Wheat is one of the most important cereals worldwide, yet significant gaps remain in our understanding of genetic variability in root traits, especially those associated with deeper rooting that support resource acquisition in challenging environments. Root traits are typically controlled by many genes with small effects and often display low heritability. Our aim was to develop a statistical approach to analyse root variation across soil depth and to determine where genetic differences in root intensity are most detectable. An experiment was conducted at the RadiMax semi-field facility, which is designed to measure deep root systems. Five years of phenotypic data recorded each June produced observations from 1500 rows. Each row captured root intensity across the soil profile from 0.6 m to 2.6 m, enabling detailed analysis of vertical root distribution. Across the five years, 513 winter wheat cultivars were grown in the facility, and among those 409 were genotyped with SNP chips. Depth-resolved regression models with random coefficients were used to quantify genetic and non-genetic variation in root intensity across soil depths, while accounting for spatial variation between rows. Random variation within rows was found to be constant across depths. The models showed that genetic variance for cumulative root intensity increased substantially below 1.1 m, with the deepest layers exhibiting the largest differences between wheat lines. Narrow-sense heritability of point measurements peaked at approximately 1.5 m ( h^2=0.13 ).

Deep-rooted crops accessing water and nutrients from deep soil layers enhance the resource base for crop production. However, studying these roots in field conditions is labour-intensive, limiting research scope. We established a field root research facility with 48 plots for replicated experiments. The facility includes 144 6-metre-long minirhizotron tubes and an AI-based pipeline for rapid root trait analysis. We also attempted to install access tubes and customized ingrowth core production for less-invasive root activity determination. Our study revealed significant differences in deep root density among species, particularly at depths of 2.5 to 4.5 m, over 5 years. The less-invasive studies using ingrowth cores reached depths of 4.2 m. Nutrient tracer 15N analysis showed marked differences in deep root activity among crop species. Time domain reflectometry sensors indicated varying water depletion in deeper soil layers, influenced by crop species and root growth patterns. We established a field facility for studying deep root growth and function, demonstrating its effectiveness in analysing diverse deep-rooted plant species. This facility provides an ideal platform for conducting meaningful research in deep soil layers, yielding statistically and biologically significant results for agricultural applications.

Root segmentation is a fundamental yet challenging task in image-based plant phenotyping. We present the first systematic comparison of Transformer and Convolutional Neural Network (ConvNet) architectures for root segmentation, evaluating 21 architectures across nine diverse datasets and comparing pre-trained models to training from scratch. Transformer-based models significantly outperform ConvNets for segmentation accuracy and root-diameter agreement. Pre-training significantly improves mean Dice from 0.623 to 0.666 ( p = 3.3 × 10 −10 ). We also find that Transformers benefit more from pre-training than ConvNets, with Dice improvements of +0.072 versus +0.022 ( p = 3.7 × 10 −4 ), supporting the hypothesis that fine-tuned Transformers transfer more effectively across large domain gaps. Among evaluated models, MobileSAM achieved the highest Dice score while maintaining computational efficiency. Dataset choice explained far more performance variance (70.9%) than model architecture (6.7%), suggesting that data curation matters more than model selection.

