Advertisement

Plant and Soil

, Volume 424, Issue 1–2, pp 233–254 | Cite as

Soil legacy effects of climatic stress, management and plant functional composition on microbial communities influence the response of Lolium perenne to a new drought event

  • Nicolas LegayEmail author
  • Gabin Piton
  • Cindy Arnoldi
  • Lionel Bernard
  • Marie-Noëlle Binet
  • Bello Mouhamadou
  • Thomas Pommier
  • Sandra Lavorel
  • Arnaud Foulquier
  • Jean-Christophe Clément
Regular Article

Abstract

Background and aims

Drought events, agricultural practices and plant communities influence microbial and soil abiotic parameters which can feedback to fodder production. This study aimed to determine which soil legacies influence plant biomass production and nutritional quality, and its resistance and recovery to extreme weather events.

Methods

In a greenhouse experiment, soil legacy effects on Lolium perenne were examined, first under optimal conditions, and subsequently during and after drought. We used subalpine grassland soils previously cultivated for two years with grass communities of distinct functional composition, and subjected to combinations of climatic stress and simulated management.

Results

The soil legacy of climatic stress increased biomass production of Lolium perenne and its resistance and recovery to a new drought. This beneficial effect resulted from higher nutrient availability in soils previously exposed to climatic stresses due to lower competitive abilities and resistance of microbial communities to a new drought. This negative effect on microbial communities was strongest in soils from previously cut and fertilized grasslands or dominated by conservative grasses.

Conclusion

In subalpine grasslands more frequent climatic stresses could benefit fodder production in the short term, but threaten ecosystem functioning and the maintenance of traditional agricultural practices in the long term.

Keywords

Plant functional traits Lolium perenne Extracellular enzymatic activities Mycorrhizae Climatic stress Soil legacy 

Introduction

Mountain climates are characterized by strong seasonal variations subjecting high-elevation ecosystems to recurrent stressful abiotic conditions (Körner 1999). These climatic conditions associated with traditional agricultural practices influence the functioning of subalpine grasslands and the ecosystem services they provide (Lavorel et al. 2011). For example, mowing and fertilization increase nutrient availability in these nutrient-poor soils, and can stimulate nutrient cycling (Robson et al. 2007), fodder production and quality (Lavorel et al. 2011), and soil carbon sequestration (Fornara et al. 2016). Changes in climatic regimes are predicted to induce more frequent and intense weather extremes such as repeated and prolonged drought periods with less frequent but more intense rainfall (Della-Marta et al. 2007; IPCC 2014). Reduced winter precipitation combined with warming are also predicted to reduce snow cover and to promote earlier snowmelt, which can increase the severity of summer drought events and their impacts on ecosystem functioning (Barnett et al. 2005; Edwards et al. 2007). These climatic changes directly or indirectly affect most organisms in ecosystems such as mountain grasslands, and can thus negatively impact the provision of numerous ecosystem services including fodder production and maintenance of soil fertility (de Vries et al. 2012; Alster et al. 2013).

Drought events can influence the responses of plant functional traits through increased leaf senescence (Gepstein 2004) and reduced vegetative growth (Benot et al. 2013). Drought also influences root functional traits through increased root dry matter content (Brunner et al. 2015; Zwicke et al. 2015), increased root biomass (Comas et al. 2013) or decreased specific root length (de Vries et al. 2016). Moreover, root exudation patterns of plants can change in response to drought and potentially affect microbial community responses to climate change through changes in resource availability (De Vries and Shade 2013). Drought also impacts microbial communities through a reduction in water availability and carbon and nutrient diffusion in soils, leading to changes in nutrient uptake and activities (Fierer et al. 2003; Henry 2013). These can result in cell death and decreased microbial biomass (Schimel et al. 1999), and changes in microbial community composition (Fierer et al. 2003; Evans and Wallenstein 2014). Similarly, extracellular enzymatic activities involved in the first steps of soil organic matter decomposition and nutrient recycling might be strongly impacted by climate change because of their strong dependence on soil temperature and moisture (Henry 2013). For instance, in a grassland transplantation experiment Puissant et al. (2015) showed that after four years soil transplanted to lower altitude had a lower enzymatic pool size, with different impacts depending on individual enzymes. Yet, other studies did not show any effect (Henry 2013), or even reported an increased soil enzymatic pool size during drought, which might influence the availability of nutrients for plants and the fate of organic matter (Acosta-Martinez et al. 2014b). These conflicting results highlight the need for a better understanding of the response of extracellular enzymatic activities to extreme weather events given their implication in detrital organic matter decomposition and mineralization. Similarly, many studies have reported the influence of climatic stress on the interactions between plants and microbes. For example, by reducing the root nitrogen (N) uptake capacity, drought could increase soil inorganic N availability and slow down N cycling by reducing the activity of soil nitrifying microbes (Hartmann et al. 2013). Contributions of arbuscular mycorrhizal fungi (AMF) and/or bacteria to enhancing plant drought tolerance under natural soil conditions have been reported in previous studies (Ortiz et al. 2015). For example, root-associated bacterial microbiome improved plant resistance to drought through water stress-induced promotion ability (Rolli et al. 2015). In addition, AMF has been shown to enhance drought tolerance in various plants through physiological mechanisms involved in nutrient uptake and biochemical mechanisms regarding hormones, osmotic adjustment, and antioxidant systems. Mycorrhizal plants also can possess direct pathways of water uptake via extraradical hyphae (Wu and Zou 2017). Drought has been shown to have both positive and negative effects on AMF colonization of host plants (Klironomos et al. 2011), but in some cases, drought has caused a reduction in extraradical hyphae or even led to the inability of AM fungi to colonize plant roots and, therefore, reduced the fungal fitness (Millar and Bennett 2016).

At the ecosystem level in subalpine grasslands diverse effects of drought have been observed, and there do not appear to be generic responses (Matteodo et al. 2016; Gritsch et al. 2016), whilst studies have shown more limited effects on plant traits, at least in the short-term (Vittoz et al. 2009; Benot et al. 2014), drought can greatly alter vegetation functional structure and its productivity by shifting dominant plants from exploitative to more conservative strategies. These changes might result from species turnover (Evans et al. 2011) as well as functional responses at the intraspecific level (Jung et al. 2014). Higher N mineralization rates (Rustad et al. 2001) and N availability, increasing plant productivity (Baldwin et al. 2014) have also been observed to lead to shift from conservative to more exploitative species (Debouk et al. 2015).

Agricultural practices such as fertilization and mowing also influence grassland communities, notably through changes in nutrient availability, and modulate the responses of individual species and communities to drought. In mountain grasslands, mowing and/or grazing may have greater short-term impacts on plant community structure than climatic variations or extremes (Benot et al. 2014; Vittoz et al. 2009). Mowing or grazing can also modify the quantity and quality of exudates released by plant roots after defoliation in the short-term (Hamilton et al. 2008), or through changes of plant community composition over the longer term (De Deyn et al. 2008). Drought and fertilization may also alter the efficiency of extracellular enzymatic activities of microbial communities because of reduced interactions between enzymes and their substrates when water is scarce (Alster et al. 2013). Combined effects of drought and N addition have also been reported to decrease microbial biomass and might induce a reduction in microbial activities involved in the recycling of nutrients (Khalili et al. 2016). Because microbial community composition and functioning, and therefore nutrient cycling, are affected by climatic stress, their response could potentially feedback to the performance of plant communities (de Vries et al. 2012; Nie et al. 2013). The expected increase in the frequency of extreme weather events could also alter the sensitivity of microbial communities to disturbances (Berard et al. 2012; Hawkes and Keitt 2015; Fuchslueger et al. 2016), and modify the response of aboveground and belowground organisms to drought as well as their interactions (Bardgett et al. 2013, 2014).

Drought, agricultural practices and plant composition as well as the response of plant or microbial communities to these drivers can have long-lasting legacy effects on ecosystem functioning. During the last decade, these legacy effects have been studied to better understand the influence of global change on ecosystem functioning (Evans and Wallenstein 2012; Cong et al. 2015; Xu et al. 2017). To unravel plant succession dynamics, several studies have focused on interspecific plant-soil feedbacks as one of these plant species composition legacy effects. van de Voorde et al. (2011).) showed that soil conditioned with Jacobaea vulgaris had no effect or increased the biomass of thirty other plant species, whereas soil conditioned with these species had neutral effect or decreased the biomass of J. vulgaris. This legacy effect can result from increased nutrient availability (Cong et al. 2015) or from the control of microbial community composition and activity through root exudates (De Deyn et al. 2008). Many years after a drought, legacy effects on plant communities were mainly characterized by a change in aboveground biomass production. Previous studies reported positive (De Vries et al. 2012) or negative (Sala et al. 2012b, Xu et al. 2017) drought effects on biomass production, due respectively to increased nutrient availability (De Vries et al. 2012), or to species reorganization within plant communities (Xu et al. 2017) and associated changes in functional traits values (Polley et al. 2014). Drought legacy effects on microbial communities have also been reported, with a greater resistance to water stress for microbial communities subject to drought than for those never previously exposed to this stress (Fierer et al. 2003). This suggests a change in microbial community structure (Evans and Wallenstein 2012) and/or functioning such as the ability to use recent plant-derived carbon (Fuchslueger et al. 2016), or a higher sensitivity of enzymatic production (Averill et al. 2016), which can influence plant-soil interactions and biogeochemical cycles. These effects, whether the results of one or combined historical legacies, remain poorly understood (Anderegg et al. 2015; Evans and Wallenstein 2012), and need to be studied to help farmers to cope with global change and maintain grassland ecosystem services such as fodder production.

To bridge this gap, we aimed at determining which combinations of past climatic stresses, management and plant community functional composition may benefit plant biomass production and quality, as well as promoting its resistance and recovery to extreme future weather events. Specifically, we tested whether there was a legacy in soils previously subjected to various combinations of climatic stresses, plant communities and management. We evaluated this legacy effect in a greenhouse experiment by planting into these soils a phytometer, Lolium perenne, which is a species often sown by farmers in mountain grasslands and in drought-damaged grasslands for its ability to grow fast and resist drought (Loucougaray et al. 2015; Zwicke et al. 2015). We measured L. perenne’s growth and biomass production under optimal conditions, and then we quantified its resistance and recovery to a new drought. We hypothesized that (1) N availability is higher in soils previously submitted to climatic stress combined with cutting and fertilization practices in exploitative plant communities, (2) after the new drought the reduction in microbial biomass and the associated release of available N would promote Lolium perenne ‘s resistance and growth, (3) soil microbial enzymatic activities and AMF that withstood past climatic stress should promote the recovery of L. perenne growth by sustaining nutrient cycling rates.

Materials and methods

Experimental design of the conditioning phase

We used soils originating from an in situ mesocosm experiment designed to test the influence of simulated drought, and early snow removal on two assembled subalpine plant community types differing in their functional composition, and subjected or not to simulated management (see Bernard 2017). Briefly, two plant community types were assembled by manipulating the abundance of three common grass species of subalpine grasslands with contrasting plant functional traits. We manipulated species abundance to obtain communities with contrasted community weighted mean traits and to avoid confounding effects linked to species identity and richness. Patzkea paniculata (L.) G.H.Loos (formerly Festuca paniculata (L.) Schinz & Tell.) is characterized by high leaf C/N and low specific leaf area (SLA); Dactylis glomerata (L.) is characterized by a fast relative growth rate, low leaf C/N and higher SLA; and Bromopsis erecta (Huds.) Fourr. (formerly Bromus erectus (Huds.)) has intermediate characteristics (Gross et al. 2007; Grassein et al. 2010). The relative abundances of the three species were manipulated to obtain two communities with 36 individuals, one community with more resource-acquisitive traits (higher SLA, and lower leaf C/N ratio) dominated by D. glomerata (27 individuals), and one community with more resource-conservative traits (lower growth rate and SLA, higher leaf C/N ratio) dominated by P. paniculata (27 individuals). The two communities included 6 individuals of B. erecta.

