Determinants of the Fynbos/Succulent Karoo biome boundary: Insights from a reciprocal transplant experiment

• We disentangled the relative roles of climate and soil type in determining the Fynbos-Succulent Karoo biome boundary.
• Karoo species are neither edaphically nor climatically constrained from establishing in fynbos soils or climates.
• Fynbos species are strongly limited edaphically and climatically under karoo soil and climate conditions.
• Fynbos fire regime is likely responsible for maintaining the sharply delineated boundaries between these biomes.

Boundaries between Fynbos and Succulent Karoo vegetation in the Greater Cape Floristic Region are frequently characterised by sharp transitions from sandy, dystrophic to loamy, mesotrophic soils, together with a more gradual climate transition from cooler, wetter conditions typical of Fynbos at higher elevations to warmer, drier conditions at lower elevations typical of Succulent Karoo. There is very high species turnover across these boundaries, providing an opportunity to disentangle the relative roles of climate and soil type in determining the biome boundary. A fully reciprocal transplant approach was employed here to investigate this question, using three species from each biome occurring naturally in close proximity at Jonaskop, Western Cape. Greenhouse-germinated and established seedlings of all species were planted into both sandy, dystrophic and loamy, mesotrophic soils typical of each biome at four sites along an elevational transect (elevations 545 m, 744 m, 953 m, 1303 m) at Jonaskop, and their growth and survivorship monitored for 7 months.At least one site on the elevation gradient represented annual climate rainfall and temperature conditions during the experimental period typical of the range edge for each of the selected species, this typically being the lowest elevation site (warm × dry) for Fynbos species (Protea humiflora, P. magnifica and P. amplexicaulis), and the highest elevation site (cool × wet) for Succulent Karoo species (Ruschia lineolata, Drosanthemum speciosum and Pteronia incana). Fynbos species showed significant adverse responses to loamy mesotrophic soil, with highly significant reductions in growth and high and rapid rates of mortality relative to their native soil. Fynbos species showed somewhat reduced growth and survival at the lowest elevation when grown in native soils, but demonstrated significant interaction between soil type and elevation with much lower growth and survival at the lowest elevation on the loamy soils. Surprisingly, all the Fynbos species showed reductions in growth and survival at the highest elevation, with significant reductions in two of the three species. Succulent Karoo species, by contrast, showed very few significant performance differences between soil types and few significant soil × elevation interactive effects, but did show significant growth and survival responses to elevation, with high growth and survival at mid-level elevations, far higher than their natural extent at the site.These results suggest that the selected Succulent Karoo species are neither edaphically nor climatically constrained from establishing and growing in sandy dystrophic soils and cool climates typical of the Fynbos vegetation along this elevation gradient, but that Fynbos species are strongly limited both edaphically and climatically from growing under conditions typical of the Succulent Karoo. We propose that Succulent Karoo elements may be excluded competitively or through disturbance from colonising sandy dystrophic soils at higher elevations in Fynbos vegetation, with fire regime most likely responsible for maintaining the sharply delineated boundaries between these biomes. This is because fire would strongly exclude non-fire adapted Succulent Karoo species at and above the biome boundary, while loamy soils and climate together would strongly exclude Fynbos species from the heavier soils of the Succulent Karoo. The relative climate and soil affinities of these biomes, accentuated by the role of fire, could therefore provide a coherent explanation for biome boundaries in the Greater Cape Floristic Region. We note however that the limited species selection in this study precludes a conclusive general result, and that several interesting questions remain about soil, climate and disturbance determinants across this biome boundary.