The Ants of Africa
Chapter 5 - Ant Ecology in Guinea - the contribution of Bernard and Lamotte
Chapter 5 - Biodiversity and Niches

For whatever reason, one has to hope not francophobia, the Section "II - Esquisse Ecologique" (Ecological Effects) found in Bernard (1952) appears never to have been appreciated by the authors of the international literature on ants. Thus, here, I have attempted to translate and summarise the work on Guinea done together by Bernard and his collector colleague, Maxime Lamotte. As I found Bernard's presentation a little confusing, I have made some rearrangement of the text.

At that time, 1952, Bernard noted the only previous ecological study of sufficient content had been that of the Imatong Mountains in Sudan by N.A.Weber. Lamotte, the field worker, had gathered complete quantitative data from two principal stations - the savanna at Kéoulenta and the montane grassland of the Mount Nimba crest. This, with other ant data enabled them to try to define the fauna of the principal local ecosystems - savanna, forest, montane grasssland - and to outline the dominant or unusual in each. Of the ant species (or forms) found some 60, of 193, were sufficiently abundant for the analysis.

Elsewhere, I have summarised the wider information on the general geography of Mt. Nimba (Click here).

Influence of ecology

Using the 60 relatively abundant species, only 15 species of were ubiquitous; 16 others were found in two of the three ecological areas; and, 29 were found in only one area.

In general terms, the savanna areas showed the lowest endemicity, with the common species being those found throughout sub-Saharan Africa. The apparently local endemic, Crematogaster lamottei was an exception, and was also common to the forest.

Ants in the Savanna areas

(a) Kéoulenta, 500-550 m, east of the massif - the best understood of the savanna areas. Survey D, made of 100 m² showed a high abundance of Crematogaster species (impressa and africana) inhabiting hollow stems and termite nests. Numbers sometimes reached 5 nests per m², each with 50-100 individuals. In decreasing order of abundance were three Pheidole species and large "Paltothyreus" (Pachycondyla tarsatus), at up to 9 per m², three Camponotus species, large and totalling perhaps 100 per m². The mass of ants present came to around 475 mg/m²; somewhat less than the mass of termites; the latter being less numerous but heavier. In the wet season only the fungus-growing termites remained generally active, the poorer structures of the Bellicosotermes usually succumbed to the rain and the inhabitants were predated by the ants.

(b) Ziéla, similar to Kéoulenta, with similar fauna but with the surprising absence of Camponotus solon, Crematogaster lamottei and Polyrhachis militaris; all common in the forest and Kéoulenta. Although the overall collection was less (27 spp cf 54 at Kéoulenta), two forest species - Dorylus stanleyi and Crematogaster bequaerti were found.

(c) Sérengbara, western drier savanna, at 590 m. In this area of low Cyperaceae several of the species present at Kéoulenta are absent, the whole ecosystem is impoverished and the species found (Pheidole speculifera, Crematogaster impressa and Crematogaster lamottei) were in lower numbers and were those which nested mainly in stems.

Ants in the Forest areas

Collections were made at seven stations, as follows;

(a) Yalanzou, dense forest, 25 km west of the massif - 30 species taken. An abundance of Camp. maculatus melanocnemis was found, disappearing above 700 m but reappearing on the montane grassland. The three species were more abundant on Nimba than in the rest of French West Africa - Pheidole picata, Crematogaster lamottei and Monomorium pharaonis again absent above 450 m but reappearing at 500 m.

(b) N'Zo, the driest forest area, from 480-500 m, 39 species were found by Lamotte. The list included forms typical of the massif (Crematogaster lamottei etc.) and also of the plain and cultivated land, such as Myrmicaria natalensis and Oecophylla longinoda.

(c) Nion, 700 m, Station 22, cleared secondary forest, on the crest, 32 species found. Similar to N'Zo, but lacking the plain species and with three previously unknown and remarkable forms - Centromyrmex constanciae, Asphinctopone lamottei and Tetramorium guineense. Although the forest had been altered, the presence of rare subterranean species showed the initial soil conditions remained.

Nion was the location for the extensive effort by Villiers, who, at the end of the wet season, sifted the leaf litter of the northeast slopes, between 700-900 m. The samples were particularly rich in new ponerines and also a new Oligomyrmex (then thought to be from a new genus, Nimbamyrma ). It was the only ecotype of the massif without the common Pheidole speculifera and Crematogaster africana "dominants elsewhere". The very small species Pheidole picata and Pheidole melancholica were common in the leaf litter.

(d) Camp IV, beautiful primary forest at 950-1000 m, above Nion. Very well wooded, this location had several species absent or rare elsewhere in the forest. Camponotus solon, for instance, seemed to be localised in insolated clearings. On the other hand, the frequency of Polyrhachis concava, several Anochetus species and members of the genera Phyracaces and Phrynoponera were reminiscent of the Congo Basin forests.

(e) Zouépo, station B, 1250 m, in the high montane zone where the common plains species (such as Crematogaster impressa, Pachycondyla tarsata and Monomorium bicolor) had disappeared. Pachycondyla soror numbers increased rapidly from here on to the montane grassland of the crests. Few new species were found, the most curious was Tetraponera nasuta, probably inhabiting stems.

(f) Ravin I of Mount Tô, 1100-1600 m. A broad ravine formed by a fault, rather less steep than the other ravines and falling east towards Ziéla. Of all Lmotte's collecting sites, this seemed to be the one most rich in Congo Basin forms. The soil, less shaded than in other sites, favoured the grassland species Hypoponera abyssinica and Pachycondyla soror. A single species known from Angola, Lepisiota monardi also appeared common.

