Genotyping-By-Sequencing diversity analysis of international Vanilla collections uncovers hidden diversity and enables plant improvement.

Genomics-based diversity analysis of natural vanilla populations is important in order to guide conservation efforts and genetic improvement through plant breeding. Vanilla is a cultivated, undomesticated spice that originated in Mesoamerica prior to spreading globally through vegetative cuttings. Vanilla extract from the commercial species, mainly V. planifolia and V. × tahitensis, is used around the world as an ingredient in foods, beverages, cosmetics, and pharmaceuticals. The global reliance on descendants of a few foundational clones in commercial production has resulted in an industry at heightened risk of catastrophic failure due to extremely narrow genetic diversity. Conversely, national and institutional collections including those near the center of cultivation contain previously undiscovered diversity that could bolster the genetic improvement of vanilla and guide conservation efforts. Towards this goal, an international vanilla genotyping effort generated and analyzed 431,204 single nucleotide polymorphisms among 412 accessions and 27 species from eight collections. Phylogenetic and STRUCTURE analysis sorted vanilla by species and identified hybrid accessions. Principal Component Analysis and the Fixation Index (FST) were used to refine relationships among accessions and showed differentiation among species. Analysis of the commercial species split V. planifolia into three types with all V. × tahitensis accessions being most similar to V. planifolia type 2. Finally, an in-depth analysis of V. × tahitensis identified seven V. planifolia and six V. odorata accessions as most similar to the estimated parental genotypes providing additional data in support of the current hybrid theory. The prevalence of probable V. × tahitensis parental accessions from Belize suggests that V. × tahitensis could have originated from this area and highlights the need for vanilla conservation throughout Central and South America. The genetic groupings among accessions, particularly for V. planifolia, can now be used to focus breeding efforts on fewer accessions that capture the greatest diversity.

Plant DNA Barcoding Principles and Limits: A Case Study in the Genus Vanilla.

Powerful DNA barcodes have been much more difficult to define in plants than in animals. In 2009, the international Consortium for the Barcoding Of Life (CBOL) chose the combination of the chloroplast genes (rbcL + matK) as the proposed official barcode for plants. However, this system has got important limits. First, any barcode system will only be useful if there is a clear barcode gap and if species are monophyletic. Second, chloroplast and mitochondrial (COI gene used for animals) barcodes will not be usable for discriminating hybrid species. Moreover, it was also shown that, using chloroplast regions, maximum species discrimination would be around 70% and very variable among plant groups. This is why many authors have more recently advocated for the addition of the nuclear ITS region to this barcode because it reveals more variations and allows the resolution of hybrid or closely related species. We tested different chloroplast genes (rbcL, matK, psaB, psbC) and the nuclear ITS region in the genus Vanilla, a taxonomically complex group and therefore a good model to test for the efficiency of different barcode systems. We found that the CBOL official barcode system performed relatively poorly in Vanilla (76% species discrimination), and we demonstrate that adding ITS to this barcode system allows to increase resolution (for closely related species and to the subspecies level) and to identify hybrid species. The best species discrimination attained was 96.2% because of one paraphyletic species that could not be resolved.

Distribución potencial y características geográficas de poblaciones silvestres de Vanilla planifolia(Orchidaceae) en Oaxaca, México / Potential distribution and geographic characteristics of wild populations ofVanilla planifolia(Orchidaceae) Oaxaca, Mexico

