Climate change predictions suggest that the continent most vulnerable to climate change is Africa. The impacts of potential changes which include increases in air temperatures and rainfall variability are negative with potential species extinctions projected throughout southern Africa. A number of climate models have been applied to examine the consequences of climate change for ranges of South African animal species. One such model frequently predicted range shifts from west to east, which is realistic considering the marked aridity gradient in an east-west direction across the country, but the authors suggested that these shifts may not be as marked if species are able to use physiological and behavioural methods to adapt to an increase in aridity. Information on the degree to which behavioural and physiological flexibility affect species range in southern Africa is scant which is surprising given its importance with regard to climate change. Thallomys nigricauda occurs along an east-west aridity gradient in southern Africa, inhabiting mesic, semi-xeric and xeric regions. One would expect phenotypic flexibility in physiological and behavioural traits in response to the diverse environmental conditions to be related to the success and range of the species. The wide distribution and arboreal habits, suggesting that T. nigricauda is exposed to greater extremes of temperature than fossorial rodents, makes T. nigricauda an ideal species to test this assumption. Hence I expected that T. nigricauda would exhibit variation in physiological and behavioural traits measured along an aridity gradient. This has important implications in predicting the survival of small mammal species in the light of climate change in southern Africa. Thallomys nigricauda were live-trapped in winter 2006 and 2007 and summer 2007 using Elliot traps in three sites: mesic site Weenen Game Reserve (KwaZuluiv Natal Province, South Africa); semi-xeric site Haina Game Farm (Botswana) on the northern boundary of the Central Kalahari Desert and xeric site Molopo Nature Reserve (southern Kalahari savannah, North-West Province, South Africa). I studied the home-range size of T. nigricauda by radiotracking 12 males and 16 females in winter 2006, 2007 and summer 2007. Home ranges were estimated using 100% and 95% minimum convex polygons and 95% and 50% fixed kernels. Home ranges varied widely, from 166 to 80199m2 for males and from 46 to 8810m2 for females. Males had larger home ranges than females, which supports a promiscuous mating system reported for the species. Although range size was reduced in both sexes in winter, this was not significant. I found no significant difference in home range size along the aridity gradient. It is suggested that a combination of precipitation, habitat productivity and breeding system influences the size of home range of the species, and that this species displays phenotypic flexibility in terms of its behavioural responses to these factors. I measured the urine concentrating ability (UCA), as indicated by urine osmolality and relative medullary thickness (RMT), and water turnover rate (WTR) of T. nigricauda. There was no significant difference in RMT between sites or sex and no difference in osmolalities when site, season and sex were taken into account. In addition, specific WTR was not significantly influenced by season. Lack of significant differences could be the result of the high degree of individual variation in the traits measured, an indication of the flexibility in UCA and WTR. However, higher urine osmolality and lower WTR’s were recorded in the dry winter months. I quantified the thermal environment perceived by a small, arboreal, mammalian endotherm using a number of methods at three study sites in winter and summer. Our area of interest was how well these methods accurately portrayed the actual temperatures that small mammals are exposed to. Temperature differences between the methods were largest during the midday, when temperatures were highest. All methods recorded a greater range of temperatures during photophase than during scotophase. Black-bulb and model temperatures produced more accurate, rapid measurements when compared to measurements produced by direct temperature recording devices, particularly during photophase, when solar radiation is the major influence of heating. Other methods lagged behind black-bulb measurements. Although the mean temperatures of some of the methods were significantly different, there was a high degree of correlation between all methods, even after randomization and generation of 25% and 10% subsamples. Computed thermal indices and blackbulb temperatures produced similar thermal profiles. In studies requiring accurate time series measurements, it is suggested that black-bulb or copper models be employed rather than direct temperature recording devices. Simpler measurement devices would suffice for studies requiring an estimate of the temperature variation and trends in the microclimate of small mammalian endotherms, particularly arboreal or cavity dwelling species. In the wild, across an aridity gradient, I measured abdominal body temperarture (Tb) of T. nigricauda using implanted iButtons®. All but three T. nigricauda displayed significant 24 h Tb rhythmicity. The Tb range for free-living T. nigricauda was 32.33-40.63 oC (n = 13) and 32.69-40.15 oC (n = 17) in winter and summer respectively. Although there was variation in Tb profiles, T. nigricauda generally displayed a bimodal distribution of Tb, with high and low Tb values during scotophase and photopase respectively. Body temperature range was significantly greater in winter, when T. nigricauda reduced its minimum Tb. It was shown that the maximum amplitude of circadian rhythms of body temperature was on average 259.6% of expected values. To determine the extent to which the microclimate of T. nigricauda cavities assists in the maintenance of Tb, I measured the temperatures of cavities across the gradient, providing an indication of the degree of buffering provided by refugia. I measured the temperatures of shallow and deep regions of cavities using iButtons® in summer and winter and recorded operative and shade temperatures for comparison. Compared with operative temperature, cavities had stable microclimates, displaying smaller ranges in temperature. Mean minimum and maximum cavity temperatures differed significantly to operative temperature and between seasons, whereas there was no significant difference between shallow and deep measurements in cavities. Differences in the buffering capacities of the cavities between seasons were not significant. To determine whether T. nigricauda alter its length of exposure in response to lower ambient temperatures in winter as a means of maintaining Tb, I measured the activity of T. nigricauda, defined as the proportion of fixes outside the home cavity of the individual. Males spent a greater proportion of the active phase away from their home cavity in summer, and significantly in winter when compared with females, but there were no differences between seasons. It is suggested that T. nigricauda realize energy savings by lowering its Tb during their rest phase during the day, allowing them to maintain nocturnal activity and overall energy balance. Thus, besides the larger male home range, a result of the reproductive pattern, the physiological and behavioural traits of T. nigricauda measured in this study did not differ between aridity sites or seasons. The results of this study, in highlighting the variation in physiological and behavioural responses of subpopulations of T. nigricauda to diverse conditions, suggest that this variation is due to phenotypic flexibility. Understanding the extent and nature of this flexibility is critical to our comprehension of the consequences climate change. By defining the presence and extent of intraspecific variation in physiology and behaviour, this study resolved the necessary first step towards this understanding for the widely distributed T. nigricauda in southern Africa. / Thesis (Ph.D.)-University of KwaZulu-Natal, Pietermaritzburg, 2008.