Handbook of Vanilla Science and Technology

Only one study has applied the resolving power of molecular systematics to characterize mycorrhizal fungi of Vanilla (Porras-Alfaro and Bayman 2007). The objectives of the study were to test differences in specificity of Vanilla for fungi at several levels: between roots in soil vs. roots on bark, between fungi isolated in pure culture vs. fungal DNA extracted directly from roots, and among species of Vanilla.

16.2 METHODS

Porras-Alfaro and Bayman (2007) collected roots of wild V. planifolia and V. poitaei in El Verde and Cambalache, Puerto Rico, roots of cultivated V. phaeantha and cultivated V. planifolia × pompona were collected in Quepos, Costa Rica, and roots of wild V. aphylla were collected in Peninsula Guanahacabibes, Cuba (Figures 16.1 and 16.3). Samples included both roots in soil and aerial roots attached to tree bark. Root sections with mycorrhizal pelotons (knots of fungal hyphae in root cortical cells, characteristic of orchid mycorrhizae; see Figure 16.2) were cultured on potato dextrose agar with antibiotics. DNA was extracted from pure cultures and also directly from mycorrhizal root tissue. The nuclear ribosomal ITS (internal transcribed spacer) region region was amplified by PCR and sequenced. This region was chosen because it is easy to amplify, phylogenetically informative, and well-represented in GenBank; it is currently the preferred candidate for DNA bar-coding of fungal species. DNA sequences were used for BLAST searches in GenBank, to identify the most similar sequence already in the database. Phylogenetic trees were constructed using both parsimony and likelihood algorithms as described (Porras-Alfaro and Bayman 2007).

Fig. 16.3 Vanilla grower Neal Byrd and Andrea-Porras Alfaro with V. planifolia × pompona plant at Cia. Agricola La Gavilana, Quepos, Costa Rica.

16.3 DIVERSITY OF MYCORRHIZAL FUNGI

Vanilla roots were colonized by at least three genera: Ceratobasidium, Thanatephorus, and Tulasnella (Figure 16.4; Porras-Alfaro and Bayman 2007). The Tulasnella sequences (all from Puerto Rico) were most similar to sequences from fungi from terrestrial orchids in Singapore (Ma et al. 2003). The Thanatephorus sequences (from Puerto Rico and Cuba) were most closely related to a sequence from Thanatephorus praticola from the United Kingdom. The Ceratobasidium sequences (from Puerto Rico and Costa Rica) were most closely related to fungi isolated from soil and plants in the United States and Japan (González et al. 2001). Figure 16.4 suggests that Ceratobasidium is paraphyletic and Thanatephorus as currently defined comprises several distinct clades within Ceratobasidium, which agrees with previous studies (González et al. 2001). A distinct Ceratobasidium has since been isolated from V. planifolia in Colombia (Mosquera-Espinosa, unpublished).

Fig. 16.4 Phylogenetic affiliations of Rhizoctonia-like fungi from Vanilla spp. based on nuclear ribosomal ITS sequences. Host species and collecting sites are shown at right. Sequences from Tulasnella and Thanatephorus are indicated; the remaining sequences are Ceratobasidium, which appears to be para-phyletic. Maximum parsimonywith 100 bootstrap replicates produced several equally parsimonious trees. Topologies of principal cladeswere identicalwhen likelihood was used. Bootstrapvalues > 70%are shown. Puccinia boroniae was used as the outgroup; otheroutgroups produced identical topologies. Length = 191 steps, CI — 0.71, RI — 0.95, RC — 0.68. From Porras-Alfaro 2004, Porras-Alfaro and Bayman 2007.

The Thanatephorus sequences from V. poitaei and V. aphylla all came from direct amplifications, not pure cultures (Porras-Alfaro and Bayman 2007). Cultures isolated from these roots had extremely limited growth in agar media or no growth at all. Addition of

vitamins, plant extracts, and other additives to the agar medium did not stimulate growth. These fungi appear to have an unknown nutritional requirement, something not previously reported for Rhizoctonia. These fungi would be overlooked in studies based solely on culturing (Bayman 2006). For this reason, frequency of fungal genera differed significantly when pure cultures were sequenced vs. direct sequencing from roots; choice of sampling method determines which fungi are detected.

16.4 DISTRIBUTION OF MYCORRHIZAE