Essential oil yield and compositions of Dracocephalum moldavica L. in intercropping with fenugreek, inoculation with mycorrhizal fungi and bacteria

Intercropping is one of the most important components of sustainable agriculture. The effects of chemical fertilizer (CF), arbuscular mycorrhizal fungi (AMF) (Glomus sp.) and AMF + nitrogen-fixing bacteria (NFB) including Azospirillum and Azotobacter (AMF + NFB) was studied on essential oil yield and compositions of Moldavian balm (Mb) (Dracocephalum moldavica L.) in sole cropping and intercropping with fenugreek (F) (Trigonella foenum-graecum L.). The experiment was conducted during 2020 and 2021 growing seasons in East Azarbayhan, Iran. The highest dry herbage yield (6132 kg ha−1) was obtained in Mb:F(4:2) and CF treatment. After sole Moldavian balm, the highest essential oil yield (15.28 kg ha−1) was obtained in Mb:F (4:2) and AMF + NFB treatment. Geranial, geranyl acetate, geraniol, neral, and nerol were the main chemical constituents of essential oil. In AMF + NFB treatments the geranial contents in intercropping patterns of Mb:F (1:1), (2:2) and (100:50), increased by 25.1, 15.5 and 34.6% compared with sole Moldavian balm. The highest LERT values were observed in Mb:F (100:50) cropping pattern in 2021 (1.70 and 1.63 for CF and AMF + NFB treatments). Generally, it can be concluded that Mb:F (100:50) intercropping and use of AMF + NFB bio-fertilizer could be recommended to medicinal plant growers in sustainable production systems.

Plant height. The mean comparison of interaction effect of year × cropping pattern ( Fig. 1) showed that at all cropping patterns the plant heights in 2020 were greater than those in 2021 (13.5 and 28.2% reduction in 2020 and 2021, respectively, compared to sole Moldavian balm). In both years the plant height in Mb:F (100: 50) cropping pattern decreased compared with those in sole Moldavian balm and other intercropping patterns (Fig. 1). The interaction of cropping pattern × fertilizer treatment indicated that in Mb:F(1:1), Mb:F(2:2) and Mb:F(4:2) cropping patterns and sole Moldavian balm the plant heights were not significantly different in CF and AMF + NFB treatments. In Mb:F(100:50) cropping pattern the plant height in AMF + NFB fertilizer treatment increased by 3.9 and 2.8% compared with that in CF and AMF treatment, respectively (Fig. 2).
Leaf area index (LAI). The mean comparison of LAI in two years showed that the LAI in 2020 (7.87) increased by 19.8% compared with 2021 (6.57) ( Table 2). In intercropping pattern of Mb:F(1:1) the LAIs in AMF and AMF + NFB fertilizer treatments increased by 21.1 and 28.9%, respectively, compared with those in sole Table 1. The combined analysis of variance for effects of cropping pattern and fertilizer treatment on growth traits, essential oil content and yield of Moldavian balm. ns, * and **: non -significant and significant at P ≤ 0.05 and P ≤ 0.01, respectively. www.nature.com/scientificreports/ Moldavian balm and the highest LAI (9.56) was observed in AMF + NFB fertilizer treatment. In Mb:F(100: 50) intercropping pattern, the lowest LAI was observed and the LAIs at all fertilizer treatments were not significantly different (Fig. 3).

