The Amphibian Research Atlas
Of 9,014 amphibian species, 3,519 (39.0 percent) have no indexed paper; fourteen laboratory, farmed and commodity species are linked to at least 18,363 of 65,133 indexed studies.
We tagged 58,963 of the 65,133 indexed studies of the amphibians, the frogs, salamanders and caecilians, 9,014 species in all, against the codebook shared across domains plus fauna-specific extensions. Two fifths of the class has no indexed paper. A large part of what exists is laboratory and farm research on a handful of species: the clawed frogs, the axolotl, the bullfrog and the common toad among them. Among the species with no indexed paper, the two largest groups are the rain frogs of the tropical Andes and the narrow-mouthed frogs of New Guinea, Madagascar and tropical Asia.
Every species in the class
We load every amphibian species in AmphibiaWeb’s taxonomy and link each to the indexed literature. A species with no linked paper is recorded as having no indexed paper.
The census holds 9,014 amphibian species in three orders: 7,954 frogs and toads (Anura), 829 salamanders and newts (Caudata) and 231 caecilians (Gymnophiona), in 77 families. We link each species, and every name AmphibiaWeb lists as its synonym or as its Red List name, to OpenAlex; tag every study with a panel of domain-expert models whose tags a second model checks; and place each study at its field site rather than its authors’ address.
5,495 species, 61.0 percent of the class, have at least one indexed paper. The remaining 3,519 have none. The gap is similar across the three orders: 3,151 frogs and toads (39.6 percent), 274 salamanders (33.1 percent) and 94 caecilians (40.7 percent) have no indexed paper. The caecilians, limbless burrowing amphibians of the wet tropics, are the smallest order and the least studied in total: 558 studies, against 15,804 for the salamanders and 50,096 for the frogs.
Extinct species stay in the census, because the Red List assesses them and they were described like any other. Of the 37 species assessed as Extinct, 23 have no indexed paper.
A count of papers is not a measure of knowledge, and “no indexed paper” means that no record in OpenAlex matched the species or one of its names. Field guides, national herpetological journals and grey literature outside OpenAlex are not seen. What the count does give is a reproducible floor, species by species, for where the indexed record is silent.
Research on laboratory, farmed and commodity species
Fourteen species are linked to at least 18,363 studies, 28.2 percent of the corpus. Seven of them were harvested only up to a fixed cap, so their counts are lower bounds.
Much of the amphibian literature is not field research on wild animals. The clawed frogs Xenopus laevis and Xenopus tropicalis are standard models of developmental and cell biology; the axolotl and the eastern newt are models of regeneration; the bullfrog, the leopard frog, the common frog and the common toad are laboratory and teaching animals; the Chinese giant salamander is farmed; and the Asiatic toad is the source of a traditional medicine. We flag these fourteen species and report them as their own stratum. They are linked to at least 18,363 studies, 28.2 percent of the corpus, and 16,352 studies name no other amphibian.
| Species | Flag | Red List | Studies linked here | Harvest |
|---|---|---|---|---|
| Rana catesbeiana | laboratory | LC | 2,536 | capped: OpenAlex reports 3,724 works for this name and 4,003 for Lithobates catesbeianus |
| Rhinella marina | laboratory | LC | 2,208 | capped: OpenAlex reports 2,252 works for this name and 2,215 for Bufo marinus |
| Rana pipiens | laboratory | LC | 2,009 | capped: OpenAlex reports 4,531 works for this name |
| Bufo bufo | laboratory | LC | 1,988 | capped: OpenAlex reports 14,075 works for this name |
| Xenopus laevis | laboratory | LC | 1,884 | capped: OpenAlex reports 28,915 works for this name |
| Rana temporaria | laboratory | LC | 1,753 | capped: OpenAlex reports 4,859 works for this name |
| Ambystoma mexicanum | laboratory | CR | 1,353 | complete |
| Xenopus tropicalis | laboratory | LC | 1,208 | capped: OpenAlex reports 2,320 works for this name |
| Notophthalmus viridescens | laboratory | LC | 1,071 | complete |
| Ambystoma tigrinum | laboratory | LC | 843 | complete |
| Bufo gargarizans | commodity | LC | 747 | complete |
| Andrias davidianus | farmed | CR | 510 | complete |
| Cynops pyrrhogaster | laboratory | NT | 495 | complete |
| Necturus maculosus | laboratory | LC | 398 | complete |
The cap is deliberate. A literature of tens of thousands of papers on cell biology in Xenopus eggs says nothing about where wild amphibians are studied, and harvesting it in full would have spent the free daily allowance for days on a few names while thousands of species waited. Each of the seven capped species has more than 2,000 indexed papers. The count shown here is a truncated sample of that literature, not a complete count, and every figure in this atlas that includes these species is a lower bound for them.
