Isognathotermes Sjöstedt, 1926

Josens, Guy, Deligne, Jean, Harry, Myriam, Roy, Virginie, Akama, Pierre D., Coulibaly, Tenon, Dosso, Kanvaly, Goergen, Georg, Hasson, Michel, Kasangij, Patrick Kasangij A, Kifukieto, Carmel, Ru, Bruno Le, Loko, Laura Estelle Yêyinou, Ndiaye, Abdoulaye Baila, Roisin, Yves, Sion, Noémie, Šobotnik, Jan, Stiblik, Petr, Kuenda, Soki Kue Di, Traoré, Saran, Viage, Manuela, Wango, Solange Patricia, Kaymak, Esra, Bourguignon, Thomas & Hellemans, Simon, 2025, An integrative revision of the genus Isognathotermes (Termitidae: Cubitermitinae) with description of seven new species and four new subspecies, European Journal of Taxonomy 1024, pp. 1-197 : 11-39

publication ID	https://doi.org/10.5852/ejt.2025.1024.3099
publication LSID	lsid:zoobank.org:pub:2A22C9AF-0E2D-46D1-A086-6CBE166F0A77
persistent identifier	https://treatment.plazi.org/id/038B9768-241F-FFFA-FDF4-FAD8FDC1FE34
treatment provided by	Plazi
scientific name	Isognathotermes Sjöstedt, 1926
status

Treatment

Genus Isognathotermes Sjöstedt, 1926 View in CoL

Figs 1–33, distribution map: Fig. 32; Table 1

Isognathotermes Sjöstedt, 1926: 214 View in CoL , by original designation.

Type species

Eutermes ( Cubitermes) minitabundus Sjöstedt, 1913 View in CoL .

Diagnosis

As in the other Cubitermitinae , the worker’s, soldier’s and imago’s guts are characterized by a mixed segment between mid and hind gut; four Malpighian tubules fused at base and by a blind diverticulum, or caecum, on the paunch of the hind gut ( Noirot & Kovoor 1958; Noirot 2001) ( Figs 2–3).

The genus Isognathotermes is best and fully characterized by its worker’s EVA which takes two main forms and an intermediate form (workers of no other Cubitermitinae genera show such EVAs.):

- in the fungifaber pattern ( Figs 4, 6), the odd-numbered primary cushions (slightly longer than the even primary cushions) are frequently narrowed and crested in their downstream parts (the crests, however, are only visible if the PCs are twisted and viewed in profile), in some species, they are only slightly protruding or even not at all in I. acristatus sp. nov., the crested and distal parts of the PCs bear long curved or hooked bristles; the transition from the part with supporting bristles to the bulbous part is relatively gradual; see also Noirot & Kovoor (1958: 458).

- in the finitimus pattern ( Figs 5–6), the odd primary cushions are more robust, less or not at all narrowed; a rather swollen part (viewed in profile) bears numerous short, stiff, non-curved bristles and the distal part bears some long curved to hooked bristles; the transition from the part with supporting bristles to the bulbous part is relatively abrupt.

- in the intermediate pattern ( Fig. 6) the odd-numbered primary cushions have relatively small crests, are not swollen in their downstream part, and bear numerous slightly curved bristles; the distal parts bear some long curved to hooked bristles (it is closer to the fungifaber pattern than the finitimus pattern and occurs only in species of the fungifaber pattern, mainly I. severus and I. ugandensis ).

The secondary cushions are wide with a homogeneous spine scattering.

The worker is also characterized by the shape and spinosity of its front legs: the fore coxa is carinated and bears two to seven spines on the ventral carina ( Fig. 7); the fore trochanter bears three to eight spines. Mandibles as in the imagines.

The soldier is dark-gutted (its gut contains some soil); it has a well-developed frontal gland and a frontal pore (fontanelle) on the frons surrounded by a tuft of setae; it has a deeply bifurcated labrum with straight, convex or, more generally sinuous sides (lyre-shaped labrum), almost always with subtruncated lobes ending in fine, whitish or translucent tips, and bearing three to ten large setae on each lobe; its EVA, however, is not as well characterized as that of worker, especially in the fungifaber pattern where it can be confused with a bilobatodes or a muneris pattern; in the finitimus pattern, the soldier’s EVA is generally recognizable ( Fig. 6).

