Sensitivity of Leafy Vegetables to Simulated Mesotrione Residues in the Soil

Abstract

Mesotrione is a triketone herbicide widely used for weed control in maize (Zea mays L.). In a bioassay conducted under controlled conditions, the simulated residual effects of mesotrione on leafy vegetables, including chard, lettuce, spinach, and endive were evaluated. The herbicide was applied at nine concentrations (0–240 µg a.i./kg soil), with the highest corresponding to the recommended field application rate. Nonlinear regression analysis was used to describe the relationship between morphological (shoot fresh weight) and physiological (pigment content) parameters as a function of herbicide dose. Shoot fresh weight was a more sensitive parameter than pigment content with mean EC₅₀ ± SE values of 23.9 ± 3.5 (chard), 34.3 ± 7.7 (lettuce), 13.2 ± 2.4 (spinach), and 990.3 ± 3921.5 (endive) µg a.i./kg soil, indicating that spinach is the most sensitive and endive the most tolerant species. A mesotrione residue level equivalent to EC₂₀ for shoot fresh weight corresponds to approximately 2, 4, 6, and 29% of the recommended application rate of mesotrione at which spinach, chard, lettuce, and endive (respectively) can be safely sown. Therefore, spinach, chard, and lettuce are not suitable substitutes for maize when the latter fails and should not be sown after silage maize. In such cases, only endive appears to be a viable alternative without the risk of crop injury.

1. Introduction

Mesotrione (2-(4-mesyl-2-nitrobenzoyl) cyclohexane-1,3-dione) is a herbicide that belongs to the triketone family. It is used for selective pre-emergence (PRE-EM) and post-emergence (POST-EM) control of broadleaf weeds and certain annual grasses in maize, including field maize, seed maize, sweet maize, and maize grown for silage [1,2,3,4]. Mesotrione affects carotenoid biosynthesis by inhibiting 4-hydroxyphenylpyruvate dioxygenase (4-HPPD), an essential enzyme in this biosynthetic pathway [1]. By inhibiting 4-HPPD, mesotrione reduces plastoquinone and tocopherol levels, leading to carotenoid depletion, which causes bleaching symptoms, necrosis of meristematic tissue, and ultimately plant death [5,6,7].

Since its introduction, this herbicide has become a valuable alternative for selective weed control in maize, offering numerous advantages such as broad-spectrum weed control, low application rates, excellent crop selectivity, a favorable environmental profile, and high flexibility in POST-EM treatments [2,8,9].

Mesotrione not only exhibits good efficacy but can also cause injury to sensitive crops grown in rotation, potentially leading to significant economic losses [4,10]. It has a half-life of 2 to 50 days, primarily depending on soil pH and organic matter content [5,11,12,13]. However, several studies have shown that mesotrione can persist longer under certain field conditions and that its residues can cause economic damage to highly sensitive rotational crops [5,7,10,12,14,15].

The phytotoxicity of mesotrione residues has been observed mainly in vegetables and legumes, as well as in certain field crops (sugar beet and soybeans) commonly grown in rotation with maize [10,14,16,17,18,19,20,21,22]. Previous studies have shown that mesotrione residues can cause injury to vegetable crops such as lettuce, snap beans, common beans, cucumbers, cabbage, peppers, tomatoes, peas, broccoli, carrots, and onions one year after application [10,14,16,17,20,21]. Several studies have characterized sugar beet, common beans, and peas as plants which are very sensitive to mesotrione residues [10,16,18,19,20,21]. For example, Riddle et al. [10] observed a complete yield reduction in sugar beet at a mesotrione application rate of 56 g a.i./ha. Similarly, Pintar et al. [19] found that sugar beet plants were completely damaged at 18 g a.i./ha in humofluvisol and at 24 g a.i./ha in hipogley soil. These findings emphasize that sugar beet should not be sown in the same field for at least 24 months after the application of mesotrione. Mesotrione at doses of 175 and 350 g a.i./ha, applied PRE-EM, caused 42% visual injury, a 31% reduction in shoot dry weight, and a 42% reduction in yield in cranberry and kidney beans, while black and white beans showed minimal visual injury (6% or less), which did not affect shoot dry weight or yield [16]. Therefore, the authors recommend a replant interval of 12 months for white and black beans and 24 months for cranberry and kidney beans. However, there is little published information on the residual effects of mesotrione on leafy vegetable crops (with the exception of lettuce and cabbage), while the short production cycle of these crops allows their use in rotation systems with maize [14,18,23]. Felix et al. [14] reported 34% cabbage injury from a PRE-EM application rate of 210 g/ha applied in the previous year. For lettuce, the results of the studies are largely consistent, and thus it can be classified as a highly susceptible species [18,24,25]. Riddle et al. [18] reported a reduction in dry weight of lettuce between 0 and 100% at mesotrione application rates of 70–560 g a.i./ha in the previous year. To our knowledge, there are no data or official recommendations for leafy vegetable crops such as spinach, chard, and endive. However, growers can use information such as EC₂₀ and EC₁₀ values to minimize the risk of crop injury by selecting the least sensitive crop that is not expected to lose yield. Given their short growth cycle, it is interesting to investigate whether these species are suitable for sowing, especially in situations where maize fails or after maize has been grown for silage.

