Sampling Methods

Sampling methodologies


This section is divided into the various habitats in which hypogean Crustacea can be found, with suggestions of potential sampling methods given for each.

Many of the nets mentioned in this section can be obtained from the Natural History Book Service (NHBS) ( or from various other suppliers of ecological field equipment. Less expensive small hand nets can also be obtained from aquarist shops. Choosing the size of the net mesh aperture requires a balance between capturing smaller sized specimens and maintaining a good flow of water through the net, preventing it from becoming clogged up by silt and debris. This is not usually a major problem in most subterranean environments where organic matter is low. The standard Environment Agency kick sampling / pond net with a mesh of 1mm is likely to retain most larger specimens, but small Niphargus glenniei, bathynellids and juveniles might be lost. A mesh of 250μm is likely to be more suitable for hypogean Crustacea sampling. 


Nets for use in sampling hypogean aquatic habitats. A: Drift net. B: Pipe Net (with draw-string). C: Aquarist’s hand net. D: Zooplankton trawl net (with lead weights). E: Hand net (with optional extended handle section, more than one can be fitted if required).
Nets for use in sampling hypogean aquatic habitats. A: Drift net. B: Pipe Net (with draw-string). C: Aquarist’s hand net. D: Zooplankton trawl net (with lead weights). E: Hand net (with optional extended handle section, more than one can be fitted if required).


The best time to sample for hypogean crustaceans (and all groundwater fauna in general) is after heavy rainfall following a prolonged dry period. This tends to mobilise the groundwater fauna and wash them out of interstices and fissures in the rock strata into caves and wells etc. Repeat visits to wells and caves might sometimes be required, as in the past visits in different weather conditions have produced significantly different results.




The shallow interstitial habitat consists of flooded alluvial (riverine) gravels and the hyporheic zone (zone of interaction between up-welling groundwater and down-welling river water) beneath and beside [the parafluvial habitat] watercourses. Water-logged gravels are also present at the sources of many springs.

Being the most superficial of all the niphargids, Niphargus aquilex will occasionally turn up in samples of the epigean aquatic benthic macro-invertebrate fauna collected by kick sampling. N. aquilex can often be a significant component of the benthic fauna in previously dry winterbourne sections of rivers, after heavy rain has elevated the groundwater levels and re-started the surface flow. Over time N. aquilex will retreat back into the hyporheic zone as it faces predation and competition with Gammarus pulex and other epigean invertebrates. Recently Niphargus fontanus has also been identified in two Welsh Environment Agency kick samples, a species far less likely to be collected on the surface. Both N. aquilex and N. fontanus have been obtained by kick sampling an area of up-welling groundwater in a pond on chalk geology near Havant, Hampshire.

Bou-Rouch pump in action
Bou-Rouch pump in action


The Bou-Rouch Method


Probably the main method used to sample the shallow interstitial by groundwater ecologists is the Bou-Rouch method (Bou & Rouch,

1967). The equipment consists of a zinced iron pipe, 2-3 cm in diameter and length. At one end is a spiked tip, with rows of holes above this. At the other end is a lip or thread. The pipe is driven into sediments with a hammer (a cap is used to protect the thread or lip) and once at the desired depth a piston pump is fitted on the end for pumping up the water, which is then either pumped into a bucket for later filtering (this allows a given volume of water to be sampled at a time, providing an estimate of species numbers per set volume) or is filtered directly through a sieve / net.  The pipe net in the illustration above, fitted over the nozzle works particularly well. The pipes can be extended to several meters for deeper penetration into the interstitial, but their removal would be difficult, if not impossible. The piston pump can be replaced by a tube running to an electric or diesel motorised pump.



Bou-Rouch sampling equipment. A: Perforated zinced iron pipe. B: Piston pump attached to pipe. C: Protective striking cap. D: Hammer.
Bou-Rouch sampling equipment. A: Perforated zinced iron pipe. B: Piston pump attached to pipe. C: Protective striking cap. D: Hammer.


