Morphological and allozyme variation of
|
|  Yüklə 0.91 Mb.
|
|
Sexes showed slight differences in their patterns of differentiation. Females are more distinct among populations in wing, skull and dental measurements, as is evident from their larger accumulative values of the Lawley–Hotelling trace. In the PCAs, the separation of Annobo´ n along the PC1 for females points to size as the main source of variation. Instead, the large percentage of explanation accounted for by PC2 (25.1%) for males, indicates relatively important differences among populations not only in size but also in shape (Fig. 2).
1993) and for Rousettus egyptiacus (P=0.177) reported from the Guinea group of islands ( Juste, Machordom & Iba´ n˜ ez, 1996). Average heterozygosity (H=0.055) for the populations studied was identical to the values found for Aethalops alecto (Kitchener et al., 1993) and Cynopterus titthaecheilus (Schmitt, Kitchener & How, 1995). Eidolon’s heterozygosity is higher than the values reported for other fruit bats, e.g. C. sphinx (H: 0.028; Schmitt et al., 1995); H. fischeri (H=0.034; Peterson & Heaney, 1993), or R. egyptiacus (H=0.038; Juste et al., 1996). Although the island populations of E. helvum show slightly lower values of polymorphism, the average heterozygosity was higher in these populations. It seems, therefore, that there is no reduction of genetic variability in the islands with respect to the mainland. The expected pattern of reduction of genetic variability in islands does not hold for other fruit bats studied in the islands ( Juste et al., 1996, 1997) nor for other fruit bats from Indonesia (Kitchener et al., 1993). However, in a study involving a larger number of islands, Peterson & Heaney (1993) found a trend (albeit not significant) towards a reduction of variability in smaller islands, and Schmitt et al. (1995) found a significant, negative relationship between heterozygosity and ‘level of isolation’. In populations with limited gene flow, reduction of variability would result mainly from genetic drift on low effective population sizes (Schmitt et al., 1995). Effective population sizes of E. helvum are quite high in the Gulf islands ( Juste & Iba´ n˜ ez, 1994). This fact would have helped maintain levels of genetic variability relatively high, even in Annobo´ n.
1995). Pairwise comparisons show that this value is due mainly to the differences between Annobo´ n and all the other populations. The difference between Annobo´ n and the mainland is close to 20% (FST=0.187), and over 10% with respect to the closest population, that of Sa˜ o Tome´. This pattern also appears in the genetic distances, which are over 2% (DN) between Annobo´ n and all the other populations except Sa˜ o Tome´. The low values of genetic distance and FST between the oceanic islands of Sa˜ o Tome´ and Pr´ıncipe suggest little restriction to gene flow between them. The rate of theoretical migrants per generation varies considerably among islands. Gene flow is highest between the ‘land-bridge’ island of Bioko and the mainland, with a value (Nm=7.51) similar to the average (Nm=7.53) reported for non-isolated- by-distance populations of Cynopterus (Peterson & Heaney, 1993). Nm values between the mainland and the oceanic islands decrease with distance, which suggests that this factor is acting as a barrier to gene flow from the mainland. Particularly for Annobo´ n, the value of Nm approaches the minimum required (Nm=1) for in- dependent divergence by random drift under the neutral island model (Wright, 1965). Gene flow is high (Nm=7.33) between Sa˜ o Tome´ and Pr´ıncipe. Nonetheless, the value of Nm obtained by the private alleles method is considerably smaller (Nm= 2.05). Estimates of Nm obtained from F-statistics and private allele methods also show some differences for other pairwise comparisons (Table 6). These differences could indicate the role of selection in determining diversity among islands (Schmitt et al., 1995), and suggest some caution in the interpretation of the values. Interestingly, the most important differences for the two estimates of Nm values occur in between- island comparisons. Variation patterns The highly significant correlations between morphological and geographic dis- tances for both sexes, as well as the phenograms, indicate that morphological variation of E. helvum populations in the Gulf islands clearly follows a geographic pattern. Mantel’s test shows concordance of morphological and genetic patterns. The fact that some alleles for Pgm-1 and Pep-A are shared only among