1.3.2 Current state of knowledge on tipping points in the biosphere

In this section we assess available scientific literature relating to tipping points in the Biosphere, as summarised in Figure 1.3.1 and Table 1.3.1. We focus on the following biomes: forests, savannas, drylands, lakes, coastal ecosystems and marine environments.

Figure: 1.3.1
Figure 1.3.1: Map of biosphere systems considered in this chapter. Systems are marked by the coloured areas, with terrestrial biomes and mangroves based on biogeographic biomes (Dinerstein et al., 2017), and lakes and ocean biomes on IUCN functional biomes (Keith et al., 2022) (lakes are shown over other biomes for tundra only; fisheries are spread across the global ocean, but are marked only on key coastal seas for simplicity). Labels indicate which of the systems are in this report considered a tipping system (+++ high confidence, ++ medium confidence and + low confidence), which are not (- – – high confidence, – – medium confidence and – low confidence), and which are currently uncertain (▽).

Table 1.3.1: Summary of evidence for tipping dynamics, key drivers and biophysical impacts in each system considered in this chapter.

Key: +++ Yes (high confidence), ++ Yes (medium confidence), + Yes (low confidence), – – – No (high confidence), – – No (medium confidence), No (low confidence), ? Uncertain Primary drivers are bolded, DC: Direct Climate driver (via direct impact of emissions on radiative forcing); CA: Climate-Associated driver (including second-order & related effects of climate change); NC: Non-Climate driver, PF: positive (amplifying) feedback, NF: negative (damping) feedback. Drivers can enhance (↗) the tipping process or counter it (↘)

(and potential tipping point)
Key driversKey biophysical impacts (see S2 for societal impacts)Key feedbacksEvidence baseAbrupt /  large rate change?Critical threshold(s)?Irreversible? (decadal / centennial)Tipping system? 
Amazon rainforest (dieback)DC: atmospheric warming (↗)
NC: deforestation / degradation (↗)
DC: drying (↗)
CA: fire frequency/intensity increase (↗)
DC: heatwaves (↗)
CA: ENSO intensification (e.g. Amazon, SE Asia) (↗)
CA: AMOC / SPG weakening / collapse (e.g. Amazon) (↗)
CA: terrestrial greening (↘, declining)
• Biodiversity loss
• Regional rainfall reduction (e.g. from Amazon dieback across Amazon Basin & Southern American Cone)
• Carbon emissions (amplifying global warming)
• Remote impacts on rainfall patterns all over the planet
Moisture recycling, fire, albedo• Models
• Observations (local scale)
++1000-1250mm annual rainfall
-400 to -450mm max. accumulated water deficit
7-8m dry season length
~20-40% deforestation
~3.5oC (2-6oC) global warming
+++++ (local)
++ (partial dieback / regional)
+ (full dieback / continental)
Congo rainforest (dieback)+~1350mm mean annual rainfall; climate change increasing rainfall++ (local)
SE Asia rainforest (dieback)~1550mm mean annual rainfall+? (local)
– – (regional)
Boreal forest (southern dieback)DC: drying (↗)
CA: fire frequency/intensity increase (↗)
DC: atmospheric warming (↗)
CA: permafrost thaw (↗)
CA: insect outbreaks (↗)
NC: deforestation / degradation (↗)
DC: heatwaves (↗)
CA: terrestrial greening (↘)
CA: vegetation albedo (↗)
CA: sea ice albedo decline (↗)
DC: precipitation change (↘,↗)
• Biodiversity loss
• Carbon emissions from dieback, carbon drawdown from expansion
• Complex regional biogeophysical effects on warming – dieback = higher albedo (cooling) but less evaporative cooling (warming) & vice versa for expansion
Fire, albedo, moisture recycling• Models
• Observations
• Experiments
++~4oC (1.4-5oC)+
[~100 yr]
++ (partial / regional)+ (continental)
Boreal forest (northern expansion) Fire, albedo, moisture recycling• Models
• Observations
• Experiments
+~4oC (1.5-7.2oC)+
[~100 yr]
+ (partial / regional)
Temperate forests (dieback)DC: atmospheric warming (↗)
DC: droughts (↗)
DC: heatwaves (↗)
CA: insect outbreaks (↗)
CA: windthrow (↗)
NC: deforestation & fragmentation (↗)
CA: fire frequency increase (↗)
• Biodiversity loss
• Carbon emissions
• Regional warming in summer due to less evaporative cooling, less cloud cover
• Less atmos. water supply
• Less groundwater recharge
Moisture recycling, soil moisture -atmosphere, interacting disturbances, albedo• Models
• Observations
• Experiments
++Widespread thresholds uncertain
? (partial / regional)
Savannas, Grasslands & Drylands
Savanna & Grasslands (degradation)NC: fire suppression (↗)
NC: overgrazing (↗)
DC: increased precipitation intensity (↗)
CA: terrestrial greening (↗)
NC: afforestation (↗)
CA: ocean circulation shift (e.g. Sahe), ↗)
• Biodiversity loss
Groundwater depletion (with encroachment)
• Nutrient cycle disruption
• Reduced fires (with encroachment)
Fire, grazing• Models
• Observations (remote sensing & fieldwork)
+Regionally variable mean annual rainfall; thresholds highly localised;

