Soil Sampling Strategies: The First Rule
When it comes to adopting a soil sampling strategy it is important to bear in mind the purpose of the exercise; the cost in terms of field and laboratory time, and the equipment and consumables that will be used. It is easy to both over and under sample with the latter producing non representative or untrustworthy data whereas the former wastes time and resources.
Therefore before any measurements are taken it is important that:
1) a suitable soil sampling strategy has first been devised
2) all samples are representative and part of a considered strategy
3) the limitations and destructive nature of soil sampling is considered
Devising a suitable sampling strategy
There is no hard fast rule here and picking the right strategy and sampling frequency is as much down to luck and judgement as to any precise scientific formula. A good soil scientist uses his intuition and judgement when trying to assess the soils properties. He cannot sample every point; he cannot even sample a significant percentage of the sample area yet he must gather sufficient representative data to make a judgement of the whole.
A good soil scientist therefore begins by using the most advanced and high tech remote sensing equipment he has: his eyes. For a good soil scientist first begins by looking at the landscape. He/She tries to read the patterns in the soil from the patterns in the landscape; a colour change in a crop, a species change in a meadow; the undulations, aspect and slope all give clues as to what lies underfoot.Thus the first thing one should do is look.
Making a Reconnaissance Study
Determining the plot boundaries
Simple Soil Sampling Strategies
The Classic ‘W’
The simplest and most common method to agriculture is the ‘W’ sampling method where five samples from five sites arranged in a ‘W’ formation across the field site are taken.
These may further be kept as 25 separate samples or combined to produce 5 bulked samples. If more than one site is being measured or the site breaks into smaller units (i.e. terraces) then each should be treated as a separate site and have a separate ‘W’ sampling strategy.
Sampling should be performed away from the field margins or any significant features such as a tree, rock outcrop or spring unless these features predominate: the intention is to sample areas that are ‘representative’ of the whole and not unique!
Once a strategy has been devised it’s important that it is implemented on a relatively dry day and when the soil is not in a rain saturated state. Waterlogged soils are not ideal for sampling and should be avoided. Make sure too that you take the tools you need, a trowel, sample bags, marker pen, 2mm sieve*, strong carry bag.
It is a good idea too to have prepared in advance a rough map of the field site and to mark on that map the positions of the sampling points and any notable features (i.e. trees, rock outcrops, springs, etc). Whilst accuracy is of benefit and a detailed OS map and a GPS unit are useful tools they are in no way essential so the map need only be a rough sketch and the sampling positions approximated upon that sketch. Make sure you have sufficient bags to put your samples in and you label and number each sample as you take it.
Alternatively, and this method applies very well to sloping ground such as a valley side, a transect may be taken. A transect consists of a series of sample points taken at regular intervals (i.e. every 4m) along a line. Its principal benefit is to identify changes along that line such as increasing depth or clay content down the hill/valley sides.
the grid and quadrat
a representative sample
When a sample is taken the sampling site is destroyed; any properties, particularly structural are irreversible changed. Similarly any measurements we take from our sample are measurements of the sample and whilst we are hoping it relates to it is not the field it was taken from.
So whilst it is likely that the sample will have similar properties to the whole field it is just as likely that it will have properties that are unique, spatially or temporally dependent, property(s) that do not reflect the field site. Thus we cannot rely on a single measurement to assess the field site and nor, for both destructive and economic considerations is measuring the entire site an option. Therefore we must use a strategy somewhere in between.
limitations and destructive nature
Sampling in the field
At each point in the ‘W’ or along the transect 3 or 5 samples should be taken, 3 is the minimum required for statistical analysis and 5 the most economical. The samples should not be taken from the surface, which is often ‘contaminated’ with organic debris (leaves, twigs, etc) but from just below the soil profile (approximately 8cm/3inch down). A sample of approximately 150g should be taken and if necessary passed through the 2mm sieve. The sample should be bagged and labelled and the process repeated at each site until 15/25 samples have been collected. If the stone content is greater than 10% then it should be properly calculated for whilst it still has little influence on the soil processes, it affects the total soil volume and so needs to be known and accounted for in future calculations.
Stone content is determined by passing a much larger sample of soil through a 2mm sieve. The best way to do this is to use a spade and a piece of plywood (20cm x 8cm) as a template. Place the plywood on the sample site and use the spade to cut around the plywood template. Then use a trowel to carefully remove the soil from under the plywood to a depth of approx 6cm and pass this through a 2mm sieve. We have now separated approximately 1 litre or 1000cm3 of soil into two fractions and can either estimate the volume of soil occupied by the stones visually or we can measure both the mass, by weighing the stones and the volume by displacement.
To do this take, two 1 litre jugs. Fill one with water to the 1 litre mark and put the stones from the sieve in the other jug. Now pour the water from one jug into the other until it covers the stones. Give the jug a gentle shake to displace any trapped air in the stones and then top up to the 1 litre mark. The amount of water that now remains in the first jug is equal in volume to the stones in the second and as the sample was approximately 1 litre it is similarly the percentage of stone content. If the stones are large then doubling the size of the plywood template to 20cm x 16cm and removing 10cm of soil might prove more practical. In this case the volume of soil sampled will be 3 litres and the calculations should be adjusted accordingly.
Preparation for analysis
Once all the samples have been taken the first thing to do is to air dry them to remove all the freely available water. This can be done at room temperature where the atmosphere is dry however it should ideally be done with some gentle heat assistance that can be provided by a few 60w incandescent light bulbs.
The samples should first be transferred from their plastic bags to open containers, such as the disposable foil trays used for take away foods, the foils should be marked with the sample id and these placed under the heat lamps or in a suitable warm and dry place for 48 hours. Technically the sample should be heated to 400C for 48 hours, but 2 days in a warm conservatory in mid July is just as good. Once air dry the samples are then ready for sieving.
Soil Horizons and the Agricultural Landscape
In soil science much is placed on the role and importance of soil Horizons. These are layers within the soil that result out of accumulations and sorting of material over time in an undisturbed environment. However in Agriculture the practice of cultivation destroys and homogenises these layers to produce a far more simplistic system of :
Top soil (where the majority of the plant soil interactions take place)
Sub Soil (accessible to the plant and important in supplying water)
Deep Sub soil (inaccessible to the plant)
Thus for the agronomist there are just three horizons, the depth of which is determined more by the plough and the mattock than anything else. So whilst academic soil science distracts itself by labelling profiles the agronomist need know only three.
Free Cultural Works (CC BY-NC-SA) : Malcolm McEwen (2012)