Assessment of aflatoxin in milk and feed samples and impact of seasonal variations in the Punjab, Pakistan

Abstract The present study was designed to assess the incidence of aflatoxin contamination in animal feed and raw milk samples (total 240 each) collected from dairy farms during the complete year of 2015. These samples were collected through a cluster random sampling technique by dividing the province of Punjab, Pakistan into five clusters (north, south, east, west and central). Factors (environmental & physical) affecting aflatoxin contamination in milk and animal feed at farms were also studied. The AFM1 levels in raw milk & AFB1 levels in feed samples were analyzed by using the ELISA technique. Results demonstrated that overall about 53% raw milk samples from dairy farms were contaminated beyond the US MRL (0.50 µg/L) for AFM1 with than average level of 0.59 µg/L, while the 95% farm feed samples were exceeding the FDA MRL (20 µg/kg) of AFB1 with average level of 43 µg/kg. During winter season, the concentration of AFM1 was higher in all clusters with avg 0.68 µg/L, while the AFB1 contamination was highest in the spring season with avg 54 µg/kg. Market feed prices had negative correlation with AFB1 contamilevels, which were further supported by the positive correlation between quantity of feed at farms with AFM1 and AFB1 contamination. Results exhibited significantly positive impact of environmental factors on milk and feed aflatoxin contamination levels, whereas temperature showed an inverse relationship with AFM1 and AFB1 levels. The study recommends need of synergistic extension work to support dairy farms and highlight the contamination levels for regulatory bodies to introduce strategic policies for control measures.


| INTRODUC TI ON
Mycotoxins are the group of naturally prevailing secondary metabolites, mainly produced by filamentous fungi (Varga, Frisvad, & Samson, 2011). This is the global problem and a quarter of the world's food and food products are affected (Lawlor & Lynch, 2005). There are many types of mycotoxins among which AFB1 is the most abundant and toxic due to teratogenic and mutagenic effects (Kang'ethe & Lang'a, 2009). The chief source of aflatoxins are Aspergillus flavus, A. parasiticus and A. nomius (Mostrom & Jacobsen, 2011).
AFM1 is documented as a metabolite of AFB1 and is concealed in the milk of those animals' which are fed on contaminated feed (Ruangwises & Ruangwises, 2010). It starts appearing in milk after 12-24 hr of contaminated feed ingestion (Rahimi et al., 2012). It must be remembered that conversion factor of AFB1 from animal feed to AFM1 in raw milk is 0.30%-6.2%, depending upon the genetics, lactation stage, milk production and heathe animalsdition of animals (Unusan, 2006). Milk and its products are imperative for sustained human health (Bașkaya, Aydın, Yıldız, & Bostan, 2006). Aflatoxin toxicity thought to be the one of the major causes of liver cancer (Omata et al., 2010). Hepatocellular carcinoma (HCC) risk reported to increase 30 folds in the presence of aflatoxin and hepatitis B virus.
Initially the International Agency for Research on Cancer (IARC) based on the toxicity, classified AFM1 as agent in Group 2B possessing potentially carcinogenic influence on human health while further it was reclassified to Group 1, along with AFB1 carcinogenic agent (IARC, 2002). Furthermore, thermal processing like pasteurization and even ultra-high temperature (UHT) treatments are unproductive in abolishing or reducing harm of contamination because of its stability at high temperatures (Prandini et al., 2009).
Livestock sector is inevitable for the economic development of any country. Pakistan is blessed with variety of animals and graced with third position among world largest milk producing nations with annual production of 54 million tons (Iqbal, Iqbal, Akbar, Khan, & Abbas, 2015). It contributes about 46.8% to agriculture with 10%-25% income generated by rural people through livestock (Iqbal, Ahmad, & Jehangir, 1999). Livestock plays vital role in alleviating poverty of rural areas by providing food and income (Mahmood, Khalid, & Kouser, 2009). Animals fed on these contaminated sources were observed with a decrease in growth rate, milk production, milk quality attributes and ultimately with compromised desired immunity against infections (Akande, Abubakar, Adegbola, & Bogoro, 2006). Aflatoxicosis in cattle leads to lethargy, ataxia, rough hair coat, enlarged pale fatty liver, less feed intake by loss of appetite, diarrhea, blindness, teeth grinding, frothing at the mouth, abortion, lameness, ovarian cyst and other reproductive disorders (Nibbelink, 1986;Pirestani, Tabatabaei, Fazeli, Antikchi, & Baabaei, 2011).
In order to minimize the risk of contamination in feed and milk, regulatory restrictions have been imposed across the world. In China, the limit of AFB1 in feed is 10 µg/kg while limits for milk is 0.5 µg/L (Wang & Liu, 2008). Most of the other countries including US Food and Drug Administration (FDA) established 0.5 µg/L MRL (maximum residual limit) for milk. European Commission Directive of 2004 described that 0.05 and 0.025 µg/L limits for liquid and dried milk, respectively (Commission, Programme, & Organization, 2007).
Pakistan's province of Punjab has also set the levels of AFM1 in raw milk for processing as 5 and 0.5 µg/L milk for consumption under the legal frame work of Punjab Food Authority Regulations 2018 (PFA, 2018). Pakistan has also set the legal limits of Animal feedstuffs in the Punjab Feed Stuff and Compound Feed Act, 2016. According to that act the compound feed or concentrate manufactured for lactating dairy cow should not contain AFB1 above 50 µg/kg, while for other commonly used feed ingredients, AFB1 levels are different in Act as presented in Table 1.
Pakistan has been facing huge economic losses so far due to the lack of aflatoxin action plan and its implementation. Investigation report of PCSIR, Karachi shared a report during 2017, when approximately five-hundred dairy animal died and twelve hundred fell sick due to the feeding of highly contaminated mycotoxin feed (Sultana & Hanif, 2009). Similarly, few years ago many of Pakistani export consignments were rejected and banned owing to high aflatoxin contamination like custard powder from South Korea, peanut from UK and chili powder from Europe.
Keeping in view the above details, the present study are designed (a) to highlight the presence of AFM1 & AFB1 contamination in raw milk and animal feed samples from different farms of province of Punjab, Pakistan (b) to investigate the environmental impact (temperature, season, humidity) on AFM1 & AFB1 levels in milk and feed samples (c) to compare the samples which exceeds the permissible level of location and international regulations and (d) to construct a database that could help policy maker in local governments, law enforcement agencies, dairy farmers and consumers to focus immediate attention to prevent or minimize the health risks associated from these toxins. publication (Akbar et al., 2019). Farmer having milk yield more than forty liters per day was selected from tehsils (an administrative unit of district) by using simple random sampling. In Punjab province corn, concentrate feed mix, bread pieces, cotton seed cake and rice polish are the most common animal feed sources. So, keeping in view the usage of feed in a sampling month, high consuming feed of that month is selected for AFB1 analysis.

