Red List of South African Species

Alternatively, Explore species
Vulnerable (VU)

Rationale (Changed due to New Information|Incorrect data used previously)

The White-tailed Rat is a Highveld grassland specialist (whilst also marginally occurring in Succulent Karoo and Fynbos biomes; Dean 1978) and is endemic to South Africa and Lesotho. Although it occurs widely across the assessment region, it has an extremely patchy and fragmented area of occupancy due to its preference for microhabitats within vegetation types and transitory habitats created after fire. Out of a total area of 417,452 km2 broadly available to the species, we inferred an effective occupancy of only 3,719–12,061 km2. They are one of the rarest species in the small mammal community, as demonstrated by consistently low trapping records. Given estimated densities ranging from 0.9–37 colonies (breeding pairs) / km2, the most likely mature population size is 6,997–13,648 individuals. Using grassland vegetation type as a proxy for subpopulation delineation, subpopulation size averages 178 ± 177 colonies and ranges from 1–760 colonies, which means the largest subpopulation is likely to be 712–1,521 mature individuals. There is an inferred continuing decline from grassland habitat loss (due to expansion in crop agriculture, urban and industrial development, and climate change) and grassland habitat degradation, primarily from suppression of natural fire regimens. This may represent an emerging threat to this species if fire is increasingly suppressed or controlled with the rise of intensive wildlife breeding. Grasslands are the most threatened biome in the assessment region with at least 33% transformed already. Thus, we list this species as Vulnerable C2a(i) under a precautionary purview. Although it was listed as Endangered A3c in previous assessments, there is no evidence for a population decline of >50% and overgrazing is a less severe threat for this species than originally thought. Further field surveys are needed to estimate population size and trend more accurately, and this species should be reassessed once more data are available. Key interventions include protected area expansion of grassland habitat and the implementation of suitable grazing and fire management systems for nature reserves, livestock and wildlife ranches. Research and monitoring is desperately needed to test the effectiveness of such interventions and to produce baseline information on its ecology and biogeography.

Distribution

White-tailed Rats are endemic to South Africa and Lesotho, where they inhabit Highveld grasslands primarily, but also Succulent Karoo and Fynbos (Dean 1978). Within South Africa, it occurs in the following provinces: southern Mpumalanga, Free State, the high-lying areas of KwaZulu-Natal (KZN), Eastern Cape, southeastern North-West, and marginally into Northern Cape (Skinner and Chimimba 2005, Avery and Avery 2011). There is a potentially relict subpopulation in the Fynbos of the Western Cape where it probably once occurred extensively throughout the southern grasslands, heathlands and renosterveld of the Cape. There are no records from Eswatini, despite extensive surveys (Monadjem pers. comm. 2015). No recent range expansions or contractions have been documented. However, it is likely that the population is fragmented and will contract in the future as grasslands continue to decline (Mucina and Rutherford 2006, Driver et al. 2012). Additionally, there is a suggestion that, historically, its range had already contracted southwards in the Northern Cape region: owl pellet analysis reveals Pleistocene records from Equus and Wonderwerk Caves, as well as records of the species in 10 other localities in the Kuruman area and surrounds from the Holocene period, while more recent owl pellet records (1930 to present) only reveal localities from the previously-mentioned regions of the province (Avery and Avery 2011), which suggests a long-term range contraction.

