本文範囲 書籍 pp. 329–336(PDF 340–347ページ)。本文は p. 334 まで。
本文と節見出しを収録。原文のページ番号は書籍のノンブルに従う。
p. 329|段落 V2-P329-01
原文
Millipedes occur on all continents except Antarctica. Their known distribution was mapped in great detail by Shelley & Golovatch (2011) who painstakingly accumulated information on the distribution of higher taxa of millipedes, down to the level of order, and presented their findings on numerous maps. They further presented detailed accounts of distribution patterns and used these, together with extensive paleogeographic material, and taking known millipede fossils into consideration, to construct hypotheses about origin and dispersal of higher millipede taxa.
MILLIPEDES AS PROVIDERS OF BIOGEOGRAPHICAL EVIDENCE
p. 329|段落 V2-P329-03
原文
Most species of millipedes have very small distribution areas, i.e., the level of endemicity is high. An example from the millipede fauna of Europa, incontestably the best studied major part of the World, is provided by volume 1 of the atlas of European species (Kime & Enghoff, 2011). The atlas covers all orders except Julida and Chordeumatida and includes 492 species — roughly a third of the known European fauna. Of these, 183 are known from only one 50 × 50 km square, and 252 — more than half — are known from only 1-3 adjacent 50 × 50 km squares.
Limited distribution — endemicity — also obtains at higher taxonomic levels. For example, 10 out of the 15 families of the order Julida recognised by Enghoff (1991, 1993) are known only from western North America, eastern North America, Europe (including Macaronesia, North Africa and the Middle East), or eastern Asia. Even the largest julidan family, Julidae, is limited to temperate Eurasia (the Palaearctic region).
Millipede endemism is particularly rampant in mountains and on oceanic islands. A mountain example is provided by the endemic Afrotropical family Oxydesmidae, of which 21 out of the 27 species described from the Tanzanian Eastern Arc mountains are endemic to just one mountain massif (Hoffman, 1990, 1993, Table 6.1). A study of one genus (Chaleponcus) of another endemic Afrotropical family (Odontopygidae) in the Udzungwa Mountains of Tanzania (Enghoff, 2014) has produced 21 species, all endemic to the Udzungwas. Island faunas are treated below.
The following summary of major millipede faunas is by no means complete. Emphasis is on larger groups of mainly larger forms. For example, several families of small Polydesmida (e.g., Fuhrmannodesmidae and Pyrgodesmidae) are very richly represented in all tropical regions.
The tendency of higher millipede taxa to have limited distributions is obvious even at ordinal level. Table 13.1 shows the occurrence of the 16 currently recognised millipede orders in each of the six major, non-Antarctic biogeographical regions. Only 3 orders occur in all 6 regions (3 further orders marginally do so). In contrast 30 out of 33 insect orders, and 4 out of 5 centipede orders, occur in all 6 biogeographical regions. Each millipede order on average occurs in 3.9 regions if marginal occurrences are excluded.
Phylogeny-based (cladistic) interpretation of millipede distributions got an early start by Kraus (1966) who analyzed family-level relationships within what is now known as suborder Spirostreptidea of order Spirostreptida. Based on a classical Hennigian analysis, Kraus found the endemic Afrotropical superfamily Odontopygoidea to be the basal clade within Spirostreptidea, sister to a dichomy between Spirostreptidae (Neotropical + Afrotropical) and Harpagophoridae (Afrotropical + Oriental). Kraus argued that the current distribution of Spirostreptidae and Harpagophoridae can be seen as a consequence of the fragmentation of Gondwana and the overlap between both families in Africa as 'secondary'. As to the Odontopygoidea, he argued that their occurrence in Africa is secondary and that they must have arisen as a northern vicariance 'partner' to Spirostreptidae + Harpagophoridae. This, although fully logical, however requires extinction of Odontopygoidea in the Holarctic region and thus appears little likely, but an alternative explanation should await a renewed phylogenetic analysis of Spirostreptida.
The Julida is one of the (almost) exclusively Holarctic millipede orders. A biogeographical analysis of Julida was made by Enghoff (1993), based on previous phylogenetic analyses. One of Enghoff's results concerns the sister generaNemasoma and Orinisobates (fam. Nemasomatidae) the species of which are allopatrically and disjunctly distribution across the Holarctic region (Fig. 13.1). Based on the cladogram of Enghoff (1985) the author concluded that the North American species of Orinisobates most likely derive from a trans-Beringian immigrant from East Asia, the opposite direction of dispersal to that of most amphi-Beringian millipede groups according to Golovatch (1979).
Wesener & VandenSpiegel (2009) made a modern cladistic analysis of the families of order Sphaerotheriida (giant pill-millipedes). The resulting cladogram is shown on Fig. 13.2, superimposed on a distribution map. Interestingly, Indian giant pill millipedes are most closely related to Malagasy ones, whereas the SE Asian family Zephroniidae is not particularly closely related to the Indian forms.
By far the best studied swarms are those of the Atlantic Macaronesian archipelagos, notably the Canary Islands and Madeira. Enghoff (2011) summarized earlier studies on the largest two of those which are in the genera Cylindroiulus and Dolichoiulus (both Julida, fam. Julidae). Reboleira & Enghoff (2014) and Enghoff (2012) provided updated information on both genera.
Besides ~100 species from the European continent, North Africa and east to Central Asia, Cylindroiulus has 30 endemic species on the island of Madeira, a further endemic on the satellite island Porto Santo and a 32nd endemic on the Canary Islands. Preliminary phylogenetic studies suggest that this swarm is monophyletic and that the species from Porto Santo and the Canary Islands stem from immigrants from Madeira. There is considerable morphological diversity within the swarm, and most species specialize in a particular (micro)habitat type. Co-existing species mostly are of different sizes.
Dolichoiulus has 50 endemic species on the Canary islands, two further endemics on Madeira and Porto Santo, still another endemic on the tiny Salvage islands situated between Madeira and the Canaries, and three additional endemics on the Cape Verde Islands further south. With one exception, the Canarian species are each confined to one of the seven main islands. Monophyly of the Macaronesian Dolichoiulus swarm cannot be ascertained on the available evidence but a preliminary molecular phylogenetic study by I. Sanmartín suggests that the "root" of the Canarian radiation is on the island of Tenerife which has the highest number of species (20). Canarian Dolichoiulus species are morphologically as diverse as the Madeiran Cylindroiulus and also are habitat specialists, but there is less co-existence at the micro-habitat level.