*3.2. Soil Nmin Content*

The mass of Nmin at the four dissected depths (Figure 3) was expectedly different in Reference and lab-rain cores, with clear depletion of soil mineral N after 200 mm of simulated rain in all treatments. Likewise, total mineral N contents in Reference and lab-rain cores show significant N losses during the leaching experiment in all tillage and plant cover treatments, except for NT-V (Table 2). These losses, approximately 22 mg N per core on average, closely resemble the total NNO3 amounts recovered in the leachate for B in NT and both B and V in AuT.

Pairwise comparisons show that Nmin contents did not differ significantly between bare fallow and volunteer cores for either Reference or lab-rain cores in AuT. In contrast, pairwise comparison of NT Reference cores revealed a significantly higher total Nmin amount in B compared to V (13.6 mg N, SE = 4.96, *p* = 0.017) and, given that the volunteer cover crops were not incorporated in NT, this difference is attributable exclusively to a reduced cover crop N uptake. Interestingly, the amount of Nmin lost from NT-V cores was considerably smaller than in the AuT-B and AuT-V treatments. This supports our hypothesis that, through a combination of reduced uptake and re-mineralization, autumn tillage can hinder the intended function of cover crops.

**Figure 3.** Mineral nitrogen (Nmin) amounts determined in four 5 cm layers in each intact core. Lab-rain cores underwent 200 mm of simulated rain in a laboratory lysimeter before dissection, while Reference cores were dissected without simulated rain. AuT and NT represent, respectively, inversion tillage carried out in late October and no-till. B and V represent bare fallow and winter rye volunteers as cover crops, respectively. The solid lines indicate linear interpolations of Nmin changes with depth.

**Table 2.** Total mineral nitrogen (Nmin) in intact cores dissected after 200 mm of simulated rain in laboratory lysimeters (lab-rain) and without simulated rain (Reference). AuT represents cover crop termination by inversion tillage approximately one month before sampling. NT represents no-till and continued cover crop residence until sampling. B and V represent bare fallow and winter rye volunteers as cover crops, respectively. Model estimates and 95% confidence intervals (CI) for Nmin were obtained from linear mixed effects models with experimental set tillage and plant cover as main effects and experimental block as random effect.


<sup>1</sup> Grouping by pairwise comparisons restricted to Reference and lab-rain treatments of the same tillage and plant cover treatments. Significance threshold, *p* = 0.05.

Unexpectedly, lab-rain NT cores retained on average approximately 5 mg N per dissected layer after the leaching experiment. In contrast, lab-rain AuT cores retained an average N mass per layer of 2 mg N in B and 3.4 mg N in V (Figure 3). The difference between AuT and NT lab-rain cores is likely due to a bypass effect, where soil solutes dispersed in the matrix are partially protected from leaching in soils experiencing heavy precipitation. This effect has been found to be more prevalent in no-till and reduced till soils, as resident solutes in the matrix are increasingly bypassed by macropore water flow [33,34]. Importantly, this effect appears to be independent from cover crops treatment in our results, with no significant difference between B and V lab-rain total N content, suggesting that the soil N which is most strongly bypassed by macropore flow at high precipitation rates, is also poorly available for uptake by cover crops. Further research into the plant availability of matrix-associated Nmin in NT systems is necessary, particularly in the context of preferential exploration of pre-existing macropores by plant roots in more compact soils [35].

Organic elemental analysis of topsoil samples showed no significant main effects of cover crop or tillage on either total C or total N contents (mean values 15.13 mg C g−<sup>1</sup> and 1.37 mg N g<sup>−</sup>1), in spite of near-significant trends of greater total C (3.6 mg C g−1, SE = 1.58, *p* = 0.052) and total N (0.29 mg N g−1, SE = 0.149, *p* = 0.088) in NT compared to AuT. The trend of greater total C content in NT resembles results by Gómez-Muñoz et al. [19], who found significantly higher soil total C contents in the upper 25 cm of NT soil in the same field experiment. However, lack of significant differences in our total C and N measurements and the fact that leaching was high in NT-B as well as AuT-B and AuT-V suggest that the results from the leaching experiment are not related to any underlying differences in total C and N soil contents. In Denmark, inversion tillage is typically carried out with a plough depth of 20–30 cm. It is therefore possible some volunteer plant material was buried below the sampling depth of the soil cores, in which case a portion of the corresponding remineralization of cover crop N would not have been captured by the leaching experiment. However, examination of the mineral N contained in the dissection layers at different depths (Figure 3) shows that Nmin is more evenly distributed across depths in AuT-V compared to AuT-B, suggesting that cover crop residues were mixed throughout the plough layer and the taken-up N had begun to re-mineralize in the one month span between tillage and sampling. Indeed, incubation studies have previously found that between 20% and 60% of the total N content in cover crop residues can mineralize within 1 month of incorporation at temperatures similar to those found in the field in autumn and winter [36,37].
