Optimizing Lavender (Lavandula angustifolia Mill.) Yield and Water Productivity with Deficit Irrigation in Semi-Arid Climates

Abstract

Irrigation optimization is critical for sustainable agriculture in water-scarce regions, particularly for drought-tolerant crops like lavender (Lavandula angustifolia Mill.), where strategic water management can enhance productivity. This study evaluated the impact of different irrigation regimes on yield, yield components, essential oil content, water productivity, and irrigation water productivity of drip-irrigated lavender under the semi-arid conditions of Niğde, Turkey, over five growing seasons (2020–2024). Treatments included full irrigation (T₁), moderate deficit irrigation (T₂, 33% water deficit), severe deficit irrigation (T₃, 66% water deficit), and rainfed conditions (T₄). Results demonstrated that T₂ achieved fresh yields (144–227 kg da⁻¹) and oil yields (2.7–6.9 kg da⁻¹) comparable to T₁, with crop water consumption ranging 781.5–923.6 mm. Rainfed conditions significantly reduced yields but maximized water productivity, highlighting the potential for efficient water use even under substantial water deficits. Essential oil content remained stable (1.88–3.04%) across treatments, except in 2022 (p < 0.05). Lavender exhibited low drought sensitivity, with a yield response factor (k_y) of 0.25, indicating adaptability to controlled water deficits. Regression analyses revealed significant positive linear relationships between crop water consumption and fresh yield. Overall, the findings emphasize that improving water productivity through moderate-to-severe deficit irrigation strategies can support sustainable lavender production under semi-arid conditions, even when absolute yields are partially compromised. Results indicated that T₂ is recommended for optimizing water productivity with minimal yield reduction. However, in water-limited regions, T₃ provided viable productivity, offering a suitable balance for sustainable lavender production in semi-arid climates.

1. Introduction

Irrigation is fundamental to global agriculture, accounting for about 70% of freshwater withdrawals worldwide [1]. The water scarcity exacerbated by climate change and increasing food demand highlights the need to optimize irrigation for sustainable agriculture [2]. By 2050, the demand for irrigated crops is expected to rise significantly; however, water availability will not increase at the same rate [3]. Therefore, improving water productivity (WP), defined as the yield produced per unit of water consumed, is paramount, often necessitating a shift in focus from maximizing absolute yield to maximizing the efficiency of water use. One approach to enhancing WP is deficit irrigation, where water is applied at levels below full irrigation conditions, thereby reducing water consumption while maintaining acceptable yield levels [4]. Although deficit irrigation can result in slight yield reductions, research indicates that it substantially improves WP, making it a viable strategy for sustainable water management in agriculture [5].

Lavender (Lavandula angustifolia Mill.), a perennial aromatic shrub, is widely cultivated for its essential oil, medicinal properties, and ornamental value [6]. It is well adapted to semi-arid Mediterranean climates and is often classified as a drought-tolerant crop [7]. Once established, lavender can survive with minimal irrigation in regions with moderate precipitation; for maximizing fresh and oil yield, supplementary irrigation remains crucial. The water requirements of lavender exhibit considerable variability depending on irrigation strategies, environmental conditions, and cultivation practices. Reported seasonal applied irrigation amounts vary between 190.1 and 597.1 mm [8]. In controlled lysimeter studies, the total water use of lavender was recorded to be as high as 1191.4 mm, demonstrating its capacity for substantial water uptake when moisture availability is not limited [9]. Due to its deep-rooted system and xerophytic adaptations, lavender exhibits a relatively low yield response factor (k_y), typically around 0.3, indicating a moderate sensitivity to water deficits [10]. This characteristic suggests that lavender may be particularly well-suited for deficit irrigation approaches focused on enhancing water productivity, as significant yield decreases might be avoided even under reduced water application.

Efficient irrigation management is crucial to optimizing lavender cultivation in semi-arid environments. Farmers typically avoid daily irrigation, instead employing strategic irrigation scheduling during critical growth stages. Optimizing irrigation schedules during key phenological stages, such as pre-flowering and flowering, is crucial, as water availability at these stages significantly influences gas exchange parameters in lavender [11]. Despite its natural drought tolerance, lavender can still benefit from controlled irrigation, which has been shown to enhance yield and oil concentration. Lower soil moisture content negatively affected the physiological traits of lavender, leading to reductions in leaf area, leaf number, and biomass accumulation in terms of fresh and dry shoot weights [12]. Research by Sałata [13] demonstrated that supplementary irrigation significantly increased dry herb yield compared to rainfed conditions, with irrigated plants producing higher biomass and greater essential oil content. Similarly, field trials in Turkey indicated that fully irrigated lavender yielded 5.3 t ha⁻¹, whereas rainfed crops exhibited substantially lower yields. Notably, even under a deficit irrigation regime with a 33% reduction in water application, yield losses remained moderate, while WP improved considerably [8]. Reported WP values for lavender vary depending on irrigation levels and environmental conditions, generally ranging from 0.9 to 2.5 kg m⁻³ [14,15,16].

Existing research on the water requirements and deficit irrigation strategies of lavender remains limited, particularly regarding long-term impacts under semi-arid conditions. Since lavender is a perennial crop, prolonged experimental durations are essential to reliably capture yield responses and water management outcomes. Therefore, this study distinctly spans five consecutive growing seasons, providing comprehensive insights into the long-term yield stability and essential oil content of lavender, with a particular emphasis on the evaluation of water productivity and irrigation water productivity based on essential oil yield. The findings of this study are expected to provide valuable insights for optimizing irrigation management, with a particular focus for implementing effective water-saving approaches in lavender production across similar climatic regions.

2. Materials and Methods

2.1. Presentation of the Experimental Area

Field studies were conducted over a five-year period between the growing seasons of 2020 and 2024 at the research area located at the Faculty of Agricultural Sciences and Technologies, Niğde Ömer Halisdemir University, Turkey (Figure 1) (latitude, 37°56′35.9″ N; longitude, 34°37′52.8″ E).

