Search

Sorghum is an important and versatile food crop in the Sudan savannah environments of West Africa. Sorghum production doubled in the last 20 years and its yield remained stable and low in Mali. Molecular markers were used to dissect quantitative variation to increase efficiency of breeding programs. From 2009–2014, marker-assisted recurrent selection (MARS), an integrated process to identify favorable alleles and to monitor their pyramiding throughout the breeding course, was implemented. Two large bi-parental populations derived from complementary recent elite inbreds, Tiandougou, Keninkeni, and Lata-3, were developed to improve grain yield productivity and stability while maintaining traditional food processing quality. Populations were phenotyped multi-locally. Quantitative Trait Loci (QTL) analyses identified, in each genomic region, the most favorable morphologic, yield components, grain quality, and flowering time parental alleles. Marker-assisted recurrent selection approach is detailed for one of the populations. Several genotypic ideotypes corresponding to ideal recombination patterns between the parents were defined. Molecular marker-guided recurrent crossing allowed favorable QTL accumulation for a large range of traits. Three marker-assisted recurrent selection cycles were accomplished. Four stable and superior elite lines were identified, leading to their registration in 2018 in national and ECOWAS (Economic Community of West African States) catalogs.

Cell walls in plants are mainly composed of polysaccharides. Among them, pectin, cellulose and hemicellulose are composed of various sugars residues (acidic: galacturonic acid in pectins; or neutral: rhamnose, galactose, arabinose, xylose, glucose in cellulose, hemicellulose and pectic side chain). This fine composition affects the structure but also the behavior and mechanical properties of cell walls and consequently the textural properties of RTB foods. In raw products and/or after cooking some cell wall components are soluble (eg. some pectin). This procedure seeks to determine the galacturonic acid level extracted from pre-dried (dried or freeze-dried) RTBs. This method gives the galacturonic acid content of cell walls extracted with different solvents, independently of the neutral sugars content. Pectin are first extracted at intermediate temperature (55°C) to facilitate their extraction while limiting their chemical and enzymatic degradation and the extraction of starch (temperature below starch gelatinization). Extracted pectin are then hydrolyzed by hot sulphuric acid in order to obtain sugar residues and their furfural derivatives. Neutral sugars gave brown derivatives whereas free galacturonic acid is quantified by colorimetry after reaction with MHDP which produces a specific fuschia pink complex with a maximum absorption at 520 nm (Blumenkrantz and Asboe-Hansen, 1973). Galacturonic acid content is calculated after subtracting the brown colour (at 520 nm) generated by neutral sugar hydrolysis.

During Period 4, RTBfoods partners involved in WP2 implemented the standard operating protocols (SOPs) developed in previous periods to characterize the target product profiles: boiled roots and tubers (cassava, Yam, sweet potato, plantain), paste (Fufu, pounded Yam, matooke) and granulated products (gari-eba, attiéké). For sensory characterizations by trained panels, 10 SOPs are now available: boiled cassava (2), gari-eba, attiéké, Fufu, boiled Yam, pounded Yam, boiled plantain, matooke, and boiled sweet potato. Key sensory attributes common to several products include firmness/hardness, stickiness, colour, smoothness, moldability and sweetness / sourness, fibrousness, moisture/mealiness, stretchability and bitterness. Several of these sensory attributes were correlated with instrumental texture parameters, in particular sensory firmness (or hardness or softness) assessed either in the mouth or by hand can be reliably predicted by the hardness parameters measured by penetrometry, TPA, and texture-extrusion. For the first time, mealiness of boiled cassava was correlated with a parameter of the texture-extrusion test, as well as with water absorption. Mouldability and stretchability were correlated with TPA measurements especially on destructured products (Gari/Eba and pounded Yam). For textural characterizations, 7 SOPs were developed and validated: boiled cassava (2), matooke, boiled Yam, pounded Yam, gari-eba, Fufu. These were implemented for several product profiles and several genotypes of various RTBs (typically 20 to 50 genotypes per product profile). Multivariate statistical analyses (e.g. PCA) of the resulting datasets allowed the classification of RTB genotypes into good, intermediate and poor quality, thus demonstrating the ability of postharvest quality measurements to contribute to varietal selection for breeding. Several correlations between textural parameters and sensory attributes were also identified (as discussed above), which also confirmed that instrumental texture is a valid approach to predict how consumers perceive RTB products. SOPs for several kitchen tests were also finalized, which bring additional information on the functional quality traits of RTB products in complement to texture measurements, including water absorption, cooking time, swelling of gari upon addition of hot water, and pounding time of paste products (pounded Yam, Fufu). Processing ability criteria were also investigated such as ease of peeling, ease of sieving the dewatered cassava mash, retting ability (softening, foaming, turbidity, pH). In the case of boiled cassava, the water absorption protocol in particular offers relatively short analysis time (30 minutes/sample) and is well correlated with sensory mealiness, cooking time and textural parameters, thus qualifying as medium-throughput phenotyping method for screening cassava breeding populations. In 2021, more than 4000 genotypes were characterized by water absorption, including from Nextgen and HarvestPlus populations. This development constitutes an important contribution of RTBfoods to improving cassava breeding and selection operations of other projects, and demonstrates the potential of the RTBfoods approach to develop breeding selection tools for a variety of post-harvest quality traits. Finally, biochemical proof of concept studies provided further evidence of the role of pectins in determining textural characteristics of RTB products, in particular non-destructured such as boiled or steamed. Starch and amylose contents also contributed to texture of boiled sweet potato and boiled Yam.

