CO2 Flux and C Balance due to the Replacement of Bare Soil with Agro-Ecological Service Crops in Mediterranean Environment

Intensive agriculture practices often results in decomposition of organic matter, thus causing soil CO2 emissions. Agro-ecological service crop could be profitably cultivated to improve soil characteristics and reduce CO2 emissions under Mediterranean environment. Two-year field trials were conducted in central Italy. The treatments were three agro-ecological service crops (hairy vetch, oat, and oilseed rape) and a no-service cover. Plant development, soil characteristics, and CO2 emissions were measured. Oat and oilseed rape showed a rapid growth, while hairy vetch started to grow rapidly only after the cold period. Soil CO2 emissions trend was similar among the agro-ecological service crops and tended to decrease during the cold period, then gradually increased until April when warm temperatures were observed. The high soil CO2 emissions and respiration index observed in hairy vetch probably stimulated mineral nutrients, especially nitrogen, to become more available in the soil compared to oat and oilseed rape throughout the decomposition of soil organic matter. These results confirmed that the cultivation of agro-ecological service crops, especially hairy vetch, could represent a suitable strategy for enhancing carbon sequestration and lead to a mitigation of CO2 emissions during the fallow period and could thus contribute to the climate change mitigation.


Introduction
Agriculture is considered one of the human activities that results in high production of CO 2 [1]. In particular, conventional agricultural practices contribute about 25% of the total anthropogenic CO 2 emissions and cause about 50% of soil organic carbon (SOC) loss [2]. Conventional land management is based on intensive agriculture practices, such as deep tillage, high inorganic nitrogen fertilizer addition, and irrigation, and often results in the rapid decomposition of organic matter through microbial activities, thus causing high soil CO 2 emissions and losses of SOC [2]. However, soil microbes play a number of vital functions in the soil and the microbial breakdown of organic matter of the soil not only affects CO 2 emission, but also contributes to the release of nutrients into the soil, such as nitrogen, phosphorous, and others, that are essential for plant establishment and development [3]. Thus, CO 2 emissions due to soil respiration are also an important indicator to assess nutrient cycling.
For these reasons, sustainable soil management is needed in order to enhance SOC in cropland, based on the chemical, physical, and biological properties of the soils [3][4][5]. The reduction of CO 2 emissions by soil carbon sequestration is of primary importance as agricultural activities could remove atmospheric carbon by sequestration and thus mitigate climate change by maintaining and enhancing the amount of carbon stored in the soil and plant material. The challenge of sustainable agriculture is to develop crop rotation schemes and increase soil carbon (C) stocks by the adoption of management strategies that could include adding manure or other organic amendments [6], leaving crop residues on the field [7], reducing soil disturbance [8], and including cover crops [5].
Cover crops are commonly identified as agro-ecological service crops because they may bring multiple benefits to agro-ecosystem [9], including suppression of weeds and pests [10], improvement of soil and water quality, and stimulation of nutrient cycles [11]. Among the several benefit of agro-ecological service crops, the enhancement of SOC and nitrogen sequestration through the production of fresh organic matter has the potential to mitigate climate changes [12][13][14][15]. In fact, the aboveground and belowground biomass from agro-ecological service crops is considered as net primary production of C and represent a net flux of CO 2 -C from the atmosphere into the biomass tissues [16]. Therefore, replacing bare soil with agro-ecological service crops has been widely accepted as a sustainable agricultural practice, even if the species of agro-ecological service crop to cultivate in crop rotation should be based on the biological, environmental, social, cultural, and economic factors where farmers operate [17]. Under Mediterranean climate conditions, agro-ecological service crops are commonly cultivated from early autumn to spring, when they achieve the highest level of biomass production and are then suppressed mechanically or chemically before the cultivation of the main summer vegetable or grain crops [18,19]. The most extensively used agro-ecological service crops are legumes and grasses that are able to accumulate nitrogen and produce a large amount of biomass. However, there is an increasing interest in crucifers which may act as biofumigants and suppress soil pests, especially root pathogens and parasitic nematodes [11]. Grass used as agro-ecological service crop include the annual cereal that are very useful for scavenging nutrients, mainly nitrogen, due to its developed root systems and produce high amount of biomass that can help to add organic matter to the soil [3]. Although several studies evaluated the impact of agro-ecological service crops on pest management and nutrient cycling, little is known about their effect on soil CO 2 emissions and C balance during their cultivation period. This current study hypothesizes that in Mediterranean environment the replacement of bare soil with the adoption of adequate agro-ecological service crop can be profitably used for sustainable agriculture to improve carbon accumulation and soil characteristics also during their growing season. The evaluation of this assumption is of particular interest to improve the knowledge on biomass production, soil characteristics, and soil carbon emissions in the Mediterranean environment. Therefore, the main objectives of this study were to evaluate the effects of agro-ecological service crops at their termination on: (i) Soil CO 2 emissions trend and short-term carbon balance, (ii) biomass production and characteristics, and (iii) soil properties.

