Abstract Background Shrub encroachment (SE) drives rangeland degradation, threatening biodiversity and forage productivity. The reversal of the catastrophic effects of SE depends on the efficacy of restoration efforts. We evaluated the potential of manual removal of the shrub Pteronia incana (Burm.f.) DC. with or without reseeding mixtures of grasses ( Panicum maximum Jacq. and Eragrostis curvula Schrad. Nees) on rangeland restoration at Peddie, Eastern Cape Province, South Africa. Methods The treatments were (1) uprooting and reseeding (UPR), (2) uprooting only (UPO), (3) cutting and reseeding (CR), (4) cutting only (CO) and (5) untreated control (CTL), each replicated three times in North and South aspects. Results Graminoid cover was higher in all treatments than in CTL, with CR showing relatively higher cover in North and South slopes. Restoration treatments increased plant density, richness and diversity compared to CTL. Slope aspect, treatment and seeded species interacted significantly on biomass production. CO in the North aspect attained two‐fold higher biomass production of nonreseeded grasses than other treatments. Responses of reseeded species varied by aspect, with P. maximum attaining two‐fold higher BP in CR and UPR in the North than the South slope, whereas E. curvula had higher biomass production in CR in the South than the North slope. Conclusions Shrub clearing facilitated restoration of diversity and plant densities regardless of the restoration method. However, biomass production and vegetation cover depended on the interaction of the shrub clearing method, reseeding and slope aspect. Thus, the decision on restoration of herbaceous vegetation should be informed by the vegetation attribute of restoration priority.
Makanya et al. (Wed,) studied this question.