*3.1. Meteorological Conditions*

During the GHG emissions monitoring periods, monthly average temperatures higher than 20 ◦C were recorded in summer 2015 (average of June, July and August 24 ◦C) and summer 2016 (average of June and July 22 ◦C) (Figure 2).

**Figure 2.** Daily rainfall (mm), daily maximum, average and minimum air temperature (◦C) from October 2014 to July 2016.

The monthly average temperature was lower than 10 ◦C in January–February 2015 and January 2016 (8 ◦C). The rainiest month was November 2014 (290 mm), while the driest month was July 2015 (3 mm). Along the whole monitoring period, the rainiest periods were August 2015 (232 mm), October 2015 (254 mm), the period between January and February 2016 (372 mm) and in June 2016 (138 mm).

#### *3.2. Soil Water Content, Temperature and Nitrate Dynamic*

Water filled pore space (WFPS) values did not differ significantly among INT, ORG and ORG+ systems, in either in F1 or F2, with exceptions of (i) 20 May 2015 in F1, where ORG+ and INT had higher WFPS than ORG, and (ii) the period between May and June 2016 in F2, where ORG+ showed significantly higher WFPS values (Figures 3a and 4a).

*Agronomy* **2019**, *9*, 446

**Figure 3.** Data recorded in F1: (**a**) Soil water filled pore space (WFPS); (**b**) soil temperature; (**c**) soil nitrate (N–NO3) concentration for each treatment. Simple arrows indicate fertilization events, and dashed arrows the primary tillage of each crop. On field 1 (F1) the temporal crop sequence was: Fennel, summer lettuce, cabbage, then spring lettuce. Significance was as follows: *n.s.* is not significant; \* is significant at the *p* ≤ 0.05 level; \*\* is significant at *p* ≤ 0.01 level; \*\*\* is significant at *p* ≤ 0.001 level.

**Figure 4.** Data recorded in F2: (**a**) Soil WFPS; (**b**) soil temperature; (**c**) soil nitrate (N–NO3) concentration for each treatment. Simple arrows indicate fertilization events; dashed arrows the primary tillage of each crop. On field 1 (F1) the temporal crop sequence was: Cabbage, spring lettuce, fennel, then summer lettuce. Significance was as follows: *n.s.* is not significant; \* is significant at the *p* ≤ 0.05 level; \*\* is significant at *p* ≤ 0.01 level.

The highest WFPS values were registered in both fields in winter, with maximum values in February 2015 (71% in F1 and 81% in F2) and minimum values in May 2015 (12% in F1 and 23% in F2). Indeed, average WFPS values were high in summer period due to irrigation (36% in F1 and to 46% in F2).

Soil temperature was not different among treatments in both fields. The lowest soil temperature (9 ◦C) was recorded in December 2014 and the highest soil temperature (39 ◦C) in June and August 2015 (Figures 3b and 4b).

Soil nitrate concentration showed values ranging from 0 to 163 mg kg−<sup>1</sup> in F1 and up to 295 mg kg−<sup>1</sup> in F2. Nitrate concentration was higher than 60 mg N–NO3 kg−<sup>1</sup> in 17 sampling dates out of 33 in F1 and in 17 sampling dates out of 28 in F2. In F1, nitrate concentrations were significantly higher in INT than the other treatments in five dates from July 2015 to October 2015 (average 112 mg N–NO3kg−1); and in ORG in three dates in April 2015 and in June 2015, with summer lettuce (average 36.2 mg N–NO3kg−1). In F2, nitrate concentration was higher in cabbage INT on one date in October 2015 (107.8 mg N–NO3N kg−1) and on one date in July 2016 (average 85.4 mg N–NO3 kg−1). It was higher in ORG+ during August and September 2015; in this case after organic nitrogen fertilization for cabbage (93.1 mg N–NO3 kg−1) (Figures 3c and 4c).

#### *3.3. Daily Flux of N2O, CH4 and CO2*

Pattern of N2O, CH4 and CO2 fluxes throughout the study period are show in Figure 5a, b, c for F1 and Figure 6a, b, c for F2, while the ANOVA results are reported in Table 4.

