user: pass:

Knight, M.H.; Hitchins, P.; Erb, K.P., 1993. An aerial survey of rhino and elephant in Chobe National Park and surrounding areas, Northern Botswana, September 1992. Pachyderm 17: 64-74, figs. 1-4, tables 1-2

Location: Africa - Southern Africa - Botswana
Subject: Ecology - Census Methods
Species: Black Rhino

Original text on this topic:
Flying and counting procedure. The principle aim of the survey was to locate black rhinos. The survey was undertaken in the late dry season (September) when the visibility is least impaired by the vegetation. The selection of areas most likely to have black rhinos was based upon past sighting records of animals or their tracks (M. Slowgrove, L. Wilmot, G. Calef, M. van der Waller and D. Joubert pers. comm.), the availability of dry season drinking water supplies, and the absence of human settlements. Aerial coverage did not extend beyond 30 km from waterholes, the maximum foraging distance black rhinos were noted to forage water in the dry season (Joubert & Eloff, 1971).
In order to maximize the chances of detecting black rhinos, it was decided to use two aircraft simultaneously to scan along each transect. Two Cessna 210 aircraft were used in the survey, with the flying formation consisting of the leading aircraft flying at low level (70 m above ground level [AGL]). The second aircraft flew at 140 m AGL and maintained a track approximately 300 in to the right of the leading aircraft and trailed by 600-700 m (or a 10 sec separation). Both aircraft maintained a ground speed of approximately 90 knots. It was intended that the lower flying lead aircraft would flush any black rhino for the trailing aircraft to detect.
The leading aircraft was manned by a pilot, navigator, and three or four observers, two of whom were always seated on the right-hand side of the aircraft looking into the track of the trailing aircraft. The leading aircraft's objectives were to navigate the predetermined transects for both aircraft, flush any rhinos potentially obscured by vegetation or directly under the aircraft for the second aircraft to detect, as well as observe animals (both alive and dead), All navigation was done with the aid of a Global Positioning System (Garmin 100 ADV). The objective of the second aircraft with pilot, observer recorder (seated directly behind the pilot), the three or four observers (one of whom was always seated on the left-hand side behind the recorder) was to record all sightings of rhinos, noting their positions on a separate GPS system. The positioning of the second aircraft was designed to cover the obscured zone beneath the leading aircraft. The flying formation thus allowed four observers to scan the 300 m between the two aircraft.
As the estimation of animal densities was not a specific task of the survey, all black and white rhinos and elephants seen from the second aircraft within the unbounded transects were recorded. Thus the survey could be described as a low intensity total count. The maximum detection range for large conspicuous species such as elephants in the open, leaf- less savanna was estimated to be about 900 m on either side of the second aircraft. This was estimated on the ground by measuring the distance to recognizable features during trial runs. As transects were 2 km apart, the chances of double counting were therefore considered remote.
The survey area of 15057 km? was subdivided into 18 strata, ranging in size from 547 to 1452 km? (Figure IB). Transects were orientated in an east to west direction and the strata were flown in sequential order from cast to west. On spotting a rhino, the second aircraft would maintain its altitude and circle the position while guiding the lower aircraft in to determine the sex and age of the animal and take photographs.
Visibility biases
Although the counts were not within defined strips, an indication of potential visibility biases between observers, particularly with respect to counting elephants, was determined using the Petersen mark-recapture method (Seber, 1982), as modified by Marsh & Sinclair (1989). This entailed the observer/ recorder of the second aircraft noting what elephant groups he detected (SF), what groups the second observer directly behind the recorder saw (Sr) and what they both observed (B). A perception correction factor (C) for groups of elephants per stratum was calculated by: C=((Sf+B) (Sr+B))/(B(Sf+Sr+B)
with the coefficient of variation (Cp) calculated by:
Cp=((Sf+Sr)/(Sf+Sr+B))* ((Sf*Sr) (B((Sf+B)(Sr+B)))).
An elephant herd was defined as any aggregation showing some form of cohesion that was separated from other groups or individuals by a clear break that was greater than the diamiter of the herd in question.
The coordinates were recorded for every rhino and elephant carcass seen by either of the two aircraft. Communications between the two aircraft reduced the chances of double counting carcasses. The carcasses were categorised as:
1 'Old': white scattered and bleached bones.
2. 'Fresh': skin covering the skeleton. With a closer inspection from the air, the age was more precisely estimated in either weeks or months depending upon: the presence of scavengers (vultures, jackals and spotted hyenas); signs of blood or body fluids around the carcass; the degree of bloat and the open patch around the carcass through trampling and the concentration of body fluids. Where possible, the presence or absence of ivory, or horns in the case of rhinos, was also noted.
A carcass ration, calculated as a percentage of the combined total number of carcasses and the alive animals counted per strata was calculated. As no correction factor for the under-counting of elephant carcasses was estimated, the three times correction suggested by Dublin & Douglas-Hamilton (1987) was used.
Distribution of pans
All pans with a diameter greater than 50 m across their flat unvegetated surface were noted for the presence of drinking water. The coordinates of those holding water were recorded on the GIS system.
Tree damage
An estimate of 'tree damage' (defined as felled trees only), assumed to have been caused by elephant, was determined during the latter half of the survey. A roughly five minute scan interval,was used in which the proportion of felled trees were categorised into one of four subjective categories: nil (0- 10 %), light (11-25 %), moderate (26-50 %), and heavy (>50 %), depending upon the percentage of felled trees to those standing. A tree was defined as being greater than 5 m in height.

[ Home ][ Literature ][ Rhino Images ][ Rhino Forums ][ Rhino Species ][ Links ][ About V2.0]