During Period 4, RTBfoods partners involved in WP2 implemented the standard operating protocols (SOPs) developed in previous periods to characterize the target product profiles: boiled roots and tubers (cassava, Yam, sweet potato, plantain), paste (Fufu, pounded Yam, matooke) and granulated products (gari-eba, attiéké). For sensory characterizations by trained panels, 10 SOPs are now available: boiled cassava (2), gari-eba, attiéké, Fufu, boiled Yam, pounded Yam, boiled plantain, matooke, and boiled sweet potato. Key sensory attributes common to several products include firmness/hardness, stickiness, colour, smoothness, moldability and sweetness / sourness, fibrousness, moisture/mealiness, stretchability and bitterness. Several of these sensory attributes were correlated with instrumental texture parameters, in particular sensory firmness (or hardness or softness) assessed either in the mouth or by hand can be reliably predicted by the hardness parameters measured by penetrometry, TPA, and texture-extrusion. For the first time, mealiness of boiled cassava was correlated with a parameter of the texture-extrusion test, as well as with water absorption. Mouldability and stretchability were correlated with TPA measurements especially on destructured products (Gari/Eba and pounded Yam). For textural characterizations, 7 SOPs were developed and validated: boiled cassava (2), matooke, boiled Yam, pounded Yam, gari-eba, Fufu. These were implemented for several product profiles and several genotypes of various RTBs (typically 20 to 50 genotypes per product profile). Multivariate statistical analyses (e.g. PCA) of the resulting datasets allowed the classification of RTB genotypes into good, intermediate and poor quality, thus demonstrating the ability of postharvest quality measurements to contribute to varietal selection for breeding. Several correlations between textural parameters and sensory attributes were also identified (as discussed above), which also confirmed that instrumental texture is a valid approach to predict how consumers perceive RTB products. SOPs for several kitchen tests were also finalized, which bring additional information on the functional quality traits of RTB products in complement to texture measurements, including water absorption, cooking time, swelling of gari upon addition of hot water, and pounding time of paste products (pounded Yam, Fufu). Processing ability criteria were also investigated such as ease of peeling, ease of sieving the dewatered cassava mash, retting ability (softening, foaming, turbidity, pH). In the case of boiled cassava, the water absorption protocol in particular offers relatively short analysis time (30 minutes/sample) and is well correlated with sensory mealiness, cooking time and textural parameters, thus qualifying as medium-throughput phenotyping method for screening cassava breeding populations. In 2021, more than 4000 genotypes were characterized by water absorption, including from Nextgen and HarvestPlus populations. This development constitutes an important contribution of RTBfoods to improving cassava breeding and selection operations of other projects, and demonstrates the potential of the RTBfoods approach to develop breeding selection tools for a variety of post-harvest quality traits. Finally, biochemical proof of concept studies provided further evidence of the role of pectins in determining textural characteristics of RTB products, in particular non-destructured such as boiled or steamed. Starch and amylose contents also contributed to texture of boiled sweet potato and boiled Yam.

Cell walls in plants are mainly composed of polysaccharides. Among them, pectin, cellulose and hemicellulose are composed of various sugars residues (acidic: galacturonic acid in pectins; or neutral: rhamnose, galactose, arabinose, xylose, glucose in cellulose, hemicellulose and pectic side chain). This fine composition affects the structure but also the behavior and mechanical properties of cell walls and consequently the textural properties of RTB foods. In raw products and/or after cooking some cell wall components are soluble (eg. some pectin). This procedure seeks to determine the galacturonic acid level extracted from pre-dried (dried or freeze-dried) RTBs. This method gives the galacturonic acid content of cell walls extracted with different solvents, independently of the neutral sugars content. Pectin are first extracted at intermediate temperature (55°C) to facilitate their extraction while limiting their chemical and enzymatic degradation and the extraction of starch (temperature below starch gelatinization). Extracted pectin are then hydrolyzed by hot sulphuric acid in order to obtain sugar residues and their furfural derivatives. Neutral sugars gave brown derivatives whereas free galacturonic acid is quantified by colorimetry after reaction with MHDP which produces a specific fuschia pink complex with a maximum absorption at 520 nm (Blumenkrantz and Asboe-Hansen, 1973). Galacturonic acid content is calculated after subtracting the brown colour (at 520 nm) generated by neutral sugar hydrolysis.

African locust bean (Parkia biglobosa) is a legume widely spread in west Africa. The major consumed part is the seeds after processing in the so-called Soumbala in Ivory Coast. It is rich in proteins and lipids and used as a condiment to enhance sauce flavour. The seeds processing can be generalized as a first cooking to allow the dehulling of the seeds, a second cooking of the almond, an alkaline spontaneous fermentation of 2 to 3 days before final sun-drying. During fermentation, microorganisms belonging to the genus Bacillus are responsible of an extensive proteolysis (Odunfa et Adewuyi, 1985) and lipolysis (Ouoba et al. 2003). This gives the specific flavour to the product but may lead to toxic compounds such as biogenic amines. Biogenic amines accumulate by microbial decarboxylation of free amino acids and are thus susceptible to be present in fermented protein rich foods. They were reported in wine, cheese and fermented fish while they have not been studied yet in processed African locust bean seeds. The objective is to evaluate amino acids decarboxylation activities of Bacillus spp. isolates from Soumbala and biogenic amines content. Soumbala samples were collected among 6 producers in 3 regions of Ivory Coast. 52 Bacillus spp. strains were isolated and identified using API 50 CHB kits. Histidine, lysine, arginine, phenylalanine and ornithine decarboxylation activities and arginine dehydrogenase activities were determined in order to study their ability to produce histamine, cadaverine, phenylethylamine and putrescine. Soumbala samples were analysed for histamine, cadaverine, phenylethylamine and putrescine content. Bacillus spp. strains were identified as followed: B. subtilis (16); B. lentus (3), B. pumilus (1), Geobacillus stearothermophillus (5), B. megaterium (9), B. firmus (2), B. circulans (2), Brevibacillus non reactive (5), B. licheniformis (4), Aneurinibacillus aneuriniliticus (4), B. cereus (1). 16 strains did not show arginine deshydrogenase activities and histidine, lysine, ornithine, or phenylalanine decarboxylation activities. Biogenic amines content of Soumbala was in average 82 mg.kg-1 for histamine, 36 mg.kg-1 for cadaverine, 45 mg.kg-1 for putrescine and 272 mg.kg-1 for phenylethylamine, some samples reaching high levels. Strategies for the control of biogenic amines production in African locust bean fermented seeds are discussed.