(g) The crest of Nion, at 1300 m, woody scrub, with intense fogs, trees and shrubs, containing a grassland with Polyrhachis schistacea.and other montane grassland species (see below), but also with forms of the Congo Basin forest (Phrynoponera gabonensis, Crematogaster pulchella, Pseudolasius weissi, etc.).

Ants of the Montane Savannah

A single location provided most of the ants. this was Camp I at 1600 m, very well surveyed by Lamotte. Some 38 species were found in an area resembling the summit areas of Mount Cameroun - low grasses, with two-thirds of the vegetation mass rotting (which favoured Technomyrmex andrei), shallow black soil and dominated by dense fog or mist.

Near Camp I, stones or rocks were abundant and provided cover for nests of four species of Pheidole (notably Pheidole picata); soil without rocks was colonised by Camponotus congolensis and Camponotus traegaordhi (previously known only from Natal, South Africa). Other species were subordinant but commonly the nomadic species Pachycondyla soror and Dorylus nigricans were encountered, less common was Dorylus lamottei. This zone was notable for the absence of termite hills and to this was attributed the absence of most Crematogaster species, other, that is, than Crem. africana. The low minimum temperature, around 12°C also played a part in limiting the species of ant, hence the absence of the Camponotus and Polyrhachis species so abundant in the savanna. In the dry season Lamotte found some 100-200 ants/m², some half to a quarter of the numbers in rich savanna. The wet season had an even more extreme influence than in the plains, with Oligochaetes and molluscs dominating. On an eastern slope, at Site E, an area without rocks, ant levels did not exceed 40-60/m², mostly being Camponotus traegardhi, Camponotus congolensis and Camponotus abjectus.


Overall, the savanna species have to be resistant to fire, to flooding, and tolerant of high variations in temperature. Crematogaster species were seen to have great vitality, dominant in the savanna, while three Pheidole and Ponerines were quantitatively more important in the forest and on the crests.

The forest zone had the greatest number of species, as, thank to the humus, there was a rich reservoir of terrestrial Ponerines. Also here was the majority of endemics, with some 30% of the total forms being previously unknown (as compared to 14-18% in the other biogeographic zones). Altitude had little effect, with similar numbers from each of the several locations. Clearings providing drier conditions showed mixtures of savanna and montane grassland species. True forest species were mainly those known from the Congo Basin forests, confirming the subequatorial nature of Nimba.

In the montane crest zone the endemic Ponerines and Myrmicids, so varied in the forest, were absent; other than Pachycondyla senegalensis and Hypoponera abyssinica. Although no new species were found, there was a surprise as three species hitherto known only from southern Africa were found - Pheidole picata, Camponotus traegaordhi and Lepisiota fervida. So, other than these three peculiarities, the species found were with greatest affinity for the plains savanna. Crematogaster species with their forest origins were generally absent.

Ant density as influenced by slope

The key factor here is the impact of the seasonal rainfall. In the dry season, the ants increase in dense savanna, are evenly abundant in the forest and freshly dominate the montane grasslands, provided there are rocks for shelter. Some 40-400 individuals can be found per m² of excavated soil. Numbers drop in the wet season; the ants are inundated and the surface is invaded by Oligochaetes, molluscs and collembolans. Numerically, it is the small species which prevail (three Pheidole and three Crematogaster; but the larger species, such as Pachycondyla tarsatus and Camponotus maculatus, can number more than 100 workers per m². The lowest abundances are on the very steep slopes and the impoverished hot savanna (Sérengbara, etc.)

Thanks to their ability to spread, the ants are widespread over much of the massif and the localisation of each species depends mainly on the ecology (insolation, humus, soil permeability, etc.). The eastern slopes towards the Ivory Coast seem little different in this respect to the western Guinean slopes. However, the easterly slopes have a preponderance of species.


West and north-west areas (Yalanzou, Nion, Sérengbara) - only one species, Odontomachus assiniensis appeared particularly favoured by this sector. The species was abundant also on the mountain crests but became subordinate in the areas which follow.
Eastern areas (N'Zo, Ziéla, Kéoulanta) - the species common here, both in forest and savanna, were rare or absent in the drier areas. They were Pheidole picata, Pachycondyla caffraria, Tetramorium humerosum, Cataulacus pygmaeus, Camponotus vividus, Camponotus chrysurus and Camponotus orthodoxus. Arguably, this pattern owes more to the origins of the species than to the biotype, as the seven species appear to be of more easterly origin and may be recent arrivals on the massif. This contrasts with Odontomachus assiniensis which clearly has westerly origins. Either way, the other 54 common species appeared highly indifferent to the slope (NOTE: B&L used "dominante" but I use "common" in preference to the literal translation of "dominant").

Ant density as influenced by termite colonies

Lastly, Bernard and Lamotte emphasised the importance of termite colonies in the distribution of all the ants. To start, the majority of ants feed on termites, especially the Ponerines and Camponotus species, even occasionally Crematogaster, Pheidole, etc. Lamotte's findings were that the mass of termites per m² usually exceeded that of ants, provided there were permanent resources for the termites. this was especially so in the wet season when savanna termites disperse throughout. Next, in localities where rocks are rare as sources of shelter for ant nests, old termite nests (hills?) frequently provide a useful habitat. Such use of termite nests, however, was less common in primary forests, even less in flood prone land and almost nil on the montane grasslands (where mound-building termites were not found).

©1998-99 - Brian Taylor CBiol FSB FRES
11, Grazingfield, Wilford, Nottingham, NG11 7FN, U.K.