After our report on a study giving insights into how to minimize the risk of toxic vegetables from urban gardens, TreeHugger reader Craig wrote to ask:
I'm researching soil test kits - for pollution, not nutrition. I want to test soil, water, and food itself. Care to recommend any test kits?
Craig wants to test the soils on high ground and compare the results with the tests of soils in the valley that channels runoff from roads -- where the wild blackberries may be less brain food than drain fruit. It sounds like a great project!Although our answer may not be what Craig wanted to hear, we hope that sharing it can help TH readers save money on tests that are not reliable.
Gold Standard for Soil TestingUnfortunately, it is unlikely that a test kit available on the consumer market will reliably and accurately test soil contamination. The "gold standard" for testing metals in soil is to extract the metals and analyze the extract by atomic absorption or atomic emission spectrometers. These instruments (which are sufficiently expensive that only well-equipped laboratories can justify them) can detect the "fingerprint" of individual atoms: each atom absorbs or emits light at specific wavelengths that are unique to that atom. Alternatively, the even more expensive and highly sensitive ICP-mass spectrometer can identify individual metal ions by their atomic weights.
Another technique that has recently been gaining attention is XRF (X-ray fluorescence spectrometers), because certain consumer organizations have started to scan products for the presence of toxic materials by XRF. These devices also need to be used only by highly trained personnel, as much due to the safety of the use of x-ray sources as for the accuracy of the technique. Generally, the cheaper the device, the less able it is to distinguish between various metals in a complex sample like soil.
Although there are reliably certified lead test kits on the market (certified not to produce more than 5% false negatives), these kits are intended to operate in the range of 5000 ppm, well above the level of interest for contaminants in soil.
Soil Test Kit DownfallsSoil is notoriously hard to test, because the contaminant absorbed to the soil must be extracted into a liquid carrier in order to be available for feeding into a spectrometer or for reacting with a reagent that can indicate the presence of the contaminant by color change, one of the most common tricks of test kits.
This extraction process significantly affects the test results. Soil composition and pH, the presence of multiple contaminants, and other factors can all influence the completeness of the extraction. It is necessary for a repeatably consistent percentage to be extracted in order to quantify how much contaminant exists in the amount of soil used for the test or no numerical estimate of contamination can be reached.
Since some lead is typically present in soil (up to 20ppm may be considered "natural"). Also, lead is not necessarily hazardous even when present as a low-level contaminant: levels up to 100 ppm in soil are considered safe by most everyone, while up to 400 ppm lead in soil is safe for a child's play area, even considering that the child will eat some soil, according to the EPA. Therefore, a test that merely indicates "yes" or "no" is not meaningful. The test must give a quantitative result; the importance of the extraction step cannot be neglected.
Small sample sizes -- which are typically necessary to keep the costs down in consumer test kits -- further complicate the testing, because it is very difficult to get a "homogenous sample" of soil (a sample which would give the same results no matter where you pull out the little bit that will actually get tested).
Finally, the colorimetric testing common to test kits relies on the contaminant reacting with another chemical which changes color. These tests are susceptible to false positives -- indicating the presence of a contaminant when there is really some other, often benign, chemical in the soil that can also react with the color-changing reagent -- as well as false negatives -- indicating no contaminant, most often because the contaminant was insufficiently extracted from the soil or because the contaminant is part of a larger molecule that fails to react with the color changing agent.
Constructive Advice on Testing SoilsWe cannot leave the topic on such a negative note. To be a bit more constructive for anyone out there with a potentially interesting hypothesis about soil contamination to test: we would suggest networking a bit with local universities. See if anyone in the chemistry department would be interested in teaming up on such a project. Grant money can be available to help finance such studies, and of course the university chemistry lab is likely to be well equipped to pursue such questions.
This kind of project is a great way for students to learn about the methods, techniques, and limitations of chemical analysis. The scope of the study might even integrate the question of test kits. Doing the analysis by one or more "gold standard" methods and comparing results from consumer test kits would probably demonstrate what other studies have demonstrated: A low correlation of results.