WWF · ZSL · Global Biodiversity Indicator
The Living Planet
Index
A barometer of the health of Earth’s vertebrate populations — tracking wild species abundance across freshwater, marine, and terrestrial ecosystems since 1970.
73%
avg. wildlife decline
1970–2020
1970–2020
5,495
species monitored
(LPR 2022)
(LPR 2022)
32,000+
population
time-series
time-series
1970
baseline
year (LPI = 1)
year (LPI = 1)
Scroll to explore
What Is the Living Planet Index?
Description & Overview
The Living Planet Index (LPI) is a measure of the state of global biological diversity based on population trends of vertebrate species from terrestrial, freshwater, and marine habitats. Published biennially by the World Wildlife Fund (WWF) in collaboration with the Zoological Society of London (ZSL), the LPI acts as a vital-sign monitor for Earth’s wildlife, much as a doctor’s thermometer reads body temperature.
The index aggregates time-series data on the abundance of thousands of vertebrate populations worldwide. Each population is tracked over time; the relative change in size is calculated; and these changes are averaged first within species, then across taxonomic groups, and finally across the three realms — to produce a single global index value.
The LPI is anchored to 1970 = 1.00 (the baseline). A value of 0.73 in 2020, for instance, means that monitored vertebrate populations were, on average, 73% below their 1970 levels — a staggering 73% average decline in just 50 years.
Scope: The LPI covers vertebrates (mammals, birds, reptiles, amphibians, and fish). It does NOT directly measure species extinction, plant diversity, or invertebrate populations — but it is the world’s most widely cited biodiversity trend indicator and a core metric of the Kunming-Montreal Global Biodiversity Framework.
Global LPI Trend (1970–2020, Schematic)
Index value relative to 1970 baseline (1.00)
Global LPI (with CI band)
Freshwater LPI
Terrestrial LPI
Historical Development
A History of the Living Planet Index
The LPI emerged from a convergence of conservation biology, longitudinal ecological data, and the growing need for a headline biodiversity indicator that policymakers could readily understand.
1970 — Baseline Year
The Chosen Anchor. 1970 was selected as the LPI baseline not because it was ecologically pristine, but because it marks the earliest year for which sufficiently broad, reliable population data could be assembled across all three realms (terrestrial, freshwater, marine). It predates many major conservation frameworks and represents a pre-globalisation ecological state.
1998 — First Publication
WWF Living Planet Report 1998. The LPI was formally introduced by Stuart Pimm, Gretchen Daily, and WWF/ZSL scientists in the inaugural Living Planet Report. The first version covered 4,000+ populations of ~1,400 vertebrate species. It used a simple geometric mean of population indices. The publication was a landmark — translating complex biodiversity trends into a single, communicable number for the first time.
2000s — Methodological Refinement
Realm-Specific Indices. Separate LPI values were computed for terrestrial, freshwater, and marine systems, acknowledging their different trajectories and data availabilities. Species-level averaging (before realm-level aggregation) was introduced to prevent highly monitored species from dominating the global index.
2005 — WWF / ZSL Formal Partnership
ZSL Living Conservation Programme. The Zoological Society of London formalised its partnership with WWF, taking primary scientific responsibility for LPI database management, methodology, and peer review. ZSL developed the rlpi R package for transparent, reproducible LPI computation, making the method open to independent researchers.
2006 — Convention on Biological Diversity (CBD)
Adopted as an Official CBD Headline Indicator. The CBD adopted the LPI as one of its core 2010 Biodiversity Target headline indicators. This formal adoption elevated the LPI from a WWF communication tool to a globally recognised policy metric, used by governments to track progress on biodiversity commitments under the Convention.
2010 — Aichi Biodiversity Targets
Strategic Plan 2011–2020. The LPI was embedded in the Nagoya Protocol era CBD monitoring framework. Despite the Aichi targets aiming to halt biodiversity loss by 2020, the LPI showed continued steep decline throughout the decade.
