Litter contribution to soil organic carbon in the processes of agriculture abandon

The mechanisms of litter decomposition, translocation and stabilization into soil layers are fundamental processes in the functioning of the ecosystem, as they regulate the cycle of soil organic matter (SOM) and CO2 emission into the atmosphere. In this study the contribution of litters of different stages of Mediterranean secondary succession on carbon sequestration was investigated, analyzing the role of earthworms in the translocation of SOM into the soil profile. For this purpose the δC difference between meadow C4C soil and C3-C litter was used in a field experiment. Four undisturbed litters of different stages of succession (45, 70, 100 and 120 since agriculture abandon) were collected and placed on the top of isolated C4 soil cores. The litter contribution to C stock was affected by plant species and it increased with the age of the stage of secondary succession. One year after the litter position, the soil organic carbon increased up to 40 % in comparison to soils not treated with litter after 120 years of abandon. The new carbon derived from C3 litter was decomposed and transferred into soil profile thanks to earthworms and the leaching of dissolved organic carbon. After 1 year the carbon increase attributed to earthworm activity was 6 and 13 % in the soils under litter of fields abandoned for 120 and 45 years, respectively.


Introduction
The major input of vegetative C to forest soil is represented by litter; hence changes in litter inputs are likely to have important consequences for soil C dynamics (Sayer et al., 2007).Generally, it has been recorded that an accumulation of litter corresponds to an increase of the carbon storage in the soil; for instance, an accumulation of litter and a consequent increase in the carbon content of the soil has been recorded following the processes of abandonment (Costa and La Mantia, 2005).
Therefore, the mechanisms of litter decomposition, translocation and stabilization into soil layers are fundamental processes in the functioning of the ecosystem as they regulate the cycle of soil organic matter (SOM), CO 2 emission into the atmosphere, carbon sequestration into the soil and nutrients mineralization (Maisto et al., 2011;Parras-Alcántara et al., 2015;Smolander et al., 2008;Fioretto et al., 1998Fioretto et al., , 2005)).
The decomposition of litter is affected by the quality of the residues (Smith et al., 2008), which determines different mineralization rates.Soluble substances and labile compounds of litter are rapidly degraded in the early stages of decomposition by fast-growing microorganisms that may require a high concentration of nitrogen (Swift et al., 1979).Cellulose and lignin, the most abundant components of forest litter, are decomposed slowly (Fioretto et al., 2005).Together with bacteria and fungi, invertebrates are responsible for the main functions of the soil ecosystems, including C cycle (Dix and Webster, 1995;Peng et al., 2015;Schimel et al., 1999).Several authors have attributed earthworms to the creation of favorable conditions for microbial activity through the fragmentation of litter and mixing of organic matter with soil mineral portion (Tiunov et al., 2001;Wurst et al., 2004).
Earthworms also affect both amount and distribution of SOM and cause an increase in the rates of SOM decomposition.Earthworms, in fact, transport large quantities of C from the surface of the soil to the lower horizons, effectively mixing the soil and significantly increasing both the rates of the humification trough litter fragmentation and of the overall decomposition (Lee, 1985;Alban and Berry, 1994;Edwards and Bohlen, 1996;Burtelow et al., 1998;Li et al., 2002, Pulleman et al., 2005).Contrarily, Alban and Berry (1994) and Burtelow et al. (1998) found that an earthworm invasion resulted in a C loss in the upper soil layer.Other fundamental processes for the stabilization of SOM are the leaching of fresh litter compound and of recently formed dissolved organic matter (DOM) from organic layers to mineral soil and the sorption of DOM into mineral surfaces (Sollins et al., 1996;Kaiser and Guggenberger, 2000;Kalbitz and Kaiser, 2008).In case of prolonged leaching, however, the litter can become more resistant to decomposition, as a consequence of the significant loss of soluble organic compounds, and readily degradable (Mangenot et al., 1980).
In this study, the objectives were (i) termining the contribution of litter in soil organic carbon (SOC) sequestration, (ii) analyzing the mechanisms of C translocation from litter to soil and (iii) singling out the amount of C leached and the role of earthworms in this process through isotopic analysis.