The flag does not move the research-gap figure. None of the fourteen lacks papers, so the 3,519 species with no indexed paper are all among the 9,000 unflagged species, 39.1 percent of them. Two unflagged species also carry more than a thousand studies each: the marsh frog Pelophylax ridibundus (1,427) and the wood frog Rana sylvatica (1,134), both with large field and physiological literatures. They are counted with everything else.
Research and species richness by family
The five most studied families hold 65.6 percent of the studies. The second largest family, the rain frogs (Strabomantidae), holds 1.0 percent.
The five most studied families, the true frogs, the toads, the tree frogs, the newts and the mole salamanders, carry 42,722 studies between them, 65.6 percent of the corpus. Three of them hold capped laboratory species, so their totals are lower bounds.
| Family | Studies | Share | Species | With no indexed paper |
|---|---|---|---|---|
| Ranidae | 15,864 | 24.4% | 467 | 109 |
| Bufonidae | 11,909 | 18.3% | 666 | 237 |
| Hylidae | 8,032 | 12.3% | 1,083 | 307 |
| Salamandridae | 6,590 | 10.1% | 148 | 29 |
| Ambystomatidae | 3,814 | 5.9% | 32 | 2 |
| Pipidae | 3,480 | 5.3% | 41 | 4 |
| Plethodontidae | 2,975 | 4.6% | 519 | 220 |
| Microhylidae | 1,342 | 2.1% | 765 | 435 |
| Strabomantidae | 629 | 1.0% | 824 | 437 |
The last two rows are where the class’s unstudied species are. The rain frogs, Strabomantidae, are the second largest amphibian family with 824 species, most of them small direct-developing frogs of the Andes and Central America, and they carry 629 studies, 1.0 percent of the corpus; 437 of their species have no indexed paper. The narrow-mouthed frogs, Microhylidae, with 765 species across the tropics and above all in New Guinea and Madagascar, have 435 with none. Between them these two families hold a quarter of the species in the class with no indexed paper.
Research effort by IUCN category
7,984 species carry a Red List category, 88.6 percent of the class. 1,277 species assessed as threatened have no indexed paper, 429 of them Critically Endangered.
Amphibians are among the most completely assessed groups on the Red List: 7,984 of the 9,014 species, 88.6 percent, carry a category, and 2,906 of them are assessed as threatened. That is the setting in which the research gap matters most. Among the threatened species, 1,277 have no indexed paper. 429 of those are Critically Endangered, more than half of the 818 Critically Endangered amphibians.
The share without an indexed paper rises with the threat category, from 22.8 percent of Least Concern species to 52.4 percent of Critically Endangered ones. Rarity and research effort pull in opposite directions here: the species most at risk are mostly narrow-range species that few people have reached. Data Deficient species are the least studied of all, which is consistent with what the category means, since assessors could not find enough information to place them.
Threats and pathogens recorded in the studies
The threat tags most often recorded are habitat loss and pollution. 1,457 studies name the chytrid fungi in the title or abstract.
Every study is tagged for the threats it measures or discusses. The threat recorded most often is habitat loss and degradation, followed by pollution and contamination, invasive species and disease.
Disease is the threat that set amphibian conservation apart. The chytrid fungus Batrachochytrium dendrobatidis, and since 2013 its salamander-killing relative B. salamandrivorans, have driven declines and extinctions on several continents. 1,457 studies in the corpus name a chytrid in the title or abstract, 125 of them B. salamandrivorans, and 315 name a ranavirus. The pathogen tags agree: 1,376 studies are tagged for a fungal pathogen.
Study-site countries
Study sites cluster in the United States and Brazil. A quarter of the corpus names a field site that can be placed on a map.
For where research sits against species ranges, coverage cells and access distances, see section 7; for the same sites set against environmental change and split by decade, section 8.
The map above is fixed. The interactive version plots the same study sites and lets you filter them by this atlas’s own tags, country, year and citations, then read the papers behind any cluster.
Open the interactive map →The United States leads with 3,377 studies that name a study site inside it, followed by Brazil (2,364), China (847), Australia (661), Japan (505), India (483), Colombia (458), Mexico (454) and Canada (454). The ordering follows where amphibian biologists work at least as much as where amphibians live.
