2.1 Conceptual Case Study—Sampling Design Outline

Water quality progressively deteriorates in a river supplying a lake, but the upstream catchment lacks hydrological infrastructure and monitoring points. The land use is diversified, and discharged pollution that impacts the lake water quality may originate from one or all activities conducted in the catchment, including dairy production, chicken farms, crops, fruit cultivation or mining. A sampling programme designed to identify pollution sources and their relative contributions to lake water quality requires optimised selection of sampling points across the whole catchment (see also the case study presented in Chap. 7).

The potential contributions from various sources of pollution are determined by dividing the catchment into a few subcatchments with expected different types and levels of pollution based on land use and infrastructure (Fig. 2.1). Each major tributary is sampled above and below potential sources of pollution and above and below the mouths of various order tributaries (mixing points). In this conceptual example, a total of 15 surface and 4 groundwater sampling points are used. The samples collected above mixing points (e.g., b2, b3) represent tracer concentrations and stable isotope compositions in the water outflowing from each subcatchment. The sample collected below the mixing point (e.g., b1) reflects a mean value that is proportional to the values of the samples (b2 and b3) collected above the mixing point and the volume of water arriving from each subcatchment.

Fig. 2.1
A sample design outline of different types and levels of pollution based on land use and infrastructures. It includes a total of 15 surface and 4 groundwater sampling points. B 1 denotes the sample collected below the mixing point, B 2, and B 3 denotes the sample collected above mixing.

Map of the conceptual sampling design on the catchment scale. This framework can serve as a template for selecting sampling points, although it requires accounting for local conditions, land use and research objectives

Full size image

Mean Signature of the Catchment

Concentrations of pollution at sampling point a1 represent the mean value for the entire catchment. Some parts of catchments can be heavily polluted, whereas others can be pristine, and the pollution concentrations can be distinctive at different locations. Therefore, the mean concentrations will reflect the pollution loads and the volumes of water inflowing from different subcatchments that are characterised by different levels of pollution. The stable isotope compositions observed in water and water solutes reflect a mean value proportional to the stable isotope compositions and contributions from all significant pollution sources across the whole catchment. The stable isotope compositions could also be further modified by secondary processes that cause stable isotope fractionation during the chemical transformation of pollution or its removal.

Baseline Background Values

All major ion concentrations and stable isotope compositions of chemical compounds that may occur in waters naturally or can be delivered with rainfall should be considered to understand the baseline values and to separate pollution originating from land use from natural concentrations and inputs from outside the catchment (Chap. 3). The baseline values can be established by sampling water in the section of tributaries above the expected sources of pollution (e.g., above a mine site, point e1) or from pristine areas (e.g., points g1, h1, c1, Fig. 2.1). Points h1 and g1 are located in the pristine area of a nature reserve and are not directly impacted by either agricultural or industrial activity. If these data are consistent with those obtained for other likely unpolluted points e1 and c1, they constitute a general baseline for this area and characterise the natural variability in solute concentrations and their stable isotope compositions.

The sampling of spring water or groundwater could also be considered. Additionally, the sampling of rainfall may provide useful information about the δ(2H)H2O and δ(18O)H2O values delivered to the study sites and for estimating evaporative losses. Rainwater chemical and stable isotope compositions can be useful for partitioning airborne pollution in regions where air pollution is a serious concern.

Subcatchment Division with Respect to Expected Pollution Sources

The studied catchment can be divided into a few subcatchments that separate sections of the major creek lines impacted by a specific type of pollution. These creek sections are named AAA, BBB to FFF. Surface water sampling points are labelled a1, b1 to k2; groundwater sampling points x1 to x4 (Fig. 2.1).

The river section EEE flows through the orchard, and if polluted, it can be expected to carry primarily fertilisers, pesticides and the other agrochemicals used in this area. The water quality of EEE can be analysed at the bottom of the subcatchment (k2) before it mixes with FFF to account for the pollutants originating from the orchard only. To understand the actual net input of pollution from the orchard to EEE, the quality of water entering the subcatchment should be considered and point h1 located in pristine area can be used to establish a baseline for water hydrochemistry entering this catchment. The difference between concentrations and stable isotope compositions between h1 and k2 reflects the added pollution loads from the orchard.

