Construction and Installation of Dendrometer Bands for Periodic Tree-Growth Measurements
By Bobby D. Keeland,
U.S. Geological Survey, National Wetlands Research Center, Lafayette,
Dendrometer bands are commonly used to make short-term repeated measurements of tree-stem growth. These bands are fairly easy to make and install on most trees. Due to the numerous inquiries we have received regarding the construction and use of dendrometer bands, we have provided here a general history of their development, theory of operation, and construction and installation details. We have also provided a partial list of suppliers for dendrometer band materials and a list of the tools required for installation. Most importantly, we try to share some of the lessons we’ve learned over the years of working with these bands, and we hope you find them helpful and informative.
Table Of Contents
Many studies in forestry, forest ecology, and tree physiology require taking radial growth measurements. Measurements are often taken with a variety of contact or noncontact instruments. Contact methods generally involve the use of calipers or diameter tapes, while noncontact methods include the use of an optical caliper, a rangefinder, or optical forks. A discussion of these methods and the relative accuracy of each can be found in Clark and others (2000). The measurement accuracy required in any study depends on several factors. In some cases where measurements are taken at long intervals of up to 5-10 years, such as in periodic forest inventories (Vissage and others, 1992) or long-term studies of forest growth response to environmental changes (King and others, 1998), the need for extreme accuracy is less important than that required, for example, by a study such as the cambial response to changing temperature over a 24-hour period. When measurements are infrequent, the amount of error introduced by diameter tapes is acceptable in relation to the expected amount of growth. However, for studies where measurements are to be taken more frequently, diameter tape errors may be great in relation to the expected growth. Although diameter tapes are quick and easy to read, their use does include some potential for error. It is impossible to position the tape in the exact same place on a tree stem during repeated measurements, resulting in placement errors (Bower and Blocker, 1966; Cameron and Lea, 1980). In addition, the scale on diameter tapes makes extremely accurate measurements (to less than 1 mm or 0.04 in resolution) difficult (Yocom, 1970). Therefore, when short measurement intervals are needed an alternative measurement technique is required.
MacDougal (1921, 1924) introduced an alternative and extremely accurate measurement technique called the dendrograph. Although this instrument eliminated some of the errors associated with diameter tapes, it was not considered accurate enough for slow growing hardwoods, was unduly complicated, expensive, and difficult to mount to trees (Fritts and Fritts, 1955).
In 1932, L. H. Reineke introduced a much simpler and precise instrument called the dial-gauge dendrometer. This dendrometer consisted of a positioning bolt and setscrew that were fastened to the tree stem and a machinist dial gauge. A measure of radial growth was read when the dial gauge was placed against the positioning bolt and setscrew. Daubenmire (1945) and Reineke (1948) discussed potential improvements in positioning the dial gauge to improve accuracy. The dial-gauge dendrometer allowed for very accurate measurements but provided for only one measurement point on the stem (Hall, 1944; Bormann and Kozlowski, 1962). A potential problem with this method was discussed by Fritts (1976) who showed that trees do not always achieve growth on all sides of the stems throughout the growing season. In fact, the selection of the measurement point on the stem could result in highly different measures of growth. This problem can be partially avoided by placing multiple measurement points around each stem. Although this method provides much more data, it requires more effort in the field collecting the data, and in the lab managing the data and calculating the average growth.
A measurement device that avoids the need for multiple measurement points around the stem is the dendrometer band introduced by Hall's (1944). The dendrometer band consists of a thin strap of metal that is wrapped around the tree and secured back to itself with a spring. Hall's (1944) band includes a vernier scale to allow measurement accuracy to 0.01 inch. Measurements taken at one point, therefore, represent an average of growth around the entire stem. While the bands do not provide as accurate data for individual growth points as the dial-gauge dendrometers, they do provide a better overall average of growth (Bormann and Kozlowski, 1962).
Dendrometer bands consist of thin straps of metal placed around a tree, with one end passed through a collar and then connected back to itself with a spring (fig. 1). As the tree grows (or as it shrinks and swells diurnally), the spring allows the band to move with the circumference of the tree. As the stem expands, the band slips through the collar and the spring is stretched.
