Winter wheat provides the economic foundation for the Wilmersdorf organic farm and has, thus far, been cultivated solely according to standard procedures. These standard organic farming procedures entail sowing winter wheat on fields prepared with a two-year multispecies legume-grass (LGS), which is mulched two to four times a year and ploughed under in autumn of the second harvest year. Directly following this virgin tillage (i.e. ploughing the sward without prior shallow soil processing), winter wheat is sown in mid-October. The subsequent crop is usually winter rye. This standard cultivation practice is practiced on large, rolling fields averaging 40 hectares, interspersed with hills and hollows. Such small-scale heterogeneity with varying soil types (see Section 2.2), further complicates ploughing and seedbed preparation. Poor soil contact and uneven germination are common problems. Particularly during dry periods, a common occurrence at Wilmersdorf (Section 2.3), hilltops are low yield (problem) areas and ploughing the shallow topsoil there is limited. In contrast, the hollows are often waterlogged, particularly in spring and in years with heavy precipitation, which also severely limits ploughing. The farm manager reports an annual fluctuation in winter wheat yields, between 0.9–5.8 t ha - 1 (Ø 3.5 t ha - 1 ). Therefore, to attain high winter wheat yields with the standard practice, the nitrogen (N) supply from LGS residuals and the N-up-take vital to winter wheat development must occur synchronously [3]. This process is closely linked to the water supply and the soil’s microbial activity, which could be strongly influenced by climate changes projected for Brandenburg [4]. Such climatic developments include frequent mild and wet winters, increasing spring and summer droughts, recurrent extreme weather events (heavy rainfall and drought), and a rise in the average annual temperature [5]. Figure 1 depicts projected climatic changes from 2062 to 2092, illustrating their effects on the Wilmersdorf organic farm.

In view of these projected climate changes, the current standard cultivation practice reveals several weaknesses:

LGS virgin tilling in mid-October, followed by a mild and rainy winter, may lead to N-mineralisation, nitrate leaching and erosion, as the low development of winter wheat mass is unable to take up sufficient nitrogen at this time [6]. On the other hand, increasing drought in early spring reduces microbial nitrogen release [7]. This can result in an N-deficiency during developmental phases, when winter wheat has high nitrogen requirements (Stem elongation; Zadoks scale 30–32) [8]. Furthermore, ploughing in autumn heightens erosion susceptibility and the risk of plough sole compression [9]. As a result, root depth and infiltration can be retarded, hindering the soil’s buffering capacity in extreme weather events (heavy rain and drought).

These issues are compounded by the influence dry periods and increased precipitation in late autumn and winter has on the number of suitable fieldwork days for ploughing and sowing. In the future, winter crop sowing conditions could deteriorate in direct proportion to the instability of appropriate fieldwork days [10].

Since LGS processes great quantities of water via transpiration, increasing temperatures can potentially lead to a water shortage in the soil, inhibiting subsequent crop growth [9]. Despite uncertainties in climate predictions, the above listed weaknesses illustrate how strongly climatic changes may influence growth, site and weather conditions, and so of course crop yields. The question is, which strategies and measures can minimise yield risk in future winter wheat cultivation on the Wilmersdorf organic farm?

The extent of climate risks can be offset by increasing diversification within an agricultural ecosystem. This means not only introducing a greater diversity in crops and varieties, but also more flexibility in cultivation and tillage methods [12, 13]. Such alternative tilling and/or seeding methods help farmers more effectively adapt to changing weather conditions, extending the number of available fieldwork days. To date, the Wilmersdorf organic farm has used only one method for producing winter wheat.

This work is therefore dedicated to the following questions: In addition to the standard practice, which winter wheat cultivation methods are most effective in offsetting climate change influences? Which methods contribute most to diversification on a specific farm?

In collaboration, the Wilmersdorf farm manager and the INKA BB field trial project a) developed new cultivation techniques and, b) tested the viability of these methods as alternatives to standard procedures. When evaluating various courses of action, i.e. cultivation methods, for their adaptability to uncertain climate change conditions, robustness is an important criterion [14]. Robustness is a system’s immunity to a wide range of influences [15]. In this case, the robustness of a new cultivation method is measured by its yield stability over a three-year period under varying field conditions. Hallmarks of these new cultivation methods are reduced tillage and alternative crop rotation. They should be able to improve the N and water supply to winter wheat. These alternative cultivation methods, (Figure 2), and the reduced tillage agricultural tool, the ring-cutter, will be described in detail.