This species is widespread across the assessment region but patchily distributed, and so estimating area of occupancy (AOO) is challenging. Since they are habitat specialists, we extracted all vegetation types in which they are recorded to occur (Mucina and Rutherford 2006), using the distribution data shown in Figure 2 (see Supplementary Material PDF attached). This yielded an AOO of 417,452 km2. However, we suspect this is a huge overestimate as they often occur in just one of the sampled sites in a region and often are not found in sites where they are expected to occur (for example, Vermeulen 2005; O’Farrell et al. 2008; Avenant 2011; Kok et al. 2012; N.L. Avenant, G. Palmer, B. Wilson unpubl. data). For example, although Rondeberg Private Nature Reserve shares similar habitat types with the Blaauwberg Conservation Area (BCA) in the Western Cape Province, and was identified as a site where the species should occur, they have only been recorded at BCA (Vermeulen 2005). More worryingly, a recent province-wide survey in the North-West did not reveal a single record for the species, where the last museum record was from Boskop Dam Nature Reserve in 1990 (Power 2014). This is most likely due to the heavy transformation of the grassland biome in the province (Driver et al. 2012). Similarly, although an owl pellet indicating the presence of the species in Tygerberg Nature Reserve, Western Cape, was found in 1992, a recent survey did not find any individuals (Vermeulen 2005). Furthermore, they are suspected to exist in transitory habitats created after fire, having been found equally in burnt and unburnt patches in the Western Cape Province and in low to mid-successional habitats following fire in the Free State Province (Kuyler 2000, Kaiser 2006, Specker 2006, Avenant and Cavallini 2007, Avenant 2011, Avenant and Schulze 2012, Morwe 2013). For example, although they are predicted to occur widely in Gauteng Province, they have not been found in the Cradle of Humankind region, possibly because the area does not burn. Thus, this species likely occurs in microhabitats (see Habitats and Ecology) that are too fine-scale to be approximated by vegetation types and that occur in a shifting mosaic of patches following fire.

Given the information above, we subsequently revised the AOO estimate by summing the area of the occupied quarter degree grid cells (using both historical and current records), which yielded 78,470 km2. We then subtracted the amount of transformed habitat as of 2014 (agricultural and urban areas, 57% of the total) and non-grassland habitat, which yielded 34,026 km2 (GeoTerraImage 2015). Finally, we used recent survey data from game farms, wildlife ranches and private conservation areas (A. Taylor unpubl. data) within White-tailed Rat habitat to estimate the proportion of natural habitat that is never burned, burned occasionally and burned using a mosaic or patch approach (which is suspected to be the best fire management system for the species, see Habitats and Ecology). Using data from Free State Province alone (within the core range of the species), only an estimated 14% of the land occasionally burns (30 km2 of a total 212 km2 sampled, N = 13 total properties). Extrapolating this proportion to the total estimated natural grassland habitat available to the species overall yields 4,764 km2. Similarly, if we use the nationwide sample, the proportion of land that burns occasionally increases to 35% (4,094 km2 of a total 11,546 km2 sampled, N = 204 total properties). This yields an effective AOO of 12,061 km2 (extrapolating across the range of the species). Lastly, using only the properties that employ mosaic burning (11% of total sampled area, N = 6 properties; 1,262 km2 of a total 11,546 km2 sampled), the effective AOO is reduced to an extrapolated 3,719 km2. We feel that measuring AOO at the quarter degree grid scale and using both old and recent records approximates the dynamic and patchy distribution of the species better than coarse vegetation types. Further research should determine in more detail its habitat preference and distribution so as to improve the measurement of AOO.

Population trend

Trend

Dean (1978) regards the White-tailed Rat as a relict species. Although conservationists have been concerned with it for over forty years, it still persists at low densities. While no empirical population estimates or trends are available, they are consistently one of the rarest species encountered and infrequent capture rates during surveys indicate that population size is low. Friedmann and Daly (2004) reported 5–10 specimens in 15–20,000 trap nights / year in the Free State Province (but only a small percentage of those trap nights took place in habitats where one expected to find the species). Similarly, with over 7,500 small mammal records collected over the current Northern Cape (concentrating in the old demarcated provincial boundaries) and current North-West (the western areas which used to be Northern Cape) provinces, only two records (from Benfontein Farm, near Kimberley) have been archived in the McGregor Museum (0.03% reporting rate, B. Wilson unpubl. data). The most recent museum record from North-West Province was from Boskop Dam Nature Reserve (de Beer 1990), but a trapping effort of 240 trap nights in 2013 did not detect the species (Power 2014). Recently, in the Blaauwberg Conservation Area, Western Cape, two studies had a trap success of 1.9% and 1.3% respectively (900 and 800 trap nights corresponding to 17 and 10 individuals respectively) (Vermeulen 2005, Specker 2006). The species may have a low sighting record as it is naturally rare and uses a burrow system.

