Search

The efficient global optimization method is popular for the global optimization of computer-intensive black-box functions. Extensions exist, either for the optimization of noisy functions, or for the conditional optimization of deterministic functions, i.e. the search for the values of a subset of parameters that optimize the function conditionally to the values taken by another subset, which are fixed. A metaphor for conditional optimization is the search for a crest line. No method has yet been developed for the conditional optimization of noisy functions: this is what we propose in this article. Testing this new method on test functions showed that, in the case of a high level of noise on the function, the PEQI criterion that we propose is better than the PEI criterion usually implemented in such a situation.

International audience; Rainfed crop production predominates in West Africa. Rice is an important staple food, especially in Senegal. The scope for increase in rice production under irrigated conditions is uncertain. Rainfed rice is therefore a key component for regional food security impelling agronomists to assess climate change impact on rainfed rice yield and to design rainfed rice ideotypes suited to future climate scenarios. The DSSAT CSM-CERES-Rice model was thus calibrated and evaluated on 19 agronomic experiments conducted in 2012, 2013, and 2014, in 6 locations, with 21 cultivars and two fertilization levels (20 and 80 kg N ha−1). Simulations were then carried out with the crop model forced with the downscaled projections of seven climate models, with and without considering the impact of an increase in atmospheric [CO2], using an ensemble of global circulation models and two Representative Concentration Pathways (RCP2.6 and RCP8.5). Simulated rice yield was divided by two over the century with RCP8.5 and stagnated with RCP2.6. Elevated [CO2] significantly increased yields, but this effect could not offset the yield decline due to elevated temperatures. Cultivars with longer vegetative phases and greater temperature tolerance were better adapted to climate change than current cultivars. Using these new cultivars with the recommended fertilization rate (80 kg N ha−1) could offset the yield decline due to climate change. For the first time, we bring together a study based on a process-based crop model handling crop response to elevated [CO2], a large set of field experiments and up-to-date climate projections (i) to provide useful insights into plausible impacts of climate change on rainfed rice in Senegal and (ii) to identify cultivar characteristics relevant for adaptation to future possible climates. Our findings will help set priorities for breeding resilient cultivar in the region.

In the central highlands of Madagascar, blast, a serious fungal disease of rice caused by Magnaporthe oryzae, was shown to be less severe in conservation agriculture (CA) than in conventional tillage (CT) cropping systems. To assess the effects of CA cropping systems on rice susceptibility to blast and to understand the mechanisms involved, an experiment was conducted over three growing seasons in two sites (at high and mid-altitude). Two fertilization treatments, one with mineral nitrogen (N) fertilizer, and one without, were studied in interaction with two different cropping systems (CA/CT) and two rice varieties that differ in their susceptibility to blast. The performances of the two cropping systems were assessed, and results showed that yields were no higher in the CA or were lower than in CT cropping systems. The effects of the treatments on upland rice crop growth were measured by leaf N content as an indicator of N crop status, and leaf area index (LAI) as an indicator of canopy density. In all years at both sites, leaf N content during early growth stages was higher in the CT than in the CA cropping system. The LAI was lower in the CA than in the CT cropping system at the high altitude site in all three years, and in two out of three seasons at the middle altitude site. Overall, when the level of blast was high, blast was less severe in the CA cropping system and with no N fertilizer. Leaf N content and LAI were used as intermediate explanatory variables in the statistical analyses, and one or the other or both always masked the effect of the cropping system or of N fertilization. Our results show that leaf N content during early growth stages masks the effect of the cropping system on leaf blast severity. When the LAI is strongly affected by the cropping system, it also masks the effect of the cropping system on the severity of leaf and panicle blast. Very often both variables masked the effects of the cropping system and of N fertilization. These findings should help improve disease management by modifying fertilization practices and crop nutrition in conservation agriculture cropping systems.

When classing cotton fiber quality, the results are governed by three sets of conditions: 1) production conditions, which may potentially be heterogeneous (variety, soil, climate, etc.); 2) ginning conditions at each ginning step, which have a positive or negative impact on final fiber quality. 3) testing conditions in the testing and classing laboratory. Our objective was to develop a way of measuring the specific impact on fiber quality of ginning operations alone, which therefore cannot be deduced easily from classing results. A sampling and testing protocol was tested in Mali in four industrial ginning mills while processing seed-cotton modules over a cotton cropping season. During the ginning of each module, samples were taken at four stages: seed-cotton samples before and after seed- cotton cleaning, and fiber samples before and after lint-cleaning. To obtain a reference ginning, a micro-gin was used to gently process seed-cotton samples. The resulting fiber samples, as well as the corresponding fiber samples collected downstream in the industrial ginning process, were characterized within a single randomized design in two blocks including controls every twenty samples. The fiber quality results thus made available at four stages of the ginning process were used to estimate the impact of each step, as well as the overall impact of ginning on the fiber quality characterization results. When significant overall ginning impacts were detected, the proposed protocol helped in finding the ginning step that lay behind fiber quality degradation, independently from the production and characterization sources of variations in classing results. Some technical and technological developments are still necessary for applying this protocol.

