The castration of male piglets, a longstanding and widespread practice in most countries, aims to prevent the occurrence of boar taint in the meat of certain male pigs (Bonneau & Weiler, 2019). However, the practice has attracted criticism and has been identified as an animal welfare issue (Sodring et al., 2020). Boar taint is an unpleasant odour found in the carcasses of entire males (EFSA, 2004; Fredriksen et al., 2011) caused mainly by the accumulation of androstenone and skatole (Claus et al., 1994), and commonly found in the meat of approximately 5 to 10% of entire male pigs (Baek et al., 1997; Aluwe et al., 2020). The accumulation of these compounds in the adipose tissue during pubertal development of growing animals leads to off-odours and off-flavours in pork, creating a negative perception of pork quality among consumers (Bonneau & Weiler, 2019).

In prominent pig-producing nations like Germany, France, and Denmark, approximately 75% to 93% of commercial male piglets undergo surgical castration (Van Ferneij, 2022). In other countries, most of the male pigs are raised without undergoing castration (Spain, 80%; Portugal, 85%; Ireland, 100%; United Kingdom, 98%), and are slaughtered before reaching puberty, with around five months of age and 100 kg of body weight. (De Briyne et al., 2016). This practice is seen as the most sustainable in reducing androstenone levels, by feeding rapid-growth diets allowing entire males to reach market weight before becoming sexually mature (Squires et al., 2020). However, there is the risk of some animals reaching puberty earlier and tainting pork in consequence.

The concentrations of skatole and androstenone vary between individuals, with a dispersion caused by several factors such as sex, age, feed, and genotype, among others (Bonneau & Weiler, 2019; Squires et al., 2020). If the concentrations of these compounds are small, the average consumer will not detect the off-odour. But this aspect is still under debate in the literature, with authors proposing different limit values for consumers not being able to detect boar taint (Lundstrom et al., 2009; Prusa et al., 2011; Bonneau & Chevillon, 2012; Borrisser-Pairo et al., 2016; Verplanken et al., 2017; Aluwe et al., 2018; Christensen et al., 2019). These limits can be defined by the sensory detection thresholds. The detection threshold is the lowest concentration of a substance in a medium relating to the lowest perception at which a stimulus is detected (ASTM:E679-04, 2011); however, sensory thresholds may not be generalised to cover different groups of people over extended periods of time (Annor-Frempong et al., 1997).

The objectives of this study were: a) to gain insights into the occurrence of boar taint within the Portuguese pork production industry by analysing levels of skatole and androstenone in the fat; b) to estimate thresholds for skatole and androstenone among Portuguese consumers. With these results, it was also intended to determine if a significant portion of samples exhibit off-odours by cross linking collected boar taint data with determined perception thresholds.

The skatole and androstenone levels in fat are shown in Table 1. A total of 140 samples were assessed and categorised based on the sampling location (abattoirs A, B, and C) and gender. In this survey, 140 pigs (100%) had measurements of skatole above the limit of detection (LoD), and 72 pigs (51.4%) had androstenone below the LoD. The peak concentrations detected were 169 ng/g for skatole and 1215 ng/g for androstenone, respectively.

Mean and confidence interval (95%) for skatole concentration for all samples was 36.0 ± 4.7 ng per gram of liquid fat. The average concentration of androstenone in all the 140 samples was 64.5 ± 21.3 ng per gram of liquid fat.

Skatole levels were significantly different (p<0.05) between males and females (40.3 ± 5.9 ng/g and 24.7 ± 5.4 ng/g respectively). As expected, there was a substantial discrepancy in the androstenone values obtained between males and females (84.1 ± 28.4 ng/g and 11.7 ± 0.7 ng/g respectively). In fact, the median value for females aligns with the calculated value for cases falling below the detection limit (Fig. 1).

Discrepancies were observed in individual odour thresholds, with detection limits ranging from 17.7 to 141 ng/g for skatole (Table 2) and 106.1 to 3394.1 ng/g for androstenone (Table 3). The group’s BET for skatole was determined to be 35.4 ng/g, while for androstenone, it was 566.3 ng/g.

Considering the 140 samples under study and using the determined BET group values, it was found that the determined skatole content exceeded the BET values in 46 samples (28.8%), while only one sample (0.7%) exceeded the BET value for androstenone (Fig. 2).

