Agronomic Practices Shape Tissue-Specific Antioxidant Capacity and Metabolic Profiles in Achillea millefolium L.

Abstract

This study investigates the influence of agronomic management on the accumulation of bioactive compounds and the antioxidant capacity of Achillea millefolium L., a medicinal species of increasing relevance for pharmaceutical and nutraceutical applications. Different cultivation strategies were applied, including controlled drought stress, foliar fertilization, and inoculation with plant growth–promoting rhizobacteria (PGPR), in order to evaluate their impact on tissue-specific metabolic responses. The total antioxidant capacity (TAC) of flowers and roots was determined using FRAP, DPPH, and ABTS spectrophotometric assays, while metabolite profiling was performed by UHPLC–MS/MS analysis. Clear differences in antioxidant activity were observed among plant organs and cultivation treatments. Flower extracts showed intermediate antioxidant capacity, with FRAP values ranging from 55.86 to 66.55 mg TE g⁻¹ extract and the highest activity consistently recorded for treatment F_010 (addition of K, P fertilizers under water stress conditions and PGPR absence) across all assays. Root extracts exhibited substantially lower antioxidant values (FRAP 19.40–33.69 mg TE g⁻¹), although samples R_000 (no foliar fertilization, under water stress conditions and PGPR absence) and R_100 (no foliar fertilization, under water stress conditions and presence of PGPR) displayed comparatively higher radical scavenging activity. Metabolic profiling revealed a shared presence of caffeic acid derivatives and flavonoids, including mono- and di-caffeoylquinic acids and apigenin-related compounds, with marked quantitative differences among tissues. Overall, the results demonstrate that agronomic practices significantly influence the accumulation and distribution of antioxidant metabolites in A. millefolium L., highlighting the importance of cultivation strategies for optimizing the production of bioactive phytochemicals.

1. Introduction

Achillea millefolium L. (Asteraceae) is a perennial aromatic herb widely distributed throughout the Mediterranean basin and temperate regions of Europe, where it grows spontaneously in grasslands, marginal lands, and semi-natural agroecosystems. Its high ecological plasticity and adaptability to different pedoclimatic conditions make this species particularly relevant from an agronomic perspective, especially in the context of sustainable valorization of wild medicinal and aromatic plants [1].

Within the Mediterranean flora, A. millefolium L. is considered an important spontaneous officinal species with a long history of traditional use and increasing scientific interest [2,3]. Its relevance in applied sciences is closely related to the complexity of its secondary metabolism and the wide spectrum of biological activities associated with its specialized metabolites [1,4]. Roots, leaves, and flowers function as distinct metabolic compartments, each contributing differently to the overall phytochemical profile of the plant, with direct implications for agronomic management, harvesting strategies, and quality assessment [5,6].

Achillea millefolium L. is a medicinal plant widely recognized for its therapeutic potential, largely attributed to its rich content of flavonoids and phenolic compounds [7]. Extracts have demonstrated significant antioxidant, anti-inflammatory, antimicrobial, and hepatoprotective activities, supporting their traditional use in gastrointestinal and inflammatory disorders. Moreover, bioactive constituents such as apigenin and luteolin exhibit neuroprotective effects and may modulate pathways involved in neurodegenerative diseases.

From a chemical–analytical standpoint, A. millefolium L. is characterized by a diverse array of secondary metabolites, including phenolic acids, flavonoids, sesquiterpene lactones, and volatile terpenoids. Non-volatile phenolic compounds are commonly identified and quantified using high-performance liquid chromatography (HPLC), whereas the essential oil fraction is typically characterized by gas chromatography coupled with mass spectrometry (GC–MS) [8,9,10]. These analytical approaches have revealed substantial qualitative and quantitative variability in metabolite composition, influenced by plant organ, phenological stage, genotype, and environmental conditions [6,11].

Among the functional properties of A. millefolium L., antioxidant activity represents one of the most extensively investigated aspects. Several studies have reported strong radical scavenging capacity and high total phenolic content in extracts obtained from different plant organs, highlighting a close relationship between antioxidant potential and phenolic accumulation [12,13]. The biosynthesis of these compounds is mainly associated with the shikimate and phenylpropanoid pathways, which are known to be highly responsive to environmental cues and agronomic factors such as growth conditions and harvesting time [14,15].

In addition to phenolic constituents, the essential oil of A. millefolium L. significantly contributes to its biological activity [9,16,17]. GC–MS analyses have consistently identified monoterpenes and oxygenated derivatives, such as 1,8-cineole, camphor, borneol, and α- and β-pinene, as major components of the volatile fraction. However, the relative abundance of these compounds varies considerably depending on plant organ, genetic background, and environmental conditions, reflecting a strong interaction between agronomy and secondary metabolism [18].

As previously anticipated, the increasing demand for natural antioxidants and plant-derived bioactive compounds has renewed interest in A. millefolium L. as a valuable resource for pharmaceutical, nutraceutical, cosmetic, and food applications. In this context, integrating agronomic approaches with robust chemical–analytical techniques such as HPLC and GC–MS is essential to elucidate organ-specific metabolite accumulation patterns and to support the sustainable exploitation and standardization of this Mediterranean medicinal species [19,20].

This study builds upon our recently published work [21] by further investigating the impact of different agronomic practices on the phenolic profile of three distinct organs of A. millefolium L., namely leaves, flowers, and roots. In the following sections, the applied cultivation practices are described and discussed, together with their effects on the antioxidant capacity of extracts obtained from the plant parts investigated. In addition, variations in the distribution and abundance of phenolic metabolites among the different tissues are reported.

Evaluating the antioxidant properties of extracts from different parts of the same plant (e.g., roots, leaves, and flowers) is crucial because the distribution of secondary metabolites is tissue-specific. Phenolic compounds, one of the main classes of plant antioxidants, are unevenly accumulated depending on physiological roles, environmental responses, and metabolic pathways [22]. This variability directly influences the bioactivity of extracts. Several studies demonstrate that antioxidant capacity differs significantly among plant organs. For example, in Achillea species, leaves showed higher radical scavenging activity than inflorescences and stems, correlating with higher phenolic content [23]. Similarly, in Sanguisorba officinalis, leaves exhibited markedly greater antioxidant activity than roots and stems, with strong correlations between phenolic levels and activity [24]. This approach is highly valuable in early-stage development of nutraceutical or cosmetic ingredients, as it enables the selection of the most bioactive plant part, optimizing extraction efficiency, efficacy, and cost-effectiveness. Therefore, targeted screening of plant tissues is a key step in ingredient standardization and functional product design.

