Field sampling program
You should never begin a field sampling program until you have:
After the basic outline of a sampling program has been agreed, you need to implement the design in the field.
Your monitoring team can follow our guidance to apply appropriate protocols for field measurements and sample collection, preservation, preparation and storage prior to any laboratory analyses.
Also, check with relevant local authorities in your jurisdiction to find out if they have already established approaches for developing field sampling programs.
Based on the specific data requirements identified in the study design process, you should consider the sample collection methods for water, sediment, biota or any other lines of evidence, including any requirements for analysis-specific sample containers, sample preservation and storage and field measurements.
Quality assurance and quality control (QA/QC) for field and laboratory measurements should be considered, together with work health and safety requirements.
Box 1 presents a checklist to help your monitoring team design an appropriate field sampling program.
Box 1 Checklist for designing field sampling programs
- Check the monitoring study design to locate relevant program elements: sample locations, sampling frequency, sample replication and monitoring parameters (biological, chemical and physical).
- Have the specific measurement parameters been identified and the data requirements stated?
- Can the required data be obtained by field measurements?
- Have appropriate field measurement techniques been selected, including calibration procedures?
- How are the positions of sampling sites to be recorded?
- What ancillary field observations are to be taken?
- Will the sampling device collect a representative sample?
- Do disturbances occur in the environment being sampled?
- Will the sample be altered by contact with the sampling device?
- Will the sampling device contaminate the sample? If yes, how is the sampling device to be cleaned?
- What are the effects of the sampling device being in contact with media other than the sample of interest?
- How are samples to be collected to prevent field contamination?
- Will the sample container contaminate or affect the stability of the sample? If so, how are these problems to be overcome?
- What size sample containers are required?
- How are samples to be preserved before analysis?
- Are procedures in place to track samples and field data?
- What process is in place to identify, measure and control errors?
- Have sampling protocols been written?
- How are sampling staff to be trained?
- How is the sampling staff’s competence to be tested?
- Can the integrity of the sample be guaranteed?
- Have blanks, duplicates and replicates been incorporated into protocols?
- How are problems to be rectified?
- Are there enough resources to prevent any bottlenecks occurring in the field or laboratory that would hinder analyses and compromise data quality?
- How are data to be stored and accessed?
- Have all reasonable steps been taken to protect the health and safety of employees?
- Have possible hazards been identified and documented?
- Have sampling staff been made aware of possible hazards, and have risk minimisation plans been developed?
- Have sampling staff been trained to ensure that sampling is done safely?
- Will sampling staff be appropriately supervised during sampling activities?
Field measurements and observations
Some parameters (e.g. flow, temperature) can only be measured in the field.
Field measurements are highly desirable for other parameters (e.g. dissolved oxygen, redox potential, pH) because their values might change in the sample after collection.
Reliable sensors make it convenient to measure many parameters in the field. Field monitoring gives on-the-spot values and allows the results to be checked immediately, which can help your team to rapidly refine the choice of sampling sites if necessary.
Whether sensing or field sampling, QA/QC is important, and it requires careful planning because it’s not easy to achieve in the field.
Field data can be obtained automatically and by remote sensing, and the data can be logged or transmitted to laboratories by telemetry. This has the advantage of providing measurements that are either continuous or at fixed intervals, allowing very cost-effective studies of temporal trends.
Field sampling is adequate for parameters that do not change during transport and storage. For example, macroinvertebrates collected in a water sample can immediately be stored in vials of alcohol and kept until they are identified and counted.
Samples that must be field sampled and then analysed in a laboratory could be stored in fixative or preservatives, and chilled during transport, to minimise changes.
Guidance on the various biological, chemical and physical measurement parameters that can be field sampled is provided in Standard Methods for the Examination of Water and Wastewater (APHA 2012) and USEPA (1996a).
Recording site details, measurements and observations
So that sites can be re-used in subsequent studies, it is important to record:
- position of each sampling site
- careful and thorough descriptions of each site and its means of access
- exact spots from which samples were taken.
Identify key onshore reference points, or the site location, using the global positioning system (GPS). If exact positioning of sample sites is necessary, then basic GPS will be inadequate and you will need a more accurate method (e.g. differential GPS). With high quality receivers and differential GPS, the accuracy can be to within a few centimetres of the actual location.
Use a single coordinate system and record which coordinate system was used, especially the datum and projection. A site identified by an easting and northing based on one datum can be up to 200 m away from a site identified by the same easting and northing numbers based on a different datum.
At each visit, note the condition of the water body and the weather conditions because these factors may influence the variables being measured. For example, changes in the wind speed and cloud cover may affect the temperature and subsequently the dissolved oxygen concentration in the water column.
Other field observations might include descriptions of odour, colour and floating material, and riverine vegetation or other conditions relevant to water quality. Take videos or photographs for future reference.
Water and sediment samples
Choosing suitable methods and equipment
You should select the most appropriate ways to collect samples or data from each sampling station and water body. Methods include:
- collection of a sample by hand
- collection by automatic sampler
- samplers that collect and integrate samples over a given time
- real-time measurement by automatic means
- measurements in the field by hand
- proximal and remote sensing
- field observation.
Your choice of sampling method depends on the parameter to be measured and the nature of the information required. For example, grab samples could be easier to preserve, less liable to contamination or a better size than samples integrated over time (or flow) by automatic devices.
All methods and equipment used should meet any relevant Australian, New Zealand and International Organization for Standardization (ISO) standards, as well as local standards like Queensland’s Water Monitoring and Sampling Manual.
Select a sampling method based on:
- objectives of the monitoring program
- local conditions (the need to obtain representative samples)
- safety of operation (safety of sampling staff is critical)
- acceptability of the method
Continuous water sampling methods and equipment can operate reliably in some areas to provide information on significant short-term variations in water quality parameters that are usually missed by discrete samples.
Time-integrated sampling reduces analysis costs and enables mean values to be calculated simply. But integrated sampling is not recommended if the objective of the monitoring program is to assess variations in water quality.
Samples can be taken at the water surface, or at specific depths in the water column, or integrated over depths.
For particular analytes (e.g. trace metals), the equipment must have a specific composition and be cleaned in certain ways to avoid sample contamination. Standard Methods for the Examination of Water and Wastewater (methods 1060A and 1060B) is a standard text for general procedures and principles of collecting water samples.
You will find additional guidance on sampling lakes, rivers, streams, marine waters, groundwaters and sediments in the Australian and New Zealand water quality standards (AS/NZS 5667 series).
Sampling operations include:
- preparation and labelling of containers
- appropriate selection of sampling sites
- collection of samples
- good housekeeping and field records
- photographic and video records
- use of boats and cars
- recording of parameters, such as depth and light intensity.
Spatial and temporal variability
Green (1979), in his 10 principles of sampling, said:
‘verify that the sampling device is sampling the population you think it is sampling with equal or adequate efficiency over the entire range.’
To do this, your monitoring team must specify the population that is to be sampled and its likely spatial and temporal variability.
