Baum, 1995, A Systematic Revision of Adansonia (Bombacaceae), Ann Miss Bot Gard, 82:440-470
Adansonia za Baill., Bull. Mens.Soc.Linn.Paris, 2: 844. 1890. TYPE: Madagascar. Mahafaly plateau, Grevi 37 (lectotype, designated by Perrier de la Bāthie (1952a), P).
Adansonia za var. boinensis H. Perrier, Notul. Syst-, 14: 304. 1952. TYPE: Madagascar. Ankara, Kamakama, Oct. 1902, Perrier de la Bdthie 960.9 (lectotype, here designated, P).
Adansonia bozy Jum. & H. Perrier, Ann. Mus. Colon. Marseille, 18: 447-451. 1910. Adansonia za Baill. var. bozy (Jum. & H. Perrier) H. Perrier, Notul. Syst., 14: 304. 1952 (= Adansonia za var. bozo). TYPE: Madagascar. Andranomandevo, Sambirano valley,' Jan. 1909, Perrier de la Bāthie 8159 (holotype, P).
Adansonia alba Jum. & H. Perrier, Matieres Grasses, 1909: 1511. 1909. TYPE: Madagascar. Upper basin of the Andranomalaza to the N of Bezofo (Bezofo), in dense forest by a waterfall close to 500 m, on rocks (syenite), Oct. 1908, Perrier de la Bāthie 8166 (holotype, P).
Usually tall (up to 30 m), rarely stunted (less than 5 m when mature), deciduous trees with usually single, cylindrical or slightly tapering trunks, often with irregular swellings; crowns rounded; primary branches usually ascending and tapering. Bark gray and ± smooth. Leaves 5-8-foliolate; petiole 5-15 cm x 1-4 mm; stipules caducous; leaflets sessile to long -petiolulate (up to 3 cm), broadly elliptic to lanceolate, apex usually apiculate, rarely mucronate; medial leaflets up to 20 x 8 cm in the north, usually less than 10 x 4 cm
in the south, with 10-20 pairs of regularly spaced secondary veins (more in north), usually alternating with prominent intersecondaries, glabrous or somewhat scabrous; margins entire. Flowers emerging simultaneously or soon after the leaves; buds erect to horizontal, elongated-cylindrical, 15-24 cm x 1.5-2.5 cm; flower stalk 2-3 cm, green. Calyx lobes (3-)5, reflexed and twisted at the base of the flower, 15-22 cm x 10-12 mm, green and scabrous outside, dark red villous within. Calyx tube fitting tightly around the petal bases, with a marked annular swelling (ca. 2 mm wide). Petals yellow, sometimes with a diffuse reddish medial streak on the adaxial surface, linear, at least 10 times as long as wide, 14-24 cm x 10-16 mm, exceeding the style and androecium. Androecium pale yellow, comprising a long cylindrical or tapering staminal tube, 4-6.5 cm, surmounted by 100-120 free filaments, up to 12 cm long. Ovary conical to ovoid, covered in dense, upward-pointed hairs. Style dark red, straight, 16-22 cm glabrous above, dense, upward-pointing hairs below, usually fitting loosely in the staminal tube and persisting after floral abscission (rarely to fruit). Stigma red, 3-5 mm diam. with irregular lobes. Fruit subglobose, oblong or ovoid, 10-30 cm long, 6-15 cm wide, rarely curved; usually with longitudinal ridges and, except in the north, a distinctly swollen peduncle. Pericarp thick and tough with many longitudinal fibers, usually blackish with a sparse indumentum. Seeds reniform, laterally flattened, variable in size, up to 12 x 11 x 8 mm. Germination phanerocotylar.
