Pollen productivity estimates from KwaZulu-Natal Drakensberg, South Africa

Abstract

Pollen assemblages from sedimentary sequences reflect the landscapes surrounding the basin at the time of deposition, however the translation of pollen records into estimates or maps of past landcover is complex. Models of the pollen-vegetation relationship can help, but need to be calibrated first against modern data on pollen and vegetation for the pollen types of interest. This study presents initial estimates of Relative Pollen Productivity in the KwaZulu-Natal Drakensberg, South Africa, a palaeoecologically significant region of southern Africa, using the Extended R-Value approach. Modern pollen spectra from soil surface samples were extracted from three dominant vegetation communities, and vegetation data were collected around each sample point using a 3-tiered ring surveying approach. Fall speeds of key taxa were calculated using Stokes Law for spherical pollen grains and Falck’s assumption for ellipsoidal grains. Pollen and vegetation data were subjected to Extended R-value analysis to calculate Relevant Source Area of Pollen (RSAP) and Relative Pollen Productivity estimates (RPP). RSAPs were found to be approximately 100 – 150 m in all three communities. Asteraceae and Proteaceae have low RPPs relative to Poaceae, the RPP of Ericaceae is comparable to Poaceae, and Pteridophyte spores and Rosaceae and Podocarpaceae pollen have significantly higher RPP values than Poaceae. These results imply that the cover of some important shrub taxa may well be underestimated in palaeoecological records from the area relative to forested communities. The number of modern pollen samples collected and analysed was small, and therefore these results can only be considered initial estimates, however they demonstrate that it is possible to use soil surface samples to obtain RSAP and PPE estimates in this important ecological and cultural region of southern Africa.