In Australian rivers, discharge can change by 2 orders of magnitude or more, and the effectiveness of sampling devices may vary over this flow range.
Device-related sampling errors cannot be removed or accounted for by statistical methods or by replication. In many cases, such errors will be undetectable unless specific tests have been made.
The sampling device should not significantly disturb the environment being sampled or alter the samples taken otherwise the samples will not reflect what ‘was’ or ‘is’. This presents a difficulty when sampling sediment.
Grab samplers and coring devices
Problems in sediment sampling illustrate these difficulties (Simpson and Batley 2016). Blomqvist (1991) reviewed the problems of using several types of grab samplers and coring devices to obtain sediment samples. Grab samplers often do not enter sediments perpendicularly, and the sediment layers mix when they close. Most grab samplers have jaws that close semi-circularly, and sediment layers below the initial penetration are only semi-quantitatively sampled.
For quantitative sampling, it is necessary to know the area and depth sampled. Coring devices must be designed to ensure that easily resuspended surface materials are not washed away. If rotation of the core occurs, shear stress may mix the sediment and cause core shortening.
Contamination and sampling errors
You must consider the environment traversed by the sampling device so that no sampling errors are caused by the device being in contact with media other than the sample of interest.
For example, when collecting subsurface water samples for hydrocarbon analysis, the sample collection device must enter the water closed or it will pick up hydrocarbons from the water surface microlayer.
When shallow water is being sampled, take care not to stir up bottom sediment.
Sampling devices should be tested under controlled conditions to check that they quantitatively collect the sample of interest. In lieu of this, refer to studies that have compared the efficiency and limitations of sampling devices: water samplers (Harris & Keffer 1974, Lane et al. 2003), sediment samplers (Simpson and Batley 2016, Blomqvist 1991, Schneider & Wyllie 1991) and biota samplers (Devries & Stein 1991).
Using this information, choose a sampling device based on the matrix to be sampled and the unique conditions at the chosen sample site.
It’s become a requirement for many monitoring studies to sample water for trace and ultratrace contaminants, often to match a guideline. Take more care than usual to avoid sample contamination by using non-contaminating equipment, which should be cleaned with acids for sampling metals, or with detergents and solvents for sampling organic compounds. Ahlers et al. (1990) described rigorous preparation of containers and sampling.
For trace metal surveys, avoid samplers with components that may contribute trace metals (Batley 1989). Use poly(methyl methacrylate) (Perspex®) poles with all-plastic fittings to hold polytetrafluoroethylene (PTFE, or Teflon®) or high-density polyethylene (HDPE) bottles for sampling shallow surface waters. Avoid depth samplers with rubber closures.
For nutrient sampling, take care that samplers are free from nitric acid or phosphate-containing detergent residues left from cleaning.
Consider using more experienced staff for trace contaminant sampling, where the possibility of sample contamination is high, as well as for field sampling of filtered nutrients.
Many of the contaminants to be measured, particularly in relatively pristine marine or alpine waters, will be present at extremely low concentrations, which may influence:
- volume of sample required (and type of sampling device that may be suitable)
- precautions required to avoid contamination (could include use of a suitable vessel that can work away from the mother vessel, such as a dinghy, and sampling devices constructed from non-contaminating materials)
- suitability of analytical methods.
Local conditions will further dictate the method and equipment used. A bucket can be used to sample from a bridge but a telescopic pole would be more useful to sample from a river bank.
In estuarine waters, the experimental sampling design needs to account for the complex and highly variable nature of the water body.
Estuarine waters that intermix have very different chemical composition and physical and chemical properties, producing great variation — vertically, horizontally and temporally (with tidal stage).
Large numbers of samples or stratified sampling may be required, which in turn is likely to have a bearing on the selection of the sampling method.
Specific sampling techniques for estuarine and marine waters were described by Grasshoff et al. (1999) and Crompton (1989), and can be found in the Australian and New Zealand water quality standards (AS/NZS 5667 series).
Sampling surface waters
Equipment for sampling surface waters (as opposed to groundwater) falls into 5 categories:
- bottle samplers for shallow waters
- pumping systems for surface-to-medium (10 m) depths
- depth samplers (50 m to > 100 m, depending on design)
- automatic samplers
- integrating samplers.
Bottle samplers for shallow waters
Many water bodies are shallow and well mixed, only requiring surface (0 to 1 m) water sampling.
Immerse a sample bottle by hand to just below the surface (typically 0.25 to 0.50 m depth), and hold the sampler downstream of where the sample is to be collected. If your hand is covered with a plastic disposable glove, then any contribution from surface films is avoided.
Sampling may be done from the shore, by standing in shallow water, or from a boat.
Whatever vessel is used, face it into the current and take water samples from the front of the vessel, preferably while moving slowly forwards (Apte et al. 1998). This procedure minimises contamination from the boat itself.
To maintain an adequate distance between the sampling point and the sampling vessel or fixed structure, the sample bottle can be held in acrylic jaws at the end of a 1 to 2 m long polycarbonate pole, 2 cm in diameter. Or use a clean bottle or bucket fixed to a clean plastic rope, and keep equipment clean between uses by storing in a plastic bag that is then placed in a clean plastic sealable container.
Pumping systems are effective for sampling and suitable for low or trace metal concentrations of micrograms per litre and for all general water quality parameters.
Do not use pumping systems for sampling ultratrace contaminants because the tubing gives the apparatus a large surface-area-to-volume ratio that increases the chance of adsorption of the analyte (Lane et al. 2003).
These sampling systems typically involve a vacuum pump, or a peristaltic pump for shallower depths, with lengths of pre-cleaned tubing made from polyethylene, silicone, PTFE or polyvinyl chloride (PVC), approximately 1 cm in diameter. Water is pumped to the surface via the tubing into a large acid-washed Pyrex conical flask (for a vacuum pump) or a plastic sample bottle (for a peristaltic pump). Condition the tubing by pumping a large volume of water to waste prior to sampling.
This procedure has been widely used for mercury sampling (USEPA 1996b) during which online filtration is also applied.
A range of purpose-built samplers are available for depth sampling (Batley 1989).
Basic operation of a depth sampler involves a bottle, which can be opened at both ends via a wire or plastic line, deployed to the required depth. A weight is then sent down the line to trigger closure of the bottle and collection of the water. The bottle closures should be considered when determining the suitability of depth samplers. For example, rubber closures are not recommended for low metal concentrations.
In samplers for trace analysis, it is important to blank test the samplers by filling with them with clean water for the same length of time as the sampling event and then analysing the water for the analytes of interest.
Contamination often increases with the age of the sampling device so blank tests should be completed at regular intervals. The Mercos water sampler uses a group of Teflon bottles and is ideal for obtaining water column depth profiles of trace metals, including mercury at depths below 100 m (Freimann et al. 1983).
For unattended water sampling, an automatic sampler can be pre-programmed to collect samples continuously or on a flow-related or time-related basis. Such an arrangement is ideal for collecting stormwater runoff, for example, and collection can be triggered by the commencement of water flow.
