An update of the Genus Gagea Salisb. (Liliaceae) in the Iberian Peninsula

7 Lagascalia 29, 2009 AN UPDATE OF THE GENUS GAGEA SALISB. (LILIACEAE) IN THE IBERIAN PENINSULA J. M. Tison 14 promenade des Baldaquins, F-38080 L’Isle d’Abeau (Recibido el 12 de Marzo de 2009) Summary. A new Gagea species from Spain and Morocco, Gagea subtrigona J.-M. Tison, is described; an overlooked species from the southern Iberian Peninsula, Gagea lusitanica A. Terrac., is cleared; an updated key to the Iberian species of the genus Gagea is provided. Resumen. Se describe Gagea subtrigona J.-M. Tison, nueva especie de España y Marruecos, y se reconoce Gagea lusitanica A. Terrac., especie propia del Sur de la Península Ibérica. También se aporta una clave actualizada de las especies ibéricas del género Gagea. INTRODUCTION The taxonomy of the genus Gagea has been studied in the Iberian Peninsula by Terracciano (1904a, 1904b, 1905a, 1905b) mainly on dried adult plants, then by Bayer & López (1988a, 1988b, 1989, 1991) including living plants with highly suitable results; during studies on the Moroccan species I published a small complement on the section Didymobolbos in southern Spain (Tison, 2004a). After ten years of investigations on Iberian Gagea including cultivation of many clones, two species previously quoted as exceptionally polymorphic, G. algeriensis Chabert and G. elliptica (A. Terrac.) Prain (Tison, 2004a, 2004b) proved to hide each one misappreciated taxon. The two cryptic species are described here, one of them as new. To summarize the current state of knowledge I propose an updated key of the Gagea species of this region. RESULTS AND DISCUSSION Gagea subtrigona J.-M. Tison, sp. nova A Gagea algeriensis Chabert (= G. wilczekii Br.-Bl. & Maire) differt colore semper viridescente, juvenilibus et immaturis plantis bulbillorum gregem ferentibus, immaturis plantis plurifoliatis, fl orescentibus plantis rarissimis, folio basali primo plus aut minus canaliculato, infl orescentia saepe bulbillifera, pedicellis semper pilosis. Lagascalia 29: 7-22 (2009) 8An update of the genus Gagea Salisb. (Liliaceae) in the Iberian Peninsula Similar to Gagea algeriensis Chabert (= G. wilczekii Br.-Bl. & Maire), but always greenish; juvenile and immature plants with a cluster of sister bulbils; immature plants with several leaves; fl owering plants very rare; fi rst basal leaf more or less canaliculate; infl orescence often bulbilliferous; pedicels always hairy. Holotypus.SPAIN: Gagea wilczekii Br.-Bl. & Maire, Hs, Alacant: (Alcoi), Serra de Mariola, Bc. de Bocairent, 30SHY1388, 880 m, 11.3.2001, L. Serra 5702, A. Bort & L. Serra Cremades (VAL 193871), inferior plant (bulbilliferous) (Fig. 1, arrow). Fig. 1. Gagea subtrigona, holotype (arrow). 9 Lagascalia 29, 2009 Epitype: Fig. 2. Other specimens examined. Because its sporadic fl owering and diffi cult recognition when dried, the species is very rare and inconspicuous in herbaria and we take into account only recent collections with anatomic study of leaves. SPAIN. Alacant: (Alcoi), Serra de Mariola, Bc. de Bocairent, 30SHY1388, 880 m, 11.3.2001, Serra 5702, Bort & Serra Cremades (Herb. Serra n° 6231, transferred to VAL), superior plant (isotypus); (Alcoi), El Puig, 30SYH2182, 850 m, 17.3.1992, Serra 2171 (Herb. Serra n° 2413); (Xixona), La Carresqueta, 30SHY1977, 1000 m, 17.3.1992, Serra 2172 (Herb. Serra n° 2414); (Agres), Serra de Mariola, Molí Mató, 30SYH1494, 850 m, 10.4.2000, Serra 5584, Laguna & Raeymaekers (Herb. Serra n° 6109); (Confrides), Serra d’Aitana, pr. Cases d’Aitana, 30SYH3683, 1190 m, 20.3.2004, Serra 6626 & al. (Herb. Serra n° 7289); (Vall de Gallinera), pr. Tossal del Molló, 30SHY4198, 615 m, 20.3.2003, Peris (Herb. Serra n° 7516). Albacete: Robledo: col de la N322 à la sortie W de la ville, c. 1000 m, pelouses sèches sur calcaire, 18.3.2005 & ex cultis 29.3.2007, Tison (Hb. privé). Córdoba: Priego-de-Córdoba: plateau de Navazuelo, c. 