Craniolaria integrifolia von Chamisso (1832: 725–726)

Hassemer, Gustavo, 2017, Typification, revision of the synonymy and range extension of Craniolaria integrifolia (Martyniaceae), Phytotaxa 291 (2), pp. 101-115 : 102-108

Craniolaria integrifolia von Chamisso (1832: 725–726)

≡ Martynia integrifolia ( von Chamisso 1832: 725–726) von Steudel (1841: 104)

Lectotype (designated here): — BRAZIL. “Brasilia meridionali”, 1814–1831, F. Sello s.n. (HAL-0098136! [ Fig. 1 ]).

= C. argentina Spegazzini (1896: 382–384)

Lectotype (designated here):— ARGENTINA. SALTA: Cafayate , December 1896, C. L. Spegazzini s.n. (LP-12556! [ Fig. 2 ]).

Epitype (designated here):— ARGENTINA. SALTA: La Viña, January 1897, C. L. Spegazzini s.n. (LP-10449! [one specimen prepared in three sheets; Figs. 3–5 ]) .

= C. integrifolia f. longiflora Chodat & Hassler (1903: 547) , syn. nov.

Lectotype (designated here): — PARAGUAY. In regione vicine Igatimí, in campo, 1898–1899, E. Hassler 5527 (G-00076895! [one specimen prepared in three sheets; Figs. 6–8 ]).

Illustrations: — Fig. 2 in van Eseltine (1929); Fig. 268 in Sorarú (1979); Fig. 110 in Kiesling (1993); illustration 2 in Zapater (1995).

Photograph: —Photograph 2 in Zapater (1995).

Distribution: —This species occurs in Argentina, Bolivia, Brazil, Paraguay and Uruguay. In Brazil this species is recorded for the following states: Mato Grosso, Mato Grosso do Sul, Minas Gerais, Piauí, Rio Grande do Sul and São Paulo ( Fig. 9 ).

Conservation Status: —Least concern (LC). This species has a wide distribution and, despite being somewhat rare, there is no evidence that it may be currently threatened with extinction.

Notes: —Thorough revision of the literature has revealed that a holotype was not designated for C. integrifolia in its protologue ( von Chamisso 1832: 725–726) and, to the best of my knowledge, no lectotype has been designated for this name in any subsequent published work. The only specimen that could be located and that can be considered original material for this name is kept at HAL (HAL-0098136 [ Fig. 1 ]); therefore, I designate this sheet as the lectotype of C. integrifolia . There is no original material for this name kept at B (Robert Vogt, pers. comm.), despite most plants collected by Friedrich Sello (1789–1831) in southern Brazil were sent to Berlin and deposited in this herbarium. It is possible that original material of C. integrifolia was housed at B before this herbarium was bombed in 1943, but there is no available evidence to confirm this.

Some authors recognise C. argentina as a synonym of C. integrifolia , either explicitly ( Kiesling 1993) or implicitly, by not counting it among the accepted species of Craniolaria ( Bretting & Nilsson 1988, Zapater 1995, Gormley et al. 2015). This is not the case in works such as van Eseltine (1929), Harley et al. (2003) and Ihlenfeldt (2004), in addition to some Argentine regional taxonomic treatments ( Abbiatti 1939, Sorarú 1979) and the country’s most important botanical catalogue ( Zuloaga et al. 2008; also available online at http://www2.darwin.edu.ar/proyectos/floraargentina/ fa.htm). However, there is uncertainty surrounding this name in all works consulted, and with the exception of Abbiatti (1939), Kiesling (1993) and Zapater (1995) there is no evidence that any of these authors had the opportunity to examine original material for this name.

Spegazzini (1896) mentioned three localities when describing C. argentina : La Viña, Guachipas and Cafayate, all in Salta province, north-western Argentina. There are four collections at LP (LP-10449 [one specimen prepared in three sheets; see Art. 8.3, and Gautier et al. 2016], LP-10450, LP-12555 and LP-12556, all collected by Spegazzini and evidently belonging to the same species) that were until now considered syntypes of C. argentina (Laura Iharlegui, pers. comm.). I have studied high-resolution images of all these specimens and it became clear that: 1. C. argentina is clearly a synonym of C. integrifolia , and thus I completely agree with the synonymisation proposed by Kiesling (1993); and 2. only one of these collections (LP-12556) can be considered original material for C. argentina . The collection date of two of these four collections (i.e. LP-10449 and LP-10450) is 1897, therefore posterior to the publication of the name C. argentina ; furthermore, two of these collections (LP-10450 and LP-12555) were collected in localities distinct from those listed by Spegazzini (1896). Therefore, I designate here the collection LP-12556 ( Fig. 2 ) as the lectotype of C. argentina . Nevertheless, because the lectotype lacks flowers and fruits, I considered appropriate to also designate the collection LP-10449 (one specimen prepared in three sheets; Figs. 3–5 ) as the epitype of C. argentina , to facilitate the application of this name (Art. 9.8). This specimen was collected by Spegazzini in La Viña, a locality mentioned in the protologue.