2. Materials and Methods

2.1. Study Area

The experiment was carried out at the Terminillo Apennine Center “Carlo Jucci” in the Province of Rieti, a University of Perugia agricultural experimental center in a controlled environment to avoid water inputs from rainfall but with temperatures identical to the outdoor environment. Rieti is situated at an altitude of about 405 m above sea level in central Italy’s Lazio region. It lies within the Rieti plain surrounded by mountains such as Mount Terminillo to the east. This geographical setting influences its climatic conditions by moderating temperature extremes compared to coastal areas. The climate of the study area is characterized by a “humid temperate climate” with a Köppen classification of Cfb, with cool winters and hot summers with significant precipitation throughout the year [25]. Overall, Rieti’s climate combines elements typical of both temperate inland regions and Mediterranean influences due to its location near mountainous terrain but still within a broader Mediterranean climatic zone. Temperature during the study period was evaluated by calculating 10-day mean temperature values and through the GDD amounts with a base temperature of 0 °C for analyzing the relationships of climatic effects on phenological stages for each species.

2.2. Experimental Design

Seeds of Achillea L. provided by the seed bank of Terminillo Apennine Center, were vernalized by immersion in 20 mL of sterile distilled water and incubated for 48 h at 4 °C in the dark. After vernalization, seeds were sown in multi-cell trays (40 pots, 2.5 cm diameter) under controlled temperature conditions. Following germination, seedlings were transplanted in square pots measuring 20 cm wide and 15 cm high and grown outdoors under natural environmental conditions. During the water stress treatment, plants assigned to the drought-stress condition were placed beneath a rain-exclusion shelter consisting of a fixed overhead transparent impermeable cover (polycarbonate roofing), designed to prevent any rainfall input while allowing natural light penetration and maintaining ambient outdoor temperature conditions. This setup ensured that water availability was strictly controlled without altering the thermal environment.

Plants were cultivated in an organic substrate (pH 5.5), the soil was supplemented with nitrogen at 180 mg N L⁻¹, magnesium at 100 mg Mg L⁻¹, phosphorus at 130 mg P₂O₅ L⁻¹ and potassium at 235 mg K₂O/L⁻¹. Each replicate consisted of a pot containing 0.78 dm³ of substrate. A four-factor factorial design was adopted, with two levels per factor, resulting in sixteen treatment combinations. Factor levels were coded as 1 (maximum) and 0 (minimum), with five replicates per treatment. The factors included water availability, inorganic fertilization, and the presence of plant growth-promoting rhizobacteria (PGPR) (

Table 1. Experimental factors (PGPR application, fertilization, and irrigation) combined in the different cultivation protocols.

Combination	PGPR	Fertilization	Water Supply
	Inoculation	(K2O-P2O5)	(Field Capacity)
0,0,0	0	0	0
0,1,0	0	1	0
0,1,1	0	1	1
0,0,1	0	0	1
1,1,0	1	1	0
1,0,0	1	0	0
1,0,1	1	0	1
1,1,1	1	1	1

PGPR: 0 = absence; 1 = presence. Fertilizer: 0 = low (just soil nutrient); 1 = high (with addition of foliar fertilization. WS: 0 = 40%; 1 = 100% field capacity.

).

2.3. Agronomic and Stress-Mitigation Treatments

Drought stress was imposed prior to and throughout the early flowering stage. Soil moisture was monitored daily using an electronic depth probe (Aicevoos Digital Soil Tester). Water-stressed plants were maintained at 40% relative soil moisture for five weeks, while well-watered plants were kept at field capacity (~80% soil moisture). Relative soil moisture values were calibrated against field capacity using pot weighing and in situ probe measurements.

Foliar fertilization with phosphorus pentoxide (P₂O₅) and potassium oxide (K₂O) was applied four times during the growth cycle. The fertilizer solution was prepared by dissolving 2.5 g of a commercial formulation (PK 30-20) in 2 L of water and applied to the foliage using a pressure sprayer in treatments including fertilization.

PGPR were applied using a commercial inoculum containing a consortium of endomycorrhizal bacteria and fungi, including Bacillus spp., Streptomyces spp., and Pseudomonas spp. (1.6 × 10⁸ CFU g⁻¹), and Trichoderma spp. (5 × 10⁵ CFU g⁻¹). The inoculum was incorporated into the substrate at a rate of 10 g per 100 seedlings, with the first application performed immediately after transplanting (30 days after sowing) and the second 20 days later.

2.4. Sample Preparation and Hydro-Alcoholic Extraction

Flowers of A. millefolium L. were collected in July, while roots were harvested in September. All plant materials were immediately oven-dried at 40 °C for 3 days to obtain constant dry matter. Dried samples were then pulverized using an IKA^® Tube Mill Control (IKA^®-Werke GmbH & Co. KG, Staufen, Germany) at 25,000 rpm for 40 s.

Powdered plant material (30 mg) was extracted with 4 mL of an ethanol–water solution (50:50, v/v) by sonication for 30 min at room temperature. The resulting suspension was centrifuged at 5000 rpm for 20 min at 4 °C, and the supernatant was collected and stored at 4 °C until further analysis.

2.5. Determination of Total Antioxidant Capacity

2.5.1. Evaluation of the TAC Using the FRAP Method

The ferric reducing antioxidant power (FRAP) assay was carried out with minor modifications of previously described protocols [26,27,28,29]. The FRAP reagent was freshly prepared by mixing 2.5 mL of a 10 mM 2,4,6-tripyridyl-s-triazine (TPTZ) solution in 40 mM HCl, 2.5 mL of a 20 mM FeCl₃ aqueous solution, and 25 mL of 300 mM sodium acetate buffer (pH 3.6).

The assay was performed by combining 100 μL of the ethanol/water plant extract with 100 μL of bi-distilled water and 1.5 mL of the FRAP reagent. The reaction mixture was incubated for 4 min at room temperature in the dark, after which absorbance was measured at 570 nm.

Total antioxidant capacity (TAC) was quantified using a calibration curve constructed with Trolox as the reference standard. Trolox solutions were prepared and analyzed following the same procedure as the samples. The calibration curve ranged from 0.01 to 0.25 mg mL⁻¹ (R² = 0.996). Absorbance measurements were carried out using 0.2 mL aliquots of each solution. All analyses were performed in triplicate for each extract.

2.5.2. Evaluation of the Radical Scavenging Capacity by the DPPH Method

The 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical scavenging assay was performed with minor modifications of previously reported methods [13,26]. DPPH was dissolved in ethanol to obtain a working solution with an absorbance of 0.65 ± 0.02 at 517 nm, which was allowed to stabilize for 2 h prior to analysis.

