The Rubiaceae Research Atlas
Of 14,400 accepted Rubiaceae species, 10,803 (75.0 percent) have no indexed paper; the genus Coffea accounts for 34.6 percent of 41,317 indexed studies.
We tagged 39,980 of the 41,317 indexed studies of the Rubiaceae, the family of coffee, quinine and kratom, 14,400 accepted species in all, against 52 dimensions shared across domains plus flora-specific extensions. Three quarters of the family has no indexed paper. What research exists is concentrated on one crop: coffee alone carries more than a third of the literature, while 74 of the 134 accepted Coffea species, the crop’s own wild relatives, have none.
Every species in the family
We load every accepted Rubiaceae species from Kew’s World Checklist of Vascular Plants and link each to the indexed literature. A species with no linked paper is recorded as having no indexed paper.
The Rubiaceae hold 14,400 accepted species in Version 16 of the World Checklist of Vascular Plants, which makes them the fourth largest family of flowering plants: coffee and quinine, the kratom tree, the madders and bedstraws of temperate roadsides, and several very large tropical genera of forest shrubs that almost nobody studies. We link each accepted name, and every name the checklist treats as its synonym, 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.
3,597 species, 25.0 percent of the family, have at least one indexed paper. The remaining 10,803 have none. That share is close to the orchid family’s, but it arrives a different way: the orchids are a very large family studied thinly and evenly, while the Rubiaceae are a very large family in which one crop has absorbed most of the attention and the rest of the family has had almost none.
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 synonyms. Regional floras, agronomic bulletins 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.
Concentration of research on coffee
One genus holds 14,298 of the 41,317 studies. Within it, two cultivated species hold almost everything and 74 of the 134 accepted species have no indexed paper.
The genus Coffea accounts for 14,298 studies, 34.6 percent of the family’s whole indexed record. Almost all of that sits on two species: Coffea arabica with 10,811 studies and Coffea canephora, robusta, with 3,953. Coffea liberica has 298. The remaining 131 accepted species of the genus share a few hundred studies between them.
The point of reading coffee as its own stratum is what it leaves behind. 74 of the 134 accepted Coffea species, 55 percent of the genus, have no indexed paper at all. These are the crop’s wild relatives, most of them narrow endemics of Madagascar, the Mascarenes and West Africa, and they are the pool that coffee breeding will have to draw on for heat tolerance and disease resistance. The cultivated species are among the most studied plants on earth; the genus they belong to is not.
Research and species richness by genus
The twenty most studied genera hold 83.7 percent of the studies. The largest genus in the family, Psychotria, holds 2.0 percent.
Research and diversity point in opposite directions in this family. The five most studied genera carry 25,185 studies between them, 61.0 percent of the corpus, and the twenty most studied carry 34,601, or 83.7 percent. Every one of those genera is economically useful: coffee, the Morinda of noni juice, the Uncaria of traditional Chinese and Amazonian medicine, the gardenias of horticulture, and kratom.
| Genus | Studies | Share | Accepted species | With no indexed paper |
|---|---|---|---|---|
| Coffea | 14,298 | 34.6% | 134 | 74 |
| Morinda | 4,724 | 11.4% | 39 | 21 |
| Uncaria | 2,401 | 5.8% | 38 | 12 |
| Gardenia | 1,908 | 4.6% | 123 | 75 |
| Galium | 1,896 | 4.6% | 678 | 480 |
| Rubia | 1,556 | 3.8% | 87 | 47 |
| Mitragyna | 1,353 | 3.3% | 10 | 1 |
| Psychotria | 835 | 2.0% | 1,645 | 1,385 |
The last row is the family in miniature. Psychotria is the largest genus in the Rubiaceae and one of the largest genera of flowering plants anywhere, with 1,645 accepted species. It carries 835 studies, 2.0 percent of the corpus, and 1,385 of its species, 84 percent, have no indexed paper. Palicourea, the second largest at 706 species, has 432 with none. These are tropical forest shrubs with no crop value, and the indexed record barely registers them.
Two smaller cases are worth naming because they are often assumed to be larger. Kratom, Mitragyna speciosa, carries 1,034 studies on its own, almost all of the genus total, and that literature is recent and pharmacological. Cinchona, the source of quinine and the reason the family was economically important for three centuries, carries 353 studies, 0.9 percent of the corpus. The history is larger than the current science.
