Spatial Variation And Causal Factors Of Heavy Metals And Phosphorus In Surface Sediment From The Catchment Of The Ningxia-inner Mongolian Reaches, Yellow River

The Ningxia-Inner Mongolian reaches of the Yellow River(NIMYR) belongs to the middle to lower reaches of the upper Yellow River in terms of its geology for 730 km with a drainage area of 164,746 km2. Fifty-six riverbed surface sediment(RSS) samples were collected along the NIMYR. These samples were analyzed to determine their heavy metal concentrations(Co, Cr, Ni, Cu, V and Zn), grain sizes, sediment sources and the causes of their heavy metal contamination. The cumulative distribution functions of the heavy metals in RSS of these reaches are plotted to identify the geochemical baseline level(GBL) of each element and determine the average background concentration of each heavy metal. Principal component analysis and hierarchical cluster analysis are conducted based on the grain sizes of the RSS and the samples are classified into two groups: coarse grained samples(CGS) and fine grained samples(FGS). The degree of heavy metal contamination for each sample is identified by its enrichment factor(EF). The results reveal that the coarse particle component(medium sand and coarse sand) in the bed materials is chiefly from the bordering deserts along the Yellow River. The clay and silt in the bed materials chiefly originate from the upper reaches of the Yellow River, and the fine sand is identif ied as hybrid sediment derived from the upper reaches of the Yellow River and the bordering deserts. The CGS primarily appeared in the reaches bordering deserts, and the sites are near the confluence of gullies and the Yellow River. The FGS are located adjacent to cities with especially strong industrial activity such as Wuhai, Bayan Nur, Baotou and Togtoh. The Cr, Ni, Cu, V and Zn concentrations(mg kg-1) are 84.34 ± 49.46, 30.21 ± 7.90, 25.01 ± 7.61, 73.17 ± 18.92 and 55.62 ± 18.93 in the FGS and 65.07 ± 19.51, 23.86 ± 6.84, 18.04 ± 3.8, 53.47 ± 10.57 and 34.89 ± 9.19 in the CGS respectively, and the concentrations of Co in the CGS(213.40 ± 69.71) are notably higher than in the FGS(112.02 ± 48.87) and greater than the Co GBL(210). The most contaminated samples in the NIMYR are adjacent to the cities of Wuhai(EFCr = 5.19; EFNi = 1.96), Bayan Nur(EFCr = 5.88; EFNi = 2.08) and Baotou(EFCu = 1.55; EFZn = 1.68) where the Cr, Ni, Cu, V and Zn concentrations are above the correlated GBLs(85, 34, 27, 75 and 62 mg kg-1, respectively), which are mostly affected by industrial processes, and samples that are only moderately contaminated by heavy metals are found in the reaches bordering desert(Wuhai- Baotou) because contaminated sediments are diluted by uncontaminated desert sand. In contrast, all of the Cu, Cr, Ni, V and Zn concentrations in RSS of the Qingtongxia-Wuhai reach are lower than the correlated GBLs of elements.Surface sediment samples concurrently collected in the catchment of the NIMYR were analyzed to determine the total phosphorus and correlated physiochemical properties of s ediments. Samples were obtained from three main areas: the RSS, the surface sediment of floodplains(FSS) and the surface sediment of deserts(DSS). A cumulative distribution function in this catchment was plotted to identify the geochemical baseline level and the background concentration of phosphorus. Principal component analysis and hierarchical cluster analysis were conducted based on the physiochemical properties of RSS and FSS. Through the analysis, we found that the heaviest phosphorus contamination(average phosphorus concentration: 562.40 mg kg-1) occurred in FSS. The next heaviest contamination was in RSS, which was only moderately contaminated by phosphorus overall. However, the highest phosphorus concentration(749.40 mg kg-1) appeared in RSS due to the point-source pollution by humans. In contrast, the phosphorus-contaminated level of DSS was the lowest due to the shortage of anthropogenic activities. This is the first study to present a qualification of the effect of the sediment‘s physicochemical properties on phosphorus by the multivariate regression tree analys is. The co-precipitation of phosphate with calcite [i], phosphate absorbed onto ferric hydroxides [ii] and the grain-size effect [iii] were the three main control mechanisms for phosphorus distribution in the fluvial surface sediment of the NIMYR. The contributions of the three mechanisms to the RSS and FSS were, respectively, [ i](74.7%) > [iii](1.2%) > [ii](1.1%) and [i](75.3%) > [ii](9.1%) > [iii](1.6%). The heaviest phosphorus-contaminated group in RSS was primarily controlled by mechanism [ i], whereas the heaviest phosphorus-contaminated group in FSS was controlled by mechanisms [i] and [ii], indicating that the FSS had a higher probability and potential r isk of releasing phosphorus from the sediment to overlying water.