Sequences deposited in GenBank are from the strain IMI 052255, also from Kenya, but Cupressus forbesii (Hesperocyparis forbesii) is given as host in the CABI database. This may represent Lepteutypa cupressi, but is probably not conspecific with isolates of Seiridium cupressi from Australia (Cunnington 2007), South Africa and New Zealand (Barnes et al. 2001). Although Swart (1973) interpreted the high morphological variation of Seiridium on Cupressus only as variants of a single species, Barnes et al. (2001) separated S. cardinale, S. cupressi and S. unicorne as distinct species using histone and tubulin sequences and determined that the major pathogens are S. cardinale and S. cupressi. Cunnington (2007) correlated tubulin sequences with conidial morphology and determined S. cupressi as the common cause of cypress canker in south-eastern Australia. He also maintained that the sexual morph of S. cupressi is Lepteutypa cupressi, and that the sexual morphs of the other Seiridium species are not known. However, there is still no evidence of a direct molecular connection of the African L. cupressi and the Australian S. cupressi and its sexual morph. Unfortunately, noAfrican material of the sexual morph, Lepteutypa cupressi, was included in these analyses, and no histone and tubulin sequences are yet available for the African L. cupressi. For the strainATCC 48158 of S. cupressi originating from New Zealand, ITS as well as histone and tubulin sequences are available. In the phylogenetic analyses of Barnes et al. (2001) and Cunnington (2007), this strain is contained within the S. cupressi clade. However, in our phylogenetic analyses based on ITS (Fig. 1), S. cupressi strain ATCC 48158 is distinct from L. cupressi IMI 052255, indicating that they are not conspecific. If this is the case, it may have the undesirable consequence that the current species concept of S. cupressi is inappropriate, as both S. cupressi and L. cupressi are based on material from the same hosts collected in Kenya. This group therefore requires substantial taxonomic revision, including a representative sampling of well-identified strains as well as additional sequence regions other than ITS-LSU.

In lack of other markers for the great majority of taxa the current phylogenies of Xylariales are primarily based on rDNA sequence data, and resolution of these phylogenies is simply insufficient to allow far-reaching subdivisions into formal higher taxa (Fig. 1). A subgrouping also appears inappropriate, because the Sporocadaceae obviously represent a natural grouping, which has the following common basis: (most) asexual morph genera are acervular coelomycetes with the same type of conidiogenesis and conidium, the latter only differing in colour (hyaline, pale or dark brown), septation (phragmo- or dictyosporous) and the number, orientation and type of appendages (exogenous/cellular or endogenous/cell elongation, branched/unbranched).