Molecular taxonomic identification in the absence of a ‘barcoding gap’: a test with the endemic flora of the Canarian oceanic hotspot


We use a comprehensive subset of Canarian angiosperms corresponding to 23 families, 35 genera and 60 Canarian endemic taxa to test whether this flora is suitable to taxonomic identification with the two proposed plant DNA barcode sequences and whether these sequences may reveal the existence of cryptic species overlooked by morphology. The rate of discrimination success between the insular congeneric samples using the rbcL+matK combination and a ‘character-based’ approach (where we use only the combination of nucleotide positions in an alignment that allows unambiguous species identification) is higher (82.29%) than that obtained with the ‘distance-based’ approach (80.20%) used by the CBOL Plant Working Group in 2009 and also when compared with tests conducted in other floras.

This suggests that the molecular identification of the Canarian endemic flora can be achieved as successfully as in other floras where the incidence of radiation is not as relevant. The facts that (i) a distance-based criterion was unable to discriminate between congeneric and conspecific comparisons and (ii) only the character-based discrimination criterion resolved cases that the distance-based criterion did not, further support the use of a character discrimination approach for a more efficient DNA barcoding of floras from oceanic islands like the Canaries. Thus, a barcoding gap seems not to be necessary for the correct molecular characterization of the Canarian flora.

DNA barcodes also suggest the possible existence of cryptic taxa to be further investigated by morphology and that the current taxonomic status of some of the taxa analysed may need revision.


To provide an universal framework for the routine use of DNA sequence data in species identification is important from a wide range of biological perspectives, encompassing (i) the rapid assignation of specimens to their correct species when morphological traits are not sufficient (Kress et al. 2005; Savolainen et al. 2005); (ii) the identification of possible cryptic species overlooked by morphology (Hebert et al. 2003, 2004; Ragupathy et al. 2009; Gao et al. 2011); (iii) the assistance to classical taxonomy in the elaboration of censuses of plant biodiversity (Hajibabaei et al. 2007; Lahaye et al. 2008; de Vere et al. 2012) or (iv) the application of these data to conservation and management strategies (Hollingsworth 2008; Kress et al. 2009; Bruni et al. 2012). These and other distinctive areas of interest have triggered ‘DNA barcoding’ for species-level identification (i.e. the use of short-standardized DNA sequences to tag species; Barret & Hebert 2005).



*Jardín Botánico Canario “Viera y Clavijo”-Unidad Asociada CSIC, Cabildo de Gran Canaria, Apartado de correos 14 de Tafira Alta, 35017, Las Palmas de Gran Canaria, Spain, †Fundación Canaria Amurga Maspalomas, Avda. Tirajana 39, II 6, 35100 San Bartolomé de Tirajana, Spain, ‡Unidad de Botánica Aplicada, Jardín de Aclimatación de La Orotava (ICIA), c/Retama 2, 38400. Puerto de la Cruz, Tenerife, Spain