Plant Ecology and Evolution 150(1): 35-44, doi: 10.5091/plecevo.2017.1260
Soil chemical properties and plant species diversity along a rainfall gradient in semi-arid grassland of South Africa
M Dingaan‡, Mitsuru Tsubo, Sue Walker, Terry Newby‡ University of South Africa, Johannesburg, South Africa
© Botanic Garden Meise and Royal Botanical Society of Belgium. all rights reserved Citation:
Dingaan M, Tsubo M, Walker S, Newby T (2017) Soil chemical properties and plant species diversity along a rainfall gradient in semi-arid grassland of South Africa. Plant Ecology and Evolution 150(1): 35-44. https://doi.org/10.5091/plecevo.2017.1260 |  |
Abstract
Background and aims – Plant diversity is affected by several biotic and abiotic influences that include the availability of soil nutrients and precipitation regimes. Our study investigated the relationship between soil chemical properties and plant diversity in semi-arid grasslands of South Africa.
Methods – We collected plant species data and determined soil chemical compositions in Kimberley, Bloemfontein and Bethlehem. The three locations lie along an increasing rainfall gradient from around 450 to 750 mm per annum. We conducted ANOVA to examine differences in diversity indicators [species richness ( N 0 ), the inverse Simpson diversity index ( N 2 ) and species evenness ( N 20 = N 2 / N 0 )] among the locations, and differences in soil chemical properties (Ca, K, Mg, Na, P, N, C and pH). We further performed correlation analysis between soil variables and diversity indicators.
Key results – The geographical gradients of N 0 and N 2 were positively associated with the rainfall gradient and correspondingly with N and C. However, N 0 and N 2 were negatively correlated with soil pH. N 20 did not show any relationships with the geographical gradient and soil properties. Our results further revealed a hump-shaped model of N 0 with soil pH, characterised by an increase in N 0 with increasing soil pH in acidic grassland (Bethlehem) but a decrease and levelling-off in the lower rainfall, alkaline grasslands of Bloemfontein and Kimberley. The negative effect of soil pH on N 0 is presumably a result of less intense leaching of the base cations in the lower rainfall areas of Bloemfontein and Kimberley.
Conclusions – Our results indicate soil pH as the main variable determining plant species diversity in semi-arid grasslands.
Methods – We collected plant species data and determined soil chemical compositions in Kimberley, Bloemfontein and Bethlehem. The three locations lie along an increasing rainfall gradient from around 450 to 750 mm per annum. We conducted ANOVA to examine differences in diversity indicators [species richness ( N 0 ), the inverse Simpson diversity index ( N 2 ) and species evenness ( N 20 = N 2 / N 0 )] among the locations, and differences in soil chemical properties (Ca, K, Mg, Na, P, N, C and pH). We further performed correlation analysis between soil variables and diversity indicators.
Key results – The geographical gradients of N 0 and N 2 were positively associated with the rainfall gradient and correspondingly with N and C. However, N 0 and N 2 were negatively correlated with soil pH. N 20 did not show any relationships with the geographical gradient and soil properties. Our results further revealed a hump-shaped model of N 0 with soil pH, characterised by an increase in N 0 with increasing soil pH in acidic grassland (Bethlehem) but a decrease and levelling-off in the lower rainfall, alkaline grasslands of Bloemfontein and Kimberley. The negative effect of soil pH on N 0 is presumably a result of less intense leaching of the base cations in the lower rainfall areas of Bloemfontein and Kimberley.
Conclusions – Our results indicate soil pH as the main variable determining plant species diversity in semi-arid grasslands.
Keywords
BIODIVERSITY; INVERSE SIMPSON INDEX; SOIL CARBON; SOIL NITROGEN; SOIL PH; SPECIES RICHNESS; THEMEDA TRIANDRA