The quality of cotton fiber observed at classing depends on the production conditions (soil, variety, climate, and crop management practices, including harvesting technique) and on the ginning conditions (seed-cotton cleaning, ginning, and lint cleaning). Thus, when lint quality is monitored, variations in seed-cotton quality may be confused with the impact of variations in ginning conditions. So, rather than measuring the quality of the cotton leaving the ginning mill, we propose measuring the impact of ginning conditions on the difference in quality between a batch of cotton ginned in the mill and a sample of the same batch ginned using micro-ginning equipment. By comparing 11 industrial ginning mills with the micro-ginning equipment used as a control, this feasibility study evaluated the impact of ginning conditions in an experiment on 119 seed-cotton modules in Mali throughout the 2014–2015 season. Based on this comparison, data analysis showed that the ginning process did have an impact on cotton quality, that the impact varied with the date, and that it was only linked to ginning conditions. Conversely, no alert concerning the impact of industrial ginning conditions was issued based on the results of the evaluation of the quality of fiber samples from industrial ginning mills alone.

Improving sucrose yield is one of the main objectives of sugarcane breeding. Splitting this complex trait into yield components should make this task easier, as each component may be influenced in its own way by environmental factors and by genetic background. Abiotic conditions experienced by sugarcane across its cropping areas differ in many respects; among them, water availability and photo-thermal conditions particularly affect sucrose yield formation. In this study, sucrose yield was divided into seven component traits and studied in a panel of 155 sugarcane accessions phenotyped at two sites under contrasting photo-thermal conditions: one in low altitude and the other in higher altitude. The accessions were hybrids developed during the last century and representing the worldwide cultivated genetic diversity. The proportion of Saccharum spontaneum genome in the genome of each accession was estimated by analyzing the genetic structure of the panel associated with two outgroups formed by 19 S. spontaneum and 29 S. officinarum accessions genotyped with 419 DArT markers and using a Bayesian clustering method implemented in STRUCTURE software. A K = 2 number of clusters clearly separated S. spontaneum from S. officinarum , while the estimated proportions of the S. spontaneum genome in the genome of hybrid accessions ranged from 0.5 to 0. Multivariate mixed model of log transformed yield components was adjusted to estimate each component's contribution to sucrose yield genetic variance, taking into account interrelationships among components. Each component's contribution to sucrose yield variance was site-dependent. On the low altitude site with high photo-thermal conditions, stalk section was the main contributor to yield variance, while on the high altitude site with low photo-thermal conditions, stalk height was the main contributor. A linear regression showed that the estimated proportion of S. spontaneum genome in the hybrids’ genome had significant effects on sucrose yield and its components. These effects also varied with the site: under low altitude conditions, the estimated proportion of S. spontaneum genome in the hybrid's genome exerted a significant negative effect on sucrose yield, whereas no significant effect was found under high altitude conditions. These results suggest that both efforts toward introgression and selection on yield components for sugarcane breeding purposes should depend on the targeted cropping environment.

SUMMARYSeed cotton yield in Northern Cameroon has been declining since the 80s despite breeding efforts. In order to evaluate the impact of genetic improvement on this decline, we conducted field experiments in two locations with 10 widely grown cotton cultivars released in Cameroon between 1950 and 2009. The rate of genetic gain (GG) was estimated with a linear regression of the cultivar mean on its year of release (YR). Contrasts between rates of GG observed with different planting dates were estimated and tested. Our results revealed a rate of GG on fibre yield of 3.3 kg ha−1 year−1 due to increased ginning out-turn (3.9% and 6.2% in 60 years in Garoua and Maroua, respectively). There was no GG on leaf area index (LAI), radiation use efficiency (RUE), aerial biomass, harvest index and on seed cotton yield. We concluded that cotton breeding efforts in Cameroon have successfully improved cotton fibre yield but there is still some room for seed cotton yield improvement.

In Africa, it is mostly the informal seed system that ensures farmers’ seed supply. This is partly because the formal seed systems are not always effective in meeting demand for new seed varieties. Sometimes informal seed recycling and exchange of improved sorghum varieties will take place alongside formal initiatives, as is the case in southern Mali. Focusing on one particular village in the Dioila district, we analyze the efficacy of farmers’ strategies for preserving varietal seed purity and genetic integrity of an improved inbred-line (Soumba variety). Six seed lots of Soumba, recycled for two to six years by farmers using different practices, were collected and assessed in on-station trials in order to compare their agronomic performance and phenotypic purity (off-type plant frequencies) with control versions of the variety. Additionally, 30 panicle samples were randomly collected from five farmer fields sown with recycled Soumba and assessed for phenotypic purity in a progeny nursery and investigated for molecular diversity using 12 SSR markers. A total of 150 panicles from five other non-Soumba varieties were collected in the village in order to investigate eventual gene flow and its potential genetic consequences for the Soumba variety. In fields sown with recycled Soumba seed, between 2% and 14% of plants showed phenotypic deviations from the typical Soumba variety. The progeny nursery and SSR marker analysis verified the presence of the off-type plants observed in the field. The STRUCTURE program revealed admixtures with other varieties in 23% of Soumba plants, confirming the presence of gene flow. Gene diversity values in Soumba samples ranged from 0.006 for the commercial sample to 0.257 for recycled samples. Introgression and contamination were best minimized when (1) farmers had received specific training in seed production, (2) they could take advantage of isolated fields and (3) they could practise true-to-type panicle selection. Farmers were generally able to maintain the phenotype, as well as sustain or even improve yield performance of their Soumba variety while at the same time genetically enriching their seed stock.