| Sample preparation & method validation for AFM1 analysis
The competitive enzyme immunoassay using Elisa kit Ridascreen ® Fast Aflatoxin M1, R5812 (R-Biopharm AG) was used for quantitative analysis of AFM1 in the milk samples. Sample preparation was done by following the manual instructions, while method validation was performed as details are published in our previous study (Akbar et al., 2019). Abbreviations: SD, standard deviation, r, repeatability, SE, Standard Error, CV, coefficient of variation). a homogenizer (e.g., Ultra-Turrax T18). Sample allowed to settled for 2 min to 3 min after extracting to enable some of the sample to settled before filtering the extract. Filtered the extract by pouring at least 5 ml through Whatman filter paper no 1 and stored the filtrate into a 125-ml amber glass Erlenmeyer.

| Procedure and validation for AFB1 analysis by ELISA
Manufacturer's recommendations were followed throughout the analysis. Blue-labelled bottles having conjugate were pipetted 100 μl in each red-marked mixing well. Standards (0.00, 5.00, 15.00 and 50.00 μg/kg) were added 100 μl by using a new pipette tip for each one was transferred into the wells. Liquids were mixed by using a of standards and samples at 650 nm using a microplate reader.
A standard curve was constructed each time using series of standards of AFB1 solutions provided with the test kit (

| Statistical analysis
All data were presented as mean by using simple excel. Correlation was applied to calculate significant impact of temperature, precipi-  (126) were considered statistically significant at p < .05. Correlation and regression analysis were applied to calculate R 2 by using statistical package Q Stat.net.

