**3. Results**

#### *3.1. Weather Conditions*

As shown in Figure 1, the weather conditions in the three experimental years differed from the normality for the area in many cases.

Monthly mean maximum temperatures were higher than multiannual values in most cases during the three experimental years. In particular, the winters were always warmer than usual, and so, it was also for the summer seasons in 2015 and 2017. Values below the multiannual means were registered in summer 2014 and in winter 2016/2017, instead. The hottest months in terms of maximum mean temperature were July 2015 and August 2017 with 32.4 ◦C. The lowest maximum mean temperature was registered in January 2017 (10.3 ◦C).

Likewise, the mean minimum temperatures were higher than multiannual values in the experimental period. Only in winter 2017, we observed values below the normality of the period. The coolest month was January 2017, indeed, with −0.8 ◦C, and the hottest was July 2016 (+20 ◦C).

The three experimental years were also characterized by high levels of rainfall compared to the multiannual trend. The rainiest months were January 2014 (355 mm), November 2014 (290 mm), October 2015 (254 mm), and September 2017 (234 mm). Unusual high peaks of rainfall occurred also in July 2014, August 2015, June 2016, September 2016, and September 2017. The driest month was July 2017 when the experiment ended, with no rainfall registered at all.

**Figure 1.** Monthly total rainfall (mm) and mean maximum and minimum air temperature (◦C) from January 2014 to December 2017 compared to multiannual mean values (1993–2017).

#### *3.2. Living Mulch, Dead Mulch, and Green Manure Biomass Production and NP Accumulations*

The dry biomass produced by the spring cover crops never exceeded 3 Mg ha−1, and it varied over years (Table 1). The mixture of field pea and pigeon bean grown in the ORG system clearly produced higher biomass than the red clover living mulch (ORG+), which was productive only in the first year. Red clover produced very low biomass especially in the second year, likely because of the low rainfall occurred in spring.


**Table 1.** Dry matter production of spring and summer cover crops and their weeds at termination dates in organic farming (ORG) and organic and conservative system (ORG+) in the three years. Means ± SE.

1 n.a. is not available; † Tp is *Trifolium pratense*, Vf is *Vicia faba* var. minor, Ps is *Pisum sativum*, Vu is *Vigna unguiculata*, Fe is *Fagopyrum esculentum*, Pm is *Panicum miliaceum*, Si is *Setaria italica*, Tot is total biomass of cover crop mixtures, and We is total weed biomass.

The summer cover crop mixture was more productive and stable than the spring one. This was mainly because of the constantly high biomass production of foxtail millet and grain millet. Buckwheat biomass was very scarce, especially in the last two years, whereas red cowpea was steadily present over years at around 1 Mg ha−<sup>1</sup> (Table 1).

#### *3.3. Field Vegetable Biomass Production and NP Accumulations*

The results of the statistical analysis of savoy cabbage, fennel, spring lettuce, and summer lettuce yield, biomass production, produce dimension, and NP concentration and accumulation are reported, respectively, in Tables S1–S4.
