Occurrence of Total Aatoxin and Zearalenone in Dairy Cattle Concentrate Feeds in Malawi

Background The study was conducted to determine occurrence and levels of total aatoxin and zearalenone in concentrate feedstuffs for dairy animals from selected farms in Malawi. Methods A total of 130 concentrate feedstuff samples were collected in paper bags from 113 randomly selected farms in the three agroecological zones (representing high, mid and low (lakeshore) altitudes) from October and December 2019. Each feed sample was ground and analyzed for total aatoxin and zearalenone using VICAM Fluorometer Method procedures. Results About 85% of samples comprised of corn (maize) bran (CB), 8% for pigeon pea (PP), 2% for dairy mash, 1% for soybean and <1% for sunower, rice bran (RB) and rice bran mixed with maize bran respectively. 75% of corn bran and 100% of dairy mash and rice bran samples were positive for zearalenone (i.e. above 100 μg/kg) while other concentrates tested negative. Incidences of total aatoxin were 32%, 67% and 9% in CB, DM and PP bran samples respectively and negligible in other concentrates. Overall, 32% and 23% of feedstuffs had total aatoxin concentrations above regulatory limits set by Malawi’s neighboring countries (Tanzania and Mozambique) and US respectively whereas only 6% of all the concentrate feedstuffs had zearalenone concentrations above the regulatory limit enforced by EU. Levels of total aatoxin and zearalenone in CB were not affected by agroecological zones (P=0.17 and P=0.87) respectively. Mean total aatoxin concentration was 35.4, 28.1 and 38.2 μg/kg in High, Lakeshore and Mid agroecologies respectively whereas zearalenone concentration was 249.7, 226.8 and 243.9 μg/kg respectively. Conclusion To the best of our knowledge, this is the rst study to report prevalence of various mycotoxins in dairy concentrate feedstuffs in Malawi. Aatoxin and zearalenone contamination in dairy concentrates exist but mostly in levels within tolerable limits at least from the limited timeframe of this study. However, presence of total aatoxin above regulatory limits in 26 - 34% of


Introduction
Contamination of livestock feeds with mycotoxins remains a global challenge. Several studies have reported presence of one or more mycotoxins in feeds and feedstuffs meant for livestock in all regions of world (Driehuis et al., 1992;Schmidts et al., 2014;Ji et al., 2014;Gutleb et al., 2015). Mycotoxins are a group of chemically diverse compounds originating from secondary metabolism in fungal molds (Fink-Gremmels, 1999;Njobeh et al., 2012). A atoxin and zearalenone (ZEN) that are mainly produced by Aspergillus and Fusarium species of fungi respectively are among mycotoxins of great concern in livestock production (Hussein and Brasel, 2001;Whitlow and Hagler, 2016). Among other effect, zearalenone is associated with reduced reproductive e ciencies in livestock (Weaver et al., 1986a;Weaver et al., 1986b;Kordic et al., 1992) whereas a atoxin B 1 and its metabolite a atoxin milk 1 (M1) are mutagenic and carcinogenic to both animals and humans (Wogan et al., 1974;IARC, 1993;Boudra et al., 2007).
Prevalence of mycotoxins is known to be highly enhanced by environmental factors such as warm temperatures and moist conditions that promote growth of toxigenic fungi (Lacey, 1991;Magan et al., 2003;Ribeiro et al., 2006;Murphy et al., 2006). These conditions would typically be expected to exist in small scale dairy production and in the feed storage facilities. Earlier studies have reported greater presence of various mycotoxins including a atoxins, zearalenone and fumonisins in cereal and legumes for human consumption in Malawi (Matumba et al., 2009;Matumba et al., 2011;Monyo et al., 2012;Matumba et al., 2015;Mwalwayo and Thole, 2016). The dependence of dairy animals on crop-based supplemental products such as corn (maize) bran, rice bran, soybean, peanuts, pigeon pea apart from dairy mash would likely increase the intake and exposure to mycotoxins hence increased consumer risk. However, in Malawi, the occurrence of mycotoxins and the extent of this problem is not well understood at least from the small scale dairy subsector standpoint and particularly on feeds. Absence of knowledge on mycotoxins prevalence and the extent of contamination of livestock feeds puts the health of both animals and humans at a higher risk of exposure to intolerable levels of mycotoxins with potential for suppressing immune function, decreasing the response to resist infectious diseases and reducing animal performance and e cacy of vaccines and drugs (Jiang et al., 2005;Oswald et al., 2005;Berek et al., 2001;Sharma et al., 1993). Therefore, the objective of this study was to determine incidences and levels of a atoxins and zearalenone in concentrate feedstuffs among small scale dairy farms located in three agroecological zones of Malawi.