Experimental layout, soil and litter sampling
The experiment was carried out in the fields of the Department of Agricultural and Forestry Sciences, University of Palermo, Italy (38 • 06 N, 13 • 20 E, 50 m a.s.l.).According to World Reference Base for Soil Resources (WRB, 2006), the soil used was shallow Aric regosol, rich in limestone (46 % of CaCO 3 ) with a pH value of 7.61, a sandy-clay-silty texture (53.9 % sand, 22.6 % silt and 23.4 % clay) and an organic matter content of 1.40 %.The climate was semiarid Mediterranean with a dry period of 4-5 months (mean temperature: minimum 13.7 • C, maximum 22.1 • C; mean annual rainfall: 531 mm).
The field plot used in the experiment was a Cynodon meadow.The soil under Cynodon was a C 4 soil under isotopic steady state because it had been covered with Cynodon (C 4 photosynthetic pathway plant) for more than 15 years.The δ 13 C of the experimental soil was −14.5±1.8.Cynodon meadow was established with an inter-specific bermuda grass hybrid (C.dactylon x C. transvaalensis), cv Tifway 419.
Agronomic management of the turf grass included monthly application of 50 kg ha −1 of N, 10 kg ha −1 of P and 40 kg ha −1 of K fertilizer from April to October.Irrigation was carried out during the spring-summer season with a sprinkler system in order to reinstate evapotranspiration (determined by a Class A evaporimeter and rainfall).The turf grass was maintained at a height of 30-35 mm using a reel lawn mower two to three times a week.The cuttings were removed without grasscycling or mulching.
Plastic cores (n.30 samples), 20 cm diameter and 40 cm height, were installed in the meadow soil after a careful removal of the grasses in March 2013 (Fig. 1).The cores were buried 30 cm, with a 10 cm surface collar.In 15 of the installed cores, a grid (0.1 mm) was placed on top of the soil core to avoid the earthworms crossing.Undisturbed different litters (four litters of C3 plant) were placed on top of soil.In all, 30 cores were placed (5 litter treatments (4 litters + 1 no litter) × 2 grid (grid and no grid) × 3 replicas).Soil samples were collected in February 2014.The 30 cm soil core was divided in four subsamples (each 7.5 cm soil thickness).The soil was dried and 2 mm sieved, and the organic fragments were removed.
Litters were collected with cores (20 cm diameter) in four different successional stages of a secondary succession in Pantelleria island, Italy (Sicily, 36 • 44 /36 • 50 N, 11 • 57 /12 • 03 E).The selected stages for litter collection were: maquis 45 years since abandon (L45), maquis 70 years since abandon (L70), maquis 100 years since abandon (L100) and forest 120 years since abandon (L120).The abandonment age of the sampled successional stages was determined by evaluating aerial photographs taken in 1955 and 1968 (produced by Istituto Geografico Militare, Florence) and 1987 (Regione Siciliana) (La Mantia et al., 2008).The sampled areas were located in direct proximity to each other and were characterized by comparable abiotic conditions (aspect, slope, soil type, rock outcrop, stone cover, etc.).The land covers where litters were placed are described in Table 1.

