Tree shrews have attracted considerable interest because of the possibility that they might be related to primates. The family Tupaiidae was originally included within the order Insectivora. Then, in Simpson's seminal classification of 1945, tree shrews were formally transferred to the order Primates. This was particularly because of Le Gros Clark's reports of a series of morphological similarities in the skull and brain and because of Carlsson's account of similarities in the musculature. However, the postulated link between tree shrews and primates was increasingly questioned from 1965 onwards and, following Butler (1972), it is now customary to allocate tree shrews to their own order Scandentia. This change in interpretation was partly due to recognition of the fundamental principle that reliable reconstruction of phylogenetic relationships depends on identification of novel, derived characters to the exclusion of retained primitive features. In fact, the undoubted morphological similarities between tree shrews and primates are arguably attributable to retention of many primitive features augmented by a number of convergent adaptations for arboreal life. For
instance, tree shrews resemble primates in possessing a cecum in the digestive tract, whereas this feature is lacking from typical insectivores. However, it is highly likely that ancestral pla-cental mammals already possessed a cecum, so retention of this feature does not indicate a specific link between tree shrews and primates. For certain other features, independent development of similar characters may have occurred. For example, both tree shrews and primates have a bony strut (postorbital bar) along the outer margin of the eye socket. This is not a primitive feature, as it was undoubtedly lacking from ancestral placental mammals, but it has been developed independently in several mammalian groups, including many hoofed mammals (ungulates), some carnivores, and hyraxes. The likelihood of convergent similarity between tree shrews and primates is increased by the observation that arboreal tree shrew species show closer similarity to primates than do terrestrial tree shrew species. All tree shrews lack all of the clearly defining features of primates that are identifiable in the skull, brain, and reproductive system. Moreover, given the general reliance on dental similarities in reconstructing mammalian evolution, it is surprising that any resemblance between the molar teeth of tree shrews and those of primates has never been proposed. Superficially, tree shrews do resemble strepsirrhine primates (lemurs and lorises) in possessing a tooth comb in the lower jaw. However, in tree shrews the comb is formed by six incisors, whereas in strepsirrhine primates it is formed by two canines and four incisors, so convergent evolution is again the most likely explanation. Analyses of sequences for both nuclear and mito-chondrial DNA have consistently failed to indicate any link
between tree shrews and primates. Instead, there have been several indications that tree shrews may be related to Lagomor-pha (rabbits, hares, and pikas).
When tree shrews were still included in the order Insectivora, they were united with the elephant shrews (Macroscelidea) in the suborder Menotyphla, partially because of shared possession of a caecum, while all other insectivores lacking a cecum were placed in the suborder Lipotyphla. Subsequently, it was suggested that the superorder Archonta should be established for Menotyphla, Chi-roptera (bats), Dermoptera (colugos), and Primates together. Now that the tree shrews have generally been excluded from the order Primates, there have been several attempts to resurrect the superorder Archonta (while discarding the elephant shrews from this assemblage). However, this alternative attempt to link tree shrews to primates along with certain other mammals is subject to the same problems as inclusion of tree shrews in the order Primates. The characters supposedly linking primates, tree shrews, colugos, and bats can be attributed to a combination of retained primitive features and convergent arboreal adaptations. DNA sequences provide no convincing evidence for any link between bats, primates, or tree
shrews, although there is some indication that colugos may be related to primates (but not tree shrews or bats).
The fossil record for tree shrews is very sparse, and many proposed fossil relatives have now been excluded from the group. It was originally believed that the Oligocene Anagale from North America was directly allied to tree shrews, but detailed examination instead suggested a link to lagomorphs (rabbits and their allies). However, some fragmentary material (partial skulls, isolated teeth, and possibly a ribcage) from Miocene deposits of India (Palaeotupaia), Pakistan (unnamed genus), and China (Prodendrogale) has been convincingly attributed to tree shrews. Furthermore, a possible early fossil relative of tree shrews (Eodendrogale) has also been reported from Eocene deposits of China.
Was this article helpful?