Study area and sample size
The study was carried out in Malawi between October and December 2018. Within Malawi, the study focused on three agro-ecological zones (AEZ) namely Highlands (< 1,300 m above sea level), Midelevation (760-1,300 m above sea level), and Lakeshore, Middle and Upper Shire (200-760 meters above sea level) where dairy farming is predominantly practiced. In order to obtain a representative sample of the population, a multi-stage sampling technique was used to obtain a total of 113 dairy farms from which various concentrate feedstuff samples were collected in these three AEZs as follows; ve milk-bulking groups were randomly selected within each agroecological zone. Milk bulking groups are associations or cooperatives comprising of up to 100 members. They are established to promote collective milk marketing and ease accessibility to various extension services including dairy management trainings (IFS, 2013).
Approximately 8-11 farms per milk bulking group were randomly selected from the provided list of dairy farms. An exception was with Lakeshore AEZ where all the 20 participating dairy farms were selected from one milk-bulking group because it was the only available group at the time of the study. Therefore, 20 farms were drawn from lakeshore, middle and upper shire (Lakeshore) agroecological zone while 51 and 42 farms were included in the study from Highlands and Mid elevation agroecological zones respectively making a total of 113 dairy farms. From these dairy farms, a total of 130 concentrate feedstuff samples were collected in paper bags. For each feedstuff, multiple samples were collected and then mixed to make one sample weighing 0.5-1 kg for preparation and analysis as described in the following sections.

Mycotoxin extraction and analysis
Page 4/15 Every feed sample was ground to pass through a 1 mm screen to provide ne and homogeneous samples for mycotoxin analysis. The ground samples were stored in plastic containers in a cool, dry place until analyzed. Total a atoxin and zearalenone were extracted and determined using VICAM Fluorometer Method (VICAM, 2014), according to the analytical procedures detailed below.

a) A atoxin extraction and analysis
A sub sample of ground feed (50 g) was mixed with 10 g NaCl and place in blender jar. Then, 200 mL methanol and water at the ratio of 80:20 was added into a blender jar. The mixture was blended at for 1 minute and thereafter the extract was ltered through uted lter paper. The ltrate was collected in a clean vessel from which 10 mL was drawn using a pipette and placed into another clean vessel. The extract was diluted with 20 mL puri ed water and then ltered through 1.5 µm glass micro bre lter into a clean vessel. A total of 2 mL of ltered extract was added (1 mL at a time) to the A aTest column headspace and passed through the column at a rate of about 1 drop/second until air came through the column. The column was washed with 1 mL of puri ed/distilled water at a rate of 1-2 drops/second. And this step was repeated until air comes through column. High Performance Liquid Chromatography (HPLC) grade methanol (1 mL) was passed through the column at a rate of 1 drop/second in order to elute the toxin and the sample elute was collected in a glass cuvette (VICAM part # 34000). A aTest Developer solution (1 mL) was added to the sample elutes in the cuvette and mixed thoroughly. The cuvette was placed in a calibrated VICAM Series 4EX uorometer and a atoxin concentration was determined after 60 seconds. The detection range of the assay was 0-50 µg/kg (VICAM, 2014). The sample elutes containing concentration of a atoxin greater than 50 µg/kg were diluted ten folds with (1:1) volume ratio of HPLC grade methanol and A aTest Developer solution and retested.

b) Zearalenone extraction and analysis
A sub sample weighing 20 g was mixed with 2 g of salt and placed in a blender jar. Then, 50 mL of methanol:water (80:20) was added to the mixture and blended at high speed for 2 minutes. The mixture was ltered through uted lter paper and collected in a clean vessel. The extract was diluted with 40 mL 1 × 0.1% Tween PBS Buffer and ltered through 1 µm glass micro ber lter. Diluted and ltered extract (10 mL) was passed through ZearalaTest column at a speed of 1-2 drops per second. The column was washed by letting 10 mL 1 × 0.1% Tween PBS Buffer pass through at the speed of 1-2 drops per second. Thereafter, the column was washed by 10 mL of distilled water at a speed of 1-2 drops per second. ZEN was eluted by passing 1.0 mL HPLC grade methanol through the column at a speed of 1-2 drops per second and the elutes was collected in a glass cuvette. ZearalaTest Developer (1.0 mL) was added to elute in a cuvette and mixed thoroughly. The cuvette was placed in a calibrated VICAM Series 4EX uorometer and readings were taken after 5 minutes. The detection range of the assay was 0-2000 µg/kg (VICAM, 2014). Sample elutes containing concentration of zearalenone greater than 2000 µg/kg were diluted ten folds with (1:1) volume ratio of HPLC grade methanol and ZearalaTest Developer solution and retested.

