Metadata Factsheet

3. Goals And Targets Addressed

3a. Goal

Goal A. The integrity of all ecosystems is enhanced, with an increase of at least 15% in the area, connectivity and integrity of natural ecosystems, supporting healthy and resilient populations of all species, the rate of extinctions has been reduced at least tenfold, and the risk of species extinctions across all taxonomic and functional groups, is halved, and genetic diversity of wild and domesticated species is safeguarded, with at least 90% of genetic diversity within all species maintained.

3b. Target

N/A

5. Definitions Concepts And Classifications

5a. Definition

The Living Planet Index is a multi-species indicator which tracks average changes in the relative abundance of species populations over time. The method of aggregating the index is the geometric mean, an established method for biodiversity indicators both globally, for example the Wetland Extent Index (Dixon et al, 2016; Darrah et al, 2019) and the Wild Bird Index (Gregory and van Strien 2010), and nationally, for example, the UK Priority Species Indicator (Eaton et al, 2015) and the Canadian Species Index (Marconi et al, 2021). The unit of measurement are index values starting from a value of 1 in 1970 with the current global LPI running to 2018. The index values are calculated for each subsequent year relative to the previous one, so any given year show the cumulative change in relative abundance compared to the value in 1970.

The index is calculated from almost 32,000 population time-series of vertebrate species from over 7,500 sites globally. These time-series are repeated measures of population size (or a proxy) using the same method (or inconsistent method plus correction factor) for a minimum of two years, not necessarily consecutive. The term population refers to a collection of individuals of the same species that have been monitored in a defined location. The data sources are listed for each data set and come from published scientific literature, government reports, national species monitoring schemes and direct submissions.

The population data are collected from terrestrial, freshwater and marine habitats around the world. Data are collected and added to the database underpinning the index on a continual basis. As such, the LPI can be updated as frequently as annually if needed and if the data are available.

Strengths

The LPI is time sensitive. Annual data points are collected so changes in populations can be picked up relatively quickly. In addition to time series data, a wealth of information on threats, management, geographic and taxonomic information is collected which allows further disaggregation of the LPI and analysis on drivers behind the trends. The LPI can be disaggregated to different scales including national and regional levels and for different habitats. Available data from published and unpublished sources, using a range of data types, are used to build the index, meaning that it is quite cost effective to develop; the underlying database also provides a valuable resource which can be used for other applications beyond the production of the indicator. The LPI is easy to communicate by providing a simple message which is easy for a range of audiences to understand.

Limitations

Data used is currently limited to vertebrates, but efforts are being made to extend this to plants and invertebrates which will be available within the next five years. The taxonomic coverage of vertebrates is also not comprehensive, and the data set is sometimes over-represented by well-studied species and regions. This is taken into account by applying weighting when calculating the index as detailed in McRae et al, 2017. Sufficient data are not currently available to produce a national index for all countries. This is being addressed methodologically by developing techniques for extrapolating species trends in countries where data are lacking. Additionally, targeted data collection is underway to fill gaps for countries where gaps exist. The geometric mean used to calculate the LPI can be sensitive to outliers (Leung et al, 2020; Gregory et al, 2019). However, sensitivity tests are incorporated into the method to check and account for any significant effect on the index and outliers are removed if they are unduly affecting the index.

5b. Method Of Computation

The LPI adopts two different methods to generate index values: a chain method, and a generalized additive modelling technique. There is an R package for using the Living Planet Index method which is fully accessible and has examples of how the method can be applied at the global and regional level.

Box 1: Pseudocode outlining the algorithm for constructing the global Living Planet Index (McRae et al 2017).

Change in abundance is calculated between the baseline year and the end year using a two-stage modelling process, using mgcv 1.8-0 package in R (R Development Core Team 2018), to obtain annual time-series level trends. Long time-series are modelled through a Generalised Additive Modelling (GAM) framework. For short time series (fewer than 6 data points) and time series that fail the GAM, the chain method (log linear interpolation) is used. Annual time series are aggregated geometrically following a standardised method to produce the final index. Within the indices calculated using the unweighted approach, each population carries equal weight within each species and each species carries equal weight within the overall trend. Unless otherwise specified, all populations of a species are weighted equally, and an average value is calculated for a species with more than one population within the data set. This means that species represented by several populations within the data set will have the same influence on the trend as species with only one population within the data set.

where nt is the number of populations, dt is the annual rate of change for a population in a given year, given by

where N is the population measure and t is the year.

Having constructed species, group, regional or global trends, these can be converted back to index values by:

In order to address unevenness in data from different taxonomic groups or regions, a weighted approach to the calculation of the global Living Planet Index was introduced in 2017 (McRae et al, 2017), which can also be used for countries. This approach uses a proportional weighting system which is based on the estimated species richness of a taxonomic group within a region, placing greater weight on trends from those groups which contain the most species even if they are not the groups which contain the most data. Box 1 shows the levels of aggregation and where weightings are applied in the global LPI.