Research effort by IUCN category
4,345 species carry a Red List assessment, 30.2 percent of the family. 1,477 species assessed as threatened have no indexed paper, 290 of them Critically Endangered.
Assessment reaches further into this family than into the orchids, but it still covers less than a third of it: 4,345 of the 14,400 accepted species, 30.2 percent, carry an IUCN Red List status. The other 10,055 have never been assessed, and 75 percent of those have no indexed paper either, so for most of the family there is neither an assessment nor a study to base one on.
Among the species that have been assessed as threatened, 1,477 have no indexed paper. 290 of those are Critically Endangered: species that assessors have judged to be at the highest risk of extinction and for which the indexed literature holds nothing at all. That is the clearest actionable list this atlas produces, and it is a list of names, not a statistic.
Study-site countries
Study sites cluster in Brazil, Indonesia and India. 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 6; for the same sites set against environmental change and split by decade, section 7.
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 →Brazil leads with 2,512 studies that name a study site inside it, followed by Indonesia (1,259), India (572), Ethiopia (426), China (423), Colombia (399), Mexico (338) and Peru (317). The ordering is a coffee map as much as a Rubiaceae map: Brazil, Indonesia, Ethiopia and Colombia are the places the crop is grown and studied. Ethiopia is the exception that proves the point, because it is where Coffea arabica is native and where its wild populations are worked on.
The species with no indexed paper are elsewhere
Map instead where the species with no indexed paper occur, using each species’ native range countries from the same WCVP distribution table, and a different geography appears: the forest floras of Madagascar, New Guinea, Borneo and tropical West Africa, where the large unstudied genera are concentrated and where almost no located fieldwork on the family has been done.
Geographic distribution of study sites
Of the 14,390 species with known range countries, 12,247 (85.1%) 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 14,400 accepted 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 the Kew World Checklist of Vascular Plants distribution table (TDWG level-3 areas mapped to countries, native and introduced), so a country counts as range wherever the checklist records the species; in-range is a country-level test, generous within a large country and conservative for a species studied only just across a border. Of the 14,390 species with range data, 12,247 (85.1%) 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 (10,803 of 14,400 have no indexed research) and species whose papers name no place a geoparser can locate (1,340 of the 3,597 studied species). A further 114 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 (Indonesia, Brazil, Colombia, Malaysia, Venezuela); they also head the table below.
| Country | Range species | With an in-country study site | Shortfall | Study sites in country |
|---|---|---|---|---|
| Indonesia | 2,065 | 131 | 94% | 949 |
| Brazil | 1,494 | 328 | 78% | 835 |
| Colombia | 1,248 | 112 | 91% | 248 |
| Malaysia | 1,200 | 69 | 94% | 66 |
| Venezuela | 885 | 41 | 95% | 49 |
| Brunei | 822 | 1 | 100% | 1 |
| Papua New Guinea | 858 | 37 | 96% | 10 |
| Peru | 844 | 44 | 95% | 180 |
| Madagascar | 848 | 54 | 94% | 14 |
| Ecuador | 808 | 77 | 91% | 183 |
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 (75°E to 129°E, 9°S to 19°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. Because the range lists include cultivated and museum records, a country can appear here through garden and herbarium holdings rather than wild populations; read the list as a prompt to check, not a verdict.
