 20.7 ± 1.6% vol.) and autumns (mean the risk of nutrient loss (Schaeffer et al., 2017). Microbes release stored temperature of 12.7 ± 0.6 ◦ C, total precipitation of 198.9 ± 23.9 mm osmolytes during rewetting periods and reactivate the mineralization of and mean soil moisture of 17.6 ± 2.1% vol.) are usually rainy (mean these and other organic substrates accumulated in soil during dry pe­ values from 1998 to 2014, Figs. S1 and S2). riods, therefore releasing assimilable nutrients for plant growth (Fierer and Schimel, 2002; Schimel, 2018). Microbial biomass may therefore 2.2. Experimental design buffer the seasonal oscillations in soil nutrients and act as a nutrient reservoir for plant nutrition during favorable periods (Aponte et al., The drought experiment began in 1999 and consisted of partial 2010; Bardgett et al., 2005). rainfall exclusion that simulated a 30% reduction in precipitation for the Chronic drought conditions, however, may impair the role of mi­ forest (hereafter chronic drought treatment). Eight plots (10 × 15 m) crobial biomass in regulating the availability of soil nutrients. Low levels were established at the same altitude (930 m a.s.l.) along the southern of soil water content reduce the connectivity among nutrient-rich soil face of the mountain (25% slope). Half of the plots received the chronic micropores and limit the mobility of soluble substrates and extracellular drought treatment, and the other four plots served as controls (natural enzymes within the soil matrix (Borken and Matzner 2009; Moyano conditions). At the start of the experiment, just before the rainfall et al., 2013). The lower substrate accessibility for soil microbes increases exclusion, soil water content measured during the previous autumn and the resources allocation of soil microbes to enzyme production and in­ winter was similar in the plots that received the chronic drought treat­ creases their C costs for resources acquisition (Schimel 2018; Asensio ment and the control plots (24.7% and 24.8%, respectively). Soil total N et al., 2021). Drought also forces microbes to synthesize structures to concentration in 1999 was 0.37 ± 0.07%. Precipitation was reduced a resist the effects of drought, to accumulate C- and N-rich compounds for 30% in the chronic drought treatment by the installation of plastic strips osmotic regulation (Schimel 2018) and to increase the allocation of re­ 0.5–0.8 m above the ground and covering 30% of the ground surface of sources to sustain basal metabolic rates rather than to growth, thus the plots, and by the interception of water runoff by ditches. The pre­ decreasing microbial C use efficiency (CUE) (Tiemann and Billings 2011; cipitation intercepted by the plastic strips and ditches was conducted Manzoni et al., 2012). These elevated C costs for maintenance when outside the plots. As a result, soil moisture was reduced a 15% in the substrate mobility and accessibility are low can potentially reduce mi­ chronic drought treatment throughout the entire duration of the crobial biomass and consequently soil C stabilization and the capacity of experiment (Fig. S2). Further details of the chronic drought treatment 2 S. Marañón-Jiménez et al. Soil Biology and Biochemistry 165 (2022) 108515 are described by Ogaya et al. (2020). (K), manganese (Mn) and molybdenum (Mo) were measured in the unfumigated NaHCO3 extracts following the procedures described 2.3. Soil sampling above. One aliquot of the NaHCO3 extracts was used for determining the concentration of inorganic extractable P (Pi) using the Olsen’s method We sampled the soil in 2014 on three consecutive days in the middle (Watanabe and Olsen, 1965). The correction for P adsorption was also of each season: winter (25–27 January), spring (16–18 May), summer applied to this fraction. The concentration of organic extractable P (Po) (31 July to 2 August) and autumn (24–26 October). We sampled five soil in the extracts was calculated as the difference between Pt and Pi. The cores per plot and season (5 cm diameter and 10 cm deep). The cores elemental concentrations in the extracts were expressed as μg of were collected in all seasons around 15:00, after soil respiration was extractable element per unit of soil dry mass. measured (see below). The samples were sieved in the laboratory (2-mm Total organic C (TOC) and N (TN), ammonium and nitrate were sieve) and stored at 4 ◦ C until analysis. Subsamples were used for the determined from 0 to 15 cm depth soil samples taken in the same way as determination of gravimetrical soil water content (swclab). Additional above in 2005, six years after the experiment started. For the determi­ subsamples were used for the analysis of microbial biomass C, N, P, K, nation of TOC in soil samples we used the Walkley-Black method Mn and Mo concentrations, soil extractable C, N, P, K, Mn and Mo (Walkley and Black 1934). TN concentrations in soils were determined concentrations and potential activities of soil extracellular enzymes. by organic elemental analysis using combustion coupled to gas chro­ matography. For that, we used a Thermo Electron gas chromatograph 2.4. Microbial elemental concentrations and extracellular enzymatic model NA 2100 (C.E. Instruments-Thermo Electron, Milano, Italy). activities Ammonium and nitrate were extracted with a 2 M KCl solution and analyzed by colorimetrically using a spectrophotometer (Spectronic 20 Elemental C, N and P concentrations in the microbial biomass were Genesys; Spectronic Instruments, Inc., Rochester NY) against the reagent measured using chloroform fumigation-extraction (Jenkinson and blank. We analyzed ammonium in soil extracts by a modified Berthelot Powlson, 1976). Two subsamples (10 g fresh weight) of sieved soil were reaction (Schinner et al., 1996) and nitrate by the cadmium reduction used for each sample. One subsample was fumigated for 24 h with method (US Environmental Protection Agency, 1979). ethanol-free chloroform and extracted using 0.5 M K2SO4 (4:1 v:w) for C We calculated the C:N, C:P and N:P stoichiometric imbalances be­ and N and 0.5 M NaHCO3 (10:1 v:w) for P, after shaking for 30 min. The tween the extractable fractions and microbial biomass following Moos­ other subsample was directly extracted following the same protocol. The hammer et al. (2014) as the ratios of the extractable fractions (EOC:ETN, extracts were then filtered through Whatman 42 equivalent paper. Total EOC:Pt and ETN:Pt) to the corresponding ratios of the microbial frac­ organic C and total N concentrations in the K2SO4 extracts were tions (Cmicro:Nmicro, Cmicro:Pmicro and Nmicro:Pmicro, respectively). The measured using a Multi N/C 3100 Total Organic Carbon/Total Nitrogen stoichiometric imbalance was thus a measure of the divergence between analyzer (Analytik Jena AG, Jena, Germany). Total P concentration in the stoichiometries of the soil microbes and the soil organic substrates. the NaHCO3 extracts was determined after the digestion of 2.5 g of an For example, a decrease in the C:N imbalance would indicate an aliquot with HNO3; a correction for the adsorption of P during extraction increasing microbial demand of C from the soil extractable fractions, was applied after measuring the recovery of samples spiked with P as whereas an increase in the C:N imbalance would indicate an increasing described by Zuccarini et al. (2020). Elemental P concentrations in the microbial