The field experiment was conducted over three consecutive seasons (2022, 2023, and 2024) at two Agroscope sites in Switzerland: Reckenholz (Zürich-Affoltern; altitude 444 m; 47°42'87"N, 8°51'67"E, canton Zürich) and Tänikon (Aadorf; altitude 539 m; 47°47'93"N, 8°90'68"E, canton Thurgau). Different fields were selected each year at both sites for a total of six different environments.

Soil types and field coordinates were identified using the AGSs WebGIS mapping system provided by Heller et al. (2023) based on the World Reference Base (WRB) soil classification. Detailed information regarding soil type, previous crops, and specific field operations for each experimental condition (site × year) is provided in Supplementary Material (SM) (Table A).

Throughout the experimental seasons, Reckenholz experienced overall warmer temperatures and lower precipitation than Tänikon (Table B, SM). However, during the growing seasons, cumulative temperatures were lower at Reckenholz than at Tänikon (Fig. 1; Table C, SM).

The first season (2022) was notably drier compared with historical cumulative precipitation records (1990–2020; Fig. 1; Table B, SM). In contrast, the subsequent seasons (2023 and 2024) recorded above-average precipitation during the early stages of the growing period, followed by drier summer conditions when compared with historical averages (Table B, SM).

A randomized complete block design with four blocks was employed across the entire experiment. The study included four lentil cultivars, Anicia, Beluga, Grüne Berry Linsen, and Château Linsen, and three spring hull-less barley cultivars, Oak Ruby, Golijat, and AF Cesar, as companion cereals.

Each crop was cultivated both as a sole crop and in mixtures. Planting density was standardized at 240 plants/m², with sole cropping representing 100% of the species and mixtures sown at a 3:1 lentil-to-barley ratio. Sole cereal treatments were established under unfertilized and fertilized conditions. Fertilized treatments received a single application of mineral NPK fertilizer with 110 kg N, 95 kg P, 220 kg K ha^-1, following Swiss fertilization guidelines to achieve a target of 90 kg N ha^-1.

In total, 22 treatments were established: three sole unfertilized barley cropping, three sole fertilized barley, four sole lentil, and 12 mixed. To minimize border effects, all treatments were established in triplicate adjacent plots, with only the central plot used for data collection (excluding destructive sampling) and harvesting (Fig. 2).

For all treatments and years, agronomic management was identical for all plots. Year-specific management details are presented in Supplementary Material (Table A). Sowing was carried out simultaneously each season at a depth of 1.5 to 3 cm, depending on soil conditions. Each individual plot measured 9 m² (1.5 m × 6 m), and a 1.00 ± 0.05 m buffer zone was mown between plots to reduce inter-plot interference, resulting in an effective harvested area of around 8.5 m² (1.5 m × 4.75 m). No herbicides, fungicides or insecticides were applied during the cropping cycle. Harvesting of the central plots was performed using a small-scale experimental combine (Wintersteiger, Ried Austria). Grains were dried until grains reached a maximum humidity of 12%. Subsequently, seeds were cleaned using a gravity separator which separates particles based on specific weight differences, effectively removing lighter or damaged seeds from the sample and then separated using a small-scale seed separator machine (Indented Cylinder, model: LA-T, Westrup, Slagelse, Denmark) with mantle wells of 5.5 cm diameter.

Data collection consisted of two phases. A pre-harvest phase where data were collected directly on the field with weekly assessment for the observations reported in Table 1. The post-harvest phase consisted in drying, cleaning and separating grains from mixed cropping to quantify total dry grain yield.

Different years, cropping systems and cultivars showed significant effects on grain production, whereas different sites showed no significant effect per se for total grain production in lentil but only when interacting with different cropping systems and years (Table 2).

Total grain production decreased steadily throughout the three years (Fig. 3). When pooled across both sites and all treatments, the average median for total productivity peaked in 2022 (2.90 ± 0.06 t ha^− 1), fell in 2023 (1.40 ± 0.09 t ha^− 1) and reached its lowest in 2024 (0.95 ± 0.05 t ha^− 1).

Different cultivars showed a significant effect on total and on components yield for HL barley and lentils.

