How sampling effort affects biodiversity measures in an arid succulent karoo biodiversity hotspot

Understanding the implications of sampling effort on biodiversity measures is important where conservation planning takes place in areas with patchy collecting coverage. This study analysed plant diversity data (i) to explore the effect of sampling effort on biodiversity measures at a landscape level and (ii) to test the usefulness of perennial plant data as indicators of biodiversity trends and patterns. Using published sources, databases and own field surveys, plant distribution data from ten grids (15-min intervals of a latitude-longitude grid) in the Sperrgebiet in the Succulent Karoo Biome in south-west Namibia, which were surveyed repeatedly, were analysed. Species-survey curves, simple biodiversity measures and ordination were used to test the effect of sampling intensity on biodiversity measures and floristic relationships. With well over 350 plant species, grids that contained mountainous terrain emerged as the most species-rich. Based on species-survey curves, only two grids were well collected. Floristic relationships showed different patterns with incomplete data, but patterns of perennials were almost identical to those with near complete data. Sampling effort affected community measures more than broad patterns of species richness and diversity in the arid succulent karoo. In reasonably species-rich areas, even if there was only medium collecting status, trends such as the ranking of poor to rich areas were maintained. This also holds for using perennials or endemics as indicators. Floristic relationships showed higher sensitivity to the completeness of sampling. Thus, while species-rich areas (biodiversity hotspots) are also likely to be identified with incomplete data, floristic relationships may be wrongly depicted, when incomplete data are used. Repeated samplings, even if only the perennial components of the vegetation can be recorded, yields more representative results. Perennial plants thus provided a suitable indicator for floristic relationships at the landscape level in the arid succulent karoo.