A Survey of Aflatoxin-Producing Aspergillus sp. from Peanut Field Soils in Four Agroecological Zones of China

Peanut pods are easily infected by aflatoxin-producing Aspergillus sp.ecies from field soil. To assess the aflatoxin-producing Aspergillus sp. in different peanut field soils, 344 aflatoxin-producing Aspergillus strains were isolated from 600 soil samples of four agroecological zones in China (the Southeast coastal zone (SEC), the Yangtze River zone (YZR), the Yellow River zone (YR) and the Northeast zone (NE)). Nearly 94.2% (324/344) of strains were A. flavus and 5.8% (20/344) of strains were A. parasiticus. YZR had the highest population density of Aspergillus sp. and positive rate of aflatoxin production in isolated strains (1039.3 cfu·g−1, 80.7%), the second was SEC (191.5 cfu·g−1, 48.7%), the third was YR (26.5 cfu·g−1, 22.7%), and the last was NE (2.4 cfu·g−1, 6.6%). The highest risk of AFB1 contamination on peanut was in YZR which had the largest number of AFB1 producing isolates in 1g soil, followed by SEC and YR, and the lowest was NE. The potential risk of AFB1 contamination in peanuts can increase with increasing population density and a positive rate of aflatoxin-producing Aspergillus sp. in field soils, suggesting that reducing aflatoxigenic Aspergillus sp. in field soils could prevent AFB1 contamination in peanuts.


Introduction
Peanuts (Arachishypogaea L) are an important economic crop in China, with its annual production being the highest on a global level at 16 million tons in 2015. China also accounts for more than 40% of world peanut production [1,2]. The four agroecological zones, namely the Southeast coastal zone (SEC), Yangtze River zone (YZR), the Yellow River zone (YR) and the Northeast zone (NE), are the major peanut producing regions in China [3].
Peanuts are often infected during pre-harvest by Aspergillus sp. which produces aflatoxins that are carcinogenic to humans and animals [4,5]. Soil is the main source of inoculum for aflatoxigenic Aspergillus species, and since peanut pods grow underground, they are in direct contact with the soil fungal population [5,6].
Chemically, aflatoxins belong to the bisfuranocoumarin group, with aflatoxins B 1 (AFB 1 ), B 2 , G 1 , G 2 being the most common contaminants. Aflatoxins have the potential to cause outbreaks of acute hepatitis and even liver cancer in animals and humans [7,8]. Of the naturally occurring aflatoxins, AFB1 is known for its toxic and carcinogenic nature [9,10]. In China, AFB 1 predominantly contaminates peanuts, with an average rate of 86.2% of the total aflatoxins and a correlation coefficient of 0.99 [11]. A significant difference in the population density and positive rate of aflatoxin-producing Aspergillus sp. across the agroecological zones was observed (Table 2). YZR zone (1039.3 cfu·g −1 , 80.7%) was the highest, followed by SEC (191.5 cfu·g −1 , 48.7%) and YR (26.54 cfu·g −1 , 22.7%). But the population density and positive rate of aflatoxigenic Aspergillus sp. in the NE zone (2.4 cfu·g −1 , 6.6%) was significantly less than the other three reported zones in this study. The population densities of aflatoxigenic Aspergillus sp. varied among the districts from all the zones, ranging from 1.1 to 2749.3 cfu·g −1 . In SEC zone, the Yunfu district recorded the highest population densities of aflatoxin-producing Aspergillus with high positive rate (481.1 cfu·g −1 , 66.7%), but was significantly less in Shantou district (19.9 cfu·g −1 , the positive rate 36.7%). When compared to SEC zone, population densities and positive rate of aflatoxigenic Aspergillus sp. in YZR zone was significantly higher, with Xiangzhou district (2749.3 cfu·g −1 , 90%) being the highest and the lowest was Xiaogan district (55.5 cfu·g −1 , 33.3%). But in YR zone, they were significantly lower when compared to SEC and YZR zone, with the Qingdao district (52.2 cfu·g −1 , 30.0%) being the highest and Linyi (5.7 cfu·g −1 , 16.7%) as the lowest. NE zone recorded significantly lower population densities of aflatoxin-producing Aspergillus and positive rate among all the zones studied, the largest district was Tieling (4.4 cfu·g −1 , 13.2%), and the least districts were Dalian (1.1 cfu·g −1 , 3.31%) and Shenyang (1.1 cfu·g −1 , 3.31%). The population density had a significant positive correlation with the positive rate (r = 0.65). Also, the positive rate had a significant positive correlation with temperature (r = 0.61), but was negatively correlated with longitude (r = −0.71), and pH had a great influence on the positive rate (r = 0.48) ( Table 3).  -8473b  Zhanjiang  0-25,812  5536d  238-20,172  10,324b  5788d  Yunfu  0-82,083  6493c  70-1380  725c  6173cd  Shantou  0-16,501  4178f  --4178e  shaoguan  0-26,293  3092h  --3092f  Total  5805A  -5465C  5795A   YZR   Huanggang  0-17,512  1220j  --1220gh  Xinzhou  0-1089  73m  --73i  Xiaogan  0-28,970  7611b  --7611bc  Yicheng  0 Table 3. Pearson's correlation coefficients of relationships among the quantity of AFB 1 producing Aspergillus in soil (cfu·g −1 ), positive rate %, the proportion of Aspergillus isolates with AFB 1 producing ability >1000 ng·mL −1 , 100-1000 ng·mL −1 , 0-100 ng·mL −1 , the proportion of L-stain, S-strain and NS-strain, the average AFB 1 quantification, the potential risk of AFB 1 contamination (ng·mL −1 /g soil), average temperature (TEM), the longitude (LONG), latitude (LAT), pH and soil moisture (%). The isolates were further grouped according to sclerotium production and size: L-strain (average diameter >400 µm), S-strain (average diameter <400 µm) and NS (non-sclerotial)-strain ( Figure 1). The NS-strains were recorded in highest percentage, representing nearly half of all the isolated strains from all the soil samples. In contrast, the recoveries of L-and S-strains were only 31.7 and 26.7%, respectively. Statistical analysis showed a significant difference in percentage of L, S and NS-strains (p < 0.05). Figure 2 shows the distribution of different aflatoxin-producing Aspergillus morphotype among the sampling districts. Frequencies in the percentage of L, S and NS-strains significantly varied within the districts in each zones. L-strains were not observed in Xiaogan district from YZR zone, Yantai district from YR zone, and NE zone, S -strains were not observed in Yantai district from YR zone, but NS strains were observed in all the zones. NS strains were found to be significantly lower in Guangzhou district from SEC zone than in other districts. S strains were found to be significantly lower in Xiaogan district from YZR zone than in other districts. However, not much of a significant difference was found among the L strains between the zones. The average production of AFB 1 by different Aspergillus phenotypes (Figure 1) was significantly different, with the highest being S-strains (5047.5 ng·mL −1 ), followed by L-strains (2963.4 ng·mL −1 ), and NS-strains (1304 ng·mL −1 ).

Distribution of Aflatoxin-Producing Aspergillus Chemotypes
Frequencies of AFB 1 production varied among the sampling districts ( Figure 3) and agroecological zones ( Figure 4). The Yantai district from YR zone had a significantly larger proportion (93.3%) of aflatoxigenic strains (>1000 ng·mL −1 ) with higher percentage of A. parasiticus (Table 1), compared to all the districts in all zones. The second largest population of highly aflatoxigenic isolates was seen in Yunfu district (69.4%) in SEC zone. Xinzhou and Xiangzhou districts in YZR zone recorded the lowest aflatoxigenic populations with 4.3% and 3.1%, respectively. Among the four agroecological zones, the highest proportion of isolates producing >1000 ng·mL −1 level of aflatoxin was observed in SEC zone (55.4%) which was 5.4 times higher than other zones. The YZR zone had the lowest percentage of isolates (10.2%) producing >1000 ng·mL −1 of aflatoxin.   Frequencies of AFB1 production varied among the sampling districts ( Figure 3) and agroecological zones ( Figure 4). The Yantai district from YR zone had a significantly larger proportion (93.3%) of aflatoxigenic strains (>1000 ng•mL −1 ) with higher percentage of A. parasiticus (Table 1), compared to all the districts in all zones. The second largest population of highly aflatoxigenic isolates was seen in Yunfu district (69.4%) in SEC zone. Xinzhou and Xiangzhou districts in YZR zone recorded the lowest aflatoxigenic populations with 4.3% and 3.1%, respectively. Among the four agroecological zones, the highest proportion of isolates producing >1000 ng•mL −1 level of aflatoxin was observed in SEC zone (55.4%) which was 5.4 times higher than other zones. The YZR zone had the lowest percentage of isolates (10.2%) producing >1000 ng•mL −1 of aflatoxin. Variation of among districts in the percent of aflatoxin-producing Aspergillus isolates producing quantities of AFB1. ND, none AFB1 detected; 0-100, producing of AFB1 by isolates was 0-100 ng•mL −1 ; 100-1000, producing of AFB1 by isolates was 100-1000 ng•mL −1 ; >1000, production of AFB1 by isolates was >1000 ng•mL −1 ; lines not sharing a common letter are significantly different (p ≤ 0.05) according to Tukey's HSD test on ranks of AFB1 concentrations. Variation of among districts in the percent of aflatoxin-producing Aspergillus isolates producing quantities of AFB 1 . ND, none AFB 1 detected; 0-100, producing of AFB 1 by isolates was 0-100 ng·mL −1 ; 100-1000, producing of AFB 1 by isolates was 100-1000 ng·mL −1 ; >1000, production of AFB 1 by isolates was >1000 ng·mL −1 ; lines not sharing a common letter are significantly different (p ≤ 0.05) according to Tukey's HSD test on ranks of AFB 1 concentrations.

Figure 4. Variation of among four agroecological zones in the percent of aflatoxin-producing
Aspergillus isolates producing quantities of AFB1 in China. ND, none AFB1 detected; 0-100, production of AFB1 by isolates was 0-100 ng•mL −1 ; 100-1000, producing of AFB1 by isolates was 100-1000 ng•mL −1 ; >1000, producing of AFB1 by isolates was >1000 ng•mL −1 ; a line not sharing a common letter is significantly different (p ≤ 0.05) according to Turkey's HSD test on ranks of AFB1 concentrations.
Aflatoxin-producing potential varied among isolates, species, districts, and agroecological zones ( Table 2). All the tested A. parasiticus isolates produced AFB1. The average production of AFB1 by the A. flavus isolates was 3420 ng•mL −1 (range: 0-82,083 ng•mL −1 ), which was significantly lower than AFB1 production of A. parasiticus isolates with an average of 14,780 ng•mL −1 (range: 5.8-61,899 ng•mL −1 ). A. flavus contributed the most to the average aflatoxin-producing potential in fungal community resident in twelve districts (Table 2). A. parasiticus made a greater contribution to the aflatoxin-producing potential of fungal communities in the districts of Zhanjiang, Qingdao, and Yantai, and in NE zone. According to Pearson's correlation analysis (Table 3), there was a significant positive correlation between the average aflatoxin-producing potential of fungal communities and percentage of the isolates which produced high levels of AFB1 (>1000 ng•mL −1 ) (r = 0.87, p = 0.01).
The average AFB1 producing potential varied widely among the zones. Average AFB1 producing potential of Aspergillus isolates in the SCE zone (5795 ng•mL −1 ) showed higher significance than the YZR zone (1566 ng•mL −1 ), but not much of a significant difference from the YR zone (5453 ng•mL −1 ). The average AFB1-producing potential of Aspergillus isolates in the NE zone (1886 ng•mL −1 ) was the least significant compared to other zones. Within the zones, the highest average of AFB1 concentration was observed in Yantai district (18,784 ng•mL −1 ) from YR zone, and the least in Xinzhou district (73 ng•mL −1 ) from YZR zone ( Table 2).
The potential risk areas with respect to aflatoxigenic Aspergillus sp and their AFB1 production varied among four agroecological zones (Table 4). Based on the AFB1 production, its highest level was seen in SEC zone, where the average production from the isolates taken from 1 gram of soil was 5795 ng•mL −1 ( Table 4). The second was YR zone, with an average AFB1 production of 5453 ng•mL −1 . The least were NE and YZR zones, where the average aflatoxin production level was lower in Figure 4. Variation of among four agroecological zones in the percent of aflatoxin-producing Aspergillus isolates producing quantities of AFB 1 in China. ND, none AFB 1 detected; 0-100, production of AFB 1 by isolates was 0-100 ng·mL −1 ; 100-1000, producing of AFB 1 by isolates was 100-1000 ng·mL −1 ; >1000, producing of AFB 1 by isolates was >1000 ng·mL −1 ; a line not sharing a common letter is significantly different (p ≤ 0.05) according to Turkey's HSD test on ranks of AFB 1 concentrations.
