Managing biodiversity information: the good, the bad and the ugly. Jon Hutton SANBI BIMF 2012

Size: px

Start display at page:

Download "Managing biodiversity information: the good, the bad and the ugly. Jon Hutton SANBI BIMF 2012"

Sabina Rice
6 years ago
Views:

1 Managing biodiversity information: the good, the bad and the ugly Jon Hutton SANBI BIMF

2 Disclaimer! You are about to hear my own personal views, not those of UNEP SANBI BIMF 2

3 This presentation 1. Introduce you to UNEP-WCMC (briefly) 2. The Good, The Bad and The Ugly w.r.t biodiversity data Data gathering Data organising Data sharing Data synthesis Analysis Tools for decisions 3. Data shortcomings; citizen science ; remote sensing National infrastructures; international stuff, taxonomists 5. GBIF; Open Access; Barriers to Sharing 6. Semantic web; Mash-ups and quality control 7. Understanding needs; Perfect vs Good SANBI BIMF 3

4 SANBI BIMF 4 UNEP-WCMC UNEP-WCMC is UNEP s specialist Biodiversity Assessment Centre Started by IUCN 1980 Then IUCN, WWF & UNEP Became an integral part of UNEP in 2000 Pioneered much of the global conservation infrastructure we work with today.

5 SANBI BIMF 5

6 What do we do these days? Data gathering Data organising Data sharing Data synthesis Analysis Tools for decisions SANBI BIMF 6

7 A few current examples SANBI BIMF 7

8 Biodiversity Indicators Partnership UNEP-WCMC led the review of the evidence in 2010 Butchart et al., Science 328, , 28 th May 2010

9 SANBI BIMF 9 Secretariat of the SGA Network Over 80 SGAs around the world

10 UN REDD Multiple Benefits SANBI BIMF 10

11 SANBI BIMF 11 Support to biodiversity MEAs N No of paragraphs incl WDPA, 2010 BIP & CC 5 No of decisions incl WDPA, 2010 BIP & CC Conferences of the Parties

12 Support to the business sector SANBI BIMF 12

13 Biodiversity Modeling, Scenarios etc SANBI BIMF 13

14 Intergovernmental Platform on Biodiversity and Ecosystem Services (IPBES) SANBI BIMF 14

15 Biodiversity data and information. Data gathering Data organising Data sharing Data synthesis Analysis Tools for decisions SANBI BIMF 15

16 Biodiversity multi-dimensional & complex Variety Quantity - the number of different types. - how much there is of the different types. Composition Distribution - how are the types combined? - where is it located. Quality - how intact/degraded is it? Measuring biodiversity is not a trivial task...

17 Plus there are different and competing views of what important biodiversity is

18 Except for vertebrates, data is sparse... MEASURE WITHIN SPECIES AMONG SPECIES ECOSYSTEMS Variability usually number of types Number of different: genes/genomes, individuals, or populations. Megafauna & Flora Number of species: measures of richness and of diversity. Number of different ecosystems. No classification Global distribution Global amount Mapped distribution of genes/genomes, individuals or populations. Total amount of genetic material, number or biomass of individuals or populations. Some populations Good for narrow sub-set Mapped distributions of species richness and diversity. Good for narrow sub-set Total species population count or biomass. Maps of ecosystem distribution. Improving Poor, esp marine Biomass of ecosystems Trends over time Changes in number, distribution and amount of genetic Some material, populations individuals or populations Loss of species (extinctions). Loss Good of distribution for narrow (range loss). Loss of amount sub-set (population declines). Habitat alteration. Land use change. Fragmentation. Remote sensing coming into play

19 Data gathering 1. Generally unsystematic & disorganised 2. Not designed to answer questions 3. Has no agreed framework 4. Not assembled to allow comparisons over time SANBI BIMF 19

20 The stirrings of change? SANBI BIMF 20

21 Quantity vs Quality? Role of citizen science/crowd-sourcing/wiki? Its not new e.g. birdwatchers and their observations Opportunities extended by mobile technology e.g. beach/air quality in Europe (Eye on Earth) SANBI BIMF 21

22 Worth look. SANBI BIMF 22

23 Lessons learned Its not just about technology It is hard to create communities of users Best if build on existing communities Create reliability thresholds or rankings Think twice before creating a new Wiki! Significant issues arise - storage and attribution May need some new skills marketing! SANBI BIMF 23

24 Emergence of competing standards? SANBI BIMF 24

25 Remote Sensing SANBI BIMF 25

26 Some marked successes - LIDAR SANBI BIMF 26

27 Detection of deforestation Terra-I machine learning approach SANBI BIMF 27

28 But where is the automatic monitoring? SANBI BIMF 28

29 Remote sensing still hasn t delivered on its early promise 1. Interests and incentives are not aligned. 2. Skill-sets often wrong. 3. Changing priorities and funding are significant. 4. But also, we haven t defined what we want! The Economist May 12 th SANBI BIMF 29

30 Data gathering Data organising Data sharing Data synthesis Analysis Tools for decisions SANBI BIMF 30

31 National Data initiatives SANBI BIMF 31

32 There aren t enough high quality national biodiversity centres SANBI BIMF 32

33 In contrast? SANBI BIMF 33

34 Taxonomy (and Taxonomists!) SANBI BIMF 34

35 Are taxonomists helping or hindering? SANBI BIMF 35

status of biodiversity at national or regional priority for the community?