Dry herbage yield.
Moldavian balm dry herbage yield in 2020 (5467 kg ha −1 ) increased by 22.1% compared with than that in 2021 (4480 kg ha −1 ) ( Table 2). The mean comparison for interaction effect of year × cropping pattern (Fig. 4) indicated that at all cropping patterns, the dry herbage yields in 2020 were greater than those in 2021. Among the intercropping patterns the highest dry herbage yield was obtained in Mb:F(4:2) cropping pattern. The results of interaction effect of year × fertilizer treatment (Fig. 5) indicated that dry herbage yield in AMF + NFB treatment was not significantly different with that in CF treatment. The mean comparison for interaction effect of cropping pattern × fertilizer treatment showed that at all intercropping patterns the dry herbage yield decreased compared with that in sole Moldavian balm (Fig. 6). The highest dry herbage yields were observed in Mb:F(4:2) intercropping pattern. In Mb:F(100:50) intercropping pattern the highest dry herbage yield (5322 kg ha −1 ) was observed in AMF + NFB treatment, while in Mb:F(4:2) intercropping pattern the highest dry herbage yield (6132 kg ha −1 ) was observed in CF treatment.
Essential oil content (%). The mean comparison indicated that the Moldavian balm essential oil content in 2020 (0.261%) increased by 15.5% compared with that in 2021 (0.226%) ( Table 2). The highest essential oil content (0.261%) was obtained in Mb:F(2:2) intercropping pattern that was not significantly different with that  Essential oil yield. The mean comparison (Fig. 7) indicated that at all cropping patterns the essential oil yield in 2020 were higher than those in 2021. The greatest essential oil yield was observed in Mb:F(4:2) intercropping pattern. The mean comparison of interaction effect of year × fertilizer treatment (Fig. 8) indicated that in 2020 the essential oil yield in AMF + NFB treatment was not significantly different with that in CF treatment and in 2021 the essential oil yield in AMF + NFB treatment increased by 21.9% compared with that in CF treatment. The mean comparison of interaction effect of cropping pattern × fertilizer treatment ( Fig. 9) indicated that the essential oil yield decreased at all intercropping patterns compared with that in sole Moldavian balm. In Mb:F(2:2), Mb:F(4:2) and Mb:F(100:50) intercropping patterns the essential oil yield in AMF + NFB and CF treatment were not significantly different but in Mb:F(1:1) intercropping pattern, the essential oil yield in AMF + NFB treatment increased by 40.7% compared with that in CF treatment (Fig. 9).

Essential oil composition (GC-MS).
Moldavian balm essential oil compositions was assessed through GC-MS analysis and it was found that the essential oil was consisted of 19 components, accounting for 98.3-99.2% of the total compounds ( Table 3). Results of the GC-MS analysis indicated that the main essential oil    www.nature.com/scientificreports/ patterns of Mb:F(2:2) and (4:2) the nerol contents in AMF and AMF + NFB fertilizer treatments decreased compared with those in CF fertilizer treatments (Table 3).

LER.
In both years the LER T indices of all intercropping patterns were higher than 1.0 (  (Table 4).