Dendrometer bands can be made from straps of stainless steel, aluminum, or any of several other metals. Aluminum has the advantage of being very flexible and easy to work with. It conforms easily to the shape of the tree and should adjust to the stem quickly. Disadvantages include the ease with which animals such as squirrels and birds or falling debris can damage the bands. In addition, in marine environments, such as mangrove swamps, the aluminum bands deteriorate very quickly and cannot be used. Stainless steel is much more resistant to animals, falling debris damage, and corrosion, but it is less flexible than aluminum and so may not fit to the stem as quickly. Another disadvantage of both aluminum and stainless steel is their coefficient of thermal expansion. For some studies where extreme accuracy in measurements is necessary this expansion could be a serious problem. Using invar banding, with a lower expansion coefficient, you can avoid this problem, but invar is more expensive and is subject to excessive rusting. In general, we have found that stainless steel labeling tape is relatively inexpensive, strong enough to resist damage and rust, thin enough to form to the stem, and easy to work with (see table 1 for material sources).
Generally, banding material is 0.3 mm (0.0115 inch) thick and approximately 12-13 mm (0.5 inch) wide and can be obtained from any of several manufacturers of metal strapping. Sources can be found through a quick search on the Internet. In addition, we have used metal labeling tape available from suppliers like Forestry Suppliers or Ben Meadows Company (table 1) (just make sure you don’t get the tape with adhesive). The springs used on dendrometer bands must be made of stainless steel or any other metal that will resist rusting. Most readily available springs will quickly rust and possibly break, leaving the band ineffective. Spring dimensions can vary, but we have found that a spring length of 76.2 cm (3 inch), outside spring diameter of 6.35 mm (0.25 inch), and wire diameter of 0.66 mm (0.026 inch) work very well. These springs provide an initial tension of 1.48 N (0.333 lb), a rate of 0.087 N/mm (0.5 lb/inch), and a maximum extension of 190.5 mm (7.5 inch).
The first step in installing a dendrometer band is to prepare the surface of the tree bark. First, the stem of the selected tree should be examined to make sure that the potential banding height is free from obstructions such as branches, knots, burls, or any other deformation. Bands are often placed at 1.4 m (4.5 feet) so that measurements correspond to diameter at breast height (dbh) measurements in many other studies. It may be necessary to place the band higher or lower on the stem depending on potential obstructions (limbs, knots, etc.) or expected flood heights (it is hard to read a band when it is under water, and band measurements taken on the buttressed area of the lower stem can be meaningless). Take care to place the band at a good height for the individual that will be making the measurements. A band that is too low or too high is difficult to read in the low light of a closed canopy forest.
Next, smooth the circumference of the tree at banding height with some type of rasp, file, or sandpaper. A file is used to eliminate all high and low spots on the bark so that the banding will extend around the stem in a uniform manner while sitting flush with the bark surface. A surform file, available at many home improvement or hardware stores, works very well (fig. 2). If you use the surform file, be sure to have extra cutting heads available. The cutting teeth become dull after a while, and the cutting heads have been known to occasionally pop off the handle and disappear.
Great care must be taken to insure that the cambium is not damaged by the filing process. Some species, such as beech (Fagus sp.), have very thin bark that the file can easily cut through, and so you should use the file sparingly. Other species such as oaks (Quercus sp.) have a hard, thick bark that can be smoothed quite effectively. Species with thick spongy bark, such as redwood (Sequoia sp.), may present additional problems as compression of the bark between the xylem and the band may affect the measurement of stem growth. In this case the thickness of the bark should be reduced to insure band movement with any growth. When smoothing the fibrous bark of species like baldcypress (Taxodium distichum (L.) Rich.), the teeth of the file can quickly become clogged, but the bark material is easily cleared from the file.
Coarse sandpaper belts (36-50 grit) approximately 2.5 cm (1 inch) wide and long enough to surround the stem also work fairly well for some species. Just cut the belt to form a long strip of sandpaper, place it around the tree stem, and pull back and forth on each end.
