Transcript Document
Informing biodiversity monitoring & reporting designs • A coordinated system for biodiversity monitoring • Trustworthy biodiversity measures • Species occupancy: uses and abuses • Solutions for standardising and mobilising data A coordinated system for biodiversity monitoring Peter Bellingham Multiple reporting obligations International • Convention on Biological Diversity National • NZ Biodiversity Strategy….. ‘maintain and restore a full range of remaining natural habitats and ecosystems to a healthy functioning state’ Internal • Assessing DOC’s performance with respect to achieving its stated outcomes Effective management requires Information on • Where biodiversity outcomes are being achieved • How management interventions can be used to improve outcomes Annual monitoring mortality and rates of all tree Biodiversity in recruitment 2000 3.5 Pirongia Pureora Kaimanawa Tararua Okataina Kaweka Networks of biodiversity information 3.0time-series data with Biased assessments 2.5 No coordination among sites Mostly in managed sites 2.0report losses & gains nationally Can’t Mt Arthur 1.5 Kokatahi Craigieburn Caples Greenstone Murchisons Waitutu 1.0 0.5 A national monitoring system For public conservation lands: 1. National and regional reporting of status and trend in ecological integrity 2. Evaluating the effectiveness of conservation management and policy 3. Informing prioritisation for resource allocation 4. An early-warning system Evaluating ecological integrity Indigenous dominance ‘Are the ecological processes natural?’ Species occupancy ‘Are the species present what you would expect naturally?’ Ecosystem representation ‘Are the full range of ecosystems protected somewhere?’ Biodiversity measures Vegetation 1. Distribution and abundance of exotic weeds considered a threat 2. Size-class structure of canopy dominants 3. Representation of plant functional types Animals 1. Distribution and abundance of exotic pests considered a threat 2. Assemblages of widespread animal species – Birds Sampling framework • 8 x 8 km grid Standardised field surveys • Vegetation • Mammal pests • Birds 5-year rotating-panel design • Unique subset of locations sampled each year Building trustworthy biodiversity measures A proof-of-concept using birds Catriona MacLeod Users of monitoring information Organisation Citizen scientists Iwi Industry Regional councils Central government NZ public Overseas markets International policy Users of monitoring information Organisation Identify data needs Provide data Citizen scientists Iwi Industry Regional councils Central government NZ public Overseas markets International policy Users of monitoring information Organisation Identify data needs Provide data Process & report data Use report Citizen scientists Iwi Industry Regional councils Central government NZ public Overseas markets International policy Meeting multiple stakeholders’ expectations Kakapo Citizen scientist Iwi Industry Regional councils DOC NZ Public International Kiwi Kereru Kaka Tui Skylark Magpie Rosella Meeting multiple stakeholders’ expectations Kakapo Citizen scientist Iwi Industry Regional councils DOC NZ Public International Kiwi Kereru Kaka Tui Skylark Magpie Rosella Meeting multiple stakeholders’ expectations Kakapo Citizen scientist Iwi Industry Regional councils DOC NZ Public International Kiwi Kereru Kaka Tui Skylark Magpie Rosella Data sources and use POWER TO DETECT CHANGE Strong Weak Local SPATIAL ZONE OF INFERENCE Specific landscape National Data sources and use POWER TO DETECT CHANGE Strong NZ bird atlases: national scale Museum collections: national scale Weak Local SPATIAL ZONE OF INFERENCE Specific landscape National Data sources and use POWER TO DETECT CHANGE Strong Traditional Ecological Knowledge: taonga species NZ bird atlases: national scale Historic 5MBC database: Specific study sites NatureWatch & eBird: Locations of interest to observer Weak Local Museum collections: national scale SPATIAL ZONE OF INFERENCE Specific landscape National Data sources and use Strong POWER TO DETECT CHANGE DOC BMRS