Abstract Background Root segmentation is a fundamental yet challenging task in image-based plant phenotyping. Accurate segmentation is a prerequisite for extracting root traits relevant to plant physiology, breeding, and agronomy. While U-Net and other convolutional neural network (ConvNet) architectures have been applied to root segmentation, no systematic comparison of multiple Transformer and ConvNet architectures has been conducted across diverse root imaging conditions. Results We evaluated 21 segmentation architectures across nine diverse root image datasets, training 1511 models to assess all combinations of architecture, dataset, pre-training strategy, and learning rate, producing over 3 million segmentations for evaluation. Transformer-based models significantly outperformed ConvNets for Dice (mean Dice 0.679 vs 0.659; $$p = 3.0 \times 10^{-3}$$ ). Root-diameter and root-length correlation were also higher for Transformers, but the differences were not statistically significant ( $$p = 0.054$$ and $$p = 0.198$$ respectively). Pre-training significantly improved mean Dice from 0.623 to 0.666 ( $$p = 6.6 \times 10^{-10}$$ ), with Transformers benefiting more from pre-training than ConvNets (Dice improvement + 0.072 vs + 0.021; $$p = 3.7 \times 10^{-4}$$ ), supporting the hypothesis that fine-tuned Transformers transfer more effectively across large domain gaps. MobileSAM achieved the highest Dice score (0.693) while maintaining computational efficiency. Both architecture families underestimated thin root length compared to manual annotations. Dataset choice explained 70.9% of performance variance, far exceeding model architecture (6.7%). Purpose Transformer architectures significantly outperform ConvNets for root segmentation accuracy, and pre-training significantly improves performance, particularly for Transformers. Pre-trained MobileSAM offers the best accuracy at competitive computational cost. Dataset choice dominates performance variance, suggesting practitioners should prioritize data curation over architecture selection.

Topographic depressions within agricultural fields contribute disproportionately to regional nitrous oxide (N2O) emissions. These low-lying areas accumulate nutrients and fine particles through water inflow and erosion, creating conditions conducive to elevated N2O emissions. Because depressions remain part of fields, understanding how crop presence and management influence their N2O source strength is essential. A greenhouse mesocosm experiment was conducted using soil collected from an agricultural depression under either drained or partially waterlogged conditions (water table maintained 10 cm below the soil surface). Wheat was sown at three dates (57, 43, and 29 days before waterlogging) to represent different vegetative stages and capacities for N uptake. Dissolved organic carbon (DOC), dissolved nitrogen (N), root growth, and N2O emissions were continuously monitored. A 15N-labelled fertilizer was applied during the waterlogging period to trace fertilizer-derived N in emitted N2O, soil, and plant biomass. Early-sown wheat was more strongly impaired by waterlogging than later-sown treatments but nevertheless significantly reduced N2O emissions compared with the unplanted control, irrespective of water regime. Early-sown plants also acted as stronger N sink, indicating that greater plant N acquisition contributed to reduced N availability for N2O production. In contrast, trends in DOC and cumulative N2O emissions across seeding dates were less consistent, particularly under waterlogged conditions. Plant N uptake emerged as the primary mechanism reducing N2O emissions under the tested conditions. Although a general trend of lower N2O emissions with increasing plant age and N uptake was observed, the relationship was not strictly linear, due to plant age-specific interactions with waterlogging. Nevertheless, the results suggest that establishing crops in depression areas can mitigate N2O emissions by strengthening plant competition for available N, while the magnitude of this mitigation depends on both seeding date and waterlogging conditions.

Background Dual-purpose wheat systems-where crops are grazed during vegetative stages and later harvested for grain-offer flexibility in mixed farming, particularly under variable rainfall. However, the effects of grazing-induced defoliation on root development, plant water status, stress response and final grain yield have been underexplored, especially under multi-season field conditions. We hypothesized that grazing-induced defoliation would temporarily suppress root growth but conserve subsoil water through reduced transpiration, potentially alleviating terminal drought stress to improve yield in dry seasons, although outcomes would be season-dependent. Objective The study aimed to quantify the effects of grazing (defoliation) on root growth dynamics, soil water availability, canopy thermal stress responses, and grain yield of early-sown winter wheat across variable field seasons Methods Field experiments were conducted over three growing seasons (2021-2023) in southern Australia using paired grazed and un-grazed treatments in early-sown winter wheat. Root depth progression and root length density were monitored throughout the season, together with soil moisture profiles, canopy temperature, and yield components. Seasonal conditions ranged from relatively wet (2021-2022) to dry (2023), enabling assessment across contrasting water availability scenarios. Results Grazing consistently delayed root descent by similar to 200 degrees C days but root depth had generally recovered by anthesis, with only transient reduction in root length density (0.4-0.8 m). In the dry 2023 season, grazed crops conserved subsoil water, lowered canopy temperatures, and reduced grain-filling stress, while effects were minimal in wetter years. Despite more rapid regrowth during the critical period, grazed crops had consistently lower yield (-0.6 t ha(-)& sup1;), likely due to more biomass allocation to leaf and stem and delayed phenology. Conclusion Grazing caused significant but transient changes in root growth and water use which had no effect on grain yield, which was primarily affected by post-grazing shoot regrowth dynamics. Grazing related stress reduction during spring drought did not translate into yield gains.