Its fore trochanter and carinated fore coxa often bear some spines (but less than the worker’s). It has a long head (SHdL = 2.18–3.77 mm), a high head elongation index (SHdL/SHdW = 1.22–1.67), and the sabre-like mandibles are always shorter than the head (SMlL/SHdL = 0.66–0.92); antennae of 14–15 articles. All these characters taken together make it possible to recognise the Isognathotermes soldiers.

The imago cannot be distinguished easily from other, related Cubitermitinae genera; it is always rather dark (C6–C8), its head bears some large setae overhanging a generally dense mat of short bristles. Its meso- and metanotum generally bear some very fine, pale setae, arranged in a medio-longitudinal strip, generally visible at 40–80 × (without any visible small setae in 6.5% of the imagines examined), and frequently with three to four larger setae. The fontanelle is always small; antennae of 16 articles (but shortened in functional queens and kings). Mandibles with apical tooth longer than marginal teeth, three on the left mandible, two on the right. Legs: tibial spurs 3: 2: 2, middle tibia with two extra subapical spines. The enteric valve is weakly developed and can hardly be used for identification ( Fig. 6).

Etymology

From Greek ΙΣΟΣ ( isos, equal, even), γναΘΟΣ ( gnatho s, mandible), and Latin termes (termite), the name Isognathotermes probably refers to the type species (see the chapter I. minitabundus ), either to the resemblance between both mandibles or to the fact that their diameter varies little between both ends.

Historical review

The first mention in the literature of a termite that could be an Isognathotermes probably dates back to Smeathman (1781) under the name Termes atrox : see the historical review of I. severus .

This genus was created by Sjöstedt (1926) for a species previously described (soldier only) as Eutermes ( Cubitermes) minitabundus by himself in 1913. Indeed, Sjöstedt (1926: 214) considered, in his identification key, that this monotypic genus was significantly different from all other Cubitermes species by their soldiers’ mandibles which are straight with slightly curved tips. It shared this feature with Megagnathotermes but in the latter genus, soldiers have longer mandibles and a less deeply emarginated labrum. In the same work, Sjöstedt (1926: 216) described the imago and worker of I. minitabundus which was then the only species of the genus, so these descriptions applied also to the genus itself.

According to Sands (1965: 98) and to Krishna et al. (2013: 1858, 1927), “ fide Emerson card catalog (unpublished)”, the imago of I. minitabundus had already been described erroneously under the name Eutermes ( Trinervitermes) carbo by Sjöstedt (1924a: 42).

The species was later reverted to the genus Cubitermes by Snyder (1949: 161). The genus Isognathotermes was recently restored as valid by Hellemans et al. (2021: 233) based on a phylogenetic analysis. It now gathers several previous Cubitermes species sharing the same kinds of Enteric Valve Architectures (EVAs), i.e., of the fungifaber or finitimus patterns as defined by Josens & Deligne (2019). Among the genera derived from Cubitermes (sensu Wasmann) , Isognathotermes is probably the most confusing with the most numerous synonymies and misidentifications in museum collections.

The mitogenomes of 12 out of the 15 valid species have been sequenced.

Species included

However, 13 of these 22 species are now considered to be junior synonyms of other species, as explained below. In addition, one species previously housed in the genus Cubitermes (sensu Wasmann) , namely C. congoensis ( Emerson 1928) and three possibly new species could not be classified for sure; they are housed in the Incertae sedis chapter.

The genus Isognathotermes also includes seven new species: I. acristatus sp. nov., I. magniplanifrons sp. nov., I. modicus sp. nov., I. phallicaecalis sp. nov., I. phalloides sp. nov., I. rectimalatus sp. nov., and I. similifinitimus sp. nov.

Redescription of the genus

Imago

COLOUR. Head capsule well sclerotised, dark, C6–C8, paler (up to C4) in long preserved samples. Fontanelle concolorous or almost so with head capsule. Postclypeus C5–C7 concolorous or somewhat paler (up to two levels) than head capsule. Antennae C4–C6 without any difference between proximal and distal articles. Thorax: pronotum C5–C7, somewhat paler than head capsule; meso- and metanotum C5–C7, as pronotum. Legs C3–C5, tibia usually slightly (one level) darker than femur. Wings hyaline with brown to grey tinge (Cf2–Cf4), anterior veins darker. Abdomen: tergites C5–C7, as pronotum. Sternites appreciably paler in middle (C3–C5) with both sides darker (C4–C6); posterior sternites darker (C4–C6) than anterior.

