For size comparisons it is important to compare like things to like things. If you mix both single and multitrunk trees together you are mixing different things. A tree for champion purposes needs to be defined as a single trunk, meaning it has a single pith at ground level. Multitrunk trees are worth measuring and documenting, but they should not be lumped together with single trunk trees for comparison purposes.
If you start talking about tree genetics and growing from the same root, then that begins a myriad of complexities that make the situation even worse. Two trunks from a rootstock may be genetically the same, but so are all of the clonal trees in colonies like the Pando Aspen Colony. Since they are all genetically the same and may be interconnected through the roots, should a tape be wrapped around all 47,000 stems covering 106 acres and call that the girth? In multitrunk trees there typically is a pinched section of bark between the trunks, clearly indicating they are separate trunks.
It is better in both practical terms and conceptually to define a tree as a single stem, even if the larger organism may have multiple trunks. The examples of unusual multitrunk specimens, trees like banyans, clonal colonies, self grafted series, fallen trees with limbs sprouting, etc. should certainly be documented, but each on their own merits, rather than lumping them in with measurements of single trunks.
If there was one aspect I would want to see cleaned up in champion tree lists, it is the persistent inclusion of multitrunk trees. They should not be on lists that are designed to compare the biggest individual - read single trunk - trees. This is something that could be resolved with better adherence to a champion guidelines that specifies single trunk trees only. I would even be in favor of a separate list for multitrunk trees, or trees with other unusual forms, but the two categories should not be mixed.
This something that can be fixed on champion tree lists. Multitrunk trees should be removed from consideration. This action does not require any expensive equipment on the part of those people measuring the tree. It does not require any special knowledge on the part of the measurers. It does not exclude anyone interested in measuring trees from the process. It would assure the integrity of the lists and reward people who find the actual giants of a tree species, rather than game playing by people who would nominate unworthy multitrunk trees as champions. Nothing annoys me as much with champion tree programs as allowing multitrunk trees to be included in the same listing category as single trunk trees.
There are examples of individuals or groups of individuals using faulty tangent based height measurement processes simply because these have yielded taller heights than more reliable sine top/sine bottom laser rangefinder/clinometer measurements available to them, but these are examples of cheating on the part of these individuals rather than a problem. with the champion program itself.
Will Blozan recently posted some examples of multitrunk trees:
- Ohio champ cottonwood.jpg (56.97 KiB) Viewed 868 times
Ohio champion cottonwood
And here are some pith trace examples:
Ohio champion sycamore
Seven sisters live oak clump
The pith lines need to merge before ground level for something to be considered a single trunk tree. If there is more than one pith line at ground level it is a multitrunk tree. If there is only one pith at ground level, then it is a single trunk tree. Low branches could come out below 4.5 feet, but above the ground level and the tree still be a single trunk tree.
In the tree measuring guidelines, (all three of the documents, the original version, the one published in the Bulletin, and the updated version) NTS SP #1a Tree Measuring Guideline of the Eastern Native Tree Society -Revised http://www.nativetreesociety.org/measure/Tree_Measuring_Guidelines-revised1.pdf Will Blozan writes:
"I use a “pith test” to define what a multitrunk tree is. If the tree has more than one pith at ground level it is a multiple-stemmed tree. Note I did not say 4.5 feet above the ground. This is because the 4.5 foot height is a forestry standard and is an arbitrary and convenient place for most people to measure a tree. Some trees, like flowering dogwood or rhododendrons, may branch well below 4.5 feet but have a single pith at ground level. In the case of such trees, I would measure the narrowest point below the lowest fork. More detailed discussions of how to measure multitrunk trees and trees with other odd forms is presented on the ENTS website."
As for the question of whether a particular tree is a double or single trunk, there will be arguments between experienced measurers about whether a particular tree is a double or a single. Many old doubles have grown together so that the trunk is regular in form and on the face of everything no longer appear to be doubles. The opposite situation s where there is a large low protruding branch. If the tree and branch grow large enough, the low branch appears to look much like a second trunk. When faced by wind and weather it is possible that these may split along the attachment line to look as if they are two trunks. In many cases there is sufficient doubt that the only way to know for sure would be to cut the tree down at ground level and see what the cross section shows.
Some people consider it being conservative to consider something a double if they can't tell for sure otherwise. I think this corrupts the data set more so than an occasional misclassified tree. For anyone measuring trees in the field, I would recommend they make detailed observations in the field, and then go with the best guess as to whether the tree is a single or double, and report that. Field inspection trumps photos except in the most egregious examples. This is not to say that if someone else goes out and looks at the tree will reach the same conclusion, but we hope so. Measurers should try to build in their mind characteristics that might distinguish singles from double or multitrunk trees, and apply these mental lists to what they are seeing in the field.
We are not defining whether something is a single or multitrunk tree based on genetics. The multitrunk tree may be growing from the same root mass and have identical DNA in all of its trunks. For measurement purposes we are classifying a multitrunk tree as a different measurement category than a single trunk tree because of its growth pattern, not because of different genetics. There may be some cases where there actually are two different specimens of the same species of tree growing together to form a fused mass, but these would be I would guess an extremely rare circumstance. There are occasional examples of two different species growing together - the Hugging Trees in the multitrunk tree classification scheme I previously proposed http://www.nativetreesociety.org/multi/index_multi.htm. I would expect that hugging trees of different species would be more common than two different trees from the same species. In any case these should not be considered in the same measurement category as single trunk trees.
Are we becoming splitters or lumpers when it comes to tree measurements? I think I am a splitter as needed to maintain what I see as a valid data set. I want to make sure the big tree lists maintain an internal integrity. On the other hand, I have championed the idea that we should be collecting data on multitrunk trees and trees of other weird forms. That was the point of the article I wrote: Multitrunk Trees, Woody Vines, and Other Forms: http://www.nativetreesociety.org/multi/index_multi.htm I want to include these other forms in our dataset, even if they are not the idealized single trunk model and have proposed ways to measure them. The columns for inclusion of multitrunk trees are in the spreadsheet I wrote, and I have been working with Mitch Galehouse in his implementation of the NTS Trees database so that the multitrunk specimens can be properly recorded. So I would counter that you can be both a splitter and also be pushing for a broader inclusion and representation in the dataset.
There needs to be a balance between lumping and splitting when looking at sets of data. If you lump too many things together then they become a mish-mash of different objects that lack a coherent theme that is useful for expanding your understanding of the set. If there is too much splitting, then each individual is its own class and you can't look at relationships between objects as easily. So really I don't think it is a matter of splitters versus lumpers. We are splitting the data only to the degree needed to make it useful, and further lumping would only hurt the overall goals. I want to keep records for and acknowledge the superlatives of the different forms, but see it as a detriment to mix different form trees together in a single list.
Edward Frank
