Question: How does the floristic diversity of Afromontane rainforests change along an altitudinal gradient? What are the implications for conservation planning in these strongly fragmented forest areas that form part of the Eastern Afromontane Biodiversity Hotspot? Location: Bonga, southwestern Ethiopia. Methods: Based on evidence from other montane forests, we hypothesized that altitude has an effect on the floristic diversity of Afromontane rainforests in southwestern Ethiopia. To test this hypothesis, detailed vegetation surveys were carried out in 62 study plots located in four relatively undisturbed forest fragments situated at altitudes between 1600 m and 2300 m. Floristic diversity was evaluated using a combination of multivariate statistical analyses and diversity indices. Results: Ordination and indicator species analyses showed gradual variations in floristic diversity along the altitudinal gradient with a pronounced shift in species composition at ca. 1830 m. Upper montane forest (41830 m) is characterized by high fern diversity and indicator species that are Afromontane endemics. Lower montane forest (o1830 m) exhibits a greater diversity of tree species and a higher abundance of the flagship species Coffea arabica. Conclusions: Our results provide crucial ecological background information concerning the montane rainforests of Ethiopia, which have been poorly studied until now. We conclude that both forest types identified during this study need to be considered for conservation because of their particular species compositions. Owing to the high degree of forest fragmentation, conservation concepts should consider a multi-site approach with at least two protected areas at different altitudinal levels.

443 Diversity and Distributions, (Diversity Distrib.) (2005) 11 , 443-452 BIODIVERSITY RESEARCH ABSTRACT The diversity and distribution of lianas were studied in five Afromontane rain forests of Ethiopia. Quadrats of 20 × 20 m were laid down along transects in the Bonga, Berhane-Kontir, Harenna, Yayu and Maji forests. In all forests, 30,917 liana individuals belonging to 123 species in 87 genera and 40 plant families were recorded. The most species-rich families were Asclepiadaceae (14), Fabaceae (9), Annonaceae (7) and Cucurbitaceae (7). The top 10 dominant families represented 56% of the total number of species. Over 400 other plant species representing different life forms were recorded growing together with lianas. The lianas accounted for over 30% of the total woody plant diversity and over 20% of the total floral diversity in the study areas. The analysis of floristic composition of the forests indicates that the Berhane-Kontir Forest had the highest Fisher's diversity index α , and Yayu the lowest. Generally, there were low similarities between the forests in terms of species composition. Although lianas were abundant in almost all forests, there was a considerable variation among the forests in terms of density and spatial distribution. The major dispersal modes of lianas were anemochory (30%), mammaliochory (30%), ornithochory and autochory, and the four mechanisms of climbing of lianas were twining (54%), hooking (24%), rooting and use of tendrils. Altitude and human disturbance were found to be important factors affecting liana distribution. The need for sustainable management and use of lianas in the Afromontane rain forests is emphasized.

No abstract

Flowering pattern and seedling establishment of umbrella bamboo (Fargesia murieliae (Gamble) Yi) were studied in its native habitat, Mount Shennongjia in Central China. Here in 1996-2000, over 95% of the bamboo plants simultaneously flowered and died, extending from lower elevations to the higher mountains along the altitude and from southwest to northeast along the mountain settings. Bamboo seedlings emerged after the simultaneous flowering, achieving an average density of 5 460 seedlings· m -2 in the autumn of the year following the flowering. After a high mortality throughout the first winter, bamboo seedlings remained a stable density in following 2-4 years (1130-1230 seedlings· m -2 ). Seedling density positively related to the coverage of parent bamboo, but negatively to the herb layers.

No abstract