Background and aimsEastern Denmark’s agricultural landscapes feature numerous topographic depressions that are frequently flooded during late winter and spring. These poorly drained, carbon- and nitrogen-rich depression soils receive eroded material from adjacent slopes. Fertilization and water saturation create N2O emission hotspots. However, the potential legacy effects of these topographic locations on microbial communities involved in N2O production and reduction remain unclear. One approach to mitigating high denitrification rates (as a source of N2O) is to alter microbial pathways by adding nonhazardous levels of copper.MethodsWe conducted an incubation study using upland and depression soils from the same site, incorporating varying Cu levels (0, 130, and 260 mM) and water levels (60% and 90% water holding capacity).ResultsDepression soils emitted eight times more N2O than upland soils at 90% WHC. Cu addition did not reduce cumulative N2O emissions but delayed or lowered the flux peak. Depression soils exhibited 3,000- and 4,000-fold higher 16S rRNA and nosZ clade I abundances, respectively, compared to upland soils. Cu addition significantly decreased 16S rRNA abundance, eliminated AOB amoA in upland soils, and slightly reduced the tested gene abundances in depression soils. The nosZ gene community structure differed significantly between the two soils.ConclusionsOverall, our study suggests that erosional differentiation of soil properties, together with frequent waterlogging conditions, can result in distinct microbial communities, fostering legacy effects that lead to differences in N2O emissions between upland and depression soils. Adding Cu to these intensively managed soils is unlikely to be an effective strategy for mitigating N2O emission hotspots in arable fields.

Deep-rooted crops accessing water and nutrients from deep soil layers enhance the resource base for crop production. However, studying these roots in field conditions is labor-intensive, limiting research scope. We established a field root research facility with 48 plots for replicated experiments. The facility includes 144 six-meter-long minirhizotron tubes and an AI-based pipeline for rapid root trait analysis. We also attempted to install access-tubes and customized ingrowth-core production for less-invasive root activity determination. Our study revealed significant differences in deep root density among species, particularly at depths of 2.5 to 4.5 meters, over five years. The less invasive studies using ingrowth-cores reached depths of 4.2 meters. Nutrient tracer 15N analysis showed marked differences in deep root activity among crop species. TDR sensors indicated varying water depletion in deeper soil layers, influenced by crop species and root growth patterns. We established a field facility for studying deep root growth and function, demonstrating its effectiveness in analyzing diverse deep-rooted plant species. This facility provides an ideal platform for conducting meaningful research in deep soil layers, yielding statistically and biologically significant results for agricultural applications. ### Competing Interest Statement The authors have declared no competing interest.

BACKGROUND AND AIMS:There is growing interest in the production of ancient grains including emmer, einkorn and spelt, particularly in low-input systems. Differences in their root systems and how these affect water and nitrogen uptake are not well known, but can offer important insights into the effects of plant breeding on resource use and root physiology, which can inform breeding of future crops. METHODS:In this study, we used imaging in minirhizotron tubes to evaluate root development in emmer, einkorn, spelt and modern wheat growing under field conditions, taking images to 2.2 m soil depth. We evaluated water stress in the different species using carbon isotope discrimination and used a nitrogen tracer to compare N uptake over time. KEY RESULTS:The results show that modern wheats have deeper and more efficient root systems. Modern wheats showed less water stress in late developmental stages due to their deeper roots which allow access to deep soil water, and can therefore sustain high grain yields. They were also able to translocate N more efficiently to the grain. The results contradict previous hypotheses that modern wheat has shallow rooting systems due to high inputs, showing that where more nutrient resources are available, deeper roots have become important for water uptake to support higher yields. CONCLUSIONS:This is the first field study of roots of ancient and modern wheats, where we clearly see that there are substantial differences between the root systems. These results help to explain how past selection for yield has affected below-ground crop physiology.

