Tropical landscapes are dominated by land‐use systems, but their contribution to the conservation of biodiversity is largely unknown. Since changes in biodiversity in response to human impact are known to differ widely among taxonomic groups and guilds, there is a need for multidisciplinary collaboration of plant, vertebrate, and invertebrate experts. We used inventories of trees, understory plants, birds (subdivided into endemics, insectivores, frugivores/nectar feeders), butterflies (endemics, fruit feeders), and dung beetles in Sulawesi (Indonesia) to characterize a gradient from near‐primary to secondary forests, agroforestry systems, and annual crops. As expected, overall species richness tended to decrease within this gradient of increasing habitat modification, but, in contrast to previous studies, we found the species richness between most taxonomic groups to be significantly correlated (36 out of 38 pairwise comparisons). However, on average only 48% of the variance could be explained (within the five main groups), and only a few taxonomic groups/guilds turned out to be good predictors for others: for example, trees for fruit‐ and nectar‐feeding birds (88% explanation) and fruit‐feeding butterflies (83%), endemic birds for endemic butterflies (72%), and frugivorous/nectar‐feeding birds for fruit‐feeding butterflies (67%). Although biodiversity of land‐use systems showed taxonomic group‐ and guild‐specific differences, most groups were affected in a similar way by habitat modification. Near‐primary forest sites proved to be of principal importance for conservation; however, land‐use systems such as secondary forests (for understory plants, birds, and butterflies) and agroforestry systems (for butterflies) supported relatively high numbers of species and might play a significant role for biodiversity conservation in tropical landscapes.

Although the Guinea-Congolian rain forest region is an important focal point for conservation in Africa, very little information is available on the effects of forest modification and land use on the region's biodiversity. We studied bird communities and vegetation characteristics in 24 sampling stations distributed over two near-natural forests (near-primary forest, secondary forest), and two land use types (agroforestry, annual cultures) in the lowlands of the Korup region, Cameroon. Repeated sampling was used to establish near-complete inventories of bird assemblages for each site. Despite a 90% average drop in tree basal area from forest to farmland, overall bird species richness did not decrease significantly with increasing habitat modification. However, different groups of birds responded in different ways. Frugivorous and omnivorous bird species richness did not differ between habitats, whereas richness in granivorous, flower-visiting, and nonbreeding species was higher in land use systems compared to forests. In contrast, insectivorous birds, especially terrestrial and large arboreal foliage gleaning insectivores, and ant followers showed a declining species richness from forest to farmland. Also, richness in species of those restricted to the Guinea-Congolian forest biome and of the family Pycnonotidae showed a pronounced decline with increasing habitat modification. Species richness of overall insectivores, terrestrial insectivores, large-and medium-sized arboreal foliage gleaners, ant followers, as well as pycnonotids and biome-restricted species, were strongly or even very strongly positively correlated with overstory tree density and, in most cases, also with basal area. In contrast, tree density and basal area were strongly negatively correlated with species richness of nonbreeding visitors and flower-visiting bird species. Species composition was most distinct between near-primary forest and annual culture sites, and the abundance of 23 out of 165 species was affected by habitat, suggesting considerable partitioning of habitat niches along the habitat gradient. Our results stress the importance of tree cover in tropical land use systems for the maintenance of resident forest bird populations and confirm that natural forest management is more beneficial for global bird conservation compared to other forms of forest exploitation, including agroforestry systems.

There is still much debate over the potential value of land-use systems for the maintenance of tropical biodiversity. An increasing number of studies indicate that much forest biodiversity can also be found in the agricultural landscape matrix. Because there is little information on the potential value of land-use systems for tropical forest bird species, we conducted repeated point counts in near-primary forest, adjacent young secondary forest, modernized cacao agroforestry systems, and annual cultures at submontane elevations in central Sulawesi, Indonesia. Species richness decreased from natural forest and young secondary forest to agroforestry systems and annual cultures. Although species richness was similar between natural and young secondary forest, the number of endemic bird species was significantly lower in second-growth forest. Species composition gradually changed as the habitat changed from natural to secondary forest, agroforestry systems, and annual cultures. Despite close proximity to near-primary forest, the agroforestry systems studied supported only a few small frugivorous-nectarivorous species. Our results suggest that secondary forest could play an important role in the conservation of many Sulawesi bird species, but, although suitable for colonization, its potential to sustain populations over the long term is unknown. Improvement of the landscape matrix for biodiversity conservation through secondary habitats therefore seems desirable to enlarge the ranges of forest species, but the fight against land conversion within protected areas of the region should be of much higher importance, at least at present.Resumen: Aun hay debate sobre el valor potencial de los sistemas de uso de suelo para el mantenimiento de la biodiversidad tropical. Un creciente número de estudios indica que mucha de la biodiversidad del bosque también se encuentra en la matriz agropecuaria del paisaje. Debido a que hay poca información sobre el valor potencial de los sistemas de uso de suelo para especies de aves de bosque tropical, realizamos conteos por puntos repetidos en bosque casi primario, bosque secundario joven adyacente, sistemas agroforestales modernizados de cacao y cultivos anuales en elevaciones sub-montanas en Sulawesi central, Indonesia. La riqueza de especies decreció del bosque natural y bosque secundario joven a los sistemas agroforestales y cultivos anuales. No obstante la cercanía de bosque casi primario, los sistemas agroforestales estudiados soportaron solo a unas cuantas especies frugívoro-nectarívoras. Nuestros resultados sugieren que el bosque secundario podría jugar un papel importante en la conservación de muchas especies de aves de Sulawesi pero, aunque adecuado para colonización, se desconoce su potencial para sostener poblaciones a largo plazo. Por tanto, parece deseable un mejoramiento de la matriz del paisaje para la conservación de biodiversidad mediante hábitats secundarios para ampliar los rangos de las especies forestales, pero la lucha contra la conversión de suelos dentro deáreas pr...