SUMMARYSorghum (Sorghum bicolor(L.) Moench) is a major staple crop of Burkina Faso where farmers continue to cultivate photoperiod-sensitive guinea landraces as part of the strategy to minimize risk and ensure yield stability. In the Boucle du Mouhoun region, however, sorghum farmers appear to have insufficient varietal choice due to cropping systems having shifted towards more intensive cultivation of cotton and maize, and rainfall patterns having decreased over the past decade. In search for new varietal options that can respond to this changing context, researchers decided to give farmers access to ex-situ national collections along with the opportunity to evaluate recent improved varieties. From 2002 to 2007, researchers and farmers worked closely together to implement on-farm testing, including varietal selection trials, crop management and multi-locational trials. Farmers’ choices tend to differ among groups, villages and years, with the exception of four particular landraces: two originating from a collection carried out in the Mouhoun region more than 30 years previous to this research, and two other landraces that came from the dissimilar agro-ecological zones of Burkina Faso. These four were the most commonly selected landraces out of 36 cultivars that covered both improved and landrace varieties. Farmers’ selection criteria were focused on adaptation to agro-climatic conditions as well as specific grain qualities for processing and consumption. The potential usefulness of each variety was verified via multi-locational trials. The paper also shows that wide dissemination of experimental seed, not just across the Mouhoun region but also at a national scale, was largely achieved through collaboration with a strong farmer organisation in conjunction with farmer training programs focused on the on-farm seed production and the commercialisation of this seed.

The possible application of conclusions from a published study concerning cottons from West and Central Africa (WCA), involving an evaluation of the within-bale variability of fiber Micronaire, Length, Uniformity, Strength, Reflectance and Yellowness in cottons from Eastern and Southern Africa (ESA) was investigated. We took eight cotton samples per bale from 240 bales produced by 32 ginning mills in six ESA countries in two crop seasons. Our representative sample comprised 1920 fiber samples that were centrally analyzed under controlled conditions using standardized instruments for testing cotton (SITC). We evaluated within-bale variability levels for both saw- and roller-ginned cottons and checked the applicability of the published conclusions to ESA. We found that (1) sampling variance levels were comparable in ESA and in WCA for saw-ginned cottons, (2) WCA recommendations for saw-ginned cottons would also apply in ESA for most fiber characteristics measured by SITC, and (3) for roller-ginned cottons, the higher within-bale variability of roller-ginned cotton fibers compared to saw-ginned cotton would require the definition of a specific sampling and testing method based on an experiment to be designed.

SUMMARYIn most cotton-growing regions of West and Central Africa where rainfed conditions prevail, cropping conditions are highly diversified since the crop is planted over an extended period. We studied production and development patterns in 10 cotton (Gossypium hirsutum) varieties to determine the most efficient strategies that could be transformed into breeding traits. Four trials were carried out between 2002 and 2003 in two cotton-growing areas in Benin to compare the 10 varieties at three stand densities and two planting dates. The parameters monitored were the mean first flower opening date (FF), effective flowering time (EFT), plant height at harvest (HH), height to node ratio (HNR), length of fruiting branch (LFB), number of vegetative branches (NVB) and average boll retention at the first position of the fruiting branches (RP1). We identified two ideotypes that yielded better than the others: (i) Mar 88-214 performed well under late planting–high stand density conditions and was characterized by low vegetative growth and early flowering onset, a short flowering period and low RP1; (ii) H 279-1 performed especially well under early planting–low stand density conditions and was characterized by high vegetative growth, late flowering, long EFT and high RP1. We propose a breeding strategy for both ideotypes based on seven indicators with high heritability (FF, HH, HNR, and LFB) or medium heritability (NVB, EFT, and RP1).