The assay was carried out by adding 50 μL of the ethanol/water plant extract to 2.95 mL of the stabilized DPPH solution. The reaction mixture was incubated at room temperature in the dark for 30 min, after which absorbance was measured at 517 nm.

Radical scavenging activity was quantified using a calibration curve constructed with Trolox as the reference standard. Trolox solutions were prepared and analyzed following the same procedure as the samples. The calibration curve ranged from 0.005 to 0.25 mg mL⁻¹ (R² = 0.993). Absorbance measurements were performed using 0.2 mL aliquots of each solution. All analyses were conducted in triplicate for each extract, and results were expressed as mg of Trolox equivalents (TEs) per gram of dry extract.

2.5.3. Evaluation of the Radical Scavenging Capacity Using the ABTS Method

The ABTS radical cation (ABTS•⁺) decolorization assay was performed with minor modifications of previously described methods. The ABTS•⁺ stock solution was generated by mixing two volumes of an aqueous ABTS solution (0.36% w/v) with one volume of an aqueous potassium persulfate (K₂S₂O₈) solution (0.2% w/v). The mixture was incubated overnight at room temperature in the dark, protected from light with aluminum foil.

Prior to analysis, the ABTS•⁺ solution was diluted with ethanol to obtain an absorbance of 0.70 ± 0.05 at 690 nm. The assay was conducted by adding 4.0 mL of the ABTS•⁺/ethanol solution to 0.06 mL of the ethanol/water plant extract. The reaction mixture was incubated in the dark at room temperature for 6 min, after which absorbance was measured at 690 nm.

Radical scavenging activity was quantified using a calibration curve prepared with Trolox as the reference standard. Trolox solutions were treated following the same procedure as the samples. The calibration curve ranged from 0.01 to 0.75 mg mL⁻¹ (R² = 0.998). Absorbance measurements were performed using 0.2 mL aliquots of each solution. All analyses were carried out in triplicate for each extract, and results were expressed as mg of Trolox equivalents (TEs) per gram of dry extract.

Reagents Employed for the Evaluation of the Total Antioxidant Capacity (TAC) with the Three Spectrophotometric Assays

Methanol (MeOH), ethanol (EtOH), potassium persulfate (K₂S₂O₈), 2,2′-Azino-bis-(3-ethylbenzothiazoline-6-sulphonate) diammonium salt (ABTS), 2,4,6-tris(2-pyridyl)-striazine (TPTZ), 6-hydroxy-2,5,7,8-tetramethyl-2-carboxylic acid (Trolox), 2,2-diphenyl-1-picrylhydrazyl (DPPH), ferric chloride (FeCl₃), and sodium acetate (NaOAc) were purchased from Merck Life Science S.r.l. (Milan, Italy). Water was purified with a New Human Power I Scholar water purification system (Human Corporation, Seoul, Republic of Korea).

2.6. Instrumentation

UV/Vis spectrophotometric assays were performed using a Sunrise™ Absorbance microplate reader (TECAN, Männedorf, Switzerland). All the spectrophotometric experiments were performed using disposable optical Corning^® 96-well plates from Merck Life Science (Merck KGaA, Darmstadt, Germany) and a Sunrise microplate reader (Tecan Italia S.r.l., Milan, Italy).

Liquid chromatographic separation and mass spectrometric analysis were performed on a UHPLC–MS/MS system consisting of an Agilent 1290 Infinity II combined with the Agilent 6560 mass spectrometer (Agilent Technologies Inc., Santa Clara, CA, USA). The chromatographic separation was done by a Gemini^® C18 column (100 mm × 2 mm, 3 µm, 110 Å, Phenomenex, Torrance, CA, USA).

HPLC eluent A was water (LC-MS grade, LiChrosolv, Supelco, Bellefonte, PA, USA) with 0.01% (v/v) formic acid (LC-MS grade, LiChropur, Supelco) and 0.1 mM HCOONH₄ (LC-MS grade, LiChropur, Supelco), while eluent B was methanol (LC-MS grade, LiChrosolv, Supelco) with 0.1 mM HCOONH₄ (LC-MS grade, LiChropur, Supelco).

The chromatographic method followed the gradient program listed below: 0–2 min, 5–22% B; 2–7 min, 22–45% B; 7–20 min, 45–60% B; 20–28 min, 60–97% B; 28–29 min, 97% B; 29–30 min, 97–5% B; 30–31 min, 5% B; followed by column reconditioning for 2 min. Column temperature was kept at 30 °C and the flow rate was 0.25 mL min⁻¹. The injection volume was 5 µL. The chromatograms were recorded at 280, 325, and 370 nm.

For MS and MS2 detection, the Dual AJS ESI source operated in negative ion mode. The Gas temperature was set at 300 °C with a flow of 5 L min⁻¹ while the Sheath Gas Temperature was 350 °C with a flow of 12 L min⁻¹. The nebulizer pressure was set at 35 psi and the Capillary and Fragmentor voltages were 3500 V and 400 V, respectively.

In the case of Auto-MS2 analysis, the fragmentation patterns of the compound were recorded at 20 eV collision energy with an isolation width of 4 m/z. The Masshunter Workstation Data Acquisition 10.0 (Agilent Technologies Inc., Santa Clara, CA, USA) program was used for data acquisition, while the Masshunter Qualitative Analysis 10.0 (Agilent Technologies Inc., Santa Clara, CA, USA) software was used for data processing.

2.7. Statistics

All the analyses were performed in triplicate, and the values were reported as the mean ± standard deviation (SD) of the three independent analyses. The Statistica 12.0 software (StatSoft GmbH, Hamburg, Germany) was used to perform the statistical analyses. The significant differences among the parameters studied were analyzed using a one-way ANOVA, followed by a post hoc Tukey’s Honestly Significant Difference (HSD) test at a significance level of p ˂ 0.05. The correlation between the total polyphenol content, DPPH antioxidant capacity, and FRAP-reducing power was evaluated through a Pearson correlation test.

3. Results and Discussion

3.1. Phenological Observations

The seeds of the experimental plants were sown on 3 March 2025 in multi-cell trays and maintained under controlled temperature conditions. Following germination, seedlings were transplanted into larger pots and subsequently placed outdoors for approximately 20 weeks under natural environmental conditions.

Table 2. BBCH phenological scale of Achillea millefolium with days following sowing (DFS).

