Malvaceae Juss. is a family of about 243 genera and over 4225 species, mainly distributed in tropical areas, but also occurring in temperate regions (Stevens, 2001–).

Many Malvaceae species are used for various purposes, such as natural fibers (e.g. cotton, genus Gossypium L.), ornamentals (e.g. genera Alcea L., Hibiscus L., Lavatera L.), foods (e.g. cacao, genus Theobroma L.), vegetable oils [e.g. from baobab (genus Adansonia L.)], beverages (e.g. colas, genus Cola Schott. & Endl.), timber (e.g. genus Tilia L.), medicinals [e.g. for the demulcent nature, genus Althaea L., or astringent (genus Hibiscus), or diuretic (genus Abutilon Mill.)] (see e.g. Mitchell, 1982; Forrest, 2006; Wakelyn et al., 2007; Benchasri, 2012; Ogbu, 2014; Ţîţei & Teleuță, 2018; Tumpa & Islam, 2019). Some species are able to escape from cultivation, impacting both natural (loss of biodiversity) and agricultural (economically with reduction in productivity and crop quality) systems (Spencer, 1984; Pasiecznik, 2016).

As part of an ongoing study on the family Malvaceae (e.g. Iamonico, 2010, 2014a , b , 2016, 2018; Iamonico & Peruzzi, 2014; Iamonico & Del Guacchio, 2017; Iamonico & Valdés, 2017) and the alien Flora of Tunisia (El Mokni & Iamonico 2017, 2018a , b , 2019; Iamonico & El Mokni, 2017, 2019a , b ), field surveys allowed to find some populations of three Malvaceae species which were not previously reported from both the country and the whole of northern Africa [Hibiscus rosa-sinensis L., H. syriacus L., and Lagunaria patersonii (Andr.) G. Don]. Distribution of the African populations in Tunisia, and notes on their habitats are provided. The nomenclature of the untypified names Hibiscus acerifolius Salisb., H. chinensis DC., H. patersonius Andr., and H. rhombifolius Cav. is also discussed.

The research is based on field surveys, analysis of relevant literature, and searching/examination of specimens preserved at BM, G, HFLA, MA, LINN, and P [herbarium codes following Thiers, 2020 (continuously updated)].

The Articles cited through the text follow the International Code of Nomenclature for algae, fungi, and plants (Turland et al., 2018, hereafter ICN).

Hibiscus rosa-sinensis

The genus Hibiscus is represented in the flora of Tunisia by just one species, i.e. H. trionum L. (Le Floc’h et al., 2010: 268; Valdés, 2011a ). Five Hibiscus species (H. micranthus L. f., H. palustris L., H. sabdariffa L., H. trionum L., and H. vitifolius L.) are currently recorded in North Africa (Valdés, 2011b ; SANBI, 2012 and literature therein). Our finding of H. rosa-sinensis L. represents the first record for both the Tunisian and the North African floras.

Hibiscus rosa-sinensis L., Sp. Pl. 2: 694 (1753) (Fig. 1A).

Lectotype (designated by Borssum Waalkes, 1966: 72): Herb. Hermann 3: 4, No. 260 (BM, barcode BM-000621802 [photo!], image available at https://data.nhm.ac.uk/object/a481219f-0364-42d4-8b9c-ddb192d0ef11/1587427200000).

Habitat in Tunisia: sidewalks in urban areas.

Elevation: 5 m a.s.l.

Distribution in Tunisia: we found one population at Bizerta-City (NE-Tunisia) comprising five individuals which occupy an area of about 21 m². At the current state of knowledge, we tentatively consider Hibiscus rosa-sinensis as casual alien species for Tunisia and North Africa. No seedlings were observed.

Chorology: the origin of Hibiscus rosa-sinensis is unknown, but it could have originated in China (Tang et al., 2007). It has also been thought to have originated in Africa (Borssum Waalkes, 1966) or the New World Tropics (Prendergast, 1982). However, since H. rosa-sinensis includes a wealth of hybrids and other selections and has a complicated history in cultivation, its origin remains uncertain at present (Tang et al., 2007; Braglia et al., 2010). This species is considered as alien in Asia (Tang et al., 2007; Valdés, 2011b ), Europe (only in Italy, in Sardinia region; Galasso et al., 2018), Central and South America (Blanchard, 2015), and eastern Australia (Atlas of Living Australia, 2020a ).

