A two-year experiment (2014-2015 and 2015-2016) was performed at the Research Field of Seed and Plant Improvement Institute (SPII), Karaj, Iran, to study the response of rapeseed genotypes to foliar application of zinc and selenium at different sowing dates. The experiment site (Karaj) is located at an altitude of 1321 meters above sea level with 35˚ 59΄ N latitude and 50˚ 75΄ E longitude. The daily climatic data of the experiment location during the rapeseed growth period are presented in Figure 1.

The experiment was conducted as a factorial-split plot in a randomized complete block design (RCBD) with three replicates. In the present study, three sowing dates of 7 (well-timed sowing), 17, and 27 October (delayed sowing dates) and two levels of foliar application with pure water (control), selenium (1.5%), zinc (1.5%), and selenium+zinc (1.5%) were factorial in the main plots and five genotypes of SW102 (early-maturity line), GKH2624 and Okapi (mid-maturity cultivar), and Ahmadi and GK-Gabriella (late-maturity cultivar) were randomized in the subplots (a total of 30 treatments). Sodium selenate (as selenium source) and chelated zinc (as zinc source) were applied in two stages of (i) before rosette (ii) and stem elongation for each sowing date and genotype, respectively.

In this study, experimental plots consisted of six 6 m lines with 30 cm spacing between the lines, with two lateral lines considered as margins. The inter-plant spacing in each line was also 5 cm. It should be noted that the distance between the blocks and main plots in each block were 7 and 2.4 m, respectively.

In order to determine physical and chemical properties of the soil at the experiment site, soil samples were randomly taken at 0-30 and 30-60 cm depths. The experiment field soil was clay loam (Table 1). Fertilization was performed according to the results of soil analysis and rapeseed fertilizer recommendations. Accordingly, 50 kg·ha^-1 nitrogen and 70 kg·ha^-1 P₂O₅ from two sources of urea and ammonium phosphate and 10 kg·ha^-1 K₂O (as pre-plant) and 46 kg·ha^-1 nitrogen (equivalent to 100 kg·ha^-1 urea) were used at the beginning of the stem elongation stage. The weeds were controlled by applying 2.5 liters per hectare of trifluralin before sowing as well as hand weeding during the rapeseed growth period. Finally, the seed of rapeseed genotypes was sown on the mentioned sowing dates. Irrigation was carried out throughout the growing period based on 80 mm evaporation from a Class A evaporation pan.

Rapeseed seed yield was determined by hand-harvesting 3.6 m² in the center of each plot and weighed using a precise scale. It should be noted that the final harvest for each plot over two years was performed when 50% of the seeds in the main siliques and primary branches turned brown. In order to measure and determine the oil content in the rapeseed, three grams of seeds were selected from each experimental plot and the oil content was measured using a Nuclear Magnetic Resonance (NMR) German Broker Brand minispec mq20 model according to the international standard ISON.5511. For this purpose, the calibration of the device was performed with a reference sample and the product calibration with pre-prepared standard samples, and then three grams of rapeseed seeds were transferred to the device-specific cell. The cell containing the specimen was placed at a specific location and the amount of oil was monitored in less than 1 minute. After determining the oil content in the rapeseed seed, oil yield was obtained by multiplying the seed yield by oil content.

In the present study, rapeseed seed oil samples were extracted according to the method proposed by Azadmard-Damirchi et al. (2005) and Fathi-Achachlouei and Azadmard-Damirchi (2009). In summary, rapeseed seed samples (approximately 10 g) with 30 ml hexane/isopropanol (3:2, v/v) were kept at room temperature for one hour in metal tubes containing four metal balloons to facilitate homogeneity. In order to extract oil from 100 g of rapeseed seed, 10 tubes were used, each tube containing 10 g of rapeseed seed. After one hour of shaking the samples, the extract was filtered through filter paper with a Buchner funnel under vacuum. The remained defatted cake was washed twice with the same solution to extract the entire potential oil content. Then, 35 ml of sodium sulfate was added to the solvent containing the oil and thoroughly mixed. Subsequently, the organic solvent layers containing the oil were separated and evaporated at 35 °C under reduced pressure. Finally, the extracted oil was stored at -20 °C for subsequent analyses.

