Uduba

Griswold, Charles, Ubick, Darrell, Ledford, Joel & Polotow, Daniele, 2022, A Revision of the Malagasy Crack-Leg Spiders of the Genus Uduba Simon, 1880 (Araneae, Udubidae), with Description of 35 New Species from Madagascar, Proceedings of the California Academy of Sciences 67, pp. 1-193 : 97-99

Geography of Uduba and Udubidae

A principal aim of the Fisher-Griswold Arthropod Team and the Schlinger et al. Malaise Trap Surveys was to sample widely in Madagascar, and especially in the western, drier parts of the island that had been less well sampled in recent years than have the eastern forests and vicinities of the port cities. In addition our knowledge of Uduba spiders has benefitted from the effort of Dr Steve Goodman and members of the Field Museum of Natural History and Madagascar collaborators who have sampled widely for small vertebrates, catching many spiders at the same time, by Frontier Madagascar, and by the efforts of Vincent and Barbara Roth, who traveled extensively though the island in the early 1990’s.

Uduba are known from most parts of Madagascar and seem to be absent only from the driest parts of the far south ( Fig. 85 ). Although the Fisher-Griswold and Schlinger et al. surveys did make collections in the driest parts of the island, even at Cape Saint Marie in the far south, no Uduba were collected there. The vast areas of burned-over grassland in the central uplands of the island may also lack Uduba . The earliest Uduba species discovered, which are U. madagascariensis ( Vinson, 1863) (Maps 4 and 21), U. dahli Simon 1903 (Maps 2 and 13) and U. funerea Simon 1906 (Maps 2 and 15), seem to have been collected in semiarid, inhabited parts of the island. At least U. madagascariensis occur in a variety of semiarid and also disturbed habitats including native forest on the plateau, dry forests in the west and southwest, in Uapaca forest and even in disturbed areas near the large city of Antananarivo (Map 4). A few species are recorded from dry forests and spiny forest in semi-open formations, though no species is clearly restricted to such localities. The majority of Uduba species have been found in wet forests, from low elevation to montane habitats. At least Uduba barbarae have been collected at near sea level in Lokobe Forest (Map 4) and individuals of U. woodae (Map 8) and U. taralily (Map 10) have been collected at less than 50 m elevation, although neither of these species is restricted to low elevation. By contrast many species have been collected in montane forest: records of U. goodmani at 1990 meters in Reserve Andringitra (Map 6) represent the highest elevation for Uduba yet known.

The species Uduba balsama and U. woodae are predominantly known from dry parts of Madagascar (Map 14). Uduba balsama are widely distributed in the drier parts of western Madagascar with at least one outlier from wet rainforest in the northeast ( Map 11 ). Individuals of U. balsama have been collected in dry habitats in southwestern Madagascar, i.e., from spiny forest at 825 m elevation in Zombitse National Park, tropical dry forest at 100m elevation in Forêt de Kirindy and from yellow pan traps in gallery forest on sandy soil in the Mackay Mountains. Some individuals of presumed U. balsama have been collected in lowland rainforest from 400-800m near Marojejy in northeastern Madagascar, which indicates that this species may not be restricted to dry habitats. The species U. woodae also appears to occur mainly in dry parts of Madagascar (Map 8) and is recorded from dry deciduous forest, tropical dry forest, in transitional dry forest at Andohahela, Uapaca forest and even spiny forest: this may be a unique dry forest specialist. But, as is the case in U. balsama , there is a unique collection of a U. woodae from a wet area at Anjaraharibe-Sud. The species pair Uduba rinha and U. irwini occurs predominantly on the western side of the great escarpment ( Maps 3 , 18). Uduba rinha are recorded across the northern half of Madagascar and Uduba irwini are recorded across the central and southern half of Madagascar on the high plateau and in western dry forests.

Because we have no cladogram for Uduba species it is difficult to discuss the distributions of related species. Within our provisional species groups most species are allopatric. There is one apparent case of sympatry (Map 16) for the closely-related species U. evanescens ( Map 7 ) and U. pseudoevanescens (Map 4). At the rich Uduba site of Ranomafana in southeastern Madagascar ( Fig. 87 ), females of U. pseudoevanescens have been collected on 5 March and 19-20 July 1992 by V. and B. Roth, whereas males and females of U. evanescens have been collected in the Ranomafana area virtually throughout the year. The absence of collections U. evanescens in August and September seems only to be an artifact of fieldwork. Sympatry between sister species pairs of spiders has been recognized previously from Madagascar. For example, the phyxelidid sister species pairs Ambohima zandry and A. avaratra and Rahavavy ida and R. fanivelona have been recorded in sympatry ( Griswold et al., 2012: 754-755).