This study is grounded in an observational study approach, utilising various carcasses sourced from the slaughterhouse. Its primary objective is to establish the prevalence of boar taint, although with a limited number of samples, without controlling for prior feeding programs, production systems during the growing-finishing period, or the specific crossbreeds included in the sample. Very low concentrations were not possible to measure due to limitations in chemical analysis techniques, and are typically referred to as being below the LoD. These data points are frequently censored in statistical assessments and replaced with a fixed value. According to Croghan & Egeghy (2003), opting to replace zero or LoD values with LoD divided by the square root of 2 (LoD/) stands out as the most appropriate choice when addressing measurements that fall below the LoD. This approach is favoured due to its tendency to yield a diminished bias. Given that the values acquired for skatole and androstenone do not conform to a normal distribution supplementary to the mean and standard deviation, this study also presented median values and interquartile ranges. The low sample size (140 adipose tissue samples collected from 140 animals) can also be addressed, representing a limitation of this study and indicating its small-scale nature. A study conducted by Prusa et al. (2011), aimed at defining the prevalence of skatole and androstenone in the fat of pigs within the United States. The findings revealed an average of 197.3 ng/g of skatole and 2363 ng/g in entire males, while female pigs exhibited levels of 34.1 ng/g for skatole and 120 ng/g for androstenone. When comparing these results to the ones obtained in the current study, significantly higher levels of boar taint were observed in the male pigs. The pigs processed in the Iberian market are usually slaughtered between five and six months old, which is insufficient time to reach sexual maturity. In contrast, the pigs examined in the United States were sourced from farms and had attained full maturity, hence the expected presence of elevated quantities of skatole and androstenone. Concerning the skatole values observed in female pigs, they seem to be closely aligned with the results from our study. However, it is noteworthy that the androstenone levels in the females appear unusually elevated. In a study by Walstra et al. (1999), skatole and androstenone levels were assessed in male and female pigs across six European countries (Spain, France, Netherlands, United Kingdom, Denmark and Sweden). The results from this study indicated higher values when compared to the results obtained in our study. In average skatole and androstenone values were reported to be 70 ng/g and 30 ng/g, respectively. For entire males, the skatole averages ranged from 100 ng/g to 170 ng/g, and androstenone levels varied between 800 ng/g and 1270 ng/g. Borrisser-Pairo et al. (2016) published a study examining the occurrence of boar taint in commercial pigs from Spain. They specifically included entire males with comparable ages and weights to those in our study sample. In the study, involving 903 animals, the average skatole level in the fat was 40 ng/g, identical to the values found in the males analysed in our study. However, concerning androstenone, the overall values reported by this author were 200 ng/g, higher than those detected in 102 males examined in this study. Assessing boar taint in adipose tissue, melted fat, or pure fat may yield disparate values (with lower values typically in adipose tissue, attributed to the presence of additional constituents). This disparity complicates direct comparison of values found in literature and underscores the importance of harmonising boar taint detection methods, since additional validation of current methods and standardisation of methodologies are required to accurately quantify boar taint compounds (Haugen et al., 2012). The presence of androstenone doesn’t seem to be limited exclusively to entire males. It is expected that, androstenone should only be present in males, but there are reports suggesting its presence in females as well, as noted by van Oeckel et al. (1996) and Prusa et al. (2011). This phenomenon may be attributed to the conversion of progesterone into androstadienone and subsequently into androstenone (Robic et al., 2014). Also, while skatole levels are typically higher in intact male pigs, it’s worth noting that skatole can also be found in barrows and females, albeit in smaller quantities (Aldal et al., 2005). The primary reason for the elevated skatole levels in intact male pigs compared to barrows or gilts is the reduced hepatic degradation of skatole. This reduction is attributed to the inhibition of hepatic metabolism by androstenone, testosterone, or 17β-estradiol (Zamaratskaia & Squires, 2009; Bonneau & Weiler, 2019). Additionally, environmental conditions and housing facilities can play a role in skatole levels, since it can be absorbed through the skin via contact with faeces or through the lungs by inhalation (Thomsen et al., 2015). Moreover, the dietary composition plays a crucial role in regulating intestinal skatole synthesis. Adding fermentable carbohydrates can reduce skatole levels by changing gut microflora activity and composition, leading to less L-tryptophan availability in the colon and decreased intestinal lumen apoptosis (Bee et al., 2020; Squires et al., 2020). In this study, it was also possible to observe a positive correlation between androstenone and skatole concentrations (r=0.5382, p<0.01). This correlation has also been reported by Walstra et al. (1999) and Borrisser-Pairo et al. (2016). Certain studies have reached the conclusion that the existence of androstenone hampers the metabolism of skatole (Doran et al., 2004; Panella-Riera et al., 2008), potentially elucidating the observed positive correlation between these two compounds responsible for boar taint.