The Karaman-Chappius method


Karaman (1935) and Chappuis (1942) both developed a method for sampling the fauna in the water beneath gravel banks at the margins of rivers and streams. It simply involves digging a hole into the gravel below water level, then bailing the pit dry with a jug.  After the initial emptying of the pit, water will continue to  flow into the pit from the surrounding sediments, potentially carrying specimens of hypogean Crustacea along with the flow.  The pit is then continually bailed and poured through a sieve or net. A combination of the two methods above seems to work quite well: initially dig and sample using the Karaman-Chappuis pit and then hammer the Bou-Rouch pipe into the bottom of the pit and pump the pit dry, as shown in the pictures below. 



Other methods


Pospisil (1992) describes several other methods for sampling groundwater in shallow sediments. These include freeze coring, stand-pipe corers and the burying of artificial substrates. 

Freeze coring involves freezing the sediment around a stand-pipe and then removing the whole as a block, frozen fauna included.  

Stand-pipe corers remove small amounts of sediments from depths of about 1m but can yield uncertain results, due to their intrusive nature and the avoidance reactions of fauna.

Artificial substrates need to be left in-situ for several days or longer, enough time to be colonised by fauna from the surrounding habitat. Perhaps the best type is the Bishop Sampler, which consists of a box with moveable sides that is buried in the sediment with the sides open and filled with sediment. When it is due to be removed the sides are closed by wires and the sampler is dug out of the sediment. Simpler artificial substrates are of the Coleman and Hynes (1970) type, consisting of two perforated PVC tubes, one inside the other and filled with sediment.




As mentioned above, kick sampling can be used at the point of issue of springs and around distinct areas of up-welling groundwater in watercourses and ponds. Similarly the Karaman-Chappuis and Bou-Rouch methods, described above, can also be used in coarse sediment [gravel, pebbles & cobbles] around the point of issue. Some springs, mostly on the sides of wooded valleys, issue from the ground as muddy seepages and the large amount of silt in these areas will negate both of these methods, which require clean gravel in order to be effective. Usable pits cannot be dug in the silt and fine sand and silt will clog the holes in the Bou-Rouch pipes.  Artificial substrates or drift nets might be a possible choice for sampling these types of springs. Knight & Penk (2010) found many of the limestone springs in Ireland to issue from fissures between strata of solid rock, negating the use of the Karaman-Chappuis and Bou-Rouch methods.  They collected specimens from such springs by using a garden fork or rake to scrape the issue point as far into the fissure as could be reached whilst a drift net or wide hand net was held in place below.

The Noll spring sampler (diagram from Pospisil, P., 1992 ‘Sampling methods for groundwater animals of unconsolidated sediments’ in ‘The natural history of biospeleology’ Monografias Museo Nacional de Ciencas Naturales, Camacho, A.I. (ed.))


Drift nets     


The placement of a drift net at the point of issue can be effective in sampling springs. The drift net consists of a rectangular framed net, with a removable filter at the apex of the mesh collecting bag, and metal side bars that can be anchored in place with stones, weights or spikes.  The net is left in situ, usually for a period of 24 to 48 hours, collecting fauna as it is washed out of the ground; it is particularly effective at times of high groundwater discharge, following prolonged heavy rain. Camacho (1992) describes a method combining drift nets and the aid of cave divers for springs and resurgences that are accessible by humans. As the divers enter the spring the drift nets are anchored in place and removed when the divers exit. As the divers advance up the flooded gallery they agitate the silt on the bottom and walls of the gallery, dislodging fauna that are then washed out and into the nets. 

Pospisil (1992) describes a spring sampler developed by Noll (1939), consisting of a double funnel of bronze wire netting, fixed to a glass flask and sealed with a rubber ring. The device is dug into the mound of the spring’s issue point and removed after several hours.

Drift net placed at the Schwyll Spring outflow in the Ogmore Valley, Glamorgan. Photos courtesy of Gareth Farr.
Drift net placed at the Schwyll Spring outflow in the Ogmore Valley, Glamorgan. Photos courtesy of Gareth Farr.













