Nitrogen Rate Increase Not Required for No-Till Wheat in Cool and Humid Conditions

An increased nitrogen (N) supply was proposed to avoid grain yield (GY) reductions and successfully implement conservation agriculture (CA). We investigated interactions effects of tillage system and N supply on winter wheat (Triticum aestivum L.) at two sites in the Swiss midlands with no (0 kg N ha−1) and high (150–160 kg N ha−1) N supply using 15N-labelled ammonium nitrate in selected treatments. Wheat’s GY, yield components, N related traits and soil mineral N content (Nmin) under conventional tillage (CT), minimum tillage (MT), and no-tillage (NT) were studied following two preceding crops: oilseed rape (Brassica napus L.) and maize (Zea mays L.). Wheat after oilseed rape had significantly higher GY and biomass than after maize while a yield decrease under NT compared with CT and MT was observed regardless of N supply level. Differences in soil Nmin among tillage systems were seldom found and were inconsistent. No differences in 15N fertilizer recovery were found between NT and CT while residual Nmin after harvest was lower under NT than CT or MT. In conclusion, we did not found consistent reductions in soil N availability and N uptake under NT that would justify an increased N supply for wheat under CA.


Introduction
Conventional farming practice, based on intensive tillage and removal of residues from the soil surface, can result in serious soil degradation [1]. Conservation agriculture (CA) has been promoted as an agricultural system that increases agricultural sustainability and has the potential to mitigate greenhouse gas emissions [2,3] and climate change [4]. CA is based on minimizing soil disturbance and using adequate levels of crop residues for soil and water conservation [5]. Reduced tillage systems have been designed to minimize soil disturbance and offer long-term benefits from an improved soil structure [6], restored soil organic carbon stocks, improved soil properties, and reduced soil erosion [7].
The winter wheat (Triticum aestivum L.) variety Runal (Breeder: Swiss Federal Research Station for Agroecology and Agriculture, FAL, Zurich, Switzerland), which is a high-quality variety with an intermediate yield potential was grown. As of 2020, Runal was still cultivated and included in the official list of recommended varieties [43]. Following the guidelines for fertilization of the Swiss federal research stations, 22 kg P ha −1 and 8 kg Mg ha −1 were broadcasted at the sowing of winter wheat. At each site winter wheat was grown succeeding maize (Zea mays L.) or oilseed rape (Brassica napus L.). These preceding crops are common crops cultivated in rotations with winter wheat in Switzerland. The following three tillage systems were studied: CT with mouldboard plough (CT, 25 cm depth), minimum tillage with chisel (MT, 15 cm depth), and no-tillage (NT). Two rates of N fertilization were compared: 0 kg N ha −1 (N0) and recommended N fertilization (N1). For the recommended N rate, N was broadcasted as ammonium nitrate (NH 4 NO 3 ) at three stages (according to BBCH scale [44]): beginning of tillering (BBCH 25), first-node stage (BBCH 31) and appearance of the flag leaf (BBCH 37). Recommended N rate applied at beginning of tillering was estimated as 120 kg N ha −1 -soil N min (0 to 90 cm) while at the first-node stage and at the appearance of the flag leaf, the amount of N fertilizer was calculated according to a N status test (Lonza, Basel, Switzerland) and applied up to a total supply of 30 and 50 kg N ha −1 , respectively. The N status test was based on the nitrate concentration in the basal stem sap as proposed by Justes et al. [45]. The total amount of N fertilizer for the recommended rate ranged from 150 to 160 kg N ha −1 (N1).
Wheat was sown at Zollikofen on 14 October 1995, 4 November 1996, 28 October 1997 December 1998 and at Schafisheim on 4 November 1996, 24 October 1997, 9 November 1998 October 1999. In the CT and MT treatments, wheat was sown with a 'Rototiller' rotary harrow-drill combination. In the NT treatment, wheat was sown with a disc-opener, John Deere 'NT 750 A' planter (Deere and Co., Moline IL, USA) directly into the dead mulch. In all the tillage systems, seeding rate was 400 grains m −2 . The distance between plant rows was 14.3 cm at Zollikofen and 12.5 cm at Schafisheim in the CT and MT systems; the distance between the rows was 16.6 cm in the NT systems at both sites.
Long-term climatic data from meteorological stations at Berne-Liebefeld (near Zollikofen) and Buchs-Suhr (near Schafisheim) were obtained from the Swiss Meteorological Institute (SMI, Zurich). In the 20 years prior to the experiments, mean annual temperature and precipitation at Zollikofen were 8.7 • C and 1075 mm, respectively. The corresponding values at Schafisheim were 9.2 • C and 1047 mm, respectively. Growing conditions in the two sites were representative of average conditions for winter wheat cropping in Switzerland.
The experimental design was a split-plot with completely randomized blocks and two treatments (tillage systems and N supply). Tillage systems (CT, MT, and NT) and nitrogen supply (N0 and N1) were assigned to the main and the subplots, respectively. The tillage main-plots were 12 × 35 m 2 and separated from each other by a one meter border. The N subplots were 6 × 35 m 2 . Each combination of main-plot and subplot was replicated three times while each different combination of year, site, and preceding crop was considered a single environment in the statistical analysis.