SETATION. Head capsule with some prominent setae, mainly near the eyes, overhanging a generally dense mat of short bristles. Labrum and postclypeus with some prominent setae mixed with shorter ones. Antennae with some prominent setae, some more numerous smaller setae and, mainly distally on most articles, a bunch of very fine, bent setae (visible only at high magnification, 50 × or more). Thorax: pronotum with prominent setae mainly on margins and many shorter ones in middle; meso- and metanotum with some fine, pale setae arranged in a medio-longitudinal strip, mainly on mesonotum and the anterior lobe of metanotum, generally visible at 20–80 ×, but sometimes without any visible small setae; frequently 1–4 larger setae on metanotum. Legs very pilose, furnished (among numerous fine setae) with some large setae becoming sometimes spine-like: 0–18 on the carina of fore coxa and 0–8 on the anterior side of fore coxa; tibia pilose and furnished with 15–30 spines; fore, mid, and hind tibia bearing 3, 2, 2 apical spurs and 0, 2, 0 subapical spurs respectively. Abdomen: tergites with many large and small setae. Sternites with long setae, erect or directed slightly forward, and many long and smaller setae directed backwards.

STRUCTURE (measurements in Table 1; Figs 8–12, 23). Size: the imagines of the genus Isognathotermes (with those of the genus Polyspathotermes ) are among the largest of those derived from the former genus Cubitermes (sensu Wasmann) ; however, with large overlapping with Ternicubitermes and Nitiditermes . This can be seen in most raw data, exemplified by the head width ( Fig. 23); in contrast, indices differ little between genera. Head capsule large; fontanelle generally a tiny round or elongate marking ( Fig. 8). Compound eyes round to shortly oval (IEy D /IEy d = 1.00–1.25, Fig. 9). Ocelli round to oval (IOc D /IOc d = 1.00–1.80). Antennae: 16 articles on alate individuals (one out of 278 alate imagines with 17 articles), always shortened by amputation of 1–4 articles in queens and kings. Labrum: cupola shaped, generally wider than long. Left mandible with apical tooth longer and more prominent than first marginal; marginal teeth three in number but second one only suggested by an undulation of edge between first and third marginal teeth (disappearing in worn mandibles); only the apical tooth is acute in unworn specimens; premolar tooth with proximal end obscured or partly obscured by molar prominence in dorsal view; molar tooth bearing a rounded molar prominence dorsally and ending posteriorly in a tiny acute apophysis. Right mandible with apical tooth longer and more prominent than first marginal; marginal teeth two in number; first marginal tooth well developed with a sharp tip when fresh; second marginal tooth smaller and with a blunt tip even when fresh; molar tooth bearing a ventral rounded flange and ending posteriorly in a kind of heel ( Fig. 10). Thorax: pronotum appreciably wider than long ( Fig. 11) and narrower than head width, straight to very weakly sellate with anterior lobe short and very slightly elevated. Legs: the fore coxa is flanged ventrally resulting in a more or less sharp edge called here ‘carina’. Wings: R1 fused entirely with costal margin, sclerotised; Rs simple, sclerotised; M and Cu not or weakly sclerotised with 2–6 and 7–15 branches respectively ( Fig. 12). Abdomen of imagines shows 10 fully visible undivided tergites in both sexes. The first sternite is absent or vestigial; fully visible undivided sternites are 8 in male and 6 in female. In the male, the 8 visible undivided sternites are followed by a reduced one divided into two paraprocts; in female the sixth visible sternite (genital plate) is elongated and often covers the three following ones which are divided and reduced, the third having become the two paraprocts. In both male and female each paraproct bears a two-jointed small cercus. Gut: The gut of the imago is similar to that of workers with notably a well-defined mixed segment and a caecum ( Fig. 3), but the caecum is proportionally less developed.