In cotton (Gossypium hirsutum), it is hard to determine the exact date when reproductive growth ceases on the basis of field observations, as compared to more visible factors such as the onset of flowering or boll opening. It is, however, essential to characterize the growth cycle in order to determine what varieties are suitable for planting in different climatic and local cropping conditions. We estimated the end of the effective flowering period on the basis of the opening date of the last flower giving rise to a first-position boll on fruiting branches (LFP1), and propose a simple method for estimating this date. This study, conducted in 2002 and 2003 at Okpara, Benin, involved a comparison of six cotton varieties planted at two different dates (June and July). Plants were monitored to determine the dates when flowers opened at each position on fruiting branches. The LFP1 indicator made a clear distinction between varieties. This highly heritable trait, which was found to be closely correlated with other earliness criteria, could be used to characterize the length of the growth cycle in cotton varieties.

We studied and modeled damage caused by Helicoverpa armigera larvae on cotton with the aim of developing a coupled crop pest model. Two damage components were studied: the voracity (quantity of fresh matter and number of organs consumed) and feeding preferences (type of organ infested). The laboratory no-choice study of voracity on excised squares and bolls revealed that an H. armigera larva consumes 2,856 mg of fresh matter throughout its larval life, with the sixth instar consuming 86% of this quantity. This consumption rate corresponded to 23.6 squares, or 7.8 bolls. We developed equations to predict the quantity of fresh matter uptake from an individual plant organ, according to the organ mass and the larval instar. The field study of feeding preference confirmed previous findings that larvae prefer squares to bolls, with this preference decreasing as the larval age increases. However, no significant relationship was noted between the age of larvae and the size of infested organs within each organ class (square or boll). We developed a logistic model to predict the probability of a larva infesting a boll rather than a square. According to this model, the relative organ availability in the field and the larval instar were found to be significant factors.

Ralstonia solanacearum is responsible for bacterial wilt affecting many crops worldwide. The emergent population of R. solanacearum (phylotype IIB/4NPB) wilts previously resistant varieties and has rapidly spread throughout Martinique. No conventional method is known to control it. In this study, previous crops used as sanitizing crops were investigated as an environmentally safe alternative method of control. The ability of the emergent population of R. solanacearum to persist in planta and in the rhizosphere of Brassicaceae, Asteraceae and Fabaceae grown as previous crops was evaluated in controlled conditions, and the incidence of bacterial wilt was assessed in the following tomato crop. Results showed that all species carried R. solanacearum latently. Among Brassicaceae and Asteraceae, the highest density of R. solanacearum was found in planta and in the rhizosphere of Tagetes erecta. The density of the R. solanacearum population in the rhizosphere of Raphanus sativus cv. Karacter was significantly higher than that in Raphanus sativus cv. Melody. In Fabaceae, the density of R. solanacearum population in planta was statistically similar in all species. The density of the R. solanacearum population in the rhizosphere of Crotalaria juncea was significantly higher than that in Crotalaria spectabilis. This study showed for the first time that Crotalaria spectabilis and Raphanus sativus cv. Melody grown as previous crops improve the performance of the following tomato with similar effects on R. solanacearum populations in the soil as bare soil. The incidence of the disease in tomato decreased by 86% and 60%, after R. sativus cv. Melody and C. spectabilis, respectively, and the proportion of infected plants also decreased. These results suggest that C. spectabilis and R. sativus cv. Melody can be used as previous crops to help bacterial wilt control in ecological management strategies without drastic suppression of R. solanacearum population in stem tissues and in the rhizosphere. (Resume d'auteur)

The main pests of cotton in Cameroon are the bollworms Helicoverpa armigera , Diparopsis watersi and Earias spp. The implementation of integrated management of these pests requires sampling plans for population monitoring. For practical reasons, these sampling plans have to enable the monitoring of the three bollworm species together rather than each separately. Based on our examination of 2083 samples, 80 or 100 plants each, this study shows that the number of larvae per plant fits a negative binomial distribution (NBD), with a common k value of 5.70, although the numbers of larvae of the three species considered separately fit different NBDs with k values ranging from 1.37 to 3.64. However, we show that possible fluctuations of the distribution resulting from fluctuations in the relative abundance of the three bollworm species would have little or no effect on either the fixed size or sequential sampling plans. We propose the design of a pegboard that allows the extension of sequential sampling in a smallholder context.

Seed coat fragments (SCF) are extremely difficult contaminants to remove from cotton fibers because of the fibers attached to them. Literature data suggest that carding is the most effective cleaning step during which a significant proportion of the SCF is eliminated. In the study described here, different detection methods are used to evaluate the ability of carding to clean SCF-Contaminated cottons. Variations in SCF number and size distribution are evaluated. The amount of fibers attached to the SCF is measured by image analysis, and it appears to have a significant effect on the ability of the card to remove SCF.

Whereas Part I describes the effects of stickiness on productivity, Part II is devoted to an assessment of its impact on quality. Samples of sliver, roving, and yarn taken during the spinning process are examined to determine the effects of stickiness on product quality from the card to the spinning frames. The first effects of stickiness on regularity are noted at the roving frame. Sliver regularity (of the card and drawing frame) does not appear to be affected by stickiness. The quality of ring spun yarn is more sensitive to stickiness than rotor spun yarn. This difference involves several important criteria, notably unevenness and imperfections.

