Various textbooks have described the diversity of trees (Oldfield et al., 1998;Grandtner, 2005; Tudge, 2005) and another series of books have treated theinternal anatomy, physiological functioning and defence mechanisms of trees(Blanchette and Biggs, 1992; Butin, 1995; Wagner et al., 2002; Schweingruber etal., 2006). In this chapter we describe trees from a different angle –how different tree properties have strong implications for the speciescomposition of saproxylic species after the tree has died.

Like other topics in ecology, the host-tree associations of saproxylics must beunderstood in an evolutionary context. In Chapter 10 we examine the evolution ofwoody plants with an emphasis on structural innovations. Here we simply mentionthat the origin of coniferous trees dates from about 310 million years back intime, while different broadleaved trees first evolved 100–120 millionyears ago (mya). Thus, coniferous trees and broadleaved trees represent distinctplant groups which differ in many ways.

Conifers versus broadleaved trees

There is a striking lack of scientific review publications that providequantitative information about host-tree associations among woodinhabitingorganisms. In a recent book on the ecology of wooddecaying basidiomycetes (Boddyet al., 2008 ), the topic of host-tree associations was only superi ciallytreated. Only one chapter touched on this subject and quantified the proportionof fungi in Denmark that were specific, strongly selective, or weakly selectivefor different broadleaved tree species (Boddy and Heilmann-Clausen, 2008 ). Thecorresponding chapter on fungal communities in boreal, conifer- dominatedforests did not mention host-tree associations at all. Similarly, a quite recentFrench book on forest insects (Dajoz, 2000 ), with a broad treatment ofsaproxylic insects, did not deal with host-tree association patterns explicitly.Dajoz was, of course, aware of such associations, since his chapter on communitydevelopment during the decomposition process was subdivided into sectionstreating different tree species individually. Such treatments of communitycomposition in wood from separate tree species are quite common (see Chapter 6). But it is only when information is brought together from many sources that wecan get a broader overview of host-tree ranges and the specific preferences ofwood-inhabiting species.