Bioassays are generally used for the quantitative determination of herbicide residues and to assess their potential phytotoxic effects by measuring the phytotoxic fraction or the amount of herbicide available for biological activity [4]. In some cases, they can detect very small amounts of herbicide residues in the soil that cannot be measured by instrumental methods (even less than 1 µg/kg) [10,26,27]. Most authors have used field bioassays to determine the carryover potential of mesotrione. On the other hand, bioassays conducted in controlled conditions can also provide valuable information, as they represent a worst-case scenario by simulating conditions that favor herbicide damage. Additionally, results are obtained in a shorter time frame compared with field bioassays [4,28]. In the available literature, plant responses to mesotrione carryover have been evaluated mainly on the basis of parameters such as visual injury, plant dry mass, height, and fresh weight [10,17,21,28]. In addition to fresh weight, recent studies have emphasized the high sensitivity of physiological parameters, including carotenoid content and chlorophyll fluorescence [19,20,21]. This approach is well justified, as the highest test sensitivity is usually achieved when the selected measurement parameter is directly related to the herbicide’s specific mode of action.

Therefore, the objectives of this study were: (1) to determine the sensitivity of leafy vegetables (lettuce, spinach, chard, and endive) to simulated mesotrione residues by measuring morphological (shoot fresh weight) and physiological (pigment content) parameters, and (2) to determine EC₅₀ values as indicators of plant species sensitivity, and EC₂₀ and EC₁₀ values to estimate the threshold values required for safe sowing after mesotrione application in the previous crop.

2. Materials and Methods

2.1. Plant Material, Soil, and Herbicide

Various species were used as test plants: lettuce (Lactuca sativa L. var. “Daguan”), spinach (Spinacia oleracea L. var. “Matador”), chard (Beta vulgaris L. var. “Liscia verde da taglio”), and endive (Cichorium endivia L. var. “Gialla a cuore pieno”). The commercial formulation of mesotrione (Callisto, 480 g/L, Syngenta Crop Protection, Basel, Switzerland) was used to prepare standard solutions of the herbicide. The application rates of mesotrione were established at 0, 1.875, 3.75, 7.5, 15, 30, 60, 120, 240 µg a.i./kg soil with the highest concentration corresponding to the recommended application rate of mesotrione in the field. The soil was a loam with the following properties: pH 7.38, 42.88% sand, 34.28% silt, 22.84% clay, and 3.18% organic matter. The soil was collected in Stara Pazova (44°59′8.207″ N, 20°9′24.757″ E) from a field where no herbicides had previously been applied. The bioassay procedure has already been described in detail earlier [29,30].

2.2. Experimental Procedures for Bioassay Under Controlled Conditions

For each dose of herbicide, 600 g of soil was treated with 6 mL of an appropriately diluted herbicide solution. The herbicide solutions were applied using a thin-layer chromatography sprayer connected to a compressor, ensuring a constant application pressure of 120 kPa. Each soil sample was first mixed manually and then in a rotating mixer (60 rpm, 5–7 min). After homogenization, soil was divided into three plastic pots, i.e., 200 g per pot. Each pot was planted with 7–10 seeds (depending on the species) of the selected test plant, and watered daily to 70% of field capacity. Plants were grown in a growth chamber with a 16 h photoperiod (300 μE/m²s) at 24 °C and 8 h of darkness at 18 °C. The treatments were arranged in a completely randomized block design with four replications for each herbicide rate per plant species.