The region is characterized by a semi-arid climate, experiencing hot and dry conditions during the summer, and cold, snowy weather in the winter. According to long-term meteorological data from 1935 to 2024, the average annual precipitation is 342.8 mm, with mean temperature and relative humidity values at 11.3 °C and 54%, respectively [17].

The soil characteristics of the study site are given in

Table 1. Physical and chemical properties of different soil depths of the study area.

Soil Depth (cm)	Sand (%)	Silt (%)	Clay (%)	Texture	Bulk Density	Field Capacity	Permanent Wilting Point (%)
					(g cm−3)	(%)
0–30	22.5	31.4	46.1	C	1.31	30.17	17.07
30–60	20.9	30.1	49.0	C	1.36	34.01	19.03
60–90	22.1	33.6	44.3	C	1.34	33.56	18.75
Soil Depth (cm)	EC (dS m−1)		pH		Available (kg da−1)		Organic Matter (%)
					P	K
0–30	0.008		7.88		2.9	14	1.72
30–60	0.011		7.94		1.7	16	1.43
60–90	0.012		7.91		3.1	11	1.57

. The textural classification content of the soil was an average of 21.8% sand, 31.7% silt, and 46.5% clay for a depth of 0–90 cm. The irrigation water was sourced from a nearby irrigation pond. Laboratory tests conducted for the irrigation water revealed that its pH was 6.37, and its electrical conductivity was recorded at 551 micromhos cm⁻¹ [(EC × 10) at 25 °C]. The concentrations of the major cations were determined as Ca²⁺ = 5.40 mEq L⁻¹, Mg²⁺ = 2.80 mEq L⁻¹, and Na⁺ = 2.18 mEq L⁻¹. Based on these concentrations, the Sodium Adsorption Ratio (SAR) was calculated as 1.08 mEq L⁻¹. This value classifies the irrigation water as C₂S₁, indicating medium salinity and low sodium hazard, suitable for most plants under proper management practices. According to Paraskevopoulou et al. [18], lavender growth remains satisfactory under salinity levels below 25 mM NaCl, which corresponds to considerably higher salinity than the levels observed in this study. Therefore, the irrigation water quality was not expected to induce salinity stress or adversely impact lavender growth or essential oil production throughout the experimental period.

2.2. Experimental Setup and Management

2.2.1. Plant Material

Lavender (Lavandula angustifolia Mill.), belonging to the Lamiaceae family, was used as the plant material. The seeds were germinated in organically enriched peat within open plastic trays, covered with vermiculite to facilitate aeration, in a greenhouse setting. Subsequently, the seedlings were moved into rectangular-shaped 60 × 40 cm plastic containers and then transplanted to the treatment plots on 3 June 2018. The plants were allowed to establish under conventional farming practices for two growing seasons (2018–2019) prior to the initiation of experimental treatments to align with the physiological maturity and industry-standard harvesting practices for lavender oil production. Based on soil test recommendations, 110 kg of nitrogen per hectare and 90 kg of phosphorus pentoxide (P₂O₅) per hectare were administered.

2.2.2. Water Treatments

The study utilized a randomized complete block design, incorporating three replicates. Each experimental plot measured 9 m in length and 2.4 m in width (21.6 m²). Each plot consisted of three rows, and the spacing between the rows was 80 × 50 cm. A 1 m distance was provided between the plots and 2 m between the blocks (Figure 2). Four different irrigation regimens were established to assess the effects of varying water availability on plant growth, based on pan evaporation (E_pan) rates. The treatments consisted of varying E_pan replenishment levels: complete replenishment (1.00 × E_pan − T₁), moderate replenishment (0.66 × E_pan − T₂), and reduced replenishment (0.33 × E_pan − T₃), along with a control group receiving no irrigation (0.00 × E_pan − T₄) [19,20]. The United States Weather Bureau (USWB) Class A evaporation pan, provided by the General Directorate of Niğde Meteorology Services and Research Area, was used to measure daily evaporation. Irrigation was applied twice a week to ensure consistent moisture levels throughout the soil profile. The study employed a drip irrigation system, utilizing 16 mm lateral polyethylene pipes with in-line emitters spaced 50 cm apart, delivering water at a rate of 2 L h⁻¹ under a pressure of 1 atm.

2.2.3. Irrigation Management

The amount of irrigation water applied was determined using the Equation (1) from the open water surface, as detailed in Cetin and Bilgel [21].

$$I=A x E_{pan} x K_{pc} x P$$

(1)

where I represents the irrigation water (L), A is the area of the plot (m²), E_pan is the evaporation recorded from the Class A pan during the irrigation interval (mm), and K_pc is the coefficient of Class A pan (determined as 0, 0.33, 0.66, and 1.0 for different irrigation levels) [22,23].

Values for crop water consumption (ET, measured in mm) for different irrigation treatments within the study were determined by employing the soil water budget equation, as denoted by Equation (2). This approach is based on comprehensive calculations that integrate multiple factors influencing soil moisture levels [24,25].

$$ET=I+P-R-D\pm \mathsf{\Delta }S$$

(2)

where I represents the amount of irrigation water applied (mm); P stands for precipitation (mm); R stands for runoff (mm); D signifies the drainage below the effective root depth (90 cm) (mm); and ΔS represents the change in soil water content, calculated as the difference between two measurements at a depth of 90 cm. To determine ΔS, soil samples were collected using the gravimetric method at various depths (0–30, 30–60, 60–90, and 90–120 cm) two days prior to the initial irrigation and just after the harvest each year. These samples aimed to capture the soil moisture levels at the beginning and end of the growing seasons, providing a comprehensive view of the water dynamics of soil. The analysis specifically focused on the 0–90 cm layer of the soil for calculating ET, whereas any increase in water content in the 90–120 cm layer was considered deep percolation and thus excluded from calculations. Runoff was not considered in the soil water budget calculations due to the precision application of water through the drip irrigation system [26].

