It is correct to construct tree species-based rarefaction curves for birds?

Question

If I register all the species of birds that uses sp1 tree, then, all the species of birds that use sp2 tree, and so on. With this dataset, I can made a rarefaction curve using "number of tree species" on X axis of the rarefaction graph, instead of using "number of samples" Is this incorrect? Does anyone knows a paper that made something similar?

Answer 1

Unfortunately, I have to say "no". A rarefaction curve is used to answer the question "how many species are in this ecosystem" without having to count all of the species (which is impossible). The surveyor takes samples of different sizes, meaning they survey 10 individual birds, then 20, then 30, and so on, which represent the whole ecosystem, rather, what you would measure if you were surveying 10, 20. and 30 individuals from the entire ecosystem. I stress this because every tree is its own ecosystem and only represents the ecosystem of that tree. The ecosystem of two tree species is different from the ecosystem of that one. Not one of your tree species nor any combination of tree species should be the same as what you would find for another. Each sample must represent the whole. Therefore, you can't use one taxon to build a rarefaction curve for another taxon.