A new agricultural tool, the ring-cutter, was used for the first time during the field tests in Brandenburg. This agricultural instrument has cutting rings running diagonally to the driving direction, allowing for an overall, non-turning, shallow tillage (see:http://www.heko-landmaschinen.de). The special ring-cutter construction is intended for soil processing when dryness or sogginess renders ploughing unsuitable. As opposed to the plough, the ring-cutter could be applied to both the dry hilltops and damp hollows on Wilmersdorf farm. The tool should also enable unploughed LGS processing, so that grain mulch seeding (early sowing and oat/ring-cutter test methods) or catch crops can be seeded directly following ring-cutter processing. If ring-cutter processing proves successful, summer crops and early sowing can stay on schedule, despite wet or dry soil. At the same time, shallow processing can reduce the water loss caused by evaporation in ploughed soil. Reduced tillage with the ring-cutter should also retard N-mineralisation, reducing the risk of N-leaching in winter (see test method oat/ring-cutter) [16].

The stockless organic farm Wilmersdorf (Bioland growers’ association member) is located in the north German lowlands in the northeast of Brandenburg (Uckermark; community Angermünde, district of Wilmersdorf: 53.11431 ∘ N; 13.90660 ∘ E). The farm has over 1,108 hectares of arable land and 5 hectares of grassland. Almost half of the arable land is used to grow legumes. Forage legume cultivation accounts for the main area (surface area 29%), followed by grain legume cultivation (19%). The remaining half of the arable land is reserved for growing grains, winter rye (28%) and winter wheat (20%), spelt (16%) and spring wheat (15%), as well as winter and spring barley (12% and 9%, respectively). Wilmersdorf has predominately sandy loamy soil and implements the following crop rotation: Alfalfa clover-grass / winter wheat / winter rye / field peas / spelt under sown with alfalfa clover-grass.

Wilmersdorf, located in the upper moraine region, is dominated by a hilly to undulating landscape of hills and hollows (40 to 75 meters above sea level) [32]. Sub-glacial till, lime-free loam or loamy sand prevail [32]. The topsoil is characterised by soil erosion and a low decalcification depth. Predominant is calcite luvisol/gley luvisol (brown soil family), known for both its fertility and its tendency to compress the topsoil foundation [32]. Typical, indigenous catena sequences are, from the hilltop to hollow, calcarite regosol > eroded calcite luvisol > gley luvisol > eutric gley [33]. In dry years, exposed hilltops are prone to yield reductions, while in wet years the loss occurs on the lower slopes and in the hollows [32]. At the same time, the eroded peaks exhibit higher pH values than the hollow depressions (Table 1). Sandy loam was the predominant soil type in all sub-plots of the pilot facility. Considering a main root zone of 100 mm, sandy loam has a field capacity of 110 mm. In contrast to the hollows, the hilltops are carbonaceous, and average a higher proportion of clay and higher pH values (Table 1). The carbonate level indicates the initial till substrate which topsoil erosion has brought close to the surface. Soil samples found compressed topsoil foundation (soil depth 20–25 cm) on the shallow calcarite regosol hilltops (Ah/C soil). In the samples taken in 2011, the above-mentioned soil conditions were particularly pronounced. Ploughing was limited on the hilltops due to the shallow topsoil and the initial substrate’s proximity to the surface.

The Wilmersdorf climate is characterised by low annual rainfall and frequent dry periods in early summer and autumn. With 517 mm annual rainfall and an average annual temperature of 8.8 ∘ C, it is one of the driest regions in Germany [32]. The three experiment years (2010–2013) were partially influenced by extreme weather patterns and high annual precipitation that exceeded the long-term average in all three years (Figure 3).

In 2010, high precipitation in summer (August: 117 mm; annual rainfall: 640 mm), made most fields impassable during both harvest and autumn sowing. 2011 was marked by a very warm and dry April and heavy rains in July (203 mm). Annual rainfall was 693 mm. Winter 2012 brought black frost and a succession of frost/thaw cycles, leading to considerable winterkill. The early sowing test method plots were so badly damaged by winterkill that they had to be re-established with spring wheat. After a warm and a partially dry spring, above average precipitation fell in July (annual rainfall 626 mm). A prolonged and snowy 2012/2013 winter was followed by a pre-summer drought in April, then a dry and warm July and August. With an annual rainfall of 583 mm, 2013 was the driest project year.