Statistics
The percentage incidences of total a atoxin and zearalenone, minimum, maximum and median values for each type of concentrate feedstuff were calculated using Microsoft Excel 2010. Comparison of total a atoxin and zearalenone levels across agroecological zones was only performed on corn bran, as it was the widely available concentrate feedstuff in dairy farms across all three agroecological zones. The total atoxin and zearalenone data in corn bran samples were not normally distributed and were logtransformed for statistical analysis (log A atoxin + 1, log Zearalenone + 1) and analyzed using SAS version 9.2 (SAS Institute Inc, Cary, NC). Means were separated using Tukey's HSD. The reported means of a atoxin and zearalenone data in corn bran samples across agroecological zones are back transformed concentrations. The level of con dence required for signi cance was set at p ≤ 0.05.

Results
Incidence and concentration of a atoxin and zearalenone in feedstuff samples Incidence and concentration of total a atoxin and zearalenone in concentrate feedstuffs are given in table 1. The majority of collected concentrate feedstuff samples was corn bran. From a total of 130 concentrate feedstuff samples collected, 111 (85%) were for corn bran, 11 (8%) for Pigeon pea bran, 3 (2%) for dairy mash, 2 (1%) for soybean and <1% each for sun ower, rice and mixture of corn bran and rice bran.
Zearalenone levels above the limit of detection (100 μg/kg) was present in 75% of corn bran feedstuff samples, 9% of pigeon peas bran samples, 100% of dairy mash and rice bran samples but not detected in soybean, sun ower and a mixture of corn bran and rice bran. The concentration of zearalenone in positive samples ranged from 100 -2400 μg/kg with a median of 240 μg/kg in corn bran, 130 -410 μg/kg with the median of 160 μg/kg in dairy mash and at a concentration of 280 μg/kg in one and only positive pigeon pea bran feedstuff sample. Total a atoxin was detected above limit of detection of 10 μg/kg in 36%, 67% and 9% of corn bran, dairy mash and pigeon pea bran feedstuffs samples respectively and not detected in soybean, sun ower, rice bran and a mixture of corn bran and rice bran. Levels of total a atoxin in positive feedstuff samples of corn bran ranged from 10 -410 μg/kg with a median of 37 μg/kg and 11 -110 μg/kg with a median of 60.5 μg/kg in dairy mash. Total a atoxin was only detected in 1 out of 9 pigeon pea bran samples at concentration of 16 μg/kg.  Feedstuff samples with a atoxin and zearalenone above regulatory limits Number and proportions of concentrate feedstuff samples with either total a atoxin or zearalenone concentration above various regulatory limits are presented in table 2. Overall, 32% of concentrate feedstuff samples had total a atoxin concentrations above regulatory limit of 10 μg/kg set by Malawi's neighboring countries Tanzania and Mozambique, 23% did not comply with total a atoxin regulatory limit of 20 μg/kg in dairy feeds set by US and only 6% of all the concentrate feedstuff samples had zearalenone concentrations above 500 μg/kg, the regulatory limit enforced by EU. Thirty four percent (34%) and 26% of corn bran samples had total a atoxin concentration levels above regulatory limits set by Tanzania plus Mozambique and US respectively whereas 7% had concentrations levels of zearalenone above EU regulatory limit. Sixty seven percent (67%) of dairy mash samples (n=3) had total a atoxin concentration levels above regulatory limits of 10 μg/kg enforced by Tanzania and Mozambique and only one sample had levels of total a atoxin above 20 μg/kg limit followed by US.

TAF = Total aflatoxin
Total a atoxin and zearalenone concentration in corn bran across agroecologies The mean concentration of total a atoxin and zearalenone in positive corn bran feedstuff samples across the High, Mid and Lakeshore agroecological zones are presented in table 3. Levels of total a atoxin and zearalenone were not affected by agroecological zone (p=0.17) and p=0.87) respectively. The mean total a atoxin was 35.4, 28.1 and 38.2 μg/kg in High, Lakeshore and Mid agroecological zones whereas mean concentrations of zearalenone were 249.7, 226.8 and 243.9 μg/kg respectively.

