Julian Slade
Wed, 17 Dec 2003 00:24:25 PST
Dear all

Recent scientific studies using DNA sequencing technology have shown that
what we know as Scilla actually appear to have multiple evolutionary
origins. Therefore it seemed reasonable to recognise each coherent group as
separate genera. The following lists these new or resurrected genera that
have been proposed.

Firstly, Scilla lazulina from Zimbabwe appears to be the most primitive
'Scilla'. Was put into Merwilla (see below) before DNA sequencing showed
otherwise. No genus name has yet been proposed.

Pseudoprospero: only 1 species, P. firmifolium. Instantly recognisable in
frequently having a side branch on the flower scape. Summer-growing.
Apparently the second-most primitive of the 'Scillas'.

The remaining 'Scillas' belong to 2 geographical groups: sub-Saharan
African/Indian (which also includes Lachenalia, Ledebouria, Massonia,
Daubenya, Drimiopsis, etc.) and North African/European/Asian (also including
Hyacinthus, Bellevalia, Hyacinthella, Muscari, etc.).

The following belong to the first group:

Merwilla: S. natalensis, S. dracomontana, S. kraussii: summer-growing,
winter-dormant. The most primitive genus in the African/Indian group. One
scape per season. Seeds whitish and papery. Unfertilised flowers drop off.

Spetaea: only 1 species, S. lachenaliiflora. Until 2003 it was known as
Scilla plumbea (actually a mystery plant that may be a colour form of S.
natalensis). An evergreen, mountain-dwelling species from the Cape. Found to
be most closely related to Daubenya!

Schizocarphus: S. nervosus, which may actually represent more than one
species. Summer-growing. Multiple scapes per season. Seeds black, ovoid.
Unfertilised flowers drop off.

The Eurasian genera (which follow), fall into 4 groups:

Most primitive is Barnardia: B. numidica (N Africa) and B. japonica (east
Asia, which may or may not represent more than 1 species). Both flower in
the autumn before the leaves have developed. Bracts and bracteoles (smaller,
secondary bracts) present. Compare to Prospero.

Next is a group of genera mainly based in the western Mediterranean region.
This group also includes Brimeura and Hyacinthoides.

Oncostema: the so-called Cuban or Peruvian lilies, such as O. peruviana, O.
sicula, O. hughii, etc. (possibly as many as 10 species). Bracts large but
bracteoles tiny. Usually coarse plants.

Tractema: such as T. verna, T. monophyllos, T. liliohyacinthus. Bracts
largish, bracteoles absent, otherwise like a miniature Oncostema; may or may
not be closer to Brimeura.

Autonoe: such as A. haemorrhoidalis, A. madeirensis, etc. Large plants,
distinct in having orange to purple, fleshy fruits and unique in the
Hyacinthaceae in not producing nectar. Seeds quite large.

Following is a group mainly from the Middle East; it also includes
Hyacinthus, Hyacinthella, Puschkinia, and probably Alrawia.

Prospero: such as P. autumnale, P. obtusifolium, P. hanburii.
Autumn-flowering before the leaves have fully developed. Bracts and
bracteoles absent. Many new species have been described that are difficult
to tell apart.

Othocallis: such as O. siberica, O. miczenkoana, O. rosenii. Tepals fall off
when spent. Multiple scapes per season, semiterete (semicircular in
cross-section), floppy in fruit. Capsule succulent; seeds black to
yellowish, warty, sometimes with an appendage.

Fessia: such as F. greilhuberi, F. hohenackeri. Similar to Othocallis but
with very different seeds: always glossy black, lacking any appendages.
Anthers rather large. Tepals persistent.

Pfosseria; only 1 species, P. bithynica. Similar to Othocallis, but seeds
without an appendage; flowers dense and starry. Tepals persistent.

Zagrosia: only 1 species, Z. persica. Related to the previous 3 (as well as
to Hyacinthus), but scapes remaining rigid, capsules dry and papery.

(Hyacinthus is most similar to Othocallis but has thick scapes and anthers
with pointed extremities; the flowers have persistent tepals and their
distinctive hyacinth shape.)

The final group has a similar geographical distribution to the previous
group. It also includes Bellevalia and Muscari.

Nectaroscilla: only 1 species, N. hyacinthoides. Scape 1 per season. Bracts
and bracteoles small. Plants large, bulbs with woolly extensible threads.
Seeds medium-large. Scape smooth, terete (circular in cross-section).

Chouardia: C. litardierei and C. lakusicii. Like a small Nectaroscilla but
with membranous bulb tunics and much smaller seeds. Scape ribbed.

Schnarfia: S. messeniaca and S. albanica. Capsules somewhat succulent.
Scapes semiterete and floppy in fruit. Bracts and bracteoles small. Very
distinctive seeds, glossy brown-yellow with a solid appendage. Several
scapes per season.

Finally, what is left of Scilla includes species such as S. bifolia as well
as the now defunct genus Chionodoxa (which itself forms 2 unrelated groups).
One scape per season, terete but flopping over in fruit. Bracts small to
absent, bracteoles absent. Capsules slightly succulent. Seeds glossy,
yellowish to black, with a soft appendage.

Furthermore, no evidence exists to support the separation of Muscari into
the genera Muscarimia, Leopoldia, or Pseudomuscari.

The differences between these genera may seem minor, but the only thing they
all have in common is the bluish, star-shaped flowers (which appears to be a
primitive characteristic). Any attempt to reduce the number of genera (by
coalescing related ones) would cause the disappearance of many
well-established ones, as well as making these new super-genera difficult if
not impossible to define.

Definitely controversial!


Julian Slade

More information about the pbs mailing list