BACKGROUND AND AIMS:Deep roots may help plants adapt to climate change by allowing them to access deeper soil layers where water is still available, reducing water stress and increasing nitrogen (N) uptake. Water stress significantly affects yield during later developmental stages, but methods are lacking for phenotyping for deep rooting under field conditions and at maturity. METHODS:Over 3 years, we used minirhizotron root imaging in the RadiMax semi-field facility to compare deep rooting in winter wheat genotypes grown in field soil to 2.7 m depth. We related this to deep soil uptake of water and N using isotopic tracers injected into the soil at 1.6-1.8 m depth. Carbon isotope discrimination was used to evaluate water stress levels. KEY RESULTS:Deep rooting was positively correlated with uptake of deep-placed N and water, and uptake of deep-placed N was three times higher in the genotype with deepest roots compared with the shallowest. Deep rooting was negatively correlated with water stress, measured using carbon isotope discrimination. This correlation was strongest in 2023, a dry year, highlighting the role of deep roots in mitigating water stress. Some genotypes had consistently deeper or shallower roots over the three experimental years, and there were strong correlations of isotopic measurements between genotypes across years. CONCLUSIONS:Our findings show strong relationships between deep rooting and deep root functions, which indicate that deep rooting is a desirable trait that should be targeted. The significant genotypic variation observed, which can be phenotyped for even under field conditions, indicates that deep rooting is a trait that can be incorporated into breeding programmes. Furthermore, the methods used in this study are effective and should be developed for further application.

Context or problem: Roots have been neglected in crop research, and in particular deep roots which are more difficult to access. Yet they play a crucial role in water stress tolerance, especially in later developmental stages. Methods for phenotyping roots are needed in order to breed for deeper rooting. While field phenotyping methods are costly and laborious, smaller scale methods are often cheaper and more easily replicated, but do not necessarily represent field conditions. Existing studies have not found strong relationships between small-scale and field grown roots, especially in later developmental stages. Objective or research question: This study aimed to investigate whether similar genotypic differences can be seen in deep rooting of winter wheat in field soil and in tube studies, and if tubes could therefore be used to predict deep rooting in the field. Methods: We used root imaging to compare deep rooting characteristics of eight modern Danish winter wheat cultivars using three different methods: field experiments assessing roots with minirhizotron tubes; the semi-field facility, RadiMax; and 1.5 m tall rhizotron tubes. Results: While deep rooting genotypes showed mostly positive correlations across all methods, significant correlations between methods were observed only in one year, specifically between the tubes and semi-field. Furthermore, deep rooting exhibited significant correlations across years and months within the RadiMax method, suggesting consistent deep rooting patterns over time. The increase in variability as experiments became more field-like highlights the complexity of soil-root interactions. Conclusions: While this study suggests that under certain conditions, small-scale phenotyping methods can indicate deep rooting genotypes, the correlations were not consistent enough to be used to predict deep rooting in the field. This underscores the challenge of using small-scale experiments to extrapolate root measurements to the field. Implications: This study demonstrates the need for caution when interpreting small-scale root experiments, and underlines the need for continued developments in root research generally. Further studies are needed to improve the quality of methods, to evaluate the effects of different soil types and environmental conditions on root growth, and to relate these to field-grown roots.

