14
views
0
recommends
+1 Recommend
1 collections
    0
    shares
      • Record: found
      • Abstract: found
      • Article: found
      Is Open Access

      Machine Learning Using Digitized Herbarium Specimens to Advance Phenological Research

      research-article

      Read this article at

      Bookmark
          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

          Abstract

          Machine learning (ML) has great potential to drive scientific discovery by harvesting data from images of herbarium specimens—preserved plant material curated in natural history collections—but ML techniques have only recently been applied to this rich resource. ML has particularly strong prospects for the study of plant phenological events such as growth and reproduction. As a major indicator of climate change, driver of ecological processes, and critical determinant of plant fitness, plant phenology is an important frontier for the application of ML techniques for science and society. In the present article, we describe a generalized, modular ML workflow for extracting phenological data from images of herbarium specimens, and we discuss the advantages, limitations, and potential future improvements of this workflow. Strategic research and investment in specimen-based ML methods, along with the aggregation of herbarium specimen data, may give rise to a better understanding of life on Earth.

          Related collections

          Most cited references41

          • Record: found
          • Abstract: not found
          • Article: not found

          Deep learning models for plant disease detection and diagnosis

            Bookmark
            • Record: found
            • Abstract: found
            • Article: not found

            Warming experiments underpredict plant phenological responses to climate change.

            Warming experiments are increasingly relied on to estimate plant responses to global climate change. For experiments to provide meaningful predictions of future responses, they should reflect the empirical record of responses to temperature variability and recent warming, including advances in the timing of flowering and leafing. We compared phenology (the timing of recurring life history events) in observational studies and warming experiments spanning four continents and 1,634 plant species using a common measure of temperature sensitivity (change in days per degree Celsius). We show that warming experiments underpredict advances in the timing of flowering and leafing by 8.5-fold and 4.0-fold, respectively, compared with long-term observations. For species that were common to both study types, the experimental results did not match the observational data in sign or magnitude. The observational data also showed that species that flower earliest in the spring have the highest temperature sensitivities, but this trend was not reflected in the experimental data. These significant mismatches seem to be unrelated to the study length or to the degree of manipulated warming in experiments. The discrepancy between experiments and observations, however, could arise from complex interactions among multiple drivers in the observational data, or it could arise from remediable artefacts in the experiments that result in lower irradiance and drier soils, thus dampening the phenological responses to manipulated warming. Our results introduce uncertainty into ecosystem models that are informed solely by experiments and suggest that responses to climate change that are predicted using such models should be re-evaluated.
              Bookmark
              • Record: found
              • Abstract: found
              • Article: found
              Is Open Access

              An explainable deep machine vision framework for plant stress phenotyping

              Significance Plant stress identification based on visual symptoms has predominately remained a manual exercise performed by trained pathologists, primarily due to the occurrence of confounding symptoms. However, the manual rating process is tedious, is time-consuming, and suffers from inter- and intrarater variabilities. Our work resolves such issues via the concept of explainable deep machine learning to automate the process of plant stress identification, classification, and quantification. We construct a very accurate model that can not only deliver trained pathologist-level performance but can also explain which visual symptoms are used to make predictions. We demonstrate that our method is applicable to a large variety of biotic and abiotic stresses and is transferable to other imaging conditions and plants.
                Bookmark

                Author and article information

                Contributors
                Journal
                Bioscience
                Bioscience
                bioscience
                Bioscience
                Oxford University Press
                0006-3568
                1525-3244
                01 July 2020
                13 May 2020
                13 May 2020
                : 70
                : 6
                : 610-620
                Affiliations
                California Polytechnic State University , San Luis Obispo, California
                Florida Museum of Natural History , Gainesville, Florida
                Department of Ecology, Evolution, and Natural Resources, Rutgers, the State University of New Jersey , New Brunswick, New Jersey
                AMAP, the University of Montpellier and with The French Agricultural Research Centre for International Development, Centre National de la Recherche Scientifique, Institut National de la Recherche Agronomique, Institut de Recherche pour le Développement, Botanique et Modélisation de l’Architecture des Plantes et des végétations in Montpellier, France
                Florida Museum of Natural History, the University of Florida , Gainesville, Florida
                Harvard University Herbaria , Cambridge, Massachusetts
                US National Phenology Network and with the University of Arizona , Tucson, Arizona
                Natural History Museum of Los Angeles County, La Brea Tar Pits and Museum , Los Angeles, California
                AMAP, the University of Montpellier and with The French Agricultural Research Centre for International Development, Centre National de la Recherche Scientifique, Institut National de la Recherche Agronomique, Institut de Recherche pour le Développement, Botanique et Modélisation de l’Architecture des Plantes et des végétations in Montpellier, France
                Carnegie Museum of Natural History , Pittsburgh, Pennsylvania
                Inria Sophia-Antipolis, Zenith team, Laboratoire d’Informatique, de Robotique et de Microélectronique de Montpellier (LIRMM), Montpellier, France
                Inria Sophia-Antipolis, Zenith team, Laboratoire d’Informatique, de Robotique et de Microélectronique de Montpellier (LIRMM), Montpellier, France
                Department of Ecology, Evolution, and Marine Biology, the University of California , Santa Barbara, Santa Barbara, California
                Department of Entomology and Nematology, the University of California , Davis, Davis, California
                Florida Museum of Natural History, the University of Florida , Gainesville, Florida
                Yale Peabody Museum of Natural History , New Haven, Connecticut
                Department of Botany and the Data Science Lab, the Smithsonian Institution , Washington, DC
                Florida Museum of Natural History and with the University of Florida Biodiversity Institute, the University of Florida , Gainesville, Florida
                Author notes
                Author information
                http://orcid.org/0000-0003-4947-7662
                http://orcid.org/0000-0003-0756-5090
                Article
                biaa044
                10.1093/biosci/biaa044
                7340542
                32665738
                a0f1f79e-ad0c-40b6-a001-ab84e47db497
                © The Author(s) 2020. Published by Oxford University Press on behalf of the American Institute of Biological Sciences.

                This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.

                History
                Page count
                Pages: 11
                Funding
                Funded by: iDigBio;
                Funded by: National Science Foundation, DOI 10.13039/100000001;
                Award ID: DBI-1547229
                Categories
                Biologist's Toolbox
                AcademicSubjects/SCI00010
                AcademicSubjects/SOC02100

                phenology,machine learning,biodiversity,climate change,deep learning

                Comments

                Comment on this article