EFFECTS OF EXCLUDING GOAT HERBIVORY ON ACACIA-TORTILIS WOODLAND AROUND PASTORALIST SETTLEMENTS IN NORTHWEST KENYA

Browsing by goats is considered to cause poor tree regeneration and re duced tree growth around settlements throughout the arid zones of sub- Saharan Africa. This study investigated whether excluding goats from A cacia tortilis woodlands increased tree regeneration, current season's shoot growth rates and browse production over a period of 52 months b etween 1986 and 1990. The study also investigated the effects of clima tic variability on tree growth and browse production. Excluding goal h erbivory provided no advantage over continuous browsing for juvenile A . tortilis. Trees on the unbrowsed and on browsed transects increased by 22.2 (standard error [SE] +/- 0.53) cm.yr(-1) and 25.0 (SE +/- 0.58 ) cm.yr(-1) respectively. Fewer but longer shoots were produced by tre es on the unbrowsed transects, while trees on the browsed transects in vested more in shorter shoots. Net total browse production was lower o n unbrowsed (1.73 [standard deviation (SD) +/- 4.3] t.ha(-1).yr(-1)) t han on the browsed (3.03 [SD +/- 3.6] t.ha(-1).yr(-1)) transects. Biom ass production on unbrowsed and browsed transects was closely correlat ed with rainfall and presumably soil moisture during wet seasons. Rela tive growth rates (RGR) of current season's shoots in the two treatmen ts did not differ, implying goat herbivory at moderate stocking densit y (i.e. 13.0 tropical livestock units [TLU] .km(-2)) stimulated shoot growth. RGR remained positive except on the browsed transects during 1 990, a dry year. Goat browsing pressure was moderate. Total biomass lo ss on unbrowsed transects was 15.5 %.yr(-1) compared with 27.7 %.yr(-1 ) on the browsed transects. These findings do not support the notion t hat goats always destroy young tries around settlements. Goat herbivor y at moderate intensity stimulated shoot productivity. However, the re sults should not be used to generalize all conditions throughout sub-S aharan Africa, let alone the arid zones of northern Kenya. Rather, the re is a need to emphasize individual case studies that ultimately can be used for managing degraded woodlands near pastoralist settlements. (C) Elsevier, Paris.