Southern Africa

REGIONAL OVERVIEW

General Statistics

Total Area: 5,973,020 km2

Range area (% of region): 1,305,140 km2 (28%)

Protected area coverage (% of region): 12%

Protected range (% of known and possible range in protected areas): 28%

Information Quality Index (IQI): 0.48

Current Issues

Southern Africa holds the largest elephant populations on the continent, and many of the management challenges associated with high elephant densities in large populations are common to a number of countries in the region. As elephant numbers continue to increase in the region's largest populations, the debate on the need for management action has continued in a number of countries, particularly in South Africa.

In 2004, the Southern African Range States embarked on a process to develop a regional strategy for the management of elephants, and a draft strategy document was produced following a workshop held in 2005. The draft strategy aims to foster regional cooperation in elephant management and monitoring, calling for coordinated surveys across international boundaries. It is noteworthy that only three countries in Africa fund their regular elephant survey programmes out of their national budgets, and all three – Botswana, Namibia and South Africa – are in Southern Africa. Every other Range State in the region, and indeed on the continent, depends on external aid for elephant survey work.

Four countries in the region, namely Botswana (DG Ecological Consulting, 2003b), Namibia (Ministry of Environment and Tourism, 2005c), South Africa (South African National Parks, 2004b) and Zambia (Ministry of Tourism, Environment and Natural Resources, 2003) have recently developed or updated their respective national elephant management policies.

The development of transfrontier conservation areas has continued in Southern Africa. A Memorandum of Understanding was signed in late 2006 by Ministers from five Southern African countries to facilitate the development the Kavango-Zambezi Transfrontier Conservation Area (KAZA TFCA). An area the size of Italy which holds nearly half of the continental elephant population, the KAZA TFCA spans some of the most important populations in Angola, Botswana, Namibia, Zambia and Zimbabwe.

The international sale of 50 tons of legally acquired ivory from Botswana, South Africa and Namibia, approved in 2002 at the 12th Meeting of the Conference of the Parties to CITES held in Santiago, Chile, was still pending at the time of writing, awaiting the finalization of an agreed baseline of data from the CITES MIKE Programme. At the 13th Meeting of the Conference of the Parties, held in Bangkok, Thailand, in 2004, a proposal was approved to allow Namibia non-commercial trade in individually marked and certified ekipas, a type of traditional ivory amulet (CITES, 2004).

In January 2007, a number of proposals were submitted for consideration at the 14th Meeting of the Conference of the Parties to CITES in The Hague, The Netherlands, in June 2007. Botswana and Namibia submitted a proposal to maintain their elephant populations, as well as those of South Africa and Zimbabwe, in CITES Appendix II, and to establish annual export quotas for these four countries to trade in raw ivory to approved trading partners (Government of Botswana & Government of Namibia, 2007). Botswana submitted an additional proposal to be allowed to trade in hides and leather goods for commercial purposes; to be allocated an annual export quota of up to eight tons of Government-owned raw ivory to approved destinations; and to conduct a one-off sale of no more than 40 tons of raw ivory stocks to similarly approved destinations (Government of Botswana, 2007).

At the same time, Kenya and Mali submitted a proposal to bar any international trade in raw or worked ivory from any of the above four Southern African countries for a period of 20 years, except for raw ivory from hunting trophies for non-commercial purposes and the export of ivory pursuant to the conditional sale of government-owned ivory stocks from Botswana, Namibia and South Africa agreed at the 12th Meeting of the Conference of the Parties to CITES (Government of Kenya & Government of Mali, 2007).

Range Data

Savanna elephants (Loxodonta africana africana) predominate throughout the region, although small populations of forest elephants are found in the Angolan exclave of Cabinda and perhaps also in northwestern Angola. With an estimated total elephant range spanning over 1.3 million km2, Southern Africa is the region with the largest elephant range area, and accounts for 39% of the continental total. While this is 22% less than the region's estimated range area in the AESR 2002, the difference is largely a result of better information, rather than a recent reduction in actual elephant range. Southern Africa is the region with the smallest proportion of elephant range in protected areas (28%).