For the first three environments (Reckenholz and Tänikon 2022 and Reckenholz 2023), Oak Ruby had the highest yield recorded as a cultivar in sole cropping. In the first environment (Reckenholz 2022), the sole unfertilized Oak Ruby averaged a median of 3.20 (± 0.08) t ha-¹. In the second (Tänikon 2022) and the third environment (Reckenholz 2023), Oak Ruby showed again the highest yield for sole fertilized cropping, averaging median yield respectively of 3.95 (± 0.17) t ha-¹ and 4.90 (± 0.30) t ha^− 1, resulting in significantly higher values.

For environments 4 (Tänikon 2023), 5 (Reckenholz 2024), and 6 (Tänikon 2024), dry grain yields dropped (Fig. 3, figure A in SM). In environment 4 - thus same year but different site compared to environment 3 - mean yield dropped by 49% and the highest grain yield was registered for mixtures between Golijat-Beluga, averaging 2.03 (± 0.1) t ha^− 1.

In the third year, hull-less barley production dropped (Fig. 3). The highest dried grain yield production was reached by sole Grüne Berry (1.7 ± 0.16) t ha^− 1 in Reckenholz (environment 5) and by Oak Ruby/Beluga in Tänikon (environment 6), averaging 1.7 (± 0.11) t ha^− 1.

Post-hoc test on regression model for total grain yield demonstrated that sole cropping for lentil cultivars Anicia, Grüne Berry, and Beluga yielded significantly more than all other cultivar–cropping system combinations (p < 0.05). By contrast, Château showed no significant yield difference between sole and mixed cropping. Nonetheless, Château Linsen suffered a marked yield reduction (p < 0.0001; Table E in SM) compared with the other lentil cultivars.

For the first two environments (Reckenholz and Tänikon 2022), a higher mean temperature (1.44 and 1.33°C higher) and lower precipitation (271 mm and 411 mm, − 34% and − 15%) were recorded compared with historical means (table B, table D in SM), resulting in the highest yields across all cropping systems (Fig. 3).

The subsequent year (2023) showed significant site-specific differences in yield (Table 2, Fig. 3), which we attribute to persisting rain leading to a 22-day sowing delay at Tänikon. Although the growing season lengths were similar (140 days in Reckenholz and 139 days in Tänikon), productivity differed considerably between sites for both HL barley and lentils (Table F in SM). At Reckenholz (environment 3), HL barley obtained the highest yields of the entire experiment, increasing by 26% and 24% in sole fertilized and unfertilized cropping, respectively, and 28% when mixed compared with 2022 (Table F in SM). In contrast, lentil yields declined sharply by 76% (mixed) and 68% (sole) compared with 2022. At Tänikon (environment 4), the dynamics were reversed. HL barley yields dropped by 73% (sole) and 87% (mixed) compared with 2022, while the decline in lentil yield was lower with a reduction of 30% in mixed cropping and 25% in sole cropping.

In 2024, the precipitation was slightly above the historical mean (3% at Reckenholz and 6% at Tänikon), yet overall yields were the lowest, reaching a median average of 1.1 t ha^− 1. Again, frequent precipitation resulted in an even longer sowing delay of 38 days (at Reckenholz) and 20 days (at Tänikon) compared with 2023, shortening the growing season by 19 to 25 days, respectively, compared with 2022.

The drop in production observed in 2024 was not an isolated episode. As reported by swissgranum, the Swiss branch organization for cereals, oilseeds and protein plants, plant growth in 2024 was overall suboptimal due to the frequent precipitation and significantly fewer sunshine hours compared with the average of recent years, explaining the reduction in quantity and quality of the harvested goods (swissgranum, 2024).

Indeed, temperature and precipitation are major factors influencing inter-annual yield variation (Matiu et al., 2017). We assumed that the major source of yield variations was defined by the combined effect of sowing time and precipitation intensity recorded along the study period.

Late sowing and high precipitation seemed to have a greater effect in reducing HL barley performance than in lentils, highlighting HL barley’s sensitivity to wet conditions and shortened growth period and, conversely, lentil’s superior resilience under the same conditions.

