An interactive visualization of 4.5 billion years of biodiversity — from the earliest microbial life through the great radiations and mass extinctions, to the modern biodiversity crisis.
What you're seeing The diversity curve shows the relative richness of life on Earth over deep time, normalized to today's estimated ~8.7 million species (1.0 on the Y-axis). It is built from the Paleobiology Database (PBDB) — the largest open repository of fossil occurrence data, with over 1.6 million occurrences contributed by hundreds of researchers worldwide. The Phanerozoic portion (last 538.8 Ma) blends marine invertebrate diversity corrected for numerical sampling bias (Shareholder Quorum Subsampling) with terrestrial vertebrate, insect, and plant fossil records to represent total biosphere diversity at geological stage resolution (136 points; 151 across the full curve). The Precambrian portion (4,567–538.8 Ma) is a qualitative complexity proxy — not a species count — reflecting the stepwise emergence of biological complexity.
Important caveat: All reconstructions of deep-time biodiversity are shaped by the fossil record's incompleteness and uneven preservation. Use the Spatial Correction toggle (in the Layers menu) to see an alternative view based on recent research (Close et al. 2020; Phillipi et al. 2024) showing that much of the apparent long-term increase in Phanerozoic diversity may reflect changes in where fossils were collected — not real biodiversity change. The corrected view shows a flatter trajectory with a stepwise increase at the Cretaceous–Paleogene boundary. Both views preserve the major extinction events. The true history of biodiversity likely lies somewhere between these two interpretations.
How to use Zoom & pan — scroll to zoom horizontally, drag to pan. The non-linear time axis allocates more space to the Phanerozoic, where the fossil record is richest.Search — click the magnifying glass. Event and extinction results zoom to that age on the curve.Composition — use the Composition dropdown to overlay stacked areas under the curve showing how biodiversity breaks down by evolutionary fauna, taxonomic group, habitat, body plan, or skeletal record. Seven overlays are available; hover for a breakdown at any age. Use the opacity slider to adjust transparency.Click any extinction marker or milestone dot for a detailed card with kill mechanisms, references, and age uncertainties.Key Events — filter evolutionary milestones by category (Origin, Metabolic, Body Plan, etc.). Pin the panel to keep it open while exploring. Use the Solo button to isolate a single category.Stories — click the book icon to explore narrated deep-time stories. Each story guides you through a sequence of evolutionary milestones, connecting the dots across billions of years. Climate band The thin horizontal strip below the geological timescale shows global mean surface temperature through Earth history, color-coded from blue (ice ages) through green (Holocene-like conditions) to orange and red (greenhouse and hothouse worlds).
Phanerozoic temperatures (last 485 million years) are based on the PhanDA reconstruction (Judd et al. 2024), which statistically integrates over 150,000 proxy estimates with HadCM3L climate model simulations via data assimilation. PhanDA shows that Earth's temperature has ranged from 11°C to 36°C, with greenhouse climates being the dominant mode (41% of time) and coldhouse conditions like today's being relatively uncommon (13%).
Precambrian temperatures are estimated from geological proxies (oxygen isotopes, glacial deposits, paleomagnetic evidence for ice latitude). The color scale is anchored at the pre-industrial baseline (~14°C) as green — the stable interglacial climate that enabled human civilization.
Hover over the climate band for temperature values and explanations of the physical mechanisms driving each climate state — from Milankovitch orbital cycles to volcanic CO₂ forcing.
Climate states follow PhanDA's five-state classification: Coldhouse (11–18°C), Coolhouse (18–22°C), Transitional (22–25°C), Warmhouse (25–28°C), and Hothouse (28–36°C).
Atmospheric O₂ band The oxygen strip shows the concentration of free oxygen (O₂) in Earth's atmosphere through time, from essentially zero in the Archean to today's 20.95%.
Phanerozoic values (last 570 million years) are based on the GEOCARBSULF model (Berner 2009), which couples carbon and sulfur biogeochemical cycles to reconstruct atmospheric composition. Key features include the Permo-Carboniferous peak at ~32% (310 Ma) — when giant insects with 70 cm wingspans thrived — and the sharp crash to ~16% at the Permo-Triassic boundary.
Precambrian estimates are qualitative reconstructions from Lyons, Reinhard & Planavsky (2014) , based on mass-independent sulfur fractionation (MIF-S), redox-sensitive trace elements, and iron speciation. The Great Oxidation Event (~2,450–2,300 Ma) marks the permanent rise of atmospheric O₂ — before this, any oxygen produced by cyanobacteria was consumed by reduced volcanic gases and dissolved iron.
Hover over the O₂ band for concentration values and explanations of the mechanisms driving each oxygenation or deoxygenation event.
Sea level band The sea level strip shows global eustatic sea level relative to modern (0 m) through the Phanerozoic, from glacial lows 130 m below modern to the Cretaceous peak 250 m above.
Mesozoic and Paleozoic values are from Haq (2014) and Haq & Schutter (2008) , based on sequence stratigraphy and backstripping. The Cretaceous Cenomanian–Turonian interval (~100–90 Ma) represents the highest sea levels of the last 250 million years, when rapid seafloor spreading, submarine volcanism, and the absence of polar ice flooded ~40% of modern continents.
Cenozoic values are from Miller et al. (2020) , using deep-sea benthic foraminiferal oxygen isotopes calibrated to the New Jersey margin record. Key features include the dramatic ~70 m drop at the Eocene–Oligocene Transition (~34 Ma) when the Antarctic ice sheet formed, and the 120 m oscillations of Pleistocene glacial cycles .
No data is available for the Precambrian — the strip is empty before 540 Ma. Hover over the sea level band for values and explanations of the mechanisms driving each sea level state.
Continental fragmentation band The fragmentation strip shows the degree of continental dispersal through Earth history, from supercontinents (warm red, index near 0) to maximum fragmentation like today (cool blue, index near 1).