#### 3.3.1. Trend of Daily N2O Flux in the Three Cropping Systems

Measured N2O daily flux ranged from −0.4 to 53.3 mg N2O m<sup>−</sup><sup>2</sup> day−<sup>1</sup> in F1 and from −1.7 to 20.2 mg N2O m<sup>−</sup><sup>2</sup> day−<sup>1</sup> in F2 (Figures 5a and 6a). Notably, high N2O fluxes were observed in F1 in June 2015 in ORG system after green manure incorporation into the soil (20.2 mg N2O m<sup>−</sup><sup>2</sup> day−1), in August 2015 (53.3 mg N2O m<sup>−</sup><sup>2</sup> day−1) in the ORG+ system just after organic fertilization on cabbage, and in April 2016 in the ORG system (37.3 mg N2O m<sup>−</sup><sup>2</sup> day−1) after tillage and organic nitrogen fertilization for spring lettuce.

In F2, N2O peaks were halved compared to F1 and the highest were registered after the organic nitrogen fertilization of fennel in September 2015 on ORG+ (16.4 mg N2O m<sup>−</sup><sup>2</sup> day−1), in October 2015 on ORG (on average 15.5 mg N2O m<sup>−</sup><sup>2</sup> day−1) and after green manure incorporation into soil in June 2016 in ORG (8.3 mg N2O m<sup>−</sup><sup>2</sup> day−1).

In P1 (Jan 2015-Dec 2015), average daily N2O flux (Table 4) in F1 was significantly lower in ORG+ (2.21 ± 1.18 mg N2O m<sup>−</sup><sup>2</sup> day−1), while no differences were observed between INT and ORG (on average 2.85 ± 0.32 mg N2O m<sup>−</sup><sup>2</sup> day−1). In F2, no differences were detected among the three cropping systems (on average 2.36 ± 0.29 mg N2O m<sup>−</sup><sup>2</sup> day−1).

During P2 (Jan 2016–Jul 2016), in F1 the effect of the cropping systems on the average daily N2O flux was the same as that in P1, with INT equal to ORG, and the highest values were recorded (on average 3.89 ± 1.15 mg N2O m<sup>−</sup><sup>2</sup> day−1) and ORG+ with the lowest value (0.47 ± 0.12 mg N2O m<sup>−</sup><sup>2</sup> day−1). In F2 N2O daily flux was significantly higher in ORG (2.63 ± 0.59 mg N2O m<sup>−</sup><sup>2</sup> day−1) than in ORG+ (1.39 ± 0.52 mg N2O m<sup>−</sup><sup>2</sup> day−1).

#### 3.3.2. Trend of Daily CH4 Flux in the Three Cropping Systems

Measured CH4 daily flux ranged from –0.7 to 0.45 mg CH4 m<sup>−</sup><sup>2</sup> day−<sup>1</sup> in F1 and from −0.47 to 0.43 mg CH4 m<sup>−</sup><sup>2</sup> day−<sup>1</sup> in F2 (Figures 5b and 6b). In F1, CH4 fluxes were positive (<0.2 mg CH4 m<sup>−</sup><sup>2</sup> day−1) in 12, 9 and 11 sampling days out of 50 in INT, ORG and ORG+, respectively. In F2, CH4 fluxes were positive (<0.5 mg CH4 m<sup>−</sup><sup>2</sup> day−1) in seven, two and six sampling days out of 48 in INT, ORG and ORG+, respectively. In particular, in F2, CH4 fluxes were significantly lower in ORG than in INT and ORG+ in two sampling dates in March 2016 and in May 2016; during which CH4 fluxes in ORG were equal to −0.35 ± 0.04 mg CH4 m<sup>−</sup><sup>2</sup> day−<sup>1</sup> and −0.34 ± 0.07 mg CH4 m<sup>−</sup><sup>2</sup> day−1, respectively. In F1 the average daily CH4 flux (Table 4) was slightly negative, with no significant differences (*p* > 0.05) among the cropping systems in both periods (on average P1: −0.10 ±0.22 mg CH4 m<sup>−</sup><sup>2</sup> day−1; P2: −0.08 ± 0.02 mg CH4 m<sup>−</sup><sup>2</sup> day−1). In F2, significantly lower flux was recorded in P1 in ORG (−0.23 ± 0.02 mg CH4 m<sup>−</sup><sup>2</sup> day−1) compared to INT and ORG+ (−0.18 ± 0.01 mg CH4 m<sup>−</sup><sup>2</sup> day−1), while in P2 ORG and

ORG+ showed similar values equal to –0.10 ± 0.02 mg CH4 m<sup>−</sup><sup>2</sup> day−1, significantly lower than INT (−0.001 ± 0.05 mg CH4 m<sup>−</sup><sup>2</sup> day−1).

