"d/Ddom is relative stem size, where d is the current stem diameter at breast height and DDom is the average stem diameter of the 100 thickest trees ha-1". You could replace tree diameter directly with tree height.

Reference: Piispanen, Heinonen, Valkonen, Mäkinen, Lundqvist, Saranpää: "Wood density of Norway spruce in uneven-aged stands". CJFR 2014, 44(2): 136-144, 10.1139/cjfr-2013-0201.

If species variation is large including species from various genera, the situation is more complex. Probably you could ask Sauli Valkonen.

Sincerely,

Riikka Piispanen

I would agree with Riikka.  Defining dominance in terms of height is commonly done in many silvicultural thinning studies and easily justified by the literature.  Assuming that you have tree canopy width and depth measurements, you could probably do something similar with canopy volume.

As for expressing dominance for each species, the mean dominance would probably be enough if you are only interested in tree-level patterns.  If, instead, you are interested in more community-level patterns, you could probably calculate something akin to an importance value.  Traditionally, this is defined as relative basal area + relative density for a plot (+ relative frequency across plots if for a stand) scaled to 100.  I don't see why you couldn't replace relative basal area with relative aggregate height (i.e., the sum of heights of species A/the sum of height of all trees in the plot).  Others can comment, but as a reviewer for manuscripts in this area, I would find that approach justifiable.

Thank you all, but to Michael Saunders, do you have any reference to work that have done it on playing around with the traditional Importance value index?

I agree with the previous contributors. Canopy area can be used as a measure of dominance just like basal area. I only wish to state that the relationship between the diameter at breast height and other parameters such as height and volumes can get complicated when dealing with savanna vegetation which then to be multi-stemmed and branches some 50 cm above the ground. DBH will most likely be somewhere in the canopy for some species.

I second Riikka. Tree height per dominant height would be my choice. The sum of heights of trees larger than the subject tree would be another alternative. It has worked very well in describing competition (=dominance) in uneven-aged mixed stands in the temperate region.But Riikka already pointed towards an important feature here: These are static measures, not accounting for dynamic components like late-successional species emerging from understories and gradually taking over.

Thank you all for the insightful ideas