Our results are in line with previous works (Cammarano et al., 2019; Hakala et al., 2012, 2020; Borrego-Benjumea et al., 2018) showing barley as a susceptible crop under wet conditions and sowing delay in temperate environments. In contrast, lentils appeared to be less susceptible to cool and wet conditions as well as sowing delay. A study by Wang et al. (2013b) on HL barley-lentil intercropping confirmed higher yields for early sown mixtures and a yield decline after late sowing for both crops. However, unlike Wang et al. (2013b), it was observed that early-sown lentils yielded poorly, particularly when HL barley exhibited vigorous growth, which might have outcompeted lentils for light and nutrients, as was the case for Reckenholz in 2023 (Fig. 3; Table F in SM).

Barley and lentils are often proposed as highly resilient crops against abiotic and biotic stresses (Witzel et al., 2009; Mahmud Al Noor et al., 2024), especially under drier conditions. However, it is often underestimated how changing rainfall patterns and weather extremes may negatively affect such crops (IPCC, 2012).

Few studies observed mixed cropping performances between lentils and spring barley under temperate areas (Schmidtke et al., 2004; Wang et al., 2012, 2013a, 2013b). In all the previously cited studies, mixed cropping reported improved yields compared to sole-cropped systems.

Our values obtained for the land equivalent ratio (LER) support the advantage of adopting mixed cropping systems for lentils paired with HL barley as a companion species.

Across all data (n = 269), the mean LER was 1.23 (± 0.30), and all mixtures except Golijat / Grüne Berry (0.88 ± 0.40, p > 0.05) exceeded a LER of 1 (Table 3). The partial LER was overall higher for lentils (0.78 ± 0.26) than for HL barley (0.46 ± 0.17), confirming previous findings that lentil-barley combinations often outperform sole cropping (Tosti et al., 2023; Wang et al., 2012).

Significant sources of variation for LER were found for different HL barley and lentil cultivars, as well as for lentil cultivars-by-year. Different years and sites did not have an effect per se on LER values, but only a marginal combination with different HL barley cultivars.

Year 2022 reported significantly higher values (+ 0.14) for LER with an average mean of 1.29 (± 0.05).

Among lentil cultivars, Château Linsen showed a significant positive effect in increasing LER (+ 0.31) while a negative effect was observed for Anicia (− 0.12). For HL barley cultivars, a significant increasing effect was given by AF Cesar (+ 0.08) while mixtures with Golijat tended to reduce significantly LER (− 0.09).

LER values obtained are comparable to the meta-analysis from Yu et al. (2015), showing an advantage in cereal/legume intercropping systems, averaging 1.22 (± 0.22). Specifically, considering a similar experimental setting, our results correspond to the one obtained by Wang et al. (2012) for lentil / HL barley mixtures averaging 1.38 (± 0.05) at a 3:1 sowing ratio. Similarly, average yields obtained were similar, as well as variation due to weather conditions (Wang et al., 2012).

The lower LER obtained for mixtures using Anicia as a lentil cultivar and Golijat as HL barley is a consequence of the higher productivity recorded as sole cropping for these two specific cultivars.

The significantly higher values obtained by Château Linsen (Table 4) are assumed to result from a possible positive effect given by a contamination in the seed lot with vetches (Vicia sativa), which potentially influences mixed cropping results. The improvement in LER shows the high benefit for Château Linsen when grown under mixed cropping, as reported also by average general higher partial LER for Château compared with the other cultivars.

The year effect obtained in 2022 resulted from the overall higher yield performance recorded during the first year of the experiment. As mentioned in section 3.1, lower precipitation and warmer temperature combined with a regular sowing time (beginning of March) resulted in overall higher productive performances, which have benefits as well as mixed cropping values compared with sub-optimal recorded in 2023 and 2024.

Positive NE was observed in all mixtures except Golijat/Grüne Berry (NE < 0). The main contribution to NE came from the HL barley component, which showed consistently positive average NE and a significant cultivar effect (Table 4). Mixtures with Oak Ruby significantly increased NE (+ 0.14) while Golijat showed a reduction effect on NE (− 0.15). Lentil cultivars Anicia and Château Linsen showed also a significant effect on NE. Combinations with Anicia tended to reduce NE (-0.18) while Château Linsen increased (+ 0.14)

Year, site, and their interaction also had a significant effect on overyielding, showing higher estimates in 2022 (+ 0.27) and Reckenholz (+ 0.09).