Fatty acid methyl esters were prepared from oil samples according to the method proposed by Savage et al. (1997). Briefly, 2 ml of 0.01 M NaOH solution were added to a tube containing the oil sample (about 10 mg) dissolved in 0.5 M hexane and then placed in a water bath at 60 °C for 10 minutes. Thereafter, a boron trifluoride solution in methanol (20% of boron trifluoride in methanol) was added, and the samples were kept in a 60 °C water bath for 10 minutes. The sample was cooled under running water, and 2 ml of sodium chloride 20% and 1 ml of hexane were added. After complete mixing, the hexane layer containing fatty acid methyl esters was separated by centrifugation.

Fatty acid methyl esters were analyzed using gas chromatography (GC) based on the method proposed by Azadmard-Damirchi and Dutta (2006). The GC instrument was equipped with a flame ionization detector and a Split/Splitless injector. The injector and detector temperatures were 230 and 250 °C, respectively. The oven conditions were such that from 158 to 220 °C, the temperature rose to 2 °C per minute and was maintained at each temperature for five minutes. Helium was used as carrier gas and nitrogen was used as an auxiliary gas at a flow rate of 30 ml per minute. Fatty acid methyl esters were identified by comparing their shelf-life with standard fatty acid methyl esters, and peak areas were reported as the percentage of total fatty acids. In this study, palmitic, stearic, oleic, erucic, linoleic and linolenic fatty acids were measured.

Rapeseed seed oil glucosinolate content was measured using a Varian Spectrophotometer Cary 100 equipped with a 50-m-long CP-Sil 88 capillary, 0.25 mm inner diameter, and 0.2 μm static phase thickness (Harinder et al., 2007).

A combined analysis of variance was performed using SAS software (version 9.2). The mean comparison was performed by the least significant difference (LSD) at 5% probability level, and interactions among treatments were determined with by-processing.

The results of mean comparison showed that the foliar application of selenium + zinc produced the highest seed yield with an average of 3692 kg·ha^-1 and had 17.7, 7.6, and 11.2% higher seed yield compared to the control, selenium, and zinc treatments, respectively (Figure 2).

The foliar application of selenium, zinc, and selenium + zinc increased the seed yield of rapeseed on three sowing dates, indicating their desired effect on increasing the photosynthesis rate of rapeseed and seed yield (Shoja et al., 2018). Sieprawska et al. (2015) reported that the selenium element is important due to activating the antioxidant system in plant cells against environmental stresses such as drought, ultraviolet rays, and high/low temperatures. The formation of reactive oxygen species (ROS) is intensified when zinc deficiency has occurred; while the application of zinc provides protection against cold stress in plants (Cakmak, 2006). It is worth noting that the protective range of micronutrients such as selenium and zinc varies depending on the stress and genotype. Fang et al. (2008) in a study investigated the effect of foliar application of selenium, zinc, and iron and reported that rice seed yield increased significantly compared to the none-foliar application of micronutrients.

The interaction between sowing date and genotype showed that the rapeseed seed yield (on average across genotypes) was reduced by 45.7 and 52.5% on the sowing dates of 17 and 27 October (delayed sowing dates) compared to 7 October (well-timed sowing date), respectively (Table 5). GK-Gabriella and Okapi on 7 October with the averages of 5355 and 5273 kg·ha^-1, GKH2624 and GK-Gabriella on 17 October with the averages of 3068 and 2936 kg·ha^-1, and GKH2624, GK-Gabriella, and Okapi on 27 October with the averages of 2682, 2567, and 2504 kg·ha^-1 were known as the superior genotypes, respectively (Table 5).

As previously mentioned, delayed sowing (on average across genotype and foliar application treatments) reduced rapeseed seed yield. Delayed sowing led to a more rapid development of the crop, a reduction in the number of days from emergence to flowering and maturity, and the duration of flowering and seed filling, which has a negative impact on final performance and reduces seed yield (Faraji et al., 2009). The higher seed yield due to optimum sowing date can be attributed to better plant growth conditions which help the plant grow faster and reach the rosette stage before the winter frost. On the other hand, timely sowing reduces drought and heat stresses at the end of the season and increases seed yield (Turner 2004; Gunasekera et al., 2006). Obviously, seed yield losses are greater when the delay in sowing is longer. Moradi Aghdam et al. (2019) reported that delayed sowing of rapeseed cultivars on 1 November compared to the well-timed sowing date on 7 October reduced seed yield by 38.4%. Faraji et al. (2009), in a study evaluating different sowing dates including 9 November (optimum sowing date), 6 December, 5 January, 4 February, and 6 March and concluded that the sowing of rapeseed genotypes after optimum sowing date reduced rapeseed seed yield by 18, 19.6, 31.3, and 77.5%, respectively.