Discussion
Occurrence of mycotoxins was high in some dairy concentrate feedstuffs and low to non-detectable in others. To the best of our knowledge, this is the rst study to report prevalence of various mycotoxins in dairy concentrate feedstuffs in Malawi. It must be noted from the onset however that corn bran (CB) was apparently the major form of concentrate commonly found in these small scale dairy farms representing 85% of all feedstuffs sampled followed by pigeon pea bran at 8% and 2% for dairy mash while the rest were at < 1-1%. With this in mind, 75% of CB feed samples, and 100% of dairy mash (DM) samples were positive for zearalenone, and only 9% for pigeon peas bran (PPB) whereas zearalenone was non detectable in soybean, sun ower and a mixture of corn bran and rice bran. Similarly, total a atoxin was detected in 36%, 67% and 9% in CB, DM and PPB feedstuffs respectively and non detectable in the other feedstuffs. These results are consistent with those from others studies that have reported high incidence of a atoxin and zearalenone in various feed samples (Kang'ethe and Lang'a, 2009;Schmidt et al., 2014;Kosicki et al., 2016;Gutleb et al., 2015;Schatzmayr and Streit, 2013;Anukul et al., 2013). The high incidence of zearalenone and a atoxins contamination in corn bran and dairy mash feedstuff samples observed in this study could be attributed to availability of favorable conditions for growth of Aspergillus and Fusarium fungi. The growth of toxigenic fungi and the subsequent production of mycotoxins are reported to be enhanced by environmental factors like temperature and moisture (Lacey, 1991;Magan et al., 2003;Ribeiro et al., 2006;Murphy et al., 2006). Optimal production of a atoxin is known to occur at temperatures between 25 and 33 o C (Ribeiro et al., 2006;Murphy et al., 2006;OBrian et al., 2007) while zearalenone is optimally produced at temperatures ranging from 25 to 30 o C (Murphy et al., 2006). Located within the tropics, these conditions are prevalent in Malawi hence growth of Fusarium and Aspergillus fungi and subsequent high incidences of a atoxins and zearalenone contamination. The low incidence of total a atoxin and or zearalenone in pigeon pea bran, and non-detection in soybean meal and sun ower is a notable result in this study. While the factor(s) responsible for this low incidence in these feedstuffs is not known and was not the core scope of this study, some mycotoxins are reported to be more associated with certain agricultural commodities than others. For example, soybeans are generally considered resistant to Aspergilli colonization and a atoxins contamination (Njobeh et al., 2009;Gupta & Venkitasubramanian, 1975) whereas zearalenone is commonly detected in corn, wheat and corn-based feeds (Driehuis et al., 1992;Schmidts et al., 2014;Lee and Ryu, 2017). Nevertheless, considering that a atoxins and zearalenone exert synergistic effects on livestock including cattle (Jovaišienė et al., 2016;Huang et al., 2018), the high incidences of these mycotoxins in major concentrate feedstuffs for dairy animals like CB is cause for concern among small scale dairy farmers in Malawi.
Considering a wide array of adverse health effects associated with exposure to mycotoxins, several countries have established regulatory limits. Among southern Africa countries, Tanzania and Mozambique have set regulatory limits at 10 µg/kg for a atoxins in animal feeds (FAO, 2000). The European Union (EU) has no regulatory limits for total a atoxin but zearalenone in cattle feeds while United States of America (US) has established limits for total a atoxin and none for zearalenone in animal or speci cally dairy animal feeds (FAO, 2000;FDA, 2000). However, Malawi is yet to set regulatory limits for mycotoxins in animal feeds. Regulations are set to prevent exposure of either livestock or humans to levels of mycotoxins that can cause adverse health effects. This study determined prevalence of feedstuff samples with total a atoxin or zearalenone levels above several regulatory limits. Findings of this study indicate small to moderate proportion of concentrate feedstuffs with zearalenone and total a atoxin above various regulatory limits. Overall, 32% of all concentrate feedstuff samples had total a atoxin concentrations above regulatory limits of 10 µg/kg set by Malawi's neighboring countries of Tanzania and Mozambique while 23% was above US set limit of 20 µg/kg for dairy feeds. Contrary, only 6% of all the concentrate feedstuffs had zearalenone concentrations above 500 µg/kg regulatory limit enforced by EU. Although Malawi has not yet established these regulatory limits for mycotoxins and speci cally a atoxin or zearalenone in animal feeds, results from this study and others done in food crops from which concentrate feeds are derived (Matumba et al., 2009;Matumba et al., 2011;Monyo et al., 2012;Matumba et al., 2015;Mwalwayo and Thole, 2016) do suggest the need for setting such policy guidelines. The nding of 26-34% of CB, 10% of PPB and 67% of DM samples with total a atoxin concentrations above various regulatory limits (FAO, 2000) raises an alarm regarding the adverse impacts these may have on small scale dairy production and safety of dairy products for human consumption in Malawi. Since corn bran is apparently the most predominant concentrate supplement among small scale farms in Malawi, their daily consumption by animals calls for urgent follow-up studies to determine residual traces in milk and dairy products commonly found on the market and associated risk to consumers.
Concentrations of total a atoxin and zearalenone were not affected by agroecological zones from where these feedstuffs samples were taken. The mean total a atoxin were 35.4, 28.1 and 38.2 µg/kg in High, Lakeshore and Mid agroecological zones whereas mean concentrations of zearalenone were 249.7, 226.8 and 243.9 µg/kg respectively. These results contrast those of other studies that have reported variations in a atoxin levels in feeds and foods across micro and macro environmental conditions (Kaaya et al., 2006;Matumba et al., 2015;Sirma et al., 2016). For example, a atoxin contamination levels in corn samples signi cantly decreased from moist zone (9.73 µg/kg) to dry zone (7.72 µg/kg) and lowest in highland zone (3.92 µg/kg) in Uganda (Kaaya et al., 2006). Similarly, Sirma et al. (2016) reported signi cantly higher mean levels of a atoxin in corn and millet from the humid and sub-humid agroecological zones than those from the temperate agroecological zone. In Malawi, Matumba et al. (2015) reported highest a atoxin level and prevalence in corn samples from hottest agroecological zones while zearalenone was most prevalent in corn samples collected from cool agroecological zones. Differences in micro or macro environmental conditions may favor or prohibit growth of certain fungi compared to others and this may account for the agroecological zone variations in prevalence between a atoxin and zearalenone. The lack of agroecological zone effect on levels of total a atoxins and zearalenone in corn bran feedstuffs observed in this study could be attributed to inter agroecological zone trading and movement of concentrate feedstuffs by dairy farmers. For example, dairy farmers in High agroecological zone usually buy corn bran from Lakeshore and Mid agroecological zones which are major corn producing areas in Malawi. However, this assumption has not been extensively veri ed. Probably a multi seasonal evaluation of total a atoxin and zearalenone in concentrate feeds may establish an insightful presentation of the trend across agroecological zones. Regardless of the reasons for the similarities in mycotoxins concentrations, the high occurrences of a atoxins and zearalenone in feedstuffs across agroecological zones observed in this study may compromise dairy production and also puts in question the safety of milk and dairy products to consumers. Further, this suggests that dairy animals are unduly exposed to higher levels of mycotoxin contamination that in part may affect dairy production hence need for concerted efforts to mitigate the resultant effects.