| RE SULTS AND D ISCUSS I ON
In the present study it was observed that almost 53% of the milk samples collected from the dairy farms situated in the five clusters of Punjab province of Pakistan, were contaminated higher than safe limit of US < 0.5 µg/L (Figures 2 and 3). AFM1 contamination was observed highest in Eastern cluster (0.59 ± 0.03 µg/L) followed by Northern (0.51 ± 0.30 µg/L), Western (0.51 ± 0.30 µg/L) and Central (0.46 ± 0.30 µg/L) cluster, while Southern cluster (0.46 ± 0.30 µg/L) remained lower throughout the year as compare to the Eastern cluster (Table 4). Though the number of samples obtained from different clusters were same but availability of AFM1 was speckled.
Average AFM1 contamination levels and number of samples exceeding US permissible limits were decreasing north to south as topography, temperature, rainfall, humidity and weather condition varies  Limits (Rastogi et al., 2004).
Moreover, results showed that winter season is more critical entailing higher contamination of AFM1 and maximum number of samples exceeding permissible limits (Figures 2 and 3). Common months unveiling higher contamination in entire province regardless of division and clusters were December to March. Hence, severity of AFM1 in winter season collateral to Eastern cluster (Table 4). In current study, it is observed that farmer's offer contaminated stored and concentrate feed sources during winter due to the shortage of green fodder. Farmers in greed of more milk production during winter season increase the quantities of these available stored contaminated feed ingredients (i.e., corn, based, which ultimately increase the levels of AFM1 contamination (Asi et al., 2012). Therefore, quantities of feed are positively correlated to AFM1 (Table 6) in winter (p < .01). Feed consumed by animals is indeed the prime factor in this regard. Our study is supported by many previous studies of AFM1 contamination levels in winter season (Asi et al., 2012;Fallah, Rahnama, & Saei-Dehkordi, 2011;Nemati, Mesgari Abbasi, Parsa Khankandi, & Masoud, 2010). Contrarily limited studies (Bahrami, Shahbazi, & Nikousefat, 2015;Fallah et al., 2011) reported no significant difference of season on AFM1 contamination levels.
It was extracted from current research that AFB1 concentration emerged in entire samples collected from five clusters, though concentration varied across the year. Monthly trend line of AFM1 and AFB1 levels showed that contamination of milk and feed are inter-linked across the year (Figure 4). This concentration of AFB1 is solemnly associated with the feed consumed by the animals. Bread pieces, concentrate feed mix, maize and cotton seed cake were the often-consumed feeds by the animals in all clusters. During informal discussion, it is observed that traditionally farmers assumed that use of cited feeds is helpful in elevating milk production and prefer to feed these one whenever available at cheaper rates.
Eastern cluster exhibited 92% samples followed by 85%, 81% and 81% ( Figure 2)   Maize is preferred feed for farmers at dairy farms. Maize samples used for animals feed also appeared source of aflatoxin (Ahsan, Bhatti, Asi, Bhatti, & Sheikh, 2010). Maize crop through fungi could develop aflatoxin concentration of varied level (Sanchis & Magan, 2004). Several researches (Anjum, Khan, Sahota, & Sardar, 2012;Bhatti, Talat, & Sardar, 2001) found highest contamination of aflatoxin in corn, which is common feedstuff for animals. Reddy and Salleh (2011) reported 23% AFB1 contaminated samples ranging from 21 to 135 µg/kg. Similar findings were reported 61% maize samples AFB1 contamination above the permissible limits (Anjum et al., 2012). Farmers also offer bread pieces to animals, which are waste material and high source of aflatoxin contamination. Asi et al. (2012) reported that those animals fed on bread pieces and concentrates, evolve higher amount of aflatoxin in milk. Chauhan, Washe, and Minota (2016) also revealed that dairy concentrates were highly contaminated (64%) feedstuff for animals.
A significant association (p < .05) between precipitation and emergence of aflatoxin in milk and feed during winter, spring and summer season was observed (Table 6). However, during autumn season precipitation did not exhibited any influence on aflatoxin in milk and feed. In overall context, association appeared significant.
Temperature indicated significant but negative association with aflatoxin in milk during autumn season. Lower temperature promotes fungal growth and AFB1, hence, findings do not corroborate with those of Pratiwi et al. (2015). Humidity, another determinant of environment was significantly associated (p < .05) with aflatoxin generation in feed during winter, spring and summer season. Findings regarding humidity are in alliance with of Pratiwi et al. (2015) where they unveiled AFB1 incidences pertinent to humidity. In Pakistan feed contamination variation were attributed to persistent relative humidity and rainy season (especially hot monsoon) which usually persist between June to September (Anjum et al., 2012;Rashid et al., 2012;Yunus, Nasir, Aziz, & Böhm, 2009). The season has a significant role on fungal growth. In the current study data (Figure 3) depicts that rainy season (Jun to Sep) effected the crops especially corn & cotton. Later on animals feed on these sources were significantly raised the levels of AFB1 & AFM1 in upcoming months of winter ( Figure 3). The AFB1 contamination levels are influenced by environmental conditions like humidity and temperature (Hanif, 2009;Hanif et al., 2008 (2013) concluded that consumption of cotton seed cake increased milk production and fat concentration in milk. Cotton seed cake is pleasant source of protein and encouraging milk production. Cotton seed cake served as appetizer for animals (Yasmeen et al., 2007). Positive impact of cotton seed cake on weight of lactating animals has been proven by number of research studies as well (Chowdhury, 2001;Jabbar, Anjum, Rehman, & Shahzad, 2006;Jabbar & Marghazani, 2009). Despite of extensive significance in health improvement and milk production, cotton seed cake produce AFM1 in milk of animals. Ullah et al. (2016) identified highest concentration of AFB1 in lactating animals fed on cotton seed cake. Cotton seed cake happened to be a source of fungi because of excessive proteins and lipids inside. Chauhan et al. (2016) found highest contamination (68%) of aflatoxin in cotton seed cake. Saleemi, Khan, Khan, and Javed (2010) reported that cotton seed cake is most susceptible to fungal attack, which is prime source of aflatoxin.
In addition to above details, the average prices of feed sources had negative correlation with AFB1 contamination levels ( Figure 5) which were further supported by the positive correlation between quantities of feed at farms. This indicated that in case of high prices, farmers opt an alternative cheaper source of feed regardless of aflatoxin contamination.

ACK N OWLED G EM ENTS
The authors are thankful to IHFS, Government College University Faisalabad for providing technical support and laboratory facilities during research work.

CO N FLI C T O F I NTE R E S T
Authors declare that they have no conflict of interest.

E TH I C A L A PPROVA L
This article does not contain any studies with human participants or animals performed by any of the authors.

I N FO R M E D CO N S E NT
For this type of study, formal consent is not required.