16 samples of sweet potatoes produced by CIP in Uganda were analysed by CIP in Uganda (cooking time, firmness, dry matter) then freeze-dried. Freeze-dried samples were than analysed by CIP-Peru (free sugars, starch, proteins by NIRS), JHI (beta-amylase activity, PME activity, cell wall characterization), CIAT (free sugars by HPLC, starch, amylose, RVA) and CIRAD (free sugars by HPLC, gelatinization and amylose by DSC and pectin contents by colorimetry). Unfortunately, no correlation between cooking time and firmness were observed and almost no significant correlation between cooking behaviour (cooking time or firmness) and the biophysical analyses (dry matter, starch characteristics, beta-amylase and PME activities) could be evidenced. In addition, no correlation with pectin level, assessed using the same procedure as for cassava, could be evidenced on a first test of 6 samples. Several correlations were however evidenced between RVA and biophysical analyses: pasting temperature measured with inhibitor can routinely be used to predict the true gelatinization temperature measured by DSC, and pasting viscosity with inhibitor was clearly correlated with starch content. However, the activity of beta-amylase did not appear correlated with RVA viscosities. There is clearly a need of more significant samples to test this proof of concept; cooking time and firmness of the studied samples were not evaluated with a suitable procedure; a new SOP for cooking and texture evaluation is now available and new samples characterized with this SOP should be used in year 4 for testing the proof of concept.

It is hypothesized that the level and/or structure (particularly the methylation level) of pectins play a major role on the cooking behaviour of RTBs. Hence the need for a medium throughput procedure to assess pectin level and/or structure. Different extraction procedures (water, addition of Ca++, from 30 to 70°C, from 15 to 120 min, pH 10, EDTA) of pectins have been tested, and released galacturonic acid assessed by colorimetric reaction with meta-hydroxydiphenyl (MHDP) on 6 cassava samples selected for their large range of cooking behaviour. Based on the results, non-methoxylated pectins had a great impact on cooking behaviour of the 6 cassava samples. A relatively rapid procedure of extraction and assessment of methoxylated (pH 10) and non-methoxylated (EDTA 0.05 M) pectins can be used to predict cooking behaviour. A procedure using a continuous flow analyser has been developed for rapid analysis of extracted galacturonic acid. The next step is to analyse a large number of cassava samples and to test the procedures on other RTBs of the project (yam, sweet potato, and banana).

It is hypothesized that the level and/or structure (particularly the methylation level) of pectins play a major role on the cooking behaviour of RTBs. Hence the need for a medium throughput procedure to assess pectin level and/or structure. Different extraction procedures (water, addition of Ca++, from 30 to 70°C, from 15 to 120 min, pH 10, EDTA) of pectins have been tested, and released galacturonic acid assessed by colorimetric reaction with meta-hydroxydiphenyl (MHDP) on 6 cassava samples selected for their large range of cooking behaviour. Based on the results, non-methoxylated pectins had a great impact on cooking behaviour of the 6 cassava samples. A relatively rapid procedure of extraction and assessment of methoxylated (pH 10) and non-methoxylated (EDTA 0.05 M) pectins can be used to predict cooking behaviour. A procedure using a continuous flow analyser has been developed for rapid analysis of extracted galacturonic acid. The next step is to analyse a large number of cassava samples and to test the procedures on other RTBs of the project (yam, sweet potato, and banana).

16 samples of sweet potatoes produced by CIP in Uganda were analysed by CIP in Uganda (cooking time, firmness, dry matter) then freeze-dried. Freeze-dried samples were than analysed by CIP-Peru (free sugars, starch, proteins by NIRS), JHI (beta-amylase activity, PME activity, cell wall characterization), CIAT (free sugars by HPLC, starch, amylose, RVA) and CIRAD (free sugars by HPLC, gelatinization and amylose by DSC and pectin contents by colorimetry). Unfortunately, no correlation between cooking time and firmness were observed and almost no significant correlation between cooking behaviour (cooking time or firmness) and the biophysical analyses (dry matter, starch characteristics, beta-amylase and PME activities) could be evidenced. In addition, no correlation with pectin level, assessed using the same procedure as for cassava, could be evidenced on a first test of 6 samples. Several correlations were however evidenced between RVA and biophysical analyses: pasting temperature measured with inhibitor can routinely be used to predict the true gelatinization temperature measured by DSC, and pasting viscosity with inhibitor was clearly correlated with starch content. However, the activity of beta-amylase did not appear correlated with RVA viscosities. There is clearly a need of more significant samples to test this proof of concept; cooking time and firmness of the studied samples were not evaluated with a suitable procedure; a new SOP for cooking and texture evaluation is now available and new samples characterized with this SOP should be used in year 4 for testing the proof of concept.

L'igame, Dioscorea sp., est un tubercule multi-espèces cultivé en Afrique, Asie, Amérique du Sud, Caraïbes et îles du Pacifique Sud. C'est un aliment de base en Afrique de l'Ouest où il est cultivé largement. Jusqu'à maintenant, le suivie de la répartition spatiale des composés majeurs de l'igname se fait par les méthodes physico-chimiques traditionnelles (Graham-Acquaah et al., 2014). Cependant, la caractérisation par ces dernières demandes beaucoup de temps et d'argent. C'est pour cette raison que nous avons décidé avec nos partenaires du Bénin de mettre en place une méthode d'analyse rapide et non destructive qui détecte la distribution spatiale des composés dans l'igname. Dans la présente étude, 4 variétés d'origine du Bénin ont été utilisées. Les tubercules ont été découpés longitudinalement en deux tranches, ensuite une seule tranche a été utilisée pour découpé 5 cubes de la zone proximale à la zone distale. Chaque cube a été découpé à son tour en tranches fines comme le montre le schéma ci-dessous. Enfin 3 spectres ont été acquis en mode réflectance sur chaque tranche de chaque cube à l'aide d'un spectromètre ASD LabSpec (350-2500 nm). Une fois les données spectrales collectées, une ACP a été appliquée afin de discriminer les différentes zones du tubercule. Ensuite, des modèles prédictifs ont été développés pour évaluer les performances de quantification de MS, polyphénols et sucres sur des tranches d'igname frais. Les résultats obtenus ont montré une différence significative entre les zones proximale, centrale et distale mais pas de différence significative entre les tanches radiales.