The field trial was designed as a three-year serial experiment. Complying with crop rotation, the experiments were carried out on a different field each year. Two homogeneous fields (one hilltop, one hollow) were selected on each of these fields by interfacing digital plot, soil and yield maps. The trial cultivation systems were set upon the two sites in fully randomised blocks. Each block contained four plots divided into five sub-plots (20 sub-plots per site/40 sub-plots total). Except for reaping, the 15 m long by 5 m wide plots were worked with the customary combine harvester. The individual steps, deadlines and applied techniques are listed in Table 2. The selected ring-cutter tool had a three-meter working width, six cutting rings and a leaf spring roller for reconsolidating the soil. After the winter wheat harvest, the trial plots were completely ploughed and winter rye was cultivated.

The yields of oat, winter wheat, and subsequent winter rye (2012 and 2013) were assessed by documenting the number of ears per square meter, the thousand kernel weight and the grain yield. The crude protein content in winter wheat, an indication of grain quality, was measured as well as the oat hectolitre weight. The development and structure of crops were determined by visually appraising cultivated plants, weeds and catch crop mixture coverage [34]. Weed appraisal focused on pre-eminent weeds displaying a high degree of both consistency and coverage.

Biomass growth was compiled by taking manual cuts (0.5 m 2 ) from LGS, the oat crop and the summer catch crop. Depending on the test method (Figure 2), two manual LGS cuts were taken from plots of the early sowing and the catch crop method, and three cuts were taken from standard practice plots. All LGS cuts were taken independently, either before the usual mulching dates or prior to using the ring-cutter for the first time, respectively (Table 2). The LGS cuts from oat fields (Zadok scale 69) and from the summer catch crop were taken after buckwheat bloom [8]. To determine the respective yield share, the LGS and oats samples were sorted by hand. The N-input of the second harvest year was determined by calculating LGS biomass yield, leguminosae and N-content. With the N-balance calculator the N-input count was taken from N 2 -fusion minus gaseous N-losses that arise from mulching [35]. Soil samples to determine Nmin- content were extracted as weather conditions allowed (Pürkhauer drills, soil layer 0–60 cm), i.e. in autumn after winter wheat seeding, at the beginning of vegetation in spring and after the winter wheat harvest. For the extraction of the soil mineral nitrogen, NH 4 -N and NO 3 -N 2M KCl solution was used. In 2010, due to drought, one shallower sample (0–30 cm soil layer) was taken after the oat harvest in late summer. The statistical variance analysis was carried out with MIXED procedure (SAS 9.1 statistical software) and subsequently compared to the mean values (Tukey HSD test).

In accordance with test method design, the same LGS seed mix was sown on all plots over all three years. Over all three years, no substantial differences in the quantity of biomass growth on hilltops and hollows could be discovered (Table 3). However, there was a great difference in growth quality regarding which species grew on which site. On the hilltops, LGS growth regularly exhibited a higher percentage of alfalfa; in the hollows, more grasses, herbs, white clover and weeds (Table 4).

In 2010 and 2011 until mid-October, catch crop test plots achieved growth between 2.5 and 3.9 t DM ha - 1 , (N up-take from 72 to 100 kg N ha - 1 ). In contrast, in 2012, low rainfall in August and unevenly distributed precipitation in September, caused summer catch crop plots to fail (Table 3).

In 2010 and 2012, oat yields were 21% and 54% higher in the oat/plough test method than in the oat/ring-cutter test method, respectively (Table 5). In each trial year, the yield difference between the two test methods was more pronounced in the hollows than on the peaks. Only on the hilltops in 2011, where ploughing was limited, (see Section 2.2), did the unploughed oat ring-cutter method achieve a 17% higher oat grain yield.

Among test methods, the winter wheat grain harvest varied considerably. This variance was most pronounced in the years with extreme weather conditions (2011 and 2012; see Section 2.3.).

