How many ant species are there on Earth?
The myrmecological community has been interested for decades in all those questions associated with: How many extant ant species are there, really? We contacted authors of the ten top-cited papers with descriptions of new ant species from the last five years as well as the author of AntCat, the online catalogue of the ants of the world (antcat.org). We always asked the same four questions and received replies from seven researchers. I would love living to see which of the answers got closest to reality!
Flash interviews compiled by Florian M. Steiner 
As always, researchers in random order, and answers not edited
MNB: antcat.org lists 13,379 valid extant ant species – how many do you believe remain undiscovered?
SC: It is hard to tell. The total number might be between 20,000 and 40,000. So my vote is on the middle = 30,000.
MNB: Will new species rather be added by discovering cryptic species or by analysing underexplored regions or underexplored habitat strata?
SC: Despite the fact there are many undiscovered cryptic taxa in such a well-studied region like Europe (or W-Palaearctics) indeed, the vast majority of undiscovered species are expected to come from the more exotic places. In Madagascar, in the last decade several hundred species were described, and many more are still awaiting description. For example, in the Malagasy Nesomyrmex fauna, 29 of the total 33 have been described in 2015-2016. Thanks to the last decades’ efforts to explore Malagasy biodiversity, our knowledge of that fauna has increased considerably, but this unbelievable increase in new species can still be expected in other tropical territories, such as Central Africa, South-America, or SE-Asia which are not yet subjects of similarly focused bodiversity research.
MNB: From which region(s) should we expect the strongest increase in species numbers?
SC: Please see above.
MNB: What portion of the undiscovered species has already been shelved in museums?
SC: I believe the majority of extant ant species are represented by at least a few (or single) individuals in museums. So – I believe – the coverage is not that bad, but the processing of such frequently old material is often poor. They are not properly databased, unidentified, and the older (some collected in the early 1900’s) standard of collecting methods may not always match today’s requirements.
MNB: antcat.org lists 13,379 valid extant ant species – how many do you believe remain undiscovered?
MRS: Ants are a dominant group of arthropods, ubiquitously distributed worldwide, and occurring almost in all terrestrial ecosystems. Ants represent the greater portion of the known animal biomass of our planet. In fact, as a famous biologist and myrmecologist once wrote, “the total biomass of all the ants in the world is approximately equal to the total biomass of the entire human race.” The number of species mentioned by antcat.org is relatively low if compared with the vast areas of unexplored regions of the World. Therefore, this number is highly likely to increase by additional taxonomic investigations and biodiversity studies, and I expect that the total number of extant ant species may reach 22,000 to 25,000 species, especially with the separation of cryptic taxa using molecular techniques. For example, a recent paper by Oberprieler et al. (2018) just published on the ants of Australia, using CO1 barcoding, indicates that the ant fauna has been grossly underestimated.
MNB: Will new species rather be added by discovering cryptic species or by analysing underexplored regions or underexplored habitat strata?
MRS: I think the discovery of new species will be accomplished by taking into account the three opportunities together. I think it would be useful to have a combination of analyzing named taxonomic groups morphologically and genetically. However, I believe the initial approach should be analyzing the underexplored regions. This approach is supported by my ant research in Saudi Arabia. The number of species recorded from the country was 210 before I initiated my comprehensive surveys, especially in poorly collected regions. My research has indicated more than 300 species. Most of the new species and new country records were discovered in the southwestern Asir Mountains of the country. The faunal affinity of these mountains are the Afrotropical Region. This area harbors the highest number of endemic species of ants in the largest peninsula in the world, the Arabian Peninsula at 3,237,500 km2 (1,250,000 sqare miles). For the Arabian Peninsula, the total number of ant species recorded from the region is about 300 species, but the estimated number to be recorded may reach nearly 500.
MNB: From which region(s) should we expect the strongest increase in species numbers?
MRS: As well documented for other groups of insects, the biodiversity hotspots are the tropical and subtropical regions (Neotropical, Afrotropical, and Oriental regions). Most of the new taxa of the Arabian Peninsula are expected to occur in mountainous regions (Asir Mountains, as mentioned above) and also in, Yemen (Hajar Mountains), UAE, and Dhofar Region (Oman).
MNB: What portion of the undiscovered species has already been shelved in museums?
