Methods and Madness




Subterranean ants: the last frontier

Hamish Robertson
South African Museum, Iziko Museums of Cape Town, P.O. Box 61,
Cape Town 8000, South Africa;
tel. +27 21 4243330; fax +27 21 4246716;
e-mail hrobertson@iziko.org.za

     A couple of weeks ago I saw Leptanilla workers for the first time. They were from a Winkler sample from grassland at Underberg in KwaZulu-Natal, South Africa and encountering them did not come as a surprise as much as relief at seeing them at last. The lack of surprise was because we have been finding male Leptanilla specimens in all our surveys round southern Africa, most of them being in malaise trap samples. In a survey of the Brandberg Massive in Namibia, nine species of Leptanilla males were collected, putting it in the top three most diverse genera in the study (Robertson 2000). Other localities have each yielded 4-5 species. These males are minute but species can be relatively easily distinguished on the basis of their genitalia which are relatively large and can be examined in situ.

      The reason that Leptanilla workers are so difficult to encounter is that they are entirely subterranean so that even sifting leaf litter and superficial soil tends to miss them. They are also minute so that it is highly unlikely one would see them while excavating soil. Finding the workers at Underberg might have been because we have been steadily perfecting our sifting technique for obtaining Winkler samples. From Brian Fisher we learnt the benefits of using a panga to chop the vegetation and soil before sifting it. After chopping it, you use a hand rake to collect the debris together for sifting. It is essential to keep the panga sharp and this is best done with a metal file. Underberg is a typical rural village servicing the farming community. Shops are sparse but the one thing it does have is an excellent hardware store. It was here that we found some broad-bladed pangas with right-angled tips called cane knives which work much better at chopping grass and soil than do the normal sized pangas. This type of blade enables one to chop well into the soil where all those Leptanilla and other interesting goodies are lurking.

      We have found that if you are sifting in habitats such as grassland and open scrub, which can have very little leaf litter, it is essential for a good return to do the sampling soon after rain when the soil is moist. The more arid the area you are sampling, the more likely it is that your field trip will not coincide with recent rains, but the more important it is that you do sample after rains because the upper soil levels become so dry. There is nothing more disheartening than sifting dry soil and litter - you find yourself asking at regular intervals "why am I doing this?".

      So sifting superficial soil and litter and placing it in Winkler sacks is one way of accessing subterranean ants. However, one constantly get the impression that one is still missing many of the subterranean species. One can see this from the strange alates in light trap and malaise trap samples. Also, examining overturned soil for ants and brood (the latter of which show up better than the workers) often reveals species that have not been picked up in Winkler surveys at the same site. Brian Fisher and I did surveys of ants in grassland and forest at a site near Ambositra in Madagascar (Fisher and Robertson 2002) in which we used pitfalls, Winklers, beating, and digging holes. The latter method involved, for each transect, digging 25 30x30x30 cm holes and collecting samples of the ants in the soil that was dug out. As this method was my idea, I landed up digging all the holes and I must admit that in view of the extra species we obtained using this method, it was not worth the time and energy expended. Nevertheless, it yielded a few species that were really interesting. This method was inspired by my earlier experiences of digging holes at Mkomazi Game Reserve in Tanzania. My non-quantitative approach at this site was to start off by digging up a nest I had discovered (e.g. some large species such as Camponotus) and while doing so to keep an eagle eye out for smaller subterranean species. In addition, I would sieve some of the soil through a fine sieve which helped greatly in picking out minute ants. I would often carry on digging long after I had finished collecting the nest I was originally after and would only stop after the number of species started tailing off. In this way I found the first record of Anillomyrma for Africa, two new species of Amblyopone, a new species of Anochetus (quite a large yellow species with small eyes that was in vacated termite or rodent channels underground) and numerous other interesting species, often in nests with associated alates. I suspect that the hole digging method we used in Madagascar would have worked better at Mkomazi because ant species diversity was much richer at the latter locality.

      Digging holes and picking out the ants one finds, clearly can be very rewarding but it is almost impossible to spot the minute species such as those of Leptanilla. Perhaps it would be an idea to put some of the dug soil into a Winkler bag. I have tried some other methods of obtaining subterranean ants. For instance, using an auger, I ground up soil to a depth of 1 metre, immersed the soil in salt water, and skimmed off the debris that rose to the surface. This debris was preserved in ethanol and then examined under a microscope. I found very few specimens using this method and it really wasn't worth the effort. Perhaps I should try it at some other localities before giving it up entirely.

      If you have found any useful ways of finding subterranean species, I would be pleased to learn about them. Subterranean ants really are the last frontier of ant collecting. We have good methods of collecting canopy, ground surface and leaf litter species but the rare species in ant surveys generally are the subterranean species because there is no collecting method that samples them well. It is almost a cave-like existence underground where life is buffered against the environmental extremes up above. Subterranean species in Africa generally have very wide distributions and this is perhaps because of this buffering effect. For instance, Plectroctena subterranea is found from West Africa right down into southern Africa and I have recorded it from the Brandberg in an arid environment where most of the large ponerine species have faded out because of the dryness. This buffering effect probably has meant that over geological time, subterranean species have been less likely to go extinct so that ancient lineages are preserved just as they are in cave-dwelling insects. For this reason, perhaps the really interesting ant discoveries of the future will be found in this realm.

References

Fisher, B.L. & Robertson, H.G. 2002. Comparison and origin of forest and grassland ant assemblages in the high plateau of Madagascar (Hymenoptera: Formicidae). Biotropica 34: 155-167.

Robertson, H.G. 2000. Formicidae (Hymenoptera: Vespoidea). In: Dâures - Biodiversity of the Brandberg Massif, Namibia (eds A.H. Kirk-Spriggs & E. Marais). Cimbebasia Memoir 9: 371-382.