The effect of different fire regimes on plant diversity in southern African grasslands
Introduction
Fire is an important component in the maintenance of grassland diversity (Freeman, 1998; Lunt and Morgan, 2002), but its effects in southern African grasslands are poorly understood. This is because research in these grasslands has focussed on the few dominant grass species that are important for livestock production. While grasses generally constitute most of the plant biomass in grasslands around the world, the diversity of these systems lies in their forb taxa (e.g., Freeman, 1998; Lunt and Morgan, 2002). As forbs have rarely been included in studies of disturbance effects in southern African grasslands, we have a poor understanding of how these grasslands should be managed to maintain and promote their diversity. Fire is a central component in this management and we therefore need to understand the effect that different burning regimes are having on this diversity for it to be conserved.
Theory suggests that species richness should be highest at intermediate levels of disturbance (Connell, 1978; Huston, 1979). However, this is not always the case. In the Scottish Highlands, plant diversity was found to increase with fire frequency (Hobbs et al., 1984) while, conversely, in the North American prairies, plant diversity decreased with increasing fire frequency (Collins et al., 1995). In the Mediterranean-climate shrublands of South Africa (fynbos), plant diversity response to fire frequency was found to be scale dependent (Schwilk et al., 1997). In the smallest fynbos plots sampled (1 m2), the pattern matched that of the Scottish heathlands, where sampling was also performed using 1 m2 quadrats (Hobbs et al., 1984). However, at larger scales (100–1000 m2), the Intermediate Disturbance Hypothesis was directly contradicted, with species richness being lowest at the intermediate burning frequency. Further confounding this situation, Australian temperate grasslands showed no clear trends in plant diversity in relation to fire frequency (Lunt and Morgan, 2002). With this array of conflicting results it becomes obvious that local studies are required for us to understand how fire influences plant diversity in a particular system. Observations from two southern African fire trials seem to mirror the Australian situation, but these results were from a short-term study (Robinson et al., 1979) and casual observation (Le Roux, 1989). A more detailed study was therefore necessary to provide insight into the effects that different burning regimes have on the plant diversity of South African grasslands. We attempt this here at different spatial scales of diversity by sampling long-term fire trials located in three different grassland types.
Section snippets
Grasslands of southern Africa
Southern African grasslands are dominated by C4 grass species (which have a greater efficiency at higher temperatures and light intensities than C3 taxa (Bjorkman, 1976)), characterised by Themeda triandra, which is both abundant and widely spread across the biome. From the eastern seaboard, across the escarpment, into the interior, the biome encompasses a range of short to tall grasslands, spanning broad gradients of rainfall (∼400 to ∼1600 mm yr−1), temperature (frost-free to snow bound in
Alpha-diversity and small plot species richness
The importance-score method (Outhred, 1984; Morrison et al., 1995) was selected for sampling alpha-diversity in the small plot trials as it provides measures of both species richness and abundance, and was compatible with the size of the plots. We used a semi-geometric series for the nested concentric sub-quadrats (1, 2, 5, 10, 20, 50 and 100 m2) based on the smallest plots (10 m × 10 m at the semi-arid site) of the three trials. The method works by awarding scores according to the smallest of
Species richness at the small plot scale
The effects of frequency and season of burn on species richness were generally consistent at both scales (1 and 100 m2) sampled at each site, but not between sites (Fig. 1). Levels of species richness were highest in the mesic grasslands (141 taxa recorded from 27 × 100 m2 quadrats) followed by the montane sites (89 taxa recorded from 27 × 100 m2 quadrats) and semi-arid sites (70 taxa recorded from 18 × 100 m2 quadrats) (Table 1). Forbs (mostly dicots) dominated the composition at each site, with 5.1
Effects of burning on alpha- and beta-diversity
Compared at two scales, across three trials, and a range of treatments, species richness in the small plots showed no consistent, clear responses to either frequency or season of burn. This study concurred with two previous southern African studies (Robinson et al., 1979; Le Roux, 1989) and therefore showed that burning in southern African grasslands does not support the Intermediate Disturbance Hypothesis (Connell, 1978; Huston, 1979). Instead, plant diversity in southern African grasslands
Acknowledgements
We thank the following organisations and their staff for facilitating this study by maintaining and allowing us access to the trials: the University of Natal, School of Applied and Environmental Sciences (Mesic site); Ezemvelo KZN Wildlife (Montane site); and Prof. Winston Trollope at the University of Fort Hare’s Department of Livestock and Forage Management (Semi-arid site). We also thank Craig Morris for his assistance with the ordination. The project was funded by the National Research
References (26)
- et al.
Effects of intensity and height of shading on the tiller initiation of six grass species from the Highland Sourveld of Natal
South African Journal of Botany
(1988) - Anon, 1993. Releve Manager 1.08 program, Cornell Ecological...
- Arnold, T.H., deWet, B.C.T., 1993. Plants of southern Africa: names and distributions. Memoirs of the Botanical survey...
Adaptive and genetic aspects of C4 photosynthesis
- et al.
Experimental analysis of intermediate disturbance and initial floristic composition: decoupling cause and effect
Ecology
(1995) Diversity in tropical rain forests and coral reefs
Science
(1978)The flora of Konza Prairie: a historical review and contemporary patterns
- et al.
Studies on fire in Scottish heathland communities: III. Vital attributes of the species
Journal of Ecology
(1984) A general hypothesis of species diversity
The American Naturalist
(1979)Die invloed van brand en maai op onbeweide lang grasveld in Natal
Journal of the Grassland Society of South Africa
(1989)