For the CR, a significant effect was found for year and lentil cultivars as independent factors, while two-way interactions year-by-site and year-by-HL barley cultivars were found to have a significant effect on CR (Table 4). AF Cesar and Golijat showed a reverse effect depending on the season. In 2022, Golijat showed a significant lowering effect (-0.32) while AF Cesar tended to increase CR by 0.46. Contrary, in 2023, AF Cesar had a negative estimate (-0.28) on CR, while Golijat had a positive (+ 0.42), meaning that for the first year (2022), combinations using Golijat tended to have higher yields compared to the lentil component, lowering lentil presence. A reverse effect was observed in 2023 with significantly lower Golijat presence. Such effects are in line with the yield dynamics registered, which presented a lowered fit for HL in 2023 and 2024. Concerning lentil cultivars, Beluga and Château Linsen increased CR significantly, meaning that those two mixtures were better at keeping HL barley contained, reducing competition.

Comparing the yield performances between mixed and sole lentil systems, lentil yielded significantly lower in mixed cropping compared with sole (− 22%). However, considering total production, a significant positive estimate was found for mixed cropping, showing an increase in overall yield of 13%. Total yields for systems using Anicia and Beluga were significantly higher compared with Château by 12% and 11.7%. The year effect was also found to be significant. Total yields obtained in 2022 were 22% higher compared with the general mean, while in 2023, they decreased by 17%.

Despite not having a significant site effect, the advantage of mixed cropping was most pronounced under Reckenholz conditions. A significantly higher total yield (+ 21%) for mixed cropping was obtained in 2023 at Reckenholz. Considering the interaction between high total grain yield and LER, the best combinations were Beluga with Oak Ruby or with AF Cesar as companion HL barley cultivars (Fig. 4).

Analysis of variance for site, cropping system and year reported significant variation on both weed volume and lodging (Table 5).

Weed pressure was significantly lower in season 2023 (− 45%) and at Reckenholz (− 43%) compared with season 2024 and Tänikon. Mixed system reduced significantly the weed pressure by 30% while no significant effect was found for any specific lentil cultivars.

Similarly, lodging varied significantly during the experiment period. Years effect reported significantly lower values for 2022 (− 31%) and positive for 2023 (+ 17%). Location effect was also found significant, with lower values obtained at Reckenholz (− 11%) compared with Tänikon. Cultivar effect found significantly lower lodging for mixture using Beluga (− 7.8%), although no significant effect was found for cultivar-by-cropping system, suggesting a missing specific improvement for any cultivar when grown under sole or mixed cropping. Finally, the cropping system reported a significant effect, showing a reduction of 19% of plot lodging in lentil-HL barley mixtures compared with sole lentil cropping.

The main role of the cereal in mixed cropping is to reduce weed pressure and to provide mechanical support to avoid lodging. Overall, our results confirm this assumption. Earlier studies reporting higher weed suppression ability for mixtures between legume and small cereal grains compared to sole crop legume systems (Gu et al., 2021), particularly when barley was used as a companion plant (Corre-Hellou et al., 2011; Wang et al., 2012), correspond to our observations. The significantly higher weed pressure observed in Tänikon likely reflected differences in field management practices and a larger soil seed bank at this site. Still, we suggest the importance of mechanical weeding as the main method to reduce weed pressure during early crop development (Gardarin et al. 2022).

Results for lodging support our assumption that combinations for lentils-HL barley can reduce plot lodging as reported in previous studies (Anil et al., 1998; Podgorska-Lesiak & Sobkowicz, 2013; Bedoussac et al., 2015).

Expression of the morphological traits varied by year, by cropping systems and by cultivars (Table 6).

Overall, our results suggest an increase in canopy structure for lentil and HL barley under mixed cropping. In HL barley, tiller numbers were significantly higher in mixtures compared to sole cropping, while height increased under sole fertilized systems compared to unfertilized and mixtures (Table 6).