Description of the Study Area and Experimental Design
Field trials were conducted in the 2009/2010 and 2010/2011 growing seasons of the agro-ecological service crops at the experimental farm of the University of Tuscia located in Central Italy (45 • 25 N, 12 • 04 E and 310 m above sea level). The soil was classified as Typic Xerofluvent of volcanic origin with the following average characteristics in the 0-30 cm layer: 10.9% clay, 14.3% silt, 74.8% sand; pH 6.5 (water, 1:2.5 w:v); organic matter 1.2%; total N 0.89 g kg −1 ; available P 35 mg kg −1 ; and exchangeable K 569 mg kg −1 .
The climate of the study area is moderate thermo-Mediterranean with average temperature of 14.3 • C. The lowest mean temperature (on average, 0 • C) was observed in winter period, especially in January and February, while the highest mean temperature (on average, 35 • C) was observed in summer period, especially in July and August. Annual rainfall was 752 mm, on average, based on the 30-year period and was mainly concentrated from September to May.
The field experiments were carried out in two adjacent fields of the same site previously cropped with durum wheat (Triticum durum Desf.). Each experimental field was 4800 m 2 (60 m × 80 m), which made it possible to perform all farming operations with agricultural machinery. The experimental treatments consisted in: Three agro-ecological service crops (hairy vetch (Vicia villosa Roth. var. Capello), oat (Avena sativa L. var. Donata), and oilseed rape (Brassica napus L. var. Licapo)) and a bare soil with no cover. The experimental treatments were arranged in a randomized block design with three blocks for a total of 12 plots.

Field Set-Up and Crop Management
In September, at the beginning of each growing season of the agro-ecological service crops, the soil was ploughed at a depth of 30 cm and fertilized with triple super phosphate at the rate of 100 kg P 2 O 5 ha −1 . The day after, the soil was harrowed to a depth of 10 cm in order to prepare the seedbed. About one week from the seedbed preparation on 24 September 2009, and 13 September 2010, the seeds of hairy vetch, oat and oilseed rape were hand broadcast at the rates of 60, 100, and 15 kg ha −1 , respectively, then the seeds were buried by harrowing to a depth of 2 cm. The bare soil plots were treated similarly to the other plots and it was kept bare throughout the agro-ecological service crop growing season by hand weeding when the weeds start to emerge. The growing season of all agro-ecological service crops was mechanically terminated in 20 May 2010, and 4 May 2011, when hairy vetch was at the flowering stage and oat and oilseed rape were in the milk stage.