ResumenLos especímenes silvestres de Vanilla planifolia G. Jack forman parte del acervo genético primario, los cuales solo se han reportado en Oaxaca, México. Por ello se evaluó la distribución de esta especie con el objetivo de ubicar y describir características ecológicas en zonas potenciales de distribución. La metodología empleada consistió de cuatro etapas: 1) Elaboración de una base de datos con registros de herbario;2) Construcción de la distribución potencial basado en los registros históricos de herbario para la especie, mediante el modelo de máxima entropía (Maxent), con el uso de 22 variables bioclimáticas como predictoras; 3) Realización de búsquedas sistemáticas de individuos in situ con base en los registros de herbario y las áreas de distribución potencial en 24 municipios, para conocer la situación y la distribución del hábitat actual, y 4) Descripción mediante factores ambientales de los nichos ecológicos potenciales generados por MaxEnt. La revisión de las colecciones de herbarios reportó un total de 18 registros de V. planifolia, comprendidos entre 1939 y 1998.La búsqueda sistemática de individuos en campo ubicó 28 plantas distribuidas en 12 sitios sobre 95 364 Km2. Las variables que contribuyeron con mayor valor porcentual para determinar la estimación del modelo de distribución potencial en vainilla son precipitación del periodo más lluvioso (61.9 %), régimen de humedad del suelo (23.4 %) y precipitación del cuatrimestre más lluvioso (8.1 %). El hábitat potencial de la especie se distribuyó en cuatro zonas; trópico húmedo del golfo de México, templado húmedo, trópico húmedo y templado húmedo del pacifico. La precipitación anual osciló de 2 500 a 4 000 mm, con lluvias en verano y porcentaje de precipitación invernal de 5 a 10 %. El régimen de humedad y clima predominantes fueron údico tipo I (330 a 365 días de humedad) y cálido húmedo (Am/A(C) m). Las plantas se ubicaron en altitudes de 200 a 1 190 msnm, en laderas accidentadas, que por lo general están al pie de sistemas montañosos de 1 300 a 2 500 metros de altitud. En condiciones naturales la distribución de la especie no se limita a selva alta perennifolia, dado que se ubicó en bosque mesófilo de montaña y bosque tropical perennifolio. La ubicación de nuevos especímenes de V. planifolia en condiciones silvestres reduce un 66 % del área potencial de distribución, y la fragmenta, al pasar de ser una zona continua a convertirse en tres zonas geográficamente separadas. La reducción del hábitat se debió a un aumento en el número de plantas ubicadas, lo que define las condiciones ambientales a un nivel más exacto. Por lo anterior, se pueden emprender o diseñar acciones de conservación enfocadas a áreas más específicas dentro del estado de Oaxaca, México.

AbstractWild specimens of Vanilla planifolia represent a vital part of this resource primary gene pool, and some plants have only been reported in Oaxaca, Mexico. For this reason, we studied its geographical distribution within the state, to locate and describe the ecological characteristics of the areas where they have been found, in order to identify potential areas of establishment. The method comprised four stages: 1) the creation of a database with herbarium records, 2) the construction of the potential distribution based on historical herbarium records for the species, using the model of maximum entropy (MaxEnt) and 22 bioclimatic variables as predictors; 3) an in situ systematic search of individuals, based on herbarium records and areas of potential distribution in 24 municipalities, to determine the habitat current situation and distribution; 4) the description of the environmental factors of potential ecological niches generated by MaxEnt. A review of herbarium collections revealed a total of 18 records of V. planifolia between 1939 and 1998. The systematic search located 28 plants distributed in 12 sites in 95 364 Km2. The most important variables that determined the model of vanilla potential distribution were: precipitation in the rainy season (61.9 %), soil moisture regime (23.4 %) and precipitation during the four months of highest rainfall (8.1 %). The species potential habitat was found to be distributed in four zones: wet tropics of the Gulf of Mexico, humid temperate, humid tropical, and humid temperate in the Pacific. Precipitation oscillated within the annual ranges of 2 500 to 4 000 mm, with summer rains, and winter precipitation as 5 to 10 % of the total. The moisture regime and predominating climate were udic type I (330 to 365 days of moisture) and hot humid (Am/A(C) m). The plants were located at altitudes of 200 to 1 190 masl, on rough hillsides that generally make up the foothills of mountain systems, with altitudes of 1 300 to 2 500 masl. In natural conditions, distribution of the species is not limited to high evergreen forests, since it was also found in mountain mesophyll and tropical evergreen forests. The location of new specimens of V. planifolia in its wild condition reduces the potential distribution area by 66 %. This area is fragmented into three geographically separated areas. Habitat reduction was due to the increased number of located plants that define the environmental conditions into a more accurate level. Conservation actions can thus be designed and implemented, focusing on more specific areas within the state of Oaxaca, Mexico.

[Potential distribution and geographic characteristics of wild populations of Vanilla planifolia (Orchidaceae) Oaxaca, Mexico]. / Distribución potencial y características geográficas de poblaciones silvestres de Vanilla planifolia (Orchidaceae) en Oaxaca, México.