As shown in Table 4, the oil content in rapeseed (averaged across foliar application treatments) was reduced by delayed sowing dates (17 and 27 October) compared to well-timed sowing date (7 October). The average across foliar application treatments, with rapeseed plants sown on 7 October had the highest oil content with an average of 41.97%; while the oil contents in rapeseed plants sown on 17 and 27 October were 39.39 and 35.92%, respectively (Table 4). The slicing interaction of sowing date × foliar application showed that selenium + zinc treatment on three sowing dates 7, 17, and 27 October produced the highest oil content with averages of 42.89, 40.23, and 37.77%, respectively.

The oil contents in the studied genotypes were reduced when rapeseed plants sown on 17 and 27 October were compared to 7 October. The highest oil content (42.4%) on 7 October was observed in the GK-Gabriella genotype and the lowest oil content (35.45%) on 27 October was observed in SW102 genotype (Table 5). The average seed oil content of rapeseed across genotypes on 7 October (41.94%) was higher than those of 17 and 27 October (39.34 and 35.92%, respectively) (Table 5).

Obviously, quality and quantity yields are determined by genetic and environmental factors as well as management practices (Ashrafi and Razmjoo, 2010; Eyni-Nargeseh et al., 2020). In general, genetic factors are important and effective parameters for crop oil content and environmental factors have little effect on this trait (Robertson and Holland, 2004).

In the present study, the sowing dates and foliar application of micronutrients as different environmental and management factors and genotypes as a genetic factor were investigated and a significant difference was observed between the levels of each treatment in terms of oil content. In this regard, the results of Eyni-Nargeseh et al. (2020) showed a statistically significant difference among 17 rapeseed genotypes in terms of oil content. Nazeri et al. (2018) evaluated the response of rapeseed genotypes to a delay in sowing date and reported that the oil content in genotypes was significantly different and the delayed sowing of rapeseed reduced the oil contents in rapeseed genotypes. The application of zinc can increase oil content and yield due to an increase in auxin biosynthesis, chlorophyll content, nitrogen, and phosphorus uptake as well as a reduction in sodium concentration in plant tissues (Moinuddin and Imas, 2008). The results of Shoja et al. (2018) showed that the seed oil content in rapeseed was increased under the foliar application of zinc compared to the control treatment (non-foliar application). The combined application of micronutrients can lead to more useful results. Eskandari Zanjani et al. (2012) concluded that the combined use of selenium and zeolite resulted in increased oil content and yield in rapeseed.

The selenium+zinc treatment had the highest seed oil yield with an average of 1512 kg·ha^-1 and produced 21.8, 12, and 11.2% more seed oil compared to the control, zinc, and selenium treatments, respectively (Figure 2). Increased seed oil yield under the foliar application of micronutrients can be the result of further assimilation during the seed filling period, which is associated with an increased photosynthesis of siliques (Shahbaz et al., 2018). Zaman Fashami et al. (2018) investigated the response of different rapeseed cultivars to a foliar application of selenium and reported that the oil yield of rapeseed cultivars under selenium foliar application conditions was 5.8% greater than the control treatment (no foliar application). Shahsavari (2019) studied the effects of zeolite and zinc on the quantitative and qualitative yields of rapeseed and showed that although the foliar application of zinc increased the seed oil yield of rapeseed, the combined use of zeolite and zinc produced even greater seed oil yield.

