Proper sampling of ponds, lakes, reservoirs and estuaries is a complex and commonly misunderstood process.  One of the largest misconceptions is that a single sample will be representative of a still-water body.  The USGS's National Field Manual for the Collection of Water-Quality Data states it best, "The probability is small that any body of still water (lake, reservoir, pond, lock, storage pool) is relatively homogeneous with regard to any water-quality characteristic.  Therefore, a single sampling point generally is not adequate to describe the physical and chemical properties of the water body, or the distribution and abundance of the inhabiting biological community.  The number of sampling locations selected and the depths where samples will be collected depend on study objectives and the physical, chemical, and biological characteristics of the water body."

1.  Spatial Variability

Still-water bodies commonly exhibit three-dimensional water quality variations that result from a number of interrelated physical, chemical and biological processes.  These include circulation patterns; thermal, chemical and density-driven stratification; water-body geometry; climatic influences, such as sun and wind exposure across the water-body's surface; stream confluences; water-body outfalls and climatic seasonality.  The sampling design should always consider these variables.  This means that collecting samples along the water body's edge or from the surface are almost certainly NOT representative of the respective water quality.  The adjacent figure to the right (click to enlarge) illustrates a 35 acre lake with a dispersed sampling scheme.  Another method is to select regularly-spaced sampling points along one or multiple transects throughout the water-body.  For a pond, one centered sampling location may be sufficient, but this should not be assumed.  With respect to conducting a water quality baseline in response  to oil & gas drilling activities, it is desirable to consider the associated activities orientation, proximity and associated drainage characteristics with respect to the water-body when planning a sampling scheme.

2.  Depth-Dependent Variability

In addition to spatial variability, it is also necessary to consider depth-dependent variability.  Throughout the seasons, ponds, lakes and other similar water bodies undergo cycles of turnover and stratification.  The adjacent figure to the lower right (click to enlarge) illustrates this phenomena.

Winter

As water cools, it increases in density up to approximately 4 degrees C (39 degrees F).  Beyond this temperature, it becomes less dense, eventually forming ice.  This unique property allows ice to float on water; otherwise, water bodies would freeze from the bottom-up!  Therefore, as winter sets in, surface water is cooled below 4 degrees C, eventually forming ice cover, which prevents wind from circulating any lake water.  Very cold, less dense water (between 0 and 4 degrees C) resides immediately below the ice, while the remaining lake beneath is typically around 4 degrees C.  Although slight, stratification is present and the lake is said to be in a "stagnant" state.  Dissolved oxygen is present predominantly in the upper portion of the lake, while the lower portion exhibits significantly reduced levels of oxygen that can drop to zero (anoxic environment).

Spring

As spring approaches, ice cover melts from the surface, exposing a relatively homogeneous water-body.  Wind across the surface induces circulation, commonly known as "turnover", which helps oxygenate the water-body throughout.  During this period, the lake, pond, etc. is relatively homogeneous.  However, as the water surface warms, it begins to stratify or float on top of the colder, denser water below, which leads into summer.

Summer

As the surface water continues to warm, the upper layer of water (epilimnion) continues to thicken and becomes distinctly different from the principal layer at depth (known as the hypolimnion).  A relatively narrow region resides in between, called the metalimnion or thermocline, which is characterized by rapidly changing temperature with depth.  Drawing from our previously-discussed decreasing temperature/increasing density relationship, the increasing density within the metalimnion zone helps create a barrier, from which circulation does not pass below.  This means that for a few months out of the summer, the deep hypolimnion zone is isolated from water above and exhibits distinctly different characteristics.  It is colder, more dense and has less oxygen present.  In many cases, oxygen is completely absent.

Fall

As cooler weather lowers the temperature of the epilimnion, its density begins to increase and winds begin to mix the water-body at greater depth. Eventually, the upper and lower zones of water approach the same temperature and density, at which time turnover occurs and oxygen is once again circulated at depth.  The cycle continues...