Different years, sites, cropping systems and two-way interaction year-by-site and year-by-cropping systems were found to significantly affect tiller numbers along the study period (Table 6).

Higher counts were found for AF Cesar and Oak Ruby compared with Golijat, despite no specific significant effect on any specific cultivar (Table 6) being noted. On the other hand, the number of tillers observed in mixtures was significantly higher compared to sole systems, with an average median of 7 tillers compared to 5 tillers from sole cropping.

Higher values were counted at Reckenholz (8 versus 5 recorded in Tänikon) and during the 2023 season, getting 17% more compared to the average (8 versus 7 in 2022 and 5 in 2024). Significant negative reduction was observed at Tänikon 2023, recording a median of 5 tillers versus 10 at Reckenholz 2023 and for sole systems in 2022, reporting the lowest average with 4 tillers, explaining the year-by-site and year-by-cropping system effect. We assume that the improved vigor registered corresponds to the better productive performance recorded for Reckenholz 2023.

The higher tiller count found in mixed cropping (7 versus 5 for sole cropping) resonated with the review article written by Mmbando (2025). Tillering can be impacted by planting density, which can change the resources available to individual plants (Yang et al. 2019; Yu et al. 2020; Veenstra et al. 2023a). A lower tiller number and less tillering are, in general, results of increased competition between plants for resources (Huang et al., 2013; Yu et al., 2020; Alipour Abookheili & Mobasser, 2021). Reduced planting density might encourage greater tillering and possibly higher yields (Zhang et al. 2021; Alipour Abookheili & Mobasser 2021). Similarly to Koskey et al. (2022), they found an increasing percentage of tiller index under low wheat-lentil density intercropping. We argue that the higher number of tillers found in the mixture can be a consequence of the lower density and possible lower intraspecies competition for similar resources, which may have enhanced spatial growth for HL barley.

Besides tiller numbers, plant height also presented significant variation along the study period with year, site, cropping system and cultivar, resulting in significant effect on plant height variation, as well as for two-way interaction year-by-site (Table 7). For HL barley, site-by-cropping system and cultivar-by-year reported significant variation, while these significant interactions were not observed in lentils.

Considering HL Barley, a significant effect for cropping systems is related to taller plants observed under fertilized sole cropping systems (Table 8), presenting an increase in height on average of 4 per cent compared with unfertilized systems and mixed stands. The cultivar AF Cesar was significantly shorter (50.0 ± 13.3 cm) compared with Golijat (64.5 ± 19.2 cm). Also, the year 2022 (+ 8.9%) and 2023 (+ 3.15%) showed significantly taller plants compared with 2024, as well as values for Reckenholz (+ 2.46 cm), which were higher compared with Tänikon. Year-by-site interaction confirmed significantly taller HL barley in Reckenholz in 2023 (+ 8.0 cm) and the significantly taller Golijat observed in 2022 (+ 7.45 cm).

Lentil vigor also showed improvement under mixed systems with significant decrease in height under sole cropping by 49%. In mixtures, they gain on average up to 10 cm in height from an average of 23 cm in sole cropping to 33 cm in mixtures. Anicia was significantly the shortest lentil cultivar (Table 8). Yet, when mixed with crops, its height increased by 8% compared with the others. On the other hand, Beluga and Château were found to be significantly taller cultivars, with Beluga being the tallest (35 ± 14.8 cm) among the four lentil cultivars (Table 8). Year-by-site effect recorded referred to significantly shorter lentils recorded in 2024 at Tänikon (22.0 ± 4.4 cm).

In contrast, the height ratio, calculated as HL barley height divided by lentil height, was not directly influenced by HL barley–lentil interactions but was determined independently by each species.

Mixtures with Golijat were on average higher (+ 0.35), while combinations with Beluga (-0.26) and Château (-0.24) reduced the differences between species. Variation in plant height for different seasons resulted in a significant year effect on height ratio as well. Significantly higher values were found for 2023 (+ 0.26) and 2024 (+ 1.79).