The species with no indexed paper are elsewhere
Map instead where the species with no indexed paper are native, using AmphibiaWeb’s list of native countries for each species, and a different geography appears: Papua New Guinea (339 of 435 native species), Colombia (297 of 843), Ecuador (273 of 719), Peru (269 of 690) and Brazil (255 of 1,210), then China, Indonesia and Madagascar, where the rain frogs and narrow-mouthed frogs are concentrated.
Geographic distribution of study sites
Of the 8,982 species with known range countries, 4,773 (53.1%) have no locatable study site inside any of them; the sites that exist cluster in a few cells and close to cities.
1 · Species ranges against study sites
For each of the 9,014 species we compared two country lists: where it is known to occur, and where its indexed papers were carried out. Known range countries come from AmphibiaWeb (the snapshot of 1 April 2026 the census uses), native countries only, so a country counts as range only where the species occurs naturally; in-range is a country-level test, generous within a large country. Of the 8,982 species with range data, 4,773 (53.1%) have no locatable study site inside any range country. That figure mixes two things that only a paper-level check can separate: species with no research (3,519 of 9,014 have no indexed research) and species whose papers name no place a geoparser can locate (1,099 of the 5,495 studied species). A further 176 species were studied only outside their known range countries.
Range shortfall by country
Each country is shaded by the share of its range species that have no locatable study site inside it: darker means a larger share of the species known to occur there have not been studied in the field there. Unshaded land has no census species in range. The five countries with the most range species lacking a site are labelled (Colombia, Ecuador, Peru, Brazil, Papua New Guinea); they also head the table below.
| Country | Range species | With an in-country study site | Shortfall | Study sites in country |
|---|---|---|---|---|
| Colombia | 843 | 300 | 64% | 264 |
| Ecuador | 719 | 229 | 68% | 151 |
| Peru | 690 | 244 | 65% | 147 |
| Brazil | 1,210 | 784 | 35% | 955 |
| Papua New Guinea | 435 | 54 | 88% | 36 |
| China | 678 | 304 | 55% | 458 |
| Venezuela | 369 | 95 | 74% | 79 |
| Indonesia | 410 | 145 | 65% | 299 |
| Madagascar | 433 | 174 | 60% | 81 |
| India | 468 | 238 | 49% | 265 |
2 · Where the study sites are
Each cell is one hexagon of the H3 grid, drawn on its true boundary so that the cells tile without overlap; darker cells hold more distinct study sites. Most cells hold one site; the fieldwork of the whole corpus concentrates in a small number of them.
The densest window, enlarged
The dashed rectangle on the map above marks the window that holds the most study sites (124°W to 70°W, 19°N to 47°N). Here it is at a larger scale and at the grid's own resolution 3 (about 12,400 km2 per cell), with its own log ramp.
Many species, few studies
The ten range countries with the most census species and two or fewer locatable study sites from this corpus. The range lists are wild distributions, so a country here holds wild populations of these species with little located field research; read the list as a prompt to check, not a verdict.
| Country | Species in range | Study sites in country |
|---|---|---|
| Zambia | 92 | 2 |
| Congo (Brazzaville) | 75 | 2 |
| Sierra Leone | 60 | 1 |
| Central African Republic | 58 | 1 |
| eSwatini | 38 | 0 |
| South Sudan | 36 | 2 |
| Burundi | 28 | 1 |
| The Gambia | 27 | 0 |
| Jamaica | 25 | 2 |
| Mali | 22 | 2 |
3 · How far the sites are from a city
Distance from each study site to the nearest populated place of 100,000 people or more is a rough measure of how reachable the fieldwork was. The busiest access points, counted greedily so that no site is counted twice, are Alajuela (Costa Rica, 60 sites within 100 km), Suwon (South Korea, 58 sites within 100 km), Wollongong (Australia, 53 sites within 100 km); 5.1% of all sites lie within 100 km of one of the ten busiest.
These are straight-line distances, not travel times; roads, terrain, sea ice and protected-area access are not modelled. Study-site coordinates are geoparsed from abstracts, and region-level places sit at their centroid.
4 · Method note
Sources. known range countries = AmphibiaWeb isocc, native countries only, introduced countries excluded (entity_places role=range, source aweb; no polygons); access points = Natural Earth ne_10m_populated_places_simple, GitHub master 5.2.0-pre (nvkelso/natural-earth-vector, downloaded 2026-09-29) (populated places with 100,000 people or more), public domain; grid = H3 resolution 3 (h3-js 4.5.0), drawn on the true cell boundaries from h3-js and aggregated one level (h3 cellToParent) on the main maps. A study site is a place a paper's abstract names as the location of the work, resolved to coordinates by a gazetteer and kept only when it came from one of the six trusted geocode sources; 17,250 of 65,133 studies (26.5%) have one. Density and distance use precise sites only (not a whole country, not a rounded centroid): 9,523 distinct places.