The experiment was conducted at the Station Alpine Joseph Fourier at the Lautaret Pass (French Alps - 45°02′N. 6°20′E.). In September 2012, plant communities were established in mesocosms (Ø 45-cm, 50-cm deep) filled at the bottom with 5-cm quartz gravel (Ø 45-cm) and soil previously sieved (5-mm) and homogenized. The soil used to fill the mesocosms was excavated from subalpine grasslands next to the field station between 10 and 40-cm depth, avoiding the top 10-cm and its root mats and seed bank. This soil is classified as a brown earth soil with 23% sand, 30% clay and 46% loam; pH: 6.3; soil organic matter: 18.4%, soil N: 0.68%. Plant individuals were collected in nearby subalpine grasslands ranging from 1700 to 2100-m a.s.l. Single tillers of each species were separated from large tussocks and transplanted after careful root washing in a standardized loam-sand mixture to homogenize microbial communities. After one month of growth seedlings were planted in the mesocosms after a brief root wash to limit the transfer of soil microbes from the loam-sand mixture. Half of the mesocosms received 150 kg N ha−1 (e.g. 17.55-g) in the form of a urea-based slow release N:P:K fertilizer (Nutricote® 13:13:13 + 2 MgO + oligoelements), applied after snowmelt in May 2014. Organic fertilization was chosen to simulate organic manure fertilization by local farmers. We used a higher N quantity than what is usually sprayed in the fields (about 50 kg N ha−1) to maximize the plant trait gradient as well as the fertilization effect. Vegetation of these fertilized mesocosms was cut to 10-cm shoot height in July 2014 to simulate mowing at peak biomass, consistent with local management practices. A climate manipulation was assigned to a half of the mesocoms. This treatment consisted of a snow removal by shoveling in April 2014, one month earlier than the end of natural snowmelt, which was assessed using a combination of snowpack measurements and models (see Bernard 2017). In addition to the early snow removal, several weeks without watering were applied to simulate periods of drought during summers 2013 (4 weeks) and 2014 (8 weeks) (Fig. S1). For this, precipitation was intercepted by transparent shelters installed after snowmelt and removed in late September. During this period plant communities were watered manually each week to compensate for water losses due to evapotranspiration (see Bernard 2017). In September 2014, after 26 months (two growing seasons), aboveground vegetation was harvested for plant trait measurements and soil was sampled for physico-chemical and microbial analyses (Bernard 2017). In total this made for 8 treatments: 2 functional compositions × 2 climate treatments (with / without) × 2 management treatments (with / without), each with 4 replicate mesocosms (Table 1).
Table 1

Previous climatic and management treatments and plant functional composition characterizing the eight soil treatments with their abbreviations

Climate treatment

Management treatment

Plant community

Soil code

Snow Removal and Drought (SRD)

Cut and Fertilized (CF)

Patzkea paniculata

SRD-CF-Patz

Snow Removal and Drought (SRD)

Cut and Fertilized (CF)

Dactylis glomerata

SRD- CF-Dact

Snow Removal and Drought (SRD)

Not Cut and not Fertilized (NCF)

Patzkea paniculata

SRD-NCF -Patz

Snow Removal and Drought (SRD)

Not Cut and not Fertilized (NCF)

Dactylis glomerata

SRD –NCF-Dact

Control (Cont)

Cut and Fertilized (CF)

Patzkea paniculata

Cont-CF-Patz

Control (Cont)

Cut and Fertilized (CF)

Dactylis glomerata

Cont-CF-Dact

Control (Cont)

Not Cut and not Fertilized (NCF)

Patzkea paniculata

Cont-NCF-Patz

Control (Cont)

Not Cut and not Fertilized (NCF)

Dactylis glomerata

Cont-NCF-Dact

After this conditioning for two growing seasons (26 months), all mesocosms were left in place without climate or management treatments during 13 months (Fig. S1). Soils were collected in September 2015 and used for the present study in a pot experiment to test for soil legacies on the growth of Lolium perenne (L.), used as an independent phytometer not present at the site, and subsequently on its resistance and recovery to a new drought event. The 20-cm of top soil from the four replicates of each initial treatment were homogenized and sieved at 2-mm to remove roots. Before planting, soils were analyzed for each mesocosm and showed consistent values of soil organic matter (7.08 ± 0.38%), pH (4.15 ± 0.2), and inorganic N content (2.78 ± 0.43 μg N. g−1 dry soil). For each of the 8 initial treatments (Table 1), 15 pots (9 × 9 cm; depth, 9.5 cm) were filled with 500-g of soil giving a total of 120 pots (8 soil treatments × 3 harvests × 5 replicates). The surface and the depth of the pots were sufficient so that they were not fully colonized by the root system at the end of the experiment (third harvest).

Legacy effect experiment

Lolium perenne was chosen as a phytometer to test the soil legacy on plant growth because, although this species is usually sown in intensively managed mountain grasslands, it was neither used in the previous mesocosm experiment, nor present in the subalpine grasslands where the original soil was sampled. This precluded any “home-field advantage” from the previous plant communities established on these soils. Seeds of Lolium perenne were sourced from Arbiotech® (Saint Gilles, France), sterilized using ethanol and 1% bleach and sown in vermiculite in a controlled environment with air temperature kept at 20/16 °C (day/night) with a 14/10 h photoperiod cycles. After seedling establishment (c. 3 weeks), three individuals were planted per pot and arranged in a triangle to limit edge effects. Pots were placed in a greenhouse (20/16 °C and 16/8 h (day/night)) and watered once a week in a completely randomized design. Pots were left to grow in optimal conditions for 8 weeks and watered daily to maintain soil moisture at 30% wt/wt, equal to 50% water-holding capacity (WHC) until the first harvest of 40 pots (5 per treatment - Harvest 1). A drought period of 5 weeks was imposed on all remaining pots until soil moisture was 8.5% (14% WHC), after which 40 pots were harvested to evaluate resistance (5 per treatment – Harvest 2). Remaining pots were rewetted up to 30% wt/wt (50% WHC) and watered daily during 5 weeks to allow a recovery period before the final harvest of the 40 last pots (5 per treatment - Harvest 3) (Fig. S1). Because of insufficient soil material we could not include control pots without impacting the number of replicates for each treatment and thus the validity of our statistical analyses.

Plant biomass and trait measurements

At each harvest, shoots and roots were harvested separately. To calculate specific leaf area and leaf dry matter content, the last mature leaf of the three individuals of each pot was sampled to measure fresh leaf area (LICOR Li-3100), fresh biomass and dry biomass after 72 h at 70 °C (Perez-Harguindeguy et al. 2013). Each of the three plants from each pot was dried at 70 °C for 72 h, and weighed with its respective last mature leaf to determine its above-ground biomass. Roots were gently rinsed with water and sieved sequentially on a 5.6, 2 and 0.5-mm mesh to avoid any loss of fine roots. The root biomass was split into three aliquots, one aliquot was dried at 70 °C for a week, the other two were kept in an alcoholic solution (ethanol 10%, acetic acid 5% v: v) until root trait analyses and determination of arbuscular mycorrhizal fungi (AMF) colonization. Root length and diameter were measured by suspension in 1-cm of demineralized water in a clear acrylic tray and scanned at 300 dpi with an Epson Expression 10000XL flatbed scanner (Long Beach, California, USA). Digital root images were processed using the WINRHIZO software (WINRHIZO software (Regent Instruments Inc., Canada)). Roots were weighed thereafter to obtain fresh biomass and dried at 70 °C for a week in order to calculate root dry matter content (RDMC) specific root length and belowground biomass. Above- and belowground biomasses were finely ground (5-μm diameter) for the analysis of N and C using an elemental analyzer (FlashEA 1112: Fisher Scientific, Waltham, Massachusetts, USA). Belowground biomass N and C contents were measured on composite root samples made of the five replicates from each treatment. To determine root colonization by AMF roots were digested in KOH solution and then stained with trypan blue in lactophenol (Phillips and Hayman 1970). For each root system AMF colonization was estimated by optical microscopy from sixty root fragments of approximately 1-cm. Mycorrhizal development was evaluated according to the method by Trouvelot et al. (1986) using the MYCOCALC program (http://www.dijon.inra.fr/mychintec/Mycocalc-prg/download.html) to estimate the proportion of colonized root cortex in the whole root system (%M).

Abiotic soil parameters and soil microbial properties

Fresh soil subsamples were taken and stored at 4 °C and processed within 48-h for the determination of chemical and microbial soil parameters. Soil subsamples were stored at −20 °C before the quantification of extracellular enzymatic activities. Fresh soil subsamples were oven-dried at 70 °C for 1 week and weighed followed by 4-h at 550 °C to determine soil water content and soil organic matter respectively. Soil subsamples were air dried and ground to determine soil pH in a 1:2.5 (soil:distilled water) solution, and to measure soil C and N with a FlashEA 1112 elemental analyser (Fisher Scientific Inc., Waltham, MA, USA). Soil nutrients (nitrate (NO3 ), ammonium (NH4 +), phosphate (PO4 ), total dissolved nitrogen, dissolved organic nitrogen (DON) and total dissolved phosphorus) were determined from soil extracts of 0.5 M K2SO4 (Jones and Willett 2006) using a photometric analyzer (Gallery Plus, Thermo Fisher Scientific, Waltham, Massachusetts, USA).

Soil N immobilized in the microbial biomass was calculated from the difference of N before and after chloroform-fumigation extraction technique of Vance et al. (Vance et al. 1987). Microbial biomass was calculated using a correction factor of 0.45 (Brookes et al. 1985). Potential activity of seven extracellular enzymes that degrade common compounds constituting soil organic matter were estimated using standard fluorimetric techniques as described in Bell et al. (Bell et al. 2013). These microbial enzymes are involved in the acquisition of carbon (α-Glucosidase, β-1,4-Glucosidase, β-D-Cellobiosidase, and β-Xylosidase), nitrogen (β-1,4-N-acetylglucosaminidase and leucine aminopeptidase) and phosphorus (phosphatase). The sum of the activities of enzymes involved in the degradation of C-rich substrates, N-rich substrates and P-rich substrates are designated hereafter as enzyme C activity, enzyme N activity and enzyme P activity respectively. Briefly, 2.75-g of soil were homogenized in 200-ml of sodium acetate buffer solution (pH 4.5) in a Waring blender for 1-min. The soil slurries were added in duplicate to 96-deep-well microplates followed by the addition of substrate solution. For each soil samples, additional control replicates were prepared by mixing 800-ml of soil slurry with 200-ml of 7-amino-4-methylcoumarin or 4-methylumbellfferone standard curves (0–100-μM concentration) in separated 96-deep-well microplates. All microplates were incubated during 3-h (dark, 175 rpm, 20 °C), then centrifuged at 2900-g for 3-min. Soil slurries (250-μL) were transferred from each well into black Greiner flat-bottomed 96-well plate (into corresponding wells) and then scanned on a Varioskan Flash (Thermo Scientific) microplate reader using excitation at 365-nm and emission at 450-nm. All enzymatic activities are expressed as nmol activity g−1 dry soil h−1. The calculation of the ratios between total enzyme C activity, enzyme N activity and enzyme P activity allowed estimating the microbial demand for C, N or P. An increase in the N:P enzymatic ratio resulting from a preferential increase in the production of β-1,4-N-acetylglucosaminidase and leucine aminopeptidase (enzyme N activity) over phosphatase (enzyme P activity) reflects a higher microbial demand for N in comparison with P, called thereafter in the manuscript “microbial N demand” (Sinsabaugh et al. 2009).

