Detection of 11 carbamate pesticide residues in raw and pasteurized camel milk samples using UHPLC-MS/MS: Method development, method validation, and health risk assessment

This study aimed to detect 11 carbamate pesticide residues in raw and pasteurized camel milk samples collected from the UAE using UHPLC-MSMS. A method was developed and validated by evaluating LOD, LOQ, linearity, extraction recovery, repeatability, intermediate precision, and matrix effect. Due to the high protein and fat content in camel milk, a sample preparation step was necessary to avoid potential interference during analysis. For this purpose, 5 different liquid-liquid extraction techniques were evaluated to determine their efficiency in extracting carbamate pesticides from camel milk. The established method demonstrated high accuracy and precision. The matrix effect for all carbamate pesticides was observed to fall within the soft range, indicating its negligible effect. Remarkably, detection limits for all carbamates were as low as 0.01 μ g kg −1 . Additionally, the correlation coefficients were >0.998, demonstrating excellent linearity. A total of 17 camel milk samples were analyzed, and only one sample was found to be free from any carbamate residues. The remaining 16 samples contained at least one carbamate residue, yet all detected concentrations were below the recommended MRLs set by Codex and the European Union pesticide databases. Nonetheless, it is worth noting that the detected levels of ethiofencarb in 3 samples were close to the borderline of the MRL. To assess the health risk for consumers of camel milk, the HI values of carbofuran, carbaryl, and propoxur were calculated. The hazard index (HI) values for these 3 carbamate pesticides were all below one, indicating that camel milk consumers are not at risk from these residues.


