Page 11 - Passalacqua_Peruzzi_Pellegrino2008

TAXON 57 (3) • August 2008: 893–906 Passalacqua & al. • Biosystematics of the Jacobaea maritima group

evidence, it is rather difficult to develop a well founded Jacobaea maritima (L.) Pelser & Meijden subsp. sicula,
hypothesis, but our suggestion is that genetic drift is the subsp. nov.
main evolutionary factor, even if it only started working Diagnosis. – Differt ab J. maritima typica utrinque
in recent times, thus the differentiation is very weak. minore pubescentia, foliis supra saepe non araneosis, mi-
From a morphological point of view, J. maritima seems nore involucris longitudine; differt ab J. maritima subsp.
to be well differentiated from J. gibbosa and J. bicolor (Figs. bicolor maiore involucris latitudine, minore capitulorum
7A–B, 8), but not in the case of the Levanzo population numero, foliis magis septatis.
of J. bicolor, which is effectively intermediate in linking Holotype. – SICILY: Sicilia, Isole Egadi, Levanzo, rupi
J. maritima with J. bicolor (Figs. 4–7, 9; Tables 5–6). The marittime, 17.VI.2002, Peruzzi & Passalacqua s.n. (CLU).
Levanzo population tends to be distinguishable from J. mar- Paratypes. – SICILY: Sicilia, 1831, Gussone s.n. (G-DC);
itima by the lower stem and capitula hair covering (Table Isola di Linosa (Sicilia), rupi marittime verso Cala della
6), the less segmented leaves (Fig. 5A–B), and the smaller Pozzolana, VII.1873, Aiuti s.n. (FI); Favignana, 15.IV.1973,
involucre (Table 5). It is distinguishable from J. bicolor Brullo s.n. (CAT); Isola Grande dello Stagnone, 20.VII.1974,
by the more branched inflorescence (Fig. 9) and its size Brullo s.n. (CAT); Marettimo 15.V.1982, Brullo s.n. (CAT);
(Table 5), and the shorter stem (Fig. 6). Jacobaea bicolor Madonnina Marettimo, 15.V.1982, Brullo s.n. (CAT); Capo
and J. gibbosa retain their distinctness both in their overall Libeccio Marettimo, 7.V.1982, Brullo s.n. (CAT); Linosa,
morphology (Figs. 7–8), and using specific characters (stem 28.VI.1985, Spampinato s.n. (CAT); Lèvanzo, s.d., Todaro
hairiness, capitula hairiness, Table 6). The morphological s.n. (RO sub Senecio bicolor Tod.); MALTA: Dragunara
similarity that links all investigated populations can prob- Malta, VI.1973, Brullo s.n. (CAT); Mgarr ix-xihi, VI.1973,
ably be explained by the recent genetic differentiation of Brullo s.n. (CAT); Rampla Tar-Torri Malta, 26.VI.1973,
units. The Levanzo population, showing affinity with both Brullo & Ronsisvalle s.n. (CAT); Malta, Mjiebah (presso Se-
J. bicolor and J. maritima, could be linked to genetic drift linum Palace), 25.IX.1985, Brullo s.n. (CAT); Malta, Gozo,
in the Levanzo population from J. maritima (or its ancestor) Dwejra Point, 15.XI.1986, Brullo & Pavone s.n. (CAT); Ras
overlapped by a recent genetic introgression from J. bicolor, l-injieya, VI.1973, Brullo s.n. (CAT).
but we did not find any molecular evidence in support of Distribution. – Egadi Islands and Canal of Sicily
this hypothesis and further research is needed. (Malta, Gozo, Linosa). Preliminary observations carried
out on plants from Croatia and Greece lead us to suspect
that this name may also be applicable to Eastern Medi-
terranean plants until now referred to J. maritima subsp.
CONCLUSIONS bicolor (Matthews, 1975). However, this matter requires
The combination of morphometric and molecular data further investigation.
showed little discriminating power. One of the aims of the Habitat. – Coastal cliffs and rocks.
work was to test the taxonomic setting of the group through
morphometric and molecular research at population level. Key to the Jacobaea maritima group in the Cen-
The results highlighted the fact that the populations refer- tral Mediterranean
able to J. maritima s.l. (Table 2, populations 1–7) and to (modified after Peruzzi & Passalacqua, 2003)
J. ambigua s.l. (populations 8–9) are well differentiated 1. Stems and involucres glabrous to glabrescent; involu-
at the molecular level (Figs. 2–3), but less well so at the cres 5 mm long or less . . J. maritima subsp. gibbosa
morphological (Figs. 4–6, 9, Tables 5–6). Nevertheless, 1. Stems and involucres more or less hairy; involucres
our observations on the auto-ecology and floral phenology usually more than 5 mm long . . . . . . . . . . . . . . . . . 2
of these taxa suggest that interbreeding between these two 2. Leaves septate (sometimes entire in young individu-
units is unlikely at present, so that it is preferable to retain als), with 5–12 segments; preferred habitat on coastal
two separate species: Jacobaea maritima (L.) Pelser & Mei- cliffs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
jden and Jacobaea ambigua (Biv.) Pelser & Meijden. 2. Leaves lyrate to entire, with 0–5 segments; preferred
Within the two species, the delimitation of subspecies habitat on hilly-mountainous screes (J. ambigua s.l.) 5
is justified mainly based on geographical distribution, 3. Leaves hardly septate, adaxial surface often with
which is related to a certain degree of morphometrical arachnoid hairs; involucres 5–6 mm wide; plants
differentiation. usually up to 70 cm tall, with fewer than 50 capitula
The overall taxonomic arrangement proposed by to each stem. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Peruzzi & al. (2006; see Introduction), therefore, can be 3. Leaves septate to lobate, adaxial surface always
maintained with one exception: the Levanzo population, without arachnoid hairs; involucres 4–5 mm wide;
previously referred to J. bicolor (under Senecio gibbosus plants usually more than 70 cm tall, with more than
subsp. bicolor, Peruzzi & Passalacqua, 2003), should be 50 capitula for each stem. . . . . . . . . . . . . . . . . . . . .
regarded as an independent subspecies. . . . . . . . . . . . . . . . . . . . .J. maritima subsp. bicolor

903