Orobanche andryalae C. J. Thorogood, M. Hernández González, Rumsey & Reyes-Bet., 2025

Thorogood, Chris J., Hernández González, Matías, Rumsey, Fred J. & Reyes-Betancort, Jorge Alfredo, 2025, Orobanche andryalae (Orobanchaceae): a new species from the Canary Islands, PhytoKeys 252, pp. 275-285 : 275-285

Orobanche andryalae C. J. Thorogood, M. Hernández González, Rumsey & Reyes-Bet. sp. nov.

Description.

Stems 6–16 (25) cm, glandular-hairy, pale orange to light reddish-brown. Stem scarcely swollen below; subterranean bracts broadly triangular, yellow; those above (reduced leaves) rather sparse, brown, 8–15 mm. Flowers 5–15 (20), arranged on the upper quarter or third of the stem, lax. Bracts 10–12 mm, rather shorter than the corolla, broadly triangular, brown, glandular-hairy. Calyx 5–7 mm with segments fused, strongly unequal (rarely entire), not exceeding the corolla tube. Corolla 10–15 mm, pale yellow with faint reddish veins and scattered glandular hairs, strongly cernuous when mature, remaining so in fruit, sometimes abruptly geniculate; upper lip bilobed; lower lip 3 - lobed, the lateral lobes slightly exceeding the central; all lobes minutely-toothed. Filaments sparsely hairy below, glabrous above; inserted conspicuously (c. 5 mm) above the corolla base; anthers ± glabrous. Stigma lobes touching, mid to dark red-orange.

Type.

Lanzarote, Canary Islands, Haría (29°08'07.4"N, 13°30'18.2"W); material grown from seed at the University of Oxford Botanic Garden, Oxford, United Kingdom; April 22, 2024. (holotype ORT 48576 !), (isotype OXF! Barcode 00227715 O) .

Distribution, ecology and IUCN Red List status.

We examined O. andryalae in two locations on the island of Lanzarote, with five specimens at a population near Yé (29°11'46.1"N, 13°29'34.1"W), and about 40 in the other near Haría (29°08'07.4"N, 13°30'18.2"W) in 2020, 2021 and 2022. The plant was also observed in the Valle de Guerra and Teno regions of north Tenerife. Because O. andryalae is an annual, and based on observations of related species, we anticipate that numbers may fluctuate markedly from year to year ( Rumsey and Thorogood 2023). Orobanche seed banks can, however, remain viable for decades ( Rumsey and Thorogood 2023). Based on our current observations the species is likely to qualify for a threat status because of its restricted distribution, few locations and very low observed numbers and given that it co-occurs with its narrowly endemic host, Andryala perezii on Lanzarote, and on A. pinnatifida on Tenerife (records on other endemic Asteraceae including Asteriscus intermedius (DC.) Pit. & Proust , A. sericeus , and Crepis canariensis (Sch. Bip.) Babc. ex Jenkins require further investigation). Andryala perezii – the predominant host, is locally common within its restricted range ( Ferreira et al. 2014) and further parasite populations are to be expected on Lanzarote (the apparent stronghold for the parasite). Andryala perezii also occurs on Fuerteventura, where we have observed O. andryalae (but here it was recorded on Asteriscus sericeus ; this too, requires further examination). Andryala pinnatifida , a recorded host in Tenerife, occurs across the western Canary Islands, again indicating O. andryalae may be under-recorded. We suggest that in the absence of long-term surveys, O. andryalae should, for now, be treated as DD (Data Deficient) ( IUCN 2001).

Etymology.

Orobanche andryalae is named in accordance with its main host species, Andryala perezii .

Taxonomic remarks.

Orobanche andryalae can readily distinguished by its strongly cernuous corolla (Figs 1 C , 2 C , 3 A , 4 A ), coloration, filament insertion (Figs 1 K , 2 D ), and distinct host and ecology. Importantly, these characteristics remain stable under cultivation (Fig. 3 A ). The plant is distinct from O. amethystea subsp. castellana , which it is now clear does not occur in the Canary Islands, and we do not discuss further here. Rather, the closest taxon to O. andryalae appears to be O. calendulae , which is a somewhat confused taxon, originally documented from Algeria ( Pomel 1874), and since recorded from Madeira, Morocco, Portugal and Spain ( Chater and Webb 1972), and Tunisia ( El Mokni et al. 2023). Orobanche calendulae has marked host specificity and ecology, growing on relatives of the Calendula suffruticosa aggregate on sea cliffs. Beck-Mannagetta (1930) also recognised a similar entity, Orobanche mauretanica Beck , mainly on the basis of distinct calyx characteristics: connate in O. mauretanica and entire, or bifid and free in O. calendulae . He described a variety of O. mauretanica that he named var. calendulae, from the Algarve region of Portugal, which is parasitic on Calendula suffruticosa . Greuter et al. (1989), however, synonymised O. mauretanica under Orobanche calendulae , a decision that has been followed rather inconsistently; since that time, floras have also differed somewhat in their descriptions of Orobanche calendulae . For example, both Flora Europaea ( Chater and Webb 1972) and Flora Iberica ( Foley 2001), describe equal calyx segments that are fused at the base; we note that the calyx segments in the holotype from Algeria are bifid but somewhat unequal; in the type description, they are reported to be entire or bifid ( Pomel 1874).

