Long-term compositional responses of a South African mesic grassland to burning and mowing

Abstract

Question: What is the long-term compositional response of grass and forb species to various combinations of burning and mowing? Can these responses be predicted from simple plant traits? Location: Ukulinga research and training farm (24°24'E, 30°24'S), Pietermaritzburg, KwaZulu-Natal, South Africa. Methods: Grass species relative abundance in 1996 in various burning and mowing treatments of a long-term (> 50 a) experiment was calculated from data obtained using a point sampling method, whereas forb species abundance in 1999 was determined using the importance score method. The experiment consisted of different frequencies (annual, biennial and triennial) of burning and mowing in winter or spring in combination with different frequencies of summer mowing (none, early, late or both). Results: Grasses responded to the type of disturbance (burn or mow) and frequency of burning, whereas forbs responded primarily to the presence or absence of any form of disturbance and secondarily to the timing of burning (spring versus winter). Summer mowing and annual or biennial dormantperiod burning maintained communities dominated by short grasses, whereas tall grasses dominated under annual dormant-period mowing, triennial burning or protection from disturbance. Grass tillering strategy (below- or above-ground) influenced response to burning frequency. Many erect herbaceous dicot species with aerial meristems were reduced in abundance by summer mowing whereas most small creeping herbaceous dicot species appeared to be dependent upon summer mowing. Conclusions: This long-term experiment demonstrated that: (1) grasses and forbs responded differently to burning and mowing; (2) simple plant traits, such as height, position of tiller initiation, and position of meristems have potential for predicting the response of species to the timing and frequency of burning and mowing.