|dc.description.abstract||Grasslands cover about a quarter of the Earth and are often characterised by wetlands (McAllister, 2000). A large diversity of animal life is found there. In South Africa, there are 21 species of grassland birds, of which I2 species are endemic (Anon, 2000). About 20% of South Africa's threatened land mammals are found in grasslands (Anon, 2000; Anon, 2003b; McAllister, 2000). In South Africa the grassland region is of great economic importance, since it is ideal for farming (Le Roux, 2002). Whether controlled or not, fire has an effect on the environment. Destruction of vegetation can affect animals, as there is little place to hide or perch on (MacArthur, 1958; Kozlowski & Ahlgren, 1974). Birds that feed on seeds may also be affected until vegetation produces seed again (Kozlowski & Ahlgren, 1974). Fire is such a part of the environment that it
is regarded as a natural factor. Fire is thought to play an important role in plant communities (Tainton & Mentis, 1984). According to evidence, early humans modified the natural fire regime to suit their own purposes and by the eighteenth century veldt burning occurred regularly to stimulate new growth in plants to improve grazing and hunting (Edwards, 1984). Birds adapted to their environment and became one of the planet's most successful creatures with many examples of their adaptation to their environment. Through hunting. habitat
destruction, pollution and introduction of exotic species, man threatens about 1000 bird species globally (Davies, 2003). Distribution may be related to environmental change, making the birds, in many instances, a good indicator of environmental change (Morrison, 1986). Mills (2004)explained that changes in the savannah could be predicted based on the alteration of habitat. Species that feed on the ground may be attracted to a fire whereas those that prefer the cover of
vegetation would stay in the control sites (Mills, 2004). According to Parr and Chown (2003), most of the research on birds and fire in South Africa has been done at species level and only the responses of a few species to fire have been studied. Barberspan is a bird sanctuary in the North-West Province of South Africa. It is situated in an area of seasonally filled pans, which makes this perennial pan very important to waterfowl
and conservation (Beyer & Duggan. 1997; De Beer, 2001). Barberspan was declared a Ramsar site in 1975 according to the Ramsar Convention of 1971 (Dini, 1999). 1 decided on using controlled bums at Barberspan to achieve the aims of this study. A point count method (Bibby, Burgess, Hill & Mustoe, 2000) was used at each of the three burns. Counts were taken on a regular basis or as predetermined by the observer for two years at Burn 1 and one year for Burns 2 and 3. Birds that were seen or heard were counted, as not all birds may be visible to the observer (Bibby, eta!., 2000). Birds hiding in the grass were identified according to their calls. Birds that flew over were ignored. If the birds perched in the grass, on the ground, or utilised the site in some other way, they were counted. The computer programmes Excel 2000, Primer 5 and Mapviewer 5 were used to analyse and present the data. The densities increased after the fire and started to decrease at the end of the year. The
number of species increased after the fire and returned to pre-fire levels at the end of the year. The species that normally used the habitat (fire-sensitive), left after the fire, and other species(fire-colonizers) colonized the area. Some of the original species remained in the area after the fire (site-tenacious). The species that had left after the fire, started to return at the end of the year. There was little change in the diversity, as measured by the Shannon index. The densities of the birds. the species composition and number of species changed, but this was not reflected in the Shannon index. Therefore, the Shannon index is not be a good measure to quantify the effect of fire on birds, at least in cases such as at Barberspan. The biomass increased after the fire, because larger birds were attracted to the fire. Smaller species were also attracted to the Burn Sites, but in much larger numbers than the Control Sites. The size of the burned areas had an effect on the birds. More birds per hectare were attracted to the larger fires (Burns 1 and 3), when compared with the smaller fire at Burn 2. The guilds showed that the birds were probably attracted to the fire because of the availability of food rather than habitat or breeding opportunities. Therefore, optimal foraging probably played a significant role in explaining the differences found between larger and smaller fires. Some of the larger birds only took the opportunity to feed on the prey killed immediately after the fire, and left after a day or two. The island effect was also seen in the Burned Sites. The burned areas could be considered as "islands"
in the "sea" of grassland. The larger islands had a higher density in birds, greater species richness and more total avian biomass. The fires probably had an effect over an area larger than the burned areas itself. There were no apparent increases in density or species richness in the Control Sites
after the fire, and therefore the birds that left had to move out further. The number of birds and number of species in the Control Sites did not decrease at the same rate as the increases seen at the Burn Sites, showing that the birds probably were attracted from a larger area than that immediately surrounding the burned areas.||