| Country | Species in range | Study sites in country |
|---|---|---|
| Brunei | 822 | 1 |
| Macao | 369 | 0 |
| Hong Kong | 369 | 2 |
| Guyana | 358 | 2 |
| Central African Republic | 294 | 1 |
| French Guiana | 254 | 2 |
| Suriname | 201 | 2 |
| Laos | 187 | 1 |
| Burundi | 160 | 1 |
| Trinidad and Tobago | 131 | 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 Salatiga (Indonesia, 91 sites within 100 km), Sukabumi (Indonesia, 67 sites within 100 km), Probolinggo (Indonesia, 57 sites within 100 km); 9.7% 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 = Kew WCVP distribution table (TDWG level-3 areas mapped to countries, native and introduced; entity_places role=range, source wcvp / wcvp_introduced; 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; 10,228 of 41,317 studies (24.8%) have one. Density and distance use precise sites only (not a whole country, not a rounded centroid): 5,599 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 2026-10-19 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. 66 of 532 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 −0.055 across 532 ecoregions and 0.015 across 1,846 admin1 regions; for drying it is −0.097 across 237 ecoregions and −0.113 across 757 admin1 regions. 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 and each first-level administrative region 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 |
|---|---|---|---|---|
| Piney WoodsTemperate Conifer Forests | United States | 75.4%90% forested in 2000 | 0 | 0.00 |
| Esperance malleeMediterranean Forests, Woodlands & Scrub | Australia | 69.3%21% forested in 2000 | 0 | 0.00 |
| California montane chaparral and woodlandsMediterranean Forests, Woodlands & Scrub | United States, México | 65.7%30% forested in 2000 | 0 | 0.00 |
| Western Guinean lowland forestsTropical & Subtropical Moist Broadleaf Forests | Liberia, Sierra Leone, Côte d'Ivoire, Guinea | 62.0%95% forested in 2000 | 0 | 0.00 |
| Northern Khorat Plateau moist deciduous forestsTropical & Subtropical Moist Broadleaf Forests | Thailand, Laos | 61.3%25% forested in 2000 | 0 | 0.00 |
| Naracoorte woodlandsMediterranean Forests, Woodlands & Scrub | Australia | 59.3%18% forested in 2000 | 0 | 0.00 |
| Madagascar succulent woodlandsDeserts & Xeric Shrublands | Madagascar | 58.4%11% forested in 2000 | 0 | 0.00 |
| Peninsular Malaysian peat swamp forestsTropical & Subtropical Moist Broadleaf Forests | Malaysia | 58.0%95% forested in 2000 | 0 | 0.00 |
| Chilean MatorralMediterranean Forests, Woodlands & Scrub | Chile | 55.8%12% forested in 2000 | 0 | 0.00 |
| Khangai Mountains conifer forestsTemperate Conifer Forests | Mongolia | 55.5%9% forested in 2000 | 0 | 0.00 |
| Region | Country | Forest lost | Papers with a site | Per 10,000 km2 |
|---|---|---|---|---|
| Bình Dương | Vietnam | 96.3%37% forested in 2000 | 0 | 0.00 |
| Bong | Liberia | 95.5%100% forested in 2000 | 0 | 0.00 |
| Aveiro | Portugal | 93.7%62% forested in 2000 | 0 | 0.00 |
| Santarém | Portugal | 91.3%24% forested in 2000 | 0 | 0.00 |
| Castelo Branco | Portugal | 91.0%35% forested in 2000 | 0 | 0.00 |
| Porto | Portugal | 90.9%33% forested in 2000 | 0 | 0.00 |
| Montserrado | Liberia | 90.4%88% forested in 2000 | 0 | 0.00 |
| Coimbra | Portugal | 90.2%68% forested in 2000 | 0 | 0.00 |
| Bomi | Liberia | 89.3%99% forested in 2000 | 0 | 0.00 |
| Melaka | Malaysia | 88.7%83% 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 and each first-level administrative region 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 |
|---|---|---|---|---|
| Southern Andean steppeMontane Grasslands & Shrublands | Argentina, Chile | −13.7%dry sub-humid | 0 | 0.00 |
| Sudd flooded grasslandsFlooded Grasslands & Savannas | South Sudan, Ethiopia | −10.4%semi-arid | 0 | 0.00 |
| 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 |
| EspinalTemperate Grasslands, Savannas & Shrublands | Argentina | −9.4%dry sub-humid | 0 | 0.00 |
| Knysna-Amatole montane forestsTropical & Subtropical Moist Broadleaf Forests | South Africa | −9.2%dry sub-humid | 0 | 0.00 |
| High MonteMontane Grasslands & Shrublands | Argentina | −9.2%arid | 0 | 0.00 |
| Namaqualand-Richtersveld steppeDeserts & Xeric Shrublands | South Africa, Namibia | −8.6%arid | 0 | 0.00 |
| Chilean MatorralMediterranean Forests, Woodlands & Scrub | Chile | −8.6%semi-arid | 0 | 0.00 |
| Region | Country | Aridity trend | Papers with a site | Per 10,000 km2 |
|---|---|---|---|---|
| Santiago Metropolitan | Chile | −15.6%dry sub-humid | 0 | 0.00 |
| Valparaíso | Chile | −13.9%semi-arid | 0 | 0.00 |
| Coquimbo | Chile | −13.7%arid | 0 | 0.00 |
| Mendoza | Argentina | −12.8%semi-arid | 0 | 0.00 |
| Biskra | Algeria | −12.8%arid | 0 | 0.00 |
| San Juan | Argentina | −12.7%semi-arid | 0 | 0.00 |
| Gabès | Tunisia | −12.3%arid | 0 | 0.00 |
| Santiago del Estero | Argentina | −12.3%semi-arid | 0 | 0.00 |
| Córdoba | Argentina | −12.2%dry sub-humid | 0 | 0.00 |
| Govi-Altay | Mongolia | −12.1%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 7,670 located papers, 77.5% were published in the 2010s or 2020s and 57.7% since 2016. The whole corpus of 41,317 papers, located or not, is 75.4% from those two decades, so located research is more recent than the corpus. Of the 419 ecoregions with a located paper, 40 have none published since 2010; of the 689 admin1 regions, 80. 10 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 1,000 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 area means of a systematic sample 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), computed in Google Earth Engine; boundary pixels are weighted fractionally. Layers are remote-sensing and reanalysis products, not field measurements.