INTRODUCTION
To deal with the impact of overpopulation and the growing consumption of food products around the world, it is extremely significant to have a more sustainable agricultural system (Martin et al., 2020;Boudebbouz et al., 2022).Thus, agriculture relies mainly on the extensive use of pesticides to increase crop yield by shielding crops from unwanted pests and weeds (Dasriya et al., 2021;Boudebbouz et al., 2022).Despite the many advantages offered by using pesticides in agriculture, their residues have created serious risks due to their impact on the ecosystem and human health.In fact, less than 0.1% of the applied pesticides reach the targeted species, while the remaining can enter into the food chain and accumulate in plants and animals, posing a high risk to human health (Arias-Estévez et al., 2008;Akhtar, 2017;Xu et al., 2017).Pesticide residues in food products, owing to their high toxicity, can lead to a wide range of short and long-term harmful effects on human health.These effects may include genetic disorders, neurological damage, congenital disabilities, endocrine system disruption, and an increased risk of cancer (Docea et al., 2017;Tajdar-oranj et al., 2021;Ramezani et al., 2022).To safeguard human health and ensure food safety, international organizations such as Codex Alimentarius (CA) and European Union (EU) have established databases for the determination of maximum residue limits (MRLs) for pesticide residues.These limits serve as a comprehensive guide to assessing the potential risks to human health arising from pesticide residues (Neme and Satheesh, 2016;Dasriya et al., 2021;Ramezani et al., 2022).
Detection of 11 carbamate pesticide residues in raw and pasteurized camel milk samples using UHPLC-MS/MS: Method development, method validation, and health risk assessment Among different types of pesticides, carbamate pesticides have been used extensively in the past few decades due to their broad spectrum of biological activity, low mammalian toxicity, and low persistence (Fernández et al., 2000;Chowdhury et al., 2014).However, carbamates are inhibitors of the acetylcholinesterase enzyme which may result in neurotoxicity (Chowdhury et al., 2014).In addition to being toxic to the nervous system, some carbamates are mutagenic and carcinogenic (Ballesteros and Jurado-Sánchez, 2013).The remaining residues of carbamate in animal feed can deposit in fat and muscles and might be found in different food products like milk and cheese (Liu et al., 2013).
In recent years, the consumption of camel milk has increased at a global scale due to its well-known benefits to human health.Camel milk has a unique composition different from other ruminant milk, such as cow, sheep and goat milk, and it is closer to human milk (Oselu et al., 2022).It is low in cholesterol, saturated fat, and lactose; while high in minerals, antioxidants, unsaturated fatty acids, and vitamin C (Cheikh Ismail et al., 2022).It has natural medicinal and therapeutic properties such as antibacterial, antiviral, and antiallergic properties.Also, it is a booster for the immune system as it enhances the effectiveness of the immune system due to the higher quantities of lactoferrin, lactoglobulins, and lysozyme compared with other types of milk.Additionally, studies have suggested that camel milk may have potential benefits in relation to diseases such as cancer, diabetes, autism spectrum disorders (ASDs), hypertension, and skin diseases (Mihic et al., 2016;Khan et al., 2021;Oselu et al., 2022).Camel milk is a vital food product that holds a substantial popularity across the Gulf communities, including the United Arab Emirates (UAE).A study reveals that in the UAE, 57% of respondents consume less than a cup of camel milk daily, 24% consume 1-2 cups daily, 12% have 2-3 cups daily, and 7% drink more than 3 cups daily (Cheikh Ismail et al., 2022).Given this robust demand, camel milk farming has experienced significant growth, with many families now establishing their own camel farms.However, camel milk is like other food products that can be contaminated with pesticide residues.The accumulation of pesticides in camel milk and other food products occurs due to the remaining residues in feed, grass, foodstuff, water, and soil (LeDoux, 2011;Ramezani et al., 2022).The accumulation of pesticide residues can be magnified through the food chain leading to harmful effects.The high quantities of unsaturated fatty acids in camel milk increase the possibility of finding carbamate residues in the milk.Consuming contaminated milk with carbamate residues can pose serious threats to human health; therefore, it is essential to evaluate the levels of these residues.
Various analytical techniques have been used to determine and detect pesticide residues in milk, but the most used techniques are Gas Chromatography (GC) and Liquid Chromatography (LC) coupled to different detectors such as GC coupled to electron capture detector (GC-ECD) (Nag et al., 2007;Bulut et al., 2011), GC coupled to mass spectrometry (GC-MS) (Deti et al., 2014;Tsakiris et al., 2014;Shaker and Elsharkawy, 2015), and LC with tandem mass spectrometry (LC-MSMS) (Saito et al., 2008).Milk is rich in proteins and fats, which can interfere with the analytical process making it difficult to detect pesticide residues (Imamoglu and Oktem Olgun, 2016;Ramezani et al., 2022).Therefore, a sample preparation step is needed before the analysis to extract pesticide residues from the milk matrix and concentrate them, making the analytical process simple and easy.Different extraction techniques have been reported like solid-phase extraction (Jaraczewska et al., 2006), liquid-liquid extraction (Hassine et al., 2012), QuEChERS technique (Jeong et al., 2012), Soxhelt extraction (Zhou et al., 2011), and solid-phase microextraction (Röhrig and Meisch, 2000).
The literature contains numerous studies on the detection of pesticide residues in cow, buffalo, and goat milk.However, a very limited number of studies are available on the detection of pesticide residues in camel milk (Shahzadi et al., 2013;Al-hawadi et al., 2021;Philip et al., 2022).In the UAE, the consumption of camel milk (whether unpasteurized or pasteurized) is immense, and some families consume it daily (Cheikh Ismail et al., 2022).Yet, there is a lack of information on levels of carbamate residues in the camel milk of the UAE.Therefore, the present study aimed to investigate the presence of 11 different carbamate pesticides (carbofuran, carbaryl, propoxur, aminocarb, phenmedipham, ethiofencarb, desmedipham, fenoxycarb, pirimicarb, bendiocarb, and methiocarb) in various camel milk samples.These samples were either purchased from local markets (pasteurized) or collected from different farms in the UAE (unpasteurized).The pasteurized samples underwent the high-temperature short time (HTST) pasteurization method.For sample preparation, different liquid-liquid extraction techniques were tested to determine the most efficient extraction technique.Levels of carbamate residues were detected using ultra-high-performance liquid chromatography coupled to a triple quadrupole mass spectrometer (UHPLC-MS/MS).Detected levels of carbamate residues were compared with the MRLs established by the CA and EU pesticide databases.Among the studied carbamates, carbaryl, desmedipham, and phenmedipham have an MRL of 0.05 mg kg −1 , while the remaining ones have an MRL of 0.01 mg kg −1 (Codex Alimentarius, 2021;European Commission, 2022).Additionally, the health risk of some carbamate pesticides was assessed for camel milk consumers.