Discussion
The Moldavian balm height in 2020 was higher than that in 2021, which can be attributed to more precipitation in 2020 than 2021 (Table 5). It was found that the Moldavian balm height increased in Mb:F(1:1) and (2:2) intercropping patterns with application of CF and AMF + NFB fertilizers (Fig. 2). According to Yang et al. 17 in shade conditions caused by intercropping, increasing plant height is expected by reducing the ratio of red to far red light (FR/R) and reducing the amount of photosynthetic active radiation. Increasing the plant height in intercropping conditions can be attributed to competition for more light interception and lack of light penetration through the canopy and non-decomposition of auxin hormone in these conditions 25 . Also, Ebwongu et al. 26 in corn (Zea mays L.)-potato (Solanum tuberosum L.) intercropping observed that the height of potato increased in intercropping compared to sole cropping. Also, increasing plant height in intercropping with legumes can be associated with high nitrogen availability 27 . The Moldavian balm height decreased in Mb:F(100:50) additive intercropping pattern compared to sole crop, that could be attributed to competition between crops for water, nutrients and biospace as a result of increasing crop density in unit area 27 . Agegnehu et al. 28 also reported that the faba bean height reduced significantly in intercropping with barley (Hordeum vulgare L.) due to interspecific competition. In Mb:F(100:50) intercropping pattern the greatest plant height was observed in AMF + NFB treatment that may be explained by better absorption of phosphorus and nitrogen which ultimately has led to improve the growth traits such as height 29 .
The highest LAI was observed in Mb:F(1:1) cropping pattern with application of AMF + NFB bio-fertilizer. The increase in Moldavian balm LAI in this treatment could be explained by improving the nutrients (N + P) availability through bio-fertilizer (AMF + NFB) treatment and biological nitrogen fixation by fenugreek. Nitrogen is one of the most important nutrients which affects crop growth, especially the development of leaf area by At all cropping patterns the Moldavian balm dry herbage yields in 2020 were greater than those in 2021 that may be related to higher precipitation in 2020 that has led to increase in growth and yield of Moldavian balm. After sole Moldavian balm, the highest dry herbage yield was obtained in Mb:F(4:2) cropping pattern, in which the proportion of Moldavian balm was higher than other cropping patterns. Also in dragonhead (Dracocephalum moldavica)-soybean (Glycine max) intercropping 33 observed that crop yields decreased in replacement intercropping patterns compared with those in sole crops. In 2020 the dry herbage yield in AMF + NFB fertilizer treatment was not significantly different with CF treatment and in 2021 increased compared with that in CF treatment. The increase in dry herbage yield of Moldavian balm as a result of bio-fertilizers application could be attributed to facilitative effect of arbuscular mycorrhizal fungi on fenugreek N-fixation and N transfer to Moldavian balm as reported by Li et al. 6 . In arbuscular mycorrhizal inoculation treatment, the faba bean N fixation and transferring to the intercropped wheat (Triticum turgidum L.) increased by 20% 22 .
The essential oil contents in Mb:F(1:1) and (2:2) intercropping patterns (0.252 and 0.261%) increased compared with those in Mb:F (100:50) and sole Moldavian balm (0.228 and 0.232%). Rezaei-Chiyaneh et al. 34 reported that the essential oil contents in all intercropping patterns of fennel (Foeniculum valgare, Mill.)-common bean, were higher than that of sole fennel. The essential oil content in AMF + NFB fertilizer treatment was higher than those of CF and AMF fertilizer treatments. Since essential oil is one of the secondary metabolites, its production is influenced by nutrients availability and photosynthesis rate 35 . Increasing the content of essential oil with using AMF + NFB bio-fertilizer treatment could be attributed to improvement of plant growth characteristics and essential oil-forming glands due to the increase in the level of absorption and access to nutrients 36 . It seems   www.nature.com/scientificreports/ that the intercropping of Moldavian balm with fenugreek especially in 1:1 and 2:2 patterns has increased the essential oil content by providing the main essential nutrients such as nitrogen which is affecting the activity of photosynthetic enzymes 37 and essential oil content. On the other hand, using the bio-fertilizers improves the availability of macro-and micro-nutrients, as a result, affect the amount of essential oil 38 . Also, Kapoor et al. 39 concluded that the use of bio-fertilizers caused changes in concentration of plant phytohormones which increased the formation of essential oil secreting glands and ultimately led to the production of more secondary metabolites. Increasing the content of essential oil due to use of bio-fertilizers in medicinal plants such as fennel 37 and Moldavian balm 2,9 has been specified. The interaction effect of cropping pattern × fertilizer treatment ( Fig. 9) on essential oil yield showed that the greatest values were observed in Mb:F(4:2) intercropping pattern, which can be attributed to the high proportion of Moldavian balm led to higher dry herbage yield in this cropping pattern. At all intercropping patterns except the Mb:F(1:1), the essential oil yield in AMF + NFB and CF treatments were not significantly different. Weisany et al. 40 also found that in coriander-soybean intercropping, application of arbuscular mycorrhizal fungi (Glomus intraradices) significantly increased the essential oil yield of coriander. Vafadar-Yengeje et al. 2 reported that in 1:1 faba bean-Moldavian balm intercropping pattern, the essential oil yield of Moldavian balm in chemical fertilizer and bio-fertilizer application was not significantly different. Also in previous studies increasing the essential oil yield of Moldavian balm 2,9,10 and basil (Ocimum basilicum L.) 41 as a result of bio-fertilizers application are specified. At all cropping patterns and fertilizer treatments the essential oil yield in 2020 were higher than those in 2021 that can be explained by higher precipitation in 2020 growth season which in turn increased the dry herbage yield, essential oil content and yield of Moldavian balm in this year (Table 5, Fig. 7 and 8).