With respect to the average across sowing dates, the highest seed oil yield was related to 7 October with the averages of 1942.75 and 2257.43 kg·ha^-1 in 2014-2015 and 2015-2016 growing seasons, respectively, and the sowing dates of 17 and 27 October showed a significant yield reduction compared to 7 October in terms of oil yield (Table 6). The results of the slicing interaction between sowing date and genotype showed that the seed oil yield varied from 711 kg·ha^-1 for the SW102 genotype on 27 October to 2275 kg ha^-1 for the GK-Gabriella genotype on 7 October. When averaged by genotypes, the sowing of rapeseed plants on 17 and 27 October compared to 7 October resulted in 49 and 52% reductions in seed oil yield (Table 5). The difference in the oil yield by genotype is due to the unique characteristics of each genotype which give different yields (Naseri et al., 2012; Eyni-Nargeseh et al., 2020). Oil yield differences between genotypes have been reported for rapeseed by Safavi Fard et al. (2018). The reduction in seed oil yield under delayed sowing dates is probably due to higher temperatures during the seed filling period, which can reduce seed size and consequently oil yield (Nazeri et al., 2018). Of course, the seed oil yield is directly affected by seed yield and oil content; the reasons for which were earlier discussed. The reduction in rapeseed oil yield due to delayed sowing dates has been reported by Adamsen and Coffelt (2005) and Moradi Aghdam et al. (2019).

After foliar application treatments, the average palmitic acid contents were 4.17, 5.49, and 4.95% on 7, 17, and 27 October, respectively (Table 4). The results showed that the foliar application of zinc resulted in the highest palmitic acid contents on three sowing dates of 7, 17, and 27 October with averages of 4.49, 5.87, and 5.45%, respectively (Table 4); whereas the foliar application of selenium produced the lowest palmitic acid contents with averages of 3.87, 5.13, and 4.48% under such conditions.

Across genotypes, the average palmitic acid contents were 3.97, 5.46, and 4.92% on 7, 17, and 27 October, respectively (Table 5). The results showed that the highest palmitic acid content on 7 October belonged to the GKH2624 genotype (4.43%) followed by GK-Gabriella (4.36%) and Okapi (4.29%) genotypes (Table 5). On the sowing date of 17 October, the GK-Gabriella and Okapi genotypes had the highest palmitic acid content with an average of 5.6%. The highest content in palmitic acid on 27 October belonged to the GK-Gabriella and Okapi genotypes with an average of 5%.

The contents in saturated and unsaturated fatty acids depend on different factors such as genotype, management, and environmental conditions and these compounds may be increased or reduced under different conditions (Eyni-Nargeseh et al., 2020; Safavi Fard et al., 2018). The results of Eyni-Nargeseh et al. (2020) showed that the palmitic acid content in 17 rapeseed genotypes was significantly different due to genetic differences among genotypes. As one of the management practices, sowing date can have different effects on fatty acid composition (Maradi Aghdam et al., 2019; Nazeri et al., 2018). Long periods of high temperatures during the seed filling stage can produce seeds with low oil content and quality (Omidi et al., 2010). In one study, Turhan et al. (2011) investigated the effect of sowing dates (10, 20, and 30 October, and 10 November) on rapeseed fatty acid compositions and reported that delayed sowing dates reduced palmitic acid content. Also, the composition of rapeseed fatty acids is significantly affected by fertilizer management such as micronutrients (Shoja et al., 2018). The study results of Bybordi and Mamedov (2010) showed that the application of micronutrients (zinc and iron) increased the content in saturated fatty acids. Zaman Fashami et al. (2018) analysed the effect of foliar application of selenium on the quantitative and qualitative yields of different rapeseed cultivars and concluded that palmitic acid content showed an increasing trend under foliar application treatments.

As shown in Table 4, linoleic acid varied from 19.21% on 7 October when treated with selenium + zinc to 13.02% on 27 October in the control treatment. After foliar application treatments, the linoleic acid content was 18.70% on the sowing date of 7 October and reduced by 16.7 and 25.2% on 17 and 27 October, respectively (Table 4). Shoja et al. (2018) investigated the effect of the foliar application of micronutrients on the quantitative and qualitative properties of rapeseed seed oil and showed that the lowest linoleic acid content was related to a non-foliar application treatment (41.41 mg·g^-1); while the highest linoleic acid content was detected in the foliar application of zinc + boron (55.55 mg·g^-1).

The highest content in linoleic acid among the studied sowing dates in both years of the experiment was detected on 7 October (Table 6). When compared to 17 and 27 October, the linoleic acid content was 20 and 28.1% higher on 7 October in the 2014-2015 growing season, and 13 and 21.8% higher in the 2015-2016 growing season, respectively. The results showed that the linoleic acid content of the studied genotypes was reduced with delay in sowing so that the amounts of this trait on 17 and 27 October were 16.6 and 25.2% less than 7 October (Table 5).

