Overall sampling strategy

Taking a soil sample using an auger

Strategies for sampling need to adequately represent a farm field. 5 to 10 sampling locations within each field are usually adequate, though as few as three are sometimes used if a large number of fields are being sampled. In general soil is cored to the depth of the plow layer (0-15 cm or 0-20 cm) where roots are most abundant – thus the zone that plants depend most on in terms of biological processes and soil fertility. It is also possible to substitute a traditional soil corer with a square-tipped shovel or even a machete, that can slice the edge of a hole cleanly, followed by trimming the slice with a knife to create a “core-like” square section of the soil e.g. 4 x 4 cm. This is more time-consuming but no worse for rigor than a soil core from a corer (which can cost hundreds of dollars).
However if assessing soil macrofauna is of interest (see section 6 below), it is necessary to excavate larger volumes (e.g. 20×20 area, x 20 cm depth) and so this becomes the strategy for also gathering some soil for testing of soil carbon, physical health, and fertility – no corer or corer is needed, just some straight spades or square-tipped shovels, for precisely excavating.

Finding Reagents
It is likely that finding chemical reagents will be a serious challenge to address in many regions. Starting early and identifying potential suppliers in larger cities is fairly important. Here are some descriptions of the five major chemicals required for the P and active C tests, contemplating that a typical kit would test about 300 samples – of course the amounts can be scaled up from there.
For the phosphorus testing using the Olsen Extraction:
Below reagents are in addition to small molybdate reagent packages that are supplied by Hanna Instruments, the company making the colorimeter (Hanna Low Range Phosphate reagent , cat. number 93713-0)
Sodium Bicarbonate (NAHCO3) (clean baking soda from the supermarket usually works but needs to be tested) If analytic grade NaHCO3 is available, this is excellent. 1.05 grams are required per sample analysis. That is, 42 g per liter to make 0.5 M NaHCO3 solution. Each liter can test 40 samples. So each kit for 300 samples, figuring a little extra for calibration, testing etc., requires 8 x 42 or 336 g, or rounding up to be safe, 350 g..

Sodium Hydroxide (NaOH) (lye from a hardware store may work but needs to be tested using the soil P kit – we used hardware lye successfully in the U.S. Very little of this reagent is needed, about 1 g per liter of extraction solution which can test 40 samples. So just under 10 g per kit for 300 samples.

Sodium bisulfate – also called Sodium Hydrogen Sulfate (NaHSO4) — this is a much safer and easier to dose alternative to sulfuric acid or hydrochloric acid. It is sold very cheaply as a swimming pool chemical in the U.S, and worth checking that avenue in a capital city in Africa, like Nairobi etc. The analysis uses 600 mg per test (i.e. 180 g for 300 tests, say), so not out of the question to import it, but could be good to find locally. We could adapt to using dilute sulfuric acid, but I would hope not to need to do this.

Phosphate salts: (optional but could be very good to know about) — potassium phosphate (monobasic or dibasic) or sodium phosphate — to prepare standards if we wanted to conduct a local calibration or check test for the phosphorus test.

For Active Carbon Testing (no additional reagents required like the P reagent packets above):
Potassium Permanganate — not much is needed per test. E.g. 64 mg per test, or 20 g for 300 tests. This means that if absolutely necessary, a small plastic container of KMnO4 could be flown in with travelers to the region. In the Andes, KMnO4 can be restricted as a drug manufacture precursor.

Calcium Chloride (Magnesium chloride may also work) — note the CaCl2 is the reagent we need in the most quantity within this test = 300 mg per test, so finding it locally would be good. I should test whether Mg++ flocculates clays as well as Ca++, since flocculation is its only role in the analysis. We would need 90 g CaCl2 for 300 tests.