Commercially available automatic sampling devices consist of a pump system, a controller and an array of sample bottles within a housing. Most instruments have a fixed number of purpose-made glass or polyethylene sample bottles fitted around the circumference of the housing. Glass is preferable for organic compounds and general water quality parameters. High-density polyethylene is better for metals.
Bottles and all surfaces are washed in acid or detergent/solvent before use.
To check that the bottles are suitable for trace contaminant analyses, analyse clean water blanks after they have been in the bottles for a suitable test period.
Very large numbers of samples can be acquired using continuous samplers (refer to Australian and New Zealand water quality standards (AS/NZS 5667 series). Such samples can be bracketed so that only a subset reflecting the conditions of particular interest need be analysed.
Be mindful that delayed preservation may compromise the integrity of samples collected by automatic devices. This would normally require that samples be collected and processed as soon as possible after the relevant event. Refrigerated samplers are available to assist in sample preservation.
Automatic samplers may not be appropriate for sampling bacteria, pH or other variables that are likely to change significantly between the time of collection by the automatic sampler and retrieval from the field for analysis.
For some sampling programs, single samplings are poorly representative of a site because water quality can vary considerably with time. Samplers that integrate water samples over a fixed time period or volume are preferred.
Two types of integrating samplers have been developed since the 1970s:
- Equilibrium samplers are designed to mimic the potential for persistent nonpolar contaminants to concentrate in aquatic organisms by comparison of the substance’s octanol–water equilibrium partition coefficient (KOW) and the sampler–water partition coefficient (KSW) and a known time to reach equilibrium (Schulz et al. 1988, Müller et al. 2001, Schäfer et al. 2010).
- Kinetic samplers do not reach equilibrium during use, which allows prediction of time-averaged concentrations of substances over the period of deployment (Södergren 1987, Huckins et al. 1993, Huckins et al. 2002, Kingston et al. 2000, Alvarez et al. 1999, Davison 2016, Huckins et al. 2006).
The use of kinetic samplers has increased since the early 2000s. Their efficacy has been boosted by the use of performance reference compounds introduced into the sampler to enable adjustment of field data from the samplers by comparison with kinetic data from the laboratory. This assumes that uptake over the exposure period is linearly related to the exposure concentration (first-order diffusion model, including resistance at the boundary layer).
A number of different types of kinetic samplers have been developed to target particular groups of substances; for example, semipermeable membrane devices (SPMD), polar organic chemical integrative samplers (POCIS) and diffusive gradients in thin-films (DGT) techniques.
Passive samplers are less influenced by short-term fluctuations in concentrations than spot sampling. They can be used to identify sources of contaminants where extremely low levels have to be detected or when the input and presence of the contamination is not constant.
Mortimer et al. (2014) summarised how these samplers work and when to use them.
Automatic samplers, described previously, can perform as integrating samplers, although they are generally designed to collect a number of discrete samples rather than a fully integrated sample; they can miss pulses of contaminant.
Monitoring of the quality of groundwater involves techniques different from those used for surface water investigations. Groundwater, by its very nature, cannot be sampled without some disturbance from the construction of a bore or other access hole and the effects of sampling devices and procedures (NEPM 2011, USEPA 2013).
Sampling techniques and equipment may cause chemical and biological contamination unless stringent precautions are taken. Sampling staff must make extreme effort to ensure that the samples are representative of the water in the aquifer. Groundwater sampling should generally be carried out by more experienced field staff or in close consultation with experts, to ensure sample integrity.
To retrieve a representative sample, consider these principles (QDME 1995):
- Sampling equipment should not change the water quality in any way; particular effort should be made to avoid cross contamination between bores and sampling equipment.
- Sufficient water should be removed to ensure the sample is newly derived from the aquifer itself rather than from water that sits in the bore.
- Methods of collection and storage in bottles and transportation to the laboratory should suit the type of analysis required.
For guidance, refer to:
- Australian and New Zealand Standard for Groundwater Sampling (AS/NZS 5667.11:1998 (R2016)
- National Environment Protection (Assessment of Site Contamination) Measure 1999.
Groundwater sampling may produce a large volume of purged water. If necessary, this should be stored onsite in drums for proper disposal.
Low-flow groundwater sampling generates low volumes of purged water, which is recommended in most groundwater sampling situations.
To gain representative samples without any groundwater purging, use other sampling techniques, such as passive sampling and no-purge groundwater sampling.
QDME (1995) commented on different types of groundwater sampling:
- Displacement pumps (positive or gas bladder displacement, mechanical displacement) provide a gentle pumping action suitable for purging. Produce a high quality sample for all reasonable purposes. If the pump is made from sufficiently high quality materials, then the types of media used to work the pump are not important. These pumps are very suitable for well purging, although they are often slow. Specific design is very important in performance.
- Submersible pumps are available in several varieties, from very cheap to quite expensive. They rely on a motor below water level powering a pump to push water to the surface via a delivery line. These pumps are efficient in purging bores and can produce an acceptable sample for most purposes, although minor gases may be lost.
- Low-flow pumps (refer to Schalla et al. 2001).
- Suction pumps (centrifugal) are an excellent method for purging wells, but limited to about 6 m depth. The ‘suction’ type action has little effect on the water remaining in the well, although the samples pumped may lose some gases or organic compounds. After purging with a suction pump, high quality samples can be taken with balers, with care.
- Down-hole grab samplers are of little use in subartesian bores. Can obtain high quality samples from different depths in flowing bores. Quality control must be exercised in the transfer of samples.
- Balers are difficult or impossible to use for purging bores. Require extreme care to prevent contamination because samples have to be pulled to surface. Rope needs to be sterilised, cleaned or replaced frequently. Good quality samples can sometimes be obtained if bores are purged by other approved means. Quality control must be exercised in transfer of samples.
- Air-lifts are generally considered to be a poor method of obtaining high quality samples. The air-lifting method strips gases and organic compounds from the water, changes pH and may cause minor chemical changes. It is efficient in purging bores, and has little if any effect on major ion chemistry. Results of analyses of air-lifted samples are generally considered acceptable for monitoring major ions. Samples should not be submitted for higher quality types of analyses. This method is not considered suitable for well purging for high quality samples even if sampled with a baler. Air-lifting has an effect on the whole water column and the whole volume would have to be removed before a high quality sample could be obtained.
Collection of deposited rain, snow and airborne particulates is not covered in the Water Quality Guidelines in detail but we do describe analytical methods applicable to these types of samples.
Refer to guidance in the Manual on Water-Quality Monitoring: Planning and Implementation of Sampling and Field Testing (WMO 1988). For additional information, consult the Australian Government Bureau of Meteorology and other relevant jurisdictional agencies that undertake sampling for these measurement parameters.
