[There are more tree-ring papers and documents here]
[Note: This page is reproduced from a short article I wrote in 1985 for the newsletter, Cambial Activities in response to repeated questions I was receiving about the techniques I used to analyze wood for trace metals. Through the years I have continued to receive similar inquiries and so I am placing this article on the World Wide Web. Advances in sampling and analysis techniques since I originally wrote this article may make some of the material appear outdated.]
Elemental Analysis of Tree Rings
...I first became acquainted with the technique of coring trees in 1973 while I was a graduate student in Biology at Emory University in Atlanta, Georgia. In those days very little had been written about sampling, preparation and analysis techniques for trace metals in wood. Since that time more literature has become available, and my own techniques have improved considerably.
I have organized my comments below under headings which correspond to queries received by [investigators]. Readers wishing more detailed information on my early work should consult my master's thesis, "Age-Specific Lead Distribution in Urban Forest Tree Rings in Atlanta, Georgia", available from Emory University and the open literature paper based on that work, "Age-Specific Lead Distribution in Xylem Rings of Three Tree Genera in Atlanta, Georgia", coauthored by H. L. Ragsdale, Environ. Pollut. Ser. B2 (1981): 21-35.
Sampling and Sample Preparation
In most of my early work with elemental analysis of tree rings I used a 4 or 5-mm steel or teflon- coated increment corer for sampling. Most recently, I have been using a 12-mm diameter corer which has been chrome or nickel-plated. The advantage of the smaller diameter corers, of course, is that they are more easily inserted into trees, especially oaks and hickories. They do, however, break more easily than the 12-mm diameter corer. The advantage of the Larger corer is that more sample is obtained, and this can be critical if trace metals are at a low concentration or the time step (3, 4 or 5 year increments) for the analysis is small. Based on simple geometric considerations, It would take 9 cores of 4-mm diameter or 6 cores of 5-mm diameter to provide as much sample mass as a single 12-mm core. The major drawback to the larger corer is that it is nearly impossible to insert into hardwood species by mere mortals.
Through the years I have done a number of small informal experiments designed to test for contamination of the sample from either the corer itself or from bark during the insertion of the corer. (This question of contamination always seems to come up during discussion of results). I have never found any evidence of such contamination. However, it's always wise when analyzing for trace metals to minimize any chances of sample contamination. For tree-ring analysis this means keeping the corer inner and outer surfaces free from "gunk" [sap, oily resins, etc.) and rust. This is one reason why I now use a chrome or nickel-plated corer. The plating eliminates rusting of the corer, makes it easier to clean, and actually makes the corer easier to insert into the tree. My experience with teflon coated corers is that the teflon soon wears off the bit and inside of the corer, and it soon rusts without frequent attention. Also, teflon corers "scream" during insertion. The plated corers are much quieter.
Back in the lab I freeze the cores until preparation for analysis. Even refrigerated, fungi and mold would soon appear. Never dry the cores before sampling for analysis, because when dried they crack and are difficult to clean and cut up. I use the setup depicted in Fig. 1 for sample preparation. The Fred C. Henson Co. core mounting stage has been modified so that it can be opened far enough to accept the teflon sleeves and the core. The idea here is that the sleeves secure the core so that pieces will not fly off like "tiddlywinks" when the cut is made. I clean the core by cutting parallel to the core axis with a stainless steel scalpel as a carrot would be peeled. The thumb screw is loosened, the core rotated, thumb screw re-tightened, and another parallel cut is made. In this way the entire surface which was exposed to the corer is removed. The object here is to both eliminate any surface contamination, and also bring out the rings for crossdating, measuring and cutting. The teflon sleeves and stainless steel scalpel blades are used to minimize trace metals contamination. Unfortunately, stainless steel quickly loses its edge, so more than one blade per core may be necessary, particularly if the wood is hard. I use stainless steel forceps or a pair of hemostats to handle the core while cleaning and cutting.
Figure 1. Setup for cleaning and cutting up tree cores.
I usually count the rings back from the most recently formed ring under 10X (rarely is a higher power necessary). In my work I have used time steps of 4, 5 and 10 year intervals. Even with a 12-mm diameter core this can mean 0.1 gram of wood at times of slow growth. However, most samples of 5-year time steps range from 0.1 to 0.3 grams. The time step selected really depends on the objective of the study and the funds available. Gross trends can be observed with larger time steps, but statistical analyses of the historical trends are facilitated by a shorter time step. In my most recent work on changes in wood chemistry since the 1950's, I have used a 5-year time step since 1950 and 10-year time step previous to 1950. The investigator will have to make an individual determination of the appropriate time step for a specific study.
I have tried wet digestion techniques for putting the wood into solution for analysis without success. The problem with wet digestion techniques (e.g., HNO3-HClO4 digests) is that they require a minimum of 10 to 15 ml of concentrated acid for digestion and this results in a tremendous dilution of the trace quantities of metals in the 0.1 to 0.3 g of wood digested. Also, for most analysis techniques, the acid digests are brought up to a larger volume, further diluting the metals concentration. I understand that there are now procedures employing Parr-type digestion bombs which use as little as 5 ml of 50% H2O2 under high temperature and pressure to digest the wood. However, production of a large number of samples with such techniques requires fabrication of a special multichambered digestion vessel.
For processing a large number of samples, I have found that dry ashing is a good approach. Small (10 ml) Pyrex beakers or quartz ashing vessels can be kept chemically "clean" in a 6 N HNO3 bath, and they serve as both drying and ashing vessels. The procedure I currently use is to place the cut ring section in a numbered 10-ml Pyrex beaker. (A set of 50 beakers can easily be inscribed with a diamond-tipped marking tool). I record the number of the beaker next to the sample identification and follow that sample number (rather than the sample name) thereafter. After all samples have been placed in the beakers, I place them in a tray, cover the tray with aluminum foil, and dry them for 24 hours at 100°C in a forced-air drying oven. The beakers are then placed in a desiccator to cool, and they are weighed on an analytical balance. Either the beaker plus sample may be weighed (and the tare weight of the beaker subtracted) or the sample may be removed from the beaker with stainless steel forceps or hemostats and weighed by itself. I prefer the latter approach with samples of less than 1 gram. After recording the weights next to the sample number, I place the beakers and sample in a muffle furnace and ash for 24 hours at 400°C. Higher temperatures are unnecessary and may cause volatilization of the metals; lower temperatures may result in an incomplete ash. After removing the beakers from the furnace and allowing them to cool, the ash is taken up in 5.0 ml 10% HNO3 (for a 0.1 g sample, and dilution factor is 50). Finally, one drop of 30% H2O2 is added to eliminate any color or precipitate left in the beaker.
The above procedure may be modified depending on the analysis performed. I use 5.0 ml of acid solution because that is the minimum amount necessary for inductively coupled plasma (ICP) analysis. The same amount would probably be necessary for flame atomic absorption techniques (AA). With graphite furnace AA techniques 1.0 ml or less would probably suffice. Depending on the type of analysis performed, the acid and amount used could be varied. It is important to remember that the concentration of most trace metals is less than 1 ppm in wood. Thus, the detection limits of the analysis technique have to be kept in mind. For trace metals in wood, graphite furnace AA techniques are likely the best.
I have used flame AA, graphite furnace AA, and ICP for trace metals analysis of wood. Flame AA would be appropriate only for analysis of Na, Mg, possibly Al, K, Ca, Mn, Sr, and Ba. ICP will typically give results for roughly 30 elements including Ag, Al, As, B, Ba, Be, Ca, Cd, Co, Cr, Cu, Fe, Ga, Hf, K, Li, Mg, Mn, Mo, Na, Ni, P, Pb, Sb, Se, Si, Sr, Ti, V, Zn, and Zr. However, rarely are Ag, As, Co, Cr, Ga, Mo, Sb, Se, V, and Zr, detected in wood by ICP. Graphite furnace techniques are probably the best for any given single element, but its chief disadvantages are the time necessary to set up and calibrate the tube, single element analysis only, and the short life of the graphite tubes. For what it's worth, I have found that acid type and molarity are not important considerations for ICP. For atomic absorption, HNO3 and HClO4 were the best acids, but molarity did have an effect on absorbance. For graphite furnace AA, I prefer perchloric acid, but it is harder on the tubes than nitric acid and is, thus, a little costlier.
In my experience with trace metals analysis of wood, the biggest difference in temporal patterns is between ring-porous and diffuse-porous species with the ring-porous species showing patterns of historical significance. I believe that the difference is mainly due to the differences in water conduction and functional lifetime of the xylem tubes in the two. For every diffuse-porous tree I have examined, temporal metals concentration patterns have shown a "bell-shaped" pattern, with highest concentrations somewhere in the middle of the bole. Such patterns are difficult to relate to historical events, and I suggest that only ring-porous trees or conifers be used to reconstruct pollution histories.
Patterns of the alkali metals (Li, Na, K, etc.), alkaline-earth metals (Mg, Ca, Sr, etc.), Mn, and Zn in [Smoky Mountains] conifers suggest that either their uptake rate changes with time1 or that they are translocated from younger to older rings. Their temporal concentration patterns tend to be very similar to each other and somewhat similar to trace metals patterns in diffuse-porous species, but the bell shape is much flatter. Patterns of Al, B, Cu, Cd, Fe, and Ni, and Pb suggest that these metals do not translocate to any significant degree (a small amount of translocation of Pb in hickory was suggested in my early work). Detection of ppm levels of Cd, Cr, Co, Pb, Mo, or V in wood is rare, and their occurrence in wood is likely an indication of anthropogenic pollution.
It is really difficult to determine if translocation has occurred in tree rings from a single observation in time. The exception to this is when there is a historical "marker" such as the introduction of tetraethyl-leaded gasoline in the United States in 1923 or the beginning or termination of a pollution source. Some of my recent work with short-leaf pine downwind of Copperhill, Tennessee, suggests that iron is immobile in that species. Increased iron accumulations between 1863 and 1910 (attributed directly or indirectly to combustion emissions from copper smelting) have apparently remained in rings formed during that time. Also, recent unpublished work by Larry Ragsdale at Emory University has shown little or no translocation of Pb in hickory since my original work ten years ago. Ragsdale recently cored the same trees I did back in 1973-74. He found that the patterns I observed were essentially unchanged in the last ten years.
Life Sciences Division
Oak Ridge National Laboratory
1060 Commerce Park
Oak Ridge, Tennessee 37830
1It is my opinion that the declining concentrations of Ca, Mg, Al, etc. that we saw in more recently formed rings in Smoky Mountain conifers is indicative of the unprecedented changes to the high-elevation forests brought about by air pollution and acid deposition since the 1950s.