Two weeks after sowing, all plants were cut off at the soil level. The fresh shoot weight of all plants and the pigment content were measured. The leaf fragments (5 mm in diameter), weighing up to 0.1 g, were taken from the youngest leaves and placed in vials, each containing 3 mL of dimethylformamide (DMF). The vials were stored in a refrigerator at 4 °C for 24 h (so that the extraction of pigments could take place in the dark). After extraction, the absorbance values of the extracts were measured with a Libra S22 spectrophotometer (Biochrom Ltd., Cambridge, UK) (with a range of 190–1100 nm) at specific wavelengths: 480, 664, and 647 nm for carotenoids, chlorophyll a, and chlorophyll b, respectively. The equations used to calculate the pigment content were as follows [31]:

chlorophyll a: c_a = 11.65A₆₆₄ − 2.69A₆₄₇,

(1)

chlorophyll b: c_b = 20.81A₆₄₇ − 4.53A₆₆₄,

(2)

carotenoids: c_c = (1000A₄₈₀ − 0.89 c_a − 52.02 c_b)/245.

(3)

To obtain the concentration of pigments in fresh leaves (mg/g) the following Equation (4) was used:

$$C={\displaystyle \frac{c\cdot V\cdot R}{m\cdot 1000}}$$

(4)

where C is the pigment concentration (mg/g); c is the pigment concentration (µg/mL); V is the total extract volume; R is the dilution factor (in cases where dilution was required); m is the fresh material weight (g) and 1000 is the factor for converting µg to g.

2.3. Statistical Analyses

All data were subjected to ANOVA using IBM SPSS software (Statistical Package for the Social Sciences, Version 25). The significance of differences between treatments and the control was determined using Tukey’s test at a 95% significance level (p < 0.05). Each bioassay was repeated twice. Since the results were not significantly different, the two data sets for the measured parameters (expressed as the percentage of inhibition of the measured parameter) were pooled and subjected to nonlinear regression analyses to calculate EC₅₀, EC₂₀, and EC₁₀. For nonlinear regression analyses, the following four parameter log-logistic model was used (Equation (5)) [32].

$$Y=C+{\displaystyle \frac{D-C}{1+\exp \left\{b\cdot \left[\log \left(X\right)-\log \left(E\right)\right]\right\}}}$$

(5)

where Y is the test plant response (i.e., inhibition of measured parameter) as a function of the herbicide dose X, C is the lower limit of plant response (lower asymptote), D is the upper limit of plant response (upper asymptote), b is the proportional slope of the curve around EC₅₀ (the inflexion point), and E is the herbicide dose required to achieve half the plant response between the lower and upper limits, i.e., EC₅₀. All statistical analyses and graphs were performed in the R Software Program (version 3.1.1.) using the dose–response curve (drc) statistic package [33].

3. Results

Plant responses to simulated concentrations of mesotrione (1.875–240 μg a.i./kg soil) were evaluated in accordance with its mode of action by measuring morphological (shoot fresh weight) and physiological (carotenoid, chlorophyll a, and chlorophyll b content) parameters. In general, shoot fresh weight and pigment content varied depending on the plant species and mesotrione dose.

The first visual changes appeared immediately after the emergence of the first true leaves in all investigated plant species. These changes initially manifested as chlorosis, which progressively transitioned into bleaching, followed by necrosis, and ultimately, in sensitive species, complete plant destruction within two weeks. The progression of visual symptoms varied depending on species sensitivity and was observed at mesotrione concentrations ranging from 30 to 240 µg a.i./kg soil.

Chard and lettuce showed a similar response, with chlorosis occurring within the concentration ranges of 60–240 μg a.i./kg and 30–240 μga.i./kg soil, respectively. In spinach, visual symptoms were observed at concentrations ≥30 µg a.i./kg soil, while complete mortality was observed at the three highest concentrations. In contrast, endive showed only mild chlorosis, which was observed at the two highest concentrations.

3.1. Influence of Simulated Mesotrione Residues on the Shoot Fresh Weight

The inhibition of shoot fresh weight ranged from 12.86 to 100% for chard, 30.35 to 100% for lettuce, 11.10 to 100% for spinach, and 9.77 to 78.97% for endive. A hormesis effect was observed at low mesotrione concentrations (1.875–15 µg a.i./kg soil) in all species. Complete inhibition (100%) was observed at the two highest concentrations in chard and lettuce, while in spinach it occurred at the three highest concentrations (Figure 1).

ANOVA results showed significant differences between treatments and the control for all tested vegetables (Tables S1b–S4b). Significant differences were observed at concentrations above 60 µg a.i./kg soil for lettuce and endive and at concentrations above 15 µg a.i./kg soil for chard and spinach (

Table 1. ANOVA analysis for the inhibition of shoot fresh weight in leafy vegetables.