Sediments are often surveyed to determine their composition and the concentration of contaminants, as well as the numbers of organisms located at various depths. There are 2 broad-based sediment classifications:
- Suspended sediments, in water quality terms, are generally dealt with as part of the water column, although specialised sampling techniques are required for obtaining representative samples (USEPA 1991, AS/NZS 5667.12:1999 [R2016]).
- Bottom sediments allow investigation of benthic organisms as measures of aquatic health, pollution or contamination, and as part of the ecology of aquatic systems.
Here we are only concerned with sampling the sediment, not the benthic organisms in the sediment.
Your monitoring team must decide what to include in the sample before beginning sampling. For example, when the objective is to sample sediments from within a seagrass bed, it is normal practice to remove the rhizomes of the seagrasses and sieve sediments to remove small molluscs.
We recommend that sampling for sediments should follow internationally accepted protocols and procedures. Your choice of sampling method will be dictated largely by the nature of the investigations being undertaken. It may be necessary to consult with relevant jurisdictional agencies about their internal guidance and guideline values for sampling sediments.
Sediment sampling methods that can be used both for biological and for nonbiological parameters include:
- corer method (EPAV 1992)
- core samplers (WMO 1988)
- dredging method (APHA 2012)
- grab method (EPAV 1992)
- grab samplers (WMO 1988)
- integrating samplers (EPAV 1992)
- Method 9060A Sample collection (APHA 2012)
- Method 10500B Soft-bottom dredge (APHA 2012)
- Method 10500B Hard-bottom dredge (APHA 2012)
- Method 10500B Rocky-bottom samplers (APHA 2012)
- protocols for sampling (Environment Canada 1995).
Sampling for determination of sediment transport is not covered in the Water Quality Guidelines.
For most applications, sediment coring is recommended (Simpson & Batley 2016, Mudroch & Azcue 1995). With this technique, samples can be taken to a measurable depth and then subsampled to provide depth profile information. Corers generally vary from 2.5 to 5.0 cm in diameter, and range from long PVC pipe (2 to 3 m) that can be immersed in shallow waters from a boat to shorter Perspex, polycarbonate or other tubes that can be immersed by hand by divers. Tubes have a bevelled leading edge to ease their movement through the sediment.
In shallow waters, cores can be extruded by gas pressure and subdivided, but this is not recommended where sediment oxidation is an issue.
If divers are not available, vibrocorers are essential for use from vessels in waters deeper than about 3 m. Vibrocorers usually contain plastic liners that protect the sample from contamination. Some corers enable in situ freezing of the sample.
A grab sampler or dredge is a useful alternative for obtaining large volumes of surface sediments. Most types of dredge sampler are suitable for sampling in shallow water depths (< 20 m). In deeper waters, take care to ensure that fines are not lost during the passage of the sampler to the surface (APHA 2012) because it is these particles that are most enriched in trace contaminants (Mudroch & Azcue 1995).
If chemical forms of contaminants and their associations with sediment phases are to be determined, then you must ensure that the redox state of the sediments (oxic or anoxic) is not altered because oxygenation (or reduction) will cause irreversible changes.
Sediments become oxygenated on contact with air so sediment samples must be capped immediately at sampling and stored in a nitrogen glove box. Oxidation can be minimised if samples are frozen at –20°C (Simpson & Batley 2016).
Choosing sample containers
Your monitoring team must decide on appropriate sample containers (type, volume) and how to clean them before use. A sample container can affect the sample’s composition by adsorbing some of its constituents (Batley 1989). For example, glass containers tend to adsorb phosphate.
Containers can be a source of contamination unless they are carefully prepared. Metals can be present at trace concentrations both on glass and on plastic surfaces, while organic compounds are more likely to be found on plastic containers.
Bacteria on container walls may use nutrients from solution (Maher & Woo 1998).
Caps of containers often contain inserts of cardboard, cork or rubber that should be removed because they can cause contamination.
A range of plastic materials have been used and their propensity for sample contamination has been thoroughly reported (Hunt & Wilson 1986, Hall 1998, Reimann et al. 1999).
Preferred sample containers for metals are fluorocarbon polymers, Teflon or fluorinated ethylene propylene (FEP), as well as high-density polyethylene. Bottles made from FEP are usually only used for mercury analysis because they are so costly.
If your budget permits, select high quality bottles with good closures that prevent sample leakage.
For samples to be analysed for selenium, bottles made of polycarbonate and some types of polyethylene are not suitable.
For nutrients, polyethylene (low density or high density) sample bottles are preferred. Glass is not favoured because there can be high concentrations of trace metals in the glass and there is potential for adsorption losses.
Preparation and cleaning
Some degree of cleaning is usually applied to bottles before they are used. This often involves soaking in acid but the rigour of the procedure varies from laboratory to laboratory.
Some authors have advocated direct use of certain bottle types without any cleaning (Reimann et al. 1999). Such sweeping statements are ill-advised because the quality of sample bottles often changes between batches.
In some laboratories, the minimum cleaning procedure involves soaking sample bottles in 10% nitric acid for at least 24 hours and then rinsing them with copious quantities of deionised water. Such precautions are worthwhile on most occasions, given the cost of sampling (especially if helicopters or boats are involved in the sampling programs). Acid washing is sometimes carried out in a dedicated dust-free room, and the acid baths are stored in a bunded, well-ventilated area similar to a very large domestic shower recess.
Emptied bottles are double-bagged using 2 zip-lock polyethylene bags.
For waters for zinc analysis, Ahlers et al. (1990) advocated that Nalgene™ bottles be soaked in hot 50% nitric acid for 2 days, rinsed with high purity water, then leached in 1% nitric acid for 2 weeks. The value of such extreme care was clearly demonstrated in the reliability of the resultant analytical data.
Sampling staff should always check with their analytical laboratory to ensure bottles have been appropriately prepared before use.
Following sampling protocols for trace analysis
When sampling waters containing trace metals, nutrients or organic compounds, a single person wearing plastic disposable gloves and taking appropriate care can carry out the operation without sample contamination if they are alert to potential sources of contamination. Generous use of polyethylene sheeting to wrap equipment and cover work areas on boats and river banks is a good practice.
Dust, powder, skin and hair are obvious external sources of metals, and rigorous care is required to minimise their effects.
Detailed protocols for ultratrace sampling are described in the literature (Ahlers et al. 1990, Nriagu et al. 1993, Nolting & de Jong 1994, Apte et al. 1998).
The recommended sampling protocol for ultratrace analysis uses the ‘dirty hands–clean hands’ approach. This involves 2 people wearing powder-free polyethylene gloves sequentially unwrapping double-bagged bottles. The first ‘dirty’ assistant removes the outer bag and hands the bottle to the ‘clean’ assistant, who removes the inner bag. The ‘clean’ assistant then immerses the bottles by hand or from the end of a pole, or fills them with the sample collected using the depth sampler. Bottles are usually rinsed with sample material first — filled, capped, shaken and emptied — before being refilled with the sample to be analysed.
