Machine Learning Using Digitized Herbarium Specimens to Advance Phenological Research

The promise of digital herbarium specimens in large-scale phenology research

The online mobilization of herbaria has made tens of millions of specimens digitally available, revolutionizing investigations of phenology and plant responses to climate change. We identify two main themes associated with this growing body of research and highlight a selection of recent publications exemplifying: investigating phenology at large spatial and temporal scales and in understudied locations and testing long-standing theories and novel questions in ecology and evolution that were not previously answerable. We explore strengths and limitations of using digitized herbarium specimens in phenology research, including: issues of sampling; reliability, transferability, and biases; and ethical and social justice considerations. This field will see further breakthroughs as herbaria around the world continue to mobilize and digitally interlink their collections. New developments will likely come from advances in technology, international collaborations, and including understudied plant taxa and regions such as the Arctic and the tropics. Advances in technology are already improving digitization workflows and speeding the collection of phenology data from digital specimens.

Long-term trends in global flowering phenology

Flowering phenology is an indicator of the impact of climate change on natural systems. Anthropogenic climate change has progressed over more than two centuries, but ecological studies are mostly short in comparison. Here we harness the large-scale digitization of herbaria specimens to investigate temporal trends in flowering phenology at a global scale. We trained a convolutional neural network model to classify images of angiosperm herbarium specimens as being in flower or not in flower. This model was used to infer flowering across 8 million specimens spanning a century and global scales. We investigated temporal trends in mean flowering date and flowering season duration within ecoregions. We found high diversity of temporal trends in flowering seasonality across ecoregions with a median absolute shift of 2.5 d per decade in flowering date and 1.4 d per decade in flowering season duration. Variability in temporal trends in phenology was higher at low latitudes than at high latitudes. Our study demonstrates the value of digitized herbarium specimens for understanding natural dynamics in a time of change. The higher variability in phenological trends at low latitudes likely reflects the effects of a combination of shifts in temperature and precipitation seasonality, together with lower photoperiodic constraints to flowering.

Tracking hidden dimensions of plant biogeography from herbaria

Plants are diverse, but investigating their ecology and evolution in nature across geographic and temporal scales to predict how species will respond to global change is challenging. With their geographic and temporal breadth, herbarium data provide physical evidence of the existence of a species in a place and time. The remarkable size of herbarium collections along with growing digitization efforts around the world and the possibility of extracting functional traits and geographic data from preserved plant specimens makes them invaluable resources for advancing our understanding of changing species distributions over time, functional biogeography, and conserving plant communities. Here, I synthesize core aspects of plant biogeography that can be gleaned from herbaria along changing distributions, attributes (functional biogeography), and conservation biogeography across the globe. I advocate for a collaborative, multisite, and multispecies research to harness the full potential of these collections while addressing the inherent challenges of using herbarium data for biogeography and macroecological investigations. Ultimately, these data present untapped resources and opportunities to enable predictions of plant species' responses to global change and inform effective conservation planning.

The digitisation workflow of the herbarium of the State Museum of Natural History of the NAS of Ukraine (LWS)

The digitisation workflow currently applied at the Herbarium of the State Museum of Natural History of the National Academy of Sciences of Ukraine (LWS) differs from other similar by cascade ('object-to-data-to-image') multilevel organisation. Its application is predicted by the need to preselect specimens by taxon and region, as well as by batched digitisation, which occurs with significant interruptions. Focusing on certain taxonomic groups from specific regions allows us to digitise specimens that could be more valuable for early scientific processing. At the same time, the herbarium benefits from such a digitisation model by revising the existing collection classification and keeping the initial ID system. The presented digitisation workflow can be easily reproduced in any herbarium with a limited budget. The purpose of this paper is to provide detailed description and schemas of the principal digitisation stages applied at the LWS Herbarium and to briefly discuss the steps crucial for a successful result. Provided information should help to maintain the digitisation and choose appropriate equipment and materials. We can conclude that, despite its general complexity, the described workflow demonstrated itself as viable and relevant due to its robust design and focus on data quality. Despite a focus on specialists' involvement, it maintains flexibility that allows combining volunteers and, if needed, outsourced efforts. Moreover, its modularity promotes independence of principal digitisation stages and allows long interruptions between the digitisation batches.

Ten lessons learned from the mass digitisation of a herbarium collection

Worldwide, herbaria maintain collections of reference specimens representing global plant diversity. These collections are a valuable resource for fundamental botanical research and applied scientific research across various disciplines, and play a significant role in addressing major societal challenges such as biodiversity conservation. The digitisation of herbarium specimens and their online dissemination is one of the most important recent developments in the curation of these collections. Digitisation significantly enhances access to the collections for the research community and facilitates large-scale analysis of biodiversity data. Digitisation also provides a means for preserving the physical specimens, as it reduces the need for handling and transportation. Rapid technological developments have greatly accelerated the rate of databasing and digital imaging of collections. Meise Botanic Garden recently completed a six-year project to mass digitise its herbarium collections of about 3 million specimens mounted on sheets and through this process we have learned valuable lessons. We have captured our experience in 10 recommendations for other collection-holding institutions to take inspiration from as they start planning their own digitisation efforts. We also present case studies where we delve deeper into certain topics as examples.