Aflatoxin-producing potential varied among isolates, species, districts, and agroecological zones ( Table 2). All the tested A. parasiticus isolates produced AFB 1 . The average production of AFB 1 by the A. flavus isolates was 3420 ng·mL −1 (range: 0-82,083 ng·mL −1 ), which was significantly lower than AFB 1 production of A. parasiticus isolates with an average of 14,780 ng·mL −1 (range: 5.8-61,899 ng·mL −1 ). A. flavus contributed the most to the average aflatoxin-producing potential in fungal community resident in twelve districts (Table 2). A. parasiticus made a greater contribution to the aflatoxin-producing potential of fungal communities in the districts of Zhanjiang, Qingdao, and Yantai, and in NE zone. According to Pearson's correlation analysis (Table 3), there was a significant positive correlation between the average aflatoxin-producing potential of fungal communities and percentage of the isolates which produced high levels of AFB 1 (>1000 ng·mL −1 ) (r = 0.87, p = 0.01).
The average AFB 1 producing potential varied widely among the zones. Average AFB 1 producing potential of Aspergillus isolates in the SCE zone (5795 ng·mL −1 ) showed higher significance than the YZR zone (1566 ng·mL −1 ), but not much of a significant difference from the YR zone (5453 ng·mL −1 ). The average AFB 1 -producing potential of Aspergillus isolates in the NE zone (1886 ng·mL −1 ) was the least significant compared to other zones. Within the zones, the highest average of AFB 1 concentration was observed in Yantai district (18,784 ng·mL −1 ) from YR zone, and the least in Xinzhou district (73 ng·mL −1 ) from YZR zone ( Table 2).
The potential risk areas with respect to aflatoxigenic Aspergillus sp. and their AFB 1 production varied among four agroecological zones (Table 4). Based on the AFB 1 production, its highest level was seen in SEC zone, where the average production from the isolates taken from 1 gram of soil was 5795 ng·mL −1 ( Table 4). The second was YR zone, with an average AFB 1 production of 5453 ng·mL −1 . The least were NE and YZR zones, where the average aflatoxin production level was lower in isolated strains. As soil is the main reservoir of aflatoxin-producing Aspergillus inoculums and aflatoxin contamination in peanuts, the survey of Aspergillus sp. and their aflatoxin production from 1 gram of soil could reflect the potential risk of contamination. The potential peanut production can be at a higher risk of AFB 1 contamination in YZR zone based on the survey, as no isolates in 1 gram of soil were reported as the highest, and least in the NE zone. Table 4. Variation among agroecological zones of China for colony-forming units of aflatoxin-producing Aspergillus, positive rate, the mean quantities of producing AFB 1 of isolated strains, potential producing AFB 1 ability in soil a .

Discussion
This study provided the first comprehensive documentation of the potential risk of aflatoxin contamination on peanuts by aflatoxigenic Aspergillus sp. in soils from the major peanut producing regions of China. Although aflatoxigenic Aspergillus strains have been reported from various crops and agricultural commodities, agricultural soil serves as the main reservoir of these fungi all over the world [15][16][17][18]. In the present study, A. flavus was the dominant species of Aspergillus fungal population in peanut soil of all districts, which was 94.2% of all strains. Aflatoxin-producing Aspergillus sp. was seen in all districts. In Brazil, A. flavus was the most frequent species of the genus Aspergillus in soil samples from four peanut production regions (13.4%) [19]. In Argentina, the Aspergillus population recovered from peanut seeds showed A. flavus as the most frequently isolated (79%) strain [20]. Not only in peanuts, A. flavus is the predominant species in the soils and vegetables of corn, cotton and other crops [7,[21][22][23][24]. Also A. flavus was widely seen reported in temperate and tropical regions [25,26].
A. parasiticus was isolated in all four agroecological regions. A. parasiticus is usually found reported in sugar cane, grapes or cassava from tropical and subtropical regions. In SEC zone, which is a tropical and subtropical region, A. parasiticus strains had been isolated from Zhanjiang (Sugar cane as rotating crop) and Yunfu districts (Cassava as rotating crop), which was consistent with the previous studies. But A. parasiticus isolated in the field soils of Yicheng district from YZR zone, Qingdao, Yantai, and Weifang districts from YR zone and Fuxin, and Shenyang from NE zone had maize as the rotating crop. We speculate that A. parasiticus may have had a transmission route with corn as the medium, and the transmission extended from tropical and subtropical to the warm temperate zone and temperate zones.