Other precautions that help avoid contamination include:
- storing reagents for use in the field in decontaminated containers
- transporting sample and reagent containers in separate sealed plastic bags
- pre-cleaning all field equipment to the same standard as the containers
- uncapping or removing containers from their transport bags for minimum amounts of time
- emptying containers that were filled with water as part of the preparation protocol well away from and downstream of the sampling location before rinsing and refilling them with sample.
Aquatic organism samples
For aquatic organisms, selection of a sampling method should be guided by:
- objectives of the monitoring program
- local conditions (the need to obtain representative samples)
- safety of operation (safety of the sampling staff is critical)
- acceptability of the method
Choosing an appropriate method
Your monitoring team will have decided which organisms to collect at the study design stage, and now must choose the appropriate equipment and procedures to use.
Methods for sampling organisms typically sampled are described in APHA (2012) and Hellawell (1986), and include:
- algae — grab samples/nets, hose-pipe samplers (APHA 2012, Falconer 1994, Hotzel & Croome 1998)
- bacteria — grab sample (APHA 2012, Ward & Johnson 1996)
- benthic macroinvertebrates and algae — bottom grabs/samplers, diver-held cores/nets (APHA 2012, Growns et al. 1999)
- bivalves — cage, basket sampler, by hand (APHA 2012)
- crayfish and shrimps — gee-traps, net traps (APHA 2012)
- fish — nets, traps, electro-fishing (APHA 2012, Harris & Gehrke 1997)
- fungi — grab samples (APHA 2012)
- macroinvertebrates — sweep nets, hand search, quadrats, long-handled pole with net (deep waters)
- macrophytes — various (APHA 2012)
- periphyton — artificial and natural substrates (ice block sticks, modified brushes) (APHA 2012)
- plankton — grab sample/plankton (cone) nets, hose-pipe sampler, Patalas–Schindler plankton trap (APHA 2012, Hellawell 1986)
- protozoa — grab samples (APHA 2012).
We provide more specific guidance in Protocols for Biological Assessment.
Several devices may be needed to ensure quantitative sampling of all required organisms. You may need to compromise between quantitative sampling and more rapid methods.
In a comparison of the efficiency of 3 sampling devices (tube sampler, vertical tow net, Schindler–Patalas trap) for collecting zooplankton, Devries & Stein (1991) found there was no best method. Zooplankton consist of a mixture of copepods, cladocerans and rotifers. Generally copepods and cladocerans were best collected using the tube sampler, while rotifers were best collected using the Schindler–Patalas trap. Some species were best collected using the vertical tow net.
Selection of devices should account for the fact that all biological sampling devices have biases of some sort. Ensure that the biological indicators required are collected with minimal bias, or at least known and consistent bias, between sites and seasons.
Decide which biota are to be collected before sampling begins. Specialist aquatic biologists should be consulted for specific guidance about the best sampling method given the circumstances of concern. As indicated above, there is no universal method for collecting biota. If a range of biota is collected, several sampling procedures can be needed.
Sample preservation and storage
Clear and distinctive sample labelling is important, especially when samples are collected for later chemical or biological analyses.
After collection, it is important to maintain the integrity of each sample and to ensure that it does not become contaminated or change between collection and analysis. It is usually necessary to preserve the samples to retard biological, chemical and physical changes.
Protocols must specify both the appropriate sample container and the preservation technique.
Preservation choices vary depending on the parameter to be measured. Suitable preservation strategies or storage procedures to account for some possible changes are given in Table 1.
Table 1 Preservation and storage strategies for physical, chemical and biological samples
|Type of change||Change||Preservation techniques|
|Biological||Cell degradation||Freeze, add fixing agent (e.g. ethanol, glutaraldehyde, glyoxal)|
|Microbial action||Reduce pH, filter, add bactericide (e.g. for sulfide add zinc acetate, if chlorine present add thiosulfate, leave small airspace to reserve viability, avoid light, refrigerate at 4°C)|
|Chemical||Photochemical action||Use dark containers|
|Precipitation||Lower pH, avoid use of chemicals that cause precipitation (e.g. sulfates)|
|Speciation||Refrigerate at 4°C|
|Physical||Adsorption/absorption||Inorganic (reduce pH on storage)|
|Diffusion||Choose correct container type and cap liners|
|Volatilisation||No head space|
Matters for consideration to ensure successful preservation and storage include:
- selection and decontamination of sample containers
- selection of a preservation technique
- time lapse acceptable between sample collection and analysis.
Choices available will depend on the variable to be measured. Comprehensive information on the selection of containers and preservation of water samples for chemical and microbiological analysis can be obtained by consulting the Australian and New Zealand water quality standards (AS/NZS 5667 series).
Complete and unequivocal preservation of samples is a practical impossibility. At best, preservation techniques only retard the chemical and biological changes that inevitably take place after sample collection.
Normally, to prevent chemical and biological changes, water samples are cooled to 4°C, frozen, filtered or given a chemical additive.
Freezing (–10°C) reduces but does not eliminate biological activity in samples. All biological activity is only effectively eliminated at –40°C and below. Chemicals such as chloroform and mercury (II) acetate have been used to prevent biological activity.
Acid is often added to prevent adsorption of metals from water samples to containers and precipitation of insoluble salts.
Chemical preservatives should be avoided, if possible, because they may contaminate samples or interfere in chemical or biological analysis, and they can be harmful to people. One example of this is formaldehyde which is a carcinogen so has been replaced by glyoxal or glutaraldehyde (Huang & Yeung 2015). Another is mercury which can interfere in the colorimetric determination of phosphate.
If preservatives are used, then they should be taken into account in the analysis of blanks.
Even if a sample is frozen or a preservative is added, samples can be stored only for a finite time. In some cases, this period may be years (e.g. phosphorus in seawater). In other cases, it may be much shorter (e.g. 6 hours for Escherichia coli samples).
The preservation time needs to be determined before samples are collected, and protocols must be developed to ensure that samples are analysed before a significant change in composition occurs.
Quality assurance and quality control
A quality assurance and quality control (QA/QC) program for field sampling is intended to control sampling errors at levels acceptable to the data user and provide evidence that this has occurred.
A QA/QC program includes procedures designed to prevent, detect and correct problems in the sampling process and to characterise errors statistically, through quality control samples.
Major errors to be avoided are:
- faulty operation of the sampling device
- changes in the sample before measurement (contamination, chemical or biological)
- incorrect sample labelling.
Field staff should be competent in sampling and making field measurements even though they may have other important qualities unrelated to the assurance of sample integrity (e.g. vehicle handling or bush skills).
The importance of appropriate training of sampling staff cannot be understated.
Before staff are permitted to do reportable work, they should demonstrate competence in field sampling procedures. At a minimum, this would include being able to adhere to protocols, avoid contaminating samples calibrate field instruments and make field observations.
Unfortunately it is common to find that the task has been assigned to inexperienced or insufficiently trained staff, resulting in poor quality monitoring data. In many such cases, the sampling team is unable to determine when loss of quality has occurred, which can result in years of worthless monitoring data being collected, at very substantial cost.