Les légumineuses, comme le Niébé, sont riches en α-galactosides (verbascose, stachyose et raffinose), ce qui crée des désordres intestinaux après consommation. Le but de ce travail est de réduire la quantité d'α-galactosides dans le Niébé cuit, en modélisant le transport et les réactions cinétiques de ces molécules durant le trempage et la cuisson. Ces derniers sont effectués à différentes températures (30°C, 60°C et 95°C) et a un ratio eau-graine de 4 :1 (poids/ poids). Un modèle prenant en compte le phénomène de diffusion moléculaire (de la graine vers l'eau de trempage), ainsi que la production-dégradation des alpha-galactosides (réaction supposée d'ordre 1), nous a permis de décrire et prédire les fractions des α-galactosides diffusées, dégradées et produites en fonction des conditions de trempage-cuisson. A 30°C et après un temps de 24h de trempage, la fraction diffusée représente seulement 1 a 5% du contenu initial dans la graine, alors que 30% du stachyose initial est dégradé. Un trempage de 4h a 60°C résulte en une importante fraction diffusée (jusqu'à 50-70%), mais aucune dégradation significative n'apparait. A une température de cuisson de 95°C, la diffusion est deux fois plus rapide qu'a 60°C (environ 60% de chaque α-galactosides diffuse après 90 minutes de temps de cuisson), sans dégradation. Ces résultats confirment la stabilité thermique des α-galactosides et suggèrent que la fraction dégradée, observée a 30°C peut être attribuée a l'action de l'α-galactosidase endogène. Le simulateur construit à partir de cette étude peut être utilise pour identifier des voies optimales pour minimiser le contenu en α -galactosides. Ainsi, une cuisson directe supprime la plupart des α-galactosides (par diffusion). Cependant, il serait intéressant de modéliser l'activité de l'enzyme endogène en fonction de la température, du pH etc. pour savoir si un trempage pourrait maximiser la dégradation des α-galactosides. L'approche pourrait être utilisée pour identifier les conditions de procédé optimales pour d'autres légumineuses riches en α-galactosides.

International audience; Age of plant and environmental conditions during growing may affect texture (friability) and taste of boiled cassava roots. We investigated the quality of boiled roots of seven cultivars harvested at 10, 12 and 14 months in three different seasons. Sensory taste (sweet or bitter) of boiled cassava root could not be correlated with sugar content and/or cyanide potential, which both interfere with taste perception; hence, bitterness is not a good indicator of the poisonous character of cassava roots. Improved cultivars generally showed lower friability scores, independently of plant age (10-14 months) or season. Rainfall before harvest directly lowers dry matter and mealiness of boiled roots. Cultivar and rainfall effects are discussed in relation to pectins (higher content for improved cultivars) that are suspected to be the major biochemical cause of vegetable mealiness.

Aromatic quality of rice grains is known to vary greatly with environmental factors and cultivation methods. Among the environmental factors, soil salinity is thought to have a positive impact on the content of 2-acetyl-1-pyrroline (2AP) in grains, the key volatile compound of rice aroma. This study compared 2AP content in grains of three improved fragrant rice ( Oryza sativa L.) varieties grown in two fields, differing mainly in their soil salinity level. The impact of salinity on yield and main yield components was also investigated to understand the relationship between aromatic quality and yield build-up. Soil salinity was monitored by measuring the electrical conductivity (EC) of soil solution samples extracted every week. 2AP content in grains was determined by a newly developed stable isotope dilution method involving solid-phase microextraction (SPME) and GC MS/MS analysis. The results showed an increase of 2AP content in grains with salinity for the three varieties. The relationship between 2AP and mean EC of the crop fitted a single model for the three varieties ( R 2 = 0.728). In contrast, the impact of salinity on yield and yield components differed greatly between the three varieties. One variety appeared to be very sensitive to salt stress, with significant yield loss up to 40%, while the two other varieties proved to be resistant to the salinity levels experienced by the plants, with no significant yield loss or even higher yields in saline conditions. Nevertheless, the three varieties presented a significant negative correlation between 1000 grain weight (TGW) and the mean EC of the crop, and between TGW and 2AP content. It was concluded that the increase of 2AP content with salinity resulted partially from a 2AP concentration mechanism in smaller size grains. The direct effect of salinity on 2AP synthesis through stimulation of the proline metabolism is further discussed.

Akpan is a traditional ready-to-drink fermented yoghurt-like cereal beverage consumed in urban and rural areas in Benin. With the aim of adapting the product to new local and export markets, this work maps African and European consumer preferences for different types of Akpan. RESULTS A sensory profile of Akpan was created and consumer tests were conducted with 103 consumers of African origin and 74 consumers of European origin. Consumer acceptance was significantly correlated with fermented odour (r = −0.94) and milky taste (r = 0.92–0.97) attributes. Cluster analysis revealed different behaviour by African and European consumers with respect to acceptability of Akpan; European consumers did not like the sour taste and African consumers liked an intense sweet milky taste. CONCLUSION This study provides information on how Akpan, and other fermented yoghurt-type cereal products, could be adapted to African and European consumer preferences. © 2014 Society of Chemical Industry.

Cooking bananas are staple food for million inhabitants worldwide. Starch is the main component of cooking bananas and plantains at green stage of maturity. the cooking process needed for the consumption of cooking banana is often applied in excess water, leading to an irreversible order-disorder transition of starch, namely gelatinization. Both water and thermal conditions applied are known to affect the functional quality of starch and of the final product, including its nutritional quality. Thus, various thermo-hydric conditions were applied to investigate the thermal transitions linked to starch gelatinization, with temperature and water content ranging 55-120°C and 1.4-2.0 kg.kg-1 of dry starch. A banana flour-water mixture was characterized using DSC while taking into account the effective glass transition temperature, and using a hermetically sealed test cell to get an equivalent extent of starch gelatinization. The coresponding digestibility was evaluated using an in vitro technique. A powerful predictive mathematical tool was developed for monitoring the starch digestibility during process while optimizing its nutritional quality. The degree of plantain starch gelatinization was revealed being mainly dependent on temperature in non-limiting water conditions (> 70% moisture content, wet weight basis). Different thermo-hydric conditions resulted in the same digestibility values, i.e. higher temperatures compensated lower moisture contents, and vice-versa, to reach similar digestibility properties.