Reagents and chemicals
Chemical standards of all carbamate pesticides were purchased from Dr. Ehrenstorfer (Augsburg, Germany).Primicarb d6 internal standard (IS), ammonium formate, and acetic acid were purchased from Sigma-Aldrich (St. Louis, MO, USA).Solvents used in preparation of the standards and extraction, such as acetone, cyclohexane, dichloromethane, and ethyl acetate, were supplied by Sigma-Aldrich (St. Louis, MO, USA).Acetonitrile and methanol used in UHPLC-MS/ MS were purchased from Honeywell (Seelze, Germany).Millex PTFE syringe filters were procured from Merck Millipore (Carrigtwohill, Ireland).

Sample collection
Eight pasteurized camel milk samples (including 3 low-fat camel milk) belonging to 2 different brands were purchased from local markets in Al Ain, UAE.Nine raw camel milk samples were collected from different farms in the UAE.A total of 17 camel milk samples were analyzed by UHPLC-MS/MS in this study.

Standards
Individual stock solutions of each standard (250 mg kg −1 ) were prepared in suitable organic solvents and stored at −18°C.Working standard solutions were then prepared by diluting the stock solutions in methanol.

Extraction procedures
Five different liquid-liquid extractions were tested for their ability to extract carbamate residues from camel milk samples.In all extractions, samples were vortexed for 5 min, then centrifuged at 6000 rpm for 15 min at 5°C.The first 3 extraction techniques differ only by the type of solvent used for extraction.The general scheme of the first 3 extractions was carried out according to Imamoglu and Oktem Olgun (2016) with some modifications.

Extractions 1, 2, and 3
A volume of 15 mL of camel milk sample was transferred to a 50 mL polypropylene centrifuge tube.Then, 100 μL of 10 ppm IS was pipetted into the milk.Next, 30 mL of organic solvent or a mixture of solvents was added to the milk.Finally, 600 μL of acetic acid was pipetted into the mixture.The sample was vortexed and centrifuged.The upper phase was taken into a 50 mL glass tube and evaporated.The residue was reconstituted with 1 mL of mobile phase B (methanol: acetonitrile 1:1).Then, the extract was filtered through a 0.45 μm PTFE syringe filter and was ready for the UHPLC-MS/MS analysis.
The types of solvents used for extractions 1, 2, and 3 were as follow: 30 mL of acetonitrile for extraction 1, a combination of 15 mL acetone and 15 mL cyclohexane for extraction 2, and a combination of 15 mL acetonitrile and 15 mL ethyl acetate for extraction 3.

Extraction 4
A volume of 15 mL of camel milk sample was transferred to a 50 mL polypropylene centrifuge tube.Then, 100 μL of 10 ppm IS was pipetted into the milk.After that, 30 mL of acetonitrile was added to the milk.Finally, 5 g of NaCl was added to the mixture.The sample was vortexed and centrifuged.The upper phase was taken into a 50 mL glass tube and evaporated.The residue was reconstituted with 1 mL of mobile phase B. Then, the extract was filtered through a 0.45 μm PTFE syringe filter and was ready for the UHPLC-MS/MS analysis.

Extraction 5
The fifth tested extraction was carried out according to Fagnani et al. (2011) with some modifications.A volume of 15 mL of camel milk sample was transferred to a 50 mL polypropylene centrifuge tube.Then, 100 μL of 10 ppm IS was pipetted into the milk.Next, a mixture of 20 mL acetonitrile and 5 mL methanol was added to the milk.The mixture was vortexed and centrifuged.The upper phase was partitioned with 50 mL dichloromethane.Dichloromethane extract was collected in a glass tube and evaporated.The residue was reconstituted with 1 mL of mobile phase B. Then, the extract was filtered through a 0.45 μm PTFE syringe filter and was ready for the UHPLC-MS/MS analysis.