Soldier

COLOUR. Head capsule generally Cd4–Cd6 becoming fader and darker (e.g., Cf5–Cf7) in long preserved samples; frons frequently 1–2 levels darker than back (e.g., C6–C7) sometimes with a sharp contrast giving the impression of a bicolour head; this two-tone colouring has no specific value. Gulamentum concolorous with head or somewhat darker. Antennae and labrum one level paler than head capsule. Mandibles dark (C7–C8) generally with an abrupt clearing on their bases (two levels) which are generally of the same colour as frons. Thorax, nota and legs paler than head capsule (C3–C5) somewhat darker in long preserved samples. Abdomen grey to red-brown owing to digestive bolus, sometimes with a yellow tinge on tergites.

SETATION. Head capsule with few scattered fine setae; on frons a dense bunch of setae surrounds and overhangs fontanelle. Antennae with some prominent setae, more numerous smaller setae and at distal extremity of distal articles, a bunch of fine, bent setae (visible only at high magnification, 50 × or more). Labrum always with 4–9 large setae on each lobe. Thorax: pro- and mesonotum with some setae mainly located on fore and hind margins. Legs: fore coxa furnished with 0–3 spines on carina and not any (or rarely one) spine on ventral side; trochanter generally with six or seven long lined-up strong setae or spines; fore, mid, and hind tibia bearing 3, 2, 2 apical spurs and 0, 2, 0 subapical spurs respectively; all tibiae furnished with a row of 6–15 spines. Abdomen: tergites with few large setae, mainly or only on their posterior margins. Sternites with long setae, erect or slightly directed forward, often coloured, and smaller setae directed backwards.

STRUCTURE (measurements in Table 1; Figs 13–17, 24). Size: the soldiers of Isognathotermes (with those of Polyspathotermes ) are among the largest of those derived from the former genus Cubitermes (sensu Wasmann) ; however, with some overlapping with Ternicubitermes and Nitiditermes . This can be seen in most raw data that vary with genus, exemplified by the head length ( Fig. 24); in contrast, most indices differ little between genera. Head capsule always clearly sclerotised and appreciably longer than wide. Dorsal view: lateral sides mostly subparallel with a more or less strong narrowing near posterior third or fourth ( Fig. 13); however, in the soldiers of incipient colonies, lateral sides are slightly convex without any narrowing; posterior side variable: regularly rounded or sometimes with a short straight or even concave part in the middle, or with three short straight or even concave parts. In profile: upper profile always more or less concave, also in the case of soldiers from incipient colonies; angle between extended mandibles and frons varies from almost right to 130°; frons with or without any small frontal hump overhanging the fontanelle ( Fig. 14). Gulamentum: profile generally with a flat posterior part; in ventral view always more or less constricted in its posterior half, with sides of anterior part variable, from smoothly rounded ( Fig. 15) to extended into a kind of angular ears. Antennae generally of 15 articles (14.5 articles in 9%, 14 articles in 6% and only 13 articles in one soldier from an incipient colony). Labrum: always deeply bifurcate and generally wider than long, generally (in 82%) with sinuous sides (lyre-shaped labrum) (with convex sides in 11%, or with straight sides in 7%) and almost always with subtruncated lobes ending in fine, whitish or translucent tips; anterior margin concave ( Fig. 16). Mandibles: sabre-like to almost hooked distally; inner edges smooth with one distinct but generally very small marginal tooth near molar tooth on each mandible; mandibles clearly shorter than head; entire surface of both mandibles smooth and glossy. Right mandible generally slightly more curved than left. Thorax: pronotum sellate, narrower than head width, with generally entire anterior and posterior margins ( Fig. 17). Fore coxa flanged ventrally resulting in a more or less sharp carina. Gut: enteric valve seating on left side, best seen in ventral view, situated in posterior half of abdomen. Arrangement of enteric valve cushions generally showing trilateral symmetry: the odd primary cushions are on average 16% longer than the even cushions; the odd cushions generally without any crest or with crests weakly developed; in the species with finitimus EVAs the pilosity becomes abruptly very dense showing the place where a crest would be expected (in comparison with the worker’s EVA); secondary cushions wide at the upstream end narrowing noticeably downstream with a homogeneous spine scattering. Caecum rather small, visible in ventral view, near centre of abdomen, either shortly lobed (in most species) or developed as a finger-like process that is sometimes capped or swollen distally.