Sampling and Measurements
The photosynthetic photon flux density (PPFD, µmol m −2 s −1 ) transmitted by the agro-ecological services crops was measured at ground level. The measurements were performed every seven days from the emergence to mechanical suppression of the agro-ecological services crops by using a linear ceptometer (SS1-UM-2.0, DELTA-T devices LDT, Cambridge, England). The ceptometer was placed horizontally at ground level five times in the central part of each plot. All measurements were carried out under full sunlight on clear days between 12:00 p.m. and 2:00 p.m. The fraction of PPFD intercepted (FiPPFD) was calculated using the following formula: where the I o is the average of five measured PPFD on the surface of the ground and I t is the radiant flux density on the top of the agro-ecological service crop canopy. The FiPPFD equal to 1 or 0 indicates all or no PPFD intercepted, respectively. Just before the suppression of the agro-ecological service crops, the biomass samples were collected. The aboveground biomass was hand-clipped at the soil surface and collected by placing a quadrat 50 cm × 50 cm (0.25 m 2 ) randomly four times over the central part. The entire aboveground biomass samples were separated into leaf and stem fractions. The stem fraction included stems and inflorescences. Each fraction was dried, after which they were weighed to determine leaf/stem ratios. The belowground biomass was determined collecting roots from each plot at 0-30 cm depth with hand-operated soil core sampler (8 cm diameter and 15 cm height). The roots were removed from soil sample by pouring slurry of distillated water through a 50 µm screen and the retained roots were rinsed under water. Subsamples of boot shoot and root biomass were dried to calculate the shoot/root ratio, and were then homogenized with a mill for carbon and nitrogen content determination using an elementary analyzer (Thermo Soil NC -Flash EA1112, Lakewood, NJ, USA). The total N or C accumulation were calculated by multiplying the biomass of agro-ecological service crop with the corresponding N or C concentration value, respectively.
At the suppression of the agro-ecological service crops, the sampling of soil was performed. Five soil cores (0-30 cm depth) were taken in the middle part of each plot and then pooled together for the physic-chemical characterization analysis. The collected soil samples were air dried and sieved (<2 mm). The total organic carbon (TOC) and the total nitrogen (TN) contents of the soil samples were determined using an elementary analyzer (Thermo Soil NC -Flash EA1112, Lakewood, NJ, USA). Moreover, soil NO 3 -N [20] and soil NH 4 -N [21] concentrations were determined after soil sampling in an aliquot of sieved fresh soil in order to assess the soil mineral nitrogen.
In both growing cycles, at the emergence of agro-ecological service crops, soil CO 2 emissions were measured in each plot on a daily basis until their mechanical suppression using a portable dynamic closed-chamber infrared gas analyzer system [22]. All soil CO 2 emission measurements were carried out at the same time of day in order to reduce diurnal variation and performed by placing the closed-chamber of a sensitive infrared gas analyzer instrument (EGM-4, PP Systems, Stotfold, UK) in a permanent PVC collar in the central plot of each agro-ecological service crop. Further information regarding the operation methods adopted for the CO 2 measurements are reported by Mancinelli et al. [5]. The measured CO 2 emission suggested the net carbon mineralization value. In both growing cycles, soil CO 2 emission was estimated as the amount of C-CO 2 accumulated [14,23] throughout the study period. The calculations were made by means of the linear interpolation of the two neighboring measured emissions and the numerical integration over time (trapezoid rule) as reported in the following equation: where i = date of first measurement of CO 2 rate taken, n = date of last measurement of CO 2 rate taken; x = CO 2 rate (kg ha −1 day −1 ), and N = number of days between the two consecutive CO 2 rate measurements. The soil respiration index was estimated as soil C stock/soil CO 2 emission rate (t ha −1 /t ha −1 ). The soil C content was calculated by applying the formula suggested by Batjes [24]. Carbon input/output ratio was calculated as the ratio between the C added to the system by cover crop biomass at their termination (C input) and the total C content in the CO 2 emissions measured throughout the growing season of the service crops [5]. Soil temperature close to the chamber was measured simultaneously to the measurement of CO 2 emissions at a depth of 5 cm by using the STP-1 Soil Temperature Probe connected to an EGM-4 instrument. Similarly, soil volume water content was measured near the chamber at the same time of CO 2 measurement at a depth of 30 cm using the TDR 300 Soil Moisture Meter (Spectrum Technologies, Inc., Plainfield, IL, USA).

Statistical Analysis
The data on agro-ecological service crops and soil were analyzed with analysis of variance (ANOVA) using JMP statistical software package, version 4.0 [25]. The data analysis was carried out applying a randomized block design with three blocks and two years of experimentation. Data were analyzed using the ANOVA model with the treatment as fixed factor, the three blocks were included as a random factor and the year (growing cycle) was considered as random effect to account for the repeated measure across time [26]. Percentage data were arcsin transformed before analysis in order to homogenize the variance [27]. The data reported in the tables were back transformed for meaningful data interpretation. Fisher's protected least significant difference (LSD) at 0.05 probability level (p < 0.05) was used for comparing the treatments. Data on the FiPPFD, CO 2 emissions, soil temperature and soil volume water content are presented as the means value ± standard error (SE).