In contrast to EEE, the river section FFF may be impacted by pollution from two potential sources: the dairy factory and the orchard. Therefore, the difference between the results obtained for k3 and g1 (Section FFF) reflects the potential inputs from these two sources. The results for FFF can be further compared with the results for EEE. If the baseline values at h1 and g1 are not significantly different, then the difference between k2 and k3 could reflect the potential input from the dairy factory to FFF. The direct signature of potential pollution from the dairy factory can also be verified using groundwater (bore x4).

The proportional contribution of FFF and EEE to KKK and to lake pollution could be further calculated using mass balance calculations and the results for individual tracers from the triple sampling point k3 + k2 = k1. The concentrations and stable isotope compositions at sampling point k1 will reflect those at k3 and k2 with respect to the contribution from FFF and EEE. A simple verification procedure can be designed using the mass balance model (see calculation examples in Chap. 3) and the following equation (Eq. 2.1):

$$\delta_{k1} = x \times \delta_{k2} + y \times \delta_{k3} ,$$
(2.1)

where δk1, δk2 and δk3 are stable isotope compositions of analysed samples collected at k1, k2 and k3. Assuming contributions from only two water sources, the proportional contributions x and y will equal 1 (Eq. 2.2).

$$x + y = {1}.$$
(2.2)

Solving these simultaneous equations, we can calculate the relative contributions x and y that reflect the inputs from EEE and FFF, respectively. This calculation can be repeated for various tracers, including stable isotope composition, ion concentrations, to calculate the relative contributions of various pollutants, ions and volumes of water. Combining these mass balance calculations allows estimation of the relative loads of pollution from both subcatchments, even without directly measuring the volumes of inflowing water, which is often challenging in the field.

A similar approach can be adopted for other subcatchments of the same order. The River Section BBB flows through an area with multiple industrial poultry farms. The potential influence of the farms on water quality in the river section BBB can be verified by comparing the results from b3 located downstream and c1 located upstream of the farms. The potential impact of the farms and the stable isotope signature of pollution can also be verified by sampling local drains (b4) or shallow groundwater bores (x1, x2 and x3). The potential impact of the mining site on water quality in the river section DDD can be verified by analysing the difference in water quality between sampling point e1 located upstream in relation to the mine and point d3. The influence of agrochemicals used for crop farming can be analysed by comparing the results from b2 and d1. Knowing the water quality change through each of the selected creek and river sections, the contributions to the catchment pollution budget can be further calculated using other triple mixing points (d3 + d2 = d1 for mixing between DDD and KKK and b3 + b2 = b1 for mixing CCC and BBB).

2.2 Other Input Data

The ranges of the stable isotope compositions typical of the various types of pollutants can be obtained from the literature; however, these values should be verified locally. If possible, the signatures of potential pollution sources should be analysed directly in the study area at the source of pollution (e.g., by obtaining fertilisers and discharged wastewater directly from animal farms or water draining from mining sites, etc.) (see also, Chap. 6).

The use of this simplifying sampling design template requires some understanding of the study area, particularly its hydrology and land use, and the acquisition of basic GIS information about the locations of potential sources of pollution. Any hydrochemical or hydrogeological information, if available, will help improve the interpretation of the results. The suggested multiple mixing models that cover different orders of catchments will allow the calculation of the relative contributions of the pollution from each listed source if the stable isotope signatures in the sources are significantly different. Understanding the local climate and obtaining rainfall records are also important factors. Fieldwork for hydrochemical studies is usually recommended during dry periods, at least a few weeks following substantial precipitation events. High-volume precipitation may dilute solute concentrations after the initial flushing down of pollutants accumulated and retained in water pools or soil. However, the major ion ratios and stable isotope signatures will not be directly impacted by dilution with rainwater if the concentrations of solutes are a few orders of magnitude higher than those in precipitation.