Magnolia floribunda (Finet & Gagnep.) Figlar

Little known in gardens, this large forest tree of Section Michelia has a great deal in common with M. doltsopa. The two can (in theory at least) be distinguished by the longer stipular scar of M. floribunda (>⅓ the length of the petiole, as opposed to <⅓ in M. doltsopa – Xia, Liu & Nooteboom 2008). It has a more easterly range in China than that better known species, but they meet in NW Yunnan where some collections are hard to identify and may be hybrids (Chen & Nooteboom 1993). The glaucous underside of the leaves in some forms is very attractive, while the silky new growth makes this a very tactile tree. Young specimens have an open, airy appearance and form a pyramidal shape, but become denser and more rounded with age (Hogan 2008).

This may prove to be one of those michelias which evaded cultivation until the late 20th century, and hence falls into the category of experimental garden trees for milder parts of our area. At Tregrehan, Cornwall, several trees of Vietnamese origin are growing steadily and getting better each year, having reached 7–9 m by 2014 (The Tree Register 2021). In Oregon trees of 5–6 m were beginning to flower by 2007 (S. Hogan pers. comm.). A 1996 collection made by Dan Hinkley (DJHC 548) in NW Yunnan is proving vigorous in Vancouver, BC (Wharton 2007) and was not affected by –9 °C in November 2006, when its branches bowed but did not break under a deep load of wet snow (P. Wharton, pers. comm. 2007). Individuals are unlikely to be uniformly hardy. In North Carolina, Tony Avent germinated 3 seeds from Kunming Botanic Garden in 1996, only to lose two seedlings in the first winter. The survivor has formed an open, upright small tree which flowers from mid-January at Juniper Level Botanic Garden; plants raised from cuttings have been distributed under the cultivar name ‘Bridal Bouquet’ (Avent 2021).

However, one large tree at Caerhays Castle, Cornwall, almost certainly from a George Forrest collection, has historically been labelled Magnolia floribunda (Williams 2018), although Owen Johnson (The Tree Register 2021) claims a further three in the same garden. Forrest certainly made collections under this name, but the identification is contested, some believing that all the Caerhays michelias from that era represent variation within M. doltsopa (Williams 2018; Williams 2022). The possibility that one of Forrest’s introductions from Yunnan included ambiguous or hybrid material, as mentioned above, should not be ignored. These are difficult plants to identify.

The species seems a reasonably good long-term bet for suitable mild sites throughout maritime Europe and along the American West Coast, as well as in the milder parts of the southeastern United States.