Fire percolation threshold ~ 60% flammable cover
++++ (local to landscape)
? (regional)
(land degradation)
DC: drying (↗)
DC: atmospheric warming (↗)
NC: land use intensification (e.g. livestock, agriculture, urbanisation)(↗)
DC: extreme events (heatwaves, floods) (↗)
DC: increased rainfall variability (↗)
CA: terrestrial greening (↘)
CA: insect outbreaks (↗)
CA: invasive species (↗)
• Biodiversity loss
• Aridification / Desertification
• Groundwater depletion (with encroachment)
• Regional rainfall changes
• Shift in species composition (e.g. shrub encroachment)
• Vegetation recruitment
Soil fertility, / moisture / microbes, vegetation structure, veg-rainfall, fire, herbivory• Models
• Observations (current & historical)
• Field experiment
++Aridity index (0.54,0.7 and 0.8) (limited reliability of aridity measures; lack of temporal evidences for some thresholds)+
(shorter timescales possible, e.g. via active restoration)
++ (local to landscape)+ (regional)
Lakes (eutrophication-driven anoxia)NC: nutrient pollution (↗)
DC: atmospheric warming (↗)
DC: precipitation changes (↗)
• Biodiversity loss
• Water quality decline
• Increased GHG emissions
Anoxia-driven P release, trophic cascades• Observations
• Models
• Experiments
+++20-30 mg P/l

