1. The relationship between biomass and species richness in temperate veget
ation has been described as a hump-back response model. The hump-back model
predicts that herbaceous species richness is highest at an intermediate le
vel of biomass. However, this has not been investigated in arid-zone grazin
g lands.
2. We tested the hump-back prediction in an and tropical grazing region in
northern Kenya where a seasonal grazing exclosure system is practised. We c
ompared vegetation structure, species richness and composition on an open r
ange and exclosures at five sites to elucidate the potential mechanisms beh
ind variation in species richness.
3. More biomass was accumulated within seasonal exclosures than in continuo
usly grazed areas. Species richness in exclosure plots varied from 5.3 to 8
.3 species In 2, while that in open plots varied from 5.1 to 7.5 species m(
-2). A pair-wise test showed no difference in two of the five sites with re
spect to both total and herbaceous species richness.
4. The primary floristic gradient illuminated through ordination was relate
d to biomass, while the secondary gradient was related to species richness.
The exclosure plots had more abundant species, especially compared with op
en plots, which had more rare and occasional species. A total of 37 herbace
ous species was recorded; 22% were indifferent to grazing, 30% grazing into
lerant and 48% promoted by grazing.
5. The relationship between biomass and herbaceous species richness showed
(i) no trend within the exclosures (maximum biomass 800 g m(-2)); (ii) a po
sitive trend in the open grazing land (maximum of 500 g m(-2)); and a hump-
back pattern when (i) and (ii)were analysed together. Optimum richness corr
esponded to a biomass level of 400-500 g m(-2). Species richness declined w
ith increase in age of exclosures.
6. We confirmed that species richness will decline when biomass exceeds 500
g m(-2), as predicted by the hump-back model, even in and grazing lands. S
easonal grazing exclosures may increase species richness to a certain level
, but the decline in species richness with age of exclosures indicates that
long-term exclusion of grazing may not necessarily increase species richne
ss in arid-zone grazing lands.