15 N labelled Fertilizer Assessment
This assessment was conducted for two years at each site. In the CT and NT treatments, two zones per plot were defined to apply and trace the fate of 15 N-labelled fertilizer for the recommended N rate treatment.
Each zone was defined at the center of the plots to avoid the dispersion of 15  At the first-node stage (BBCH 31) and at the appearance of the flag leaf (BBCH 37), the required amount of N fertilizer in each plot was calculated using the above described N status test and applied as non-labeled NH 4 NO 3 to complete the N rate of the treatment.

Plant Sampling and Analysis
Shoot samples were taken at tillering (BBCH 25), shooting (BBCH 31), and anthesis (BBCH 65) from four areas (0.25 m 2 ) per plot. At maturity (BBCH 92), an area of 15 m 2 per plot was harvested by combine harvester (Wintersteiger, Ried, Austria). Straw was collected and its fresh weight determined on the field; a random sample was taken to determine the dry weight and the N content. Plant material was dried at 65 • C for 48 h and then weighed. The grains were ground with a mill ('A10 , Janke & Kunkel Labortechnik, Staufen, Germany) prior to analysis. The milled grains were analyzed for N with a LECO CHN-1000 autoanalyzer (LECO Corporation, St. Joseph, MI, USA). This instrument involves dry combustion according to the Dumas principle to extract N from the samples.
Shoot samples from the subplots for 15 N-labelled fertilizer assessments were taken from two places per 15 N application zone at the end of tillering (BBCH 29) after 15 N application, at shooting (BBCH 31), at anthesis (BBCH 65), and at physiological maturity (BBCH 92). At each sampling, plants were cut at soil surface level. Shoots were divided into grains and vegetative parts (chaff and straw). Dried shoot material (65 • C for 48 h) was weighed. Vegetative parts were ground with two mills, through a 3-mm sieve (Wolf Mühle, Wien, Austria) and then through a 1-mm sieve ('Cyclotec Tecator 1093 mill, Tecator AB, Höganäs, Sweden), the grains with an 'A10 mill (Janke & Kunkel Labortechnik, Staufen, Germany). Then, the plant and grain material was ground with a 'MM2 ball mill (Retsch, Arlesheim, Switzerland). N and 15 N concentrations were determined with a bench-top isotope ratio mass spectrometer (Europa Scientific Integra, Cambridge, UK) in the stable isotope facility of the University of California at Davis, USA. %Ndff calculations (percentage of plant N derived from N fertilizer) were made as follows [48]: %Ndff = (%N excess in sample/%N excess in fertilizer) × 100 (1) The resulting percentage was multiplied by the total N quantity in the plant sample to obtain a value of plant derived from N fertilizer in kg ha −1 (N R ).

