Fig 1.
Maps of (A) the biomes of South Africa with the Fynbos biome separated into fynbos, renosterveld and strandveld, and (B) the ratio of winter half-year (Apr–Sep) to summer half-year precipitation.
The conservation areas that were sampled using a 1 km2 grid of points are indicated.
Table 1.
Climate characteristics associated with vegetation types of South Africa.
Fig 2.
Correlation of annual average normalised difference vegetation index (NDVI) with (A) the average annual precipitation (MAP) and (B) water availability measured as the average of monthly precipitation (mm)–potential evapotranspiration (PET, mm) of each vegetation type.
The vegetation types are indicated by codes (see Table 1). The coefficient of determination (r2) is shown for significant relationships (P < 0.05). Points and error bars indicate mean ± SE (for n see Table 1) and the grey bands indicate the 95% confidence limits.
Fig 3.
Time lag between maximum monthly precipitation or mean monthly temperature and maximum normalised difference vegetation index (NDVI) for each vegetation type.
Different letters indicate significant (P < 0.05) differences between vegetation types as determined by one-way ANOVA followed by post-hoc Tukey tests. Vegetation types are arranged in order of increasing NDVI.
Fig 4.
Correlation of annual average normalised difference vegetation index (NDVI) with base saturation (BS) of each vegetation type.
The vegetation types are indicated by codes (see Table 1). The coefficient of determination (r2) is shown for significant relationships (P < 0.05). Points and error bars indicate mean ± SE (for n see Table 1) and the grey band indicates the 95% confidence limits.
Fig 5.
Correlation between base saturation (BS) and (A) the average annual precipitation (MAP) and (B) water availability measured as the average of monthly precipitation—potential evapotranspiration (PET) of each vegetation type (Fynbos separated into strandveld, renosterveld and fynbos).
The vegetation types are indicated by codes (see Table 1). The coefficient of determination (r2) is shown for significant relationships (P < 0.05). Points and error bars indicate mean ± SE (for n see Table 1) and the grey bands indicate the 95% confidence limits.
Fig 6.
Water availability measured as the average of monthly precipitation—potential evapotranspiration (P-PET) for the vegetation types in the month when annual NDVI was maximal.
A positive value indicates that precipitation exceeded potential evaporation when NDVI was maximal. Different letters indicate significant (P < 0.05) differences between vegetation types as determined by one-way ANOVA followed by post-hoc Tukey tests. Vegetation types are arranged in order of increasing NDVI.
Table 2.
Multiple regression analysis of the climatic and edaphic predictors of NDVI and BS retained following stepwise model simplification based on AIC.
Table 3.
Comparison of the multiple regression standardized coefficients for average annual NDVI regressed on water availability (P-PET) and base saturation (BS) for the vegetation types of South Africa, as defined by Mucina and Rutherford (2006), but with the Fynbos separated into strandveld, renosterveld and fynbos.
Fig 7.
SEM representing the standardised path coefficients and r2 values for significant (P < 0.001) paths from exogenous environmental variables including water availability measured as the average of monthly precipitation—potential evapotranspiration (P-PET), precipitation seasonality, precipitation concentration index (PCI), base saturation (BS), total exchangeable bases (TEB), cation exchange capacity (CEC), organic carbon (OC) and clay through two composite variables (Water and Soil) to NDVI.
The direct influence of water availability on NDVI (0.86) is greater than the indirect influence through the soil composite variable (product of coefficients = 0.13). The Chi2 P value was 0.97 with d.f = 1.
Fig 8.
Contour plot of the average annual NDVI versus water availability (P-PET) and base saturation (BS) with a superimposed scatterplot for the entire dataset.
The colours of the contour surface correspond to the NDVI values of the points (key on right). The solid, dashed and dotted lines represent the 95%, 50% and 5% quadratic quantiles, respectively, estimated using the ‘quantreg’ package [60] in R [57].