Data analysis

To explore the relationships within plant traits, soil abiotic parameters and microbial properties we conducted three principal component analyses (PCA) on samples harvested after 8 weeks (Harvest 1). The soil legacy of the previous treatments (Climate: Snow-Removal and Drought (SRD) vs Control (Cont); Management: Cut and Fertilized (CF) vs None (NCF); Plant community: More exploitative community dominated by Dactylis glomerata (Dact) vs more conservative community dominated by Patzkea paniculata (Patz)) were tested using three-way ANOVAs followed by least square difference post hoc tests. As roots were pooled per treatment for N and C analysis and thus could not be used for ANOVAs, we calculated the plant N and C quantity in total plant biomass to take into account the variability of root nutrient concentrations between treatments. When necessary, data was transformed to meet the assumptions of normality and homoscedasticity required for analyses. An index of relative response to drought and an index of relative recovery for plant traits and microbial properties was calculated for each parameter (p) (plant biomass and functional traits, microbial biomass, mycorrhizal intensity and extracellular enzymatic activities) as follows:
$$ {\displaystyle \begin{array}{l}\mathrm{Relative}\ \mathrm{Drought}\ \mathrm{Response}\ \left(\mathrm{p}\right)=\left[\left(\mathrm{Drought}\ \left(\mathrm{p}\right)-\mathrm{Optimal}\ \mathrm{Growth}\ \left(\mathrm{p}\right)\right)/\mathrm{Optimal}\ \mathrm{Growth}\ \left(\mathrm{p}\right)\right]\hfill \\ {}\mathrm{Relative}\ \mathrm{Recovery}\ \mathrm{Response}\ \left(\mathrm{p}\right)=\left[\left(\mathrm{Recovery}\ \left(\mathrm{p}\right)-\mathrm{Drought}\ \left(\mathrm{p}\right)\right)/\mathrm{Drought}\left(\mathrm{p}\right)\right]\hfill \end{array}} $$

For the relative Drought Response (DR), the value for Drought is the individual observation of parameter (p) at the end of the second harvest and the value for Optimal Growth is the mean of the parameter calculated from the five replicates of the same treatment at the end of the first harvest. For the relative Recovery Response (RR), the value for Recovery is the individual observation of the parameter at the third final harvest and the value for Drought is the mean of the parameter calculated from the five replicates of the same treatment at the end of the second harvest. Because we lacked control pots, this calculation of resistance and resilience did not take into account temporal dynamics of plants and microorganisms in our second experiment. We are aware of this uncertainty but our primary objective was to compare the influence of historical treatment on the responses of plants and microbial activities. We also assumed that the resistance and resilience to drought of Lolium perenne individuals growing on a same soil will be similar (even if some plasticity differences between individuals could occur), and that differences observed in our experiment were mainly due to the soil legacy of the previous treatments.

The soil legacies of the previous treatments (Climate, Management and Plant community) were tested on the two relative response indices using three-way ANOVAs. Linear regression models were developed using a bidirectional stepwise selection procedure (R library MASS using the stepAIC function) with relative drought and recovery response indexes as response variable to identify which plant traits, microbial properties or soil abiotic parameters best explained the relative response of Lolium perenne and microbial communities to a new drought. All the analysis were performed in R version 3.2.2 (R Core Team 2015), using the ade4 (Dray et al. 2007), lsmeans (Lenth 2016), nlme (Pinheiro et al. 2016), mass (Venables and Ripley 2002) and vegan (Oksanen et al. 2015) packages.

Results

Growth period – Harvest 1

After 8 weeks of growth in optimal conditions the principal component analysis of plant traits (PCA 1, Fig. S2) separated pots of Lolium perenne by their plant biomass, N content and more generally by their use of nutrient resources reflected by their traits (e.g. root N content, RDMC, specific leaf area) (axis 1: 44.3%). The principal component analysis of microbial properties (PCA 2, Fig. S3) separated pots of Lolium perenne by their enzyme P activity (axis 2: 26.7%) and by the N demand of the microbial community reflected by the enzyme N:P and C:N activities (axis 1: 35.2%). The principal component analysis of soil abiotic parameters (PCA 3) showed overall that more P (phosphate and total dissolved P) and less inorganic N were available in CF soils, whereas the opposite was true for soils from Dactylis glomerata-dominated mesocosms (‘Dact soils’ hereafter) that were not cut or fertilized (Fig. S4).

Only aboveground plant traits (e.g. biomass, leaf dry matter content, specific leaf area) of Lolium perenne were influenced by the previous climate manipulation, where individuals had more exploitative traits in SRD soils. All plant traits, except root diameter, were influenced by the previous management with more exploitative traits (e.g. higher specific root length, biomass, N content) for plants grown in CF soils (Table 2; Fig. S2). The previous plant community only influenced aboveground biomass which was highest in soils from mesocosms dominated by Patzkea paniculata (‘Patz soils’ hereafter) (Table 2). The interaction of previous management and plant community resulted in a higher plant biomass productivity and quality (i.e. aboveground biomass and shoot N content) in CF and Patz soils (Table 2). Aboveground biomass was lowest in all NCF soils and increased from CF-Dact soils to CF-Patz soils (Fig. 1). Microbial properties were differently affected by previous soil treatments. No significant difference was found for microbial biomass or enzyme P activity. Previous climate manipulation increased microbial N demand relative to C and P demands in SRD soils (e.g. Enzyme C:N and N:P activities – Table 2). Previous plant community also influenced microbial N demand with an increase of enzyme N activity and a higher enzyme N:P activities in Patz soils. The microbial N demand was also affected by the interaction between previous composition and management, and was highest in CF-Patz soils (Table 2; Fig. S3). Mycorrhizal intensity varied between soil treatments and was slightly higher in Cont-CF-Dact soils (Table 1). Soil abiotic parameters were also strongly influenced by previous soil treatments and their interactions, leading to complex and unclear patterns (Table 2). Soil pH varied among soil treatments but values were restricted between 3.99 and 4.11, and thus without influence on the observed variation of plant traits and microbial enzymatic activities. Overall, inorganic N forms were lowest in Cont soils, CF soils and Patz soils, while P availability (PO4 3− and total dissolved P) was highest in CF soils and Patz soils (only for total dissolved P). These availabilities of soil nutrients were correlated with plant biomass production as showed by the negative correlation between aboveground biomass and inorganic N concentration (R2 = 0.129, p = 0.023) and by the positive correlation between aboveground biomass and total dissolved P (R2 = 0.190, p = 0.005). Microbial nutrient demand was also related to total dissolved nutrient concentration as showed by the positive correlation between enzyme N:P activities and total dissolved P (R2 = 0.352, p < 0.001) and enzyme N:P activities and total dissolved N (R2 = 0.239, p = 0.001).
Table 2

Effect of previous climatic and management treatments, plant functional composition, and all possible interactions on Lolium perenne biomass and functional traits, microbial properties and soil abiotic parameters at the end of the growth period under optimal conditions

 

Climate

Management

Community

Cli*Man

Cli*Com

Man*Com

Cli*Man*Com

F

p

F

p

F

p

F

p

F

p

F

p

F

p

Plant traits

 Aboveground biomass

4.22

0.048

91.07

<0.001

6.33

0.017

3.03

0.091

0.20

0.657

8.55

0.006

0.03

0.856

 Leaf Dry Matter Content

6.52

0.016

37.38

<0.001

1.60

0.215

3.63

0.066

2.21

0.147

1.14

0.293

3.26

0.080

 Specific Leaf Area

6.87

0.013

11.42

0.002

1.00

0.326

9.19

0.005

0.25

0.624

0.01

0.929

1.60

0.215

 Shoot N Content

3.52

0.070

48.11

<0.001

1.75

0.195

1.64

0.210

0.50

0.486

11.37

0.002

2.23

0.145

 Shoot C Content

8.07

0.008

35.19

<0.001

0.37

0.546

4.00

0.054

0.57

0.457

3.40

0.075

0.94

0.339

 Root Dry Matter Content

0.02

0.893

11.55

0.002

1.85

0.184

0.67

0.421

1.50

0.230

1.50

0.230

0.07

0.790

 Specific Root Length

0.71

0.405

6.42

0.016

0.35

0.559

0.82

0.372

0.78

0.385

0.18

0.677

0.91

0.348

 Root Diameter

1.04

0.315

0.42

0.520

1.46

0.236

0.01

0.943

0.16

0.690

0.00

0.959

2.24

0.145

Microbial properties

 Microbial Biomass

0.36

0.553

1.12

0.299

0.10

0.757

3.00

0.093

0.85

0.363

0.21

0.650

0.00

0.952

 Mycorrhizal Intensity

1.78

0.191

10.38

0.003

0.96

0.336

0.19

0.664

16.96

<0.001

0.47

0.496

14.69

<0.001

 Enzyme C activity

0.80

0.378

8.39

0.007

3.03

0.091

1.30

0.263

0.56

0.459

1.65

0.208

0.25

0.623

 Enzyme N activity

1.62

0.212

1.43

0.241

9.55

0.004

6.12

0.019

2.19

0.148

1.57

0.219

4.19

0.049

 Enzyme P activity

2.19

0.149

2.25

0.144

0.92

0.344

0.48

0.491

1.94

0.174

0.11

0.741

0.25

0.621

 Enzyme C/N activities

8.31

0.007

0.79

0.381

4.13

0.051

4.07

0.052

1.10

0.303

0.05

0.818

10.70

0.003

 Enzyme C/P activities

1.25

0.271

2.81

0.103

0.82

0.373

7.06

0.012

0.70

0.410

0.86

0.360

0.46

0.500

 Enzyme N/P activities

4.46

0.043

0.09

0.769

9.51

0.004

12.59

0.001

0.03

0.867

0.22

0.645

9.78

0.004

Soil parameters

 Soil Water Content

0.10

0.751

0.95

0.337

7.30

0.011

3.03

0.091

0.10

0.754

0.00

0.967

1.62

0.210

 pH

30.25

<0.001

12.79

0.001

6.64

0.015

0.33

0.568

0.12

0.731

0.65

0.425

5.87

0.021

 Total Soil N

4.48

0.042

1.84

0.184

1.03

0.319

14.80

<0.001

0.59

0.449

0.45

0.509

18.95

<0.001

 Total Soil C:N ratio

2.46

0.127

5.65

0.024

4.11

0.051

2.33

0.137

0.35

0.556

0.91

0.348

4.63

0.039

 Inorganic N

25.62

<0.001

9.88

0.004

25.22

<0.001

25.43

<0.001

8.14

0.008

3.01

0.093

12.00

0.002

 Phosphate

0.17

0.687

18.24

<0.001

1.80

0.189

3.45

0.072

5.53

0.025

5.64

0.024

4.23

0.048

 Total Dissolved N

0.33

0.567

0.30

0.588

1.00

0.325

20.67

<0.001

0.16

0.687

0.10

0.750

27.58

<0.001

 Dissolved Organic N

16.57

<0.001

5.23

0.029

12.27

0.001

0.85

0.360

4.82

0.036

1.53

0.230

10.92

0.002

 Total Dissolved P

1.97

0.170

26.39

<0.001

9.83

0.004

18.42

<0.001

0.31

0.580

1.87

0.180

37.92

<0.001

For each treatment, F-values and p-values of ANOVAs were given. Significant differences are highlighted in bold type. Abbreviations: Carbon (C), Nitrogen (N), Phosphorus (P)

