Six strains of Agaricus spp. from the Collection of germplasms of Agaricus in Bordeaux (CGAB), INRA were used: strains CA454, CA560, CA572, CA646 and CA647 for A. subrufescens and one strain Bs527 for A. bisporus. CA454 is a subculture of the collection strain WC837 in PSUMCC registered as A. blazei and assumed to be similar to ATCC 76739, which is claimed to be a copy of the original strain cultivated for the first time in Brazil in the 1960s. CA646 and CA647 are two cultivars sold in Europe by Mycelia (Belgium) under the reference numbers 7700 and 7703, respectively. CA560 was isolated in Botucatu, Brazil (1999) from a commercial farm and CA572 is a Brazilian cultivar purchased by D.C. Zied in 2007; both strains have proved to be genetically different [9]. Their original reference numbers are ABL-99/28 and ABL-07/58, respectively, in the Culture Collection of the Edible and Medicinal Mushroom Module of São Paulo State University (UNESP). An A. bisporus strain, Bs527, for which storage is commonly performed at 2–4 °C, was used as control. Bs527 was a copy of the cultivar 30A (Euromycel) placed in the CGAB in 1997.

Precultures were performed in glass Petri dishes containing 12 g dry weight of cooked and coated rye grains (spawn substrate used by spawn makers, provided by Somycel, Langeais, France) inoculated with 6 plugs (1 cm × 1 cm) of malt agar cultures. They were incubated in a climatic chamber at 25 °C and 90% RH (relative humidity) for 15 days. Then, to evaluate the effect of two factors i.e. time (15 or 30 days), and temperature (10 or 15  C) of storage, these precultures were incubated in a climatic chamber either at 10 °C or 15 °C for 15 or 30 days and were used immediately after storage to inoculate the mesocosms. For controls, the precultures were directly used to inoculate the mesocosms. These mesocosms were prepared with 450 g of fresh mushroom compost (135 g dry weight) placed in a 2-L glass pot in order to obtain a final volume of compost of 20 cm × 16 cm × 6 cm (length × width × height). They were inoculated with 3% (w/w) of rye grains (dry weight), which were thoroughly mixed with the compost under aseptic conditions. The compost was commercial compost produced by SARL Renaud et Fils at Avy, France, for the cultivation of A. bisporus. The main raw materials used were horse manure and wheat-straw, and the phase I of composting was performed indoor. Compost quality was assessed by the yields obtained in commercial cultivation facilities that reached standards. Three mesocosms were prepared for each strain and were incubated in a climatic chamber at 25 °C and 90% of RH for 15 days in darkness. An un-inoculated control (with the compost alone), UIC, was also prepared and incubated under the same conditions.

For each mesocosm after incubation at 25 °C for 15 days, enzyme extraction was performed as previously described by Velazquez et al. (2004) using 10 g of a composite sample (10 samplings per mesocosm) in a 1-L flask containing 200 mL of an extraction solution (polyvinylpolypyrolidone 5.7 g, CaCl₂, 0.2 M, Tween 80, 0.05%). These samples were subjected to axial shaking for 1 h at 120 rpm at room temperature. The solids were eliminated by centrifugation at 10,000 g for 15 min and filtration through Whatman GF/D filters (2.7 μm) and through Whatman GF/C filters (1.7 μm). The filtrates of each extract were concentrated overnight at 4 °C in dialysis tubes (15 kda cut-off) using polyethyleneglycol to a final volume of 10% of initial volume.

Laccase activities were measured by monitoring the oxidation of syringaldazine to quinone (ɛ^M = 65,000 M⁻¹·cm⁻¹) at 525 nm in acetate buffer, 0.1 M, pH 5 on a spectrophotometer Biomate 3 (Fischer Bioblock Scientific). One unit per mL (U.mL⁻¹) of laccase activity is defined as the amount of μmol of quinone of syringaldazine produced per min and per mL of sample at room temperature. Three assays were performed for each enzyme extract.

CM-cellulase activities were measured with carboxymethylcellulose (CMC) and using a simplified methodology adapted from that of Somogyi–Nelson, as previously described [10]. Solution A was composed of 50 mL of CuSO₄, 2%, Na₂SO₄, 4 g, complemented with 50 mL of a solution of Na₂CO₃, 25 g, NaHCO₃, 20 g and sodium potassium tartrate, 25 g, q.s.p. 1 L. Solution B was composed of ammonium-molybdate, 25 g, H₂SO₄, 21 mL, and Na₂HAsO₄ 7H₂O, 3 g, q.s.p. 450 mL. To measure CM-cellulase activity, 0.1 mL of the extract was incubated at 50 °C for 1 h in 0.9 mL of 50 mM acetate buffer (pH 5.0) with 0.1% CMC. 1 mL of solution A was added to 1 mL of the reaction medium and 1 mL of distilled water; then the mixture was boiled for 20 min and cooled down in a cold water bath for 15 min. Then, 1 mL of solution B was added and the mixture was left for 50 min at room temperature and then centrifuged for 2 min at 10,000 g. The absorption was measured at 870 nm. Dilutions were performed when necessary. For the calibration curves, 1 mL of solution A was added to 1 mL, 750 μL, 500 μL, 250 μL of a solution of glucose (80 mg/L) q.s. 1 mL. Three assays were performed for each enzyme extract.