**Figure 5.** Daily average fluxes recorded in F1 of: (**a**) N2O; (**b**) CH4; and (**c**) CO2 for each treatment. Simple arrows indicate fertilization events; dashed arrows the primary tillage of each crop. On field 1 (F1) the temporal crop sequence was: Fennel, summer lettuce, cabbage, then spring lettuce. Significance was as follows: *n.s.* is not significant; \* is significant at the *p* ≤ 0.05 level; \*\* is significant at *p* ≤ 0.01 level.

**Figure 6.** Daily average fluxes recorded in F2 of: (**a**) N2O; (**b**) CH4; and (**c**) CO2 for each treatment. Simple arrows indicate fertilization events; dashed arrows the primary tillage of each crop. On field 1 (F1) the temporal crop sequence was: Cabbage, spring lettuce, fennel, then summer lettuce. Significance was as follows: *n.s.* is not significant; \* is significant at the *p* ≤ 0.05 level; \*\* is significant at *p* ≤ 0.01 level; \*\*\* is significant at *p* ≤ 0.001 level.


**Table 4.** Effects of the three cropping systems on average daily flux of CO2, CH4 and N2O, during the two monitoring periods (P1: January 2015–December 2015; P2: January 2016–July 2016) in Field 1 and Field 2. System levels are INT: Integrated; ORG: Organic; ORG+: Conservation organic. Different letters represent significant

#### 3.3.3. Trend of Daily CO2 Flux in the Three Cropping Systems

Measured CO2 daily flux ranged from 3.9 to 60.9 g CO2 m<sup>−</sup><sup>2</sup> day−<sup>1</sup> in F1 and from 4.4 to 65.2 g CO2 m<sup>−</sup><sup>2</sup> day−<sup>1</sup> in F2 (Figures 5c and 6c). Daily pattern of CO2 flux varied according to that of soil temperatures. Indeed, higher values of CO2 flux were recorded from May 2015 to September 2015. Higher CO2 flux was observed in ORG+ than in the other systems during summer 2015 in both fields. In F1, significantly higher emissions were observed in ORG+ with respect to the other treatments in eight dates out of 50, while CO2 flux was higher in ORG than in the other systems in five dates out of 50. In F2, CO2 flux was significantly higher in nine dates out of 48 in ORG+ and in six dates out of 48 in ORG. Higher CO2 fluxes in ORG systems were observed in March 2015 after tillage for green manure sowing, and in summer 2015, some days after main tillage operations for summer lettuce, and cabbage in F1, and for fennel in F2. Otherwise, CO2 flux was significantly higher in INT than in the other treatments in only two dates in March 2016 in F2.

In both fields, daily average CO2 flux (Table 4) in P1 was higher in ORG and ORG+ (on average, F1: 21.51 ± 1.16 g CO2 m<sup>−</sup><sup>2</sup> day−1; F2: 23.88 ± 1.05 g CO2 m<sup>−</sup><sup>2</sup> day−1) than in INT (F1: 15.11 ± 0.98 g CO2 m<sup>−</sup><sup>2</sup> day−1, F2: 18.08 ± 1.00 g CO2 m<sup>−</sup><sup>2</sup> day−1) In P2 higher CO2 flux was higher in ORG+ in F1 compared to INT, while the opposite was recorded in F2, where INT and ORG (on average 25.40 ± 1.31 g CO2 m<sup>−</sup><sup>2</sup> day−1) recorded higher values than ORG+ (19.97 ± 0.79 g CO2 m<sup>−</sup><sup>2</sup> day−1).

#### *3.4. Relationship among the Soil Variables and GHG Fluxes*

The correlation between daily N2O flux and soil N–NO3 concentration-computed using a subset of the dataset, including only the monitoring days in which the soil samples were collected-turned out to be non-significant for all the three cropping systems (data not shown).

Considering the whole dataset, soil temperature and WFPS correlated negatively in all the three cropping systems, with a correlation coe fficient (rs) between −0.55 (ORG+) and −0.65 (INT) (Figure 7).

**Figure 7.** Correlation plot among the soil variables and GHG fluxes. Numbers indicate the correlation coe fficients, while the intensity of the colour of the boxes represents the level of correlation according to the scale reported close to each plot. Significance was as follows: \* is significant at the p ≤ 0.05 level; \*\* is significant at p ≤ 0.01 level; \*\*\* is significant at p ≤ 0.001 level.

Flux of N2O correlated positively with soil temperature in INT (rs: 0.38) and ORG+ (rs: 0.48); and with CO2 flux in ORG (rs: 0.42) and ORG+ (rs: 0.39).