DFS	BBCH Code	Description
0	0	Germination—Dry seeds
16	10	Leaf development—Cotyledons completely unfolded
52	16	Leaf development—6 true leaves, leaf pairs
56	19	Leaf development—9 or more true leaves, leaf pairs
77	40	Vegetative propagated organs begin to develop
80	43	Vegetative propagated organs have reached 30% of final size
93	51	Inflorescence or flower buds visible
107	60	Flowering—First flowers open
116	62	Flowering—20% of flowers open
128	64	Flowering—40% of flowers open
142	65	Full flowering: 50% of flowers open, first petals may have fallen

summarizes the phenological development of A. millefolium L. according to the BBCH scale, expressed as a function of Days from Sowing (DFS). The early developmental phases (DFS 0–56) comprise germination and leaf development, ranging from dry seeds (BBCH 00) to plants bearing nine or more fully developed true leaves (BBCH 19) (Figure 1). The formation of harvestable vegetative biomass begins around DFS 77 (BBCH 40) and reaches approximately 30% of the final vegetative size by DFS 80 (BBCH 43), reflecting a phase of intense vegetative growth. The transition from the vegetative to the reproductive stage occurs at DFS 93, marked by the visibility of inflorescences or floral buds (BBCH 51). Flowering initiates at DFS 107 (BBCH 60) and progresses to full flowering at DFS 142 (BBCH 65), corresponding to approximately 50% of flowers open and the presence of the first senescent petals. Overall, the table provides a chronological framework linking DFS to the main phenological stages of A. millefolium.

3.2. Phenological Development and Climate Conditions

Table 3. Temperatures and GDD amounts during the plant developmental season.

Temperature °C		Date	GDD Amount	Days from Sowing	BBCH
Max	Min
16.0	−0.2	3 March 2025	25.70	0	00
13.6	−2.9	19 March 2025	183.90	16	10
18.4	3.7	2 April 2025	329.90	30	14
19.3	7.5	24 April 2025	600.15	52	16
22.0	9.7	28 April 2025	657.65	56	19
23.1	7.7	12 May 2025	872.25	70	26
24.1	7.4	19 May 2025	971.15	77	40
20.4	12.0	22 May 2025	1018.80	80	43
28.9	11.2	4 June 2025	1249.40	93	51
30.0	14.1	18 June 2025	1564.90	107	60
34.1	16.0	27 June 2025	1776.35	116	62
34.2	16.0	3 July 2025	1927.90	122	63
25.9	13.0	9 July 2025	2069.90	128	64
32.0	13.0	23 July 2025	2385.40	142	65

shows a progressive increase in temperature from sowing (March 2025) to mid-summer, which was reflected in the steady accumulation of growing degree days (GDD). During the early developmental stages (BBCH 00–14), corresponding to germination and initial leaf formation, thermal accumulation remained relatively low, reaching approximately 330 GDD within the first 30 days after sowing. The relatively cool conditions recorded during this period, with occasional sub-zero minimum temperatures, likely contributed to the moderate growth rate observed in the initial stages.

A more pronounced increase in GDD was recorded during the vegetative development phase (BBCH 16–40), when temperatures gradually increased, and cumulative values exceeded 970–1018 GDD between 70 and 80 days from sowing. This stage corresponds to the onset of active vegetative growth and biomass accumulation. The transition to reproductive development occurred at approximately 1249 GDD (DFS 93; BBCH 51), when the first floral buds became visible. Similar thermal requirements for the initiation of flowering have been reported for A. millefolium L. and other herbaceous species, where phenological transitions are strongly regulated by cumulative heat units [30,31].

The particularly warm conditions recorded in June, with maximum temperatures exceeding 30 °C, accelerated the accumulation of thermal units and promoted the rapid progression of reproductive stages. Flowering began at 1565 GDD (BBCH 60) and progressed to full flowering (BBCH 65) at approximately 2385 GDD and 142 days from sowing. These results confirm the strong relationship between temperature dynamics, cumulative GDD and phenological development, highlighting the importance of thermal conditions in regulating the growth cycle of A. millefolium L. under field conditions [31,32].

3.3. Comparative Quantitative Assessment of the Total Antioxidant Capacity (TAC) Among Flowers, Leaves, and Roots

A comparative assessment of the total antioxidant capacity (TAC) of A. millefolium flowers, leaves, and roots revealed clear quantitative differences among the three plant matrices across all applied assays (FRAP, DPPH, and ABTS).

Antioxidant capacity of plant extracts is commonly assessed through spectrophotometric assays; however, reliance on a single method may lead to incomplete or biased estimations. For this reason, the combined use of different assays, such as FRAP, DPPH, and ABTS, is recommended to obtain a more comprehensive evaluation of antioxidant potential [21,28,29]; These methods are based on distinct reaction mechanisms and experimental conditions, which influence the response of antioxidant molecules. The FRAP assay measures the ferric reducing ability of compounds under acidic conditions and primarily reflects the reducing power of electron-donating antioxidants. In contrast, DPPH and ABTS assays evaluate the radical scavenging activity of antioxidants toward stable free radicals, although they differ in solvent compatibility and reaction kinetics. Importantly, antioxidant molecules with different chemical structures and physicochemical properties (such as polarity, solubility, and redox potential) may interact differently with each radical system. Therefore, the application of multiple spectrophotometric assays allows a more realistic and reliable characterization of the overall antioxidant capacity of complex extracts. Although the results obtained for the leaves have already been reported in a previous study [21], they are presented again in

Table 4. Results of the measured FRAP, DPPH, and ABTS values of the extracts of different parts of Achillea millefolium L. (F = Flowers; L = Leaves; R = Roots) obtained under different cultivation conditions. The triplets of numbers refer to the applied cultivation conditions as reported in the text (Table 1). All data are reported as the mean value of three independent measurements, along with the corresponding standard deviation. Different letters (_{a b c d e f}) indicate statistically different mean values (p ≤ 0.05; ANOVA one-way, Tukey’s HSD tests) from highest to lowest.