Geographic And Taxonomic Occurrence R-based Scrubbing (gatoRs): An R package and workflow for processing biodiversity data

Premise

Digitized biodiversity data offer extensive information; however, obtaining and processing biodiversity data can be daunting. Complexities arise during data cleaning, such as identifying and removing problematic records. To address these issues, we created the R package Geographic And Taxonomic Occurrence R-based Scrubbing (gatoRs).

Methods and Results

The gatoRs workflow includes functions that streamline downloading records from the Global Biodiversity Information Facility (GBIF) and Integrated Digitized Biocollections (iDigBio). We also created functions to clean downloaded specimen records. Unlike previous R packages, gatoRs accounts for differences in download structure between GBIF and iDigBio and allows for user control via interactive cleaning steps.

Conclusions

Our pipeline enables the scientific community to process biodiversity data efficiently and is accessible to the R coding novice. We anticipate that gatoRs will be useful for both established and beginning users. Furthermore, we expect our package will facilitate the introduction of biodiversity-related concepts into the classroom via the use of herbarium specimens.

Ancient DNA genomics and the renaissance of herbaria

Herbaria are undergoing a renaissance as valuable sources of genomic data for exploring plant evolution, ecology, and diversity. Ancient DNA retrieved from herbarium specimens can provide unprecedented glimpses into past plant communities, their interactions with biotic and abiotic factors, and the genetic changes that have occurred over time. Here, we highlight recent advances in the field of herbarium genomics and discuss the challenges and opportunities of combining data from modern and time-stamped historical specimens. We also describe how integrating herbarium genomics data with other data types can yield substantial insights into the evolutionary and ecological processes that shape plant communities. Herbarium genomic analysis is a tool for understanding plant life and informing conservation efforts in the face of dire environmental challenges.

From leaves to labels: Building modular machine learning networks for rapid herbarium specimen analysis with LeafMachine2

Premise

Quantitative plant traits play a crucial role in biological research. However, traditional methods for measuring plant morphology are time consuming and have limited scalability. We present LeafMachine2, a suite of modular machine learning and computer vision tools that can automatically extract a base set of leaf traits from digital plant data sets.

Methods

LeafMachine2 was trained on 494,766 manually prepared annotations from 5648 herbarium images obtained from 288 institutions and representing 2663 species; it employs a set of plant component detection and segmentation algorithms to isolate individual leaves, petioles, fruits, flowers, wood samples, buds, and roots. Our landmarking network automatically identifies and measures nine pseudo-landmarks that occur on most broadleaf taxa. Text labels and barcodes are automatically identified by an archival component detector and are prepared for optical character recognition methods or natural language processing algorithms.

Results

LeafMachine2 can extract trait data from at least 245 angiosperm families and calculate pixel-to-metric conversion factors for 26 commonly used ruler types.

Discussion

LeafMachine2 is a highly efficient tool for generating large quantities of plant trait data, even from occluded or overlapping leaves, field images, and non-archival data sets. Our project, along with similar initiatives, has made significant progress in removing the bottleneck in plant trait data acquisition from herbarium specimens and shifted the focus toward the crucial task of data revision and quality control.

Analyzing trait-climate relationships within and among taxa using machine learning and herbarium specimens

PREMISE

Continental-scale leaf trait studies can help explain how plants survive in different environments, but large data sets are costly to assemble at this scale. Automating the measurement of digitized herbarium collections could rapidly expand the data available to such studies. We used machine learning to identify and measure leaves from existing, digitized herbarium specimens. The process was developed, validated, and applied to analyses of relationships between leaf size and climate within and among species for two genera: Syzygium (Myrtaceae) and Ficus (Moraceae).

METHODS

Convolutional neural network (CNN) models were used to detect and measure complete leaves in images. Predictions of a model trained with a set of 35 randomly selected images and a second model trained with 35 user-selected images were compared using a set of 50 labeled validation images. The validated models were then applied to 1227 Syzygium and 2595 Ficus specimens digitized by the National Herbarium of New South Wales, Australia. Leaf area measurements were made for each genus and used to examine links between leaf size and climate.

RESULTS

The user-selected training method for Syzygium found more leaves (9347 vs. 8423) using fewer training masks (218 vs. 225), and found leaves with a greater range of sizes than the random image training method. Within each genus, leaf size was positively associated with temperature and rainfall, consistent with previous observations. However, within species, the associations between leaf size and environmental variables were weaker.

CONCLUSIONS

CNNs detected and measured leaves with levels of accuracy useful for trait extraction and analysis and illustrate the potential for machine learning of herbarium specimens to massively increase global leaf trait data sets. Within-species relationships were weak, suggesting that population history and gene flow have a strong effect at this level. Herbarium specimens and machine learning could expand sampling of trait data within many species, offering new insights into trait evolution.

