How best to quantify soil seed banks in arid rangelands of the Nama Karoo?

Sampling design and three sample treatments prior the application of the seedling emergence method were tested in order to find the best method for seed bank quantification in arid Nama Karoo rangelands. I analyzed species composition and seed densities by contrasting undercanopy and open-matrix samples from two soil depths and by comparing the effects of cold-, heat-, and no stratification on germination rates of species in a greenhouse setting. The soil seed bank showed minimal similarity to the standing vegetation, with only 20 plant species germinated. Spatial distribution of seeds was highly heterogeneous. Nearly 90% of germinated seeds were located in 0- to 4-cm compared to >4- to 8-cm soil depth. Undercanopy seed banks contained significantly more species and seeds than open-matrix seed banks. Neither the number nor the diversity of seeds germinated differed significantly among the three treatments. Cold stratification tended to detect more species and seeds only at >4- to 8-cm soil depth. The results highlight the importance of spatial heterogeneity in the accurate evaluation of soil seed banks in the arid Nama Karoo and the importance of considering seasonal variability in the availability of readily germinable seeds. Data also suggest that sample pretreatment in germination trials may give little return for cost and effort, which emphasizes that it is more important to choose the sampling design most likely to give a representative number of seed bank species. Further studies are needed to analyze seed bank dynamics and species-specific germination requirements to promote recruitment of plant taxa underrepresented in the seed bank.