Litter analysis
Dry biomass weight and its chemical composition (acid detergent lignin (ADL), neutral detergent fibre (NDF), cellu-  , 1991).The litter respiration rates (mg CO 2 day −1 dry litter) were measured during the incubation experiment using a method of alkali absorption in a closed chamber.Three replicates in each litter treatment with three blank samples were measured.Ten grams of litter were placed inside 1 L glass bottle.A 30 mL 0.1 N NaOH solution was used to trap the CO 2 which was released inside the bottle.The CO 2 -trapped solution titrated with HCl solution using phenolphthalein and methyl-orange as color indicator.During the 7 days of incubation, CO 2 measurements were done after 24, 48, 60 and 96 h and 1 week from the start of incubation.All flasks were ventilated for 30 min with fresh air 24 h before the CO 2 sampling, and an NaOH trap was placed inside the bottle and then sealed with rubber stoppers.The C mineralization rate was expressed in mg CO 2 -C g −1 TOC day −1 and was fitted to the following first-order decay function: where C r is the readily mineralizable C at time 0 (i.e., the intercept value), k is the decay rate constant and t is the time.The amount of total C mineralized was calculated through the linear interpolation of two neighboring measured rates and the numerical integration over time as reported in the following equation: where i is the date of the first measurement of CO 2 -C rate, n is the date of the last measurement of CO 2 -C rate, r is the CO 2 -C rate expressed as mg CO 2 -C kg −1 dry soil and d is the number of days between the two consecutive CO 2 rate measurements.
The mean residence time (MRT) in days was determined as a reciprocal of the rate constant (k) of first-order decay (Eq.1).

Chemical analysis
For each soil sample the C content and δ 13 C abundance were measured.The δ 13 C isotopic signature of litter biomass was also analyzed.For SOC and the δ 13 C analysis, an elemental analyzer isotope ratio mass spectrometer was used.The reference material used for analysis was IA-R001 (Iso-Analytical Limited standard wheat flour, δ 13 CV-PDB = −26.43‰).IA-R001 is traceable to IAEA-CH-6 (cane sugar, δ 13 CV-PDB = −10.43‰).IA-R001, IA-R005 (Iso-Analytical Limited standard beet sugar, δ 13 CV-PDB = −26.03‰) and IA-R006 (Iso-Analytical Limited standard cane sugar, δ 13 CV-PDB = −11.64‰) were used as quality control samples for the analysis.The International Atomic Energy Agency (IAEA), Vienna, distribute IAEA-CH-6 as a standard reference material.
The results of the isotope analysis are expressed as a δ value (‰) relative to the international Pee Dee Belemnite standard as follows: where δ = δ 13 C, R = 13 C / 12 C, s = sample and st = standard.

Data calculation
Natural abundance of δ 13 C was used to determine the proportion of C in SOC derived from the new C input (C 3 -C).These proportions were calculated with the mixing equation (Gearing, 1991) separately for grid and no grid plots: where NCD is the fraction of new C derived, δ 13 C new is the isotope ratio of the soil sample, δ 13 C litter is the isotope ratio of different litters and δ 13 C old is the isotopic ratio of the previous vegetation (Cynodon).Carbon derived from worms was calculated as the difference between NCD in grid and no grid treatments.
The mass of new carbon additions was calculated according to Eq. ( 5).

New carbon (g kg
The standard deviation of the δ 13 C and C values were calculated for each depth and treatment.For the average value, the Duncan test was used at p < 0.05 (SAS Institute, 2001).

Litter characteristics
The plant litter collected during the stages of secondary succession differed in the total weight and C content.The highest weight of litter biomass was in L120 with values of 1113 ± 90 g m −2 , followed by L100, L45 and L70 with values of 1027±77, 915±104 and 946±82 g m −2 , respectively.The highest C content of litter was in L45 and decreased with the increase of the age of abandon (Fig. 2); however, L120 contributed with the highest C litter input (total C litter/core) due to the higher weight in comparison to other litters of the stages of secondary succession.
The MRT was not significantly different among litter ages, except for L120 (Table 2).These findings were confirmed by the readily mineralizable C which was highest in L120 (Table 2).The composition of litter was not statistically different among successional stages regarding cellulose and ADL content (Table 2).The NDF value was, instead, significantly higher in L120 in comparison to litters of other successional stages.

