Aria (Pers.) Host

Aria contains numerous species, ranging from tall trees to shrubs, often with smooth, pale bark. It is characterised by ovoid, conical buds with many spirally arranged scales and simple, toothed, or more or less lobed leaves which often have a persistent white or greyish tomentum beneath. In the leaf the lateral veins run straight out to the leaf margin (craspedodromous) or bend just before reaching the margin, forming a loop (camptodromous). Inflorescences are paniculate or corymbose. The flowers are white, 2– to 5-merous, and with a campanulate hypanthium. The fruits (pomes) are often red but may be yellowish, greenish or russet-coloured. The calyx may or may not persist on the fruit. In the species with persistent calyx lobes the fruits are never softly fleshy. Throughout Aria the fruit flesh is heterogeneous, giving a dryish granular texture, and when ripe the fruits taste sweet without the bitterness found in the fruits of almost all species of Sorbus s.s. The type species of the genus is Aria nivea Host, best known as Sorbus aria (L.) Crantz, the Whitebeam.

In addition to the basic distinction of the leaf shapes and fruit characters, in most cases the two genera can also be distinguished by the form of their buds, and so can be told apart in winter – Sorbus having mostly elongated ovate-conical buds, one scale extending most of the length of the bud and usually few scales evident, while in Aria the buds are usually less elongated and have several bud scales evident on the surface, but no one scale extending the full length of the bud. The apomictic microspecies of hybrid origin are intermediate, with more deeply lobed leaves than in species of Aria, and those that have leaves with free pinnae have a terminal leaflet larger than, and of a different shape from, any of the free pinnae, distinguishing them from species of Sorbus, in which the terminal leaflet is more or less the same size as the laterals.

As a native genus Aria is confined to the Old World, with no representatives in North America, but occurring through temperate Eurasia from the western seaboard of Europe and North Africa to Japan. In the east of this range it is represented by a number of A. nivea-like, large-leaved, rather ill-defined diploid species in the Himalaya and western China, and a few species that form a link with those that used to be assigned to Sorbus subgenus Micromeles. The breeding systems in Aria are just like those in Sorbus s.s., as outlined above.

Aria as now interpreted includes species previously placed in Sorbus subgenus Micromeles. These have sometimes been considered sufficiently distinct to be placed in the separate genus Micromeles Decne., on the basis of their inferior ovaries and deciduous calyces (Kovanda & Challice 1981) as well as their general appearance. However, Robertson et al. (1991) argued that the heterogeneous flesh of the fruits in all species of both groups, and lack of consistency in the distribution of other characters (including those used by Kovanda & Challice), made it impossible to maintain the separation of Micromeles, even at the subgeneric level. Several familiar species (A. megalocarpa, A. folgneri, A. hemsleyi) show characters of both groups.

It is clear that modern phylogenetic studies of both Sorbus s.l. and Aria s.l. are long overdue, but at present we have only morphology and cytology to work with. Hugh McAllister, who kindly supplied text that forms most of this introduction to Aria, recognises three groups of species that appear to represent the most natural divisions within the genus. No infrageneric classification is as yet available in Aria.

Group A

Equating to Sorbus subgenus Aria Pers. section Aria (Pers.) Dumortier

European and west Asiatic species related to Aria nivea (Sorbus aria), with fruits that are wholly red when ripe and seeds that are round in cross-section. The embryos of these seeds contain germination inhibitors as they do not germinate on excision. Almost all have leaves that are white-tomentose beneath.

This group occurs from western Europe to the Caucasus, is the only group found in this area, and does not appear to overlap in distribution with the other two groups. Within the group, diploids (only A. nivea, possibly S. umbellata and Chamaemespilus alpina) are sexual and polyploids apomictic. Apomicts have hybridised with A. nivea, S. aucuparia, Torminaria torminalis and Chamaemespilus alpina to produce a vast range of apomictic microspecies (100 named in Flora Europaea), which are often grouped with the pure aria types under Sorbus subgenus Aria or treated as the arranensis and latifolia aggregates. New taxa within this group are described regularly, with every crag, gorge and ridge in some areas supporting distinct microspecies. The Avon Gorge near Bristol in southwestern England is the most intensively studied sorbiferous part of the world, with nine native and four introduced taxa growing on its limestone cliffs (Rich & Houston 2006). Five of the natives are British endemics, three of these known only from the Avon Gorge, including White’s Whitebeam S. whiteana T.C.G. Rich & L. Houston, described in 2006. The abundance and diversity of such local taxa seems to depend somewhat on research effort (Avon Gorge is under the beady eyes of researchers from the universities of Bristol and Cardiff, for example).*