C vs. Applied Treatment (μg a.i./kg soil)	Chard	Lettuce	Spinach	Endive
	Shoot Fresh Weight
	p
C vs. 30	0.004		<0.0005
C vs. 60	0.009		<0.0005
C vs. 120	<0.0005	<0.0005	<0.0005	<0.0005
C vs. 240	<0.0005	<0.0005	<0.0005	<0.0005
F	41.68	62.895	55.09	45.75

C—control; F—test at 0.005 significance level; p—Tukey’s test at a 95% significance level (p < 0.05).

).

Nonlinear regression analysis was used to describe the relationship between shoot fresh weight and mesotrione dose, and the regression parameters EC₅₀, EC₂₀, and EC₁₀ were calculated (Figure 2,

Table 2. Regression parameters (Equation (5)) and mesotrione doses (μg a.i./kg soil) that caused 50%, 20%, and 10% inhibition [EC₅₀ (±SE), EC₂₀ (±SE), and EC₁₀ (±SE)] of shoot fresh weight for all crops tested.

Plant Species	Shoot Fresh Weight
	Regression Parameters (±SE)			EC50	EC20	EC10
	B	D	C
Chard	−1.6 (0.4)	104.4 (8.8)	−37.1 (7.3)	23.9 (3.5)	10.2 (2.3)	6.2 (2.0)
Lettuce	−1.6 (0.5)	108.9 (16.0)	−48.2 (9.2)	34.3 (7.7)	14.6 (3.9)	8.9 (3.3)
Spinach	−1.4 (0.3)	107.2 (6.8)	−41.3 (12.9)	13.2 (2.4)	4.8 (1.7)	2.7 (1.3)
Endive	−0.5 (0.3)	353.1 (498.4)	−51.6 (16.5)	990.3 (3921.5)	68.6 (186.4)	14.4 (29.4)

B—the slope of the line; D—upper limit; C—lower limit.

).

The results in Table 2 show a clear and significant difference between the species, with spinach being the most sensitive (EC₅₀ = 13.2 µg a.i./kg soil) and endive the most tolerant (EC₅₀ = 990.3 µg a.i./kg soil). Chard and lettuce can be characterized as moderately sensitive, with EC₅₀ values of 23.9 µg a.i./kg soil and 34.3 µg a.i./kg soil, respectively.

3.2. Influence of Simulated Mesotrione Residues on the Pigment Content (Carotenoids, Chlorophyll a and Chlorophyll b)

The inhibition, correlated with the applied mesotrione concentration, varied as follows: chard (carotenoids: 6.20–100%; chlorophyll a: 6.66–100%; chlorophyll b: 8.68–100%), lettuce (carotenoids: 14.87–100%; chlorophyll a: 18.69–100%; chlorophyll b: 17.94–100%), spinach (carotenoids: 1.58–100%; chlorophyll a: 19.31–100%; chlorophyll b: 14.84–100%), and endive (carotenoids: 10.07–55.28%; chlorophyll a: 13.25–63.08%; chlorophyll b: 11.34–63%) (Figure 3). A hormesis effect was observed only in spinach at the two lowest concentrations (1.875 and 3.75 µg a.i./kg soil), suggesting that pigment content is more sensitive to low mesotrione concentrations than morphological parameters.

Significant differences between applied treatments and the control for all measured pigments were confirmed by ANOVA (Tables S1a–S4a). As shown in

Table 3. ANOVA analysis for the pigment content in leafy vegetables.

C vs. Applied Treatment (μg a.i./kg soil)	Chard			Lettuce			Spinach			Endive
	Carot	Chl a	Chl b	Carot	Chl a	Chl b	Carot	Chl a	Chl b	Carot	Chl a	Chl b
	p
C vs. 1.875				<0.0005	<0.0005	0.003
C vs. 3.75				<0.0005	<0.0005	<0.0005				0.025	0.047
C vs. 7.5		0.009	<0.0005	0.008	<0.0005	0.002	0.001	0.011			0.007
C vs. 15	<0.0005	<0.0005	<0.0005	0.0001	<0.0005	<0.0005	<0.0005	0.001	0.018	0.018	0.001	0.005
C vs. 30	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	0.01	<0.0005	0.001
C vs. 60	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005
C vs. 120	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005
C vs. 240	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005	<0.0005
F	128.03	113.84	122.69	197.98	239.73	149.09	131.30	163.37	164.47	35.60	47.93	27.31

C—control; Carot—carotenoids; Chl a—chlorophyll a; Chl b—chlorophyll b; F—test at 0.005 significance level; p—Tukey’s test at a 95% significance level (p < 0.05).