2012 — Weighted Aggregation Introduced
Geographic Weighting. To correct for biased data coverage (far more monitored populations in Europe and North America than in tropical regions where biodiversity loss is greatest), geographic weighting schemes were introduced. This significantly changed the global LPI value and led to higher estimated declines, especially in tropical freshwater systems.
2022 — LPR Report: 73% Decline Announced
Living Planet Report 2022. This edition reported that monitored vertebrate populations had declined by an average of 69% (later revised to 73% in updated analyses) between 1970 and 2018/2020. Latin America and the Caribbean showed 94% decline. The report galvanised negotiations for the Kunming-Montreal Global Biodiversity Framework (GBF).
2022–Present — Kunming-Montreal GBF
Post-2020 Biodiversity Framework. The LPI was retained as a core monitoring indicator for the Kunming-Montreal Global Biodiversity Framework, adopted in December 2022, with its “30×30” target (protect 30% of land and sea by 2030). The LPI now feeds directly into national reporting and global conservation policy at the highest level.
Why It Matters
Significance & Relevance
The LPI transcends a simple count — it is a dynamic barometer of ecological integrity that connects biodiversity science to global policy, education, and resource management.
Global Policy Anchor
Formally adopted by the CBD and the Kunming-Montreal GBF as a headline indicator. LPI values directly influence international biodiversity agreements and national reporting obligations.
Trend Detection
By tracking populations over decades, the LPI reveals directional trends (increasing, stable, or declining) that snapshot species counts cannot show. Rate of change matters as much as the state.
Ecosystem Health Proxy
Vertebrate population trends act as proxies for broader ecosystem health. Declines in apex predators, herbivores, or keystone species signal structural ecosystem degradation.
Identifies Drivers
Disaggregated LPI data (by realm, region, taxonomic group, or threat) allow identification of primary drivers of wildlife loss — habitat destruction, overexploitation, invasive species, climate change.
Conservation Effectiveness
Protected area LPI can be compared to unprotected area LPI, directly measuring whether conservation investments are translating into population recoveries on the ground.
Communication Tool
A single number (e.g. “73% decline since 1970”) communicates a complex global reality to the public, politicians, and media with immediate impact — something tables of species data cannot achieve.
Biodiversity Hotspot Monitoring
Regional LPIs (e.g. Tropical LPI, Freshwater Tropical LPI) pinpoint where declines are most severe, directing conservation resources and priority-setting in the most threatened biomes.
Reproducibility & Openness
The ZSL’s open-source rlpi R package makes LPI computation transparent and reproducible, enabling regional and national organisations to calculate their own LPIs using standardised methods.
Limitations to note: The LPI is not an extinction metric — it measures population abundance trends, not species losses. It is biased toward well-monitored species in data-rich regions (Europe, North America). Taxonomic coverage excludes plants, invertebrates, and fungi. These constraints are acknowledged by WWF/ZSL and are being addressed through geographic weighting and database expansion.
Core Methodology
LPI Methodology — Step by Step
How thousands of wildlife population time-series are transformed into a single global index value
The LPI methodology involves four hierarchical stages of data aggregation, moving from individual population records to a global composite index. Each stage applies specific statistical transformations to ensure that the final index is not dominated by a few well-monitored species or regions.
① Data Collection
- Time-series data on vertebrate population size (counts, density, index measures) are collected from peer-reviewed literature, monitoring programmes, and institutional databases.
- Each population must have ≥2 data points ≥4 years apart.
- Data cover mammals, birds, reptiles, amphibians, and fishes across terrestrial, freshwater, and marine realms.
- As of LPR 2022: >32,000 population time-series for 5,495 species.
② Population Index Calculation
- For each population, an annual index is computed relative to its own first-year value (set to 1.00).
- A linear interpolation fills gaps between measured years (log-scale interpolation).
- The log ratio of consecutive values is used: dt = log(Nt+1/Nt)
- This removes units and allows comparison across species with vastly different absolute abundances.
