This paper documents a continuous ∼ 44,000-yr pollen record derived from the Mfabeni Peatland on the Maputaland Coastal Plain. A detailed fossil pollen analysis indicates the existence of extensive Podocarpus-abundant coastal forests before ∼ 33,000 cal yr BP. The onset of wetter local conditions after this time is inferred from forest retreat and the development of swampy conditions. Conditions during the last glacial maximum (∼ 21,000 cal yr BP) are inferred to have been colder and drier than the present, as evidenced by forest retreat and replacement of swampy reed/sedge communities by dry grassland. Forest growth and expansion during the Holocene Altithermal (∼ 8000–6000 cal yr BP) indicates warm, relatively moist conditions. Previous records from Maputaland have suggested a northward migration of Podocarpus forest during the late Holocene. However, we interpret a mid-Holocene decline in Podocarpus at Mfabeni as evidence of deforestation. Forest clearance during the mid-Holocene is supported by the appearance of Morella serrata, suggesting a shift towards more open grassland/savanna, possibly due to burning. These signals of human impact are coupled with an increase in Acacia, indicative of the development of secondary forest and hence disturbance.

The aim of this study was to determine the minimum conservation area needed to conserve vegetation types and their landscape and to apply it to an area in KwaZulu-Natal, South Africa, which is within the Maputaland Centre of Plant Endemism and part of the Maputaland–Pondoland–Albany biodiversity hotspot. Outside conservation areas this Centre of Plant Endemism is under threat from human utilization. We used a method initially designed to determine minimum conservation areas for rare plant species, which we adapted from its original country and context, to determine minimum conservation areas for landscape species in Maputaland’s little-documented environment. The minimum area required for conservation was established for the Sand Forest and Woodland vegetation types in the region. We found that sufficient habitat is presently conserved to preserve the Sand Forest but not the Woodlands. The method holds promise to provide answers to critical conservation issues in lesser-known environments and, although relatively difficult to establish for the first time, is an efficient and easy to use tool that can be refined once more knowledge becomes available.

Systematic conservation planning is widely considered the most effective approach for designing protected area and other ecological networks. However, many conservation practitioners still ignore these methods and we suggest that five perceived limitations of this process are affecting its uptake. These perceptions are that (1) systematic conservation planning software is difficult to use, (2) the process requires extensive biodiversity distribution data, (3) setting targets for representing conservation features is not possible, (4) the advantages of systematic conservation planning do not outweigh the costs, and (5) the resulting plans often identify unsuitable areas. Here we review these perceived limitations and argue they are all misplaced, although we recognize difficulties in the target setting process. We then illustrate the value of systematic conservation planning to practitioners using a case study that describes a low-cost exercise from Maputaland, South Africa. This preliminary conservation assessment measured the effectiveness of the existing reserve system and identified a number of candidate areas that could be the focus of community- or privately-run ecotourism or game ranching ventures. Our results also emphasize both the importance of producing planning outputs that are specifically targeted for stakeholders, and the role of systematic conservation planning in providing a framework for integrating different provincial, national and transnational conservation initiatives.