These provide access to the deeper interstitial habitats and the phreatic groundwater. If a property is still known to be connected to a water supply from a well or borehole, one of the easiest sampling methods is to attach a pipe net (a simple net with a draw string or clamp) around the end of a tap or hydrant and leave it flowing for a while. Niphargus glenniei was first discovered in West Cornwall when specimens turned up in the water of a farm connected to a subterranean well.  If the water from a well or borehole is pumped to a header tank then sweeping the sediment in the bottom of the tank with a long-handled net will sometimes produce results. 

Piezometers (of varying diameters from 2.5-20cm) are installed at many places for groundwater table and quality monitoring purposes. A tube can be inserted into a piezometer or borehole and connected to a pump; the flow generated by this is then passed through a sieve or net, or into a tank for holding sediments and fauna. A variety of pumps can be used, from hand-operated piston pumps to motorised pumps. However, the latter, whilst having a higher extraction power, can consequently damage delicate specimens.

Old accessible well shafts (pit wells) are fast disappearing as many properties are now on mains water and the old shafts, deemed to be unsafe, are either filled in or capped. However, such wells can provide good results, especially if they have been in existence for some time.  Several methods for sampling wells and boreholes are described below.


Net sampling


The photo below shows various designs of net (manufactured on request by NHBS) used for sampling boreholes and wells. The net is attached at the top to a length of rope or cable by a link and is then lowered to the bottom of a well or borehole.  Hypogean Crustacea generally occur on the sides of the shaft or are concentrated in the sediment at the bottom.  Once on the bottom, the net is jerked up and down several times and the weight suspended from the bottom of the net disturbs the sediment causing it to become suspended in the water, along with any fauna that might be present.  The net is then drawn rapidly up through the water column to capture the specimens.  The net also has a detachable filter that screws into a collar, attached to the bottom of the net, enabling the easy removal of specimens.


A selection of borehole / well sampling nets.  They range in size from 30cm in diameter (top net) to 10cm (bottom).  The two long tubes are groundwater bailers used for collecting water quality samples from narrow boreholes.  Even smaller nets 5cm in diameter (middle) can be constructed from the cut off tops of these bailers.   


In large wells the zooplankton trawl net (net D in the first photo at the top of this section) can be lowered on a  rope into the water at the bottom of the shaft several times and trawled around the bottom, disturbing the sediment if possible. However, whilst these nets are light and therefore easy to carry, they might require some weighting in order to sink properly.  The trawl, net  also requires more space to manoeuvre in comparison to the specially designed nets in the photo above. There is the risk whilst net sampling in boreholes and especially wells  that they can become snagged on obstacles at the bottom , such as large stones, branches and the remains of old piping etc.

Many of the Environment Agency monitoring boreholes are very deep (some go over 100m into the Chalk), but they tend to be very narrow, many about 5cm in diameter.  To sample these narrow boreholes, nets can be constructed from the cut off tops of groundwater bailers (as in the photo above), essentially long, thin tubes used to collect water quality samples from the base of boreholes.  The bailers themselves can also be used to collect specimens, bringing water to the surface which is then filtered through a sieve or net.   


Sampling a narrow diameter borehole on Jersey using an adapted net design.
Sampling a narrow diameter borehole on Jersey using an adapted net design.


Bait traps


During the 1950s and 60s Glennie achieved very good results from wells using bait traps. Sometimes these consisted of nothing more complicated than jars lowered on a rope and left in-situ. A more effective trap can be made from plastic bottles by simply cutting off the top, inverting it and re-inserting it into the lower half and sealing the join with glue or heavy-duty tape.

Traps are probably best left in place for 24 hours and certainly not longer than 48 hours, as the bait itself can become a pollutant in the water if left for too long. There are various discussions on what constitutes good bait, although fish, liver and smelly (non-processed) cheeses seem to be the best. Some workers advocate fresh bait and some have achieved better results with slightly rotten bait.  