Fig. 1

Aboveground biomass of Lolium perenne after 8 weeks in optimal growth conditions in soils previously subjected to different climatic events (Snow-Removal and Drought (SRD) vs Control (Cont)), management practices (Cut and Fertilized (CF) vs None (NCF)) and plant communities (More exploitative community dominated by Dactylis glomerata (Dact) vs more conservative community dominated by Patzkea paniculata (Patz)). Bars indicate mean ± 1SE (n = 5) and were compared using least square difference post hoc tests. Values sharing the same letter are not significantly different (p > 0.05)

Drought period – Harvest 2

During the drought period aboveground biomass increased similarly in all soils as observed with the relative drought response for aboveground parts of Lolium perenne (Table 3). However at whole plant scale more N accumulated in SRD soils. Belowground parts showed significant differences in their relative drought response between soil treatments with a higher production of belowground biomass in SRD soils, and particularly in Dact soils. In SRD soils, the root diameter of L. perenne remained stable and was associated with the lowest increase in RDMC. On the contrary, root diameter decreased in Cont soils and this was associated with the highest increase in RDMC (Table 3). The previous management practices modulated these RDMC responses with the highest increase of RDMC and the highest decrease of root diameter in NCF-Cont soils. Specific root length decreased in all soils regardless of treatment. These belowground responses to the drought period were illustrated by the positive correlation between root diameter and belowground biomass (R2 = 0.290, p < 0.001). Linear models showed that the responses of belowground traits were correlated with changes in root nutrient concentrations and morphology through an increase of root diameter and an increase of root C concentration with the increase of belowground biomass and RDMC respectively (Table 4). Microbial biomass decreased in all soils during drought period but its relative drought response indicated a stronger decrease in SRD soils. Enzyme P activity showed an overall increase in all treatments whereas enzyme N and enzyme C activities decreased in all treatment and particularly in NCF soils. Microbial demand for P increased in all treatments as shown by the decrease of enzyme C:P and enzyme N:P activities (Table 3). However the microbial demand in SRD-CF-Patz soils differed from other treatments with an increase in enzyme N activity (Fig. 2), the lowest decrease in enzyme C:P activities and no decrease in enzyme N:P activities (Fig. 2). Mycorrhizal intensity was significantly influenced by previous soil treatments and increased significantly in Cont-Dact soils (Table 3). Linear models showed that the response of microbial biomass and N demand were overall explained by root functional traits (e.g. root C:N ratio and root C content) and total activity of N-, C- and P-acquiring enzymes (Table 4). These results were confirmed by the fact that changes in microbial nutrient demand were driven by variations in N-acquiring enzymes activity as suggested by their positive correlation with enzyme N:P activities (R2 = 0.778, p < 0.001) and their negative correlation with enzyme C:N activities (R2 = 0.206, p = 0.003). The responses of root functional traits, microbial biomass and N demand were all related to soil N availability as suggested by positive correlations between root diameter (R2 = 0.167, p = 0.009) and the drought response of enzyme C:N activities (R2 = 0.123, p = 0.026) and the negative correlation between the drought response of root C:N ratio (R2 = 0.249, p = 0.001) and the drought response of microbial biomass (R2 = 0.131, p = 0.022).
Table 3

Effect of previous climatic and management treatments, plant functional composition, and all possible interactions on relative Drought Response (DR) and Recovery Response (RR) of Lolium perenne biomass and functional traits and microbial properties

 

Climate

Management

Community

Cli*Man

Cli*Com

Man*Com

Cli*Man*Com

F

p

F

p

F

p

F

p

F

p

F

p

F

p

Plant traits

 Aboveground biomass DR

2.81

0.103

0.03

0.861

3.46

0.072

0.66

0.424

0.06

0.801

0.16

0.691

0.08

0.777

 Belowground biomass DR

5.99

0.020

0.29

0.594

5.75

0.023

3.20

0.083

0.00

0.976

0.02

0.888

0.67

0.419

 Shoot N Content DR

2.36

0.134

0.95

0.335

0.01

0.929

3.41

0.074

0.22

0.643

0.00

1.000

2.78

0.105

 Root Dry Matter Content DR

14.99

<0.001

7.89

0.008

2.12

0.155

6.35

0.017

0.01

0.909

0.26

0.613

1.88

0.180

 Specific Root Length DR

0.01

0.943

1.45

0.237

0.02

0.896

0.00

0.996

0.10

0.759

1.50

0.229

0.00

0.982

 Root Diameter DR

20.50

<0.001

1.14

0.293

2.26

0.143

5.91

0.021

0.01

0.934

1.92

0.176

0.06

0.803

Microbial properties

 Microbial Biomass DR

26.05

<0.001

0.72

0.404

3.76

0.061

2.01

0.165

0.90

0.349

0.17

0.680

0.69

0.411

 Mycorrhizal Intensity DR

38.50

<0.001

141.21

<0.001

90.39

<0.001

33.97

<0.001

13.85

<0.001

85.37

<0.001

11.87

0.002

 Enzyme C activity DR

0.96

0.333

7.87

0.008

0.87

0.357

0.61

0.442

6.85

0.013

0.90

0.349

0.02

0.895

 Enzyme N activity DR

1.95

0.173

5.70

0.023

5.04

0.032

1.49

0.230

16.11

<0.001

2.41

0.130

18.99

<0.001

 Enzyme P activity DR

0.00

0.972

1.29

0.265

0.93

0.342

10.35

0.003

0.98

0.329

1.07

0.309

0.27

0.606

 Enzyme C/N activities DR

4.98

0.033

3.42

0.074

8.08

0.008

15.12

<0.001

0.07

0.399

12.47

0.001

66.79

<0.001

 Enzyme C/P activities DR

4.14

0.051

11.07

0.002

0.07

0.799

0.75

0.393

4.24

0.048

0.03

0.875

0.42

0.523

 Enzyme N/P activities DR

11.47

0.002

14.11

<0.001

13.43

<0.001

38.94

<0.001

21.58

<0.001

2.49

0.124

56.20

<0.001

Plant traits

 Aboveground biomass RR

2.82

0.103

0.15

0.703

0.01

0.914

0.48

0.492

0.21

0.648

3.66

0.065

4.79

0.036

 Belowground biomass RR

0.48

0.492

5.99

0.020

5.79

0.022

0.32

0.574

1.73

0.200

3.04

0.090

0.00

0.965

 Shoot N Content RR

38.05

<0.001

53.12

<0.001

9.94

0.003

1.93

0.180

0.01

0.940

7.54

0.009

0.11

0.741

 Plant N Quantity RR

2.59

0.117

10.05

0.003

1.41

0.243

0.07

0.789

0.7

0.407

0.05

0.823

6.92

0.013

 Root Dry Matter Content RR

39.62

<0.001

57.91

<0.001

66.58

<0.001

5.36

0.027

0.06

0.802

0.08

0.785

32.94

<0.001

 Specific Root Length RR

0.005

0.944

9.98

0.003

1.33

0.257

13.22

<0.001

0.09

0.767

4.91

0.034

1.13

0.296

 Root Diameter RR

43.87

<0.001

6.64

0.015

1.38

0.248

6.33

0.017

0.48

0.494

2.88

0.099

15.00

<0.001

Microbial properties

 Microbial Biomass RR

2.56

0.119

3.17

0.084

0.00

0.942

2.81

0.103

0.00

0.954

2.28

0.141

0.46

0.503

 Mycorrhizal Intensity RR

31.88

<0.001

54.63

<0.001

9.78

0.004

34.91

<0.001

19.71

<0.001

13.82

<0.001

16.56

<0.001

 Enzyme C activity RR

0.54

0.468

0.56

0.462

3.58

0.068

2.60

0.116

10.73

0.003

0.01

0.919

1.09

0.304

 Enzyme N activity RR

10.43

0.003

0.67

0.418

1.29

0.264

5.84

0.022

36.91

<0.001

1.94

0.173

8.84

0.006

 Enzyme P activity RR

3.46

0.072

0.00

0.945

10.35

0.003

0.05

0.826

2.11

0.156

0.15

0.701

1.73

0.198

 Enzyme C/N activities RR

1.3

0.262

6.38

0.017

0.07

0.798

1.08

0.307

0.01

0.924

0.10

0.754

0.00

0.971

 Enzyme C/P activities RR

0.78

0.385

2.91

0.097

1.97

0.171

10.35

0.003

17.39

<0.001

0.68

0.415

0.37

0.545

 Enzyme N/P activities RR

5.72

0.023

6.64

0.015

8.21

0.007

5.01

0.032

14.30

<0.001

0.33

0.568

2.13

0.154

For each treatment, F-values and p-values of ANOVAs were given. Significant differences are highlighted in bold type. Abbreviations: Carbon (C), Nitrogen (N), Phosphorus (P)

Table 4

Percentages of variance of selected root traits and microbial parameters explained by plant traits, microbial properties and soil abiotic parameters

Response variable

Retained variables

% explained

t value

p-value

Belowground Biomass DR

Root diameter

28.95

3.9

<0.001

Root Dry Matter Content DR

Root C concentration

41.45

4.976

<0.001

Leaf dry matter content

19.72

4.83

<0.001

Mycorrhization

4.27

−2.11

0.042

Microbial Biomass DR

Root C:N ratio

28.68

3.22

0.003

Enzyme N:P activities

7.91

−2.97

0.005

Enzyme P activity

11.11

2.77

0.009

Enzyme N activity DR

Enzyme C activity

40.33

7.28

<0.001

Soil organic matter

16.10

−3.51

0.001

Mycorrhization

5.42

2.26

2.26

Enzyme C:N activities DR

Root C concentration

44.93

−10.01

<0.001

Enzyme N:P activities

18.33

−7.57

<0.001

Enzyme C:P activities

8.37

5.95

<0.001

Phosphate

10.99

−4.71

<0.001

Enzyme N:P activities DR

Enzyme N activity

53.01

15.37

<0.001

Root C concentration

20.48

7.29

<0.001

Enzyme P activity

13.93

−6.11

<0.001

pH

3.58

−3.73

<0.001

Belowground Biomass RR

Shoot N concentration

31.03

−4.19

<0.001

pH

21.28

4.70

<0.001

Total Dissolved N

4.68

−1.98

0.055

Specific Root Length RR

Root N concentration

27.73

4.75

<0.001

Aboveground biomass

20.45

−4.58

<0.001

Phosphate

5.68

2.42

0.021

Enzyme C:N activities

4.74

−2.00

0.053

Enzyme N:P activities RR

Mycorrhization

20.74

−2.99

0.005

Enzyme P activity

14.37

−6.88

<0.001

Enzyme N activity

17.04

5.74

<0.001

Soil pH

9.72

3.01

0.005

Shoot C concentration

7.05

2.78

0.009

Mycorrhizal Intensity RR

Plant N quantity

34.55

3.69

<0.001

Leaf dry matter content

7.60

2.66

0.012

Phosphate

5.59

2.24

0.031

Total soil N

6.08

−2.15

0.039

All treatments were used in the models. Linear correlation models for DR were carried out for measurements made in harvest 2 and those for RR with measurements in harvest 3. Abbreviations: Carbon (C), Nitrogen (N), Phosphorus (P), relative drought response (DR) and relative recovery response (RR)

Fig. 2

Relative drought response (DR) on total N-acquiring extracellular enzymatic activity and the extracellular enzymatic activity N:P ratio after 5 weeks of drought in soils previously subjected to different climatic stresses (Snow-Removal and Drought (SRD) vs Control (Cont)), management practices (Cut and Fertilized (CF) vs None (NCF)) and plant communities (More exploitative community dominated by Dactylis glomerata (Dact) vs more conservative community dominated by Patzkea paniculata (Patz)). Bars indicate mean ± 1SE (n = 5) and were compared using least square difference post hoc tests. Values sharing the same letter are not significantly different (p > 0.05)