Sclerotia have been demonstrated to be the survival structure for many fungi. Because sclerotia of aflatoxin-producing Aspergillus can germinate sporogenically, they could be a potential source of primary inoculums. In this study, nearly 60% of the isolates produced sclerotia, which suggest that it is essential for the survival of aflatoxigenic Aspergillus in the peanut ecosystem in China, together with mycelia and conidia. These results are similar to previous studies [4,29].
From the soil samples of China's peanut-growing region, the NS-strains were isolated in the highest percentage and the S-strains with the lowest percentage. Barros [4] reported the isolation of L-strains in highest percentage from soil samples of peanut-growing region in Argentina. Soil isolates of A. flavus along a transect within the United States were identified as members of either the L-strains (n = 774), or the S-strains (n = 309) [13]. Orum [30] reported ranges of S strain incidence from less than 5% to more than 90%, and the association with cotton cultivation in the southern United States. In Arizona, incidence of S strains is inversely correlated with elevation [31].
So far, researchers have done some research on the effect of crop rotation on A. flavus types. Nesci and Etcheverry [32] recovered only A. flavus L phenotype from insects, soil and debris samples from corn fields alternately cropped with peanuts and soybeans from the same region. In our study, the isolation frequency of the L-strains within the aflatoxin-producing Aspergillus was higher. The L phenotype was recovered in the highest percentage in the districts of Guangzhou, Shaoguan, Huanggang, Yicheng, Qingdao and Liaocheng, and the crop rotation in these six districts were either rice and corn (rice, rice, corn, corn, corn, and corn respectively), so the rotation between corn or rice towards peanut was conducive to the growth and reproduction of the L phenotype. On the other hand, Horn et al. [33] and Cotty et al. 12 found that the S strain was widely distributed in cotton-producing areas in the United States. The condition responsible for the S strains distribution appeared to be complex. In the fields along a transect through the peanut-growing region of the USA described by Horn and Donner 13, the S strain was highly prevalent in west central Texas and Louisiana, where cotton is grown extensively. These studies and our results suggest that the crop could be selecting the phenotype found, as was reported by Garber [3] and Barros [4].
The soil type, landform and rainfall had a greater influence on the growth of aflatoxin-producing Aspergillus in different agroecological zones. In SEC zone, the top three districts with a higher population density of aflatoxigenic Aspergillus sp. and positive rate were Yunfu (481.1 cfu·g −1 , 66.7%), Shaoguan (258.6 cfu·g −1 , 46.7%) and Zhanjiang (170 cfu·g −1 , 56.7%), and the soil types of these districts were all arid hillside, and the least was observed in Guangzhou (27.8cfu·g −1 , the positive rate 36.7%) and Shantou (19.9 cfu·g −1 , the positive rate 36.7%) where the soil type was paddy soil. The rainfall in the southeast coastal region is more due to the subtropical monsoon climate, and paddy soil has poor drainage, resulting in soil with high water content which is not conducive to the growth of aflatoxigenic Aspergillus. However, the hillside had good drainage and water retention, which produced suitable soil moisture for the growth of A. flavus. In YZR zone, the district with the highest population density and positive rate was Xiangzhou (2749.3 cfu·g −1 , 90%), the second was Huanggang district (1920 cfu·g −1 ,100%), and the least was observed in Xiaogan (55.5 cfu·g −1 , 33.3%). High temperatures and a drought period with very little rainfall were observed before two months of peanut harvest in the YZR zone. Meanwhile, the agrotype of Xiangzhou was clay loam and the landform of Huanggang district was plains, which were beneficial to maintaining soil moisture for the growth of aflatoxin-producing Aspergillus. In Xiaogan, the agrotype was sandyloam and the landform was arid hillside, which all were detrimental to water retention, so soil moisture in Xiaogan was too low to inhibit the growth of aflatoxin-producing Aspergillus sp.