Their patchy distribution may explain the lack of trapping success, but once a patch is found they are trapped easily if present. No home range or dispersal rate data are available, but records indicate a clumped distribution within suitable habitats. The two individuals trapped on Benfontein Farm, just east of Kimberley, were trapped in the same area, although several years apart, despite consecutive annual surveys at similar trapping effort (B. Wilson, unpub. data). Within the Free State Province, colonies do not seem to be in close proximity as, within a homogenous area-trap plot of 0.027 km2 (100 traps spaced 5 m apart with a border area of 25 m around the traps), a maximum of one colony is present (N.L. Avenant unpubl. data). Furthermore, a colony is found in approximately one in every 20 area-trap plots, which means colony density can be preliminarily estimated at one colony / 0.55 km2 or 1.8 colonies / km2 (N.L. Avenant unpubl. data). We can assume a minimum of two mature adults / colony (colonies are small, consisting of < 10 individuals). If we use maximum density as 1 colony / 0.027 km2 and minimum density as one colony / 1.05 km2 (assumed to be the positive interval between the maximum and medium density estimates), a range of population sizes can be estimated for the different AOO parameters described in Distribution (see Table 1 in the Supplementary Material PDF attached). Colony numbers and corresponding mature population size ranges over orders of magnitude, depending on the AOO estimate. The most likely density estimates (medium and minimum) range from 3,499 colonies (6,997 mature individuals) to 22,131 colonies (44,262 mature individuals). Given that colonies (breeding pairs) are suspected to be isolated and may effectively represent a single breeding unit, colony number is probably a more accurate measure of mature population size (sensu IUCN Standards and Petitions Subcommittee 2014).

Defining subpopulations for this species is difficult as we lack basic information about its ecology and distribution. However, we suspect that using vegetation types as proxies for subpopulation distribution may be an adequate approximation since the species may rely on fire to create suitable habitat patches and fire frequency will depend on the fuel load and climate of the vegetation type or bioregion. Subpopulations are suspected to be severely fragmented due to the fragmented nature of the grassland biome (Neke and du Plessis 2004), as well as the prevention of fire as an ecological process or cessation of burning as a management tool by landowners in many areas (A. Taylor unpubl. data). By treating each vegetation type within the grassland biome as a subpopulation (N = 71 subpopulations), and using the medium density estimate from Table 1 (as well as adjusting for the likely proportion of land that is occasionally burnt), average subpopulation size is 178 ± 177 colonies (357 ± 355 mature individuals), with a maximum of 760 colonies (1,521 mature individuals) in a subpopulation. However, we feel that these are overestimates given that the remaining area of the vegetation types were calculated using 1997 satellite data integrated with more recent regional data where possible, and thus mostly do not reflect the habitat lost over at least the past decade. The area currently available to each subpopulation is likely to be significantly less and thus subpopulation sizes are likely to be smaller or even locally extinct (Power 2014).

Threats

Habitat loss of grasslands, both from agricultural and industrial or urban expansion (accounting for 97% of the transformed land in North-West Province, NW READ 2014), as well as in the future from climate change (through bush encroachment, van Jaarsveld and Chown 2001), is the greatest threat to this species. This species is also threatened by afforestation of Highveld grasslands (Armstrong and van Hensbergen 1996b); for example, Mpumalanga has already lost at least 9% of its natural vegetation to forestry plantations (Lötter et al. 2014). Mining is also a major ongoing cause of habitat loss, where 25% of the surface of Mpumalanga is under mining application (Lötter et al. 2014), most of which is coal mining.

It appears as if this species has adapted to a mosaic of disturbances, such as that brought about through patch burns (through natural vegetation fires) and/or overgrazing (perhaps from when herds were migratory across the Highveld). Even in formally protected areas, most of these disturbances are managed, thereby prohibiting natural fluctuations. Suppression of fire is thus suspected to be the next most severe threat, as the microhabitats that the species requires are not created or recycled. Although Coetzee and Monadjem (2008) list grazing pressure as contributing to habitat loss, data from Lesotho indicate that the species occurs in overgrazed areas (Avenant 2011, N.L. Avenant unpubl. data).

Uses and trade

They are used extensively in laboratory research (Hall 3rd et al. 1967). They may occasionally be utilised as bushmeat but are not a target species.