MRS: This is a good question. Many natural history museums (World Museum Liverpool; Natural History Museum, London; Natural History Museum, Basel; King Saud University Museum of Arthropods, Saudi Arabia, etc.) and entomological collections are rich in ant material collected over the last hundred years from all over the World and they also represent a remarkable reservoir of taxa. The study of these specimens are an important aspect of any treatment of ant biodiversity and especially in the discovery of new species.
MNB: antcat.org lists 13,379 valid extant ant species – how many do you believe remain undiscovered?
BB: Being asked about the possible total count of extant ant species globally, my first thought is about the need for taxonomic revision for groups with names which are already available. AntCat lists 13,382 valid species, however, it also lists 1,919 valid subspecies as of May 9th, 2018. Almost half of these subspecies, 792 to be exact, belong to the Formicinae, and more than half that number, 465, belong to Camponotus alone; another 720 valid subspecies belong to the Myrmicinae, dispersed across dozens of genera. Anyway, making the (gross) assumption that these subspecies are species, the number as of today would be 15,301. Given that hundreds of undescribed species are already known spanning the Pseudomyrmecinae, Strumigenys, Pheidole, and other groups, a conservative estimate would be 20,000 total species globally. This number could be extended higher, particularly if it is assumed that there are a thousand or more yet-to-be-observed forms out there, however I am skeptical of numbers exceeding 25,000. The bottom line of my answer is that, for now, our estimates are hampered by the need for revisionary work across the majority of subfamilies, and for massive radiations, such as the Camponotini.
MNB: Will new species rather be added by discovering cryptic species or by analysing underexplored regions or underexplored habitat strata?
BB: It is certain that hundreds of morphologically-distinguishable forms will be accumulated in museums via leaf litter sampling in underexplored tropical regions. Now, the fundamental assumption beyond phenotypic disparity is our methods of differentiating individuals into taxa, and the criteria we use to delimit species (i.e., “species concepts”). It is increasingly being found that morphological “smears” comprise discrete lineages which can be visually distinguished—once the phylogenetic population structuring of the focal taxon is known. Phylogenomic methods are crucial for detecting these clades, and will also be crucial for modeling morphological evolution. Efforts to phylogenetically model morphological evolution which will provide insights into the rate of diversification, extinction, and transformation across the ant Tree of Life. Fieldwork, scopework, and benchwork are all necessary components for the discovery and delimitation of new species, and I don’t think any one should be given greater emphasis, except, perhaps, the basic requirement of museums and curatorial staff for this research.
MNB: From which region(s) should we expect the strongest increase in species numbers?
BB: My intuition is that the Indomalayan region is top of the list, followed by the Neotropics, Australasia, and the Afrotropics.
MNB: What portion of the undiscovered species has already been shelved in museums?
BB: I can’t give a solid estimate for the number of undiscovered species in museums because I don’t have the data. Plus, the distinction between “undiscovered” and “undescribed” is a bit confounding. A back of the envelope guesstimation is 15-33%. I’d like to emphasize that many of these undescribed or currently invalid species are sorted into taxa which need revision, many desperately so.
Barry Bolton, author of antcat.org 
MNB: antcat.org lists 13,379 valid extant ant species – how many do you believe remain undiscovered?
BB: The question of how many ant species are in existence, or more exactly, how many valid ant species are alive and walking about the earth today, has been a recurring field of friendly speculation among myrmecologists. Back in 1960, Richards & Davies estimated that about 3500 species had been described, a number first proposed by W. M. Wheeler some years earlier. By the time of Hölldobler & Wilson’s (1990) classic volume, 30 years later, the estimated number had risen to 8804, a 2.5 × increase in 30 years. Now, almost 30 years after their book appeared, the number of species recorded in AntCat stands at 13,379, an increase of 1.5 × over the 1990 figure, and of 3.8 × the 1960 figure. Imagining that these numbers reflect some version of reality, can we expect a further 0.5 × increase by the year 2050? I would say probably yes, in which case the number of described species will have risen to over 15,000. Strangely, this is the number of ant species that several taxonomists, myself included, have informally quoted in recent years.