Regularly confronted with the problem of cotton stickiness during spinning, those in the international market wish to see the implementation of a cotton bale classification system to describe the degree to which a given cotton is contaminated. Thermodetection equip ment such as the H2SD high speed stickiness detector may be suitable for such a task. With a working capacity equivalent to that of HVI lines, this instrument evaluates the stickiness of cottons contaminated by insect honeydew. This paper discusses the feasibility of a qualitative classification based on within-bale distribution of the number of sticky points. A detailed example obtained with the H2SD illustrates the qualitative classification proce dure. Sticky bales can be separated from uncontaminated bales for the latter to be marketed with a guaranteed stickiness level below a set threshold and a litigation risk selected in advance.

'Sticky cotton' causes substantial economic loss in central Africa. The phenomenon is the result of honeydew excreted by sapsucking insects, mainly the aphid Aphis gossypii Glover. The chemical protection currently available to farmers does not limit this type of damage effectively. From exhaustive counting of infested leaves and quantitative measurement of stickiness with a thermodetector, a positive relationship can be established between the seriousness of fibre stickiness and aphid outbreaks at the end of the season. A control method involving the topping of plants after boll opening reduces the number of leaves on which aphids feed and thus the abundance of the aphids.

Rice is a crop prone to drought stress in upland and rainfed lowland ecosystems. A deep root system is recognized as the best drought avoidance mechanism. Genome-wide association mapping offers higher resolution for locating quantitative trait loci (QTLs) than QTL mapping in biparental populations. We performed an association mapping study for root traits using a panel of 167 japonica accessions, mostly of tropical origin. The panel was genotyped at an average density of one marker per 22.5 kb using genotyping by sequencing technology. The linkage disequilibrium in the panel was high (r2>0.6, on average, for 20 kb mean distances between markers). The plants were grown in transparent 50 cm × 20 cm × 2 cm Plexiglas nailboard sandwiches filled with 1.5 mm glass beads through which a nutrient solution was circulated. Root system architecture and biomass traits were measured in 30- day-old plants. The panel showed a moderate to high diversity in the various traits, particularly for deep (below 30 cm depth) root mass and the number of deep roots. Association analyses were conducted using a mixed model involving both population structure and kinship to control for false positives. Nineteen associations were significant at P

Binding and competition among Cry1Aa, Cry1Ac, and Cry1Ba toxins were analyzed quantitatively in vitro by using (sup125)I-labeled activated toxins and brush border membrane vesicles isolated from Chilo suppressalis larval midguts. The three toxins bound specifically to the midgut brush border membrane vesicles. Direct binding experiments showed that Cry1Aa and Cry1Ba recognized a single class of binding sites with different affinities, whereas Cry1Aa recognized two classes of binding sites, one with a high affinity and a low concentration and the other with a lower affinity but higher concentration. Competition experiments showed that toxins Cry1Ac and Cry1Ba shared a binding site in the C. suppressalis midgut membranes and that this site was also the low-affinity binding site for Cry1Aa.

Around 50% of the cotton bales produced in the world are sold based on the analysis of their technological characteristics using a standardized instrument for testing cotton (SITC). In the United States of America, periodical studies of variability of the results allow the results to be accompanied by commercial tolerances to limit the frequency of claims. However, to our knowledge, no such study has been conducted in Africa. For this reason, we studied within-bale variability of fiber micronaire, length, uniformity, strength, reflectance and yellowness as measured by the SITC. We took eight samples per bale from 215 cotton bales produced by 27 ginning mills in eight Sub-Saharan countries in Africa in two cropping seasons. Our representative sample was composed of 1720 fiber samples that were analyzed in controlled conditions using the SITC in a laboratory in which international recommendations are fully respected. We measured within-bale variability and suggest a sampling method and procedure to perform technological and instrumental tests in West and Central Africa.

Mutations at the codon 132 in the isocitrate dehydrogenase 1 (IDH1) gene occur early, with a high frequency, in World Health Organization (WHO) grade II gliomas. We investigated the impact of IDH1 mutations on spontaneous glioma growth rate, known to be an early prognostic factor.The mean tumor diameter was assessed on the first MRI performed at diagnosis and on a second MRI, performed immediately before surgery, in a series of 64 WHO grade II gliomas. The patients did not undergo treatment before surgery. Because of a frequent association, we jointly analyzed the 1p19q co-deletion and IDH1 mutations effects on tumor velocity of diameter expansion (mm/year) during preoperative spontaneous growth period. 1p19q co-deletion had a significant slowing effect (p = 0.0133) on tumor growth estimated at −1.7760 ± 0.711 mm/year (95% CI −3.154, −0.366), whereas IDH1 mutations estimated effect of +0.036 ± 0.833 mm/year (95% CI −1.668; +1.596) was not significant (p = 0.9654). Our results provide first evidence that IDH1 mutations are not significantly involved in tumor growth rate. By contrast, we confirm that 1p19q co-deletion decreases growth velocity.

