Les mycotoxines sont des molécules toxiques produites par de nombreuses espèces fongiques. Les seules mycotoxines avérées aujourd’hui cancérigènes pour l’homme sont les aflatoxines. Elles sont produites par le genre Aspergillus principalement et sont retrouvées tout au long de la chaine alimentaire (champs, stockage, transformation, etc.). A cause du réchauffement climatique, la France devient de plus en plus exposée à la présence de ces mycotoxines. Afin de limiter l’exposition des consommateurs, de nombreuses stratégies de prévention ou de décontamination sont développées. Dans ce contexte, nous avons recherché à mettre au point un système de lutte biologique permettant de prévenir la production d’aflatoxines sur le maïs au champ. Pour cela, nous avons choisi des bactéries issues du sol et déjà connues pour être commercialisées pour la lutte biologique, les actinomycètes. Nous avons étudié l’interaction in vitro sur boites de Pétri entre Aspergillus flavus, principal producteur d’aflatoxines, et certains actinomycètes. Nous avons démontré que l’interaction peut réduire la concentration en aflatoxines mesurée par HPLC. De plus, certains isolats bactériens sont aussi capables de réduire, en culture pure, la concentration d’aflatoxine B1 dans le milieu. Des premiers tests d’adsorption ont été réalisés pour comprendre la nature de ce mécanisme. Par ailleurs, une étude approfondie via RT-qPCR sur 6 souches bactériennes du genre Streptomyces sp. A montré que celles-ci étaient capables d’impacter l’expression de différents gènes impliqués dans la voie de biosynthèse chez A. flavus et A. parasiticus. Enfin, nous avons complété les données déjà existantes sur l’impact de facteurs environnementaux (température, disponibilité en eau et du temps d’incubation) sur la production d’aflatoxines. / Mycotoxins are toxic contaminants of foodstuffs produced by a wide range of fungal species. Aflatoxins are the only mycotoxins carcinogenic for humans. They are mainly produced by the Aspergillus genus and can be found at each step of the agrofood chain (e.g. field, storage, process). Due to climate changes, France is starting to be exposed to aflatoxins. In order to limit the consumer exposure, many prevention or decontamination techniques have been developed. To this aim, we started the development of a biocontrol against aflatoxins accumulation for maize field application. Actinomycetes, are soil-borne bacteria that has already been commercialized as biocontrol. In Petri dishes, we studied the in vitro interaction between some actinomycetes and Aspergillus flavus, the main aflatoxins producer. We revealed that the interaction reduced the aflatoxins content (monitored by HPLC). Moreover, some bacterial isolates were able to reduce pure-aflatoxin B1 added in the medium. To understand this mechanism, adsorption tests has been conducted. Otherwise, RT-qPCR methodology was used to study the impact of Streptomyces-Aspergillus sp. on aflatoxin gene expression. Finally, the current knowledge of the impact of environmental factors (temperature, water activity and incubation time) on aflatoxins production was supplemented.

>Magister Scientiae - MSc / Subsistence farmers may contribute significantly to food production, food security, and employment in South Africa. However poor storage practices and contamination with mycotoxins, particularly fumonisins and aflatoxins impacts adversely on production, food safety and food security. Mycotoxins are toxic natural food-borne compounds which frequently contaminate agricultural produce worldwide. They are hazardous to humans and animals and result in significant production losses for farmers. This study focused on former Bantustans in Northern South Africa, namely Vhembe District Municipality (Limpopo) and Gert Sibande District Municipality (Mpumalanga). The aim was to assess mycological and mycotoxin contamination of crops grown by subsistence farmers. A semi-structured questionnaire was administered to randomly thirty-nine households. Data on demographics, storage practices and production during period of 2011 and 2012 cropping seasons were collected. One hundred and fifteen (115) crop samples (maize, beans and peanuts) were collected for analysis. Standard mycological methods and validated mycotoxin analysis methods (HPLC and LC- MS/MS) were used. It was found that maize was the staple food in both provinces, with a significant difference (p = 0.0184) in its production between the two districts; Vhembe produced 0.6 tonnes compared to 2.4 tonnes in Gert Sibande. The majority of the farmers for storage used traditional open wooden cribs (15/20) and steel tanks (5/20) while VDM farmers used sealed store houses 5/19 and 15/19 used polystyrene sacks. Aflatoxin occurrence was low with <1% of GSDM samples contaminated compared to 11% of VDM samples. No significant difference (p > 0.05) was observed in the aflatoxin contamination in VDM samples between the year 2011 and 2012. Samples from VDM households had higher Aspergillus fungal infection (maximum incidence 69%) compared to GSDM (27%) over both seasons. The most frequently isolated Fusarium species in VDM samples was F. verticillioides (92%; 93%), and F. subglutinans (97%; 80%) in GSDM samples over seasons 2011 and 2012, respectively. Highest levels of fumonisins (FB1+ FB2) ranged between 1010 μg/kg and 12168 μg/kg with less than 30% extremely contaminated above the regulated limit in 91% of samples from Limpopo over both seasons (2011 and 2012). Fumonisin levels between the two seasons in VDM showed no significant difference (p>0.05). Only three (less than 5%) from 68% GSDM contaminated maize samples were above the FB1 and FB2 limit. In 2011, there were two highly contaminated maize samples (1762 μg/kg and 4598 μg/kg) with the other samples less than 600 μg/kg, whereas in season two (2012) all samples were below 200 μg/kg, except one highly contaminated sample (26115 μg/kg). None of the beans and peanuts from Mpumalanga was contaminated with mycotoxins above the recommended limit, but from Limpopo 1/5 peanuts was found contaminated with aflatoxin G1 (41 μg/kg). Natural occurrence and contamination of both fumonisin and aflatoxin in stored home-grown maize from VDM was significantly (p < 0.0001) higher than GSDM over both seasons. In general, Limpopo farmers’ experience lower harvests and greater mycotoxin contamination of agricultural produce. This may be attributed in part to poor storage practices and environmental and climatic conditions in that agro-ecological zone.