In 2011 and 2012, winter wheat on the oat/plough and oat-ring-cutter test plots could not be harvested due to frequent heavy precipitation and a rampant weed infestation (Tripleurospermum perforatum). In all years and on all test plots, when harvesting was possible, winter wheat grown with the oat/ring-cutter test method produced the lowest grain yield of all test methods—significantly lower on hilltops in 2012 and 2013; and in hollows in 2013 (Table 6). In addition, the hilltop winter wheat in oat/ring-cutter test methods exhibited decidedly lower crude protein in 2011 and 2012. There was no great difference in grain yields between the standard procedure and early sowing and catch crop test methods. Compared to standard practices, however, the early sowing method had consistently lower grain yields. In 2011, the crude protein content in early sowing was markedly lower than in standard and catch crop methods. In contrast, in the dry year, 2013, early sowing exhibited a much higher thousand kernel weight than standard and catch crop methods. The catch crop method came closest to matching the standard practice in yield and quality. With the exception of catch crop in 2013, none of the alternative test methods were able to achieve a higher yield than the standard practice.

In 2012 and 2013, winter rye was a subsequent crop to winter wheat, bringing in high grain yields, especially in the hollows. With the exception of the thousand kernel weight in the hollows in 2013 there was no discernible difference between the test methods (Table 7).

The N-dynamic in cultivation methods was estimated based on the LGS N-input, N-up-take in the grains and N m ⁢ i ⁢ n -contents in the soil. Interdependent on tilling date, biomass yield and the leguminous crop, the LGS provided an N-input from 145 to 352 kg N ha - 1 in the second harvest year (Table 8). Oats attained the highest nitrogen up-take after LGS tillage and ploughing, taking up more than 100 kg N ha - 1 in the hollows in 2010 and 2011.

In late summer 2010 (Table 9), N m ⁢ i ⁢ n sampling from the early sowing and catch crop test plots at the hollow exhibited significantly higher N m ⁢ i ⁢ n values than the other test methods. Prior to sampling, the LGS plots of both test methods were killed in July via root separation with the ring-cutter (see Table 2).

In the autumn of 2011 and 2012, oat/plough and partly oat/ring-cutter wheat plots exhibited significantly higher N m ⁢ i ⁢ n -values than those of the standard practice (Ø +45 kg N ha - 1 ; Table 10). Oat plough had the highest N m ⁢ i ⁢ n -values of all test methods. In spring, the highest N m ⁢ i ⁢ n -values were always observed on standard practice and catch crop wheat plots, which had been ploughed at a later date. Compared to the standard practice, oat/ring-cutter and oat/plough methods had significantly lower N m ⁢ i ⁢ n -values in the spring (Hollow 2011–2013). There were no remarkable differences in N m ⁢ i ⁢ n -values in samples taken from beneath the subsequent winter rye crop (08.11.2011, 22.10.2012), or taken immediately after the winter wheat harvest (08.02.2013; Table 11). The five highest N m ⁢ i ⁢ n -values of all sampling dates were found in the subsequent winter rye crop (max. value, 148 kg Nha - 1 ; oat/ring-cutter test method).

Weed distribution varied on the different sites and test method plots. In all experimental years, the weed population in oats and winter wheat was more pronounced in the hollows than on the hilltops.

On average, the standard practice and catch crop test method (autumn ploughing and late seeding) had the lowest weed dockage, the weed coverage ratio (WCR) in both methods being nearly identical. In comparison, the early sowing WCR was from distinctly to significantly higher in each year. On the oat/ring-cutter plots there was usually a higher WCR (Table 12) and/or a higher weed biomass (Table 13) in established oat and winter wheat cultures than on the oat/plough (Op) test plots. In addition to these quantitative differences, specific differences were found in weed group composition on the varying sites and in the varying test methods.

The prior LGS crops generated weed flora—alfalfa (Medicago sativa L.), red and white clover (Trifolium repens L. and Trifolium pratense L.) as well as perennial ryegrass (Lolium perenne L.) and couch grass (Agropyron repens L.)—on all locations and test methods. Comparing standard practices and ploughed test methods (catch crop, oat/plough) to the unploughed test methods (early sowing and oat/ring-cutter) (Table 13), alfalfa did not regrow in ploughed methods, neither in the preceding oat crop nor or in winter wheat. In 2011, however, alfalfa did regrow on standard practices and catch crop hilltop plots because the plough could not optimally turn the soil in autumn. Besides alfalfa, the weed population was dominated by, in descending order, creeping thistle (Cirsium arvense L.), scentless chamomile (Tripleurospermum perforatum), poppy (Papaver rhoeas L.), common chickweed (Stellaria media L.), larkspur (Consolida regalis Gray) and veronica (Veronica chamaedrys L.). These species varied within sites and procedures. In the hollows and in the unploughed test methods, creeping thistle was much more dominant (Table 14), particularly in the oat/ring-cutter winter wheat. In the hollows and on one hilltop, scentless chamomile dominated, mainly in the early sowing, oat/ring-cutter and oat/plough winter wheat test methods (cover ratio, 75–90%). In 2011 and 2012, a rampant odourless chamomile infestation caused winter wheat crop failure in oat/plough and oat/ring-cutter hollows.