2.2.4. Agronomic Practices

Detailed agronomic evaluations were conducted annually during the harvest periods from 2020 to 2024. Each year, plants were manually harvested to assess their growth outcomes and physiological responses. The specific harvest dates were as follows: 11 September 2020; 4 September 2021; 10 September 2022; 19 September 2023; and for the final year, 15 September 2024. Throughout each growing season, key growth parameters, such as plant diameter, height, number of primary branches, total branch count, spike diameter, and stalk length (all measurements in cm except spike diameter in mm), were meticulously recorded with the fresh yield (kg da⁻¹) on the harvest dates. Yield measurements and productivity calculations were expressed on a per-decare basis (da; 1 da = 0.1 hectare). Plant samples were carefully processed to segregate different components, facilitating precise evaluations of each part, as detailed by Paraskevopoulou et al. [18].

2.3. Determination of Essential Oil Yield

Dry lavender flowers (20 g) were subjected to hydro distillation using a Clevenger-type apparatus. The essential oil was co-distilled with 400 mL of distilled water over a period of 3 h. Following distillation, the extracted oils were collected and immediately stored in amber glass vials at 4 °C. The yield of the essential oil was determined by measuring the weight of the oil collected, expressed as a percentage of the dry weight of the lavender flowers used. The yield of essential oil was calculated using the methodology described in Pirzad and Mohammadzadeh [4] (Equation (3)).

$$Y_o=Y_d\times EO$$

(3)

where Y_o is the yield of essential oil (kg da⁻¹), Y_d is the yield of dried aerial parts (kg da⁻¹), and EO is the percentage of essential oil (%).

2.4. Water–Yield Functions

The yield response factor (k_y), which quantifies the correlation between the relative decrease in yield and the relative deficit in ET, was calculated using the method described by Stewart et al. [27] and Doorenbos et al. [28], as outlined in Equation (4).

$$\left(1-{\displaystyle \frac{Y_a}{Y_m}}\right)=k_y(1-{\displaystyle \frac{{ET}_a}{{ET}_m}})$$

(4)

where k_y is the yield response factor, Y_a is the actual yield achieved under different irrigation conditions (kg da⁻¹), and Y_m is the maximum yield possible under full irrigation conditions (kg da⁻¹). ET_a refers to the actual seasonal evapotranspiration observed with the treatment, while ET_m signifies the maximum seasonal evapotranspiration achievable under optimal irrigation conditions.

The water productivity and irrigation water productivity metrics were applied to evaluate both fresh yield (WP_f, kg m⁻³) and oil yield (WP_o, g mm⁻³) for each treatment, which were derived using Equations (5) and (6) [29,30].

$$WP={\displaystyle \frac{Y}{{ET}_a}}$$

(5)

$$IWP={\displaystyle \frac{Y-Y_0}{I}}$$

(6)

where Y is the fresh (kg da⁻¹) or essential oil yield (g da⁻¹) achieved under different irrigation conditions; Y₀ is the fresh (kg da⁻¹) or essential oil yield (g da⁻¹) obtained from the non-irrigated treatment; ET_a is the actual seasonal evapotranspiration (mm); and I indicates the total amount of irrigation water applied during the season for each treatment (mm).

2.5. Statistical Analysis

Statistical evaluations were conducted using the SPSS 22 software package. Data collected from the experiment were subjected to variance analysis separately for each group according to the Randomized Block Design. Before conducting means comparison, the normality of the data distribution was tested using the JMP Pro statistical software version 14. The results indicated that the data for all variables followed a normal distribution. To determine the effects of irrigation treatments on yield, agronomic parameters, and water productivity, variance analysis was performed at the 0.01 and 0.05 probability levels. Significant findings were further analyzed and grouped using the Duncan Multiple Range Test at a significance level of 0.05. Additionally, regression analyses were employed to investigate the relationships between irrigation water amount, crop water consumption (ET), and fresh yield. The significance of each regression coefficient was evaluated using Student’s t-test, and the overall significance of the regression models was tested using the F-test.

3. Results

3.1. Meteorological Conditions During the Experimental Periods

Table 2. Seasonal (May–September) meteorological data for the experimental years.

Parameters	2020	2021	2022	2023	2024
Avg. min temperature (°C)	9.20	8.02	9.36	8.25	9.20
Avg. max temperature (°C)	34.48	34.12	34.08	33.32	33.62
Avg. relative humidity (%)	44.26	45.70	46.62	45.62	45.70
Avg. wind speed 1 (m s−1)	1.91	1.91	1.73	1.73	1.75
Total precipitation (mm)	39.8	84.2	125.8	113.6	101.8

¹ Average wind speed (at 2 m height).

presents the average meteorological data for the May–September growing seasons, collected over the experimental years from the Niğde Meteorology Services and Research Area meteorological station.

Following the seasonal meteorological data presented in Table 2, relatively stable climatic conditions were recorded throughout the experimental years. Average minimum temperatures ranged between 8.0 and 9.4 °C, while average maximum temperatures varied from 33.3 to 34.9 °C, indicating consistent, moderately warm growing seasons. Relative humidity was stable at approximately 44.3–46.6%, and wind speeds were consistently low, ranging from 1.7 to 1.9 m s⁻¹. However, notable variability occurred in total seasonal precipitation, which ranged from 39.8 mm in 2020 to 125.8 mm in 2022.

3.2. Irrigation Water Amount and Crop Water Consumption

Table 3. Seasonal irrigation water amounts and evapotranspiration (mm) for different treatments.

