Biological collections are essential for biodiversity research and conservation. However, these collections face challenges, including limited financial, human and spatial resources, inadequate infrastructure and competing curatorial responsibilities. The UFMG-AC Acarological Collection, housed at the Centro de Coleções Taxonômicas (CCT) at the Universidade Federal de Minas Gerais (UFMG), serves as a significant repository of mite diversity, containing over 16,000 catalogued specimens. This study aimed to standardise and publish the occurrence data from the UFMG-AC collection using the Darwin Core (DwC) standard to facilitate integration with global biodiversity databases. The workflow included a manual review of data to correct common errors, followed by the implementation of automated Python scripts to clean, standardise and restructure the data. A taxonomic validation pipeline was also developed to verify hierarchical consistency using the NCBI Taxonomy database and Taxallnomy API. The result is a comprehensive, high-quality dataset, compliant with DwC standards and ready for publication on platforms such as the Brazilian Biodiversity Information System (SiBBr) and the Global Biodiversity Information Facility (GBIF). By improving data quality and accessibility, our study ensures the preservation and global dissemination of the UFMG-AC’s invaluable records. This effort highlights the importance of metadata digitisation, automation and standardisation in ensuring data reliability and accessibility. The integration of the UFMG-AC data into public platforms enhances its visibility and utility for biodiversity research.
Natural history museums worldwide curate millions of preserved specimens and provide a vast untapped resource for scientific research. Advances in computer vision, particularly deep learning, now enable automated analysis of image data and several museums have made significant strides in digitising their collections. These efforts demonstrate the potential of digitised specimen images to support computer vision applications, including deep learning-based object detection and trait recognition. However, relatively few studies have explored the potential of preserved museum specimens as training data for automated trait detection. This study explores whether images of preserved fish specimens can be used to effectively train a deep-learning model, You Only Look Once (YOLO), to recognise morphological traits with key functional relevance in selected fish species from Zimbabwe. We also evaluated whether differences in model size (i.e. nano, small and medium) affected detection accuracy. Additionally, we evaluated these models’ performance on fish images from outside the museum setting to assess their potential for broader applications. Our results showed that all model sizes achieved high detection performance on museum specimens, with mean recall exceeding 0.95 and mAP@0.5 values ranging from 0.75 to 0.78. The medium model showed slightly higher precision and stricter localisation (mAP@0.5:0.95) and the nano model slightly outperformed it in overall detection (mAP@0.5). However, performance differences between the nano, small and medium models remained small and were not statistically significant across repeated validation subsamples. This suggests that lightweight models can provide performance comparable to larger models for detecting traits, making them viable for museums with limited computational resources. When applied to external images from the Global Biodiversity Information Facility (GBIF), the models showed a noticeable decline in precision (0.073–0.166) and recall (0.153–0.251), highlighting the limited ability of models trained on a small set of curated museum specimens to generalise to real-world images. Overall, this study demonstrates that curated museum specimen images provide a structured and valuable data source for training computer vision models to extract morphological traits, support trait-based taxonomic interpretation and advance digital curation efforts. While the performance of the object detection models was robust under museum-specimen imaging conditions, limited transferability to GBIF images suggests that broader application will require expanded datasets that include images from additional institutions and data sources.
The genus Megasorex is monotypic and endemic to Mexico, comprising the Mexican shrew, M. gigas (Mammalia, Soricidae). It was described in 1897 by Clinton H. Merriam, based on three specimens collected by Edward W. Nelson and Edward A. Goldman in the State of Jalisco. We have noted that, due to homonymous place names, the precise locality of the type series has been misinterpreted. Through scrutiny of their field journals and other historical records, here we investigated the origin of the type series for this species. To achieve this, we reconstructed the detailed field itinerary associated with the collection of the type series of M. gigas and restricted the type locality to Las Milpillas, approximately 4 km southwest of San Sebastián del Oeste, Jalisco, with a high degree of certainty. The data obtained in this study are available to support taxonomic studies and spatial analyses for biodiversity conservation.
The virtual 3D-reconstruction of a fragmented skull of a representative of Peramelemorphia (bandicoots and bilbies) using high-resolution surface scans proved to be a useful method for correcting its identification. Originally identified as and accordingly inscribed with Chaeropus castanotis (syn. Chaeropus ecaudatus ), i.e. a southern pig-footed bandicoot, it turned out to be the skull of a southern long-nosed bandicoot ( Perameles nasuta ). The provenance of the skull remains obscure.
