Drought Effects on Seed Yield Stability and Oil Quality Traits in Different Rapeseed Genotypes: Toward Adaptive Sustainability of Crops in Semi-Arid Regions

Abstract

Rapeseed (Brassica napus L.) is a major oilseed crop worldwide, particularly valued for its high yield potential, favorable fatty acid composition, and its bioactive compounds that improve nutritional and industrial quality. However, its productivity and oil quality are increasingly compromised by climate change-induced water scarcity, particularly in semi-arid Mediterranean regions. In this study, the performance of 17 genotypes was evaluated under well-watered (irrigated) and rainfed (stressful) conditions across two contrasting locations, Douyet Experimental Station (DYT) and Ecole Nationale d’Agriculture de Meknès (ENAM), during the 2023/2024 growing season. The assessment concerned seed yield, oil traits, and nutraceutical quality. The results obtained show that drought stress significantly reduced seed yield by about 26% (from 2.29 to 1.69 t ha⁻¹) and decreased oil content by about 8.3% (from 41.1% to 37.7%). The highest reductions were observed for sensitive genotypes, particularly ‘IND23’, ‘IND82’, and ‘H2M-5’, while ‘Moufida’, ‘Nap9’, and ‘TP2’ maintained seed and oil yield above the overall average across both water regimes. Drought also impacted the accumulation of secondary metabolites, with mean total phenolic content increasing from 5.41 to 9.98 mg GAE g⁻¹ (+84.5%) and total flavonoid content rising from 25.25 to 34.93 mg QE g⁻¹ (+38.3%) under rainfed conditions, accompanied by marked increases in antioxidant activity (DPPH, ABTS), particularly for ‘Moufida’, ‘Nap9’, ‘TP2’, and ‘Marina’. Oil quality indices remained within Codex Alimentarius standards, with a slight increase in acidity values accompanied by a decrease in peroxide values, thus suggesting improved oxidative stability. Principal component analysis grouped genotypes into distinct clusters, with an elite group (‘Moufida’, ‘Nap9’, and ‘TP2’) characterized by yield stability, accumulation of phenolics, and high antioxidant activity, thus confirming their strong adaptation to the dry conditions of the Mediterranean region. These drought-tolerant lines, with high agronomic performance combined with good nutritional quality, can be recommended as valuable cultivars for sustainable and high-quality rapeseed production in dry Mediterranean regions.

1. Introduction

Rapeseed (Brassica napus L.) is a globally important oilseed crop cultivated across diverse agroecological regions, ranging from temperate to semi-arid zones [1,2]. Its global production reached 68.9 million tonnes in 2024, covering approximately 33.7 million hectares [3]. Rapeseed contributes substantially to human as well as animal nutrition, owing to its high seed oil content and the important nutritional value of its protein-rich meal [4,5]. Its oil is prized for its favorable fatty acid profile with about 60% monounsaturated (oleic acid, C18:1), 20% polyunsaturated (linoleic and α-linolenic acids), and only ~7% saturated fatty acids, alongside a beneficial ω-6/ω-3 ratio, making it particularly advantageous for cardiovascular and metabolic health [6,7,8,9]. Additionally, the presence of sinapic acid derivatives and tocopherols enhances its oxidative stability and antioxidant capacity [10]. However, sustaining rapeseed productivity under semi-arid environments, such as those in the Mediterranean basin, including Morocco, is increasingly challenged by climate-induced drought events [11,12,13]. Drought stress during critical phenological stages, particularly flowering, disrupts physiological processes such as photosynthesis, stomatal conductance, and nutrient uptake, ultimately reducing seed yield, oil content, and oil quality [14,15]. Furthermore, drought alters fatty acid biosynthesis and impairs the production of antioxidant compounds, thereby degrading the oil’s nutritional value and oxidative stability [16,17].

Although advances in canola breeding, such as the development of high-oleic, low-linoleic (HOLL) varieties, have improved oil quality [18,19], the combined evaluation of drought-tolerance and oil quality remains little explored, especially under field conditions reflective of semi-arid climates. Brassica juncea and B. rapa, for example, have demonstrated superior tolerance to abiotic stress in warm and dry environments [20]; however, limited research has addressed the dual challenge of maintaining both yield stability and oil quality in B. napus under water-deficient conditions. To date, only a few integrated studies have evaluated B. napus genotypes under drought by simultaneously considering yield, oil quality, and biochemical indices in semi-arid field environments [9,21,22,23,24,25]. The detrimental impacts of water stress, particularly the reduction in biomass, floral fertility, pod and seed formation, and alterations in oil composition, are well established [26,27,28]. However, the interactions between genotype and water availability for the oil’s biochemical attributes remain insufficiently investigated.

In Morocco, over the last 20 years, many rapeseed germplasms and inbred lines have been developed, of which nine varieties with good agronomic performance, high oil content, and canola quality have been registered [13,29,30]. Very recently, it was reported that some of those inbred lines and varieties, including ‘Nap10’, ‘Nap9’, and ‘Baraka’, exhibited drought tolerance at early growth stages under in vitro [31] and controlled greenhouse conditions [32]. However, despite reported variation in drought responsiveness in controlled or early-stage evaluations, the oil quality of these materials has rarely been assessed under water deficits in multi-environment field conditions. Therefore, we selected a panel of 17 genotypes representative of the genetic and agronomic diversity currently used or targeted in Moroccan/Mediterranean semi-arid production systems (including released cultivars and advanced lines) [9] and evaluated their performance under two water regimes (irrigated vs. rainfed) across two contrasting environments. The assessment covered several parameters, including seed yield, oil content, total phenolic content (TPC), total flavonoid content (TFC), antioxidant activities (DPPH and ABTS), and lipid quality indices (acidity, peroxide, iodine, and refractive index). The subsequent objective of the study was to identify genotypes that combine drought tolerance with high oil quality, thereby supporting the development of cultivars adapted to dry environmental conditions and enhancing the nutritional value, sustainability, and utilization of this strategic oilseed crop.

2. Materials and Methods

2.1. Experimental Sites and Climatic Conditions

A field experiment was conducted during the 2023–2024 growing season at two distinct agro-climatic locations in Morocco under two contrasting water regimes (irrigated and rainfed). The first site is the INRA Douyet Experimental Station (DYT), located approximately 10 km from Fez city (34°04′ N, 5°07′ W; 416 m a.s.l.). The site receives an average annual rainfall of 350 mm and has a Mediterranean climate characterized by cold and Sirocco winds, wet winters, and hot, dry summers, with an average annual temperature of 18.5 °C [33]. The soil is classified as a vertisol with a texture composed of 29% clay, 58% silt, and 12% sand. The second site is the experimental farm of the National School of Agriculture of Meknes [34], situated approximately about 10 km away from Meknes city (33°50′ N, 5°28′ W; 906 m a.s.l.). This site presents a warm temperate climate with precipitation concentrated during the winter season. The mean annual rainfall is 509 mm, and the average annual temperature is 17.4 °C. The soil is moderately deep, with a clay loam texture, neutral pH, and a gentle slope.

2.2. Plant Material

Seventeen rapeseed genotypes were evaluated, including six pure lines, two synthetic varieties registered in Morocco, three elite breeding lines, two accessions from Australia, two accessions from Pakistan, and two commercial hybrid cultivars from Germany, also registered in Morocco. The selected genotypes differ in plant architecture, phenological behavior (early to late flowering and maturity), branching pattern, leaf and silique morphology, biomass production, and reported yield stability under contrasting water availability, which are traits commonly associated with drought adaptation. A detailed description of genotypes, including origin, registration year, and key morphological and agronomic traits, is presented in

Table 1. Origin and description of rapeseed genotypes used in this study.