Years	Treatment	Total Irrigation (mm)	Rainfall (mm)	Seasonal ET (mm)
2020	T1	1105.6	39.8	1152.7
	T2	729.7		781.5
	T3	364.8		444.9
	T4	0.0		98.5
2021	T1	1162.4	84.2	1262.5
	T2	767.1		904.7
	T3	383.6		523.1
	T4	0.0		154.2
2022	T1	1094.4	125.8	1270.0
	T2	722.3		923.6
	T3	361.2		551.3
	T4	0.0		170.3
2023	T1	986.5	113.6	1134.2
	T2	651.1		817.9
	T3	325.5		483.5
	T4	0.0		147.6
2024	T1	1096.9	101.8	1255.5
	T2	723.9		865.7
	T3	362.0		521.2
	T4	0.0		121.3

summarizes the total irrigation, rainfall, and seasonal crop water consumption (ET) for each treatment over the five-year study period (2020–2024). The irrigation amounts varied across treatments, ranging from 0 mm in the non-irrigated control (T₄) to a maximum of 1162.4 mm applied in T₁ during the 2021 growing season. Rainfall also showed interannual variation, fluctuating between 39.8 mm (2020) and 125.8 mm (2022). Annual irrigation amounts were determined by the cumulative evaporative demand measured via a Class A evaporation pan, a metric responsive not only to precipitation events but also to temperature fluctuations, solar radiation, wind speed, and the duration of rainless intervals.

The seasonal ET values differed among treatments depending on irrigation levels, with the highest ET values consistently recorded in the T₁ treatment, which received full irrigation, ranging from 1134.2 mm (2020) to 1270.0 mm (2022). Conversely, the lowest ET values were observed in the non-irrigated control, varying from 98.5 mm in 2024 to 170.3 mm in 2021.

3.3. Yield Components and Essential Oil Content

The effects of irrigation treatments on lavender yield and yield-related parameters across five growing seasons (2020–2024) are shown in

Table 4. Impact of different irrigation treatments on agronomic and quality parameters.

Year	Trt.	Plant Diameter (cm)	Plant Height (cm)	No. of Primary Branches	Total Branch Count	Spike Diameter (mm)	Spike Length (cm)	Essential Oil (%)	Fresh Flower Yield (kg da−1)	Oil Yield (kg da−1)
2020	T1	56.6 a	52.3 a	3.7 a	20.4 a	3.3 c	30.8	1.95	151 a	2.94 a
	T2	44.4 b	48.0 a	1.8 bc	12.6 b	4.1 b	30.2	1.88	144 a	2.71 b
	T3	54.1 a	47.1 a	2.7 ab	20.5 a	5.0 a	45.1	1.93	142 ab	2.74 ab
	T4	30.1 c	40.2 b	1.0 c	7.9 b	4.2 b	41.6	1.94	131 b	2.54 b
	F-test	**	**	**	**	**	n.s.	n.s.	*	*
2021	T1	69.6 a	78.4	10.6 a	47.8 a	3.5 b	19.4	2.06	190 a	3.91 a
	T2	55.1 b	75.2	7.7 b	39.4 b	3.8 ab	18.3	2.01	185 a	3.72 a
	T3	67.3 a	73.0	9.7 a	45.7 a	4.2 a	19.1	2.03	177 b	3.59 a
	T4	38.8 c	78.1	5.4 c	36.8 b	3.7 b	17.8	1.99	147 c	2.93 b
	F-test	**	n.s.	**	**	*	n.s.	n.s.	**	**
2022	T1	96.3	103.3 a	9.7 b	59.5 a	5.0	18.3	2.35 a	221 ab	5.19 ab
	T2	95.7	108.3 a	11.3 a	54.9 ab	5.1	20.3	2.41 a	224 a	5.40 a
	T3	92.7	99.6 a	8.7 b	53.5 b	5.1	19.3	2.43 a	205 b	4.98 b
	T4	93.3	87.0 b	8.3 c	43.8 c	5.0	20.0	2.21 b	156 c	3.45 c
	F-test	n.s.	*	**	**	n.s.	n.s.	*	**	**
2023	T1	94.3	94.8 a	14.3	54.9 a	3.0	20.0	2.35	240 a	5.64 a
	T2	98.3	96.4 a	18.7	49.0 ab	2.6	18.0	2.37	227 a	5.38 a
	T3	96.3	87.6 b	15.3	51.8 a	2.9	20.3	2.34	201 b	4.70 b
	T4	98.0	87.8 b	15.7	45.5 b	3.0	19.3	2.29	165 c	3.76 c
	F-test	n.s.	**	n.s.	*	n.s.	n.s.	n.s.	**	**
2024	T1	97.3	101.4 a	15.1 b	53.6 a	3.2	22.1	2.48	227 a	5.63 b
	T2	99.4	103.2 a	17.3 a	56.6 a	2.9	20.3	3.04	227 a	6.90 a
	T3	98.8	91.2 b	14.2 b	52.1 ab	3.0	21.7	2.34	206 b	4.82 c
	T4	98.2	82.4 c	14.1 b	45.7 b	2.6	19.4	2.69	151 c	4.06 d
	F-test	n.s.	**	**	**	n.s.	n.s.	n.s.	**	**

^a–d Significant differences using the least significant difference (LSD) test; ** significance at the %1 probability level (p < 0.01); * significance at the %1 probability level (p < 0.05); and n.s., non-significant.

. Significant differences among irrigation treatments were evident for plant diameter in the early years of study. In 2020 (p < 0.01) and 2021 (p < 0.01), fully irrigated (T₁) plants exhibited significantly larger diameters (56.6 cm and 69.6 cm, respectively) compared to the rain-fed (T₄) plants, which showed markedly reduced diameters (30.1 cm in 2020 and 38.8 cm in 2021). Notably, T₃ treatment resulted in statistically comparable plant diameters to full irrigation (T₁) in these early years (54.1 cm in 2020 and 67.3 cm in 2021), indicating the ability of lavender to sustain growth effectively under water stress. The differences became statistically non-significant from 2022 onward, suggesting that the lavender plants progressively adapted to water-limited conditions over the years, diminishing the effect of irrigation levels on plant diameter. The evaluation of plant height under different irrigation treatments (T₁–T₄) revealed statistically significant differences in all years, while no significant difference was observed in 2021 (Table 4). For 2020 and 2022, treatments T₁, T₂, and T₃ were statistically grouped together, indicating that their effects on plant height were not significantly different from each other, showcasing similar efficacy across these treatments. Conversely, T₄ consistently displayed a statistically significant reduction in plant height, highlighting its distinct impact compared to the other treatments. This pattern was evident in 2020 with T₄ at 40.2 cm, significantly lower than the others, and continued through subsequent years. For instance, in 2022, while T₁, T₂, and T₃ maintained heights over 99.6 cm, T₄ was notably shorter, at 87.0 cm.

The number of primary branches per plant has shown statistically significant (p < 0.01) variations in response to different treatments across the years, with the exception of 2023, where the variations were found to be non-significant (Table 4). Notably, treatment T₂ has consistently emerged as particularly influential, especially in the years 2022 and 2024. In 2022, plants under treatment T₂ exhibited an average of 11.3 primary branches, which was the highest among the treatments for that year. This trend was even more pronounced in 2024, where T₂ plants reached an average of 17.3 primary branches. Throughout the study period, the total branch count demonstrated statistically significant differences at the p < 0.01 level in response to different irrigation treatments, except for 2023 (p < 0.05). Notably, treatment T₃ exhibited the highest average branch counts in the years 2020 and 2021, revealing that a 66% reduction in water does not necessarily impair the ability of lavender to produce branches. This resilience suggests a substantial capacity for these plants to endure moderate drought conditions without significant detriment to branch production. In contrast, plants subjected to treatment T₄, which received no supplementary irrigation, exhibited a stark inability to maintain branch production, highlighting the critical role that even minimal water supplementation can play in supporting plant growth under stress.

In terms of spike diameter, the younger lavender plants were found to be more responsive to variations in irrigation treatments (Table 4). Significant distinctions were observed particularly in the initial years, with 2020 demonstrating a high level of statistical significance (p < 0.01) and 2021 showing notable effects as well (p < 0.05). In these years, treatment T₃ was especially effective, as evidenced by spike diameters of 5.0 mm in 2020 and 4.2 mm in 2021, suggesting substantial growth under these specific irrigation conditions with 66% water deficit. Contrastingly, over the subsequent three years, the influence of different irrigation levels on spike diameter has leveled off at non-significant levels. Spike length in lavender plants exhibits no significant variation in response to differing irrigation treatments across multiple years, underscoring a physiological robustness against changes in water availability. The consistent trait obtained suggests an efficient adaptive mechanism that stabilizes essential reproductive features under variable environmental conditions, typical of drought-resistant species.

The impact of irrigation treatments on the essential oil content of lavender plants is largely non-significant, except for the year 2022, where a minor statistical difference (p < 0.05) was observed (Table 4). This exception suggests that while the oil production rate can occasionally respond to changes in water availability, generally, essential oil yield remains stable across different irrigation levels, with a small range between 1.88 and 3.04%. This stability might be attributed to lavender’s inherent biological traits, which ensure the preservation of oil production under various hydration conditions. This resilience is particularly valuable in lavender, as essential oil quality and quantity are critical factors for commercial and medicinal uses. Oil yield has demonstrated statistically significant effects across all years, with a consistent upward trend observed as the study progressed. By 2024, oil yield reached as high as 6.3 kg da⁻¹, indicating a positive response to the irrigation treatments over time. Oil yield in lavender was significantly affected by irrigation treatments in all years. Generally, the highest yields were obtained in T₁ and T₂, while the lowest values were consistently recorded under rainfed conditions (T₄). A moderate reduction in water availability does not lead to a loss of oil yield in semi-arid regions, highlighting the resilience of lavender under controlled water stress conditions.

The analysis of fresh yield data from lavender plants across multiple years consistently demonstrates significant variability. Treatment T₂, with a 33% water deficit, has consistently produced high average yields, ranging between 144 kg da⁻¹ and 227 kg da⁻¹, illustrating its effectiveness under controlled water stress conditions. Treatment T₃, with supplementary irrigation, has also shown competitive performance. In stark contrast, treatment T₄, representing rainfed conditions prevalent in arid regions like Niğde, where precipitation is minimal, recorded yields ranging from 131 kg da⁻¹ to 165 kg da⁻¹. This pattern indicates that while lavender can tolerate moderate water deficits (as seen in T₂), it struggles under severe water scarcity (T₄), leading to significant yield reductions.

3.4. Water–Yield Functions

The relationships between irrigation water amount, ET, and fresh yield of lavender were evaluated through regression analyses (Figure 3). The relationship between fresh yield and ET showed a highly significant overall model fit (F (1,18) = 11.75, p < 0.01). Student’s t-test confirmed that both the intercept (t = 12.05, p < 0.01) and the ET coefficient (t = 3.43, p < 0.01) were statistically significant. Although the model demonstrated high statistical significance, the determination coefficient (R² = 0.395) indicated a moderate explanatory power, suggesting ET alone explains approximately 40% of the variability in lavender yield. Turning to the relationship between irrigation water amount and fresh yield, after evaluating several models, a quadratic polynomial regression model was found to best represent the data, providing the highest determination coefficient (R² = 0.431) and a significant overall model fit (F (2,17) = 6.44, p < 0.01). Student’s t-tests showed the intercept (t = 13.56, p < 0.01) and the linear irrigation coefficient (t = 3.24, p < 0.01) were statistically significant, whereas the quadratic term was not significant at the 5% level (t = −1.70, p = 0.107). Despite this non-significance, the quadratic polynomial model was retained because it clearly illustrated diminishing yield increments with increasing irrigation beyond an optimal threshold. The significant linear relationship established between ET and yield serves as a basis for subsequent calculation and analysis of the yield response factor (k_y), essential for optimizing irrigation strategies under varying water-availability scenarios.

The yield response factor (k_y), derived from five-year average data (2020–2024), was found to be 0.25, with a highly significant linear relationship (p < 0.01, R² = 0.91) (Figure 4). The calculated k_y value indicates that lavender has relatively low sensitivity to water stress, as it is substantially lower than unity (k_y < 1). This result suggests lavender is well adapted to deficit irrigation practices, exhibiting moderate yield reductions even under significant reductions in water availability. Therefore, this characteristic makes lavender cultivation suitable for semi-arid regions, promoting sustainable water resource management and maximizing water productivity.