Contrary to Tosti et al. (2022), who did not find an influence of companion plants (barley and triticale) on lentil height, we support the assumption that mixed cropping improves lentil vertical growth, increasing its height. An interaction observed in the field, but not quantified, involved the number of tillers produced by HL barley and the tendrils of lentils. Fabaceae generally produce specialized leaf tendrils that twine around supports, providing a secure attachment (Hattermann et al., 2022). We assume that in mixed systems for cereals-legumes, the presence of robust cereal stems may help the vertical growth of the legume, enhancing its vertical height by creating support for the tendril anchored

Similarly to morphological traits, phenology was also directly influenced by seasons, cropping systems and cultivars used (Table 6). We considered days after sowing to reach half plant flowering (BBCH 65 in legumes) and to reach half heading (BBCH 55 in cereals) as a benchmark for the phenological status. Different cropping systems did not show a significant effect on lentils, while a significant reduction was present on sole HL barley cropping systems with an observed earlier heading under sole fertilized contexts (− 2 days on average).

AF Cesar showed significant delay in heading (+ 2 days) while Golijat required fewer growing days (-1), being then an earlier maturing cultivar. Year effect referred to longer growing days registered in 2022 (+ 3 days) and 2023 (+ 7 days), while on average, Reckenholz recorded a delay heading (+ 1 day) compared with Tänikon.

A significant cultivar effect was found for flowering in lentils and their combinations for year and sites. Also, different years and sites had an independent effect on reaching flowering stage. Year 2022 and 2023 required significantly more growing days (+ 6 days in 2022 and + 5.6 days in 2023) to reach maturity compared to 2024. Also, observations in Reckenholz reported a delay in flowering by 3 days compared to Tänikon.

Anicia is confirmed as an early-flowering cultivar (− 4.9 days) while Château Linsen was a late cultivar (+ 9 days). Year-by-site effect reported significantly fewer growing days in Reckenholz 2022 (− 3 days) and significantly more growing days (+ 5 dys) in the following season 2023 to reach flowering.

Gap in reaching maturity (Δ maturity) was estimated as the difference between days after sowing registered for HL barley (BBCH 55) and days after sowing for lentil (BBCH 65).

A significant increase in Δ maturity was found for mixtures involving the HL barley AF Cesar and the lentil Anicia. AF Cesar increased Δ maturity as a heading cultivar, while Anicia increased Δ maturity as an early-flowering cultivar. Significant year effect reported lower Δ maturity days in 2022 (− 3 days) and longer in 2024 (+ 4.3 days). Site effect registered lower Δ maturity at Reckenholz. However, year-by-site effect showed significantly larger Δ maturity at Reckenholz in 2022 (3.25 days), while significantly lower for the following season (− 5 days). Combinations using AF Cesar and Anicia significantly increased Δ maturity by 2.2 days and 4.3 days. Contrary, significant reduction effects for Δ maturity were registered for Golijat (− 1 day) and Château Linsen (− 9 days) Delay or anticipation in reaching flowering did not vary significantly in lentil when grown under sole or mixed cropping systems. Contrary, HL barley showed a significant reduction in days reaching heading in sole cropping. As observed in Hauggaard-Nielsen et al. (2001) and Galanopoulou et al. (2019), the modest delay in barley phenology when grown with legume might be attributed to two reasons. Firstly, the higher N availability in sole fertilized systems may have enhanced stem elongation and grain filling, shortening the time to reach maturity compared to mixed cropping. Furthermore, the sowing delay and reduced growing season for season 2024 and season 2023 at Tänikon might have favored lentils over HL barley during the first stages of growth, particularly related to light acquisition, leading to a delay in reaching maturity. Shorter plants recorded in these three environments (respectively averaging median of 50.0, 45.0, 39.5 cm) could further represent a second factor related to the general reduced growth performance of HL barley recorded.

Morphological disparity, measured as height ratio, significantly affected both system productivity and lodging. Estimates obtained for total and lentil grain yields as well as for LER, NE indicate a decrease in performance for mixed cropping with increasing height ratio between species (Table 9). Also, a significant positive estimate was found for height ratio and lodging (Table 10). Spearman’s correlation between height ratio and Δ maturity (r = 0.27, p < 0.0001) also supports the assumption that matching late-flowering lentils together with early-heading HL barley will likely reduce height differences, advantaging similar growth between the two partners.