800 m, pelouses rocailleuses très ouvertes sur gypse, 20.3.2005, Tison (Hb. privé). Guadalajara: Alcolea: voie express NII à 1,5 km à l’est du Puerto de Alcolea, c. 1100 m, biotopes secondaires très ouverts sur marnes, 10.4.2001, Tison (Hb. privé). Jaén: Cazorla: Puerto de las Palomas, c. 1300 m, pelouses rocailleuses sur calcaire, 18.3.2005, Tison (Hb. privé). Málaga: Ronda: entre le Puerto del Viento et la maison des gardes au N du C344, c. 1300 m, pelouse rocailleuse exp. N sur arête calcaire, 21.3.2001 & ex cultis 28.3.2005, Tison (Hb. privé); El Burgo: Puerto de las Abejas, c. 800 m, pelouses ouvertes et + secondaires sur pentes calcaires, 22.3.2005, Tison (Hb. privé). Valencia: (Fontanars), Serra de la Solana, La Replana, 30SXH99, 940 m, 14.3.1992, Serra Cremades (Herb. Serra n° 2415); (Bocairent), Naixement del Vinalopó, 30SHY0688, 780 m, 20.3.2004, Peris (Herb. Serra n° 7517). MOROCCO. Al- Hoceima. Ajdir: col du Nador 1500 m, pente marneuse exp. S à végétation ouverte, ab. 2, 25.3.2001, Tison (Hb. privé). Khenifra. Aguelmouss: près du Tizi-n-Tichka 2300 m, arête d’éboulis schisteux fi ns, 02.4.2001 & ex cultis 28.3.2005, Tison (Hb. privé). Marrakech. Mouldikht: Tizi-n-Test, c. 2300 m, pentes d’éboulis schisteux fi ns, 01.4.2001, Tison (Hb. privé). Meknes. Timhadite: Tizi-n-Tretten, c. 1900 m - doline non pâturée, 29.3.2001 & ex cultis 28.3.2005, Tison (Hb. privé). Ecology. Open grasslands, rocky, gravelly or sandy places; more tolerant to mechanical instability than the neighbouring species; c. 500-1500 m in Spain, at least to 2300 m in Morocco. On disturbed ground the clusters of non-fl owering plants with canaliculate, arched, dirty green leaves are easily mistaken for Muscari neglectum Ten. 10 An update of the genus Gagea Salisb. (Liliaceae) in the Iberian Peninsula Distribution. Spain: S.-E. quadrant, at least to Guadalajara; Morocco: at least Rif, Medium Atlas and Great Atlas. Taxonomic remarks. G. subtrigona is an apomictic taxon: immature stage very long and bulbilliferous, adult stage and seeds virtually absent, bulbils often sprouting unusually easily. Some bulbs produce an infl orescence once before death, but this infl orescence has an evident character of immaturity: it is sterile, usually few-fl owered and often bulbilliferous; hence, we can emphasize that both fl owering and non-fl owering old plants represent a similar immature stage and that this taxon actually never reaches the adult stage. Our cultivated lineages from Alcolea (Spain) and Ajdir (Morocco) regularly execute their complete ontogenic cycle to death without any fl owering. The clones having a very thick fi rst leaf look like bulbilliferous G. algeriensis Chabert (= G. wilczekii Br.-Bl. and Maire: Tison, 2004b) and have been interpreted in this way (Bayer & López, 1988a; Tison, 2004b). However, although solitary bulbilliferous plants may very exceptionally appear in natural populations of G. algeriensis, all our cultivated lineages of this species have a short immature stage (c. 2 years), a long fl owering stage (3-5 years) and never produced sister bulbils for 10 years. In brief, G. subtrigona has normally no real adult stage and is regularly bulbilliferous, while G. algeriensis has a long adult stage and its bulbilliferous form can be considered as abnormal. Since every Gagea species has a strict ontogenetic sequence (Levichev, 1999), the co-existence of such variations inside a single species is unlikely and allows to conclude to the existence of a cryptic species. G. subtrigona indeed is in many respects intermediate between G. algeriensis and G. lacaitae A. Terrac. and likely results hybridogenous. The basal leaves are strongly variable depending on the clone, sometimes almost identical to those of G. algeriensis, sometimes rather similar to those of G. dubia A. Terrac. except the thickened keel and the two layers of vascular bundles, with all the intermediate possibilities (Fig. 3, 1c and 2c). To contrast with the polymorphism of the leaves, the ontogenetic sequence and the morphology of the subterranean organs are roughly constant. Many investigated Spanish populations take in G. subtrigona and its two putative parental species growing together. G. subtrigona, still less-fl owering than G. lacaitae, usually is completely unnoticed; however, it is often the most abundant among the three taxa, especially in very dry regions with poor soil (Albacete: Bayer & López, 1988a, sub G. wilczekii; from personal observations in Albacete, Jaén, Cuenca, Ciudad Real, Córdoba, Málaga). It apparently extends further to the north than G. algeriensis, to the province of Guadalajara (pers. Obs.), and is locally abundant in the province of Valencia where G. algeriensis is to be confi rmed (P. Pablo Ferrer Gallego, pers. com.). On the contrary it seems less common in Morocco, probably reaching southernmost only the 11 Lagascalia 29, 2009 Fig. 2. Gagea subtrigona, Ronda (Málaga). A, B, C: three different clones in the same population.1,adult stage; 2, immature stage. a, habitus; b, bulb and basis of the aerial parts; c, cross section of the fi rst leaf; d, cross section of the second leaf; e, cross section of the peduncle; f, cross section of the fi rst cauline leaf; g, cross section of a pedicel; h, basal bulbil in summer; i, fl ower; j, tepals (left: external; right: internal); k, androceum and gyneceum; l, cross section of the ovary. Scale bar: a, 2 cm; b, i-l, 0,5 cm; c-h, 1 mm. 12 An update of the genus Gagea Salisb. (Liliaceae) in the Iberian Peninsula Great Atlas, while G. algeriensis is present to Anti-Atlas (sub G. cossoniana Pascher: Tison, 2004b). Whether G. subtrigona is better considered as an hybridogenous species or a simple hybrid is debatable. In my mind the second option should be unpleasant at least in the present state of knowledge. The taxon is very widespread and abundant and has been found several times without its putative parents. If we admit that both G. algeriensis and G. lacaitae and only them took place in its origin, which is not absolutely certain, it must be taken in mind that the taxa of the G. granatellii complex are often triploid (Peruzzi, 2003; Peruzzi & Aquaro, 2005; further data submitted, L. Peruzzi, pers. comm.) and that the Spanish populations of G. lacaitae always show a high percentage of pollinic abortion and almost never give viable seeds, which makes the hypothesis of numerous recent hybridizations very unlikely. As a matter of fact, the only good argument in consideration of an hybrid status is the high polymorphism of G. subtrigona, especially regarding the leaf anatomy. From my point of view this polymorphism does not allow to reject the specifi c status: I interpret this situation as an ancient introgressive phenomenon, as demonstrated for another hybridogenous taxon, G. pomeranica (Rüthe) Rüthe (Peterson et al., 2004), usually treated as a good species. On some living plants the carinate basal leaves may lead to confusions with G. dubia, present in southern Spain and Morocco (Tison, 2002 [sub G. maroccana (A. Terrac.) Sennen & Mauricio], 2004a, 2004b) and northernmost at least to Sierra de Moncayo (J.-M. Tison, personal herbarium). G. dubia, however, is devoid of infl uence of G. algeriensis: it has thin, solid basal leaves with one layer of vascular bundles; its bulb is broadly ovate to subglobose while the bulb of G. algeriensis and G. subtrigona is more obliquely elongated. In addition G. dubia regularly produces complete infl orescences, has comparatively short and erect basal leaves, and usually grows at higher altitudes (1000-2300 m in Spain, at least to 2700 m in Morocco), commonly associated with G. nevadensis Boiss., but rarely with G. algeriensis, G. lacaitae or G. subtrigona. According to its somewhat dubia-like habitus, to its reduced and irregular infl orescence, to its ecology (plant