Conservation

Due to its patchy distribution, this species often does not occur in protected areas that one would expect to find them (for example, De Graaff 1974, Vermeulen 2005). Further surveys should document the full list of protected areas in which it occurs. Currently (post-2000 records), there are confirmed records for Tussen-die-Riviere Nature Reserve (NR) (Ferreira and Avenant 2003), Maria Moroka NR, Soetdoring NR, Erfenisdam NR and De Brug Army Training Camp in the Free State Province; Blaauwberg Conservation Area (BCA) in the Western Cape Province; and Abe Bailey Nature Reserve in Gauteng Province and Boskop Dam Nature Reserve in the North-West.

Conservation of grasslands through protected area expansion and biodiversity stewardship schemes is suspected to be the most important intervention for this species. For example, the Maloti-Drakensberg Transfrontier Conservation and Development Area, established in 2001, accounts for the majority of the c. 2% of the protected grassland within the assessment region (Carbutt et al. 2011) and has higher levels of species richness than neighbouring communally grazed areas (Everson and Morris 2006).

Similarly, biodiversity stewardship schemes are suspected to be conserving grassland patches that would benefit this species. For example, a study in the Little Karoo region (overlapping with White-tailed Rat distribution) found that private conservation areas conserved almost three times the amount of land as formally protected areas for both Succulent Karoo and Renosterveld vegetation types (Gallo et al. 2009), both habitats where the species occurs (albeit marginally).

Grassland condition must also be conserved through correct grazing and fire management. Anecdotal observations suggest that overgrazing is not detrimental to White-tailed Rats (N.L. Avenant unpubl. data). Interestingly, research suggests that continuous cattle grazing systems may have the least negative effect on grasslands (as opposed to continuous sheep grazing or high stocking levels and rotational grazing systems, O’Connor et al. 2010), because continuous grazing dissipates impact across the landscape whereas rotational grazing concentrates impacts on small areas while also leading to fragmentation of the landscape through fencing and infrastructure. Thus, depending on stocking level, game and livestock can be the best land use for grasslands (as opposed to intensive dairy farming, crop agriculture, or rural settlements, O’Connor and Kuyler 2009), which supports the occurrence of this species on such land uses, and should be promoted to limit the sprawl of alternative, damaging land uses.

As this species may be connected with post-fire patches, the pristine condition of the grassland may not be as limiting as the lack of ecosystem functioning and thus phase diversity. Landscapes that burn occasionally are suspected to have greater conservation value. Thus mosaic burning, within the natural fire frequency of the landscape and within the goals of grazing management for the relevant ungulate species, should be used by managers to promote the persistence of White-tailed Rats, and the effects should be monitored.Further field research is needed to produce evidence for the conservation effectiveness of the proposed interventions for this species specifically (see Research Priorities).

Recommendations for land managers and practitioners:
  • Incorporate monitoring the species as a possible indicator of disturbed grassland patches (Avenant 2011), as its presence may indicate that the ecosystem is functioning and that patch heterogeneity exists in the landscape.
  • Managers must use fire as a management tool.
  • Managers should not over-stock ranch lands and should be encouraged to switch to cattle or wildlife.
  • Survey work, including the collection of owl pellets, is needed to determine its occurrence in different land-management areas.
Research priorities:
  • Basic ecological research on habitat requirements (and ability to exist on modified landscapes), movement/dispersal capacity between patches, and biogeography.
  • Research on the effects of different management systems on presence and abundance. This should be compared to a baseline subpopulation in suitable protected habitat (control).
  • Research to test the effectiveness of utilising best management practices as identified or hypothesised above.
  • Research to determine colony size, reproduction, dispersal/movement between patches.
Encouraged citizen actions:
  • Report sightings on virtual museum platforms (for example, iSpot and MammalMAP). As this species is easily identified from its white tail and white-grey fur, photographs and coordinates of this species, especially outside of protected areas (for example, livestock and wildlife ranches), will help us to determine its habitat preferences and ecology more accurately.
  • Collect and submit owl pellets to relevant museums.
  • If a game farm or conservancy landowner, use a mosaic fire management as a tool to create microhabitats for this species. Its occurrence should be monitored.

Lead agencies, Partners and Funders

See the partners page