Bioversity International, Via dei Tre Denari 472/a, Maccarese, Rome, Italy

Accessing the original observation data captured during plant exploration missions for collecting crop diversity

Bioversity International, Via dei Tre Denari 472/a, Maccarese, Rome, ItalyHannes Gaisberger, Massimo Buonaiuto, Federico Mattei, Andrea De Pirro, Valentina Barbiero, Simone Mori, Imke Thormann, Tom Hazekamp, Elizabeth Arnaud

Agenda

• Part 1: Safeguarding the original paper documents by scanning and digitizing the data – Hannes Gaisberger

• Part 2: Creation of a public repository of full scanned documents enabling access to the full text – Massimo Buonaiuto

Bioversity supported germplasm collecting missions

• Since 1974, Bioversity International has supported more than 550 germplasm collecting missions yielding 225,875 samples and covering 4,300 species from 137 countries

• Samples were sent to several genebanks worldwide for safety duplication, conservation and potential distribution

• Other CGIAR centers organized various collecting missions for their mandate crops

Original observation data is essential for:

• Identify duplicates between collections and gaps in diversity – value for genebank curators and collecting actions

• Tracking original sample & country of origin in pedigrees – value for Breeders and Benefit Sharing

• Collectors recorded key sample information (passport data) and other observation data in field books

Scanning of field notebooks and related documents

Original observation: a treasure for genebanks and breeders

• Genus and Species

• Collecting Number

• Site Information: Admin boundaries, Latitude, Longitude and Elevation

• Collecting Source and Sample Status

The collecting form contains the botanical classification along with localization details, environment, cultural practices, diseases and pest presence and symptoms and traditional uses

Identification and quality-checking in databases

• Different publicly available genebank inventories are checked in order to track corresponding samples and complete missing passport data

Integration of quality passport data

• Data extracted from field books and databases is integrated in a sample level database of collecting missions

Results in figures

• To date, the quality of 101,171 passport data records from 375 collecting missions has been improved through data extracted from scanned documentation

• 56,454 of these collected samples are linked to genebank accessions in 51 institutes worldwide

Priority crops/ use group

Number of collected samples

Forages 44056Rice 25022Maize 16484Beans 10976Wheat 7507Cowpea 7473Potato 7146Pearl millet 6662Barley 4429Groundnut 2928Finger millet 2850Chickpea 1467Banana 1326Pigeon pea 999Others 86550Total 225875

• A total of 43,637 scanned pages are saved as 1063 pdf-files and stored in an online repository aside the 26,000 other files scanned by CGIAR centers and partners

Publishing the data and attached information

• End of 2010: work must be finished for Bioversity supported missions

• Full text available on the online repository and publish the collection mission database

• Visualization: Map sites where diversity was collected (after georeferencing with Biogeomancer)

• Various projects to address gaps analysis and diversity analysis, like Genesys, encourage partners to perform same work and share the full text and data – links to CWR information, Museum herbaria information, Literature

Public access to the scanned collecting missions documents

A Repository that presently contains 27,000 Collecting Missions Files from CGIAR Centers and partners:• Agricultural Research Centre (ARC) of Lao People’s

Democratic Republic • AfricaRice• Agricultural Research for Development in Africa (IITA)• Bioversity International• International Rice Research Institute (IRRI)

Typology of the documents produced by Collectors

1) Mission Reports2) Summary Forms3) Sample lists4) Collecting Forms5) Accession Vouchers                                       6) Newsletters7) Factsheets8) Distribution lists9) Field Books

Documents Types Hierarchy

Analysis of Metadata (1/5)

Analysis of Metadata (2/5)

Analysis of Metadata (3/5)

Analysis of Metadata – Darwin Core for Germplasm (4/5)

Analysis of Metadata (5/5)

Darwin Core Germplasm metadata+

Collecting Missions metadata =

Metadata for Collecting Missions Documents

How users will access the Repository

Alfresco DMS

Typo3 CMS

Import of 27,000 PDF Files

Process of import PDF files in 3 phases:1. Conversion of institutional metadata in Darwin

Core Germplasm metadata2. Association of metadata to all PDFs files, using

heterogeneous sources (databases, Excel files and filenames, etc.)

3. Batch upload of all PDF files together with metadata file associated to each file in DC-Germplasm standard.

Public Search Mask (1/3)

Public Search Mask (2/3)

Public Search Mask (3/3)

How users will manage and publish documents

• Simple Workflow to publish into the Repository:1. Upload the file in private

user Home Space2. Edit metadata3. Approve the document

for public repository with a click

... the file will be and public

Summary

• Improved quality of passport data for about 100,000 collected samples from 137 countries

• 56,454 of these collected samples are linked to genebank accessions in 51 institutes worldwide

• Collected 27,000 documents classified in 9 types of documents with metadata

• Metadata extracted and parsed using Gerplasm Darwin Core standards

Open questions and challenges- Interaction with Open Archive standards and Protocol

for Metadata Harvesting- Integration with Crop Terminizer, University of

Manchester- Web Analytics for monitoring of downloads in details

(referrers, visits, etc.) and web marketing- CMIS protocol used to interact with content

management systems- Metadata validation with crop scientists, collectors

http://www.central-repository.cgiar.org/

http://www.central-repository.cgiar.org/

Guidelines for collecting samples

- Being revised and will be published in a new section of the on the Crop genebank knowledge base

- Adding guidelines for illustrating with photos that support the tentative taxonomy, captured data and GPS

THANK YOU!