demand of N from the soil extractable fractions. digested samples and blanks were determined using an Optima 4300DV inductively coupled plasma/optical emission spectrometer (ICP-OES; 2.6. Soil CO2 efflux, water content and temperature PerkinElmer, Wellesley, USA). Microbial biomass C (Cmicro), N (Nmicro) and P (Pmicro) concentrations were calculated as the difference between Soil CO2 efflux (sresp) was measured at each season, prior to each the fumigated and unfumigated samples. Due to their roles in microbial soil sampling from 09:00 to 15:00 on five soil collars permanently mineralization, the elemental concentrations of K, Mn and Mo in the inserted into the soil (2 cm) of each plot using a closed system with a microbial biomasses (Kmicro, Mnmicro and Momicro, respectively) were SRC-1 soil chamber connected to an EGM-4 portable system (PP-sys­ also determined for the 0.5 M NaHCO3 extracts using chloroform tems, Hitchin, UK) as described by Zuccarini et al. (2020). Soil tem­ fumigation-extraction and ICP-OES analysis. All elemental concentra­ perature (stemp) was measured simultaneously to the soil CO2 efflux at tions in the microbial biomasses were expressed as μg of element per one point close to the collars at a depth of 10 cm using a TO 15 digital unit of soil dry mass. soil thermometer (Jules Richard Instruments, Argenteuil, France). We The activities of acid and alkaline phosphatase (acp and akp), determined swclab in the sieved soil subsamples by weighing before and β-glucosidase (bgl), protease (prot) and urease (ure) were determined after drying at 105 ◦ C for 24 h. colorimetrically, based on original methods developed by Tabatabai and Bremner (1969) (acp and akp), Hoffmann and Dedeken (1965) (bgl), 2.7. Data analysis Ladd and Butler (1972) (prot) and Kandeler and Gerber (1988) (ure). The procedures were adapted from Schinner et al. (1996) and Sardans The effects of season and the experimental chronic drought on et al. (2008b,c) with one major modification: after the addition of sub­ swclab, stemp, microbial elemental concentrations, soil extractable el­ strate and incubation, the soil suspension was centrifuged at 11200 g for ements, microbial and soil extractable stoichiometric ratios, potential 5 min (instead of filtering), and the clear supernatant was photometri­ soil extracellular enzymes activities and the stoichiometric imbalances cally compared to blank reagent (following the steps for color devel­ were analyzed using two-way ANOVAs, with season and treatment as opment if necessary). We then measured the optical density of the fixed effects and plot as a random factor (function lme in lme4 package calibration standards, samples and controls using a Helios Alpha spec­ in R v3.5.1; Bates et al., 2015; R Core Team, 2018). The data were log- or trophotometer (Thermo Spectronic, Cambridge, UK). Enzymatic activity square-root-transformed prior to analysis if distributions were skewed to was expressed as μmol of substrate released per gram of dry soil and improve the normality of residuals. The assumptions of normality and incubation time (h). homocedasticity were checked using Shapiro-Wilk and Levene tests, respectively. The existence of a temporal autocorrelation structure be­ 2.5. Soil extractable elements tween time points (seasons) in the model residuals was checked by including an autocorrelation structure in the models and comparing the The concentrations of soil extractable total organic C (EOC) and results of the mixed linear models with and without autocorrelation extractable total N (ETN) were measured in the unfumigated K2SO4 structure. It was not detected a significant temporal autocorrelation in extracts. The concentrations of extractable total P (Pt) and of potassium any of the variables and the mixed linear models without 3 S. Marañón-Jiménez et al. Soil Biology and Biochemistry 165 (2022) 108515 autocorrelation structure were the most parsimonious model. did not change seasonally. As a result, microbial C:N ratios increase We performed a principal component analysis (PCA) to visualize the during the dry seasons (summer and autumn), C:P ratios were also correlations among the soil, microbial and environmental variables lowest in spring and N:P ratios were highest in winter. using correlation matrices with standardized variables. Vector variables Cmicro, Nmicro and Pmicro were strongly positively correlated with were grouped by category (e.g. soil and microbial elements) using swclab, particularly Nmicro (Fig. 3). The correlation between Cmicro and different colors. Cases were separated by treatment and season, and the swclab varied throughout the seasons (P = 0.0051 for the swclab × effects of the chronic drought treatment and season were then tested for Season interaction). Cmicro was more strongly correlated with swclab the scores of the first two components (PC1 and PC2) of the PCAs using during the driest seasons of summer and autumn, with similar slopes, one-way analyses of variance (ANOVAs) and Tukey HSD post hoc tests whereas the correlations were weakest during the wetter seasons (winter (Jolliffe 2002; Quinn and Keough 2009). Pairwise Pearson correlations and spring, Fig. 3a). The response of Nmicro to swclab also varied between the soil, microbial and environmental variables were per­ seasonally (P = 0.007 for the swclab × Season interaction), although formed and represented in a heat map. Strongly correlated variables slopes did not differ so much between dry and wet periods (Fig. 3b). were clustered using the VARCLUS clustering algorithm that iteratively Pmicro was correlated with swclab similarly throughout the seasons (P = splits clusters of variables and reassigns the variables to clusters until no 0.5955 for the swclab × Season interaction, Fig. 3c). more splits are possible (SAS Institute Inc. 2018). The correlations be­ tween Cmicro, Nmicro, Pmicro and swclab for each season were further 3.3. Effect of chronic drought and seasonal drought on soil nutrients, explored using Pearson’s correlations. The effect of season on the re­ potential enzyme activities and microbial to soil stoichiometric imbalances lationships between Cmicro, Nmicro, Pmicro and swclab was tested using analyses of covariance (ANCOVAs), with season as a fixed factor and Six years after the experiment establishment, TOC had higher values swclab as the covariate. in control compared to chronic drought plots, while the opposite happened for TN, although differences were not statistically significant 3. Results (Table 1). The effect of chronic drought on the soil ammonium in the same soil samples changed across seasons. Ammonium decreased in 3.1. Effect of chronic drought and seasonal drought on soil microclimate response to chronic drought during the wettest seasons (winter and spring), whereas it increased during the driest seasons (summer and Swclab was consistently lower in the chronic drought than the con­ autumn). Soil nitrate had lower values in chronic drought plots trol treatment and was highest in winter and lowest in summer (Fig. 1a). compared to control plots during spring. The chronic drought treatment, however, caused higher reduction of ETN and Mn were significantly higher in the chronic drought than swclab in winter than in the other seasons. The chronic drought treat­ the control treatment (Table 1). EOC, ETN and Mn were consistently ment did not have main effects on soil temperature although there was a highest in summer in both treatments, whereas K and Mo were highest in significant TreatmentxSeason interaction (Fig. 1b). Soil temperature winter and Pt was lowest in autumn. Soil extractable ratios were only was highest in summer and lowest in winter. significantly affected by season, with highest C:N and C:P values in summer (Table 1) and lowest N:P ratios in winter. 