Phanerozoic values (last 538.8 million years) are based on the plate fragmentation index of Zaffos, Finnegan & Peters (2017) , derived from plate model reconstructions at stage level, supplemented by Scotese (2021) PALEOMAP continent-count estimates. Key features include the assembly and breakup of Pangaea (peak cohesion ~280–260 Ma, breakup from ~200 Ma), and the progressive rise to near-maximum fragmentation in the Cenozoic.
Precambrian values are qualitative reconstructions of the supercontinent cycle from Nance, Murphy & Santosh (2014) and related literature, tracing Kenorland (~2,500 Ma), Columbia/Nuna (~1,800 Ma), Rodinia (~1,000 Ma), and Pannotia (~600 Ma).
Hover over the fragmentation band for index values and explanations of each tectonic configuration and its effects on biodiversity, climate, and ocean circulation.
LIPs & Impacts The LIPs & Impacts strip shows Large Igneous Provinces as vertical ticks growing symmetrically from the center line, and bolide impacts as purple meteor icons. Tick height and color use the Inferno colormap (dark purple = small LIPs, bright yellow = the largest), scaled logarithmically by erupted volume. LIPs are massive volcanic eruptions that released enough CO₂, sulfur, and halogens to trigger climate disruptions and mass extinctions.
LIP data are compiled from Bond & Grasby (2017) , who catalogued temporal links between LIPs and extinction events, supplemented by the LIP Commission database (Ernst 2014). Key eruptions include the Siberian Traps (252 Ma, ~4 Mkm³) — linked to the end-Permian "Great Dying" — and the Ontong Java Plateau (122 Ma, ~80 Mkm³), Earth's largest known igneous province.
Purple meteor icons mark confirmed asteroid impact craters larger than 40 km in diameter, from the Earth Impact Database (Planetary and Space Science Centre). Icon size scales with crater diameter. The most famous is Chicxulub (66 Ma, 180 km) — the impact that ended the age of dinosaurs. The largest known crater, Vredefort (2,023 Ma, 300 km), predates complex life.
Hover over any tick or meteor icon for volume/diameter, location, and the scientific context of each event.
δ¹³C Carbon isotope curve The δ¹³C strip displays a line chart of marine carbonate carbon isotope ratios (δ¹³C, reported in per mil relative to PDB standard) across Earth history. This ratio tracks the global balance between organic carbon burial (drives values positive) and volcanic/metamorphic CO&sub2; release (drives values negative).
Major positive excursions include the Lomagundi-Jatuli excursion (~2,220–2,060 Ma, up to +10‰), the largest in Earth history, driven by massive organic carbon burial after the Great Oxidation Event; SPICE (~497 Ma) and HICE (~445 Ma), linked to glaciation and ocean anoxia; Carboniferous burial (coal swamps, +4.5‰); and OAE-2 (~93.5 Ma, +5‰), recording ocean-wide anoxia.
Major negative excursions include the Shuram-Wonoka (~575–567 Ma, −12‰), the deepest ever recorded — and, on current evidence, an event of the warm interval between the Ediacaran’s two icehouses, since no glacial deposit anywhere has been found within it (Wong Hearing et al. 2026, GSA Bulletin , in press); the end-Permian (~252 Ma), linked to Siberian Traps volcanism; the Toarcian OAE (~183 Ma); the K–Pg boundary “Strangelove ocean” (~66 Ma); and the PETM (~56 Ma, −3‰) — the closest geological analogue to modern anthropogenic carbon release.
Phanerozoic data follow the compilation of Saltzman & Thomas (2012) . Precambrian values are based on Shields & Veizer (2002) and Halverson et al. (2005, 2010) . The background gradient uses a PRGn diverging scheme : purple for negative excursions, green for positive.
Hover over the strip for interpolated values and scientific context for each interval.
Snowball Earth, and the two ice ages before the animals The 200 million years before animals appear hold the most extreme climate swings in the geological record, and the timeline's Precambrian bands show them as states rather than temperatures — see Climate under Earth System, and the new Ice Extent band beneath it.
What Snowball Earth was Twice in the Cryogenian Period, ice reached the equator. The Sturtian glaciation (717–659 Ma) lasted some 58 million years; the Marinoan (~650–635 Ma) followed after a brief hot interval. Glaciers ground down to sea level in the tropics, and the ice-albedo feedback made it self-reinforcing: white ice reflects sunlight, which cools the planet, which grows more ice. What eventually broke each Snowball was carbon dioxide, accumulating from volcanoes beneath an ice lid that had shut down the weathering reactions that normally remove it. The immediate aftermath was a fierce greenhouse, recorded worldwide in the distinctive “cap carbonate” rocks that sit directly on top of glacial debris (Hoffman et al. 2017).
The point worth carrying away is where they sit in the sequence: the Cryogenian Snowballs are the last glaciations in Earth's history to reach the equator. Every ice age since — and there have been many — has stopped well short of the tropics: in this compilation none reaches closer to the equator than about 37 degrees. The Ice Extent band makes this visible in a single glance: two bars fall to zero degrees, and nothing ever does again.
Two more ice ages, in the Ediacaran Between the Marinoan thaw and the first large complex organisms lie 60 million years, and that gap is not empty. It contains a complete cold–warm–cold–warm cycle, only recently resolved (Wong Hearing et al. 2026, EGUsphere preprint, not yet peer reviewed; and GSA Bulletin , in press):
Mid-Ediacaran icehouse , ~593–579 Ma — the glaciation long called Gaskiers is its closing phase, not a standalone event.Late Ediacaran greenhouse , ~579–565 Ma — at least 14 million years with no glacial deposits anywhere. The Shuram carbon isotope excursion falls inside it.Late Ediacaran icehouse , ~565–550 Ma.Terminal Ediacaran greenhouse , ~550–538.8 Ma, running into the Cambrian. Gaskiers had been read as a single cold snap of perhaps less than a million years. Independently dated glacial records now stretch it to about 5 million years on Avalonia, 10 million on the Río de la Plata craton, and 10–15 million across North African Gondwana — three fragments of the ancient continents, whose rocks are now found in Newfoundland and Britain, in Uruguay, and in Morocco — all ending together at ~579 Ma. The authors propose retiring the name in favour of “mid-Ediacaran icehouse”; this page uses that term while keeping Gaskiers labelled so the familiar name stays findable.