UHPLC-MS/MS analysis
The analysis of carbamate residues in camel milk was performed using a Shimadzu UHPLC instrument (Nexera-i LC-2040C 3D, Kyoto, Japan) coupled to a 8030 Shimadzu triple quadrupole LCMS (Kyoto, Japan) with electrospray ionization (ESI).An ACQUITY UPLC BEH C18 column (2.1 mm x 150 mm x 1.7 μm) supplied by Waters (Milford, USA) was used.The column temperature was set at 40°C.A gradient elution program was applied, utilizing aqueous and organic mobile phases: 10 mM ammonium formate in water, pH = 3 (mobile phase A), and acetonitrile and methanol (1:1) (mobile phase B).The flow rate was set to 150 μL/min and the injection volume was 10 μL.The total run time for the analysis was 25 min.

Optimization of MS conditions for UHPLC-MS/MS
To achieve high sensitivity, MS conditions were optimized.A 1.0 mg kg −1 stock solution of each individual standard and IS was prepared and injected into the UHPLC-MS/MS system.The obtained optimum MS parameters are summarized in Table 1S.

Validation study
Method validation of the limits of detection (LOD), limits of quantitation (LOQ), linearity, extraction recovery, repeatability, intermediate precision, and matrix effect were evaluated for all 11 carbamate compounds.
In the present study, calibration curves were prepared in neat solvent and in matrix to evaluate the extraction method and the matrix effect.The number of points on each calibration curve varied for different carbamates with 8, 9, or 10 points at concentrations of: 0.00001, 0.00005, 0.0001, 0.0005, 0.001, 0.005, 0.01, and 0.05, 0.1, and 0.5 mg kg −1 .The specific concentration ranges used for constructing calibration curves of each carbamate pesticide are presented in Table 1.A fixed concentration of 1.0 mg kg −1 of pirimicarb-d6 IS was added to all prepared standard solutions.
The sensitivity and accuracy of the developed method for detecting and quantifying 11 carbamate pesticides in camel milk samples were evaluated by calculating the LOD and LOQ.The determination of LOD and LOQ involved measuring the signal-to-noise ratios (S/N) for each carbamate pesticide.The S/N ratio enables the determination of the minimum concentration level of the analyte that can be reliably detected and quan-tified.LOD and LOQ were determined as the lowest detected analyte concentrations resulting in S/N ratios greater than 3 and 10, respectively.
Matrix effect (ME) was determined by comparing the slopes of the solvent-based calibration curve and the matrix-matched calibration curve using the following formula:

%. 100
The ME can either be negative or positive and is categorized as soft, medium, or strong.Soft ME is observed when the ME value falls between −20% and 20%, whereas medium ME occurs within the range of −50% to −20% or 20% to 50%.On the other hand, strong ME is identified when the value is less than −50% or greater than 50% (Kim et al., 2023).It is essential to note that the ME is deemed significant when its value exceeds ± 20%, while it is considered negligible when the value is ≤ ± 20%.In cases where the ME is negligible, a solvent-based calibration can be utilized instead of a matrix-based calibration (European Commission, 2021).
For the recovery study, blank samples of camel milk were spiked with carbamate standards at 2 different concentrations of 0.0005 mg kg −1 and 0.005 mg kg −1 (n = 3), then extractions 3 and 5 were performed.
Precision was determined based on repeatability (intraday) and intermediate precision (interday) for both extractions 3 and 5. To assess repeatability, 6 blank camel milk samples were spiked at 2 concentration levels of 0.0005 mg kg −1 and 0.005 mg kg −1 , and analyzed on the same day.For evaluating intermediate precision, 12 blank camel milk samples were fortified at 0.0005 mg kg −1 and 0.005 mg kg −1 , and the analysis was conducted on 2 different days (6 analyses per day).