Caveat. Study-site coordinates come from automated geoparsing; a 40-site manual check found 31 plausible; a 100 to 200 pin human check is planned. Range coverage is a country-level test, so a study in the right country but far from the species' actual distribution still counts as in range.
Provenance. Generated 2026-10-29 from the published database. Figures in this section are recomputed by phase1/spatial_shortfall.js and checked by the figures audit.
5 · Explore these layers
The coverage cells are a layer on the interactive map, switched on when you open it from here, with the study sites and this atlas's tag filters underneath. A second layer, range shortfall by country, shades each country as the map above does. Any country, ecoregion, administrative region or drawn area can be queried against the whole database in the place explorer, which lists the papers inside it and the topics that are thin there.
Research against change
Three questions of the same study sites: where the environment is changing fastest with the least located research, whether research effort tracks change at all (it does not: the correlation is near zero), and when the located research was published. 51 of 551 ecoregions sit in the high-change, low-research quadrant for forest loss.
In this corpus the rate of environmental change and the density of located research are essentially unrelated: Spearman's rank correlation between the two, unit by unit, is near zero. For forest loss it is 0.080 across 551 ecoregions and −0.065 across 1,955 states and provinces; for drying it is 0.166 across 236 ecoregions and 0.008 across 768 states and provinces. These are descriptive figures: they do not show that change causes or deters research, and the units differ in area, accessibility and richness. What they do show is where fast change and little located research coincide, which is the list each block below ranks.
1 · Forest loss 2001 to 2024
Change here is the share of the area that was forest in 2000 (tree cover of 30 percent or more) with a loss event between 2001 and 2024, from Hansen Global Forest Change v1.12 at 30 m. Each ecoregion (an area of broadly similar natural vegetation and climate, from the RESOLVE 2017 ecoregion map) and each state or province (the first level of division inside a country: a state, province, region or its equivalent, from GADM 4.1) is placed in one of four quadrants: the top third of the universe by forest loss is high change, the bottom third by located research density (papers of this corpus with a precise study site inside the unit, per 10,000 km2) is low research. The high-change, low-research quadrant is the candidate gap list.
Loss is any stand-replacing event, including fire and the rotational harvest of plantations, and regrowth is not subtracted, so a high value can be a working plantation landscape as easily as a cleared forest; the 30 percent canopy threshold makes open woodland "not forest".
| Ecoregion | Country | Forest lost | Papers with a site | Per 10,000 km2 |
|---|---|---|---|---|
| Esperance malleeMediterranean Forests, Woodlands & Scrub | Australia | 57.1%22% forested in 2000 | 2 | 0.19 |
| Sundaland heath forestsTropical & Subtropical Moist Broadleaf Forests | Indonesia | 44.9%85% forested in 2000 | 5 | 0.65 |
| Madagascar spiny thicketsDeserts & Xeric Shrublands | Madagascar | 44.1%8% forested in 2000 | 0 | 0.00 |
| Peninsular Malaysian peat swamp forestsTropical & Subtropical Moist Broadleaf Forests | Malaysia | 42.9%91% forested in 2000 | 0 | 0.00 |
| Naracoorte woodlandsMediterranean Forests, Woodlands & Scrub | Australia | 41.3%19% forested in 2000 | 1 | 0.41 |
| Khangai Mountains conifer forestsTemperate Conifer Forests | Mongolia | 37.6%18% forested in 2000 | 0 | 0.00 |
| Tocantins/Pindare moist forestsTropical & Subtropical Moist Broadleaf Forests | Brazil | 36.6%78% forested in 2000 | 13 | 0.67 |
| Drakensberg grasslandsTropical & Subtropical Grasslands, Savannas & Shrublands | South Africa, Lesotho | 34.3%8% forested in 2000 | 8 | 0.69 |
| Western Guinean lowland forestsTropical & Subtropical Moist Broadleaf Forests | Liberia, Sierra Leone, Côte d'Ivoire, Guinea | 32.1%90% forested in 2000 | 9 | 0.44 |
| Maranhão Babaçu forestsTropical & Subtropical Dry Broadleaf Forests | Brazil | 31.9%72% forested in 2000 | 10 | 0.70 |