The effects of different irrigation treatments significantly influenced water productivity (WP) values for both fresh yield and essential oil production throughout the five-year study period (

Table 5. Impact of irrigation treatments on water and irrigation water productivity.

Year	Treatment	Fresh Yield		Oil Yield
		WPf (kg m−3)	IWPf (kg m−3)	WPo (g m−3)	IWPo (g m−3)
2020	T1	0.13 c	0.02	2.55 c	0.36
	T2	0.18 c	0.02	3.46 c	0.23
	T3	0.32 b	0.03	6.16 b	0.55
	T4	1.33 a	-	25.8 a	-
	F-test	**		**
2021	T1	0.15 c	0.04	3.10 c	0.85
	T2	0.20 c	0.05	4.11 c	1.03
	T3	0.34 b	0.08	6.87 b	1.74
	T4	0.95 a	-	18.97 a	-
	F-test	**		**
2022	T1	0.17 d	0.06	4.09 d	1.60
	T2	0.24 c	0.09	5.84 c	2.70
	T3	0.37 b	0.14	9.04 b	4.25
	T4	0.92 a	-	20.24 a	-
	F-test	**		**
2023	T1	0.21 d	0.08	4.97 d	1.91
	T2	0.28 c	0.10	6.58 c	2.49
	T3	0.42 b	0.11	9.73 b	2.91
	T4	1.11 a	-	25.44 a	-
	F-test	**		**
2024	T1	0.18 d	0.07	4.48 d	1.43
	T2	0.26 c	0.10	7.97 c	3.92
	T3	0.40 b	0.15	9.25 b	2.10
	T4	1.24 a	-	33.49 a	-
	F-test	**		**

^a–d Significant differences using the least significant difference (LSD) test; ** significance at the %1 probability level (p < 0.01).

). The WP values for fresh yield (WP_f) ranged between 0.13 and 1.33 kg m⁻³, whereas essential oil WP (WP_o) values varied between 2.55 and 33.49 g m⁻³. Similarly, irrigation water productivity for fresh yield (IWP_f) ranged between 0.02 and 0.15 kg m⁻³, while irrigation water productivity for oil yield (IWP_o) varied between 0.36 and 4.25 g m⁻³ (Table 5). Notably, T₄ (rainfed) mostly demonstrated superior water productivity, underscoring its advantage in enhancing water productivity, particularly under limited water-availability scenarios.

4. Discussion

4.1. Irrigation Water Amount and Crop Water Consumption

In semi-arid and arid agricultural regions, efficient irrigation management is vital for sustainable crop production [31,32]. The differential seasonal ET values observed across the treatments highlight a pivotal aspect of crop–water dynamics. Consistent with prior studies such as Ariza et al. [33] and Xing et al. [34], our findings from the T₁ treatment underscore the principle that full irrigation correlates with maximal ET rates, indicative of optimal water availability facilitating maximum physiological activity. On the other hand, our rain-fed T₄ treatment presented the lowest ET values. This is also evident in lavender; despite its moderate sensitivity to water stress, irrigation applications remain crucial for its cultivation [35]. The variability in irrigation water amounts applied across treatments in our study aligns closely with earlier findings by Akcay et al. [10], who reported irrigation volumes ranging from 197.0 to 633.0 mm for lavender cultivation under semi-arid conditions. Comparable to our findings, their study showed that incremental increases in irrigation levels correspondingly elevated seasonal ET, with values spanning from 154.9 to 739.0 mm. Additionally, Noorollahi et al. [9] reported even higher values, with seasonal ET for lavender reaching up to 1191.4 mm. The variations in annual rainfall clearly influenced the observed ET patterns, reinforcing the complex interplay between irrigation management and natural precipitation in dictating water uptake dynamics [5].

4.2. Yield Components and Essential Oil Content

Irrigation significantly affects the yield components on lavender cultivation, emphasizing the critical role of water management during periods of low rainfall [13]. Plant diameter, a crucial agronomic indicator reflecting vegetative vigor and overall crop performance, exhibited significant variation in response to differential irrigation treatments, particularly during the initial years of cultivation (2020–2021). The correlation between reduced soil moisture and the decrease in plant diameter might be linked to detrimental effects on plant cell division, as proposed by Ogbonnaya et al. [36]. The diminishing differences among irrigation treatments observed in later cultivation years (2022–2024) indicate a possible acclimatization of lavender plants to limited water conditions over time. During periods of water stress, variations in internal water reserves are scarcely perceptible in the daily fluctuations of diameter [37]. Plant height, indicative of vegetative growth vigor and directly linked to potential biomass accumulation, was significantly influenced by irrigation treatments, particularly under severe water deficit conditions (T₄). This impact is corroborated by Paraskevopoulou et al. [18], who noted the positive effects of irrigation on plant height of lavender, and further supported by Akcay et al. [10], who found that the higher irrigation levels applied using gravimetric method notably increased plant height.

The number of primary branches and total branch count significantly reflected the impact of irrigation treatments, demonstrating clear patterns of vegetative branching response to differential water availability. Under full irrigation conditions (T₁), the increased branch number is physiologically linked to optimal hormonal signaling, particularly elevated cytokinin and auxin levels, which stimulate lateral bud development and promote branching patterns [38]. Further supporting this, Zhen and Burnett [12] found that applying mild moisture stress could lead to a noticeable reduction in plant branching. Meanwhile, Abdelsadek et al. [39] reported that the number of branches per plant for lavender ranged from 24.67 to 27.00 for different treatments.