Health risk assessment study of pesticide
In the current risk assessment study, a probabilistic approach using Monte-Carlo simulation was used (Schuhmacher et al., 2001).The distribution of pesticide residues is assumed to be lognormal.One hundred thousand replications were used to compute the hazard quotients (HQ).The 95th percentile was used to assess risk.The HQ was computed for adults using the following method.First, the critical daily intake (CDI) was calculated as follow: where C represents the residue concentration of the carbamate pesticide in milk samples (measured in mg kg −1 ), IR is the milk consumption rate assumed to be 0.220 kg/day, and BW is the body weight taken as 70 kg for adults.Next, the non-cancerous risk hazard quotient HQ was computed using the formula: where RfDo represents the pesticide chronic oral reference dose obtained from US-EPA (https: / / iris .epa.gov/AtoZ/ ?list _type = alpha).Finally, the hazard index (HI) was computed by summing the HQ values of different pesticides studied.According to Boobis et al. (2008), pesticides do not cause non-carcinogenic risk when HQ or HI < 1, while pesticide residues can cause non-carcinogenic risk if HQ or HI > 1.The RfDo for carbofuran, carbaryl, and propoxur are taken to be 0.005, 0.1, and 0.005 (mg/kg/day), respectively.However, the RfDo for the remaining carbamate pesticides, for which their residues were found in camel milk samples, is not available.Hence, the HI was only calculated for carbofuran, carbaryl, and propoxur.

Extraction procedures
To determine the most efficient extraction procedure, all 5 different extraction methods were applied to the same camel milk sample.Figure 1 presents the total number of detected carbamate pesticides when employing the 5 different extraction techniques.Extractions 1 and 4 yielded the lowest numbers, with detections of 2 and 3 carbamates, respectively.Extraction 2 led to the detection of 5 carbamates.Extractions 3 and 5 demonstrated the highest efficiency, resulting in the detection of 8 and 7 carbamates, respectively.Consequently, the validation of the extraction method was conducted specifically for both extractions 3 and 5 only.
As carbamates are semi-polar compounds (Matošević and Bosak, 2020), an extraction solvent with moderate polarity appears to be more effective.In extraction 3, a combination of acetonitrile and ethyl acetate was utilized as the extraction solvent, leading to a good extraction efficiency.The intermediate polarity of ethyl acetate outperformed the use of a polar solvent alone, as observed in extractions 1 and 4. Extraction 5 also showed good extraction efficiency due to the utilization of dichloromethane, which effectively extracted the targeted analytes.On the other hand, extraction 2 utilized a combination of acetone and cyclohexane as the extraction solvent.While acetone is polar, the inclusion of cyclohexane, a strong nonpolar solvent, shifts the overall polarity toward non-polar rather than moderate polarity, leading to less effective extraction.The polarity index (PI) of the solvents used in the extraction techniques is summarized in Table 2S.Extractions 3 and 5 both detected the same types of carbamate pesticides.However, extraction 3 showed an additional carbamate residue, bendiocarb, compared with extraction 5. Further details regarding the concentration and types of detected carbamate residues using the 5 different extraction techniques are presented in Table 3S.Based on the results, it can be inferred that extraction 3 exhibited superior efficiency in terms of the number of detected carbamates compared with all other types of extractions.