Group B

Equating to Sorbus subgenus Aria Pers. section Micromeles (Decne.) Rehder OR subgenus Micromeles (Decne.) J.B. Phipps, K.R. Robertson & Spongberg, in part

Himalayan and east Asiatic species with usually more or less spherical fruits that never become uniformly red or orange, the skin of the fruit often containing much chlorophyll when ripe, and which may be densely covered with lenticels (russeted), thereby resembling an apple or pear (Malus or Pyrus) rather than fruit of European species of Group A. The seeds are always laterally flattened and radially elongated, as in Cormus, and the embryos germinate on excision. All species that have been examined are diploid and sexual.

The concept of this grouping is new and it needs refinement, as it probably contains elements from different lineages (H. McAllister, K. Rushforth, pers. comms. 2008), but this will require further study. Species with white-tomentose leaf undersides mostly have persistent calyces and free carpel apices (these were formerly referred to subgenus Aria), while those with more or less glabrous leaf undersides mostly have deciduous calyces and fused carpel apices (and were formerly referred to Micromeles); both craspedodromous and camptodromous venation are found in this latter subgroup.

* Fourteen new taxa from Britain and Ireland have been described in 2009 (Rich & Proctor 2009, Rich et al. 2009).

The existence, however, of species with a mixture of characters of these two subgroups makes it impractical to maintain separation between them: A. megalocarpa has a persistent calyx but more or less glabrous leaf undersides and fused carpel apices, for example, while A. hemsleyi has deciduous calyces but white-tomentose leaf undersides. (As pointed out by Robertson et al. (1991), Malus and Pyrus include species with and without deciduous calyces, and in some species of Malus leaves on long shoots have craspedodromous venation while short-shoot leaves have camptodromous venation.)

Group C

Equating to Sorbus subgenus Aria Pers. section Alnifoliae (T.T. Yu) Aldasoro, Aedo & C. Navarro OR subgenus Micromeles (Decne.) J.B. Phipps, K.R. Robertson & Spongberg, in part

For example, Aria alnifolia, A. folgneri, A. japonica, A. yuana, A. zahlbruckneri. East Asiatic species with oblong fruits that become a distinctive translucent red or orange on ripening, the seeds of which are pear-shaped and round in cross-section, and the embryos of which contain germination inhibitors and do not germinate on excision. Calyx lobes are persistent or deciduous; leaf venation craspedodromous. All species that have been examined are diploid and sexual.

The distribution of this group overlaps with that of Group B but no hybrids between them have been reported from the wild. (Hybrids between A. nivea and A. alnifolia have been found in cultivation: H. McAllister, pers. comm. 2008.) Aldasoro et al. (2004) provided the first published recognition of it as a natural species grouping, and combinations in Aria are available for most species (Ohashi & Iketani 1993). Aria alnifolia is considered to be one of the few sorbi to be a success in the cooler parts of the United States (Jacobson 1996, Dirr 1998). Adding the evidence from A. yuana, it seems that these members of former subgenus Micromeles are among the best options for Sorbus s.l. in American horticulture. All appear to be totally self-incompatible, so that single isolated trees bear only small fruits that are much less attractive than the large fruits they produce if cross-fertilised. They should therefore always be planted in groups containing more than one clone, with spacing to ensure that at least two clones can be grown to maturity. One option might be to graft two clones to a single stock, to ensure a free-fruiting specimen.

The protean diversity of Aria and its hybrid derivatives in Europe and southwestern Asia makes a comprehensive account almost impossible in a book of this nature; there is an almost endless pool of local taxa available to be brought into cultivation, each with its own alluring name, although perhaps not really very distinct. The value of these trees in the garden and landscape is undoubted, though they often fall just short of the first rank. Their prime moment is often the unfurling of the leaves in spring, when the white undersides show to best advantage on the whitebeams, and some Asian species (Micromeles) flush with effective red or bronzed new growth. Their autumn efforts are variable. Most of the allies of Aria nivea and its derivatives are disappointing then, but the Asian species can colour very well.