③ Species-Level Aggregation
- If multiple populations of the same species are monitored, their log ratios are averaged to give a single species-level trend.
- This ensures no single common species with many monitored populations dominates the index.
- The species-level index is then reconstructed from the averaged log ratios.
④ Realm & Global Aggregation
- Species indices are averaged using a geometric mean within each realm (terrestrial, freshwater, marine).
- The three realm-level LPIs are then averaged (often with equal weighting) to produce the global LPI.
- Geographic weighting can be applied to correct for taxonomic and spatial data bias.
- Uncertainty intervals are computed using bootstrapping.
LPI Computational Pipeline
From raw population counts to the global index value
Mathematics
Equations & Formulae
Step 1 — Annual Log Ratio (Population Rate of Change)
dᵗ = log₁₀(Nᵗ₊₁ / Nᵗ)
dᵗLog ratio of population size in consecutive years t and t+1. Measures the rate of change in one time step.
NᵗPopulation size (or relative abundance index) at time t
Nᵗ₊₁Population size at time t+1
log₁₀Common logarithm (base 10). Some implementations use natural log (ln). The choice does not change the final index value, only scale.
Why log ratios? Using log(Nᵗ₊₁/Nᵗ) rather than (Nᵗ₊₁−Nᵗ)/Nᵗ produces symmetric, additive values: a doubling (+0.301) and a halving (−0.301) cancel out correctly. It also prevents large absolute populations from dominating trends.
Step 2 — Reconstructed Population Index (Iᵗ)
Iᵗ = 10^(d₁ + d₂ + d₃ + … + dᵗ₋₁) = 10^(Σ dᵔ ) for k = 1 to t-1
IᵗIndex value for this population at year t, relative to baseline year (I₁ = 1.00 by definition)
Σ dᵔCumulative sum of log ratios from year 1 to year t−1. This is a running total of how much the population has changed in log space.
10^(…)Anti-log to convert back from log space to a ratio scale. Result: 1.00 at baseline, >1 if population grew, <1 if declined.
Step 3 — Species-Level Mean Log Ratio
Dᵗ̂ = (1/n) × Σ dᵗ₀ᵢ for i = 1 to n populations
Dᵗ̂Average log ratio across all n monitored populations of the same species at year t
nNumber of monitored populations for this species
dᵗ₀ᵢLog ratio of the i-th population at time t
Species-level averaging: This prevents a species with 50 monitored populations (e.g. a common bird) from having 50 times the weight of a species with just one population. Each species gets exactly one vote in the next aggregation step.
Step 4 — Realm-Level LPI (Geometric Mean of Species Indices)
LPIᵗ = 10^( (1/S) × Σ log₁₀(Iᵗ̂ᵎ) ) for j = 1 to S species
LPIᵗLiving Planet Index at year t for this realm (terrestrial, freshwater, or marine)
SNumber of species included in this realm’s index
Iᵗ̂ᵎReconstructed species-level index for species j at year t (computed from D̂ values)
10^(…)The exponent of the mean log is equivalent to the geometric mean — used because it is robust to extreme values and gives equal weight on a multiplicative scale.
Step 5 — Global LPI (Equal-Weighted Realm Average)
Global LPIᵗ = (LPIᵗ̂ₒₑₐ + LPIᵗ̂ᵅₐ₊ + LPIᵗ̂ᵎᵃₐ) / 3
LPÎₒₑₐTerrestrial realm LPI at year t
LPÎᵅₐ₊Freshwater realm LPI at year t
LPÎᵎᵃₐMarine realm LPI at year t
Note on weighting: The equal weighting of three realms is a deliberate methodological choice. Alternative weightings (by area, by species richness, by threat level) produce different global LPI values. The choice is explicit and reported in all LPR publications. Geographic weighting within realms is also applied to correct for data bias toward well-monitored regions.