Method validation
The proposed method was validated as per to the European SANTE/2020/12830 guidelines (European Commission, 2021) by determining linearity, LOD, LOQ, recovery, and precision.The linearity of the method was assessed using calibration curves, with 3 blank milk samples spiked with standards at concentrations ranging from 0.00001 to 0.5 mg kg −1 .Linearity was defined by the correlation coefficient (R 2 ).As shown in Table 1, our study showed excellent linearity with correlation coefficients greater than 0.999 for all analyzed carbamate pesticides, except for propoxur and carbaryl, which still demonstrated very good R 2 values exceeding 0.998.These findings outperform the results reported by Bogialli et al. (2004), who investigated propoxur, pirimicarb, carbofuran, and carbaryl.Bogialli et al. (2004) reported R 2 values for these carbamates ranging from 0.9911 to 0.9999.In contrast, our R 2 values for the same carbamates ranged from 0.9984 to 1.0000.The same carbamates were also investigated by Liu et al. (2013), and the reported R 2 exceeded 0.99 only.While our study, encompassing 11 carbamate pesticides, consistently achieved R 2 surpassing 0.998 for all analyzed compounds.This underscores the improved precision and robustness of our developed method compared with previous studies.
The calculated LOD and LOQ values are also presented in Table 1.For all analyzed carbamate pesticides, the LOD was approximately 0.01 μg kg −1 , meeting the S/N criteria of >3.The LOQ values ranged between of 0.03 and 0.04 μg kg −1 , adhering to the S/N criteria of >10.These results indicate that our developed method possesses the sensitivity required to detect and quantify trace amounts of carbamate pesticide residues in camel milk.Our LOD and LOQ values exhibit significant improvements over those reported by Bogialli et al. (2004).They reported LOD values of 5.0 μg kg −1 , 3.0 μg kg −1 , 4.0 μg kg −1 , and 4.0 μg kg −1 for pirimicarb, propoxur, carbofuran, and carbaryl, respectively, with LOQ values of 8.0 μg kg −1 , 3.0 μg kg −1 , 4.0 μg kg −1 , and 4.0 μg kg −1 .In contrast, our study achieved an approximate LOD value of 0.01 μg kg −1 for these carbamates, along with LOQ values of 0.031 μg kg −1 , 0.032 μg kg −1 , 0.034 μg kg −1 and 0.041 μg kg −1 , respectively.Our LOD and LOQ values also outperformed those reported by Liu et al. (2013).Among the carbamates they studied, carbofuran was the only one with LOD and LOQ values similar to ours.While the reported LOD values for propoxur, pirimicarb, and carbaryl were 0.023 μg kg −1 , 0.017 μg kg −1 , and 0.037 μg kg −1 , respectively, with LOQ values of 0.077 μg kg −1 , 0.057 μg kg −1 , and 0.12 μg kg −1 , respectively.This highlights the enhanced sensitivity of our developed method compared with previous studies.The recovery was determined at 2 concentrations of 0.0005 mg kg −1 and 0.005 mg kg −1 for both extractions 3 and 5.The recovery data of extraction 3 are shown in Table 2.The obtained recovery of each carbamate pesticide was satisfactory, ranging between 88% to 103%.This range falls within the acceptable range (60-120%) according to SANTE/2020/12830 guidelines (European Commission, 2021).Table 3 shows the recovery results of extraction 5. Acceptable recoveries ranging from 89% to 105% were obtained for carbofuran, carbaryl, propoxur, aminocarb, ethiofencarb, fenoxycarb, pirimicarb, bendiocarb, and methiocarb.However, phenmedipham and desmedipham showed low recovery values that fell below the acceptable range at both concentrations.At 0.0005 mg kg −1 , phenmedipham and desmedipham had recoveries of 19% and 18%, respectively.Meanwhile, at 0.005 mg kg −1 , the obtained recovery of phenmedipham and desmedipham was 27% and 31%, respectively.These results indicate that extraction 5 failed to effectively extract phenmedipham and desmedipham from camel milk samples.Thus, it can be concluded that extraction 3 is more reliable for extracting carbamate pesticides from camel milk.
The results obtained indicate that extraction 3 is more accurate, precise, and reliable compared with extraction 5. Therefore, extraction 3 was selected for monitoring carbamate pesticide levels in camel milk, which is critical for maintaining food safety and safeguarding public health.

Matrix effect
ME can significantly affect the identification and quantification of an analyte, thereby affect the overall performance of the analysis.The ME results, as shown in Table 1, indicate that the values for all carbamate pesticides ranged from −1.085% to 10.047%.While most carbamates exhibited a positive value, excep-tions were observed for aminocarb and bendiocarb, which showed negative values of −1.085 and −0.997, respectively, corresponding to soft signal suppression.Notably, all obtained ME values fall within the range of the soft ME, suggesting that the ME is not significant.Therefore, calibration with standards in solvent can be utilized, which is easier to construct, simpler, and less time-consuming compared with calibration with standards in the matrix.Matrix-induced effects can be reduced with sample preparation step.Additionally, the use of internal standards can compensate for the dependence of results obtained on the matrix and further reduce ME (Zhou et al., 2017).Consequently, it can be concluded that the extraction process used to extract carbamate pesticides from camel milk samples, coupled with the use of an internal standard, effectively eliminated matrices or other interferences in the sample extract.