Communities of sympatry of Uduba . The extensive collecting efforts undertaken over the last 25 years has enabled us to focus efforts and even to return to some sites multiple times. Some places have been particularly notable for the rich community of Uduba that occur in geographic sympatry ( Figs. 86 , 87 ). Even though we have more than 30 years of intensive sampling we still cannot account for the possibility of apparent geographic differences in species distributions that may in fact be due to elevation, season, or even to climate change. Our discoveries reveal the richest communities of Uduba sympatry along the mountain spine on the wet, eastern side of the island. A few localities in the arid west are also notable. Two species are known from Namoroka ( Fig. 87 ), ( Uduba rinha and the endemic Uduba lehibekokoa ), two from Tsingy de Bemaraha ( Fig. 87 ) ( Uduba balsama and U. irwini ), two from Forêt de Milua ( Fig. 87 ) ( Uduba taralily and U. woodae ) and three species are known from the isolated southwestern patch of semiarid forest at Analavelona ( Fig. 87 ) ( Uduba irwini , U. madagascariensis and U. schlingeri ). An isolated patch of montane forest in the far north at Montagne d’Ambre ( Fig. 86 ) revealed two sympatric species: the widespread Uduba schlingeri and the endemic U. salegy . A large area of little explored mountain forest in the northeast stretching across a river valley from Marojejy to Anjanaharibe-Sud ( Fig. 87 ) hosts at least five species: the widespread Uduba balsama , U. lamba and U. woodae and the endemic Uduba sarotra . Uduba platnicki (Map 2) occurs in this forest and also in the rich forests of the Masoala Peninsula, just to the south. The northeastern forests of Madagascar are renowned for richness and endemism and promise many more discoveries from some of the least explored parts of the island. The tiny remnant sacred forest at Ambohimanga near the capital Antananarivo ( Fig. 86 ) is home to two sympatric species, the widespread Uduba madagascariensis and the endemic Uduba volana : this forest probably represents a formation that was widespread on the central plateau of Madagascar before the arrival of people. Another small remnant forest in central Madagascar at Ambohitantely ( Fig. 87 ) hosts at least four sympatric species: the widespread Uduba evanescens (Dahl, 1901) , U. fandroana and U. irwini and the endemic Uduba danielae . There is a biogeographic pattern in Madagascar that is commonly known as the “Périnet effect” ( Lees, 1996). As part of this pattern the highest levels of richness and endemism are recorded from the middle of the island rather than in the north or south: these foci of richness may be related to mid-latitudinal and mid-elevational peaks in richness that arise from overlapping ranges of widespread species at these mid zones. This pattern is not exhibited by Uduba . There are places rich in species at and near Périnet, but higher levels of sympatric richness have been found in forests to the south. At the mountain forest at Anjozorobe ( Fig. 86 ), we have Uduba fandroana , U. fisheri and the endemic Uduba jayjay . At the famous region of Périnet itself (Analamazaotra, Mitsinjo and Périnet, ( Fig. 86 ) we know of four sympatric species. These include Uduba pseudoevanescens and U. rinha and the endemics Uduba ida and U. rakotozafy . The species Uduba lavitra is known only from two nearby localities in montane rainforest where their forest habitat is threatened by a huge chrome-ore strip mine (Map 8). Areas of the highest sympatric richness are in the south and far south of the eastern mountain spine. At the far southern locality of Andohahela ( Fig. 86 ) four sympatric species are known: the widespread Uduba halabe , U. schlingeri and U. woodae and the endemic Uduba goodmani . Ranomafana ( Fig. 87 ), a famous lemur locality ( Wright, 1997; Wright and Andriamihaja, 2003), which includes the localities Bellvue, Vatoharanana, Talatakely and Vohiparara, hosts eight sympatric species: Uduba evanescens (Dahl, 1901) , U. fandroana , U. halabe , U. kavanaughi , U. milamina , U. pseudoevanescens and U. schlingeri and the endemic Uduba ibonia . Finally, at the southern massif of Andringitra-Ivohibe ( Fig. 87 ) we find nine species in sympatry: Uduba andriamihajai , U. evanescens (Dahl, 1901) , Uduba halabe , U. lamba , U. milamina , U. schlingeri , U. valiha and the endemics Uduba hainteny and U. orona . The Périnet ( Fig. 86 ) area has the highest proportion of endemics (3/5), which is a higher proportion than at Andringitra-Ivohibe (2/9) ( Fig. 87 ).

Transcontinental Biogeography. Uduba were previously discussed as a component of a transcontinental vicariant distribution including America, Africa, Madagascar and Sri Lanka ( Griswold, 2000: 352, fig. 3). That paper on Afromontane “ Zorocratidae ” is partially obsolete: molecular phylogenetics studies have suggested that the “ Zorocratidae ” of that discussion is not a natural group. Recently analyses by Polotow et al. (2015) and Wheeler et al. (2017) suggest that Zorocrates can no longer seen as closely related to Uduba . In those same studies a close phylogenetic relationship between Raecius (Africa), Uduba and Zorodictyna ( Madagascar) has been corroborated by molecular and morphological data ( Polotow et al., 2015) to comprise the core of the family Udubidae Griswold and Polotow 2015 . The placement of the genus Campostichomma from Sri Lanka, previously allied to Uduba by morphological data ( Griswold, 1993), remains unchallenged. We still have evidence for a transcontinental distribution of a clade, here of the Udubidae , but vicariance with the Americas is no longer part of the picture.

Figures

FIGURE 85. All locality records (stars) from this study for Uduba in Madagascar.

MAP 11. Distributions of Uduba balsama and Uduba MAP 12. Distributions of Uduba lamba, Uduba lakroa heliani. and Uduba fandroana.

MAP 3. Distributions of Uduba rinha and Uduba irwini. MAP 4. Distributions of Uduba madagascariensis, Uduba barbarae, Uduba pseudoevanescens and Uduba orona.

MAP 7. Distributions of Uduba evanescens (Dahl, 1901) MAP 8. Distributions of Uduba woodae, Uduba lavitra and Uduba lehibekokoa. and Uduba milamina.

FIGURE 87. Notable communities of sympatry for Uduba in Madagascar. Ambohitantely (1/4), Analavelona (0/3), Andringitra–Ivohibe (2/9), Foret Milua (0/2), Marojejy-Anjanaharibe-Sud (1/5), Namoroka (1/2), Ranomafana (1/8), Tsingy de Bemaraha (0/2). (endemic species at site / total species at site).

FIGURE 86. Notable communities of sympatry for Uduba in Madagascar. Ambohimanga (1/2), Andohahela (1/4), Anjozorobe (1/3), Betampona (0/2), Italaviana (0/2), Montagne d’Ambre (1/2), Mt. Papango (0/2) and Perinet–Anomalozaotra (3/5). (endemic species at site / total species at site).