Nitrous oxide emissions from agricultural land largely contribute to the greenhouse gas budget worldwide. Denmark’s glacial landscape has widespread small scale topographic depressions, typically flooded for 1-3 months per year. These depressions within agricultural land are considered as hotspots of N2O emissions, because of exposure to an increased nitrate availability and labile carbon due to fertilization and deposition of eroded soil material. Temporal waterlogging in these depression areas affects plant development, thus their ability to deplete available nitrogen in soil. Additionally, living plants provide substrates for denitrification through root exudates. However, the effect of living plants and roots on N2O emissions from glacial depressions is not very clear yet. In this study, we aimed to elucidate how waterlogging influences nitrogen uptake and dissolved organic carbon (DOC) release from plants at different root growth stages, and to quantify how this would affect N2O emissions. We conducted a fully crossed mesocosm experiment with depression soils subjected to saturated or freely-drained water conditions, three different wheat growth stages to mimic possible different root N uptake, and an unplanted control. In order to differentiate how much N2O was produced from newly-added fertilizer, we applied a 15N tracer. For monitoring root development, roots were imaged through the translucent mesocosm walls on a weekly basis. The growth stage of wheat significantly influenced the fate of mineral nitrogen and the dynamic of DOC in the soil solution, thereby affecting N2O emissions from these soil systems. The interaction between DOC and mineral nitrogen explained 53.9% of the variance in daily N2O fluxes. Therefore, these findings highlight the critical role of root development and soil water conditions in regulating N2O emissions from conditions representative for glacial depressions.

The objectives of this study were; (i) to develop and test an anatomical key using histochemical staining to identify roots to the plant species level, and (ii) to demonstrate that it could use this method to quantify the proportion of individual plant roots from different intercropping combinations, and how the fraction of the roots were affected by nitrogen supply and soil depth. Roots of oilseed rape, lucerne, red beet, wheat, barley, and millet were produced in a greenhouse, washed free of soil and bundles of roots from each of the species were embedded in agarose, cut on a microtome and stained with FSA (Fuchsin, Safranin O and Astrablue). A root identification key was developed based on root cross sections from the single species root samples. From the anatomical identification key, it was possible to identify roots of individual species in root mixtures. The anatomical key also made it possible to distinguish roots from the intercropping experiment. Root cross sections followed by histochemical staining was shown to be a useful and simple method for studying root interactions as well as the fractions of roots from two intercropping combinations due to different of nitrogen supply and soil depth in intercropping.

Breeding for potato deep roots can increase water and nitrogen uptake by potatoes and it can be an option to maintain stable yields with decreased inputs. This study investigates the relationship between potato root characteristics, water stress resistance and deep soil nitrogen uptake, accessing variations among cultivars and nitrogen fertilization levels. Thirteen potato cultivars were grown during 2018 and 2020 at a semi-field root phenotyping platform in Denmark. Root growth was monitored via minirhizotron tubes down to 1.8 m soil depth. Drought treatment started in the mid-June and deep soil nitrogen uptake was tracked via 15N isotope application at 1.3–1.4 m soil depth during tuber formation. Water stress resilience was identified using 13C natural discrimination process in plants. Tuber samples were analyzed for 15N and 13C content. While drought affected potato yield (not always significantly), it did not affect nitrogen uptake. Root length and distribution varied among varieties, with deeper roots (down to 1.30 m) observed in August. Statistical differences (p < 0.05) in root length, yield and nitrogen uptake were found among varieties. Cultivars with longer growing season exhibited larger, deeper roots and increased nitrogen uptake from deep soil. High correlation (R = 0.8) between deep roots and 15N uptake was observed for all varieties. Deeper roots are contributing to deep soil nitrogen uptake, but 13C content in tubers is not a reliable indicator of water stress resilience. Despite this, the study suggests the potential for breeding potatoes with deep roots to achieve stable yields, considering differences in water and nitrogen uptake among varieties.

