Search

The release of the cassava reference genome by Prochnik et al. in 2012 has allowed cassava geneticists identifying tens of thousands genome-wide sequence variations across multiple cultivars. These genomic variations have helped to develop a new generation of molecular markers for the crop's genetics research either by re-sequencing using restriction-site associated DNA-sequencing (RAD-seq) or by the SNPtype™ allele-specific PCR assays technology. Today CIAT's cassava program using these technologies has characterized at the DNA level more than 3000 cassava cultivars used by farmers in South East Asia or Latin America; including improved cultivars, LAC landraces, and its potential wild close relatives conserved at the world's largest collections of the crop. These next generation molecular markers along with the analytical methods implemented in the Cassava Genome Hub has allowed to analyze the crop's genetic diversity, performed population and family structure analyses, unravel the crop's phylogenetic and phylogeographic history and confirm its recent introduction histories in Africa and South East Asia. Also, using these genomic and bioinformatics resources, the SNPtype™ technology has allowed us to examine the factors affecting the adoption of improved cassava varieties in the Cauca Department in southwest Colombia, as well as, six regions of Vietnam where most of the cassava is grown. These analytical approaches have showed the power of our next-generation sequencing analytical methods could have in identifying both historical population structure and recent colonization history along with the identification of clones recently adopted including its pedigree. But more importantly, it is guiding our efforts to understand the nature of complex traits in cassava such as whitefly resistance, post-harvest physiological deterioration of the roots, starch stability and content and its resistance to diseases such as frog skin disease.

We evaluated the yields of Oryza sativa L. 'Nipponbare' rice lines expressing a gene encoding an A20/AN1 domain stress-associated protein, AlSAP, from the halophyte grass Aeluropus littoralis under the control of different promoters. Three independent field trials were conducted, with drought imposed at the reproductive stage. In all trials, the two transgenic lines, RN5 and RN6, consistently out-performed non-transgenic (NT) and wild-type (WT) controls, providing 50–90% increases in grain yield (GY). Enhancement of tillering and panicle fertility contributed to this improved GY under drought. In contrast with physiological records collected during previous greenhouse dry-down experiments, where drought was imposed at the early tillering stage, we did not observe significant differences in photosynthetic parameters, leaf water potential, or accumulation of antioxidants in flag leaves of AlSAP-lines subjected to drought at flowering. However, AlSAP expression alleviated leaf rolling and leaf drying induced by drought, resulting in increased accumulation of green biomass. Therefore, the observed enhanced performance of the AlSAP-lines subjected to drought at the reproductive stage can be tentatively ascribed to a primed status of the transgenic plants, resulting from a higher accumulation of biomass during vegetative growth, allowing reserve remobilization and maintenance of productive tillering and grain filling. Under irrigated conditions, the overall performance of AlSAP-lines was comparable with, or even significantly better than, the NT and WT controls. Thus, AlSAP expression inflicted no penalty on rice yields under optimal growth conditions. Our results support the use of AlSAP transgenics to reduce rice GY losses under drought conditions.

Rice breeding has made important contributions to Latin America. More than 400 cultivars were released from 1975 to 2012, which helped to raise total production to >27 million tonnes obtained from 5.7 million hectares (average for 2010-2012). Rice production provides ?US$8.8 billion for thousands of farmers in Latin America and the Caribbean (LAC). The result of higher yields in the irrigated sector was to triple rice production in LAC while area did not grow, thus preserving more fragile environments. Several estimates on genetic gains for grain yield have been carried out in LAC. In temperate irrigated rice, the estimates are around 1.5-2.6% per year. In the tropical irrigated, it is ?1% and in the upland rice the estimate is ?1.4% per year. Different breeding strategies, including pedigree, modified bulk, recurrent selection methods, anther culture, interspecific crosses, composite populations, quantitative trait loci (QTL) introgression, and recombinant inbred lines, accompanied by shuttle breeding schemes, direct seeding, and evaluation/selection in hot spots for main diseases are being used by CIAT and NARES in the region. In this process, methods for screening for diseases and other stresses were established. Networking has been a cornerstone for success and several networks such as INGER, FLAR, and HIAAL were created. Looking forward, as farmers' yields are approaching the genetic yield potential exhibited by current cultivars, as a result of improved agronomic management, a new breakthrough is needed in terms of more productive cultivars. To achieve this goal, a strategy is needed that includes strong pipelines focused on specific environments and markets; better product profiling; integration between discovery, development, and delivery; and new breeding strategies using cuttingedge technologies and new breeding methods to accelerate genetic gains.