The herbarium of the future

The ~400 million specimens deposited across ~3000 herbaria are essential for: (i) understanding where plants have lived in the past, (ii) forecasting where they may live in the future, and (iii) delineating their conservation status. An open access 'global metaherbarium' is emerging as these specimens are digitized, mobilized, and interlinked online. This virtual biodiversity resource is attracting new users who are accelerating traditional applications of herbaria and generating basic and applied scientific innovations, including e-monographs and floras produced by diverse, interdisciplinary, and inclusive teams; robust machine-learning algorithms for species identification and phenotyping; collection and synthesis of ecological trait data at large spatiotemporal and phylogenetic scales; and exhibitions and installations that convey the beauty of plants and the value of herbaria in addressing broader societal issues.

Estimating phenological sensitivity in contemporary vs. historical data sets: Effects of climate resolution and spatial scale

PREMISE

Phenological sensitivity, or the degree to which a species' phenology shifts in response to warming, is an important parameter for comparing and predicting species' responses to climate change. Phenological sensitivity is often measured using herbarium specimens or local studies in natural populations. These approaches differ widely in spatiotemporal scales, yet few studies explicitly consider effects of the geographic extent and resolution of climate data when comparing phenological sensitivities quantified from different data sets for a given species.

METHODS

We compared sensitivity of flowering phenology to growing degree days of the alpine plant Silene acaulis using two data sets: herbarium specimens and a 6 yr observational study in four populations at Niwot Ridge, Colorado, USA. We investigated differences in phenological sensitivity obtained using variable spatial scales and climate data sources.

RESULTS

Herbarium specimens underestimated phenological sensitivity compared to observational data, even when herbarium samples were limited geographically or to nearby weather station data. However, when observational data were paired with broader-scale climate data, as is typically used in herbarium data sets, estimates of phenological sensitivity were more similar.

CONCLUSIONS

This study highlights the potential for variation in data source, geographic scale, and accuracy of macroclimate data to produce very different estimates of phenological responses to climate change. Accurately predicting phenological shifts would benefit from comparisons between methods that estimate climate variables and phenological sensitivity over a variety of spatial scales.

Evaluation of the Students’ Learning Status in the Foreign Language Classroom Based on Machine Vision

In order to improve the effectiveness of the evaluation of student’s learning status in foreign language classrooms, this paper applies machine vision to classroom teaching. Through an in-depth analysis of the relative motion relationship between the end marker points of classroom feature recognition and the center point of the machine vision system window, this paper first proposes an autonomous tracking motion algorithm of the machine vision system window based on the preset field of view parameters. Moreover, this paper realizes the motion function of the window to track the marker points autonomously, completes the simulation analysis through two sets of planned trajectories and two sets of master hands to collect the actual trajectories, and verifies the correctness and feasibility of the algorithm. The research study shows that the algorithm based on machine vision proposed in this paper can effectively judge the real-time state of students in the foreign language classroom.

Assessing the performance of historical skins and bones for museomics using wolf specimens as a case study

Advances in the field of museomics have promoted a high sampling demand for natural history collections (NHCs), eventually resulting in damage to invaluable resources to understand historical biodiversity. It is thus essential to achieve a consensus about which historical tissues present the best sources of DNA. In this study, we evaluated the performance of different historical tissues from Iberian wolf NHCs in genome-wide assessments. We targeted three tissues—bone (jaw and femur), maxilloturbinal bone, and skin—that have been favored by traditional taxidermy practices for mammalian carnivores. Specifically, we performed shotgun sequencing and target capture enrichment for 100,000 single nucleotide polymorphisms (SNPs) selected from the commercial Canine HD BeadChip across 103 specimens from 1912 to 2005. The performance of the different tissues was assessed using metrics based on endogenous DNA content, uniquely high-quality mapped reads after capture, and enrichment proportions. All samples succeeded as DNA sources, regardless of their collection year or sample type. Skin samples yielded significantly higher amounts of endogenous DNA compared to both bone types, which yielded equivalent amounts. There was no evidence for a direct effect of tissue type on capture efficiency; however, the number of genotyped SNPs was strictly associated with the starting amount of endogenous DNA. Evaluation of genotyping accuracy for distinct minimum read depths across tissue types showed a consistent overall low genotyping error rate (<7%), even at low (3x) coverage. We recommend the use of skins as reliable and minimally destructive sources of endogenous DNA for whole-genome and target enrichment approaches in mammalian carnivores. In addition, we provide a new 100,000 SNP capture array validated for historical DNA (hDNA) compatible to the Canine HD BeadChip for high-quality DNA. The increasing demand for NHCs as DNA sources should encourage the generation of genomic datasets comparable among studies.

New directions in tropical phenology

Earth's most speciose biomes are in the tropics, yet tropical plant phenology remains poorly understood. Tropical phenological data are comparatively scarce and viewed through the lens of a 'temperate phenological paradigm' expecting phenological traits to respond to strong, predictably annual shifts in climate (e.g., between subfreezing and frost-free periods). Digitized herbarium data greatly expand existing phenological data for tropical plants; and circular data, statistics, and models are more appropriate for analyzing tropical (and temperate) phenological datasets. Phylogenetic information, which remains seldom applied in phenological investigations, provides new insights into phenological responses of large groups of related species to climate. Consistent combined use of herbarium data, circular statistical distributions, and robust phylogenies will rapidly advance our understanding of tropical - and temperate - phenology.