Background and aims Defoliation triggers the remobilisation of root reserves to generate new leaves which can affect root growth until the shoot resumes net assimilation. However, the duration of root growth cessation and its impact on resource uptake potential is uncertain. Methods Winter wheat was established in a 4 m high outdoor rhizobox facility equipped with imaging panels, sensors, and access points for tracer-labelling. The wheat was defoliated in autumn at early tillering and roots were imaged at a high-time resolution and analyzed by deep learning segmentation. The water and nitrogen (N) uptake were measured using time-domain reflectometer (TDR) sensors and 2 H and 15 N isotopes. Results Root penetration of wheat paused for 269 °C days (20 days) following defoliation after which it resumed at a similar rate to un-defoliated plants (1.8 mm °C days −1 ). This caused a substantial decrease in root density with an associated reduction in water and N uptake at maturity, especially from deeper soil layers (>2 m). Conclusions Our results have significant implications for managing the grazing of dual-purpose crops to balance the interplay between canopy removal and the capacity of deep roots to provide water and N for yield recovery.

Abstract Background In drought periods, water use efficiency depends on the capacity of roots to extract water from deep soil. A semi-field phenotyping facility (RadiMax) was used to investigate above-ground and root traits in spring barley when grown under a water availability gradient. Above-ground traits included grain yield, grain protein concentration, grain nitrogen removal, and thousand kernel weight. Root traits were obtained through digital images measuring the root length at different depths. Two nearest-neighbor adjustments (M1 and M2) to model spatial variation were used for genetic parameter estimation and genomic prediction (GP). M1 and M2 used (co)variance structures and differed in the distance function to calculate between-neighbor correlations. M2 was the most developed adjustment, as accounted by the Euclidean distance between neighbors. Results The estimated heritabilities ( $${\widehat{h}}^{2}$$ h ^ 2 ) ranged from low to medium for root and above-ground traits. The genetic coefficient of variation ( $$GCV$$ GCV ) ranged from 3.2 to 7.0% for above-ground and 4.7 to 10.4% for root traits, indicating good breeding potential for the measured traits. The highest $$GCV$$ GCV observed for root traits revealed that significant genetic change in root development can be achieved through selection. We studied the genotype-by-water availability interaction, but no relevant interaction effects were detected. GP was assessed using leave-one-line-out (LOO) cross-validation. The predictive ability (PA) estimated as the correlation between phenotypes corrected by fixed effects and genomic estimated breeding values ranged from 0.33 to 0.49 for above-ground and 0.15 to 0.27 for root traits, and no substantial variance inflation in predicted genetic effects was observed. Significant differences in PA were observed in favor of M2. Conclusions The significant $$GCV$$ GCV and the accurate prediction of breeding values for above-ground and root traits revealed that developing genetically superior barley lines with improved root systems is possible. In addition, we found significant spatial variation in the experiment, highlighting the relevance of correctly accounting for spatial effects in statistical models. In this sense, the proposed nearest-neighbor adjustments are flexible approaches in terms of assumptions that can be useful for semi-field or field experiments.

Cover crops can contribute to climate change mitigation through enhanced sequestration of atmospheric carbon dioxide into soil organic carbon. Few studies, however, have estimated the total carbon (C) input to soil, i.e. derived both from plant material (shoot and root) and phyllo- and rhizodeposition. Selection of cover crop species should account for multiple objectives, such as C inputs to soil, nitrate leaching reduction and positive residual effects on the following main crop. However, trade-offs between these objectives may occur. The aim of this study was to investigate the performance of the cover crop species winter rye, hairy vetch and oilseed radish, and to assess the ability of mixtures to overcome potential trade-offs. A randomized split-plot field trial was conducted to compare cover crop treatments and a weeded control under high and low soil nitrogen (N) availability. Multiple-pulse labeling with C-14-CO2 was carried out to trace net cover crop-derived rhizodeposition C. Soil mineral N was measured to 1.5 m depth in autumn, as well as grain and N yield in the subsequent spring barley. Cover crop species accumulated between 1250 and 2580 kg C ha(-1), with significantly higher total C input (in shoot, root and phyllo- and rhizodeposits) for the mixtures compared with pure stands of either vetch or radish, while the results for rye were in between. The quantity of C lost via phyllo- and rhizodeposition (qClvPR) showed a significant positive correlation with root C and was highest for the mixtures and rye. The relative ClvPR ranged between 7% and 14% of total cover crop-derived C and tended to decrease under higher soil N availability. All cover crop treatments were able to decrease soil mineral N (0-1.5 m), with radish displaying the highest N leaching reduction potential. Despite substantial differences in cover crop total N uptake and C:N ratios, no significant differences were observed in the subsequent main crop grain or N yields. The mixtures showed the highest total C input and generally a higher or similar mineral N depletion potential than the average of the pure stands, suggesting that cover crop mixtures offer a realistic means for overcoming trade-offs among ecosystem functions.

