Identification of herbarium specimen sheet components from high-resolution images using deep learning.

Advanced computer vision techniques hold the potential to mobilise vast quantities of biodiversity data by facilitating the rapid extraction of text- and trait-based data from herbarium specimen digital images, and to increase the efficiency and accuracy of downstream data capture during digitisation. This investigation developed an object detection model using YOLOv5 and digitised collection images from the University of Melbourne Herbarium (MELU). The MELU-trained 'sheet-component' model-trained on 3371 annotated images, validated on 1000 annotated images, run using 'large' model type, at 640 pixels, for 200 epochs-successfully identified most of the 11 component types of the digital specimen images, with an overall model precision measure of 0.983, recall of 0.969 and moving average precision (mAP0.5-0.95) of 0.847. Specifically, 'institutional' and 'annotation' labels were predicted with mAP0.5-0.95 of 0.970 and 0.878 respectively. It was found that annotating at least 2000 images was required to train an adequate model, likely due to the heterogeneity of specimen sheets. The full model was then applied to selected specimens from nine global herbaria (Biodiversity Data Journal, 7, 2019), quantifying its generalisability: for example, the 'institutional label' was identified with mAP0.5-0.95 of between 0.68 and 0.89 across the various herbaria. Further detailed study demonstrated that starting with the MELU-model weights and retraining for as few as 50 epochs on 30 additional annotated images was sufficient to enable the prediction of a previously unseen component. As many herbaria are resource-constrained, the MELU-trained 'sheet-component' model weights are made available and application encouraged.

A field-based investigation of simple phenol variation in Australian Agaricus xanthodermus.

Agaricus xanthodermus and other species of the yellow-staining section Xanthodermatei are responsible for mushroom-related poisoning cases that require treatment. However, longstanding anecdotal evidence indicates that this species appears to exhibit considerable variation in toxicity, resulting in gastrointestinal irritation of varying severity in most cases. We quantified the amount of phenol, hydroquinone, and catechol in mushrooms using a novel protocol for gas chromatography-mass spectrometry (GC-MS) and investigated their levels in different basidiomatal structures, different developmental stages, and on different nutritional substrates. Phenol concentration was greater in the pileus than the stipe, in mature compared with immature basidiomata, and in basidiomata occurring in woody mulch versus lawns. Variation in toxicity is suggested to be due in part to variation in phenol concentration, developmental stage and tissue type consumed, and substrate. Variation in human sensitivity to simple phenols may also play a role but was not formally investigated in this study.

New targets acquired: Improving locus recovery from the Angiosperms353 probe set.

PREMISE: Universal target enrichment kits maximize utility across wide evolutionary breadth while minimizing the number of baits required to create a cost-efficient kit. The Angiosperms353 kit has been successfully used to capture loci throughout the angiosperms, but the default target reference file includes sequence information from only 6-18 taxa per locus. Consequently, reads sequenced from on-target DNA molecules may fail to map to references, resulting in fewer on-target reads for assembly, and reducing locus recovery. METHODS: We expanded the Angiosperms353 target file, incorporating sequences from 566 transcriptomes to produce a 'mega353' target file, with each locus represented by 17-373 taxa. This mega353 file is a drop-in replacement for the original Angiosperms353 file in HybPiper analyses. We provide tools to subsample the file based on user-selected taxon groups, and to incorporate other transcriptome or protein-coding gene data sets. RESULTS: Compared to the default Angiosperms353 file, the mega353 file increased the percentage of on-target reads by an average of 32%, increased locus recovery at 75% length by 49%, and increased the total length of the concatenated loci by 29%. DISCUSSION: Increasing the phylogenetic density of the target reference file results in improved recovery of target capture loci. The mega353 file and associated scripts are available at: https://github.com/chrisjackson-pellicle/NewTargets.

Biogeography of the monocotyledon astelioid clade (Asparagales): A history of long-distance dispersal and diversification with emerging habitats.