International audience; The applicability of conservation agriculture (CA) in sub-Saharan Africa (SSA) is poorly documented. In the ''Nord'' and ''Extre ˆme-Nord'' provinces of Cameroon, in a 2-year rotation between a cereal (maize or sorghum) and cotton, conventional techniques were compared with CA. The study was conducted from 2001 to 2006 in 662 plots in 243 farmers' fields. Cereal treatments compared were conventional management techniques and CA consisting in the production of mulch using cover crops (Brachiaria ruziziensis, Crotalaria retusa, Dolichos lablab, Mucuna pruriens, Vigna unguiculata) intercropped within the cereal. In the ''Extre ˆme-Nord'' province up to 9.7 t ha1 of vegetative biomass was produced in the CA plots with sorghum and cover crops against up to 4.8 t ha1 for sorghum alone in conventional plots. In the ''Nord'' province maize + cover crops produced up to 5.2 t ha1 of biomass against up to 2.5 t ha1 for maize alone. In both provinces, the cereal grain yields were equivalent or higher in CA compared to conventional plots. In 18 fields of the ''Extre ˆme-Nord'' province the mulch remaining the year after sorghum + B. ruziziensis is mainly comprised between 3 t ha1 and 5 t ha1. Cotton treatments compared were T (tillage), NT (no tillage), and NTM (no tillage with mulch). In both provinces these treatments differed in soil cover, number of localized herbicide sprays used, ridging, and amount of nitrogen fertilizer used. In the ''Extre the date of the first weeding. In the ''Extre ˆme-Nord'' province treatments differed also for the number of weeding and ˆme-Nord'' province cotton yieldswere 12% lower for T and 24% lower for NT than for NTM. Cotton yields were regressed on crop husbandry indicators and used inputs. After a manual backward removal in amultiple linear regression respectively no parameters were found to significantly influence yield for T, only one parameter for NT, the number of herbicide sprays used at sowing, and three parameters for NTM: difference between heavy clay and silty loam, application of NPK fertilizer, sowing date. In the ''Nord'' province no difference in cotton yield was observed between T, NT and NTM. The flowering period was longer for NTM vs NT in the ''Extre ˆme-Nord'' and the ''Nord'' provinces andNTM vs T in the ''Nord'' province, respectively 13, 9 and 8 days. Although we show that CA techniques can have benefits at field level, further studies are needed to assess their suitability at farm and village levels.

Diversity among 124 sorghum landraces from 10 villages surveyed in 3 regions of Burkina Faso covering different agroecological zones was assessed by 28 agromorphological traits and 29 microsatellite markers. 94.4% of the landraces collected belonged to the botanical race guinea (consisting of 96.6% guinea gambicum and 3.4% guinea margaritiferum), 74.2% had white kernels, 13.7% had orange and 12.1% had red kernels. Compared to the ''village nested within zone'' factor, the ''variety nested within village within zone'' factor predominately contributed to the diversity pattern for all nine statistically analysed quantitative traits. The multivariate analyses performed on ten morphological traits identified five landrace groups, and of these, the red kernel sorghum types appeared the most homogenous. 2 to 17 alleles were detected per locus with a mean 4.9 alleles per locus and a gene diversity (He) of 0.37. Landraces from the sub- Sahelian zone had the highest gene diversity (He = 0.38). Cluster analysis revealed that the diversity was weakly stratified and could not be explained by any biophysical criteria. One homogenous guinea margaritiferum group was distinguished from other guinea landraces. The red kernel type appeared to be genetically distinct from all other guinea landraces. The kernel colour was the principal structuring factor. This is an example of a homogeneous group of varieties selected for a specific use (for local beer preparation), mainly grown around the households in compound fields, and presenting particular agromorphological and genetic traits. This is the most original feature of sorghum diversity in Burkina Faso and should be the focus of special conservation efforts.

Correspondance: desclaux@supagro.inra.fr; International audience; The standardization of environments (E) encouraged by modern society and by the productivist model of agriculture has resulted in the standardization of genotypes (G) thereby reducing G × E interaction. New societal values call for the diversification of agriculture to fit contrasted environments. This process can be depicted by four models defined by two axes, one socio-economic (individual logics versus collective governance), and the other agro-ecological (reductionist versus systemic approaches). These models differ in (i) their objectives (from improvement in yield to the empowerment of farmers), (ii) their specific expectations with respect to genotypes (from inherited genetic resources to varieties that represent genetic, ethical and social progress), (iii) their specific representations of the environment (E) (from a simple interaction between the bio-physical environment (B) and the crop management (C), to a complex interaction including the competences of the actors (A), outlets (O), regulations (R), society (S)), (iv) their particular relations between G and E (from G × E to G × B × C × A under evolving constraints represented by R × O × S). Taking this diversity into account changes the way plant improvement is considered. Thus, depending on the model, the order, interest and status of the five classic stages of plant improvement (setting objectives, creating variability, selecting, evaluating and disseminating) may be called into question. Between the existing analytical model (Model I) and a holistic model (Model IV) which remains to be developed, lies the challenge of ensuring the sustainability, efficiency and acceptability of plant breeding and resulting innovations. From a simple “statistical parameter” that we, as plant breeders, attempt to reduce, the G × E interaction is becoming an “objective” that we try to predict and valorize. Structuring the different components of E, G and G × E, enables us to extend the basic concept of representivity to both the cultivation conditions and the relational socio-economic positions of the actors involved