The spike diameter significantly responded to varying irrigation treatments during the early years of study, but its responsiveness diminished in subsequent years as the plants matured. This plateau in sensitivity can be explained by the aging process of the lavender plants, where increased tissue lignification and structural consolidation reduce their responsiveness to changes in water supply [40]. Spike length represents key floral characteristics that directly influence the essential oil yield potential and commercial quality of lavender [41]. As a drought-tolerant species, lavender likely possesses an efficient root system capable of extracting moisture from soils with lower irrigation levels, thereby maintaining growth and key reproductive outputs like spike length. This acclimatization is especially beneficial in arid regions, where water conservation is crucial. While our study indicated non-significant changes in spike length across different irrigation treatments, Akcay et al. [10] reported significant effects of varying irrigation levels on spike length (ranging between 25.9 and 29.0 cm), demonstrating that irrigation can still substantially impact the morphological traits of lavender, with these effects confirmed to be statistically significant at p < 0.05.

The impact of irrigation treatments on the essential oil content of lavender plants was largely non-significant, except in the year 2022, when a minor statistical difference (p < 0.05) was observed (Table 4). Weather conditions are known to significantly influence essential oil content and composition in aromatic plant species [42]. The significant variation recorded in 2022 is likely attributable to environmental anomalies, particularly the higher-than-average seasonal precipitation (125.8 mm), compared to other experimental years. Increased water availability during this year may have modulated the secondary metabolic pathways in lavender, altering essential oil synthesis and accumulation, which aligns with known plant physiological responses to reduced water stress conditions. Furthermore, variability in essential oil concentration observed across years may also reflect the influence of fluctuating seasonal rainfall and climatic conditions on the overall stress experienced by plants. Similar patterns have been previously discussed in the literature, with studies suggesting that interannual climatic variations may influence secondary metabolite production in lavender [43,44]. Nonetheless, the generally consistent essential oil content across varying irrigation treatments (ranging narrowly between 1.88% and 3.04%) indicates inherent physiological resilience in lavender, beneficial for maintaining commercial and medicinal value despite fluctuating environmental conditions. This general stability across different water-availability scenarios aligns with previous studies. Chrysargyris et al. [45] reported no significant differences in essential oil content of lavender (averaging 3.57%) across varying irrigation treatments, suggesting intrinsic stability and genetic control over lavender essential oil biosynthesis. Similarly, Zhu et al. [46] also observed stable essential oil concentrations around 1.2% for different treatments, emphasizing inherent physiological capacity of lavender to maintain secondary metabolite production despite environmental variations.

Oil production, a primary determinant of economic profitability in lavender cultivation, exhibited a clear and consistent response to irrigation treatments, with significantly reduced yields under increased water deficit across all experimental years. Similar reductions under drought stress conditions have been widely documented in various aromatic plants, emphasizing a broader physiological trend among economically valuable species. Specifically, drought stress induces notable changes in essential oil yield and composition across aromatic plants [47,48]. The oil yields observed in this study fall within the wide range of values previously documented in lavender cultivation across diverse ecological and management contexts. For instance, Zheljazkov et al. [49] documented lavender oil yields reaching up to 5.5 kg da⁻¹, reflecting significant variability due to environmental and management factors. Minev et al. [50] similarly found substantial variations in lavender oil yields, from 5.9 kg da⁻¹ to as high as 18.3 kg da⁻¹ under optimal cultivation practices. In Turkish ecological conditions, Sönmez and Okkaoğlu [51] reported oil yields varying between 7.8 L da⁻¹ and 8.31 L da⁻¹, further indicating inherent variability based on local climatic and management practices. Notably, economic viability of lavender hinges on efficient biomass conversion, with approximately 1 kg of oil extractable from 50 to 55 kg of fresh flowers [52].

Fresh flower yield in lavender cultivation showed a pronounced sensitivity to irrigation treatments, consistently declining with increasing water deficit across all five growing seasons. These findings clearly illustrate the critical dependency of lavender plants on adequate water availability for optimal biomass production, aligning closely with previous studies reporting similar trends in aromatic plants [53]. The observed reduction in fresh flower yield under more severe deficit irrigation regimes (T₃ and T₄) can primarily be attributed to decreased photosynthetic efficiency and carbon assimilation capacity caused by drought-induced stomatal closure and reduced leaf area [54]. In support of our findings, Chrysargyris et al. [45] reported that fresh biomass of lavender significantly decreased with increasing water stress, declining from 7.63 g plant⁻¹ under full irrigation to 4.46 g plant⁻¹ under moderate stress, and further reduced to 2.23 g plant⁻¹ under severe water stress. Similarly, Akcay et al. [10] observed notable reductions in lavender yields under progressive irrigation stress, reporting yields declining sharply from 310.2 kg da⁻¹ under mild water stress to as low as 171.5 kg da⁻¹ under severe deficit irrigation conditions. Moreover, Sałata [13] found substantially higher green herb yields (0.68 kg m⁻²) in lavender plants provided with supplemental irrigation compared to non-irrigated plants, further reinforcing the economic significance of adequate water provision. Additionally, the observed interannual yield increases, especially noticeable under fully irrigated (T₁) or mildly restricted irrigation treatments (T₂), may be explained by cumulative improvements in vegetative parameters such as plant height, number of stems, and spike dimensions, typical of perennial crops reaching maturity over multiple years [55,56].

4.3. Water–Yield Functions

Optimizing irrigation strategies based on the accurate determination of water–yield relationships is critical, especially in semi-arid regions where water resources are limited, and sustainable water management is paramount for economic viability [57]. The positive linear relationship between ET and lavender fresh yield observed in this study underscores the critical physiological link between crop water consumption and biomass productivity. The increased ET values correspond to enhanced plant growth and yield due to optimized gas exchange rates [58] and nutrient transport efficiency [59] under adequate water availability conditions. Results from this study parallel those reported in lavender research by Akcay et al. [10], who identified quadratic polynomial relationships between irrigation levels and herb yields, with notably high coefficients of determination (R² = 0.98 and R² = 0.96). The lower R² values obtained from the regression analyses (R² = 0.395 for ET and R² = 0.43 for irrigation water amount) present the inherent complexity of multi-year perennial systems. The relationship between lavender yield and both evapotranspiration and irrigation levels is inherently complex and influenced by multiple interrelated factors beyond water availability alone. Significant annual variations likely contributed substantially to the observed variability. Specifically, interannual climatic differences, such as fluctuations in precipitation distribution, temperature extremes, humidity variations, and wind speed, are characteristic of semi-arid climates and considerably influence water productivity and physiological responses [60]. These environmental variables can lead to notable differences in crop yield responses across successive growing seasons, reducing the explanatory strength of regression models based solely on irrigation or ET. Additionally, the perennial nature of lavender adds another layer of complexity to the interpretation of these findings.

