Contamination,Risk Assessment And Prediction Of Heavy Metals In Soil Of Jiuquan Area

Jiuquan area is located in a key part of the Silk Road Economic Zone,which is the main component of the northwest granary(Hexi Corridor area)in China.The study area is the main production base of grain and oil in northwest China,food security is vital to the sustained and stable development of the ecology and society.With the rapid and extensive development of industry and agriculture,soil environment is inevitably contaminated by heavy metals,which cause adverse effects on the ecological environment and human health.Meanwhile,the traditional methods of combining field sampling with laboratory analysis of soil heavy metals are not only time-consuming and expensive,but also can only obtain limited information,and the spatial and temporal distribution characteristics of heavy metals in soil can not be analyzed in large area.Therefore,the research on the pollution status and prediction of heavy metals in soil of Jiuquan area is very important for food safety,human health,economic development and ecological environment protection.To identify the spatial distribution,pollution status,ecological risk and health risk to local residents of heavy metals in Jiuquan area,382 soil samples(Gobi and farmland)were collected and analyzed for their elements concentrations and particle sizes using a X-ray fluorescence spectrometer(XRF)and Mastersizer 2000.Meanwhile,based on these measured concentrations,the accuracy of using spectral data to predict the contents of heavy metals was verified.The grain characteristics,the spatial distribution and sources of heavy metals in Gobi and farmland samples were analyzed using ternary diagram,descriptive statistics,ordinary Kriging interpolation and Pearson correlation coefficient matrix.The results showed that the grain size distribution of samples in Gobi and farmland of Jiuquan area showed the characteristic concentrations of silt and sand(the sum of both was greater than 90%),and the clay content was very low,and Gobi sample particles were coarser than the farmland samples.The mean concentrations of heavy metals(Cu,Cr,Mn,Ni,V,and Zn)in the two types of samples exceeded their background concentrations(especially farmland samples),and all heavy metals concentrations in farmland soil were higher than those in Gobi soil.These heavy metals followed a similar spatial pattern(the concentrations of heavy metals were relatively high located in thesouth,east,and northwest of study area with urban and frequent industrial activities,and those of the north and middle were low)and showed significant positive correlations with each other(r >0.55,p < 0.01),implying that they have the same sources which may be traffic,industry,mining and agricultural activities.The contamination features of heavy metals and ecological risks were investigated using several quality indicators(pollution index(PI),pollution load index(PLI),and potential ecological risk index(RI)),and their health risks for population groups were quantified using the health risk model proposed by the United States Environmental Protection Agency.The results showed that the Gobi and farmland soils are uncontaminated to moderately contaminated by heavy metals,and that farmland pollution is more serious than that of Gobi(the average PLI of Gobi and farmland were 1.28 and 1.76,respectively).The ecological risk caused by heavy metals was low(RI < 50)in the study area.In addition,PI,PLI and RI of heavy metals have the same spatial distribution pattern,which are consistent with the trend of concentration distribution,relatively high pollution and risks were distributed in the south,east and northwest of the research area,the north and central areas were low.As a whole,the non-carcinogenic risk which was caused by heavy metals was low(HI < 1)for local residents,and the carcinogenic risk was within an acceptable level(1E-06 < CR < 1E-04).Comparatively speaking,children are the vulnerable population to health risk,which are more susceptible to heavy metals.The Zn and Cu pollution in the study area was relatively serious,and Cr and V have been identified as major contributors to health risks.Therefore,it is necessary to reduce the levels of these heavy metals in soils.The key to mastering the pollution status of heavy metals in soil is to quickly and accurately obtain the contents and distributions of heavy metals in monitored soil.So this study combined rapid and environmentally-friendly remote sensing(multispectral data),stepwise multiple linear regression(SMLR)and partial least-squares regression(PLS)to predict heavy metals contents in soil of Jiuquan area.Besides,we used laboratory data to verify the accuracy of the prediction results.The results suggested that the SMLR and PLS models can predict the metals contents in the study area(the coefficients of the two regression models were greater than 0.5 and p < 0.05;the correlation coefficients between the predicted and measured values of six heavy metals were above 0.55).The contents of Cr,Ni,V and Zn could be predicted by two regression models,while Cu and Mn were more accurate with the SMLR model.Moreover,the spatial distribution patterns of the two predicted values of heavy metals were generally consistent with the measured values,which was increasing from north to south.The difference is that the highest measured concentrations were mainly concentrated in the vicinity of the cities and towns,while the highest predicted values were distributed in the Qilian Mountains area.The spatial distribution features ofSMLR and PLS predicted values were basically the same,but there are certain differences in the proportion of different concentration regions to the total area of the study area.