The Investigation and Diagnostic Centre of the Plant Health and Environment Laboratory of the former Ministry of Agriculture and Forestry (now Ministry for Primary Industries) did a molecular analysis on DNA extracted from ascomata of this fungus but found no match with any of the registered sequences. A search of the literature and use of various keys led us to the revised family Acrospermataceae (Minter et al. 2007) and the genus Acrospermum. A search in Index Fungorum yielded the names of 66 species and subspecies in Acrospermum. We then began to examine the protologues of each of these names. That for Acrospermum gaubae, described by Petrak (1955) from Banksia ericifolia from Pigeon House Mountain in the Budawang Range, New South Wales, Australia, matched our collections well, except that the New Zealand specimens are larger. A comparison is given in Table 1.

This is the first record of A. gaubae in New Zealand; indeed, apart from the initial report of Petrak (1955), it appears to be the only other record of the fungus in the literature. Acrospermum gaubae is purely superficial and causes no damage to its hosts.

This fungus was originally described from Hawaii on Hibiscus sp. as Microthyriella hibisci (Stevens 1925). According to Müller and von Arx (1962) Microthyriella Höhnel is a synonym of Schizothyrium Desmazières. This view is generally accepted (Kirk et al. 2008) and most species previously placed in Microthyriella have been transferred to Schizothyrium. Müller and von Arx (1962) also pointed out that Microthyriella hibisci, did not follow the type of Microthyriella and was wrongly placed in that genus. Microthyriella rickii (Rehm) Höhnel, the type species of Microthyriella (in common with species in Schizothyrium), does not have an ostiole, the whole surface of the shield splitting and allowing the ascospores to escape. As Microthyriella hibisci has a distinct ostiole, it does not belong in Microthyriella and has to be redisposed to a genus other than Schizothyrium. Morphologically, M. hibisci belongs to the Micropeltidaceae and all keys (Batista, 1959; Müller and von Arx, 1962; Luttrell 1973; von Arx and Müller 1975) show that it is best placed in Dictyothyrium Theissen. Molecular analysis yielded no close matches to any records in Genbank.

The thyriothecia of the fungus are entirely superficial, the shield-shaped fruiting bodies merely sitting on the leaf surface without a basal wall. No superficial mycelium was seen and there is no penetration of the host tissue. The concentric arrangement of the thyriothecia (Figure 3A) makes the fungus easily recognisable. It is not uncommon in New Zealand. It has also been recorded (as Microthyriella hibisci) in Australia (Walker and McLeod 1971).

The genus Septoriella is characterised by mucoid appendage-bearing pale brown multiseptate conidia formed on holoblastic conidiogenous cells (Sutton 1980). The genus was reviewed later by Adrianova and Minter (2007) who recognised 11 species in the genus and provided a key to them. Only two of the species, S. halensis and S. phragmitis Oudemans have conidiomata (160–220 μm wide) and conidia (fusiform, 30–45 μm long and 5-7-septate) that are similar to those of the New Zealand fungus (Table 2).

A comparison of the three descriptions shows that the New Zealand fungus cannot be distinguished from S. halensis. Use of the key to accepted species of Septoriella (Adrianova & Minter 2007) also leads to S. halensis. This is the first record of a species of Septoriella in New Zealand. Most of the eleven taxa accepted in the genus are from temperate areas, the only exception being S. rockiana (Petrak) Nag Raj from Hawaii. Ten species occur on monocotyledon families and only one, S. viciae Adrianova & Minter, has been recorded on a plant (Vicia unijuga A.Braun) belonging to a dicotyledon family (Fabaceae). This record of S. halensis on P. insignis adds another dicotyledon family (Asteraceae) to the list. Most species of Septoriella have been recorded on dead or dying leaves or culms and are regarded as saprobes. The New Zealand collections of P. halensis were from leaf spots on living leaves of P. insignis but whether the fungus was acting as a pathogen or not is not known.