Notes on cultivar name: Hibiscus rosa-sinensis is widely cultivated as ornamental plant and many cultivars were created (The American Hibiscus Society, 2016). A recent molecular study on 47 cultivars (Braglia et al., 2010) showed that there are few differences among them using DNA sequences, although some clusters were moderately to highly supported (bootstrap values > 80). It is difficult to identify the variety/cultivar. On the basis of the size and color of corollas, Tunisian plant can be referred to the cultivar ‘Leuchtfeuer’ (Braglia et al., 2010).

Specimina visa from Tunisia: Tunisia, Bizerta, Bizerta-South, sidewalks near public gardens, 5 m a.s.l., 6.III.2020, El Mokni & Iamonico s.n. (Herb. Univ. Monastir!); ibidem (HFLA!).

Hibiscus syriacus

As already stated above (see under the paragraph “ Hibiscus rosa-sinensis ”) the occurrence of Hibiscus syriacus is here recorded for the first time in both Tunisia and the whole of northern Africa (Le Floc’h et al., 2010: 268; Valdés, 2011a ; SANBI, 2012 and literature therein).

Hibiscus syriacus L., Sp. Pl. 2: 695 (1753) ≡ Ketmia syriaca (L.) Scop., Fl. Carniol. (ed. 2) 2: 45 (1772) ≡ K. syrorum Medik., Malvenfam.: 45 (1787), nom. superfl. et illeg. [1] (Art. 52.2 of ICN) ≡ H. arborescens Gatereau, Descr. Pl. Montauban 122 (1789), nom. superfl. et illeg. [2] (Art. 52.2 of ICN) ≡ K. arborea Moench, Methodus: 617 (1794), nom. superfl. et illeg. [3] (Art. 52.2 of ICN) ≡ H. floridus Salisb., Prodr. Stirp. Chap. Allerton: 383 (1796), nom. superfl. et illeg. [4] (Art. 52.2 of ICN) (Fig. 1B).

Lectotype (designated by Abedin, 1979: 13): Herb. Linn. 875.24 (LINN [photo!], image available at http://linnean-online.org/6872/).

= Hibiscus rhombifolius Cav., Diss. 3, Tertia Diss. Bot.: 156 (1787)

Lectotype (designated here): [Icon] Tab. LXIX f. 3 Hibiscus rhombifolius in Cavanilles (1787), image of lectotype available at https://archive.org/details/monadelphiaeclas14cava/page/n363/mode/2up/search/LXIX).

= Hibiscus acerifolius Salisb., Parad. Lond. 1: Tab. 33 (1805).

Lectotype (designated here): [Icon] Tab. 33 Hibiscus acerifolius in Salisbury (1805), image of lectotype available at https://www.biodiversitylibrary.org/page/36898142#page/70/mode/1up).

= Hibiscus chinensis DC., Prodr. 1: 455 (1824), nom. illeg. (Art. 53.1 of ICN) non H. chinensis Roxb.

Holotype: [Icon] Tab. 24 in van Braam Houckgeest (1821), image of lectotype available at https://gallica.bnf.fr/ark:/12148/btv1b7300008x/f53.item).

– Althaea frutex Mill., Gard. Dict., ed. 8. (1768), nom. inval. pro syn. of Hibiscus syriacus (Art. 36.1b of ICN).

Habitat in Tunisia: foots of walls of an ancient building.

Elevation: 35 m a.s.l.

Distribution in Tunisia: one population was found at Lahouemdeya (Tabarka, Jendouba, NW-Tunisia) which comprises six individuals occupying an area of about 3 m². At the current state of knowledge, we tentatively consider Hibiscus syriacus as casual alien species for Tunisia and North Africa. No seedlings were observed.

Chorology: Hibiscus syriacus is native to China (Provinces of Anhui, Guangdong, Guangxi, Jiangsu, Sichuan, Taiwan, Yunnan, and Zhejiang) but it was taken into cultivation very early and was distributed to the Middle East where it is considered as alien (Tang et al., 2007). This species is also naturalised in Europe (France, Italy, Hungary, Serbia, and Greece; Valdés, 2011b ), North America (eastern USA.; Swearingen & Bargeron, 2016), and southeastern Australia (Atlas of Living Australia, 2020b ).