, in most of the examined plant species, pigment inhibition was significantly different from the control at concentrations ranging from 7.5 to 240 µg a.i./kg soil, with statistical significance confirmed only in lettuce for all applied concentrations.

Nonlinear regression analysis was used to describe the relationship between pigment content and mesotrione dose, and the regression parameters EC₅₀, EC₂₀, and EC₁₀ were calculated (Figure 4,

Table 4. Regression parameters (Equation (5)) and mesotrione doses (μg a.i./kg soil) that caused 50%, 20%, and 10% inhibition [EC₅₀ (±SE), [EC₂₀ (±SE), and EC₁₀ (±SE)] of all physiological parameters measured for all crops tested.

Plant Species	Parameter Measured	Regression Parameters (±SE)			EC50	EC20	EC10
		B	D	C
Chard	Carotenoids	−1.9 (1.7)	113.3 (26)	14.1 (10.2)	66.9 (20.2)	32.2 (19.3)	21 (20.1)
	Chlorophyll a	−1.1 (0.9)	131.9 (60.7)	7.8 (17.1)	71.1 (63.2)	19.4 (12.9)	9.1 (10.6)
	Chlorophyll b	−1.2 (0.7)	125.4 (36.1)	8.3 (11.7)	63.0 (34.5)	19.1 (9.8)	9.5 (8.0)
Lettuce	Carotenoids	−10.4 (18.2)	100.1 (1.9)	17.7 (0.9)	67.4 (13.9)	59.0 (1.9)	54.6 (9.1)
	Chlorophyll a	−6.7 (5.2)	100.9 (4.4)	23.0 (2.0)	68.2 (7.3)	55.4 (4.4)	49.1 (8.1)
	Chlorophyll b	−5.8 (5.4)	101.5 (7.4)	25.8 (3.4)	69.9 (10.1)	55.1 (7.4)	48.0 (11.8)
Spinach	Carotenoids	−2.3 (0.9)	103.4 (6.6)	3.5 (7.4)	22.7 (3.7)	12.5 (4.0)	8.8 (3.9)
	Chlorophyll a	−2.2 (0.8)	104 (6.6)	−2.1 (7.5)	22.4 (3.6)	11.8 (3.7)	8.1 (3.5)
	Chlorophyll b	−1.8 (0.8)	107 (10.5)	−12.3 (12.3)	23.9 (5.6)	11.0 (5.3)	7.0 (4.7)
Endive	Carotenoids	−1.7 (0.4)	81.8 (22.9)	10.8 (0.9)	176.1 (73.0)	77.4 (16.8)	47.9 (6.4)
	Chlorophyll a	−0.9 (0.3)	251.1 (392.3)	10.0 (2.6)	1043.1 (2975)	206.8 (485.2)	80.2 (165.3)
	Chlorophyll b	−0.6 (0.3)	274.2 (377.6)	1.0 (9.1)	2194.4 (8262.7)	183.8 (552.9)	43.1 (11.6)

B—the slope of the line; D—upper limit; C—lower limit.

).

The calculated EC₅₀, EC₂₀, and EC₁₀ values for pigment content are in the same order of magnitude for each plant species and reflect the same sensitivity of the tested plants as determined on the basis of shoot fresh weight inhibition, with spinach being the most sensitive (EC₅₀ in the range between 22.4 and 23.9 μg a.i./kg soil, for all pigments) and endive being the most tolerant (EC₅₀ in the range between 176.1 and 2194.4 μg a.i./kg soil) (Table 4).

4. Discussion

Typical plant injury symptoms, such as leaf chlorosis and bleaching of newly formed leaves, followed by necrosis of leaf tissue, were observed in all tested plants. However, differences in chlorosis intensity were detectable only at mesotrione concentrations above 15 µg a.i./kg soil, indicating that visual assessments may not always be reliable or relevant. Significant plant injury can occur even at lower herbicide concentrations before visible symptoms appear. Therefore, it is crucial to include various measurable parameters, especially those resulting from the mode of action of herbicide itself.