Soil Sampling and Analysis
At the end of winter, four soil cores (0.0 to 90 cm) per plot were taken using a 'Pürckhauer' auger (Eijkelkamp, Giesbeek, The Netherlands) and separated according to depth (0 to 15 cm, 15 to 30 cm, 30 to 60 cm, and 60 to 90 cm). Samples were taken right after shoot sampling from the centre of the shoot sampling areas. Additional, soil samples were taken in the autumn, at the end of the vegetation period, following the same protocol described above. Soil samples were frozen at −18 • C until analysis. For each sample, 100 g of soil were extracted using 100 mL of a 0.005 M CaCl 2 solution. After shaking the solutions for 90 min, the samples were filtered through N-free filters. Concentrations of NO 3 − -N and NH 4 + -N in the extract were determined using a colorimetric method with an 'Evolution II' autoanalyser (Alliance Instruments, Nanterre, France).

Statistical Analysis
Data were analyzed by means of the analysis of variance (ANOVA) according to the GLM procedure of the SAS software [49]. Briefly, the GLM procedure uses the method of least squares to fit general linear models and allows specifying interaction and nested effects. We also used this procedure to compute least square means and least square mean differences. Factors in ANOVA were tillage system and N fertilization level. Environments were considered combinations of year, site, and preceding crop. Single ANOVAs were conducted for each environment. As meaningful interactions between tillage systems and the level of N fertilization did not occur, we only present results from the recommended N fertilization level. On the other hand, data are shown separately for single environments because strong effects of environment existed which included the effects of the preceding crops oilseed rape and maize.

Soil Mineral N Content (N min )
Total soil N min (Figures 1 and 2) varied among environments (i.e., combination of year, site, and preceding crop) with a mean value of 54 kg N ha −1 and a coefficient of variation of 30%. Total soil N min contents at Schafisheim were generally higher than at Zollikofen. At Schafisheim, tillage systems had no significant effect on N min , with the exception of the assessment at the maturity of wheat ( Figure 2). In contrast, at Zollikofen soil N min contents were significantly (p < 0.05) affected by preceding crop and tillage system. Total soil N min (0 to 90 cm) tended to be higher after oilseed rape than after maize in the heavier soil (Figure 1c,e) and this was especially the case in the topsoil (0 to 30 cm). Averaged across all treatments, the soil N min content of the MT system was higher than that of NT and slightly but significantly higher than that of CT system (data not shown). Around 60% of the total soil N min content was found in the topsoil, regardless of the tillage system and site (Figures 1 and 2). Before winter (i.e., December in Figures 1 and 2), the average soil N min content was 10 kg N ha −1 higher under MT than under CT and 8 kg N ha −1 higher than under NT. After winter (i.e., March in Figures 1 and 2), soil N min contents were still the highest under MT, but differences among tillage systems were usually below 10 kg N ha −1 . Changes in soil N min content between autumn and winter occurred mainly in the topsoil while the soil N min content in the subsoil (30 to 90 cm) was less variable. Net increases in soil N min contents from the beginning of the vegetation period until shooting were, on average, slightly higher under CT and . At shooting, the average N min content was nearly twice as high as after winter; and although average differences among tillage systems were small, the highest values were still recorded under MT (Figure 1a,d, and Figure 2b). From shooting to anthesis, soil N min contents decreased by around 30 kg N ha −1 , and significant differences disappeared among tillage systems (Figures 1 and 2). The subsoil (30 to 90 cm) contained only marginal levels of mineral N with values below 20 kg N ha −1 . Between anthesis and maturity, few changes in the content of soil N min were observed (Figures 1 and 2). Therefore, N mineralization on the one hand and the sum of N uptake and N losses on the other were probably counterbalanced. Average soil N min at maturity (an indicator of residual N) was significantly lower (by about 8 kg N ha −1 ) under NT than under CT and MT, with a significantly higher percentage of N min in the topsoil than in the subsoil (Figure 1b fit general linear models and allows specifying interaction and nested effects. We also used this procedure to compute least square means and least square mean differences. Factors in ANOVA were tillage system and N fertilization level. Environments were considered combinations of year, site, and preceding crop. Single ANOVAs were conducted for each environment. As meaningful interactions between tillage systems and the level of N fertilization did not occur, we only present results from the recommended N fertilization level. On the other hand, data are shown separately for single environments because strong effects of environment existed which included the effects of the preceding crops oilseed rape and maize.