4-Hexen
Based on the results of GC-MS analysis, it was found that geranial, geranyl acetate, geraniol, neral and nerol were the main constituents of essential oil. In previous studies, it has been found that geraniol, geranial, geranyl acetate, neral and neryl acetate were the main constituents of essential oil in Moldavian balm 2,8,9,33 . Hussein et al. 42 reported that in Moldavian balm the geranial and linalool were the main constituents of essential oil. The effects of fertilizer treatment and cropping pattern on major constituents of essential oil were different. The highest amount of geranial (61.4%) was observed in AMF treatment in sole Moldavian balm, which increased by 18.3 and 19.6% compared with that in AMF + NFB and CF treatments, respectively. Our results did not match with those of Vafadar-Yengeje et al. 2 in intercropping of faba bean-Moldavian balm which reported that the highest geranial content was observed in chemical fertilizer treatment. The highest content of geraniol (26.2%) was related to the AMF treatment in the Mb:F(1:1) cropping pattern. Vafadar-Yengeje et al. 2 found that in 1:1 faba bean-Moldavian balm intercropping the geraniol content was higher in bio-fertilizer and vermicompost  www.nature.com/scientificreports/ treatments compared with chemical fertilizer. Also it has been found that the bio-fertilizers improved the content of geraniol in essential oil of Moldavian balm in comparison with control treatment 10 . The AMF + NFB fertilizer treatment in Mb:F(4:2) cropping pattern showed the highest content of geranyl acetate (47.1%). Karimzadeh Asl et al. 10 also found that the greatest geranyl acetate content of Moldavian balm essential oil was obtained in bio-fertilizer treatment. On the other hand, the geranyl acetate content in CF fertilizer treatment of Mb:F(1:1) cropping pattern (31.2%) was higher than other cropping patterns. Also, the lowest contents of geranyl acetate in CF and AMF treatments were related to sole Moldavian balm (3.1 and 10.2%, respectively). In this regard, in faba bean-Moldavian balm intercropping, the content of geranyl acetate in chemical fertilizer treatment increased in all intercropping patterns compared to sole Moldavian balm 2 . Ganjewala and Luthra 43 reported that a negative correlation was observed between the amount of geraniol and geranyl acetate in lemongrass (Cymbopogon flexuosus L.); so that as geraniol increased, the content of geranyl acetate decreased. They also reported that this correlation could be related to conversion of geranyl acetate to geraniol during the leaf growth stage. It has been found that the differences in ecological factors, genetic origin and cropping patterns cause quantitative and qualitative differences in essential oil compounds 44 . In previous studies the positive effect of intercropping and use of bio-fertilizers has been reported on improving the essential oil composition of Moldavian Balm 2,45 and fennel 34 . Even though many studies indicated the positive effect of bio-fertilizer on essential oil constituents of medicinal plants such as white horehound (Marrubium vulgare L.) 46 and Moldavian Balm 2,9,10 , on the other hand, a number of studies showed the advantageous effect of chemical fertilizers on the essential oil compositions 33 . Different biosynthesis pathways of these components may be responsible for different responses to fertilizer application 47 . The biosynthesis of essential oils is influenced by the amount of nutrients in the soil such as nitrogen and phosphorus 48 . It can be expected that the use of bio-fertilizers through the provision of more nutrients and creation and division of new essential oil-containing cells, glandular trichomes and essential oil channels increase essential oil in medicinal plants 16 . According to the results of present study, inoculation with AMF and NFB and intercropping with fenugreek could be used in order to improve the yield and quality of essential oil in Moldavian balm.
The LER is one of the intercropping evaluation indices that indicate the intercropping advantage in terms of land use. The results of LER values (Table 4) showed that the highest LER Mb was obtained in the Mb:F(4:2) cropping pattern and CF fertilizer (0.87 and 0.85 in 2020 and 2021, respectively). Also, the highest LER F values were obtained in Mb:F(100:50) cropping pattern and CF fertilizer treatment (0.93 and 0.98 in 2020 and 2021, respectively). In general, the partial LERs for fenugreek (LER F ) were higher than those of Moldavian balm (LER Mb ) (except for Mb:F(4:2) cropping pattern), that indicated the intercropping had more positive effect on fenugreek. Previous studies 49,50 indicated that increasing the partial LER greater than 0.5 depends on the complementary degree of the intercropping components. Also, LER T values higher than 1.0 indicate the superiority of intercropping over sole crop. Previous researches have shown that superiority in intercropping is due to different morphological properties and growth and the tendency of intercropping components to make optimum use of sources such as soil, moisture, light and nutrient elements 2,4,32,34 . Also there is a difference in the root structure, distribution of the canopy cover and the nutritional needs of the plants in intercropping 4,51 .
The highest LER T (1.70) was obtained in Mb:F(100:50) cropping pattern and CF treatment in 2021, accounted for which was equivalent to an increase of 70% in land productivity compared to sole cropping of two species. The results of previous studies have also shown that when the legume species beside the other species are planted as intercropping, nitrogen fixation is stimulated due to the complementary effect, which increases the growth and yield of the legume species through increasing the number of active nodes 6,24,30 . Although the intercropping increases the competitiveness to absorb environmental resources, if one species has nitrogen fixation ability, competitive pressure will be reduced because the legume species will have less competition with other species in nitrogen absorption 4,51 . More distribution of nitrogen in the soil through nitrogen fixation and lower leaves fall of legumes improve soil fertility and adjacent plant growth in intercropping 24,30 . Fenugreek also appears to acidify the rhizosphere through nitrogen stabilization, thereby increasing the solubility and absorption of phosphorus by Moldavian balm 52 . At all cropping patterns the greatest essential oil yield of Moldavian balm was observed in AMF + NFB fertilizer treatment and the highest LER T values were obtained in Mb:F(100:50) intercropping pattern. It seems that in our study the inoculation of AMF increased the N transfer from fenugreek to Moldavian balm as reported in maize (Zea mays) and alfalfa (Medicago sativa) intercropping system 53 . Also the nitrogen availability could be improved through inoculation of NFB which in turn increased the Moldavian balm growth and essential oil yield.