The FBA Well Sampler


The former director of the FBA Hugh Gilson designed a simple device for effectively sampling wells. It consists of two plastic bottles, the upper inverted and connected to the lower, upright bottle by a length of copper tube, with a lead collar on the base of the lower bottle to maintain the sampler in an upright position when in use. A gauze mesh disc is fitted on the top of the upper bottle. An air-pipe, terminating in a ceramic aquarium diffuser enters the lower bottle at the shoulder. When in use, the sampler sits with the bottom of the lower bottle partially buried in the sediment at the bottom of the well. Air is pumped (by a suitable pump connected to a hose) into the lower bottle to agitate the sediment and wash out fauna, which is then washed into the upper bottle and trapped there by the mesh disc.

A: Brass sleeve and gauze. B: Sampler bottles (in sampling position. C:Nylon air pipe with union and connector. D: Connector to attach bicycle or foot pump.
A: Brass sleeve and gauze. B: Sampler bottles (in sampling position. C:Nylon air pipe with union and connector. D: Connector to attach bicycle or foot pump.
A: Brass sleeve and gauze. B: Sampler bottles (in sampling position. C:Nylon air pipe with union and connector. D: Connector to attach bicycle or foot pump.
FBA Well Sampler (pictures from Driver, D.B., 1963 ‘Some simple techniques and apparatus for the collecting and preservation of animal from cave habitats’ Transactions of the Cave Research Group of Great Britain, Vol. 6, No. 2, pp91 - 101















Schematic drawing of the grab (1, rope; 2, hook; 3, springs release string; 4, coupling bar; 5, holes covered by nylon net; 6, springs; 7, block of hinged doors; 8, closing string; 9, hinged doors)
Schematic drawing of the grab (1, rope; 2, hook; 3, springs release string; 4, coupling bar; 5, holes covered by nylon net; 6, springs; 7, block of hinged doors; 8, closing string; 9, hinged doors)
The Pagliani grab 


Pagliani (1997) designed a small grab to take quantitative samples of sediment and fauna from the bottom of wells. The grab consists of a metal tube with spring-loaded doors at the bottom, weighs 6.5kg and samples an area of 86.5cm2. Its function is similar to that of the Ekman Grab but it is considerably less bulky and thus easier to deploy in the confines of a well shaft. The grab is lowered to the base of a well on a rope in the open position and when it reaches the bottom, the rope is loosed for about 1m, which allows a hook to release the coupling bar of the grab. At this point the grab is lifted by pulling on the rope, which is connected to the hinged doors. The traction of the rope unlocks the doors, which are pulled up by the action of the springs and hence close the grab with water and sediment inside. As the grab is hauled to the surface, water can exit via holes along the walls of the tube, which are covered with mesh to retain fauna.


Other methods

Pospisil (1992) describes two other methods for investigating deep interstitial habitats.  

The Double-packer Sampler is used in small diameter (2.5-7.5cm) monitoring wells (piezometers) which are perforated throughout their length. A sampling head is attached to a tube and fixed within the pipe at a set depth by inflatable tubes. Water is then pumped through the packerhead and tube by a piston or motorised pump. 

The second method involves inserting transparent pipes of 5cm diameter into the sediment. A small video camera, equipped with lights and a watertight casing and attached to a monitor is then introduced for in-situ observation of the sediment and its fauna. 




Caves especially contain a variety of aquatic habitats that can be sampled for hypogean fauna.