Soil Mineral N content (Nmin)
Total soil Nmin (Figures 1 and 2) varied among environments (i.e., combination of year, site, and preceding crop) with a mean value of 54 kg N ha −1 and a coefficient of variation of 30%. Total soil Nmin contents at Schafisheim were generally higher than at Zollikofen. At Schafisheim, tillage systems had no significant effect on Nmin, with the exception of the assessment at the maturity of wheat ( Figure 2). In contrast, at Zollikofen soil Nmin contents were significantly (p < 0.05) affected by preceding crop and tillage system. marginal levels of mineral N with values below 20 kg N ha −1 . Between anthesis and maturity, few changes in the content of soil Nmin were observed (Figures 1 and 2). Therefore, N mineralization on the one hand and the sum of N uptake and N losses on the other were probably counterbalanced. Average soil Nmin at maturity (an indicator of residual N) was significantly lower (by about 8 kg N ha −1 ) under NT than under CT and MT, with a significantly higher percentage of Nmin in the topsoil than in the subsoil (Figure 1b

Establishment and N Uptake of Wheat
The emergence of wheat plants was lower and less regular under NT than under CT and MT. There was a significant reduction (20%) in wheat plant density under NT compared to CT at Zollikofen in 1996 and at both sites in 1998, while the number of wheat plants per area was intermediate under MT (data not shown). Despite this initial difference, wheat under NT compensated for the reduced plant stand with a higher number of fertile tillers.
Shoot N uptake at the early vegetative stages was the highest under CT or MT, while values under NT were 25% lower (Figures 3 and 4). Wheat after oilseed rape had 25% higher shoot N uptake than wheat after maize and N uptake was relatively higher at Schafisheim than at Zollikofen (Tables 2 and 3). Differences among environments were pronounced until the end of tillering, with generally lower rates of shoot N uptake at Zollikofen. At stem elongation, wheat under NT had the highest shoot N concentrations, followed by MT. Significant differences in the shoot N concentration were found among environments at this stage, mainly caused by higher N concentrations at Schafisheim than at Zollikofen (Figures 3 and 4).
From stem elongation onwards, differences among tillage systems largely disappeared while differences between environments dominated. The shoot N concentration increased slightly until stem elongation and then decreased steadily until physiological maturity (Figures 3 and 4). Shoot N uptake at the beginning of stem elongation was, on average, lower under NT than under CT. Shoot N uptake under MT was also slightly lower than under CT in most environments. However, these differences disappeared by anthesis (Figures 3 and 4) and after grain filling, shoot N uptake was similar in all tested environments and under the three tested tillage treatments. uptake than wheat after maize and N uptake was relatively higher at Schafisheim than at Zollikofen (Tables 2 and 3). Differences among environments were pronounced until the end of tillering, with generally lower rates of shoot N uptake at Zollikofen. At stem elongation, wheat under NT had the highest shoot N concentrations, followed by MT. Significant differences in the shoot N concentration were found among environments at this stage, mainly caused by higher N concentrations at Schafisheim than at Zollikofen (Figures 3 and 4).  At maturity, differences in shoot N uptake due to preceding crop were often associated to differences in shoot biomass (Tables 2 and 3). Wheat after oilseed rape produced a significantly higher shoot biomass than wheat after maize at nearly all the investigated growth stages (data not shown). This was reflected in shoot N uptake which was higher after oilseed rape than after maize, with the exception of plots under NT. At maturity, the shoot biomass was only slightly, but not significantly lower under NT than under CT and MT (Tables 2 and 3). These effects were, however, different across environments. The shoot dry matter of wheat was 15% higher after oilseed rape than after maize and it was slightly lower at Zollikofen than at Schafisheim but not significantly different (Tables 2 and 3). From stem elongation onwards, differences among tillage systems largely disappeared while differences between environments dominated. The shoot N concentration increased slightly until stem elongation and then decreased steadily until physiological maturity (Figures 3 and 4). Shoot N uptake at the beginning of stem elongation was, on average, lower under NT than under CT. Shoot N uptake under MT was also slightly lower than under CT in most environments. However, these differences disappeared by anthesis (Figures 3 and 4) and after grain filling, shoot N uptake was similar in all tested environments and under the three tested tillage treatments.
At maturity, differences in shoot N uptake due to preceding crop were often associated to differences in shoot biomass (Tables 2 and 3). Wheat after oilseed rape produced a significantly higher shoot biomass than wheat after maize at nearly all the investigated growth stages (data not shown). This was reflected in shoot N uptake which was higher after oilseed rape than after maize, with the exception of plots under NT. At maturity, the shoot biomass was only slightly, but not significantly lower under NT than under CT and MT (Tables 2 and 3). These effects were, however, different across environments. The shoot dry matter of wheat was 15% higher after oilseed rape than after maize and it was slightly lower at Zollikofen than at Schafisheim but not significantly different (Tables 2 and 3).