Conclusions
The Moldavian balm height, LAI, dry herbage and essential oil yield were affected significantly by interaction effect of cropping pattern × fertilizer treatment. In Mb:F(1:1) and (100:50) intercropping patterns the highest dry herbage yields were observed in AMF + NFB fertilizer treatment. Among the fertilizer treatments the highest essential oil content (0.259%) was obtained in AMF + NFB treatment. At all cropping patterns except the Mb:F(1:1), the essential oil yields in AMF + NFB fertilizer treatments were not significantly different with those in CF treatments. Geranial, geranyl acetate, geraniol, neral and nerol were the main chemical constituents of Moldavian balm essential oil and affected differently by cropping patterns and fertilizer treatments. The LER T indices in all intercropping patterns were greater than 1.0 in both years, which indicates more efficient use of land and resources in intercropping patterns than sole crop. In both years the LER T values in Mb:F(100:50) intercropping patterns and AMF + NFB fertilizer treatments (1.55 and 1.63) were higher than those in other intercropping patterns and fertilizer treatments. Therefore, according to our findings, additive intercropping of Moldavian balm with fenugreek (Mb:F(100:50)) and the combined use of arbuscular mycorrhizal fungi + nitrogen-fixing bacteria (AMF + NFB) bio-fertilizers could be introduced as a eco-friendly and alternative methods to sole cropping and www.nature.com/scientificreports/ chemical fertilizer application in line with the goals of sustainable and cleaner production of Moldavian balm. For more studies, evaluating the other types of bio-fertilizers such as phosphorus-releasing bacteria and plant growth-promoting rhizobacteria (PGPR) could be recommended. Also there is a need for investigating the effects of intercropping of Moldavian balm with other legume and non-legume crops in different cropping systems in order to improve the efficiency and productivity of intercropping.