Linoleic acid content ranged from 19.1% for the GK-Gabriella genotype on 7 October to 13.5% for the SW102 genotype on 27 October (Table 5). On the sowing dates of 17 and 27 October, the highest linoleic acid content belonged to the GKH2624 genotype with the averages of 16 and 14.3%, respectively. In separate studies, Eyni-Nargeseh et al. (2020) and Nazeri et al. (2018) stated that different genotypes showed statistically significant differences in terms of linoleic acid content. The results of Turhan et al. (2011) showed that delay in sowing significantly reduced the linoleic acid content. Moradi Aghdam et al. (2019) evaluated the response of different rapeseed cultivars to delayed sowing and concluded that linoleic acid content was reduced by 15.2% due to delayed sowing date (1 November) compared to the well-timed sowing date (7 October).

After the foliar application treatments, the average linolenic acid contents were 4.3, 6.8, and 5.87% on 7, 17, and 27 October, respectively (Table 4). It should be noted that the foliar application of zinc on the three sowing dates (7, 17, and 27 October) produced the highest linolenic acid contents (4.5, 6.72, and 6.30%, respectively) (Table 4). Zaman Fashami et al. (2018) concluded that the foliar application of selenium significantly increased the linolenic acid contents in different rapeseed cultivars. Shoja et al. (2018) reported that the amount of linolenic acid in rapeseed was significantly increased when zinc, boron, and sulphur were applied compared to the non-foliar application treatment.

The results showed that in the 2014-2015 and 2015-2016 growing seasons, the sowing date of 17 October showed the highest linolenic acid contents with the averages of 6.22 and 6.73%, respectively (Table 6). The average linolenic acid contents across genotypes were 4.3, 6.42, and 5.82% on 7, 17, and 27 October, respectively (Table 5). The difference between genotypes in terms of linolenic acid content has been reported by Safavi Fard et al. (2018) and Eyni-Nargeseh et al. (2020). Omidi et al. (2010) studied rapeseed fatty acid compounds as affected by environmental, genotype, and agronomic parameters and found that the linolenic acid content of the studied genotypes showed an increasing trend due to delayed sowing. The results of Moradi Aghdam et al. (2019) showed a 28.8% increase in linolenic acid content as a result of delayed sowing dates. Nazeri et al. (2018) also investigated the response of rapeseed genotypes to two sowing dates including 7 October (well-timed sowing date) and 17 October (delayed sowing date) and showed that linolenic acid content increased by 30.7% with delayed sowing date.

The oleic acid content varied from 63.84% on 7 October and treated with selenium + zinc treatment to 50.98% on 27 October for the non-foliar application treatment (control) (Table 4). the average oleic acid content across foliar application treatments on 7 October was 62.45% and was reduced by 4.4 and 13.8% on 17 and 27 October, respectively (Table 4). In a study, Hashem et al. (2013) showed the protective role of selenium foliar application on rapeseed and reported that oleic acid content increased when selenium was applied compared to non-foliar application. The study results of Shoja et al. (2018) showed that the foliar application of micronutrients including boron, zinc, and sulphur resulted in increased oleic acid content. Their results also showed that the combined application of these micronutrients had more beneficial effects.

The results illustrated that the rapeseed plants sown on 7 October had the highest amount of oleic acid with the averages of 61.52 and 63.39% in the 2014-2015 and 2015-2016 growing seasons, respectively (Table 6). Averaged by genotypes, the highest oleic acid content was detected on 7 October with an average of 62.4%, and was reduced by 4.4 and 13.8% on 17 and 27 October, respectively (Table 5). The difference in the studied genotypes in terms of oleic acid content is related to genetic differences. Farahani et al. (2019) and Eyni-Nargeseh et al. (2020) reported differences among genotypes in terms of oleic acid content in separate studies. When compared to 17 and 27 October, higher oleic acid content on 7 October could be attributed to cooler temperatures at the seed filling stage (Fayyaz-UI-Hassan et al., 2005). Nazeri et al. (2018) investigated the response of different rapeseed cultivars to well-timed (7 October) and delayed (27 October) sowing dates and found that oleic acid content on 7 October was 2.1% more than 27 October. Moradi Aghdam et al. (2019) showed in another study that the oleic acid content of different rapeseed genotypes was significantly reduced with a delay in sowing.

