Monitoring wild nature for the 2010 target

Slide 1: 

Beyond extinction rates: Monitoring wild nature for the 2010 target 21 - 22 July 2004, The Royal Society, London Final report from the group working on “Habitats and biomes”  Group members: Andrew Balmford, Isabelle Côté, Rob Ewers, Simon Ferrier, Tom Lovejoy, Philippe Mayaux, Doug Muchoney, Richard Norris, Carmen Revenga, Marc Steininger, Holly Strand, Woody Turner

Measures of habitat change : 

Measures of habitat change Measures of extent and distribution Obtained mainly through remote sensing Change in extent of habitat is a raw measure, therefore the translation into indicators requires integration with other sources of information Measures of quality Will be ecosystem-specific

Properties of habitat change measures : 

Properties of habitat change measures Repeatable Representative Assumptions and weaknesses explicit Hierarchical collection Fit for purpose Understandable Cost-effective

Initial thoughts on coastal habitats : 

Initial thoughts on coastal habitats Coral reefs: rough remote estimates of extent (unlikely to be delivered by 2010) and percentage live coral as a measure of quality (possibly deliverable by 2010) Seagrasses and mangroves: remote estimates of extent (possibly deliverable by 2010) Kelp beds: rough remote estimates of extent (unlikely to be delivered by 2010) Estuaries: remote estimates of urban development (can be delivered by 2010) and eutrophication (empirical and modelled) and a measure of quality (possibly deliverable by 2010)

Initial thoughts on freshwater habitats : 

Initial thoughts on freshwater habitats Rivers, lakes and wetlands: global remote estimates of extent for large water bodies (can be delivered by 2010); remote estimates for smaller water body extent (unlikely to be delivered by 2010); and as quality indicators water abstraction, riparian vegetation and modelled composite measure (all could possibly be delivered by 2010). Peatlands: global remote estimates of extent (possibly deliverable by 2010).

Initial thoughts on marine habitats : 

Initial thoughts on marine habitats Open ocean: population data as an indicator of quality (could be delivered by 2010) Shelf: indicators of quality might include trawling data (possibly deliverable by 2010); trophic level index (can be delivered by 2010); pollution (frequency of red tides, anoxic events etc) (unlikely to be delivered by 2010). Sea mounts: Remote data on extent? (can be delivered by 2010).

Initial thoughts on forests: Data available : 

Initial thoughts on forests: Data available Maps 1 km Land Cover classes and tree canopy cover 0.5 km tree canopy percentage (TCP) Change FAO National Statistics: 1980, 1990, 2000 FAO Deforestation in Tropics: 1980, 1990, 2000 TREES: 1990, 1997 AVHRR Time Series: 1980-2001 (8 km) National: wall-to-wall for some countries

Forest data on maps: Weaknesses (1) : 

Forest data on maps: Weaknesses (1) Non-hierarchical Resolution: strength and issue Classification system: not optimum system / resolution Validation Data limited Variable accuracy Limited historical knowledge Does not discriminate (always) between natural and anthropogenic classes (land cover not land use)

Forest data on maps: Weaknesses (2) : 

Forest data on maps: Weaknesses (2) Calling land cover habitat Classifications generalize; hard to extract parameters Difficult to detect change Not multi-scale, nested sampling

Forest data on maps: Solutions : 

Forest data on maps: Solutions Parameter-based classification More validation Field data collection and integration Need local input and linking of national and global efforts Importance of continued low- or no-cost satellite data delivery to partners Need input from local institutions on natural versus anthropogenic

Forest change data: Weaknesses : 

Forest change data: Weaknesses Only FRA have 3 points, but forest definition changed between the two studies Sampling limitations (stratification, intensity…) Very coarse resolution of AVHRR time-series (8 km)

Forest change data: Solutions : 

Forest change data: Solutions Revisit the deforestation measurement sites Medium-resolution data for mapping changes Maintain same forest definitions

Forest data gaps : 

Forest data gaps Temperate and boreal forests Accurate map of forest changes Degradation (logging, hunting…) Changes in woodlands, grasslands, shrublands, agriculture…

Ideal scenario for forests : 

Ideal scenario for forests Coarse resolution global yearly land cover (300-500 m) High resolution sampling yearly global forests (including degradation) High resolution 5-yearly wall-to-wall for priority regions (including degradation)

Indicator development : 

Indicator development The following slides illustrate the thinking of the group on development of indicators for the following: forest extent and quality shrublands, grasslands and desert extent and quality croplands extent freshwater catchment condition, extent of wetlands and large water bodies, and fragmentation (dams) coral cover, and location and extent of mangroves, seagrasses and macroalgal beds extent and quality of estruaties marine trophic index

Indicator of forest extent : 

Indicator of forest extent How: Remote sensing with aerial and ground validation By whom: National Agencies, NGOs, Universities £$: < $5M per year to support multiple efforts and resolutions Repeatability: 3+measures by 2010 Weaknesses & caveats: General class scheme (e.g. IGBP), assumes data continuity and low-cost delivery to users (e.g. next Landsat? but next Landsat launch will probably not take place before 2009, making 2010 a challenge. There might be other data available??) Data nearly in hand; indicator deliverable by 2010

Indicator of forest quality : 