15 N-Labelled Fertilizer Recovery
Labelled-15 N fertilizer allows to trace the fate of N supplied as fertilizer and differentiate it from other N sources (Table 4). About 50% and 40% of the basal N application were recovered at physiological maturity in the shoot and grains, respectively. Recovery of 15 N was significantly higher at Zollikofen (silty-loam soil) compared to Schafisheim (loamy soil) and in 1998 than in 1997. Differences between sites were higher than those between years for the same site (Table 4). Table 4. Effects of conventional (CT) and no-tillage (NT) on the percentage of nitrogen in plants derived from labelled-15 N from the basal N application (Ndff) and labelled fertilizer-N recovery (NR) in plants (kg ha −1 ) at four growth stages of wheat and four environments.

Environment
Tillage

Grain Yield and Yield Components
In most environments, there were no significant differences in grain yield among tillage systems, with the exception of Zollikofen in 1998 and 1999 when oilseed rape was the preceding crop. On average, wheat's grain yield was 8% higher after oilseed rape than after maize (Tables 2 and 3).
There were only isolated effects of tillage system on yield components and often depended on the environment. Tables 2 and 3 show selected yield components at Zollikofen and Schafisheim, respectively. When oilseed rape was the preceding crop, there were on average 10% more spikes m −2 than after maize. In the few cases that there were significant differences due to tillage on yield components (i.e., after maize in Zollikofen in 1998 and 1999, Table 3), compensations between yield components prevented these differences to be reflected on the grain yield. For example, spike density was significantly higher under MT than under NT while grains spike −1 , were also significantly different, showing a reversed ranking (i.e., NT > MT). Although it was often not significant, thousand-kernel weight was consistently lower under NT than under CT and MT (Tables 2 and 3) and this contributed to lower grain yields under NT compared to CT at Schafisheim (Table 3). Similarly, grain N content tended to be higher under CT and MT than under NT, however the differences were small and insignificant.
The lack of significant and consistent effects of tillage system and N supply on grain yield and grain yield components suggest that a dependence of tillage system on the rate of N fertilization can be excluded in the environments.