Described more fully below, though with abbreviated cultivation notes, S. hajastana, S. takhtajanii and S. tamamschjanae are drought-tolerant small trees from eastern Turkey and Armenia, initially introduced to cultivation through seed sent to Ness Botanic Gardens by Dr Eleanora (Nora) Gabrieljan of Yerevan, Armenia, who named them (H. McAllister, pers. comm. 2007). They were distributed to other botanical gardens in Britain from Ness, and all three may be seen in the collections of Kew and Edinburgh, for example. Sorbus armeniaca Hedl., S. kusnetzovii Zinserl. and S. subfusca Boiss. are whitebeams from the same general area also occasionally found in botanical gardens. All do well in the British Isles, and while not particularly outstanding, could be grown where smallish, slow-growing whitebeams are required, for their attractive red fruits and leaves with white undersides.

Allen Coombes has made us aware of an interesting range of Aria species (or hybrids) being cultivated commercially in Hungary and evidently valued as good ornamentals and street trees. The following taxa were offered by the nursery Alsótekeresi Faiskola Kft. of Balatonvilágos, Hungary, in their 2006–2007 catalogue (the brief descriptions are derived from this source): S. bakonyensis Jáv. (a small, bushy tree with orange-red fruits); S. bodajkensis nom. nud. (a medium-sized, robust tree suitable for landscaping or as a street tree); S. borbasii Kárpáti (bushy, with early-ripening purplish fruits; can be grown as a hedge); S. borosiana Kárpáti ‘Alba Regia’ (a state-approved selection, forming a small tree with very dark green leaves for landscaping; the fruits sweet and delicious); S. decipientiformis Kárpáti ‘Vállus’ (also a state-approved selection, forming a slender conical tree, again with delicious fruits); S. degenii Jáv. (a Hungarian native, allied to Torminaria torminalis and producing edible fruits); S. eugenii-kellerii Kárpáti (a small tree native to the Vértes Mountains of western Hungary, producing ‘marvellous little vermilion apples’); S. ‘Hainburg’ (selected near Hainburg, Austria; medium-sized, very tolerant of calcareous soils and dry conditions); and S. incana Hedl. (a small, thick-crowned tree of hybrid origin, from Scandinavia). Several of these are in cultivation at Wageningen in Holland, from where they have been distributed over many years through the botanic gardens seed exchange. Some are growing well at Ness, near Liverpool (H. McAllister, pers. comm. 2008), where the Sorbus form a National Plant Collection. Another, for the whole group, is at the Jodrell Bank Science Centre, Cheshire. There are also National Plant Collections of British Sorbus maintained by Lord Ridley at Blagdon, Northumberland and by Glyndwr Marsh at Cadnam, Hampshire, while Aria and Micromeles collections are maintained by the East Durham & Houghall Community College, Durham, and at Winkworth Arboretum, Surrey.

TROBI includes records on a number of taxa belonging to Aria Group A in cultivation in British arboreta, mostly represented by solitary specimens in the Sir Harold Hillier Gardens. Their names are noted here, but information on them will have to be sought in regional floras: S. adamii Kárpáti, S. barthae Kárpáti, S. danubialis (Jáv.) Kárpáti, S. latissima Kárpáti, S. leptophylla E.F. Warb., S. pseudobakonyensis Kárpáti, S. semi-incisa Borbás, S. simonkaiana Kárpáti, S. thaiszii (Soó) Kárpáti, S. zolyomii (Soó) Kárpáti. A further selection is cultivated in the Netherlands, according to the catalogue of Plantentuin Esveld (2004–2006), which mentions the following (not all of which are offered by the nursery): S. dacica Borbás, S. dubia Hedl., S. pontica Zaik., S. pseudolatifolia Boros, S. redliana Kárpáti, S. retroflexis (author not traced), S. stankovii Juz., and S. taurica Zinserl. No doubt a more diligent search of arboretum catalogues would reveal many further names.

An attractive species of unknown origin is S. croceocarpa P.D. Sell, a member of the S. latifolia group with small, weakly lobed leaves and bright orange-yellow fruits in autumn. This has become widely naturalised in Britain. All of the named and some of the un-named native British Sorbus species are in cultivation at Ness, where S. hibernica and S. porrigentiformis from the Menai Straits in northern Wales are particularly attractive as heavy-fruiting small versions of A. nivea that (being apomictic) can be propagated by seed. Also particularly promising horticulturally are slow-growing small trees from southern Europe of apomictic tetraploid origin, including, for example, S. aff. graeca from Sicily and the Sierra de Cazorla in Andalucia, southern Spain. Like the Armenian microspecies noted above, these are very heat- and drought-tolerant.