But is this 15,000 total in the realm of possibility? Or is it out by several orders of magnitude? Recently, Fisher & Bolton (2016) made an amateurish attempt to calculate the total number of ant species that could be present in the Afrotropical and Malagasy zoogeographical regions. Using a simple, and rather shaky, assumption that a modern revision of a relatively large, previously unrevised, mostly hand collected genus, generally increases the number of valid species by about 65% over the species count recorded at the start of the revision. With numerous provisos and exceptions, for example under-representation of cryptic or entirely hypogaeic species, under-sampling of particular ecoregions, and so on, a very rough guess of 5000 species maximum for the two regions together was the result. If this even approaches reality then the Oriental + Malesian (Indo-Australian) region will most likely approach, or may even excede, the same number. The Neotropical region and the Austral region will not be far behind, and the vast but mainly temperate Holarctic region will add a couple of thousand more. Speculating furiously, that would most likely take the total to well over 15,000 species. Which, if correct, would mean that about 2000 ant species remain to be discovered.
MNB: Will new species rather be added by discovering cryptic species or by analysing underexplored regions or underexplored habitat strata?
BB: In recent years, the close morphological examination of mainly Palaearctic species, many of them supposedly well known and common, has disclosed the existence of numerous cryptic species. By which I mean that a supposedly single species, often a common and widely distributed taxon whose single-species status has been unchallenged for years, has turned out to be an assemblage of closely related but discrete species. These investigations have noticably increased the number of Palaearctic ant species, but the methods have not yet been applied to the obviously vastly larger tropical faunas. There is little doubt in my mind that when the techniques are applied to widespread, apparently single, common species in the tropics, similar results will be obtained. Add to this the possibilities presented by DNA analysis to disclose cryptic species, and the potential number of new species begins to increase quite strikingly. On top of the numbers of new species produced by these methods, detailed sampling in under-explored zoogeographical regions, or intense investigation of discrete ecoregions within them, will undoubtedly produce hundreds of undescribed species. One merely has to look at the results obtained in a couple of recent surveys to get some idea of this. Taking the genera Strumigenys and Pheidole as examples, the number of described species of the former genus rose, between the years 1995 and 2000, from 339 to 790 (Bolton, 2000), and the NewWorld species of the latter genus increased, between the years 1995 and 2003, from 263 to 624 (Wilson, 2003). Brian Fisher’s intensive sampling on Madagascar increased the number of Strumigenys species alone from six to 90. Obviously, the Madagascan fauna was poorly known compared to most other regions, but one cannot help but suspect that similar intensive surveys in other tropical regions will also produce striking increases in the numbers of species represented. Then consider the possibilities presented by close investigations of restricted or highly specialised habitats. For instance, deep in various types of soil, among the matted root systems of grasses, within the topmost twigs of tall trees, under bark, in mosses, fungi and the bases of epiphytes, in the walls of subterranean, terrestrial and arboreal termitaries (both inhabited and abandoned); you can probably imaging many other restricted habitats. Some specialised habitats have been examined in the Holarctic, but in the tropical realms, where ants are overwhelmingly abundant, their potential has barely been touched upon, and certainly never explored in great detail. In conclusion, new species will be abundantly exposed both by the application of refined techniques, and by the exploration of under-investigated habitats.
MNB: From which region(s) should we expect the strongest increase in species numbers?
BB: It is already well established that the tropical zoogeographical regions contain by far the greatest numbers of ant species. Despite the fact that the Palaearctic and Nearctic regions have, historically, received much more attention than the tropics, the number of species they contain is considerably less than those already discovered in tropical zones of equal area. Also, recent taxonomic revisions of tropical genera emphasise the point that there is an abundance of new species awaiting description. By comparison, revisions of Holarctic genera produce relatively few new species, and many of these are only as a result of applying the new discriminatory techniques. Just imaging the number of tropical species that could arise from the application of those new techniques to a fauna that is already considered to be extremely species-rich.