Recovery period – Harvest 3

An increase of plant biomass was observed in all treatments, and no difference was observed in the recovery relative response of aboveground biomass across treatments. The relative recovery response of belowground biomass showed an increase in root production which was highest in CF soils and Patz soils (Table 3). The shoot N concentration of L. perenne growing in SRD soils, CF soils and Patz soils was reduced. This decrease was concomitant with the lowest plant N accumulation during this period in the individuals growing in SRD-CF-Patz soils (Table 3). The relative recovery response of root traits differed also between treatments with an overall decrease of RDMC in Cont-NCF soils whereas it remained stable in SRD-CF soils. Specific root length decreased only in SRD-CF-Patz soils whereas it increased or remained stable in all other soils. Root diameter increased in all soils during the recovery period with the highest increase observed in Cont-NCF soils (Table 3). Linear models revealed that response of belowground biomass and specific root length were mainly explained by variations in plant nutrient concentrations and soil abiotic parameters and only marginally by microbial properties (Table 4). The calculation of relative recovery response for microbial properties showed that microbial biomass increased similarly in all soil treatments. All enzyme activities decreased in all soils during the recovery period with a stronger decrease in SRD soils and Patz soils for enzyme N and C activities, and only in Patz soils for enzyme P activity. Microbial demand for N relative to P decreased in SRD-CF soils and remained stable or increased in other soils (Fig. 3), while the microbial demand for N relative to C decreased in CF soils but not in NCF soils (Table 3). During this recovery period, the intensity of mycorrhizal root colonization strongly increased in comparison to the two previous periods, and was significantly higher in SRD-CF soils (Table 4, Figs. 3 & 4). Linear models revealed that the recovery response of enzyme N:P activities was mainly explained by the activity of each respective enzyme but also by the intensity of the mycorrhizal colonization, whereas the recovery response of mycorrhizal intensity was mainly explained by the variation of plant N quantity and soil nutrient availability (Table 4). Soil nutrient availability differed between soil treatments. Overall inorganic N (NH4 + and NO3 ) were highest in SRD soils and in NCF soils whereas phosphate was higher in CF soils (Fig. 4). Throughout the experiment, and particularly during the recovery period, we observed an overall decrease of inorganic N availability in SRD soils with the exception of SRD-NCF-Patz soils, while it remained stable in Cont soils. Dissolved organic N concentration showed an opposite pattern with an increase in SRD soils whereas it remained stable in SRD-NCF-Patz and Cont soils. Soil phosphate concentration increased during the recovery period in all SRD soils with the exception of SRD-None-Dact and in all Cont-None soils, whereas it remained stable in Cont-CF soils (Fig. 4).
Fig. 3

Relative recovery response (RR) on the extracellular enzymatic activity N:P ratio and mycorrhizal root colonization rate 5 weeks after rewetting in soils previously subjected to different climatic stresses (Snow-Removal and Drought (SRD) vs Control (Cont)), management practices (Cut and Fertilized (CF) vs None (NCF)) and plant communities (More exploitative community dominated by Dactylis glomerata (Dact) vs more conservative community dominated by Patzkea paniculata (Patz)). Bars indicate mean ± 1SE (n = 5) and were compared using least square difference post hoc tests. Values sharing the same letter are not significantly different (p > 0.05)

Fig. 4

Soil inorganic N, dissolved organic nitrogen, phosphate contents and mycorrhizal root colonization rate over time. Symbols and bars represent mean ± 1SE (n = 5). Harvest 1: after 8 weeks of growth in optimal conditions; Harvest 2: after 5 weeks of drought; Harvest 3: 5 weeks after rewetting

Discussion

The primary aim of this study was to determine how previous climatic stress, plant community functional composition and management practices have interactive long-lasting effects on soil functioning which could affect the biomass production and quality of Lolium perenne. We hypothesized that (1) N availability is higher in soils previously subjected to climatic stress combined with cutting and fertilization practices on exploitative communities, (2) after a new drought the reduction in microbial biomass and the associated release of available N promote Lolium perenne resistance and growth, (3) soil microbial enzymatic activities and AMF that withstood past climatic stress should promote the recovery of L. perenne growth by sustaining nutrient cycling rates.

Soil legacy effects on Lolium perenne growth

In agreement with our first hypothesis, we observed a strong positive soil legacy of previous climatic stress (SRD) and cutting and fertilization practices (CF) on the biomass production and quality of Lolium perenne. However our first hypothesis was not fully supported as we did not observe a positive soil legacy of the exploitative community (Dact soils), since plant biomass production was higher in soils previously dominated by conservative community (Patz soils). This was not expected since exploitative plant communities are associated with higher rates of soil nutrient cycling and higher nutrient availability, whereas conservative plant communities generally drive slower rates of nutrient cycling (Grigulis et al. 2013; Kastovska et al. 2015). However the higher biomasses in Patz soils detected under CF conditions suggested that previous cutting and fertilization could have reduced nutrient limitation, which eventually promoted the growth of Lolium perenne. Moreover studies have shown that microbial communities in grasslands dominated by Patzkea paniculata have lower N uptake abilities than in grasslands dominated by exploitative species (Robson et al. 2010; Legay et al. 2013). This lower microbial competitive ability might be a positive soil legacy of Patzkea soils. These results suggest that Lolium perenne had access to available nutrients for its growth only if previous SRD communities were managed. Soils supporting the highest plant biomass showed also a reduced soil inorganic N availability but similar microbial biomass, which suggests that competition with microbial communities was similar to other soils. However, microbial communities in SRD soils were more N limited (higher enzyme N:P activities and lower enzyme C:N activities) which could reflect a climatic stress legacy on the microbial community. In a previous study Dijkstra et al. (2015) showed that microbial communities affected by drought had higher requirements for N relative to P. In our study, these N-limited microbial communities increased their enzymatic investments into N-rich substrates degradation which could be explained by the lower availability of DON in SRD soils. Altogether these results suggest a strong soil legacy of previous climatic stress and management which reduced L. perenne competition with microorganisms for inorganic N acquisition and promoted its growth. Under these conditions, Lolium perenne was most likely not colonized by AM fungi because more nutrients in the soil were directly available to the plant. Under these conditions, carbon investment in the AM symbiosis was therefore less necessary to improve plant nutrient acquisition (Fig. 5).
Fig. 5

Conceptual schema summarizing the most significant patterns observed. Each plant represents a Lolium perenne individual planted in pre-treated soil and grown under optimal conditions (harvest 1), and then afterwards being exposed to a drought (harvest 2), and subsequently a recovery period (harvest 3) during the glasshouse experiment. Dashed arrows represent a negative legacy effect and plain arrows represent a positive legacy effect. Legend: Snow-removal and drought (SRD), cutting and fertilized (CF) or not (NCF), plant community dominated by Patzkea paniculata (Patz) or Dactylis glomerata (Dact). Colors of the arrow represent the effects of one treatment or combination of treatments at each harvest with colour choices being independent from one harvest to the other

Soil legacy on Lolium perenne resistance to drought

During the drought period, we observed an increase in plant biomass and N acquisition associated with an increase in RDMC and a decrease of specific root length, a response commonly observed and which could reflect a drought resistance strategy (Brunner et al. 2015; Zwicke et al. 2015; De Vries et al. 2016). However, a higher root biomass production and plant N acquisition associated with a lower RDMC was observed in SRD soils, suggesting that the response of L. perenne could reflect phenotypic plasticity. The lower increase of RDMC in SRD soils could result from a lower increase of nonstructural carbohydrates in roots (Zwicke et al. 2015), and reflect favorable soil conditions for the growth of L. perenne under drought conditions. These plant trait variations could have lowered root exudation (Baptist et al. 2015), and influenced nutrient availability and enzymatic activities of the microbial community (De Vries et al. 2016). This was suggested by root traits being involved in the resistance response of microbial enzymatic activities.

This new drought also strongly impacted microbial properties with a decrease in all enzyme activities and microbial biomass in all soils. The soil drying led to a decrease in enzyme activities and reduced nutrient uptake (Henry 2013) and likely caused the death of less resistant microbial populations (Schimel et al. 1999). As suggested by the decrease in the ratios of enzyme C:P and enzyme N:P activities during the new drought, the microbial demand for P relative to C and N increased. Although microbial P demand increased soil phosphate content increased suggesting either a reduction of P plant demands (Sardans et al. 2006), or that drought reduced the mobility of soil phosphate, and thus its availability for plants and microbes (Peuke and Rennenberg 2004). Inorganic nutrient concentration was not influenced by new drought but soil DON concentration overall increased in SRD soils. We argue that this difference reflected a legacy of previous climatic stresses since in SRD soils microbial biomass was strongly reduced and might have released their organic N cell contents (Lipson et al. 1999). In Cont soils where microbial biomass reduction was the lowest, DON concentration was stable. A possible explanation for this stable DON concentration is that remaining soil microbes used the excess DON to accumulate osmolytes (e.g; proline, glutamine) that reduce their internal water potential and limit further dehydration (see review of Schimel et al. 2007). Soil microbes could also have synthetized extracellular polymeric substances to protect their cells integrity as well as their local environment during the drought period (Berard et al. 2015). Hence climate legacy could have been beneficial for the growth of L. perenne since microbial communities in SRD soils were more impacted by the new drought reducing their competitive ability for inorganic nutrients acquisition.

However, microbial communities in soils previously submitted to climatic stresses, management practices and dominated by P. paniculata (SRD-CF-Patz) showed a singular response. The microbial demand for P was lower but stronger for N (highest enzyme N:P activities and microbial N limitation) suggesting a negative management legacy through a potential decrease of N-cycling related microbes (Hartmann et al. 2013). This effect was likely associated with a negative (or selective) plant community legacy of Patzkea paniculata on the microbial communities (Robson et al. 2007; Binet et al. 2017), resulting in a further decrease of microbial N competitive abilities for inorganic N acquisition in this treatment. Hence previous climatic treatment could have promoted microbial communities which are more sensitive to drought. These results differ from previous studies which overall found more drought-resistant microbial communities in soils frequently exposed to climatic stresses (Fierer et al. 2003; Berard et al. 2012; de Vries et al. 2012; Evans and Wallenstein 2014). This apparent increased sensitivity of microbial communities previously exposed to climatic stresses (i.e. SRD) might in fact reflect an increase in the proportion of drought-sensitive copiotroph microorganisms such as Proteobacteria as already reported in other studies (Acosta-Martinez et al. 2014a; Barnard et al. 2013; Berard et al. 2015). In these soils, the occurrence of freeze-thaw (associated with early snow removal treatments) events and drought followed by rewetting might have promoted the release of labile organic carbon and nutrients from lysed cells (Schimel et al. 2007). Such repeated inputs of labile material could favour opportunistic or copiotroph organisms that preferentially use labile organic carbon and have high nutrient requirements (Fierer et al. 2007; Litchman et al. 2015). The well-known lower resistance of copiotroph microorganisms to environmental stress in comparison with oligotrophs could explain the higher reduction in microbial biomass observed in SRD soils (Fierer et al. 2007). This lack of resistance of the microbial communities could also reflect the particular environmental conditions of subalpine grasslands from which soils originate. Seasonal variations induce strong microbial community shifts between spring and peak biomass in summer (Lipson and Schmidt 2004; Zinger et al. 2009; Legay et al. 2013), and a low resistance but high recovery could characterize these microbial communities (Mills et al. 2014). Hence, drought-resistance of Lolium perenne might have been favored in soils where the decline in biomass of sensitive microorganisms resulted in lower competition with plants for N acquisition. Altogether these results also revealed differences in the response of microbial communities to drought. Indeed, microbial communities never exposed to drought were more resistant and appeared to display mechanisms of adaption, whereas microbial communities already exposed to drought were more sensitive and did not display similar mechanisms. We suggest that the drought event in 2014 (8 weeks) had exceeded a critical threshold of drought duration leading to a drastic biomass die-off (Banning and Murphy 2008). This duration could have exceeded the resistance and resilience abilities of some microbial groups, explaining why mechanisms of adaptation occurred in microbial communities which have never experienced drought. Also, caution should be exercised concerning the legacy effect of soils observed since the size of the pots in this second experiment could have exerted a confinement effect on the response of plant and microbial communities.