Antarctica, not a Snowball These were ordinary ice ages by modern standards, and that is the surprise. Reconstructions of where the ice actually sat put the mid-Ediacaran glaciers no closer to the equator than about 30 to 40 degrees , depending on which map of the ancient continents is used. Climate and ice-sheet models reproduce a matching geometry — a central dome thinning toward the margins — resembling the Antarctic ice sheet today, or North America's Laurentide sheet during the last glaciation.
More striking still: ice could begin to grow with carbon dioxide as high as 16 times pre-industrial levels , around 4,480 parts per million. Earlier work had concluded that Ediacaran ice needed carbon dioxide far below today's level and a planet hovering near freezing — one nudge from runaway glaciation. If ice can instead form at 16 times pre-industrial carbon dioxide, the mid-Ediacaran was a broad, stable climate state rather than a knife-edge.
One caution on that number. The 21.2°C global mean quoted alongside it is a model input — the temperature before ice-sheet feedbacks are allowed to act — not a reconstruction of Ediacaran temperature. No proxy temperature record exists for this interval, which is why the climate band shows state rather than degrees before 538.8 Ma.
What the biosphere was doing The reason this belongs on a page about life is that the fossil record tracks those four climate states. The Ediacaran biota are conventionally divided into three successive assemblages, and each transition falls near a climate boundary (Wong Hearing et al. 2026, GSA Bulletin , in press):
The Avalon assemblage appears ~579–575 Ma, immediately after the mid-Ediacaran ice retreats — in a greenhouse, and only in deep water. Around 565 Ma, as the second icehouse begins, those organisms appear in shallow water for the first time. Animals with a distinct left and right side follow, in the White Sea assemblage. At ~550 Ma, as that icehouse ends, the Nama assemblage brings biomineralized tubes — the first hard skeletons — and the first deep burrows. For the middle step the authors propose a mechanism: cold water holds more dissolved oxygen, and cold animals consume it more slowly, so cooling may have let deep-water-adapted generalists colonise the shallows. They call it polar emergence , after the Antarctic record of the last few million years, and note that it runs exactly opposite to what is happening now, as warming drives species out of the shallows and toward the poles.
A detail that persuaded the authors their signal is real: the highest Ediacaran diversity falls inside an icehouse , at a time of low sea level when comparatively little rock survives to be searched (the sea-level curve is Bowyer et al. 2024). Sampling bias alone would predict the opposite, so the peak is unlikely to be an artefact of how much rock there is to look at.
A modern Earth before the Cambrian None of this says the ice made the animals. What the two papers establish is an ordering in time, plus a proposed mechanism for one step of it — not cause and effect. Read carefully, though, the ordering carries a larger claim than any single correlation: a climate–biosphere relationship of the kind that characterises the whole Phanerozoic was already running some 40 million years before the Cambrian , under a glaciation style Earth still uses. The stage on which animals appeared was not an alien world recovering from a global freeze. It was a recognisably modern one.
The deep time story Archean (4,031–2,500 Ma) Life emerged in a world without free oxygen. The oldest putative stromatolites at Isua, Greenland (~3,700 Ma; Nutman et al. 2016) and the Dresser Formation, Australia (~3,480 Ma; Allwood et al. 2006) record microbial mat communities. Molecular biomarkers suggest cyanobacteria by ~2,700 Ma, though some early evidence (Brocks et al. 1999) remains debated (French et al. 2015). The Late Archean saw the first large-scale biological production of oxygen, setting the stage for the Great Oxidation Event.
Proterozoic (2,500–538.8 Ma) The Great Oxidation Event (~2,400 Ma; Lyons et al. 2014) transformed Earth's surface chemistry. While this caused a massive loss of anaerobic biomass, the biosphere at this stage was predominantly microbial, making direct comparison to later extinction events problematic — fossil data support biomass collapse rather than quantifiable species loss. Eukaryotic cells with membrane-bound organelles appear by ~1,650 Ma (Javaux et al. 2001). Multicellularity evolved independently in multiple lineages. The Cryogenian closed with two Snowball Earth glaciations — the Sturtian (717–659 Ma) and the Marinoan (~650/639–635 Ma), the last times in Earth's history that ice reached the equator (Hoffman et al. 2017). Sixty million years then separate the Marinoan thaw from the first large, complex organisms, and that gap is not empty: it holds the mid-Ediacaran icehouse (~593–579 Ma), whose closing phase is the glaciation long known as Gaskiers. Its ice sheets sat at middle and high latitudes rather than sealing the tropics — the style of glaciation Earth has had ever since (Wong Hearing et al. 2026, EGUsphere preprint, not yet peer reviewed). The Ediacaran biota (575–538.8 Ma; Narbonne 2005, Droser & Gehling 2015) — Earth's first large, complex organisms — appear in the warm interval that followed, from ~579–575 Ma. A second cold interval, the late Ediacaran icehouse (~565–550 Ma), coincides with their spread into shallow water and with the highest diversity of the period; biomineralized skeletons and deep burrows follow its end at ~550 Ma (Wong Hearing et al. 2026, GSA Bulletin , in press). The disappearance of most Ediacaran organisms near the Precambrian–Cambrian boundary (~538.8 Ma; Darroch et al. 2015, Laflamme et al. 2013) was likely a complex transition involving ecological displacement, environmental change, and gradual decline — not a simple extinction event — paving the way for the Cambrian explosion.
Paleozoic (538.8–252 Ma) The Cambrian explosion (~538.8–520 Ma; Marshall 2006) produced most modern animal phyla within ~20 million years. The Burgess Shale and Chengjiang Lagerstätten preserve exquisite soft-bodied faunas, including the first chordates (Haikouella , Myllokunmingia ; Shu et al. 1999). The Cambrian itself was punctuated by several significant extinction events — including the Sinsk Event (~513 Ma), which devastated archaeocyathid reefs, and the Dresbachian extinction (~502 Ma), which eliminated most ptychopariid trilobites (Bond & Grasby 2017). Plants colonized land by ~470 Ma (Kenrick & Crane 1997), followed by arthropods, then vertebrates — Tiktaalik (Shubin et al. 2006) documents the fish-to-tetrapod transition at ~375 Ma. The late Paleozoic saw the rise of forests (Stein et al. 2007), the first amniote egg, and coal-swamp ecosystems. Three major extinctions punctuate this era: the End-Ordovician (~443 Ma), Late Devonian (~372 Ma), and the catastrophic End-Permian.