Before beginning any work on dendrometer bands be aware that the edges of the banding material are relatively sharp and can easily cut through skin. The worst cuts happen when placing the band around the tree. It is difficult to make bands while wearing protective gloves, so it is important to be careful during all phases of band construction and installation. Never slide your fingers along under the band or slide the band across your hand while pulling it around the tree.
The length of a band is highly dependent on stem circumference but generally should be equal to the circumference of the tree plus about 20-30 cm (8-12 inches) (on large trees) or about 50% of the tree circumference (on small trees), if possible. The amount of excess banding may be limited for small diameter trees because the length of the slightly stretched spring plus the excess banding material (the part protruding beyond the collar when the band is tight around the tree stem - see text below) must be less than the circumference of the tree. If too little banding is left beyond the collar, the band will have to be replaced before the end of the growing season. This need to replace bands is especially a problem for small or young stems of fast growing species. If sufficient excess banding is provided, the band may not have to be replaced for several years.
One end of the band should be shaped into a collar following the methods described by Hall (1944) and shown by Cattelino and others (1986). To make a collar you must first make a simple tool from a small length of banding material (fig. 3). Take approximately 10 cm (4 inches) of banding and bend it double with the ends together and sides parallel. This double thickness of banding is required to allow each fold of the collar to be crimped tight, yet also allow enough space so that a single thickness of banding material will slip through the collar easily after removal of the tool and installation is complete. You will also need to cut a small length of banding that is about 2.5+ cm (1 inch) long (just short of three times the width of the banding material).
Hold the tool under the band with the ends of the tool and banding flush (fig. 4), and then hold the short piece of banding perpendicular to and about 1 cm (0.4 inch) from the end of the other pieces (fig. 5). Fold the ends of the short piece over, around the banding and tool (fig. 6). Use a pair of pliers to crimp the fold (fig. 7). Make sure that the folds are square and tight or the collar will be loose and measurement errors could result. Next, fold the band on either side of the short piece (fig. 8). Again, each fold should be square, and pliers must be used to make the fold tight. Remove the tool and the collar is complete.
Before installing the band you should measure the circumference of the stem at the location where the band will be installed. This step will allow a starting point for circumference changes measured during the study. Any smoothing of the bark should be done before taking the initial circumference measurement. Make sure that you have all required tools and a good field vest for carrying the necessary tools and equipment (table 1). In addition, if you will be working over water it is a very good idea to tie a short length of field flagging to each and every tool. When you drop a tool into the water (and eventually you will) the flagging will make it much easier to find.
Place the band around the tree, making sure that the free-running end is always wrapped around the tree in the same direction. If some bands have the free end wrapped around the tree in a clockwise direction and others have the band wrapped around the stem in a counter-clockwise direction, it will be difficult (awkward or confusing) to make the measurements on a regular or periodic basis. The key word here is consistency. These right-handed authors find that wrapping the free-running end of the band around the stem in a clockwise direction makes reading the band very easy. Left-handed individuals may find the counter-clockwise direction to be more convenient, but the ultimate choice may depend more on the ruler to be used to make measurements. With a clockwise orientation of the band, measurements are made with the ruler held horizontally and with the numbers along the top increasing to the right. With a counter-clockwise orientation, the ruler will be held with the numbers upside down on the bottom and increasing to the left.
The free-running end of the band is then slipped through the collar and the band tightened around the stem. Make sure that the free-running end of the band does not extend around the stem past the collar and that there is enough room between the end of the band and the collar for the slightly stretched spring. Use a hole punch such as the Roper Whitney junior hand punch to make a hole in the end of the band (fig. 9). Any heavy duty punch designed to cut through steel or aluminum by exerting approximately 2,400 lbs of punching power will work. A paper punch will not work. Next, attach a spring through the hole at the end of the band, pull the band tight around the tree stem, and stretch the spring to produce a slight tension. Note the position of the end of the spring on the banding (fig. 10), and then release the tension on the spring. Another hole is punched in the banding at the noted position (fig. 11). The completed dendrometer band can then be installed on the tree stem with the spring supplying the tension needed to hold the band tight to the tree. Make sure that the band is flush to the bark surface and extending uniformly around the stem at a consistent height. It is common for the band to droop at some point around the stem circumference after initial placement. Care must be used to straighten the band and reposition any sagging sections. The band should appear as the intersection of the stem surface and a plane perpendicular to the stem. The spring should have some but not excessive tension and should rest on the band, not on the bark.