To predict which zoogeographical region will produce the most new species, and which region will prove to have the greatest number of species in total, must contain a large element of guesswork. For the simple historical reason, the length of time that the fauna has been intensively studied, it seems apparent that in the Palaearctic and Nearctic regions most new species will arise from the re-assessment of pre-existing taxa. The number of absolute novelties will be relatively small. Conversely, in the tropics it is obvious from recent revisions that hundreds of novelties, previously entirely unknown species, and indeed whole new genera, await description. The Malagasy region has produced a startling abundance of new taxa in recent years, but this just reflects the concerted modern sampling of the ant fauna in a region that had previously been almost ignored. I will be brave and predict that two regions in particular will produce the greatest number of novelties: the Afrotropical and the Indo-Australian. Both regions span enormous areas, both contain a great diversity of tropical ecologies, both have had a relatively small number of dedicated taxonomists analysing their entire ant faunas, and both have enormously speciose genera whose taxonomy sadly remains at a pre-1950 level. Why not the Neotropics? Because among the tropical regions this is the one that has been best served by myrmecologists, and particularly taxonomists, during the past 150 years. So perhaps it is safe to assume that its representation of species is more accurate than any other.
MNB: What portion of the undiscovered species has already been shelved in museums?
BB: Without doubt the main entomological museums of the world already contain hundreds of undescribed ant species. Some specimens are already sitting in collections, mounted, labelled and ticketed as undescribed (among which are some real rarities, such as the world’s second Ishakidris species, represented by a single specimen in London); others have been roughly sorted to species but await further analysis; others languish in limbo in the “unsorted accessions” drawers that accompany many collections; still others remain undetected in vials of alcohol-preserved samples that accumulate over time in all busy museums. The problem is that usually there is no-one available to carry out even an initial sorting of the vials’ contents, nor anyone with the time or expertise to give these samples a preliminary sorting, and distribute them to their appropriate places in the collection, where eventually a taxonomist may get to look at them. In short, there is an insufficiency of myrmecologists in general, and an utter dearth of ant taxonomists in particular.
Perhaps I can try to illustrate the inertia in the system by reference to the small collection that I have accumulated at home since my retirement in 2004. Considering only Afrotropical species in genera that have been revised since 1970, I have here certified new species in the genera Anochetus (1), Calyptomyrmex (1), Leptogenys (2), Ocymyrmex (2), Probolomyrmex (1), Strumigenys (3), and Tetramorium (> 10). Beside these there are other novelties, certainly new because their genera, though unrevised, contain so few described species, such as Discothyrea (4), Parvaponera (1), and Stigmatomma (1). All of these are from a collection of only a few hundred species. Now scale this up in your mind to the magnitude of a major collection, about the size of the Natural History Museum in London, or the Museum of Comparative Zoology in Cambridge, Massachusetts, where the accretion rates can be truly overwhelming, and the number of new species awaiting description becomes frightening.
In conclusion, I am tempted to concede that the world’s total number of ant species will excede 15000, and may approach 20000, about the same as the number of bee species. Only time will tell, but I wish there were 20 or 30 more full-time ant taxonomists scattered about the world, who were free to work on the groups of their choice.
References
Bolton, B. 2000: The ant tribe Dacetini. – Memoirs of the American Entomological Institute 65: 1-1028.
Fisher, B.L. & Bolton, B. 2016: Ants of Africa and Madagascar. A guide to the genera. – University of California Press, Oakland, CA, 503 pp.
Hölldobler, B. & Wilson, E.O. 1990: The ants. – Harvard University Press, Cambridge, MA, 732 pp.
Richards, O. W. & Davies, R. G. 1960: A. D. Imms. A general textbook of entomology (ninth edition). – Methuen & Co, London & New York, 886 pp.
Wilson, E.O. 2003: Pheidole in the New World: a dominant hyperdiverse ant genus. – Harvard University Press, 794 pp.
MNB: antcat.org lists 13,379 valid extant ant species – how many do you believe remain undiscovered?
BF: Based on the number of undescribed species versus described species from recent collections, especially those collections from tropical regions, I think the total number of ant species in the world is closer to 30,000. But it is not just the tropical regions where you will find new species. Undescribed species are also present in the temperate areas such as California or Greece.
MNB: Will new species rather be added by discovering cryptic species or by analysing underexplored regions or underexplored habitat strata?
BF: There are many regions of the world where ants have not been carefully explored using modern techniques, particularly in tropical and subtropical areas such as the Congo Basin or SE Asia.
At some level, most ant species are “cryptic” – that is, in general, ant species are difficult to delimit and often challenging to identify. Understanding phenotypic and molecular variation in the context of a species level revision is time-consuming, challenging, and in the end, only as good as the completeness of the collections.