Average erucic acid contents across foliar application treatments were 0.175, 0.418, and 0.381% on 7, 17, and 27 October, respectively (Table 4). Erucic acid is one of the most important fatty acids for rapeseed oil and its edible use, which is needed in small quantities (Gecgel et al., 2007), and its content should not exceed 2% (Moradi Aghdam et al., 2019). The results revealed that the foliar application of selenium provided the minimum erucic acid contents on 7, 17, and 27 October with averages of 0.156, 0.399, and 0.348%, respectively (Table 4). Shahsavari et al. (2014) investigated the effect of the foliar application of zinc and zeolite on the qualitative yield and composition of the fatty acids in rapeseed and found that the erucic acid content was significantly affected by foliar application and increased when zinc and zeolite were applied compared to a non-foliar application treatment. Shoja et al. (2018) investigated the response of rapeseed fatty acid compounds to foliar application of micronutrients (sulfur, boron, and zinc) and concluded that it led to a decrease in erucic acid content compared the non-foliar application treatment.

The results showed that the highest erucic acid content was obtained on 17 October with averages of 0.413 and 0.428% in 2014-2015 and 2015-2016 growing seasons, respectively (Table 6). Averaged by genotypes, rapeseed plants sown on 7 October had the minimum erucic acid content with an average of 0.175%, while erucic acid contents were 0.422 and 0.383% on sowing dates of 17 and 27 October, respectively (Table 5). Plant growth and development as well as oil quality and quantity in rapeseed are affected if maximum daily temperatures are greater than 27 ˚C (Morrison and Stewart, 2002). When rapeseed is sown late in the fall, long periods of high temperatures during the seed filling stage can produce seeds with low oil content and quality (Omidi et al., 2010; Nazeri et al., 2018). The study results of Samadzadeh Ghale Joughi et al. (2018) showed that the delayed sowing of rapeseed increased erucic acid contents in all studied genotypes. The difference in erucic acid content was due to genetic differences between genotypes, and in agreement with the current study results, Moradi Aghdam et al. (2019) reported that studied cultivars were significantly different in terms of erucic acid content.

After foliar application treatments, the average glucosinolate content was 8.87 μmol·g^-1 on 7 October; while the glucosinolate content for rapeseed plants sown on 17 and 27 October were 15.17 and 13.43 μmol g^-1, respectively (Table 4). It should be noted that for all treatments the glucosinolate content did not exceed the standard level for edible consumption (30 μmol·g^-1). Increasing glucosinolate content resulted in a reduction in the quality and nutritional value of rapeseed oil (Sulisbury et al., 1987), which is affected by hereditary and environmental factors (Fieldsend et al., 1991). The results of Shahsavari et al. (2014) in evaluating the quantitative and qualitative responses of rapeseed to foliar application of zeolite and zinc showed that glucosinolate content was significantly reduced when zeolite and zinc were applied compared to non-foliar application.

The highest glucosinolate content was observed on 17 October with averages of 15.23 and 15.11 μmol·g^-1 in the 2014-2015 and 2015-2016 growing seasons, respectively (Table 6). Averaged across genotypes, glucosinolate content was 8.84 μmol g^-1 on 7 October, and increased by 71 and 51.4% on 17 and 27 October, respectively (Table 5). Glucosinolate content is dependent on environmental and genetic factors (Fieldsend et al., 1991) and its content might vary between genotypes (Farahani et al., 2019). The difference in glucosinolate content due to genetic differences between cultivars has been reported by Li et al. (2019) and Safavi Fard et al. (2018). A delay in sowing date also has a significant effect on glucosinolate content in rapeseed as an environmental factor. In this regard, Samadzadeh Ghale Joughi et al. (2018) investigated the effect of sowing date on the qualitative yield of different rapeseed cultivars and found that a 10-day delay in sowing resulted in a 69% increase in the seed glucosinolate content in rapeseed. Increased glucosinolate content in rapeseed cultivars with delayed sowing dates has also been reported by Moradi Aghdam et al. (2019).