The quality of range data varies considerably across the region. Although three-quarters of the range information is less than 10 years old, only 53% of total range currently falls within the known category. Although elephant range is expanding in Botswana and spreading into neighbouring countries such as Angola and Namibia, the overall range area may decline in future as more detailed information is obtained, particularly from Angola and Mozambique, where range data are least reliable, but which together account for 57% of the regional range estimate.

Population Data

Southern Africa is the region with the highest overall quality of elephant information, as measured by the IQI. There is, however, wide variation amongst countries, with nearly perfect information available for Swaziland, South Africa, Zimbabwe and Botswana, but virtually no reliable information for Angola.

This report features 96 new estimates for Southern Africa, 84 of which originate from systematic counts. Overall, elephant population estimates are available for some 690,000 km2, or 53% of estimated elephant range in Southern Africa, with estimates from systematic surveys covering two-thirds of that area. Some countries such as South Africa, Botswana and Zimbabwe have complete coverage, in contrast with Angola, where estimates are only available for 5% of estimated elephant range.

Southern Africa also holds by far the largest number of elephants on the continent. At nearly 321,000, the number of definite plus probable elephants, is nearly twice as high as the corresponding number for Eastern Africa, the next most populous region with about 166,500 definite plus probable elephants. Over three-quarters of the regional definite plus probable population in Southern Africa occur in just two countries, namely Botswana and Zimbabwe. These two countries together also account for nearly 47% of the continental definite plus probable population.

The estimated number of definite elephants in Southern Africa has increased by over 51,000 (19%) compared to the previous report, largely as a result of higher estimates from recent, methodologically comparable surveys in Botswana, Namibia, South Africa, Zambia and Zimbabwe. The total number of elephants estimated in such surveys (namely from those sites marked repeat survey or rs in the national tables of estimates; see Appendix II for a list of sites) accounts for over three-quarters of the region's definite plus probable estimate. A comparison of estimates for these sites between this and the previous report, as described in Blanc et al. (2005), indicates a significant increase of 46,354 ± 30,588 in the pooled comparable estimate since the AESR 2002 (t = 2.97; p < 0.01). This translates into an overall average annual rate of increase of 3.88% (95% CI of rate 1.06% to 6.39%) in the comparable populations.

SUMMARY TOTALS FOR SOUTHERN AFRICA

Data Category DEFINITE PROBABLE POSSIBLE SPECULATIVE
Aerial or Ground Total Counts 18,431 0 0 0
Direct Sample Counts and Reliable Dung Counts 278,283 23,137 23,137 0
Other Dung Counts 27 49 9 0
Informed Guesses 977 0 1,588 201
Other Guesses 0 0 0 9,552
TOTAL 2006 297,718 23,186 24,734 9,753
TOTALS 2002 246,592 23,722 26,098 7,508

INTERPRETATION OF CHANGES IN ESTIMATES FROM PREVIOUS REPORT

Cause of Change DEFINITE PROBABLE POSSIBLE SPECULATIVE
Repeat Survey +47,808 +105 -137 0
New Population +300 -48 +181 0
Different Technique +448 +167 -709 -44
Different Area +2,697 -951 -94 0
New Guess -73 0 +85 +1,442
New Analysis +26 -9 -11 -67
Population Lost 0 0 -28 0
Data Degraded -79 +201 -651 +914
TOTAL CHANGE +51,126 -536 -1,364 +2,245

AREA OF RANGE COVERED BY EACH DATA CATEGORY (km2)

Data Category Known Range Possible Range Total Range
Aerial or Ground Total Counts 43,478 0 43,478
Direct Sample Counts and Reliable Dung Counts 397,261 23,300 420,561
Other Dung Counts 482 0 482
Informed Guesses 14,801 10,988 25,789
Other Guesses 134,788 69,029 203,817
Unassessed Range 102,833 508,179 611,012
TOTAL 693,643 611,497 1,305,140