The hormesis effect observed at low mesotrione concentrations in all tested species is a phenomenon previously reported for several other herbicides, such as glyphosate, 2,4-D, paraquat, and imazethapyr [30,34,35,36]. Pestemer and Günther [37] explained this effect as an optimal adaptation process in which the organism compensates for environmental changes. The occurrence and degree of this effect are influenced by various factors, including plant growth stage, physiological state, and environmental conditions [36,38]. Since the physiological parameter has shown sensitivity to lower mesotrione concentrations it appears to be suitable for detecting small amounts of mesotrione in the soil (1.875 µg a.i./kg soil). A similar conclusion was reached by Pintar et al. [19], Pintar et al. [20], and Pismarović et al. [21].

In the scientific literature, most studies on mesotrione carryover, conducted in the field or greenhouse, have assessed the effects of residues on parameters such as visual symptoms, plant dry mass, height, fresh weight, and yield [10,17,21,28]. Allemann and Molomo [28] and Brankov et al. [25] indicated that plant dry mass or fresh biomass provides a much better indication of plant susceptibility to mesotrione than plant height. The results of this study show that shoot fresh weight and pigment content are also reliable indicators of mesotrione residue activity. Interestingly, shoot fresh weight was somewhat more sensitive than the pigment content parameter (Table 2 and Table 4), which was an unexpected response considering the mode of action of mesotrione. As far as we know, there are only a few studies on mesotrione carryover effects based on physiological parameters [19,20,21,39]. Pismarović et al. [21] found that soil pH significantly influences the variability of physiological parameters, such as carotenoid content, whereas no such effect was observed for fresh weight reduction (in bean plants grown in soils with pH values of 7.5, 6.5, or 5.5, the ED₅₀ for fresh weight reduction remained similar, while the ED₅₀ for carotenoid content reduction varied with pH: 5.3 g a.i./ha at pH 7.5, 9.6 g a.i./ha at pH 6.5, 13.1 g a.i./ha at pH 5.5, and 15 g a.i./ha at pH 4.5). This suggests that measuring total carotenoid reduction is a more reliable method for assessing plant susceptibility to herbicide–soil interactions than fresh weight reduction [21]. Furthermore, it could be hypothesized that differences between the plants and their cultivars in terms of sensitivity and metabolism could be another factor influencing the variability of this parameter, making this topic an interesting possibility for further research.

The EC₅₀ value is most commonly used to compare different plant species for their sensitivity to herbicides in greenhouse bioassays [10,22,40,41]. However, there is only data on the EC₅₀ value for lettuce in relation to the measured parameters. Our results for lettuce sensitivity are consistent with the findings of Riddle et al. [18], Maeghe et al. [24], and Torma et al. [42]. In greenhouse experiments, Riddle et al. [18] found that lettuce was highly sensitive (I₅₀ = 14.9 g a.i./ha based on dry weight as a percentage of the control). The value of 14.9 g a.i./ha agrees well with the EC₅₀ of 34.3 µg a.i./kg soil determined in this study (equivalent to 17.15 g a.i./ha). These results are also consistent with earlier results from Torma et al. [42], in which 100% of lettuce plants died from residues of mesotrione, applied at a dose of 336 g a.i./ha in the previous year. We observed similar plant injury but at much lower mesotrione rates, comparable to 120 g a.i./ha and 60 g a.i./ha. The difference in lettuce susceptibility between these two studies which used soils with similar properties (pH and organic matter content) may be related to the possibility that freshly applied herbicide has a greater effect on susceptible plants in a simulated carryover study than the same amount of herbicide applied to soil the previous year. In this context, and due to the limited volume of soil available for root growth and the fact that the roots are in constant contact with the herbicide-treated soil, bioassays in pots are usually more rigorous than the same trials conducted in the field. However, this does not mean that pot trials can replace field trials. Rather, the two trials should be used in a coordinated manner to predict and confirm the risk of herbicide damage [28]. Furthermore, Brankov et al. [25] found that lettuce exhibited high sensitivity to mesotrione, even when applied in a POST-EM treatment as a drift simulation (application rate of 1.8 g a.i./ha caused up to 80% visual injury, with EC₅₀ values ranging from 0.22 to 0.50 g a.i./ha, depending on the evaluated parameters).

Besides the EC₅₀ value, other EC values, such as EC₂₀ and EC₁₀ (known as the No Observable Effect Level) have also played a crucial role in herbicide phytotoxicity studies. For instance, Jovanović-Radovanov and Rančić [30] proposed EC₂₀ as an acceptable imazethapyr residue level (ARL) at which maize, sunflower, and wheat can be safely sown. A mesotrione residue level equivalent to EC₂₀ for shoot fresh weight corresponds to approximately 2, 4, 6, and 29% of the recommended application rate of mesotrione, at which spinach, chard, lettuce, and endive can be safely sown, respectively. Therefore, these EC₂₀ values can be considered a useful guide for farmers wishing to sow these crops safely in rotation.