No clear warming/rainfall thresholds
+++ (localised, widespread)
(DOM loading – ’browning’)
CA: terrestrial greening (↗)
NC: afforestation (↗)
DC: atmospheric warming (↗)
• Biodiversity loss
• Increased GHG emissions
Anoxia-driven P release• Observations
• Models
+>10 mg DOC/l++
++ (localised, widespread in boreal)
Lakes (appearance / disappearance)CA: permafrost thaw-related thermokarst formation / drainage (↗)
CA: glacier lake formation / drainage (↗)
• Biodiversity loss
• Increased GHG emissions
(can be driven by thermokarst)• Observations+++As for permafrost thaw+++ (centennial)– (localised, widespread on tundra)
Lakes (N to P limiting switch)NC: nutrient pollution (atmos. deposition) (↗)• Biodiversity lossN/A• Observations++Related to elemental ratio++
– (localised, regions with high N-deposition)
Lakes (salinisation)DC: atmos. warming (↗)
DC: drought (in arid regions) (↗)
CA: water use intensification (↗)
• Biodiversity loss
• Reduced GHG emissions
Salt release from sediment• Observations+Species-specific salinity threshold++
– (localised, arid regions)
Lakes (invasive species)CA: warming-driven range expansion (↗)
NC: human-mediated introduction (↗)
• Biodiversity lossN/A• Observations
• Models
+Cannot be defined++ (decadal – centennial)– (localised, widespread)
coral reefs (die-off)
DC: ocean warming (↗)
DC: marine heatwaves (↗)
CA: disease spread (↗)
CA: ocean acidification (↗)
NC: pollution (nutrient / sediment) (↗)
NC: disruption (ships, over-harvesting) (↗)
CA: invasive species (↗)
DC: storm intensity (↗)
CA: sea level rise (↗)
• Biodiversity loss (ecosystem collapse, ~25% marine species have life stages dependent on coral reefs)
• Loss of commercial & artisanal fisheries, and other sectors
• Coastal protection loss
Thermal stress leading to symbiont expulsion, decarbonisation, loss of structure (habitat)• Observations
• Models
+++Region and reef dependent:~1.2oC (1.0-1.5oC) GW
Temporally variable heat stress (8-12 Degree Heating Weeks)
Long-term consequences of >350 ppm atmospheric CO2
Acidification threshold uncertain
+++ (localised)
+++ (regionally
CA: sea level rise (↗)
DC: increased climate extremes (tropical cyclones, El Niño-related heat, drought, & flooding, drops in sea level )(↗) 
NC: habitat loss (to agri/aquaculture) (↗)
DC: increased regional drought (↗)
NC: shoreline change (erosion, sedimentation) (↗)
NC: nutrient pollution (↗)
• Biodiversity loss 
• Loss of coastal protection 
• Loss of carbon sink / increased GHG emissions
• Loss of water quality
• Sediment salinisation
• Subsidence
• Enhanced sediment sulphide and methane releases
• Hypoxia (seagrasses)
• Reduced nutrient recycling
Failed recovery between increasingly frequent extreme events; coastal subsidence and erosion preventing re- establishment• Palaeo (mangroves)
• Observations
• Models
+ (sea level induced shifts more gradual than for drought or recurrent extremes)Region dependent (see text for priority regions):
~1.5-2oC global warming
≥ 4-7 mm.yr-1 relative sea level rise rate
Recurrent cyclonic exposure  (e.g  return period below a decade)
Soil pore hyper salinisation (site dependent) 
++(decadal) ++ (regional;
region dependent)
Seagrass (die-off)DC: marine heatwaves (↗)
NC: nutrient pollution (↗)
DC: ocean warming (↗)
CA: disease spread (↗)
NC: sedimentation (↗)
NC: turbidity(↗)
NC: invasive species (↗)
NC: fishing practices (benthos damage) (↗)
CA: sea level rise (↗)
++Region dependent:
~1.5oC global warming
Degrees heating weeks with higher sensitivity in temperate (site dependent. e.g. 6 -12 °C)
Turbidity ( <10-30% of PAR)
Nutrient load (site dependent e.g. N loads >1-5 mg/l; P loads >0.03-0.1 mg/l)
++ (regional;
region dependent)
Marine ecosystems and environment
NC: over-exploitation (↗)
DC: ocean warming (↗)
• Keystone species collapse
• Trophic cascades
Trophic cascades, Allee effect• Observations
• Palaeo/historical records
• Models
++Warming & over-fishing; thresholds highly localised+ to +++ (decades)+++ (cod, regional)+ (large fish, regional)– (small fish, regional)
Marine communities
(regime shift)
NC: over-exploitation (↗)
DC: ocean warming (↗)
NC: nutrient pollution (↗)
CA: sea ice loss (↗)
• Regime shifts
• Trophic cascades.
Trophic cascades, eutrophication• Observations
• Models
++Multiple drivers (warming, nutrients, overfishing); thresholds highly localised+ (decades)+ (local)
Kelp forests
(die off)
NC: urchin overgrazing (linked to overfishing)  (↗)
NC: habitat loss (↗)
NC: pollution (nutrient/sediment)  (↗)
DC: ocean warming (↗)
DC: marine heatwaves  (↗)
• Regime shift to more barren stateSea urchin recruitment &  grazing• Observations
• Models
++Multiple drivers, thresholds highly localised+++ (months – decades)+++ (local)
Biological pump
DC: ocean warming (↗)• Regime shift
• Changes to carbon sink
• Impacts on ocean biogeochemistry
Diatom recruitment• Models
• Theory
+?Ocean warming & stratification; thresholds unknown+ (decades)? (lipid, regional)
– – (gravitational, regional to global)
Marine oxygenation (hypoxia)NC: nutrient pollution (↗)
DC: ocean warming (↗)
• Major changes in ocean productivity, biodiversity and biogeochemical cyclesDecomposition, sediment P release• Models
• Observations
+Nutrient load, warming; thresholds highly localised++ (months / years to centuries)+ (local)? (regional to global)
Bezos Earth Fund University of Exeter logo
Earth Commission Systems Change Lab logo Systemiq logo
Global Tipping Points logo
Share this content