The yield response factor (k_y) quantifies the relationship between relative yield reduction and relative water deficit, providing a key indicator of crop sensitivity to water stress. In our study, lavender exhibited a k_y value of 0.25. This relatively low value, as previously reported by [10] (k_y of 0.27 and 0.35), signifies that lavender experiences a smaller proportional yield reduction under water stress compared to more sensitive crops. For instance, drought-sensitive basil (Ocimum basilicum L.) shows much higher ky values, such as 1.18 reported by Şenyiğit et al. [16] and 0.70–0.76 reported by Goldani et al. [61], indicating substantial yield decreases under deficit conditions. The comparatively low k_y of lavender is likely underpinned by morphological and physiological acclimatization, such as deeper or more extensive rooting systems capable of accessing water in lower soil layers [62], thicker cuticular layers, and finely tuned stomatal regulation that limits transpiration [63], collectively enhancing water-use efficiency and minimizing yield losses even under limited water availability.

However, this relative tolerance to stress, as indicated by the low k_y, does not imply that lavender yield is insensitive to water availability or that water management is unimportant. On the contrary, effective water management remains crucial in lavender cultivation for optimizing both fresh yield and essential oil production, which are vital for commercial viability [64]. Our experimental results demonstrate this clearly: water productivity (WP) and irrigation water productivity (IWP), essential criteria for assessing agricultural sustainability and efficiency, showed significant sensitivity to the different irrigation treatments applied across the experimental years. This statistically significant response underscores the complex but direct interaction between water availability and lavender yield components. While absolute yields were lower under reduced water conditions, the marked increase in WP and IWP points toward a substantial improvement in water-use efficiency. It highlights the potential of strategically applied deficit irrigation to enhance agricultural sustainability, especially in water-limited environments, by optimizing the output per unit of water consumed rather than maximizing total biomass or flower production. Similar results have been reported in various aromatic and medicinal plants, where deficit irrigation, though reducing overall biomass production, significantly enhances water productivity [65]. Specifically, Akcay et al. [10] reported WP values ranging from 0.70 kg m⁻³ under full irrigation to 2.30 kg m⁻³ under non-irrigated conditions, while IWP increased from 0.82 kg m⁻³ (full irrigation) to 2.10 kg m⁻³ at higher deficit levels. Their second experimental year showed a similar trend, with WP increasing from 0.81 kg m⁻³ under full irrigation up to 2.82 kg m⁻³ without irrigation, and IWP rising from 0.92 to 2.57 kg m⁻³, respectively. Previous studies, such as that by Yetik and Candoğan [5], reported that moderate water deficits could enhance water productivity due to physiological adjustments that plants exhibit under limited water availability. Similarly, our findings indicated that deficit irrigation treatments improved WP compared to full irrigation, highlighting physiological acclimatization of lavender and efficient water utilization under water-limited conditions. Differences observed between our study and previous research in terms of WP magnitudes can primarily be attributed to varying levels of drought tolerance among plant species, distinct physiological responses, and contrasting soil and climate conditions. Additionally, oil yield-based WP_o values presented a clear increase as irrigation levels decreased, achieving maximum values in rainfed treatments, thus indicating the economic potential of strategic deficit irrigation practices. Pirzad and Mohammadzadeh [4] indicated that certain levels of water deficit could specifically enhance oil-based water productivity in Lamiaceae species through improved stomatal control of transpiration, thus effectively balancing plant water loss and carbon assimilation rates.

5. Conclusions

This study demonstrates that strategic irrigation management significantly influences lavender cultivation under semi-arid conditions, highlighting the potential for sustainable agricultural practices through optimized deficit irrigation. Lavender exhibited considerable adaptability to varying irrigation regimes, displaying moderate sensitivity to water stress, which was quantified by a low yield response factor (k_y = 0.25). Although maximum yields were consistently achieved under full irrigation, moderate deficit irrigation (33% water deficit) proved particularly effective, offering an optimal balance between yield performance and water conservation. This treatment consistently produced comparable fresh and essential oil yields, underlining the resilience of lavender to mild water deficits. Importantly, the essential oil content (%) itself remained relatively stable across the different irrigation levels, indicating the plant’s ability to maintain oil quality even under varying water availability. Furthermore, the study emphasized the complex relationship between irrigation water amount, evapotranspiration, and productivity. Fresh yield displayed significant linear relationships with evapotranspiration, reinforcing the importance of precise water management to maximize productivity. Notably, rainfed conditions resulted in significant yield declines, highlighting the limitations of relying solely on natural precipitation in semi-arid climates. Water productivity and irrigation water productivity analyses further supported the effectiveness of controlled deficit irrigation. Despite lower absolute yields under reduced water conditions, water productivity significantly increased, indicating improved efficiency in water use. Therefore, moderate deficit irrigation emerges as an environmentally and economically advantageous strategy, maximizing water productivity without severely compromising crop performance.

In conclusion, for drip-irrigated lavender cultivation under semi-arid conditions, moderate deficit irrigation (T₂; 33% water deficit) is recommended to achieve an optimal balance between yield performance and water conservation. However, in regions with limited or insufficient water resources, supplementary irrigation of approximately 300–350 mm per growing season (T₃, 66% water deficit) can also be recommended. Future studies should investigate long-term gas exchange parameters and economic feasibility to further support and optimize these irrigation strategies.