Soil carbon content and distribution
The total amount of SOC differed under the two treatments (grid and no grid) and time of abandon.The SOC was significantly higher in soils where L120 was placed on the top of soil cores, followed by the other litter treatments (Table 3).Comparison between grid and no grid treatments showed the highest C content in soil cores without grid for all litters.
After 1 year of litter permanence, the SOC under L120 increased on average (0-30 cm) by 26 and 40 % in grid and no grid treatment, respectively, in comparison to the no litter treatment.
Such C increase was smaller in grid treatment for the other litters (L45, L70 and L100) with a value of about 12 %.In the no grid treatment, the SOC increased by 22, 23 and 15 % in soil under L100, L70 and L45, respectively, in comparison to the no litter treatment.SOC decreased with the increase of the soil depth, but on average the difference between the first and the deepest soil layer was more pronounced in the no grid treatment (Table 3).

13 C isotopic signature in soil profile
The value of soil δ 13 C changed significantly after litter positioning (Fig. 3).The baseline is represented by soil without litter, where the δ 13 C values were −14.0 ± 0.3 and −16.0 ± 0.4 ‰ in the top and deepest soil layer, respectively.After litter position, δ 13 C was depleted due to C 3 litter input.The most depleted soil was L120 with average (grid and no grid treatment) values of −18.6 and −21.6 ‰ in the top and deepest soil layer, respectively (Fig. 3).For the others litter treatments the value ranged between −15.0 and −20.5 ‰.
The effect of litter input on C stock was highlighted by estimates of C derived from litter (C 3 plant) in the meadow soil (C 4 soil).After 1 year of litter permanence, C originated from litter input was 32.4,34.2, 38.5 and 49.8 % of total SOC in L45, L70, L100 and L120, respectively.
The new derived soil C (C 3 -SOC) was lower for all litter treatments in soil with grid.The portion of C 3 -C in soil with grid was, in fact, 12.4, 23.1, 23.4 and 40.7 % of total C in L45, L70, L100 and L120, respectively (Fig. 4).Considering Solid Earth, 6, 425-432, 2015 www.solid-earth.net/6/425/2015/average average 0-7.5 1.5 1.8 1.8 1.8 1.9 1.6 1.9 1.8 2.0 2.4 1.8 a 1.9 a 7.5-15 1.4 1.5 1.6 1.5 1.7 1.4 1.9 1.9 2.0 2.1 1.5 b 1.9 a 15-22.only the C 3 -C of SOC for each litter treatment, the contribution of earthworms to the incorporation of new C 3 -SOC was in percentage higher in L45, decreased with the age of litter and decreased for each treatment with the increase of the soil depth.The difference of C 3 -C between no grid, grid treatment and depth assesses the earthworm contribution to soil C increase and distribution.

Litter contribution to SOC stock
Previous studies on the island of Pantelleria demonstrated the potential of land cover in the change of C stocks (Novara et al., 2014;Saiano et al., 2013).In fact, land abandonment determines the increase in litter layer and SOC.In natural ecosystems, unlike ecosystems on arable lands, litter is not incorporated into the soil.For this reason it was hypothesized that SOC increase is due to C leaching and/or to earthworm contribution.This hypothesis was confirmed by the present experiment, in which the effect of plant litter contribution to SOC stock was isolated from other soil and environmental parameters.In line with several reports in other ecosystems (Lal, 2005;Yu and Jia, 2014), we recorded that the SOC stock depends on C litter input as well as on litter qual-ity.The incubation experiment of litters showed differences in readily mineralizable C, litter composition (NDF %) and consequently C litter mineralization rate.The litter of L120 had a higher amount of readily mineralizable C in comparison to other litters, and it was easily decomposed and transferred to SOC pool.The faster mineralization rate of L120 could be attributed both to a different composition of plant species (lower content of sclerophyll) (Gianguzzi, 1999) and to a variation in the micro-climatic conditions (Wang et al., 2010;Sheffer et al., 2015) due to a higher accumulation layer on the soil surface.As far as the effect of plant species on the litter mineralization rate is concerned, several studies found a lower litter decomposition rate in Q. ilex in comparison to other Mediterranean species, like Myrtus and Cistus (Berg at al., 1996;Fioretto et al., 2005).Likewise, Maisto et al. (2011) found a slower decomposition of Q. ilex in comparison to Ph. angustifolia, while no significant difference in the decomposition rate was recorded between Q. ilex and Pistacia lentiscus.In these studies the lower decomposition of Q. ilex was attributed to higher lignin content.Our results confirm those of other research with regard to the higher lignin content of Q. ilex, but this was not closely associated to lower decomposition rates.In fact, L120, where the main species was Q. ilex, was the litter with a higher decomposition rate.Therefore, other aspects could explain the differences in the  decomposition rates, like the percentage of a species in each stage of succession, the age of litter and the thickness of litter.