Subterranean rivers and streams

Many of the sampling methodologies used to sample the aquatic macro-invertebrate benthos in surface waters can also be used underground, although the main factor to consider is getting the equipment in and out of a difficult environment. These methods can include kick sampling, cylinder sampling, surber sampling, artificial substrates (both in the water and the sediment) and the use of drift nets. In allogenic streams (originating on the surface and flowing into cave systems) epigean invertebrates, washed in from the surface, will often be present, especially near the entrance. Autogenic streams (originating underground, within the cave system) are likely to produce better results with regards to hypogean fauna. Wood used a combination of kick, cylinder and surber sampling, along with drift nets and small aquarium nets to sample streams and pools in the Peak-Speedwell system of the Derbyshire Peak District (Gunn et al. 2000; Wood et al. 2002; Wood et al. 2008).  Knight (2011) achieved good results kick sampling allogenic and autogenic streams in Swildon’s Hole on the Mendip Hills of Somerset and a similar survey has recently been completed in the extensive Ogof Draenen system of South Wales (Knight et al. 2018) .

The Karaman-Chappuis and Bou-Rouch methods can be used to sample gravel beds at the margins of underground watercourses.


Kick sampling in one of the many streams in Ogof Draenen, South Wales
Kick sampling the main stream in Swildon's Hole, Mendip Hills















Small pools, gour pools, dripping / trickling water on calcite slopes etc.


These habitats are probably best sampled using manual searching, simply looking for fauna and collecting specimens using forceps or a pipette. It is recommended that entomological or feather-light forceps are used to avoid damaging specimens. Animals in thin / trickling films of water can be collected using small paint brushes, minimising damage. Care should be taken to avoid disturbing the sediment at the bottom of pools as this can quickly cloud the water and make manual searching impossible. This also means that manual searching is only effective in small pools in which the whole of the bottom can be viewed from the edges.   In slightly larger pools an aquarium hand-net (or similar), fitted with extended handle sections if required, can be used to sweep through the water and substrate at the bottom. This method can also be used to supplement the manual search in smaller pools if no taxa have been found, they might be hiding in the substrate at the bottom of the pool.

Care should be taken when sampling for any hypogean fauna not to over-collect and damage small, isolated populations. The populations seen in caves are really only the “tip of the iceberg”, compared to the populations in the surrounding rock strata and meso-cavernous spaces, but as with all natural history study, only collect the minimum number of specimens required for identification or none at all.

Most hypogean Crustacea seem to prefer pools with fine silt at the bottom, although they are occasionally seen in crystal-lined gour pools. Whilst feeding, Niphargus species often leave small trails on the surface of silt and these can be good indicators that taxa are present. However, due to the humidity in caves, it is not unusual for them to exit the water and migrate between pools or back into fissures in the surrounding rock over damp sediments and rocks.


Niphargus feeding trails in a pool in Pridhamsleigh cavern, Devon (photo courtesy of Chris Proctor).


A lot of water can enter cave systems through cracks and fissures in the roof and walls. After heavy rainfall these can become quite active and might contain fauna washed out of the surrounding rock strata. Jars, or small nets fixed in place beneath the fissures, can serve to collect specimens. Camacho (1992) states that such diffuse filtrations can be sampled by placing a mesh, slightly detached from the ceiling, and set in such a way that it channels the water towards a group of funnels with a net to retain fauna. Such an apparatus could be placed over a group of fissures and left as a semi-permanent fixture, as only the final net would need to be checked and regularly emptied.


Larger pools / underground lakes, sumps and flooded passages


Manual searching or the use of nets (kick and sweep sampling in the margins), especially those fitted with additional handle sections for extra length, can be used to sample the margins of such waterbodies but will obviously not sample the deeper sections.

The deeper sections can be sampled using a trawl net (zooplankton net) attached to a length of rope. The net will probably need to be weighted in order to be thrown from the edge of the waterbody effectively and sink to the bottom. There is also the risk that the net will become snagged on submerged rocks.

Bait traps can also be used in the larger water bodies, as discussed in the section on pit-wells. However, the use of bait traps is not recommended in caves as they involve two visits to place and retrieve and the bait can cause organic contamination of the water. Also there is the risk that a sudden flood can displace traps so that they cannot be recovered. In the cave environment, where food is scarce, the bait will then become a long-lasting focus of attraction which could ruin small, localised populations of hypogean fauna. Predators might also enter the traps and devour the fauna that has gathered.

