Assessing DNA for fish identifications from reference collections: the good, bad and ugly shed light on formalin fixation and sequencing approaches

Australia's marine fishes DNA barcode reference library for integrated taxonomy, metabarcoding & eDNA research

Over 15 000 species of fishes are found globally in the marine environment and DNA barcodes are used extensively to describe, catalogue, understand and manage this diversity. The dataset outlined here represents a DNA barcode reference library of the mitochondrial cytochrome c oxidase subunit 1 gene (COI) from 9767 voucher specimens (representing at least 2220 species and 288 families) of marine fishes. This publicly available dataset in the Barcode of Life Data System (BOLD) represents 17 years (2005-2022) of barcoding of marine fishes identified from Australian territorial waters. Tissues targeted for sequencing with their matching physical specimens (and extracted DNA), obtained via a multi-agency sampling effort, are mostly maintained and curated by the CSIRO Australian National Fish Collection (ANFC) in Hobart, Australia. Species-level integrated taxonomy (assigned after combined morphological and genetic assessment) has been determined for 91% of the dataset. The library represents the most complete COI barcode reference dataset for marine fishes from Australian waters and is currently utilised for integrated taxonomy, (meta)barcoding and eDNA studies.

Mitochondrial genome data provide insights into the phylogenetic relationships within Triplophysadalaica (Kessler, 1876) (Cypriniformes, Nemacheilidae)

Due to the detrimental effect of formaldehyde on DNA, ethanol has replaced formalin as the primary preservative for animal specimens. However, short-term formalin fixation of specimens might be applied during field collection. In an increasing number of studies, DNA extraction and sequencing have been successfully conducted from formalin-fixed specimens. Here the DNA from five specimens of Triplophysadalaica (Kessler, 1876) were extracted and performed high-throughput sequencing. Four of the specimens underwent short-term fixation with formalin and were subsequently transferred to ethanol. One was continuously stored in ethanol. No significant difference of DNA quality and amount were observed among these samples. Followed by assembly and annotation, five mitochondrial genomes ranging in length from 16,569 to 16,572 bp were obtained. Additionally, previously published data of other individuals or species were included to perform phylogenetic analyses. In the reconstructed trees, all eight individuals of T.dalaica form a monophyletic group within the Triplophysa branch. The group is divided into three clades: (1) samples from the Yellow River, (2) those from the Yangtze River, and (3) those from the Haihe River, and the Lake Dali Nur. This study sheds initial light on the phylogeographic relationships among different populations of T.dalaica, and will support the research about its evolutionary history in the future.

SNP data reveals the complex and diverse evolutionary history of the blue-ringed octopus genus (Octopodidae: Hapalochlaena) in the Asia-Pacific

The blue-ringed octopus species complex (Hapalochlaena spp.), known to occur from Southern Australia to Japan, currently contains four formally described species (Hapalochlaena maculosa, Hapalochlaena fasciata, Hapalochlaena lunulata and Hapalochlaena nierstraszi). These species are distinguished based on morphological characters (iridescent blue rings and/or lines) along with reproductive strategies. However, the observation of greater morphological diversity than previously captured by the current taxonomic framework indicates that a revision is required. To examine species boundaries within the genus we used mitochondrial (12S rRNA, 16S rRNA, cytochrome c oxidase subunit 1 [COI], cytochrome c oxidase subunit 3 [COIII] and cytochrome b [Cytb]) and genome-wide SNP data (DaRT seq) from specimens collected across its geographic range including variations in depth from 3m to >100m. This investigation indicates substantially greater species diversity present within the genus Hapalochlaena than is currently described. We identified 10,346 SNPs across all locations, which when analysed support a minimum of 11 distinct clades. Bayesian phylogenetic analysis of the mitochondrial COI gene on a more limited sample set dates the diversification of the genus to ∼30mya and corroborates eight of the lineages indicated by the SNP analyses. Furthermore, we demonstrate that the diagnostic lined patterning of H. fasciata found in North Pacific waters and NSW, Australia is polyphyletic and therefore likely the result of convergent evolution. Several "deep water" (> 100m) lineages were also identified in this study with genetic convergence likely to be driven by external selective pressures. Examination of morphological traits, currently being undertaken in a parallel morphological study, is required to describe additional species within the complex.