Tier 2: Managed sites Traditional Ecological Knowledge: taonga species DOC BMRS Tier 1: Public conservation lands NZ Garden bird survey: Urban landscapes NZ bird atlases: national scale Historic 5MBC database: Specific study sites NatureWatch & eBird: Locations of interest to observer Weak Local Museum collections: national scale SPATIAL ZONE OF INFERENCE Specific landscape National Extent of knowledge Numbers, ranges and trends Knowledge development & survey design Rare or concentrated Intermediate Abundance & distribution Widespread & common Knowledge development & survey design Site & species surveys Extent of knowledge Numbers, ranges and trends Generic surveys Atlases Rare or concentrated Intermediate Abundance & distribution Widespread & common Improving data sources and use Strong POWER TO DETECT CHANGE DOC BMRS Tier 2: Managed sites DOC BMRS Tier 1: Public conservation lands Traditional NatureWatch & eBird: Ecological Locations of regional Knowledge: interest NZ Garden bird survey: taonga Urban landscapes species NZ bird atlases: national scale Historic 5MBC database: Specific study sites NatureWatch & eBird: Locations of interest to observer Weak Local Museum collections: national scale SPATIAL ZONE OF INFERENCE Specific landscape National Improving data sources and use Strong POWER TO DETECT CHANGE DOC BMRS Tier 2: Managed sites DOC BMRS Tier 1: Public conservation lands Traditional NatureWatch & eBird: NatureWatch & eBird: Ecological Locations of regional Locations of regional Knowledge: interest interest NZ Garden bird survey: taonga Urban landscapes species NZ bird atlases: Historic 5MBC database: national scale Specific study sites NatureWatch & eBird: Locations of interest to observer Weak Local Museum collections: national scale SPATIAL ZONE OF INFERENCE Specific landscape National Key steps for monitoring design Why? 1. Knowledge focus 2. Action focus Key steps for monitoring design What? 1. Identify target indicators 2. State or dynamic variables? 3. Scale you want to inform? Why? 1. Knowledge focus 2. Action focus Key steps for monitoring design How? 1. 2. 3. 4. Study sites Sampling effort/site Sampling events Sampling method What? 1. Identify target indicators 2. State or dynamic variables? 3. Scale you want to inform? Why? 1. Knowledge focus 2. Action focus Key steps for monitoring design Report 1. 2. 3. 4. Database structure & management Data analysis skills Audit results Report results How? 1. 2. 3. 4. Study sites Sampling effort/site Sampling events Sampling method What? 1. Identify target indicators 2. State or dynamic variables? 3. Scale you want to inform? Why? 1. Knowledge focus 2. Action focus Research aims GOALS & VALUES OF INTEREST Research aims MECHANISMS TO ENHANCE DATA SOURCES GOALS & VALUES OF INTEREST Research aims TRUSTED & USEFUL INDIVIDUAL INDICATORS MECHANISMS TO ENHANCE DATA SOURCES GOALS & VALUES OF INTEREST Research aims EASILY COMMUNICATED AGGREGATED MEASURES TRUSTED & USEFUL INDIVIDUAL INDICATORS MECHANISMS TO ENHANCE DATA SOURCES GOALS & VALUES OF INTEREST Trustworthy biodiversity measures to benefit NZ Process for aggregating & scaling measures Trustworthy biodiversity measures to benefit NZ Process for building engagement & trust Process for aggregating & scaling measures Trustworthy biodiversity measures to benefit NZ Process for building engagement & trust Process for aggregating & scaling measures Ways to improve data sources & reporting PROCESS FOR AGGREGATING & SCALING MEASURES Indicator characteristics reflect goals & values Critical goals to NZ PROCESS FOR AGGREGATING & SCALING MEASURES Benefits & limitations of harmonised reporting Relative value & contributions of different data sources Indicator characteristics reflect goals & values Critical goals to NZ PROCESS FOR AGGREGATING & SCALING MEASURES Benefits & limitations of harmonised reporting Comparable indicators for different scales & needs Relative value & contributions of different data sources Indicator characteristics reflect goals & values Critical goals to NZ PROCESS FOR AGGREGATING & SCALING MEASURES Aggregating & scaling measure for tailored reporting Comparable indicators for different scales & needs Relative value & contributions of different data sources Indicator characteristics reflect goals & values Critical goals to NZ Benefits & limitations of harmonised reporting PROCESS FOR BUILDING ENGAGEMENT & TRUST PROCESS FOR AGGREGATING & SCALING MEASURES Aggregating & scaling measure for tailored reporting Comparable indicators for different scales & needs Relative value & contributions of different data sources Biodiversity values of interest Range of monitoring & reporting goals Indicator characteristics reflect goals & values Critical goals to NZ PROCESS FOR BUILDING ENGAGEMENT & TRUST PROCESS FOR AGGREGATING & SCALING MEASURES Aggregating & scaling measure for tailored reporting Comparable indicators for different scales & needs Data awareness & sharing barriers Data credibility & understanding criteria Biodiversity values of interest Range of monitoring & reporting goals Relative value & contributions of different data sources Indicator characteristics reflect goals & values Critical goals to NZ PROCESS FOR BUILDING ENGAGEMENT & TRUST PROCESS FOR AGGREGATING & SCALING MEASURES Aggregating & scaling measure for tailored reporting Individual indicators are useful & trusted Data awareness & sharing barriers Data credibility & understanding criteria Biodiversity values of interest Range of monitoring & reporting goals Comparable indicators for different scales & needs Relative value & contributions of different data sources Indicator characteristics reflect goals & values Critical goals to NZ PROCESS FOR BUILDING ENGAGEMENT & TRUST PROCESS FOR AGGREGATING & SCALING MEASURES Aggregated measures are easily communicated & understood Aggregating & scaling measure for tailored reporting Individual indicators are useful & trusted Comparable indicators for different scales & needs Data awareness & sharing barriers Data credibility & understanding criteria Biodiversity values of interest Range of monitoring & reporting goals Relative value & contributions of different data sources Indicator characteristics reflect goals & values Critical goals to NZ PROCESS FOR BUILDING ENGAGEMENT & TRUST PROCESS FOR AGGREGATING & SCALING MEASURES Aggregated measures are easily communicated & understood Aggregating & scaling measure for tailored reporting Individual indicators are useful & trusted Comparable indicators for different scales & needs Data awareness & sharing barriers Data credibility & understanding criteria Biodiversity values of interest Range of monitoring & reporting goals WAYS TO IMPROVE DATA SOURCES & REPORTING Relative value & contributions of different data sources Indicator characteristics reflect goals & values Critical goals to NZ Communication strategies to cross social boundaries Mechanisms to collaborate on shared goals PROCESS FOR BUILDING ENGAGEMENT & TRUST PROCESS FOR AGGREGATING & SCALING MEASURES Aggregated measures are easily communicated & understood Aggregating & scaling measure for tailored reporting Individual indicators are useful & trusted Comparable indicators for different scales & needs Data awareness & sharing barriers Data credibility & understanding criteria Biodiversity values of interest Range of monitoring & reporting goals WAYS TO IMPROVE DATA SOURCES & REPORTING Relative value & contributions of different data sources Cost-effective ways to address gaps & improve data Indicator characteristics reflect goals & values Communication strategies to cross social boundaries Critical goals to NZ Mechanisms to collaborate on shared goals PROCESS FOR BUILDING ENGAGEMENT & TRUST PROCESS FOR AGGREGATING & SCALING MEASURES Aggregated measures are easily communicated & understood Aggregating & scaling measure for tailored reporting Individual indicators are useful & trusted Data awareness & sharing barriers Data credibility & understanding criteria Biodiversity values of interest Range of monitoring & reporting goals Comparable indicators for different scales & needs Relative value & contributions of different