SOUTHERN AFRICA: COUNTRY AND REGIONAL TOTALS & DATA QUALITY

COUNTRY ELEPHANT NUMBERS RANGE AREA (km2) % OF REGIONAL RANGE % OF RANGE ASSESSED IQI1 PFS2
DEFINITE PROBABLE POSSIBLE SPECULATIVE
Angola 818 801 851 60 406,946 31 5 0.03 1
Botswana 133,829 20,829 20,829 0 100,265 8 99 0.87 2
Malawi 185 323 632 1,587 7,538 1 89 0.17 3
Mozambique 14,079 2,396 2,633 6,980 334,786 26 77 0.48 1
Namibia 12,531 3,276 3,296 0 146,921 11 55 0.46 2
South Africa 17,847 0 638 22 30,455 2 100 0.96 2
Swaziland 31 0 0 0 50 0 100 1.00 5
Zambia 16,562 5,948 5,908 813 201,247 15 61 0.47 1
Zimbabwe 84,416 7,033 7,367 291 76,931 6 99 0.91 2
TOTAL* 297,718 23,186 24,734 9,753 1,305,140 39 53 0.48 1

* Note that totals for the Definite, Probable and Possible categories are derived by pooling the variances of individual estimates, as described under the Data Types and Categorization section. As a result, totals do not necessarily match the simple sum of the entries within a given category.

1 IQI: Information Quality Index. This index quantifies overall data quality at the national and regional levels based on the precision of estimates and the proportion of assessed elephant range (i.e. range for which estimates are available). The IQI ranges from zero (no reliable information) to one (perfect information). See the Introduction section for a detailed explanation of how the IQI is calculated.

2 PFS: Priority for Future Surveys, ranked from 1 to 5 (highest to lowest). Based on the IQI and the proportion of continental range accounted for by the country in question, the PFS is a measure of the importance and urgency for future population surveys, particularly in areas of unassessed range and areas not surveyed in the last 10 years or more. See Introduction for a more detailed explanation of how the priority ranking is derived.

The numbers of elephants in the probable and possible categories have declined by about 540 and 1,360 respectively, reflecting a marginal overall improvement in the precision of estimates. The figure under the speculative category, on the other hand, has increased by 2,245, primarily as a result of new guesses and the degradation of old estimates to the category of other guesses.

Cross-border Movements

Movements of elephants are known to occur between Mozambique and Tanzania (Eastern Africa). The only other area in Southern Africa where cross-border movement may take place is between northern Angola and the southwest of the Democratic Republic of Congo, but there are no reliable reports of such movements.

ANGOLA

General Statistics

Country area: 1,246,700 km2

Range area (% of country): 406,946 km2 (53%)

Protected area coverage (% of country): 7%

Protected range (% of known and possible range in protected areas): 9%

Information Quality Index (IQI): 0.03

CITES Appendix: I

Listing Year: 1989

Current Issues

Although conservation and monitoring work has begun after the end of hostilities in Angola, it has concentrated largely on the southeast, with little or no assessment effort taking place elsewhere. A detailed nationwide study of elephant distribution and density needs to be undertaken.

A project to clear landmines from the Luiana Partial Reserve got underway in 2005 as part of a plan to establish the Kavango Zambezi Transfrontier Conservation Area, which will span five countries and will cover part of the Upper Zambezi Basin, the Okavango Basin and the Okavango Delta.

A recent survey found over 1,500 kg of ivory openly for sale in Luanda. Although most of the stock is believed to originate in neighbouring Democratic Republic of Congo, it is thought that some of the ivory may be sourced within Angola, and particularly from the northwest and southeast of the country. Although Angola formally approved its membership of CITES in 2001, the decision has yet to be formally gazetted, and it therefore officially remains a non-party to the convention (Milliken et al., 2006).

A recent study found human-elephant conflict in the province of Cabinda to be widespread, and the severity of the problem is reported to be increasing (Heffernan, 2005). There have also been recent reports of crop raiding and destruction of property by elephants between Malengue and the Bicuar National Park (Anon., 2004).

Range Data

Elephants are believed to be present in the far north and far south of the country. Forest elephants (Loxodonta africana cyclotis) are believed to occur in the northwest and savanna elephants (Loxodonta africana africana) in the northeast and south (Enock, quest. reply, 2002).