Determination of carbamate residues in camel milk
After validating the developed method, the concentrations of 11 carbamate pesticide residues were evaluated in 17 camel milk samples and compared with the MRLs established by CA and the EU.The analysis of all camel milk samples was conducted after extracting carbamate pesticides using extraction procedure 3. The concentrations of the detected carbamate residues in the 17 camel milk samples are presented in Table 4.These samples can be categorized as follow: a) Samples 1-6 are pasteurized camel milk samples from Brand 1, with 1-3 being low-fat camel milk and 4-6 being full-fat camel milk, b) Samples 7 and 8 are pasteurized full-fat camel milk samples from Brand 2, and c) Samples 9-17 are raw camel milk samples collected from different farms.Among the carbamates studied, phenmedipham, desmedipham, and methiocarb, were not detected in any of the samples and were consequently excluded from the table.
The concentrations of the detected carbamate residues in the camel milk samples ranged from 0.345 μg kg −1 to 9.509 μg kg −1 , all falling below the MRLs set by CA and the EU.However, it's important to highlight that in 3 raw camel milk samples, ethiofencarb concentrations were nearly at the threshold of its MRL (10 μg kg −1 ), with values of 9.272 μg kg −1 in Sample 11, 9.076 μg kg −1 in Sample 14, and 9.509 μg kg −1 in Sample 17.Moreover, our findings indicate that raw camel milk samples (unpasteurized) contain a higher number and concentration of carbamate residues compared with the pasteurized camel milk samples.This observation suggests that the pasteurization process most likely led to the thermal degradation of some pesticides.Previous studies have reported similar effects of temperature during pasteurization, leading to the degradation of different types of pesticides and a decrease in their concentrations (Neme and Satheesh, 2016;Singh and Nelapati, 2017;Pardío Sedas et al., 2021).Moreover, the variations in residual levels of carbamates may also be attributed to differences in the quality control of animal feed.Smaller farms owned by local families appear to be subject to less control and regulation regarding the animal feed's quality, unlike larger farms operated by different companies that supply pasteurized milk.
Among the pasteurized camel milk samples, only Sample 2 was found to be free from any carbamate residues.In all the remaining pasteurized camel milk samples from Brand 1, carbofuran was detected, indicating its persistence.The detected concentrations of carbofuran were notably higher in full-fat camel milk samples (1.672 μg kg −1 in Sample 4, 2.805 μg kg −1 in Sample 5, and 2.118 μg kg −1 in Sample 6) compared with low-fat camel milk samples (1.335 μg kg −1 in Sample 1 and 0.872 μg kg −1 in Sample 2).This variation in concentrations can be attributed to the higher fat content in full-fat camel milk, which promotes the accumulation of carbofuran due to its lipophilic properties.Furthermore, in Sample 6, alongside carbofuran, fenoxycarb was detected at level of 1.251 μg kg −1 , suggesting the possibility of an outlier or variations arising from different batches or feed sources.Samples 7 and 8, both belonging to pasteurized camel milk from Brand 2, contained ethiofencarb at a concentration of 3.716 and 8.201, respectively.The variations in residual levels of carbamates within the same brand can likely be attributed to differences between batches within a single brand.Such variations may arise from differences in processing methods, storage conditions, and sourcing.
Various studies have reported the presence of carbamate residues in milk.For instance, Shahzadi et al. (2013) detected carbofuran residues in several types of milk, including cow, buffalo, goat, sheep, and camel, in Lahore, Pakistan.The residual levels of carbofuran in these samples ranged from 0.10 -0.84 mg kg −1 in cow milk, 0.09 -0.60 mg kg −1 in buffalo milk, 0.02 -0.74 mg kg −1 in goat milk, 0.11 -0.623 mg kg −1 in sheep milk, and 0.01 -0.55 mg kg −1 in camel milk.Notably, most of these samples exceeded the carbofuran MRL of 0.01 mg kg −1 .In a different study from China, Liu et al. (2013) investigated 7 carbamate residues in bovine milk.Out of these, only 4 residues (pirimicarb, propoxur, carbaryl and fenobucarb) were found in 6 samples of the 20 samples studied.Their concentrations ranged between 0.0013 -0.0032 mg kg −1 , all below the MRL set by Japan and the EU.Additionally, Chandrakar et al. ( 2020) assessed carbaryl residues in 200 milk samples sourced from India.Out of these samples, 55 (or 27.5% of the total) tested positive for carbaryl residues.The concentrations varied with some samples containing trace amounts below the MRL of 0.05 mg kg −1 set by CA, while others presented concentrations as high as 0.287 mg kg −1 , notably exceeding the MRL.Fagnani et al. (2011), on the other hand, have investigated milk samples obtained from 11 distinct farms and one milk cooler in Brazil.The pesticides that were identified included carbofuran (present in 25% of the samples), aldicarb (present in 16.67% of the samples), and carbaryl (present in 8.33% of the samples).It is worth noting that in certain samples, the simultaneous