Finally, during this new drought period in soils previously dominated by D. glomerata and without management (NCF-Dact), mycorrhizal colonization increased slightly suggesting that L. perenne could also develop drought-avoidance strategies through an association with AM fungi (Brunner et al. 2015) to increase phosphorus uptake and reduce water stress (Mohan et al. 2014). The mycorrhizal colonization found only in Dact soils could also reflect a plant community legacy since P. paniculata has been shown to reduce AMF colonization probably through the release of endophytic alcaloïds in soils especially in unmown conditions, where alkaloids concentration were higher due to high plant biomass and litter accumulation (Binet et al. 2017). Another explanation for the absence of AMF colonization could be the specific adaptation of some AMF strains to drought conditions or the loss of the ability of AMF to colonize plants internally (Millar and Bennett 2016). Disentangling the mechanisms or shifts in community composition explaining the response observed would require the molecular characterization of the soil microbial communities and arbuscular mycorrhizal communities, and this deserves further investigation.

Altogether, these results suggest that L. perenne was able to cope with the drought period through the development of a drought-resistance strategy characterized by higher root biomass and RDMC, and promoted by climate legacy lowering microbial N competitive abilities. Depending on the different legacies across plant community types, L. perenne could avoid more efficiently the negative effects of the drought period through AMF root colonization (Fig. 5).

Soil legacy on Lolium perenne recovery after drought

In comparison to the drought period, the recovery period was characterized by increases in aboveground and belowground biomass as well as in plant N acquisition. Root response consisted of a reduction of RDMC and an increase in root diameter and changed root morphology. Except in SRD-CF-Patz soils, specific root length increased in all soils suggesting that L. perenne plants were under suitable conditions to develop a root morphology associated with its resource-acquisitive strategy (Bardgett et al. 1999; Legay et al. 2014). As for the drought period, changes in root traits potentially affected microbial community resilience to drought (De Vries and Shade 2013). Although the rewetting period allowed the recovery of the microbial biomass in all soils, enzyme activities decreased again and particularly in SRD soils and Patz soils. The absence of microbial activity recovery is consistent with other studies (Acosta-Martinez et al. 2014a, 2014b and Averill et al. 2016) showing that enzyme activities were very sensitive to climate manipulation after past drought. This microbial pattern also suggests that in these subalpine grasslands previous climatic stress selected opportunistic (or copiotroph) bacterial ecological strategies which negatively influence the functional potential of microbial communities (Evans and Wallenstein 2014). These microbial communities would have been promoted by the release of labile compounds from lysed cells without investment of energy in the production of extracellular enzymes (Litchman et al. 2015). Although it was shown that differences of soil physico-chemical properties such as soil organic matter or pH can impact the resilience (and resistance) of microbial communities (Berard et al. 2015), these properties had limited influence in our study since soil originated from the same grassland with the same abiotic properties before the first treatment. In fact, soil organic matter and soil pH were implied in the resistance or resilience response of extracellular enzymatic activities (Table 3) but were not the main explanatory variables in contrast to plant traits.

In soil previously subjected to climatic stresses, management practices and dominated by P. paniculata (SRD-CF-Patz), the microbial demand for P relative to N increased. Although not significant a similar pattern was observed in soil previously dominated by D. glomerata (SRD-CF-Dact), whereas the microbial demand for P relative to N decreased in all other treatments. This greater microbial investment in enzymes degrading P-rich substrates could explain the relatively stable soil phosphate concentration in both SRD-CF treatments while plant biomass increased during this recovery period. These responses may reflect the mycorrhizal benefits and could be related to the high level of the mycorrhizal colonization in the root of L. perenne growing especially in SRD-CF soils upon rewetting. High mycorrhizal colonization occurred also in Patz soils where mowing has been shown to strongly decrease and limit the negative influence of endophytic alkaloids on AMF (Binet et al. 2017). The increase of mycorrhizal colonization may also explain the observed increase of phosphatase activity (Gianinazzi-Pearson and Gianinazzi 1976) for phosphate acquisition. Moreover, the mycorrhizal benefits for plants were also reflected by the positive correlation between AM colonization and plant N quantity, which may confirm the role of AM symbiosis in nutrient acquisition (through increased soil prospection) and transfer to the host plant (Hodge and Storer 2015).

Altogether, these results suggest a soil legacy of previous climatic stress and management practices (Fig. 5) which could have selected fungal communities able to colonize roots during the rewetting period and increased the enzymatic activity of P relative to N and C (Bell et al. 2014). This mycorrhizal root colonization also allowed plants growing in SRD-CF soils to cope with the lower soil inorganic N availability thanks to a more effective access to soil nutrients and water (van der Heijden et al. 2008; Allen 2009).

Conclusion

In subalpine grasslands the soil legacy of climatic stress had positive effects on plant production and tissue quality through an increase in soil inorganic N content. This increase in nutrient availability was the result of a negative soil legacy of climatic stress that reduced the ability of microbial communities to compete with plants for inorganic N acquisition. The positive effect on plant growth was modulated by previous management practices and plant functional composition legacies, which both reduced the competitive abilities of microbial communities. The resistance of L. perenne to the new drought period, characterized by an increase in root biomass and dry matter content and a decrease of specific root length, resulted from differences in soil legacies such as the lower microbial N competitive abilities (climate legacy) or the facilitation of AMF root colonization (plant functional composition legacy). The rewetting period did not result in the recovery of microbial activities. We suggest that past drought periods probably promoted drought-sensitive copiotroph microorganisms, which were less resistant to drought, but resilient in terms of microbial biomass. In the absence of recovery of microbial functioning high L. perenne growth rates under lower nutrient availability were allowed by AMF fungal colonization of plant roots. This soil legacy of climatic stress was modulated by the legacy of plant functional composition. In subalpine grasslands the increased frequency of stressful weather events such as earlier snow melt and drought could impair soil microbial communities and functioning, and alter critical nutrient recycling that sustains traditional agricultural practices and associated ecosystem services. However, caution should be exercised concerning the legacy effect of soils since the observed response of plant and microbial community occurred in a pot experiment and needs to be confirmed in the field. Although further characterization of microbial communities is required to better understand drought responses of subalpine grassland soils and functioning, our study provides new insights into aboveground and belowground ecosystem responses to global changes.

Notes

Acknowledgements

This study was conducted as part of the ECO-SERVE project through the 2013–2014 BiodivERsA/FACCE-JPI joint call for research proposals, with the national funders ANR, NWO, FCT (BiodivERsA/001/2014), MINECO, FORMAS, and SNSF. The conditioning phase of soils was a part of BiodivERsA project REGARDs with funding from the French Agence Nationale pour la Recherche (ANR). We thank Sophie Périgon for help and assistance for the measurements of arbuscular mycorrhizal colonization. We thank Karl Grigulis for his critical reading of the manuscript and his correction of the language. We also thank the referee for their constructive comments which improved the manuscript substantially.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.

Supplementary material

11104_2017_3403_MOESM1_ESM.docx (593 kb)
ESM 1 (DOCX 592 kb)