Two major changes have been made to the southern portion of the range map for Angola for this report, both based on information provided by Chase (2006). The area of known range around the Luiana Partial Reserve has been expanded northward and eastward, and the area of southern Angola north of the 15th parallel has been re-categorized as doubtful range. Nevertheless, a herd of elephants was reported in August 2004 in the village of Capembe, to the southwest of the town of Malengue (Anon., 2004), and this is shown as a cross on the map. Further southeast, also shown as a cross on the map, is a sighting of 17 elephants (in four herds) along the Cuito River, 120 km north of the Namibian border during a reconnaissance flight in November 2005 (Chase & Griffin, 2005b). The straight boundaries of these changes in the range map are a reflection of limited knowledge in this area and do not represent actual elephant distribution. The hard boundaries of elephant range along the border with the Democratic Republic of Congo to the north, and with Namibia to the south, on the other hand, reflect marked differences in human population densities on either side of the border.

The area of known range in the exclave of Cabinda has been extended to the southwest based on information in Heffernan (2005). Despite these updates, the range map for Angola remains, for the most part, highly speculative, with over three-quarters of the available range data being over 10 years old.

Population Data

A series of surveys of the Luiana Partial Reserve have been conducted since 2003 by Chase and Griffin (2005b). The estimates from these surveys have consistently increased from 263 in 2003 to 1583 in 2005, suggesting that elephants are moving into Angola from neighbouring countries in increasing numbers. In all three surveys, elephants were found invariably in the southeastern sectors of the park, with no elephants detected in the western sector. In consequence, two entries are shown for Luiana in the table of estimates, one for Luiana (West) with an estimate of zero and another for Luiana (East) with an estimate of 1,583 ± 801 from the 2005 survey. These replace an informed guess of 100 by Hanks (J. Hanks, pers. comm., 2003). The resulting increases in the definite, probable and possible categories reflect better information as well as the movement of elephants from other countries into Angola.

All other estimates for Angola remain unchanged from the previous report. Despite the considerable reduction in elephant distribution shown in this report, nearly 95% of Angola's estimated elephant range remains unsurveyed.

Cross-border Movements

As mentioned above, there is evidence that elephants are moving into southeastern Angola from Botswana through neighbouring Namibia in increasing numbers. Cross-border movement is also possible between Luiana and Sioma Ngwezi National Park (Zambia).

Movement across Angola's northern border with the Democratic Republic of Congo is unlikely to take place, as elephants are no longer thought to occur in that part of the DRC, where human population densities are relatively high.

Elephant distribution in the exclave of Cabinda is spread along the border with Congo, and there is likely to be movement of elephants across that border (Heffernan, 2005).

SUMMARY TOTALS FOR ANGOLA

Data Category DEFINITE PROBABLE POSSIBLE SPECULATIVE
Aerial or Ground Total Counts 36 0 0 0
Direct Sample Counts and Reliable Dung Counts 782 801 801 0
Informed Guesses 0 0 50 0
Other Guesses 0 0 0 60
TOTALS2006 818 801 851 60
TOTALS 2002 36 0 150 60

INTERPRETATION OF CHANGES IN ESTIMATES FROM PREVIOUS REPORT

Cause of Change DEFINITE PROBABLE POSSIBLE SPECULATIVE
Different Technique +782 +801 +701 0
TOTAL CHANGE +782 +801 +701 0

AREA OF RANGE COVERED BY EACH DATA CATEGORY (km2)

Data Category Known Range Possible Range Total Range
Aerial or Ground Total Counts 468 0 468
Direct Sample Counts and Reliable Dung Counts 10,510 3 10,512
Informed Guesses 0 8,396 8,396
Other Guesses 0 1,502 1,502
Unassessed Range 51,254 334,814 386,068
TOTAL 62,232 344,714 406,946

ANGOLA: ELEPHANT ESTIMATES

INPUT ZONE CAUSE OF CHANGE1 SURVEY DETAILS2 NUMBER OF ELEPHANTS SOURCE PFS3 AREA (km2) MAP LOCATION
TYPE RELIAB. YEAR ESTIMATE 95% C.L. LON. LAT.
Bongola Area IG3 E 1992 60 Anstey, 1993 3 1,505 17.6 E 17.3 S
Cáua Camp IR1 A 2002 36 P. Vaz Pinto, pers. comm., 2003 2 9,500 13.3 E 9.3 S
Luiana (East) Partial Reserve DT AS2 B 2005 1,583 801 Chase & Griffin, 2005b 3 4,032 22.8 E 17.4 S
Luiana (West) Partial Reserve DT AS2 B 2005 0 Chase & Griffin, 2005b 4,368 22.3 E 17.4 S
Quiçama National Park IG3 D 2002 50 P. Vaz Pinto, pers. comm., 2003 2 9,500 13.6 E 9.8 S

