Study Site

This study was conducted between February 2003 and August 2004 on the farm Velloor approximately 80 km south of Karasburg and 20 km north of the Orange River (28°34′S, 19°11′E, elevation 800 m) in southern Namibia. Although the farm is 40,000 ha in size, the tortoises were located and studied close to the farm house in an area approximately 500 ha in size. The general vegetation type is classified as Dwarf Shrub Savanna (Giess 1971) or Nama Karoo biome (Lovegrove 1999) and is dominated by Stipagrostis species grasses, Rhigozum trichotomum and Eriocephalus spp. shrubs, and Acacia erioloba trees along the ephemeral drainage lines. Annual precipitation occurs primarily as thundershowers during summer months with an annual average rainfall of 50–100 mm (Van der Merwe 1983; Mendelsohn et al. 2002). Average daily minimum and maximum temperatures for the coldest and hottest months range between 3°–4°C and 36°–37°C, respectively (Van der Merwe 1983; Mendelsohn et al. 2002). The dominant soils in the area are eutric Leptosols with a low water holding capacity and often subject to droughts (Mendelsohn et al. 2002).

Methods

Four female tortoises were located on 28 February 2003 by scouring the general area specifically targeting likely resting places. Radiotransmitters each weighing 10 g (2% of average body weight) were attached using glue and clear silicon to the side of the carapace so as not to affect any mating attempts. Tortoises were visited monthly (dependent on authors' availability) to determine home range with foraging observations being conducted for a few days each month during the warmer months. These transmitters were left on until 31 August 2004, when they were removed.

Tortoise movement was estimated in 2 ways. The first was during direct observations while conducting foraging observations—i.e., the total distance traveled was measured with a measuring tape (m) after the tortoise had become active until assuming a resting position. The second was by plotting the positions of the resting positions and then estimating the direct distance between resting positions with the aid of a GPS.

The home range was determined by plotting the positions of the resting places (♀1, n = 14; ♀2, n = 28; ♀3, n = 26; and ♀4, n = 16) with the aid of a GPS and then using the Minimum Convex Polygon option in the animal movement analysis ARCVIEW extension to determine the home range.

Direct observations were conducted during summer (i.e., when tortoises were most active—March, April, September, October, November, and December 2003 and January to April 2004) to determine their activity pattern. They were also visited during the winter months, but very little movement and activity was evident during this time. Tortoise activity was categorized into active (moving and foraging) versus passive (resting). The temperature was measured at the start and end of a particular activity period (during morning and afternoon observations) with a thermometer at 30-cm height in the shade.

Daily foraging data (during morning and afternoon observations) were obtained from one study animal at a time after first locating the individual before it became active and then following it throughout the day at a distance of approximately 5 m. When a tortoise was observed feeding, the plant species and number of bites per species was recorded. Binoculars (8 × 40) were used to facilitate observations. Plants utilized were immediately identified where possible or samples collected for later identification or verification. A total of 590 minutes (♀1 = 10 days, ♀2 = 22 days, ♀3 = 28 days, and ♀4 = 14 days) were spent on foraging observations.

Data on resting places (i.e., plant species or terrain features used as shelter) and resting orientation were collected throughout the year. The latter was included to determine if tortoises were selecting for a certain orientation for thermoregulatory purposes. A resting place was classified as an “open area” if the tortoise was not physically in contact with or underneath vegetation (i.e., > 10 cm away).

A rapid survey of the vegetation (species composition) was conducted within the general area utilized by the tortoises by identifying plant species at 1-m intervals along 2 × 200 m transects in the area.

Results

Four female P. tentorius verroxii were followed from sunrise to sunset on 5 consecutive days during 1–5 March 2003. During this period they moved an average distance of 81.2 m (SD = 71.57 m, range 6.5–190 m) and 91.1 m (SD = 76.44 m, range 12–169 m) during morning and afternoon observations, respectively. There was a significant difference between morning and afternoon foraging distances (one-way ANOVA, df = 9, F = 0.05, p = 0.84). The longest distances covered in a session were 190 m (morning) and 169 m (afternoon). The mean average distance (straight line) that the 4 females had moved away from their respective release sites (i.e., after fitting the transmitters) after a period of 18 months was 855 m (SD = 642 m, range 350-1750 m). The mean (n = 82) average distance that the 4 females moved between resting places was 100 m (SD = 178 m, range 0–1117 m) with no significant difference observed between individuals (one-way ANOVA, df = 2, F = 2.18, p = 0.096); although, the sample size was low.

The mean average home range as determined throughout the study period (i.e., summer months, as tortoises did not move much during the winter months) for the 4 females with radiotransmitters (n = 84 resting places) was 31.97 ha. It is certain that some movements were missed, but it was clear that the tortoises stayed within a certain area and did not venture far from this area. As a result of male tortoises not initially being found, no formal study was conducted on them, but subsequent limited data indicated a home range of 6.65 ha (SD: 4.5 ha) (n = 4) for males.