Sample	Type of Assay
	FRAP (mg TE/g Extract)	DPPH (mg TE/g Extract)	ABTS (mg TE/g Extract)
F_000	60.43 ± 0.48 abc	45.81 ± 2.12 bc	48.07 ± 1.23 ab
F_001	55.86 ± 2.51 c	42.76 ± 1.12 c	43.69 ± 2.25 b
F_011	59.73 ± 2.14 bc	45.59 ± 1.67 bc	47.66 ± 5.3 ab
F_010	66.55 ± 1.19 a	51.89 ± 1.96 a	54.11 ± 0.72 a
F_100	60.65 ± 0.86 abc	45.26 ± 0.44 c	46.44 ± 1.57 ab
F_110	60.87 ± 2.18 abc	47.52 ± 2.44 abc	48.11 ± 2.61 ab
F_101	60.86 ± 1.9 abc	47.65 ± 2.32 abc	45.64 ± 2.31 ab
F_111	61.87 ± 2.25 abc	45.26 ± 1.64 c	49.48 ± 2.56 ab
Sample	Type of Assay
	FRAP (mg TE/g Extract)	DPPH (mg TE/g Extract)	ABTS (mg TE/g Extract)
L_000	86.47 ± 3.85 bc	24.64 ± 0.43 d	49.90 ± 2.89 c
L_111	107.31 ± 3.66 a	29.46 ± 1.34 ba	64.74 ± 7.80 a
L_011	95.76 ± 5.31 b	26.79 ± 2.00 dcb	60.60 ± 0.04 ba
L_001	82.19 ± 3.72 cd	28.06 ± 0.18 ba	57.85 ± 1.23 ba
L_110	71.67 ± 0.06 e	27.65 ± 1.15 ba	60.09 ± 4.16 ba
L_100	88.52 ± 5.31 bc	24.77 ± 0.54 dc	55.92 ± 1.27 cb
L_101	75.79 ± 3.00 ed	27.44 ± 0.59 cba	60.60 ± 2.34 ba
L_010	81.04 ± 2.43 cd	30.02 ± 1.41 a	53.85 ± 5.07 cb
Sample	Type of Assay
	FRAP (mg TE/g Extract)	DPPH (mg TE/g Extract)	ABTS (mg TE/g Extract)
R_000	33.69 ± 1.71 a	28.29 ± 4.61 a	27.60 ± 1.02 ab
R_001	27.93 ± 0.83 bcb	23.56 ± 1.38 ab	30.72 ± 0.76 a
R_011	27.30 ± 0.15 cd	25.87 ± 0.34 ab	30.25 ± 2.82 a
R_010	24.64 ± 0.63 e	23.80 ± 0.42 ab	26.43 ± 2.41 ab
R_100	32.74 ± 0.82 a	25.85 ± 1.44 ab	31.35 ± 0.3 a
R_110	25.65 ± 0.68 de	20.59 ± 0.26 bcd	25.39 ± 0.75 ab
R_101	30.07 ± 0.82 b	14.88 ± 6.43 de	31.03 ± 1.04 a
R_111	19.40 ± 0.21 f	16.67 ± 1.98 cde	23.70 ± 2.41 b

to facilitate a direct comparative evaluation. The results of this investigation clearly indicate a matrix-dependent distribution of antioxidant compounds, with TAC values varying significantly among plant organs (Table 4). Moreover, a graphical representation of this organ-specific difference is given in Figure S1 (Supporting Information).

Leaf extracts consistently exhibited the highest TAC, particularly in the FRAP assay (which measures ferric reducing antioxidant power), with values reaching approximately 107 mg TE/g extract, which were markedly higher than those recorded for flower and root extracts [21]. In addition, leaf extracts showed high ABTS radical scavenging activity and moderate-to-high DPPH activity, suggesting a substantial presence of compounds characterized by strong reducing power and effective radical scavenging properties. In contrast, root extracts displayed the lowest TAC values across all assays.

The consistently elevated antioxidant responses observed for leaf extracts across multiple assays support the conclusion that leaves represent the most valuable matrix in terms of antioxidant capacity among those examined.

Flower extracts demonstrated intermediate antioxidant capacity. Specifically, FRAP values for flowers were significantly lower than those observed in leaves but clearly higher than those measured in roots, reflecting an intermediate reducing capacity. A pronounced decrease in FRAP values was observed in root extracts compared with both flower and leaf extracts, indicating a limited electron-donating capacity of the phytocomplexes derived from this plant organ.

Results from the DPPH assay (which evaluates the capacity to scavenge lipophilic radicals) revealed more pronounced radical scavenging activity in flower extracts, whereas several root extracts exhibited the weakest radical neutralization capacity. The ABTS assay (which evaluates the capacity to scavenge both hydrophilic and lipophilic radicals) confirmed the highest radical scavenging activity in leaf extracts, with some samples showing performance comparable to that of flower extracts.

Taken together, these findings highlight substantial differences in metabolic activity and in the accumulation of phenolic and other bioactive compounds among the different plant organs. More detailed evaluations are provided below for the flower and root extracts, while repetition of the findings already described [21] for the leaf extracts is intentionally avoided.

3.3.1. Comparative Evaluation of Antioxidant Activity in A. millefolium Flowers by FRAP, DPPH, and ABTS Assays

A significant variation among the samples was observed in all of the assays (p < 0.05), indicating that the antioxidant response was strongly influenced by the applied agricultural practices.

In the FRAP assay, values ranged from 55.86 to 66.55 mg TE/g extract (Table 4). The sample F_010 exhibited the highest reducing capacity (66.55 ± 1.19 mg TE/g), which was significantly higher than the majority of other samples. The majority of the remaining samples (F_000, _011, _100, _110, _101, and _111) displayed comparable FRAP values, with a tendency to cluster around 60–62 mg TE/g and sharing the same statistical group. In contrast, F_001 exhibited the lowest FRAP value (55.86 ± 2.51 mg TE/g), indicating the lowest electron-donating capacity.

The DPPH radical scavenging assay revealed a broader differentiation among the samples, with values ranging from 42.76 to 51.89 mg TE/g extract. Once more, F_010 exhibited the strongest scavenging activity (51.89 ± 1.96 mg TE/g). Intermediate DPPH activity was observed for F_110 and F_101, while several samples, including F_001, F_100, and F_111, exhibited significantly lower scavenging capacity. The results obtained from this study indicate significant variations in the capacity of the extracts to neutralize stable free radicals.

The ABTS assay showed values ranging from 43.28 to 54.11 mg TE/g extract. In accordance with the FRAP and DPPH results, F_010 produced the highest ABTS value (54.11 ± 0.72 mg TE/g), thereby significantly outperforming all other samples. The majority of extracts exhibited intermediate ABTS activity (approximately 46–49 mg TE/g), with no significant differences observed among them. In contrast, F_001 presented the lowest ABTS scavenging capacity, suggesting a weaker overall antioxidant response in this assay.

When the three assays are compared, F_010 consistently represents the most potent antioxidant sample, showing superior performance across all methods. This finding suggests an optimal co-existence of compounds capable of both electron transfer and radical scavenging. In contrast, F_001 consistently exhibited the lowest antioxidant activity, regardless of the assay employed. The remaining samples showed moderate and assay-dependent variations, highlighting the importance of employing multiple antioxidant tests to elucidate the multifaceted mechanisms of action of plant-derived antioxidants.