Conversely, tiller numbers did not influence lentil yield, lodging, or weed volume significantly. Yet, significant coefficients were obtained for total and HL barley grain yields as well as NE and CR, suggesting that mixtures with more tillers generally exhibit greater crop dominance (lower CR) and overall higher productivity (Table 10).

The results support the assumption that canopy structure has considerable implications in intercropping systems (Willey, 1990). Preferably, combinations with reduced height differences and a higher number of tillers demonstrated higher production performances while reducing lodging (Fig. 5; Table 9). The negative estimates between height-ratio and indexes indicate a reduction in LER and NE with the increase in height differences.

The reduction in LER and NE might be explained by the higher yields obtained by taller HL barley cultivars grown in mixtures (r = 0.80, p < 0.0001). Significant height disparities can create competition or even suppression if one crop is overtaking the other (Wang et al., 2021). Correlation between HL barley height and CR showed a significant negative coefficient (r = − 0.43, p < 0.0001), suggesting that taller HL barley tends to reduce lentil competition.

Reduction in performance for lentil resulted in a reduction of performance for the entire systems, considering that the lentil sowing share was higher compared with that of HL barley (180 lentil seeds versus 60 HL barley seeds per m^− 2).

Several studies proposed canopy heterogeneity as a key aspect to enhance spatial niche differentiation, increasing light capture and resource utilization (Willey 1990; Hauggaard-Nielsen & Jensen 2001; Bedoussac & Justes, 2010). Frequently, this is proposed in maize intercropping systems (Ren et al., 2025), where C4 and C3 plants are combined. However, given our results, we argue that using combinations of short-stemmed HL barley with lentils may reduce competition (ρ = -0.21, p = 0.001) while enhancing total production and land ratio.

Finally, the overall improvement in mixed cropping performances for cereal cultivars with a higher number of tillers is in line with Haug et al. (2023), indicating also in HL barley-lentil mixtures, barley genotypes with traits that provide above-average competitive ability suitable to attain a high general mixing ability. The assumption that a higher number of tillers would reduce plot lodging was partially confirmed by the correlation coefficient (ρ = − 0.28, p < 0.0001) but was not found to be significant by ANOVA type II (Table 10).

Disparity in days between reaching flowering (BBCH 65 in lentils) and heading (BBCH 55 in HL barley) growth stage was found to significantly affect mixed cropping performances.

A Type II ANOVA (Wald χ² test) on the best-fit regression models showed that Δ maturity days, height ratio, and number of tillers significantly influenced total and partial (lentil and hull-less barley) production, LER, NE, and lodging, but not weed volume (Fig. 6; Table 9).

An increase in Δ maturity days had a significant effect on total and lentil grain yields, as well as on lodging and indexes (LER, NE) (Table 9). The effect size of changes in Δ maturity days on total grain yield differed across seasons, with an average marginal increase of 2.1%. Notably, inter-annual variability further modulated this effect. In 2022, Δ maturity days had a significant positive effect compared to the grand mean with increase of 3.78%. In 2023, Δ maturity days increased by 10.3% while dropped in 2024 (-7.7%). Also, lentil production registered higher values for Δ maturity days, boosting yield by 21% when early-flowering lentils grew with late-heading HL barley. In contrast, LER and NE significantly reduced their values with increasing Δ maturity days, with a general reduction of 19% for LER and − 84.5% for NE. Plot lodging was also significantly increased with Δ maturity days (Table 9).

The positive estimates suggest that pairing early-flowering lentils together with late-heading HL barley increased total grain production, boosting lentil yields, even though there was an increased risk of lodging. Contrary to improved yields, combinations between early-flowering lentils and late-heading HL barley showed negative coefficients for the two indexes LER and NE as well as for total grain production, suggesting that combinations of early-heading HL barley with late-flowering lentils tended to improve more under mixed cropping compared to the reverse phenological mixtures. Spearman’s correlations between days to reach heading in HL barley-lentil grain yield and between days to reach flowering in lentils- HL barley grain yield reported significant increases in production for HL barley cultivars paired with late-flowering lentils (r = 0.67, p < 0.0001) while no effect was found between HL barley phenology and lentil grain yield (p > 0.05).