Confidence intervals are calculated by bootstrapping the species trends 10,000 times to generate indices for the upper and lower bounds. These intervals represent the variability in the underlying species trends but do not propagate uncertainty from the population estimates themselves.

5c. Data Collection Method

The primary data in the LPI is collected using a variety of methods which are appropriate to the species and location. These include transect counts, boat surveys, trapping, aerial surveys and visual counts. These data are collated by the LPI team and other organisations involved in producing national or regional LPIs; countries can also submit their data for inclusion into the LPI (see 6.d.3).

The data collation is an ongoing activity to capture data as soon as it is available. This includes the addition of new population time-series to the database and new annual data points to existing population time-series.

Supporting documents are available on the Living Planet Index website which outline the data collection protocols and standards

http://www.livingplanetindex.org/documents/data_standards.pdf and http://www.livingplanetindex.org/documents/lpi_data_guide_data_input.pdf

5d. Accessibility Of Methodology

The LPI methodology has been published in the peer-reviewed literature for both the unweighted (Loh et al, 2005 and Collen et al, 2009) and diversity-weighted versions (McRae et al, 2017). The methodology for the national calculations of the LPI e.g., Australian Threatened Species Index (Bayraktarov et al, 2021) and Canadian Species Index (Marconi et al, 2021) also have been peer- reviewed and are readily available (see References).

All data added to the Living Planet Index Database are publicly available. An exception is when they are provided to the LPI team under the agreement that they should be kept confidential, usually if they contain data for protected species or it is a temporary confidentiality agreement while the data are being used in other publications (in which case they are used as part of the calculation of global trends, but are never shown individually). In case of data sources providing large amounts of data, an MoU is usually signed between the parties. The database has an online form system which allows individuals to enter and store their own data after receiving some training.

After every global analysis of the data, the entire public data set is published in an Excel spreadsheet and can be replicated using the same methodology which is available online https://github.com/Zoological-Society-of-London/rl.... The public data set does not include confidential data (about 4,500 populations – 16%) so an anonymised version of this part of the data set is also made available.

5e. Data Sources

Data are collated from scientific literature, government reports, NGO reports, national species monitoring schemes, conservation practitioners and scientists. Data sets do not have to be published or peer-reviewed to be included in the database. Details of individual data sources are available for each record on the Living Planet Index Database website. Compiling the national data set and producing indices is done by either national agencies – government or non-governmental organisations or by the LPI team, who maintain the global database.

5f. Availability And Release Calendar

The indicator is available now and is updated every two years.

The database is hosted and accessible online (www.livingplanetindex.org) and has a search interface that registered users can operate to access and download the data. Updated versions of the database are made available on the website twice a year. A global Living Planet Index and the accompanying data set is released every two years with the publication of WWF’s Living Planet Report. National LPIs can be updated with the same frequency, depending on the regularity of data updates.

5g. Time Series

The global Living Planet Index is currently available from 1970 to 2018. Usually there is a time lag of 3-4 years for the global LPI. However, national LPIs could close this gap if data is made available as soon as it has been collected.

5h. Data Providers

There are several thousand data sources, mostly scientific papers, which cannot all be listed here. The main data providers are national and regional databases of species monitoring data. Examples of national data providers to the database are the National Biodiversity Data Bank of Uganda and the Open Government Portal of Canada.

5i. Data Compilers

The LPI team based at the Zoological Society of London are responsible for maintaining the database behind the Living Planet Index and for calculating regular updates to the global index. The results are published every two years (WWF 2022; https://www.livingplanetindex.org/latest_results) and the LPI team are responsible for providing regular updates to the Biodiversity Indicators Partnership and for global assessments such as the Global Biodiversity Outlooks.

5j. Gaps In Data Coverage

The LPI is focussed on vertebrate species currently. The underlying data set is subject to similar geographic and taxonomic biases found in other indicators and data sets (Troudet et al, 2017; Donaldson et al, 2016); however, the representation of certain taxa is increasing. For example, in the 2022 LPI, the largest percentage of species were fishes (40%), followed by birds (34%), mammals (14%), and amphibians and reptiles (11%).

The geographic spread of the data shows that Africa and Asia-Pacific are less well-represented than Europe-Central Asia and the Americas. Most species came from Latin America & Caribbean (31%), followed by North America (23%)Europe-Central Asia (15%), Asia-Pacific (18%) and Africa (13%).

5k. Treatment Of Missing Values

At country level

Missing values are interpolated as described in the Computation Method section. No extrapolation of data occurs at time-series level before the start year or beyond the end year. Where insufficient data are available to calculate a Living Planet Index at the national level, techniques are in development (based on Spooner et al, 2019 and Leclère et al, 2020) to calculate spatially extrapolated values.