Maize is the principal staple food/feed crop exposed to mycotoxins, and the co-occurrence of multiple mycotoxins and their metabolites has been well documented. Dietary (co)-exposure to mycotoxins is associated with human and animal health concerns as well as economic losses. The present thesis aims to apply a holistic approach for the risk assessment of mycotoxin mixtures in food and feed, i.e. from fungal production and occurrence to harmonised risk characterisation. This was done in three folds. Firstly, available environmental, ecological, and agronomic factors that may affect the relative abundance of co-occurring mycotoxins in the contaminated crops were collected from peer-reviewed literature, with focus on maize (Chapter I). Secondly, (co-)occurrence data on mycotoxins in core cereals was extracted from available articles in the scientific literature and analysed to estimate potential pattern of co-exposure in humans and animals (Chapter II). Finally, Chapter III investigates the applicability of the EFSA guidance to multiple mycotoxins through a scenario of possible co-exposure in humans and animals, using maize as a case study. In particular, a human and animal risk assessment to mycotoxin mixture in maize was conducted using a modelled component-based approach for selected mixture of mycotoxins, that, according to our data, co-occur in maize based feed and food products. / Maize is the principal staple food/feed crop exposed to mycotoxins, and the co-occurrence of multiple mycotoxins and their metabolites has been well documented. Dietary (co)-exposure to mycotoxins is associated with human and animal health concerns as well as economic losses. The present thesis aims to apply a holistic approach for the risk assessment of mycotoxin mixtures in food and feed, i.e. from fungal production and occurrence to harmonised risk characterisation. This was done in three folds. Firstly, available environmental, ecological, and agronomic factors that may affect the relative abundance of co-occurring mycotoxins in the contaminated crops were collected from peer-reviewed literature, with focus on maize (Chapter I). Secondly, (co-)occurrence data on mycotoxins in core cereals was extracted from available articles in the scientific literature and analysed to estimate potential pattern of co-exposure in humans and animals (Chapter II). Finally, Chapter III investigates the applicability of the EFSA guidance to multiple mycotoxins through a scenario of possible co-exposure in humans and animals, using maize as a case study. In particular, a human and animal risk assessment to mycotoxin mixture in maize was conducted using a modelled component-based approach for selected mixture of mycotoxins, that, according to our data, co-occur in maize based feed and food products.

<p>  </p> <p>Food systems are facing unprecedented challenges and need reforms to be more efficient and provide safe and nutritious food from farm to fork. In this dissertation, I present new empirical evidence on different strategies to tackle two critical problems of food systems in sub-Saharan Africa (SSA): food contamination and the inefficient delivery and use of subsidized agricultural inputs. The first essay measures demand and network effects for a newly available technology called Aflasafe that delivers credence attributes (i.e., benefits difficult or impossible to observe and measure even after frequent use of the technology). The second essay exploits a policy change in Malawi to estimate the efficiency of public and private channels at distributing inputs to smallholder farmers. I use a panel dataset covering the period from 2010 to 2016 and a difference-in-differences (DID) framework to estimate the impact of the policy change on procurement of subsidized fertilizer, efficiency and service quality, agricultural productivity, and the household’s decision to purchase commercial fertilizer. The third essay identifies a series of constraints faced by smallholder farmers in Senegal and assesses which ones are associated with chemical and microbial contaminants in groundnut (peanut).</p>

No description available.