In this deliverable the main results of sensory profile of reengineered products from group I are presented. Group I included Akpan and Gowé (Benin), Kenkey (Ghana) and Kishk Sa'eedi (Egypt). In this stage all products were submitted to reengineering process to overcome food safety and product quality issues (identified in the first stage of the project in traditional food) to facilitate their promotion and introduction to EU markets. Although all products were tested in Europe for consumer acceptance, in the case of SENSORY, for Kenkey and Akpan in Europe, the results although not obtained by a trained panel, were generated by Consumers through CATA test. The use of check-all-that-apply (CATA) questions for sensory product characterizations with consumers has been gaining acceptance by several authors (G. Ares and S.R. Jaeger, 2015) and for that we considered an equivalent test to replace the classical descriptive analysis by the trained panel that would take longer time. Due to the restrictions of delayed reengineering processes, we did not have an enough frame time to train adequately the panel. However, the characterisation of product attributes was also achieved. In the case of Gowe, due to restrictions in the productions of reengineered product due to contaminations of raw material, these results for sensory profile could not be obtained, but this was the only case in group I. However, the differences in main attributes were not so far from the traditional ones. In sensory studies of reengineered products, we´ve tried to integrate quantitative characterization of the perceivable product attributes and in consumer studies hedonic profiles (Worch, Lê, and Punter, 2010). So, in the sensory profiling of a specific group of reengineered products, once performed by a trained panel for a specific group of attributes by a specific African or European country panel, it must be considered unique and reproducible for all other panels that can test the product, afterwards. So in our sensory study of Kishk Sa'eedi as it was performed in Africa or Europe by a trained panel performing quantitative and reproducible characterization of the perceivable (Kishk Sa'eedi) attributes, herein will be used the same results for European sensory description of the product. Akpan: In terms of consumer perception of sensory attributes performed in Europe for the three Akpan products did not significantly. If we remove the terms such as “artificial”, “strong in taste”, “floral” due to a manufacturing error (use of a few drops of citronella essential oil instead of citronella infusion as a traditional flavouring of Akpan in Benin), it remains the terms “mealy”, “liquid” “drinking yoghurt” that better describe the product and were previously used for describing traditional Akpan product. This means that sensory properties of traditional Akpan were not totally improved during re-engineering to meet French consumer taste. Kenkey: The results of this study based on CATA testing allowed to provide a basis of understanding on how the reengineered Kenkey (RK) compares to standard commercial Kenkey (SK) in terms of its sensory properties. An interesting finding was that the sensory properties of RK were very different from that of the standard Ga Kenkey (SK): consumers scored differently the new product from the standard one. RK was globally perceived as too weak whilst SK was too strong. RK was perceived as bland and crumbly and SK as sour, salty and fermented. Kishk Sa'eedi (KS) was evaluated by a trained panel by QDA and analysed using QFD and PCA analysis. The results of KS sensory characteristics provide in depth understanding of the sensory quality criteria as perceived by the sensory trained panel. The present study showed that substantial differences in sensory character were noted between the traditional and re-engineered KS in particular, differences in colour, fresh odor, KS taste, fracutability and mouth coating. This work showed that the application of techniques could provide the useful information to KS and helped to identify the importance of product attributes. So in general, it can be observed that sensory properties of reengineered products of Akpan and Kenkey did not differed significantly from traditional ones, however the opposite was observed for the new KS compared to the traditional one.

This deliverable concerns the sensory evaluation of the reengineered group 1 African products in the AFTER project. Specifically, it related to reengineered akpan and gowe from Benin, kenkey from Ghana and Kishk Sa'eedi in Egypt. Concerning reengineered akpan from Benin, the sensory evaluation was undertaken in Montpellier, France. Re-engineering of akpan has focused primarily on improvement of sanitary properties of the product, which was a great achievement and will allow producing Akpan on a larger scale in SMEs in Africa. Sensory evaluation of the Akpan products was carried out using CATA and JAR techniques that have been developed for use with consumers instead of a trained panel. Three Akpan products were tested by 102 consumers: Akpan added with 10% sugar (AS10), Akpan added with 3% spray-dried milk and 8.7% sugar (AMS8.7) and Akpan added with 3% spray-dried milk and 15% sugar (AMS15). Independently of the Akpan tasted, Acidity or Sweetness attributes were scored “Just About Right, as I like” by 56 to 77% of consumers. Odour perception was perceived differently, depending on consumers. However, Texture was found “Too weak”, too liquid by the majority of consumers (49 to 55%) and Taste “too strong” (46 to 54%). The most frequently CATA descriptors checked by consumers which better described Akpan products were: “Artificial”, “Floral”, “New/Different”, “Strong in Taste”, “Mealy”, followed by “Liquid”, “Drinking yoghurt”, “Sweet”, “Acidic”, and “Rough”. At the opposite, an ideal-yoghurt was described as Creamy, Natural, Good for health, Refreshing, Homogeneous, with a texture of a Bulgarian yoghurt-type, Thick, Sweet, Attractive, Nutritious and Milk taste. In terms of sensory evaluation, the three Akpan products did not significantly. If we remove the terms such as “artificial”, “strong in taste”, “floral” due to a manufacturing error (use of a few drops of citronella essential oil instead of citronella infusion as a traditional flavouring of Akpan in Benin), it remains the terms “mealy”, “liquid” “drinking yoghurt” that better describe the product and were previously used for describing traditional Akpan product. This suggests that sensory properties of the reengineered Akpan may not be acceptable to French consumers who prefer a product with a creamy, homogeneous, Bulgarian yoghurt-type taste. Gowe in Benin was not tested using sensory evaluation. Sensory testing of Gowe in Benin was not undertaken because this was planned to be undertaken in Europe. The reason is because the methodology used in sensory evaluation is independent of the location provided the samples are the same. However, the particular samples provided for French sensory testing contained a concentration of aflatoxin that was higher than the minimum EU allowable limit. It was not possible to repeat the sensory test in France because it would have taken too long to obtain a replacement supply from Benin and to repeat the processing (takes one week). In which case the samples would have been took different to enable a comparison. The sensory evaluation of kenkey was carried at the Food Research Institute, Ghana. Current trends in urbanization, and the increasing popularity of kenkey among consumers, require larger scale production with consistent quality. Testing was conducted to determine the sensory profile of white reengineered kenkey made using the optimum pre-process conditions of steeping time (30 and 45h), steeping temperature (30ᵒC and 35ᵒC) and dough fermentation time of 12 hours. The qualitative descriptive analysis showed that the sensory profile of white kenkey was dependent on preprocessing variables. Thus merely optimizing the pre-processing variables with regards to acid production and other readily measurable constituents though could shorten the production process could not guarantee the best product sensory quality. The results show that all the descriptors generated were appropriate for differentiating sensory qualities among samples and could be used for basic research and product development for white kenkey. Soft and sticky texture in white kenkey was highly appreciated. Sensory evaluation of Kishk Sa'eedi (KS) was undertaken in Egypt. KS is an Egyptian indigenous wheat-based fermented food prepared traditionally according to the method applied by Upper Egyptians. This work is done to characterize sensory properties and sensory profile of the reengineered KS. Quantitative descriptive analysis (QDA) coupled with principal component analysis (PCA) was used to study the interrelationship among and between sensory attributes. 14 terms regarding appearance, odour, flavour and texture of the samples, was selected and a glossary describing each descriptor was developed. Three KS samples were profiled by 11 assessors using the chosen 14 sensory descriptors. Mean intensity ratings of the descriptive attributes showed that there were significant differences (p