Material and methods
Experimental site. This research was conducted in Maragheh city in East Azarbaijan province, Iran (latitude 37˚4 N, longitude 46˚26 E, altitude 1478 m above sea level) in 2020 and 2021 growth seasons. The climatic data of monthly total precipitation and mean temperature of the experimental site during the growth seasons of 2020 and 2021 are presented in Table 5. The soil physicochemical characteristics of the experimental field at a depth of 0-30 cm are presented in Table 6. ) and additive intercropping of Moldavian balm + fenugreek Mb:F(100: 50) (100% optimum density of Moldavian balm + 50% optimum density of fenugreek planted between Moldavian balm rows). The fertilizer treatment (second factor) was consisted of 100% chemical fertilizer (CF), application of arbuscular mycorrhizal fungi (AMF) and combined application of AMF and nitrogen-fixing bacteria (AMF + NFB). The CF treatment was 50 kg ha −1 urea and 80 kg ha −1 triple superphosphate (according to soil test results); which were used at planting time. Myco-Root bio-fertilizer contains arbuscular mycorrhiza fungi (AMF) of Glomus mosseae, Glomus intraradices and Glomus etunicatum with count 10 7 to 10 8 CFU/gr is provided by Zist Fanavar Pishtaz Varian Company, Karaj, Iran. This bio-fertilizer is an easy-to-use powder form that is used for crops as seed inoculation. According to the manufacturer's instructions, the seeds of Moldavian balm and fenugreek required for the experimental area were placed in the shade on nylon or a clean surface, and after spraying a small amount of water on them, the required amount of AMF bio-fertilizer was added and mixed thoroughly; so that all the seeds were covered with a uniform layer of bio-fertilizer and after drying the seeds in shade, the planting was done. In order to inoculation of NFB, Biofarm bio-fertilizer contains Azospirillum and Azotobacter bacteria with a population of 2 × 10 7 CFU/gr is provided by Nature Biotechnology Company Plant height. In both years in order to measure the height of Moldavian balm, at flowering stage (on 29 July 2020 and 09 August 2021) 10 plants from each plot were randomly selected after removing the marginal effects (side rows and half a meter from the sides of the middle rows) and plant height was measured.

Leaf area index (LAI).
In both years the LAI was determined at flowering stage (on 29 July 2020 and 09 August 2021). In order to measure the leaf area of Moldavian balm, the plants in 1 m −2 area of each plot were cut from the soil surface, transferred to the laboratory and the leaf area of the all plants were measured by Delta-T leaf area meter (Delta-T Devices, Cambridge, England). By dividing the leaf area to the ground area (1 m −2 ) the LAI was obtained for all treatments 54 . www.nature.com/scientificreports/ Moldavian balm dry herbage yield and fenugreek grain yield. Moldavian balm plants were harvested manually from each plot (1 m −2 area) at the flowering stage to determine dry herbage yield on 29 July 2020 and 09 August 2021, and after drying in shade and room temperature (for 7 days), the herbage yield was measured as kg ha −1 2 . Fenugreek plants were harvested manually from each plot (1 m −2 area) at maturity stage on 02 September 2020 and 08 September 2021. In order to obtain the grain yield per unit area the harvested plants were dried at 30 °C for 48 h and after hand-threshing the grain yield was measured.
The content and yield of essential oil. In flowering stage the Moldavian balm herbage was harvested, dried and powdered. Then 100 g from powder was hydrodistilled for 3.0 h by a Clevenger device 9,55 . In order to dehydrate the obtained essential oil, the anhydrous sodium sulphate was used; and by measuring the amount of essential oil, it was expressed as a percentage (EO content). The samples were stored at 4 °C until GC-MS analysis. Also, the yield of essential oil (kg ha −1 ) was calculated as follows 8 .
Gas chromatography-mass spectrometry. Gas  LER assessment. In Moldavian balm-fenugreek intercropping patterns the LER values were assessed as follows 2 as: Where Y Mb and Y MbI are the Moldavian balm dry herbage yields in sole cropping and intercropping, respectively, and Y F and Y FI are the fenugreek grain yields in sole cropping and intercropping patterns, respectively. Also, LER Mb and LER F represent the partial LER of Moldavian balm and fenugreek, respectively, and LER T is the total LER. When LER T is equal to 1.0, indicates the same competitive ability of both crops in intercropping. This situation occurs if the inter-specific competition is equal to the intra-species competition in the intercropping components. In addition, if the increase in the yield of one crop is accompanied by a decrease in the yield of another crop in intercropping, the LER T value will be equal to 1.0. If the value of LER T is less than 1.0, sole crop is superior to intercropping, and if the value of LER T is higher than 1.0, intercropping is more beneficial than sole crop 2,4 .
Statistical analysis. For analysis of variance (ANOVA) the SAS Version 9.0.3 was used. For two growing seasons of 2020 and 2021 and all traits the combined ANOVA was done based on complete randomized block design with 15 treatments and three replicates. The data of LER M , LER F , and LER T were not subjected to analysis of variance. The experimental data met the assumptions of normality and variance homogeneity and no transformation was needed. For comparison of the means the Duncan´s multiple range test was used at p ≤ 0.05.

Data availability
The necessary information is available from the corresponding author on reasonable request.