Genotypes	Type/Genetic Structure	Origin	Registration Year	Main Attributes
Narjisse	Pure line variety	Morocco	2008	Tall plant featuring fully grown leaves, robust lobes, and a bright green leaf surface; early-to-mid spring flowering and well-developed branches [24].
Moufida	Pure line variety	Morocco	2009	The plant has a moderate height with deeply lobed, green, serrated leaves; elongated siliques and anthem-yellow flowers [9].
Alia	Pure line variety	Morocco	2017	A tall plant with underdeveloped leaves, robust appendages, and a small, vibrant green blade [9].
Adila	Synthetic variety	Morocco	2015	The leaves have an intermediate green blade, robust margins, and moderate serration; moderate stem height. The primary stem attains a moderate stature. Long beaks characterize siliques, which present pale yellow flowers [24].
Lila	Synthetic variety	Morocco	2015	Moderately developed with robust lobes and moderate serration; intermediate green blade; moderate stem height; long-beaked siliques and pale-yellow flowers [24].
Baraka	Pure line variety	Morocco	2018	Large plant with light green leaves, sturdy appendages, late flowering, and abundant branching [9].
Marina	Pure line variety	Morocco	2023	A moderately tall plant with light-green, lobed leaves, early flowering, and high seed yield [31].
Redana	Pure line variety	Morocco	2023	A tall plant with dark-green, lobed leaves, early flowering, and thin siliques prone to shattering but with high seed yield [31].
CZKN	Inbred line	Morocco	2025	A moderately tall plant with deep-green, lobed leaves, early flowering, thick siliques resistant to shattering, and a high seed yield [31].
H2M-5	Mutant line	Morocco	Not yet	Large leaves, a strong root system, early flowering, and elevated oleic acid content [24].
TP2	Mutant line	Morocco	Not yet	A late-flowering line with large leaves and high biomass and seed yield.
Nap9	Pure line variety	Australia	In progress	This short plant has dark green, underdeveloped foliage, with early flowering and maturation, demonstrating strong drought adaptation [31].
Nap10	Pure line variety	Australia	In progress	The plant has dark green leaves, early flowering, good drought adaptation at the germination stage, and a satisfactory yield [31].
IND23	Pure line variety	Pakistan	Not yet	A moderately tall plant with light-green, lobed leaves, a slender stem, and a significant reduction in height under stress; early flowering; moderate seed yield with a high thousand-seed weight.
IND82	Pure line variety	Pakistan	Not yet	Medium-sized plant with dark-green, serrated leaves, a compact structure, and few branches. It produces a high thousand-seed weight, flowers early, develops elongated siliques, and has low pod production under stress conditions.
Traper	Hybrid Variety	Germany	2014	Dark green, sparse foliage; early flowering and rapid maturity; maintains a satisfactory yield under drought conditions [35]
Solar	Hybrid Variety	Germany	2020	A semi-early and Clearfield-type variety, with high biomass and seed yield [36].

.

2.3. Experimental Design and Water Regimes

A completely randomized design (CRD) with three replications was used at each site. Sowing was performed manually in the second week of November 2023, following regional technical recommendations for rapeseed cultivation. Plots consisted of 3 m-long rows spaced 0.6 m apart. A basal fertilizer of 60 kg ha⁻¹ nitrogen (N), 80 kg ha⁻¹ phosphorus (P₂O₅), and 80 kg ha⁻¹ potassium (K₂O) was applied before sowing. Thinning was performed at the two-leaf stage to ensure a uniform interplant distance of 4–5 cm. Manual weeding was carried out at the rosette stage, and insecticide treatments were applied based on pest monitoring thresholds. A topdressing of 60 kg ha⁻¹ nitrogen was split equally between the rosette and flowering stages.

A drip irrigation system was used to manage water delivery, consisting of 16 mm-diameter polyethylene pipes fitted with emitters rated at 2 L/h. Field performance measurements showed the emitters maintained an average flow rate of 1.6 L/h over a continuous two-hour irrigation period, with a calculated uniformity coefficient (UC) of 92.81%. Irrigated plots were watered twice per week, and the irrigation volume was determined based on daily weather data. The amount of water required was calculated as the difference between crop evapotranspiration (ETc) and effective rainfall to offset moisture losses. ETc (mm/day) was estimated using the FAO Penman–Monteith method [37], according to the equation ETc = ET₀ × Kc, where ET₀ is the reference evapotranspiration, and Kc is the crop coefficient. In contrast, rainfed plots received no supplemental irrigation and relied exclusively on natural rainfall throughout the growing season. All plots were mechanically harvested at full physiological maturity in late June, and grain yield was measured after adjusting seed moisture content to 10%.

2.4. Seed and Oil Yields

After harvest, seed yield (kg ha⁻¹) was determined for each genotype under both water regimes. Grain moisture content was adjusted to 10% before yield calculation, as described by Channaoui et al. [24]. Oil extraction was carried out using a semi-automated solvent extractor (SER 148/6, VELP Scientifica, Usmate Velate, Italy) operating according to the Randall (hot solvent extraction/Soxtec) technique (immersion–rinsing–recovery). For each sample, 20 g of crushed seeds were extracted with 150 mL of n-hexane for 3 h using a Soxhlet apparatus under reflux at atmospheric pressure. The heating mantle set-point (130 °C) was adjusted to maintain a constant reflux rate at the hexane boiling point (≈69 °C). The solvent was removed by evaporation at 40 °C using a rotary evaporator (HAHNVAPOR, HAHNSHIN, Gimpo, Republic of Korea).

Oil content (%) was calculated using the following equation:

Oil content (%) = [(M₁ − M₀)/M₂] × 100

(1)

where

• M₀ = mass of the empty flask (g)
• M₁ = mass of the flask after solvent evaporation (g)
• M₂ = initial mass of the seed sample (g)

Extracted oil samples were stored at 4 °C in the dark until further physicochemical and biochemical analyses. Oil yield (kg ha⁻¹) was calculated by multiplying oil content (%) by seed yield (kg ha⁻¹).

2.5. Drought Tolerance and Yield Stability Indices

Drought tolerance indices were employed to discriminate the canola genotypes evaluated under well-watered and drought conditions. Two widely used indices were calculated to assess yield stability and drought tolerance according to the following formulas:

• Stress Tolerance Index (STI) [38]:

STI = (Yp × Ys)/(Ȳp)²

(2)

• Yield Stability Index (YSI) [39]:

YSI = Ys/Yp

(3)

where Ys is the seed yield under drought stress conditions, Yp is the seed yield under well-watered conditions, and Ȳp represents the mean seed yield of all genotypes under non-stress conditions.

2.6. Physicochemical Analysis

The physicochemical properties of rapeseed oil were determined following international standards, as presented by Guirrou et al. [9]. The acidity index (mg KOH/g oil) was measured by titration of an oil–ethanol solution with ethanolic KOH, according to NF EN ISO 660:2009 [40]. The peroxide value (mEq O₂/kg oil) was assessed by iodine titration, following NF ISO 3960:2004 [41]. The iodine value (g I₂/100 g oil) was determined using ISO 3961:2018 [42], in which oil samples were reacted with Wij’s reagent, and excess iodine was titrated with sodium thiosulfate. The refractive index at 20 °C was measured with an Abbe refractometer (BK-R2S, BIOBASE, Jinan, China), following ISO 6320:2017 [43].

2.7. Biochemical Analysis

2.7.1. Phenolic Extraction

Phenolic compounds were extracted following the protocol of Tsimidou et al. [44], with slight modifications. Briefly, 2 g of oil were dissolved in 10 mL of n-hexane and mixed with 4 mL of 60% (v/v) methanol. The mixture was vortexed and kept under continuous agitation for 2 h at room temperature in the dark. After extraction, the samples were filtered through Whatman No. 1 filter paper. All extractions were performed in triplicate. The filtrates obtained from the three extractions were pooled, washed with 10 mL of hexane to remove residual lipids, and stored at 4 °C in amber vials until analysis.