island populations, and the increasing frequency of the slowest allele of Sod-1 along the oceanic islands, give support to a geographic pattern of differentiation. Genetic and geographic matrices show a weak correlation, which probably reflects the effect of the short distance between Sa˜ o Tome´ and Pr´ıncipe on the genetic matrix. The estimate of gene flow and the presence of a diagnostic allele at Est-1 in these two populations sustain the inter-island genetic resemblance. Colonization of the Gulf islands, according to a stepping-stone model, is the most parsimonious explanation for this geographic pattern. Independent colonization events from the mainland are expected to have occurred due to the relatively long distances between islands and their small size. However, numerous examples, including plants (Figueiredo, 1994), mollusks (Gascoigne, 1994), birds (Peet & Atkinson, 1994) and shrews (Heim de Balzac & Hutterer, 1982), indicate close evolutionary relationships between the Gulf islands. The association is particularly tight between Sa˜ o Tome´ and Pr´ıncipe, which share diverse endemisms even of land snails or amphibians. Like E. helvum, the endemic populations of the fruit bat R. egyptiacus from Sa˜ o Tome´ and Pr´ıncipe show close affinities ( Juste et al., 1996). However, the two populations of R. egyptiacus—a less vagile species—are more differentiated morphologically than those of E. helvum. Effects of migration and dispersal E. helvum is commonly acknowledged as a migratory species. A seasonal movement takes place at the beginning of the rainy season following births which is accomplished by large groups of males and lactating females (Thomas, 1983). We found adult individuals of E. helvum in both wet and dry seasons in all our visits to the islands. In addition, lactating females were also found in all the islands. These observations, together with information gathered from the locals, strongly suggest that none of the insular populations of E. helvum undertakes seasonal migrations to the mainland, as Eisentraut (1964) and Bergmans (1990) had suggested for the population of Bioko. Nevertheless, we have detected important differences in colony size between dry and wet seasons in the islands (e.g. ‘Cacahual’ colony in Bioko drops from tens of thousands of individuals to a few hundred during the rainy season). This indicates intra-island movements, which are probably associated with fruit availability, the trigger factor suggested for the migration in the mainland (Thomas, 1983). In the absence of migration, the quite high levels of gene flow found among the islands would result from dispersing individuals. E. helvum’s wing morphology shows the largest aspect ratio among African fruit bats ( Juste & Iba´ n˜ ez, 1994), which allows for its high dispersal capacity. Vagrant specimens of E. helvum have been reported as far as 400 km off Africa (Varona, 1975). In the Gulf of Guinea, the dominant southwesterly movement of wind (Wauthy, 1983) would favour inter-island contact. Nevertheless, the role of dispersal in structuring the populations (Thomas, 1983), or in determining the distribution patterns of this species (Bergmans, 1990), is still obscure and remains to be investigated. Evolutionary and taxonomic inferences The variation found in the populations of Bioko, Pr´ıncipe and Sa˜ o Tome´ can be considered within the range of the nominate subspecies, E. helvum helvum. The levels of gene flow between the mainland and the islands—and among the latter—seem to prevent differentiation. Unlike those populations, E. helvum shows remarkable morphological and genetic differentiation in Annobo´ n. Founder effect is considered the major evolutionary force for determining gene pool differentiation in insular populations (Kilpatrick, 1981), and that may be the case for Annobo´ n. In addition, the observed morphological trend towards a reduction of size on the three oceanic islands suggests the existence of a selective pressure in this direction. The role of natural selection is also suggested by the differences in estimates of Nm by the F- statistics and private allele methods, and the gradient of frequencies for Sod. A reduction in size in island populations seems to be a general pattern for bats (Krzanowski, 1967), and it has been explained as a result of change in selective pressures on trophic conditions (Palmeirim, 1991) or flight environments (Iliopoulou- Georgudaki, 1986). In the Gulf islands, the combination of selective forces in harsher oceanic environments and restricted