Discussion
Using soil, shoot and 15 N-labelled fertilizer measurements, we did not find evidence of significant interactions between tillage system and N supply that would justify an increase of the N supply for NT cropping of wheat.
In general, the soil N min content at the beginning of the vegetation period varied in similar ranges of about 25 to 65 kg N ha −1 for the different tillage systems (Figures 1 and 2). The soil N min contents were usually higher after oilseed rape than after maize, as was also found by other authors [50,51]. Similarly, less soil nitrate was released after different cereals than after oilseed rape and higher soil N min contents were found during spring when wheat followed oilseed rape than when it followed cereals [52,53]. This was presumably the consequence of: (i) a higher mineralization potential of soil N after the harvest of oilseed rape, (ii) the uptake of N by volunteers of oilseed rape competing with the succeeding crop, and (iii) a higher N release from decomposition of oilseed rape residues. Higher N release from residues of oilseed rape may be explained by a lower C/N ratio than cereals [54]. We did not observe the expected soil N min peak with CT [55] presumably due to N leaching. In spring, MT soils had on average the highest soil N min contents. Similarly, in experiments conducted by Hansen and Djurhuus [56], higher N mineralization in autumn resulted in a higher soil N min content in spring. The decrease in N min in the topsoil (0 to 30 cm) and the constant content in the subsoil (30 to 90 cm) that we observed during winter (Figure 1d,e and Figure 2b,c) suggests a significant amount of N lost by leaching or denitrification. N min in soil under MT seemed to have been the most susceptible to N leaching losses since it tended to have the highest mineral N contents in the subsoil before winter. The net decrease of soil N min during winter was higher for wheat after maize than after oilseed rape because average soil N min contents were higher before winter for wheat grown after maize. The soil N min content did not generally change between anthesis and maturity, suggesting that N mineralization and N uptake were counterbalanced at a level of 30 kg N ha −1 in all tillage systems during this period. Thus, a more intense N mineralization during later growth stages under NT as sometimes observed elsewhere [57,58] was not observed in our study. In any case, soil N min values obtained at different stages throughout the growing season of wheat do not suggest consistent N limitations in NT compared to the other tillage systems.
Wheat plant density in autumn increased with tillage intensity, as was also observed by other authors [59]. This may be attributed to suboptimal soil physical conditions for germination and emergence that were found under NT systems. To overcome this limitation and to increase germination and emergence ratios, Sunderman [60], recommended the use of wheat varieties with higher grain weights with NT systems. This justifies partly the choice of the cultivar 'Runal' in our experiments, since it had a thousand-kernel weight above average as compared to other wheat cultivars registered in the Swiss variety catalog. Accordingly, several other authors have recommended the cultivation of genotypes with early vigor traits under CA systems [61][62][63][64]. Similar to Rao and Dao [65] and Cornish and Lymbery [66], we observed a lower shoot N uptake at NT than at CT during the early stages (Figures 3 and 4), but from stem elongation to physiological maturity and with the exception of wheat after oilseed rape at Zollikofen in 1998, significant tillage effects on shoot N uptake were no longer found. Despite the lower total shoot N uptake at tillering, from stem elongation to physiological maturity under NT, wheat showed an increased rate of shoot N uptake compared to other tillage systems. This suggests that an increased N mineralisation potential in the topsoil (0-30 cm) under NT, may have prevented limitations on N uptake due to a lower root length density in the same soil layers [67]. In addition, since the formation of tillers is positively correlated with N supply [68] and a higher number of fertile tillers were observed under NT, N availability seemed not to have been a limiting factor under NT. In the present experiment, differences in net N uptake between consecutive sampling dates among tillage systems were very small throughout the growing season. Thus, it seems unlikely that a severe decrease in N availability under NT occurred. This is supported by the observation that the N concentration in aboveground vegetative parts during the vegetation period was not lower under NT than under the other tillage systems. This result was consistent with what we found using a chlorophyll meter (SPAD) readings which were measured on the flag leafs of wheat (data not shown). Therefore, shoot N uptake also supports the finding that the N fertilisation strategy, which was developed for CT systems, was not detrimental to the N availability and, ultimately, to the productivity of wheat under NT.