References

  1. Acosta-Martinez V, Cotton J, Gardner T et al (2014a) Predominant bacterial and fungal assemblages in agricultural soils during a record drought/heat wave and linkages to enzyme activities of biogeochemical cycling. Appl Soil Ecol 84:69–82.  https://doi.org/10.1016/j.apsoil.2014.06.005 CrossRefGoogle Scholar
  2. Acosta-Martinez V, Moore-Kucera J, Cotton J et al (2014b) Soil enzyme activities during the 2011 Texas record drought/heat wave and implications to biogeochemical cycling and organic matter dynamics. Appl Soil Ecol 75:43–51.  https://doi.org/10.1016/j.apsoil.2013.10.0 CrossRefGoogle Scholar
  3. Allen MF (2009) Bidirectional water flows through the soil-fungal-plant mycorrhizal continuum. New Phytol 182:290–293CrossRefPubMedGoogle Scholar
  4. Alster CJ, German DP, Lu Y, Allison SD (2013) Microbial enzymatic responses to drought and to nitrogen addition in a southern California grassland. Soil Biol Biochem 64:68–79.  https://doi.org/10.1016/j.soilbio.2013.03.034 CrossRefGoogle Scholar
  5. Anderegg WRL, Schwalm C, Biondi F et al (2015) Pervasive drought legacies in forest ecosystems and their implications for carbon cycle models. Science 349:528–532.  https://doi.org/10.1126/science.aab1833 CrossRefPubMedGoogle Scholar
  6. Averill C, Waring BG, Hawkes CV (2016) Historical precipitation predictably alters the shape and magnitude of microbial functional response to soil moisture. Glob Chang Biol 22:1957–1964.  https://doi.org/10.1111/gcb.13219 CrossRefPubMedGoogle Scholar
  7. Baldwin AH, Jensen K, Schoenfeldt M (2014) Warming increases plant biomass and reduces diversity across continents, latitudes, and species migration scenarios in experimental wetland communities. Glob Chang Biol 20:835–850CrossRefPubMedGoogle Scholar
  8. Banning NC, Murphy DV (2008) Effect of heat-induced disturbance on microbial biomass and activity in forest soil and the relationship between disturbance effects and microbial community structure. Appl Soil Ecol 40:109–119.  https://doi.org/10.1016/j.apsoil.2008.03.011 CrossRefGoogle Scholar
  9. Baptist F, Aranjuelo I, Legay N et al (2015) Rhizodeposition of organic carbon by plants with contrasting traits for resource acquisition: responses to different fertility regimes. Plant and Soil 394:391–406.  https://doi.org/10.1007/s11104-015-2531-4
  10. Bardgett RD, Mawdsley JL, Edwards S et al (1999) Plant species and nitrogen effects on soil biological properties of temperate upland grasslands. Funct Ecol 13:650–660.  https://doi.org/10.1046/j.1365-2435.1999.00362.x CrossRefGoogle Scholar
  11. Bardgett RD, Manning P, Morrien E, De Vries FT (2013) Hierarchical responses of plant-soil interactions to climate change: consequences for the global carbon cycle. J Ecol 101:334–343.  https://doi.org/10.1111/1365-2745.12043 CrossRefGoogle Scholar
  12. Bardgett RD, Mommer L, De Vries FT (2014) Going underground: root traits as drivers of ecosystem processes. Trends Ecol Evol 29:692–699.  https://doi.org/10.1016/j.tree.2014.10.006 CrossRefPubMedGoogle Scholar
  13. Barnard RL, Osborne CA, Firestone MK (2013) Responses of soil bacterial and fungal communities to extreme desiccation and rewetting. ISME J 7:2229–2241.  https://doi.org/10.1038/ismej.2013.104 CrossRefPubMedPubMedCentralGoogle Scholar
  14. Barnett TP, Adam JC, Lettenmaier DP (2005) Potential impacts of a warming climate on water availability in snow-dominated regions. Nature 438:303–309.  https://doi.org/10.1038/nature04141 CrossRefPubMedGoogle Scholar
  15. Bell CW, Fricks BE, Rocca JD, Steinweg JM, McMahon SK, Wallenstein MD (2013) High-throughput fluorometric measurement of potential soil extracellular enzyme activities. J Vis Exp 81:e50961.  https://doi.org/10.3791/50961
  16. Bell CW, Tissue DT, Loik ME et al (2014) Soil microbial and nutrient responses to 7years of seasonally altered precipitation in a Chihuahuan Desert grassland. Glob Chang Biol 20:1657–1673.  https://doi.org/10.1111/gcb.12418 CrossRefPubMedGoogle Scholar
  17. Benot M-L, Saccone P, Vicente R et al (2013) How extreme summer weather may limit control of Festuca Paniculata by mowing in subalpine grasslands. Plant Ecol Divers 6:393–404.  https://doi.org/10.1080/17550874.2013.784818 CrossRefGoogle Scholar
  18. Benot M-L, Saccone P, Pautrat E et al (2014) Stronger short-term effects of mowing than extreme summer weather on a subalpine grassland. Ecosystems 17:458–472.  https://doi.org/10.1007/s10021-013-9734-4 CrossRefGoogle Scholar
  19. Berard A, Ben Sassi M, Renault P, Gros R (2012) Severe drought-induced community tolerance to heat wave. An experimental study on soil microbial processes. J Soils Sediments 12:513–518.  https://doi.org/10.1007/s11368-012-0469-1 CrossRefGoogle Scholar
  20. Berard A, Ben Sassi M, Kaisermann A, Renault P (2015) Soil microbial community responses to heat wave components: drought and high temperature. Clim Res 66:243–264.  https://doi.org/10.3354/cr01343 CrossRefGoogle Scholar
  21. Bernard L (2017) Mécanismes fonctionnels de résilience des prairies subalpines au changement global. PhD thesis, Université Grenoble-AlpesGoogle Scholar
  22. Binet M-N, van Tuinen D, Souard F et al (2017) Responses of above- and below-ground fungal symbionts to cessation of mowing in subalpine grassland. Fungal Ecol 25:14–21.  https://doi.org/10.1016/j.funeco.2016.10.001 CrossRefGoogle Scholar
  23. Brookes P, Landman A, Pruden G, Jenkinson D (1985) Chloroform fumigation and the release of soil-nitrogen - a rapid direct extraction method to measure microbial biomass nitrogen in soil. Soil Biol Biochem 17:837–842.  https://doi.org/10.1016/0038-0717(85)90144-0 CrossRefGoogle Scholar
  24. Brunner I, Herzog C, Dawes MA et al (2015) How tree roots respond to drought. Front Plant Sci 6:547.  https://doi.org/10.3389/fpls.2015.00547 CrossRefPubMedPubMedCentralGoogle Scholar
  25. Comas LH, Becker SR, Cruz VMV et al (2013) Root traits contributing to plant productivity under drought. Front Plant Sci 4:442.  https://doi.org/10.3389/fpls.2013.00442 CrossRefPubMedPubMedCentralGoogle Scholar
  26. Cong W-F, van Ruijven J, van der Werf W et al (2015) Plant species richness leaves a legacy of enhanced root litter-induced decomposition in soil. Soil Biol Biochem 80:341–348.  https://doi.org/10.1016/j.soilbio.2014.10.017 CrossRefGoogle Scholar
  27. Core Team R (2015) R: a language and environment for statistical computing. R Foundation for Statistical Computing, ViennaGoogle Scholar
  28. De Deyn GB, Cornelissen JHC, Bardgett RD (2008) Plant functional traits and soil carbon sequestration in contrasting biomes. Ecol Lett 11:516–531.  https://doi.org/10.1111/j.1461-0248.2008.01164.x CrossRefPubMedGoogle Scholar
  29. de Vries FT, Liiri ME, Bjornlund L et al (2012) Legacy effects of drought on plant growth and the soil food web. Oecologia 170:821–833.  https://doi.org/10.1007/s00442-012-2331-y CrossRefPubMedGoogle Scholar
  30. de Vries FT, Shade A (2013) Controls on soil microbial community stability under climate change. Front Microbiol 4:265.  https://doi.org/10.3389/fmicb.2013.00265
  31. de Vries FT, Brown C, Stevens CJ (2016) Grassland species root response to drought: consequences for soil carbon and nitrogen availability. Plant Soil 409:297–312.  https://doi.org/10.1007/s11104-016-2964-4 CrossRefGoogle Scholar
  32. Debouk H, de Bello F, Sebastia M-T (2015) Functional trait changes, productivity shifts and vegetation stability in mountain grasslands during a short-term warming. PLoS One 10:e0141899.  https://doi.org/10.1371/journal.pone.0141899 CrossRefPubMedPubMedCentralGoogle Scholar
  33. Della-Marta PM, Haylock MR, Luterbacher J, Wanner H (2007) Doubled length of western European summer heat waves since 1880. J Geophys Res-Atmos 112:D15103.  https://doi.org/10.1029/2007JD008510 CrossRefGoogle Scholar
  34. Dijkstra FA, He M, Johansen MP et al (2015) Plant and microbial uptake of nitrogen and phosphorus affected by drought using N-15 and P-32 tracers. Soil Biol Biochem 82:135–142.  https://doi.org/10.1016/j.soilbio.2014.12.021 CrossRefGoogle Scholar
  35. Dray S, Dufour AB, Chessel D (2007) The ade4 package-II: two-table and K-table methods. R News 7:47–52Google Scholar
  36. Edwards AC, Scalenghe R, Freppaz M (2007) Changes in the seasonal snow cover of alpine regions and its effect on soil processes: a review. Quat Int 162:172–181.  https://doi.org/10.1016/j.quaint.2006.10.027
  37. Evans SE, Wallenstein MD (2012) Soil microbial community response to drying and rewetting stress: does historical precipitation regime matter? Biogeochemistry 109:101–116.  https://doi.org/10.1007/s10533-011-9638-3
  38. Evans SE, Wallenstein MD (2014) Climate change alters ecological strategies of soil bacteria. Ecol Lett 17:155–164.  https://doi.org/10.1111/ele.12206
  39. Evans SE, Byrne KM, Lauenroth WK, Burke IC (2011) Defining the limit to resistance in a drought-tolerant grassland: long-term severe drought significantly reduces the dominant species and increases ruderals. J Ecol 99:1500–1507.  https://doi.org/10.1111/j.1365-2745.2011.01864.x
  40. Fierer N, Schimel JP, Holden PA (2003) Influence of drying-rewetting frequency on soil bacterial community structure. Microb Ecol 45:63–71.  https://doi.org/10.1007/s00248-002-1007-2
  41. Fierer N, Bradford MA, Jackson RB (2007) Toward an ecological classification of soil bacteria. Ecology 88:1354–1364.  https://doi.org/10.1890/05-1839
  42. Fornara DA, Wasson E-A, Christie P, Watson CJ (2016) Long-term nutrient fertilization and the carbon balance of permanent grassland: any evidence for sustainable intensification? Biogeosciences 13:4975–4984.  https://doi.org/10.5194/bg-13-4975-2016
  43. Fuchslueger L, Bahn M, Hasibeder R et al (2016) Drought history affects grassland plant and microbial carbon turnover during and after a subsequent drought event. J Ecol 104:1453–1465.  https://doi.org/10.1111/1365-2745.12593
  44. Gepstein S (2004) Leaf senescence - not just a “wear and tear” phenomenon. Genome Biol 5:212.  https://doi.org/10.1186/gb-2004-5-3-212
  45. Gianinazzi-Pearson V, Gianinazzi S (1976) Enzymatic studies on the metabolism of vesicular-arbuscular mycorrhiza. I. Effect of mycorrhiza formation and phosphorus nutrition on soluble phosphatase activities in onion roots. Physiologie Vegetale 14:833–841Google Scholar
  46. Grassein F, Till-Bottraud I, Lavorel S (2010) Plant resource-use strategies: the importance of phenotypic plasticity in response to a productivity gradient for two subalpine species. Ann Bot 106:637–645.  https://doi.org/10.1093/aob/mcq154
  47. Grigulis K, Lavorel S, Krainer U et al (2013) Relative contributions of plant traits and soil microbial properties to mountain grassland ecosystem services. J Ecol 101:47–57.  https://doi.org/10.1111/1365-2745.12014
  48. Gritsch A, Dirnboeck T, Dullinger S (2016) Recent changes in alpine vegetation differ among plant communities. J Veg Sci 27:1177–1186.  https://doi.org/10.1111/jvs.12447
  49. Gross N, Nash SK, Sandra L (2007) Leaf dry matter content and lateral spread predict response to land use change for six subalpine grassland species. J Veg Sci 18:289–300.  https://doi.org/10.1111/j.1654-1103.2007.tb02540.x
  50. Hamilton EW, Frank DA, Hinchey PM, Murray TR (2008) Defoliation induces root exudation and triggers positive rhizospheric feedbacks in a temperate grassland. Soil Biol Biochem 40:2865–2873.  https://doi.org/10.1016/j.soilbio.2008.08.007
  51. Hartmann AA, Barnard RL, Marhan S, Niklaus PA (2013) Effects of drought and N-fertilization on N cycling in two grassland soils. Oecologia 171:705–717.  https://doi.org/10.1007/s00442-012-2578-3
  52. Hawkes CV, Keitt TH (2015) Resilience vs. historical contingency in microbial responses to environmental change. Ecol Lett 18:612–625.  https://doi.org/10.1111/ele.12451
  53. Henry HAL (2013) Reprint of “Soil extracellular enzyme dynamics in a changing climate.”. Soil Biol Biochem 56:53–59.  https://doi.org/10.1016/j.soilbio.2012.10.022
  54. Hodge A, Storer K (2015) Arbuscular mycorrhiza and nitrogen: implications for individual plants through to ecosystems. Plant Soil 386:1–19.  https://doi.org/10.1007/s11104-014-2162-1
