First record of Erysiphe elevata on Catalpa bignonioides in the UK

In 2002, a powdery mildew caused extensive patches of fine white bloom on the upper surfaces of leaves of a mature Indian bean tree (Catalpa bignonioides) in the Royal Horticultural Society's Garden, Wisley. In the absence of ascomata (chasmothecia), the assumed pathogen was Erysiphe catalpae, already known in Britain. However, in October 2003 abundant chasmothecia appeared with branched Microsphaera-like appendages, unlike the simple ones of E. catalpae. It conformed to E. elevata (syn. Microsphaera elevata), previously only known in North America (Braun, 1984), but now reported in Hungary (Vajna et al., 2004). The appressoria were mostly single and conspicuously lobed, some being paired, simple and kidney shaped (Fig. 1). Somewhat flexuous foot cells, 18-34 x 5-9 µm, bore 1-2 other cells, 12-22 x 5-10 µm. Cylindrical to elliptical conidia, 21-32 (43) x 9-16 (18) µm (smaller than E. catalpae) matured singly, being typical of Oidium subgen. Pseudoidium, the anamorph of Erysiphe. Scattered or grouped chasmothecia, (84) 95-132 µm bore (1) 4-8 equatorial appendages, 360-700 x 6 µm (up to 9 µm wide at base), becoming flexuous in the upper half, but with bases stiff and straight, hyaline or brownish, thick walled with a rough surface, occasionally with a septum close to the fruiting body. Their apices branched dichotomously, 1-4 times, with simple or knob-like tips when sparsely branched, but recurved when more complex, compact or widely forked (Fig. 2). Chasmothecia contained 3-8 short stalked asci, 45-63 x 25-39 µm with 4-6 elliptical ascospores, 16-20 x 8-11µm.

DNA was extracted as described by Saenz & Taylor (1999), from chasmothecia on Catalpa collected in the UK, Hungary and the USA. The first amplification of the ITS region was with primers ITS5 and P3 and the second with ITS1 and P3 using the thermal cycling conditions described by Hirata & Takamatsu (1996). The ITS sequences of the UK and Hungarian isolates (accession numbers AY587012 and AY587013 respectively) were identical and showed only one base pair substitution with the American isolate (accession number AY587014). As these are the first submitted to GenBank, comparisons with other E. elevata isolates was not possible. The Wisley material was virtually identical to E. elevata collected in 2002 in both Hungary and Wisconsin, USA, supporting the contention that E. elevata has only recently spread to Europe from the USA (Vajna et al., 2004).