Significant impacts on cave microclimate from large populations of the bat Rousettus aegyptiacus have been documented in three simple caves in pyroclastic rock of Mount Elgon National Park, Kenya, one of which, Kitum Cave, with few bats, acts as a control, indicating microclimatic variations in the absence of significant biological activity. Seven days of temperature logger records, and on-site mapping of rock and air temperature, humidity, and air flow provide the basis for modeling of heat, water, and CO 2 production and dispersion. Internal temperatures in the presence of bats in Mackingeny Cave and Ngwarisha Cave rise to ~18 °C above ambient (from ~12 °C to ~30 °C), but in the control site by only ~2 °C. Excess bat-generated energy is dissipated by conduction to rock and by ongoing air circulation, the strongest of which accompanies bat entry and exit flights. In Kitum Cave, temperatures that are substantially lower than bat thermo-neutral zone raise concern for Allee effects on long-term colony fitness: Modeling indicates that a population of at least 100,000 bats should promote colony vitality. Metabolic outputs were modeled to yield corrosional potential: At these population densities, were the caves in limestone, rates of surface denudation caused directly by metabolic outputs would be 1 m in ~80,000 yr. These results confirm that tropical bats can be effective niche constructionists, by optimizing microclimatic roost conditions, by longer-term bioerosional optimization of rock surfaces for roosting, and by long-term niche engineering through zoo-speleogenetic enlargement of roost volume.

The eastern flanks of Mount Elgon, an early Miocene stratovolcano, host caves (∼150 m long, ∼60 m wide, ∼10 m high) of debatable origin. Many animals, primarily elephants, “mine” the pyroclastic bedrock for sodium-rich salts. Speleogenesis has been argued to be primarily zoogeomorphic, or primarily dissolutional with only minor zoogeomorphic modification. This report provides the first detailed mapping and geomorphological study of the caves. Speleogenesis is polygenetic and strongly related to lithology. Geological units are, from the top down, ∼2 m of dense pyroclastic agglomerate cap rock over which water falls, ∼10 m of more permeable agglomerate, up to ∼0.2 m of discontinuous impermeable lava, ∼2 m of very soft and permeable agglomerate, and >2 m of impermeable swelling-clay tuff. Caves develop behind waterfalls under surface stream valleys by sapping of the incompetent agglomerate above the clay, and failure of the clay (aquiclude and base level for speleogenesis), followed by collapse of harder agglomerate layers above. The dominant passage shape is breakdown dome, with abundant fresh collapse. Geophagy by elephants and other species, and human mining significantly modify and enlarge the caves and remove collapse debris. These activities, focused on accessible and salt-rich units, create quasi-horizontal undercuts (up to ∼4 m tall and deep), the loci of which move upward as collapse raises the floor. Significant erosion also occurs by incongruent dissolution, corrosion, pressure release, efflorescence flaking, and biogeochemical activity from huge bat colonies. No evidence was found of channeled flow, or of phreatic or vadose activity. These caves are probably no older than Holocene.