  55. IPCC (2014) IPCC climate change 2014: synthesis report. Contribution of working groups I, II, and III to the fifth assessment report of the intergovernmental panel on climate change. Cambridge University Press, CambridgeGoogle Scholar
  56. Jones DL, Willett VB (2006) Experimental evaluation of methods to quantify dissolved organic nitrogen (DON) and dissolved organic carbon (DOC) in soil. Soil Biol Biochem 38:991–999.  https://doi.org/10.1016/j.soilbio.2005.08.012 CrossRefGoogle Scholar
  57. Jung V, Albert CH, Violle C et al (2014) Intraspecific trait variability mediates the response of subalpine grassland communities to extreme drought events. J Ecol 102:45–53.  https://doi.org/10.1111/1365-2745.12177 CrossRefGoogle Scholar
  58. Kastovska E, Edwards K, Picek T, Santruckova H (2015) A larger investment into exudation by competitive versus conservative plants is connected to more coupled plant-microbe N cycling. Biogeochemistry 122:47–59.  https://doi.org/10.1007/s10533-014-0028-5 CrossRefGoogle Scholar
  59. Khalili B, Ogunseitan OA, Goulden ML, Allison SD (2016) Interactive effects of precipitation manipulation and nitrogen addition on soil properties in California grassland and shrubland. Appl Soil Ecol 107:144–153.  https://doi.org/10.1016/j.apsoil.2016.05.018 CrossRefGoogle Scholar
  60. Klironomos J, Zobel M, Tibbett M et al (2011) Forces that structure plant communities: quantifying the importance of the mycorrhizal symbiosis. New Phytol 189:366–370.  https://doi.org/10.1111/j.1469-8137.2010.03550.x CrossRefPubMedGoogle Scholar
  61. Körner C (1999) Alpine plant life; functional plant ecology of high mountain ecosystems. Springer-Verlag, Berlin HeidelbergGoogle Scholar
  62. Lavorel S, Grigulis K, Lamarque P et al (2011) Using plant functional traits to understand the landscape distribution of multiple ecosystem services. J Ecol 99:135–147.  https://doi.org/10.1111/j.1365-2745.2010.01753.x CrossRefGoogle Scholar
  63. Legay N, Grassein F, Robson TM et al (2013) Comparison of inorganic nitrogen uptake dynamics following snowmelt and at peak biomass in subalpine grasslands. Biogeosciences 10:7631–7645.  https://doi.org/10.5194/bg-10-7631-2013 CrossRefGoogle Scholar
  64. Legay N, Personeni E, Slezack-Deschaumes S et al (2014) Grassland species show similar strategies for sulphur and nitrogen acquisition. Plant Soil 375:113–126.  https://doi.org/10.1007/s11104-013-1949-9 CrossRefGoogle Scholar
  65. Lenth RV (2016) Least-squares means: the R package lsmeans. J Stat Softw 69(1):1–33.  10.18637/jss.v069.i01 CrossRefGoogle Scholar
  66. Lipson DA, Schmidt SK (2004) Seasonal changes in an alpine soil bacterial community in the Colorado Rocky Mountains. Appl Environ Microbiol 70:2867–2879.  https://doi.org/10.1128/AEM.70.5.2867-2879.2004 CrossRefPubMedPubMedCentralGoogle Scholar
  67. Lipson DA, Schmidt SK, Monson RK (1999) Links between microbial population dynamics and nitrogen availability in an alpine ecosystem. Ecology 80:1623–1631CrossRefGoogle Scholar
  68. Litchman E, Edwards KF, Klausmeier CA (2015) Microbial resource utilization traits and trade-offs: implications for community structure, functioning, and biogeochemical impacts at present and in the future. Front Microbiol.  https://doi.org/10.3389/fmicb.2015.00254
  69. Loucougaray G, Dobremez L, Gos P et al (2015) Assessing the effects of grassland management on forage production and environmental quality to identify paths to ecological intensification in mountain grasslands. Environ Manag 56:1039–1052.  https://doi.org/10.1007/s00267-015-0550-9 CrossRefGoogle Scholar
  70. Matteodo M, Ammann K, Verrecchia EP, Vittoz P (2016) Snowbeds are more affected than other subalpine-alpine plant communities by climate change in the Swiss alps. Ecol Evol 6:6969–6982.  https://doi.org/10.1002/ece3.2354 CrossRefPubMedPubMedCentralGoogle Scholar
  71. Millar NS, Bennett AE (2016) Stressed out symbiotes: hypotheses for the influence of abiotic stress on arbuscular mycorrhizal fungi. Oecologia 182:625–641.  https://doi.org/10.1007/s00442-016-3673-7 CrossRefPubMedPubMedCentralGoogle Scholar
  72. Mills RTE, Gavazov KS, Spiegelberger T et al (2014) Diminished soil functions occur under simulated climate change in a sup-alpine pasture, but heterotrophic temperature sensitivity indicates microbial resilience. Sci Total Environ 473:465–472.  https://doi.org/10.1016/j.scitotenv.2013.12.071 CrossRefPubMedGoogle Scholar
  73. Mohan JE, Cowden CC, Baas P et al (2014) Mycorrhizal fungi mediation of terrestrial ecosystem responses to global change: mini-review. Fungal Ecol 10:3–19.  https://doi.org/10.1016/j.funeco.2014.01.005 CrossRefGoogle Scholar
  74. Nie M, Pendall E, Bell C et al (2013) Positive climate feedbacks of soil microbial communities in a semi-arid grassland. Ecol Lett 16:234–241.  https://doi.org/10.1111/ele.12034 CrossRefPubMedGoogle Scholar
  75. Oksanen J, Blanchet FG, Kindt R, Legendre P et al (2015) Vegan: community ecology package. R package version 2.3-1. http://CRAN.R-project.org/package=vegan
  76. Ortiz N, Armada E, Duque E et al (2015) Contribution of arbuscular mycorrhizal fungi and/or bacteria to enhancing plant drought tolerance under natural soil conditions: effectiveness of autochthonous or allochthonous strains. J Plant Physiol 174:87–96.  https://doi.org/10.1016/j.jplph.2014.08.019 CrossRefPubMedGoogle Scholar
  77. Perez-Harguindeguy N, Diaz S, Garnier E et al (2013) New handbook for standardised measurement of plant functional traits worldwide. Aust J Bot 61:167–234.  https://doi.org/10.1071/BT12225 CrossRefGoogle Scholar
  78. Peuke A, Rennenberg H (2004) Carbon, nitrogen, phosphorus, and sulphur concentration and partitioning in beech ecotypes (Fagus Sylvatica L.): phosphorus most affected by drought. Trees-Struct Funct 18:639–648.  https://doi.org/10.1007/s00468-004-0335-x CrossRefGoogle Scholar
  79. Phillips J, Hayman D (1970) Improved procedures for clearing roots and staining parasitic and vesicular-Arbuscular Mycorrhizal fungi for rapid assessment of infection. Trans Br Mycol Soc 55:158–161CrossRefGoogle Scholar
  80. Pinheiro J, Bates D, DebRoy S, Sarkar D, R Core Team (2016) nlme: linear and nonlinear mixed effects models. R package version 3.1–126. http://CRAN.Rproject.org/package=nlme
  81. Polley HW, Derner JD, Jackson RB et al (2014) Impacts of climate change drivers on C-4 grassland productivity: scaling driver effects through the plant community. J Exp Bot 65:3415–3424.  https://doi.org/10.1093/jxb/eru009 CrossRefPubMedGoogle Scholar
  82. Puissant J, Cecillon L, Mills RTE et al (2015) Seasonal influence of climate manipulation on microbial community structure and function in mountain soils. Soil Biol Biochem 80:296–305.  https://doi.org/10.1016/j.soilbio.2014.10.013 CrossRefGoogle Scholar
  83. Robson TM, Lavorel S, Clement J-C, Le Roux X (2007) Neglect of mowing and manuring leads to slower nitrogen cycling in subalpine grasslands. Soil Biol Biochem 39:930–941.  https://doi.org/10.1016/j.soilbio.2006.11.004 CrossRefGoogle Scholar
  84. Robson T, Baptist F, Clement J-C, Lavorel S (2010) Land use in subalpine grasslands affects nitrogen cycling via changes in plant community and soil microbial uptake dynamics. J Ecol 98:62–73.  https://doi.org/10.1111/j.1365-2745.2009.01609.x CrossRefGoogle Scholar
  85. Rolli E, Marasco R, Vigani G et al (2015) Improved plant resistance to drought is promoted by the root-associated microbiome as a water stress-dependent trait. Environ Microbiol 17:316–331.  https://doi.org/10.1111/1462-2920.12439 CrossRefPubMedGoogle Scholar
  86. Rustad LE, Campbell JL, Marion GM et al (2001) A meta-analysis of the response of soil respiration, net nitrogen mineralization, and aboveground plant growth to experimental ecosystem warming. Oecologia 126:543–562CrossRefPubMedGoogle Scholar
  87. Sala OE, Gherardi LA, Reichmann L et al (2012) Legacies of precipitation fluctuations on primary production: theory and data synthesis. Philos Trans R Soc B-Biol Sci 367:3135–3144.  https://doi.org/10.1098/rstb.2011.0347 CrossRefGoogle Scholar
  88. Sardans J, Penuelas J, Estiarte M (2006) Warming and drought alter soil phosphatase activity and soil P availability in a Mediterranean shrubland. Plant Soil 289:227–238.  https://doi.org/10.1007/s11104-006-9131-2 CrossRefGoogle Scholar
  89. Schimel JP, Gulledge JM, Clein-Curley JS et al (1999) Moisture effects on microbial activity and community structure in decomposing birch litter in the Alaskan taiga. Soil Biol Biochem 31:831–838.  https://doi.org/10.1016/S0038-0717(98)00182-5 CrossRefGoogle Scholar
  90. Schimel J, Balser TC, Wallenstein M (2007) Microbial stress-response physiology and its implications for ecosystem function. Ecology 88:1386–1394.  https://doi.org/10.1890/06-0219 CrossRefPubMedGoogle Scholar
  91. Sinsabaugh RL, Hill BH, Shah JJF (2009) Ecoenzymatic stoichiometry of microbial organic nutrient acquisition in soil and sediment. Nature 462:795–U117.  https://doi.org/10.1038/nature08632 CrossRefPubMedGoogle Scholar
  92. Trouvelot A, Fardeau J, Plenchette C et al (1986) Nutritional balance and symbiotic expression in Mycorrhizal wheat. Physiologie Vegetale 24:300–300Google Scholar
  93. van de Voorde TFJ, van der Putten WH, Bezemer TM (2011) Intra- and interspecific plant-soil interactions, soil legacies and priority effects during old-field succession. J Ecol 99:945–953.  https://doi.org/10.1111/j.1365-2745.2011.01815.x
  94. van der Heijden MGA, Bardgett RD, van Straalen NM (2008) The unseen majority: soil microbes as drivers of plant diversity and productivity in terrestrial ecosystems. Ecol Lett 11:296–310.  https://doi.org/10.1111/j.1461-0248.2007.01139.x CrossRefPubMedGoogle Scholar
  95. Vance E, Brookes P, Jenkinson D (1987) An extraction method for measuring soil microbial biomass-C. Soil Biol Biochem 19:703–707.  https://doi.org/10.1016/0038-0717(87)90052-6 CrossRefGoogle Scholar
  96. Venables WN, Ripley BD (2002) Modern applied statistics with S, Fourth edn. Springer, New YorkCrossRefGoogle Scholar
  97. Vittoz P, Randin C, Dutoit A et al (2009) Low impact of climate change on subalpine grasslands in the Swiss northern alps. Glob Chang Biol 15:209–220.  https://doi.org/10.1111/j.1365-2486.2008.01707.x CrossRefGoogle Scholar
  98. Wu QS, Zou YN (2017) Arbuscular Mycorrhizas and stress tolerance of plants. Springer-Verlag, Singapore, pp 25–41CrossRefGoogle Scholar
  99. Xu X, Polley HW, Hofmockel K, Wilsey BJ (2017) Species composition but not diversity explains recovery from the 2011 drought in Texas grasslands. Ecosphere 8:e01704.  https://doi.org/10.1002/ecs2.1704 CrossRefGoogle Scholar
  100. Zinger L, Shahnavaz B, Baptist F et al (2009) Microbial diversity in alpine tundra soils correlates with snow cover dynamics. ISME J 3:850–859.  https://doi.org/10.1038/ismej.2009.20 CrossRefPubMedGoogle Scholar
  101. Zwicke M, Picon-Cochard C, Morvan-Bertrand A et al (2015) What functional strategies drive drought survival and recovery of perennial species from upland grassland? Ann Bot 116:1001–1015.  https://doi.org/10.1093/aob/mcv037 CrossRefPubMedPubMedCentralGoogle Scholar

Copyright information

© Springer International Publishing AG 2017

Authors and Affiliations

  • Nicolas Legay
    • 1
    • 2
    • 3
    Email author return OK on get
  • Gabin Piton
    • 1
  • Cindy Arnoldi
    • 1
  • Lionel Bernard
    • 1
  • Marie-Noëlle Binet
    • 1
  • Bello Mouhamadou
    • 1
  • Thomas Pommier
    • 4
  • Sandra Lavorel
    • 1
  • Arnaud Foulquier
    • 1
  • Jean-Christophe Clément
    • 1
    • 5
  1. 1.Laboratoire d’Ecologie AlpineUMR CNRS-UGA-USMB 5553, Université Grenoble AlpesGrenoble cedexFrance
  2. 2.Ecole de la Nature et du PaysageINSA Centre Val de LoireBloisFrance
  3. 3.CNRS, CITERES, UMR 7324ToursFrance
  4. 4.Ecologie MicrobienneINRA UMR1418, CNRS UMR5557, Université de LyonVilleurbanne CedexFrance
  5. 5.CARRTEL, INRA, Université Savoie Mont BlancThonon-Les-BainsFrance

Personalised recommendations