Specimen alignment with limited point-based homology: 3D morphometrics of disparate bivalve shells (Mollusca: Bivalvia)

Background

Comparative morphology fundamentally relies on the orientation and alignment of specimens. In the era of geometric morphometrics, point-based homologies are commonly deployed to register specimens and their landmarks in a shared coordinate system. However, the number of point-based homologies commonly diminishes with increasing phylogenetic breadth. These situations invite alternative, often conflicting, approaches to alignment. The bivalve shell (Mollusca: Bivalvia) exemplifies a homologous structure with few universally homologous points-only one can be identified across the Class, the shell 'beak'. Here, we develop an axis-based framework, grounded in the homology of shell features, to orient shells for landmark-based, comparative morphology.

Methods

Using 3D scans of species that span the disparity of shell morphology across the Class, multiple modes of scaling, translation, and rotation were applied to test for differences in shell shape. Point-based homologies were used to define body axes, which were then standardized to facilitate specimen alignment via rotation. Resulting alignments were compared using pairwise distances between specimen shapes as defined by surface semilandmarks.

Results

Analysis of 45 possible alignment schemes finds general conformity among the shape differences of 'typical' equilateral shells, but the shape differences among atypical shells can change considerably, particularly those with distinctive modes of growth. Each alignment corresponds to a hypothesis about the ecological, developmental, or evolutionary basis of morphological differences, but we suggest orientation via the hinge line for many analyses of shell shape across the Class, a formalization of the most common approach to morphometrics of shell form. This axis-based approach to aligning specimens facilitates the comparison of approximately continuous differences in shape among phylogenetically broad and morphologically disparate samples, not only within bivalves but across many other clades.

Caveat emptor: On the Need for Baseline Quality Standards in Computer Vision Wood Identification

Computer vision wood identification (CVWID) has focused on laboratory studies reporting consistently high model accuracies with greatly varying input data quality, data hygiene, and wood identification expertise. Employing examples from published literature, we demonstrate that the highly optimistic model performance in prior works may be attributed to evaluating the wrong functionality—wood specimen identification rather than the desired wood species or genus identification—using limited datasets with data hygiene practices that violate the requirement of clear separation between training and evaluation data. Given the lack of a rigorous framework for a valid methodology and its objective evaluation, we present a set of minimal baseline quality standards for performing and reporting CVWID research and development that can enable valid, objective, and fair evaluation of current and future developments in this rapidly developing field. To elucidate the quality standards, we present a critical revisitation of a prior CVWID study of North American ring-porous woods and an exemplar study incorporating best practices on a new dataset covering the same set of woods. The proposed baseline quality standards can help translate models with high in silico performance to field-operational CVWID systems and allow stakeholders in research, industry, and government to make informed, evidence-based modality-agnostic decisions.

Deep Learning in Plant Phenological Research: A Systematic Literature Review

Climate change represents one of the most critical threats to biodiversity with far-reaching consequences for species interactions, the functioning of ecosystems, or the assembly of biotic communities. Plant phenology research has gained increasing attention as the timing of periodic events in plants is strongly affected by seasonal and interannual climate variation. Recent technological development allowed us to gather invaluable data at a variety of spatial and ecological scales. The feasibility of phenological monitoring today and in the future depends heavily on developing tools capable of efficiently analyzing these enormous amounts of data. Deep Neural Networks learn representations from data with impressive accuracy and lead to significant breakthroughs in, e.g., image processing. This article is the first systematic literature review aiming to thoroughly analyze all primary studies on deep learning approaches in plant phenology research. In a multi-stage process, we selected 24 peer-reviewed studies published in the last five years (2016-2021). After carefully analyzing these studies, we describe the applied methods categorized according to the studied phenological stages, vegetation type, spatial scale, data acquisition- and deep learning methods. Furthermore, we identify and discuss research trends and highlight promising future directions. We present a systematic overview of previously applied methods on different tasks that can guide this emerging complex research field.

The Herbarium 2021 Half-Earth Challenge Dataset and Machine Learning Competition

Herbarium sheets present a unique view of the world's botanical history, evolution, and biodiversity. This makes them an all-important data source for botanical research. With the increased digitization of herbaria worldwide and advances in the domain of fine-grained visual classification which can facilitate automatic identification of herbarium specimen images, there are many opportunities for supporting and expanding research in this field. However, existing datasets are either too small, or not diverse enough, in terms of represented taxa, geographic distribution, and imaging protocols. Furthermore, aggregating datasets is difficult as taxa are recognized under a multitude of names and must be aligned to a common reference. We introduce the Herbarium 2021 Half-Earth dataset: the largest and most diverse dataset of herbarium specimen images, to date, for automatic taxon recognition. We also present the results of the Herbarium 2021 Half-Earth challenge, a competition that was part of the Eighth Workshop on Fine-Grained Visual Categorization (FGVC8) and hosted by Kaggle to encourage the development of models to automatically identify taxa from herbarium sheet images.