In the protologue of C. integrifolia f. longiflora ( Chodat & Hassler 1903: 547) two gatherings were mentioned as original material (therefore, syntypes): Hassler 5527 and Hassler 3305. One collection of E. Hassler 5527 in the personal herbarium of Emil Hassler (1864–1937), housed at G, consisting of a specimen prepared in three sheets (see Art. 8.3, and also Gautier et al. 2016) ( Figs. 6–8 ), all with the same barcode (G-00076895), is here designated as the lectotype of C. integrifolia f. longiflora .

Chodat & Hassler (1903) described C. integrifolia f. longiflora based mainly on the longer corolla tubes (to 200 mm) and the wider corolla lobes (40–42 mm). Nevertheless, leaf and flower size is highly variable in C. integrifolia , even within populations; furthermore, these characters do not follow a coherent distributional pattern, and do not appear to be taxonomically informative. Therefore, I propose here the synonymisation of C. integrifolia f. longiflora under C. integrifolia .

Although the protologue of C. integrifolia ( von Chamisso 1832: 725–726) clearly indicates southern Brazil as the origin of the type specimens, the Flora of Brazil ( Gutierrez 2015) does not report the presence of this species for any state of the South region of Brazil. This is rectified here by the recognition of four overlooked gatherings, all from Rio Grande do Sul: C. Gaudichaud-Beaupré s.n. (P-02966951), B. Rambo s.n. (PACA-44942), B. Rambo s.n. (PACA-45154) and F. Gonzatti 1392a (HUCS-42559). These specimens are morphologically very similar to the type of C. integrifolia and a better match than any other specimens from any other known population of this species. The Gaudichaud-Beaupré gathering indicates only “Rio-Grande” as collection locality; however, both Rambo gatherings and the Gonzatti gathering originated from low-elevation sandy areas in eastern Rio Grande do Sul (Osório and Viamão municipalities). Therefore, I consider it very likely that Sello collected the type of C. integrifolia in this area. Also, this locality is close to the state capital, Porto Alegre, where Sello is known to have spent some time.

During the revision of specimens from herbaria in the Northeast region of Brazil I discovered a gathering of C. integrifolia (Del’Arco et al. s.n., EAC-10634 [ Fig. 10 ]) collected in December 1980 in Ribeiro Gonçalves municipality, Piauí state. This gathering represents the first record of the genus Craniolaria in the Northeast region of Brazil, and also first record of the family Martyniaceae in Piauí ( Fig. 9 ), extending the distribution of C. integrifolia by ca. 1,400 km from the nearest recorded occurrence, in Campina Verde municipality, Minas Gerais. This specimen is lacking fruits and seeds, and is much less pilose than those collected from other populations, but overall it matches very well with C. integrifolia . Rediscovery of this outlier population and collection of fruiting specimens would be of great value to improve the taxonomic understanding of Craniolaria .

I must remark that a gathering from Bahia state, north-eastern Brazil, L.A.M. Silva et al. 2297 (HUEFS-187463 and MBML-43263) was, prior to my revisions, identified as Craniolaria sp. Closer study of these specimens revealed that they belong to Holoregmia viscida von Esenbeck (1821: 300) , a species endemic to Bahia, and until now the only member of the Martyniaceae recorded for this Brazilian state ( Harley et al. 2003, Giulietti & Harley 2013). Furthermore, a gathering from Brasília (Distrito Federal territory) (P.C.M. Ramos 239, UB-3896-2), was also erroneously identified as Craniolaria sp. ; however, this material does not belong to the Martyniaceae . Identification of these specimens as members of other genera means that Craniolaria does not occur in Bahia and Distrito Federal.

Figures

FIGURE 1. Lectotype of Craniolaria integrifolia (F. Sello s.n., HAL-0098136).

FIGURE 2. Lectotype of Craniolaria argentina (C.L. Spegazzini s.n., LP-12556).

FIGURE 3. Epitype of Craniolaria argentina (C.L. Spegazzini s.n., LP-10449), sheet 1/3.

FIGURE 4. Epitype of Craniolaria argentina (C.L. Spegazzini s.n., LP-10449), sheet 2/3.

FIGURE 5. Epitype of Craniolaria argentina (C.L. Spegazzini s.n., LP-10449), sheet 3/3.

FIGURE 6. Lectotype of Craniolaria integrifolia f. longiflora (E. Hassler 5527, G-00076895), sheet 1/3.

FIGURE 7. Lectotype of Craniolaria integrifolia f. longiflora (E. Hassler 5527, G-00076895), sheet 2/3.

FIGURE 8. Lectotype of Craniolaria integrifolia f. longiflora (E. Hassler 5527, G-00076895), sheet 3/3.

FIGURE 9. Distribution of Craniolaria integrifolia in Brazil, including the new record for Piauí (green star) and the overlooked gatherings from Rio Grande do Sul (blue stars).

FIGURE 10. The first record of Craniolaria integrifolia in the Northeast region of Brazil (Del’Arco et al. s.n., EAC-10634).