Mesozoic (252–66 Ma) Recovery from the End-Permian took ~10 million years (Benton & Twitchett 2003). Dinosaurs arose by ~233 Ma (Langer et al. 2010), diversified through the Jurassic, and dominated terrestrial ecosystems for 165 million years. The first mammals appear at ~225 Ma (Luo et al. 2002), remaining small and nocturnal. Flowering plants (angiosperms) radiated from ~130 Ma (Friis et al. 2011), reshaping terrestrial ecology. The End-Triassic extinction (~201 Ma; Blackburn et al. 2013) cleared competitors, enabling dinosaur dominance. Feathered dinosaurs bridge the gap to modern birds (Xu et al. 2011).
Cenozoic (66 Ma–present) The Chicxulub impact ended the Cretaceous (Alvarez et al. 1980; Schulte et al. 2010; Renne et al. 2013), eliminating non-avian dinosaurs and ~76% of species. Mammals radiated explosively: within 10–15 Myr, most modern orders had appeared (O'Leary et al. 2013). Grasslands spread from ~25 Ma (Strömberg 2011), driving ungulate evolution. The genus Homo appears at ~2.8 Ma; anatomically modern Homo sapiens at ~300 ka (Hublin et al. 2017). The Pleistocene megafauna extinction (~50 ka) coincided with human expansion across continents.
The Big Five mass extinctions Mass extinctions are identified by genus-level loss rates exceeding background by an order of magnitude (Raup & Sepkoski 1982; Bambach, Knoll & Wang 2004). The Big Five are:
End-Ordovician (~443 Ma) — 85% species lost. Gondwanan glaciation and sea-level drop devastated tropical marine faunas (Finnegan et al. 2011; Brenchley et al. 1994).Late Devonian (~372 Ma) — 75% species lost over multiple pulses spanning ~25 Myr. Anoxic oceans and the rise of land plants disrupted marine nutrient cycles (McGhee 1996; Algeo et al. 1995).End-Permian (~252 Ma) — 96% species, 81% genera lost. The greatest catastrophe in the history of life. Siberian Traps volcanism triggered ocean anoxia, acidification, and thermal stress (Burgess et al. 2014; Shen et al. 2011; Erwin 1994). Recovery took ~10 Myr.End-Triassic (~201 Ma) — 80% species lost. Central Atlantic Magmatic Province volcanism drove rapid warming and ocean acidification (Blackburn et al. 2013; Hautmann et al. 2008).End-Cretaceous (~66 Ma) — 76% species lost. The Chicxulub asteroid impact combined with Deccan Traps volcanism (Alvarez et al. 1980; Schulte et al. 2010; Renne et al. 2013). A sixth extinction is now underway. Current extinction rates are 100–1,000× the background rate (Ceballos et al. 2015; de Vos et al. 2015). The IPBES Global Assessment (2019) estimates ~1 million species are threatened with extinction. The companion page Wildlife Population Index takes this apart driver by driver, and lets you switch individual pressures on and off to see which ones bend the curve hardest.
What changes a biosphere Extinctions are the losses. But if you ask a broader question — what has ever changed the biosphere, in either direction — the record sorts into two kinds of event, and the difference between them is the most useful thing deep time has to say about the present (Wong Hearing, Williams, Zalasiewicz et al. 2026, preprint, not yet peer reviewed).
Companion page Biosphere Disruptors The whole argument as an interactive reader: every disrupting episode in the record, how long it lasted, what it did, and a control that lets you move the duration assumption the rankings below rest on. Transient disruptors Short-lived, mostly non-biological, and arriving from outside the biosphere: asteroid impacts, the vast volcanic outpourings of large igneous provinces, the opening and closing of ocean gateways. Across 25 examples in the record their average duration is about 800,000 years . They damage, and then the damage sets: the biosphere can be locked for millions of years in a state that is stable but impoverished — what the authors call undesirable resilience. A depleted world is not necessarily a fragile one, and that is the problem.
Persistent disruptors Long-lived, biological, and arising from inside the biosphere: oxygen-producing photosynthesis, the origin of complex cells, the first animals to burrow and mix seafloor sediment, the colonisation of land by plants, the spread of flowering plants. Only five are identified, and their average duration is about 160 million years — two hundred times longer. Every one of them made the planet more habitable than it had been. Several did so only after harming the living world that produced them: the oxygen that made animals possible was, to the anaerobic microbial biosphere that released it, a poison.
This is the durable finding, and it does not depend on any contested number: every persistent disruptor in Earth's history came from inside the biosphere, and every one increased the planet's capacity to support life.
Where humans sit We are an anomaly. We arose from inside the biosphere, as persistent disruptors do. We are behaving, so far, like a transient one — fast, damaging, and over in a geological instant. And we are the only disruptor in the whole record that can be aware of which of the two it is becoming. Oxygenic photosynthesis could not choose. That is not a comforting observation so much as an unusually literal one: the categories are descriptive, drawn from rock, and we are not yet fully inside either.
How much to trust the rankings The same paper ranks disruptors by rate of biosphere change, and places humanity second-fastest for loss, behind only the asteroid that ended the Cretaceous. That ranking is worth understanding rather than repeating, because it is far softer than it sounds.
A rate is a magnitude divided by a duration, and the durations are the least certain quantities in the entire dataset. The impact holds first place on the strength of a chosen duration of 30 years , drawn from a published range spanning seconds to 185,000 years. Humanity's second place rests on a chosen 200 years , from a range of 70 to 50,000. Choose 50,000 instead and the gap all but disappears: humanity's rate falls by more than two hundredfold, to roughly the level of the end-Permian — the worst extinction in the record. The two most quotable numbers in the paper are the two with the widest uncertainty.