Background and Aims: West African sorghum (Sorghum bicolor) varieties are generally highly photoperiod-sensitive, which is a necessary adaptation to the variable onset date of the rainy season and the variable dates of sowing in the savannah zone. Depending on sowing date, plants can produce from 12 to >40 leaves on the main culm, with height varying from 1 m to more than 5 m. The present study aimed to better understand the complex phenology of these variables.Methods: A 2-year series of monthly sowings of three West African sorghum varieties was conducted near Bamako, Mali. Drought stress was avoided by supplemental irrigation. Rate of initiation of primordia at the stem apex was recorded, together with rate of leaf emergence and increase in plant height.Key Results: Leaf initiation and appearance rates (plastochron-1 and phyllochron-1) were constant for a given sowing date in cases where less than 20 leaves were produced (generally observed with late sowing dates). In contrast, rates were bilinear for early sowing dates, for which plants produced more than 20 leaves. The secondary rates, which occurred from the 20th leaf onwards, were only half of the initial rate. Plastochron and phyllochron showed large variations among sowing dates, and were correlated with the rate of plant height increase. The initial plastochron and phyllochron were positively correlated with soil temperature and negatively correlated with both day length and day-to-day change of day length prevailing at plant emergence, but these factors explained only half of the variation observed.Conclusions: Although they belong to different genetic groups and have different height and photoperiod sensitivity, the three varieties studied exhibited similar response patterns of development rates among phenological phases and seasons, with the local landrace showing the greatest variation due to its longer vegetative phase and longer stem internodes. The possible adaptive advantages in African savannah environments of bilinear development rates and the associated limitation in height increase are discussed. (Resume d'auteur)

Effective flowering time in Gossypium hirsutum cotton plants was studied with the aim of enhancing decision making on the best varieties to plant according to the planting date under rainfed cropping conditions. Trials were conducted at two sites in a cotton-growing area of Benin in 2002 and 2003. A split-split plot design with three replicates was used to compare 10 cotton varieties, with different growth cycle lengths and morphology, at three stand densities (42 000, 125 000, 167 000 plants ha−1) and two planting dates (standard planting in June and late planting). The flowering period was characterized by the mean first flower opening date (FF), which is an indicator of flowering earliness, and by the opening date of the last flower giving rise to a first-position boll on fruiting branches (LFP1). Effective flowering time (EFT) was calculated as the difference between LFP1 and FF. EFTs differed markedly in the 10 cotton varieties tested and this parameter could not be predicted on the basis of flowering earliness. Late planting and high planting rates delayed first-flower opening, accelerated last-boll development and reduced the effective flowering time. This latter factor should be taken into account in cotton breeding programmes so that varieties adapted to local rainfall constraints can be recommended to growers while also enhancing crop management sequences.

2. Genotyping The genotyping of all panels (hybrids, S. officinarum and S. spontaneum) using the DArT technology produced 1472 polymorphic dominant arkers (Heller-Uszynska et al., 2011) coded 0 for absence and 1 for presence. Low or high frequency markers ( 0.95) or markers ith more than 10% missing data were removed. Marker frequency was used to impute missing data. To homogenize genotyping and void over-representation of certain genomic regions, a subsample of independent markers was chosen from the entire marker dataset. arkers were selected as independent based on Fisher’s exact test between each pair of markers. The threshold value was defined using onferroni’s correction for multiple testing. The final number of DArT markers used in the analyses was 429. The Section 2.7 should be read as follows

Une nouvelle méthode pour modéliser les interactions génotype X environnement: APLAT. Le rendement de génotypes prédit par un modèle de simulation de cultures est développé en série de Taylor à l'ordre 1 au voisinage du vecteur de paramètres d'un génotype de référence. À l'aide de cette linéarisation locale, l'estimation des paramètres de ces génotypes se fait par régression linéaire des rendements observés sur la sensibilité des sorties du modèle de simulation de cultures par rapport aux paramètres.