1 Key to Causes of Change: DA: Different Area; DD: Data Degraded; DT: Different Technique; NA: New Analysis; NG: New Guess; NP: New population; PL: Population Lost; RS: Repeat Survey (RS' denotes a repeat survey that is not statistically comparable for reasons such as different season); —: No Change

2 Key to Survey Types: AS: Aerial Sample Count; AT: Aerial Total Count; DC: Dung Count; GD: Genetic Dung Count; GS: Ground Sample Count; GT: Ground Total Count; IG: Informed Guess; IR: Individual Registration; OG: Other Guess. Survey Type is followed by an indicator of survey quality, ranked from 1 to 3 (best to worst). Survey Reliability is keyed A–E (best to worst)

3 PFS: Priority for Future Surveys, ranked from 1 to 5 (highest to lowest). Based on the precision of estimates and the proportion of national range accounted for by the site in question, PFS is a measure of the importance and urgency for future population surveys. All areas of unassessed range have a priority of 1. See Introduction for details on how the PFS is derived.

BOTSWANA

General Statistics

Country area: 600,370 km2

Protected area coverage (% of country): 18%

Range area (% of country): 100,265 km2 (17%)

Protected range (% of known and possible range in protected areas): 19%

Information Quality Index (IQI): 0.87

CITES Appendix: II

Listing Year: 1997

Current Issues

Botswana's elephant population continues to grow, and controversy over its impact on vegetation and biodiversity has become increasingly heated in recent years. There have been numerous calls for control measures to be put in place to prevent biodiversity loss, as changes to vegetation, particularly along the river fronts, have become more apparent. Some authors, however, believe that no action should be taken, and that the ecosystem is reverting to its condition before hunting reduced elephant numbers in the 19th century (Skarpe et al., 2004). Still other experts believe that Botswana's population is now beyond feasible control measures (Cumming & Jones, 2005), and that only a drought or disease outbreak can reduce the population.

These and other issues were addressed in a review of the 1991 Elephant Conservation and Management Plan (Department of Wildlife and National Parks, 1991), which resulted in a new draft policy and strategy for the conservation and management of elephants in Botswana (DG Ecological Consulting, 2003a,b). The primary objectives of the strategy are to reduce human-elephant conflict to acceptable levels; to prevent, reduce or reverse unacceptable elephant-induced environmental changes; to maximize benefits from sustainable utilization of elephants; and to protect elephants through law enforcement.The strategy sets targets to reduce human-elephant conflict, which is reportedly intensifying as the elephant population expands (DG Ecological Consulting, 2003a), with up to 40% of the potential annual harvests of subsistence farmers being destroyed by elephants in some areas (Mosojane, 2004). As the strategy awaits final approval and implementation, a number of measures aimed at reducing human-elephant conflict and the impact of the growing elephant population have been planned or taken. These include the granting in 2005 of citizen hunting permits for elephant in the Tuli area (Mojaphoko, pers comm. 2005), and a plan to translocate 500 elephants from Botswana to Gorongosa National Park in Mozambique, scheduled for 2004, but which did not take place.

A number of TFCAs are in the process of being established to provide for elephant dispersal beyond Botswana's borders. The largest of these will be the Kavango-Zambezi (KAZA) TFCA, which is expected to cover over 287,000 km2, and which will include parts of northern Botswana, the Luiana Partial Reserve in Angola, Sioma Ngwezi National Park in Zambia and Babwata National Park in Namibia. This TFCA is in its conceptual phase, and a Memorandum of Understanding between the Governments of these countries was signed in late 2006.The signing of a similar Memorandum of Understanding to pave the way for the creation of the Limpopo-Sashe TFCA met with unexpected delays in 2005 (Peace Parks Foundation, 2006). This TFCA will encompass the Tuli Game Reserve, the new Mapungubwe National Park in South Africa and the Tuli Circle Safari Area in Zimbabwe, and will include a large proportion of privately owned land in all three countries.