At regional and global levels

Because the Living Planet Index does not rely on comprehensive coverage of taxonomic groups, there are no specific missing values. However, the weighted approach to the method is employed to compensate for disparity in taxonomic and geographic coverage, including gaps. A weighted value is assigned to a species group based on how diverse (how many species) it is. In addition, where insufficient data are available to calculate a Living Planet Index at the national level, techniques are in development (based on Spooner et al, 2019 and Leclère et al, 2020) to calculate extrapolated values.

6. Scale

6a. Scale Of Use

Scale of application (please check all relevant boxes): Global, Regional, National

Scale of data disaggregation/aggregation:

Global/ regional scale indicator can be disaggregated to national level: Yes

National data is collated to form global indicator:

National data can be collated to produce a global indicator. For example, many national species trends for Canada are used in the global LPI. Conversely, because the data comes from monitoring sites, the global LPI can be disaggregated for national use. However, there is not always data readily available for every country (see 6.b).

6b. National Regional Indicator Production

Recently, the second version of guidelines for national applications of the Living Planet Index was produced (Marconi and McRae 2021). This technical document provides detailed steps to collecting and analysing data through to interpretation and diagnostics and is available from the LPI website. A manuscript outlining the national application of the LPI using the example for Canada has recently been published (Marconi et al, 2021). In addition to the tools for analysis and guidance document, experts from the LPI partnership are happy to provide advice and support for the development of national LPIs.

Underlying data can be accessed by countries and supplemented with their own data if appropriate. Data are available for 199 countries (161 are Parties to the CBD) and the amount of data varies for each country – see national portal for more details: https://livingplanet.shinyapps.io/national_map/. National indices have been produced for Canada, Netherlands, Uganda, Belgium, Austria and are in development for Indonesia (http://ibi.wwf.id/biodiversity/index.php/about/). Australia have developed a national indicator for threatened species of birds, mammals and plants using the LPI method.

We estimate that over 40 countries currently have sufficient data for over 50 species, but many data sources remain untapped so this could be increased relatively quickly. Where insufficient data are available to calculate a Living Planet Index at the national level, techniques are in development (Spooner et al, 2019; Leclère et al, 2020) to calculate extrapolated values.

6c. Sources Of Differences Between Global And National Figures

There shouldn’t be any discrepancy between country estimates as the underlying data will be the same and are a nested part of the global data set. However, as the global LPI is a weighted average, it won’t represent the trends for all countries. The weight each country has in the index varies according to how biodiverse the country is, and how much available data there is for that country.

Disaggregation will highlight differences in biodiversity trends in different areas of the globe (especially between temperate and tropical areas). However, these are genuine differences and do not represent a discrepancy as national-level data are a nested part of the global data set.

6d. Regional And Global Estimates And Data Collection For Global Monitoring

6d.1 Description Of The Methodology

The calculation is the same as for the global LPI (see 5.b.) – this includes the modelling of the population and species trends and the calculation of confidence intervals.

6d.2 Additional Methodological Details

This process is the same as in the calculation section (see 5.b) as the underlying data are the same for the version of the indicator at different scales.

6d.3 Description Of The Mechanism For Collecting Data From Countries

7. Other MEA And Processes And Organisations

7a. Other MEA And Processes

CMS – the LPI is listed as one of the indicators listed in the CMS strategic plan and has been used as an indicator for migratory species and species listed on CMS appendices: Deinet, S. The Living Planet Index (LPI) for species listed on the CMS Appendices. Technical summary submitted to UNEP-WCMC and the CMS Secretariat. 11 (ZSL, UNEP/CMS/COP13/Doc.24/Annex 5, 2019); Latham, J., Collen, B., McRae, L., & Loh, J. (2008). Final Report for The Convention on the Conservation of Migratory Species.

IPBES – use of global LPI, threat summary and LPI for utilised species: IPBES (2019): Global assessment report on biodiversity and ecosystem services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. E. S. Brondizio, J. Settele, S. Díaz, and H. T. Ngo (editors). IPBES secretariat, Bonn, Germany. 1148 pages. https://doi.org/10.5281/zenodo.3831673; IPBES (2022). Thematic Assessment Report on the Sustainable Use of Wild Species of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (1499 pages). Fromentin, J. M., Emery, M. R., Donaldson, J., Danner, M. C., Hallosserie, A., and Kieling, D. (eds.). IPBES secretariat, Bonn, Germany.

Ramsar – use of terrestrial, freshwater and marine LPIs. Finlayson, C. M., Davidson, N., Fennessy, S., Coates, D., Gardner, R. C., Darwall, W., ... & Stroud, D. (2018, October). Section 2: Status and trends. In: Gardner, RC and Finlayson, CM (eds), Global Wetland Outlook: State of the World’s Wetlands and their Services to People. Gland, Switzerland: Ramsar Convention Secretariat. Ramsar Convention Secretariat.

7b. Biodiversity Indicator Partnership

Yes

10. Data Reporter

10a. Organisation

Institute of Zoology, Zoological Society of London

10b. Contact Person

Louise McRae louise.mcrae@ioz.ac.uk

Robin Freeman robin.freeman@ioz.ac.uk