NG/26 Wildlife Management Area Okavango Delta Botswana

Transcription

1 NG/26 Wildlife Management Area Okavango Delta Botswana Aerial Wildlife Survey September 2009 Petri Viljoen Environmental & Wildlife Consultant PO Box 3124 White River 1240 South Africa Cell: Tel: Fax:

2 2 Introduction Botswana s Okavango Delta, with Moremi Game Reserve and its surrounding concession areas, forms a large protected area exceeding 68,640 km 2 1. The Okavango Delta is a hydrologically unique region which supports an important, well-established tourist industry and many local communities. It is also a biologically diverse area (Ramberg et al. 2006) with a wide range of animal and plant species. A general aerial wildlife survey was undertaken in the NG/26 Wildlife Management Area (WMA) between 28 and 30 September The purpose of this survey was to provide information on the dry-season distribution and abundance of wildlife species. Survey Area The area surveyed (2,048 km 2 ) is situated between E22 o and S19 o 18 53, and E22 o and S19 o (Appendix I). This includes the approximate 1,736 km 2 NG/26 wildlife management area (WMA). NG/26 lies in the upper-central part of the Okavango Delta, south of where the Okavango River divides into main channels that flow into the rest of the Delta. The Boro River flows to the east of NG/26 while the Thaoge River is to the west. NG/26 is bordered by NG/25 in the north, NG/8 in the west and NG/29 in the south. Moremi Game Reserve (NG/28) together with NG/27A forms the eastern boundary (Appendix I). The survey area is entirely in the Zambezian Flooded Savanna Ecoregion (WWF 2004). Vegetation can be divided into three main types: the perennial swamp area, seasonally inundated floodplain grassland, and drier vegetation in the south (Bekker & de Wit 1990). A total of six basic vegetation types have been identified for the survey area, primarily based on satellite imagery 2 : Dry Floodplain and Island Interiors, Swamp Vegetation, Shrubbed Grassland on Former Floodplains, Shrubbed Woodland of Riparian Zones, Tree Shrubland with Acacia and Shrubbed Woodland with Mixed Mopane. A sandveld tongue stretches into NG/26 from the south-eastern boundary. 1 The Ramsar Convention Secretariat, Gland, Switzerland. 2 HOORC, Maun, Botswana.

3 3 The area is known to undergo significant changes between seasons, and from year to year as a result of flooding variations (Kays pers comm.). The flood peak in this part of the Okavango Delta is usually between about April to August. NG/26 is used for photographic safaris. Abu Elephant Camp and Macatoo Camp are operated as tourist camps in the area. Earlier countrywide, low sample intensity aerial surveys 3, have recorded a number of ungulates species in, and in the immediate vicinity of, NG/26. These were African elephant Loxodonta africana, common zebra Equus quagga, hippopotamus Hippopotamus amphibious, warthog Phacochoerus aethiopicus, giraffe Giraffa camelopardalis, African buffalo Syncerus caffer, sitatunga Tragelaphus spekei, greater kudu Tragelaphus strepsiceros, eland Taurotragus oryx, bush duiker Sylvicapra grimmia, steenbuck Raphicerus campestris, southern reedbuck Redunca arundinum, lechwe Kobus lechwe, waterbuck Kobus ellipsiprymnus, impala Aepyceros melampus, tsessebe Damaliscus lunatus and wildebeest Connochaetus tarinus 4. Methodology Survey methodology was based on a standardised aerial wildlife sampling approach (Gwynne & Croze 1981; ILCA 1981; Norton-Griffiths 1978; TWCM/FZS 1996). This methodology, usually referred to as SRF (Systematic Reconnaissance Flight), has been the basis of the majority of aerial wildlife surveys in many regions on the continent. The survey area boundaries and survey flight path positions were determined on a GIS 5 prior to the survey and the data were then transferred to the survey GPS unit 6. A single-engine Cessna 206 aircraft was used for the survey. The survey team consisted of a pilot, FSO (front-seat observer/data recorder/survey co-ordinator) and two rear-seat observers (LSO = left-seat observer; RSO = right-seat observer). Temporary wing-strut markers were fitted on the aircraft and calibrated with ground markers along the airfield at Tubu Tree Camp during a calibration flight. Calibration methodology was based on Norton-Griffiths (1978) and TWCM/FZS (1996). The calibration was done to allow for a targeted 150 m sample strip on each side of the aircraft when flown at 106 m (350 ft) above ground level 7. The positions of the wing- 3 Unpublished data DWNP, Gaborone, Botswana. 4 Naming follows Kingdon ArcView, ESRI, Redlands: USA. 6 Garmin GPS V, Garmin Ltd. 7 The aircraft was equipped with a radar altimeter.