The hypothesis that the ring-cutter can effectively kill and regulate LGS in both wet (spring) and dry (summer) soils was confirmed—LGS secondary growth coverage on unploughed oat and wheat plots was the same height as on ploughed plots. The tool’s cutting technique - a shallow, vertical undercutting of the upper topsoil - separates the alfalfa sprout shaft below the root crowns, effectively impeding regrowth. In contrast to the plough, ring-cutter tractive power requirements are low and operating speed (11 km h - 1 ) is remarkably fast. Nonetheless, the tool required two to four working cycles to prepare the plots for mulch seeding after the two-year LGS cultivation. Further drawbacks to the ring-cutter were much lower grain yields in two of the three years on oat/ring-cutter and winter wheat plots and less N-availability than oat/plough plots. Ring-cutter implementation only reached higher oat yields than oat/plough on the shallow hilltop in the 2011 test method. Numerous other studies substantiate that reduced tillage in organic farming results in lower yields and lower N-availability [36, 37, 38, 39, 19]. Reduced tillage limits N-availability by delaying soil organic matter (SOM) mineralisation, due to minimal aeration [19] and poor soil warming in the spring, particularly when densely mulched [40]. Especially in spring, unprocessed earth is often cooler and soggier than ploughed soil, hindering both germination and the initial growth of summer grains [41]. Reduced soil processing can also lead to increased compression in untreated layers [39, 18] and to accumulated SOM in the upper soil layer [42], as confirmed by ring-cutter processing research conducted at the ZALF Müncheberg experimental station where much higher soil compression limited root penetration of winter grains [43]. Limited N, compressed soil and thwarted root growth impede a plant’s ability to absorb nutrients, which may have caused the much lower oat and winter wheat harvests [37]. In addition to these factors, the ring-cutter plots displayed dense weed growth (Tables 12 and 13); typical for reduced tillage in organic farming [44, 19]. Shallow ring-cutter processing causes weed seeds to remain close to the surface, providing excellent germination conditions. Perennial species such as couch grass and creeping thistle thrive under reduced tillage [45].

Bearing all this in mind, it becomes clear that the results of reduced tillage are highly dependent on the time of ring-cutter processing, on crop rotation selection and on weather conditions, as confirmed by the completely unploughed oat/ring-cutter and early sowing test methods [18]. While oat/ring-cutter did not generate stable oats or winter wheat grain yields, the early sowing wheat yields of two trial years did not deviate greatly from standard procedure yields. Hence, reduced soil processing can be successful when applied to pertinent sites and adapted to the entire cultivation system, i.e. alterations in crop rotation, seeding and the catch crop [19].

In the test methods early sowing and catch crop, LGS processing takes place in the summer. The ring-cutter was used to sever LGS roots by shallowly working the topsoil, even in dehydrated soil (i.e. in June 2010). August 2010 N m ⁢ i ⁢ n -samples proved that early ring-cutter processing releases significant N-quantities in late summer (up to 109 kg N ha - 1 ) for early sowing and summer catch crops (see Table 9). Heavy rains in August and November 2010 and in July 2011 caused above average precipitation, possibly leading to nitrate leaching and loss in late summer and autumn, especially in 2010, in these two test methods. Further indication of such losses is the significantly lower N m ⁢ i ⁢ n -value in early sowing and partly in catch crop in spring 2011, when compared to standard practices.

On the other hand, in 2010 and 2011, catch crop plots developed vigorous cover crops that were able to absorb a portion of the mineralised nitrogen by October (up to 86 kg N ha - 1 ). Other studies came to similar results with an early catch crop after clover-grass tilling [6]. Compared to standard procedures, the summer catch crop organic material provided a slightly improved N-supply for winter wheat in spring. This is supported by increased N m ⁢ i ⁢ n -values on catch crop plots in spring 2012. Despite the possible catch crop nitrate transfers, yields were comparable to standard yields, confirming the hypothesis that the catch crop winter wheat test method ploughed in autumn produces equivalent annual grain yields, on hilltops as well as in hollows, where the other test methods reported yield losses. As to location and weather conditions, the catch crop test method proved the hardiest.