2.7.2. Total Phenolic Content

The total phenolic content (TPC) was determined using the Folin–Ciocalteu method [45]. A total of 50 μL of extract was mixed with 3.0 mL of distilled water, 250 μL Folin–Ciocalteu reagent, and 750 μL of 7% sodium carbonate. After 8 min, 950 μL of distilled water was added, and the solution was incubated in the dark for 2 h. Absorbance was measured at 765 nm using a UV–Vis spectrophotometer (V-530, Jasco GmbH, Pfungstadt, Germany). Results were expressed as mg gallic acid equivalents per 100 g of oil (mg GAE/100 g).

2.7.3. Total Flavonoid Content

The total phenolic content (TFC) was determined using the method of Favati et al. [46], with modifications. An extract aliquot 0.5 mL was mixed with 0.5 mL of 2% aluminum chloride solution in methanol. After 15 min of incubation at room temperature, absorbance was recorded at 430 nm using methanol as a blank. Quercetin was used for the standard curve, and results were expressed as mg quercetin equivalents per 100 g of oil (mg QE/100 g).

2.7.4. Free-Radical Scavenging Activity (FRSA)

Antioxidant activity was assessed using both DPPH and ABTS^•+ assays, as originally described by Brand-Williams et al. [47] and Re et al. [48], respectively, with slight modifications.

• DPPH assay: A total of 50 μL of oil extract was mixed with 950 μL of 0.030 mg/mL DPPH methanolic solution. After 60 min of dark incubation at room temperature, absorbance was measured at 515 nm. The percentage of inhibition was calculated as follows:

FRSA (%) = [(Ac − Ae)/Ac] × 100

(4)

where Ac and Ae represent the absorbance of the control and sample, respectively.

• ABTS assay: ABTS^•+ was prepared by mixing 7 mM ABTS with 2.45 mM potassium persulfate and incubating for 18 h in the dark. The resulting solution was diluted with methanol to an absorbance of 0.700 ± 0.05 at 734 nm. Then, 10 μL of extract or water (control) was added to 990 μL of ABTS^•+ solution and incubated in darkness for 30 min. Absorbance was read at 734 nm, and inhibition was calculated as follows:

FRSA (%) = [(Ac − Ae)/Ac] × 100

(5)

where Ac represents the absorbance of the control, and Ae represents the absorbance of the sample.

2.7.5. Quality Control and Data Expression

All extractions and colorimetric assays were performed in triplicate for each sample. Reagent blanks were included for each batch, and calibration curves were prepared using gallic acid (TPC) and quercetin (TFC) standards. Results are reported as “extractable” phenolics/flavonoids under the present extraction conditions (methanolic fraction after hexane defatting). Absolute recovery/extraction efficiency was not determined; therefore, values should be interpreted comparatively among genotypes and treatments rather than as absolute totals.

2.8. Statistical Analyses

All statistical analyses were performed using R software (version 4.3.2). The experiment was arranged in a completely randomized design (CRD) with three replicates. Data were analyzed using a three-way ANOVA with Genotype (G), Water regime (W), and Site (S) as fixed factors, including their interactions. Mean differences among irrigation regimes, canola genotypes, and their interactions were compared using the Tukey HSD test (p ≤ 0.05). Principal component analysis (PCA) was performed to classify genotypes and identify key traits contributing to drought tolerance and oil quality.

3. Results

3.1. Climatic Data Analysis

In both environments, meteorological variables were recorded continuously using an automated weather stations Figure 1 summarizes rainfall and temperatures registered during the 2023–2024 cropping season. Climatic conditions confirm that rapeseed genotypes at both experimental sites experienced substantial water limitation. At DYT, cumulative rainfall reached only 229.5 mm, with very low precipitation during sowing and emergence (6.2–17.2 mm in November–December), well below the 30–40 mm required for uniform establishment in Mediterranean environments [49]. Vegetative growth also occurred under moisture deficit, as January rainfall (44.9 mm) remained below the 60–80 mm/month typically needed for canopy development [50]. The most severe shortage occurred during bolting and the beginning of flowering (February), which received only 12.6 mm, a drought-sensitive phase in rapeseed development [51]. Also, the last developmental stages (April–May), corresponding to pod onset and seed-filling, received less than 18 mm/month combined with rising temperatures (>19–23 °C), thus generating pronounced late drought [52]. Nevertheless, during the full-flowering stage, occurring in March, the experimental trials were well-watered, as they registered 130.4 mm of rain in DYT and 112 mm in ENAM. On the other hand, ENAM recorded higher seasonal rainfall (296 mm compared to 229.5 mm at DYT), yet early-season precipitation remained limited (<50 mm), and late-season rainfall (<45 mm) was insufficient to meet crop water demand under progressively warmer conditions. Across both sites, cumulative rainfall remained below the 282–399 mm minimum reported for rapeseed in semi-arid regions [50] and under the 400–450 mm required for optimal spring rapeseed performance in Mediterranean climates [52].

3.2. Results of Analysis of Variance

The ANOVA results (

Table 2. Mean squares from the combined analysis of variance for the 17 rapeseed genotypes evaluated under two water regimes across two contrasting sites.

Source of Variation	DF	Seed Yield	Oil Content	Oil Yield	TPC	TFC	% DPPH	% ABTS	Acidity Index	Peroxide Index	Iodine Value	Refractive Index
Site (S)	1	13,798,721.25 ***	9.00	2,307,498.24 ***	66.28 ***	1152.35 ***	123.04 ***	30.93	0.45	0.63 ***	1286.05 ***	0.00002
Water regime (WR)	1	18,129,368.86 ***	607.77 ***	4,715,495.30 ***	1090.08 ***	4783.56 ***	5943.54 ***	9094.78 ***	5.08 ***	1.52 ***	5536.13 ***	0.001 ***
S × WR	1	841,084.22 **	12.88	122,695.92 **	5.83 ***	3.77	111.47 **	26.65	0.12	0.023 ***	376.95 ***	0.00001
Genotype (G)	16	1,058,636.30 ***	75.85 ***	282,649.65 ***	6.67 ***	196.66 ***	146.56 ***	69.33 ***	0.60 ***	0.44 ***	291.19 ***	0.0004 ***
G × WR	16	202,486.79 *	1.30	38,563.70 **	2.64 ***	18.69 ***	44.88 ***	53.52 ***	0.02	0.01 ***	15.82	0.000002
S × G	16	278,045.46 ***	7.02	48,204.67 ***	4.44 ***	68.46 ***	80.49 ***	47.17 ***	0.11	0.003	30.06	0.00004
S × WR × G	16	149,857.66	25.63 **	23,947.49	2.36 ***	84.12 ***	32.57 ***	24.94 ***	0.15	0.002	25.90	0.00003

DF: Degree of freedom. *: significant at 0.05. **: significant at 0.01. ***: significant at 0.001.

) showed that water regime, genotype, and their interaction significantly affected seed yield, oil content, oil yield, total phenolic content (TPC), total flavonoid content (TFC), antioxidant activities (%DPPH and %ABTS), acidity index, peroxide index, iodine value, and refractive index. Water regime was the main driver of the observed variation, explaining more than 90% for TPC, %DPPH, and %ABTS, 70–85% for oil content, TFC, iodine value, acidity index, peroxide index, and refractive index, and more than 53% for seed yield. The environment strongly impacted seed yield (≈41%) and oil yield (≈31%), and also affected TFC and iodine value (≈18%). On the other hand, Genotype contributed most to refractive index (~23%) and explained 4–10% variation in oil content, acidity index, iodine value, and seed yield. Interaction effects, though significant for several traits, accounted for only 0.2–4% of variation, indicating stable genotype ranking across environments and water regimes. These results highlight opportunities to select high-performing, stress-resilient cultivars that combine drought tolerance with superior oil quality, adapted to arid and semi-arid regions, and supporting sustainable production of this strategic oilseed.