gene flow would have favoured morphological differentiation of E. helvum in Annobo´ n, which undergoes a pronounced long dry season (Herna´ ndez Pacheco, 1943) and seasonal fruit shortage. In these conditions, an earlier achievement of sexual maturity—and consequently smaller size—may be advantageous in the absence of migratory behaviour. The more marked differ- entiation among females also sugests that selection could be linked to the reproductive pattern. Similar environmental pressures may be acting on the small subspecies E. helvum sabaeum, which experiences comparable isolation and harsh conditions in the Arabian Peninsula. A´ lvarez (1961) noticed the particularly small size of the population of E. helvum from Annobo´ n, and suggested its recognition at the subespecific level. The extent of DN genetic distances for this population (0.016–0.024) is within the range of distances described for conspecific subspecies in other fruit bats like R. egyptiacus (DN=0.019; Juste et al., 1997), or Aethalops alecto (DN=0.011; Kitchener et al., 1993), or the vespertilionid Myotis lucifugus (DN=0.011; Herd, 1987). Therefore, the population of Annobo´ n is described hereafter as Eidolon helvum annobonensis subsp. nov. Eidolon helvum annobonensis subsp. nov. Holotype. Female adult (EBD 17603) (skin and skull), collected January 12, 1987 in San Antonio de Pale´ (1°24′S, 5°38′E), Annobo´ n island, by Javier Juste. Geographical distribution. The subspecies is endemic to the island of Annobo´ n (West Central Africa). Diagnosis. Small size combined with weak dentition. Description. Eidolon helvum annobonensis is a large fruit bat with a large fox-like head. Externally, it resembles the nominate subspecies, and shows the typical colour pattern of E. helvum in both sexes: light brown on the back turning to greenish grey on the rump. The flanks are yellowish. This colour extends along the dorsal side of arms and forearms, contrasting against the darkness of the wing membrane. Ventrally, the fur is yellowish. The subspecies shows marked individual variability and some specimens have a darker appearance. The holotype presents 4+3+3 palatal ridges. The skull shows the ‘typical’ Eidolon morphology, although it is smaller than in the nominate species. It shows a more delicate general appearance with narrower postorbital processes, slender rostrum, and a more marked interorbital constriction than E. h. helvum. The jaw exhibits quite individual variation in the shape of the coronoid and angular processes, but it is always less massive than in the nominate subspecies. The teeth are similar in shape than those of E. h. helvum but notably smaller, particularly the premolars and molars. Measurements. Average values for the subspecies are given in Appendix 2 by sex. Measurements (mm) of the holotype (EBD 17603) are: Body measurements: FA: 109.6; Ear: 27.5; Total length: 172.0; Tail length:, 19.2; IIMC: 53.7; IIF1: 15.3; IIF2: 7.9; IIIMC: 75.6; IIIF1: 48.6; IIIF2: 70.0; IVMC: 73.5; IVF1: 40.2; IVF2: 43.9; VMC: 67.5; VF1: 31.3; VF2: 28.8. Cranial measurements: GSL: 50.2; CBL: 48.2; RL: 19.9; PL: 24.7; ZB: 27.1; IOB: 8.1; POB: 11.5; BCB: 19.3; C1-M1: 17.3; C1-M2: 19.4; C1-C1: 9.2; M2-M2: 14.6; ML: 39.5; MH: 15.1; C1-M2: 20.4; C1-M3: 22.1; MA: 134°. Dental measurements (length and breadth respectively): P3: 3.1–2.3; P4: 3.8–2.4; M1: 3.8–2.4; M2: 1.7–1.6; P3: 2.9–2.0; P4: 3.6–2.1; M1: 4.3–1.8; M2: 2.6–1.7; M3: 1.3–1.0.
sabaeum by its lighter fur colour and slender dentition and from E. dupreanum by its much smaller size and less elongated muzzle. ACKNOWLEDGEMENTS We are grateful to Leandro Mbomio, former Culture Minister of the Republic of Equatorial Guinea and to Jose Luis Xavier Mendes former Minister of Agriculture of the Democratic Republic of Sa˜ o Tome´ and Pr´ıncipe. We thank the following: R. E. Strauss for statistical advice and for making available the Matlab functions; C. Lo´ pez-Gonza´ lez for her suggestions; A. Ayong Nguema and C. Ru´ız for their assistance with the field work. The work was supported by the Instituto de Coop- eracio´ n para el Desarrollo (ICD) of the Spanish Ministerio de Asuntos Exteriores, with the cooperation of the Estacio´ n Biolo´ gica de Don˜ ana (CSIC), and Asociacio´ n de Amigos de Don˜ ana. Travel expenses were partially financed by the Agencia Espan˜ ola de Cooperacio´ n Internacional, the Junta de Andaluc´ıa (RNM-158), and by DGICYT (PB90-0143). REFERENCES A´ lvarez J. 1961. Impresiones de un viaje a la isla de Annobo´ n. Archivos Instituto Estudios Africanos 57: 53–70.