3.2. Effect of chronic drought and seasonal drought on soil microbial Chronic drought decreased potential ure, prot and acp enzyme ac­ biomass and nutrients tivities and seasonal drought caused a declines in akp activities in autumn (Table 1). The chronic drought treatment caused a significant reduction of The stoichiometric imbalance between the soil extractable fractions Nmicro (42%) and in Pmicro (65%) in particular, and a marginally sig­ and microbial biomass was affected by seasonal but not by chronic nificant reduction of Cmicro (32%) (Fig. 2a–c). These changes, however, drought (Fig. 4). The C:N imbalance was lower during summer and were not reflected into significant alterations on microbial C:N, C:P and autumn, as a result of the lower C:N ratios in the extractable fraction N:P ratios (Fig. 2d–f). The effect of season on Cmicro and Nmicro was relative to the C:N ratios of microbial biomass during the drier seasons greater than the effect of the chronic drought treatment, where Cmicro (Fig. 4a). In contrast, the N:P imbalance increased from the wettest was lowest in spring and Nmicro was highest in winter. In contrast, Pmicro seasons (winter and spring) to the driest seasons (summer and autumn, Fig. 1. Soil microclimatic variables in the treatments and seasons measured in 2014 (16 years after the beginning of the experiment). Mean (±standard error) of A) gravimetric soil water content and B) soil temperature. Chronic drought samples are in brown, control samples are in green. Significant effects of treatment and season are indicated. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) 4 S. Marañón-Jiménez et al. Soil Biology and Biochemistry 165 (2022) 108515 Fig. 2. Elemental composition of microbial biomass in the treatments and seasons measured in 2014 (16 years after the beginning of the experiment). Mean (±standard error) of microbial A) carbon, B) nitrogen and C) phosphorus concentrations and microbial D) C:N, E) C:P and F) N:P molar ratios. Chronic drought samples are in brown, control samples are in green. Significant effects of treatment and season are indicated. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Fig. 3. Correlations between microbial C, N and P concentrations and soil water content measured in 2014 (16 years after the beginning of the experiment). P values and R2 coefficients for each season are indicated. Shaded areas are the 95% confidence intervals of the regression lines. Chronic drought samples are represented by filled dots, control samples are represented by empty dots. Different seasons are represented by different colors. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Fig. 4c), and the C:P imbalance was highest in spring (Fig. 4b). the chronic drought and control treatments along PC2 (P < 0.0001), with the highest scores and therefore extractable C:P and N:P ratios, EOC and ETN in the chronic drought treatment and in summer and the lowest 3.4. Multivariate effects of chronic drought and seasonal drought scores in the control treatment and in winter. Iterative variable clustering allowed us to identify groups of vari­ The first (PC1) and second (PC2) principal components of the PCA ables that were strongly correlated, explaining 70.1% of the total vari­ performed with the soil and microbial variables explained 26.6 and ation in the data (Fig. 6). N-acquiring enzymes, bgl, akp, Nmicro, Momicro 17.2% of the total variance, respectively (Fig. 5). The extracellular and Mnmicro were strongly positively correlated and constituted a well enzymatic activities akp, prot, ure, bgl, Nmicro and Pmicro contributed distinguished cluster (cluster 1). Soil ETN and the C:P and N:P ratios more to PC1, and the C:P and N:P ratios, EOC and ETN contributed more formed a second strongly correlated cluster of variables (cluster 2). A to PC2 (Fig. 5c). Samples differed between the chronic drought and third cluster was composed of Pt, Pi and Po. The microbial C:N ratios, control treatments along PC1 (P < 0.0001). The chronic drought sam­ swclab, stemp and sresp were also strongly correlated, although the ples had the lowest scores in PC1 and thus are associated with the lowest correlations were positive between microbial C:N and stemp and be­ extracellular enzymatic activities, Nmicro and Pmicro. The samples also tween sresp and swclab and negative between the other combinations differed seasonally along PC1 (P = 0.0002), with higher scores in winter (cluster 4). The microbial C:P and N:P ratios were also strongly and spring than in summer and autumn. Samples also differed between 5 S. Marañón-Jiménez et al. Soil Biology and Biochemistry 165 (2022) 108515 Table 1 Soil elemental concentrations and potential enzymatic activities in the control and experimental drought treatments across seasons. All variables were measured 16 years after the start of the experiment (2014), except TOC, TN, NH4+ and NO3− that were measured 6 years after the start of the experiment (2005). P values for the two-way ANOVAs are included in the last three columns. Significant differences at α = 0.05 are in bold. Values are means ± standard errors. Different letters indicate significant differences among seasons within treatments according to one-way ANOVAs, * indicates significant differences between treatments within seasons ac­ cording to one-way ANOVAs. Concentrations are in % for TOC and TN, μg g− 1 of dry soil for the rest of soil fractions and in μmol g− 1 of dry soil h− 1 for potential enzyme activities. TOC: Total organic carbon; TN: Total nitrogen; EOC: Extractable organic carbon; ETN: Extractable total nitrogen; Pt: Extractable total phosphorous; ure: urease activity; prot: protease activity; bgl: β-glucosidase activity; acp: acid phosphatase activity; akp: alkaline phosphatase activity. Variable Treatment Season P Winter Spring Summer Autumn Season Treatment Season × Treatment TOC Control 1.78 ± 0.10 – – – - 0.0822 – Drought 2.28 ± 0.13 – – – TN Control 0.25 ± 0.02 – – – - 0.1214 – Drought 0.18 ± 0.02 – – – TOC:TN Control 9.71 ± 1.18 – – – – 0.0002 – Drought 18.78 ± 2.21 – – – NH4+ Control 4.28 ± 0.27a* 3.30 ± 0.54a 0.96 ± 0.22b 2.89 ± 0.45a* <0.0001 0.6598 <0.0001 Drought 2.23 ± 0.13a* 2.28 ± 0.10a 2.01 ± 0.28a 4.92 ± 0.28b* NO3− Control 1.46 ± 0.72a 6.82 ± 1.83b* 1.34 ± 0.77a 2.39 ± 0.45a,b 0.0214 0.5935 0.0024 Drought 2.35 ± 0.17 1.73 ± 0.65* 1.28 ± 0.22 2.02 ± 0.36 EOC Control 308.25 ± 82.26a 441.87 ± 65.26a,b 646.22 ± 116.65b 439.49 ± 79.51a,b <0.0001 0.1423 0.9852 Drought 507.99 ± 96.92a 652.58 ± 110.00a,b 829.16 ± 108.56b 569.55 ± 85.30a,b ETN Control 32.02 ± 5.51 43.67 ± 5.27 53.92 ± 8.79 44.43 ± 7.35 0.0008 0.0475 0.7557 Drought 45.06 ± 7.49a 72.59 ± 10.87b 78.82 ± 8.95b 66.06 ± 9.30a,b Pt Control 57.40 ± 5.27a 57.31 ± 7.58a 57.56 ± 7.31a 39.52 ± 4.64b 0.0001 0.4440 0.4629 Drought 44.27 ± 6.33a,b 39.37 ± 6.14a,b 49.16 ± 6.73a 34.97 ± 6.57b C:N Control 9.93 ± 0.66a 11.59 ± 0.55a 13.72 ± 0.31b 11.36 ± 0.40a <0.0001 0.8865 0.0020 Drought 12.17 ± 0.72 11.63 ± 0.54 12.44 ± 0.52 10.78 ± 0.38 C:P Control 13.55 ± 2.61a 24.94 ± 3.97b 34.18 ± 4.72b 29.38 ± 4.69b <0.0001 0.2001 0.1266 Drought 33.42 ± 5.47a 58.06 ± 8.52b 53.99 ± 8.22b 43.00 ± 5.56b N:P Control 1.30 ± 0.16a 2.06 ± 0.27a,b 2.48 ± 0.33b 2.59 ± 0.37b <0.0001 0.1391 0.3833 Drought 2.58 ± 0.35a 4.83 ± 0.58b 4.38 ± 0.62b 4.33 ± 0.46b K Control 460.41 ± 39.83a 373.47 ± 42.37a,b 299.40 ± 17.47b 265.38 ± 23.21b <0.0001 0.0913 0.3733 Drought 337.85 ± 31.65 320.24 ± 20.44 279.73 ± 17.95 252.42 ± 26.05 Mn Control 1643.74 ± 383.98a,b 2648.34 ± 493.35b,c 3230.16 ± 525.37c 963.71 ± 141.72a <0.0001 0.0010 0.8877 Drought 3488.47 ± 704.78a,b 3617.99 ± 847.40a,b 4155.81 ± 272.14a 1978.87 ± 364.19b Mo Control 26.27 ± 2.05a 16.80 ± 0.81b 21.51 ± 1.11a 16.73 ± 0.78b <0.0001 0.4519 0.0702 Drought 22.71 ± 1.67 20.39 ± 1.37 22.20 ± 1.44 19.54 ± 1.32 ure Control 6.25 ± 0.40a* 5.83 ± 0.63a,b 4.85 ± 0.51a,b 4.40 ± 0.50b 0.0940 0.0009 0.2211 Drought 3.16 ± 0.30* 3.52 ± 0.29 3.05 ± 0.39 3.35 ± 0.43 prot Control 0.22 ± 0.03 0.27 ± 0.03 0.27 ± 0.02 0.24 ± 0.02 0.0602 0.0319 0.9591 Drought 0.17 ± 0.02 0.22 ± 0.02 0.21 ± 0.02 0.18 ± 0.02 bgl Control 11.57 ± 1.42a 9.76 ± 1.13a,b 8.53 ± 0.97a,b 7.28 ± 0.64b 0.2438 0.3230 0.0733 Drought 6.95 ± 0.81 8.01 ± 0.66 8.22 ± 0.98 7.09 ± 0.79 acp Control 30.13 ± 2.12* 27.88 ± 2.30 27.87 ± 1.59* 26.15 ± 1.94* 0.5755 0.0035 0.3875 Drought 16.56 ± 1.22* 17.97 ± 1.16 14.08 ± 2.02* 17.21 ± 1.15* akp Control 26.63 ± 2.63 24.23 ± 3.23 24.25 ± 1.62 19.09 ± 1.61 0.0442 0.1710 0.4507 Drought 18.79 ± 2.23 21.28 ± 2.23 17.99 ± 1.75 14.99 ± 1.40 Fig. 4. Stoichiometric imbalances between the soil extractable fractions and soil microbes measured in 2014 (16 years after the beginning of the experiment). Mean (±standard error) of A) C:N imbalance, B) C:P imbalance, C) N:P imbalance. Stoichiometric imbalances are calculated as the ratio of soil extractable stoichiometric ratios to the respective ratios of microbial biomass. Chronic drought samples are in brown, control samples are in green. Significant effects of treatment and season are indicated. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) 6 S. Marañón-Jiménez et al. Soil Biology and Biochemistry 165 (2022) 108515 Fig. 5. Two-dimensional representation of the areas defined by the first two principal components (PC1 and PC2) of the principal component analysis of season, treatment, microbial activities (soil respiration and enzymatic activities), soil extractable fractions, microbial elemental concentrations and soil microclimatic variables. All variables included were measured in 2014 (16 years after the beginning of the experiment). A) Values assigned to each soil sample for PC1 and PC2 grouped by treatment. B) Values grouped by season. C) Relative contribution of each variable to PC1 and PC2. Chronic drought samples are represented by filled dots, control samples are represented by empty dots. Different seasons are represented by different colors. Vector colors correspond to the variable categories. The percent contributions to the total variability are in parentheses. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Fig. 6. Heat map of pairwise correlation co­ efficients between the soil, microbial and environ­ mental variables measured in 2014 (16 years after the beginning of the experiment). The colors repre­ sent the coefficient of correlation between each pair of variables, where red represents a positive corre­ lation and blue represents a negative correlation. The brackets and numbers indicate clusters of strongly correlated variables. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) positively correlated (cluster 5). Cmicro was strongly correlated with consistent with previous studies of the effects of chronic drought on Pmicro, Kmicro and acp (cluster 6). Extractable K and Mo were also extractable N (He and Dijkstra 2014; Matías et al., 2011). This result, strongly correlated (cluster 7), as were EOC, C:N and Mn (cluster 8). and the consistently higher values of EOC (although not statistically significant) and TOC (marginally significant) in the chronic drought 4. Discussion plots compared to control plots (Table 1), are evidences of the accu­ mulation of organic matter in upper soil layers. The chronic conditions 4.1. Effect of chronic drought on the C:N:P stoichiometry of microbial of low water availability may have impaired substrate mobility and biomass and soil accessibility (Borken and Matzner 2009; Moyano et al., 2013), thus reducing the release of enzymes from soil microbes. Persistent decreases Chronic drought altered soil N availability and the capacity of soil in the potential mineralization of N and P and the depolymerization microbial community (as a whole) to retain N and P. Soil ETN increased activities of soil microbes (decrease in prot, ure and acp in our study; in the chronic drought treatment (Table 1), particularly in summer, Table 1) may also slow the rates of SOM decomposition and the capacity 7 S. Marañón-Jiménez et al. Soil Biology and Biochemistry 165 (2022) 108515 of microbes to uptake N and P (Dijkstra et al., 2015). The impaired drought) that are generally released later during periods of growth microbial SOM mineralization and substrate accessibility under chronic (Blackwell et al., 2010; Schaeffer et al., 2017), although its contribution drought most likely caused the significant and persistent reductions in has not been quantified in Mediterranean ecosystems. A lower capacity microbial N and particularly microbial P and the marginal reductions of of SOM decomposition and nutrient retention in the microbial biomass microbial C in chronic drought plots across all seasons (Fig. 2), sup­ pool during unfavorable periods for plant growth (Fig. 2), together with porting first and second hypotheses. These data, together with previous an increasing frequency of torrential and irregular rains and higher rates results from the same site (Asensio et al., 2021), provide clear evidences of erosion, can lead to the reduction of assimilable soil nutrients for of restricted microbial activity and N and P retention in the microbial plants during the plant growing season (Table 1, Bloor and Bardgett, biomass pool under chronic drought and of the accumulation of organic 2012; Peñuelas et al., 2018). Mediterranean ecosystems are exposed to matter in the soil. frequent and short cycles of drying and rewetting. Soil rewetting rapidly Despite the evidences of lower SOM mineralization under chronic activates soil microbes and the mineralization of organic compounds drought, soil ammonium accumulated during the driest seasons (sum­ accumulated during seasonal drought, releasing nutrients to the soil mer and autumn) (Table 1). The lowest N stored in the microbial solution (Barnard et al., 2020; Borken and Matzner 2009). Microbes biomass pool, together with the lower plant biomass growth and N later activate the synthesis of proteins and growth (Barnard et al., 2013), storage under chronic drought in the same sites (Sardans et al. 2008a, which immobilizes microbial nutrients (Gao et al., 2016; Yevdokimov 2020), may explain the accumulation of NH4+ in chronic drought plots et al., 2016). Plant uptake generally takes longer to begin after soil is during the driest seasons. By contrast, chronic drought caused a reduc­ rewetted (Wraith et al., 1995), so microbial nutrient immobilization can tion of soil NH4+ during the wettest seasons (spring and winter) and of represent a buffering mechanism to ensure the availability of nutrients NO3− during spring, the period of highest plant growth in this Medi­ to plants during periods of growth (Bardgett et al., 2005; Qiu et al., terranean ecosystem (Ogaya and Peñuelas, 2004). The lower amount of 2009). Long-term drought, however, may reduce the N and P stored in mineralized N released from microbial biomass probably contributed the microbial biomass pool and the supply of mineralized forms of these more to the lower availability of NH4+ and NO3− during spring than the nutrients during favorable periods for vegetation. increase in plant N uptake, since total vegetation N storage was lower Strategies of the soil microbes to overcome drought may have also under chronic drought conditions compared to control plots (Sardans contributed to the smaller effects of chronic drought on microbial C and et al. 2008a, 2020). Moreover, a recent study in the same sites shows N compared to microbial P. Chronic drought requires the development that both plant and soil have lost substantial amounts of N under chronic of microbial adaptations, such as the accumulation of osmolytes to drought conditions compared to control plots (Sardans et al., 2020). An overcome water stress and the energy limitation imposed by low ac­ impaired capacity of the microbial biomass pool to retain N due to cessibilities of organic substrates (Schimel et al., 2007, Gutiérrez-Girón chronic drought conditions may therefore reduce plant N availability et al., 2014) or the creation of resistance structures (e.g. cell-wall pep­ during the growing period and increase the risks of N losses from the tidoglycans) to resist osmotic stress (Marchus et al., 2018; Yadav et al., ecosystem. 2018). Specifically, bacteria typically use C- and N-rich amino com­ Unlike N, extractable P compounds did not accumulate in the soil, pounds as primary osmolytes (Csonka 1989), whereas fungi use C-rich but even decreased (although not significantly) (Table 1), consistent polyols (Witteveen and Visser 1995) that do not contain N. Soil microbes with consistent with previous results showing lower potential phos­ can thus increase their total cytoplasmic constituents under conditions phatase enzymatic activity (Table 1, Sardans and Peñuelas 2005; of extreme water stress, by as much as 30–40% of total C for both bac­ Asensio et al., 2021). Phosphorous generally has low mobility in Medi­ teria and fungi and by nearly 60 and 20% of total N for bacteria and terranean soils, where it can precipitate with carbonates, mainly at the fungi, respectively (Schimel et al., 1989). Microbes also have additional early and intermediate stages of soil chronosequences, and with Fe/Al–P C costs to provide energy for the synthesis of osmolytes, which can minerals in older soils (Carreira et al., 1997; Sardans and Peñuelas represent two-to three-fold the C cost for the simple accumulation of C 2013). The accumulation of SOM under chronic drought may also in­ compounds (Schimel et al., 2007). These increasing C demands of soil crease the adsorption capacity of soil and further reduce the mobility of microbes under chronic drought conditions were also indicated by the P in these soils (White 1981; Missong et al., 2018). However, despite enzymatic activities, with a higher production of C-catalyzing enzymes total soil P did not decrease significantly in response to chronic drought, relative to P- and N-catalyzing enzymes (Asensio et al., 2021), pointing microbial P decreased drastically compared to microbial N and C to an increase in microbial C limitation. The need to develop mecha­ (Fig. 2c), leading to consistent but not significant trends toward higher nisms of drought resistance together with the lower mobility and microbial C:P and N:P ratios across all seasons (Fig. 2e and f). Physio­ accessibility of organic substrates in the chronic drought treatment logical adaptations of the microbial community to save P under the probably increased the relative demand of C-catalyzing enzymes and the conditions of the low P availability exacerbated by chronic drought may C costs of acquiring resources. The equivalent accumulation of C- and account for the drastic decrease in microbial P (Merchant and Helmann, N-rich compounds as a mechanism to resist drought stress may account 2012). Recent evidence has shown that microorganisms can minimize for the lack of effect of chronic drought on the microbial C:N ratios and their P requirements when soil P availability is low by switching from may have contributed to the lower sensitivity of microbial C and N to phospholipids to betaine lipids when building biomass (Warren 2020). chronic drought compared to microbial P. This mechanism could account for a decrease of around 10% in the amount of P required for microbial growth. Other P sparing and recy­ 4.2. Effect of seasonal drought on the C:N:P stoichiometry of microbial cling strategies, such as the increase in the ribosome and phospholipids biomass and soil degradation (Merchant and Helmann, 2012) or shifts to microbial communities towards more efficient P recycling have been described Unlike chronic drought, seasonal drought caused marked but tran­ (Bergkemper et al., 2016), although still poorly investigated in soils that sient fluctuations in the C:N:P stoichiometry of the soil microorganisms undergo chronic drought. The generally low availability of soil P in among seasons (Fig. 2d–f). Microbial C:N ratios increased during the dry Mediterranean soils may therefore be exacerbated by chronic drought seasons to much higher values (C:N = 11.1–14.6) than those reported for conditions, causing physiological adaptations or community shifts in temperate forests (C:N = 8.6) and global averages (C:N = 7.6) (Xu et al., soil microbes to save P and imbalances in soil microbial P stoichiometry. 2013). Seasonal shifts toward a microbial community more tolerant to The role of microbial biomass in retaining N and P in the soil is drought stress, with the increasing dominance of fungi compared to important in Mediterranean ecosystems (Qiu et al., 2009; Aponte et al., bacteria (Strickland and Rousk 2010; Preece et al., 2019), together with 2010), because microbial biomass might represent a reservoir of nutri­ the primary accumulation of C-rich osmolytes in fungi and the synthesis ents for vegetation during periods of lower activity (e.g. summer of resistance structures to cope with water stress (Kakumanu et al., 8 S. Marañón-Jiménez et al. Soil Biology and Biochemistry 165 (2022) 108515 2019), may have contributed to the higher microbial C:N ratios and the work reported in this paper. lower respiration rates during periods when water was limited. This was also indicated by the cluster of strong correlations among microbial C:N, Acknowledgments microbial respiration and soil water content (Fig. 6). In contrast, soil extractable C:N ratios (C:N = 11.0–13.1) were lower than for temperate This research was supported by the Spanish Government grant forests (C:N = 18.7) and globally (C:N = 16.4) (Xu et al., 2013). The PID2019-110521GB-I00. The European Research Council Synergy grant increase in microbial C:N during the drier seasons consequently lowered ERC-SyG-2013-610028 IMBALANCE-P, and the Catalan Government the C:N imbalance between the extractable soil fractions and the mi­ grant SGR 2017-1005. SMJ was