Taken together, these findings demonstrate that all investigated traits were primarily affected by water regime (irrigated vs. non-irrigated, followed by environmental conditions (location) and, to a lesser extent, genotype. This multilevel variability provides a robust basis for identifying promising cultivars with stable performance and desirable seed and oil traits under contrasting water availability.

3.3. Seed and Oil Yields

Under irrigated conditions, the overall mean seed yield was 2289.14 kg ha⁻¹, with the highest yields obtained from ‘Redana’ (2759.03 kg ha⁻¹), ‘Solar’ (2699.58 kg ha⁻¹), and ‘Moufida’ (2686.22 kg ha⁻¹) while the lowest yields were recorded for ‘IND82’ (1560.30 kg ha⁻¹) and ‘H2M-5’ (1533.00 kg ha⁻¹). Under rainfed (stressful) conditions, the overall mean yield declined to 1692.92 kg ha⁻¹, with ‘TP2’ (2058.66 kg ha⁻¹), ‘Moufida’ (1972.86 kg ha⁻¹), and ‘Nap9’ (1960.13 kg ha⁻¹) being the highest-performing, whereas ‘Lila’ (1414.88 kg ha⁻¹), ‘IND23’ (1360.78 kg ha⁻¹) and ‘H2M-5’ (1359.03 kg ha⁻¹) exhibited the lowest values (Figure 2). Yield reductions under drought are generally attributed to impaired water uptake, photosynthesis, and assimilate partitioning, which shorten the seed-filling period and reduce biomass accumulation [14]. Despite these constraints, genotypes such as ‘TP2’, ‘Moufida’, and ‘Nap9’ consistently maintained the highest seed yields under rainfed conditions, confirming their superior performance and relative resilience. These findings are consistent with earlier studies that reported ‘Nap9’ and ‘Moufida’ as drought-tolerant at germination and seedling stages [12,31]. For irrigated conditions, the average yield (2.29 t ha⁻¹) was very close to the value reported by Anastasi and Scavo [53] (2.27 t ha⁻¹). However, under rainfed conditions, the mean seed yield recorded in the present study (1.69 t ha⁻¹) was higher than those reported by Delgado [54] (1.45 t ha⁻¹) and Morsi et al. [55] (1.54 t ha⁻¹). Overall, these comparisons demonstrate that Moroccan genotypes, particularly ‘TP2’, ‘Moufida’, and ‘Nap9’, maintain competitive yields and resilience under contrasting water regimes, thus supporting their potential for cultivation in semi-arid Mediterranean environments.

As indicated above, the environment strongly influenced seed yield, with a significant difference detected between DYT and ENAM (p < 0.05) (Table 2). While yields were generally higher at ENAM, genotypes such as ‘TP2’, ‘Moufida’, and ‘Nap9’ maintained stable performance with only moderate reductions, indicating adaptability to contrasting environments (

Table 3. Seed yield, biochemical, and oil quality traits mean values for 17 rapeseed genotypes evaluated across two contrasting environments (DYT and ENAM).

Genotypes	Seed Yield		Oil Content		Oil Yield		TPC		TFC		FRSA% (DPPH)		FRSA% (ABTS)		Acidity Index		Peroxide Index		Iodine Value		Refractive Index
	ENAM	DYT	ENAM	DYT	ENAM	DYT	ENAM	DYT	ENAM	DYT	ENAM	DYT	ENAM	DYT	ENAM	DYT	ENAM	DYT	ENAM	DYT	ENAM	DYT
Adila	2511.03 ab	1582.25 a–d	40.74 abc	39.81 ab	1022.80 a	631.82 a–e	8.42 ab	6.49 ab	35.25 a	32.41 a–d	38.83 ab	44.47 a	28.64 ab	27.37 b	2.38 ab	2.28 def	2.33 cde	2.19 d–g	103.01 abc	107.37 a	1.473 a	1.475 a
Alia	2261.75 ab	1662.6 a–d	38.90 abc	38.52 ab	876.97 ab	655.04 a–e	7.96 ab	8.18 a	34.09 a	27.10 b–e	35.53 ab	41.62 ab	29.65 ab	29.12 ab	2.42 ab	2.41 de	2.36 cde	2.21 def	96.97 abc	105.02 a	1.465 bcd	1.464 bcd
Baraka	2157.87 b–e	1968.33 abc	40.72 abc	40.52 ab	877.76 ab	809.30 a–d	8.63 ab	7.92 a	35.33 a	37.82 ab	39.75 ab	43.54 a	28.32 ab	33.44 a	2.48 ab	2.37 de	2.36 cde	2.21 def	101.74 abc	104.18 a	1.470 ab	1.464 a–d
CZ-KN	2307.22 ab	1628.90 a–d	40.37 abc	39.47 ab	940.20 ab	646.38 a–e	9.03 a	8.47 a	31.64 b	28.57 a–d	40.16 ab	42.38 ab	31.67 ab	25.87 b	2.57 ab	2.48 ab	2.41 cd	2.31 cd	95.81 bc	107.22 a	1.467 bc	1.462 bcd
H2M-5	1802.01 b	1090.02 d	37.81 abc	38.18 ab	682.40 b	415.37 e	8.05 b	5.82 b	32.10 ab	22.35 cde	31.21 b	38.64 b	35.21 a	31.24 ab	2.89 a	2.58 a	2.67 ab	2.5 ab	95.83 bc	98.08 a	1.459 cde	1.460 cd
IND23	1762.60 b	1246.80 cd	34.43 c	33.67 b	622.12 b	420.54 de	7.31 b	6.56 ab	28.02 b	18.59 e	38.69 ab	40.65 ab	27.30 ab	30.28 b	2.69 a	2.69 a	2.80 a	2.66 a	95.87 bc	98.29 a	1.458 de	1.460 cd
IND82	1672.77 b	1457.80 a–d	36.41 bc	33.50 b	607.22 b	486.92 cde	6.84 b	7.18 ab	24.39 b	20.36 de	40.39 ab	42.64 ab	31.08 ab	30.83 b	2.65 ab	2.45 ab	2.46 bcd	2.39 bc	93.89 bc	98.81 a	1.459 cde	1.458 cd
Lila	1891.25 b	1508.77 a–d	39.12 abc	40.87 ab	742.48 b	620.81 a–e	8.25 b	6.51 ab	32.37 ab	22.41 de	44.90 ab	43.82 a	26.69 b	33.48 a	2.27 ab	2.19 b	2.23 d–g	2.17 efg	99.54 abc	108.92 a	1.461 b–e	1.461 bcd
Marina	2252.08 ab	2002.33 abc	40.47 abc	42.02 a	909.46 ab	847.09 abc	9.17 ab	8.18 a	32.10 ab	32.15 abc	38.19 ab	43.44 a	28.81 ab	30.20 ab	2.19 ab	2.31 def	2.08 fg	2.00 gh	99.86 abc	106.61 a	1.462 b–e	1.459 cd
Moufida	2417.52 ab	2241.57 a	44.19 a	41.10 ab	1085.86 a	913.44 a	9.27 a	6.43 ab	36.58 a	37.48 a	48.21 a	43.66 ab	26.73 b	30.71 ab	1.18 b	2.13 b	2.01 g	1.93 g	106.47 ab	109.70 a	1.470 ab	1.471 ab
Nap10	2389.65 ab	1357.82 c	39.97 abc	37.82 ab	958.90 ab	514.03 b-e	9.32 a	4.94 b	36.64 a	21.26 de	36.26 ab	39.98 ab	24.68 b	27.19 b	2.73 a	2.56 a	2.49 bc	2.49 bc	92.19 bc	94.23 a	1.463 bcd	1.455 d
Nap9	2412.77 ab	2152.95 a	42.78 ab	43.87 a	1035.67 ab	947.44 a	9.19 ab	8.79 a	33.66 ab	36.18 ab	44.61 ab	44.45 a	29.49 ab	35.61 a	2.43 ab	2.11 b	2.23 d-g	2.13 fgh	111.39 a	110.93 a	1.471 ab	1.468 abc
Narjisse	1986.70 b	1704.87 a-d	37.27 abc	38.20 ab	744.55 b	657.46 a-e	8.27 ab	7.85 ab	28.72 b	28.74 a-d	31.30 b	39.56 ab	27.52 ab	28.11 b	2.45 ab	2.20 b	2.15 efg	2.04 gh	96.26 bc	100.88 a	1.455 de	1.458 cd
Redana	2733.43 a	1908.77 a-d	39.84 abc	38.95 ab	1097.09 a	750.20 a-e	8.38 ab	6.61 ab	36.19 a	36.09 ab	43.85 ab	44.64 a	37.53 a	30.77 ab	2.41 ab	2.29 def	2.32 c-f	2.18 efg	99.35 abc	104.28 a	1.467 bc	1.461 bcd
Solar	2479.60 a	2140.40 ab	38.28 abc	35.93 b	952.69 ab	781.19 a-e	7.60 b	7.48 ab	30.07 b	19.45 de	34.84 ab	32.18 b	23.85 b	27.45 b	2.92 a	2.67 a	2.53 bc	2.39 bc	89.09 c	97.63 a	1.451 e	1.455 d
TP2	2486.77 a	2290.58 a	41.69 abc	42.61 a	1037.20 a	900.55 ab	8.97 a	8.59 a	33.43 ab	26.44 cde	47.10 ab	33.76 b	30.06 ab	33.61 a	2.38 ab	2.47 ab	2.34 cde	2.25 de	96.04 bc	103.20 a	1.464 bcd	1.468 abc
Traper	2743.88 a	1671.40 a-d	40.49 abc	41.29 ab	1110.88 a	690.68 a-e	8.07 ab	5.79 b	32.87 ab	27.37 a-e	39.08 ab	39.88 ab	29.93 ab	30.59 ab	2.34 ab	2.61 a	2.24 d-g	2.16 efg	97.41 abc	98.57 a	1.453 de	1.456 d
Average	2251.11 a	1730.95 b	39.62 a	39.20 a	900.25 a	687.55 b	8.29 a	7.15 b	32.47 a	27.71 b	39.59 b	41.14 a	29.25 a	30.03 a	2.50 a	2.41 a	2.36 a	2.25 b	98.16 b	103.17 a	1.462 a	1.463 a

