The utility of Trichilia dregeana leaves as a bioindicator of air pollution within selected industrial areas in the eThekwini Municipality, South Africa.

Abstract

Increased anthropogenic activities worldwide have led to dangerously high levels of gaseous emissions. Air pollution levels within industrial areas in South Africa, such as South Durban Basin (SDB), are monitored daily at a few monitoring stations; however, limited coverage of, and data generated by, these stations necessitates alternative strategies such as biomonitoring. In this regard, the use of tree leaves as bioindicators of air pollution can generate valuable data on environmental health and pollutant levels. The present study investigated the utility of Trichilia dregeana Sond. leaves as a bioindicator of air pollution within selected industrial areas in SDB. The first part of the study focussed on effects of SO₂ pollution on leaf morphological (leaf area), physiological (leaf chlorophyll content and leaf fluorescence) and biochemical (intracellular superoxide and hydrogen peroxide production, total aqueous antioxidant activity, and electrolyte leakage) biomarkers of environmental stress. Leaves were sampled over four seasons from (four) trees growing at three industrial (treatment) sites (Prospecton, Ganges and Southern Works) within SDB and from greenhouse-located trees, which served as an ex situ control. Results indicated annual SO₂ concentrations ([SO₂]) were high by global standards and significantly different (p<0.001) across sites, with levels being highest at Southern Works. All biomarkers, except leaf chlorophyll fluorescence, could discriminate between SO₂-exposed and -unexposed leaves. Seasonal data for many of these biomarkers were significantly (p<0.001) correlated with seasonal [SO₂]; however, none of them reflected differences in [SO₂] across treatment sites. The second part of the study, partial least squares regression (PLSR) was used to quantify the relationship between two air pollution biomarkers (chlorophyll content and leaf area) and hyperspectral data. Trichilia dregeana leaves (n=28) were sampled in spring and summer only. Spectral reflectance data were able to distinguish between SO₂-exposed and -unexposed leaves and PLSR was able to relate the hyperspectral dataset to both biomarkers. However, the interaction between biomarkers suggests simultaneous prediction of these, using an algorithm such as PLS-2, may be more suitable. The variable importance in projection method identified wavebands within the red-edge region of the electromagnetic spectrum that showed promise in identifying stress in the leaves of T. dregeana. Collectively, the results provide ample motivation for the establishment of T. dregeana leaves as a bioindicator of air pollution.