Natural history collections are structured knowledge systems in which taxonomic and historical information is embedded within specimens, labels and their physical arrangement. However, in large entomological collections, much of this information remains inaccessible because it is not captured in standardised digital formats. Whole-drawer imaging offers a scalable approach to digitisation, but existing workflows primarily focus on specimen-level extraction and often neglect the curatorial structure encoded in drawer organisation. Here, we present an open-source workflow that transforms whole-drawer images into structured, machine-readable inventory data, while preserving their spatial arrangement and taxonomic context. The pipeline integrates low-cost imaging, barcode linkage, deep learning-based object detection (YOLOv.11) and optical character recognition (OCR) to reconstruct taxonomic series and estimate specimen counts per taxon from drawer images. The approach was developed for the Coleoptera collection of the Senckenberg Research Institute Frankfurt, encompassing approximately 1.9 million specimens which are stored in more than 5,000 drawers. Model training on Carabidae collection drawers overall achieved high performance (precision 99.2%, recall 97.9%, mAP0.5 98.5%), with reasonable transferability to other beetle families and Hymenoptera drawers tested. By treating the drawer as the primary unit of digitisation, this workflow provides a scalable intermediate layer between physical collections and specimen-level databases. It enables rapid assessment of taxonomic composition, supports collection management and digitisation planning and contributes to the mobilisation of biodiversity data from large historical collections.
Fluid-preserved specimens housed in museums represent a wealth of potential genomic data. Historically, museum wet specimens have been difficult to use as a source for genomic DNA due to the damage that fixative mediums may cause to the DNA. In this study, we explore several methods for DNA extraction to evaluate their efficacy when applied to fluid-preserved specimens, considering yield of DNA and the quality of short-read sequences that result from the extractions. We evaluated these methods as well as specimen age and starting tissue mass on the axes of DNA yield and sequencing quality. We found that across extraction methods, as sample age increased, DNA yield was reduced. Starting sample mass had no significant effect on DNA yield. Lastly, although each extraction was capable of yielding sequences with accurate base calls, loci recovery was highly variable across our tested extraction methods. Our results demonstrate the viability of wet museum specimens as a source of genomic DNA when processed with extraction and sequencing methods that account for their unique characteristics.
Natural history collections are invaluable sources of biological information for research, education, conservation, and public outreach but often lack the resources, programming expertise, and infrastructure needed to create custom digital tools for data accessibility. Large scale data aggregators facilitate data access but partially address common institutional needs. Artificial Intelligence (AI) has rapidly changed the way developers create, maintain, optimize, and test code, lowering the technical barrier to software creation. AI-assisted coding can accelerate the development of tools for digitization, data management, visualization, and public accessibility in natural history collections. Here I demonstrate the use of AI-assisted coding in the development of two custom web applications designed to address common institutional needs. Two web applications were built using GPT-5.3-Codex. Both were based on data exported from the Cornell University Museum of Vertebrates database. The first tool is a specimen search portal enabling multi-field filtering, data export, and mapping of localities. The second contains an interactive map to visualize the sampling of localities for four collections and a cumulative graph of specimens and lots sampled through time. Both applications performed the intended functions accurately, demonstrating that AI-assisted coding can be used to generate operational, lightweight applications for natural history collections. The development of web applications through AI-assisted coding reduces technical barriers for collections, enabling the creation of tailored tools with minimal infrastructure. As application development becomes more accessible, the challenge shifts to identifying tools for specific needs and determining how they can be shared in standardized ways across institutions and users. Careful prompt design, refinement, and attention to data security are essential for robust results and safe adoption of this technology.
Effective sealing of glass jars used in natural history wet collections is critical for long-term specimen preservation. The selection of an appropriate sealant is a long-known challenge in wet collections management. For a new permanent exhibition at the Croatian Natural History Museum, we decided to use ground-glass-stopper jars, but faced limited time at the outset of the project to select a suitable sealant. We therefore compared nine different sealants that included silicone, wax-based sealants, as well as several commercially available greases and rubbers. Our evaluation focused on key criteria relevant to long-term exhibition use: aesthetic appearance, handling properties and sealing performance. Based on its advantages over the other sealants and its established use in other wet collections, we selected the beeswax–colophony mixture to seal exhibition jars. Although highly effective, this sealant requires skill and is relatively time-consuming to apply. Therefore, we refined the application protocol for the commercial beeswax–colophony mixture, which will serve as practice-based guidance for sealing glass jars and cylinders with ground-glass stoppers. Our comparative assessment also offers practical guidance for curators and collection managers seeking appropriate sealants for exhibition purposes.