All equipment and field instruments should be kept clean and in good working order, and calibrations and preventative maintenance should be recorded carefully. All repairs to equipment and instruments should be noted, as well as any incidents that could affect the reliability of the equipment.
When automatic sampling devices are used, their timing mechanisms must be calibrated to ensure that the samples are acquired at the specified intervals. This is especially important where hydrological or other conditions result in significant short-term concentration variations.
Tracking samples and field data
During sampling or field measurements, it is important to complete a field data sheet or similar record that describes the samples taken, their labels and other details (refer to our field sampling record sheet example). Record all field data and instrument calibration data on this sheet before leaving a sampling station.
Any observations or information on the conditions at the time of sampling that may assist in interpretation of the data should be noted on a field record sheet or in a field notebook. This information may explain unusual data that otherwise might be attributed to problems in sampling or analysis.
If samples are to be the basis for future legal proceedings, questions likely to be asked include:
- Exactly where was the sample taken?
- Was the person who took the sample competent to do so?
- How was the sample labelled to ensure no possibility of mix up or substitution?
- Was there any possibility of contamination (e.g. of the container) of the sample during filling, or later?
- Did the sample deteriorate after collection?
Chain of custody documentation (Table 2) ensures that these questions can be answered.
Table 2 Example of chain of custody documentation
|Process step||Quality assurance procedure|
|Field sampling||Field register of sample number, site, type/technique, time, date, technician, field data sheet|
|Sample storage and transport||Field register of transport container and sample numbers, time, date|
|Laboratory receipt of samples||Laboratory register of transport container number and samples numbers, time date|
|Laboratory storage of samples||Laboratory register of storage location, type, temperature, time, date|
|Sample preparation||Analysis register of samples (laboratory) number, pre-treatment, date, technician|
|Sample analysis||Analysis register of instrument, calibration, technician, standard method, date, result|
Documenting sampling protocols
Sampling errors can be minimised by ensuring that correct procedures have been followed during the field sampling, transport and storage.
Develop and adhere to sampling protocols for each matrix and constituent that must specify:
- detailed procedures for collecting, labelling, transporting and storing samples
- detailed procedures for collecting, transmitting and storing ancillary field data
- sample collection device
- type of storage container
- sample preservation procedures
- types and numbers of quality control samples to be taken.
Before such a protocol can be written, the nature of errors (systematic and random) and the level of accuracy desired must be assessed.
Common sources of error include:
- reactions with the sample or the sample container
- contamination (field, sampling device, containers)
- chemical and physical instability
- biological changes.
The exact locations of sampling sites and any subsites must be recorded in the sampling protocol.
Field notes must accurately describe where samples were collected, to allow crosschecking with the sampling locations specified in the sampling protocol. If transects are to be sampled, then the location range must be specified if this is within the precision of the positioning instrument.
Noting the time when samples are taken (standard or daylight-saving time) is an obvious but frequently overlooked requirement of rigorous sample definition.
Protocols should specify how sampling staff are to be trained to use sampling equipment.
Anticipate problems that may occur in the field. Sample containers may be lost. Sample volumes may be low. Should foreign objects be included? On the basis of what criteria is foreign matter rejected? What happens if sites cannot be sampled?
One of the major challenges of sampling is to prevent contamination.
Protocols must include basic precautions for avoiding contamination:
- Field measurements should be made on separate subsamples of water.
- New or reused sample containers must be appropriately cleaned (use of containers supplied by the analytical laboratory is recommended).
- Only the sample bottles recommended for each parameter should be used.
- Container lids should be checked for liners that may cause contamination or adsorb particular analytes.
- Containers that have already been used for other purposes should be discarded.
- Insides of containers and lids should not come in contact with hands or objects.
- Sample containers and filter units should be kept in a clean environment away from potential contaminants, including dust, dirt and fumes.
- Preservatives should be tested for contamination.
- Care should be taken to avoid cross contaminating samples when adding preservatives.
- Sample containers used for collecting samples for microbiological analyses must be sterilised.
- Sampling staff should use plastic disposable gloves when handling sample containers at every stage during sampling (to avoid touching the sample and the insides of caps or containers).
Dealing with blanks and quality control practices
Blanks to check on field procedures, containers, equipment and transport
If contamination is possible during the sampling process, then appropriate sets of blank samples should be used to detect and measure the contaminant sources.
Field blanks mainly detect contamination from sample handling, dust and other atmospheric fallout. Take extra containers (with suitable contents) to the site for field blanks. At the site, the containers are opened and closed and the contents are handled just as if these were real samples during transfer and storage. Use sample bottles filled with deionised water as field blanks for freshwater sampling, and water of the appropriate salinity for marine water sampling. Prepare 1 blank per 10 samples, adding any preservative in the field.
Filter blanks allow estimation of contamination by filtration in the field. They are prepared in the field by passing a sample of distilled water through a pre-cleaned filter and adding preservative to the water sample.
Container blanks determine the contamination from the container. Containers of each type to be used for sampling (about 1 in 10) are selected at random and filled with deionised water and preserved in the same manner as field samples. Analysis of these blanks detects contamination by the container washing process. If this is measured as a rinse blank, then the last of several distilled water rinses of sampling equipment in the field is analysed.
Equipment blanks measure contamination introduced through contact with sampling equipment or a sampler. They consist of the water or solvent that is used to rinse the sampling equipment between samples.
Trip blanks can be used to assess gross cross contamination of samples during transport and storage. The simulated samples are similar to the samples to be collected but the analyte (substance to be measured) is at background or low concentrations.
Often it is not possible to achieve no contamination. Aim for stable contamination levels instead. When levels of contamination are outside the agreed acceptable limits, the contamination is likely to be coming from new sources.
Checks for QA/QC are partly reactive. If changes in samples are detected by using standard additions or blanks, a specified procedure is devised to determine and rectify the problem. The water is re-sampled if possible.
Another sample QA/QC measure is to use multiple samples.
Duplicate samples reveal the magnitudes of errors (contamination, random and systematic) occurring between sampling and sample analysis.
Obtain duplicate samples by:
- dividing a sample into 2 or more subsamples
- collecting 2 or more samples simultaneously to establish the reproducibility of sampling.
Ideally, 3 samples are required to enable testing of inter and intralaboratory accuracy and precision.
Another alternative is to ‘spike’ subsamples in the field to detect change. Add a known amount of the analyte to the subsample and then measure it. Samples for QA/QC should be labelled so they are indistinguishable from other samples in the batch.
Quality control in biological sampling
The main question to be addressed for biological sampling is whether or not it is quantitative and representative. That the sampling device has been used in a consistent manner to ensure equal representativeness between replicates is often a critical concern. Alternative sampling strategies need to be devised and tested to establish the suitability of any preferred sampling technique.
Quality control in data storage and access
Transfer of results from the field to a database should be automated where possible, and the printout of the entry should be checked against the field record sheet and the laboratory register.
Entries can be validated by electronic screening against the expected range and against other analytes for the same site and sampling date, and against field measurements.