Representative specimens examined. MADAGASCAR. Antsiranana Province: Ambanja District: Ambodibozo near Ambanja, alt. ca. 5 m, Mar. 1951 (1, fr), Humbert & Capuron 25500 (K, P). Mahajanga Province: Maromandia District: Antanandava I village, ca. 11 km E of Maromandia, isolated tree ca. 1.5 km W of village, 14°16'5, 48°08'E, 1 Nov. 1991 (1, fl), Baum 332 (MO, P, TAN); Ambondrona village (ca. 2 km E of Ambaibo), 500 m E of village near the Andranomaiaza river, 14°13'30"S, 48°22'30"E, alt. ca. 100 m, 4 Nov. 1991 (l, fl), Baum 336 (MO, P, TAN). Marovoay District: Ankaboka near Marovoay, Dec. 1909 (1, fl), Perrier de la Bdthie 8189 (BR, G, P). Ambato-Boeny District: Ankara, Kamakama, Oct. 1902 (1, fl, fr), Perrier de la Bāthie 960A & 960C (P). Kandreho District: Kariza forest on Kelifely tablelands, Nov. 1904 (fl), Perrier de la Bāthie 960E (P). Maevatanana District: Andriba, isolated trees at foot of Mount Andriba on Gneiss, 17 Jan. 1942 (1, fl), Decary 17088 (P). Toliara Province: Mcrondava District: Kirindy forest, 4.5 km E of Morondava to Belo-sur-Tsiribihina road 42 km from Morondava, 50 m from the Kirindy river bed, 20°02'S, 44°39'E, alt. 25 m, 24 Dec. 1991 (1, fl), Baum 344 (MO, P, TAN); near Analaiva, E of Morondava, 28 Dec. 1962 (1, fl), SF (Capuron) 22141 (A, BR, G, K, bf0, P, TEF). Mahabo District: 21 km E of Mahabo, 4 Dec. 1920 (fl), Keraudren-Aymonin & Aymonin 25917 (P). Miandrivazo Dis trict: Sakeny River, Aug. 1910 (1, fl), Perrier de la Bathie 8180 (P). Morombe District: Morombe, Oct. 1940 (fr), Decary 16426 (P). Sakaraha District: Analamarina, SE of Sakaraha, alt. 500 m, 28 Dec. 1961 (l, fl), SF (Capuron) 20595 (BR, G, MO, P, TEF). Toliara District: limestone plane of the basin of the Fiherana, Jan. 1910 (1), Perrier de la Bāthie 8182 (P). Betioky District: near Beza Mahafaly reserve, near Betioky, valley of the Ehazoara river, E of Sakamena, 23°40'S, 44°39'E, alt. 170 m, 29 Nov. 1987 (1, fl), P. B. Philippson 2638 (MO, P). Ampanihy District: between Beloha and Ampanihy, 19 Feb. 1949 (1, fl), SF 455 (P, TEF). Bekily District: Ambatofotsy, 15 Jan. 1957 (1, fl), SF 3376 (P, TEF). Ambovombe District: Ambovombe, 24 Jan 1925 (fl, 1), Decary 3554 (BR, G, P,
TAN). Amboasary District: Fangidraty NE Androy, 21 Nov. 1931 (I, fl), Decary 9319 (K, P); Andohahela parcelle 2, NE of Amboasary near Hazafotsy, 24°50'5, 46°32'E, 100 m, 12 Dec. 1991 (l, fl), S. T. Malcomber 1141 (MO, P).
s.n. (P). MADAGASCAR. Jardin Tsimbazaza, 7 May 1941 (I), Jardin Tsimbazaza 4896 (TAN).
Immature or incomplete material with uncertain determinations. Near Radama [= Sahamalaza], Maromandia, 11 Oct. 1922 (imm. 1), Decary 1146 (P); N of Fiherana, dry forest of Fandrare, 40 km N of Tulear, Mar. 1934 (juv. shoot), Humbert 14348bis (P); Besevo, Oct. 1899 (1), Perrier de la Bāthie 960 (P); Ambolibozo, Sambirano May 1924 (seedlings), Perrier de la Bāthie 16340a (P).
 |
Figure 6. The distribution of Adansonia za in
Madagascar. Collection sites are marked with a circle.
When more than one collection is made from the same
locality the number of separate collections is indicated.
|
Distribution (Fig. 6). Adansonia za occurs in dry deciduous forest, spiny forest, savannas,
and scrubland from Andohahela and the Mandrare River in the southwest through southern and western Madagascar to the Boina region and the Sambirano River basin. In some deciduous forests in the south it constitutes the dominant tree species (e.g., Analava west of Betioky and to a lesser extent Sakaraha abd the forests east of Ankilizato). In the northwest it is less abundant, concentrated near rivers.
Phenology. In leaf throughout the wet season. Flowering November to February, earlier in the north, later in the south. Ripe fruit from end of dry season.
Cytology.Miege (1974) reported 2n = 48 for specimens collected in the southwest of Madagascar. Baum & Oginuma (1994) found 2n = 88 from both south and north Madagascar.
Taxonomy and nomenclature. As discussed above, there are few fixed differences between Adansonia za and A. madagascariensis. The case for exclusivity is weaker in A. za than it was in A. madagascariensis in view of its wide geographical range and the fact that some individuals in the north share features with A. madagascariensis that are absent from southern individuals assigned to A. za. One possibility is that A. za is a "metaspecies" (Donoghue, 1985), i.e., that the organisms constituting A. za are not members of any exclusive species. This situation could have arisen, for example, under the following scenario, which assumes for the sake of illustration that A. madagascariensis and A. za are sister groups. (1) A small subset of the common ancestral taxon became genetically isolated. (2) This subset continued to have a small population size and/or went through a severe bottleneck, becoming the exclusive species A. madagascariensis. (3) The remainder of the common ancestor, constituting the putative taxon "A. za," retained a large population size and did not become exclusive. If this scenario were correct, the organisms in "A. za" would not be in any basal, exclusive group, because the smallest exclusive group that they are members of is +A. rnadagascariensis +A. za, which contains the exclusive group A. madagascariensis. We do not know if such a series of events occurred, and thus we cannot determine if A. za is a species or a metaspecies. Given this uncertainty, what is the prudent course? In the absence of guidelines for dealing with metaspecies in a taxonomy, and in order to avoid unecessarily destabilizing the taxonomy of Adansonia, I here advocate recognition of A. za as a species, while flagging it as a taxon worthy of more detailed analysis.
