Josephine Smit, who studies elephant behavior as a researcher with the Southern Tanzania Elephant Program, says that among the female elephants she tracks at Ruaha National Park, an area that was heavily poached in the 1970s and 1980s, 21 percent of females older than five are tuskless. As in Gorongosa, the numbers are highest among older females. About 35 percent of females older than 25 are tuskless, she says. And among elephants ages five to 25, 13 percent of females are tuskless. (Smit, a doctoral candidate at the University of Stirling, in Scotland, says the data have not yet been published, though she presented the findings at a scientific wildlife conference last December.)
Despite the wave of human-influenced tusklessness in recent decades, elephants missing their tusks are surviving and appear healthy, according to Poole. Scientists say that the significant proportion of elephants with this handicap may be altering how individuals and their broader communities behave, and they want to find out if, for example, these animals have larger home ranges than other elephants because they might need to cover more ground to find recoverable foods.
Video: elephants making more elephants in Addo
Nowadays the park is a sanctuary to am opulent of wildlife and superabundant birdlife, including more than 600 elephants, 400 Cape buffaloes, zebras, spotted hyenas, a variety of antelope species and many other animals including the unique flightless dung beetle.
I first went to Addo in 2003 after photographing a rugby Test in what was then known as Port Elizabeth. I added an extra day to my journey to go to Addo. From then on, whenever I had a photo shoot in Gqeberha, I added at least three days so that I could go to Addo. Now, living so close to it, I regularly photograph elephants and other wildlife in the national park.
My picture of the two aggressive bulls was taken while I was sitting in the passenger seat, shooting a rather funny standoff between some elephants and zebras wanting to drink at the same waterhole. For at least an hour, various elephants stopped about 25 zebras from getting close enough to drink until they finally let them quench their thirst.
People need to use common sense. If you see a huddle within a herd walking along, the chances are that there is a baby in the group. The elephants can be on high alert and become protective, even aggressive. You need to be careful and show respect.
Sometimes you can be watching several elephants play in the water when suddenly the matriarch (the female head of the family/herd) orders the herd to leave. A massive rush ensues as they obey and race to get out of the dam.
Despite their size, elephants can run up to 40km/h. It is incredible to watch a herd that is within about 75m of a waterhole or dam on a really hot day. Suddenly, some members (often the younger ones) start sprinting to the water. It is also amazing to watch what they can do with their trunks. With more than 40,000 muscles in their trunks, elephants can lift extremely heavy objects as well as a single grain of rice.
Different attempts towards the automatic detection of elephants have been made in the past. Techniques such as satellite tracking (efficient but also invasive) using global positioning system (GPS) [7], or light sensors with laser beams [8], and systems based on vibration sensors in the ground [9] are cost-intensive and might not scale to large populations. Today, non-invasive techniques such as acoustic and visual monitoring represent promising low-cost alternatives to sample populations and to obtain reliable estimates of species presence and, potentially, abundance [10, 11]. To approach the challenges of non-invasive monitoring, biologists and computer scientists have joined forces in an interdisciplinary research project to establish the scientific foundations for a future non-invasive early warning and monitoring system for elephants.
Elephants are the largest terrestrial herbivores and have a well-distinguishable visual appearance. Thus, visual information may be useful for the automated detection of elephants. In this project we investigated the suitability of visual cues for the automatic detection of elephants in wildlife video recordings and developed a first visual detection algorithm.
Overview of the envisioned elephant early warning and monitoring system. The first step of analysis is the detection of elephants either visually through video and thermal cameras or acoustically through a microphone (array). Different detection mechanisms may be combined in a multimodal approach. Automatically recognized detections can be directly input to an early warning systems or serve as a basis for higher-level tasks. The analysis layer contains the most important higher-level tasks in our context. Highlighted tasks have been investigated and automated in the course of the project
We annotated 2199 vocalizations (of different call types) from free-ranging elephants at the Addo Elephant National Park, and 681 vocalizations recorded at Bela Bela. 633 of these vocalizations were rumbles (see Table 1 for a detailed overview on the number of calls recorded for each call type during each field session). The complete and precise annotation of the recorded data formed the basis for objectively evaluating the automatic analysis methods developed in the project.