4 strut markers were adjusted for each observer prior to the survey as part of the calibration exercise. A total of 36 north/south parallel survey strips were flown to sample the total area of approximately 2,048.7 km 2 (Appendix III & IV). Planned strip separation was 1.4 km, giving an expected sample of ±21.0%. The lengths of the 36 survey strips varied between 5.4 and 57.3 km, with a total length of 1,469.5 km (Appendix II). The calculated area sample was 22.7%. 4 Mobile GIS 8 on a pocket PC 9 with a Bluetooth GPS 10 was used with preloaded survey sample strip and survey area boundary to ensure accurate in-flight navigation. This system was also used to capture tracklog data at 2-second intervals. A second Pocket PC and Bluetooth GPS with SurveyCapture software 11 ensured real-time data capture of all observations, including height-above-ground data. Oblique, geotagged 12 digital 13 aerial photos were taken of all large buffalo and elephant herds. The average flying speed (ground speed) during the survey was km/hr (99.6 knots). Four survey sessions were flown over three days. Survey session duration varied between 01:40 and 03:09 hrs and the total survey time was 09:24 hrs. Weather conditions were favourable during all survey sessions. Wind velocity increased towards the end of second day s first session. No survey flights were conducted that afternoon because of the windy conditions. Tubu Tree Camp s airfield (Hunda Airfield), in the south-western part of NG/25, was used as a base for all operations (Appendix I). All large mammal species and human activities inside the survey strips were recorded. Observations of larger bird species and crocodiles Crocodylus niloticus were also included. GIS was used to map animal distribution data obtained during the survey for all the wildlife species recorded. A spreadsheet analysis of the aerial sample data was done using Jolly s Method for unequal-sized sampling units (Norton-Griffths 1978). Geotagged photos of large herds (elephant and buffalo) were mapped 14 and matched with recorded observations in GIS to allow the correction of aerial estimates. Elephant and buffalo counted both inside and outside the sample strips were combined. The totals for these two species were therefore treated as total area count (TAC) data. 8 ArcPad, ESRI, Redlands: USA. 9 ipaq hx2950, Hewlett-Packard Company, Palo Alto: USA. 10 Holux Slim240, Holux Technology Inc, Hsinchu, Taiwan. 11 SurveyCapture, Petri Viljoen/Hawk Systems. 12 Blue2CAN GPS interface, Red Hen Systems, USA. 13 Nikon D300 dslr, Nikon Corporation. 14 MediaMapper software, Red Hen Systems, USA.

5 5 Results A total of 20 wildlife species were recorded in the survey area and the total number of wildlife observations was 651 (Appendix V). These were elephant, zebra, hippopotamus, warthog, giraffe, buffalo, greater kudu, southern reedbuck, lechwe, impala, wildebeest, gemsbuck Oryx gazella and ostrich Struthio camelus. Lions Panthera leo and crocodiles were also seen during the survey. Other bird species recorded included Wattled Crane Grus carunculata and Ground Hornbill Bucorvus leadbeateri. The gemsbuck sighting was of a single animal on the edge of the sandveld tongue in the south-eastern part of the survey area (Appendix I). No sitatunga or waterbuck were recorded within the actual survey flight strips, although there were two sitatunga sightings (in the north-eastern area) during ferry flights. In addition to wildlife species, human activities and livestock were also seen (Appendix V). These included cattle, donkeys and horses. Survey results are summarised in Table 1 and Appendix V. The sample sizes for some species were considered too small for population estimates (warthog, reedbuck, wildebeest and ostrich). Population estimates were calculated for zebra (901), giraffe (328) kudu (573), lechwe (1,618), impala (3,092) and tsessebe (307). The total number of elephant and buffalo counted were 1,351 and 2,562 respectively. Table 1. NG/26 results: population estimates (NG/26 aerial survey Sep 2009). Zebra Giraffe Kudu Lechwe Impala Tsessebe Σy Σz² or Σy² Σz.y Ř = Σy/Σz sy² sz² szy Ŷ = Ř.z (population estimate) ,618 3, Var(Ŷ) SE(Ŷ) t.se(ŷ) % of Ŷ (95% confidence limits of Ŷ) The total number of Wattled Cranes, Saddlebilled Storks and Ground Hornbills seen were 42, 18 and 10 respectively. Two of the Wattled Cranes and three Saddebilled Storks counted were outside the sample strips.