Worker

COLOUR. Head capsule pale (C1–C3) turning grey in long preserved samples. Antennae: proximal articles pale (C2–C3), distal articles always 1–2 levels darker (C4–C5). Thorax, nota and legs pale (C1–C3). Abdomen grey to red-brown owing to digestive bolus.

SETATION. Head capsule and postclypeus with few, erect, scattered setae. Labrum with few, robust, scattered setae. Antennae with some prominent setae, some more numerous smaller setae and at distal extremity of distal articles, a bunch of fine, bent setae (visible only at high magnification, 50 × or more). Thorax: nota with scattered setae. Legs: fore coxa always carinated, bearing one fine seta and furnished with 3–7 spines on carina (only 2 spines in two out of 400 samples) and 0–4 spines on ventral side; fore trochanter with 5–7 spines; fore, mid, and hind tibia bearing 3, 2, 2 apical spurs, and 0, 2, 0 subapical spurs respectively and a row of spines. Abdomen: tergites with scattered setae. Sternites with long setae, erect or slightly directed forward, often coloured, and smaller setae directed backwards. STRUCTURE (measurements in Table 1; Figs 18–19, 25). Size: the workers of Isognathotermes (with those of Polyspathotermes ) are among the largest of those derived from the former genus Cubitermes (sensu Wasmann) ; however, with large overlapping with Ternicubitermes and Nitiditermes . This can be seen in most raw data that vary with genus, exemplified by the hind tibia lengths ( Fig. 25); in contrast, indices differ very little between genera. Head capsule weakly sclerotised (except mandibles) ( Fig. 18). Antennae 15 articles, more rarely 14.5 or 14, and only 13 in a worker from an incipient colony. Labrum: cupola shaped, generally wider than long. Left mandible: apical tooth well developed with a sharp tip when fresh; marginal teeth three in number, first marginal well developed, second marginal faint (visible as an undulated edge but disappearing in worn mandibles), third marginal with a blunt tip; premolar tooth with its proximal end partly generally hidden under molar prominence; molar tooth bearing a rounded molar prominence dorsally and ending posteriorly in a tiny acute apophysis ( Fig. 19). Right mandible: apical tooth well developed with a sharp tip when fresh; marginal teeth two in number; first marginal well developed with a sharp tip when fresh; second marginal smaller and with a blunt tip even when fresh; molar tooth bearing a ventral rounded flange and ending posteriorly in a kind of heel ( Fig. 19). Thorax: pronotum sellate. Fore coxa flanged ventrally resulting in a sharp carina. Gut: enteric valve seating on left side, best seen in ventral view, situated in posterior half of abdomen. The enteric valve with either a fungifaber or a finitimus EVA or an intermediate EVA (but in one species, I. acristatus sp. nov., with a muneris EVA), always with a triradial symmetry; the odd PCs, in their downstream part, either bear crests that are as high as or higher than they are wide, with long and strong bristles ( fungifaber EVA , Fig. 4) or bear at that place a higher density of rather short bristles on a globular bulge ( finitimus EVA , Fig. 5); supporting bristles are generally numerous: 13–38 on each side of the odd PCs (with an exception in I. acristatus ); secondary cushions are wide at the upstream end, narrowing noticeably downstream with a homogeneous spine scattering. Caecum rather small, visible in ventral view, near centre of abdomen, frequently with two or three small lobes or developed as a finger-like process that is sometimes capped or swollen distally.

Morphologic analysis

The worker’s valve remains the best criterion to recognise the genus Isognathotermes , with the exception of I. acristatus sp. nov. whose valve is similar to the Ternicubitermes pattern. Additionally, the spinosity of workers’ fore coxae characterises well the genus Isognathotermes (including I. acristatus ), although this feature is shared with Polyspathotermes and some species of Nitiditermes .

While the soldier caste allows the best separation between species, several soldier characteristics need to be brought together to delimit the Isognathotermes genus (see diagnosis).

The imagines provide few discriminating criteria and are sometimes difficult to distinguish from other Cubitermitinae imagines. As a notable exception, however, four subspecies of I. ugandensis , practically indistinguishable from the soldiers, could be evidenced from the imagines’ overall dimensions, as well both eyes and ocelli relative dimensions.

Ideally, therefore, all three castes should be available for proper identification, bearing in mind that Isognathotermes termite mounds, especially older ones, often contain several species and can lead to misidentification.

An initial distinction between finitimus – type and fungifaber – type EVAs is essential: a Principal Component Analysis (PCA) on the soldiers’ morphology involving all the species seems indeed confusing ( Fig. 20). However, applying a PCA separately to species with a fungifaber – type EVA ( Fig. 21) and those with a finitimus – type EVA ( Fig. 22) results in better (though not yet complete) separation of the species. Nevertheless, it is important to note that creating sub–genera based on EVA types is not feasible because they do not delineate two distinct clades.

Colour does not provide any diagnostic criterion: it depends on both the age of the individual at the time of capture and the time spent in the alcohol before being studied.

Chorology-ecology

This genus has been collected in forests and savannahs of West, Central, and East Africa, from Gambia to Kenya and from Angola to Malawi ( Fig. 32). The species with a finitimus EVA are almost restricted to forested areas of central Africa while the species with a fungifaber EVA or an intermediate EVA can be found in forests and savannahs and is therefore present in a much larger area, from west to east equatorial Africa. There are only a few localities where species with finitimus and fungifaber EVAs coexist ( Fig. 32). All species are soil feeders.

Phylogenetic analyses

Our phylogenetic tree integrates the mitogenomes of 101 samples from the genus Isognathotermes ( Fig. 33). Mitogenome similarities between all members of Isognathotermes are larger than 91.23% (Table ST2). These analyses revealed that the analysed samples belonged to 12 lineages (out of the 15 morphological valid species reported herein), excluding the unidentified species represented by the samples DJ P141 and DJ 0937. All species were monophyletic, with the exception of the intricate paraphylies of I. similifinitimus sp. nov. and I. phallicaecalis sp. nov. Note that our exclusive use of (maternally-inherited) mitogenomes may have limited our ability to fully resolve the evolutionary relationships among species of Isognathotermes .

The 12 species could further be subdivided into five main clades. The first four clades are monospecific, all supported by SHalrt> 99 and UFB = 100, and respectively composed of: the species (i) I. acristatus sp. nov.; (ii) I. rectimalatus sp. nov.; (iii) I. bulbifrons ; and (iv) I. fungifaber . The fifth clade combines the eight remaining species: I. severus , I. minitabundus , I. ugandensis , I. finitimus , I. phallicaecalis sp. nov., I. phalloides sp. nov., I. planifrons , and I. similifinitimus sp. nov., While the relationships among species of the first four clades are highly supported (SHalrt> 99 and UFB = 100), the relationships between species of the fifth clade are unresolved.

Interestingly, the clades evidenced from the mitogenomic tree did not perfectly match EVA patterns. While clades I and IV are limited to the fungifaber EVA , and clades II and III to the finitimus EVA ; the last unresolved clade V groups species exhibiting both EVAs.

Additionally, we reconstructed a phylogenetic tree based on the COII gene (Supp. file 2) to study the placements of samples from Roy et al. (2006) and Hellemans et al. (2021) in the herein presented larger sampling of Isognathotermes . The COII tree included 120 samples of Isognathotermes , and recovered similar patterns as the mitogenome-based tree including the paraphyly of I. similifinitimus sp. nov. and I. phallicaecalis sp. nov. In addition, both I. phalloides (DJ 0459) and I. aff. planifrons (DJ P141) were inserted within I. severus . We specifically detail our reanalyses of samples within I. fungifaber and I. planifrons in their respective sections.

Identification key

Isognathotermes View in CoL species show very high intraspecific variability: Figs 26–31 clearly show the extent to which the measurements of different species overlap. In the absence of molecular data, the identification of Isognathotermes View in CoL species thus remains difficult and it was not possible to produce an accurate key based on a single caste. We also remind the reader that several species may live together in the same termite mound, thereby complexifying the identification task. As few selected measures are rarely sufficient to recognise a species, a multidimensional approach (PCA) has been used in this study; however, such multidimensional approach cannot be easily incorporated into a dichotomous key.

The dissection of a worker is essential to recognise its enteric valve and caecum types; precise measurements of soldiers’ head and mandible length and curvature (in ventral view) are required as well as geographic origin.

The identification key begins with the genera derived from Cubitermes sensu Wasmann View in CoL (items 1–4) and then gives the species of Isognathotermes View in CoL (items 5–18). Incipient colonies (one or two soldiers with some workers and a non-physogastric queen) are excluded from the key. Abbreviations: see the chapter Material and methods.

