Advancing techniques to fingerprint sediment sources in disturbed watersheds

土壤侵蚀已被确定为导致世界各地水质恶化和作物减产的最严重威胁之一。土壤侵蚀和与沉积物相关的影响--如水库和鲑鱼产卵砾石的沉积，以及沉积物上附着的污染物的运输--每年给加拿大经济造成数十亿美元的损失。随着人口的增加，气候的变化以及相关的野火和洪水，这些成本预计将增加。确定侵蚀的来源将有助于减轻土壤和河岸侵蚀，从而节省成本。沉积物指纹识别是一项新兴技术，它通过使用沉积物颜色、地球化学和放射性沉降物等关键诊断特性或示踪剂，直接将有问题的沉积物与其来源联系起来。这项研究计划的长期目标是通过推进一些最近确定的示踪剂的使用，并通过以一种新的方式测试其他示踪剂的使用，来改进和扩大沉积物来源指纹技术。工作将集中在两个流域Quesnel(不列颠哥伦比亚省，卑诗省)和南烟溪(马尼托巴省)，这两个流域代表着不同的条件(例如，有森林/草原的山区和有农田的平地)，并且存在大量补充数据。一个项目将改进化合物特定稳定同位素(CSSI)作为示踪剂的使用，以确定来自特定植被类型的沉积物，如某些草本植物或农作物。第二个项目将评估持久性有机污染物(持久性有机污染物，如多氯联苯和遗留农药)和多环芳烃(PAHs)作为示踪剂的应用，以确定来自特定土地使用活动地区的沉积物和来自野火的沉积物，例如2017年夏天发生在不列颠哥伦比亚省的野火。第三个项目将改进经常被称为混合或混合模型的模型的使用，这些模型能够定量地将沉积物分配到来源。这项工作将具体使用基于贝叶斯的模型(MixSIAR)，目的是改进该模型如何利用与沉积物和土壤等来源材料中的示踪剂相关的可变性的信息，以及该模型如何纳入CSSIs、持久性有机污染物和多环芳烃等示踪剂特性。这项工作将通过测试和评价几组新的和正在开发的示踪剂以及提高混合模型定量查明问题沉积物来源的能力，加强沉积物来源指纹方法的使用。这些改进的技术将促进沉积物指纹的使用，并通过提供保护我们的土壤和水资源的准确手段，直接造福加拿大的经济。

Soil erosion has been identified as one of the most serious threats leading to deteriorating water quality and lowering of crop yields throughout the world. Soil erosion and associated sediment-related impacts - such as sedimentation of reservoirs and salmonid spawning gravels, and the transport of pollutants attached to the sediment - cost billions of dollars each year to the Canadian economy. With an increasing population, and changes in climate and related wildfires and floods, these costs are expected to increase. Determining the origin of the erosion will allow for the mitigation of soil and channel bank erosion, leading to cost savings. Sediment fingerprinting is an emerging technique that directly links problematic sediment to its source by using key diagnostic properties or tracers like sediment colour, geochemistry and fallout radionuclides. The long-term objective of this research program is to improve and expand the sediment source fingerprinting technique by advancing the use of some recently identified tracers and by testing the use of other tracers in a novel way. Work will focus on two watersheds Quesnel (British Columbia, BC) and South Tobacco Creek (Manitoba) which are representative of contrasting conditions (e.g. mountainous with forests/grassland vs flat with cropland), and where a considerable amount of complementary data exists. One project will improve the use of compound-specific stable isotopes (CSSIs) as tracers to identify sediment coming from specific vegetation types, such as certain grasses or crops. A second project will evaluate the application of persistent organic pollutants (POPs, like PCBs and legacy pesticides) and polycyclic aromatic hydrocarbons (PAHs) as tracers to determine the sediment coming from areas with specific land use activities and from wildfires, such as those that occurred in BC in summer 2017. A third project will improve the use of models often referred to as mixing or unmixing models that are able to quantitatively apportion sediment to sources. This work will specifically use a Bayesian-based model (MixSIAR) with the goal of improving how the model utilises information on the variability associated with tracers in sediments and source materials like soils, and how the model incorporates tracer properties such as CSSIs, POPs and PAHs. This work will strengthen the use of the sediment source fingerprinting approach by testing and evaluating several novel and developing groups of tracers and by improving the ability of unmixing models to quantitatively identify the sources of problematic sediment. These improved techniques will advance the use of sediment fingerprinting and will directly benefit Canada's economy by providing an accurate means to protect our soil and water resources.