In this deliverable, the reengineered products relating to groups 1 to 3 are tested by consumers in Africa in order to test whether the acceptance meets consumer needs. The summaries are by group. These products were widely different and tested in differing cultures and situations. A common feature, however, is that all products are already commonly consumed by the communities and accepted, particularly with respect to the baseline products. Some products were found to be more acceptable (Gowe and Akpan in Benin, Kenkey in Ghana, Kong and Lanhouin in Senegal, Bissap and Baobab in Senegal), some of similar acceptance (Kitoza in Madagascar) and some were less acceptable (Kishk Sa'eedi in Egypt and Jaabi in Cameroon). Ways forward are discussed in all cases. Group 1: The reengineered group 1 products had mixed acceptance to consumers. Ones that were highly accepted were the reengineered Akpan which was highly accepted regardless of age, gender, socio- professional category and possessions (goods) and the majority (84.9%) expressed an interest for purchasing the reengineered Akpan and the reengineered gowe (n=100 consumers) in Benin, which was strongly accepted and more so than the traditional gowe. However, the reengineered Kenkey in Ghana while being comparable acceptance to the traditional white kenkey this would only meet a niche market. However, for the reengineered Kishk Sa'eedi, it was less acceptable and the difference was due to sour taste and spicy flavour. Group 2:The group 2 products had mixed acceptance among consumers. In the case of Kong in and Lanhouin in Senegal, the reengineered products had improved acceptance while for kitoza in Madagascar, there was no improvement in acceptance apart from those in terms of quality and safety. More specifically, Kong made using liquid smoke novel technique (RLK) had by far the highest acceptance and could easily be integrated as part of the Senegalese cuisine. This shows that there is a potential for RLK on the Senegalese market. Two types of reengineered Lanhouin types (Lanhouin dry powder, Lanhouin dry fillet) were greatly acceptable and the powdered form more than the fillet. In the case of Kitoza, consumer acceptance studies showed that there was no significant difference between all samples tested. Group 3: Group 3 reengineered products had mixed acceptance among consumers. Re-engineered bissap (hibiscus drinks) in Senegal were positively appreciated by consumers in Senegal and better accepted than the traditional commercial infusion used as baseline comparison. However, the new drinks were not perceivably different from the baseline drink to consumers concerning both the traditional character and the nutritional and healthy attributes. The reengineered baobab drinks were positively appreciated by consumers in Senegal and significantly better appreciated than the traditional commercial ones earlier tested. Jaabi (Ziziphus mauritiana). The result is the proposition of a reengineered Yaabande presenting fine texture well appreciated by consumers. However, particle size was critical and this needed to be lower than 125μm. Due to this limit, processors have adopted the manual production of flour of particle size 250μm>Ø>125μm, which is less tiresome, though it appears in second rank in the preference of consumers. The detailed methodology for each product is given for Groups I, II and III respectively. Common to all of the methodologies is the Ethical assessment and consent which is listed as follows: Ethical assessment and consent. The studies have been assessed and approved by the University of Greenwich Research Ethics Committee and the Ethics Committee at CIRAD. Consent was sought from sensory panellists and from adult consumers participating in this study. Enumerators informed participants about the study and explained that their participation was entirely voluntary, that they could stop the interview at any point and that the responses would be anonymous.