data sources Indicator characteristics reflect goals & values Critical goals to NZ WAYS TO IMPROVE DATA SOURCES & REPORTING Ways for stakeholders to identify ‘fit-for-purpose’ indicators Cost-effective ways to address gaps & improve data Communication strategies to cross social boundaries Mechanisms to collaborate on shared goals PROCESS FOR BUILDING ENGAGEMENT & TRUST PROCESS FOR AGGREGATING & SCALING MEASURES WAYS TO IMPROVE DATA SOURCES & REPORTING Aggregated measures are easily communicated & understood Aggregating & scaling measure for tailored reporting Benefits & limitations of harmonised reporting Individual indicators are useful & trusted Comparable indicators for different scales & needs Data awareness & sharing barriers Data credibility & understanding criteria Biodiversity values of interest Range of monitoring & reporting goals Relative value & contributions of different data sources Indicator characteristics reflect goals & values Critical goals to NZ Ways for stakeholders to identify ‘fit-for-purpose’ indicators Cost-effective ways to address gaps & improve data Communication strategies to cross social boundaries Mechanisms to collaborate on shared goals Harmonised system for different needs International National Regional Site/farm Occupancy: Uses and abuses Andrew Gormley Landcare Research What is occupancy? • Occupancy is a robust measure of distribution of plants or animals in the landscape • Key indicator of ecological integrity • Metrics: 1. Probability that a site is occupied 2. Proportion of area occupied (PAO) Presence-only data • Locations of species – Specimens, sightings etc Presence data Date Species Lat Long 15/10/13 Kea 43.8 S 172.9 E 16/10/13 Kea 43.7 S 172.7 E Presence-only data • Locations of species – Specimens, sightings etc • Draw a shape around points to indicate its distribution Presence data Date Species Lat Long 15/10/13 Kea 43.8 S 172.9 E 16/10/13 Kea 43.7 S 172.7 E Presence-only data • Locations of species – Specimens, sightings etc • Draw a shape around points to indicate its distribution • Can determine habitat suitability Habitat suitability Presence-only data • Locations of species – Specimens, sightings etc • Draw a shape around points to indicate its distribution • Can determine habitat suitability • Subject to sampling bias • No estimate of uncertainty Didn’t look here Presence-absence data • • • • Sample proportion of possible sites Record where species is present and absent Occupancy = proportion of sites that are occupied More reliable estimates of potential distribution Iratus roseii was at 72 % of sites. Occupancy = 0.72 Presence-Absence Data Date Species Lat Long Status 15/10 I.roseii 43.8 S 172.9 E Present 16/10 I.roseii 43.7 S 172.7 E Absent Measuring occupancy • Can stratify by land cover or other covariates I. roseii at 89% of forest and 56% of non-forest sites Forest: Present in 8 out of 9 Occ. = 0.89 Pasture: Present in 5 out of 9 Occ. = 0.56 Issue 1: Imperfect detection • Species is present but you miss it – Detections might be related to habitat/landcover Forest: Present in 8 Observed in 3 Pasture: Present in 5 Observed in 4 Issue 1: Imperfect detection • Species is present but you miss it – Sampling and statistical methods available • Presence-Absence Data w/repeat surveys Date Species Lat Long Survey1 Survey2 Survey3 15/10/13 Kererū 43.8 S 172.9 E Present Absent Present 16/10/13 Kererū 43.7 S 172.7 E Absent Absent Absent Issue 2: Data from managed areas only • Managed areas are not representative of entire region – Low due to hence mgmt. of predators (?) – High due to mgmt. of predators (?) Pest control No control – not measured Issue 3: Different Data Formats • Presence Data Date Species Lat Long 15/10/13 Kererū 43.8 S 172.9 E 16/10/13 Kererū 43.7 S 172.7 E • Presence-Absence Data • Other data Date Species Lat Long Status 15/10/13 Kererū 43.8 S 172.9 E Present 16/10/13 Kererū 43.7 S 172.7 E Absent – – – – – • Presence-Absence Data w/repeat surveys Date Species Lat Long Survey1 Survey2 Survey3 15/10/13 Kererū 43.8 S 172.9 E Present Absent Present 16/10/13 Kererū 