36 Descriptive Taxonomy ACTION 8: Complete a comprehensive coverage of information Why hasn t on known alpha species taxonomy at the global level, been then the use it by 2020 to assess the status of biodiversity at national or regional priority for the community? levels including information on abundance, distribution, functioning and conservation in the ecosystem. SANBI BIMF 36

37 Data gathering Data organising Data sharing Data synthesis Analysis Tools for decisions SANBI BIMF 37

38 GBIF SANBI BIMF 38

39 SANBI BIMF 39

40 Quality data behind pay-walls SANBI BIMF 40

41 On the other hand there are issues about Open Access and simplistic attitudes to data sharing All your data should be freely available for anyone and everyone to use SANBI BIMF 41

42 A case study. World Database on Protected Areas /- 45,000 records /- 60,000 records Restricted for Commercial use /- 197,000 records SANBI BIMF 42

43 Sometimes is immoral NOT to pay! SANBI BIMF 43

44 SANBI BIMF 44

45 SANBI BIMF 45 Barriers to the free interchange of data and information are complex and may be: 1. Related to the nature of the information itself: languages, taxonomies, standards etc 2. Technical and/or practical: Discovery, storage, transfer etc 3. Resource constraints: State or donor funding etc 4. Behavioural

46 SANBI BIMF 46 Behavioural reluctance to disclose or share 1. Career advancement not just academics 2. Concerns about inappropriate use 3. Sensitivity of data e.g. endangered species 4. Issues of Brand and competition 5. Concerns about fairness, recognition, attribution 6. Issues of value

47 Aggregating Tools WDPA (PAs) UNEP-WCMC processes WBDB (KBAs) BL & CI Processes $24m IUCN WCPA & NGOs Governments, national & regional processes RedList (Species) IUCN SSC Species - Experts and Processes $80m 47

48 SANBI BIMF 48 A coordinated effort to lift the Barriers 1. Governments: Public goods need public funding 2. Donors/Funding Agencies: Build data management into projects Accept long term data commitments Research permissions tied to open data 3. Publishers: Should all require underlying data to be made available 4. Academics: Find a way to credit the publication of data 5. Institutions: Need to find a consistent licensing approach to branded datasets 6. Business: Has to internalise the cost of biodiversity data

49 Data gathering Data organising Data sharing Data synthesis Analysis Tools for decisions SANBI BIMF 49

50 Semantic Web and Technologies. A vision for converting the current web of unstructured documents into a "web of data SANBI BIMF 50

51 Bringing data together - mashups etc. SANBI BIMF 51

52 Overlaying biodiversity and carbon 1. Endemic Bird/Plant Areas? 2. Protected Areas? 3. Threatened RedList Species? 4. KBAs? 5. National Priorities (SCP)? 6. Threatened Ecosystems? 7. Global 200 Ecoregions? 8. Hotspots? 9. Wilderness Areas? 10. Etc. x30 SANBI BIMF 52

53 A little knowledge is a dangerous thing Alexander Pope SANBI BIMF 53

54 1. Differencing two carbon stock maps derived from land-use from Globcover 2005 and 2009 is simple intuitive and. wrong! 2.. reliable change detection can never be achieved by comparing two independent successive land cover maps. Therefore, the GlobCover 2009 land cover map cannot be used for any change detection application. In particular, the direct comparison with the previous GlobCover 2005 product should be avoided since it cannot guarantee accurate land cover change quantification (Bontemps et al. 2010). SANBI BIMF 54

55 This analysis is totally misleading SANBI BIMF 55

56 Quality control? 1. How do tackle this growing problem? 2. Peer Review is either missing or not picking up on it 3. Encourage publication of methodology and assumptions 4. Standards and guidelines? 5. Training? SANBI BIMF 56

57 Data gathering Data organising Data sharing Data synthesis Analysis Tools for decisions SANBI BIMF 57

58 Most of us don t understand the needs of decision-makers SANBI BIMF 58

59 SANBI BIMF 59

60 SANBI BIMF 60

61 Understanding Biodiversity Standards SANBI BIMF 61

62 SANBI BIMF 62

SANBI BIMF 63 The perfect is the enemy of the good (Voltaire) Tittensor et al. 2010.

63 SANBI BIMF 63 The perfect is the enemy of the good (Voltaire) Tittensor et al Global patterns and predictors of marine biodiversity across taxa. Nature 466:

64 Few understand the needs of decision-makers But SANBI clearly does! SANBI BIMF 64

65 Thanks for your attention SANBI BIMF

National Biodiversity Information System. Brenda Daly South African National Biodiversity Institute

National Biodiversity Information System Brenda Daly South African National Biodiversity Institute Data workflows Specify Custom National data store FBIP IPT 11 Museums queries ispot Spatial BGIS NBIS