Distribution of Aspergillus and Fusarium ear rot causative fungi in soils under push-pull and maize monocropping system in Western Kenya

It is imperative to establish the distribution and density of soil fungal communities as a requisite for formulating strategies for management of ear rot infections and mycotoxin contamination. In a two seasons study, short (SR) and long rainy (LR) seasons, we investigated the distribution of Aspergillus and Fusarium fungi causing ear rots and producing mycotoxins from 120 soil samples collected from maize fields under push-pull (PP) and maize monocrop (MM) systems in Western Kenya. Cultural methods were used for identification of Aspergillus and Fusarium species, while molecular techniques were used for confirmation of Fusarium section Liseola. Detection of total aflatoxins in cultures of section Flavi isolates was carried out by enzyme-linked immunosorbent assay (ELISA). A total of 338 fungi were isolated; 80% were identified as Aspergillus and 4.4% Fusarium. The distribution of fungi was significant with season but not cropping systems. The frequency of occurrence was higher during the LR (68.4%) than the SR (31.6%). In cropping systems, the frequency of occurrence of Aspergillus flavus was higher in MM (60.2%) than PP (39.8%). However, Aspergillus parasiticus was more frequent in PP (71.4%) than MM (28.6%); and during the SR (78.6%) than the LR (21.4%). Majority (81.3%) of A. flavus and A. parasiticus were toxigenic. There was low recovery of Fusarium species in soil samples. These findings show that soils from both cropping systems are potential for Aspergillus infection and aflatoxins contamination; however, low Fusarium distribution in soil suggest external inoculum source for Fusarium ear rot infections common in most maize fields in Western Kenya.


INTRODUCTION
Fungi are part of diverse living components of soil, with several of them living as saprophytes and symbionts contributing to various soil services including structure formation, organic decomposition, recycling of major elements (for example carbon, nitrogen and phosphorus) and toxic removal (Aislabie and Deslippe, 2013). Pathogenic fungi also exist as major causal agents of soil borne diseases affecting roots, stalks, leaves and ears of *Corresponding author. E-mail: maxwelljuti@gmail.com.
Author(s) agree that this article remains permanently open access under the terms of the Creative Commons Attribution License 4.0 International License various crops including maize (Shurtleff, 1980). Nevertheless, the presence of certain non-pathogenic (mainly saprophytes) or pathogenic fungi on grains, soils and other reservoirs are potential for ear rot infection and mycotoxin production, especially species in the Aspergillus and Fusarium genera (Horn et al., 1995;Pereira et al., 2011).
The Aspergillus genus is divided into sections (or subgenus groups) of which Flavi is most important in agriculture as cause of ear rot diseases and producer of aflatoxins (Gnonlonfin et al., 2011). Several species are classified under Flavi, but Aspergillus flavus Link, Aspergillus parasiticus Speare and Aspergillus nomius Kurtzman (Rodrigues et al., 2007) are prominent isolates in maize and soil samples. Amongst these species, A. flavus and A. parasiticus are prolific producers of aflatoxins with the former being the most abundant in both air and soil (Hedayati et al., 2007), hence affecting more of aerial crops like maize. On the other hand, A. parasiticus is mostly reserved in soils with high isolation frequency of peanuts fields (Garber and Cotty, 2014).
The filamentous fungus with equal importance in maize production is Fusarium. Most of its members are producers of three important agricultural mycotoxins which include: fumonisins, deoxynivalenol and zearalenone (D'Mello et al., 1999). They are also causative agents of root and ear rots in maize resulting in yield losses (Sutton, 1982). Three Fusarium species with high frequency of isolation in maize include: Fusarium graminearum Schwabe, Fusarium verticillioides (Sacc.) Nirenberg, Fusarium proliferatum (T. Matsushima) Nirenberg and Fusarium subglutinans (Wollenweb and Reinkings) P. E Nelson, T.A. Toussoun and Marasas (Leslie and Summerrel, 2006).
Soil is the primary habitat for Fusarium and Aspergillus species. The population of Aspergillus and Fusarium propagules in the soil (field) increases the risk for maize infections and mycotoxin contamination (Horn, 2003). In order to safeguard against losses, fungal distribution in food and soil ecology is imperative for effective formulation of prevention and control measures. In soil fungal ecology, cultural practices greatly encourage or discourage fungal distribution. For instance, rotation of susceptible crops like wheat with maize together (Schaafsma et al., 2005) increase fumonisin incidence. Addition of organic matter either by cultural practice through minimum tillage, or application of organic amendments increases Aspergilli propagules (Zablotowicz et al., 2007) while decreasing those of Fusarium in soil (Alakonya et al., 2008).
Among several strategies used for soil management in western Kenya, push-pull technology (PPT), a companion cropping system where maize or sorghum is intercropped with moth repellent (push) forage legumes in the genus Desmodium, edged with moth attractive (pull) grasses such as Napier grass (Pennisetum purpureum Schumach) or Brachiaria, is mostly adopted by smallholder farmers (Khan et al., 2011). Although, PPT is known for insect pest management, it contributes to soil health improvement which is potentially impactful on soil fungal community. The technology improves organic matter content of the soil, nitrogen fixation, overall improvement in soil macro-and micro arthropods and conservation of soil moisture (Khan et al., 2011). However, distribution of Aspergillus and Fusarium ear rot fungi in soil under push-pull remains unknown. In this context, the aim of this study was to investigate the level of soil-borne Aspergillus and Fusarium species in pushpull and maize monocrop plots in western Kenya.