For most ant groups, one can not go on a field trip to the tropics and then return home and describe the “new species” collected. One may not be able to identify many of the species, and these are then assumed to be candidate new species. But the process to actually describe them is a long, slow process.
Alternatively, with advances in developing a species-level phylogeny of ants, we are almost at the stage where you can return from a trip and then sequence new species with the goal of discovering new unique ant lineages. I think documenting all ant lineages at the species complex level is a more immediately attainable goal than the description of all ant species. The documentation of lineages will also help future taxonomic revision which are best focused on monophyletic lineages.
MNB: From which region(s) should we expect the strongest increase in species numbers?
BF: Tropical regions are where ant diversity shines, especially humid forest below 1500 m.
MNB: What portion of the undiscovered species has already been shelved in museums?
BF: I estimate that 1/3 of the undescribed species are already in museums. The CAS collection has over 1000 undescribed species of ants, but to describe these species requires careful study of phenotypic variation across related species.
MNB: antcat.org lists 13,379 valid extant ant species – how many do you believe remain undiscovered?
GF: I believe that there are in excess of 20,000 extant ant species in total. In the past two decades, new ant species have been described at an average rate of almost 180 species per year, summing up to an increase of more than 3500 species over that period. There is no sign of a slow-down according to the statistics on www.antwiki.org. Given large amounts of undescribed museum specimens and a lack of large-scale inventories in several diversity hotspots, we could still be describing new species at today’s rate for the next 50 years.
MNB: Will new species rather be added by discovering cryptic species or by analysing underexplored regions or underexplored habitat strata?
GF: Many genera across ants are in comparatively good taxonomic order, especially those in the northern Hemisphere’s temperate regions. However, I know from my own experience with tropical ant communities that a lot of genera, many of which are in the two largest subfamilies Myrmicinae and Formicinae, are going to grow considerably. There are large numbers of unidentified, undescribed species from inventories and collections of the past 30 years or so stored in various museum collections. Many of the new species could be cryptic or hiding in drawers with already identified species. However, the majority of new diversity should be morphologically discrete taxa that hasn’t been studied yet.
In general, we might expect higher proportions of undiscovered species from subterranean and canopy strata than from the ground. However, leaf-litter and pitfall trap studies keep producing previously unknown species, even in already sampled areas. Colony collections will most likely contribute to a marked increase in parasitic species, which are still vastly underrepresented in tropical regions.
MNB: From which region(s) should we expect the strongest increase in species numbers?
GF: The vast majority of new ant diversity being discovered originates from the tropics, i.e., Central and South America, Africa, Madagascar, India, Southeast Asia, New Guinea and Australia. For the genus Pheidole alone, I expect an increase in total species diversity of around 500 species worldwide, mainly from the Old World tropics. Hundreds of undescribed, endemic Pheidole species from Africa, Madagascar and Asia (especially Papua New Guinea) for example are already documented and waiting for their names. More are likely to be collected in regions that haven’t been sampled yet or only with inadequate effort. Other diverse genera are in similar situations.
MNB: What portion of the undiscovered species has already been shelved in museums?
GF: I’m convinced that there may be several thousand new species located in museum collections, but more are being found all the time. It’s hard to make a good estimate. The ratio could be somewhere around 2 additional species for every one that’s already been collected.
MNB: antcat.org lists 13,379 valid extant ant species – how many do you believe remain undiscovered?
BS: 30,000.
MNB: Will new species rather be added by discovering cryptic species or by analysing underexplored regions or underexplored habitat strata?
BS: 65% by discovering cryptic species, 30% by analyzing underexplored regions, 5% by analyzing underexplored habitat strata.
MNB: From which region(s) should we expect the strongest increase in species numbers?
BS: In absolute numbers, definitely from the tropics. Within tropical biomes, rain forests will clearly have the first position at the current stage, but the finally expected figure will sadly be reduced by large-scale habitat destruction.
MNB: What portion of the undiscovered species has already been shelved in museums?
BS: Maybe one third.
Featured image gratefully received from myrmecos.net (© Alex Wild)
I LOVE THESE INTERVIEWS!
One of the best of collections of answers on an important question. The Diversity article is very worth reading! Thanks!!! And a very good photo of Barry Bolton….