Freshwater fish are amongst the most threatened species globally, a situation that is particularly acute in the Iberian Peninsula and Portugal. Natural History Collections (NHCs) are biodiversity data sources with unparalleled taxonomic, geographic and temporal scope. These collections are essential for modern scientific research and the conservation of species; however, the accession of voucher specimens has faced a steep decline in the last decades. As part of an ongoing effort to review the Portuguese vertebrate collections in the national NHCs, freshwater fish data from major museums and some smaller collections have been gathered and analysed to assess the coverage and representation of the species known from Portuguese inland waters. Overall, the collections do not fully represent the country’s freshwater biodiversity, neither at a taxonomic, geographic or temporal level. Nearly 97% of the species are found in the national NHCs (64 out of 66), but close to 40% of taxa are represented by less than 50 specimens and nearly 70% from less than 50 lots. From a geographic standpoint, the collections cover less than 40% of Portuguese territory and one fifth of species has less than 10% of their known range represented in the assessed institutions. Additionally, 60% of native species are threatened with extinction, but they represent only one quarter of all extant specimens. We draw attention to the risks of incomplete coverage, especially when dealing with such sensitive ecosystems as freshwater.
Biodiversity data from African countries, including São Tomé & Príncipe, is often held outside the continent and remain difficult to access. Despite recent data mobilization efforts in Portuguese-speaking regions, significant gaps persist. Addressing these issues is crucial to support research, conservation and decision making. In this study, we compile and consolidate all available data on the herpetofauna of São Tomé & Príncipe from bibliography, natural history collections and citizen science platforms, with the goal of creating the most comprehensive and up-to-date database. We gathered a total of 6185 records of all amphibian and reptile species from the country and provided an overview of current taxonomic and distribution maps for each species, as well as taxonomic, distribution, and temporal sampling biases. Our results highlight the value of combining traditional scientific and citizen science data, while underscoring the continued importance of voucher specimens and historical data. This integrative approach not only advances knowledge of São Tomé & Príncipe’s unique herpetofauna but also provides a model for similar efforts in other under-studied regions and taxonomic groups.
In this report, we detail the process undertaken for preparing and mounting a sperm whale skeleton (Physeter macrocephalus Linnaeus, 1758) specimen for exhibition at the Natural History Museum of Oslo, University of Oslo, which was displayed in 2022. We provide a detailed account of materials applied and techniques used in the specimen for skeletonisation, including defleshing, carcass composting and maceration. We highlight a successful and cost-effective application of horse manure as a bulking agent in carcass composting and readily available detergents in maceration. Furthermore, we describe the skeleton mounting process, addressing architectural limitations restricting the possible loads on the ceiling. Consequently, the adjusted mounting procedure had to be undertaken, which included the construction of a custom-made podium and a steel support structure. Lastly, we present the result of the skeleton mount, which is supplemented with a 3D photogrammetric model.
Octocorals are noted for their high diversity, with the identification of species and sometimes even genera often difficult based solely on morphological characteristics. Thus, many recent studies have employed DNA barcoding techniques to acquire diversity data. There are many octocoral specimens already deposited within zoological collections that could provide potentially important taxonomic and historical information, but due to their age, often the DNA contained in such specimens has degraded severely. One potential approach to overcome this issue is DNA mini barcoding, amplifying a shorter subset of the original barcode region. This study aimed to utilize a mini barcoding method using novel octocoral-specific primers for validating octocoral species from the Okinawa Churaumi Aquarium Zoological Collection, with specimens dating back to 1982. Based on in silico evaluations using reference sequence data including nucleotide variations and match ratio values of species delimitations, using newly designed primers, mini barcodes performed as well as full-length barcodes. Subsequently, mini barcodes were generated by these novel primers, amplifying a 215 base pairs (bp) fragment of the mitochondrial MutS-like protein (mtMutS), successfully amplifying 92.7% of 166 specimens, and sequences could be used to identify to genus level, while full-length barcoding primers amplified only 45.8% of specimens. Moreover, mini barcodes identified 67 different operational taxonomic units (OTUs) of octocorals within the specimens based on the species delimitation method of Assemble Species by Automatic Partitioning (ASAP) and unique sequences. With most octocoral specimens from the Okinawa Churaumi Aquarium Zoological Collection identified only to genus level, these additional molecular data were significantly beneficial in identifying specimens. Our study shows this DNA mini barcoding approach can cheaply and quickly acquire molecular data from comparatively older octocoral specimens, helping to identify such specimens, and adding further value to historical museum specimen collections.