The majority of aflatoxin-producing Aspergillus isolates from peanut soil across the four agro-ecological zones was aflatoxigenic. In previous studies, the average aflatoxin-producing potential of fungal communities highly varied. In the southern USA [13,34] and in Argentina's peanut the majority of Aspergillus isolates produced aflatoxins [35]. While in Iran, only 27.5% of aflatoxin-producing Aspergillus isolates from corn soil were toxigenic [36], and in Nigeria [27], 44% of aflatoxin-producing Aspergillus isolates were aflatoxigenic. Different results of aflatoxigencity among Aspergillus section Flavi population may be attributed to differences in prevailing climatic conditions, the cultivar grown, and local agricultural practices.
We observed a positive association between aflatoxin production and sclerotia phenotype in A. flavus isolates from China peanut soil since the S strains produce higher levels of aflatoxin than the L strain isolates and similar results were found in Argentina [4,37]. These results are supported by previous studies that showed an evident interrelationship between regulation, biosynthesis and sclerotia morphogenesis [4,38].
Orum et al. [30] postulated that temperature, soil condition, day length, crop sequence history, insect levels, rainfall frequency and management practice may influence aflatoxin-producing Aspergillus communities. All these factors and many other micro-climatic factors are different between these four agroecological zones of China. In the present study, the population density had a significant positive correlation with positive rate (r = 0.65). Positive rate of Aspergillus sp. had a significant positive correlation with temperature (r = 0.61), and a significant negative correlation with longitude (r = −0.71), had positive correlation with pH (r = 0.48). The incidences of the L-strain had significant positive correlation with soil pH and had a negative correlation with latitude. In Nigeria, the incidences of the L-strain also had a significant negative correlation with latitude [27].
The AFB 1 producing potential isolates in field soils significantly varied among the four peanut production areas. YZR had the largest number of AFB 1 producing potential isolates, while the least was NE. Ding [39] researched the distribution of AFB 1 contamination in post-harvest peanut in China, and the highest was observed in the Yangtze River ecological region and the lowest in NE. In our present study, the Yangtze River ecological region had the largest AFB 1 -producing potential isolates in 1 g soil, and AFB 1 contamination in post-harvest peanut in this region was also reported to be higher. Meanwhile, NE had the least AFB 1 producing potential isolates in 1 g soil, and AFB 1 contamination in post-harvest peanut in this region was also lowest. AFB 1 contamination in post-harvest peanuts was closely related to the AFB 1 producing potential of peanut fields. Therefore, we drew the conclusion that AFB 1 contamination risk mainly came from aflatoxin-producing Aspergillus sp. in peanut field soils. We can predict that the highest risk zone for AFB 1 contamination in peanuts is YZR zone, followed by SEC and YR zone. However, NE zone tends to be highly safe for peanut cultivation.
The potential risk of AFB 1 contamination in peanuts will increase with an increase in population density and positive rate of aflatoxin-producing Aspergillus strains in peanut field soils. YZR had the higher population density and positive rate of aflatoxin-producing Aspergillus strains (1039.3 cfu·g −1 , 80.7%), the next highest were SEC (191.5 cfu·g −1 , 48.7%) and YR (26.5 cfu·g −1 , 22.7%), and the last was NE (2.4 cfu·g −1 , 6.6%). These results suggest that the reduction in the number of aflatoxin-producing Aspergillus strains in field soil is crucial for controlling AFB 1 contamination in peanuts. Novel biological control technology has been developed in recent years that can prevent AFB 1 contamination in peanuts. The application of non-toxigenic A. flavus strains to the peanut crop seems to be one of the most efficient strategies [40,41].

Conclusions
This study has shown that A. flavus is the dominant species in peanut soil fungal population in all the agroecological zones, with widespread aflatoxigentic strains. The YZR zone is highly prone to AFB 1 contamination risk in peanuts, as the study has shown that it has the highest Aspergillus sp. population density with a positive rate of aflatoxin production. However, the number of A. parasiticus identified was lower compared to A. flavus, though, their presence should not be overlooked, as they indicate the possibility of high-risk exposure due to their high level of AFB 1 production. The influence of AFB 1 through peanuts on human populations in China over the past decade demonstrates a clear need for tools to manage contamination of locally produced peanuts. Given the widespread nature of AFB 1 -producing strains, any control strategy should include field interventions.

Survey Sites
Soil samples were collected from peanut fields across four agroecological zones: Southeast coastal (SEC), the Yangtze River (YZR), the Yellow River (YR), and the Northeast (NE). Five districts were selected for study sites within each agroecological zones, and these which districts were separated geographically by at least 50 km. Field ecology information for each sampling region are shown in Table 5.