Field survey
The study sites included Kisumu, Siaya and Vihiga sub-counties (districts) of Western Kenya ( Figure 1). Soils in these sites are generally vertisols, ferralsol and nitosols showing a natural decline in soil fertility predominantly manifested by occurrence of purple witch weed, Striga hermonthica (Del.) Benth. (Orobanchaceae), soil erosion and deficient nitrogen and phosphorus. However, heterogeneity in soil fertility exists amongst smallholder farms in the region where there is less investment in external inputs to restore soil fertility. Push-pull technology has been disseminated for pest control and soil fertility improvement for over 10 years in the region (Khan et al., 2011).

Sample collection
The sampling method of Horn and Dorner (1998) was adopted for soil sampling with slight modifications. Briefly, transect which runs 5 km from one push-pull cluster to the next was made. A total of 60 fields were sampled at maize silking period during the short and long rainy seasons of 2014 and 2015, respectively. In a cluster, four push-pull and maize monocrop fields were sampled by removal of 4 subsamples of soil with a sterile trowel from the top 4 to 6 cm of soil at intervals of 2 to 4 m. The soil subsamples collected from each field were mixed and placed in a paper bag and air dried at 25°C for 1 week. The soil was then carefully mixed and sieved through a no.10 USA standard sieve (2.00 mm opening) (Dual Manufacturing Company, Franklin Park, IL 60131, USA) and stored at 4°C.

Isolation of fungi
The dilution plate technique by Cotty (1994) and Leslie and Summerell (2006) were used for Aspergillus and Fusarium recovery, respectively. One gram of thoroughly mixed soil samples was suspended in 9 ml of distilled water. These resultant solutions were serially diluted to 10 -3 . One milliliter of 10 -2 and 10 -3 were plated in quadruplicate in Petri dish (90 x 15 mm) containing a quarter strength potato dextrose agar (PDA) (HiMedia Laboratories Pvt. Ltd) amended with 30 mg chloramphenicol. The plates were then incubated at 31°C for 6 days in the dark for Aspergillus recovery, and at 25°C for 14 days for Fusarium. Colonies of Aspergillus and Fusarium that grew on each plate were counted and their population determined as colony forming unit (CFU) per gram and calculated as follows: Total fungal colonies = Number of colonies x dilution factor/weight of soil (1 g) Colonies of Aspergillus and Fusarium were then sub-cultured on full strength PDA amended with 30 mg chloramphenicol.