3.3.2. Comparative Evaluation of Antioxidant Activity of A. millefolium Root Samples by FRAP, DPPH, and ABTS Assays

A substantial variation was observed for all three assays (p < 0.05), indicating a strong dependence of antioxidant activity on sample composition (Table 4).

The FRAP assay revealed substantial variability among the root extracts, ranging from 19.40 to 33.69 mg TE/g extract. The highest ferric reducing antioxidant power was observed for R_000 (33.69 ± 1.71 mg TE/g) and R_100 (32.74 ± 0.82 mg TE/g), both belonging to the top statistical group and indicating superior electron-donating capacity over the other extracts. Intermediate FRAP values were recorded for R_101, while R_001 and R_011 showed slightly lower but comparable reducing activity. In contrast, R_111 exhibited the lowest FRAP value (19.40 ± 0.21 mg TE/g), reflecting a markedly reduced reducing power.

The DPPH radical scavenging assay revealed pronounced variability among samples, ranging from 10.82 to 28.29 mg TE/g extract, highlighting pronounced differences among the samples in their ability to neutralize free radicals. R_000 exhibited the highest DPPH scavenging activity (28.29 ± 4.61 mg TE/g), followed by a group of samples including R_001, R_011, R_010, and R_100, which displayed statistically similar intermediate values. In contrast, R_101 showed notably lower DPPH activity, while the weakest scavenging capacity was observed for R_111, confirming its limited antioxidant performance in this assay.

The ABTS assay yielded values between 23.70 and 31.35 mg TE/g extract, indicating a narrower but still significant range of antioxidant activity. The highest ABTS scavenging capacity was observed for R_100 (31.35 ± 0.30 mg TE/g), closely followed by R_101, R_001, and R_011, all of which belong to the highest statistical group. Most of the remaining samples showed intermediate ABTS values with no significant differences among them. In contrast, R_111 displayed the lowest ABTS value (23.70 ± 2.41 mg TE/g), confirming its weaker radical scavenging ability.

When comparing the three assays, R_000 and R_100 emerged as the most effective samples in terms of reducing power and overall radical scavenging activity, although their performance varied depending on the assay. In contrast, R_111 consistently showed the lowest antioxidant activity across all assays, indicating a limited presence of antioxidant compounds. Several samples exhibited assay-dependent behavior, underscoring the importance of employing multiple analytical approaches to obtain a comprehensive evaluation of antioxidant potential.

3.4. UHPLC-MS/MS Analysis of the Three Extracts from Flowers, Leaves, and Roots

Based on the spectrophotometric assays, the extracts showing the best TAC profile were chemically characterized through UHPLC-MS/MS analysis. Although the analyzed tissues originated from plants grown under different agronomic conditions, the LC–MS/MS profiles revealed a clear organ-specific metabolic specialization. The LC–MS peak areas therefore provide an indication of relative metabolite distribution rather than an absolute quantitative comparison among organs. However, the metabolic profiling of A. millefolium L. reveals a highly pronounced distribution of phenolic compounds (

Table 5. The metabolic profiling of Achillea millefolium L. in the different investigated plant parts: flowers (F), leaves (L), and roots (R). Compounds are reported according to their elution time (Rt).

Putatively Identified Compound	Formula	m/z	MS/MS Fragments (m/z)	Retention Time (Rt, min)	Flowers, F	Leaves, L	Roots, R
					Peak Area (Arbitrary Units)
Quinic acid isomer	C7H12O6	191.0564	85.0297/93.0345	2.33	3 × 106	150,791,941	24,770,888
Malic acid isomer	C4H6O5	133.0142	71.0136	2.8		3,280,103	2,420,610
Hydroxybenzoic acid or isomer	C7H6O3	137.0247	93.255	3.45	94,458	122,362	591,153
Vanillic acid	C8H8O4	167.0347		4.04		72,589	185,662
Hydroxybenzoic acid or isomer	C7H6O3	137.0247	93.255	4.84		87,438	62,395
Coumaric acid or isomers	C9H8O3	163.0401	119.0493	5.62		229,030
Hydroxybenzoic acid or isomer	C7H6O3	137.0247	93.255	6.07		150,610	94,986
Quercetin rutinoside derivative/diglycoside	C33H40O21	771.1988	609.1461/301.0356	6.33		133,367	23,856
Caffeic Acid or isomer	C9H8O4	179.0351	135.179	6.4	132,201	568,047	413,240
Caffeoyilquinic acid isomer	C16H18O9	353.088	191.0558/173.0419	6.7	4 × 106	5,402,994	794,086
Apigenin diglucoside isomer *	C27H30O15	593.1504	473.1071/353.0658	6.99	281,893	130,523	1,790,870
Caffeic Acid or isomer	C9H8O4	179.0351	135.179	7.12		414,194	853,777
Feruloylquinic acid	C16H18O10	369.0825	191.1561/133.0561/93.0347	7.12	205,255	1,199,828	2,140,487
Malic acid isomer	C4H6O5	133.0142	71.0136	7.16		817,443	194,729
Quinic acid isomer	C7H12O6	191.0564	85.0297/93.0345	7.2		77,147,188	16,907,395
Caffeoyilquinic acid isomer	C16H18O9	353.088	191.0558/173.0419	7.21	155,270	170,499,252	34,654,534
Apigenin-7-O-rutinoside or isomer ***	C26H28O14	563.1407	269.01	7.71	191,399	271,557	55,021
Apigenin diglucoside isomer ***	C26H28O14	563.1407	269.01	7.84	412,488	144,234	32,517
Rutin or isomer	C27H30O16	609.1461	301.0350/271.8790	8.21	63,425
Dicaffeoylquinic acid isomer	C25H24O12	515.1197	191.0560/353.0865	8.44	926,586	537,913	1,113,596
Rutin or isomer	C27H30O16	609.1461	301.0350/271.8790	8.44	202,816	280,804
Apigenin diglucoside isomer ***	C26H28O14	563.1407	269.01	8.45	159,053	359,120	986,309
Caffeoyilquinic acid isomer	C16H18O9	353.088	191.0558/173.0419	8.6	1 × 106	9,228,865	1,102,692
Coumaric acid or isomers	C9H8O3	163.0401	119.0493	8.69		181,161
Naringenin	C15H12O5	271.0613	151.0054;	8.81	31,315	1,220,016
Apigenin diglucoside isomer *	C27H30O15	593.1504	473.1071/353.0658	9.05	814,189	11,354,819	106,093
Luteolin/kaempferol O-hexoside isomer **	C21H20O11	447.0935	285.0404	9.21	6 × 106	22,224,927	47,482
Rutin or isomer	C27H30O16	609.1461	301.0350/271.8790	9.67	23,378	60,494,431	2,625,290
Apigenin diglucoside isomer ***	C26H28O14	563.1407	269.01	9.68	740,530	6,397,613	65,029
Quercetin O-hexoside ****	C21H20O12	463.0888	301.0337	9.7	209,517	882,240
Apigenin diglucoside isomer *	C27H30O15	593.1504	473.1071/353.0658	10.12	11,144	569,244	88,954
Apigenin-O-hexoside isomer	C21H20O10	431.1006	269.0449/225.0549	10.38	497,847		244,177
Caffeic Acid or isomer	C9H8O4	179.0351	135.179	10.55		1,048,662	904,099
Caffeoyilquinic acid isomer	C16H18O9	353.088	191.0558/173.0419	10.57	147,639	3,912,788	571,447
Dicaffeoylquinic acid isomer	C25H24O12	515.1197	191.0560/353.0865	10.57	3 × 106	18,327,875	17,195,067
Ferulic acid	C10H10O4	193.051	178.5153	10.6		30,856
Caffeoyilquinic acid isomer	C16H18O9	353.088	191.0558/173.0419	10.78		13,099,716	2,128,451
Dicaffeoylquinic acid isomer	C25H24O12	515.1197	191.0560/353.0865	10.79	1 × 106	20,850,557	21,165,576
Apigenin-O-hexoside isomers	C21H20O10	431.1006	269.0449/225.0549	10.88	9 × 106	1,326,630	238,625
Luteolin/kaempferol O-hexoside isomer **	C21H20O11	447.0935	285.0404	11.1	637,189	295,734	4,668,805
Cinnamic acid	C9H8O2	147.0463		11.64	396,936	5,282,039	8,870,933
Luteolin/kaempferol O-hexoside isomer **	C21H20O11	447.0935	285.0404	12.26	115,605	134,538
Apigenin diglucoside isomer ***	C26H28O14	563.1407	269.01	13.3		304,895
Dicaffeoylquinic acid isomer	C25H24O12	515.1197	191.0560/353.0865	13.34	62,882	9,158,058	12,306,451
Apigenin-O-hexoside isomers	C21H20O10	431.1006	269.0449/225.0549	13.92			20,657
Flavone aglycone *****	C15H10O6	285.0408	151.0037/133.0294	14.89	9 × 106	1,803,822	114,007
Dicaffeoylquinic acid isomer	C25H24O12	515.1197	191.0560/353.0865	15.71	27,667	698,907	411,150
Apigenin or isomer	C15H10O5	269.0458	117.012	17.45	1 × 107	252,258
Apigenin or isomer	C15H10O5	269.0458	117.012	19.99	3 × 106	58,900