Weather Conditions
The growing period of agro-ecological service crops lasted for 239 days in 2009/2010 and 233 days in 2010/2011 growing cycles. Generally, the average air temperature was slightly higher in the first year than in the second (on average 10.5 vs. 10.2 • C, respectively), particularly from December to February, when the temperature dropped below 0 • C several times. Furthermore, the first year was drier compared to the second (864 vs. 694 mm, precipitation in 2009/2010 and 2010/2011, respectively), while the evapotranspiration was 420 mm in the first year and 441 mm in second.

The Fraction of PPFD Intercepted (FiPPFD) during the Agro-Ecological Service Crop Growing Cycle
The fraction of photosynthetic photon flux density intercepted (FiPPFD) from the emergence to mechanical suppression of the agro-ecological service crop in 2009/2010 and 2010/2011 is reported in Figure 1. In both growing cycles, the values of FiPPDF in the agro-ecological service crops increased gradually in time, however, this trend was different among the treatments. Indeed, the first stages of the agro-ecological service crops immediately after their germination until to January, when it was observed the lowest soil temperature, the values of FiPPDF were generally similar and higher in oat and oilseed rape, while hairy vetch grew more slowly showing lower values of FiPPDF compared to oat and oilseed rape ( Figure 1). Subsequently, the FiPPDF values in oilseed rape tended to decrease notably, while in hairy vetch they increased considerably and reached similar values observed in oat ( Figure 1).

Weather Conditions
The growing period of agro-ecological service crops lasted for 239 days in 2009/2010 and 233 days in 2010/2011 growing cycles. Generally, the average air temperature was slightly higher in the first year than in the second (on average 10.5 vs. 10.2 °C, respectively), particularly from December to February, when the temperature dropped below 0 °C several times. Furthermore, the first year was drier compared to the second (864 vs. 694 mm, precipitation in 2009/2010 and 2010/2011, respectively), while the evapotranspiration was 420 mm in the first year and 441 mm in second.

The fraction of PPFD Intercepted (FiPPFD) during the Agro-Ecological Service Crop Growing Cycle
The fraction of photosynthetic photon flux density intercepted (FiPPFD) from the emergence to mechanical suppression of the agro-ecological service crop in 2009/2010 and 2010/2011 is reported in Figure 1. In both growing cycles, the values of FiPPDF in the agro-ecological service crops increased gradually in time, however, this trend was different among the treatments. Indeed, the first stages of the agro-ecological service crops immediately after their germination until to January, when it was observed the lowest soil temperature, the values of FiPPDF were generally similar and higher in oat and oilseed rape, while hairy vetch grew more slowly showing lower values of FiPPDF compared to oat and oilseed rape ( Figure 1). Subsequently, the FiPPDF values in oilseed rape tended to decrease notably, while in hairy vetch they increased considerably and reached similar values observed in oat ( Figure 1).

Soil CO 2 Emissions during the Agro-Ecological Service Crop Growing Cycle
The trends of soil CO 2 emissions dynamics, throughout growing period of the agro-ecological service crops are shown in Figure 2. Generally, during the growing period the four adopted treatments affected soil CO 2 emissions with a similar trend in both years. In fact, the CO 2 flux tended to increase immediately after agro-ecological service crop emergence until December, when it was observed low soil temperature (Figure 2). After the cold days in each year, the soil CO 2 flux tended to increase gradually in all treatments reaching a high value in April, just before the agro-ecological service crop suppression. In both years, few differences among the treatments until February were observed, while in the following months the trend of soil CO 2 fluxes among the agro-ecological service crop treatments were largely different. In particular, the main differences were observed between hairy vetch and the other treatments in the 2009/2010 growing cycle, and between hairy vetch and oat compared to oilseed rape and bare soil in the 2010/2011 growing cycle.

