Decomposition of litter is a key biochemical process that regulates the rate and magnitude of CO2 fluxes from biosphere to atmosphere and determines soil nutrient availability. Although several studies have shown that plant litter decomposes faster in their native compared to a foreign environment, i.e. a home field advantage (HFA) for litter degradation, to date HFA has only been considered in terms of respiration or litter mass loss.

The ecological success will be determined by the decomposer microorganism’s ability to transform used OM into population growth, and therefore we hypothesized that HFA for microbial growth would be more pronounced than for decomposition, by driving the feedback between community alignment and environmental resource conditions. We also expected that HFA for microbial processes would increase with lower quality litter, that the fungal role in litter decomposition would be more dominant than that of bacteria, and that HFA effects would increase with more pronounced environmental differences between sites.

We designed a 2-month microcosm reciprocal transplant experiment with litter from two sites with contrasting climates (Atlantic and Sub-Mediterranean climates) and including 3 tree species (Quercus robur, Pinus sylvestris and Fagus sylvatica).

We found a stronger HFA for microbial growth than for decomposition, that the nutrient content and C-quality of litter influenced the microbial HFAs, and that interactions between bacterial and fungal communities during litter decomposition modulated the HFA for litter degradation.

Low litter nutrient content, strong nutrient limitations and low C-qualities all favoured fungal over bacterial decomposers, and our results suggest a dominant functional role of the fungal community and gave rise to HFA for fungal growth but that translated to only marginal implications for overall decomposition of litter.

2.1 Experimental design and sample collection

We designed a 2-month microcosm reciprocal transplant study with litter and soil from 2 sites with contrasting climatic conditions (M-site was in a Sub-Mediterranean climate and A-site was in an Atlantic climate) and including 3 different tree species (Quercus robur L., Pinus sylvestris L. and Fagus sylvatica L.) at each of the two sites. The laboratory microcosm study was carried out at controlled temperature and moisture to standardise any direct influence by climate on the litter decomposition processes. We measured respiration, bacterial growth and fungal growth throughout the experiment to determine the responses of the microbial community to different treatments and estimated the HFA effect for each process. To elucidate putative factors that could drive HFA effect, we determined the litter quality of all studied litters, and the physiochemistry of the studied soils. In addition, to link HFA to the perceived nutrient quality of the studied litters, enzymatic activities related with C, N and P cycles were also measured. Finally, PLFA analysis were performed to screen for differences between microbial community structures resulting from different combinations of soils with litter.

The O layer (approx. 5-10 cm deep) and leaves for the microcosm experiment were collected within the forest at Artikutza (43°13'N 1°48'W, Goizueta) and Montoria (42°38'N, 2°44'W, Urizaharra), located in the northern and southern Basque County, respectively, in north Spain. Both study sites have acid soils, with pH-H₂O 3.9 in Artikutza (“A-site”) and 4.4 in Montoria (“M-site”). Soil sampled from the A site has a sandy clay loam soil texture with 13% of organic matter and that sampled from the M site also had a sandy loam soil texture, with 9.4% of organic matter. Both soils were classified to be Umbisols (WRB, 2015). These two study sites differ in climatic conditions. While the northern site (Artikutza, orA-site), is located in an Atlantic climate with annual mean precipitations about 2500 mm, 16^◦C annual mean temperature and annual 31 frost days, the southern M-site is in a sub-Mediterranean climate in which mean annual precipitation is approximately 650 mm, i.e. 4 times lower than in A site, and the mean annual temperature is 10.5^◦C, i.e. about 5.5 °C cooler than the A-site, and the frost days are approximately double than in A-site (58 days). We selected three different tree species, similar in age, density and structure, which were present in the two areas: common oak (Quercus robur L.), scots pine (Pinus sylvestris L.) and European beech (Fagus sylvatica L.).

Naturally senesced leaves were collected with litter traps (composited from 4 litter traps per forest type) between late October 2017 and early January 2018 from oak, pine and beech stands within each study site. They were oven-dried (40˚C) until dry (>48 h) and ground. O layer samples (top 5-10 cm) were collected in early January 2018 during frost free conditions in both sites. Composite samples were collected in each stand around each litter trap. O layer samples were sieved (4 mm) and kept at 4˚C for a week until microcosm experiment started.

