Sediment and Erosion subproject

Montclair State University REU 2012

Gregory A. Pope

Introduction

Northwest New Jersey has concerns for its water resources, given the land use tensions of growing urbanization as well as the need to conserve natural resources and open space. Stream sediments reveal many aspects of stream quality, and we attempt here a study similar to that of Gellis et al. (2009, USGS Sci. Inv. Rpt. 2008–5186) and Devereux et al. (2010, Hydrol. Proc. 24), who used trace element fingerprinting to determine the origin of stream sediments. Sediment sources relate to land uses/land cover, and are important in assessing the impacts of agriculture, deforestation or reforestation, and urbanization on erosion and sedimentation of rivers and wetlands. The results presented here illustrate an initial analysis of stream bed sediments only. What trace elements are best related to stream stability measures? Can trace elements indicate proximal sources of sediment, such as stream banks and reworked channel sediments?

Methods

Field locations involved a selection of forested, agricultural, and urban settings. REU team members used the Rapid Geomorphic Assessment, observing stream channel, bank, and floodplain criteria (Kline et al. 2009), to assess channel stability. Samples were obtained from locations along the Flat Brook, Wallkill, and Rockaway River watersheds. Core samples of sediments and soils employed a PVC pipe tool, sectioning the core into 2-3cm increments into storage bags in the field.  Samples were oven dried, then pulverized to a fine powder, combined with LiBo flux and fused in a muffle furnace at 1050°C.  The molten sample is then dissolved in a 7% HNO3 solution to a 500x dilution factor, then diluted again with 2% HNO3 for a 10,000x dilution suitable for inductively coupled plasma mass spectrometry (ICP-MS) analysis.  Samples as well as USGS rock standards were analyzed with ICP-MS for W, V, Cr, Co, Ni, Ga, Rb, Sr, Y, Zr, Nb, Cs, Ba, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu, Hf, Ta, Pb, Th, and U.

Results

Cobalt, Lead, and Strontium are examples of trace elements exhibiting a linear relationship with the RGA value: as RGA increases (increasing instability), trace element content increases. These elements are more mobile when bank erosion is higher. Trend lines with respective correlations are shown in Figure 1.  Gallium, Rubidium, Yttrium, and Niobium exhibit a complex nonlinear relationship with the RGA value. Trace element content peaks at middle RGA values, and decreases at the RGA extremes, a relationship that approximates a polynomial function. Reasons for this relationship are unknown, but may have to do with particle size of the sediment, such as clay content.

Figure 1: sediment project chart

sediment project box and whisker plots‌Several of the trace elements differentiated sediment source with good (p<0.05) to marginal (p<0.10) statistical significance: Tungsten (p=0.052), Strontium (p=0.024), and Cobalt (p=0.090). (Tungsten’s distribution is shown in Figure 2.) Stream bed sediments are distinctly lower in tungsten, compared to other sources. Agricultural soils, and to a lesser extent urban sediments, preferentially contribute Sr. No single trace element is capable of distinguishing all sediment sources.

sediment project box and whisker plotsThe predominant land use or land cover in each drainage basin had some influence on both stream quality and the trace element composition. RGA was shown to correlate to land use.  RGA value increases (more unstable) with increasing urban land coverage in the respective watershed. RGA value decreases with increasing forested land coverage in the respective watershed.  Land use class is differentiated by tungsten (ANOVA p=0.044) and lead (ANOVA p=0.022). The boxplot of lead is shown in Figure 3, indicating elevated lead in urban areas. There is a field data bias, however. An accidental outcome of the data gathering strategy showed a covariance of land use with geology. Urban areas were sampled only in granitic or shale bedrock areas; conglomerate and dolomite dominated in forested watersheds.

Discussion and Conclusions

Trace elements respond differently to within-channel erosion potential. Several elements, notably Cobalt, Lead, and Strontium, indicated a good correlation to the RGA stability assessment. In other words, an increase in these elements is indicative of increased within-channel and/or bank erosion. Other elements, such as Gallium, Rubidium, Yttrium, and Niobium, indicate a more complex relationship with stream stability. Both very stable (low RGA) and very unstable (high RGA) conditions seem to limit the abundance of these trace elements, while middle values of stability have higher trace element content. It is clear that a better understanding of the trace element and sediment budget is needed, a topic of ongoing research. There is already indication that the RGA is well correlated to land use, such as coverage of forested and urban land in the watersheds.  
Sediment sources are important, and some trace elements reveal the origin of the sediment (within the channel, in the banks or floodplain, from upland forested or agricultural areas, or urban sources). No single element is able to differentiate all sediment sources, and a combination of elements would be necessary to discern all sediment sources.

Our preliminary chemical analysis indicates that some trace elements are also related to land use, relating to the sediment source. Chromium, for instance, is statistically significant in determining a difference between forested, agricultural, and urban influences. These relationships will be explored further as additional samples are analyzed. Currently, the existing samples co-vary land use with local geology, a confounding issue. We hope to gain a better understanding of the trace element flux and budget such that it relates to erosion and sediment transport and storage. Samples are being processes for Pb-210 and Cs-136 activity levels that will serve to differentiate recent versus older sediments.