Dataset Information

Going deeper in the automated identification of Herbarium specimens.

ABSTRACT: Hundreds of herbarium collections have accumulated a valuable heritage and knowledge of plants over several centuries. Recent initiatives started ambitious preservation plans to digitize this information and make it available to botanists and the general public through web portals. However, thousands of sheets are still unidentified at the species level while numerous sheets should be reviewed and updated following more recent taxonomic knowledge. These annotations and revisions require an unrealistic amount of work for botanists to carry out in a reasonable time. Computer vision and machine learning approaches applied to herbarium sheets are promising but are still not well studied compared to automated species identification from leaf scans or pictures of plants in the field.In this work, we propose to study and evaluate the accuracy with which herbarium images can be potentially exploited for species identification with deep learning technology. In addition, we propose to study if the combination of herbarium sheets with photos of plants in the field is relevant in terms of accuracy, and finally, we explore if herbarium images from one region that has one specific flora can be used to do transfer learning to another region with other species; for example, on a region under-represented in terms of collected data.This is, to our knowledge, the first study that uses deep learning to analyze a big dataset with thousands of species from herbaria. Results show the potential of Deep Learning on herbarium species identification, particularly by training and testing across different datasets from different herbaria. This could potentially lead to the creation of a semi, or even fully automated system to help taxonomists and experts with their annotation, classification, and revision works.

SUBMITTER: Carranza-Rojas J

PROVIDER: S-EPMC5553807 | biostudies-other | 2017 Aug

REPOSITORIES: biostudies-other

ACCESS DATA

Similar Datasets

Project description:The paucity of sequence data from pelagic deep-ocean microbial assemblages has severely restricted molecular exploration of the largest biome on Earth. In this study, an analysis is presented of a large-scale 454-pyrosequencing metagenomic dataset from a hadopelagic environment from 6,000 m depth within the Puerto Rico Trench (PRT). A total of 145 Mbp of assembled sequence data was generated and compared to two pelagic deep ocean metagenomes and two representative surface seawater datasets from the Sargasso Sea. In a number of instances, all three deep metagenomes displayed similar trends, but were most magnified in the PRT, including enrichment in functions for two-component signal transduction mechanisms and transcriptional regulation. Overrepresented transporters in the PRT metagenome included outer membrane porins, diverse cation transporters, and di- and tri-carboxylate transporters that matched well with the prevailing catabolic processes such as butanoate, glyoxylate and dicarboxylate metabolism. A surprisingly high abundance of sulfatases for the degradation of sulfated polysaccharides were also present in the PRT. The most dramatic adaptational feature of the PRT microbes appears to be heavy metal resistance, as reflected in the large numbers of transporters present for their removal. As a complement to the metagenome approach, single-cell genomic techniques were utilized to generate partial whole-genome sequence data from four uncultivated cells from members of the dominant phyla within the PRT, Alphaproteobacteria, Gammaproteobacteria, Bacteroidetes and Planctomycetes. The single-cell sequence data provided genomic context for many of the highly abundant functional attributes identified from the PRT metagenome, as well as recruiting heavily the PRT metagenomic sequence data compared to 172 available reference marine genomes. Through these multifaceted sequence approaches, new insights have been provided into the unique functional attributes present in microbes residing in a deeper layer of the ocean far removed from the more productive sun-drenched zones above.

Project description:Background and aimsHabitat degradation and landscape fragmentation dramatically lower population sizes of rare plant species. Decreasing population sizes may, in turn, negatively affect genetic diversity and reproductive fitness, which can ultimately lead to local extinction of populations. Although such extinction vortex dynamics have been postulated in theory and modelling for decades, empirical evidence from local extinctions of plant populations is scarce. In particular, comparisons between current vs. historical genetic diversity and differentiation are lacking despite their potential to guide conservation management.MethodsWe studied the population genetic signatures of the local extinction of Biscutella laevigata subsp. gracilis populations in Central Germany. We used microsatellites to genotype individuals from 15 current populations, one ex situ population, and 81 herbarium samples from five extant and 22 extinct populations. In the current populations, we recorded population size and fitness proxies, collected seeds for a germination trial and conducted a vegetation survey. The latter served as a surrogate for habitat conditions to study how habitat dissimilarity affects functional connectivity among the current populations.Key resultsBayesian clustering revealed similar gene pool distribution in current and historical samples but also indicated that a distinct genetic cluster was significantly associated with extinction probability. Gene flow was affected by both the spatial distance and floristic composition of population sites, highlighting the potential of floristic composition as a powerful predictor of functional connectivity which may promote decision-making for reintroduction measures. For an extinct population, we found a negative relationship between sampling year and heterozygosity. Inbreeding negatively affected germination.ConclusionsOur study illustrates the usefulness of historical DNA to study extinction vortices in threatened species. Our novel combination of classical population genetics together with data from herbarium specimens, an ex situ population and a germination trial underlines the need for genetic rescue measures to prevent extinction of B. laevigata in Central Germany.

Dataset Information

Going deeper in the automated identification of Herbarium specimens.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets