The yew tree (Taxus baccata) and a related species common to gardeners, Japanese yew (Taxus cuspidata) is known to be one of the most poisonous woody plants in the world, with all components of the tree, excepting the fleshy red part of the berry containing lethal amounts of taxine, a toxic alkaloid found in the yew. There are numerous references in the literature of yew poisoning (for a paper citing many examples, see Wilson, Sauer, and Hooser).
One point not especially clear in the case studies is whether or not the root of yew has the same level of toxicity as the other parts of the tree, and whether that toxicity, if present, remains in a root system for extended periods of time after the tree has been removed. While above-ground parts of yew are concerns because of potential exposure to humans, livestock, and wild animals and fish, root systems are rarely concern for exposure.
However, recent use of plant root systems by hobbyists for uses in taxidermy, floral arrangement, and aquascaping have created concern for potential exposure. In particular, the root system of the low, spreading Taxus cuspidata is an especially tempting candidate for use in these applications due to its propensity to remain attractive and intact upon removal from the soil.
In one such case by this author, a several root systems from Taxus cuspidata bushes were excavated after seven years of aging dead in the ground. There was a natural question of whether these attractive root systems were still poisonous or whether the taxine has leached or otherwise dissipated from the root material.
To test the question, several root samples were removed from the root systems in question and compared to freshly-cut specimens of root, twig, and needle of live Taxus cuspdiata via mass spectrometry. Mass spectrometry works by ionizing the molecules of the sample, usually by adding either a positively charged hydrogen atom (H, sometimes called a proton) or a sodium (Na) ion to the molecules, thus causing the molecules of the sample to become positively charged. The charged molecules then travel through a magnetic field towards a detector, and reach the detector in varying amounts of time depending upon their charges and masses. The detector detects the masses of the incoming particles, and these masses are shown in graphical representation of the resulting spectra. Because the samples are charged, usually by addition of either a proton or a sodium ion, we have to adjust our interpretation of the results based on this knowledge.
Each yew sample (fresh root, seven-year-old dead root, fresh twigs, and fresh leaves) was ground in a Wiley mill to obtain fine sawdust, then extracted by refluxing 0.2 grams of sample in a ten-milliliter mixture of 50% diethyl ether and 50% acetone. This solvent system was chosen because taxine is soluble in organic, non-polar solvents, but isn't soluble in polar solvents such as water. The extract was then concentrated and submitted to mass spectrometry.
The molecular weight of taxine A is 641.8 atomic mass units (amu), and the molecular weight of taxine B is 583.7 amu. During the process of mass spec, a proton (molecular weight of 1 amu) is often added to these molecules, so a positive identification of the toxic alkaloid is 641.8 + 1 in the case of Taxine A, or 583.7 + 1 in the case of Taxine B. We do in fact observe these peaks in some of the spectra displayed in the images.
In fact, the mass spectrum for the leaf extract (see figure, panel D) shows a huge taxine B + proton peak, indicating that yew leaf contains a lot of taxine B, but no detectable taxine A. Because sodium (molecular weight of 23 amu) is so pervasive in the environment it is also quite common to observe sodium adducts in mass spectra. The sodium adducts would be 641.8 + 23 = 664.8 for taxine A and 583.7 + 23 = 606.7 for taxine B. The mass spec technique employed here is not truly quantitative, but simply by "eye-balling" the spectra we can see similar quantities of taxine A and taxine B sodium adducts in both the fresh root sample as well as the seven year old root sample (see figure, panels A and B), indicating that both taxines A and B are present in yew root tissue even after extended periods of time following the death of the plant. This might be expected, as taxine is almost completely insoluble in water and thus probably not susceptible to leaching. If taxine is stable then, and this data suggests that it is, it will remain present in the material, and in ever-increasing concentration as the tree matter dessicates. Interestingly, the twig material showed almost no taxine, with peaks potentially being ascribable to taxine only visible down in the noise (see figure, panel C).
Our conclusion is that the toxic alkaloid taxine is in fact present in yew root as it is in above-ground components of the tree, and that it remains present in the root material for as long as that root material remains undecomposed. Use of yew root in hobbyist or other uses is not recommended unless ingestion of the material is carefully prevented.