Perrier de la Bathie (1955) proposed an infraspecific classification of Adansonia za, recognizing three varieties. Although there is significant variation from the north to the south of the range of A. za, this variation is gradual and I see no value in recognizing subspecific taxa. The southern populations of A. za are the most distinctive because of their long- petiolulate leaflets and ovoid-cylindrical fruit with swollen peduncles. These characters are only gradually replaced by sessile leaflets and normal peduncles as one goes north. Some other characters also vary clinally, such as leaflet size (larger in north) and flowering season (earlier in north), but again I see no use in forcing a continuum of variation into discrete taxa.
There is some nomenclatural confusion as to the taxon to which the name Adansonia za should apply. Hochreutiner (1908) pointed out that the description given by Baillon (1890b) contradicts the illustration labeled "A. za" (79C) in Histoire Naturelle des Plantes (Baillon, 1893) and the description given by Drake del Castillo (1902). Hochreutiner concluded that the plate predated Baillon's description and hence had priority. However, Jumelle & Perrier de la Bāthie (1910) pointed out that although plates 79A and 79B predate Baillon (18906), the others, 79C to 791, were not issued until 1893, and hence the written description has priority. Of the three plates (79C, D, and I) labeled A. za, two (79C and I) appear to be A. suarezensis, whereas plate 79D is an accurate representation of the fruit of A. za Baill. Thus, the original description of A. za is Baillon (1890b). The lectotype (Greve 37) was selected by Perrier de la Bāthie (1952a), but he incorrectly gave the collection locality as Morondava.
The first mention of Adansonia bozy is in Jumelle & Perrier de la Bāthie (19096), but the taxon was only recognized provisionally and hence this name was not validly published until 1910 (Jumelle & Perrier de la Bāthie, 1910). Perrier de la Bāthie (1952a) subsumed this taxon as a variety of A. za. However, in this publication Perrier de la Bāthie used the spelling bozo, an error which was corrected by Perrier de la Bāthie (1955). Only one specimen of A. za var. bozy was collected early enough to have been consulted by Jumelle & Perrier de la Bāthie (1910), and this is, therefore, assumed to be the holotype.
In selecting a lectotype for Adansonia za var. boinensis I chose Perrier de la Bāthie 960A, the most complete specimen from one of the localities mentioned by Jumelle & Perrier de la Bāthie (1909a, 1910) that predates these articles.
The putative species Adansonia alba was named by Jumelle & Perrier de la Bāthie (1909a) to accommodate a specimen Perrier de la Bāthie collected along the Andranomalaza River. This species was said to be diagnosed by its white flowers and short staminal tube. In the course of my observations and collections along the Andranomalaza River, all the baobabs I encountered could be assigned to A. za. Furthermore, the type specimen, though of poor quality, shows no features that distinguish it from A. za. Thus, I believe that A. alba is synonymous with A. za. The characters that supposedly separate the species, white petals and short stamina] tube, were probably incorrectly scored by Perrier de la Bāthie, being based upon fallen flowers.
Ethnobotany.
 |
Adansonia za fruits.Image courtesy:
Dr.Pavel Hoek, CZ,
who headed a
scientific expedition to Madagascar.
Go to Lemuria for more images.
|
Adansonia za has several common names, for example, "za" or "zabe" in the south and "bozy" or "bozybe" in the north, and "ringy" or "boringy" in the Ambongo region (Perrier de la Bāthie, 1955).
Little is known about the human exploitation of Adansonia za, but Jumelle & Perrier de la Bāthie (1912) reported that the seeds are eaten and the trunk is sometimes hollowed out as a cistern for storing water. Perrier de la Bāthie (19526) mentioned that the seedling roots are an edible vegetable, and Miege (1974) reported that A. za is destroyed by ranchers so that their cattle can feed on the moist wood.
Conservation. In view of the extensive geographical range, Adansonia za is conservationally secure, despite the fact that several local populations are endangered by forest clearance. Adansonia za has diverse interactions with animals: it provides nesting sites for birds, holes for carnivores and lemurs (M. Nicoll, pers. comm.); perches for territorial surveillance and display by birds and the sportive lemur (Lepilemur sp.); many insects feed on the leaves, sap, nectar, and pollen; sunbirds (Nectarinia souimanga) feed on nectar; sifaka (Propithecus verrauxi verrauxi) feed on flower buds; and fork-marked lemurs (Phaner furcifer) feed on exuded gum (Petter et al., 1975).
