Genetic diversity among economically important bamboo species of India as revealed by ISSR markers

Bamboo taxonomy rests largely on vegetative and culm-sheath characters, which are diagnostic on mature culms but show marked phenotypic plasticity across sites and culm ages, while reproductive characters are seldom available because flowering is gregarious, unpredictable and separated by cycles of 3–120 years. Molecular markers therefore provide a complementary rather than a substitute route to characterisation. In the present study, an inter-simple sequence repeat (ISSR) marker system employing 21 primers was used to study the genetic relationships among 15 bamboo accessions representing 13 species in four genera, collected from different regions of India. The 21 primers amplified 316 scoreable bands, of which six primers yielded 15 or more bands each; the maximum was recorded for UBC826 (20 bands), followed by UBC827 (19). Polymorphism reached 99.7 %, and polymorphism information content averaged 0.312 per primer against a theoretical maximum of 0.5 for a dominant marker. Similarity coefficients based on simple matching varied between 0.52 and 0.96. The phylogenetic tree based on the genetic similarity matrix, constructed using the unweighted pair group method with arithmetic average (UPGMA) algorithm, resolved the 15 accessions into seven clusters. The highest similarity (0.96) was recorded between Arundinaria falcata and the unidentified hill bamboo accession, supporting its provisional placement in Arundinaria, followed by Bambusa vulgaris var. yellow and var. wamin (0.81). Lowest genetic similarity was observed between B. tulda and D. strictus (0.50), followed by B. tulda and B. balcooa (0.57). Clustering was broadly, though not entirely, congruent with the conventional classification, the six Bambusa accessions other than B. tulda forming a single group. Because ISSR markers are dominant and woody bamboos are predominantly polyploid, the results are interpreted as phenetic similarity among clonal accessions rather than as allele-frequency-based population diversity. Results suggest the possibility of selection among, and hybridisation between, accessions drawn from divergent clusters.