Follow agreed procedures for handling and tracking updates and corrections to data, including provision for handling censored data (refer to Data analysis). There should be fields for all necessary identifiers, for traceability purposes (e.g. sample and laboratory numbers).
With respect to security, specify those personnel who have read-access or write-access to the data. Daily data backup is always essential in case of system or file failures.
Work health and safety
Hazards or risks involved with field sampling need to be identified and documented on a preliminary site visit, to resolve some important questions:
- Can staff reach the site in safety?
- Can a sample be safely taken? Is the water fast flowing? Is a boat to be used? Is there safe boat access? Is the site prone to flash floods? Is the bank stable? Are tidal changes likely?
- Will sampling staff be exposed to toxic or other hazardous substances?
- Will sampling staff be exposed to any pathogens? (e.g. Ross River virus, malaria)
- Will any potentially dangerous fauna be encountered? (e.g. spiders, ticks, snakes, leeches, crocodiles, sharks, pigs)
- Are weather conditions likely to endanger personal safety? (In alpine areas especially, weather patterns are extremely variable.)
Personnel who are to conduct sampling should be physically and mentally able to undertake field work. For example, if sampling staff fall into a water body, they must be physically fit enough to get out without assistance (although staff should never work alone in the field).
Sampling staff working near water must be able to swim and climb up river banks.
In proper professional practice, risks must be reduced as much as possible, and staff must not be required to operate in conditions that are unsafe.
Educating about hazards
All staff must be appropriately trained as part of the formal risk-minimisation strategy. Training will include:
- familiarisation with environmental hazards that may be encountered
- familiarisation with sampling protocols (e.g. sampling procedures, chain of custody considerations)
- use of sampling equipment
- qualifications to drive appropriate vehicles (e.g. off-road 4-wheel-drive vehicles, bikes, tractors or boats)
- familiarisation with safety procedures
- qualifications in advanced first aid.
Developing risk minimisation plans
Include clear directives in a risk minimisation plan to reduce risks during sampling operations.
Limit continuous driving. If sampling sites are at a considerable distance, do not drive there without a stop. Take breaks of at least 15 minutes every 2 hours, and sample for no more than 10 hours in one 24-hour period.
Choose safe sites with safe access. Visit potential sites and check them after they have been tentatively selected from map surveys. They should have reasonable access, be free of dangerous animals or prickly or poisonous plants, have no steep, slippery or unstable banks, and not be prone to rapid flooding or tidewater rise without warning.
Wear appropriate clothing. Obtain weather forecasts for an area to be sampled. Be prepared, for example, to wear raincoats if there is likelihood of rain, warm clothing if it is cold, hat and sunscreen at all times, and footwear with a good grip for wet rocks (do not go barefoot, risking injury from sticks or broken glass). Be aware that sunscreen can be a source of contamination and should be used with due care for this reason. Take extra clothes and a towel in case someone falls in the water.
Take appropriate safety gear and a first aid kit. Wear life jackets when sampling near deep water with poor footing or from a boat. Plastic gloves are essential to anyone who has an open or bandaged wound when handling chemicals or contaminated water, or even if the water quality at the site is unknown. Take a fully stocked first aid kit to the monitoring site. Ideally, someone in the monitoring team should have first aid training.
Maintain contact with help and never sample alone. Work with at least 2 others and stay in contact with someone who can raise the alarm. Carry a mobile phone and car charger for emergency telephone calls. In remote areas, carry maps, compass, mirror and waterproof matches and inform a responsible person of intended movements. There must be written procedures describing how emergency services are to be accessed.
Never wade into deep water. Sampling in deep water requires the use of an appropriate boat with the necessary safety equipment (life jackets, flares). Preferably, sample from a bridge or use a cableway if installed at the site.
Avoid contact with contaminated water. Carry drinking water. Do not drink from the source being monitored. Always wear plastic gloves when water quality at the site is unknown, particularly when collecting samples in which the presence of algae, pathogenic organisms or toxins can be expected. (Blue–green algae can cause skin and eye irritations.) Wash hands after monitoring and before eating. Treat all bacterial cultures as pathogenic.
Professional practice requires sampling staff to:
- obtain approval as required, such as permits to collect fauna and flora or take water samples
- have access to sites (landholders’ permission may be required to enter private land)
- use appropriate etiquette.
It is good practice to inform local authorities and park rangers about your sampling program, even if formal permission is not required. Local people can give useful information that helps in the choice of safe sampling locations and warns of local hazards.
Individual sampling staff have a duty of care to other personnel, including:
- if one person cannot carry out all aspects of field work, then colleagues must assist
- there should be no discrimination
- privacy of individuals should be respected.
There is a duty of care to avoid damaging the environment during sampling:
- do not litter
- observe fire restriction requirements
- do not wash in waterways (streams, lakes, estuaries)
- remove human wastes
- do not feed native animals
- minimise environmental damage by keeping to paths, tracks and roads.
Ahlers WW, Reid MR, Kim JP & Hunter KA 1990, Contamination-free sample collection and handling protocols for trace elements in natural fresh waters, Australian Journal of Marine and Freshwater Research 41(6): 713–720.
Alvarez DA, Huckins JN, Petty JD & Manahan SE 1999, Progress toward the development of a passive in situ SPMD-like sampler for hydrophilic organic contaminants in aquatic environments, presented at 20th Annual Meeting of Society of Environmental Technology and Chemistry, November 1999, Philadelphia.
APHA 2012, Standard Methods for Examination of Water and Wastewater, 22nd Edition. American Public Health Association, Washington DC.
Apte SC, Batley GE, Szymczak R, Rendell PS, Lee R & Waite TD 1998, Baseline trace metal concentrations in New South Wales coastal waters, Marine and Freshwater Research 49(3): 203–214.
Batley GE 1989, Collection, preparation and storage of samples for speciation analysis, in: Batley GE (ed.), Trace Element Speciation: Analytical Methods and Problems, pp. 1–24, CRC Press Inc, Boca Raton.
Batley GE, & Simpson SL 2016, Sediment sampling, sample preparation and general analysis In Sediment Quality Assessment: a Practical Guide. CSIRO Press, Clayton, Vic, Australia. http://www.publish.csiro.au/book/7383/
Blomqvist S 1991, Quantitative sampling of soft-bottom sediments: problems and solutions, Marine Ecology Progress Series 72: 295–304.
Crompton TR 1989, Analysis of Seawater, Butterworths, London.
Davison W.(ed) 2016, Diffusive gradients in thin-films (DGT) for environmental measurements. Cambridge University Press, Cambridge, UK.
Davison W & H. Zhang H 1994, In-situ speciation measurement of trace components in natural waters using thin-film gels, Nature 367: 546–548.
Devries DR & Stein RA 1991, Comparison of three zooplankton samplers: a taxon-specific assessment, Journal of Plankton Research 13: 53–59.
Environment Canada 1995, Guidance Document on Collection and Preparation of Sediments for Physicochemical Characterisation and Biological Testing, Report EPS 1/RM/29.