| State or province | Country | Forest lost | Papers with a site | Per 10,000 km2 |
|---|---|---|---|---|
| Bình Dương | Vietnam | 78.9%36% forested in 2000 | 0 | 0.00 |
| Otdar Mean Chey | Cambodia | 63.9%43% forested in 2000 | 0 | 0.00 |
| Melaka | Malaysia | 61.0%69% forested in 2000 | 0 | 0.00 |
| Krâchéh | Cambodia | 56.2%58% forested in 2000 | 0 | 0.00 |
| Kâmpóng Cham | Cambodia | 55.3%16% forested in 2000 | 0 | 0.00 |
| Amnat Charoen | Thailand | 53.2%12% forested in 2000 | 0 | 0.00 |
| Bong | Liberia | 53.0%98% forested in 2000 | 0 | 0.00 |
| Braga | Portugal | 52.0%41% forested in 2000 | 0 | 0.00 |
| Atlantique | Benin | 52.0%5% forested in 2000 | 0 | 0.00 |
| Tbong Khmum | Cambodia | 50.2%29% forested in 2000 | 0 | 0.00 |
Forest loss 2001 to 2024, cell by cell
2 · Aridity trend 1990 to 2024
Change here is the linear trend of the annual aridity index (precipitation over potential evapotranspiration, P/PET) from 1990 to 2024, as percent change per decade of the 1990 to 1999 mean, from TerraClimate at about 4.6 km; a negative number means drying. Each ecoregion (an area of broadly similar natural vegetation and climate, from the RESOLVE 2017 ecoregion map) and each state or province (the first level of division inside a country: a state, province, region or its equivalent, from GADM 4.1) is placed in one of four quadrants: the top third of the universe by drying is high change, the bottom third by located research density (papers of this corpus with a precise study site inside the unit, per 10,000 km2) is low research. The high-change, low-research quadrant is the candidate gap list.
TerraClimate is an interpolated climate surface, so a 35-year trend at 4 km carries station-density and interpolation error, and a relative trend is unstable where the baseline is near zero; the ranking is restricted to drylands (mean P/PET under 0.65) and excludes hyper-arid units (mean under 0.05) for that reason.
| Ecoregion | Country | Aridity trend | Papers with a site | Per 10,000 km2 |
|---|---|---|---|---|
| Sudd flooded grasslandsFlooded Grasslands & Savannas | South Sudan, Ethiopia | −10.4%semi-arid | 1 | 0.05 |
| Khangai Mountains alpine meadowMontane Grasslands & Shrublands | Mongolia | −9.8%dry sub-humid | 0 | 0.00 |
| Gobi Lakes Valley desert steppeDeserts & Xeric Shrublands | Mongolia | −9.6%arid | 0 | 0.00 |
| Western Australian Mulga shrublandsDeserts & Xeric Shrublands | Australia | −9.5%arid | 0 | 0.00 |
| Khangai Mountains conifer forestsTemperate Conifer Forests | Mongolia | −8.5%semi-arid | 0 | 0.00 |
| North Arabian highland shrublandsDeserts & Xeric Shrublands | Saudi Arabia | −8.1%arid | 0 | 0.00 |
| Great Lakes Basin desert steppeDeserts & Xeric Shrublands | Mongolia, Russia | −7.9%semi-arid | 0 | 0.00 |
| Karakoram-West Tibetan Plateau alpine steppeMontane Grasslands & Shrublands | Pakistan, India, China, Afghanistan | −7.7%dry sub-humid | 1 | 0.07 |
| Nullarbor Plains xeric shrublandsDeserts & Xeric Shrublands | Australia | −6.9%arid | 1 | 0.05 |
| Southwest Australia savannaMediterranean Forests, Woodlands & Scrub | Australia | −6.9%arid | 2 | 0.11 |
| State or province | Country | Aridity trend | Papers with a site | Per 10,000 km2 |
|---|---|---|---|---|
| Lakes | South Sudan | −13.0%dry sub-humid | 0 | 0.00 |
| Biskra | Algeria | −12.8%arid | 0 | 0.00 |
| Gafsa | Tunisia | −12.0%arid | 0 | 0.00 |
| Kalmyk | Russia | −11.3%semi-arid | 0 | 0.00 |
| Hovd | Mongolia | −11.1%semi-arid | 0 | 0.00 |
| Sidi Bou Zid | Tunisia | −11.1%arid | 0 | 0.00 |
| Warap | South Sudan | −10.9%dry sub-humid | 0 | 0.00 |
| Hilmand | Afghanistan | −10.8%arid | 0 | 0.00 |
| Balqa | Jordan | −10.8%arid | 0 | 0.00 |
| Madaba | Jordan | −10.6%arid | 0 | 0.00 |
Aridity trend 1990 to 2024, cell by cell
3 · When the research happened
The same study sites split by decade of publication. Of the 14,290 located papers, 63.9% were published in the 2010s or 2020s and 42.1% since 2016. The whole corpus of 65,133 papers, located or not, is 49.0% from those two decades, so located research is more recent than the corpus. Of the 662 ecoregions with a located paper, 47 have none published since 2010; of the 1,572 states and provinces, 187. 25 located papers have no publication year.