* Most likely Apigenin-7-O-rutinoside-4′-O-glucoside; ** Most likely Luteolin/Kaempferol-7-O-glucoside; *** Most likely Apigenin-7-O-rutinoside; **** Most likely Quercetin 7-O-hexoside; ***** Most likely Luteolin or Kaempferol.

), with the chemical signature of each tissue, flowers F (F_010), leaves L (L_111), and roots R (R_000), being tailored to its specific biological role [33].

A common metabolic signature was observed in all extracts, characterized by the presence of caffeic acid and its derivatives, mono- and di-caffeoylquinic acids, and several flavonoid aglycones and glycosides [34,35,36]. Despite this shared qualitative profile, noticeable differences in the relative abundance of metabolites were observed among the three plant parts in terms of metabolite abundances, suggesting a tissue-specific distribution pattern (Figure 2, Table 5) [33,34]. The semi-quantitative and comparative evaluation was performed by considering the peak areas of the individual identified compounds, rather than the relative distribution of metabolites among the extracts rather than to absolute quantitative differences. This approach was possible because all samples were analyzed at the same concentration (mg/mL), allowing a semi-quantitative comparison of the chromatographic responses among the analyzed extracts.

Exploratory pairwise comparisons based on log₂ fold change (log₂ FC) revealed a strong enrichment of hydroxycinnamic acid derivatives, particularly caffeoylquinic acids, in leaves, whereas flowers were characterized by the accumulation of flavonoids such as apigenin. Roots displayed a comparatively simpler phenolic profile, with moderate en-richment in cinnamic acid derivatives. This information is detailed in

Table 6. Log₂ Fold Changes in Metabolite Abundance Among Leaves, Roots, and Flowers of Achillea millefolium L.

Metabolite (Rt, min)	log2FC (Leaves vs. Flowers)	log2FC (Roots vs. Flowers)	log2FC (Leaves vs. Roots)	Putative Interpretation
Caffeoylquinic acid isomer (7.21)	+10.1	+7.8	+2.3	Strongly enriched in leaves
Quinic acid isomer (7.20)	+6.3	+2.8	+3.5	Predominantly accumulated in leaves
Rutin (quercetin rutinoside) (9.67)	+11.3	+6.8	+4.5	Major leaf-specific flavonoid
Caffeoylquinic acid isomer (8.60)	+3.2	+0.1	+3.1	Moderately enriched in leaves
Dicaffeoylquinic acid isomer (10.79)	+4.4	+4.3	+0.1	Highly abundant in both leaves and roots
Cinnamic acid (11.64)	+3.7	+4.5	−0.8	Slightly higher accumulation in roots
Apigenin (17.45)	−5.3	NA	NA	Flower-specific metabolite
Flavone aglycone (14.89)	−2.3	−6.3	+4	Enriched in flowers relative to roots
Luteolin/Kaempferol O-hexoside isomer (9.21)	1.9	−7.0	+8.9	Strong depletion in roots
Caffeic acid isomer (6.40)	+2.1	+1.6	+0.5	Moderately enriched in leaves
Feruloylquinic acid (7.12)	+2.5	+3.4	−0.9	Slightly higher in roots than leaves
Naringenin (8.81)	+5.3	NA	NA	Predominantly detected in leaves

Data represent exploratory comparisons based on single measurements. log₂FC: log₂-transformed fold change between groups. Positive values indicate higher abundance in the first group of comparison. Negative values indicate higher abundance in the second group. “NA” indicates that this differences between groups is not available due to undetected species in a specific tissue.

.

In Figure 2, the hierarchically clustered heatmap visualizes the relative abundance of putatively identified metabolites across the F, L, and R samples. Each row corresponds to a specific compound, while the columns represent the analyzed samples. The color gradient reflects the log10-transformed peak areas, where warmer colors (red) denote a higher relative abundance of the metabolite, and cooler colors (blue) indicate lower abundance or absence. Furthermore, the dendrogram on the left side clusters the compounds based on the similarity of their quantitative profiles across the experimental conditions, highlighting groups of metabolites sharing comparable distribution patterns across the analyzed tissues [37].