Wild specimens of Vanilla planifolia represent a vital part of this resource primary gene pool, and some plants have only been reported in Oaxaca, Mexico. For this reason, we studied its geographical distribution within the state, to locate and describe the ecological characteristics of the areas where they have been found, in order to identify potential areas of establishment. The method comprised four stages: 1) the creation of a database with herbarium records, 2) the construction of the potential distribution based on historical herbarium records for the species, using the model of maximum entropy (MaxEnt) and 22 bioclimatic variables as predictors; 3) an in situ systematic search of individuals, based on herbarium records and areas of potential distribution in 24 municipalities, to determine the habitat current situation and distribution; 4) the description of the environmental factors of potential ecological niches generated by MaxEnt. A review of herbarium collections revealed a total of 18 records of V. planifolia between 1939 and 1998. The systematic search located 28 plants distributed in 12 sites in 95 364 Km(2). The most important variables that determined the model of vanilla potential distribution were: precipitation in the rainy season (61.9 %), soil moisture regime (23.4 %) and precipitation during the four months of highest rainfall (8.1 %). The species potential habitat was found to be distributed in four zones: wet tropics of the Gulf of Mexico, humid temperate, humid tropical, and humid temperate in the Pacific. Precipitation oscillated within the annual ranges of 2 500 to 4 000 mm, with summer rains, and winter precipitation as 5 to 10 % of the total. The moisture regime and predominating climate were udic type I (330 to 365 days of moisture) and hot humid (Am/A(C) m). The plants were located at altitudes of 200 to 1 190 masl, on rough hillsides that generally make up the foothills of mountain systems, with altitudes of 1 300 to 2 500 masl. In natural conditions, distribution of the species is not limited to high evergreen forests, since it was also found in mountain mesophyll and tropical evergreen forests. The location of new specimens of V. planifolia in its wild condition reduces the potential distribution area by 66 %. This area is fragmented into three geographically separated areas. Habitat reduction was due to the increased number of located plants that define the environmental conditions into a more accurate level. Conservation actions can thus be designed and implemented, focusing on more specific areas within the state of Oaxaca, Mexico.

Volatile composition and sensory properties of Vanilla × tahitensis bring new insights for vanilla quality control.

BACKGROUND: Vanilla × tahitensis produced in French Polynesia has a unique flavour among vanilla species. However, data on volatiles and sensory properties remain limited. In this study, the volatile composition and sensory properties of V. × tahitensis from three Polynesian cultivars and two origins (French Polynesia/Papua New Guinea) were determined by gas chromatography-mass spectrometry and quantitative descriptive analysis, respectively, and compared to Vanilla planifolia. RESULTS: Vanilla species, origins and cultivars were differentiated by their volatile and sensory profiles using principal component analysis. The V. × tahitensis flavour from French Polynesia was characterized by a well-balanced sensory profile, having strong anise and caramel notes due to high levels of anisyl compounds. V. × tahitensis from Papua New Guinea was distinct from that of French Polynesia, having strong spicy, fruity, brown rum notes due to p-vinylguaiacol, p-cresol and esters. Vanilla planifolia showed stronger phenolic, woody, smoky notes due to guaiacol, creosol and phenol, which were found to be biomarkers of the species. Vanilla sensory properties were linked by partial least squares regression to key volatile compounds like guaiacol or creosol, which are indicators of lower quality. CONCLUSION: This study brings new insights to vanilla quality control, with a focus on key volatile compounds, irrespective of origin.

Authenticity and traceability of vanilla flavors by analysis of stable isotopes of carbon and hydrogen.

Authenticity and traceability of vanilla flavors were investigated using gas chromatography-isotope ratio mass spectrometry (GC-IRMS). Vanilla flavors produced by chemical synthesis (n = 2), fermentation (n = 1), and extracted from two different species of the vanilla orchid (n = 79) were analyzed. The authenticity of the flavor compound vanillin was evaluated on the basis of measurements of ratios of carbon stable isotopes (δ(13)C). It was found that results of δ(13)C for vanillin extracted from Vanilla planifolia and Vanilla tahitensis were significantly different (t test) and that it was possible to differentiate these two groups of natural vanillin from vanillin produced otherwise. Vanilla flavors were also analyzed for ratios of hydrogen stable isotopes (δ(2)H). A graphic representation of δ(13)C versus δ(2)H revealed that vanillin extracted from pods grown in adjacent geographic origins grouped together. Accordingly, values of δ(13)C and δ(2)H can be used for studies of authenticity and traceability of vanilla flavors.

Comparative analysis of volatiles in traditionally cured Bourbon and Ugandan vanilla bean ( Vanilla planifolia ) extracts.