The loss figure also rests on a single measurement — an 83% fall in the total mass of wild mammals — and wild mammals are a small fraction of life on Earth. The paper's own figures show human estimates spread across nearly four orders of magnitude; a rank order compresses all of that into one place on a list.
The mirror claim deserves the same scepticism. Humans could become the greatest driver of biosphere gain, the paper argues — but again by rate, not by amount. The Great Oxidation Event raised the planet's primary productivity by up to 188,000%. The stewardship scenario imagines 5% more species by the year 3100. Humanity wins that comparison only because 3,300 years is roughly 150,000 times shorter than 500 million. And 5% more species than today , reached after the losses already recorded, is a gain measured against a diminished baseline — not against the world of ten thousand years ago.
None of which makes the framing useless. It makes it a question rather than a headline: the biosphere has been remade from within before, repeatedly, and always slowly. Nothing in the record says it cannot be remade well. Nothing in the record says it can be done fast.
All of this is laid out episode by episode, chart by chart, on Biosphere Disruptors — including a control that sweeps every duration across its published range so you can watch the ranking re-sort.
Data sources & methodology Diversity curve The Phanerozoic curve is a composite of four data streams, weighted to represent total biosphere diversity:
Marine invertebrates (primary signal) — stage-level Shareholder Quorum Subsampling (SQS, q=0.7) from the divDyn/ddPhanero analysis (Kocsis et al. 2019), correcting for uneven sampling. The established Alroy et al. (2008) envelope constrains the overall curve shape.Terrestrial vertebrates — raw genus counts from the Paleobiology Database (PBDB, CC-BY 4.0), capturing post-K-Pg mammalian radiation and other non-marine signals.Insects — PBDB genus counts, reflecting the Carboniferous diversification and Cretaceous co-radiation with angiosperms.Plants — PBDB genus counts, capturing land colonization (Silurian) and angiosperm radiation (mid-Cretaceous). Non-marine groups modulate the marine SQS signal to capture radiations the marine record misses. Modern species richness is calibrated to Mora et al. (2011): ~8.7 million eukaryotic species (1.0 on the Y-axis). Big Five extinction magnitudes follow Bambach et al. (2004).
Spatial sampling bias The default curve uses numerical sampling standardization (SQS), which corrects for how many fossils were collected per time bin. However, spatial bias — changes in where fossils were collected — is a separate and significant issue. Close et al. (2020) showed that 50–60% of the variation in "global" diversity can be explained by changes in the geographic spread of fossil localities alone. Phillipi et al. (2024) confirmed that a simple spatial resampling model reproduces much of the apparent Phanerozoic diversity trend without invoking any change in true biodiversity.
The Spatial Correction toggle applies a transformation based on these findings: the pre-Cenozoic curve is flattened (using a square-root compression) while the Cenozoic retains a modest increase, reflecting the consensus that the main real temporal signal is a ~2× stepwise increase at the Cretaceous–Paleogene boundary (Close et al. 2020). All major extinction events remain clearly visible in both views. The truth likely lies between the two curves — the standard view may overestimate the long-term increase, while the corrected view may slightly underestimate it.
Geological timescale 193 intervals from the International Chronostratigraphic Chart v2024/12 (ICS/IUGS; Cohen et al. 2025). Colors follow the official CGMW/ICS color codes.
Mass extinctions Genus-level loss from Bambach, Knoll & Wang (2004). Species-level estimates from Raup & Sepkoski (1982) and event-specific literature. Each extinction card includes kill mechanisms, taxonomic selectivity, recovery timescales, and full references with DOIs.
Evolutionary milestones 273 milestones across 16 categories, from the origin of the Solar System to the present. Every milestone carries an age uncertainty range and cites its primary reference. 38 of them take their dates from molecular divergence estimates in TimeTree (Hedges et al. 2015; Kumar et al. 2022) rather than from fossil first appearances; molecular dates often run older than the earliest fossil, and both are shown where they differ.
Skeletal composition Marine skeletal composition data from Singh et al. (2025, Current Biology , DOI: 10.1016/j.cub.2025.06.006). The STAR thin-section point-count method quantifies the relative abundance of major skeletal groups (corals, brachiopods, trilobites, mollusks, echinoderms, etc.) through geological time.
Composition overlays Seven optional stacked-area overlays can be activated via the Composition dropdown. Each fills the area under the diversity curve with a compositional breakdown, sampled at each curve data point for precise alignment. An opacity slider at the bottom of the panel controls transparency. All overlays use the Turbo colormap (Mikhailov, Google Research 2019) for perceptual uniformity.
Sepkoski Faunas — Sepkoski's Three Great Evolutionary Faunas (Sepkoski 1981, 1984), a foundational framework in macroevolution. The Phanerozoic marine biosphere is divided into three successive faunal assemblages, each dominated by different taxonomic groups:
Cambrian fauna — trilobites, inarticulate brachiopods, hyoliths, monoplacophora. Dominant in the Cambrian, declining through the Ordovician.Paleozoic fauna — articulate brachiopods, rugose and tabulate corals, crinoids, stenolaemate bryozoans, cephalopods. Dominant from the Ordovician to the Permian. Devastated by the End-Permian extinction.Modern fauna — bivalves, gastropods, echinoids, gymnolaemate bryozoans, bony fish, decapod crustaceans. Rose after the End-Permian and dominates to the present. Data: 35,826 marine genera from the sepkoski R package (Jones 2023), which updates the original Sepkoski Compendium (2002) to ICS 2023 boundaries. Each genus is counted as present in every geological stage its first-to-last appearance range spans (range-through method). Fractions are normalized per stage.
Life Groups — 15 major taxonomic groups spanning the full biosphere, from marine invertebrates to terrestrial plants. Unlike the Sepkoski Faunas (which are marine-only), this overlay includes terrestrial vertebrates, insects, and plants, showing the changing balance of life across all environments.
