Dataset Information

Toward a large-scale and deep phenological stage annotation of herbarium specimens: Case studies from temperate, tropical, and equatorial floras.

ABSTRACT: Premise of the Study:Phenological annotation models computed on large-scale herbarium data sets were developed and tested in this study. Methods:Herbarium specimens represent a significant resource with which to study plant phenology. Nevertheless, phenological annotation of herbarium specimens is time-consuming, requires substantial human investment, and is difficult to mobilize at large taxonomic scales. We created and evaluated new methods based on deep learning techniques to automate annotation of phenological stages and tested these methods on four herbarium data sets representing temperate, tropical, and equatorial American floras. Results:Deep learning allowed correct detection of fertile material with an accuracy of 96.3%. Accuracy was slightly decreased for finer-scale information (84.3% for flower and 80.5% for fruit detection). Discussion:The method described has the potential to allow fine-grained phenological annotation of herbarium specimens at large ecological scales. Deeper investigation regarding the taxonomic scalability of this approach is needed.

SUBMITTER: Lorieul T

PROVIDER: S-EPMC6426157 | biostudies-literature | 2019 Mar

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Toward a large-scale and deep phenological stage annotation of herbarium specimens: Case studies from temperate, tropical, and equatorial floras.

Lorieul Titouan T Pearson Katelin D KD Ellwood Elizabeth R ER Goëau Hervé H Molino Jean-Francois JF Sweeney Patrick W PW Yost Jennifer M JM Sachs Joel J Mata-Montero Erick E Nelson Gil G Soltis Pamela S PS Bonnet Pierre P Joly Alexis A

Applications in plant sciences 20190320 3

<h4>Premise of the study</h4>Phenological annotation models computed on large-scale herbarium data sets were developed and tested in this study.<h4>Methods</h4>Herbarium specimens represent a significant resource with which to study plant phenology. Nevertheless, phenological annotation of herbarium specimens is time-consuming, requires substantial human investment, and is difficult to mobilize at large taxonomic scales. We created and evaluated new methods based on deep learning techniques to a ...[more]

PMID: 30937225

Similar Datasets

Project description:Changes in phenological events have been vastly documented in face of recent global climate change. These studies are concentrated on temperate plants, and the responses of tropical species are still little understood, likely due to the lack of long-term phenological records in the tropics. In this case, the use of herbarium specimens to gather phenological data over long periods and wide geographic areas has emerged as a powerful tool. Here, we used four Melastomataceae species endemic to the Brazilian Atlantic Forest to evaluate phenological patterns and alterations as responses to recent climate changes. Phenological data were gathered from Reflora Virtual Herbarium specimens collected between 1920 and 2018, and analyzed with circular statistics applied to the intervals 1920-1979, 1980-1999, and 2000-2018. The effects of temperature range, average temperature, precipitation, and photoperiod on flowering and fruiting of each species were tested using multiple linear regressions. Through circular statistics, we detected changes, mostly delays, in the flowering of Miconia quinquedentata, Pleroma clavatum and P. trichopodum, and in the fruiting of M. acutiflora, P. clavatum and P. trichopodum. We also found that flowering and fruiting occurrence were related to local climatic conditions from months prior to the collections. We found marked phenological variations over the decades and also that these variations are associated to global climate change, adding up to the large body of evidence from higher latitudes. Our results also support herbarium collections as an important source for long-term tropical phenological studies. The lack of consistent patterns of responses among the four species (e.g. fruiting delayed two months in P. clavatum and advanced one month in M. acutiflora) suggests that climate change has unequal effects across tropical forests. This highlights the urgent need for further research to understand and forecast the ecological implications of these changes in global ecosystems processes.

Dataset Information

Toward a large-scale and deep phenological stage annotation of herbarium specimens: Case studies from temperate, tropical, and equatorial floras.

Publications

Toward a large-scale and deep phenological stage annotation of herbarium specimens: Case studies from temperate, tropical, and equatorial floras.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure