For me, the names of few African rivers have the power to inspire a daydream like the Lukuga. It has to do with this river being the sole link between the very different fish faunas of Lake Tanganyika and the Congo River, its inaccessibility and unexplored status. The horrific death of South African kayaker Hendri Coetzee by crocodile attack on the Lukuga in 2010 has also added some Conradian dread to the river's mystique, at least in my mind.
Many of us have had the experience of arriving some place in Africa, sure we are the first ichthyologist to have been there, only to learn from locals that someone matching the description of Dr. Tyson Roberts planted his flag there years ago. No one in ichthyology provokes more reactions and elicits more stories than Tyson Roberts and few have contributed as much to knowledge of African fishes in the past half century. It seems only fitting that an ichthyologist with a mystique of his own should be the one take on the Lukuga.
Despite the fact that he was there 26 years ago, it was not until last month in Ichthyological Explorations of Freshwaters that the results of his collections were published in an article he coauthored with Dr. Sven Kullander of the Swedish Museum of Natural History. The paper, entitled "Out of Lake Tanganyika: endemic lake fishes inhabit rapids of the Lukuga River" is available as a free download from the journal's webpage.
While Dr. Roberts apparently traveled the full length of the entire river from Kalemie on Tanganyika (L.T.) to its outlet in the Lualaba (Upper Congo), the paper focuses on species collected in the vicinity of the confluence of the Niemba River into the Lukuga, just above a major falls, about 100 km from the Lake. Above these falls one finds a mix of lacustrine species, some endemic to L.T., and riverine species from the Upper Congo.
Roberts made no collections downstream of Niemba on the Lukuga itself (they do not say why), but from tributaries of the Lukuga that empty into its lower reaches, the Lufwango and the Luizi, where the fish fauna is exclusively Congolian.
In three tables, the authors present data on the 57 species recorded in previous studies of the Lukuga (notably by Boulenger and Poll) and from the 1986 collections: the 39 species collected in downstream tributaries and 43 collected at the Niemba confluence.
Of the fishes from the Niemba confluence, 13 of the 43 are species found also in L.T., 10 of the 13 being Lake endemics. Twenty-one of these species are known also from Kalemie, on L.T. at the origin of the Lukuga. Ten are cichlid species, and six of these otherwise are known only from L.T. These six are Astatotilapia burtoni, Ctenochromis horei, Simochromis babaulti, S. diagramma, Telmatochromis dhonti, and Tylochromis polylepis. They collected an undescribed Tanganicodus cichlid that they are describing elsewhere that they say is closely related to Tanganicodus irsacae, which inhabits rocky shores of L.T. Oreochromis upembae from the Lualaba is at this site, the only non-L.T. cichlid there, if you don't count the unwelcome presence of the invasive Oreochromis niloticus, that owes its presence to human introduction in the L.T. basin.
The site apparently also harbors an undescribed Chiloglanis mochokid, different from C. lukugae which also occurs there. The two clupeids were colleted at Niemba illustrate the regions mixed affinities: Stolothrissa tanganicae is otherwise endemic to the Lake and two small specimens of similar to Microthrissa minuta from the Congo.
Interestingly, the Lukuga only intermittently receives outflow from L.T., when lake levels are high and much of its flow comes from its major tributary, the Niemba River. The authors make the case that the lacustrine species they collected near the Niemba confluence 100 km from the lake represent not merely "spill over" vagrants, but a stable fish community. While the falls below the Niemba confluence may effectively prevent Congo fishes dispersing into L.T. via the Lukuga, the very different water chemistry of L.T. (hard water, high pH, high conductivity) and Congo freshwaters (medium-low pH, low conductivity and hardness) may be the reason why L.T. lacustrine forms don't make it all the way down the Lukuga to the Lualaba. Presumably, other Lukuga tribs make the water less L.T.-like and more Congo-like in the Lukuga's lower reaches.
The authors point out that because there are no collections of fishes from between Niemba and the Lualaba, the point where the L.T. influence in the Lukuga disappears remains somewhat speculative. The idea that water chemistry is major barrier to dispersal is an interesting one. I especially wonder about the mormyrid electric fish, whose electrosensory system would seemingly have to be impedence-matched to their environment. These fishes are much more diverse in the low-conductivity rivers than in the Lake, although there are some present. Could the lacustrine Cyphomyrus (Hippopotamyrus) discorhynchus' adaptation to high conductivity lake water (~700 µS/cm) prevent it from dispersing into the low-conductivity Lualaba via the Lukuga? A great future study for intrepid ichthyologists who are less intimidated than I by man-eating crocs.
For some interesting geologic history of the Lukuga and other African rivers, check out this page by geoglogist Bill Butler.