The available literature on mesotrione carryover and phytotoxicity reports a wide range of possible consequences, from severe damage at early growth stages with no impact on yield, to damage leading to significant yield reduction, and, under certain conditions, complete plant destruction. The extent of injury depends on various factors, including soil properties, climatic conditions, and other factors that influence the retention of higher residues in the soil [10,16,18,20]. In addition, the genotype of the plant species, including varietal and hybrid differences, plays a crucial role [16,22,28]. In some weed species, differential sensitivity has been shown to be due to micro-morphological characteristics [43], but subtle differences in physiology, gene expression, and other factors must also be considered, highlighting the need for further research. The sensitivity of a given plant species to the same herbicide can also vary considerably in different studies. Some authors suggest that differences in mesotrione formulation and recommended application rates in different countries may explain the different results, making direct comparisons difficult [16,17,28].

Several authors have reported that vegetables and legumes respond differently to mesotrione residues one year after application under certain conditions [14,15,17,18,19,20,21]. Felix et al. [14] reported carryover effects of mesotrione (12 months after its application at the recommended dose of 210 g a.i./ha (PRE-EM) in maize at two locations) on rotational vegetable crops such as snap bean, cucumber, bell pepper, processing tomatoes, cabbage, and red clover. Snap bean was the most sensitive crop, with visual injury and yield reductions of 75% and 94%, respectively. At the second location, there was no visual injury, and the yield reduction was insignificant. The authors attributed these varying responses to differences in edaphic factors (soil pH, organic matter, clay content) and environmental conditions. More severe injuries were observed at the site with lower pH, higher clay content, and greater organic matter content levels. Dyson et al. [12] found that mesotrione adsorption decreases as soil pH increases, leading to a slower degradation rate at lower pH. Consequently, reduced degradation after application could result in more mesotrione being available the following year to injure susceptible crops grown in rotation [14]. Similarly, Robinson [17] examined the effects of mesotrione residues one year after application at the recommended rate (140 g a.i./ha) on broccoli, carrots, cucumbers, onions, and potatoes, with injuries being 43%, 37%, 18%, 24%, and 0%, respectively.

Riddle et al. [18] determined the sensitivity of green beans, peas, lettuce, cucumber, sugar beet, and soybean to simulated mesotrione residues in a greenhouse bioassay. Based on I₅₀ for dry plant weight, sugar beet was the most sensitive (8.6 g a.i./ha), followed by lettuce (14.9 g a.i./ha), green beans (29.8 g a.i./ha), cucumber (41.6 g a.i./ha), pea (52.9 g a.i./ha), and soybean (67.9 g a.i./ha). These results are consistent with those of Riddle et al. [10], who evaluated mesotrione residues on the same plant species using a two-year field residue study (a conventional carryover study) and a one-year field study, where crops were grown in soil treated with reduced rates of mesotrione applied in the same year. Compared with our results, all tested plant species, except for sugar beet, were less sensitive than chard (EC₅₀ of 23.9 µg a.i./kg soil for shoot fresh weight) and spinach (EC₅₀ of 13.2 µg a.i./kg soil) (Table 2).

In previous studies, different bean varieties have shown different levels of susceptibility to mesotrione residues [16,28]. Soltani et al. [16] determined that black and white beans can be safely grown in rotation with maize treated with mesotrione (PRE-EM or POST-EM) at the labeled dose, but for cranberry and kidney beans the re-cropping interval of two years is recommended. Mesotrione at doses of 175 and 350 g a.i./ha applied PRE-EM caused 42% visual injury, 31% reduction in shoot dry weight, and 42% reduction in yield in cranberry and kidney beans, while black and white beans showed minimal visual injury (6% or less), which did not affect shoot dry weight or yield [16]. Pismarović et al. [21] and Pintar et al. [20] confirmed the high susceptibility of common beans and peas to mesotrione residues in soil, with a strong dependence on soil pH (the lowest rate of 1.1 g a.i./ha applied to acidic soil (pH 4.5) resulted in an insignificant reduction in carotenoid content in beans, while slightly alkaline soils (pH 7.5) caused a 35.5% reduction in carotenoids) [21]. Pintar et al. [20] observed a similar response for chlorophyll fluorescence reduction in peas. Some authors indicate that mesotrione adsorption primarily depends on soil pH and organic carbon (OC) [7,12]. Since mesotrione is a weak acid, its adsorption is more pronounced in acidic soils, where it exists as a neutral molecule with a strong affinity for adsorption. This makes it less susceptible to degradation and easily absorbed by plants [12]. In our study, an application rate of 1.875 µg a.i./kg soil (equivalent to 0.9375 g a.i./ha) resulted in carotenoid inhibition in chard and lettuce plants by 9.3% and 18.42%, respectively. However, the ED₅₀ values for beans ranged from 5.5 to 14.98 g a.i./ha depending on pH [21], indicating that chard and lettuce are somewhat less sensitive. With regard to spinach, it cannot be conclusively stated that it is more sensitive than beans, as the EC₅₀ value for carotenoid content was 22.7 µg a.i./kg soil (equivalent to 11.35 g a.i./ha) (Table 2).