1. Worker and soldier: conical fore coxae bearing some fine, soft setae but not any spines or spine-like setae. Imagines: conical fore coxae bearing many soft setae ............................................................ 2

– Worker and soldier: carinated fore coxae bearing some spines or spine-like setae (rarely none in the soldier). Imagines: carinated fore coxae bearing some spine-like setae among many other soft setae ........................................................................................................................................................... 3

2. PCs of the workers valve bearing no more than 3–6 supporting bristles on each side; SCs generally with a heterogeneous scattering of spines (the spines are lacking in some irregular spots). This criterion also applies, albeit less clearly, to soldiers............................................................................ ............................................................... Cubitermes Wasmann, 1906 View in CoL (sensu Hellemans et al. 2021)

– PCs of the worker’s valve bearing 10–30 supporting bristles on each side; SCs with either a homogeneous scattering of the spines, or with a heterogeneous scattering tending towards the shape of a spearhead ..................................................................... Ternicubitermes Josens & Deligne, 2021

3. Worker and soldier EVAs: SCs narrow (not wider or only slightly wider than the PCs). Generally, two PCs ending in a sclerified spatula but, in some samples from West Africa, not any spatula and, in this case, PC1 much longer than the others (WVP1-bil = 1.42–1.62). Soldiers on average smaller: SHdL = 1.55–2.45 mm........................................................................... Nitiditermes Emerson, 1960 View in CoL

– Worker and soldier: SCs clearly wider than the PCs; PC1 equal to or somewhat longer than the others (WVP1-bil = 1.00–1.26). Soldiers on average larger: SHdL = 2.05–3.80 mm................................. 4

4. Soldier: triangular or finger-shaped labrum lobes. Worker and soldier EVAs: three to six PCs bearing a sclerified spatula (only sketched in some soldiers)....... Polyspathotermes Josens & Deligne, 2021

– Soldier: lyre-shaped labrum or with triangular lobes. Worker and soldier EVAs: PCs not spatulated (or sometimes with soft, not sclerified spatulas) ...........................5 Isognathotermes Sjöstedt, 1926 View in CoL

5. Worker’s EVA: odd PCs distally narrowed and crested and bearing strong, long bristles: fungifaber EVA , Figs 4, 6, sometimes an intermediate EVA, Fig. 6) but one species, I. acristatus sp. nov., with a bilobatodes EVA)............................................................................................................................ 6

– Worker’s EVA: odd PCs distally barely narrowed, rather swollen and bearing numerous strong, short setae: finitimus EVA ( Figs 5–6) ........................................................................................................11

6. Sample from regions south or east of the Congolian forests; soldier’s left mandible on average less curved (SMlc = 0.08–0.24 mm); asymmetrical curvature of the mandibles: the right is generally more curved than the left (SMlc/SMrc = 0.51–1.03).................................................................................. 7

– Sample from West Africa or regions north of the Congolian forests; soldier’s left mandible on average more curved (SMlc = 0.14–0.33 mm); curvature of the mandibles less asymmetrical, SMlc/SMrc = 0.79–1.24).......................................................................................................................................... 9

7. Smaller soldier (SHdL = 2.44–2.59 mm); worker’s valve with almost non-existent crests on the odd PCs ( bilobatodes EVA); sample from Angola....................... I. acristatus Josens & Deligne sp. nov.

– Larger soldier (SHdL = 2.91–3.72 mm); worker’s valve with well-developed crests on the odd PCs; sample from a more eastern country.................................................................................................. 8

8. Two species very difficult to distinguish without sequencing. Sample from East Africa (from Uganda to Northern Zambia: Fig. 49). Four subspecies, only identifiable from imagines, see chapter I. ugandensis ........................................................................................... I. ugandensis ( Fuller, 1923)

– Sample from the DRC (Kasaï–Katanga) or Northern Zambia ( Fig. 49).............................................. ......................................................................................................... I. minitabundus ( Sjöstedt,1913) View in CoL 9. Soldier with longer mandibles (SMlL = 2.75–2.99 mm in ventral view); sample from Cameroon................................................................................................. I. zenkeri (Desneux, 1904)

– Soldier with shorter mandibles (SMlL = 1.74–2.58 mm) ............................................................... 10