Flux of CO2 correlated positively with soil temperature with rs equal to 0.69 in INT, 0.77 in ORG and 0.78 in ORG+; and negatively with WFPS with rs between −0.33 (INT) and −0.5 (ORG and ORG+). Flux of CH4 correlated positively with WFPS only in ORG (rs: 0.23) and negatively with CO2 flux in the same treatment (rs: −0.17).

Fluxes of N2O higher than 20 mg m-2 day−<sup>1</sup> were recorded with WFPS values between 38% and 70% (Figure S1a). When WFPS was lower than 38% N2O fluxes ranged between −0.04 mg m<sup>−</sup><sup>2</sup> day−<sup>1</sup> and 17.44 mg m<sup>−</sup><sup>2</sup> day−1, while when WFPS values were higher than 70%, N2O fluxes ranged between −0.07 mg m<sup>−</sup><sup>2</sup> day−<sup>1</sup> and 3.35 mg m<sup>−</sup><sup>2</sup> day−1.

The ANOVA describing the relationship between the logarithm of CO2 flux and the soil temperature highlighted that the slope of the linear regression was not different according to the treatments (0.059), while the intercept of the regression was significantly lower in INT (1.249) than in ORG (1.471) and ORG+ (1.487) (Figure 8).

**Figure 8.** Relationship between the logarithm of CO2 flux and the soil temperature.

#### *3.5. Cumulative Soil Emissions during the Two Periods*

In P1 cumulative N2O emissions showed no significant differences among the cropping systems both in F1 (average 5.5 ± 1.1 kg N–N2O ha−1) and in F2 (average 4.4 ± 0.5 kg N–N2O ha−1) (Figure 9a).

In P2 cumulative N2O emissions were significantly affected by the cropping system in both fields (*p* < 0.05). Indeed, in F1, cumulative N2O emissions were higher in INT and in ORG (average 2.0 ± 0.5 kg N–N2O ha−1) than in ORG+ (0.3 ± 0.1 kg N–N2O ha−1). In F2 N2O emissions were significantly higher in ORG (2.2 ± 0.7 kg N–N2O ha−1) than in ORG+ 1.0 ± 0.2 kg N–N2O ha−1), and INT was not significantly different from both ORG and ORG+ (1.1 ± 0.1 kg N–N2O ha−1).

There was an overall sink effect for CH4 cumulative emissions in all systems, in both periods and fields, with no significant differences among cropping systems (average in F1: −162 ± 38 g C–CH4 ha−1; average in F2: −356 ± 60 g C–CH4 ha−1) (Figure 9b).

Cumulative CO2 emissions in P1 were significantly affected by cropping system in both fields (F1: *p* < 0.01; F2: *p* < 0.05) (Figure 9c). Lower values were recorded in both fields in INT (F1: 13.0 ± 1.0 t C–CO2 ha−1; F2: 16.7 ± 0.8 t C–CO2 ha−1) than in ORG and ORG+ (average in F1: 18.3 ± 0.7 t C–CO2 ha−1; average in F2: 22.6 ± 0.6 t C–CO2 ha−1). In P2 differences were significant only in F1 (*p* < 0.01), where cumulative CO2 emissions were higher in ORG+ (8.5 ± 0.4 t C–CO2 ha−1) than in ORG and INT (5.4 ± 0.3 t C–CO2 ha−1).

**Figure 9.** Cumulative emissions of N2O (kg N–N2O ha−1), CH4 (g C–CH4 ha−1) and CO2 (t C–CO2 ha−1) for P1 and P2 in field 1 (**a, b, c**) and in field 2 (**d, e, f**), respectively. Different lowercase letters in P1, and uppercase letters in P2 indicate significant differences between the cropping systems resulting from the post-hoc test.

The estimated net GHG emissions (CO2-eq) were significantly affected by the cropping systems with exception of P2 in F2 (Table 5).

**Table 5.** Estimated cumulative CO2 emissions (t CO2 ha−1): CO2 equivalents of non-CO2 GHG as the sum of cumulative N2O and CH4 emissions (t CO2-eq ha−1), and net GHG emissions (t CO2-eq ha−1) during the two monitoring periods (P1: January 2015-December 2015; P2: January 2016– July 2016) in Field 1 and Field 2. System levels are INT: Integrated; ORG: Organic; ORG+, conservation organic. Different letters represent significant differences between the cropping systems resulting from the post-hoc test.


In P1 in both F1 and F2 the net CO2-eq were significantly higher in ORG (+40%, +33%), and ORG+ (+37%) than in INT. The CO2-eq of non-CO2 GHG were not different among INT, ORG and ORG+ in both fields and periods.