In 2006 Botswana notified CITES of an increase in its export quota for elephant trophies to 540 tusks (270 animals), up by 120 tusks from the quota it had maintained for the previous four years (UNEP-WCMC, 2006). In January 2007, Botswana submitted a proposal for consideration at the 14th Meeting of the Conference of the Parties to CITES, to be allowed to trade in hides and leather goods for commercial purposes; to be allocated an annual export quota of up to eight tons of Government-owned raw ivory of Botswana origin to approved destinations; and to conduct a one-off sale of no more than 40 tons of raw ivory stocks to similarly approved destinations (Government of Botswana, 2007).

SUMMARY TOTALS FOR BOTSWANA

Data Category DEFINITE PROBABLE POSSIBLE SPECULATIVE
Direct Sample Counts and Reliable Dung Counts 133,829 20,829 20,829 0
TOTALS2006 133,829 20,829 20,829 0
TOTALS 2002 100,629 21,237 21,237 0

INTERPRETATION OF CHANGES IN ESTIMATES FROM PREVIOUS REPORT

Cause of Change DEFINITE PROBABLE POSSIBLE SPECULATIVE
Repeat Survey +34,128 +643 +643 0
Different Technique -928 -1,051 -1,051 0
TOTAL CHANGE +33,200 -408 -408 0

AREA OF RANGE COVERED BY EACH DATA CATEGORY (km2)

Data Category Known Range Possible Range Total Range
Direct Sample Counts and Reliable Dung Counts 88,352 11,190 99,543
Unassessed Range 722 0 722
TOTAL 89,075 11,190 100,265

BOTSWANA: ELEPHANT ESTIMATES

INPUT ZONE CAUSE OF CHANGE1 SURVEY DETAILS2 NUMBER OF ELEPHANTS SOURCE PFS3 AREA (km2) MAP LOCATION
TYPE RELIAB. YEAR ESTIMATE 95% C.L. LON. LAT.
Chobe National Park & Environs RS AS2 B 2006 40,767 10,539 DWNP, 2006 2 12,195 24.6 E 18.5 S
Nxai Pan & Makgadikgadi National RS AS3 B 2006 1,436 1,950 DWNP, 2006 2 11,476 24.8 E 20.4 S
Parks Okavango Delta RS AS2 B 2006 31,191 7,191 DWNP, 2006 2 17,160 23.1 E 19.4 S
Rest of Northern Botswana RS AS2 B 2006 80,226 16,334 DWNP, 2006 1 18,303 23.5 E 19.7 S
Tuli Game Reserve DT AS2 B 2006 1,038 685 DWNP, 2006 2 3,510 29.1 E 22.2 S

1 Key to Causes of Change: DA: Different Area; DD: Data Degraded; DT: Different Technique; NA: New Analysis; NG: New Guess; NP: New population; PL: Population Lost; RS: Repeat Survey (RS' denotes a repeat survey that is not statistically comparable for reasons such as different season); —: No Change

2 Key to Survey Types: AS: Aerial Sample Count; AT: Aerial Total Count; DC: Dung Count; GD: Genetic Dung Count; GS: Ground Sample Count; GT: Ground Total Count; IG: Informed Guess; IR: Individual Registration; OG: Other Guess. Survey Type is followed by an indicator of survey quality, ranked from 1 to 3 (best to worst). Survey Reliability is keyed A–E (best to worst)

3 PFS: Priority for Future Surveys, ranked from 1 to 5 (highest to lowest). Based on the precision of estimates and the proportion of national range accounted for by the site in question, PFS is a measure of the importance and urgency for future population surveys. All areas of unassessed range have a priority of 1. See Introduction for details on how the PFS is derived.