TPC, total phenolic content; TFC, total flavonoid content; FRSA, free radical scavenging activity. Genotypes with mean values followed by the same lowercase letter, for each site, are not significantly different according to Tukey’s HSD test (p ≤ 0.05). The two sites, ENAM and DYT, are compared for all genotypes overall average (last line).

). This response is supported by high Stress Tolerance Index (STI) and Yield Stability Index (YSI) values (

Table 4. Stress tolerance index and yield stability index of 17 rapeseed genotypes evaluated under irrigated and rainfed conditions across contrasting environments.

Genotypes	Stress Tolerance Index	Yield Stability Index
Adila	0.830 ab	0.795 ab
Alia	0.707 ab	0.609 ab
Baraka	0.788 ab	0.703 ab
CZ-KN	0.709 ab	0.798 ab
H2M-5	0.422 b	0.690 ab
IND23	0.439 b	0.775 ab
IND82	0.464 b	0.795 ab
Lila	0.543 ab	0.720 ab
Marina	0.804 ab	0.718 ab
Moufida	1.020 a	1.009 a
Nap10	0.604 ab	0.699 ab
Nap9	1.030 a	1.017 a
Narjisse	0.632 ab	0.717 ab
Redana	0.992 a	0.908 ab
Solar	0.717 ab	0.560 b
TP2	1.026 a	0.993 a
Traper	0.678 ab	0.699 ab

Genotypes with mean values followed by the same lowercase letter are not significantly different according to Tukey’s HSD test (p ≤ 0.05).

), indicating limited yield loss under drought and supporting the classification of these genotypes as drought-tolerant and yield-stable, with relevance for breeding under water-limited conditions.

Their yield stability highlights their potential value for breeding programs targeting resilience under variable and water-limited conditions.

Under irrigated conditions, the mean oil content (OC) was 41.13%. The highest values were recorded for ‘Nap9’ (45.02%) and ‘Moufida’ (44.24%), whereas the lowest values were observed for ‘IND82’ (36.21%) and ‘IND23’ (35.83%). Under rainfed conditions, mean oil content declined to 37.68%, representing an overall reduction of 8.2% relative to optimal irrigation, with genotype-specific decreases ranging from 6.07% (‘Moufida’) to 12.74% (‘Solar’) (Figure 3). Overall, oil content was relatively little affected by water stress, indicating a moderate sensitivity of lipid accumulation to drought, in agreement with previous findings reported by Morsi et al. [55]. The observed reduction in OC under water deficit may be attributed to physiological constraints such as reduced carbon assimilation, impaired translocation of assimilates, and the reallocation of resources toward stress-defense mechanisms [56]. On the other hand, ‘Nap9’ (7.51% reduction) and ‘TP2’ (7.30% reduction) maintained relatively stable oil contents, which highlights their potential as promising candidates for oil production under water-limited conditions. Moreover, along with their high oil content, ‘Nap9’ and ‘Moufida’ have been identified as drought-tolerant genotypes at the germination stage [31], making them particularly suitable for cultivation in arid and drought-prone regions. These findings are consistent with previous studies on Moroccan genotypes, which reported OC values ranging from 36.7% to 39.9% and emphasized the important effect of the environment in regulating lipid accumulation [9].

Under irrigated conditions, the average oil yield (OY) was 945.93 kg ha⁻¹. The genotypes ‘Moufida’, ‘Nap9’, ‘Redana’, ‘TP2’, ‘Traper’, ‘Marina’, and ‘Solar’ achieved the highest yields with respective values of 1190.54, 1167.75, 1146.39, 1102.50, 1083.60, 1082.88, and 1068.42 kg ha⁻¹, all exceeding the mean value, whereas ‘IND82’ (564.93 kg ha⁻¹), ‘IND23’ (604.60 kg ha⁻¹), and ‘H2M-5’ (609.94 kg ha⁻¹) recorded the lowest values (Figure 4). Under rainfed conditions, the mean oil yield declined to 641.86 kg ha⁻¹, with ‘TP2’ (835.25 kg ha⁻¹), ‘Nap9’ (815.36 kg ha⁻¹), and ‘Moufida’ (808.76 kg ha⁻¹) maintaining above-average performance, while ‘IND23’ (438.06 kg ha⁻¹) showed the lowest yields (Figure 4). Oil yield losses under drought are mainly attributed to restricted carbon assimilation due to stomatal closure, shortened seed-filling duration, and disruption of lipid biosynthesis during reproductive growth [50,57]. Nevertheless, ‘TP2’, ‘Moufida’, and ‘Nap9’ maintained the highest oil yields under rainfed conditions, as compared to the other genotypes (Figure 4). Consistent with these findings, previous studies also identified ‘Moufida’ and ‘Nap9’ as drought-tolerant genotypes, reporting oil yield reductions of approximately 23% under water deficit conditions [12]. In comparative terms, the mean oil yield under well-watered conditions of the present study (945.93 kg ha⁻¹) was comparable to that reported under Mediterranean conditions in Sicily (960 kg ha⁻¹) [58], while under rainfed conditions, the average oil yield (835 kg ha⁻¹) exceeded values observed in semi-arid environments (650 kg ha⁻¹) [54]. Overall, these results demonstrate that Moroccan genotypes, particularly ‘TP2’, ‘Moufida’, and ‘Nap9’, achieve competitive oil yields and maintain stability under both irrigated and water-limited environments, emphasizing their potential for cultivation in Mediterranean and semi-arid regions.