Assessing the human health risk
The levels of carbamate residues detected in camel milk were found to be below the MRL set by CA and the EU.Despite this, it is worth mentioning that numerous families in the UAE consume camel milk on a daily basis.Therefore, it is important to assess the potential human health risk from carbamate pesticides found in camel milk samples.For this purpose, the noncarcinogenic health risk of the first 3 pesticides, carbofuran, carbaryl, and propoxur were studied.Figure 2 displays the distributions of HQ for these pesticides, obtained from Monte-Carlo simulations.Figure 2A illustrates the HQ distribution for carbofuran found in pasteurized milk, with the 95th percentile at 0.00199, which is significantly lower than the safety limit of 1. Figure 2B shows the HQ distribution for carbofuran found in raw camel milk, with the 95th percentile at 0.0022.Furthermore, Figures 2C and 2D depict the distribution HQ for carbaryl and propoxur found in raw camel milk, respectively.The 95th percentiles were 0.0002 for carbaryl and 0.0043 for propoxur.It is important to note that none of the pesticides' HQ values exceeded 1.Additionally, the HI, computed as the sum of the 3 quantiles, was equal to 0.0067, which is less than 1.0.Therefore, the consumption of raw camel milk is unlikely to pose a non-carcinogenic health risk.In a parallel analysis focused on children (taking an average body weight of 15 Kg and exposure duration of 6 years), we conducted the same health risk assessment for the same 3 pesticides (carbofuran, carbaryl, and propoxur).The results mirrored those for adults, with none of the pesticides' HQ values exceeding 1 (see Figure S1).Therefore, the consumption of both pasteurized and raw camel milk in unlikely to pose noncarcinogenic health risk for children as well.The health risk assessment for other residues found in camel milk was not performed due to the lack of evaluated RfDo values for those compounds.

CONCLUSION
This study aimed to detect 11 different carbamate pesticide residues in the camel milk of the UAE using UHPLC-MS/MS.For sample preparation, a simple liquid-liquid extraction technique was used.The developed method was validated and used to determine the presence of carbamate residues in 17 raw and pasteurized camel milk samples.The results obtained indicated that raw camel milk contained a higher number and higher concentration of carbamate residues compared with pasteurized camel milk.Among the 17 samples analyzed in the study, only one sample showed no detectable residues of carbamate pesticides, whereas the remaining samples contained at least one type of carbamate pesticide residue.Carbofuran, carbaryl, and ethiofencarb were present in all of the raw camel milk samples.However, the detected levels of residues for all these carbamate pesticides were found to be below the MRL set by CA and the EU.HI-values were calculated for carbofuran, carbaryl, and propoxur to assess their potential health risk on camel milk consumers.The calculated HI values for these 3 carbamates were below 1, indicating that they are unlikely to cause noncarcinogenic risk to consumers.

Figure 1 .
Figure 1.Number of detected carbamate pesticides by the 5 tested extraction procedures.

Figure 2 .
Figure 2. HQ due to levels of (A) carbofuran residues in pasteurized camel milk samples; (B) carbofuran residues in raw camel milk samples; (C) carbaryl residues in raw camel milk samples; and (D) propoxur residues in raw camel milk samples.