Soil CO2 Emissions during the Agro-Ecological Service Crop Growing Cycle
The trends of soil CO2 emissions dynamics, throughout growing period of the agro-ecological service crops are shown in Figure 2. Generally, during the growing period the four adopted treatments affected soil CO2 emissions with a similar trend in both years. In fact, the CO2 flux tended to increase immediately after agro-ecological service crop emergence until December, when it was observed low soil temperature (Figure 2). After the cold days in each year, the soil CO2 flux tended to increase gradually in all treatments reaching a high value in April, just before the agro-ecological service crop suppression. In both years, few differences among the treatments until February were observed, while in the following months the trend of soil CO2 fluxes among the agro-ecological service crop treatments were largely different. In particular, the main differences were observed between hairy vetch and the other treatments in the 2009/2010 growing cycle, and between hairy vetch and oat compared to oilseed rape and bare soil in the 2010/2011 growing cycle.

Soil Temperature during the Agro-Ecological Service Crop Growing Cycle
As expected, the soil temperature and climatic data observed during the cover crop growing period showed similar trends. From March to April in 2010 and from February to April in 2011, significant differences between the agro-ecological service crop treatments were observed. In

Soil Volumetric Water during the Agro-Ecological Service Crop Growing Cycle
In both years, the soil volume water content was generally influenced by weather conditions and mainly rainfall ( Figure 4). Generally, from December onward, the agro-ecological service crop treatments showed higher water volume content than the bare soil. In 2000/2010, oilseed rape treatment showed high values of soil volumetric water content until March, in correspondence of high rainfall period and when all treatments showed similar value of soil volume water content. In the 2010/2011 growing period, a high value of soil volume water content in the hairy vetch treatment was observed, even if only in April, a high difference among the treatments was observed.

Soil Volumetric Water during the Agro-Ecological Service Crop Growing Cycle
In both years, the soil volume water content was generally influenced by weather conditions and mainly rainfall (Figure 4). Generally, from December onward, the agro-ecological service crop treatments showed higher water volume content than the bare soil. In 2000/2010, oilseed rape treatment showed high values of soil volumetric water content until March, in correspondence of high rainfall period and when all treatments showed similar value of soil volume water content. In the 2010/2011 growing period, a high value of soil volume water content in the hairy vetch treatment was observed, even if only in April, a high difference among the treatments was observed.

Agro-Ecological Service Crop Biomass and Soil Characteristics
The total biomass, leaf/stem ratio (L/S), shoot/root ratio (S/R), N, and C accumulated in the total biomass of the agro-ecological service crop measured at their suppression are reported in Table 1. , and the amount of total biomass measured in oilseed rape was the lowest (on average of 518.8 g m −2 DM). Similarly, the leaf/stem ratio (L/S) was higher in hairy vetch and oat than oilseed rape (on average of 1.8 vs. 0.9, respectively), while no differences were observed between the growing cycles. Conversely, oilseed rape showed the highest value of shoot/root ratio (3.4). The nitrogen accumulated in the total biomass of the agro-ecological service crops at the time of their suppression was higher in hairy vetch than in oilseed rape and oat (on average of 33.1 vs. 7.1 g m −2 , respectively, Table 1). At the same time, the C accumulated in the aboveground biomass was highest in hairy vetch, intermediate in oat and low in oilseed rape (476.6, 336.1, and 218.9 g m −2 , respectively). The C/N ratio of the agro-ecological service crops was highest in oat, followed by oilseed rape and hairy vetch (data not shown).

Agro-Ecological Service Crop Biomass and Soil Characteristics
The total biomass, leaf/stem ratio (L/S), shoot/root ratio (S/R), N, and C accumulated in the total biomass of the agro-ecological service crop measured at their suppression are reported in Table 1. , and the amount of total biomass measured in oilseed rape was the lowest (on average of 518.8 g m −2 DM). Similarly, the leaf/stem ratio (L/S) was higher in hairy vetch and oat than oilseed rape (on average of 1.8 vs. 0.9, respectively), while no differences were observed between the growing cycles. Conversely, oilseed rape showed the highest value of shoot/root ratio (3.4). The nitrogen accumulated in the total biomass of the agro-ecological service crops at the time of their suppression was higher in hairy vetch than in oilseed rape and oat (on average of 33.1 vs. 7.1 g m −2 , respectively, Table 1). At the same time, the C accumulated in the aboveground biomass was highest in hairy vetch, intermediate in oat and low in oilseed rape (476.6, 336.1, and 218.9 g m −2 , respectively). The C/N ratio of the agro-ecological service crops was highest in oat, followed by oilseed rape and hairy vetch (data not shown). Table 1. The main effects of growing cycle and agro-ecological service crop on the total biomass, shoot/root ratio (S/R), leaf/stem ratio (L/S), N and C accumulated in the biomass. Values belonging to the same variable followed by the same letter and are not statistically different according to LSD (0.05).