Overall, genotypes such as ‘Moufida’, ‘Nap9’, and ‘TP2’ maintained comparatively higher oil yields at both sites, although they experienced clear reductions when shifting from the ENAM site to the more water-limited DYT environment (Table 3). Consistent with their STI and YSI values (Table 4), these genotypes appear relatively better adapted and more yield-consistent across environments, making them promising candidates for breeding under variable, water-limited conditions.

3.4. Drought Tolerance and Yield Stability Indices

The stress tolerance index (STI) and yield stability index (YSI) were used as formal, quantitative measures to assess drought tolerance and yield stability of rapeseed (Brassica napus L.) genotypes across irrigated and rainfed environments (Table 4). The combined indices clearly discriminated genotype responses. ‘Nap9’, ‘TP2’, ‘Moufida’, and ‘Redana’ exhibited the highest joint STI (0.99–1.03) and YSI (0.91–1.02) values, quantitatively demonstrating their capacity to maintain high productivity under both stress and non-stress conditions, consistent with established selection criteria [38,59]. Notably, ‘Nap9’, ‘TP2’, and ‘Moufida’ showed YSI values close to or exceeding unity, indicating minimal yield reduction under rainfed conditions. In contrast, ‘H2M-5’ displayed the lowest STI (0.422), reflecting poor drought tolerance, while ‘IND23’ and ‘IND82’ combined low STI (<0.47) with moderate YSI, indicating stable but low-yielding responses rather than effective adaptation. ‘Solar’ showed the lowest YSI (0.560), revealing marked yield instability under drought. Overall, the joint STI–YSI analysis provides quantitative support for genotype classification and validates ‘Nap9’, ‘TP2’, ‘Moufida’, and ‘Redana’ as drought-tolerant and stable materials for Mediterranean semi-arid environments, in agreement with recent genetic evidence in rapeseed [55,60].

3.5. Physicochemical Analyses

3.5.1. Acidity Index (AI)

Under irrigated conditions, the overall mean acidity index (AI) was 2.29 mg KOH g⁻¹. The highest values were observed for ‘IND23’ (2.76 mg KOH g⁻¹) and ‘Solar’ (2.64 mg KOH g⁻¹), whereas ‘Moufida’ (1.82 mg KOH g⁻¹) showed the lowest values (Figure 5). Under rainfed conditions, the mean AI increased to 2.61 mg KOH g⁻¹ (Figure 5). Once again, ‘IND23’ (2.97 mg KOH g⁻¹) exhibited the highest indices, while the lowest were recorded for ‘Moufida’ (2.14 mg KOH g⁻¹) (Figure 5). The drought-induced rise in AI is likely associated with accelerated lipid hydrolysis via lipase activation, promoting oxidative rancidity [61,62]. Similar trends have been reported for rapeseed and sunflower, where drought during seed filling increased free fatty acids and reduced oxidative stability [63]. In the present study, acidity index (AI) values slightly exceeded the 2.1–2.5 mg KOH g⁻¹ range reported under milder conditions by Guirrou et al. [9], likely reflecting the harsher climatic constraints of Moroccan semi-arid sites. Comparable drought-induced increases in free fatty acids have also been observed in sunflowers during seed filling [63], suggesting a conserved oilseed response to water deficit. Nevertheless, all genotypes remained below the Codex Alimentarius limit for crude virgin rapeseed oil (≤4.0 mg KOH g⁻¹; Zhang et al. [64]), confirming their suitability for food use. Notably, ‘Moufida’, ‘Marina’, and ‘Nap9’ exhibited the lowest increases under drought, reflecting more efficient antioxidant systems that limit lipolysis and oxidation, making them strong candidates for breeding programs aimed at stabilizing oil quality in Mediterranean semi-arid environments.

3.5.2. Peroxide Index (PI)

Under irrigated conditions, the overall mean peroxide value (PI) was 2.42 meq O₂/kg. Genotypes ‘IND23’ and ‘H2M-5’ exhibited the highest values (2.87 and 2.67 meq O₂/kg, respectively), whereas ‘Marina’ (2.08 meq O₂/kg) and ‘Moufida’ (2.04 meq O₂/kg) recorded the lowest (Figure 6). Under rainfed conditions, the overall mean PI decreased to 1.90 meq O₂/kg. The highest indices were registered for ‘IND23’ (2.60 meq O₂/kg), ‘H2M-5’ (2.50 meq O₂/kg), while the lowest values were observed for ‘Moufida’ (1.90 meq O₂/kg) (Figure 6). The lower PI observed under rainfed conditions indicates a reduced level of primary lipid oxidation compared to irrigated conditions. A similar decline in PI under drought stress has been reported by Guirrou et al. [9] in six Moroccan rapeseed varieties and has also been documented in sunflowers [63]. According to the literature, this response may be associated with drought-induced changes in oil composition, including reduced proportions of polyunsaturated fatty acids (PUFAs) and increased levels of monounsaturated and saturated fatty acids [65], as well as enhanced concentrations of natural antioxidants such as tocopherols and phenolic compounds that inhibit hydroperoxide formation [66].

All PI values recorded in this study were well below the Codex Alimentarius thresholds for virgin vegetable oils (≤15 meq O₂/kg for cold-pressed oils; ≤10 meq O₂/kg for refined oils; [67]), confirming the high quality of the oils produced. The consistently low PI values of ‘Marina’, ‘Narjisse’, and ‘Moufida’ highlight their superior oxidative stability and suggest their strong potential as candidates for breeding programs aimed at developing stable, high-quality rapeseed oils.

Although several rapeseed genotypes showed differences in PI between the ENAM and DYT environments (Table 3), ‘Narjisse’, ‘Moufida’, and ‘Marina’ maintained consistently low values across both sites, indicating reduced susceptibility to lipid peroxidation under variable conditions.

3.5.3. Iodine Value (IV)

Under irrigated conditions, the overall mean iodine value (IV) was 105.49 g I₂/100 g. The highest values were recorded in ‘Nap9’ (115.36 g I₂/100 g), whereas the lowest were observed for ‘Nap10’ (99.97 g I₂/100 g) and ‘Solar’ (97.36 g I₂/100 g) (Figure 7). Under rainfed conditions, the mean IV decreased to 95.68 g I₂/100 g, with ‘Nap9’ maintaining the highest value (106.97 g I₂/100 g) and ‘Nap10’ showing the lowest (86.47 g I₂/100 g) (Figure 7). This reduction under drought reflects a lower degree of oil unsaturation and is consistent with literature reports describing drought-induced modifications in fatty acid composition, notably increased saturated fatty acids and reduced polyunsaturates as a consequence of inhibited desaturase activity [65,68], in agreement with the findings reported by Guirrou et al. [9]. Comparable reductions in IV have also been observed in sunflower [63] and safflower [69]. In this study, higher IVs, as in ‘Nap9’, ‘Moufida’, and ‘Marina’, likely had greater unsaturation, which may benefit nutritional quality but can reduce oxidative stability; conversely, lower IV may suggest improved stability yet a less unsaturated profile, so IV should be interpreted cautiously.