Morphological identification of Aspergillus and Fusarium
The colonies on PDA identified as Aspergillus were transferred aseptically onto Czapek Dox Agar (CZ) (Oxoid Ltd, Basingstoke, Hampshire, England) plates and incubated at 31°C for five days. Their colony characteristics (colour and reverse) were observed. Those characterized to belong to Aspergillus section Flavi were confirmed on Aspergillus Flavus Parasiticus Agar (AFPA) base (HiMedia Laboratories Pvt. Ltd) plates incubated at 25°C for five days for positive orange reverse. Microscopic features such as: head serration, vesicle and conidia were observed in a compound light microscope (Carl Zeiss MicroImaging GmbH 37081, Gottingen, Germany) using keys by Klich (2002).
Fusarium colonies recovered were grown on PDA plates and observed for pigmentation on both top and reverse, and on Spezieller Nahrstoffarmer Agar (SNA) for macroconidial features. Further identification using species-specific primers was used for identification of F. verticillioides, F. proliferatum and F. subglutinans.

DNA extraction
Fusarium isolates, 13 in total, culturally identified to belong to Fusarium section Liseola were grown as monosporing cultures on PDA plates for seven days at room temperature. For each isolate, mycelium was harvested for total DNA extraction according to Gherbawy et al. (2001). One gram of freshly harvested mycelium was ground in liquid nitrogen with a mortar and pestle into a very fine powder. Fifty milligrams of the ground mycelium was transferred into 1.5 ml Eppendorf tube and mixed with 700 µL 2 x CTAB buffer. The contents of Eppendorf tube was incubated at 65°C for 30 min before addition of 700 µL of chloroform : isoamyl alcohol (24:1 v/v), and a brief mixing. The mixtures were then centrifuged at 10,000 g for 30 min and supernatant was transferred into another tube. Isopropanol, 700 µL in volume was added and mixed with the supernatant and left to chill overnight at -20°C. This content was centrifuged again at 10,000 g for 5 min, after which the supernatant was discarded and pellets washed twice in 1 mL of 70% ethanol and left to dry under a vacuum. The pellets were afterwards resuspended in 700 µL distilled water. The quality of DNA was evaluated in 1% agrose gel electrophoresis.

Detection of Fusarium DNA using species-specific primers
The following primer pairs, VER 1/2, PRO 1/2 and SUB 1/2 were used for identification of F. verticillioides, F. proliferatum and F. subglutinans, respectively in PCR assay according to Rahjoo et al. (2008)

Test for aflatoxigenicity
Twenty seven species identified belonging to Aspergillus section Flavi were grown on PDA at 31°C for a period of 7 days and total aflatoxin were extracted from their cultures according to method described by Rao et al. (1997). A whole sample comprising agar, mycelia and spores was ground in a blender for 5 min. Two grams of the blended mixture was used to extract total aflatoxins in 10 mL of 60% methanol. The mixture was then filtered (Whatman #1) and the filtrate analysed by using ELISA Kits for total aflatoxin (Helica Biosystem Inc.).

Data analysis
All analyses were done using R version 3.3.1. The incidences of fungi were presented in counts and percentages as score of total counts. The data for total fungi, Aspergillus and A. flavus were normalized by log x+1 transformation before mean comparison.

Distribution of fungi in soil
A total of 338 fungi in four genera were isolated from soil samples (Table 1). From these isolates, 80% were Aspergillus, 8.9% Penicillium, 5.6% Trichoderma and 4.4% Fusarium. The incidence of fungi with district was not statistically significant. However, the incidence of Fusarium was low in Kisumu (1.6%) than Siaya (6.1%) and Vihiga (6.3%). This was converse to Penicillium which had high (12.5%) incidence in Kisumu than Siaya (5.3%) and Vihiga (8.3). There was significant difference (t (118) = 4.6018, p˂0.001) in distribution of total fungi in short and long rainy season. SR had lower (29.9%) incidence than LR (70.1%). Similarly, the incidence of Aspergillus was significant (t (118) = 2.1683, p˂0.001) in SR and LR season. Incidence of total fungi and in both PP and MM was not significant. However, Penicillium had high incidence in PP (12.7%) than MM (5.2%).