Treatments
Total The soil total organic C (TOC) and soil mineral N (NO 3 + NH 4 ), soil CO 2 emission, and soil respiration index of agro-ecological service crop measured at their suppression are reported in Table 2. As expected, the TOC was similar among growing cycles and the agro-ecological service crops (on average 1.24%). The mineral nitrogen in the soil was higher in 2009/2010 than in 2010/2011, while among the agro-ecological service crops, it was higher in hairy vetch than oilseed rape, bare soil, and oat (35.6, 22.0, 19.6 and 17.4 mg kg −1 , respectively). Table 2. The main effects of growing cycles and agro-ecological service crops on the soil total organic C (TOC), soil total N (TN), total mineral N (NO 3 + NH 4 ), C by CO 2 emission from agro-ecological service crops, soil respiration index and carbon input/output ratio. Values belonging to the same variable followed by the same letter and are not statistically different according to LSD (0.05). The soil C-CO 2 emissions during the agro-ecological service crops were significantly higher in the first cycle than in the second (averagely 372 vs. 320 g C m −2 in 2009/2010 and 2010/2011, respectively), while among the agro-ecological service crops, the values recorded were highest in hairy vetch, intermediate in oat, and low in oilseed rape and bare soil (408, 355, and 312 g C m −2 , respectively, Table 2). In 2009/2010, soil respiration index was higher than in 2010/2011. Hairy vetch showed the highest value of soil respiration index compared to the other agro-ecological service crops (0.094 vs. 0.077, respectively, Table 2). The input/output ratio of C was higher in hairy vetch (1.2), intermediate in oat (0.9) and oilseed rape (0.7), and low in bare soil (0.0, Table 2).