The final step in band installation is to make a scratch across the band at the edge of the collar away from the free-running end of the band (fig. 12) . As the tree grows the scratch will move away from the collar. Be sure to make a good clean scratch with something like the tip of a sharp knife. You want a good deep scratch, but you do not want to crease the band, as this will make it difficult to obtain an accurate measurement. A scratch that is too light will be difficult to see. Multiple scratches that do not completely overlap will also make it difficult to obtain an accurate measurement.
The amount of growth (circumference growth) can be directly measured to the nearest 0.25 mm (0.01 inch) with a hand-held ruler (fig. 13). Hall's (1944) method of using a vernier scale can be used to achieve greater measurement precision among measurements (0.25 mm or 0.01 inch), but the scale takes a lot of extra time to construct (especially when a lot of trees are being banded), and for fast growing trees the scale may need to be remade before the end of the growing season. If growth measurements are determined by measuring the distance the mark has moved away from the collar, there will be no need to remake the band until the amount of excess (free-running) banding has been taken up by stem growth. The length of time required for this amount of growth will vary depending on the size of the stem, the tree growth rate, and the amount of excess banding provided during band installation. Small stems will allow for little excess banding. In a study of baldcypress growth, one of us (Young) had to replace the band on a small fast-growing tree four times during a single growing season. Each time a band is replaced it takes valuable time and introduces error into the measurement (while the band resettles onto the stem).
One author (Keeland) noted deformed growth under the banding of some trees in a study in South Carolina. Smaller stems, close to 10 cm (4 inches) dbh or less, were much more prone to deformed growth when compared with larger stems, and water tupelo (Nyssa aquatica L.) and baldcypress had more deformed growth than did swamp tupelo (Nyssa sylvatica var. biflora [Walter] Sargent). We think that the deformed growth may have resulted from springs that exerted too much tension. The springs suggested in (table 1) do not seem to cause deformed growth. Using these springs, Young has not seen any deformations in trees as small as 5-cm (2 inches) dbh.
Commercially made (or ready-made) bands are available, but they are more expensive and must be made to the length necessary for a specific tree. If bands specific to individual trees are purchased, it could be difficult to keep track of each band/tree combination in the field. If all bands are constructed to the same length, a lot of wasted band material will result when each band is cut to fit individual trees.
A potential source of error with dendrometer bands is the amount of time needed for the band to settle onto the tree stem (Auchmoody, 1976; Fuller and others, 1988; Keeland and Sharitz, 1993). During this time period, the band reading will underestimate actual growth, even if the stem is properly smoothed and the spring has sufficient tension. The length of time required for the band to settle varies with many factors that may include species, bark thickness and rigidity, canopy position, soil nutrient and/or moisture status, the rate of radial growth, etc. Growth rate seems to be the most important factor. The best way to deal with this problem is to install the bands a full growing season prior to the initiation of measurements. For fast growing trees a portion of the growing season may be sufficient for the bands to settle onto the tree stems. For some extremely slow growing trees, a period of 5 years may be insufficient for band settlement (Keeland, USGS, personal observation).
Additional issues to consider when banding trees include a clear approach to the tree and where around the circumference of the stem to place the collar. If you will be measuring numerous trees on a regular basis (we measured 600 to 700 trees weekly over a period of several years) having a clear and obvious approach to the band will speed the measurement process. The collar should be visible as you approach the tree along the route you will normally take. This may sound obvious, but if it is not thought of during band installation, you will spend time each measurement period searching around the stem for the collar.