Identification of Aspergillus and Fusarium species
Three species belonging to Aspergillus section Flavi were identified by colony reverse on AFPA agar (Plate 1). The three species further identified on CZ based on their conidial colour and head serration were A. Flavus; yellow green surface and numerously biseriated [Plate 2 a(i) and a(ii) ]; A. parasiticus, conifer green surface and mainly uniseriated There was low recovery of Fusarium species causing ear rots; however, 13 isolates recovered morphologically belong to F. verticillioides [Plate 2 g(i) and g(ii)]. Out of these isolates (13), 9 were positive ( Figure 2) for F. verticillioides after molecular characterization with TEFgene.

The population of Aspergillus and Fusarium species in soil
In general, there was no significant difference in population of total and individual fungal species between the two cropping systems (Table 4). However, low population of fungi was observed in push-pull (2,266.

DISCUSSION
There was higher (averagely 80%) incidence of Aspergillus than other fungi in all the districts, cropping systems and seasons observed in this study. This corroborates findings of other studies in different agroecological areas in Kenya that reported relatively higher incidence of Aspergillus relative to other fungi (Okoth et al., 2012;Karanja, 2013). However, insignificant difference in incidence of Aspergillus between push-pull and maize monocrop systems contradicted the finding which showed significant increases in Aspergillus population with minimum tillage and organic matter amendments (Zablotowicz et al., 2007). Thus, more Aspergillus expected on a conserved system like pushpush which improves organic matter content in the soil and reduces the amount of tillage was not observed. This observation could be explained on the basis that historically, and depending on the cropping season and amounts of rainfall, most farms in western Kenya more often have maize intercropped with food legumes such as common bean (Phaseolus vulgaris L.) and peanuts (Arachis hypogaea L.) (Mudavadi et al., 2001). Such edible legumes provide beneficial ecological services of soil improvement through addition of organic matter and nitrogen fixation that could increase Aspergillus incidence in the soil (Mudavadi et al., 2001).
Soil as the main reservoir for both A. flavus and A. parasiticus has relatively higher frequency of the former   n, number of isolates (%); raw percentages calculated based on counts within district, cropping system and season. Aspergillus species than the latter (Klich, 2007). However, the frequency of A. parasiticus is comparatively higher and more endemic in soils where peanut or sugarcane is grown relative to that under maize (Garber 0.5 Kb 1 Kb and Cotty, 2014). Although, not measured in the current study, soil temperature has been reported to influence incidence of these fungi, with lower temperatures favoring A. parasiticus relative to A. flavus (Horn, 2005). Optimally, A. parasiticus grow at temperature of 22°C, while A. flavus, at 30 to 37°C (Horn, 2005). This cool soil temperature is encouraged by cultural practices such as cover cropping, reduced tillage (Sławiński et al., 2012), and wet season (Horn et al., 1995). In the push-pull system, Desmodium provides soil cover for a longer period due to its perennial nature as compared to annual edible intercrop legumes common (bean and peanuts) in western Kenya. The push-pull system also manifest limited tillage practices during land preparation and weeding for conservation, and from cover cropping of Desmodium, respectively. This explain probable low soil temperature in PP thus higher population of 71.4% was observed on A. parasiticus in soil samples from push-pull as compared to 28.6% in maize monocrop systems. The long dry spells which increases soil temperature in long rainy season than the short rainy season (Mugalavai et al., 2008) also account for low (21.4%) population of A. parasiticus during the long rainy season relative to 78.6% during the long rainy season in this study. Several studies on aflatoxin production have reported fewer incidences of non-aflatoxin (atoxigenic) producers amongst A. parasiticus isolates (Tran-Dinh et al., 2009;Barros et al., 2006), except in few cases (Okoth et al., 2012;Salano et al., 2016). The current