In the mountains of China and northern Vietnam many interesting Aria taxa are to be found, mostly attributable ‘in old money’ to Micromeles. These are proving to be of great interest to collectors visiting the area and are trickling into cultivation. Their taxonomy is extremely uncertain, however, and names should probably not be taken as definitive. One of the best is a taxon identified as S. subulata (J.E. Vidal) T.H. Nguyên & Yakovlev (syn. S. verrucosa Rehder var. subulata J.E. Vidal) and described as being a small tree with arching branches, bearing neat, heavily veined leaves on long petioles and crops of yellowish orange fruits. The leaves emerge glossy bronze and stay that colour for a long time; in autumn they turn good shades of orange and red (Crûg Farm Plants 2007–2008). It was introduced by Keith Rushforth in 1992 and has made 5 m in Devon. It has done well in northern Wales, and also for Dan Hinkley in Indianola, Washington. In an e-mail sent from northern Vietnam (to JMG, in October 2006), Hinkley wrote: ‘It is absolutely lovely in new leaf, rather red-copper and one of the earliest to leaf out. I offered it through Heronswood for some time but only under a collection number (HWJ 99579, collected 1999; HWJ 3925, 2003). It grows nearby here in Sa Pa; the fruit are a bitch to clean and as usual with Sorbus, very few seed inside each fruit.’ On reading this Hugh McAllister wrote (pers. comm. 2006): ‘I agree with this comment if you try to clean seed from hard fruits – it is best to let them ripen and soften in a polythene bag before trying to extract the seeds.’

Several other northern Vietnamese species are noted for their attractive spring flush. Sorbus aff. granulosa, HWJ 1041, collected from a large tree at 1900 m, has excellent bronze-red new leaves and turns good colours in autumn. Sorbus ligustrifolia (A. Chev.) J.E. Vidal is a small tree to 8 m or more, often shrubby, from exposed ridge-tops, with small, thickened leaves that are evidently well adapted to this habitat, and also flush bronze. It was first introduced by Keith Rushforth in 1994. Although distributed as S. aff. ligustrifolia, the collection HWJ984 probably fits in this species. Another in this group from northern Vietnam, and extending just into Yunnan at Pingbian, is S. brevipetiolata T.H. Nguyên & Yakovlev. This small tree to 10 m is allied to S. caloneura but differs in having very short petioles. It was first introduced from Vietnam by Keith Rushforth (under KR 2169a) in 1992, and from Yunnan in 1996 (under KR 4285) from the Da Wei Shan, and is in cultivation at Ness and Tregrehan. The parent trees of another introduction (BSWJ 11771), which had been felled by local people clearing forest for crops, had reached about 8 m and bore strongly ‘corrugated’ leaves and greenish brown lenticellate fruits. Keith Rushforth (pers. comm. 2007) reports that these fruits are very tasty when bletted. A species allied to A. megalocarpa but with even larger fruits has been collected in the Mangshan, Hunan by several travellers there (see Plate 531).

From China come A. aronioides (Rehder) H. Ohashi & H. Iketani and A. thomsonii (Hook. f.) H. Ohashi & H. Iketani, lumped together by Aldasoro et al. (2004) as Sorbus thomsonii, although they are quite different in appearance and are maintained as distinct by Gu et al. (2003). Aria aronioides seems to have first been introduced as SICH 407 in 1988 (illustrated p. xvi). Three trees derived from seed collected from a small specimen of this growing at Kew have been in cultivation at Ness for many years. Two have formed neat upright single-trunked trees of over 6 m. The flowers are greenish white and not showy, and neither are the greenish fruits. Under some conditions the foliage can develop brilliant autumn coloration like that of a Nyssa, the leaves being also very similar in proportion (H. McAllister, pers. comm. 2007). Aria thomsonii was introduced from Arunachal Pradesh in 2005 (K. Rushforth, pers. comm. 2008).

A remarkable (and baffling) diversity of cultivated material was displayed and discussed at the Sorbus: Micromeles & Aria Sections Workshop held at Wisley in September 2008, organised by the Royal Horticultural Society Woody Plant Committee. The day emphasised the fascination these trees hold for gardeners, but also highlighted how much remains to be learnt about them!

All Group A whitebeams related to the European A. nivea are much more tolerant of sunny dry sites than most rowans, and in Europe and southwest Asia they are often to be found on dry limestone slopes – an ecological distinction that draws attention to the differences between the two genera. The Asian species are also apparently more adaptable than most rowans, being more tolerant of clay soils and of more southern and low-altitude origin, although often from areas with moist climates – but doing reasonably well on thin soils too. As usual, however, they will all perform much better where conditions are more benign. In general, the remarks on cultivation for Sorbus given above apply to Aria as well.