Chromosomal polytypy, morphological conservatism and absence of data have frustrated the taxonomic revision of two species of southern African-endemic laminate-toothed rats (Otomys irroratus and Otomys saundersiae s.l.). New cytogenetic (G-banding and fluorescence in situ hybridization), DNA sequence [cytochrome b (cyt b) gene] and geometric morphometric data demonstrate the synonymy of O. saundersiae from Grahamstown (Eastern Cape, South Africa) under O. irroratus, and the validity of Otomys karoensis from the Fynbos Biome of the Western Cape. Phenotypic dimorphism in pelage colour and cranial morphology in O. irroratus from the climatically unpredictable Albany Thicket (= Savanna) Biome of the Eastern Cape results from the retention of allometric paedomorphic traits in some adults (saundersiae morph) but not others. The same paedomorphic traits are associated with speciation and karyotypic and genetic differentiation in O. karoensis. Within O. irroratus, two phenotypically and genotypically (cyt b divergence = 6.4%) divergent lineages correspond with the Fynbos/Albany Thicket and Grassland biomes. Incipient speciation in O. irroratus seems to be associated with ecology rather than karyotype.

The distribution and environmental conditions of bark consumption by small rodents were compiled for the northern and southern hemispheres. Main bark consumption occurred under snow at peak cyclic density and in the dry season of annually fluctuating rodent populations, respectively. The species responsible appeared more folivorous in the north. Among woody species that are damaged, a large number were in both cases exotics. Various defences against bark consumption have evolved, including secondary compounds. However, this defence appears inadequate when indigenous trees are grown under new environmental conditions. It is pointed out that the stem and bark are very sensitive components of perennial trees and bushes and that this potential but usually well protected resource will only be used in crucial situations by various mammals. In view of the many unsuccessful attempts to prevent small rodent damage by bark consumption, costs for rodent damage should be included in any cost–benefit analyses for plantations of exotic species or indigenous species under new conditions. Screening and even selection for resistant varieties may be needed at large scale plantations.

Abstract:Altitudinal transects of biodiversity are important to understanding macro-ecological patterns. Hump-shaped altitudinal profiles in species richness are a common pattern in terrestrial small-mammal communities studied previously mostly in New World mountain ranges. Based on capture–mark–recapture live-trapping conducted over four seasons (four nights per session) along an altitudinal transect from 1000 to 1747 m asl on the southern slopes of the Soutpansberg Mountains of northern South Africa, we recorded 12 species of small mammal, including nine rodents, two shrews and one elephant shrew (sengi). Three species of rodent and the sengi dominated the communities whilst six species were recorded with three or fewer captures. Although we found no effect of season, we found a significant linear increase in terrestrial small-mammal richness (but not Simpson's diversity, D) with altitude in the Soutpansberg Mountains located just north of the Tropic of Capricorn in South Africa. This trend could best be explained by an increase in the proportion of rocky cover and a change in vegetation from savanna to grassland. Peak densities of three common rodent species and one common insectivore varied from 10 ha−1 (Elephantulus myurus) to 54 ha−1 (Micaelamys namaquensis) and density showed no significant variation with season. However, density either declined (Aethomys ineptus) or increased (Rhabdomys dilectus and Elephantulus myurus) significantly with altitude; this variation was best explained by the proportion of rocky cover. Seasonal variation had a significant effect on body mass of two rodent species. A review of small-mammal studies from major mountain ranges in Africa and Madagascar found that the hump-shaped pattern is not common. Declines in richness with increasing altitude as well as cases of no significant altitudinal change were the most common patterns noted. Tropical African mountains (including the Soutpansberg with 27 species) have far higher small-mammal diversity than temperate African mountains.

Climate change constitutes a potential threat to montane biodiversity, particularly in low-altitude, tropical mountains; however, few data exist for the Afromontane taxa. In South Africa, the temperate grassland and fynbos biomes are mostly associated with the Great Escarpment and the high-lying central plateau. Varying contractions of the grassland and fynbos biomes are predicted under different climate scenarios by 2050. Animal taxa adapted to these biomes should suffer similar range declines and can be used to independently test the vegetation models. We constructed MaxEnt models from 271 unique locality records for three species of montane and submontane vlei rats that are closely associated with grassland angoniensis) due to historically documented changes from a grassland-dominated to thicket-dominated landscape.

Forest or mouse shrews (Myosorex) represent a small but important radiation of African shrews generally adapted to montane and/or temperate conditions. The status of populations from Zimbabwe, Mozambique, and the north of South Africa has long been unclear because of the variability of traits that have traditionally been 'diagnostic' for the currently recognized South African taxa. We report molecular (mitochondrial DNA and nuclear DNA), craniometric, and morphological data from newly collected series of Myosorex from Zimbabwe (East Highlands), Mozambique (Mount Gorogonsa, Gorongosa National Park), and the Limpopo Province of South Africa (Soutpansberg Range) in the context of the available museum collections from southern and eastern Africa and published DNA sequences. Molecular data demonstrate close genetic similarity between populations from Mozambique and Zimbabwe, and this well-supported clade (herein described as a new species, Myosorex meesteri sp. nov.) is the sister group of all South African taxa, except for Myosorex longicaudatus Meester & Dippenaar, 1978. Populations of Myosorex in Limpopo Province (herein tentatively assigned to Myosorex cf. tenuis) are cladistically distinct from both Myosorex varius (Smuts, 1832) and Myosorex cafer (Sundevall, 1846), and diverged from M. varius at approximately the same time (2.7 Mya) as M. cafer and Myosorex sclateri Thomas & Schwann, 1905 diverged (2.4 Mya). Morphometric data are mostly discordant with the molecular data. For example, clearly distinct molecular clades overlap considerably in craniometric variables. On the other hand, extreme size differentiation occurs between genetically closely related populations in the Soutpansberg Range, which coincides with the bissection of the mountain range by the dry Sand River Valley, indicating the potential for strong intraspecific phenotypic divergence in these shrews.