Artificial substrates are another possible method but again would involve two visits to place and retrieve and could amount to a lot of encumbrance to carry, along with any other caving equipment that might be needed.

Camacho (1992) mentions the use of manual searching in sumps by cave divers, collecting fauna using modified nets. She also mentions a manual aspirator which works with compressed air and is very effective at collecting fauna from submerged cracks and clay walls, although it can cause considerable damage to the delicate specimens.


Net sampling the margins of the lake in Pen Park Hole, Bristol
Kick / sweep netting the margins of the lake in Dunmore Cave, Co. Kilkenny.












Trawl net being prepared for throwing, to sample the deeper reaches of the lake in Pen Park Hole.
Sampling the lake in Ogof Pant Canol, South Wales using a combination of marginal net sweeps and a trawl net
















The two main chemicals used for the preservation of aquatic fauna are formalin solution and alcohol solution. The former, usually used as a 4% solution, is good at fixing specimens, especially delicate structures such as gills and legs. However, it is also very acidic and needs to be buffered, as over time it will dissolve calcareous structures, such as snail shells and crustacean exo-skeletons. Formalin is also highly toxic and carcinogenic.  Its main advantage is that as it initially comes as a concentrated 40% solution, only a minimal amount is required in a small plastic bottle for transportation into caves, where encumbrance can be a major factor on difficult trips.  Only a small amount of the preservative is added to a sample, which is then diluted with water from a nearby stream or pool to make up a 4% solution.  The bottle should always be carried inside a waterproof container to prevent any leakages damaging the environment; great care should also be taken not to breathe in the toxic fumes in a confined space. 

Alcohol is the safer option and is normally used as a 70% solution, although this can damage specimens and make them brittle. An initial killing solution of 40% is best, with the specimens later placed in a 70% solution for long-term storage if required. The addition of a small amount of glycerol will also prevent complete dehydration of the specimens should the tubes become damaged. If specimens are required for DNA analysis then they must be preserved in 100% alcohol (preferably ethanol, although bear in mind that ethanol cannot be made 100% as it always contains 6% water).

The most commonly used alcohol solution is industrial methylated spirits (IMS), although this will require a licence from HM Customs and Excise to obtain. Other alcohols used include ethanol and isopropanol. If the solution is required for short-term storage, (i.e. for sending on to an expert for identification / confirmation) then standard, purple-dyed methylated spirits (as used in camping stoves etc.), or even a strong spirit such as vodka or rum can be used. However, note that the purple methylated spirits should be used neat and preferably only as a last resort, since it forms a white flocculent precipitate when added to water.



Most samples from underground environments should contain little organic matter, so separation of the fauna from a predominately mineral substrate should not be too difficult. They are also not likely to contain large snails and cased caddis larvae with stony cases, as found in epigean samples, which will probably be too heavy to separate out easily.

The easiest method is by decantation and sieving. The sample (or fractions of larger samples) is washed in a large tray and the sediment agitated so that fauna float to the surface. The supernatant is then poured through a sieve and fractions of the sieved matter are then sorted through on another tray with sufficient illumination and preferably some sort of low level magnification. Specimens are then picked out for further examination under a binocular microscope. It is wise to check the substrate left in the first tray, to make sure that no fauna has been missed.

Another method is to use the modified flotation technique developed by Anderson (1959), involving the introduction of sugar solution (approx. 2.5lbs granulated sugar per gallon of solution) to the sample, which will cause organic matter to float to the surface, due to differences in the specific gravities of the organic matter and the mineral substrate.

A third method is to use staining, with agents such as 1% eosine, which stains organic matter pink and hence makes it easier to identify. This staining requires at least 24 hours and the sample will need to be refrigerated to prevent deterioration.

Sometimes it is easier to sort samples live, as the movement of the animals in a tray can make them easier to spot and pick out. It is preferable to do this in a dark-coloured tray, against which the white hypogean organisms will stand out better. However, small specimens such as juvenile Niphargus and especially bathynellids are likely to be missed.