Using Convolutional Neural Networks to Efficiently Extract Immense Phenological Data From Community Science Images

Community science image libraries offer a massive, but largely untapped, source of observational data for phenological research. The iNaturalist platform offers a particularly rich archive, containing more than 49 million verifiable, georeferenced, open access images, encompassing seven continents and over 278,000 species. A critical limitation preventing scientists from taking full advantage of this rich data source is labor. Each image must be manually inspected and categorized by phenophase, which is both time-intensive and costly. Consequently, researchers may only be able to use a subset of the total number of images available in the database. While iNaturalist has the potential to yield enough data for high-resolution and spatially extensive studies, it requires more efficient tools for phenological data extraction. A promising solution is automation of the image annotation process using deep learning. Recent innovations in deep learning have made these open-source tools accessible to a general research audience. However, it is unknown whether deep learning tools can accurately and efficiently annotate phenophases in community science images. Here, we train a convolutional neural network (CNN) to annotate images of Alliaria petiolata into distinct phenophases from iNaturalist and compare the performance of the model with non-expert human annotators. We demonstrate that researchers can successfully employ deep learning techniques to extract phenological information from community science images. A CNN classified two-stage phenology (flowering and non-flowering) with 95.9% accuracy and classified four-stage phenology (vegetative, budding, flowering, and fruiting) with 86.4% accuracy. The overall accuracy of the CNN did not differ from humans (p = 0.383), although performance varied across phenophases. We found that a primary challenge of using deep learning for image annotation was not related to the model itself, but instead in the quality of the community science images. Up to 4% of A. petiolata images in iNaturalist were taken from an improper distance, were physically manipulated, or were digitally altered, which limited both human and machine annotators in accurately classifying phenology. Thus, we provide a list of photography guidelines that could be included in community science platforms to inform community scientists in the best practices for creating images that facilitate phenological analysis.

Harnessing Large-Scale Herbarium Image Datasets Through Representation Learning

The mobilization of large-scale datasets of specimen images and metadata through herbarium digitization provide a rich environment for the application and development of machine learning techniques. However, limited access to computational resources and uneven progress in digitization, especially for small herbaria, still present barriers to the wide adoption of these new technologies. Using deep learning to extract representations of herbarium specimens useful for a wide variety of applications, so-called "representation learning," could help remove these barriers. Despite its recent popularity for camera trap and natural world images, representation learning is not yet as popular for herbarium specimen images. We investigated the potential of representation learning with specimen images by building three neural networks using a publicly available dataset of over 2 million specimen images spanning multiple continents and institutions. We compared the extracted representations and tested their performance in application tasks relevant to research carried out with herbarium specimens. We found a triplet network, a type of neural network that learns distances between images, produced representations that transferred the best across all applications investigated. Our results demonstrate that it is possible to learn representations of specimen images useful in different applications, and we identify some further steps that we believe are necessary for representation learning to harness the rich information held in the worlds' herbaria.

Machine Learning Undercounts Reproductive Organs on Herbarium Specimens but Accurately Derives Their Quantitative Phenological Status: A Case Study of Streptanthus tortuosus

Machine learning (ML) can accelerate the extraction of phenological data from herbarium specimens; however, no studies have assessed whether ML-derived phenological data can be used reliably to evaluate ecological patterns. In this study, 709 herbarium specimens representing a widespread annual herb, Streptanthus tortuosus, were scored both manually by human observers and by a mask R-CNN object detection model to (1) evaluate the concordance between ML and manually-derived phenological data and (2) determine whether ML-derived data can be used to reliably assess phenological patterns. The ML model generally underestimated the number of reproductive structures present on each specimen; however, when these counts were used to provide a quantitative estimate of the phenological stage of plants on a given sheet (i.e., the phenological index or PI), the ML and manually-derived PI's were highly concordant. Moreover, herbarium specimen age had no effect on the estimated PI of a given sheet. Finally, including ML-derived PIs as predictor variables in phenological models produced estimates of the phenological sensitivity of this species to climate, temporal shifts in flowering time, and the rate of phenological progression that are indistinguishable from those produced by models based on data provided by human observers. This study demonstrates that phenological data extracted using machine learning can be used reliably to estimate the phenological stage of herbarium specimens and to detect phenological patterns.

Macrophenology: insights into the broad-scale patterns, drivers, and consequences of phenology

Plant phenology research has surged in recent decades, in part due to interest in phenological sensitivity to climate change and the vital role phenology plays in ecology. Many local-scale studies have generated important findings regarding the physiology, responses, and risks associated with shifts in plant phenology. By comparison, our understanding of regional- and global-scale phenology has been largely limited to remote sensing of green-up without the ability to differentiate among plant species. However, a new generation of analytical tools and data sources-including enhanced remote sensing products, digitized herbarium specimen data, and public participation in science-now permits investigating patterns and drivers of phenology across extensive taxonomic, temporal, and spatial scales, in an emerging field that we call macrophenology. Recent studies have highlighted how phenology affects dynamics at broad scales, including species interactions and ranges, carbon fluxes, and climate. At the cusp of this developing field of study, we review the theoretical and practical advances in four primary areas of plant macrophenology: (1) global patterns and shifts in plant phenology, (2) within-species changes in phenology as they mediate species' range limits and invasions at the regional scale, (3) broad-scale variation in phenology among species leading to ecological mismatches, and (4) interactions between phenology and global ecosystem processes. To stimulate future research, we describe opportunities for macrophenology to address grand challenges in each of these research areas, as well as recently available data sources that enhance and enable macrophenology research.

