Feedback from plant session - Biodiversity...
Transcript of Feedback from plant session - Biodiversity...
Feedback from plant e-Flora and occurrence session
The Policy
Prioritise what we can and should include in the e-Flora website for every species:
1. One morphological description
2. One image
3. Habitat and distribution information
4. Distribution map
5. A link to the SANBI Red List
6. Specimen information from BRAHMS
Markup Add In Tool
Literature R
ibbon
Display taxon
panel
Excel export button
e-Flora mock-up species page
Status of digitising plant specimen data • 44% of South Africa’s plant
specimens have been electronically encoded (NRF, 2011)
• 90% of the country’s ca. 3,263,200 plant specimens concentrated in 6 major herbaria (PRE, NBG, NH, NU, BOL, GRA)
• Network of herbaria working with SANBI
• Set-one migrate to BRAHMS • Central publishing platform
(Integrated Publishing Tools) • Curation of specimens remain
responsibility of each herbarium
e-Flora
• Markup time – depending on quality of scan of the revision / OCR some can go very fast
• Integration of indigenous knowledge into e-Flora; methods to integrate the data? – Info would not necessarily be incorporate by 2016 but
could look at incorporating with time
– SANBI tool?
– Research Chairs on IK – incorporate how?
• SANBI Sustainable devlopment?
e-Fauna
• Vernacular names should include reference citation
Speeding up digitisation of plant specimens
• Challenges:
– Speeding up
– Human resources are limited
– Training and assistance with BRAHMS, also in data capturing
– Staff; local teams (turnover issues)
Digitising plant specimens cont.
• Some solutions: – SANBI provides BRAHMS training (Hannelie Snyman) –
EWC, WITS, Shonaland
– Dedicated effort to make speed up digitisation – core team
– Crowd sourcing
– Central place for digitised images of specimens required
– Investigate OCR for typed labels
– Training to herbaria staff for a certain level of quality of digitisation to manage ongoing updates and loans.
– Students should put their research in BRAHMS (even via Excel)
Specimen Verification
• Dual process needed: 1. Incorrect IDs
- solution network of experts to share load - Users should sign contract to verify/det specimens
(loan and visit conditions); complete list of cited specimens not necessarily published – more people/capacity involved
- Photographing specimens virtual herbaria (Require capacity to scan)
2. Incorrect georeferenced data
Specimen Verification
2. Incorrect georeferenced data
• Fast tracking georeferencing work with collector itinerary
• Consolidated system of BRAHMS for all herbaria in country – fast track quality control; but we need a virtual herbarium – duplicate specimens.
Specimens
• Observe errors – feedback process and BRAHMS; IPT?? – Feedback to herbaria and specialists? – Everybody has to be on the same system – Automate feedback and entry flagged – Herbarium pull report of queries or errors (for certain
families/groups and within certain dates); BRAHMS functionality? Put request forward to Denis
– API people had a system for feedback on errors? (Should be possible)
– Users flag and comment on errors
Specimens
• SAFARIS (field expeditions)
• DigiVol (volunteers building knowledge); digitisation verification projects
• Each project has a tutorial on how or what to do
• Not in game format
• Specimen information verified by two pairs of eyes (review vs validate)
• Awarding effort of volunteers via access to gardens; BotSoc membership?
Specimens
• SAFARIS (field expeditions) • Preparing files and setting up project:
– Photograph pages/specimens – Batch upload – Specify fields that need to be checked/updated – Can include using a mapping tool – DwC fields used – Ready for import into GBIF
• Les gives feedback to every volunteer (for specimens batch feedbacks could work better) – volunteers should feel more inclined to continue if they receive feedback
Specimens
• SAFARIS (field expeditions)
• Digitisation game (French) – engage with developers and ask to translate
• Following collector’s field trail/itinerary (more interesting and more effective – georeferencing multiple specimens for single locality or close proximity)
• Verification: specimen against database (verification); georeferencing is a different process)
• Georeferencing is possible (Les showed example)
Digitisation summary
• Masses of encoding needs to be done • Look into crowd sourcing possibilities • DigiVol possibility • Gaming possibility • Continue digitisation:
– Train herbarium staff – Invest in centralised team to work on specimens
(anywhere in country) – Actual text on label digitised (separate process to
georeferencing) – Collectors and expeditions
• SANBI has a gazetteer available (could assist with localities)
Digitisation summary
• Locality issues (specific vs general location)
• Go back to existing georeferenced data using a programme to pick up inconsistencies
• How to deal with feedback of incorrect data (BRAHMS committee to investigate)
• 50% digitised records not all unique – duplicates in other herbaria; could a system be developed to integrate data and flag that information/georeferenced data captures; also process to assist with verification (note: all labels don’t have accurate data)
Digitisation summary
• Process of encoding – herbaria busy encoding search and find information that is exist in other herbaria that have completed digitisation (herbarium should be able to personalise data)
Way forward
• Opportunities to apply for funding? • Projects not coherent • SANBI coordinate under one project • Herbarium/researcher feels capacitated should apply
individually for small grants • Theme to be added to funding categories: Digitising
processes • Strongest approach theme 5:
– New technologies – Plants or animals – Digitised medicinally valuable species (combination) – Biodiversity in general important (what about utilised
species??)