Indicator of forest quality How: Fragmentation via GIS of maps of extent; degradation via samples of high-resolution mapping and ground survey samples; wood volume harvest data; distance to roads; various measures of protection and concessions (assessed by other group) By whom: National Agencies, NGOs, Universities £$: $1-2M per year? Repeatability: 3+ Weaknesses & caveats: Reliability of data on timber harvest? High cost of realistic estimates of hunting rates? Data nearly in hand; indicator deliverable by 2010

Indicator of extent of shrublands, grasslands and deserts (1/2) : 

Indicator of extent of shrublands, grasslands and deserts (1/2) How: Remote sensing By whom: USA (MODLAND science team), EU (GEOLAND), FAO, NGO-Univ consortium £$: Partially paid by Space agencies, universities, existing programs. More support needed. Repeatability: 2, maybe 3 Weaknesses & caveats: Classes of drylands are not discrete; low resolution is a problem; lack of validation; lack of integration of in situ data.

Indicator of extent of shrublands, grasslands and deserts (2/2) : 

Indicator of extent of shrublands, grasslands and deserts (2/2) Recommendations: Pull in additional in situ data and interpretive data from conservation organizations and national agencies Data could be extracted; indicator deliverable by 2010

Indicator of quality of shrublands, grasslands and deserts (1/2) : 

Indicator of quality of shrublands, grasslands and deserts (1/2) How: Meta-analysis by ecoregion; duration of growing season; fire activity from remote sensing By whom: Meta-analysis: TBI; consult with OSS, IBBP-LUCC, Desert Research Institute, space agencies £$: Low for meta-analysis; considerably higher for remote sensing measures. Repeatability: 2, maybe 3? Weaknesses & caveats: Uneven representativeness of sites; biased publications?; inaccuracies inherent with remote sensing of these habitat types

Indicator of quality of shrublands, grasslands and deserts (2/2) : 

Indicator of quality of shrublands, grasslands and deserts (2/2) Recommendations: Consider livestock density; involve national agencies, conservation orgs; interpretation of results—consider case studies/scenarios from smaller areas Data could be extracted; indicator deliverable by 2010

Indicator of cropland extent - rainfed, irrigated, shifting (1/2) : 

Indicator of cropland extent - rainfed, irrigated, shifting (1/2) How: moderate resolution land-cover change (250- 300m) for recent periods, comparison with statistics for historical periods By whom: FAO, universities , USA (MODLAND science team), EU (GEOLAND) £$: partially paid by Space agencies, universities, existing programs. More support needed Repeatability: 2 existing (but 1 old), probably 3 by 2010

Indicator of cropland extent - rainfed, irrigated, shifting (2/2) : 

Indicator of cropland extent - rainfed, irrigated, shifting (2/2) Weaknesses & caveats: Reliable with intensive agriculture, problems with shifting cultivation; spatial resolution; is it related to the 2010 target or should we limit the analysis to the marginal areas? Recommendations: Include national statistics with RS data Data nearly in hand; indicator deliverable by 2010

Indicators of catchment condition: land cover and extent of riparian vegetation : 

Indicators of catchment condition: land cover and extent of riparian vegetation – How: Remote sensing – By whom: As for forests £$: Included in forests Repeatability: 3+ Weaknesses & caveats: Spatial resolution etc. Data could be extracted; indicator deliverable by 2010

Extent of wetlands and large water bodies : 

Extent of wetlands and large water bodies How: Remote sensing By whom: As for forests £$: Included in forests Repeatability: 3+ Weaknesses & caveats: Only large bodies done Data in hand; indicator deliverable by 2010

Freshwater fragmentation indicator: numbers of dams : 

Freshwater fragmentation indicator: numbers of dams How: Remote sensing, existing watersheds database By whom: WRI, CI? £$:?? Repeatability: 3 (based on Landsat Geocover) Weaknesses & caveats: New product Data could be extracted; indicator deliverable by 2010

Percent coral cover : 

Percent coral cover How: Collating existing data By whom: University, NGO? £$: Not much ($10s k) Repeatability: Yearly over 2 decades? Weaknesses & caveats: Representativeness of sites, data availability; continue to examine capabilities of 30m global satellite datasets to detect coral cover & distinguish it from algal and other cover types Data could be extracted; indicator deliverable by 2010

Extent and location of mangroves, seagrasses and macroalgal beds : 

Extent and location of mangroves, seagrasses and macroalgal beds How: Remote sensing By whom: As for forests £$: Not much Repeatability: Maybe 3? Weaknesses & caveats: Spatial resolution Data could be extracted; indicator deliverable by 2010

Extent and quality of estruaries : 

Extent and quality of estruaries How: Remote sensing and meta-analysis of sediment loads & eutrophication By whom: ? £$: ? Repeatability: Maybe 3 Weaknesses & caveats: Spatial resolution; data availability Data could be extracted; indicator deliverable by 2010

Marine Tropic Level Index : 

Marine Tropic Level Index How: Existing FAO data By whom: Seas around us £$: ? Repeatability: 3+ Weaknesses & caveats: Can hide important shifts in community composition; can be biased by changes in fishing preferences; not widely applicable beyond marine ecosystems; length would be less sensitive to biases Data in hand; indicator deliverable by 2010