Derived — Percentage Change from Baseline
% Change = (LPIᵗ − LPI₁┹⁷₀) / LPI₁┹⁷₀ × 100
Example: If LPI₂₀₂₀ = 0.27, then % Change = (0.27 − 1.00)/1.00 × 100 = −73%. This is the headline figure reported in the LPR 2022: “vertebrate populations have declined by an average of 73% since 1970.”
Worked Example
Step-by-Step LPI Calculation
Scenario: We monitor 3 vertebrate species across a freshwater ecosystem over 5 years (Year 1 to Year 5). Each species has 2 monitored populations. We will compute the species-level index, the realm-level LPI, and interpret the result.
Raw Population Count Data
| Species | Population | Year 1 | Year 2 | Year 3 | Year 4 | Year 5 |
|---|---|---|---|---|---|---|
| A — Giant Otter | Pop. A1 | 120 | 108 | 90 | 78 | 65 |
| A — Giant Otter | Pop. A2 | 85 | 80 | 72 | 60 | 50 |
| B — Nile Tilapia | Pop. B1 | 2400 | 2200 | 1900 | 1600 | 1400 |
| B — Nile Tilapia | Pop. B2 | 3100 | 2800 | 2500 | 2300 | 2100 |
| C — Marsh Frog | Pop. C1 | 540 | 580 | 610 | 640 | 670 |
| C — Marsh Frog | Pop. C2 | 410 | 430 | 460 | 490 | 520 |
Step 1 — Compute Annual Log Ratios (dᵗ)
dᵗ = log₁₀(Nᵗ₊₁ / Nᵗ) for each population and each year interval.
| Population | d₁ (Yr1→2) | d₂ (Yr2→3) | d₃ (Yr3→4) | d₄ (Yr4→5) |
|---|---|---|---|---|
| Pop. A1 | log(108/120) = −0.0458 | log(90/108) = −0.0792 | log(78/90) = −0.0623 | log(65/78) = −0.0786 |
| Pop. A2 | log(80/85) = −0.0253 | log(72/80) = −0.0458 | log(60/72) = −0.0792 | log(50/60) = −0.0792 |
| Pop. B1 | log(2200/2400) = −0.0378 | log(1900/2200) = −0.0637 | log(1600/1900) = −0.0745 | log(1400/1600) = −0.0581 |
| Pop. B2 | log(2800/3100) = −0.0444 | log(2500/2800) = −0.0491 | log(2300/2500) = −0.0362 | log(2100/2300) = −0.0398 |
| Pop. C1 | log(580/540) = +0.0315 | log(610/580) = +0.0221 | log(640/610) = +0.0207 | log(670/640) = +0.0197 |
| Pop. C2 | log(430/410) = +0.0207 | log(460/430) = +0.0296 | log(490/460) = +0.0277 | log(520/490) = +0.0253 |
Step 2 — Species-Level Mean Log Ratios (D̂)
Average the dᵗ values across populations of the same species for each year.
| Species | D̂₁ (Yr1→2) | D̂₂ (Yr2→3) | D̂₃ (Yr3→4) | D̂₄ (Yr4→5) |
|---|---|---|---|---|
| A — Giant Otter | (−0.0458 + −0.0253)/2 = −0.0356 | (−0.0792 + −0.0458)/2 = −0.0625 | (−0.0623 + −0.0792)/2 = −0.0708 | (−0.0786 + −0.0792)/2 = −0.0789 |
| B — Nile Tilapia | (−0.0378 + −0.0444)/2 = −0.0411 | (−0.0637 + −0.0491)/2 = −0.0564 | (−0.0745 + −0.0362)/2 = −0.0554 | (−0.0581 + −0.0398)/2 = −0.0490 |
| C — Marsh Frog | (+0.0315 + +0.0207)/2 = +0.0261 | (+0.0221 + +0.0296)/2 = +0.0259 | (+0.0207 + +0.0277)/2 = +0.0242 | (+0.0197 + +0.0253)/2 = +0.0225 |
Step 3 — Reconstruct Species Index (Iᵗ)
Iᵗ = 10^(ΣD̂ᵔ) for k=1 to t−1. I₁ = 1.00 (baseline) for all species.
