TY - JOUR
T1 - Avifaunal collapse in West African forest fragments
AU - Beier, Paul
AU - Van Drielen, Maryann
AU - Kankam, Bright O.
N1 - Copyright:
Copyright 2008 Elsevier B.V., All rights reserved.
PY - 2002
Y1 - 2002
N2 - Despite the fact that West African tropical forests are the most fragmented in Africa, there has been no published research on biodiversity in these fragments. To determine how forest birds respond to five factors - patch size, patch isolation, canopy density, abundance of large trees, and proximity to forest edge - in these forests, we surveyed 60 species of forest birds on 121 transects (2.5 ha each) in 35 forest fragments in the semideciduous forest zone of Ghana. Species richness per transect increased with patch size over the entire range of patch sizes observed (3-33,000 ha). The diversity of forest birds (22 species) on a single transect in a large forest patch was similar to the cumulative diversity (25 species) on all 17 transects in 13 small patches. Twenty-two of 60 species were area sensitive, 15 of which were never found in small patches. These results suggest that only large forests will conserve many species of West African forest birds. Nine species were edge sensitive, 7 of which were also area sensitive. However, forest structure near patch edges was not consistent with bird responses to canopy and tree density, suggesting that mechanisms other than microclimate or structural changes (perhaps predation or nest parasitism) underlie the response of most species that are sensitive to both edge and area. Regression of critical patch size (the smallest patch size in which a species was detected) on logarithm of body mass (an index of home range size) for 22 area-sensitive species suggests that area-sensitive species are unlikely to occur in patches smaller than several home-range areas. Canopy density influenced 13 species (11 positively, 2 negatively), and abundance of large trees influenced 8 species (3 positively, 5 negatively). Forest birds did not respond to isolation (distance from a patch to a large forest) for isolation distances of 1-25 km, suggesting that island biogeographic mechanisms had less influence on birds than other potential mechanisms of area sensitivity. Although small patches contributed little to the conservation of forest birds - species found in small patches were well represented in large patches - small patches are probably important for supporting generalist bird species that provide ecological services in the agricultural matrix and serving as nuclei for future ecosystem recovery.
AB - Despite the fact that West African tropical forests are the most fragmented in Africa, there has been no published research on biodiversity in these fragments. To determine how forest birds respond to five factors - patch size, patch isolation, canopy density, abundance of large trees, and proximity to forest edge - in these forests, we surveyed 60 species of forest birds on 121 transects (2.5 ha each) in 35 forest fragments in the semideciduous forest zone of Ghana. Species richness per transect increased with patch size over the entire range of patch sizes observed (3-33,000 ha). The diversity of forest birds (22 species) on a single transect in a large forest patch was similar to the cumulative diversity (25 species) on all 17 transects in 13 small patches. Twenty-two of 60 species were area sensitive, 15 of which were never found in small patches. These results suggest that only large forests will conserve many species of West African forest birds. Nine species were edge sensitive, 7 of which were also area sensitive. However, forest structure near patch edges was not consistent with bird responses to canopy and tree density, suggesting that mechanisms other than microclimate or structural changes (perhaps predation or nest parasitism) underlie the response of most species that are sensitive to both edge and area. Regression of critical patch size (the smallest patch size in which a species was detected) on logarithm of body mass (an index of home range size) for 22 area-sensitive species suggests that area-sensitive species are unlikely to occur in patches smaller than several home-range areas. Canopy density influenced 13 species (11 positively, 2 negatively), and abundance of large trees influenced 8 species (3 positively, 5 negatively). Forest birds did not respond to isolation (distance from a patch to a large forest) for isolation distances of 1-25 km, suggesting that island biogeographic mechanisms had less influence on birds than other potential mechanisms of area sensitivity. Although small patches contributed little to the conservation of forest birds - species found in small patches were well represented in large patches - small patches are probably important for supporting generalist bird species that provide ecological services in the agricultural matrix and serving as nuclei for future ecosystem recovery.
UR - http://www.scopus.com/inward/record.url?scp=0036324155&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0036324155&partnerID=8YFLogxK
U2 - 10.1046/j.1523-1739.2002.01003.x
DO - 10.1046/j.1523-1739.2002.01003.x
M3 - Article
AN - SCOPUS:0036324155
SN - 0888-8892
VL - 16
SP - 1097
EP - 1111
JO - Conservation Biology
JF - Conservation Biology
IS - 4
ER -