Genomic selection is a promising breeding strategy for rapid improvement of complex traits. We investigated the accuracy of genomic estimate breeding values (GEBV) through cross validation in a synthetic population (SP) of broad genetic diversity developed for upland rice breeding for Latin America, through recurrent selection (RS). Three hundred and three S4 lines extracted from a SP that has undergone several RS cycles were phenotyped for flowering time (FL), plant height (PH) and grain weight per plant (GW), and genotyped with an average density of one marker per 22.9 kb, using genotyping by sequencing. Cross validation was performed hundred times, for four ratios of training and validation population sizes (TP/VP), using ridge regression best linear unbiased prediction (RR-BLUP), least absolute shrinkage and selection operator (LASSO) and Bayesian linear regression methods (BL). Accuracy was estimated as the correlation between GEBV and true breeding value. Linkage disequilibrium was high (average r²>0.5 at 20 kb; r²>0.25 at 1.5 Mb distance). Accuracy of GEBV increased with the size of the training population, regardless of the model. For FL and TP/VP varying from 150/153 to 270/33, the GEBV accuracy ranged from r=0.20 (0.05) to r=0.22 (0.15) under RR-BLUP, from r=0.19 (0.09) to r=0.28 (0.17) under LASSO, and from r=0.28 (0.03) to r=0.32 (0.17) under BL, before marker selection based on alleles frequency. GEBV accuracy was more sensitive to markers' allele frequency than to the total number of markers. The best selected S4 will be recombined and their progenies tested for GEBV accuracy across generations.

A decade ago CIAT began the search of what it called "High-Value Cassava". What was then only a promise has gradually became a reality. In the area of starch quality traits an amylosefree (waxy) mutation was identified and has now been successfully introgressed to be released as commercial varieties. Part of this work is financially supported by the private starch sector. A small-granule mutation was also identified and found to be particularly suitable for processes requiring starch hydrolysis (e.g. sweeteners and ethanol). Moreover, CIAT has also identified a high-amylose genotype with as much as 42% amylose (whereas cassava starch typically has around 20%). Breeding methodologies to exploit these traits have been developed and tested. On the nutritional side, during the past decade, the increases in carotenoids content in the roots have been remarkable, and there seems to be no plateau in the progress made. Protocols for efficient selection in carotenoids have also been made. These results demonstrate that Manihot esculenta has many useful traits waiting to be found and also that this crop is capable of rapid responses if proper breeding methodologies are employed. Future work will concentrate on developing herbicide resistance and improving the protocol for assessing post-harvest physiological deterioration in roots.

Since 1992, we have been carrying out an upland rice breeding program for Latin American and the Caribbean based on the improvement of synthetic populations of broad genetic diversity through recurrent selection (RS). In this breeding scheme, at each RS cycle the population (Pn) is evaluated for specific traits and the selected units undergo recombination to generate the Pn+1 population and pedigree breeding to develop new and improved breeding lines. Recombination of the selected units is facilitated by the presence of recessive nuclear male-sterility. RS constitutes an effective framework for combining the centralised improvement of a small number of complex traits with decentralised breeding for site-specific traits. It has allowed us to provide our partners with improved lines and progenitors. We are now adapting our RS strategies to the availability of cheaper genotyping tools and high throughput and precision phenotyping platforms. Established from several cycles of recombination among a large number of founder accessions, synthetic populations make-up a valuable material for association analysis, with lower LD and improved mapping resolution compared to the RILs populations. This paves the way for implementation of marker assisted recurrent selection (MARS) aimed at continuous extraction of lines for specific environments while keeping genetic diversity for long term progress for complex traits. Synthetic populations also constitute a good material for genome-wide selection (GWS). We intend to compare the relative efficiency of different GWS schemes and MARS scheme for the improvement of yield potential and drought tolerance. (Texte integral)