study supports these findings as 8.3% of A. parasiticus isolates as compared to 23.7% of A. flavus were positive for aflatoxin production. With more aflatoxigenic fungi, the merit of conserved systems in increasing soil agricultural sustainability might also expose crops to aflatoxin contamination by increasing their A. flavus propagules in soils (Zablotowicz et al., 2007). However, contamination of maize is not entirely dependent on the population of A. flavus in the soil since maize intercropping which encourages more A. flavus has shown low aflatoxin contamination as compared to sole cropping system (Mutiga et al., 2015). Therefore, as revealed in these studies, intercrops are able to reduce Aspergillus infections and contamination through other factors such as increased soil nitrogen and limiting insect damage (Brun, 2003).
The frequency of A. parasiticus or ratios of A. flavus/A. parasiticus (4:1) in this study suggest the potential levels of contamination in maize. Studies show that A. parasiticus is comparatively a poor colonizer of aerial plants like maize (Horn, 2003) and have low spore density in air (Horn et al., 1995) than A. flavus (Hedayati et al., 2007). Indeed, study by Angle et al. (1982) observed almost complete infection of maize ears with A. flavus despite high incidence of both A. parasiticus and A. flavus in soil. Therefore, increased frequency of occurrence of A. parasiticus in push-pull relative to maize monocrop warrants further investigation.
The observations of this study presented A. terreus, A. niger and A. fumigatus as equally abundant in soil, with respective 30, 45 and 69 isolate counts as compared to 98 of A. flavus. This observation corroborates reports of most studies on distribution of microflora in the soil (Horn et al., 1995;Horn, 2005). However, they contradicted study by Salano et al. (2016) which reported higher (55) count of A. niger than A. flavus (26) in eastern province of Kenya. The high presence of these species portends less impact on grain quality as they are not chief producers of agriculturally important mycotoxins (D'Mello et al., 1999;Gnonlonfin et al., 2011). However, recent studies have reported production of fumonisin and ochratoxins A by A. niger (Mogensen et al., 2010;Palencia et al., 2010) and territrem by A. terreus (El-Sayed Abdalla et al., 1998), while A. fumigatus is known causal agent of invasive aspergilosis (Hedayati et al., 2007) The observations in this study illustrated low (18) isolate count of Fusarium section Liseola and no F. graminearum isolates in soils. This was similar to the study by Okoth and Siameto (2010) on soils in maize fields. The most plausible explanation for this occurrence could be their inherent scarcity (Okoth and Siameto, 2010) or effects of organic matter in the soil (Alakonya et al., 2008) from intercropping systems common in western Kenya. But importantly, low soil Fusarium incidence indicated more Fusarium infection from aerial spores and external sources.
The cultural identification in Fusarium section Liseola is demanding and limiting (Summerell et al., 2003), thus molecular methods are used for confirmation. In molecular identification of F. verticillioides using translation elongation factor 1-alpha (TEF) gene, 140 isolates culturally identified as F. verticillioides, 133 and 4 isolates were confirmed as F. verticillioides and F. proliferatum, respectively (Rahjoo et al., 2008). Therefore, further identification of species in Fusarium section Liseola using TEF genes is more accurate and reliable. Evidentially, in this study, 13 isolates were initially identified by cultural characteristics as F. verticillioide; 9 isolates were positive for F. verticillioides using TEF gene. However, F. proliferatum and F. subgluinans were not present amongst the isolates.
In conclusion, seasons had significant influence on distribution of Aspergillus and Fusarium fungi in soil, while cropping system did not. The high Aspergillus fungi in the soil in this study show that soil fungal community within the field is a potential risk for aspergillus ear rot infection and aflatoxin contamination, while the low frequency of F. verticillioides and F. graminearum in the soil samples suggest external inoculum as important for both gibberella and fusarium ear rot infection in the field.