Although the early sowing test method yields did not differ greatly from standard practices, compared to catch crop, early sowing yield was characterised by low stability and robustness. In 2012, early sown winter wheat was particularly susceptible to winterkill and in 2011, contained a much lower crude protein. The early sowing test method also had a high weed content despite the catch crop mixture, which is typical in early sown crops [46]. In 2013, with an early drought in April and dry summer months, early sowing came closest to matching standard practice yields. Possibly the more abundant pre-winter growth, followed by a cold and snowy winter in 2012/2013, allowed early sowing winter wheat to exploit the winter moisture and take up nitrogen early, as confirmed by much higher N m ⁢ i ⁢ n -values on early sowing plots the previous autumn and decidedly lower ones in spring 2013. Due to the developmental advantage of early sowing, winter wheat was perhaps less affected by water shortages in April (stem elongation) and drought in July during the grain-filling phase [25, 26]; an idea supported by the significantly higher thousand kernel weight in this year (Table 6). Overall, the early sowing test method shows how heavily grain yields depend on cultivation procedure and weather conditions throughout the year, and that unploughed early sowing is not necessarily synonymous with minimal yield. The only early sowing location effect was weed pressure. This was less pronounced on hilltops and therefore advocates unploughed processing. Based on the grain yield, the early sowing test method is not particularly suitable for dry knoll locations, as was otherwise proposed in hypothesis 3.

Studies suggesting oat as a good LGS exploiter are confirmed by their results [28]. Especially in the oat/plough test method and in the damp hollows, oats were able to take up high levels of mineralised LGS nitrogen in spring (up to 105 kg N ha - 1 ). With a high N-supply, oat/plough strongly suppressed weed growth [29].

Oat/plough and oat/ring-cutter test methods showed high N m ⁢ i ⁢ n -values (up to 115.6 kg N ha - 1 on oat/plough; 98.7 kg N ha - 1 on oat/ring-cutter) at the beginning of November 2011 and the end of October 2012 despite high N-deprivation and an early winter wheat sowing date in 2011. In contrast to standard practice and catch crop, oat/plough and oat/ring-cutter exhibited lower N m ⁢ i ⁢ n -content in the spring than in autumn, which points to nitrate leaching into deeper soil layers after rainy and cold winter months. In both winters, 2011/2012 (205 mm) and 2012/2013 (149 mm), rainfall from November to February was above field capacity (110 mm; see Section 2.2.).

Despite summer crop cultivation after springtime LGS tilling, there may be an increased risk of nitrate leaching the following winter [27]. This risk can be offset by establishing a second subsequent crop in early autumn, ensuring a high and rapid N-up-take. Winter wheat N-up-take (10–30 kg N ha - 1 ) in autumn is less pronounced than other grains (for example, winter rye with 30–50 kg N ha - 1 ) [47]. Consequently, winter wheat sown on oat/plough and oat/ring-cutter plots at the end of September (2010 and 2011) and in mid-October 2012 only minimally prevented nitrate leaching.

In the spring, oat/plough and oat/ring-cutter winter wheat plots were unable to compete against odourless chamomile. This multiyear species has a fibrous and deep-reaching root system, asserting its dominance especially on wet and compacted terrain [48]. Without weed control, odourless chamomile can overrun a winter wheat grain yield by up to 90% [48]. In the wet years 2011 and 2012, this was irrefutably confirmed in the hollows, locations particularly vulnerable to crop failure [32]. In summary, despite spring tillage and an oat insertion, the nitrogen supply for winter wheat was not improved. Also, the oat/plough and oat/ring-cutter test methods were highly susceptible to weeds. Only on the hilltops and in the dry year 2013 was oat/plough winter wheat yield comparable to standard practices, contrary to hypothesis 2. In each trial year and compared to standard practice, the oat/ring-cutter winter wheat hilltop harvests produced a smaller deficit margin than harvests in the hollows did. According to hypothesis 3, the oat method is more suitable for hilltops. However, it is not a viable alternative to standard cultivation practices as its yield levels are much lower.