Discussion
In this study, the effects of the cultivation of different agro-ecological service crops under Mediterranean climate conditions on short-term carbon balance, soil characteristics, and biomass production were evaluated throughout the growing cycle until their mechanical suppression before the preparation of the transplanting bed of the following main summer vegetable crop. To date, few studies have focused on the effect of agro-ecological service crops on CO 2 fluxes in agricultural soils [28], especially during their growing cycle. In fact, as observed by Papp et al. [4], the majority of studies investigate the effects of agro-ecological service crops after their suppression and throughout the cultivation of the following cash crop when their residues where managed in different way, such as green manuring or mulching the soil. As a consequence, this approach evaluates the combined effects of agro-ecological service crop residues in terms of rhizo-depositions release due to species cultivation and of residue mineralization. Conversely, the approach adopted in this study is aimed to provide knowledge on changes due to the cultivation of a specific agro-ecological service crop during its growth period.
The plant species adopted in this study as agro-ecological service crops were representative of cereal (oat), legume (hairy vetch), and cruciferous (oilseed rape) species, and were chosen because they are common and suitable for the cultivation in the temperate climate of the Mediterranean environment. Although the seedling emergence and establishment were uniform and regular among the three agro-ecological service crops, without any evident gaps, their canopy development followed different trends due to their inherent growth patterns and responses to the weather conditions. Indeed, based on the data regarding the FiPPDF, oat and oilseed rape showed a rapid establishment and growth compared to hairy vetch, resulting in high soil cover and, thus, in biomass production, especially during the first steps after their sowing. Conversely, soil coverage in hairy vetch was slow and started to grow rapidly only after the cold period reaching, before mechanical suppression, a complete soil coverage similar to oat (FiPPDF = 1). The results of this study confirmed that hairy vetch and oat could be well-suited plant species for their cultivation as agro-ecological service crops, in the temperate climate of the Mediterranean environment [3,5,18]. In fact, they were frost resistant and produced the highest amount of total biomass (shoots + roots), thus their cultivation could be encouraged in the Mediterranean environment, even if oat, as grass species, should be preferable for reducing nitrogen leaching risk [29] or suppress winter weeds [18], while hairy vetch, as legume species, should be suggested for its ability to fix high amount of atmospheric nitrogen and enhancing its availability after the suppression in a sustainable way throughout mineralization process [19]. Conversely, oilseed rape that grew fast after seeding was partly damaged (data not shown) by the winter frost, as shown by the FiPPDF data, and provided low amount of total biomass at the suppression. Similarly, Mancinelli et al. [5], in a previous study, observed high sensibility of cruciferous species to cold season in a similar environment. Furthermore, the data on leaf/stem ratio showed a higher number of leaves in hairy vetch and oat compared to oilseed rape, meaning a different canopy structure, mainly composed by materials more suitable to convert, after the mechanical suppression, in organic mulches, especially in the case of the adoption of conservation agriculture practices, as confirmed by the results reported by Mancinelli et al. [5] and Radicetti et al. [19] in similar environmental conditions.
In both growing cycles, the flux of CO 2 emissions follows similar trend among the agro-ecological service crops and tended to decrease during the cold period, then gradually increased until April when warm temperatures were observed. Similarly, Lal et al. [16] showed a gradual decrease of soil CO 2 emissions from summer to winter and an increase of soil emissions with the hotter temperature. According to Jacinthe et al. [30], soil CO 2 emissions strongly depend on weather conditions, indeed, the differences among the agro-ecological crop species were low and follow the soil temperature trend. However, the high biomass production in hairy vetch with low shoot/root ratio compared to oat and oilseed rape resulted in high and denser root biomass thus increasing microbial activity near to the soil surface [4]. As a consequence, rhizosphere and root respiration probably contributed significantly to the high soil CO 2 emission values observed in hairy vetch, even if considering the high amount of carbon accumulated on hairy vetch biomass, the input/output ratio resulted more favorable terms of carbon sink. However, the high soil CO 2 emissions and respiration index observed in hairy vetch was probably due to an intensive microbial activity that began to decompose the organic material from soil and plant residues. As a consequence, mineral nutrients, especially nitrogen, became more available in the soil compared to oat and oilseed rape. Although the soil CO 2 emissions and respiration index in oat and oilseed rape were significantly low compared hairy vetch, the carbon accumulated in their biomass determined a low input/output ratio and thus less efficient in term of carbon sink. These results confirmed that the cultivation of agro-ecological service crops, especially hairy vetch, could be adopted as a suitable strategy for enhancing short-term carbon dynamic and, thus, could contribute to the climate change mitigation if integrated in an integrated management strategy [31]. Furthermore, maximizing organic matter inputs, as observed in hairy vetch, could represent a way to add carbon to the soil (carbon sink) that could be a key factor for improving crop yields and reducing the need for synthetic fertilizers in agreement with Aguilera et al. [32], even if the addition of easily mineralizable carbon from agro-ecological crop residues could results in high CO 2 emissions throughout the cultivation of the following main summer crop in agreement with Bodner et al. [33]. However, as stated by Mancinelli et al. [5,14] the soil tillage practice adopted for managing the residues of agro-ecological service crops represents an important factor that could affect the C balance in the agro-ecosystems.

Conclusions
In this study, the replacement of bare soil with agro-ecological service crops during the fallow period under Mediterranean climate represents a sustainable practice for improving carbon sequestration and contribute to mitigate CO 2 emissions, even if the responses varied among the cover crops species. In fact, hairy vetch accumulated the largest amount of total biomass and carbon (input), also resulting in a better performance when soil CO 2 emissions (output) were considered. Furthermore, the high amount of nitrogen accumulated in the total biomass of hairy vetch during the growing cycle assured an elevate availability for the following cash crop through mineralization process. Although oat was less efficient in terms of C end N soil accumulation than hairy vetch, the biomass produced and the carbon sequestered encourages the adoption of legumes in Mediterranean environment, even if the adoption of oat should be well evaluated because could sensibly reduce soil nitrogen availability. Conversely, the adoption of oilseed rape seems unable to cover the need to accumulate carbon on its biomass, due to the risk of frost damage especially in the winter when is frequent in the Mediterranean environment. However, further research should be performed for evaluating the adoption of this species for other agro-ecological service.