DNA Sequencing Historical Lichen Specimens

Biological specimens in natural history collections worldwide are increasingly being used in biogeographical, environmental, and taxonomic studies. For their meaningful use, correct species identification is crucial. For example, clarifying if a species is new to science requires an overview of what has already been described. This includes comparisons with existing authoritative specimens (types). Most type specimens are rather old and their DNA expected to be degraded to various extents. Comparative DNA sequence analysis is in regular use in taxonomic research of today and is essential for identifying and delimiting species. In this study, we focus on lichenized fungi (lichens), in which many species groups are highly inconspicuous and impossible to identify to species based on morphology alone. Our aim was to test the non-mutually exclusive hypotheses that DNA quality of lichens depends on (1) time since collection, (2) taxonomic affinity, and/or (3) habitat/ecology. We included two species from each of four different lichen genera (i.e., Cladonia, Nephroma, Peltigera and Ramalina), each species pair with a different autecology. For each species, we included samples from approximately every 25 years from present to about 150 years back in time. We used a two-step PCR-based approach followed by sequencing on an Ion Torrent PGM to produce target sequences (mtSSU) of degraded DNA. We received satisfactory DNA sequence information for 54 of 56 specimens. We recovered full-length sequences for several more than 100-year-old specimens, including a 127-year-old specimen, and retrieved enough sequence information for species identification of a 150-year-old specimen. As expected, sequencing success was negatively correlated with age of the specimens. It also varied with taxonomic affinity. We found no significant correlation between sequencing success and habitat ecology of the investigated specimens. The herein tested Ion Torrent sequencing approach outperformed Sanger sequencing with regard to sequencing success and efficiency. We find the protocol used herein highly suitable for obtaining sequences from both young and old lichen specimens and discuss potential improvements to it.