The astelioid families (Asteliaceae, Blandfordiaceae, Boryaceae, Hypoxidaceae, and Lanariaceae) have centers of diversity in Australasia and temperate Africa, with secondary centers of diversity in Afromontane Africa, Asia, and Pacific Islands. The global distribution of these families makes this an excellent lineage to test if current distribution patterns are the result of vicariance or long-distance dispersal and to evaluate the roles of Tertiary climatic and geological drivers in lineage diversification. Sequence data were generated from five chloroplast regions (petL-psbE, rbcL, rps16-trnK, trnL-trnLF, trnS-trnSG) for 104 ingroup species sampled across global diversity. The astelioid phylogeny was inferred using maximum parsimony, maximum likelihood, and Bayesian inference methods. Divergence dates were estimated with a relaxed clock applied in BEAST. Ancestral ranges were reconstructed in the R package 'BioGeoBEARS' applying the corrected Akaike information criterion to test for the best-fit biogeographic model. Diversification rates were estimated in Bayesian Analysis of Macroevolutionary Mixtures (BAMM). Astelioid relationships were inferred as Boryaceae(Blandfordiaceae(Asteliaceae(Hypoxidaceae plus Lanariaceae))). The crown astelioid node was dated to the Late Cretaceous (75.2 million years; 95% highest posterior density interval 61.0-90.0 million years) and an Antarctic-Australasian origin was inferred. Astelioid speciation events have not been shaped by Gondwanan vicariance. Rather long-distance dispersal since the Eocene is inferred to account for current distributions. Crown Asteliaceae and Boryaceae have Australian ancestral ranges and diversified since the Eocene. In Hypoxidaceae, Empodium, Hypoxis, and Pauridia have African ancestral ranges, while Curculigo and Molineria have an Asian ancestral range. Diversification of Pauridia and the Curculigo clade occurred steadily, while diversification of Astelia and Hypoxis was punctuated over time. Diversification of Hypoxis and Astelia coincided temporally with the expansion of the habitat types occupied by extant taxa, e.g., grassland habitat in Africa during the Late Miocene and alpine habitat in New Zealand during the Pliocene, respectively.

Evolution of Lomandroideae: Multiple origins of polyploidy and biome occupancy in Australia.

Asparagaceae: Lomandroideae are a species-rich and economically important subfamily in the monocot order Asparagales, with a center of diversity in Australia. Lomandroideae are ecologically diverse, occupying mesic and arid biomes in Australia and possessing an array of key traits, including sexual dimorphism, storage organs and polyploidy that are potentially adaptive for survival in seasonally arid and fire-dependent habitats. The Lomandroideae phylogeny was reconstructed using maximum likelihood and Bayesian inference criteria, based on plastome data from genome-skimming to infer relationships. A fossil-calibrated chronogram provided a temporal framework for understanding trait transitions. Ancestral state reconstructions and phylogenetic comparative trait correlation analyses provided insights into the evolutionary and ecological drivers associated with Lomandroideae diversification. Lomandroideae diverged from the other Asparagaceae ca. 56.61 million years ago (95% highest posterior density values 70.31-45.34 million years) and the major lineages diversified since the Oligocene. The most recent common ancestor of the clade likely occupied the mesic biome, was hermaphroditic and geophytic. Biome occupancy transitions were correlated with polyploidy and the presence of storage roots. Polyploidy potentially serves as an "enabler" trait, generating novel phenotypes, which may confer tolerance to climatic ranges and soil conditions putatively required for expansion into and occupation of new arid biomes. Storage roots, as a key factor driving biome transitions, may have been associated with fire rather than with aridification events in the Australian flora. This study contributes significantly to our understanding of biome evolution by identifying polyploidy and storage organs as key factors associated with transitions in biome occupancy in this lineage.

Testing efficacy of distance and tree-based methods for DNA barcoding of grasses (Poaceae tribe Poeae) in Australia.