In the low light of a closed canopy forest it may be difficult to see the scratch or the graduations on the ruler. Sometimes it helps to carry a reflector (even the white lid from a fast-food coffee cup will work) to help illuminate the scratch.
Many tree stems shrink a little during the winter. If you scratch the bands in the late fall or winter, the scratch can disappear under the collar and you will have to scratch the band again. Once the original scratch appears from under the collar (after sufficient stem growth) you will have to be careful of which scratch to measure to avoid errors.
On fast growing trees, as the stem grows and the scratch moves away from the collar it may become advantageous to make an additional scratch adjacent to the collar so that the ruler does not have to wrap so far around the stem to make the measurement. If a new scratch is made, measurements should be made on both scratches for a few weeks so that an offset between scratches can be made to reduce the possibility of measurement errors.
It is tempting to make new scratches each spring so that all measurements for a given year start from zero, but for slow growing trees this will place multiple scratches adjacent to each other, potentially causing confusion. It is best to make as few scratches as possible on each band.
Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. Government. The authors would like to express their appreciation to Dr. William Conner who taught us how to install and use dendrometer bands.
Auchmoody, L.R., 1976, Accuracy of band dendrometers: U.S. Forest Service Research Note NE-221, 4 p.
Bormann, F.H., and Kozlowski, T.T., 1962, Measurement of tree growth with dial gauge dendrometers and vernier tree ring bands: Ecology, v. 43, p. 289-294.
Bower, D.R., and Blocker, W.W., 1966, Accuracy of bands and tape for measuring diameter increments: Journal of Forestry, v. 64, p. 21-22.
Cameron, R.J., and Lea, R., 1980, Band dendrometers or diameter tapes?: Journal of Forestry, v. 78, p. 277-278.
Cattelino, P.J., Becker, C.A., and Fuller, L.G., 1986, Construction and installation of homemade dendrometer bands: Northern Journal of Applied Forestry, v. 3, p. 73-75.
Clark, N.A., Wynne, R.H., and Schmoldt, D.L., 2000, A review of past research on dendrometers: Forest Science, v. 46, p. 570-576.
Daubenmire, R.F., 1945, An improved type of precision dendrometer: Ecology, v. 26, p. 97-98.
Fritts, H.C., 1976, Tree Rings and Climate: New York, Academic Press, 567 p.
Fritts, H.C., and Fritts, E.C., 1955, A new dendrograph for recording radial changes of a tree: Forest Science, v. 1, p. 271-276.
Fuller, L.G., Cattelino, P.J., and Reed, D.D., 1988, Correction equations for dendrometer band measurements of five hardwood species: Northern Journal of Applied Forestry, v. 5, p. 111-113.
Hall, R.C., 1944, A vernier tree-growth band: Journal of Forestry, v. 42, p. 742-743.
Keeland, B.D., and Sharitz, R.R., 1993, Accuracy of tree growth measurements using dendrometer bands: Canadian Journal of Forest Research, v. 23, p. 2454-2457.
King, S.L., Allen, J.A., and McCoy, J.W., 1998, Long-term effects of a lock and dam and greentree reservoir management on a bottomland hardwood forest: Forest Ecology and Management, v. 112, p. 213-226.
MacDougal, D.T., 1921, Growth in trees: Carnegie Institute of Washington Publication No. 307.
MacDougal, D.T., 1924, Dendrographic measurements, in MacDougal, D.T., and Shreve, F. eds., Growth in trees and massive organs of plants: Washington, D.C., Carnegie Institute, p. 3-88.
Reineke, L.H., 1932, A precision dendrometer: Journal of Forestry, v. 30, p. 692-697.
Reineke, L.H., 1948, Dial gauge dendrometers: Ecology, v. 29, p. 208.
Vissage, J.S., Miller, P.E., and Hartsell, A.J., 1992, Forest statistics for Louisiana parishes: Asheville, U.S. Forest Service, Southern Forest Experiment Station Resource Bulletin SO-168, p. 65.
Yocom, H.A., 1970, Vernier
scales for diameter tapes: Journal of Forestry, v. 68, p. 725.