EPAV 1992, Guidelines for the Design of Water Quality Surveys, Environmental Protection Authority Victoria, Melbourne.
Falconer IR 1994, Comparative evaluation of methods for testing the toxin content of cyanobacterial blooms and related water supplies, in: (Steffensen DA & Nicholson BD (eds), Toxic Cyanobacteria: Current Status of Research and Development: Proceedings of an International Workshop, Adelaide, 22–26 March 1994.
Freimann P, Schmidt D & Schomaker K, 1983, Mercos — A simple Teflon sampler for ultratrace metal analysis in seawater, Marine Chemistry 14(1): 43–48
Grasshoff K, Kremling K & Ehrhardt M 1999, Methods of Seawater Analysis, 3rd Edition, WileyVCH, Weinheim.
Green RH 1979, Sampling Design and Statistical Methods for Environmental Biologists, John Wiley and Sons, New York.
Growns JE, King AJ & Betts FM 1999, The Snag Bag, a new method for sampling in situ macroinvertebrate communities on large woody debris, Hydrobiologia 405: 67–77.
Hall GEM 1998, Relative contamination levels observed in different types of bottle used to collect water samples, Explore 101: 3–7.
Harris JH & Gehrke PC 1997, Fish and Rivers in Stress: The NSW Rivers Survey, NSW Fisheries Office of Conservation and the Cooperative Centre for Freshwater Ecology.
Harris DJ & Keffer WJ 1974, Wastewater Sampling Methodologies and Flow Measurement Techniques, US Environmental Protection Agency, Kansas City.
Hellawell JM 1986, Biological Indicators of Freshwater Pollution, Elsevier, London and New York.
Hotzel G & Croome R 1998, A Phytoplankton Methods Manual for Australian Rivers, Occasional Paper no. 18/98, Land and Water Resources Research and Development Corporation, Canberra.
Huckins JN, Manuweera GK, Petty JD, Mackay D & Lebo JA 1993, Lipid-containing semipermeable membrane devices for monitoring organic contaminants in water, Environmental Science & Technology 27(12): 2489–2496.
Huckins JN, Petty JD & Booij K 2006, Monitors of Organic Chemicals in the Environment: Semipermeable Membrane Devices, Springer, New York.
Huckins JN, Petty JP, Prest HF, Clark RC, Alvarez DA, Orazio DC, Lebo JA, Cranor WL & Johnson BT 2002, A Guide for the Use of Semipermeable Membrane Devices (SPMDs) as Samplers of Waterborne Hydrophobic Organic Contaminants, American Petroleum Institute publication 4690.
Hunt DTE & Wilson AL 1986, The Chemical Analysis of Water, General Principles and Techniques, Royal Society of Chemistry, Cambridge.
Kingston JK, Greenwood R, Mills GA, Morrison GM & Bjorklund Persson L 2000, Development of a novel passive sampling system for the time averaged measurement of a range of organic pollutants in aquatic environments, Journal of Environmental Monitoring 2: 487–495.
Lane SL, Flanagan S., Wilde FD, Radtke DB, Gibs J. & Iwatsubo RT (eds) 2003, Selection of equipment for water sampling. Chapter A2 in National Field Manual for the Collection of Water-Quality Data U.S. Geological Survey Techniques of Water-Resources Investigations Book 9, Denver CO, USA.
Maher WA & Woo L 1998, Procedures for the storage and digestion of natural waters for the determination of filterable reactive phosphorus, total filterable phosphorus and total phosphorus, Analytica Chimica Acta 375(1–2): 5–47.
Mortimer M, Müller JF & Liess M 2014, Sampling methods in surface waters, in: Nollet LML & De Gelder LSP (eds), Handbook of Water Analysis, 3rd Edition, CRC Press.
Mudroch A & Azcue JM 1995, Manual of Aquatic Sediment Sampling, CRC Press, Boca Raton.
Müller J, Krishnamohan M, Mortimer M & McLaughlin MS 2001, Partitioning of polycyclic aromatic hydrocarbons in the polyethylene/water system, Fresenius' Journal of Analytical Chemistry 371(6): 816–822.
Nolting RF & de Jong JTM 1994, Sampling and analytical methods for the determination of trace metals in surface seawater, International Journal of Environmental Analytical Chemistry 57: 189–196.
National Environmental Protection Measure 2011, Guideline on site characterisation. Schedule B2 in National Environment Protection (Assessment of Site Contamination), Canberra ACT, Australia.
Nriagu JO, Lawson G, Wong HKT & Azcue JM 1993, A protocol for minimizing contamination in the analysis of trace metals in Great Lakes waters, Journal of Great Lakes Research 19: 175–182.
QDME 1995, Technical Guidelines for the Environmental Management of Exploration and Mining in Queensland, Queensland Department of Mines and Energy.
Reimann C, Siewers U, Skarphagen H & Banks D 1999, Does bottle type and acid-washing influence trace element analyses by ICP-MS on water samples? Science of the Total Environment 239: 111–130.
Schäfer RB, Hearn L, Kefford BJ, Nugegoda D & Mueller JF 2010, Using silicone passive samplers to detect polycyclic aromatic hydrocarbons from wildfires in streams and potential acute effects for invertebrate communities, Water Research 44(15): 4590–4600.
Schneider P & Wyllie SJ 1991, An efficient vibrocoring system for collecting coastal sediments: a comparison with other techniques, in: Coastal Engineering — Climate for Change, Proceedings of the 10th Australasian Conference on Coastal and Ocean Engineering, Auckland, 2–6 December 1991, pp. 357–362.
Schultz DE, Petrick G & Duinker DC 1988, Chlorinated biphenyls in North Atlantic surface and deep water, Marine Pollution Bulletin 19(10): 526–531.
Södergren A 1987, Solvent-filled dialysis membranes simulate uptake of pollutants by aquatic organisms, Environmental Science & Technology 21(9): 855–859.
USEPA 1991, Evaluation of Dredged Material Proposed for Ocean Disposal: Testing Manual, US Environmental Protection Agency and US Army Corps of Engineers, Report no. EPA-503/8-91/001, Washington DC.
USEPA 1996a, Compendium of ERT Surface Water and Sediment Sampling Procedures, Report no. EPA 540/P-91-005, US Environmental Protection Agency, Washington DC.
USEPA 1996b, Mercury in Water by Oxidation, Purge and Traps and Cold Vapour Atomic Fluorescence Spectrometry, Method 1631, Report no. EPA/821/R-96-001, US Environmental Protection Agency, Washington DC.
Ward AK & Johnson MD 1996, Heterotrophic microorganisms, in: Hauer FR & Lambert GA (eds), Methods in Stream Ecology, Academic Press, San Diego, pp. 232–235.
WMO 1988, Manual on Water Quality Monitoring — Planning and Implementation of Sampling and Field Testing, Operational Hydrology Report no. 27, Secretariat of the World Meteorological Organization, Geneva.