The decade with the most located papers, cell by cell
4 · Method note
Change layers. Forest loss 2001 to 2024 = UMD/hansen/global_forest_change_2024_v1_12 (v1.12 (2024 update)), 2001-2024, sampled at 30 m (native 30 m), CC BY 4.0; Aridity trend 1990 to 2024 = IDAHO_EPSCOR/TERRACLIMATE (TerraClimate monthly, EE catalogue copy (1958-01 to 2024-12 at run time)), 1990-2024, sampled at 4,638 m (native 4,638 m), TerraClimate: CC0 1.0 (public domain). MODIS: NASA, no restrictions; dryland classes after UNEP 1992. Values are computed in Google Earth Engine over each H3 cell of the grid (the maps above draw them one resolution coarser, as valid-area-weighted means of the child cells) and each unit outline (RESOLVE Ecoregions 2017 and GADM 4.1, simplified for display); boundary pixels are weighted fractionally. Forest loss sums every 30 m pixel by its area for cells and units up to 25,000 km2; a larger unit is estimated from 1,000 windows of about 1 km2 placed at random inside it, each measured at 30 m. The other layers are area means of a systematic sample at their native resolution. Layers are remote-sensing and reanalysis products, not field measurements.
Research. 14,290 of 65,133 papers in this corpus have a precise trusted study site (the same selection as the density map above); a paper counts once per cell and once per unit it has a site in, and a unit's density is papers per 10,000 km2 of its area. Only units of 1,000 km2 or more are ranked, and only units in a country where a census species is known to occur, so that a city-state's density is not set against a province's. 39 precise places (open sea, Southern Ocean, atoll centroids) fall in no ecoregion and 24 in no state or province; they are in the cell maps but not in the unit tables.
Reading it. A near-zero correlation says that, across these units, more change does not go with more or with less located research. It does not say that change deters research or that research prevents change, and units differ in area, accessibility, richness and national research capacity. A unit in the high-change, low-research quadrant is a candidate gap, not a proven one: papers whose text names no locatable site are invisible here, and the geoparsing caveat of the section above applies.
Provenance. Generated 2026-10-29 from the published database and the Earth Engine cache. Figures in this section are recomputed by phase1/research_vs_change.js, drawn by phase1/gen_atlas_change_section.js and checked by the figures audit.
5 · Explore these layers
These layers are on the interactive map as forest loss 2001 to 2024 and aridity trend 1990 to 2024, cell by cell with the value and the located papers on hover, and a decade selector that limits the study-site dots to one decade of publication.
Expected research Beta
How much located research would each place be expected to carry, given its size, species richness, access, environmental change, biome and national research effort, and where the record falls short of that expectation. Cross-validated pseudo-R2 0.63 for ecoregions and 0.60 for states and provinces. Experimental: a first statistical model, still under test and stated with its interval and its limits. The sections above do not depend on it.
Beta This section fits a model of research attention: the number of papers of this corpus with a precise study site inside each ecoregion (an area of broadly similar natural vegetation and climate, from the RESOLVE 2017 ecoregion map) and each state or province (the first level of division inside a country: a state, province, region or its equivalent, from GADM 4.1), predicted from the unit's area, the number of census species whose known range countries include the unit (country level, so it separates countries rather than places within one), its distance to the nearest city of 100,000 people, forest loss and the aridity trend, its biome, realm and latitude band, and how much located research the rest of its country carries. The model predicts where research tends to go. It does not predict ecological importance, threat or need, and a unit with less research than expected is a candidate gap: a place where the record is thinner than in comparable places. It is never proof of neglect. Fit quality is measured out of sample by 5-fold spatial cross-validation (whole realms or whole countries held out): pseudo-R2 0.63 for ecoregions and 0.60 for states and provinces. Each unit carries a 90 percent prediction interval from the negative-binomial count spread about its expectation. Located research is strongly clustered, so that interval is wide, and in this run no unit falls below it. The rankings therefore order units by the tail probability, the chance of a count at or under the observed one given the expectation, and carry the interval label alongside.