The use of cover crops may play an important role in ecological intensification of farming systems through their impacts on cycling and the availability of essential nutrients. However, little is known about the effect of soil fertility on the performance of cover crops and consequent nutrient dynamics. We performed a two-year field trial to quantify the effect of two soil fertility levels (low and moderate) on the growth and nutrient content of different leguminous and non-leguminous cover crop species and their mixtures, and on their fertilizing value for the succeeding crop. Overall, the growth and shoot nutrient content of cover crops in autumn was influenced by the species, mixture choices and soil fertility level. The growth of some species was less affected by fertility level than others. Among the legumes, fertility levels had no effect on the growth of lupin. Among the non-legumes, buckwheat had the highest biomass production as well as C:N ratio at a low fertility level, and in the first year was affected least by soil fertility level. However, buckwheat and lupin grown as a single species or in mixtures did not lead to yield improvements in the subsequent crop. Generally, legumes produced more biomass than non-legumes in the mixtures at the low fertility level, whereas non-legumes produced more biomass at the moderate fertility level. Growing a mixture of vetch and radish achieved the best synergetic effect and highest shoot N content. In both years, the greatest improvements in barley yields were achieved after vetch grown as a single species and in a mixture with radish (by up to 46 %). Rye was less competitive in mixtures than radish and buckwheat, and in the second year it had a negative effect on barley yield at the low soil fertility level, presumably due to pre-emptive competition for nitrogen. Soil fertility level had a greater effect on barley yields than cover crop treatments, as yield increments at the moderate fertility level amounted to 48 % in the first year and 64 % in the second year compared with the low fertility level. These results demonstrate that the cultivation of winter-persistent legumes alone or in a mixture with oilseed radish offers a promising tool for improving the yields of subsequent crops through nitrogen input and nutrient cycling on farms where soil fertility level is low.

Crops with deeper rooting is an emerging tool for better exploitation of soil resources. However, there is a need for more in-depth understanding on how the increased rooting depth may be achieved. In this study a novel approach for obtaining deeper rooting has been proposed. Crops with assumed similar capacity for subsoil exploration: sugar beet (Beta vulgaris) and chicory (Cichorium intybus var. foliosum) were intercropped. Repeated measurements of biomass, deep root growth, and nutrient uptake were conducted to monitor plant competitive dynamics in the intercrop and sole crops. It was found that the intercrop positively affected biomass production with Land Equivalent Ratio close to or greater than 1 (0.99 - 1.14). Similarly, the strongest root growth over time was observed for the intercrop (from 98 +/- 48 to 304 +/- 28 cm depth). Moreover, the effect from the interspecific interactions in the intercrop varied over time. In the first half of the season yield advantage and the observed enhanced contribution to the uptake of N, Mg, Mn, Zn, and Na in the intercrop were driven by the sugar beet. Later in the growing season, yield advantage, deep root growth, and contribution to the uptake of S, Fe, Cu, and Al in the intercrop were driven by the chicory. This has also been confirmed by the root quantification analysis, which showed that in the end of the season intercrop consisted of 84 % and 98 % roots from the chicory at 1 and 2.5 m depth, respectively. This study concluded that intercropping two crops with similar root characteristics, sugar beet and chicory, can still lead to complementary interactions showing potential for efficient deep soil exploration by roots and yield advantage in comparison with the sole crops.