The following post turned out to be much longer than intended. The short version is that this month's (February 2018)
Arborist News magazine had two separate, independent articles that touched on whether or not removing deadwood is beneficial to the health of a tree. One stated that it was (p.30-31), the other stated that it was not (p. 14). Although both provided a reference for their statement, neither gave any substantive, material evidence. Both acknowledged and discussed various environmental and ecological reasons for leaving deadwood.
A third article (p. 26) mentions the importance of "...dead, dying and decaying trees and branches" for wildlife habitat, but gives no opinion either way about how dead branches affect the health of an individual tree.
------------------
In an article that appears in this months
Arborist News magazine (Arborists and Wildlife: Retaining Trees for Wildlife Habitat,
Arborist News Volume 27 Number 1, February 2018:12-19) Brian French (p. 14) states
According to Dr. Edward Gilman, author of An Illustrated Guide to Pruning (2002), there is no evidence at this time that supports the statement "Removing deadwood improves tree health".
The quotes are French's, but he gives no reference for that specific statement by, nor the general opinion of Dr. Gilman.
In a different article in that same issue of
Arborist News (Deadwooding Tree Canopies: Cosmetic or safety surgery?
Arborist News Volume 27 Number 1, February 2018:30-33), G.M. Moore (p. 30) states
The biological advantages of deadwooding not only relate to the development of a sound canopy, but also to compartmentalization, wound closure, and the minimization of pest and disease attack.
That paragraph goes on in more detail and then provides a reference, of which Moore is the co-author: Ryder, C., and G.M. Moore. 2013. The aboricultural and economic benefits of formative pruning street trees.
Arboriculture and Urban Forestry 39(1):17-24.
But this reference is both irrelevant and misleading as support for any biological advantages of the removal of deadwood. The AUF article states that
Research was undertaken to determine the need for, and costs of, formative pruning recently planted street trees.
...Health was similar for all species, but form and structure varied. Data showed that
codominant stems (68%) and included bark (40%) in the canopy or trunk were by far the most common structural defects.
...The pruning required
to rectify these structural defects was recorded and then multiplied by a time factor for pruning with secateurs (hand pruners),
a handsaw, or a pole pruner. Total time was then converted to an economic cost using current labor market prices. (Abstract, p. 17)
The purpose of this study was to look not only at how formative
pruning can improve a tree’s structure, but to quantify the financial benefits of removing structural defects early in a tree’s life. (p. 17)
The trees selected were young (3–5 years of age) and under
6.5 m in height. (p. 17)
Structural defects on which data were collected (Table 3) were
co-dominant stems, included bark, decay, deadwood and crossing
or rubbing branches...(p. 18)
...epicormic shoots, broken stems, and deadwood occurred at low levels. The low value [i.e. low incidence] of deadwood is a reflection that the street trees were young and in good health. (p. 19)
Analysis of data indicates there is a strong rationale for pruning
when trees are young. (Discussion, p. 21)
Formative pruning is a cost effective way of reducing structural
defects in trees and improving the quality of the structure
of street tree populations. Overwhelmingly, the major structural
fault recorded was the presence of codominant stems
in the canopy or trunk, often exacerbated by the presence
of included bark. All other structural defects recorded were
in much smaller numbers. (Conclusion, p. 23)
The article by Ryder and Moore focuses on a study that investigated the time and cost benefits of doing structural pruning on young street trees, and the article itself states that deadwood was a very small part of that pruning. There is no research or discussion on any possible biological advantages of removing deadwood. Additionally, the AUM study looks only at young (3-5 year old) trees, not at established or older or mature trees, which is the focus of the AN article.
In the AN article Moore also states (p. 31)
...the removal of deadwood may reduce available habitat for wildlife and for beneficial organisms that deadwood and hollows in trees provide. There is an important balance to be maintained between removing deadwood and preserving important wildlife habitat within the urban forest. deadwooding should be done with a clear purpose rather than as a matter of habit, and other environmental values must be considered.
He provided no specific reference for the statement above. He ends the article with a balanced assessment:
Because deadwood has environmental value in terms of the provision of habitat for wildlife and microorganisms, such as fungi, unnecessary removal of deadwood should be avoided. However, there are many good reasons for implementing a deadwooding program as part of urban tree management. (Conclusion, p. 32
There were a total of 12 references for Moore's AN article, but the only one I looked at was by Ryder and Moore, discussed above.
In yet a third article (Practical Advice for Wildlife Protection During Tree Care Operations,
Arborist News Volume 27 Number 1, February 2018:326-28), Ryan Gilpin mentions the importance of "...dead, dying and decaying trees and branches" for wildlife habitat, but gives no opinion either way about how dead branches affect the health of an individual tree.
It's good to see discussion on the importance of wildlife, ecology, and the environment.