Influence of earthworm on soil carbon
Plant litter is the main source of SOM in soils under secondary succession.The transformation of C litter into SOM is caused by the decomposition of plant biomass and its incorporation into the soil profile.Responsible for this mechanisms are bacteria and fungi, forming up to 90 % of the soil microbial biomass (Dix and Webster, 1995;Schimel et al., 1999) and faunal groups.Our observations highlighted the annual contribution to SOM derived from litter and singled out the activity of decomposition through the difference of isotopic signature between previous SOC-C (C 4 soil) and the new C 3 -C input originated from litter.The 13 C litter recovery in the soil profile was higher in L120 (89 %), followed by L45 (63 %), L100 (60 %) and L70 (52 %).Firstly, the activity of microbial biomass in soil samples where the grid was placed between litter and soil was highlighted.In this case, the new C 3 -C represented the C-pool originated by fungi and bacterial decomposition, transferred into the soil depth mainly through dissolved organic carbon.Such decomposition and incorporation activity contributed to C increase up to 77.6 g core −1 year −1 in L120 treatment (Fig. 5).However, the difference between soil core with and without litter gave information about the contribution of earthworms to litter decomposition and incorporation into the soil.In several studies, the introduction of earthworms in cold temperate forests resulted in a decline of SOC (Bohlen et al., 2004;Alban and Berry, 1994).The results of the present study instead suggest that earthworms have the potential to increase SOC.After 1 year, earthworm activity increased SOC by 13.5, 11.3, 11.1 and 5 %, in L120, L100, L70 and L45, respectively.The effects of earthworm activity on the recovery of soil C released from litter could be attributed to different mechanisms: (i) the mixture of undecayed particulate C into the soil, (ii) the creation of preferential flow paths in the soil increasing nutrient transportation and (iii) protection of C in soil aggregates cre-ated by earthworm feeding (Bohlen et al., 2004;Fahey et al., 2013;Fonte et al., 2007).

Conclusions
This study highlights the effects of vegetation succession on C dynamics in soil after the termination of its agricultural use.Based on δ 13 C signature of C 3 -C of litter and C 4 -C of meadow soil, the annual contribution of vegetation input to C stock was estimated.Moreover, the effect of DOC leaching and earthworm activities on C storage in soil depth have also been evaluated.
Hence in order to understand the ecosystem processes of C sequestration in semiarid environments, a better understanding of the impact of above-ground biomass on soil community is still needed.

Figure 1 .
Figure 1.Sampling area of litter in Pantelleria secondary succession (numbers represent litter in field abandoned for 120, 100, 70 and 40 years) and experimental design in meadow field.Numbers indicate the age since abandon.

Figure 3 .
Figure 3. δ 13 C value at different depth in no grid (a) and grid (b) treatment.The green line represents no litter treatment, while blue, red, grey and black represent litter in fields abandoned for 120, 100, 70 and 40 years, respectively.

Figure 4 .
Figure 4. Contribution (%) of worm activity (black columns) and DOC (green columns) in C 3 -C portion at different soil depth.For each portion, different letters indicate differences for P ≤ 0.05.

Table 3 .
Average of soil organic carbon (%) at different soil depths.For each treatment different letters indicate differences for P ≤ 0.05.