A Multifactor Statistical Model for Analysing The Physico-Chemical Variables in the Coastal Area at St-Louis and Tamarin, Mauritius

Vandna JOWAHEER
Varunah LALBAHADOOR
Roshan RAMESSUR
Lutchmee DOSORUTH

Keywords

experimentation, statistical modeling and analysis, estuaries, trace metals, nutrients, physical parameters

Abstract

The coastal pollution is an issue of concern for Mauritius. Since the past two decades, agricultural activities have contributed to pesticide and fertilizer run off in coastal waters. Over the recent years, the major urban and industrial growth in Mauritius have also contributed to water pollution by indirect wastewater discharge containing contaminants into rivers. As polluted water is hazardous to both marine life and human beings, it is of national interest to analyse the levels of water pollutants and the factors effecting these levels. This paper aims at developing a statistical model to evaluate the extent of coastal pollution in urbanized and agricultural regions in Mauritius. The study was carried out at two stations: St-Louis River (an urbanized industrial area) and Tamarin River (an agricultural area). A multifactor statistical model was formulated and analysed for the experimental data on the chemical and physical variables collected at the two sites for the estuary and downstream at each station, during 2001 and 2005 randomly spread over summer and winter seasons. The model is highly efficient in depicting the independent as well as the interactive effects of seasonality, time-interval, strategic locations and activities on the levels of different variables. A series of interesting conclusions were drawn from the analysis of the model. One major derivation was that the seasonal factor and time-interval had a significant effect (p-values < 0.01) on the levels of chromium, lead and nitrates at both the stations. However, the direction and magnitude were different with respect to each variable over the strategic locations. Moreover, considerable interactive effect between various factors regarding salinity was detected. These conclusions among others raise concern.