Search

Human encroachment continues to degrade highland forests. Abbott's Starling (Pholia femoralis) is a little-known Eastern African endemic frugivore restricted to highland forest canopy at altitudes of 1 800-2 600m. It occurs patchily within its restricted range of about 5 100km², its global population of 1 250-5 000 pairs is declining and it is listed in the IUCN Red Data List as Globally Vulnerable. In Kenya, it inhabits Mt Kenya forests, the Kikuyu escarpment forest, and occasionally the Taita and Chyulu Hills, while in Tanzania it lives in forests of Mounts Kilimanjaro, Meru and North Pare. We studied the bird's distribution in two primary and two disturbed blocks of the Kikuyu escarpment forest from August to September 2003. Twelve transects, averaging 1.8km each, were established to determine the logging patterns and to count P femoralis at 130 points, using recorded calls. A total of 16 birds were sighted in the main Kereita block in three out of five transects, perching at a mean height of 30m. Possibly due to rainy weather, no sightings were made in Kieni block. The bird appeared to be sensitive to human disturbance, preferring secluded forest sections of tall indigenous trees with discontinuous canopy. This suggests that this species will only survive at this locality if logging and other forms of forest use do not exceed current levels. [ABSTRACT FROM AUTHOR]

Learned traits are thought to be subject to different evolutionary dynamics than other phenotypes, but their evolutionary tempo and mode has received little attention. Learned bird song has been thought to be subject to rapid and constant evolution. However, we know little about the evolutionary modes of learned song divergence over long timescales. Here, we provide evidence that aspects of the territorial songs of Eastern Afromontane sky island sunbirds Cinnyris evolve in a punctuated fashion, with periods of stasis of the order of hundreds of thousands of years or more, broken up by evolutionary pulses. Stasis in learned songs is inconsistent with learned traits being subject to constant or frequent change, as would be expected if selection does not constrain song phenotypes over evolutionary timescales. Learned song may instead follow a process resembling peak shifts on adaptive landscapes. While much research has focused on the potential for rapid evolution in bird song, our results suggest that selection can tightly constrain the evolution of learned songs over long timescales. More broadly, these results demonstrate that some aspects of highly variable, plastic traits can exhibit punctuated evolution, with stasis over long time periods. [ABSTRACT FROM AUTHOR]