Lake Bogoria, in the Rift Valley of Kenya is an extreme saline lake (conductivity 40-80 mS cm −1 , alkalinity 1500 m equ l −1 ). It is hydrologically more stable than the other, endorheic lakes in Kenya, because it is deepmaximum depth at present just over 10 m in an area of 3000 ha -and so does not have periods when it is dry. It is ecologically simple, with only one species dominating the phytoplankton -the cyanobacterium 'spirulina', Arthrospira fusiformis. Its biomass and productivity were very high -biomass between 38 and 365 µg l −1 chlorophyll 'a' and 3.4-21 × 10 3 coils ml −1 and net production between 0.24 and 1 gm C m 3 h, the latter in a narrow zone of less than a metre. There were no macro-zooplankton in the plankton and the only grazer of A. fusiformis was the lesser flamingo, Phoeniconaias minor, which occurred irregularly in very high concentrations (in excess of 1 × 10 6 ). Detritivory in the benthos was effected by a single chironomid species, Paratendipes sp., at a maximum density of 4 × 10 4 m −2 . The mean daily emergence of adult chironomids was estimated to be 1 × 10 3 m −2 , the maximum 3. There was no littoral plant community within the lake but 44 dicotyledonous and 31 monocotyledonous plant species in the drawn-down zone and adjacent to it. A diverse draw-down terrestrial invertebrate fauna, only superficially described here, processed the flamingo feathers and carcasses, with other detritus such as chironomid pupal exuviae and decaying A. fusiformis scum. About 50 bird species depended upon the chironomids, either as they emerged through the water column as flying adults or later on the shoreline as floating pupal exuvia and dead adults. The lake has high conservation value because of three bird species in particular -lesser flamingo, Cape teal and black-necked grebe. The former provides real economic value in a region otherwise impoverished, because of the spectacle of tens of thousands of flamingos set against the landscape of hot springs and fumaroles at the lake edge, which draws 15 000 visitors per annum. P. minor has experienced three periods during the past ten years when major mortalities have occurred, the last of which killed 700 birds day −1 . This could have involved as many as 200 000 birds (about 1/5th of the maximum population at this lake) if mortality was at a constant rate for the nine months it was observed. Causes of mortality have been suggested as avian tuberculosis, poisoning from cyanobacterial toxins or from heavy metal contamination at Lake Nakuru, but it is still not yet clear what contribution each makes to the problem.