The foliar profile is defined by a strong metabolic investment in antioxidant and photo-protective compounds, consistent with the leaves’ role as the primary interface with solar radiation. Leaf tissues exhibit the highest overall phenolic content, dominated by quinic acid and its caffeoylated esters. The abundance of quinic acid suggests a large pool of precursors potentially available for the phenylpropanoid pathway. Notably, several mono- and di-caffeoylquinic acid isomers show elevated relative signals in leaf extracts, suggesting a potential role in buffering oxidative stress. Furthermore, the substantial accumulation of the flavonol rutin suggests a specialized adaptation for UV screening [38]. The prevalence of glycosylated forms, such as luteolin/kaempferol-7-O-glucoside, reflects an advanced capacity for vacuolar stabilization and storage [39].

In contrast to the leaves, F is characterized by a selective enrichment of apigenin and its derivatives. While the aglycone form of apigenin is predominantly accumulated in the flowers, the concurrent massive presence of specific apigenin-O-hexoside isomers underscores the role of these distinct metabolites in plant-pollinator interactions [40]. Conversely, leaf tissues show a preferential accumulation of more complex glycosylated forms, such as apigenin diglucoside isomers.

The root fraction presents a distinct chemical signature optimized for the rhizosphere. There is a relative enrichment of di-caffeoylquinic acid isomers. This shift toward complex hydroxycinnamate esters, combined with higher relative concentrations of simpler phenolics from the early phenylpropanoid pathway (like cinnamic and caffeic acids), supports the biological requirement for diffusible antimicrobial and allelopathic agents [41]. The root profile is also characterized by a lower diversity of complex flavonoid glycosides compared to aerial tissues, reflecting a restricted flux through the downstream flavonoid biosynthetic branches in favor of a mature developmental stage focused on soil microbiota interactions.

The detection of key organic acids (specifically malic and quinic acids) is diagnostic for the coupling between primary and secondary metabolic pathways [42]. These metabolites, originating from the tricarboxylic acid cycle, reflect the plant’s energetic status. The high biosynthetic demand of the carbon-intensive phenylpropanoid pathway requires a continuous supply of precursors from central carbon metabolism.

The presence of naringenin (an essential intermediate) in F and L confirms the active operation of the core flavonoid biosynthesis in aerial tissues. However, a clear divergence in the downstream metabolic routing is evident: while leaves primarily direct the flux towards the flavonol branch (yielding rutin and quercetin derivatives), flowers exhibit a strong commitment to the flavone branch, leading to the massive and targeted accumulation of Apigenin and its mono-hexosides.

This spatial differentiation highlights the biochemical plasticity of Achillea genus, allowing it to allocate resources efficiently according to the functional demands of each organ.

The mechanistic partitioning of carbon in A. millefolium L. follows several key trajectories:

(i)

Metabolic Routing: A hallmark of Asteraceae phytochemistry is the esterification of caffeoyl-CoA with quinic acid, linking the phenylpropanoid and shikimate pathways. This process generates mono- and di-caffeoylquinic acids, which are essential for redox homeostasis. In roots, the directed flux toward di-caffeoylquinic acid may also relate to structural tissue maturity or rhizosphere defense.

(ii)

Tailoring Processes: The conversion of naringenin into a diverse suite of flavones and flavonols via sequential hydroxylation and oxidation is highly active. The pronounced accumulation of apigenin aglycones and mono-hexosides in flowers, contrasted with the preferential accumulation of complex flavonol glycosides (like rutin and luteolin/kaempferol derivatives) in leaves, demonstrates a highly specialized, tissue-specific enzymatic tailoring at the branch point of flavone and flavonol synthases.

(iii)

Stabilization and Storage: The predominance of glycosylated metabolites in L is a strategic mechanism for enhancing stability and solubility, facilitating high-density storage within the vacuoles to prevent autotoxicity while maintaining a ready supply of antioxidants.

The metabolic profiling of A. millefolium L. underscores a sophisticated spatial differentiation of secondary metabolism. While the fundamental phenylpropanoid backbone is conserved across the species, the quantitative output is finely tuned to tissue-specific functional demands: leaves act as primary hubs for antioxidant and UV protection, flowers specialize in flavone-mediated signaling, and roots prioritize phenolic acids associated with biological interactions in the soil. This biochemical plasticity demonstrates the species’ ability to dynamically regulate metabolic flux to survive diverse environmental pressures.

4. Conclusions

This study demonstrated that agricultural practices significantly influence the antioxidant activity of A. millefolium L. extracts, as indicated by the significant differences observed among samples in all assays (p < 0.05). Clear variability was detected in both aerial parts and root extracts, confirming that cultivation conditions can modulate the accumulation of antioxidant compounds.

Among the aerial parts, the flower sample cultivated with the addition of K, P fertilizers under water stress conditions and PGPR absence (F_010) consistently exhibited the highest antioxidant activity across FRAP, DPPH, and ABTS assays, suggesting a favorable profile of compounds capable of both electron transfer and radical scavenging. Conversely, the flower sample cultivated without water stress, fertilizers, and PGPR (F_001) showed the lowest activity in all tests. Root extracts generally displayed lower antioxidant values than aerial parts; however, root samples in water stress condition (R_000) and with the addition of PGPR (R_100) showed the strongest performance, while root samples with PGPR, fertilization, and no water stress condition (R_111) consistently exhibited the weakest activity.

The obtained results indicate that aerial parts represent a richer source of antioxidant compounds compared with roots, although the response varies depending on the assay and cultivation treatment. Future research should further investigate the most promising plant organs under different agronomic conditions and evaluate how leaf harvesting at different balsamic periods (young, adult leaves) may affect the accumulation of antioxidant phytochemicals, in order to optimize cultivation and harvesting strategies for enhanced bioactive compound production.

Moreover, based on the obtained results, one of the future directions of this project will certainly focus on investigating biosynthetic pathways modulated by selected agricultural conditions. Specifically, the next coming study will evaluate how these practices may influence the activation of genes involved in key metabolic routes responsible for secondary metabolite production and organ specific accumulation. A main objective will be to correlate gene expression patterns with the actual presence of corresponding metabolites in plant extracts. This combined molecular and phytochemical approach will contribute to verify whether pathway activation translates into increased metabolite accumulation. Such insights are deemed to be essential to better understand plant biochemical responses to agronomic factors and to optimize cultivation strategies aimed at maximizing bioactive compound yield.