In this article I propose that there is an opportunity for media-based storytelling to be a more prominent application of basic research functions of natural history museums. I show, through surveying the field at large, that stories about the research and collections aspects of museums are not a prominent feature on institutional social media channels. However, I also show how natural history museum science has been, and is, the foundation of many successful science communication efforts on YouTube. I highlight some of this content from different creator perspectives and share my own experience as a museum-based researcher and creator to point towards how and why more research and collections focused media can be created in the future.
An herbarium must have its specimens processed, identified and accessioned so that members of the botanical community can readily consult these materials. However, many herbaria struggle to keep up with the influx of specimens and accumulate a backlog of unaccessioned specimens. In this study, we surveyed herbarium staff about their institution’s specimen backlog, focusing on the number and taxonomic lineage of backlogged specimens as well as the challenges associated with specimen curation. Based on responses from staff members at 65 herbaria, we calculated an average of 30,372 unaccessioned specimens per herbarium across all plant and fungal taxa. According to Index Herbariorum, there were 3,426 active herbaria in 2020 and, thus, we estimate at least 104 million specimens are present in herbarium backlogs worldwide. We also asked specific questions about unaccessioned bryophyte specimens in order to explore these collections in depth. While bryophytes represent only 9.0% of the accessioned specimens reported in our survey, they are overrepresented in the specimen backlog (16.0%). More than half of the bryophyte backlog is stored in field packets without labels and approximately three-quarters of these unaccessioned bryophytes lack species level identifications. Obstacles to backlog accessioning include labour shortages, insufficient bryological expertise and challenges integrating partially curated materials. Drawing on our survey responses, we offer strategies for reducing backlogs, such as accessioning straightforward specimens first and organising backlogs according to tasks leading to accessioning.
The flat wasps subfamily Mesitiinae (Hymenoptera, Bethylidae) are rarely collected using standard entomological sampling methods, resulting in a marked scarcity of specimens in natural history collections and persistent taxonomic uncertainty within the group. Mesitius ghilianii Spinola, 1853, which is the first species historically associated with what is now recognized as Mesitiinae, has a particularly complex nomenclatural history and is known from a single surviving historical specimen, designated as lectotype. Here, we report the restoration, redescription, and historical reassessment of this specimen, preserved in the Museo Regionale di Scienze Naturali, Turin, Italy. We document an inexpensive, non-toxic restoration protocol based on ammonia cleaning, mechanical debris removal with clinical needle, and controlled rehydration, which allowed the stabilization and partial repositioning of a nearly two-century-old point-mounted specimen. Following restoration, we provide a detailed modern redescription of the female, including diagnostic morphological characters of the head, mesosoma, wings, legs, and metasoma, supported by high-resolution imaging. In addition, we reconstruct the provenance and curatorial history of the specimen, tracing its collection by Vittore Ghiliani in Sicily and its subsequent study by Massimiliano Spinola, and we review the long and often contradictory taxonomic history of M. ghilianii and related taxa. This study underscores the irreplaceable scientific value of historical museum specimens for resolving longstanding taxonomic problems in poorly collected groups and highlights the importance of affordable restoration techniques for safeguarding type-level material. Our results provide a robust morphological baseline and historical framework for future integrative revisions of Mesitius and particularly M. ghilianii, which represent the genus-type of Mesitiinae and the type-species of Mesitius respectively.