10. Two species sometimes difficult to distinguish without sequencing. On average, smaller soldier (SHdL = 2.37–3.11 mm, ST3L = 1.26–1.67 mm), smaller worker (WHdW = 0.94–1.25 mm), imago of the same size; sample of evergreen forests from southern RCI to Gabon ( Fig. 45) ....................... ............................................................................................................... I. fungifaber ( Sjöstedt,1896)

– On average, larger soldier (SHdL = 2.43–3.29 mm, ST3L = 1.52–1.93 mm), larger worker (WHdW = 1.06–1.26 mm), imago of the same size; sample from various ecosystems (savannah, orchard, woodland, forest gallery or secondary forest, or from the evergreen forest in RCI near the border with Liberia or from the Lobaye province, CAR); large geographical range across Africa from Gambia to CAR ( Figs 45, 49)...................................................................................... I. severus ( Silvestri, 1914)

11. Soldier’s and worker’s caecum extended forward in a finger-like process, sometimes swollen, or capped distally (phalloid caecum, Figs 55–56) ............................................................................... 12

– Soldier’s and worker’s caecum rather small, generally with 2–5 short lobes ( Fig. 2) sometimes a small amorphous button................................................................................................................... 13

12. Mandibles, on average, less curved SMlc = 0.18–0.26 mm, sample from the evergreen Congolian forest ( Congo Republic, Cameroon) ............................... I. phallicaecalis Josens & Deligne sp. nov.

– Mandibles, on average, more curved SMlc = 0.19–0.33 mm, sample from gallery forests in the southwest and west of the DRC of from northern Angola............ I. phalloides Josens & Deligne sp. nov.

13. Smaller species (SHdL = 2.42–2.66 mm) from Gabon........... I. modicus Josens & Deligne sp. nov. – Larger species (SHdL = 2.65–3.77 mm) ......................................................................................... 14

14. Worker’s left mandible with a larger apico-marginal distance (WMlAmD = 0.20–0.21 mm), soldier’s mandibles, especially the right one, very slightly curved (SMrc = 0.13–0.19 mm, SMrC = 0.20– 0.25 mm); sample from Angola........................................ I. rectimalatus Josens & Deligne sp. nov.

– Worker’s left mandible generally with a smaller apico-marginal distance (WMlAmD = 0.15– 0.22 mm); soldier’s mandibles generally more curved (SMrc = 0.16–0.38 mm, SMrC = 0.24– 0.49 mm); sample from elsewhere .................................................................................................. 15

15. Larger species (SHdL = 2.98–3.77 mm) from Haut-Uele or Kivu, DRC............................................ .................................................................................... I. magniplanifrons Josens & Deligne sp. nov. – Generally smaller species (SHdL = 2.60–3.76 mm) from elsewhere ............................................. 16

16. Soldier’s head generally more heavily humped (SHdC = 0.01–0.12 mm); worker’s valve with longer secondary cushions (WVS-AvL = 0.47–0.62 mm) ............................... I. bulbifrons (Sjöstedt,1924)

– Soldier’s head generally more slightly humped (SHdC = 0.01–0.08 mm); worker’s valve with shorter secondary cushions (WVS-AvL = 0.38–0.56 mm) ......................................................................... 17

17. Soldier’s mandibles somewhat more curved (SMlc = 0.19–0.35 mm); indistinguishable workers and imagines; sample from forests of western-central Africa ( Cameroon, CAR, Gabon, Congo and Kongo-Central, DRC, Fig. 47) ............................................................. I. planifrons (Sjöstedt,1924)

– Soldier’s mandibles somewhat less curved (SMlc = 0.16–0.31 mm); indistinguishable workers and imagines; sample from continental forests of central Africa .......................................................... 18

18. Two species indistinguishable without sequencing; sample from the continental forests of Central Africa ( CAR, DRC, Uganda, Fig. 38) ................................................... I. finitimus (Schmitz, 1915) – Cryptic species; only one site known to date, south of Kisangani, DRC ( Fig. 47) .............................. ........................................................................................ I. similifinitimus Josens & Deligne sp. nov.

Valid species and subspecies

Kingdom	Animalia
Phylum	Arthropoda
Class	Insecta
Order	Blattodea
Family	Termitidae