Nine tortoises observed (5 female and 4 male—including the 4 radiotracked females originally located) had a distinctive bimodal activity pattern during summer with peak foraging activity between 0730 and 0900 hours (Fig. 1). Foraging started and stopped during the morning with mean ambient temperatures of 26.2°C (SD = 2.77°C, range 22°–29°C) and 29.4°C (SD = 2.41°C, range 26°–32°C), respectively. A lesser mode of activity was observed in the afternoon between 1630 and 1900 hours (Fig. 1). Foraging started and stopped during the afternoon with mean ambient temperatures of 33.4°C (SD = 1.52°C, range 32°–36°C) and 30.2°C (SD = 1.48°C, range 28°–32°C), respectively.

View Figure

• Download as Powerpoint
• Download Figure

The 4 female tortoises spent an average 72 minutes (SD = 38.9 minutes) and 51 minutes (SD = 24.7 minutes) active during morning and afternoon observations during summer, respectively. No significant difference was observed for time spent active between morning and afternoon activity periods (one-way ANOVA, df = 2, F = 1.05, p = 0.36). Average activity levels during summer over 24 hours were 8.5% active (AM = 5% and PM = 3.5%) and 91.5% inactive. Winter activity levels were almost nonexistent except during occasional fog when tortoises became active.

A total of 18 different food items, of which 17 were plant matter (the other item eaten included weathered bone), were included in the diet of the 4 female tortoises during summer observations.

Grasses were the most utilized food item as determined by frequency of visit (51.2%) followed by herbs, bulbs, and succulents (41.9%), trees and shrubs (2.3%), other plant matter (2.3%), and bone (2.3%). However, herbs, bulbs, and succulents were the most utilized food item as determined by frequency of bite (56.9%) followed by grasses (35.2%), other plant matter (7.1%), bone (0.6%), and trees and shrubs (0.2%). The most utilized plant species as determined by frequency of visit were the grasses Stipagrostis obtusa and S. uniplumis, followed by unidentifiable herb seedlings (Table 1). The most sought after plant species determined by frequency of bite were S. obtusa, Anacampseros albissima, unidentifiable herb seedlings, and Trachyandra saltii (Table 1). The top 4 food items selected by the 4 female tortoises in terms of frequency of visit/plant and frequency of bite/plant were responsible for 74.4% and 71.3% of all the observations, respectively (Table 1).

Table 1. Food selection as determined by frequency of visit/plant and frequency of bite/plant for 4 female P. tentorius verroxii in southern Namibia. All food items were available throughout the summer.^a

View Fullscreen

Eight tortoises observed (5 female and 3 male—including the 4 females originally located) used a variety of resting sites (Table 2) and included 15 shrub species, 3 tree species, and 2 grass species. Shrubs as resting places were favored, followed by holes under trees and shrubs, and thirdly under trees alone (Table 2). Resting places used during this study indicated that 90.5% of the trees and 36.4% of the shrubs had thorns. Holes used as resting places were almost exclusively (97%) located under vegetation with thorns.

Table 2. Resting sites during summer for 5 female and 3 male P. tentorius verroxii individuals in southern Namibia.

View Fullscreen

Tortoise orientation while in a resting place varied between morning and afternoon during summer. Tortoise orientation was mostly towards the north (41%, including northeast and northwest) and east (32%, including northeast and southeast) during morning observations (Table 3 and Figure 2). Orientation during the heat of the day (i.e. after the morning foraging period) was mainly in an easterly and southerly direction (50%—including northeast and southeast). These positions were usually left during the afternoon foraging period with new resting places located during the late afternoons. Orientation during winter did not change during the day due to the inactivity of the individuals during this period, with most tortoises (60%) orientated towards the east (including northeast and southeast) and north (Table 3).

Table 3. Orientation of the posterior part of the carapace of 5 female and 4 male P. tentorius verroxii individuals in southern Namibia while resting.

View Fullscreen

View Figure

• Download as Powerpoint
• Download Figure

The vegetation in the general area utilized by the tortoises is dominated by perennial Stipagrostis grasses (S. obtusa, 30% and S. uniplumis, 12%) with only 6% of the vegetation comprised of thorny species (Rhigozum trichotomum, 3.5%; Acacia mellifera, 2%, and Protasparagus ssp., 0.5%).

Discussion

Female P. tentorius verroxii were sedentary, as indicated by the distances covered while foraging, distance from release site after 18 months, distance moved between resting places, and home range size. Foraging also does not occur on a daily basis during the summer and is possibly influenced by food availability, food consumed the previous day, and local environmental conditions.

