Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 7, 11034

Energy Cost and Return for Hunting in African Wild Dogs and Cheetahs

Affiliations

Energy Cost and Return for Hunting in African Wild Dogs and Cheetahs

Tatjana Y Hubel et al. Nat Commun.

Abstract

African wild dogs (Lycaon pictus) are reported to hunt with energetically costly long chase distances. We used high-resolution GPS and inertial technology to record 1,119 high-speed chases of all members of a pack of six adult African wild dogs in northern Botswana. Dogs performed multiple short, high-speed, mostly unsuccessful chases to capture prey, while cheetahs (Acinonyx jubatus) undertook even shorter, higher-speed hunts. We used an energy balance model to show that the energy return from group hunting and feeding substantially outweighs the cost of multiple short chases, which indicates that African wild dogs are more energetically robust than previously believed. Comparison with cheetah illustrates the trade-off between sheer athleticism and high individual kill rate characteristic of cheetahs, and the energetic robustness of frequent opportunistic group hunting and feeding by African wild dogs.

Figures

Figure 1
Figure 1. Example day's speed and track for an African wild dog.
(a) GPS data trace for one individual, dog position at the beginning of the day (0/0), GPS position colour coded based on speed bins, black arrow shows the direction of the movement, inset shows range covered by dogs during 5 months of observation and trace of example day. (b) Speed profile at local time for one individual colour coded based on speed bins (legend in a). (a,b) Movement is interspersed by six fast chases (circles 1–5 and 7). The GPS sample rate varies throughout the day and is set to capture high-resolution data when the animal is moving. When the animal is stationary, the sampling interval is once an hour (circles 6 and 9). Between 18:00 and 20:00, it is at 10-s intervals (when moving, circle 8), when high-speed chases are detected, it is increased to 5 Hz (circles 1–5 and 7). At all other times, it is once every 5 min when moving.
Figure 2
Figure 2. Detailed analysis of African wild dog movement.
Histogram of distance travelled per day including kernel density estimate for (a) 18 dogs in 13 other packs in the area (n=2,226 dog days) and (b) for the focal pack (n=708 dog days). Abscissa is cutoff at 98th percentile. The kernel density estimate for other packs (blue line) is displayed alongside the estimate for the focal pack (red line) in b after being normalized to fit the focal pack sample size. (c) Start time for all recorded chases at local time (n=1,119). (d) Distance travelled per day in kilometres by example dog ‘MJ' colour coded into different speed ranges (walk: 0–1.5 ms−1 (red), trot: 1.5–3 ms−1 (green), slow gallop: 3–6 ms−1 (yellow), fast gallop: >6 ms−1 (blue)). (e) Total distance covered in each chase (n=1,119). (f) Top stride speed recorded in each chase (n=1,119).
Figure 3
Figure 3. Hunt parameters in African wild dogs.
(a) Histogram of hunt distance covered from end of a chase to end of next chase by the same individual (the distance of 1,870 m used in the energetic calculation is the mean of the individual dog median values, n=1,119). (b) Histogram of time elapsed between end of one chase and end of next chase by the same individual (n=1,119). (c) Histogram of distance covered by individual dogs in 5 min after the end of each of their chases (n=1,119, bin width 25 m).
Figure 4
Figure 4. Preferred speed and tortuosity estimates based on instantaneous velocity measures.
(a) Histogram of instantaneous velocity measures during the daily 2 h of 10 s GPS fixes. (b) Differentiation derived velocity versus instantaneous velocity for walking and trotting (10 s GPS fixes, blue, slope 0.84) and running (5 Hz, red, slope 0.47), averaged for 30 s windows. The lower ratio of differentiated to instantaneous velocity for running (5 Hz) shows the higher tortuosity during chases.
Figure 5
Figure 5. Comparison of African wild dog and cheetah hunt parameters.
Histogram and kernel density estimate for cheetah (ac; number of cheetahs N=5; number of hunts n=381) and African wild dog (df; number of dogs N=6; number of hunts n=1,119; including cheetah kernel density estimate normalized to African wild dog count (blue)). Distance covered 5 min before the start of chase (a,d) between 5 and 10 min before chase (b,e) and 10 min before chase (c,f).
Figure 6
Figure 6. Comparison of African wild dog and cheetah stride parameters.
Histogram and kernel density estimate for cheetah (ac; number of cheetahs N=5; number of strides n=15,845) and African wild dog (df; number of dogs N=6; number of strides n=140,141; including cheetah kernel density estimate normalized to African wild dog count (blue)). Horizontal stride speed (a,d); centripetal acceleration (b,e); tangential acceleration (c,f). African wild dog chases contain about twice as many strides as cheetah chases with a considerably greater number of straight strides at steady speed.
Figure 7
Figure 7. Comparison of solitary versus group hunting scenarios.
Solitary hunter (a) and multiple hunter (b) scenarios assuming a fictional individual kill rate of 0.2. On average, five hunts have to be conducted for one kill. Those hunts can be conducted consecutively by one hunter or simultaneously by five hunters. The kill distance is the sum of all hunt distances and is equal in both scenarios. However, costs (proportional to distance travelled) for individual hunters and for followers is lower for the multiple hunter scenario (b).
Figure 8
Figure 8. Energy model outcome for African wild dogs.
(a) Maximum pack energy gain per kill (EPK_In) divided by pack energy expenditure per kill (EPK_Out) for different pack sizes and compositions (40 kg impala, 3.5 kg dog stomach capacity). (b) Maximum pack energy gain per kill (EPK_In) divided by total daily pack energy expenditure including one kill (EP_total) for different pack sizes and compositions (sustainability line at 1). (c) Number of kills necessary to fulfil daily energy requirement of pack (basal metabolic cost+ranging cost+x times kill cost). (d) Total daily hunting time based on number of kills required; based on kill rate and time between hunts (22.58 min), red line indicated 3 h daily hunting boundary.
Figure 9
Figure 9. Influence of different model parameters on model outcome.
(a) Effect of kleptoparasitism and COT error analysis. Boundary on sustainable pack composition assuming a 180 min limit on daily hunting time. Red lines indicate boundary (pack composition to the right of the line is sustainable) for different percentages of prey stolen (40 kg, impala). Blue lines indicate sustainability for higher COT (COTLow × 2.5; COTHigh × 1.5) based on ref. . (b) Sustainability boundary for 180 min daily hunting time limit calculated for different COT, prey size and abundance. Lower prey density simulated using the 3rd quartile for hunt distance (3.38 km) and hunt duration (45.3 min).

Similar articles

See all similar articles

Cited by 9 PubMed Central articles

See all "Cited by" articles

References

    1. Cresswell W. Surprise as a winter hunting strategy in Sparrowhawks Accipiter nisus, Peregrines Falco peregrinus and Merlins F. columbarius. Ibis 138, 684–692 (1996).
    1. Domenici P. The scaling of locomotor performance in predator–prey encounters: from fish to killer whales. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 131, 169–182 (2001). - PubMed
    1. Hedenström A. & Rosén M. Predator versus prey: on aerial hunting and escape strategies in birds. Behav. Ecol. 12, 150–156 (2001).
    1. Wilson A. M. et al. . Locomotion dynamics of hunting in wild cheetahs. Nature 498, 185–189 (2013). - PubMed
    1. Bateson P. & Bradshaw E. L. Physiological effects of hunting red deer (Cervus elaphus). Proc. Biol. Sci. 264, 1707–1714 (1997). - PMC - PubMed

Publication types

Feedback