Marine invertebrates — Trilobites, Brachiopods, Corals (Anthozoa), Cephalopods, Bivalves, Gastropods, EchinodermsVertebrates — Sharks & rays (Chondrichthyes), Bony fish (Osteichthyes), Amphibians, Reptiles, Dinosaurs (incl. birds), MammalsTerrestrial — Insects, Land plants Data: Paleobiology Database diversity API (CC-BY 4.0). Each group queried separately at stage resolution using sampled_in_bin genus counts. Fractions normalized per stage. Note: PBDB terrestrial records are less complete than marine, and some groups (insects, plants) have sparse Paleozoic coverage due to preservation bias.
Marine / Terrestrial — regroups the 15 Life Groups into three broad habitat categories: Marine (corals, trilobites, brachiopods, cephalopods, bivalves, gastropods, echinoderms, sharks & rays), Terrestrial (amphibians, reptiles, dinosaurs, mammals, insects, land plants), and Mixed (bony fish, which span freshwater and marine). Shows the great invasion of land from ~470 Ma onward and the shifting balance between ocean and continent.
Invertebrates / Vertebrates / Plants — regroups the Life Groups by body plan into three categories. Reveals how invertebrates dominated for most of the Phanerozoic, with vertebrates rising gradually through the Mesozoic and Cenozoic.
Familiar Groups — regroups the Life Groups into seven everyday categories that a general audience can relate to: marine invertebrates, insects, fish, amphibians, reptiles & dinosaurs (including birds, which are classified under Dinosauria in the PBDB), mammals, and plants.
Skeletal (Singh 2025) — a fundamentally different metric from the diversity-based overlays. Instead of counting genera, this shows what the seafloor was actually made of: the relative abundance of skeletal material from different organism groups in thin-section point counts of carbonate rocks. Based on the STAR method from Singh et al. (2025, Current Biology ). Includes 11 categories: microbial, algae, sponges, foraminifera, corals, brachiopods, trilobites, mollusks, echinoderms, bryozoans, and other. Epoch-level resolution (coarser than the stage-level PBDB overlays).
Non-linear time axis The X-axis uses a piecewise-linear mapping: Hadean & Archean (4,567–2,500 Ma) occupy 10% of screen space; Proterozoic (2,500–538.8 Ma) 15%; Paleozoic (538.8–252 Ma) 30%; Mesozoic & Cenozoic (252–0 Ma) 45%. This allocates space proportional to the richness of the fossil record.
References DIVERSITY & METHODOLOGY Kocsis, A.T., Scotese, C.R. & Kiessling, W. (2019). divDyn: an R package for stage-level diversity analyses. Methods Ecol. Evol. 10: 1896–1907. DOI Alroy, J. et al. (2008). Phanerozoic trends in the global diversity of marine invertebrates. Science 321: 97–100. DOI Singh, S. et al. (2025). Macroevolutionary coupling of marine biomass and biodiversity across the Phanerozoic. Current Biology . DOI Bambach, R.K., Knoll, A.H. & Wang, S.C. (2004). Origination, extinction, and mass depletions of marine diversity. Paleobiology 30(4): 522–542. DOI Mora, C. et al. (2011). How many species are there on Earth and in the ocean? PLOS Biology 9(8): e1001127. DOI Rohde, R.A. & Muller, R.A. (2005). Cycles in fossil diversity. Nature 434: 208–210. DOI Kirchner, J.W. & Weil, A. (2000). Delayed biological recovery from extinctions throughout the fossil record. Nature 404: 177–180. DOI Sepkoski, J.J. (1981). A factor analytic description of the Phanerozoic marine fossil record. Paleobiology 7(1): 36–53. DOI Sepkoski, J.J. (1984). A kinetic model of Phanerozoic taxonomic diversity. III. Post-Paleozoic families and mass extinctions. Paleobiology 10(2): 246–267. DOI Sepkoski, J.J. (2002). A compendium of fossil marine animal genera. Bulletins of American Paleontology 363: 1–560. Jones, L.A. (2023). sepkoski: an R package for accessing Sepkoski's fossil marine animal genera compendium. DOI Peters, S.E. & McClennen, M. (2016). The Paleobiology Database application programming interface. Paleobiology 42(1): 1–7. DOI Hedges, S.B., Marin, J., Suleski, M., Paymer, M. & Kumar, S. (2015). Tree of life reveals clock-like speciation and diversification. Molecular Biology and Evolution 32(4): 835–845. DOI Kumar, S. et al. (2022). TimeTree 5: an expanded resource for species divergence times. Molecular Biology and Evolution 39(8): msac174. DOI Mikhailov, A. (2019). Turbo, an improved rainbow colormap for visualization. Google AI Blog , 20 August 2019. Link SPATIAL SAMPLING BIAS Close, R.A. et al. (2020). The spatial structure of Phanerozoic marine animal diversity. Science 368: 420–424. DOI Phillipi, D.B., Czekanski-Moir, J. & Ivany, L.C. (2024). Global Phanerozoic biodiversity — can variation be explained by spatial sampling intensity? Paleobiology 50: 661–672. DOI Bond, D.P.G. & Grasby, S.E. (2017). On the causes of mass extinctions. Palaeogeography, Palaeoclimatology, Palaeoecology 478: 3–29. DOI MASS EXTINCTIONS Raup, D.M. & Sepkoski, J.J. (1982). Mass extinctions in the marine fossil record. Science 215: 1501–1503. DOI Erwin, D.H. (1994). The Permo-Triassic extinction. Nature 367: 231–236. DOI Benton, M.J. & Twitchett, R.J. (2003). How to kill (almost) all life: the end-Permian extinction event. Trends in Ecology & Evolution 18(7): 358–365. DOI McGhee, G.R. (1996). The Late Devonian Mass Extinction . Columbia University Press. DOI Algeo, T.J. et al. (1995). Late Devonian oceanic anoxic events and biotic crises. Geology 23: 115–118. DOI