Soil water deficit is a major factor influencing yield and grape berry composition, with an optimal situation corresponding to a moderate water deficit. A computer-based diagnosis tool requiring a limited number of measurements was developed and used to characterise the typology of soil water deficit experienced by grapevine in field sites over the cropping season with the aim of producing acceptable yields and achieving good quality red wine. The tool combined a simple soil water balance model parameterised for vine, to simulate the fraction of transpirable soil water (FTSW) at a daily time step, and an indicator of vine water status in terms of source/sink relationships and canopy microclimate. This indicator was developed from quantitative relationships between FTSW and maximal net assimilation (Anmax), under optimal conditions of light and temperature, the leaf emergence rate on first-order lateral branches (LER I), the leaf emergence rate on second-order lateral branches (LER II) and the final length of first-order lateral branches (BL I). These relationships were parameterised in a greenhouse on Syrah (syn. Shiraz) grafted onto Fercal rootstock. The computer-based diagnosis tool was used on 24 fields within 4 vineyards of Mediterranean southern France covering contrasting pedoclimatic conditions, cultivar/rootstock pairs and receiving different water regimes. The key soil parameter of the water balance model, the total vine's transpirable soil water (TTSW), was estimated in each field from measurements of predawn leaf water potential (Ψl) at different vine growth stages in conjunction with an optimisation algorithm of the simulated FTSW as a function of predawn Ψl obtained with an independent data set. The tool enabled classification of the fields on the basis of the changes of the indicator of vine water status across four phases of the plant cycle, from budburst to harvest, relative to an optimum change of this indicator defined for production of a good quality wine. Such a tool will be useful to farm advisers and researchers for the evaluation of management practices as for comparative analysis between vineyards.

A new method for modelling genotype x environment interaction: APLAT. The yield predicted by a crop-simulation model is developed as a Taylor series in the neighbourhood of a parameter vector of a control genotype. With this local linearisation, these genotype parameters can be estimated by a linear regression of the observed yield on the derivatives of the crop-simulation model predictions with respect to its parameters.

Field sampling plans are proposed for monitoring Helicoverpa armigera (Hubner) populations on cotton and for threshold insecticide treatment in sub-Saharan Africa. The geostatistical analysis of the spatial distribution of bollworms validates systematic sampling (along a diagonal for example); systematic sampling is easier to perform than random sampling. The same negative binomial distribution was observed in Burkina Faso and Cameroon under controlled conditions, with or without insecticide treatment. The low infestation levels prevented the use of the approximate methods of Iwao or Kuno. Hence, exact confidence intervals not available in the literature are presented in graph form. Sequential and fixed size sampling methods are presented.

Les pullulations precoces d'Aphis gossypii etaient anciennement negligeables en Afrique centrale, mais elles affectent maintenant les cultures de coton des le debut du cycle de developpement. Quatre annees d'essais de protection des cultures ont montre que ces pullulations provoquent un retard dans le developpement vegetatif ainsi qu'une perte de 5 a 23 % de production de graines de coton. Aphis gossypii est donc un des principaux ravageurs dans cette partie du monde. Le semis precoce apparait comme le plus efficace des moyens de lutte. Pour les semis tardifs, c'est la protection insecticide des graines qui semble la meilleure methode de lutte

Ecomeristem, FSPM simulating sugarcane morphogenesis and plasticity depending on its nutritional status (C, H20), was used to analyze the genetic diversity of source-sink related traits controlling structural vs. non structural carbohydrate accumulation and thus sugar production. Results are discussed with respect to the challenge of FSPM assisted phenotyping and breeding.

Madagascar is one of the biggest rice producers in Africa. About 20% of the production is rainfed rice. Since the 1990's, cold tolerant cultivars adapted to high altitude have been successfully proposed to farmers and adopted in the central highlands of the country. But rice blast pressure on the hillsides is important and is the second constraint (after cold) breeders have to take into account. In the same time, hillsides were more and more put under cultivation, conservation agriculture cropping systems have been proposed to farmers (Scopel et al., 2013), to limit erosion and to ensure the sustainability of rainfed crops. In experimental studies, these conservation agriculture cropping systems were shown to limit blast epidemics (Sester et al., in press). We hypothesized that the specificities of the conservation agriculture cropping systems (no tillage, permanent soil cover and rotations) induced changes in plant nutrition and plant nutritional balance, especially nitrogen assimilation, and that it was the reason of a differential expression of tolerance to blast. To test this hypothesis, pluriannual factorial experiments were conducted in two sites to measure blast disease severity on leaves and on panicles in two cropping systems (conservation agriculture vs conventional) and three nitrogen fertilization levels. Crop development and nitrogen assimilation were measured at four stages and yield components were assessed at harvest. A mineral diagnosis was performed at booting stage on the flag leaf. A relational database has been created to manage agronomic and disease severity rating data from the project. For statistical analysis (performed using Sas) the proportion of damaged leaf area and damaged panicles were fitted to a generalized linear model. We will present the results of the respective impacts of the cropping system environment, the nitrogen fertilization and the crop density on the leaf and panicle blast level. In some particular conditions, a significant impact of the cropping system appeared in addition to the effects explained by crop density, confirming the hypothesis of a plant susceptibility modification. We present then the way these results could be used to adapt cropping systems and to improve upland rice tolerance to blast in a sustainable way.