6 A total of 421 cattle in 24 herds were counted in the adjacent NG/8 just outside NG/26 s boundary. Twelve horses and 22 donkeys were also recorded in this zone. 6 The largest group of elephant counted was 83, and buffalo 512 (Table 2). An average group size of 6.5 has been calculated for the former, as the majority of this species was seen in relatively small groups (Fig. 1). The average group size for buffalo was 36.6, with most of the animals seen in groups of 10 or fewer animals (Fig. 2). Table 2. Ungulate group sizes recorded (NG/26 aerial survey Sep 2009). Species Min Max Average SD N Elephant Zebra Hippopotamus Warthog Giraffe Buffalo Greater kudu Southern reedbuck Lechwe Impala Tsessebe Wildebeest Gemsbuck Fig. 1. Elephant group sizes (NG/26 aerial survey Sep 2009) >50

7 7 Fig. 2. Buffalo group sizes (NG/26 aerial survey Sep 2009) < >200 Distribution maps for the various wildlife species recorded appear in Appendix VI. Many of the ungulate species were dispersed throughout most of the survey area with the exception of the sandveld tongue region (Appendix I). The larger elephant herds were mostly in the southern region, while the highest buffalo densities were in the south-western and south-eastern parts. The majority of zebra also occurred in these regions. Reedbuck and Lechwe were confined mostly to the floodplain areas in the north-east. This is also were almost all the Wattled Crane and Saddle-billed Storks were seen. Almost all the ungulates recorded in the sandveld tongue region were elephant, apart from some of the buffalo herds and a small number of zebra. Almost all the ostriches recorded were seen in the western section. All the cattle, horses and donkeys recorded (Appendix VII) were outside the NG/26, but within the narrow zone around NG/26 which was included as part of the survey area. The majority of the cattle herds were in the north-western part (NG/8; Appendix I).

8 8 Recommendations It is recommended that consideration be given for future surveys also to be done during the late dry season to compare possible numerical and spatial changes of the various wildlife species. Early September would be ideal for surveys, as the weather then generally tends to be more favourable for aerial surveys than at the end of the dry season. If feasible, annual counts - or counts at least every second or third year - would be preferable to obtain trend data. It is also highly recommended that the survey methodology be standardised as far as possible to ensure that comparative results are obtained. Additional wildlife surveys during other times of the year (at the end of the rainy season) would provide additional distribution data for more mobile species such as buffalo (which were absent during this survey), elephant and zebra. It is also recommended that consideration be given to a dedicated TAC, specifically to survey elephant and buffalo (Norton-Griffith 1978; TWCM/FZS 1996). Sample count methodology is not ideally suited to survey these two species, particularly when relatively high densities are present.

9 9 Acknowledgements I am grateful to Cathy and David Kays of Ngamiland Adventure Safaris, and Martin Kays (Tubu Tree Lodge) who provided support in numerous ways. Thanks also to David Kays who very kindly made his aircraft headsets and GPS available. Chris Roche of Wilderness Safaris is thanked for his enthusiastic support and assistance during the planning phase. The survey team members are thanked for their very valuable contributions: Graham Skinner from Sefofane Air Charters piloted the aircraft, and Martin Kays and Kate Evans (Elephants for Africa) who were the observers. Sefofane Air Charters also made the survey aircraft available. The survey team is grateful to Justin and Jackey Stevens and other Tubu Tree Camp staff for their hospitality and support during our stay at the camp. Tubu Tree Camp is also thanked for the use of their Hunda Airfield for the survey. The management of Elephant Back Safaris / NG/26 is also thanked for their enthusiasm for the survey.

10 10 References Literature Consulted and Cited. Bekker, R.P. & de Wit, P.V Vegetation Map of the Republic of Botswana. Soil mapping and Advisory Services Project AG:DP/BOT/85/011. FAO, Ministry of Agriculture, Gaborone. De Wit, P. & Nachtergaele, F Soil Map of the Republic of Botswana. Soil Mapping and Advisory Services Project FAO/BOT/85/011. FAO. East, R African Antelope Database IUCN SSC Occasional Paper No. 21. Gland: IUCN / SSC. Gwynne, M.D. & Croze, H The concept and practice of ecological monitoring over large areas of land: The systematic reconnaissance flight (SRF). In: Wildlife Management in Savannah Woodlands. Recent Progress in African Studies (eds Ajayi, S.S. & Halstead, L.B.). Pp London:Taylor & Francis Ltd. ILCA Low-level aerial survey techniques. Workshop Proceedings. International Livestock Centre for Africa (ILCA) Monograph 4. Nairobi: ILCA. Kingdon, J The Kingdon Field Guide to African Mammals. London: Academic Press. Norton-Griffiths, M Counting animals. Techniques in African Wildlife Ecology. Nairobi: African Wildlife Foundation. Ramberg, L., Hancock, P., Lindholm, M., Meyer, T., Ringrose, S., Sliva, J., Van As, J. & Vanderpost, C Species diversity of the Okavango Delta, Botswana. Aquatic Science. 68: Ringrose, S., VanderPost, C. & Matheson, W Mapping Ecological Conditions in the Okavango Delta, Botswana, Using Fine and Coarse Resolution Systems Including Simulated Spot Vegetation Imagery. International Journal of Remote Sensing, 24: Smithers, R.H.N The Mammals of Botswana. Mus. mem. Natl. Mus. Monum. Rhod, Harare. TWCM/FZS Tanzania Wildlife Conservation Monitoring Procedures Manual. Unpublished. Arusha: TWCM/FZS. WWF Terrestrial Ecoregions. Conservation Science Program (CSP).