Conclusion
There were high occurrences of a atoxin and zearalenone contamination in dairy concentrate feedstuffs commonly used in Malawi. Cereal-based concentrates (Corn bran and dairy mash) had the highest incidence of zearalenone and total a atoxin levels among all the feedstuffs sampled compared with legumes-based concentrates (PPB, soybean and sun ower) which had low mycotoxin incidences.
Although most levels were within limits, a small to moderate proportion of samples had total a atoxin and zearalenone above various regulatory limits hence cause for concern for the safety of animals and dairy products for human consumption which also calls for urgent need for mitigation measures including setting policy guidelines. Furthermore, concentrations of total a atoxin and zearalenone were not affected by agroecological zones which either suggests the similarities in environment conditions during the study time period or trading and movement of feedstuff across agroecological zones as possible contributory factors. Overall, these results bring to light the status of dairy animal concentrate feedstuffs regarding total a atoxin and zearalenone contamination and raises possibilities of exposure of animal and consumers to greater and intolerable levels of mycotoxins with higher risk potential for adverse health effects.

Declarations
Availability of data The dataset used and or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.
Authors' contributions CAN conceived the study and drafted the manuscript. JCH and MB participated in statistical analysis and helped drafting the manuscript. All authors read and approved the nal manuscript.

Funding
This study was curried out with funds from United States Agency for International Development (USAID) through Borlang Higher Education for Agricultural Research and Development (BHEARD).