Research. 7,670 of 41,317 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. 3,949 precise places (open sea, Southern Ocean, atoll centroids) fall in no ecoregion and 3,945 in no admin1 region; 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 2026-10-19 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 0.54 for ecoregions and 0.42 for administrative regions. Beta: a first model, stated with its interval and its limits.
Beta This section fits a model of research attention: the number of papers of this corpus with a precise study site inside each ecoregion and each first-level administrative region, 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 0.54 for ecoregions and 0.42 for administrative regions (admin1). Each unit carries a 90 percent prediction interval from the negative-binomial count spread about its expectation. Because located research is strongly clustered (theta under 1), that interval is wide: a unit with no located paper falls below it only when the model expects more than 2544.1 papers there (ecoregions) or 8196.1 papers there (administrative regions (admin1)), and in this run no unit of either type does. 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. 775 ecoregions in the universe, 406 with a located paper; negative binomial, dispersion theta 0.34, mean by Poisson quasi-likelihood, ridge lambda 100. 5-fold spatial cross-validation blocked by realm: pseudo-R2 0.54 (0.55 without the national-effort covariate), Spearman rho 0.61 out of fold. Labels against the 90 percent interval: 0 below, 746 within, 29 above.
| Ecoregion | Country | Observed | Expected, 90% interval | Tail probability | Label |
|---|---|---|---|---|---|
| Purus-Madeira moist forestsTropical & Subtropical Moist Broadleaf Forests | Brazil | 0 | 38.1590%: 0 to 169 | 0.204 | within |
| Taiwan subtropical evergreen forestsTropical & Subtropical Moist Broadleaf Forests | Taiwan | 0 | 18.7990%: 0 to 83 | 0.258 | within |
| Borneo montane rain forestsTropical & Subtropical Moist Broadleaf Forests | Indonesia, Malaysia | 0 | 17.1290%: 0 to 76 | 0.266 | within |
| Southern Congolian forest-savannaTropical & Subtropical Grasslands, Savannas & Shrublands | Congo - Kinshasa, Angola | 0 | 13.5590%: 0 to 60 | 0.287 | within |
| Irrawaddy moist deciduous forestsTropical & Subtropical Moist Broadleaf Forests | Myanmar (Burma) | 0 | 11.8590%: 0 to 53 | 0.300 | within |
| Puerto Rican dry forestsTropical & Subtropical Dry Broadleaf Forests | Puerto Rico | 0 | 11.4490%: 0 to 51 | 0.303 | within |
| Northern New Guinea lowland rain and freshwater swamp forestsTropical & Subtropical Moist Broadleaf Forests | Papua New Guinea, Indonesia | 0 | 10.0490%: 0 to 45 | 0.316 | within |
| Solimões-Japurá moist forestsTropical & Subtropical Moist Broadleaf Forests | Colombia, Peru, Brazil | 1 | 23.5590%: 0 to 104 | 0.318 | within |
| Beni savannaTropical & Subtropical Grasslands, Savannas & Shrublands | Bolivia | 0 | 9.4490%: 0 to 42 | 0.323 | within |
| Tocantins/Pindare moist forestsTropical & Subtropical Moist Broadleaf Forests | Brazil | 2 | 35.2890%: 0 to 156 | 0.325 | within |
| Ecoregion | Country | Observed | Expected, 90% interval | Tail probability | Label |
|---|---|---|---|---|---|
| New Caledonia dry forestsTropical & Subtropical Dry Broadleaf Forests | New Caledonia | 23 | 0.0390%: 0 to 0 | under 0.001 | above |
| Hawai'i tropical low shrublandsTropical & Subtropical Grasslands, Savannas & Shrublands | United States | 44 | 0.4890%: 0 to 3 | under 0.001 | above |
| Mascarene forestsTropical & Subtropical Moist Broadleaf Forests | Réunion, Mauritius | 13 | 0.1590%: 0 to 1 | under 0.001 | above |
