3.3.2. Fennel

For fennel, the e ffect of the cropping system was significant for all the dependent variables tested except for HI and P2O5accr (Table S2). The year e ffect was not significant only for Naccr, Pconcy, Pconcr, P2O5accr, and P2O5acct.

The data on biomass production of fennel at harvest time in all three years are reported in Table 4. For this crop, there were no significant differences between INT and ORG, although the integrated system always resulted in the highest values of fresh and dry marketable yield. For the organic conservative system, we observed encouraging results, as the fennel in ORG+ plots performed statistically equal to ORG in 2014 and 2015, although always significantly lower than INT. In 2015, the fresh dry matter of swollen bases collected in the ORG+ plots was even higher than that in the ORG plots, although not significantly. The same trend was observed also for the mean fresh weight of the swollen bases.

Overall, the yield depletion observed in the ORG+ system, compared to ORG and INT, averaged ca. 35%. The dry matter of residues in ORG+ was statistically not di fferent from ORG only in 2015 and was always lower than INT. The dry matter of the total aboveground biomass was not di fferent between INT and ORG, whereas it was significantly lower in ORG+. For HI, as mentioned, we did not observe any di fferences among the three cropping systems. The mean diameter of the swollen bases was significantly a ffected by the cropping systems and the highest values were always shown by INT. The ORG system produced bases with similar diameters to INT in 2014 and 2016. In 2015, ORG was significantly lower than ORG+ and INT, instead.

Data on N and P concentration and accumulation in biomass components of fennel under the three cropping systems are shown in Table 5. In 2014 and 2016, N concentration was higher in INT than ORG heads and in ORG than ORG+ heads. In 2015, ORG+ showed higher values than ORG. For N concentration in crop residues, the INT system showed significantly higher values than ORG and

ORG+ in 2015 and 2016. In 2014, there were no differences between INT and ORG, only with ORG+. In 2015, ORG and ORG+ showed similar results. N accumulation in marketable dry matter was higher in INT than ORG in all three years. Only in 2015, we did not observe any differences between ORG and ORG+. The same trend was observed also for N accumulation in crop residues, a parameter for which there were no significant differences between INT and ORG in 2014 and 2016. For P uptake, the differences among the three systems were less evident. ORG+ showed similar results to INT in many cases (i.e., for Pconcy in all three years, for Pconcr in 2015, for P2O5accy in 2014, and for P2O5accr and P2O5acct in all three years).

**Table 4.** Least squares means and standard errors of marketable fresh yield (Y), dry matter of marketable yield (dwy), dry matter of residues (dwr), total aboveground dry matter (dwt), mean fresh weight of swollen bases (MFW), Harvest Index (HI), and mean diameter of swollen bases (MD) in fennel. Confidence level: 95%.


Means followed by different letters are statistically different (95% confidence interval).

**Table 5.** Least squares means and standard errors of N concentration in marketable yield (Nconcy) and residues (Nconcr); N accumulation in marketable yield (Naccy), residues (Naccr), and total aboveground dry matter (Nacct); P concentration in marketable yield (Pconcy) and residues (Pconcr); and P2O5 accumulation in marketable yield (P2O5accy), residues (P2O5accr), and total aboveground dry matter (P2O5acct) in fennel. Confidence level: 95%.



**Table 5.** *Cont*.

Means followed by different letters are statistically different (95% confidence interval).