of foothills), G. subtrigona may fi t the lectotype of G. granatellii subsp. maroccana A. Terrac. (Tanger, Schousboeh 1869: Tison, 2001). Following as close as possible Terracciano’s descriptions and specimens I quoted numerous Moroccan populations of G. dubia, including G. granatellii subsp. maroccana in synonymy (Tison, 2004b); all undoubtedly belong to G. dubia with the possible exception of the lectotypical collection of G. granatellii subsp. maroccana. According to this, G. maroccana (A. Terrac.) Sennen & Mauricio might be a prioritary name for G. subtrigona. This synonymy, however, remains doubtful on account of the bad condition of Schousboeh’s collection. 13 Lagascalia 29, 2009 Gagea lusitanica A. Terrac., Boll. Soc. Ortic. Mutuo Soc. Palermo, 2 (3): 36 (1904) Taxon intermediate between Gagea elliptica (A. Terrac.) Prain and G. lacaitae A. Terrac.; differs from the former in its bulbilliferous juvenile and immature stages, its somewhat wider and fl atter basal leaves having to 7 vascular bundles (vs. 3-5), its sometimes hairy pedicels and its rare and usually empty capsules; differs from the latter in its more elongated bulb, its usually more erect, much narrower basal leaves (1,5-3 mm at adult stage, vs. 3-5 mm in G. lacaitae) without adaxial additional vascular bundles, its more regular and plentiful fl owering and its sometimes glabrous pedicels. Lectotype (here designated).PORTUGAL. Beira Baixa. Gagea polymorpha, Monte Brito, Castello Branco, Junho de 1881, A.R. Da Cunha (LISU n° 8504!), left specimen (Fig. 3, arrow). Four collections, all from the Castelo Branco region, are quoted by Terracciano (1904a) in the protologue of G. lusitanica: one by Da Cunha (1881, LISU) and three by Daveau (1885, COI, G, MPU). All have been traced and can be considered as syntypes. The fi rst collection (Da Cunha), which is in very good condition, is designated here as lectotypical; among the three plants of this collection, the most complete and well-presented one is chosen as lectotype. Epitype; Fig. 4. Other specimens examined. PORTUGAL. Beira Baixa. Castello Branco, juin 1885, Daveau (COI, G, MPU) [syntypi]. SPAIN. Cádiz. Grazalema: à l’est du Puerto de las Palomas, c. 1250 m, pelouse embroussaillée, 21.3.2005, Tison (Hb. privé). Granada. Loja: N321 au pied de la montagne de Los Infi ernos, c. 900 m, vires herbeuses sur escarpements calcaires, 19.3.2005, Tison (Hb. privé). Málaga. Monda (Málaga), UF35, 22.3.79, Diez &. Zarazaga (MGC 5510); Laguna de Fuente Piedra (Málaga), UG41, 06.3.76, Asensi & Garretas (MGC 9026); Málaga, Álora, El Sabinal, La Medionda, 04.02.89, Bootillo (MGC 31611 & 42267); Málaga: Ronda, prox. Cortijo de la Planilla (entre las piedras de una era), erizos, calizas, 730 m s.m., UF0867, 05.3.89, Montilla (MGC 40552); Málaga: Sierra de las Nieves, Tolox, Ampigal de Tolox, 03.4.2003, Cabezudo (MGC 55158); Atajate: route de Benadalid v. 700 m, pelouses embroussaillées sur éminences, 21.3.2005, Tison (Hb. privé); Ronda: versant ouest du Puerto del Viento v. 1100 m, broussailles sur pente calcaire exp. W, 21.3.2001, Tison (Hb. privé). Ecology. Known only on rocky meadows over limestone, at 600-1200 m s.m. Distribution. Spain: Andalucía (at least provinces of Granada, Malága and Cádiz); Portugal? 14 An update of the genus Gagea Salisb. (Liliaceae) in the Iberian Peninsula Fig. 3. Gagea lusitanica, lectotype (arrow). 