For each mesocosm after incubation at 25 °C for 15 days, 10 g of a composite sample (10 samplings per mesocosm) were vigorously shaken in 100 mL of a sterilized desorption-solution (sodium pyrophosphate 0.08 g/L) for 1 h and then brought to a final OD = 0.001 at 595 nm. One hundred and fifty microliters of the extracted solution were used to inoculate all 96-wells of a BIOLOG© EcoPlate (Biolog, California, USA). One EcoPlate™ was used for three mesocosm extracts, since each 96-well plate contained three replicates, each with 31 different carbon sources, one per well, and a water blank. The plates were incubated at 25 °C for 48 h and the absorbance was read at 590 nm using a microplate reader (Tecan, France). To measure the functional diversity (FD), the Shannon's diversity index [11] was calculated [12] from: FD = –Σ p_i(ln p_i), where p_i is the ratio of color development of well i to the sum of color development of all positive wells. The microbial activity of each microplate was assessed as average well color development (AWCD) calculated as follows: AWCD = Σ OD_i/31, where OD_i is the optical density for each well.

An amount of 10 g of substrate, sampled as described above, were lyophilized and then ground and sieved (1 mm). CP/MAS ¹³C NMR spectra were obtained on a Bruker DSX 400 MHz spectrophotometer operating at 100.7 MHz. Then, samples (600 mg) were spun at 10 KHz at the magic angle. Contact times of 2 ms were applied with a pulse width of 2.8 μs and a recycle delay of 3 s. Chemical shift values were referenced to glycine signal (carbonyl C at 176.03 ppm). The ¹³C NMR spectra were divided into seven chemical shift regions, according to [13]. Dmfit 2003 software was used to determine the intensity of each chemical shift region [14]. An index of decomposition (HR#1 Alkyl–C/O–Alkyl–C) was calculated as described before [15]. An aromaticity ratio was also calculated to estimate the degree of humification of organic matter and Alkyl–C/COOH–C ratio HR#2 was used to assess organic matter transformation using the alkyl chain length of organic acids.

The significance of differences (enzyme activities, AWCD and H′ from Biolog, NMR ratio) between the means of three experiments was tested by non-parametric multiple comparison of Kruskal–Wallis. Each experiment was considered as an independent sample. The optical absorbances of the 30 wells in the ECO plate were subjected to Principal Component Analysis (PCA) after 48 h of incubation, considering each substrate as a variable. Statistica Vs 6 (StatSoft, Maison-Alfort, France) was used for statistical analysis and a P-value < 0.05 is considered as significant.

Agaricus subrufescens is known to be sensitive to low temperatures, and this particular characteristic is an issue to define adequate storage conditions [16]. Thus, ideal storage conditions, with limited consequences on the ability to colonize cultivation substrate, were still to be defined. Substrate colonization can be assessed by monitoring the extracellular enzyme involved in lignocellulosic material transformation [17]. For these lignocellulolytic activities, the storage of the spawn had an effect on cellulase and laccase activities measured in the composted substrate: after certain storage conditions, these activities were indeed enhanced. Fig. 1 A and B shows the tendency observed for the A. subrufescens strains studied, using the results obtained for strains CA572 and CA 647 for cellulase and laccase activities, respectively. Cellulase activities were significantly higher than in controls (no storage of the spawn) when spawn was stored for 30 days at 10 or 15 °C (for instance with strain CA 572 in Fig. 1A). Concerning laccase activities, storage had no effect or led to higher laccase activities after spawn incubation at 15 °C for 30 days (Fig. 1B with strain CA 647). In non-inoculated compost, no laccase and cellulase activities were detected. Thus, surprisingly, storage at cold temperatures seems to have a positive effect on the production of the extracellular enzymes involved in the transformation of lignocellulosic substrate during spawn running. The chemical transformation of the substrate was monitored via solid-state NMR of ¹³C: different ratios [14,15], an aromaticity ratio (AR) and two humification ratios (HR1 = Alkyl–C/O–Alkyl–C and HR2 = Alkyl–C/COOH–C) were calculated from the NMR spectra to determine how storage conditions of the spawn may affect the capacity of transformation of the substrate. For AR, when an effect was observed, it was always a higher ratio in the case of substrate colonized by spawn previously stored at 10 or 15 °C than in control (Table 1). The same tendency was observed for HR1. Higher values in both AR and HR1 indicate that the recalcitrant fraction of organic matter (linked to alkyl–C and aromatic–C signals and thus assigned to lipids and phenolic compounds) accumulated in the substrate, while O–Alkyl–C signal, assigned to polysaccharides, decreased. These results clearly indicate that the easily biodegradable fraction of organic matter, i.e. mainly polysaccharides, is more effectively transformed in the substrate inoculated with spawn previously stored at 10 or 15 °C. Thus, our results obtained with solid-state NMR of ¹³C corroborated those found with lignocellulolytic activities, both microbial and chemical markers indicating that organic matter is actively transformed when spawn is previously stored. For HR2, a higher ratio in substrate inoculated with pre-incubated spawn than in control shows that long alkyl chain acids were predominantly observed. These molecules can be linked to phospholipids from cell membranes [18] and this suggests that fungal biomass may be more extensively produced in solid-state fermentation under these conditions, i.e. when the inoculum has been previously stored at 10 or 15 °C for two or four weeks. Spawn storage at 10 or 15 °C seems to enhance the potential for colonization of the culture substrate during spawn running, as described by both microbial and chemical markers used here. This can also be linked with the efficient polysaccharide transformation shown by the increase in both the AR and HR1 ratios.