Andersen K. 1912. Catalog of the chiroptera of the British Museum. I. Megachiroptera. London: British Museum (Natural History). Anthony ELP. 1988. Age determination in bats. In: Kunz TH, ed. Ecological and behavioural methods for the study of bats. Washington DC: Smithsonian Institution Press, 47–58. Barboza du Bocage JV. 1897. Subsidios para a fauna da ilha de Ferna˜ o do Po. Jornal de Sciencias Mathematicas Physicas e Naturaes 2: 1–15. Barboza du Bocage JV. 1903. Contribution a` la faune des quatre ˆıles du Golfe de Guine´e. IV. Iˆle de St. Thome´. Jornal de Sciencias Mathematicas Physicas e Naturaes 2: 65–97. Bergmans W. 1979. Taxonomy and biogeography of African fruitbats of the People’s Republic of Congo, with notes on their reproductive biology (Mammalia, Megachiroptera). Bijdragen tot de Dierkunde 48: 161–186. Bergmans W. 1990. Taxonomy and biogeography of African fruit bats (Mammalia, Megachiroptera). 3. The genera Scotonycteris Matschie, 1894, Casinycteris Thomas, 1910, Pteropus Brisson, 1762, and
Bernard RTF, Cumming GS. 1997. African fruitbats: evolution of reproductive patterns and delays. Quarterly Review of Biology 72: 253–274. Cabrera A. 1908. Lista de mam´ıferos de las posesiones espan˜ olas del Golfo de Guinea. Memorias de la Sociedad Espan˜ola de Historia Natural 1: 435–456. DeFrees SL, Wilson DE. 1988. Eidolon helvum. Mammalian Species 312: 1–5. Eisentraut M. 1964. La faune de chiropte`res de Fernando Po. Mammalia 54: 529–552. Exell AW. 1968. Pr´ıncipe, Sa˜ o Tome´ and Annobo´ n. Acta Phytogeographica Suecica 54: 132–136. Figueiredo E. 1994. Diversity and endemism of angiosperms in the Gulf of Guinea islands. Biodiversity and Conservation 3: 785–794. Fuster JM. 1955. Vulcanolog´ıa del Atla´ ntico meridional. Archivos del Instituto de Estudios Africanos 33: 67–69. Gaiscoigne A. 1994. The biogeography of land snails in the islands of the Gulf of Guinea. Biodiversity and Conservation 3: 794–808. Heim De Balzac H, Hutterer R. 1982. Les Soricidae (Mammiferes, Insectivores) des ˆıles du Golfe de Guine´e: faits nouveaux et proble`mes biogeographiques. Bonner Zoologische Beitra¨ge 33: 133–151. Herd RM. 1987. Electrophoretic divergence of Myotis leibii and Myotis ciliolabrum (Chiroptera: Ves- pertilionidae). Canadian Journal of Zoology 65: 1857–1860. Herna´ ndez Pacheco F. 1943. Annobo´ n, u´ nica tierra espan˜ ola en el hemisferio austral. Revista Africa 19/20: 44–51. Iliopoulou-Georgudaki J. 1986. The relationship between climatic factors and forearm length of bats: evidence from the chiropterofauna of Lesbos island (Greece–east Aegean). Mammalia 50: 476–482. Jones PJ. 1994. Biodiversity in the Gulf of Guinea: an overview. Biodiversity and Conservation 3: 772–784. Juste J, Iba´ n˜ ez C. 1994. Bats of the Gulf of Guinea islands: faunal composition and origins. Biodiversity and Conservation 3: 837–850. Juste J, Machordom A, Iba´ n˜ ez C. 1996. Allozyme variation of the Egyptian Rousette (Rousettus egyptiacus; Chiroptera, Pteropodidae) in the Gulf of Guinea (West–Central Africa). Biochemical Systematics and Ecology 24: 499–508. Juste J, Iba´ n˜ ez C, Machordom A. 1997. Evolutionary relationships among the African fruit bats:
Juste J, A´ lvarez Y, Tabare´ s E, Garrido-Pertierra A, Iba´ n˜ ez C, Bautista JM. 1999. Phylo- geography of African fruitbats (Megachiroptera). Molecular Phylogenetics and Evolution 13: 596–604. Kilpatrick CW. 1981. Genetic structure of insular populations. In: Smith MH, Joule J, eds. Mammalian population genetics. Athens, GA: The University of Georgia Press, 28–59. Kitchener DJ, Hisheh S, Schmitt LH, Maryanto I. 1993. Morphological and genetic variation |