15 Lagascalia 29, 2009 Taxonomic remarks. I found several Andalucian populations conspicuously agreeing with the syntypes of G. lusitanica and not with other taxa. This plant was quoted as a critical bulbilliferous form of G. cf. elliptica in previous papers (Tison, 2004a, 2004b), a rather troubling hypothesis considering the normally constant ontogenic cycle of every Gagea species (I.G. Levichev, pers. com.). The cultivation through several years confi rmed the total absence of bulbilliferous stage in G. elliptica, its regularity in G. lusitanica, and the existence of further distinctive features suggesting an infl uence of G. lacaitae in G. lusitanica. The two taxa often grow together (for example at Ronda, Atajate, Grazalema) and this situation brings the differences in light: in addition to the elements quoted above, G. lusitanica blooms c. 10-15 days before G. elliptica on the same place, is somewhat shorter and stouter than the latter, and usually does not give viable seeds. According to its probable hybrid origin, G. lusitanica is rather polymorphic: depending on the lineage, the basal leaves are more or less broad (Fig. 4, 1c), the cluster of bulbils at immature stage is sessile or pedunculate (Fig. 4, 2a), the pedicels are glabrous or hairy (Fig. 4, 1i), but no evident correlation is visible among these variations. G. lusitanica seems very close to G. polymorpha, a cryptic taxon known only around Granada. Although G. polymorpha was supposed intermediate between G. durieui and G. elliptica (Tison, 2004a), it is more probably intermediate between G. durieui and G. lusitanica according to its slightly hairy pedicels. The only constant difference between G. polymorpha and G. lusitanica is the anatomy of the adult basal leaves: semicylindrical with U- or V-arranged vascular bundles in G. polymorpha, fl at with plane-arranged vascular bundles in G. lusitanica. Some narrow-leaved specimens of the latter, especially at the Eastern part of the area, appear very close to the former. Despite its name and type locality, the presence of G. lusitanica in Portugal is to be confi rmed today. The labels of the type collections are unlikely and may be due to mixtures of herbarium parts. The region of Castelo Branco includes only plains and low siliceous mountains less than 1000 m high; “Monte Brito” quoted by Da Cunha is a small hill c. 6 km north-east of the city, where Gagea can hardly fl ower later than mid-April, while the labels of the type collections quote June. The only Gagea I found in this region was G. soleirolii F.W. Schultz, blooming in March and rather common down to the foothills. The Gagea taxa in Portugal are probably few-numerous (Terracciano, 1905b) and remain poorly known. The only unquestionable species are G. soleirolii, widespread and locally common, G. bohemica (Zauschner) Schultes and Schultes f., probably rare (Bayer & López, 1991) and G. pratensis (Pers.) Dumort., very rare and/or underestimated (absent in Terracciano, 1905b; sub G. nova Samp. ex Miranda: Miranda Lopes, 1929-1930; sub G. pratensis subsp. nova Samp.: COI!). G. elliptica and G. lacaitae, the putative parental 22 An update of the genus Gagea Salisb. (Liliaceae) in the Iberian Peninsula Peterson, A., H. John, E. Koch & J. Peterson (2004). A molecular phylogeny of the genus Gagea (Liliaceae) in Germany inferred from non-coding chloroplast and nuclear DNA sequences. Plant Syst. Evol. 245: 145-162. Terracciano, A. (1904a). Gagearum novarum diagnoses. Boll. Soc. Orticola di Palermo 2(3): 3-10. ––– (1904b). Per la priorità delle mie Gagearum novarum diagnoses. Boll. Soc. Orticola di Palermo 2(4): 1-7. ––– (1905a). Revisione monografi ca delle specie di Gagea della fl ora spagnola. Bol. Soc. Aragon. Ci. Nat. 4 (6, 7, 8): 188-253. ––– (1905b). Le Gagea della fl ora portoghese. Bol. Soc. Brot. 20: 200-206. Tison, J. M. (2001). Typifi cation de Gagea cossoniana Pascher, de Gagea fragifera (Vill.) Ehr. Bayer et G. Lopez et de Gagea maroccana (A. Terracc.) Sennen et Mauricio. Candollea 56(1): 197-202. ––– (2002). Gagea Salisb. In B. Valdes, M. Rejdali, A. Achhal El Kadmiri, S. L. Jury & J. M. Montserrat (eds.) Catalogue des plantes vasculaires du Nord du Maroc, incluant des clés d’identifi cation 2. C.S.I.C., Madrid. ––– (2004a). Identité et situation taxonomique de Gagea polymorpha Boiss. Candollea 59(1): 109-117. ––– (2004b). Contribution à la connaissance du genre Gagea Salisb. (Liliaceae) en Afrique du Nord. Lagascalia 24: 67-87.