Analyzing a phenological anomaly in Yucca of the southwestern United States

Yucca in the American desert Southwest typically flowers in early spring, but a well-documented anomalous bloom event occurred during an unusually cold and wet late fall and early winter 2018–2019. We used community science photographs to generate flowering presence and absence data. We fit phenoclimatic models to determine which climate variables are explanatory for normal flowering, and then we tested if the same conditions that drive normal blooming also drove the anomalous blooming event. Flowering for Yucca brevifolia (Joshua tree) and Yucca schidigera (Mojave yucca) is driven by complex, nonlinear interactions between daylength, temperature, and precipitation. To our surprise, early-season flowering odds are highest in colder and drier conditions, especially for Joshua trees, but increase with precipitation late-season. However, the models used to fit normal blooming overpredicted the number of anomalous blooms compared to what was actually observed. Thus, predicting anomalous flowering events remains a challenge for quantitative phenological models. Because our model overpredicted the number of anomalous blooms, there are likely other factors, such as biotic interactions or other seasonal factors, which may be especially important in controlling what is presumed to be rare, out-of-season flowering in desert-adapted Yucca.

Region-specific phenological sensitivities and rates of climate warming generate divergent temporal shifts in flowering date across a species' range

PREMISE

Forecasting how species will respond phenologically to future changes in climate is a major challenge. Many studies have focused on estimating species- and community-wide phenological sensitivities to climate to make such predictions, but sensitivities may vary within species, which could result in divergent phenological responses to climate change.

METHODS

We used 743 herbarium specimens of the mountain jewelflower (Streptanthus tortuosus, Brassicaceae) collected over 112 years to investigate whether individuals sampled from relatively warm vs. cool regions differ in their sensitivity to climate and whether this difference has resulted in divergent phenological shifts in response to climate warming.

RESULTS

During the past century, individuals sampled from warm regions exhibited a 20-day advancement in flowering date; individuals in cool regions showed no evidence of a shift. We evaluated two potential drivers of these divergent responses: differences between regions in (1) the degree of phenological sensitivity to climate and (2) the magnitude of climate change experienced by plants, or (3) both. Plants sampled from warm regions were more sensitive to temperature-related variables and were subjected to a greater degree of climate warming than those from cool regions; thus our results suggest that the greater temporal shift in flowering date in warm regions is driven by both of these factors.

CONCLUSIONS

Our results are among the first to demonstrate that species exhibited intraspecific variation in sensitivity to climate and that this variation can contribute to divergent responses to climate change. Future studies attempting to forecast temporal shifts in phenology should consider intraspecific variation.

The growing and vital role of botanical gardens in climate change research

Botanical gardens make unique contributions to climate change research, conservation, and public engagement. They host unique resources, including diverse collections of plant species growing in natural conditions, historical records, and expert staff, and attract large numbers of visitors and volunteers. Networks of botanical gardens spanning biomes and continents can expand the value of these resources. Over the past decade, research at botanical gardens has advanced our understanding of climate change impacts on plant phenology, physiology, anatomy, and conservation. For example, researchers have utilized botanical garden networks to assess anatomical and functional traits associated with phenological responses to climate change. New methods have enhanced the pace and impact of this research, including phylogenetic and comparative methods, and online databases of herbarium specimens and photographs that allow studies to expand geographically, temporally, and taxonomically in scope. Botanical gardens have grown their community and citizen science programs, informing the public about climate change and monitoring plants more intensively than is possible with garden staff alone. Despite these advances, botanical gardens are still underutilized in climate change research. To address this, we review recent progress and describe promising future directions for research and public engagement at botanical gardens.

A choice of persistent identifier schemes for the Distributed System of Scientific Collections (DiSSCo)

Persistent identifiers (PID) to identify digital representations of physical specimens in natural science collections (i.e., digital specimens) unambiguously and uniquely on the Internet are one of the mechanisms for digitally transforming collections-based science. Digital Specimen PIDs contribute to building and maintaining long-term community trust in the accuracy and authenticity of the scientific data to be managed and presented by the Distributed System of Scientific Collections (DiSSCo) research infrastructure planned in Europe to commence implementation in 2024. Not only are such PIDs valid over the very long timescales common in the heritage sector but they can also transcend changes in underlying technologies of their implementation. They are part of the mechanism for widening access to natural science collections. DiSSCo technical experts previously selected the Handle System as the choice to meet core PID requirements.

Using a two-step approach, this options appraisal captures, characterises and analyses different alternative Handle-based PID schemes and the possible operational modes of use. In a first step a weighting and ranking the options has been applied followed by a structured qualitative assessment of social and technical compliance across several assessment dimensions: levels of scalability, community trust, persistence, governance, appropriateness of the scheme and suitability for future global adoption. The results are discussed in relation to branding, community perceptions and global context to determine a preferred PID scheme for DiSSCo that also has potential for adoption and acceptance globally.

