Ant diversity and composition in a reforested landscape of Buffelsdraai Landfill Conservancy, KwaZulu-Natal.

Abstract

Restoration of degraded and reclaimed landscapes provide a useful framework to evaluate the recovery of biodiversity loss. A reforestation project was initiated in 2008 by eThekwini Municipality in Buffelsdraai Landfill Conservancy, aiming to offset carbon emissions over a 20-year period and increase climate change adaptation through biodiversity and ecosystem services restoration. The project offered an opportunity to evaluate to what extent reforestation for carbon sequestration can have co-benefits for biodiversity. The current study monitors the recovery of habitat restoration practices (planting of indigenous forest trees) in Buffelsdraai Landfill Conservancy, eThekwini Municipality, KwaZulu-Natal Province, in South Africa. The main aim of the study was to evaluate how biodiversity recovers following forest restoration. The study used ants (Formicidae: Hymenoptera) as a model organism as they comprise a significant component of invertebrate diversity and a keystone taxon in the terrestrial ecosystems. The study objectives were to provide ant checklist in a reforested landscape and to describe ant diversity patterns along a gradient of restoration and to identify the environmental variables which drive the diversity patterns along a reforestation gradient. Using a standardized pitfall survey, ants were sampled across eight sites, each replicated four times, which included sugarcane (unrestored), grassland and scarp forest (natural reference sites), short-term (0-2 year), medium-term (3-5 years) and long-term (6-8 years) restored sites. Ant sampling was conducted in April-May 2017 (early dry season) and December 2017 (wet season). Environmental (habitat structure) and soil surveys were conducted at each plot. A total of 27 439 ant specimens comprising of 96 species in 31 genera, and six subfamilies were collected. Sample coverage estimator was larger than 0.97, indicating that inventory completion approximated most of the ant assemblages found in the study area. Myrmicinae, Ponerinae and Formicinae were the most abundant and species-rich subfamilies, with Tetramorium, Pheidole and Monomorium as the most species-rich genera. The most numerically dominant species were Pheidole megacephala species group and Anoplolepis custodiens. Ant species richness and activities were significantly highest in the restored and grassland sites and low in forest site, and lowest in unrestored sugarcane. Species richness responded with a hump-shaped response as patterns of species richness significantly decreased with increasing bare-ground cover. High species diversity and composition was associated with open habitats with grass layer. Forest had the most distinct assemblages. Leaf litter, vegetation structure, canopy cover and bare-ground cover, were the four predictor variables which had major influences on ant assemblage structure. Four forest indicator taxa were identified (Pheidole UKZN_11 (megacephala gp.), Tetramorium UKZN_04 (squaminode gp.); Tetramorium UKZN_28 (setigerum gp.) and Leptogenys attenuate), and one indicator for grassland (Lepisiota capensis). No indicators were found for sugarcane sites. Solenopsis UKZN_01 and Pheidole UKZN_09 were potential indicator for restored sites. The restoration sites were transitioning from sugarcane plantation, and were drawing most of their colonisation from grasslands at this stage. This study shows that open woodlands are ideal habitats for maximising species diversity, as they provide a complex habitat for many species, and the availability of local natural grassland as a source of invertebrates assists restoring functioning, even if we expect the community to transition to forest species as regrowth progresses.