1 · Ecoregions: expected against observed Beta
Fit. 769 ecoregions in the universe, 638 with a located paper; negative binomial, dispersion theta 0.75, mean by Poisson quasi-likelihood, ridge lambda 1000. 5-fold spatial cross-validation blocked by realm: pseudo-R2 0.63 (0.57 without the national-effort covariate), Spearman rho 0.68 out of fold. Labels against the 90 percent interval: 0 below, 737 within, 32 above.
| Ecoregion | Country | Observed | Expected, 90% interval | Tail probability | Label |
|---|---|---|---|---|---|
| Purus-Madeira moist forestsTropical & Subtropical Moist Broadleaf Forests | Brazil | 1 | 56.0490%: 1 to 187 | 0.067 | within |
| Interior Yukon-Alaska alpine tundraTundra | United States, Canada | 0 | 20.9190%: 0 to 70 | 0.079 | within |
| Madagascar spiny thicketsDeserts & Xeric Shrublands | Madagascar | 0 | 18.7490%: 0 to 63 | 0.086 | within |
| North Tibetan Plateau-Kunlun Mountains alpine desertMontane Grasslands & Shrublands | China | 0 | 16.0590%: 0 to 54 | 0.096 | within |
| Catatumbo moist forestsTropical & Subtropical Moist Broadleaf Forests | Venezuela, Colombia | 0 | 15.4190%: 0 to 52 | 0.099 | within |
| Northern Canadian Shield taigaBoreal Forests/Taiga | Canada | 0 | 13.9990%: 0 to 47 | 0.106 | within |
| Kazakh semi-desertDeserts & Xeric Shrublands | Kazakhstan | 0 | 13.1790%: 0 to 44 | 0.111 | within |
| Beni savannaTropical & Subtropical Grasslands, Savannas & Shrublands | Bolivia | 1 | 27.0990%: 0 to 91 | 0.114 | within |
| Alaska-St. Elias Range tundraTundra | United States, Canada | 1 | 26.3290%: 0 to 88 | 0.116 | within |
| Puerto Rican dry forestsTropical & Subtropical Dry Broadleaf Forests | Puerto Rico | 1 | 26.0490%: 0 to 87 | 0.117 | within |
| Ecoregion | Country | Observed | Expected, 90% interval | Tail probability | Label |
|---|---|---|---|---|---|
| Puerto Rican moist forestsTropical & Subtropical Moist Broadleaf Forests | Puerto Rico | 67 | 0.7690%: 0 to 3 | under 0.001 | above |
| Hawai'i tropical dry forestsTropical & Subtropical Dry Broadleaf Forests | United States | 55 | 2.9390%: 0 to 11 | under 0.001 | above |
| Fiji tropical moist forestsTropical & Subtropical Moist Broadleaf Forests | Fiji | 10 | 0.3290%: 0 to 2 | under 0.001 | above |
| Nansei Islands subtropical evergreen forestsTropical & Subtropical Moist Broadleaf Forests | Japan | 72 | 4.5990%: 0 to 16 | under 0.001 | above |
| South Western Ghats montane rain forestsTropical & Subtropical Moist Broadleaf Forests | India | 107 | 7.3690%: 0 to 25 | under 0.001 | above |
The "more than expected" list doubles as a geocoding audit: a small unit with tens of papers against an expectation under one is as likely a place-name collision, a herbarium or a field station as a genuine concentration of fieldwork.
2 · States and provinces: expected against observed Beta
Fit. 2,676 states and provinces in the universe, 1,425 with a located paper; negative binomial, dispersion theta 0.73, mean by negative-binomial IRLS, ridge lambda 100, intercept recalibrated by 1.081 so that the expected total equals the observed total. 5-fold spatial cross-validation blocked by country: pseudo-R2 0.60 (0.54 without the national-effort covariate), Spearman rho 0.71 out of fold. Labels against the 90 percent interval: 0 below, 2,564 within, 112 above.