| Comoros forestsTropical & Subtropical Moist Broadleaf Forests | Comoros, Mayotte | 9 | 0.1090%: 0 to 1 | under 0.001 | above |
| Puerto Rican moist forestsTropical & Subtropical Moist Broadleaf Forests | Puerto Rico | 25 | 0.5390%: 0 to 3 | 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 · Administrative regions (admin1): expected against observed Beta
Fit. 2,642 administrative regions (admin1) in the universe, 648 with a located paper; negative binomial, dispersion theta 0.29, mean by Poisson quasi-likelihood, ridge lambda 300. 5-fold spatial cross-validation blocked by country: pseudo-R2 0.42 (0.33 without the national-effort covariate), Spearman rho 0.51 out of fold. Labels against the 90 percent interval: 0 below, 2,572 within, 70 above.
| Region | Country | Observed | Expected, 90% interval | Tail probability | Label |
|---|---|---|---|---|---|
| Kalimantan Utara | Indonesia | 0 | 14.8190%: 0 to 69 | 0.315 | within |
| Tshopo | Democratic Republic of the Congo | 0 | 12.7990%: 0 to 59 | 0.329 | within |
| Bolívar | Colombia | 0 | 12.4190%: 0 to 58 | 0.332 | within |
| Heilongjiang | China | 0 | 11.8290%: 0 to 55 | 0.337 | within |
| Córdoba | Colombia | 0 | 10.1190%: 0 to 47 | 0.352 | within |
| Coahuila | México | 0 | 10.0690%: 0 to 47 | 0.352 | within |
| Chocó | Colombia | 0 | 9.7790%: 0 to 46 | 0.355 | within |
| Amapá | Brazil | 1 | 20.5690%: 0 to 95 | 0.370 | within |
| Ayacucho | Peru | 0 | 8.1390%: 0 to 38 | 0.374 | within |
| Somali | Ethiopia | 0 | 8.0390%: 0 to 38 | 0.376 | within |
| Region | Country | Observed | Expected, 90% interval | Tail probability | Label |
|---|---|---|---|---|---|
| Sud | New Caledonia | 23 | 0.0390%: 0 to 0 | under 0.001 | above |
| Hawaii | United States | 66 | 0.6290%: 0 to 3 | under 0.001 | above |
| Banskobystrický | Slovakia | 9 | 0.0490%: 0 to 0 | under 0.001 | above |
| Okinawa | Japan | 13 | 0.1590%: 0 to 1 | under 0.001 | above |
| Olancho | Honduras | 13 | 0.3090%: 0 to 2 | 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 0.51 and 1.04, top decile 1.45; administrative regions (admin1): deciles 3 to 8 between 0.64 and 1.47, top decile 1.18. 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. 7,670 of 41,317 papers in this corpus have a precise trusted study site (18.6 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: 4,934 unit-papers over 775 ecoregions and 4,768 over 2,642 administrative regions (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 100, administrative regions (admin1) lambda 300); 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 (admin1), 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 2026-10-19 from the published database (SELECT only) and the change layers' cache. Refresh with node phase1/expected_research.js rubiaceae, 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 administrative regions, 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 Rubiaceae are the largest corpus in this series and among the least evenly distributed. One genus holds a third of the studies, two cultivated species hold a quarter of them on their own, and the family’s largest genus holds one fiftieth. Three quarters of the accepted species have no indexed paper, and 1,477 species already assessed as threatened are among them.
What the atlas does not show is why. A species may be unstudied because it is rare, because it is hard to reach, because it has no commercial use, 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, and for most of these species nobody has asked it.