Brenchley, P.J. et al. (1994). Bathymetric and isotopic evidence for a short-lived Late Ordovician glaciation. Geology 22: 295–298. DOI Finnegan, S. et al. (2011). The magnitude and duration of Late Ordovician–Early Silurian glaciation. Science 331: 903–906. DOI Stanley, S.M. & Yang, X. (1994). A double mass extinction at the end of the Paleozoic era. Science 266: 1340–1344. DOI Shen, S. et al. (2011). Calibrating the end-Permian mass extinction. Science 334: 1367–1372. DOI Burgess, S.D. et al. (2014). High-precision timeline for Earth's most severe extinction. PNAS 111: 3316–3321. DOI Hautmann, M. et al. (2008). Catastrophic ocean acidification at the Triassic-Jurassic boundary. Neues Jahrbuch für Geologie und Paläontologie 249: 119–127. DOI Blackburn, T.J. et al. (2013). Zircon U-Pb geochronology links the end-Triassic extinction with the Central Atlantic Magmatic Province. Science 340: 941–945. DOI Alvarez, L.W. et al. (1980). Extraterrestrial cause for the Cretaceous-Tertiary extinction. Science 208: 1095–1108. DOI Schulte, P. et al. (2010). The Chicxulub asteroid impact and mass extinction at the Cretaceous-Paleogene boundary. Science 327: 1214–1218. DOI Renne, P.R. et al. (2013). Time scales of critical events around the Cretaceous-Paleogene boundary. Science 339: 684–687. DOI Robertson, D.S. et al. (2013). Survival in the first hours of the Cenozoic. GSA Bulletin 125: 679–686. DOI PRECAMBRIAN & EARLY LIFE Nutman, A.P. et al. (2016). Rapid emergence of life shown by discovery of 3,700-million-year-old microbial structures. Nature 537: 535–538. DOI Allwood, A.C. et al. (2006). Stromatolite reef from the Early Archaean era of Australia. Nature 441: 714–718. DOI Brocks, J.J. et al. (1999). Archean molecular fossils and the early rise of eukaryotes. Science 285: 1033–1036. DOI French, K.L. et al. (2015). Reappraisal of hydrocarbon biomarkers in Archean rocks. PNAS 112(19): 5915–5920. DOI Lyons, T.W. et al. (2014). The rise of oxygen in Earth's early ocean and atmosphere. Nature 506: 307–315. DOI Javaux, E.J. et al. (2001). Morphological and ecological complexity in early eukaryotic ecosystems. Nature 412: 66–69. DOI Hoffman, P.F. et al. (2017). Snowball Earth climate dynamics and Cryogenian geology-geobiology. Science Advances 3: e1600983. DOI Narbonne, G.M. (2005). The Ediacara biota: Neoproterozoic origin of animals and their ecosystems. Annual Review of Earth and Planetary Sciences 33: 421–442. DOI Darroch, S.A.F. et al. (2015). Biotic replacement and mass extinction of the Ediacara biota. Proceedings of the Royal Society B 282: 20151003. DOI Laflamme, M., Darroch, S.A.F., Tweedt, S.M., Peterson, K.J. & Erwin, D.H. (2013). The end of the Ediacara biota: extinction, biotic replacement, or Cheshire Cat? Gondwana Research 23(2): 558–573. DOI Droser, M.L. & Gehling, J.G. (2015). The advent of animals: the view from the Ediacaran. PNAS 112: 4865–4870. DOI Pu, J.P. et al. (2016). Dodging snowballs: Geochronology of the Gaskiers glaciation and the first appearance of the Ediacaran biota. Geology 44: 955–958. DOI Bowyer, F.T., Wood, R.A. & Yilales, M. (2024). Sea level controls on Ediacaran–Cambrian animal radiations. Science Advances 10: eado6462. DOI Wong Hearing, T.W. et al. (2026). Ediacaran coupling of climate and biosphere dynamics. GSA Bulletin , in press. Preprint Wong Hearing, T.W. et al. (2026). A Phanerozoic-style icehouse climate in the middle Ediacaran. EGUsphere preprint — under open discussion, not yet peer reviewed. DOI Wong Hearing, T.W., Williams, M., Zalasiewicz, J. et al. (2026). Humans could become the greatest driver of biosphere net gain in Earth history, but we are currently the second fastest driver of biosphere loss. bioRxiv preprint — not yet peer reviewed. DOI MAJOR RADIATIONS & TRANSITIONS Marshall, C.R. (2006). Explaining the Cambrian "explosion" of animals. Annual Review of Earth and Planetary Sciences 34: 355–384. DOI Shu, D. et al. (1999). Lower Cambrian vertebrates from south China. Nature 402: 42–46. DOI Shubin, N.H. et al. (2006). The pectoral fin of Tiktaalik roseae and the origin of the tetrapod limb. Nature 440: 764–771. DOI Kenrick, P. & Crane, P.R. (1997). The origin and early evolution of plants on land. Nature 389: 33–39. DOI Stein, W.E. et al. (2007). Giant cladoxylopsid trees resolve the enigma of the Earth's earliest forest stumps at Gilboa. Nature 446: 904–907. DOI Langer, M.C. et al. (2010). The origin and early evolution of dinosaurs. Biological Reviews 85: 55–110. DOI Luo, Z. et al. (2002). In search of the earliest mammals. Nature 416: 847–852. DOI Friis, E.M. et al. (2011). Early flowers and angiosperm evolution. Cambridge University Press . DOI Xu, X. et al. (2011). An Archaeopteryx -like theropod from China and the origin of Avialae. Nature 475: 465–470. DOI O'Leary, M.A. et al. (2013). The placental mammal ancestor and the post-K-Pg radiation. Science 339: 662–667. DOI Strömberg, C.A.E. (2011). Evolution of grasses and grassland ecosystems. Annual Review of Earth and Planetary Sciences 39: 517–544. DOI Hublin, J.