| State or province | Country | Observed | Expected, 90% interval | Tail probability | Label |
|---|---|---|---|---|---|
| Nei Mongol | China | 3 | 119.2790%: 2 to 400 | 0.070 | within |
| Chubut | Argentina | 0 | 20.4990%: 0 to 69 | 0.085 | within |
| Xinjiang Uygur | China | 6 | 151.4790%: 3 to 508 | 0.088 | within |
| Arequipa | Peru | 0 | 13.5090%: 0 to 46 | 0.114 | within |
| Alaska | United States | 22 | 294.9490%: 6 to 988 | 0.130 | within |
| Potosí | Bolivia | 0 | 9.2090%: 0 to 32 | 0.148 | within |
| Rajasthan | India | 3 | 39.7890%: 0 to 134 | 0.152 | within |
| Bourgogne-Franche-Comté | France | 1 | 18.0990%: 0 to 61 | 0.158 | within |
| Chhattisgarh | India | 2 | 26.2490%: 0 to 89 | 0.165 | within |
| Guaviare | Colombia | 1 | 16.8590%: 0 to 57 | 0.166 | within |
| State or province | Country | Observed | Expected, 90% interval | Tail probability | Label |
|---|---|---|---|---|---|
| Okinawa | Japan | 62 | 1.9490%: 0 to 7 | under 0.001 | above |
| Kerry | Ireland | 5 | 0.0890%: 0 to 1 | under 0.001 | above |
| Distrito Federal | México | 45 | 2.6290%: 0 to 10 | under 0.001 | above |
| Bioko Sur | Equatorial Guinea | 11 | 0.4390%: 0 to 2 | under 0.001 | above |
| Hong Kong | China | 25 | 1.4490%: 0 to 6 | under 0.001 | above |
The "more than expected" list doubles as a geocoding audit: a small unit with tens of papers against an expectation under one is as likely a place-name collision, a herbarium or a field station as a genuine concentration of fieldwork.
3 · Calibration
Out of fold, by decile of expected count (observed / expected): ecoregions: deciles 3 to 8 between 0.74 and 1.35, top decile 1.24; states and provinces: deciles 3 to 8 between 0.72 and 1.08, top decile 1.21. The middle deciles are the best calibrated; the top decile is under-predicted in every model, because the busiest places (long-term field stations, provinces with a herbarium or a university) exceed what any smooth covariate model gives. That is one more reason the residuals are candidate gaps and candidate artefacts rather than findings.
4 · Method note Beta
Response and exposure. 14,290 of 65,133 papers in this corpus have a precise trusted study site (21.9 percent); the response per unit is the number of those papers with a site inside it (point in polygon, or the nearest unit within 25 km, as the place explorer counts), so a paper with sites in two units counts once in each: 16,850 unit-papers over 769 ecoregions and 16,225 over 2,676 states and provinces (of 847 and 3,686 units in all). The exposure is log(area), with log(area) also entering as a slope so the data set the area elasticity.
Covariates. log_area_km2, log_richness, log_access_km, forest_loss_pct, forest_loss_missing, aridity_trend_pct_per_decade, aridity_trend_missing, log_national_loo_density, national_loo_undefined, biome, realm, lat_band; continuous covariates standardised, categorical levels with fewer than 10 units pooled. Protected-area share is not available (no protected-area table in the database) and is stated as the natural next covariate. Richness is a country-level proxy: every unit of a country is credited with the national list.
Model and validation. Negative binomial (NB2) GLM with a log link and a ridge penalty on the standardised slopes (ecoregions lambda 1000, states and provinces lambda 100); the mean estimator and lambda are chosen on out-of-fold Poisson deviance over the same spatial folds; the dispersion is estimated by maximum likelihood. 5-fold spatial cross-validation blocked by realm (ecoregions) or country (states and provinces), never random rows; a held-out realm's own dummy is unseen in training. The 90 percent interval is the 5th and 95th percentile of NB(expected, theta) and carries no parameter uncertainty, so it is if anything too narrow. Zero rule: a unit with no located paper and an expectation under 0.5 is never labelled below and never ranked.
Limits. Located papers only: papers whose text names no locatable site are invisible here, and the geoparsing error rate of the sections above applies. The national-effort covariate carries most of the between-country signal; the fit without it is printed above. Terrestrial units only. This is a Beta layer: the numbers are honest as stated, and they will move when the geocoding, the corpus or the covariates change.
Provenance. Generated 2026-10-29 from the published database (SELECT only) and the change layers' cache. Refresh with node phase1/expected_research.js amphibia, redraw with phase1/gen_atlas_expected_section.js; every figure in this section is checked by the figures audit.
5 · Explore this layer
The same units are on the interactive map as the beta layer "Expected vs observed", ecoregions or states and provinces, with observed, expected, interval and label on hover, and a link from each unit to the place explorer. The place explorer shows the same figures for a selected unit under "Beta: expected research".
Summary of the record
What this atlas measures, and what it does not.
The amphibians are the largest corpus in this series and the first that spans a whole class. Two fifths of the species have no indexed paper, and more than a quarter of the studies sit on fourteen laboratory, farmed and commodity species. The rest of the research is concentrated in a few families and a few countries, while the largest tropical families, and the threatened species within them, are barely registered.
What the atlas does not show is why. A species may be unstudied because it was described recently from a single locality, because it is hard to find, because it lives where few herpetologists work, or because the literature that covers it is not indexed in OpenAlex. The record states where the indexed evidence is and where it is not; the reasons are a separate question.