The industrial starch market is undergoing major expansion, but certain specific industrial uses cannot be satisfied by native starches and, therefore, chemical or physical modification is necessary. These modifications are often harmful to the environment and generate additional costs. Mutations in the cassava starch biosynthesis pathways were recently discovered at CIAT, Cali, Colombia. CIRAD, in partnership with CIAT, carried out a study on the physicochemical and functional properties of these starches, which could result in interesting industrial applications and create new market for cassava starch. Among the new cassava starch types, two new mutant cassava starches with extreme amylose contents (0% and 31%) have been recently reported. These mutants are drastically different from normal cassava starch whose amylose content typically ranges between 15-25%. The new mutants were compared with normal cassava starches and commercial versions of amylose-free or normal potato, rice and maize starch. The structure of cassava amylopectin was not modified by the waxy mutation and waxy cassava starch exhibited properties similar to the ones of waxy maize starch. On the contrary, the higher-amylose mutations induced by gamma rays radiation in cassava deeply modified the branching pattern of amylopectin as well as other starch characteristics and properties. These modifications resulted in changes in starch granule ultrastructure (e.g. decreased starch crystallinity), a weak organized structure, and increased susceptibility to mild acid and enzymatical raw starch hydrolysis (fastest and most efficient hydrolysis of all studied native starches). This mutation could offer interesting advantages for the production of bioethanol. Gels from normal root and tuber starches (potato, cassava) after refrigeration and freeze/thaw had lower syneresis than cereal starches (maize, rice). Gels from waxy starches (except for potato) did not present any syneresis after 5 weeks of storage at 4°C. Waxy cassava starch was the only one not showing any syneresis after 5 weeks of storage at -20°C. The distinctive properties of the new cassava starches suggest new opportunities and commercial applications for tropical sources of starch. Supporting information: Journal of Agricultural and Food Chemistry (2007), 55(18): 7469-7476. http://dx.doi.org/10.1021/jf800603p; Journal of Agricultural and Food Chemistry (2008), 56(16): 7215-7222. http://dx.doi.org/10.1021/jf800603p; Journal of Agricultural and Food Chemistry (2010), 58(8): 5093-5099. http://dx.doi.org/10.1021/jf1001606; Food hydrocolloids (2012), 27(1): 161-174. http://dx.doi.org/10.1016/j.foodhyd.2011.07.008. (Texte integral)

The Global Cassava Strategy initiative identified weak markets for cassava products as a bottleneck limiting the capacity of the crop to help farmers out of poverty. Identifying or creating genetic variability to make cassava better suited to meet the needs of different industries could help overcoming this bottleneck. This article summarizes the results from traditional approaches (breeding, screening germplasm collections, inbreeding to expose recessive traits, mutagenesis and inter-specific crosses). Significant progress has been attained for different traits recently: 1) Tolerance to post-harvest physiological deterioration was introgressed from #Manihot wakerae# and perhaps found in mutagenized populations. This trait will drastically reduce the marketing costs of cassava; 2) Commercially relevant starch mutations (amylose-free and small-granule/high-amylose) have been identified through inbreeding and induction of mutations. These traits are relevant for the starch and bio-ethanol industries; 3) A group of cassava clones from the germplasm collection, with 2- 3 times higher levels of proteins in the roots, has been identified. This is relevant for the feed industry and, obviously, to improve nutritional status of human populations; 4) Rapidcycling recurrent selection has doubled the maximum level of [bêta]-carotene in cassava roots. This is important for human nutrition; for the feed industry and as example of cassava's capacity to respond to proper breeding approaches. High-value cassava is, therefore, no longer a dream but a reality that will have a positive impact not only in the livelihood of cassava farmers but in the scientific community: cassava offers a wealth of professional opportunities to young scientists.

There is wide variation in the description of starch quality traits for cassava and little variation had been reported until recently. CIAT, therefore, initiated a project to evaluate and screen for starch quality traits in the entire germplasm collection. High value traits (for example, amylose-free starch and high-protein content) have been identified. Up to now, starch samples from more than 4000 accessions (3272 landraces and 772 improved clones) from the cassava germplasm collection have been obtained and analyzed. The size of this sample is very large and the information it provides very robust. Average dry matter content of the landraces was 32.8% whereas for improved clones it was 36.7%, but starch content was the same (84.5%). Cyanogenic potential ranged from 14 to 3274 ppm for an average of 325 ppm. Averages for landraces were slightly higher (339ppm) than in improved clones (267 ppm). Total sugars and reduced sugars were 3.68 and 1.25, respectively (landraces) and 4.05 and 1.56 (improved clones. Average amylose content was 20.7% across all the genotypes evaluated. There was little variation between landraces and improved clones for water absorption (4.59%); water solubility (2.17%); easy cooking (2.80 min); pasting temperature (65.2°C); maximum viscosity (777 cP); breakdown (298 cP); and consistency (156 cP). Clarity of the gels was 44.5% for landraces and 48.1% for improved clones. (Texte intégral)

DNA variation was studied at the intra- and interspecific levels within the genus #Passiflora#, with particular emphasis on the subgenus #Tacsonia# and its relationship with the subgenus #Passiflora#. AFLP fingerprints were generated for 35 accessions from 21 species, 14 of which belong to subgenus #Tacsonia#, and four to subgenus #Passiflora#. The morphologically intermediate #P. manicata# (subgenus #Manicata#) was also included, as were #P. trinervia# (subgenus #Psylanthus#) and #P. alnifolia# (subgenus #Decaloba#). The two latter were used as outgroup species. Two primer combinations generated 34-80 scorable fragments per genotype for a total of 260 fragments. Polymorphism analyses spanned the levels from individual genotype to section. Subgenus #Tacsonia# presented wide genetic variability at both intra- and interspecific levels, particularly within wild species. The classification based on these markers generally agrees well with current taxonomic definitions of common species, but less with definitions of sections. A geographical component was evident in the intraspecific variation of #P. mixta#, #P. tripartita# var. #mollissima# and #P. tarminiana#, with wider variation in Ecuador and Peru than in Colombia and Venezuela. These same three species. plus #P. cumbalensis#, formed adjacent clusters, showing closer affinity with each other than with the other species of subgenus #Tacsonia#. As expected, #P. manicata# clustered at an intermediate position between subgenera #Tacsonia# and #Passiflora#.

DNA variation was studied in 35 accessions representing 14 species of Passiflora subgenus Tacsonia (including five cultivated species), P. manicata (subg. Manicata), four species of subgenus Passiflora, P. trinervia (subg. Psylanthus), and P. alnifolia (subg. Decaloba). Two primer combinations were retained out of 24 tested. They generated 34 to 80 scorable bands per genotype, for a total of 260 bands on the whole sample. The five wild accessions of P. mixta showed higher polymorphism than those of the cultivated P. tripartita var. mollissima, P. sp'curuba india', and P. pinnatistipula. DNA fingerprints of P. x rosea clearly correspond to the superposition of the bands from its putative parents, P. tripartita var. mollissima and P. pinnatistipula. Cluster analysis is consistent with taxonomical classification into subgenera, with P. manicata intermediate between subgenera Tacsonia and Passijlora, but less so concerning Tacsonia sections. Within this subgenus, P. mixta, P. tripartita var. mollissima, P. sp'curuba india', and P. cumbalensis show particular affinity, as compared with P. pinnatistipula and less common species. Distances between Tacsonia and other subgenera appear limîted as compared to distances within Tacsonia, suggesting the use of other techniques to study the relationships between subgenera.