Maximizing Molecular Data From Low-Quality Fluid-Preserved Specimens in Natural History Collections

Over the past decade, museum genomics studies have focused on obtaining DNA of sufficient quality and quantity for sequencing from fluid-preserved natural history specimens, primarily to be used in systematic studies. While these studies have opened windows to evolutionary and biodiversity knowledge of many species worldwide, published works often focus on the success of these DNA sequencing efforts, which is undoubtedly less common than obtaining minimal or sometimes no DNA or unusable sequence data from specimens in natural history collections. Here, we attempt to obtain and sequence DNA extracts from 115 fresh and 41 degraded samples of homalopsid snakes, as well as from two degraded samples of a poorly known snake, Hydrablabes periops. Hydrablabes has been suggested to belong to at least two different families (Natricidae and Homalopsidae) and with no fresh tissues known to be available, intractable museum specimens currently provide the only opportunity to determine this snake’s taxonomic affinity. Although our aim was to generate a target-capture dataset for these samples, to be included in a broader phylogenetic study, results were less than ideal due to large amounts of missing data, especially using the same downstream methods as with standard, high-quality samples. However, rather than discount results entirely, we used mapping methods with references and pseudoreferences, along with phylogenetic analyses, to maximize any usable molecular data from our sequencing efforts, identify the taxonomic affinity of H. periops, and compare sequencing success between fresh and degraded tissue samples. This resulted in largely complete mitochondrial genomes for five specimens and hundreds to thousands of nuclear loci (ultra-conserved loci, anchored-hybrid enrichment loci, and a variety of loci frequently used in squamate phylogenetic studies) from fluid-preserved snakes, including a specimen of H. periops from the Field Museum of Natural History collection. We combined our H. periops data with previously published genomic and Sanger-sequenced datasets to confirm the familial designation of this taxon, reject previous taxonomic hypotheses, and make biogeographic inferences for Hydrablabes. A second H. periops specimen, despite being seemingly similar for initial raw sequencing results and after being put through the same protocols, resulted in little usable molecular data. We discuss the successes and failures of using different pipelines and methods to maximize the products from these data and provide expectations for others who are looking to use DNA sequencing efforts on specimens that likely have degraded DNA.

Life Science Identifier (Hydrablabes periops)

urn:lsid:zoobank.org:pub:F2AA44 E2-D2EF-4747-972A-652C34C2C09D.

Unlocking inaccessible historical genomes preserved in formalin

Museum specimens represent an unparalleled record of historical genomic data. However, the widespread practice of formalin preservation has thus far impeded genomic analysis of a large proportion of specimens. Limited DNA sequencing from formalin-preserved specimens has yielded low genomic coverage with unpredictable success. We set out to refine sample processing methods and to identify specimen characteristics predictive of sequencing success. With a set of taxonomically diverse specimens collected between 1962 and 2006 and ranging in preservation quality, we compared the efficacy of several end-to-end whole genome sequencing workflows alongside a k-mer-based trimming-free read alignment approach to maximize mapping of endogenous sequence. We recovered complete mitochondrial genomes and up to 3× nuclear genome coverage from formalin-preserved tissues. Hot alkaline lysis coupled with phenol-chloroform extraction out-performed proteinase K digestion in recovering DNA, while library preparation method had little impact on sequencing success. The strongest predictor of DNA yield was overall specimen condition, which additively interacts with preservation conditions to accelerate DNA degradation. Here, we demonstrate a significant advance in capability beyond limited recovery of a small number of loci via PCR or target-capture sequencing. To facilitate strategic selection of suitable specimens for genomic sequencing, we present a decision-making framework that utilizes independent and nondestructive assessment criteria. Sequencing of formalin-preserved specimens will contribute to a greater understanding of temporal trends in genetic adaptation, including those associated with a changing climate. Our work enhances the value of museum collections worldwide by unlocking genomes of specimens that have been disregarded as a valid molecular resource.

Mining museums for historical DNA: advances and challenges in museomics

Historical DNA (hDNA), obtained from museum and herbarium specimens, has yielded spectacular new insights into the history of organisms. This includes documenting historical genetic erosion and extinction, discovering species new to science, resolving evolutionary relationships, investigating epigenetic effects, and determining origins of infectious diseases. However, the development of best-practices in isolating, processing, and analyzing hDNA remain under-explored, due to the substantial diversity of specimen preparation types, tissue sources, archival ages, and collecting histories. Thus, for hDNA to reach its full potential, and justify the destructive sampling of the rarest specimens, more experimental work using time-series collections, and the development of improved methods to correct for data asymmetries and biases due to DNA degradation are required.