Implications

Depending on the season, basin geometry & depth, locations of stream inlets & spring seeps and exposure to sun & wind, water-bodies exhibit varying degrees of stratification and therefore, need to be evaluated for such phenomena, to help guide appropriate sampling locations.  Summer months are especially of concern, since denser water with greater dissolved constituents is commonly present at depth.  Anoxic (oxygen deficient) conditions do often form in the lower hypolimnion zone during summer months.  Of particular interest, anaerobic decomposition of organic matter via. microorganisms often results in appreciable methane, carbon dioxide and hydrogen sulfide production.  Therefore, it should be not be surprising to find detectable levels of these gases under natural conditions, when sampled for.  These conditions also promote the mobilization of organometal(loid) compounds, the metals of which can originate from anthropogenic sources, including oil & gas drilling activities.  Many of them are also toxic.  For this reason, it is absolutely necessary for water quality baselines to identify the hypolimnion zone (if present) and incorporate its sampling into the baseline assessment.  If a follow-up water quality survey is conducted in response to suspected oil & gas drilling contamination, water quality depth profiling is an absolute, to identify and sample stratified layers that may contain appreciable contaminant concentrations.

1.  Collection of Spatial/Depth-Dependent Samples

Samples collected from still-water bodies must have documented locations and depths.  Also, the samples recovered from depth must remain in an unaltered state (no mixing with adjacent water above or below).  This is accomplished either by using a specialized water-sampling bottle or pump.

Thief-Type Water Sampling Bottles

The use of this sampling device is typically preferred due to its simplicity, modest cost and light weight and is capable of sampling thinly-stratified layers.  Construction materials vary; however, sampling for trace metals and organics requires special construction, such as polycarbonate, polyethylene and silicone materials.  The bottles come in vertical and horizontal configurations.  The horizontal configuration is generally preferred as it can collect from a finer-discrete interval and is shown in the illustration to the right.  The bottle is lowered to the desired sampling depth.  Afterwards, a messenger is sent down the tether line, which activates a spring-loaded mechanism and closes the bottle ends.  The bottle is raised and sampled via. the drain/sampling tube.  The bottles typically come in 2 to 5 liter volumes, so in the event that a large sample volume is required, it may be necessary to make multiple collection trips to a water stratum.

Pumps

There are a variety of pumps, including centrifugal, peristaltic and submersible, that are sometimes used for sampling still-water bodies.  The single largest advantage to using pumps is their ability to collect a large sample volume.  However, use of this equipment does allow for greater mixing between water strata, due to the fact that water moves inward from all directions towards the pump's intake.  This makes it very difficult to sample narrow stratum.  Furthermore, if the intake level is adjusted to sample another stratified zone, enough water must be cycled through the pump and tubing to properly flush the equipment.

2.  Sample Integration/Compositing

Separate analyses for samples collected both spatially and with depth throughout a still-water body is often times cost-prohibitive and unnecessary.  In certain circumstances, samples can be composited or combined and then analyzed to represent the bulk water quality.  In other cases, they can not.

Volatile Organics

Volatile constituents, which includes methane and wet natural gases, should be sampled at a discrete location and depth and not be composited due to their volatility.  The attempt at mixing and creating a representative sample would in all likelihood, result in the escape of volatiles and bias the associated laboratory analyses.  Investigators should be aware that anoxic conditions (typically present at lake bottoms in late summer & associated with the hypolimnion zone) are likely conditions for the generation of methane and other natural gases.  Organometal(loid) compounds, as previously discussed, can also be detected at times.  Proper baselines need to consider these conditions and should collect discrete samples from these zones, when present.  Collecting surface samples are unlikely to reveal significant (if any) volatile compounds due to atmospheric interaction.  Even lakes associated with heavy motoring have often revealed BTEX (benzene, toluene, ethylbenzene & xylenes) concentrations less than 1 ug/L (USGS WRIR 98-4264).

Inorganics

Compositing discrete inorganic samples is typically done with a device, commonly known as a churn-splitter.  As with the water sampling bottles, polyethylene and/or polypropylene construction materials without metal parts are used for proper sampling of trace metal constituents.  It is necessary to document the spatial and depth-dependent locations and associated volumes used for the composite sample.  If future sampling is to be conducted and compared, it must be replicable to allow for accurate comparison (apples vs. apples).  If samples are collected at stream confluences, it would not necessarily be desirable to include those locations as part of a lake composite sample.

Composite samples can be designed to represent the bulk water-quality of a water-body or alternatively, the water-quality associated with each observed stratified layer.  Naturally, the latter is more specific and would therefore require a greater number of samples to be analyzed by the laboratory.