Therefore, according to our results, HL barley grown in mixed cropping contexts received higher benefits and influence from lentils' phenological delay rather than the other way around. Yet, to maximize lentils’ yield and total grain production - based on our results - cultivar choice should pick mixtures of early-flowering lentils with late-heading HL barley. The significant reduction for LER and NE corresponded to the reduced performance of HL barley when it showed late heading.

Beside the production increase, a significant positive effect between lodging and Δ maturity days was found. Such an effect shows that the temporal discrepancy between lentils and HL barley will likely favor lodging between early-flowering lentils, together with late-heading HL barley.

These results might be related to higher biomass observed in the field by early maturing lentil cultivars that tended to push down the overall plot, increasing lodging percentage.

For instance, combinations of AF Cesar with Grüne Berry showed the highest total grain yield, averaging 2.2 t ha^− 1, positive overyielding (0.35) and the highest production for HL barley (1.1 t ha^− 1). Such a mixture had the largest Δ maturity days with lentils setting flowers 9 days in advance compared with heading for HL Barley. We assumed that an earlier flowering for lentils may increase N fixation and N use by the companion plant, increasing its production. In support of this, the correlation coefficient between HL barley yield and days to reach flowering in lentils was strongly positive (r = 0.67, p < 0.0001).

To further support this mechanism, the other two mixtures, the best performing mixtures in terms of lentil production, were Beluga mixed with AF Cesar (1.50 t ha^− 1) or with Oak Ruby (1.40 t ha^− 1). These combinations overlapped in co-growth, matching the flowering and heading period. We argue that the reduced temporal complementarity did not boost HL barley as much as for combinations with dissimilar phenology (Golijat-Château Linsen).

Yet, these two mixtures were the second (AF Cesar-Beluga) and the third (Oak Ruby-Beluga) highest in total grain production, both with LER values above 1 (1.20 for Oak Ruby-Beluga and 1.16 for AF Cesar-Beluga).

Successful intercropping aims at a complementarity between diverse functional traits rather than competition, achieved through different resource acquisition strategies (Callaway et al., 2003; Brooker et al., 2015; Zuppinger-Dingley et al., 2014).

Intercrops between early- and late-maturing species are adopted widely to exploit the length of the growing season (Lithourgidis et al., 2011) and increase light interception over time (Keating & Carberry, 1993; Zhang et al., 2008). Our results are in line with previous studies showing how temporal discrepancy affected mixed cropping performances (Yu et al., 2015). Haug et al. (2023), who tested 8 spring barley genotypes with 27 spring pea genotypes under Swiss environments, reported significant GMA (general mixing ability) improvement for pea-traits showing early vigor, onset of flowers together, higher biomass and longer stipule length. Similarly, studies from Jensen et al. (2020) and Timaeus et al. (2022) reported that combinations between early maturing field peas with cereals release N from the roots. This happens around the time when cereal flowers and require increased N for grain filling. Furthermore, as reported by several authors (Schmidtke et al., 2004; Hauggaard-Nielsen et al., 2008; Rodriguez et al., 2020), and in the context of intercropping cereals-legumes, the ability to fix N from legume crops is the reason for the yield advantage of the cereal-legume mixture. Schmidtke et al. (2004) found that intercropping barley with lentils led to increased barley productivity, primarily due to complementary N acquisition strategies between the two species.

Given our experimental design using a replacement design within mixed cropping systems and according to our results, combining early flowering lentils with late-heading HL barley seemed to enhance overall productivity, while higher mixed cropping performances could be obtained for combinations between late-flowering lentils and early ripening HL barley.

Therefore, depending on the purpose of cultivation, cultivar selection could shift for diverse temporal maturity (Ross et al., 2004) if a higher share of one species over the other is desired.

Overall ranking across multiple agronomic and yield-related traits (Fig. 7) highlighted that Oak Ruby–Beluga and AF Cesar–Château Linsen were the most balanced and stable mixtures, whereas Golijat-based mixtures generally ranked lower, confirming that cultivar choice strongly determines the performance and resilience of mixed intercropping systems.