In Australia, Poaceae tribe Poeae are represented by 19 genera and 99 species, including economically and environmentally important native and introduced pasture grasses [e.g. Poa (Tussock-grasses) and Lolium (Ryegrasses)]. We used this tribe, which are well characterised in regards to morphological diversity and evolutionary relationships, to test the efficacy of DNA barcoding methods. A reference library was generated that included 93.9% of species in Australia (408 individuals, [Formula: see text] = 3.7 individuals per species). Molecular data were generated for official plant barcoding markers (rbcL, matK) and the nuclear ribosomal internal transcribed spacer (ITS) region. We investigated accuracy of specimen identifications using distance- (nearest neighbour, best-close match, and threshold identification) and tree-based (maximum likelihood, Bayesian inference) methods and applied species discovery methods (automatic barcode gap discovery, Poisson tree processes) based on molecular data to assess congruence with recognised species. Across all methods, success rate for specimen identification of genera was high (87.5-99.5%) and of species was low (25.6-44.6%). Distance- and tree-based methods were equally ineffective in providing accurate identifications for specimens to species rank (26.1-44.6% and 25.6-31.3%, respectively). The ITS marker achieved the highest success rate for specimen identification at both generic and species ranks across the majority of methods. For distance-based analyses the best-close match method provided the greatest accuracy for identification of individuals with a high percentage of "correct" (97.6%) and a low percentage of "incorrect" (0.3%) generic identifications, based on the ITS marker. For tribe Poeae, and likely for other grass lineages, sequence data in the standard DNA barcode markers are not variable enough for accurate identification of specimens to species rank. For recently diverged grass species similar challenges are encountered in the application of genetic and morphological data to species delimitations, with taxonomic signal limited by extensive infra-specific variation and shared polymorphisms among species in both data types.

Corrigenda: A revision of infrageneric classification in Astelia Banks & Sol. ex R.Br. (Asteliaceae).

[This corrects the article DOI: 10.3897/phytokeys.52.4768.].

A revision of infrageneric classification in Astelia Banks & Sol. ex R.Br. (Asteliaceae).

Systematic investigations and phylogenetic analyses have indicated that Astelia, as currently circumscribed, is paraphyletic, with Collospermum nested within it. Further, AsteliasubgenusAstelia is polyphyletic, and Astelia subgenera Asteliopsis and Tricella are paraphyletic, as currently circumscribed. Revision of the subgeneric classification of Astelia is warranted to ensure classification accurately reflects the evolutionary history of these taxa. Collospermum is relegated to synonymy within Astelia. Astelia is dioecious or polygamodioecious, with a superior ovary, anthers dorsi- or basifixed, pistillodes or pistils that have a single short or poorly defined style, a 3 lobed stigma, and fleshy uni- or trilocular fruit with funicular hairs that are poorly to well developed. AsteliasubgenusCollospermum (Skottsb.) Birch is described. A key to Astelia sections is provided. Asteliahastata Colenso, Asteliamontana Seem., and Asteliamicrosperma Colenso pro parte are resurrected and the new combination Asteliasamoense (Skottsb.) Birch, comb. nov. is made.

Molecular phylogeny and dating of Asteliaceae (Asparagales): Astelia s.l. evolution provides insight into the Oligocene history of New Zealand.

Asteliaceae (4 genera, 36 species) are found on both continents and island archipelagos in the southern hemisphere and across the Pacific. The circumscription of Asteliaceae and intrageneric relationships are poorly understood. We generated a phylogeny including all genera and 99% of the species using DNA sequence data from chloroplast (trnL, psbA-trnH, rps16, and petL-psbE) and nuclear (NIA-i3) regions. Relaxed clock methods were applied to infer the age of the family and the timing of cladogenic events. Generic delimitations change as a result of this study. Collospermum is nested within Astelia and is recognized here only at the subgeneric level. Further, Astelia subgenera Astelia, Asteliopsis, and Tricella are paraphyletic and to achieve monophyly their recircumscriptions are proposed. Despite the presence of Asteliaceae taxa on multiple Gondwanan landmasses and proposed Cretaceous origins for the family, radiation of genera was during the Tertiary. The largest and oldest genus, Astelia s.l. (including Collospermum), radiated around the Eocene/Oligocene boundary (ca. 34.2 million years ago (Ma)). Astelia s.l. subgenera diverged from the Oligocene/Miocene boundary onwards (<24.0 Ma). These dates suggest that current distributions are most likely to be the result of long-distance dispersal. Alpine taxa in New Zealand and Australia radiated during the Late Miocene/Pliocene. These results are congruent with Astelia micro- and macro-fossil data and suggest that Astelia s.l. either persisted in New Zealand during the proposed Oligocene marine transgression or dispersed from Australia after the subsequent expansion of terrestrial habitat.