supported by the Serra Hunter fellow­ crobes (Fig. 4a), representing an increase in the microbial C limitation in ship Programme from the regional government of Catalonia. summer and autumn. An increase in C limitation of soil microbes also impairs the capacity of the microbial community to retain N (Mar­ Appendix A. Supplementary data añón-Jiménez et al., 2019), as it is shown by the lower pool of microbial N during the driest seasons (Fig. 2), which may contribute to a lower Supplementary data to this article can be found online at https://doi. availability of mineralized forms of N during the vegetation growing org/10.1016/j.soilbio.2021.108515. season (Table 1). The progressive increase in the N:P imbalance from the wetter to the References drier seasons also indicated an increase in the microbial P limitation relative to N (Fig. 4c). This finding is consistent with previous evidence Abbasi, A.O., Salazar, A., Oh, Y., Reinsch, S., del Rosario Uribe, M., Li, J., Rashid, I., of the higher production of P- versus N-catalyzing enzymes during sea­ Dukes, J.S., 2020. Reviews and syntheses: soil responses to manipulated precipitation changes – an assessment of meta-analyses. Biogeosciences 17 (14), sonal drought at the same study site (Asensio et al., 2021). Given the 3859–3873. lower mobility and solubility of soil P compared to soil N (He and Aponte, C., Marañón, T., García, L.V., 2010. Microbial C, N and P in soils of Dijkstra 2014), decreases in the availability of soil water may lead to Mediterranean oak forests: influence of season, canopy cover and soil depth. Biogeochemistry 101, 77–92. larger restrictions in microbial P uptake compared to microbial N uptake Asensio, D., Zuccarini, P., Ogaya, R., Marañón-Jiménez, S., Sardans, J., Peñuelas, J., and stimulate more the microbial production of P-catalyzing enzymes 2021. Soil enzyme stoichiometry of Mediterranean forest soils submitted to climate than the production of N-catalyzing enzymes. This Mediterranean forest change-simulated and seasonal drought indicate a rise in carbon and phosphorus demand. Soil Biology and Biochemistry 163, 108424. indeed had two distinct periods of water (Fig. 3) and nutrient (Fig. 4) Bardgett, R., Bowman, W., Kaufmann, R., Schmidt, S., 2005. A temporal approach to limitations for soil microbes. Microbial biomass depended greatly on the linking aboveground and belowground ecology. Trends in Ecology & Evolution 20, availability of water and on the accessibility of C and P (Figs. 3a and 4a 634–641. and c) during the dry periods (summer and autumn), whereas microbes Bardgett, R., Manning, P., Morriën, E., Vries, F., 2013. Hierarchical responses of plant–soil interactions to climate change: consequences for the global carbon cycle. shifted from C to N limitation during the wet periods (spring and winter) Journal of Ecology 101, 334–343. (Fig. 4a). Barnard, R., Osborne, C., Firestone, M., 2013. Responses of soil bacterial and fungal communities to extreme desiccation and rewetting. ISME J 7, 2229–2241. Barnard, R.L., Blazewicz, S.J., Firestone, M.K., 2020. Rewetting of soil: revisiting the 5. Conclusions origin of soil CO2 emissions. Soil Biology and Biochemistry 147, 107819. Bates, D., Mächler, M., Bolker, B., Walker, S., 2015. Fitting linear mixed-effects models Both seasonal and long-term periods of chronic water limitation using lme4. Journal of Statistical Software 67 (1), 1–48. Beier, C., Emmett, B.A., Tietema, A., Schmidt, I.K., Peñuelas, J., Lang, E.K., Duce, P., De caused reductions in the size of the N and P pool stored in microbial Angelis, P., Gorissen, A., Estiarte, M., de Pato, G.D., Sowerby, A., Kröel-Dulay, G., biomass, partially supporting our first hypothesis. The lower minerali­ Lellei-Kovacs, E., Kull, O., Mand, P., Petersen, H., Gjelstrup, P., Spano, D., 2009. zation activity under both seasonal and chronic water limitation led to Carbon and nitrogen balances for six shrublands across Europe. Global Biogeochem Cycles 23, GB4008. the accumulation of C- and N-rich organic compounds in the upper soil Bergkemper, F., Schöler, A., Engel, M., Lang, F., Krüger, J., Schloter, M., Schulz, S., 2016. layers. The responses of microbial biomass C and nutrients, however, Phosphorus depletion in forest soils shapes bacterial communities towards differed between the seasonal and chronic droughts. Chronic drought phosphorus recycling systems. Environ Microbiol 18 (6), 1988–2000. Blackwell, M.S.A., Brookes, P.C., de la Fuente-Martinez, N., Gordon, H., Murray, P.J., reduced microbial biomass C, N and particularly P substantially and Snars, K.E., Williams, J.K., Bol, R., Haygarth, P.M., 2010. Chapter 1 - phosphorus persistently across seasons, in agreement to our second hypothesis. The solubilization and potential transfer to surface waters from the soil microbial lower N pool in microbial biomass most likely was accompanied by a biomass following drying–rewetting and freezing–thawing. In: Sparks, D.L. (Ed.), lower availability of mineralized forms of N during the vegetation Advances in Agronomy, vol. 106. Academic Press, pp. 1–35. Bloor, J.M.G., Bardgett, R.D., 2012. Stability of above-ground and below-ground growing season (spring), in agreement to our third hypothesis. Chronic processes to extreme drought in model grassland ecosystems: interactions with plant drought may exacerbate the low mobility of P in soil due to its adsorp­ species diversity and soil nitrogen availability. Perspectives in Plant Ecology, tion in accumulated SOM. The low soil P availability probably triggered Evolution and Systematics 14, 193–204. Borken, W., Matzner, E., 2009. Reappraisal of drying and wetting effects on C and N microbial community shifts or physiological adaptations to spare P, mineralization and fluxes in soils. Global Change Biology 15, 808–824. which reduced particularly the size of the microbial P pool stored. Carreira, J.A., Laftha, K., Niell, F.X., 1997. Phosphorus transformations along a soil/ Nonetheless, microbial C:N ratios remained unaltered by chronic vegetation series of fine-prone, dolomitic, semi-arid shrublands of southern Spain soil P and Mediterranean shrubland dynamics. Biogeochemistry 39, 87–120. drought, likely associated with the equivalent accumulation of C- and N- Csonka, L.N., 1989. Physiological and genetic responses of bacteria to osmotic stress. rich osmolytes by microbes. In contrast, soil microbes experienced Microbiological Reviews 53, 121–147. transitory changes in the C:N:P stoichiometry in response to seasonal Deng, L., Peng, C., Kim, D.-G., Li, J., Liu, Y., Hai, X., Liu, Q., Huang, C., Shangguan, Z., Kuzyakov, Y., 2021. Drought effects on soil carbon and nitrogen dynamics in global drought, increasing their C content relative to N content. Therefore, natural ecosystems. Earth-Science Reviews 214, 103501. while microbial P is more sensible to chronic water stress than microbial Díaz-Raviña, M., Acea, M.J., Carballas, T., 1995. Seasonal changes in microbial biomass N and C, the last are more closely coupled to seasonal fluctuations of and nutrient flush in forest soils. Biology and Fertility of Soils 19, 220–226. Dijkstra, F.A., He, M., Johansen, M.P., Harrison, J.J., Keitel, C., 2015. Plant and water availability. These findings are consistent with the idea that mi­ microbial uptake of nitrogen and phosphorus affected by drought using 15N and 32P crobial adaptations to seasonal drought mainly involve changes in mi­ tracers. Soil Biology and Biochemistry 82, 135–142. crobial C and N contents, whereas chronic water limitation drastically Fierer, N., Schimel, J.P., 2002. Effects of drying–rewetting frequency on soil carbon and and persistently impairs the P stored in the microbial biomass pool. nitrogen transformations. Soil Biology and Biochemistry 34, 777–787. Gao, J., Feng, J., Zhang, X., Yu, F.-H., Xu, X., Kuzyakov, Y., 2016. Drying-rewetting cycles alter carbon and nitrogen mineralization in litter-amended alpine wetland Declaration of competing interest soil. Catena 145, 285–290. García-Angulo, D., Hereş, A.M., Fernández-López, M., Flores, O., Sanz, M.J., Rey, A., Valladares, F., Curiel Yuste, J., 2020. Holm oak decline and mortality exacerbates The authors declare that they have no known competing financial drought effects on soil biogeochemical cycling and soil microbial communities across interests or personal relationships that could have appeared to influence a climatic gradient. Soil Biology and Biochemistry 149, 107921. 9 S. Marañón-Jiménez et al. Soil Biology and Biochemistry 165 (2022) 108515 Gutiérrez-Girón, A., Rubio, A., Gavilán, R.G., 2014. Temporal variation in microbial and Padilla, F.M., Pugnaire, F.I., 2007. Rooting depth and soil moisture control plant biomass during summer in a Mediterranean high-mountain dry grassland. Mediterranean woody seedling survival during drought. Funct Ecol 21, 489–495. Plant and Soil 374, 803–813. Peñuelas, J., Sardans, J., Filella, I., Estiarte, M., Llusià, J., Ogaya, R., Carnicer, J., Hartmann, D.L., et al., 2013. In climate change 2013: the physical science basis. In: Bartrons, M., Rivas-Ubach, A., Grau, O., Peguero, G., Margalef, O., Pla-Rabés, S., Stocker, T.F., et al. (Eds.), Ch. 2 (IPCC. Cambridge Univ. Press. Stefanescu, C., Asensio, D., Preece, C., Liu, L., Verger, A., Rico, L., Barbeta, A., He, M., Dijkstra, F.A., 2014. Drought effect on plant nitrogen and phosphorus: a meta- Achotegui-Castells, A., Gargallo-Garriga, A., Sperlich, D., Farré-Armengol, G., analysis. New Phytologist 204, 924–931. https://doi.org/10.1111/nph.12952. Fernández-Martínez, M., Liu, D., Zhang, C., Urbina, I., Camino, M., Vives, M., Nadal- Hoffmann, G., Dedeken, M., 1965. Eine Methode zur colorimetrischen Bestimmung der Sala, D., Sabaté, S., Gracia, C., Terradas, J., 2018. Assessment of the impacts of β-Glucosidase-Aktivität in Böden. Zeitschrift für Pflanzenernährung, Düngung, climate change on Mediterranean terrestrial ecosystems based on data from field Bodenkunde 108 (3), 193–198. https://doi.org/10.1002/jpln.19651080302. experiments and long-term monitored field gradients in Catalonia. Environmental Hoover, D.L., Rogers, B.M., 2016. Not all droughts are created equal: the impacts of and Experimental Botany 152, 49–59. interannual drought pattern and magnitude on grassland carbon cycling. Global Preece, C., Verbruggen, E., Liu, L., Weedon, J.T., Peñuelas, J., 2019. Effects of past and Change Biology 22, 1809–1820. current drought on the composition and diversity of soil microbial communities. Soil Jenkinson, D.S., Powlson, D.S., 1976. Effects of biocidal treatments on metabolism in Biology and Biochemistry 131, 28–39, 2019. soil. 5. Method for measuring soil biomass. Soil Biology and Biochemistry 8, Qiu, S., Bell, R., McComb, A., 2009. Soil microbial biomass and sustainable farming- 209–213. enhancing nutrient retention in soils through management of microbial biomass. In: Jolliffe, I.T., 2002. Principal Components Analysis. In: Springer Series in Statistics. Lucero, D.P., Boggs, J.E. (Eds.), Soil Fertility. Nova Science Publishers, New York, Springer, New York. pp. 93–102. Kakumanu, M.L., Ma, L., Williams, M.A., 2019. Drought-induced soil microbial amino Quinn, G.P., Keough, M.J., 2009. Experimental Design and Data Analysis for Biologists. acid and polysaccharide change and their implications for C-N cycles in a climate Cambridge University Press, Cambridge. change world. Scientific Reports 9, 10968. Core Team, R., 2018. R: a language and environment for statistical computing. Taken Kandeler, E., Gerber, H., 1988. Short-term assay of soil urease activity using colorimetric from. https://www.r-project.org/. determination of ammonium. Biology and Fertility of Soils 6 (1), 68–72. https://doi. Rousk, J., Smith, A.R., Jones, D.L., 2013. Investigating the long-term legacy of drought org/10.1007/BF00257924. and warming on the soil microbial community across five European shrubland Kreuzwieser, J., Gessler, A., 2010. Global climate change and tree nutrition: influence of ecosystems. Global Change Biology 19, 3872–3884. water availability. Tree Physiology 30, 1221–1234. Santonja, M., Fernandez, C., Proffit, M., Gers, C., Gauquelin, T., Reiter, I.M., Cramer, W., Ladd, J., Butler, J., 1972. Short-term assays of soil proteolytic enzyme activities using Baldy, V., 2017. Plant litter mixture partly mitigates the negative effects of extended proteins and dipeptide derivatives as substrates. Soil Biology and Biochemistry 4 (1), drought on soil biota and litter decomposition in a Mediterranean oak forest. Journal 19–30. https://doi.org/10.1016/0038-0717(72)90038-7. of Ecology 105, 801–815. Li, X.Y., Contreras, S., Solé-Benet, A., Cantón, Y., Domingo, F., Lázaro, R., Lin, H., Van Sardans, J., Peñuelas, J., 2005. Drought decreases soil enzyme activity in a Wesemael, B., Puigdefábregas, J., 2011. Controls of infiltration-runoff processes in Mediterranean Quercus ilex L. forest. Soil Biology and Biochemistry 37, 455–461. Mediterranean karst rangelands in SE Spain. Catena 86, 98–109. Sardans, J., Peñuelas, J., 2007. Drought changes phosphorus and potassium Liang, C., Schimel, J.P., Jastrow, J.D., 2017. The importance of anabolism in microbial accumulation patterns in an evergreen Mediterranean forest. Functional Ecology 21, control over soil carbon storage. Nature Microbiology 2, 17105. 191–201. Lloret, F., Granzow-de la Cerda, I., 2013. Plant competition and facilitation after extreme Sardans, J., Peñuelas, J., 2010. Soil enzyme activity in a mediterranean forest after six drought episodes in Mediterranean shrubland: does damage to vegetation cover years of drought. Soil Science Society of America Journal 74, 838–851. trigger replacement by juniper woodland? Journal of Vegetation Science 24, Sardans, J., Peñuelas, J., 2012. The role of plants in the effects of global change on 1020–1032. nutrient availability and stoichiometry in the plant-soil system. Plant Physiology Manzoni, S., Schimel, J.P., Porporato, A., 2012. Responses of soil microbial communities 160, 1741–1761. to water stress: results from a meta-analysis. Ecology 93, 930–938. Sardans, J., Peñuelas, J., 2013. Plant-soil interactions in Mediterranean forest and Marañón-Jiménez, S., Castro, J., Kowalski, A.S., Serrano-Ortiz, P., Reverter, B.R., shrublands: impacts of climatic change. Plant and Soil 365, 1–33. Sánchez-Cañete, E.P., Zamora, R., 2011. Post-fire soil respiration in relation to burnt Sardans, J., Peñuelas, J., Ogaya, R., 2008a. Drought-Induced changes in C and N wood management in a Mediterranean mountain ecosystem. Forest Ecology and stoichiometry in a Quercus ilex mediterranean forest. Forest Science 54, 513–522. Management 261, 1436–1447. Sardans, J., Penuelas, J., Estiarte, M., 2008b. Changes in soil enzymes related to C and N Marañón-Jiménez, S., Peñuelas, J., Richter, A., Sigurdsson, B.D., Fuchslueger, L., cycle