DiSSCo will adopt a ‘driven-by DOI’ persistent identifier (PID) scheme customised with natural sciences community characteristics. Establishing a new Registration Agency in collaboration with the International DOI Foundation is a practical way forward to support the FAIR (findable, accessible interoperable, reusable) data architecture of DiSSCo research infrastructure. This approach is compatible with the policies of the European Open Science Cloud (EOSC) and is aligned to existing practices across the global community of natural science collections.

Estimating and Monitoring Land Surface Phenology in Rangelands: A Review of Progress and Challenges

Land surface phenology (LSP) has been extensively explored from global archives of satellite observations to track and monitor the seasonality of rangeland ecosystems in response to climate change. Long term monitoring of LSP provides large potential for the evaluation of interactions and feedbacks between climate and vegetation. With a special focus on the rangeland ecosystems, the paper reviews the progress, challenges and emerging opportunities in LSP while identifying possible gaps that could be explored in future. Specifically, the paper traces the evolution of satellite sensors and interrogates their properties as well as the associated indices and algorithms in estimating and monitoring LSP in productive rangelands. Findings from the literature revealed that the spectral characteristics of the early satellite sensors such as Landsat, AVHRR and MODIS played a critical role in the development of spectral vegetation indices that have been widely used in LSP applications. The normalized difference vegetation index (NDVI) pioneered LSP investigations, and most other spectral vegetation indices were primarily developed to address the weaknesses and shortcomings of the NDVI. New indices continue to be developed based on recent sensors such as Sentinel-2 that are characterized by unique spectral signatures and fine spatial resolutions, and their successful usage is catalyzed with the development of cutting-edge algorithms for modeling the LSP profiles. In this regard, the paper has documented several LSP algorithms that are designed to provide data smoothing, gap filling and LSP metrics retrieval methods in a single environment. In the future, the development of machine learning algorithms that can effectively model and characterize the phenological cycles of vegetation would help to unlock the value of LSP information in the rangeland monitoring and management process. Precisely, deep learning presents an opportunity to further develop robust software packages such as the decomposition and analysis of time series (DATimeS) with the abundance of data processing tools and techniques that can be used to better characterize the phenological cycles of vegetation in rangeland ecosystems.

Reversing extinction trends: new uses of (old) herbarium specimens to accelerate conservation action on threatened species

Although often not collected specifically for the purposes of conservation, herbarium specimens offer sufficient information to reconstruct parameters that are needed to designate a species as 'at-risk' of extinction. While such designations should prompt quick and efficient legal action towards species recovery, such action often lags far behind and is mired in bureaucratic procedure. The increase in online digitization of natural history collections has now led to a surge in the number new studies on the uses of machine learning. These repositories of species occurrences are now equipped with advances that allow for the identification of rare species. The increase in attention devoted to estimating the scope and severity of the threats that lead to the decline of such species will increase our ability to mitigate these threats and reverse the declines, overcoming a current barrier to the recovery of many threatened plant species. Thus far, collected specimens have been used to fill gaps in systematics, range extent, and past genetic diversity. We find that they also offer material with which it is possible to foster species recovery, ecosystem restoration, and de-extinction, and these elements should be used in conjunction with machine learning and citizen science initiatives to mobilize as large a force as possible to counter current extinction trends.

Climate, urbanization, and species traits interactively drive flowering duration

A wave of green leaves and multi-colored flowers advances from low to high latitudes each spring. However, little is known about how flowering offset (i.e., ending of flowering) and duration of populations of the same species vary along environmental gradients. Understanding these patterns is critical for predicting the effects of future climate and land-use change on plants, pollinators, and herbivores. Here, we investigated potential climatic and landscape drivers of flowering onset, offset, and duration of 52 plant species with varying key traits. We generated phenology estimates using >270,000 community-science photographs and a novel presence-only phenometric estimation method. We found longer flowering durations in warmer areas, which is more obvious for summer-blooming species compared to spring-bloomers driven by their strongly differing offset dynamics. We also found that higher human population density and higher annual precipitation are associated with delayed flowering offset and extended flowering duration. Finally, offset of woody perennials was more sensitive than herbaceous species to both climate and urbanization drivers. Empirical forecast models suggested that flowering durations will be longer in 2030 and 2050 under representative concentration pathway (RCP) 8.5, especially for summer-blooming species. Our study provides critical insight into drivers of key flowering phenophases and confirms that Hopkins' Bioclimatic Law also applies to flowering durations for summer-blooming species and herbaceous spring-blooming species.

Towards Sustainable North American Wood Product Value Chains, Part I: Computer Vision Identification of Diffuse Porous Hardwoods

Availability of and access to wood identification expertise or technology is a critical component for the design and implementation of practical, enforceable strategies for effective promotion, monitoring and incentivisation of sustainable practices and conservation efforts in the forest products value chain. To address this need in the context of the multi-billion-dollar North American wood products industry 22-class, image-based, deep learning models for the macroscopic identification of North American diffuse porous hardwoods were trained for deployment on the open-source, field-deployable XyloTron platform using transverse surface images of specimens from three different xylaria and evaluated on specimens from a fourth xylarium that did not contribute training data. Analysis of the model performance, in the context of the anatomy of the woods considered, demonstrates immediate readiness of the technology developed herein for field testing in a human-in-the-loop monitoring scenario. Also proposed are strategies for training, evaluating, and advancing the state-of-the-art for developing an expansive, continental scale model for all the North American hardwoods.