The activity pattern of P. tentorius verroxii was typically bimodal and affected by ambient temperature. Ramsay et al. (2002) indicated that Chersina angulata, a tortoise common in the southwestern Cape in South Africa, followed a bimodal activity pattern as an adaptation to high ambient temperatures, an activity that would assist the animal in maintaining its water balance. Lambert (1981) stated that the activity mode of Testudo graeca became bimodal when ambient temperatures rose above 28°C. Similarly, Kinixys spekii is inactive when ambient temperatures exceed 29°C (Hailey and Coulson 1996). This bimodal activity would thus be even more of an advantage to P. tentorius verroxii from southern Namibia due to the general higher average ambient temperatures experienced there. Female P. tentorius verroxii stop foraging and seek shade when mean ambient temperatures reach 29.4°C (SD = 2.41°C) during the morning. This is slightly higher than documented for C. angulata from the southwestern Cape (28.7°C) (Ramsay et al. 2002) and in the eastern Cape (28.5°C) (Els 1989). Afternoon foraging usually ceased around sunset; although, some individuals were observed still being active after sunset. This probably occurs when extreme daytime temperatures prevent tortoises from foraging before sunset.

During winter the 4 female P. tentorius verroxii were not very active with the exception being during fog. On one occasion when fog occurred during winter (9 August 2003) all 4 individuals were observed away from cover in the open with a lot of local activity observed as indicated by their tracks. This was possibly due to the tortoises “fog collecting” as described by Branch (1998) and Boycott and Bourquin (2000). Although fog incidences are relatively infrequent in southern Namibia (5–10 days annually) (Mendelsohn et al. 2002) the above-mentioned activity indicates that P. tentorius verroxii may actively utilize moisture from fog when available.

Holes were not utilized as hiding places by the 4 females under observation during winter as compared to 23.1% (n = 31) of the total observations as resting places during summer. Holes are possibly not as important for thermoregulation during winter as during hot summer days.

This study indicates a variety of plants being utilized by female tortoises, including the importance of grasses in their diet during periods of below-average rainfall. Rall and Fairall (1993) confirmed the tendency to utilize grasses during unfavorable periods for P. oculifer, a similar-sized species also inhabiting dry areas in southern Africa. Stipagrostis obtusa is the most important plant species included in the diet of the 4 female P. tentorius verroxii as observed by frequency/visit and frequency/bite. This small perennial climax grass is one of the best grazing grasses for livestock in the arid regions of southern Africa (Van Oudtshoorn 1999) and has a relatively high nutritive value. It remains relatively palatable and well utilized by animals even when dry (Müller 1984). Herbs, bulbs, and succulents are sought after as observed by frequency/bite. The succulent Anacampseros albissima (known as “skilpadkos”—tortoise food in Afrikaans) and Trachyandra saltii were the second and third most sought after plant species as observed by frequency/bite, although not very common throughout the area. These plants are usually consumed entirely when encountered. It is suggested that P. tentorius verroxii may feed more on annuals (e.g., forbs) when encountered, but due to low rainfall throughout the study area/period, annuals were sparse, and they thus had to utilize the available perennial grasses more.

Tortoises utilized a variety of plants as shelters when resting; although, thorny plants were used more often (36% of shrubs and 84% of trees used as resting places are thorny), which probably assists with defense and or thermoregulation as most thorny plants also present better shade (denser growth form, bigger leaves) during summer. Furthermore, thorny plants make up a small percentage (6%) of the vegetation in the general area utilized by the tortoises, suggesting that these plants are selected. Holes under bushes (23%) were also used, which would possibly further assist with defense and thermoregulation. The use of holes during this study was only observed during summer during an exceptionally warm period with daytime ambient temperatures reaching 40°C. Branch (1998) stated that tortoises are known to burrow into sandy soil at the base of low shrubs emerging after the onset of rains. The use of resting places (i.e., site selection) and holes as resting places (i.e., to what extent are these utilized, do they dig these themselves or utilize existing holes) should be investigated further.

Orientation of the carapace during summer was mainly in a northerly and easterly direction during morning observations, which could suggest that P. tentorius verroxii position themselves when going into a resting site the previous evening. This is probably to enable them to reach an optimum temperature quickly before the heat of the day to commence foraging while still relatively cool in the morning. Orientation away from the sun (i.e., easterly and southerly direction) during the heat of the day is probably to avoid overheating during this period. Orientation of the carapace during winter towards the north and east (70%) is probably similar to that as suggested during summer.

Although P. tentorius verroxii inhabits an extremely marginal habitat in southern Namibia, the species seem well adapted to handle the adverse environmental conditions; although, the effect that climate change (aridification) may have is unknown. However, the relatively sedentary lifestyle, slow reproductive rate (Branch 1998; Boycott and Bourquin 2000; Cunningham et al. 2004), and anthropogenic factors (e.g., road kills and habitat alteration by livestock overgrazing) makes the species susceptible to local extinctions. Illegal collecting and/or indiscriminate utilization by humans for food exacerbates its fate and are thus condemned.