Head yield, structure, mechanical properties and physico-chemical of rice. 4. International Rice Congress

Objectif : Evaluer trois équipements pour la fermentation de la pâte destinée à préparer l'ablo (pâte cuite à la vapeur) puis, l'influence de la température, de la durée de fermentation et du taux de levure ajoutée à la pâte. Méthodologie et résultats : La pâte semi-solide a été fermentée durant quatre heures à température ambiante (26-27 °C) dans du bac en verre, glacière et bassine en aluminium (Témoin). En outre, l'effet de la température, de la durée de fermentation et du taux de levure a été évalué à travers un dispositif factoriel complet dont les niveaux bas et hauts des facteurs rangeaient entre 2 et 6 heures, 22 et 30 °C et 1 et 3 g/Kg de farine. La pousse, le pH, l'acidité titrable et la charge en levures/moisissures ont été déterminés. A l'exception de la quantité de mousse produite et de la matière sèche, aucun effet significatif (P>0,05) du matériel de fermentation n'a été mis en évidence sur les paramètres mesurés. Le taux de levure, la température et la durée de fermentation avaient une influence positive significative (P

The Gowe is a Beninese traditional cooked paste pre pared from sorghum. It is sold as is and then diluted and consumed in as a sweet, slightly acidic drink. The traditional manufacture of Gowe includes the unit operations of malting, fermentation and cooking. A Gowé of high quality relies on a sufficient acidification during fermentation and th e hydrolysis of starch by the malt amylases which produces fermentable substrates and imparts a sweet taste. The kinetic of lactic fermentation thus depends on the malt amylasic activities that are in return inhibited by the acidification due to the production of lactic acid. In addition, the Gowe sh all present a good sanitary quality that will be promoted by a fast and sufficient acidification. Th e objective of this study is to propose a global model combining predictive microbiology and predictive biochemistry, for predicting and optimizing the preparation of a Gowe with high organoleptic an d sanitary qualities. Two strains of Lactobacilli ( Lb casei and Lb brevis ), known to have different potentialities for acidification were grown on liquid medium MRS. A pathogenic strain of Bacillus cereus was grown on nutrient broth to test the resistance to contamination of the gowe and thus ensure the sanitary quality. We modelled the growth of the strains at constant pH according to a logistic model that determines the growth rate and the duration of the lag phase. Secondly, a cardinal model was used to model the effect of temperature and pH on the gr owth. The kinetics of mortality of B. cereus with pH has also been modelled using reaction of order 1 . We then determined the speed of production of lactic acid, which increases linearly with the r ate of growth of lactic acid bacteria, and modeled the effect of lactic acid on the pH of the medium. The variation of α-amylase activity with temperature and pH was also modeled with empirical laws and the rate of the hydrolysis of starch and of the production of free sugars (mainly glucose and maltose) was modelled by enzymic kinetic with competitive inhibition by the products of reactions (sugar free). A global fermentation simulator was then built. The comparison of measured and predicted values showed that the model gave a fairly good prediction of the microbial populations (Lactobacilli and B. cereus ), but also the acidity and the rate of free sugars . This tool thus allowed proposing scenarios of fermentation to optimize both the organoleptic and sanitary qualities of gowe. Several optimized scenarios could be offered to economic operators, depending on their technological level.

Le gowé est un breuvage traditionnel du Bénin obtenu à partir de la farine de sorgho ou de maïs malté et fermenté. Il est produit à petite échelle et présenté à la vente sous forme d'une pâte cuite, molle, légèrement sucrée emballée dans des feuillesde Thalia welwichii ou Tectona grandis. La pâte est additionnée,au moment de la consommation, d'eau, de sucre, de lait et des glaçons à volonté. Le maltage et la fermentation, opérations clés du procédé de production confèrent au produit des qualités sanitaires, nutritionnelles et sensorielles variables parce qu'elles ne sont pas maîtrisées. La forme de présentation et la teneur en eau du produit final traditionnel ne permettent pas une conservation de longue durée et la conquête de marchés urbains. La valorisation du gowé exige la standardisation du procédé de production accompagné d'une meilleure présentation du produit final. Pour ce faire, le maltage a été optimisé avec le traitement des grains de sorgho par une solution de NaCl, l'optimisation de la durée et de la température de trempage, de la germination et du séchage. L'optimisation de la fermentation a été réalisée par une saccharification contrôlée de la farine maltée etl'utilisation de souches lyophilisées de Lactobacillus casei et Kluveromyces thermotolerans. Une meilleure présentation du produit final a été développée.Le lavage des grains de sorgho à l'eau salée a réduit la charge de moisissures au cours de la germinationde 5,7 log UFC/g à 4,1 log UFC/g. Le maltage suivant les conditions optimales a permis d'augmenter l'activité α-amylasique de 14,8 UC/g à 47 UC/g. La durée totale du maltage a été réduite de 6 jours à 4 jours avec le procédé amélioré. La saccharification du malt a permis d'augmenter la teneur en sucres (maltose, glucose et fructose) de 12,8% à 26,3%. Avec l'enrichissement parles souches de bactérie lactique et levure, la durée de fermentation est réduite de 24 à 12h avec une amélioration de la qualité sanitaire et sensorielle du gowé (goût sucré et acide). Les conditions améliorées de maltage, de saccharification et de fermentation permettent de réduire la durée de production de gowé de 8 à 5 jours. Une meilleure présentation du produit a été mise au point par le séchage de la pâte fermentée de gowé prête à cuire. Le gowé amélioré est sous la forme d'une farine séchée mise en sachet ayant une longue durée de conservation. La farine de gowé est aisément reconstituable en breuvage répondant à la qualité désirée par les consommateurs. (Texte intégral)

Les populations côtières Ouest – Africaines valorisent le maïs sous la forme d'un pain cuit à la vapeur appelé Aboloo au Ghana ou Ablo au Bénin. Avec l'urbanisation, l'enracinement du riz dans les habitudes culinaires de ces populations a induit une généralisation de la production de ce pain à partir du riz. Dans une perspective d'industrialisation et de valorisation des productions locales, la présente étude a déterminé le profil sensoriel du ablo de riz obtenu à partir de cinq procédés de fabrication différents. Les descripteurs sensoriels du Ablo ont été générés au cours de 16 discussions de groupe menées avec des panels de 3-5 consommateurs habituels de Ablo et quantifiés par un panel semi-entrainés de 22 dégustateurs en trois répétitions. Les discussions de groupe font ressortir que le Ablo peut être décrit principalement par ses attributs de texture (levé, consistant, collant, ferme, élastique, friable, tendre, surface lisse) et de goût (sucré, salé, acide), mais aussi par des attributs secondaires liés à l'apparence (blanc, jaunâtre, luisant, couleur uniforme, surface poreuse) et à l'odeur (odeur de fermentée, odeur de riz). L'analyse quantitative descriptive montre que le procédé caractérisé par " une non utilisation du blé et une fermentation longue d'environ 21 heures" donne un Ablo dont l'attribut " levé " est significativement plus faible que celui de l'Ablo issu des autres procédés. Le procédé caractérisé par " la précuisson hors feu " est évalué avec une couleur " blanche " plus faible que celle de l'Ablo issu des autres procédés. En dehors de ces différences, les Ablo issus des cinq procédés ont un profil sensoriel identique pour leurs attributs " jaunâtre ", " collant ", " consistant ", " tendre ", " friable ", " élastique ", " odeur de fermenté ", " acide " et " sucré ". Un tel profil sensoriel permet de définir le référentiel descriptif du Ablo de riz tel que consommé en milieu urbain au Bénin. (Texte intégral)