11 11 Appendix I. Survey area (NG/26 aerial survey September 2009). Hunda Airfield NG/25 Boro River Abu Camp NG/28 Moremi Game Reserve Macatoo Camp NG/27A NG/8 NG/26 Sandveld Tongue NG/29 Thaoge River

12 12 Appendix II. Survey parameters (NG/26 aerial survey September 2009). Survey date Sep 2009 Survey area NG/26 1,736 km 2 Area surveyed 2,048 km 2 Type of survey Systematic Reconnaissance Flight (SRF) Aircraft Cessna C206 (A2-ANT) Survey crew Pilot Graham Skinner (Sefofane Air Charters) FSO Petri Viljoen (Survey Consultant) LSO Martin Kays (Tubu Tree Lodge) RSO Dr Kate Evans (Elephants for Africa) Airfield used Hunda Airfield (NG/25) Number of survey sessions 4 Number of survey strips 36 Survey strip lengths Total 1,469.5 km Range km (average = 40.9 km) Survey height (AGL) Targeted survey height 107 m (350ft) Recoded survey height - average m Recorded survey height - range m Survey ground speed Recorded survey ground speed - average km/hr Recorded survey ground speed - range km/hr Survey strip spacing 1.5 km (48 sec) Sample strip calibration (calculated regression) Left seat observer (LSO) 0.94 R 2 Right seat observer (RSO) 0.91 R 2 Sample strip width Target sample strip width 300 m (150 m m) Calculated average sample strip width m (161.3 m m) Area sampled Targeted sample area ± km 2 (± 21.5%) Calculated area sampled ± 22.7% Survey session durations Day 1 session 1 (28 Sep 2009) 01:43 hrs (15:34 17:17) Day 2 session 1 (29 Sep 2009) 03:09 hrs (07:00 10:09) Day 3 session 1 (30 Sep 2009) 02:52 hrs (09:16 12:08) Day 3 session 2 (30 Sep 2009) 01:40 hrs (14:57 16:37) Weather conditions Day 1 session 1 (28 Sep 2009) Warm, light E wind, 0/8 cloud Day 2 session 1 (29 Sep 2009) Warm, moderate SE wind, 0/8 cloud Day 3 session 1 (30 Sep 2009) Cool, light SE wind, 8/8 cloud Day 3 session 2 (30 Sep 2009) Cool, light SE wind, 7/8 cloud

13 13 Appendix III. Survey strip data (NG/26 aerial survey September 2009). Survey Northern Boundary Southern Boundary Strip ID Longitude Latitude Longitude Latitude

14 14 Appendix IV. Survey strips flown (NG/26 aerial survey September 2009).

15 Appendix V. Wildlife species counted and recorded human activities (NG/26 aerial survey September 2009). Species 15 Total N Inside Strip Outside Strip Chacma baboon Papio ursinus (troop) 1 Lion Panthera leo 10 2 Elephant Loxodonta africana ** 1, Zebra Equus quagga Hippopotamus Hippopotamus amphibious 16 8 Warthog Phacochoerus aethiopicus 12 8 Giraffe Giraffa camelopardalis Buffalo Syncerus caffer ** 2, Greater kudu Tragelaphus strepsiceros Southern reedbuck Redunca arundinum 9 6 Lechwe Kobus leche Impala Aepyceros melampus Tsessebe Damaliscus lunatus Wildebeest Connochaetes taurinus 30 6 Gemsbuck Oryx gazella 1 1 Ostrich Struthio camelus Wattled Crane Grus carunculata Saddlebilled Stork Ephippiorhynchus senegalensis Ground Hornbill Bucorvus leadbeateri 10 3 Crocodile Crocodylus niloticus 12 8 ** Totals include animals counted both inside and outside the sample strips. N = number of observations 651 Human Activities & Livestock Total N Cattle ** Horses ** 12 2 Donkeys ** 22 1 ** Totals include animals counted both inside and outside the sample strips. 15 Naming follows Kingdon (1997).

16 16 Appendix VI. Distribution of wildlife (NG/26 aerial survey September 2009). Buffalo Elephant

17 17 Impala Zebra & Wildebeest

18 18 Kudu & Tsessebe Reedbuck & Lechwe

19 19 Warthog & Giraffe Hippopotamus & Crocodile

20 20 Ostrich & Ground Hornbill Wattled Crane & Saddlebilled Stork

21 21 Appendix VII. Distribution of human activities and livestock (NG/26 aerial survey September 2009).