Interestingly, all genotypes remained within or close to the Codex Alimentarius standard for low-erucic acid rapeseed oil (94–120 g I₂/100 g), confirming their suitability for edible oil production.

3.5.4. Refractive Index (RI)

Under irrigated conditions, the overall mean refractive index (RI) was 1.467. Genotypes ‘Adila’, followed by ‘Moufida’, exhibited the highest values (1.478 and 1.476, respectively), whereas ‘Traper’ and ‘Narjisse’ recorded the lowest, each reaching 1.456 (Figure 8). Under rainfed conditions, the overall mean RI decreased slightly to 1.462. The genotype ‘Adila’ retained the highest value (1.471), while ‘Traper’ (1.452) and ‘Solar’ (1.447) presented the lowest values (Figure 8). The slight decrease in refractive index (RI) under water deficit reflects modifications in the physicochemical properties of the oil. Since RI is commonly used as a quality and purity indicator, the narrow range of values observed in this study confirms the good quality of the oils produced under both water regimes [70,71]. Similar trends have been reported in soybeans [72] and sunflowers [63,73]. Genotypes such as ‘Adila’ and ‘Moufida’ maintained higher RI across both water regimes, which may indicate relatively higher unsaturation and potential nutritional value, but could also imply greater susceptibility to oxidation. Therefore, these genotypes appear promising for maintaining acceptable physicochemical indices under water-limited conditions, although nutritional and stability implications require confirmation by fatty acid profiling and direct stability assays.

3.6. Biochemical Analysis

3.6.1. Total Phenolic Content (TPC)

Under irrigated conditions, the overall mean TPC was 5.41 mg GAE/100 g oil, with the highest values recorded for ‘Marina’ (6.38 mg GAE/100 g), ‘CZ-KN’ (6.32 mg GAE/100 g), ‘Narjisse’ (6.28 mg GAE/100 g), and ‘Nap9’ (6.29 mg GAE/100 g). However, the lowest values were registered for ‘Adila’ (4.57 mg GAE/100 g), ‘Redana’ (4.48 mg GAE/100 g), ‘Lila’ (4.36 mg GAE/100 g), and ‘H2M-5’ (4.31 mg GAE/100 g) (Figure 9). Under rainfed conditions, the overall mean almost doubled to reach 9.98 mg GAE/100 g oil. ‘Nap9’ (11.82 mg GAE/100 g) and ‘TP2’ (11.68 mg GAE/100 g) exhibited the highest values, while ‘IND23’ (8.31 mg GAE/100 g) and ‘H2M-5’ (8.24 mg GAE/100 g) recorded the lowest ones. The marked increase in TPC under water stress can be attributed to activation of the phenylpropanoid pathway and enhanced phenylalanine ammonia-lyase activity, which promotes phenolic accumulation and reinforces oxidative stability through reactive oxygen species scavenging and inhibition of lipid peroxidation [74]. Notably, the TPC levels observed in ‘Nap9’ and ‘TP2’exceeded by 15.4% those previously reported in rapeseed oils [9,24] and also surpassed typical values reported for sunflowers (4.04–8.21 mg GAE/100 g) [63,75]. These results highlight the potential of Moroccan genotypes, particularly ‘Nap9’ and ‘TP2’, to produce rapeseed oils with enhanced antioxidant stability, which can present wide applications in functional, nutraceutical, and culinary domains.

Environmental factors significantly influenced TPC variation among genotypes (Table 2); however, ‘Nap9’, ‘TP2’, and ‘CZ-KN’ showed consistent performance across ENAM and DYT sites, maintaining higher phenolic levels than other genotypes (Table 3), supporting their potential as robust genetic resources combining drought resilience with enhanced oil antioxidant quality under Mediterranean conditions

3.6.2. Total Flavonoid Content (TFC)

Under irrigated conditions, the overall mean total flavonoid content (TFC) was 25.25 mg/100 g oil. The highest values were recorded for ‘Moufida’ (30.83 mg/100 g), ‘Baraka’ (30.07 mg/100 g), ‘Redana’ (29.85 mg/100 g), and ‘Nap9’ (28.90 mg/100 g), whereas the lowest were registered for ‘H2M-5’ (21.37 mg/100 g), ‘Lila’ (20.23 mg/100 g), ‘IND23’ (20.05 mg/100 g), and ‘IND82’ (19.75 mg/100 g). Under rainfed conditions, the overall mean TFC increased markedly to 34.93 mg/100 g oil. ‘Moufida’ (43.23 mg/100 g), ‘Nap9’ (40.93 mg/100 g), ‘Baraka’ (40.41 mg/100 g), and ‘Redana’ (39.93 mg/100 g) exhibited the greatest accumulation, highlighting their strong responsiveness to water stress. In contrast, ‘IND23’ (26.55 mg/100 g) and ‘IND82’ (25.01 mg/100 g) showed, again, the lowest values (Figure 10). This pronounced increase under drought may be related to the activation of phenylalanine ammonia-lyase (PAL), a key enzyme in secondary metabolite biosynthesis [57]. Notably, the high TFC levels of Moroccan genotypes, particularly ‘Moufida’, ‘Nap9’, ‘TP2’, and ‘Redana’, were consistently higher than those previously reported in rapeseed oils [74,76] and in agreement with the findings of Guirrou et al. [9]. They also clearly surpassed the range reported for other cold-pressed oils such as sunflower (12.7–25.6 mg/100 g) [63], underscoring their strong adaptive potential and superior oxidative stability, which confirms the importance and the good performance of the Moroccan varieties studied herein, in general, and ‘Moufida’, ‘Nap9’, and ‘TP2’ in particular.

Significant differences in TFC were observed between the ENAM and DYT experimental sites (Table 2). However, ‘Moufida’, ‘Baraka’, ‘Nap9’, and ‘Redana’ showed stable performance across both sites, consistently maintaining higher flavonoid levels, compared to the other genotypes (Table 3). This stability highlights their potential to deliver superior oxidative protection and nutraceutical quality under contrasting Mediterranean conditions.

The pronounced increase in TPC and TFC under rainfed conditions highlights genotype-dependent plasticity of secondary metabolism. In our panel, ‘Nap9’ displayed consistently high values for both TPC and TFC under water deficit, while ‘Moufida’ and ‘Baraka’ showed particularly strong flavonoid accumulation. Such responses may reflect differences in agronomic adaptation strategies (e.g., conservative water use and maintenance of functional photosynthesis under stress) and intrinsic metabolic regulation, where drought-induced ROS signaling activates the phenylpropanoid pathway and enhances the activity/expression of key enzymes such as phenylalanine ammonia-lyase (PAL), leading to increased accumulation of phenolics and flavonoids [77,78]. Importantly, these antioxidant-related traits may contribute to drought stress mitigation by improving ROS scavenging and protecting cellular structures, which is particularly relevant under the intermittent droughts typical of Mediterranean semi-arid environments. Consistently, the antioxidant activity assessed using radical-scavenging assays (DPPH and ABTS) provides a functional indicator of the redox-buffering capacity of the extracts; however, the strength of the relationship between TPC/TFC and FRSA may remain genotype-dependent, suggesting that other oil constituents can also contribute to the observed antioxidant responses [79,80].