Analyses of leopard Panthera pardus prey choice reveal a strong preference for species weighing 10-40 kg. In the Congo Basin rainforests, species within this weight range are also targeted by bushmeat hunters, potentially leading to exploitative competition between leopards and hunters. We investigated leopard prey choice along a gradient of human disturbance, hypothesizing that leopards will exploit smaller prey where competition is strong, possibly resulting in reduced leopard densities at highly hunted sites. We determined leopard diet by means of scat analysis at four rainforest sites in central Gabon, which varied according to their distance from human settlements. Camera trap data collected at each of the four study sites revealed that human hunting intensity increased with proximity to settlements, while the abundance of potential leopard prey species decreased. We found no evidence of leopards at the site nearest to settlements. At the remaining sites, the number of scats collected, mean leopard prey weight and the proportion of large prey (>20 kg) in leopard diet increased with distance from settlements. Camera trap data demonstrated that leopard population density increased with distance from settlements, from 2.7 ± 0.94 leopards/100 km2 to 12.1 ± 5.11 leopards/100 km2. Our results document an increasing use of smaller prey species and a decrease in leopard density in proximity to settlements, supporting our hypothesis. Comparison of leopard diet with hunter return data revealed a high dietary niche overlap between leopards and hunters at sites situated at similar distances from settlements. Our results suggest that bushmeat hunting may precipitate the decline in leopard numbers through exploitative competition and that intensively hunted areas are unlikely to support resident leopard populations. Conserving the leopard in the Congo Basin will rely on effective protected areas and alternative land management strategies that promote regulated human hunting of leopard prey

Protected forest areas of Sulawesi are gradually being replaced by intensively used agroforestry systems and farmland, especially in lowland and sub-montane regions. Studies on the impact of these man-induced changes on biodiversity are of urgent conservation concern. We compared the fruitfeeding butterfly assemblage of a natural hill forest to that of a disturbed hill forest, representing a mosaic of old secondary forest and recently abandoned or active subsistence farms. Overall, species richness seemed highest in the disturbed site, but both abundance and diversity of endemic butterflies were significantly higher in the natural forest. Although the butterfly assemblage showed a clear vertical structure in the natural forest, vertical stratification was no longer pronounced at the disturbed site. Comparative studies based on diversity estimates from ground samples should consider not only the scale at which sampling is carried out and influences from nearby habitat patches in the surrounding landscape mosaic, but also possible behavioural changes in stratified species after forest modification. This study shows that higher overall species richness does not imply higher species distinctiveness, and indicates that the contribution of land-use systems to global biodiversity should be evaluated with caution, even when relatively high species richness estimates are found.

Vegetation surveys were carried out at 24 sampling stations distributed over four land use types, namely near-primary forest, secondary forest, agroforestry systems and annual crop lands in the northeastern part of the Korup region, Cameroon, to assess the impact of forest conversion on trees and understorey plants. Tree species richness decreased significantly with increasing level of habitat modification, being highest and almost equal in secondary and nearprimary forests. Understorey plant species richness was significantly higher in annual crop lands than in other land use types. The four land use types differed in tree and understorey plant species composition, the difference being smaller among natural forests. Tree and understorey plant density differed significantly between habitat types. Density was strongly correlated with species richness, both for trees and understorey plants. Five tree and 15 understorey plant species showed significant responses to habitat. A 90% average drop in tree basal area from forest to farmland was registered. Our findings support the view that agroforestry systems with natural shade trees can serve to protect many forest species, but that especially annual crop lands could be redesigned to improve biodiversity conservation in agricultural landscapes of tropical rainforest regions.

No abstract

No abstract