As for field crops, the most comprehensive data are available for sugar beet, soybean, and rapeseed [10,18,19,22,39,42]. Pintar et al. [39] confirmed no sensitivity of rapeseed to one-year mesotrione residues, even in soil treated with an application rate four times higher than recommended. Torma et al. [42] came to a similar conclusion for rapeseed, as well as for sunflower, barley, and wheat, at mesotrione rates of 168–336 g a.i./ha. Researchers largely agree that sugar beet is highly sensitive to mesotrione residues in the soil, classifying it as one of the most susceptible crops [10,18,19,39]. In a simulated carryover study, Riddle et al. [10] observed a complete injury and yield loss at a mesotrione application rate of 56 g a.i./ha. Similar to these results, we observed a complete inhibition of shoot fresh weight and pigment content in spinach, but at an even lower application rate corresponding to 30 g a.i./ha.

Furthermore, Pintar et al. [19] determined that sugar beet exhibited higher sensitivity in silt loam (humofluvisol) compared with silty clay loam (hipogley). The ED₅₀ for fresh weight was 5.9 g a.i./ha (equivalent to 11.8 µg a.i./kg) in hipogley and 4.4 g a.i./ha (equivalent to 8.8 µg a.i./kg) in humofluvisol. In contrast, total carotenoid content was slightly more sensitive, with an ED₅₀ of 4.7 g a.i./ha in hipogley and 2.1 g a.i./ha in humofluvisol, corresponding to 9.4 and 4.2 µg a.i./kg, respectively. The authors attributed the differences in the observed response to the fact that hipogley contains twice as much organic matter, organic carbon (OC), and clay compared with humofluvisol, providing a greater number of adsorption sites on its surface. This enhances mesotrione adsorption and reduces its bioavailability to plants.

From everything that is written, it follows that that leafy vegetables belong to the highly sensitive risk group, which, under certain conditions, can be severely affected by mesotrione residues in the soil. Since the concentration of 30 µg a.i./kg (equivalent to 15 g a.i./ha) resulted in more than 50% inhibition of fresh shoot weight and pigment content in spinach, this species is clearly more sensitive than others, such as green beans, peppers, cucumbers, onions, carrots, and potatoes, and, under certain conditions, also peas and beans. The EC₅₀ value for shoot fresh weight in spinach is 13.1 µg a.i./kg (equivalent to 6.55 g a.i./ha), which is similar to that of sugar beet (between 4.4 and 8.6 g a.i./ha, depending on soil properties). Thus, these two species can be classified as highly sensitive to mesotrione carryover.

5. Conclusions

The overall results of this study indicate that all leafy vegetables tested, with the exception of endive, have a clear and high sensitivity to simulated mesotrione residues, in the following order: spinach > chard > lettuce > endive, with spinach appearing to be the most sensitive species. Consequently, spinach, chard, and lettuce cannot be a suitable substitute when maize fails, nor can they be sown after maize for silage production. In these cases, only endive can be a suitable choice without the risk of crop injury.

Mesotrione residues corresponding to the EC₂₀ value could be acceptable for the safe cultivation of leafy vegetables after maize, which means that a certain level of mild phytotoxicity can be expected and accepted. Bioassays conducted under controlled conditions provide valuable data on plant sensitivity and enable comparison between plants. However, it should be kept in mind that this conclusion only applies to the soil type and plant varieties used in this study. Under field conditions, environmental factors also play an important role and significantly influence herbicide persistence and the occurrence of residues in the soil. Therefore, future research should focus on field trials in soils with different properties and under different environmental conditions that may influence the resulting phytotoxicity.