Concerning re-engineering, based on deliverables from WP1 (D1.1.2.1-survey results, D1.2.4.1-results for sensory, physical and textural analysis) and WP5 (D5.1.2.1- marketing and regulatory opportunities, D 5.3.2.1 – sensory and consumer tests), a first diagnostic of product opportunities, processing problems and re-engineering options has been obtained for each product (gowé, akpan, kenkey and Kishk Sa'eedi “KS”). The main objectives and actions that will be followed for re-engineering are listed in Annexes. First results concerning re-engineering are then presented, followed by a plan of work for each product The main conclusions for reengineering options are as follows: (a) re-engineering will with in particular deal improving shelf life and sanitary level of the traditional products. This will go through controlling fermentation by using in particular specific functional inoculums combined with re-engineering of the other unit operations (malting, milling pre-cooking, etc) that interfere with fermentation control. (b) Several options of re-engineering will be targeted: improvement of the “homemade” processes and traditional products, involving low capital investment processes and few modifications of the form of the original products that are targeted to the local market, and deeper process modifications and new formulations and ways of presentations of the products for larger processors and national and international markets; this includes in particular ready to prepare dried Akpan, Gowé and kenkey and new ready to drink/eat products (for all products).

The present study aims to propose a comprehensive appropriate technological package for small-scale processing of maize flour, which meets consumer demand for quality and achieves extended shelf life. One traditional soft cultivar was processed by the conventional method (direct grinding with disc mill) and was compared with flours obtained under various degerming and grinding intensities; two types of grinder (disc mill and hammer mill, already present on the market) were tested in parallel. Consumer satisfaction scores were registered for the different products after increasing storage duration (0 to 6 months). It was shown that combining degerming and hammer milling produce high quality flour from hard grains, which can be stored up to 6 months without significant deterioration. In parallel, physicochemical and rheological characterizations showed that the main reaction occurring during storage is lipid degradation, which slightly increased fat acidity, hence decreasing taste and elasticity acceptability of the paste.

International audience; Rice from two scented (Aychade, Fidji) and one nonscented (Ruille) cultivars from Camargue area (France) together with one Asian scented rice (Thai) were cooked and volatile compounds were extracted by dichloromethane/pentane (1;2, v/v) solvent mixture. 40 odorous compounds were noticed during GC-O analysis of the organic extracts, amongst which oct-1-en-3-one and 2-acetyl-1-pyrroline were altmost constantly perceived. Hierarchical cluster analysis (HCA) of perceived odours enabled to distinguish the groups of rice cultivars. 60 compounds were identified and quantified by GC-MS. New odour-active components of cooked rice were detected for the first time including oct-3-en-2-one, 2-methylpropanoic acid, y-decalactone and o-decalactone. A principal component analysis (PCA) differentiated scented cultivars from a non-scented one and scented rice cultivars from Camargue from Thai sample.

International audience; Efforts have been undertaken for the last decades to promote the production of scented rice cultivars in Camargue, a traditional rice cultivation area in the south of France. Volatile compounds of cooked rice from two scented Camargue rice cultivars (Aychade and Fidji) were analysed and compared to that of a marketed Asian scented rice one (Thai)(1). Additionally volatile analysis from a cooked rice of a non-scented cultivar (Ruille) grown in Camargue was performed. (...)

A group of 118 representative accessions (based on dry matter content, DMC; cyanogenic compounds, CC; amylose % and RVA functional properties) from CIAT¿s germplasm cassava collection was selected for this study. Average DMC and CC were 32% and 696 ppm, respectively. Average contents (dry basis) were: starch, 84%; fiber, 3.5%; sacarose, 3.1%; glucose and fructose 1% each, and nitrogen (N) 0.64%. Gelatinization enthalpy was 15J/g; and gelatinization temperature (TG) was 61.3°C. Average amylose content (quantified through DSC) was 20%. Swelling power, solubility and Ö evaluated through RVA at 75°C were 30.0g/g, 10.6%, and 0.7, respectively. Root DMC was correlated with starch, fiber and sugars contents. TG was strongly associated with pasting temperature (Tpasting) of the gels obtained from the RVA. Percentage of amylose correlated well with gel clarity. CC correlated with swelling power. This study allows a better understanding of the preference by processors and consumer of cassava products from "bitter" germplasm. This type of cassava is used in traditional processes in the production of healthy foods which meet the required physicochemical and sensorial characteristics.

The evolution of cyanide content and colour were monitored during the processing of lafu, traditional flour and improved flour from five cassava cultivars from Benin. In addition, the total phenol, polyphenoloxidase (PPO), peroxydase (POD) and linamarase activities were assessed. The processing of cassava in lafu and improved flour proved superior for producing safe and white non-fermented and fermented cassava flours with total cyanide mean values of 16.6 and 11.4 mg HCN/kg, db and DE values of 9.2 and 12.1, respectively. Detoxification appeared to be only linked to processing, in particular to the size reduction level of cassava roots, regardless of the initial cyanide level and the linamarase activity of the fresh roots. Cassava flour yellowness was closely linked to the phenol content (r = 0.95) that decreased after steeping and pressing. The PPO and POD activities did not appear to be linked to flour discoloration.