A
Giant Otter (Species A) — Cumulative Index
Yr1: I = 1.0000 (baseline)
Yr2: I = 10^(-0.0356) = 0.9222
Yr3: I = 10^(-0.0356 + -0.0625) = 10^(-0.0981) = 0.7970
Yr4: I = 10^(-0.0981 + -0.0708) = 10^(-0.1689) = 0.6780
Yr5: I = 10^(-0.1689 + -0.0789) = 10^(-0.2478) = 0.5651
B
Nile Tilapia (Species B) — Cumulative Index
Yr1: I = 1.0000 (baseline)
Yr2: I = 10^(-0.0411) = 0.9099
Yr3: I = 10^(-0.0411 + -0.0564) = 10^(-0.0975) = 0.7987
Yr4: I = 10^(-0.0975 + -0.0554) = 10^(-0.1529) = 0.7032
Yr5: I = 10^(-0.1529 + -0.0490) = 10^(-0.2019) = 0.6281
C
Marsh Frog (Species C) — Cumulative Index
Yr1: I = 1.0000 (baseline)
Yr2: I = 10^(+0.0261) = 1.0622
Yr3: I = 10^(+0.0261 + +0.0259) = 10^(+0.0520) = 1.1270
Yr4: I = 10^(+0.0520 + +0.0242) = 10^(+0.0762) = 1.1920
Yr5: I = 10^(+0.0762 + +0.0225) = 10^(+0.0987) = 1.2548
Step 4 — Freshwater Realm LPI (Geometric Mean)
The geometric mean of the three species indices at each year = 10^(mean of log species indices).
| Year | I (Giant Otter) | I (Nile Tilapia) | I (Marsh Frog) | log(Iₐ) | log(Iₒ) | log(Iₓ) | Mean of logs | LPI (Freshwater) |
|---|---|---|---|---|---|---|---|---|
| 1 | 1.0000 | 1.0000 | 1.0000 | 0.0000 | 0.0000 | 0.0000 | 0.0000 | 1.0000 |
| 2 | 0.9222 | 0.9099 | 1.0622 | -0.0354 | -0.0410 | +0.0261 | -0.0168 | 0.9628 |
| 3 | 0.7970 | 0.7987 | 1.1270 | -0.0985 | -0.0975 | +0.0519 | -0.0480 | 0.8948 |
| 4 | 0.6780 | 0.7032 | 1.1920 | -0.1688 | -0.1529 | +0.0763 | -0.0818 | 0.8280 |
| 5 | 0.5651 | 0.6281 | 1.2548 | -0.2479 | -0.2019 | +0.0986 | -0.1171 | 0.7633 |
| LPI₅ = 10^(mean of logs) = 10^(-0.1171) = 0.7633 — a 23.7% decline over 5 years | ||||||||
5
Final Interpretation — Percentage Change
LPI at Year 1 (baseline) = 1.0000
LPI at Year 5 = 0.7633
% Change = (0.7633 - 1.0000) / 1.0000 x 100
= -0.2367 x 100
= -23.67%
Interpretation: Monitored vertebrate populations in this freshwater
ecosystem declined by an average of 23.7% over 5 years.
Note: Marsh Frog (Species C) increased by +25.5%, but this was
outweighed by Giant Otter (-43.5%) and Nile Tilapia (-37.2%)
declines in the geometric mean aggregation.
Species Indices & Freshwater LPI — Years 1–5
Index value relative to Year 1 baseline (1.00)
Giant Otter
Nile Tilapia
Marsh Frog
Freshwater LPI
Reading the Results
Interpreting the Living Planet Index
📒 What LPI Values Mean
- LPI = 1.00 — Index at the baseline year (1970). Reference point, not ecological ideal.
- LPI > 1.00 — Monitored populations are, on average, larger than in the baseline year.
- LPI < 1.00 — Average population size has declined relative to 1970.