White paper on the alignment and interoperability between the Distributed System of Scientific Collections (DiSSCo) and EU infrastructures - The case of the European Environment Agency (EEA)

The Distributed System of Scientific Collections (DiSSCo) Research Infrastructure (RI) is presently in its preparatory phase. DiSSCo is developing a new distributed RI to operate as a one-stop-shop for the envisaged European Natural Science Collection (NSC) and all its derived information. Through mass digitisation, DiSSCo will transform the fragmented landscape of NSCs, including an estimated 1.5 billion specimens, into an integrated knowledge base that will provide interconnected evidence of the natural world. Data derived from European NSCs underpin countless discoveries and innovations, including tens of thousands of scholarly publications and official reports annually (supporting legislative and regulatory processes on sustainability, environmental change, land use, societal infrastructure, health, food, security, etc.); base-line biodiversity data; inventions and products essential to bio-economy; databases, maps and descriptions of scientific observations; educational material for students; and instructive and informative resources for the public. To expand the user community, DiSSCo will strengthen capacity building across Europe for maximum engagement of stakeholders in the biodiversity-related field and beyond, including industry and the private sector, but also policy-driving entities. Hence, it is opportune to reach out to relevant stakeholders in the European environmental policy domain represented by the European Environment Agency (EEA). The EEA aims to support sustainable development by helping to achieve significant and measurable improvement in Europe's environment, through the provision of timely, targeted, relevant and reliable information to policy-making agents and the public. The EEA provides information through the European Environment Information and Observation System (Eionet). The aim of this white paper is to open the discussion between DiSSCo and the EEA and identify the common service interests that are relevant for the European environmental policy domain. The first section describes the significance of (digital) Natural Science Collections (NHCs). Section two describes the DiSSCo programme with all DiSSCo aligned projects. Section three provides background information on the EEA and the biodiversity infrastructures that are developed and maintained by the EEA. The fourth section illustrates a number of use cases where the DiSSCo consortium sees opportunities for interaction between the DiSSCo RI and the Eionet portal of the EEA. Opening the discussion with the EEA in this phase of maturity of DiSSCo will ensure that the infrastructural design of DiSSCo and the development of e-Services accommodate the present and future needs of the EEA and assure data interoperability between the two infrastructures.

The aim of this white paper is to present benefits from identifying the common service interests of DiSSCo and the EEA. A brief introduction to natural science collections as well as the two actors is given to facilitate the understanding of the needs and possibilities in the alignment of DiSSCo with the EEA.

Landscape Analysis for the Specimen Data Refinery

This report reviews the current state-of-the-art applied approaches on automated tools, services and workflows for extracting information from images of natural history specimens and their labels. We consider the potential for repurposing existing tools, including workflow management systems; and areas where more development is required. This paper was written as part of the SYNTHESYS+ project for software development teams and informatics teams working on new software-based approaches to improve mass digitisation of natural history specimens.

A New Method for Counting Reproductive Structures in Digitized Herbarium Specimens Using Mask R-CNN

Phenology-the timing of life-history events-is a key trait for understanding responses of organisms to climate. The digitization and online mobilization of herbarium specimens is rapidly advancing our understanding of plant phenological response to climate and climatic change. The current practice of manually harvesting data from individual specimens, however, greatly restricts our ability to scale-up data collection. Recent investigations have demonstrated that machine-learning approaches can facilitate this effort. However, present attempts have focused largely on simplistic binary coding of reproductive phenology (e.g., presence/absence of flowers). Here, we use crowd-sourced phenological data of buds, flowers, and fruits from >3,000 specimens of six common wildflower species of the eastern United States (Anemone canadensis L., A. hepatica L., A. quinquefolia L., Trillium erectum L., T. grandiflorum (Michx.) Salisb., and T. undulatum Wild.) to train models using Mask R-CNN to segment and count phenological features. A single global model was able to automate the binary coding of each of the three reproductive stages with >87% accuracy. We also successfully estimated the relative abundance of each reproductive structure on a specimen with ≥90% accuracy. Precise counting of features was also successful, but accuracy varied with phenological stage and taxon. Specifically, counting flowers was significantly less accurate than buds or fruits likely due to their morphological variability on pressed specimens. Moreover, our Mask R-CNN model provided more reliable data than non-expert crowd-sourcers but not botanical experts, highlighting the importance of high-quality human training data. Finally, we also demonstrated the transferability of our model to automated phenophase detection and counting of the three Trillium species, which have large and conspicuously-shaped reproductive organs. These results highlight the promise of our two-phase crowd-sourcing and machine-learning pipeline to segment and count reproductive features of herbarium specimens, thus providing high-quality data with which to investigate plant responses to ongoing climatic change.