The first author ( CJT) has observed Orobanche populations on sea cliffs in the Algarve growing on C. suffruticosa (Fig. 3 C , 4 C ) that may pertain to Beck’s O. mauretanica var. calendulae (= O. calendulae ). These plants possess pale orange stems, whitish corolla with violet pigmentation, reddish stigmas, apically filiform calyx lobes, and flowers arranged over most of the stem. These features appear to be consistent with the holotype of O. calendulae ( Pomel 1874) . But to complicate matters further, this plant ( O. calendulae s. l.) co-occurs in the Algarve with O. minor (on various hosts), O. litorea Guss. (on Plantago ), O. balsensis (J. A. Guim.) Carlón, M. Laínz, Moreno Mor. & Ó. Sánchez (on Carlina ), and interestingly, O. amethystea (on Eryngium campestre ) (Fig. 3 B , 4 B ); Beck-Mannagetta also recognised a form of O. mauritanica on Eryngium campestre in the Algarve ( ‘ forma dioristha ’). Presumably this form in fact pertains to O. amethystea , but in the absence of material to examine, this remains a mystery. Again, this demonstrates the nontrivial historical confusion surrounding the taxa in this area, and the importance of examining a range of characters, as well as ecology and host identity when describing Orobanche Subsection Minores.

Furthermore, we should note that populations of an unexamined entity in the Subsection Minores in Madeira also grow on Andryala , in this case A. glandulosa Lam. which is endemic to Madeira, Porto Santo and the Desertas. These plants have puberulent, yellowish stems, rather flat-backed corollas, long, apically filiform calyx lobes, and pinkish stigma lobes. Despite their Macaronesian distribution and host, they are clearly distinct from the Canary Island plant under consideration. They do, however, show a superficial similarity to O. litorea , and deserve further attention.

Notwithstanding the confusion surrounding the Subsection Minores , none of the taxonomic entities considered hitherto in this complex possess the stable combination of features we observe in Orobanche andryalae in the Canary Islands: a markedly cernuous corolla, high filament insertion and yellowish-orange colouration with a reddish stigma, and specificity for Andryala spp. on thermophilous volcanic substrates (Fig. 3 D ). In advance of a well-resolved phylogeny, it is important that taxonomic entities in the Subsection Minores are characterised morphologically and ecologically, to enable robust sampling and nomenclature.

Figures

Figure 1. A Orobanche minor habit B O. andryalae habit C O. andryalae corolla in profile D O. minor corolla in profile E O. andryalae whole plant F, G calyx and bract of O. andryalae H, I calyx and bract of O. minor J, K O. andryalae corolla in profile with cross section; arrow indicates filament insertion point L, M O. minor corolla in profile with cross section; arrow indicates filament insertion point N carpel and stigma of O. minor O carpel and stigma of O. andryalae P stamen of O. andryalae Q Stamen of O. minor. Scale bars: 15 mm (A, B); 7.5 mm (C, D); 15 mm (E); 15 mm (F, M); 5 mm (N – O); 4 mm (P, Q).

Figure 2. Orobanche andryalae A habit B lower stem C corolla D corolla cross section E carpel F stamen G calyx lobe H bract.

Figure 3. A Orobanche andryalae in cultivation at the University of Oxford Botanic Garden B O. amethystea in the Algarve, Portugal C O. calendulae in the Algarve, Portugal D the habitat of O. andryalae in Lanzarote: volcanic cliffs (Famara area, northern Lanzarote) E O. andryalae in Lanzarote, with its host plant Andryala perezii F O. andryalae collected from Fuerteventura (putatively parasitising Asteriscus sericeus) G, H O. andryalae in northwest Tenerife growing on Andryala pinnatifida (the corolla dorsal line of the specimen in H is atypical in being less conspicuously cernuous).

Figure 4. Corollas in profile A O. andryalae B O. amethystea C O. calendulae. Note the scale is indicative only, as corolla length is variable.