Survey Methods
Six hundred soil samples from fifteen districts (30 from each district) were collected at harvest time. Each soil sample (100 g) consisted of a pool of five subsamples taken with a trowel from the top 5 cm of soil at 5-10 m intervals. The samples were placed in plastic bags with pinholes for gas exchange and transferred to the laboratory and stored at 4-5 • C for further assay [4].

Strain Isolation and Identification
Ten grams of soil from each of the total 600 soil samples were diluted with 90 mL of 0.1% (w/v) peptone water and kept at room temperature (25 ± 2 • C) for 20 min on a thermostatic shaker (Hunan Xiangyi Instrument Co., Ltd, Changsha, China. This mixture was decimally diluted and a 0.1 mL aliquot was spread on dichloran-18% glycerol (DG18) at appropriate dilution to allow collection of isolates from plates with fewer than 10 colonies. The plates were incubated in the dark for 5-7 days at 30 • C. At the end of the incubation period, the average number of duplicate colonies was determined and the results were expressed as colony-forming units per gram (cfu·g −1 ) of soil. Isolates were sub-cultured at 30 • C on malt extract agar (MEA).
All isolates were preliminarily identified based on their characteristic growth patterns on AFPA (A. flavus and A. parasiticus agar) [43] and CYA (Czapek yeast autolysis) [44]. The identities of the strains were further confirmed by molecular analysis, which involved sequencing of their calmodulin genes [44]. The calmodulin gene in each isolate was amplified using the primers CL1 (GARTWCAAGGAGGCCTTCTC) and CL2A (TTTTTGCATCATGAGTTGGAC).

Classification of Aflatoxin-Producing Aspergillus morphotypes
To induce the production of sclerotia, the strains were incubated in Czaper medium, in triplicate, and maintained in darkness at 30 • C for 14 days. Following this period, Tween 20 (100 µL/L, 5 mL) was added, and the surface was scraped so that the mycelia could be filtered through Whatman No.2 filter paper. The sclerotia were washed in distilled water and placed in microtubes until further analysis [37].
To measure the mean diameter of the sclerotia for each strain, 10 sclerotia from each replicate were randomly selected and the arithmetic mean was calculated.

Mycotoxin Analyses
The liquid fermentation method used by Barros [4] was modified and used for qualitative determination of AFB 1 production by aflatoxin-producing Aspergillus. The strains were induced to sporulate on MEA slants at 28 • C for 5 days. After incubation, 5 mL of Sterilized distilled water was added to the slant followed by vigorous agitation to obtain a spore suspension. The spore concentration was measured using a cell counting plate and adjusted to 10 6 spores mL −1 . Approximately 10 5 spores were used to inoculate 50 mL vials containing 10 mL of liquid medium made by dissolving 150 g of sucrose, 20 g of yeast extract, and 10 g of soytone, in 1 L of distilled water; the pH of the medium was adjusted to 6.0 with HCl. One vial was incubated per isolate and performed in triplicate. The cultures were incubated for 7 days at 30 • C with 200 rpm in the dark. Vial cultures were analyzed by high-performance liquid chromatography for the production of AFB 1 . HPLC analysis was performed using Waters 2695 (Waters Corporation, Milford, MA, USA) coupled to Waters 2475 fluorescence detector (λ exc 360 nm; λ em 440 nm) and a post-column derivation system, and an Agilent TC-C18 column (250 × 4.6 mm, 5 µm particle size). The mobile phase (water:methanol:acetonitrile, 4:1:1) was pumped at a flow rate of 0.5 mL/min. AFB 1 procured from Sigma-Aldrich (St. Louis, MO, USA) was used as standard. The mean recovery of the method used was calculated by culture medium at different levels ranging from 0.5 to 100 ng/g of AFs and was estimated at 95.2% ± 8.4%. The lowest detection limit was 0.5 ng of AFB 1 per mL [45].

Statistical Analysis
Analyses were performed with SPSS (version 18.0). Analysis of variance was performed on all data with the general linear model (GLM), suitable for unbalanced data. The GLM of SPSS uses the least-squares method to fit data to a general linear model. Tukey's honestly significant difference (HSD) test was performed to compare treatment means at the 5% level. Pearson's correlation coefficients were generated to assess relationships between ecological and biological variables.