43.7 S 172.7 E Absent Absent Absent Survey ID Person Method Time Weather Issue 4: Scale of sampling unit • Occupancy decreases as your sampling unit gets smaller • Issue for pasting together different sources of data. Issue 4: Scale of sampling unit • Occupancy decreases as your sampling unit gets smaller • Issue for pasting together different sources of data. Present in 2 of 9 plots (22%) Issue 4: Scale of sampling unit • Occupancy decreases as your sampling unit gets smaller • Issue for pasting together different sources of data. Present in 2 of 9 plots (22%) Present in 7 of 9 plots (78%) Issue 4: Scale of sampling unit • Bird Atlas has 3166 × 10 km2 grid squares – Brown Kiwi in 176: occupancy = 0.06 Issue 4: Scale of sampling unit • Bird Atlas has 3166 × 10 km2 grid squares – Brown Kiwi in 176: occupancy = 0.06 • If sampling unit was up to 100 km2 (61 squares) – Brown Kiwi in 26: occupancy = 0.43 Issue 5: Data Quality • Species is misidentified • How to assess quality of the record? • Reliability of observers Issue 6: What is a presence? • What constitutes a positive detection? – Specimen? – Sighting? – Sound? • Other species… – Poo? – Sign? – Remote sensing? Level of data storage • Site Summary Data Site Species Status AA144 Bellbird Present AD156 Bellbird Absent … … … Tier 1 sampling • • • Conservation lands National and regional scales Grid, random start point Level of data storage • Site Summary Data Site Species Status AA144 Bellbird Present AD156 Bellbird Absent … … … • Summary Data Site Station Species Status AA144 A Bellbird Present AA144 D Bellbird Present AA144 M Bellbird Present AA144 X Bellbird Absent AA144 P Bellbird Absent Level of data storage • Site Summary Data Site Species Status AA144 Bellbird Present AD156 Bellbird Absent … … … • Raw Data • Summary Data Site Station Species Status AA144 A Bellbird Present AA144 D Bellbird Present AA144 M Bellbird Present AA144 X Bellbird Absent AA144 P Bellbird Absent Site Station Species Status AA144 A Bellbird Present AA144 A Bellbird Present AA144 A Bellbird Present AA144 A Bellbird Present AA144 D Bellbird Present AA144 M Bellbird Present AA144 X Bellbird Present AA144 M Bellbird Present AA144 P Bellbird Absent Have to document how raw data is summarised for analysis Distribution vs Abundance Distribution is better! • Estimating abundance is too hard and/or expensive – Easier to detect species rather than count individuals Distribution vs Abundance Distribution is better! • Estimating abundance is too hard and/or expensive – Easier to detect species rather than count individuals • Distribution is a good surrogate for abundance 2010 – As a population increases, so does its distribution 2013 Distribution vs Abundance Abundance is better! • Distribution does not provide enough detail • Abundance may change with no change in distribution – Species is widespread and then has localised increases in population – Species goes into decline but remains widespread • Distribution will not detect this Standardising and mobilising data Nick Spencer Goal: solutions for standardisation and mobilisationideally through e-federation of distribution data; What are the barriers to delivering this? Federated Bio-data Misc data Specimens Observations Databases Services Federated Bio-data Misc data Specimens Observations Databases GBIF.ORG Free and open access to biodiversity data Services Federated Bio-data Misc data Confederated Bio-data Specimens And/Or Observations Databases Services Federated Networked Bio-data Federated Bio-data Misc data Confederated Bio-data National Reporting National Modelling Specimens And/Or Observations Evidential Decision Making Databases Federated Networked Bio-data GBIF.ORG Free and open access to biodiversity data Services International Reporting Discovery | Mobilisation | Integration Barriers Issues Consequences Tier One Monitoring Vegetation Component Solutions GBIF.ORG Free and open access to biodiversity data Botanical Information and Ecology Network National Vegetation Survey Databank Indigenous Vegetation Discovery | Mobilisation | Integration