Traditionally cured vanilla beans ( Vanilla planifolia ) from Madagascar and Uganda were extracted with organic solvents, and the volatiles were separated from the nonvolatile fraction using the solvent assisted flavor evaporation (SAFE) technique. Concentrated vanilla bean extracts were analyzed using GC-MS and GC-O. Two hundred and forty-six volatile compounds were identified using the Automated Mass Spectral Deconvolution and Identification System (AMDIS) software, of which 13 were confirmed with authentic compounds from commercial sources and the others were tentatively identified on the basis of calibrated linear retention indices and the comparison of deconvoluted mass spectra with the in-house and/or NIST spectra databases. Vanillin was the most abundant constituent followed by guaiacol. The total concentration of the volatile compounds, excluding vanillin, was 301 mg/kg for Bourbon and 398 mg/kg for Ugandan vanilla bean extracts. Analytical comparison between the two vanilla bean extracts was discussed. Seventy-eight compounds were identified as odor-active compounds in the vanilla bean extracts with 10 confirmed with authentic references. It was found that there were substantial analytical differences in the odor-active compounds of the two extracts.

Evidence of transoceanic dispersion of the genus Vanilla based on plastid DNA phylogenetic analysis.

The phylogeny and the biogeographical history of the genus Vanilla was investigated using four chloroplastic genes (psbB, psbC; psaB and rbcL), on 47 accessions of Vanilla chosen from the ex situ CIRAD collection maintained in Reunion Island and additional sequences from GenBank. Bayesian methods provided a fairly well supported reconstruction of the phylogeny of the Vanilloideae sub-family and more particularly of the genus Vanilla. Three major phylogenetic groups in the genus Vanilla were differentiated, which is in disagreement with the actual classification in two sections (Foliosae and Aphyllae) based on morphological traits. Recent Bayesian relaxed molecular clock methods allowed to test the two main hypotheses of the phylogeography of the genus Vanilla. Early radiation of the Vanilla genus and diversification by vicariance consecutive to the break-up of Gondwana, 95 million years ago (Mya), was incompatible with the admitted age of origin of Angiosperm. Based on the Vanilloideae age recently estimated to 71 million years ago (Mya), we conclude that the genus Vanilla would have appeared approximately 34 Mya in South America, when continents were already separated. Nevertheless, whatever the two extreme scenarios tested, at least three long distance migration events are needed to explain the present distribution of Vanilla species in tropical areas. These transoceanic dispersions could have occurred via transoceanic passageway such as the Rio Grande Ridge and the involvement of floating vegetation mats and migratory birds.

Evaluation of chemical variability of cured vanilla beans (Vanilla tahitensis and Vanilla planifolia).

In order to establish a chemical fingerprint of vanilla diversity, thirty samples of V. planifolia J. W. Moore and V. tahitensis G. Jackson cured beans from seven producing countries were examined for their aroma and fatty acid contents. Both fatty acid and aroma compositions were found to vary between vanilla species and origins. Vanillin was found in higher amounts in V. planifolia (1.7-3.6% of dry matter) than in V. tahitensis (1.0-2.0%), and anisyl compounds were found in lower amounts in V. planifolia (0.05%) than in V. tahitensis (1.4%-2.1%). Ten common and long chain monounsaturated fatty acids (LCFA) were identified and were found to be characteristic of the vanilla origin. LCFA derived from secondary metabolites have discriminating compositions as they reach 5.9% and 15.8% of total fatty acids, respectively in V. tahitensis and V. planifolia. This study highlights the role of the curing method as vanilla cured beans of two different species cultivated in the same country were found to have quite similar fatty acid compositions.

On the value of nuclear and mitochondrial gene sequences for reconstructing the phylogeny of vanilloid orchids (Vanilloideae, Orchidaceae).

BACKGROUND AND AIMS: Most molecular phylogenetic studies of Orchidaceae have relied heavily on DNA sequences from the plastid genome. Nuclear and mitochondrial loci have only been superficially examined for their systematic value. Since 40% of the genera within Vanilloideae are achlorophyllous mycoheterotrophs, this is an ideal group of orchids in which to evaluate non-plastid gene sequences. METHODS: Phylogenetic reconstructions for Vanilloideae were produced using independent and combined data from the nuclear 18S, 5.8S and 26S rDNA genes and the mitochondrial atpA gene and nad1b-c intron. KEY RESULTS: These new data indicate placements for genera such as Lecanorchis and Galeola, for which plastid gene sequences have been mostly unavailable. Nuclear and mitochondrial parsimony jackknife trees are congruent with each other and previously published trees based solely on plastid data. Because of high rates of sequence divergence among vanilloid orchids, even the short 5.8S rDNA gene provides impressive levels of resolution and support. CONCLUSIONS: Orchid systematists are encouraged to sequence nuclear and mitochondrial gene regions along with the growing number of plastid loci available.