Ecological engineering through fire‐herbivory feedbacks drives the formation of savanna grazing lawns. Issue 1 (25th July 2017)

• Record Type: Journal Article
• Authors: Donaldson, Jason E.; Archibald, Sally; Govender, Navashni; Pollard, Drew; Luhdo, Zoë; Parr, Catherine L.
• Editors: Schultz, Cheryl
• Is Part Of: Journal of applied ecology. Volume 55:Issue 1(2018)
• Page Start: 225
• Page End: 235
• Publication Date: 2017-07-25
• Subjects: African savannas -- burn area -- burning -- fire ecology -- grass system‐state -- grazer ecology -- grazing -- grazing lawns -- herbivore; Agriculture -- Periodicals; Biology, Economic -- Periodicals; Agricultural ecology -- Periodicals; Applied ecology -- Periodicals; 577
• DOI: 10.1111/1365-2664.12956
• Languages: English
• ISSNs: 0021-8901

Abstract

Abstract: Variation in grass height is beneficial to biodiversity conservation in savanna landscapes. Theory predicts that small fires can promote short‐grass areas within savannas. We experimentally assessed the influence of fire season and size on grass height and the resultant response of wild grazer communities and tested three hypotheses: (1) repeated small fires in tall‐grass savannas increase short‐grass grazer densities in the post‐burn environment; (2) increased grazer densities maintain grass height in a short, palatable state and drive feedbacks that exclude fire; and (3) late‐dry season burns concentrate grazers more effectively than early dry season burns. We repeatedly applied annual treatments (unburned, early‐ and late‐burns) in 0.25‐, 5‐ and 25‐ha plots over a period of 3 years in a tall‐grass savanna system in Kruger National Park, South Africa. Dung counts for grazer density and grass height data were collected along 50‐m transects. Grass height was measured in paired 1‐m 2 herbivore exclosures on plots before and after applied fires. Dung data indicate that wildebeest occurred most frequently in grass heights below 5 cm. Their preference for plots regardless of fire size or season increased over time with each repeated burn. Zebra and buffalo favoured burns immediately post‐fire, but buffalo did not actively select for burnt areas over longer time periods. By the second year of treatment, herbivory maintained 28% and 91% of the grass height below 10 cm in the early‐ and late‐season burns respectively. In contrast, herbivory on the unburned treatments had no effect on grass height. Synthesis and applications . Fires less than 25 ha in size attracted sufficient grazing herbivores to shorten grass height. Repetition of the fire treatments resulted in the active selection of these areas in the longer term by wildebeest, impala and, to a lesser degree, zebra. Grazing pressure was high enough to initiate positive feedbacks and maintain lawns after only two seasons of burning and, depending on the season of burn, reduced grass height to a level that excluded repeat fires. Our study demonstrated that theory on grazer use of the post‐fire environment can be implemented practically by applying small repeated burns to promote the formation of short‐grass areas within savannas. Abstract : Fires less than 25 ha in size attracted sufficient grazing herbivores to shorten grass height. Repetition of the fire treatments resulted in the active selection of these areas in the longer term by wildebeest, impala and, to a lesser degree, zebra. Grazing pressure was high enough to initiate positive feedbacks and maintain lawns after only two seasons of burning and, depending on the season of burn, reduced grass height to a level that excluded repeat fires. Our study demonstrated that theory on grazer use of the post‐fire environment can be implemented practically by applying small repeated burns to promote the formation of short‐grass areas within savannas.