Typification of the name Hibiscus rhombifolius: Cavanilles (1787: 156) validly published H. rhombifolius through a detailed description, the provenance (“Habitat in India orientali”) and the following statement: “V. S. [Vidi Siccam] unicum exemplar apud D. de Lamarck ex collectis sonneratianis”. The latter sentence indicates that J. A. Cavanilles has seen a single specimen collected by P. Sonnerat (1748–1814), a French naturalist and explorer (Stafleu & Cowan, 1985: 744). However, we cannot consider it as the holotype of the name Hibiscus rhombifolius since Cavanilles (loc. cit.) also published an illustration (“Tab. LXIX, fig. 3”) which is part of the original material for the name Hibiscus rhombifolius (see Arts. 9.1 and 9.4 of ICN). As a consequence, a lectotypification is necessary (Art. 9.3 of ICN). Sonnerat’s collection is preserved at P (Stafleu & Cowan, 1985: 744). In the online database of the P herbarium (see https://science.mnhn.fr/institution/mnhn/collection/p/item/p00287601) appears a record form referring to a specimen of H. rhombifolius (P00287601) which would be included in the Historical Collection of Paris Herbarium. This specimen was identified by J. van Borssum Waalkes (April 1959) as H. syriacus and marked by him as “ISOTYPE” according to the record form. There also is a later annotation (“Type”) by S. Doorenbos (in 2013). Since no image of P00287601 is available on line, we requested it but we have not received a reply. Fortunately, the Cavanilles’ illustration, which morphologically matches the original description of H. rhombifolius, can be used to fix the name and it is here designated as the lectotype. According to the current concept in Hibiscus (see e.g. Tang et al., 2007; Blanchard, 2015) the lectotype of H. rhombifolius can be identified with H. syriacus and the two names are considered as heterotypic synonyms.

Typification of the name Hibiscus acerifolius: Salisbury (1805: without pagination) validly described H. acerifolius by giving a short diagnosis and a detailed description highlighting the species is “very distinct ... from the Syriacus of Linneé, in which the leaves are wedge-shaped at the base with much shorter petioles”. Salisbury (loc. cit.) also provided a colored illustration which is original material. We have not been able to trace specimens of orginal material and, as a consequence, we are forced to designate the Salisbury’s illustration no. 33 as the lectotype of the name Hibiscus acerifolius. Note that this image matches the Salisbury’s diagnosis and description. According to the current concept in Hibiscus (see e.g. Tang et al., 2007; Blanchard, 2015) the lectotype of H. acerifolius can be identified with H. syriacus and the two names are here synonymzed.

Type of the name Hibiscus chinensis: Candolle (1824: 455) validly described H. chinensis providing a short diagnosis and citing as synonym “Braam. ic. chin. 1821. t. 24” which refers to Icones plantarum sponte China nascentium, e bibliotheca braamiana excerptae by Braam Houckgeest (1821). Braam’s illustration (t. 24) is original material for the name. Note that Candolle (loc. cit.) did not reported “v. s.” (vidi siccam) under H. chinensis, which would have indicated that he examined herbarium specimens (see the Prefatio by Candolle 1824: vi). Moreover, no specimen was found by us. According to the Art. 9.1 of ICN and the considerations by McNeill (2014), the image no. 24 of Braam Houckgeest (1821) can be considered as the holotype of the name H. chinensis. According to the current concept in Hibiscus (see e.g. Tang et al., 2007; Blanchard, 2015) this holotype can be identified with H. syriacus and the two names are here synonymzed.

Taxonomic notes on the infraspecific variability: Hibiscus syriacus is a very popular frost-hardy flowering shrub commonly used as ornamental plant in gardens and public green (e.g. as street trees) and many cultivars have been developed, some of which have been given formal scientific names. According to Tang et al. (2007) these varieties can be distinguished on the basis of flowers arrangement (solitary or double), the diameter of the corollas, and the color of the petals. The Tunisian plants have solitary flowers, 5–6 cm in diameter, lavender or violet with a darkred center which would correspond to the var. grandiflorus Rehder.