The recommendation to apply greater amounts of N fertilizer to winter wheat under NT, as proposed by Francis and Knight [69] and Rasmussen and Douglas [70] is not supported by the results of 15 N-labelled fertilizer. Recovery of 15 N-labeled fertilizer at physiological maturity was on average 30 kg N ha −1 in the grains, which is about 40% of the applied N fertilizer. Thus, nearly a quarter of the total N in grains was derived from the first fertilizer application. Labelled-15 N fertilizer recovery was almost always higher under CT than under NT. However, these differences were not significant, with the exception of the measurement taken at stem elongation at Zollikofen in 1998. This differs from the findings by Carefoot et al. [71] who observed a greater %Ndff in grains at NT than in the CT. The temporal pattern of 15 N-fertilizer recovery from the basal N application (Table 4) was almost parallel to that of shoot N accumulation (Figures 3 and 4). The slightly lower uptake of 15 N-labeled fertilizer may be explained by the lower shoot development and consequently reduced N uptake under NT. Losses of 15 N-labelled fertilizer (approximately 5 kg N ha −1 ) are in the same range of previous reports using similar N rates [72]. More fertilizer N may have been immobilized initially under NT; but this could have been only temporarily as almost the same amount of the initial fertilizer application was recovered in both tillage systems. In contrast, an increased N rate of fertilization under NT would have probably increased N losses above the already high 60% of the applied N.
The preceding crop had a remarkable effect on grain yield despite identical sowing dates. In a continental climate, modifications in grain yield by NT were probably due to differences in water retention by preceding crops [73,74]. In the climatic conditions of this study, higher soil N min contents were more likely when wheat followed oilseed rape than maize. In addition, oilseed rape, with deep and dense root systems [75], leaves a good soil structure, an ideal condition for the following crop as shown in experiments in Germany [76]. Thus, the amount and C/N ratio of the residues of the preceding crop [77] are critical factors to consider while designing the N fertilization management of a CA system.
Tillage systems did not affected consistently grain yield and its components, with the exception of thousand-kernel weight. Differences among tillage systems were not highly pronounced in most environments. Similar findings with regard to tillage effects on grain yield were reported for wheat in temperate climates [58,78,79]. However, in some cases, significantly lower grain yields, under reduced or NT, were found in Scotland [80], where winter barley was grown on a clay soil, and in Germany for several cereals [81]. To attain the same grain yield as with CT, N fertilization was increased with reduced tillage in the latter two cases. In contrast, only a small overall decrease in grain yield was found in NT compared with CT and MT without N fertilization [27,82]. This may explain the missing interaction effect between N supply and tillage intensity in our environments. Reduced grain weight is a known symptom of Fusarium infection [83]. NT wheat showed the highest incidence of infection, followed by MT, whereas infection of CT wheat was negligible (data not shown). Fusarium infections are a severe problem in reduced tillage systems [84] because the survival of the pathogen is facilitated by the presence of cereals' residues [85]. Thus, crop rotations with cereals followed by cereals are problematic. We also did not find an effect of tillage system on grain N content, similar to the findings of Ditsch and Grove [86] and contrary to those of Carter et al. [87]. The lack of effect occurred at both studied N supply levels. Therefore, the results on wheat quality also do not suggest the need to increase N supply because of NT. In contrast, our results suggest that a diagnose of N availability after winter and the use of suitable crop rotations allows for utilizing N fertilizers more efficiently under CA in a cool and humid climate.

Conclusions
With crop residues left on the field and suitable rotations, GY, yield components, N uptake, N fertilizer recovery and soil N min availability of wheat under NT were comparable to those under MT or CT in a cool and humid climate. We did not find evidence of interactions between tillage system and N supply that would justify an increase of the N supply for CA.
Oilseed rape as a preceding crop improved the productivity of the subsequent wheat compared to growing maize as a pre-crop of wheat. As wheat without N fertilization under NT showed similar decreases in grain and biomass yields compared to CT or MT, it can be assumed that N was not a limiting factor for the productivity of wheat under NT. This is supported by the similar soil N min contents in all the tillage systems throughout the growing season and the similar recovery of 15 N-fertilizer. With regard to product quality, the N content of the grains was not negatively affected by reduced tillage. In addition, the generally strong influence of the environments, composed by sites, years and preceding crops, on the expression of differences between tillage systems may explain why results between different published experiments often differ drastically.