Barriers Issues Consequences Tier One Monitoring Vegetation Component Solutions GBIF.ORG Free and open access to biodiversity data Botanical Information and Ecology Network National Vegetation Survey Databank Indigenous Vegetation Barrier Issue Consequence Solution Barrier Schemas Issue Consequence Solution Barrier Issue • Incompatible schemas Schemas • Constraints • Missing elements Consequence Solution Barrier Issue • Incompatible schemas Schemas • Constraints • Missing elements Consequence • Restructuring costs • Risk of incorrectly combining elements • Risk to implying data are equivalent Solution Barrier Issue • Incompatible schemas Schemas • Constraints • Missing elements Consequence • Restructuring costs • Risk of incorrectly combining elements • Risk to implying data are equivalent Solution • • • • Standard methods Standard schema Be realistic Don’t underestimate effort Barrier Issue Consequence Solution Schemas Geographical Barrier Issue Consequence Solution Schemas Geographical • Missing or inaccurate geo-references • Cultivated specimens Barrier Issue Consequence Solution Schemas Geographical • Missing or inaccurate geo-references • Cultivated specimens • Unusable data • Distribution and rarity estimation errors • Effort to resolve issues Barrier Issue Consequence Solution Schemas Geographical • Missing or inaccurate geo-references • Cultivated specimens • Unusable data • Distribution and rarity estimation errors • Effort to resolve issues • Geo-ref’s for locations • Rules for misspellings; valid coordinates; sensible locations; consistency with location narrative; known distributions; collector routes Barrier Issue Consequence Solution Schemas Geographical Attribution Barrier Issue Consequence Solution Schemas Geographical • Tracking data source Attribution • Original records • Derived data Barrier Issue Consequence Solution Schemas Geographical • Tracking data source Attribution • Original records • Derived data Without source information integrated data is of dubious quality Barrier Issue Consequence Solution Schemas Geographical • Tracking data source Attribution • Original records • Derived data Without source information integrated data is of dubious quality • Collect metadata and method information • Ensure this remains with the data • Aggregate from source records Barrier Issue Consequence Solution Schemas Geographical Attribution Organism names Barrier Issue Consequence Solution Schemas Geographical Attribution • Misspellings Organism names • Taxonomic concepts • Taxonomically homogenous datasets Barrier Issue Consequence Solution Schemas Geographical Attribution • Misspellings Organism names • Taxonomic concepts • Taxonomically homogenous datasets • Inflates species richness • Reduces range estimates • Poor decision about protection or mitigation Barrier Issue Consequence Solution Schemas Geographical Attribution • Misspellings Organism names • Taxonomic concepts • Taxonomically homogenous datasets • Inflates species richness • Reduces range estimates • Poor decision about protection or mitigation • NZOR provides consensus of synonymy, spellings and concepts (but...) • Must be applied to data to create taxonomically homogenous datasets Concepts - Nertera dichondrifolia Until MacMillan (1995) Nertera dichondrifolia regarded as variable species distributed throughout NZ N. dichondrifolia ? Nertera Spp After 1995 N. villosa circumscribed; occurs South of latitude 37o N. villosa N. dichondrifolia Concept of N. dichondrifolia narrowed; occurs North of latitude 38o Nertera dichondrifolia Between 37o and 38o latitude uncertainty about which species you have N. villosa N. dichondrifolia ? Barrier Issue Consequence Solution Schemas Geographical Attribution Organism names Bio-data Services Stack Terrestrial Work Stream 1 Species Occupancy Where do species occur? – If data is poor our ability to answer the other questions well is limited Data challenges Bird Data with Regional Council data holders Discovery & Mobilisation • • • • Identifying, Describing, Data use agreements, Technology assistance and challenges