Similarly, for the evaluation of visual detectors it was necessary to compile a reference dataset from the collected video material. We selected a representative subset of the entire video collection to speed up subsequent annotation and further processing. During selection we rejected sequences that were too similar to already selected ones to increase the heterogeneity in the dataset. The selected sequences contained elephants (groups and individual elephants) of different sizes from two distance categories (far distance and near distance). Elephants were visible in arbitrary poses and ages, performing different activities such as eating, drinking, running, and different bonding behaviors. The sequences showed different locations, such as elephants at a water hole, elephants passing a trail, and highly occluded elephants in bushes. Selected sequences reflected different times of the day and showed different lighting and weather conditions. Furthermore, recording settings varied across the sequences (e.g. from almost static camera mounted on a tripod to shaking handheld camera). Additionally, the subset contained sequences with no elephants at all and sequences where elephants entered and left the scene. The heterogeneous settings of the selected scenes were necessary to reflect the broad variability of real-world conditions to enable the development of robust visual detectors that can cope with a variety of different situations.
To date, only little research has been performed on the automated detection of elephants. Venter and Hanekom [28] detected elephants based on their rumbles in wildlife recordings by extracting the characteristic fundamental frequency of rumbles using sub-band pitch estimation. They defined those audio segments as rumbles in which a pitch (in the typical frequency range of rumbles) can be tracked robustly for a certain amount of time. A similar approach also relying on pitch extraction has recently been proposed by Prince and Sugumar [40] for elephant detection. Wijayakulasooriya [27] proposed an alternative rumble detection method that employs the shape of formant frequency tracks as a clue. The basic assumption of the latter method is that the first and second formants are nearly stationary during a rumble. Thus, detecting audio sequences with stationary formant frequencies in the frequency range of rumbles should provide clues for their acoustic detection.
The visual detection of animals such as elephants without a distinct skin texture poses additional challenges to visual analysis. Attributes such as color, shape, and motion need to be exploited, but plants and trees often occlude elephants, revealing only body parts rather than the whole elephant. Additionally, elephants appear in different postures and sizes, as well as in groups and as individuals. Thus the typical shape of an elephant as well as its size are generally not useful clues for their visual detection. Similarly, motion is a questionable visual clue for detection, as elephants often move slowly or rest for lengthier times.
Figure 6 shows qualitative results for both scenarios. Figure 6a demonstrates that elephants of different sizes can be detected using the same detection method. The depicted frame shows three adults and one calf. The adults and the calf are detected successfully. The two adults on the right side are reported as one detection by our method. This is due to the large overlap of the two individuals. We further observe that even elephants that are highly occluded by vegetation as shown in Fig. 6b can be detected successfully. This scene further demonstrates why shape is in many situations not a useful clue for detection, as only individual body parts of the elephants are visible.
The results of our quantitative evaluation on the entire video dataset are summarized in Table 3. For elephants at near distances we obtained a detection rate of 91.7 % at a false positive rate of only 2.5 %. This means that most elephants are detected successfully (only 8.3 % are missed) and the rate of false detections is quite low (only every 40th detection is actually not an elephant).
With larger distance the detection becomes increasingly difficult. The small area covered by the elephants in the image plane requires a much finer image segmentation of the input images. This results in a stronger over-segmentation of the image and thus a large number of small image patches. The small patches exhibit fewer distinctive visual features useful for automated detection than larger segments, which impedes automated detection and results in more detection errors for far-distant elephants. Our experiments clearly show the effect of the smaller object size on the detection rate and false positive rate (Table 3). 2ff7e9595c
Comments