3.6.3. Free-Radical Scavenging Activity

Free-radical scavenging activity (FRSA) differed significantly among rapeseed genotypes and was strongly influenced by water regime (p < 0.05). In the DPPH assay, antioxidant activity under irrigated conditions ranged from 25.37% in ‘Solar’ to 42.71% in ‘Moufida’. High values were also observed in ‘Moufida’ (42.71%) (Figure 11a). Under rainfed conditions, FRSA increased, ranging from 36.13% to 50.45%. The highest activities were recorded for ‘Redana’ (50.45%) and ‘Marina’ (50.34%) (Figure 11a). In contrast, ‘Narjisse’ and ‘H2M-5’ exhibited the lowest FRSA values (38.49% and 36.13%, respectively). In the ABTS assay, inhibition under irrigation ranged from 18.86% to 28.46%, with ‘Redana’ showing the highest value (28.46%), followed by ‘TP2’ (26.54%), and ‘CZ-KN’ (26.19%), whereas ‘Marina’ and ‘Moufida’ presented the lowest values (19.05% and 18.86%, respectively). Under rainfed conditions, antioxidant activity ranged from 29.11% to 41.68%, with ‘Nap9’ (41.68%), ‘Marina’ (39.96%), and ‘Redana’ (39.84%) showing the highest values, while ‘Solar’ (29.50%) showed the lowest (Figure 11b). The consistently high antioxidant capacity of ‘Moufida’, ‘Nap9’, ‘TP2’, and ‘Marina’ may be attributed to their richness in polyphenols and flavonoids, which aligns with previous studies reporting that phenolic and flavonoid compounds are the primary contributors to the antioxidant potential of rapeseed oil extracts [9,24,81]. Although TPC/TFC likely contribute to FRSA, genotype-dependent responses suggest other constituents also modulate DPPH/ABTS; thus, FRSA should be interpreted as an integrated antioxidant response rather than a direct proxy of TPC/TFC. Moreover, other studies suggest that specific metabolites such as vinylsyringol (canolol), derived from the decarboxylation of sinapic acid, may further enhance antioxidant capacity [82]. Overall, these findings highlight the value of ‘Moufida’, ‘Nap9’, ‘TP2’, and ‘Marina’ as promising sources of high-quality, antioxidant-rich rapeseed oil for functional foods and nutraceutical applications under Mediterranean semi-arid conditions.

Environmental factors significantly influenced free radical scavenging activity (FRSA) among the rapeseed genotypes (Table 2). Nevertheless, ‘Moufida’, ‘Nap9’, and ‘Marina’ maintained high FRSA values across environments, demonstrating strong antioxidant activity despite differences in water availability (Table 3). This stability suggests that these genotypes may deliver enhanced nutritional and functional performance under varying Mediterranean conditions.

3.7. Principal Component Analysis

Principal component analysis (PCA) clearly distinguished the genotypes based on their yield performance and oil quality traits, explaining 81.98% of the total variance through the first two components (Figure 12). The first component, PC1, which accounted for 70.75%, was positively associated with seed yield, oil yield, oil content, total phenolic content (TPC), total flavonoid content (TFC), and antioxidant activities (DPPH, ABTS), and negatively associated with iodine value and refractive index, making it a combined indicator of yield and nutraceutical oil quality.

The second component, PC2 (11.23%), was mainly related to acidity and peroxide indices in opposition to refractive index, thus reflecting oil oxidative stability. The PCA biplot presented in Figure 12 classified all genotypes into four distinct groups based on their yield performance and oil-quality traits. Genotypes in Group-1 are positioned on the high positive side of PC1, while those in Group-4 exhibit high negative values for PC1. In contrast, the genotypes in Group-2 occupy intermediate positive positions on PC1, whereas those in Group-3 are mainly separated along PC2.

The genotypes of Group-1, namely ‘Nap9’, ‘Moufida’, and ‘TP2’, present the highest positive scores on PC1 for traits such as seed yield, oil yield, oil content, TPC, TFC, and antioxidant activity (DPPH, ABTS), as well as the lowest negative values for iodine value and refractive index. This indicates that these genotypes exhibit the most favorable combination of high yield potential and enhanced nutraceutical oil quality. On the other hand, genotypes in Group-4, including ‘Narjisse’, ‘Alia’, ‘Lila’, ‘IND23’, ‘IND82’, and ‘H2M-5’ are strongly and negatively correlated with PC1 and, thus, present the lowest scores for seed yield, oil yield, oil content, and antioxidant traits, indicating that they possess the lowest performance for both yield and oil-quality attributes. Group-3 includes the genotypes ‘Solar’, ‘Nap10’, and ‘Traper’, which are mainly distinguished by their high positive scores on PC2 for acidity and peroxide values, reflecting reduced oxidative stability compared to the other groups. Finally, the Group-2 genotypes, including ‘Redana’, ‘Marina’, ‘Baraka’, ‘Adila’, and ‘CZ-KN’, show moderate positive values for PC1 and low values for PC2, suggesting intermediate performance for both yield-related traits and oil-quality indices (Figure 12).

The PCA further strengthened these findings by integrating yield and oil-quality traits into a coherent multivariate structure, clearly distinguishing between high-performing and sensitive genotypes. ‘Moufida’, ‘Nap9’, and ‘TP2’ clustered strongly on PC1 due to their combined high performance for seed and oil yields, elevated phenolic and flavonoid contents, and superior antioxidant capacity, traits known to enhance oil stability and stress resilience in Brassica crops as reported by Jamshidi Zinab et al. [28] and Batool et al. [83]. In contrast, the sensitive genotypes (‘IND23’, ‘IND82’, ‘H2M-5’, ‘Nap10’) clustered on the negative side of PC1, reflecting lower productivity and weaker oil quality. These clustering patterns highlight the coordinated nature of physiological and biochemical traits that contribute to resilience under semi-arid conditions. Overall, the genotypes ‘Moufida’, ‘Nap9’, and ‘TP2’ can be proposed as valuable cultivars for drought-prone environments as well as promising parental material for breeding programs targeting yield stability and enhanced nutraceutical oil quality.

4. Conclusions

This study demonstrates that water deficit, associated with rainfed conditions across two contrasting environments, significantly impairs rapeseed yield and oil traits. Moreover, it allows identifying genotypes capable of sustaining productivity and quality under such water stress. Water regime stood out as the dominant factor driving trait variation, while site-specific conditions further shaped performance. Among the tested genotypes, ‘Moufida’, ‘Nap9’, and ‘TP2’ consistently combined yield stability with superior oil quality, characterized by lower acidity and peroxide values, higher phenolic and flavonoid contents, and stronger antioxidant activity. These findings underline the potential of these genotypes as resilient candidates for cultivation in semi-arid Mediterranean regions. Overall, these genotypes are promising for semi-arid conditions based on yield, physicochemical traits, and antioxidant indices; however, nutritional gains and oxidative stability should be confirmed by fatty acid profiling and direct stability tests (e.g., Rancimat/OSI). Promoting such genotypes can help secure oilseed production under climate change. Breeding programs should prioritize traits such as stable oil yield, high antioxidant potential, and resilience across environments. Furthermore, integrating multi-season and multi-location evaluations with genomic, transcriptomic, and metabolomic tools would accelerate the development of climate-resilient rapeseed cultivars. Complementary analyses of fatty acid composition, antibacterial activity, sensory traits, and post-storage quality will further allow strengthening the industrial and nutraceutical value. These results also provide a practical framework for screening rapeseed germplasms under contrasting water regimes in other drought-prone arid and semi-arid regions worldwide. The combined use of multi-environment field testing and antioxidant-related oil traits can support breeding pipelines aiming at stable yield and quality under climate variability. Future work should further validate these candidates across additional seasons and environments representative of global semi-arid production zones. Together, these integrated approaches will support the deployment of high-quality cultivars adapted to water-limited environments and offering superior nutritional and industrial properties.