Records of 25,298 diurnal raptors observed in 1762 spot counts and 55,577 k
m of roadside counts in Botswana during 1991-1995 were analysed as a functi
on of distance from the boundary of conservation areas. Spot counts were mo
re sensitive and efficient than roadside counts. Overall there was a dramat
ic decline in raptor abundance once > 15 km outside conservation areas, mor
e so in the dry than in the wet season. Different groups had different patt
erns: numbers of large resident eagles dropped on land > 50 km from protect
ed areas to < 10% of numbers in the core of protected areas. Numbers in the
peripheral 30 bm inside protected areas were already > 40% lower than in t
he core, indicating a massive edge effect on populations inside reserves. M
igrants were less sensitive; the Wahlberg's (Aquila wahlbergi) and steppe e
agle (A. nipalensis), which one would expect to belong to a sensitive group
because of their large size, were indifferent to land use factors. The lat
ter species had a pattern strikingly different from the resident tawny eagl
e (A. rapax). Numbers of scavenging vultures peaked near the interface betw
een conservation areas and unprotected land, but the white-headed vulture (
Trigonoceps occipitalis) followed the pattern of the large eagles. This app
ears to be the single most sensitive species, occurring anywhere > 15 km ou
tside protected areas at < 10% of its capacity. Species with a preference f
or human-modified habitat were more abundant outside conservation areas, as
expected. The widespread depletion of biomass and biodiversity (including
potential prey) following from structural changes to vegetation caused by o
vergrazing by livestock seems to be the main force behind the poor status o
f raptors on unprotected land. Large buffer zones with special management a
re needed outside parks to ensure the integrity of conservation areas. (C)
2000 Elsevier Science Ltd. All rights reserved.