Specimina visa from Tunisia: Tunisia, Jendouba, Lahouemdeya, Tabarka, 35 m a.s.l., 30.IX.2015, El Mokni & Iamonico s.n. (Herb. Univ. Monastir!); ibidem (HFLA!).

Lagunaria patersonia

Lagunaria patersonia is currently recorded in Africa only in Kenya (Lebrun & Stork, 1991), and South Africa (Coyne, 2011). Our finding represents the first record for both the Tunisian and the North African floras.

Lagunaria patersonia (Andr.) G. Don, Gen. Hist. 1: 485 (1831, as “patersonii”) ≡ Hibiscus patersonius Andr., Bot. Repos. 4: t. 286 (1803) [basionym] (Fig. 2).

Lectotype (designated here): [Icon] Plate CCLXXXVI Hibiscus patersonius in Andrews (1803), image of lectotype available at https://www.biodiversitylibrary.org/item/109451#page/144/mode/1up).

Habitat in Tunisia: sidewalks in urban area.

Elevation: 5–15 m a.s.l.

Distribution in Tunisia: two populations were found at Ariana-City and Bizerta-North (NE-Tunisia). The first population comprises five individuals which occupy an area of about 24 m², the second population includes three individuals occupying an area about 16 m². At the current state of knowledge, we tentatively consider Lagunaria patersonia as casual alien species for Tunisia and North Africa. No seedlings were observed.

Chorology: Lagunaria patersonia is native to the Pacific Islands (Lord Howe Island, Norfolk Island) and Australia (Queensland and New South Wales) and it is alien in America [San Diego County (Hanes, 2015), California (Calflora, 2020), and Hawaii islands (Coyne, 2011)], Australia [south-east (Jessop & Toelken, 1986; Atlas of Living Australia, 2020c ) and New Zealand (Coyne, 2011; New Zealand Plant Conservation Network, 2020)], Europe [e.g. in Portugal (Almeida & Freitas, 2006) and Sicily (La Mantia et al., 2015)], Africa [Kenya (Lebrun & Stork, 1991), and South Africa (Coyne, 2011)].

Typification of Hibiscus patersonius: The name Hibiscus patersonius was validly published by Andrews (1803) who provided a diagnosis and a detailed and colored illustration (Plate CCLXXXVI) that is part of the original material used by Andrews (loc. cit.) to describe the species. Stafleu & Cowan (1976: 49) stated that H. C. Andrews (1799–1830), a botanical painter and engraver, prepared his illustrations based on living plants and no herbarium specimens are known to exist. However, Andrews (loc. cit.) stated “Our figure is from a specimen received from the Right Hon. Lord Viscount Courtenay, in whose collection at Powderham, near Exeter...”. As a consequence, the Andrews’ t. 286 can not be considered as the holotype (see Arts. 9.1 and 8.4 and the consideration by McNeill, 2014), but it is available as a possible lectotype. The Courtenay’s specimen cited by Andrews (loc. cit.) in the protologue was not found. As a consequence, the Plate CCLXXXVI, which matches the diagnosis and corresponds to the current concept of the species (see e.g. Hanes, 2015), is here designated as the lectotype of the name H. patersonius.

Taxonomic notes: Craven et al. (2006) treated the populations occurring in Queensland at species rank, naming them as Lagunaria queenslandica Craven (in Craven et al., 2006: 352). This species was said to differ from L. patersonia by the lenght of the petals (47–58 mm vs. 40–46 mm), lenght of the staminal column (40–62 mm vs. 30–36 mm), and lenght of the antheriferous part (8–13 mm vs. 13–20 mm). The populations discovered in Tunisia display morphological characteristics that match those of L. patersonia.

Specimina visa from Tunisia: Tunisia, Ariana, Ariana-City, near public gardens, 6 m a.s.l., 5.XII.2018, El Mokni & Iamonico s.n. (Herb. Univ. Monastir!); ibidem (HFLA!); ibidem, 8.III.2020 (Herb. Univ. Monastir!); Bizerta, Bizerta-North, sidewalks near public gardens, 15 m a.s.l., 5.III.2020 (Herb. Univ. Monastir!).