and in soil C and N content under prolonged warming and drought in a Leblans, N.I.W., Janssens, I.A., 2019. Coupled carbon and nitrogen losses in response Mediterranean shrubland. Applied Soil Ecology 39, 223–235. to seven years of chronic warming in subarctic soils. Soil Biology and Biochemistry Sardans, J., Urbina, I., Grau, O., Asensio, D., Ogaya, R., Peñuelas, J., 2020. Long-term 134, 152–161. drought decreases ecosystem C and nutrient storage in a Mediterranean holm oak Marchus, K.A., Blankinship, J.C., Schimel, J.P., 2018. Environmental controls on forest. Environmental and Experimental Botany 177, 104135. extracellular polysaccharide accumulation in a California grassland soil. Soil Biology SAS Institute Inc, 2018. “The VARCLUS Procedure.” SAS/STAT 15.1 User’s Guide. SAS and Biochemistry 125, 86–92. Institute Inc, Cary, NC. https://support.sas.com/documentation/onlinedoc/sta Martí-Roure, M., Casals, P., Romanyà, J., 2011. Temporal changes in soil organic C under t/151/varclus.pdf. Mediterranean shrublands and grasslands: impact of fire and drought. Plant Soil 338, Schaeffer, S.M., Homyak, P.M., Boot, C.M., Roux-Michollet, D., Schimel, J.P., 2017. Soil 289–300. carbon and nitrogen dynamics throughout the summer drought in a California Matías, L., Castro, J., Zamora, R., 2011. Soil-nutrient availability under a global-change annual grassland. Soil Biology and Biochemistry 115, 54–62. scenario in a Mediterranean mountain ecosystem. Global Change Biology 17, Schimel, J.P., 2018. Life in dry soils: effects of drought on soil microbial communities 1646–1657. https://doi.org/10.1111/j.1365-2486.2010.02338.x. and processes. Annual Review of Ecology, Evolution, and Systematics 49, 409–432. Merchant, S.S., Helmann, J.D., 2012. Chapter 2 - elemental economy: microbial Schimel, J.P., Scott, W.J., Killham, K., 1989. Changes in cytoplasmic carbon and nitrogen strategies for optimizing growth in the face of nutrient limitation. In: Poole, R.K. pools in a soil bacterium and a fungus in response to salt stress. Applied and (Ed.), Advances in Microbial Physiology. Academic Press, pp. 91–210. Environmental Microbiology 55, 1635–1637. Miltner, A., Bombach, P., Schmidt-Brücken, B., Kästner, M., 2012. SOM genesis: Schimel, J.P., Balser, T.C., Wallenstein, M., 2007. Microbial stress-response physiology microbial biomass as a significant source. Biogeochemistry 111, 41–55. and its implications for ecosystem function. Ecology 86, 1386–1394. Missong, A., Holzmann, S., Bol, R., Nischwitz, V., Puhlmann, H., Wilpert, K.v., Schinner, F., Öhlinger, R., Kandeler, E., Margesin, R., 1996. In: Schinner, F., Öhlinger, R., Siemens, J., Klumpp, E., 2018. Leaching of natural colloids from forest topsoils and Kandeler, E., Margesin, R. (Eds.), Methods in Soil Biology. Springer, Berlin. their relevance for phosphorus mobility. Science of the Total Environment 634, Strickland, M.S., Rousk, J., 2010. Considering fungal:bacterial dominance in soils – 305–315. methods, controls, and ecosystem implications. Soil Biology and Biochemistry 42, Mooshammer, M., Wanek, W., Zechmeister-Boltenstern, S., Richter, A., 2014. 1385–1395. Stoichiometric imbalances between terrestrial decomposer communities and their Tabatabai, M., Bremner, J., 1969. Use of p-nitrophenyl phosphate for assay of soil resources: mechanisms and implications of microbial adaptations to their resources. phosphatase activity. Soil Biology and Biochemistry 1 (4), 301–307. Frontiers in Microbiology 5. Tiemann, L.K., Billings, S.A., 2011. Changes in variability of soil moisture alter microbial Moyano, F.E., Manzoni, S., Chenu, C., 2013. Responses of soil heterotrophic respiration community C and N resource use. Soil Biology and Biochemistry 43, 1837–1847. to moisture availability: an exploration of processes and models. Soil Biology and US Environmental Protection Agency (USEPA), 1979. Methods for Chemical Analysis of Biochemistry 59, 72–85. Water and Wastes, Method 353.3. US Environmental Protection Agency, Navarro-García, F., Casermeiro, M.A., Schimel, J.P., 2012. When structure means Washington, DC. conservation: effect of aggregate structure in controlling microbial responses to Vicente-Serrano, S.M., Gouveia, C., Camarero, J.J., Beguería, S., Trigo, R., López- rewetting events. Soil Biol Biochem 44, 1–8. Moreno, J.I., Azorín-Molina, C., Pasho, E., Lorenzo-Lacruz, J., Revuelto, J., Morán- Ogaya, R., Peñuelas, J., 2004. Phenological patterns of Quercus ilex, Phillyrea latifolia, Tejeda, E., Sanchez-Lorenzo, A., 2013. Response of vegetation to drought time-scales and Arbutus unedo growing under a field experimental drought. Écoscience 11, across global land biomes. Proceedings of the National Academy of Sciences 110, 263–270. 52–57. Ogaya, R., Peñuelas, J., 2007. Tree growth, mortality, and above-ground biomass Vicente-Serrano, S.M., et al., 2014. Evidence of increasing drought severity caused by accumulation in a holm oak forest under a five-year experimental field drought. temperature rise in southern Europe. Environ. Res. Lett. 9, 044001. Plant Ecol 189, 291–299. Walkley, A., Black, I.A., 1934. An examination of Degtjareff method for determination Ogaya, R., Liu, D., Barbeta, A., Peñuelas, J., 2020. Stem mortality and forest dieback in a soil organic matter and a proposed modification of the chromic acid titration 20-years experimental drought in a mediterranean holm oak forest. Frontiers in method. Soil Sci 37, 29–37. Forests and Global Change 2. Warren, C.R., 2020. Soil microbial populations substitute phospholipids with betaine lipids in response to low P availability. Soil Biology and Biochemistry 140, 107655. 10 S. Marañón-Jiménez et al. Soil Biology and Biochemistry 165 (2022) 108515 Watanabe, F.S., Olsen, S.R., 1965. Test of an ascorbic acid method for determining Yaalon, D.H., 1997. Soils in the Mediterranean region: what makes them different? phosphorous in water and NaHCO3 extracts from soil. Soil Sci. Soc. Am. Proc. 29, Catena 28, 157–169. 677–678. Yadav, A.K., Espaillat, A., Cava, F., 2018. Bacterial strategies to preserve cell wall White, R., 1981. Retention and release of phosphate by soils and soil constituents. In: integrity against environmental threats. Frontiers in Microbiology 9. Tinker, P.B. (Ed.), Soils and Agriculture. John Wiley and Sons, pp. 71–114. evdokimov, I.Y., Larionova, A., Blagodatskaya, E., 2016. Microbial immobilisation of Witteveen, C.F.B., Visser, J., 1995. Polyol pools in Aspergillus Niger. FEMS Microbiology phosphorus in soils exposed to drying-rewetting and freeze-thawing cycles. Biology Letters 134, 57–62. and Fertility of Soils 52, 685–696. Wraith, J.M., Baker, J.M., Blake, T.K., 1995. Water uptake resumption following soil Zaimes, G.N., Emmanouloudis, D., Iakovoglou, V., 2012. Estimating soil erosion in drought: a comparison among four barley genotypes. Journal of Experimental Natura 2000 areas located on three semi-arid Mediterranean islands. J Environ Biol Botany 46, 873–880. 33, 277–282. Xu, X., Thornton, P.E., Post, W.M., 2013. A global analysis of soil microbial biomass Zuccarini, P., Asensio, D., Ogaya, R., Sardans, J., Peñuelas, J., 2020. Effects of seasonal carbon, nitrogen and phosphorus in terrestrial ecosystems. Global Ecology and and decadal warming on soil enzymatic activity in a P-deficient Mediterranean Biogeography 22, 737–749. shrubland. Global Change Biology 26, 3698–3714. 11