With the purpose of comparing and classifying the values obtained in the study of the prevalence of boar taint, the olfactory thresholds of skatole and androstenone were determined using Portuguese consumers. Sensory thresholds cannot be universally applied to diverse groups of people over extended periods, as their success in detecting specific substances in threshold tests depends on momentary individual sensitivity, which has been shown to change over time (Annor-Frempong et al., 1997). It is important to highlight that the optimal approach for establishing boar taint sensitivity thresholds would involve utilising pork fat free of androstenone and skatole initially instead of liquid vaseline, in order to imitate the meat odour as closely as possible, taking into consideration any potential interference it may introduce. Comparing this study group BET with a publication by Annor Frempong et al. (1997), the calculated BET values were slightly lower, being 26 ng/g for skatole and 426 ng/g for androstenone. The lowest individual odour thresholds of this study were obtained by women. Worth notice is that females seem to exhibit a higher sensitivity to androstenone: a study carried out in Spain and Germany determined that 31% of the Spanish and 18% of German respondents to a survey revealed higher sensitivity to androstenone, and among these sensitive individuals, a notably larger proportion were women (Weiler et al., 2000). Roughly 50% of the adult human population cannot detect the characteristic scent when they inhale androstenone (Wysocki et al., 1989). This finding is corroborated by a recent investigation by Klyuchnikova et al. (2022), which evaluated the ability of 807 individuals residing in central Russia to perceive the odour of androstenone; it was found that 49% of these participants could be classified as potentially having a specific inability to detect this substance through smell. On the other hand, skatole is perceived by 99% of consumers (Weiler et al., 2000), thus explaining why panellists exhibit a higher level of proficiency in detecting odour.

Industrial procedures for quality control concerning meat classification according to the level of boar taint are usually based on consumer’s perception threshold values. Thus, it is essential to set threshold values for androstenone, skatole, or boar taint levels, determining whether these carcasses should be categorised as having boar taint or not. This threshold should be determined considering consumer sensitivity, acceptability, and the prevalence of boar taint in the swine population (Aluwe et al., 2018; Font-i-Furnols et al., 2020). From the results, given the lowest BET, it can be inferred that panellists with higher olfactory sensitivity would be able to detect the odour of skatole in 114 of the tested samples (constituting 81.4%), while they would detect the odour of androstenone in 19 samples (equivalent to 13.6%). Nonetheless, it’s important to note that factors such as panel training can significantly influence the odour detection thresholds by a substantial margin. Additionally, test conditions, such as the size of sniff bottles employed, can be relevant to the results and some how justifies the variability found on the published data (Annor-Frempong et al., 1997). The thresholds for evaluating boar taint, frequently discussed in the literature, exhibit variability. This divergence might arise from the fact that the current acceptability standards are mostly based on the concentrations of these odour compounds in uncooked meat (Annor-Frempong et al., 1997). Researchers have suggested a range between 500 and 1000 ng/g for androstenone as a potential limit (Prusa et al., 2011; Borrisser-Pairo et al., 2016). In cases where skatole is absent, some studies propose even higher limits, falling within the range of 2000 to 3000 ng/g (Bonneau & Chevillon, 2012). As for skatole, the acceptable limits can differ, with some sources suggesting a range of 100 to 200 ng/g (Prusa et al., 2011; Borrisser-Pairo et al., 2016). However, Aluwe et al. (2018) has reported that skatole values exceeding 100 ng/g are not considered acceptable, regardless of the androstenone level, probably caused by the synergistic relationship between androstenone and skatole when it comes to defining the intensity of odours (Annor-Frempong et al., 1997). Furthermore, Bonneau & Chevillon (2012) deems skatole values in the 200 to 250 ng/g range to be excessively high. The determination of any sorting threshold is contingent upon the tolerance for adverse consumer responses that stakeholders are prepared to accept. Hence, Aluwe et al. (2018) undertook a study to illustrate the decline in preference for boar meat patties as the levels of androstenone and skatole increased. The preference for boar meat patties diminished with higher skatole content. Additionally, for consumers sensitive to androstenone, preference decreased as androstenone levels increased, particularly when skatole levels were low. Moreover, Christensen et al. (2019) stated that the anticipated risks of aversion among an average consumer, a discerning consumer, and the general consumer population aligned with sorting thresholds averaging 250 ng/g of skatole and 4000 ng/g of androstenone.

In conclusion, a low level of boar taint was found in this study aiming at investigating the prevalence of skatole and androstenone in pork entering the food chain in Portugal. There were little worth notice number of samples classified as having high levels of boar taint (4.2% exceeded the threshold value of 100 ng/g for skatole, and 0.7% surpassed the value of 500 ng/g of androstenone), however 28.8% of the samples exceeded the estimated BET values for skatole, which underscores that the practice of slaughtering immature males alone may not be sufficient to resolve this issue. Additionally, high skatole levels in females may indicate the possibility of skatole resorption, possibly stemming from inadequate hygiene management. An inherent limitation of this study lies in its observational study design, lacking control over preceding factors such as diet, production system, or genetic crossbreeding. Nonetheless, and despite the limitation of a small sample size in this study, the knowledge about the percentage of carcasses with elevated levels of androstenone and skatole offers valuable and objective insights to the sector regarding boar taint, which directly impacts the sensory quality of pork. Nevertheless, it's essential to tailor the acceptable limits of these compounds to the particular sensitivity of a given population or culture. This adaptation can be achieved through sensory tests like the best-estimated threshold.