Use of sonic tomography to detect and quantify wood decay in living trees

Abstract

Premise of the study: Field methodology and image analysis protocols using acoustic tomography were developed and evaluated as a tool to estimate the amount of internal decay and damage of living trees, with special attention to tropical rainforest trees with irregular trunk shapes.

Methods and results: Living trunks of a diversity of tree species in tropical rainforests in the Republic of Panama were scanned using an Argus Electronic PiCUS 3 Sonic Tomograph and evaluated for the amount and patterns of internal decay. A protocol using ImageJ analysis software was used to quantify the proportions of intact and compromised wood. The protocols provide replicable estimates of internal decay and cavities for trees of varying shapes, wood density, and bark thickness.

Conclusions: Sonic tomography, coupled with image analysis, provides an efficient, noninvasive approach to evaluate decay patterns and structural integrity of even irregularly shaped living trees.

Keywords: Argus PiCUS 3 Sonic Tomograph; ImageJ; acoustic tomography; tropical trees.

Fig. 1.

PiCUS 3 Sonic Tomograph sensor placement on tree trunks of different shapes. (A) Regularly shaped tree using recommended equidistant spacing of sensors (T-shape with numbers 1–12). All chords (e.g., 1–4, 4–8, 8–12, 9–10) including the diameter chord (1–7) are appropriately complete (shown as dashed lines) within the outline of the trunk. (B) Oval trunk with decay cavity (rough shaded region) open on one side. All sensors are placed around the outline of the trunk, avoiding the decay cavity. All chords (e.g., 1–2, 1–4, 1–7, 4–8, 8–12) are complete (dashed lines) within the trunk outline. Chords 1–2, 1–4, and 1–7 cross the internal decay cavity, but are within the outline of the presumed intact trunk. Sound transit time for those chords will be longer than expected for that distance (indicating decay), because the sound must either take a longer path or move more slowly through the less dense medium in the cavity. (C) Incorrect placement of sensors on an irregularly shaped tree trunk. Chords with dashed lines are contained within the trunk outline, but chords 2–12, 3–11, 4–12, and 9–10 are incomplete (shown as dotted lines), because they all pass outside the natural tree outline. Sound transit time will be longer than expected for those chords, and can produce false positives on the tomogram as if they passed through decay cavities. Instead, (D) and (E) show correct placement of sensors for more effective scanning of the same irregularly shaped trunk as in (C) by dividing the trunk into two separate scanning regions (D: sensors 1–12 and E: sensors i–viii) to allow piece-wise scanning. Within each section, all chords are complete (dashed lines), providing robust measures of each part of the trunk. A small shaded region between the two components is not scanned. (F) Cross-section of trunk with buttresses. Fig. 1 (See p. 2). Scanning ideally excludes the buttresses and focuses on the internal core of the trunk, by carefully placing sensors so that all chords are complete (dashed lines). If required, buttresses could be separately scanned by modifying the approach shown in D and E. Buttresses and other irregularities often make measuring particular chord lengths physically difficult (e.g., chord 2–4 in G and H). If the chord (2–4) is extended to a point (marked x) that is more easily accessed with the PiCUS electronic tree calipers (G), the chord length 2–4 can be determined by measuring 2–x and 4–x and then finding the difference between the two. (H) Alternatively, posts can be extended from the sensor points at right angles to the 2–4 chord and parallel to each other, and then the distance between them measured with a ruler. Long-jaw straight calipers can sometimes be used for this measurement if the chord distance is not too great.

Fig. 2.

Images illustrating key elements of proper sensor placement for reliable sonic tomogram study of internal decay of four individuals of living Alseis blackiana trees on Barro Colorado Island, Panama. (A) Alseis blackiana (plot tag 2379) with irregular trunk shape due to buttresses. (B) Tomogram of tree 2379 with sensors improperly placed at the extreme outer and inner points of buttresses (red numbers indicate sensor positions). Note aberrant green and magenta tips of buttresses near sensors 3 and 7; wood sampled using a one-quarter-inch paddle drill bit from area indicating decay showed no sign of discoloration or decay at 0–5 cm or 5–10 cm depth, indicating a false positive. Instead, this was structurally intact wood within the buttress. Position of collection marked with a collection bag on trunk in A. (C) Proper placement of sensors on tree 2379 (placed as suggested in Fig. 1F) excludes the buttresses from the tomogram. Outline of the overlayed scan geometry is visible as the gray area in B. (D) Tomogram of A. blackiana 4528, showing a large area of decay. Wood sampled showed clear discoloration and decay at 5–10 cm, but not at 0–5 cm, as suggested by the tomogram. (E) Tomogram of A. blackiana 4014 showing distributed areas of minor decay. Wood sampled shows clear discoloration and decay in suspected areas, but apparently healthy wood from 5–10 cm depth, in accordance with the tomogram. (F) Tomogram of the apparently healthy A. blackiana 7195. Wood samples from between sensors 6 and 7 were apparently healthy from 0–10 cm deep.

Fig. A1-1.

The PiCUS 3 Sonic Tomograph attached to a tree with sensor cables.

Fig. A2-1.

Screen capture to illustrate protocol to use ImageJ to quantify decay (steps 2–9).

Fig. A2-2.

Screen capture to illustrate protocol to use ImageJ to quantify decay (steps 10–14).

Fig. A3-1.

In most cases, decay columns are greatest near the base of the tree and decline with height, as for this Casearia arborea (plot tag 2709) scanned at 10, 100, and 150 cm above ground level.

Fig. A3-2.

Examples of nickel upholstery nail (top, with head intact [right] or flattened for better sensor contact [left]) for thin-barked tree species, and larger steel roofing nails (middle and bottom) used for thick-barked tree species.