-J. et al. (2017). New fossils from Jebel Irhoud, Morocco, and the pan-African origin of Homo sapiens . Nature 546: 289–292. DOI MODERN BIODIVERSITY CRISIS Ceballos, G. et al. (2015). Accelerated modern human-induced species losses: entering the sixth mass extinction. Science Advances 1(5): e1400253. DOI de Vos, J.M. et al. (2015). Estimating the normal background rate of species extinction. Conservation Biology 29: 452–462. DOI IPBES (2019). Global assessment report on biodiversity and ecosystem services. DOI Barnosky, A.D. et al. (2011). Has the Earth's sixth mass extinction already arrived? Nature 471: 51–57. DOI WWF & ZSL (2024). Living Planet Report 2024. livingplanetindex.org . Published LPI figures reproduced under CC BY-SA 4.0. Capdevila, P. et al. (2026). Halting predicted vertebrate declines requires tackling multiple drivers of biodiversity loss. Science Advances 12, eadx7973. DOI PALEOCLIMATE Judd, E.J. et al. (2024). A 485-million-year history of Earth's surface temperature. Science 385: eadk3705. DOI Scotese, C.R. et al. (2021). Phanerozoic paleotemperatures: The Earth's changing climate during the last 540 million years. Earth-Science Reviews 215: 103503. DOI Hoffman, P.F. et al. (2017). Snowball Earth climate dynamics and Cryogenian geology-geobiology. Science Advances 3: e1600983. DOI ATMOSPHERIC OXYGEN Berner, R.A. (2009). Phanerozoic atmospheric oxygen: New results using the GEOCARBSULF model. American Journal of Science 309: 603–606. DOI Berner, R.A. (2006). GEOCARBSULF: a combined model for Phanerozoic atmospheric O₂ and CO₂. Geochimica et Cosmochimica Acta 70: 5653–5664. DOI Lyons, T.W., Reinhard, C.T. & Planavsky, N.J. (2014). The rise of oxygen in Earth's early ocean and atmosphere. Nature 506: 307–315. DOI Krause, A.J. et al. (2018). Stepwise oxygenation of the Paleozoic atmosphere. Nature Communications 9: 4081. DOI SEA LEVEL Haq, B.U. (2014). Cretaceous eustasy revisited. Global and Planetary Change 113: 44–58. DOI Haq, B.U. & Schutter, S.R. (2008). A chronology of Paleozoic sea-level changes. Science 322: 64–68. DOI Miller, K.G. et al. (2020). Cenozoic sea-level and cryospheric evolution from deep-sea geochemical and continental margin records. Science Advances 6: eaaz1346. DOI CONTINENTAL FRAGMENTATION Zaffos, A., Finnegan, S. & Peters, S.E. (2017). Plate tectonic regulation of global marine animal diversity. Proceedings of the National Academy of Sciences 114: 5653–5658. DOI Scotese, C.R. (2021). An atlas of Phanerozoic paleogeographic maps: the seas come in and the seas go out. Annual Review of Earth and Planetary Sciences 49: 679–728. DOI Nance, R.D., Murphy, J.B. & Santosh, M. (2014). The supercontinent cycle: a retrospective essay. Gondwana Research 25: 108–130. DOI δ¹³C CARBON ISOTOPES Saltzman, M.R. & Thomas, E. (2012). Carbon isotope stratigraphy. The Geologic Time Scale 2012 (F.M. Gradstein et al., eds.), pp. 207–232. Elsevier. DOI Shields, G. & Veizer, J. (2002). Precambrian marine carbonate isotope database: version 1.1. Geochemistry, Geophysics, Geosystems 3: 1031. DOI Halverson, G.P., Wade, B.P., Hurtgen, M.T. & Barovich, K.M. (2010). Neoproterozoic chemostratigraphy. Precambrian Research 182(4): 337–350. DOI Halverson, G.P., Hoffman, P.F., Schrag, D.P., Maloof, A.C. & Rice, A.H.N. (2005). Toward a Neoproterozoic composite carbon-isotope record. GSA Bulletin 117: 1181–1207. DOI LARGE IGNEOUS PROVINCES & IMPACTS Bond, D.P.G. & Grasby, S.E. (2017). On the causes of mass extinctions. Palaeogeography, Palaeoclimatology, Palaeoecology 478: 3–29. DOI Ernst, R.E. (2014). Large Igneous Provinces . Cambridge University Press. DOI Earth Impact Database. Planetary and Space Science Centre, University of New Brunswick. passc.net GEOLOGICAL TIMESCALE Cohen, K.M., Harper, D., Gibbard, P. & Car, N. (2025). The ICS International Chronostratigraphic Chart this decade. Episodes 48: 105–115. DOI | Chart v2024/12 Paleobiology Database. paleobiodb.org (CC-BY 4.0) Each extinction card and milestone in the visualization includes its specific citations with DOIs.
Story Illustrations The banner images accompanying the Deep Time Stories were generated using Midjourney . They are intended as evocative visual accompaniments inspired by natural history museum panoramas, not as scientifically accurate depictions of past environments or organisms. Deep-time imagery is inherently speculative, and these illustrations prioritize atmosphere and emotional resonance over anatomical or ecological precision.
Educational Purpose This interactive visualization is a non-commercial educational project by GLOBAÏA, a non-profit organization dedicated to making scientific knowledge accessible to a general audience. All data is drawn from peer-reviewed literature and open-access databases. The visualization, its stories, and accompanying materials are intended solely for educational and science communication purposes. They do not constitute professional scientific advice and should not be cited as primary research sources.
Credits Created by GLOBAÏA . Data from the Paleobiology Database (CC-BY 4.0) and the International Commission on Stratigraphy. Diversity curve uses stage-level SQS from divDyn (Kocsis et al. 2019) within the Alroy et al. (2008) envelope, with terrestrial, insect, and plant modulation from PBDB.
Scientific advisors Prof. Mark Williams (University of Leicester) — guidance on spatial sampling bias, Precambrian biodiversity interpretation, Cambrian extinction events, and relating deep-time biodiversity loss to contemporary rates of change.
Acknowledgements Thanks to Owen Gaffney for suggesting this page.
Suggested citation GLOBAÏA (2026). Biodiversity Timeline: The Rise and Fall of Life on Earth [interactive visualization]. globaia.org/life/. Accessed