Different microcosms were prepared for a reciprocal transplant experiment of litter from different study sites for each of the three tree species (Figure 1). Litter from the A-site was combined with its native soil (‘home’), but also with soil from the other site (‘away’), and vice versa, resulting in 4 different treatments per tree species, and thus a total of 12 treatments. The treatments were replicated in three (n=3), resulting in 36 microcosms that were monitored over time. As we wanted to investigate the soil environment’s role in affecting the decomposition of litter by microbes, capturing this necessitates a sufficiently large proportion of soil to litter. Without it, the environments in the microcosm environments would be dominated by the litter only. Therefore, in each microcosm, 60 g of soil (O-layer) were adjusted to optimal moisture, at 50% of water holding capacity, after which 12 g of dried and ground litter were added. This combination of O layer with litter was used to simulate incoming litter into the top of a forest floor. Unsterilized litter will bring with it a small amount of microbes that could contribute to seed the soil environment below. However, the dried litter samples should have carried only a negligible microbial community compared to that already active in the O-horizon sample used, which we argue natural conditions of litter fall on a forest floor well. The microcosms were incubated at 16˚C, thus falling within a range of summer soil temperatures normal for both sites, and constant moisture (adjusted as needed) for 2 months, and respiration, bacterial growth and fungal growth were monitored (10 sampling times: 1, 3, 5, 8, 12, 21, 28, 35, 42, and 55 days). In addition, biomarkers for the microbial community size and composition (PLFA) were determined 3 times (1 day, 28 days and 55 days) over the experiment. Basic soil chemical properties, pH and electrical conductivity, and enzymatic activities were measured in the first and in the last sampling times to capture the differences in microbial nutrient availability.

2.2 Microcosm microbial parameters

Bacterial growth was measured by the incorporation of ³H-Leucine (Leu) into bacteria (Bååth et al., 2001). At each time point, 1 g of soil was mixed with 20 ml demineralized water by vortexing for 3 min. The supernatant with a bacterial suspension was sampled after low-speed centrifugation (1000 g for 10 min) and the incorporation of Leu was measured in 1.5 ml aliquots of the bacterial suspension. A combination of nonradioactive and tritiated Leu (³H-Leu, 37 MBq ml 1, 5.74 TBq mmol 1, Perkin Elmer, USA) were added to yield a final concentration of 275 nM in the bacterial suspension. The extracted bacteria were incubated for 1 h at 16˚C. The samples were washed (Bååth et al., 2001) and the radioactivity of the incorporated Leu was measured on a liquid scintillator. Bacterial growth was expressed as the amount of Leu that was incorporated in the extracted bacteria per g dry soil and per h (Rousk and Bååth, 2011).

Fungal growth was assessed using the acetate-into-ergosterol-incorporation (Bååth, 2001; Rousk et al., 2009). A radioactively labeled acetate solution (20 µl 1-[14C] acetic acid (sodium salt), 37 MBq/ml, 2.10 GBq/mmol, Perkin Elmer), was combined with 30 µl unlabelled 16 mM acetate and 1.95 ml water was added to 0.5 g of soil, yielding a final concentration in the soil mixture of 0.2 mM, and incubated for approximately 4 hours at 16˚C. Growth was terminated by adding 500 µl 10% formalin, and ergosterol was extracted and separated using high-performance liquid chromatography and a fraction collector (Rousk and Bååth 2007). The incorporated ¹⁴C-acetate into ergosterol was measured using liquid scintillation and used as a proxy for fungal growth (Rousk & Bååth, 2011).

Soil respiration was measured using a gas chromatograph equipped with a methanizer and a flame ionization detector. For each sampling time, 0.5 g of soil was put into a 20 ml glass vial, which were purged with air, sealed and incubated at 16˚C (microcosm incubation temperature) for 2-6 hours (Silva-Sanchez et al., 2019). The concentration in the air used for purging was subtracted from the headspace concentrations, which was then divided by the incubation time to estimate the respiration rate. The respiration rates were expressed as µg CO₂ per g dry soil and per h.

2.3. PLFA analysis

The microbial phospholipids were extracted in a one-phase mixture of chloroform, methanol, and citrate buffer (Frostegård et al., 1991) from frozen subsamples (see above). Then, phospholipids were fractionated and methanolyzed into free fatty acid methyl esters (FAMEs) with FAME 19:0 added as an internal standard (Frostegård et al., 1993; Cruz Paredes et al., 2017). The resulting phospholipid fatty acids (PLFAs) were separated and quantified on a GC with a flame ionization detector (Frostegård et al., 1993). A total of 18 PLFAs were selected and used to indicate the microbial community composition, and to estimate total, bacterial and fungal biomass concentrations.  Fungal biomass was quantified based on 18:2w6,9 content (PLFAfun), and bacterial biomass (PLFAbact) was quantified as the sum of i15:0, a15:0, i16:0, 16:1w7c, i17:0 cy17:0, 18:1w7, 18Me18:0 and cy19:0 (Frostegård and Bååth, 1996; Ruess and Chamberlain, 2010), and the total biomass (PLFAtot) was calculated as the sum of the 18 PLFAs.