A. subrufescens grows at higher optimal temperatures than A. bisporus: for instance in previous studies [19], the authors found that optimal temperatures for mycelial growth was around 30 °C, depending on the strains studied. De Mendoca et al. [16] reported optimal temperatures ranging from 25 to 28 °C and Llarena et al. [20] optimal temperatures varying from 26 to 30 °C, with a similar range of variation for cultivars and wild strains. Long-time storage at low temperatures such as 4 °C is known to be lethal for A. subrufescens [21]. Mycelial growth rate decreased gradually over storage time at 4 °C of spawn on sorghum grains used for long-term storage of genetic resources [22]. Here, we found that storage at 10 or 15 °C for a period as long as 30 days did not strongly affect the potential of different A. subrufescens strains in their ability of early colonization and transformation of the compost used as cultivation substrate during the spawn running period.

To thoroughly investigate the consequences of mycelium storage on A. subrufescens ability to colonize compost and how it may change its biotic environment, the diversity of microbial communities of the substrate was also considered via their functional structure. Ecoplates are known to focus on the functional diversity of the bacterial communities since the reduction of tetrazolium dye cannot be accomplished by fungi [23,24]. Thus, the Agaricus strain, which colonizes the culture substrate, is not supposed to be involved in the catabolic potential measured here. Global microbial activity was measured by Average Well Color Development (AWCD) and functional diversity via the Shannon–Weaver Index, H′ [11,12]. As for lignocellulolytic activities, a similar tendency was observed for all the A. subrufescens under study and Fig. 2A and B shows the results obtained with strains CA 454 and 647 as an example. In most of the treatments, AWCD was the same as in control. When AWCD varied from the control, higher values were observed whatever the temperature of storage considered (Fig. 2A). For functional diversity characterized by H′, the same result was observed: a higher H′ value was found when spawns were stored at 10 °C or 15 °C than in control (Fig. 2B). Thus, storage of the spawn had probably an effect on A. subrufescens mycelium, which consequently influenced the microbial communities during mycelial colonization of compost, by enhancing both their global catabolic activity – and thus their potential to transform the substrate – and their biodiversity. Moreover, the high diversity of microbial communities of the substrate is of major importance in the selectivity of the mushroom culture substrate, since it prevents the colonization of potential antagonists [25]. Fig. 3 shows the result of PCA analysis using optical densities of all the substrates used as variables. Axes F1 and F2 accounted for 59.39 and 10.42% of the variance, respectively. This figure shows that the temperature used for storage (10 or 15 °C) did not globally affect microbial catabolic diversity, since the projections were not clearly differentiated on the PCA (squares vs triangles). On the other hand, the time of storage (15 or 30 days) influenced this diversity (black vs white) mainly according to axis F2, as observed for strains CA647 and CA560, and to a lesser extent according to F1, for strains CA572 and CA646. PCA was used to give additional information about how the functional diversity of these communities varied. Though the effects of both time and temperature of storage were observed, the more drastic conditions, i.e. 10 °C and 30 days, did not lead to a strong variation in the diversity structure of substrate microbial communities according to the PCA. Colonization of the substrate by the mycelium of A. subrufescens may have modified the associated microbial communities because of microbial interactions, including competition for nutrients or production of antagonistic metabolites. Here, we found that the diversity index, H′, remained constant or was even higher after spawn storage and PCA added complementary information by indicating that for certain strain, this functional diversity actually changed compared to control.

A. subrufescens can be considered as an important edible mushroom because of its medicinal and culinary properties, which culture conditions still have to be precisely described. In our previous study [17], we found that a horse manure and wheat-straw-based compost produced under physicochemical conditions commonly used for the button mushroom A. bisporus was suitable for this tropical Agaricus species. Here, our investigations focused on storage conditions of the spawn, which is a very important parameter to favor mushroom cultivation. Our study revealed that storage conditions of the spawn for up to 30 days, at relatively low temperatures for this fungus, did not further affect substrate colonization by this species. Storage under certain conditions even improved its potential of colonization and this potential seems to be strain dependent. These results are encouraging, since they demonstrate that these storage conditions at 10 to 15 °C cannot be ruled out in the cultivation process. However, further studies should determine whether these storage conditions have effects on crop yields or mushroom quality in order to confirm these first findings.