Aim: Monitoring plant communities in response to drivers of global change is essential for understanding ecosystem health and conservation over time. Natural history collections are underutilized, open-source data that can be leveraged to study how communities change over time. This informs how under surveyed organisms, like bryophytes, may be impacted by drivers of global change. Location: We used herbarium specimen data to examine bryophyte community change in the southern Appalachians before and after the arrival of the balsam wooly adelgid (BWA). Methods: We used herbarium collections to examine raw and rarefied alpha diversity, as well as rarefied dissimilarity, over time. We compared two timespans, pre-BWA invasion and initial canopy recovery from invasion, to determine if species were either lost from, persisted in, or were new to high elevation areas and if these species had any shared features. Results: We found that mountains with greater disturbance resulted in lower levels of rarefied richness and that more extreme management practices may drive greater changes to the bryophyte community. However, this result cannot be decoupled from the disturbance with our data. Finally, we found that generalists significantly persisted through time compared to expected and compared to specialist species. Main conclusions: This is a promising use of herbarium specimens as the results match those of field-based experiments. As such, herbarium collections may be used as a substitute for permanent plots for lands with limited resources or a lack of expertise to determine the impacts management practices or invasion events have on ecological communities.
Natural history collections house material from centuries of collecting efforts. In the Phyletisches Museum Jena (PMJ), ca. 1 Mio specimens are deposited, some of them dating back as far as the 17th century. Modern imaging techniques have the potential to gain new insights from this historical material. However, a large part of the PMJ insect collection has not been revised by scientists in recent times. We screened the entire Auchenorrhyncha collection and found several specimens parasitized by two different species of the genus Halictophagus (Halictophagidae, Strepsiptera) that had previously been overlooked. These historical findings represent the only evidence to date of the occurrence of these two species in Germany and therefore suggest, at least historically, a larger distribution area than was previously known. In addition, hitherto unknown females and primary larvae were morphologically documented using state-of-the-art techniques such as synchrotron-radiation-based X-ray µCT and scanning electron microscopy. The data generated in this study cover the field of collectomics and can be seamlessly used as a basis for the emerging discipline of museomics. In taxonomic and systematic research and in the context of environmental change, pinned insects may play an outstanding role in the near future, as their DNA is not damaged by formalin fixation and thus can yield remarkable results even after more than 100 years. Our results underpin the value of historical material for modern research questions, especially for species that are difficult to find in nature.
Glass plays a large role in the storage and preservation of natural history collections, from the jars and vials used to preserve specimens in fluid to the glass microscope slides holding thin sections and fragments. While the glass used for storing specimens is typically not considered part of the collection it houses, the condition and preservation of the glass directly impacts the long-term preservation of the collections as well as the time and budgetary resources of the museum. This is particularly true for large collections where consistent monitoring of the condition of every item is not realistic. The Glass Deterioration Working Group at the Smithsonian Institution conducted a survey of the condition of the fluid collections and microscope slides in the Department of Invertebrate Zoology at the Smithsonian Institution’s National Museum of Natural History. Nearly a third of microscope slides, and over half of fluid collection containers were observed to have visible indications of glass deterioration.
Specimens preserved in alcohol require ongoing monitoring to prevent deterioration, as the preservative is inherently unstable and susceptible to evaporation and concentration changes. The Natural History Museum, London, holds a collection of true flies (order Diptera ) comprising 2,596 jars of alcohol-preserved material. Between April 2024 and January 2025, this collection was systematically surveyed to evaluate container performance in maintaining preservative integrity. Sufficient data were collected to compare three jar types: jam jars with metal screw-on lids, jars with glass lids and wire bail clip closures, and jars with ground glass stoppers. A composite comparison of performance, giving equal weight to the following indicators, seized lids, preservative loss, component failure, satisfactory preservation (full jars with 70–80% alcohol concentration), and unit cost. Jam jars with metal screw-on lids demonstrated the highest overall performance. Jars with ground glass stoppers, although traditionally considered the gold standard, showed high rates of lid seizure and substantially greater cost. Jars with wire bail clip tops performed least effectively, exhibiting poor control of alcohol evaporation. These findings provide evidence-based guidance for selecting cost-effective and durable containers for alcohol-preserved collections.
The digitization and labelling of historic and new specimens is a time-consuming, error-prone process as labels are still often hand-cut. To improve efficiency and consistency, we developed a low-cost solution for high-throughput environments which employs Cricut Maker® 3 to automatically cut precision entomological labels. We optimized digital label templates for use with Cricut® software and developed custom accessories for efficient label transfer and organization. Implementation of pre-cut label batches increased workflow efficiency, reduced user strain, and improved label quality. This system is compatible with existing digitization pipelines and supports standardized labelling formats, including those for DNA barcode workflows. With minimal set-up and maintenance costs, Cricut® offers an effective suite of tools for generating specimen labels in natural history collections. Modernizing entomology labelling workflows supports data standardization, collection digitization, and the scientific value of natural history specimens.