- LPI = 0.27 (2020) — Populations are 73% below their 1970 average — NOT that 73% of species are extinct.
- The LPI measures abundance trends, not species counts or extinction rates.
⚠ Common Misinterpretations
- NOT extinction: A declining LPI does not mean species are disappearing — it means monitored populations are getting smaller.
- NOT all wildlife: The LPI covers only vertebrates — not plants, insects, fungi, or marine invertebrates.
- NOT uniform decline: The global LPI masks regional and taxonomic variation. Freshwater declines (−83%) are far worse than marine (−39%).
- NOT a random sample: Monitored species are biased toward well-studied groups in data-rich countries.
LPI Value Interpretation Scale
0.00
0.25
0.50
0.75
1.00
>1.00
0.00–0.30Critical DeclinePopulations have lost >70% of baseline abundance. Crisis-level biodiversity deterioration. Applies to Latin America freshwater (LPR 2022).
0.30–0.60Severe Decline40–70% below baseline. Major population losses in progress. Urgent intervention required. Global LPI 2022 (0.27) falls here.
0.60–0.85Moderate Decline15–40% below baseline. Significant negative trend; conservation action recommended. Our worked example: LPI = 0.76.
0.85–1.00Slight DeclinePopulations near or slightly below baseline. Monitoring and preventive conservation warranted.
>1.00Recovery / GrowthPopulations have grown beyond 1970 baseline levels. Indicates effective conservation or natural recovery. Rare at global scale.
LPI by Realm — Divergent Trajectories
Freshwater vertebrates have declined most severely; some marine groups show partial recovery in protected zones
| Realm | LPI 2020 (approx.) | % Change from 1970 | Key drivers |
|---|---|---|---|
| Freshwater | 0.17 | −83% | Habitat loss, dams, pollution, overextraction |
| Terrestrial | 0.33 | −67% | Land-use change, deforestation, agriculture |
| Marine | 0.61 | −39% | Overfishing, ocean warming, pollution |
| Global LPI | 0.27 | −73% | Combined across all three realms |
🌏 Regional Extremes: The Neotropical region shows the sharpest decline among all biogeographical regions — an average of 94% decline in monitored vertebrate populations since 1970, primarily driven by deforestation in the Amazon basin, wetland drainage, and freshwater overexploitation. Africa shows −66%; Europe/Central Asia shows −18% (with some recoveries due to targeted conservation policies).
LPI vs Other Biodiversity Indices
| Attribute | Living Planet Index (LPI) | Species Richness / Red List Index |
|---|---|---|
| What is measured | Population abundance trends of vertebrate species over time | Number of species present / extinction risk category changes |
| Baseline | 1970 = 1.00 (anchored year) | Species count at assessment date / IUCN category |
| Detects early warning | Yes — population decline precedes extinction by decades | No — captures extinction/near-extinction, not trend |
| Taxonomic scope | Vertebrates only (mammals, birds, reptiles, amphibians, fish) | All assessed species (but <10% of known species assessed) |
| Publication cycle | Biennial (WWF Living Planet Report) | Continuous (IUCN Red List updated 4×/year) |
| Policy linkage | CBD, Kunming-Montreal GBF (core headline indicator) | CBD, IPBES, national legislation |
| Geographic weighting | Yes — corrects for data bias toward N. Hemisphere | Not directly applicable |
| Our worked example | Freshwater LPI = 0.76 (−24% over 5 yrs) | Separate calculation required; different metric |
Try It Yourself
Interactive LPI Calculator
Enter your own population count data to compute annual log ratios, species indices, and a realm-level LPI.
LPI Computation Tool
─ Enter population counts for up to 4 species over up to 6 years · Minimum: 2 species, 2 years
SPECIES NAMES (optional)
POPULATION COUNTS PER YEAR
Final LPI
—
Compute to see result
% Change
—
from baseline year
Species Tracked
—
monitored populations
Status
—
Results will appear here after computation...
No comments:
Post a Comment