Range Data

Most of the country's elephant range is situated in the north and spans around 100,000 km2. The expansion of elephant range in Botswana continues, and portions of known range have been added to the southern sector of Nxai Pan National Park and along the watercourse marking the western boundary of Magkadigkadi National Park, where elephants were sighted in recent surveys and have become permanent residents (G.C. Craig, pers. comm., 2006). An area of range to the southeast of Maun, featured in the previous report, has been removed for lack of recent evidence of elephant presence (G.C. Craig, pers. comm., 2006). Nevertheless, elephant spoor has been regularly seen in a small holding close to Maun (D. Gibson, pers. comm. 2006).

The only other area where elephants occur in Botswana is considerably smaller than the northern range and lies in the eastern tip of the country, covering around 1,000 km2 at the confluence of the Shashe and Limpopo rivers. Two small areas of known range have been added to the northwest of Tuli: a narrow strip along the Sashe River on the border with Zimbabwe and a small area south of Francistown (DG Ecological Consulting, 2003a; G.C. Craig, pers. comm., 2006; M.E. Gadd, pers. comm., 2006). In 2004, two elephants were shot on Wayside Farm, outside Francistown (Chase & Griffin, 2005a), and this is shown as a cross on the map. Another cross is shown further south, where elephants were seen in 2000 (M.E. Gadd, pers. comm., 2006).

Population Data

Most of Botswana is covered by regular aerial sample counts that include the entire northern elephant range. Five surveys have been conducted since 2002, but the 2005 survey did not cover the entire survey area due to logistical problems (C. Taolo, pers. comm. 2006). The national estimates for the other surveys were 123,152 ± 17,152 in 2002 (Department of Wildlife and National Parks, 2002), 109,471 ± 18,443 in 2003 (Department of Wildlife and National Parks, 2003), 151,000 ± 20,004 in 2004 (Department of Wildlife and National Parks, 2004) and 154,658 ± 21,253 (Department of Wildlife and National Parks, 2006). The estimates presented here, obtained from the 2006 survey report, have been split to show separate figures for the survey blocks containing national parks and the Okavango Delta. These replace a 1999 combined aerial sample count estimate of 120,604 ± 21,237 (Department of Wildlife and National Parks, 1999). The estimate for the Tuli Game Reserve, also from the 2006 national survey, replaces an aerial total count estimate of 1,262 (Selier et al., 2002).

The number of elephants in the definite category has increased by over 33,000 as recorded by methodologically comparable surveys. The numbers in the probable and possible categories have dropped by approximately 410 elephants each due to an overall increase in precision of the national estimate.

Cross-border Movements

Elephants in northern Botswana are part of a larger population that stretches east into Zimbabwe (Craig, 1996b), north into the Caprivi Strip in Namibia and possibly into Zambia and Angola as well (Chase & Griffin, 2005a). It constitutes the largest known population of elephants in Africa, and one of the largest continuous stretches of known range on the continent. As part of a continuing westward expansion in northern Botswana, elephants are reported to be crossing into the Khaudom and Nyae Nyae areas of Namibia. Elephants also move from the eastern range portion of Botswana to Zimbabwe and into private reserves in northern Limpopo Province (South Africa).

MALAWI

General Statistics

Country area: 118,480 km2

Range area (% of country): 7,538 km2 (7%)

Protected area coverage (% of country): 9%

Protected range (% of known and possible range in protected areas): 84%

Information Quality Index (IQI): 0.17

CITES Appendix: I

Listing Year: 1989

Current Issues

Elephant populations in Malawi are small and fragmented, and are further threatened by encroachment. Human-elephant conflict is prevalent as a result, and poaching for meat and ivory is believed to be further reducing certain populations (Bhima et al., 2003). The long-term viability of several of Malawi's elephant populations is already in doubt.

In July 2006, 70 elephants from Liwonde National Park and Mangochi Forest Reserve were translocated to Majete Wildlife Reserve, where elephants had been absent for over 10 years (Sherry & Tattersall, 1996). The African Parks Foundation, which conducted the translocation, had taken over the management of Majete in 2003 (African Parks Foundation, 2006b).

Range Data

Because of its small size and high human population densities, Malawi's elephants are almost entirely confined to protected areas, including national parks and forest reserves, with only a small part of range outside protected areas in the southeast (Bhima, 1996). The main elephant populations occur in Liwonde and Kasungu National Parks and the Nkhota-Kota and Vwaza Marsh Wildlife Reserves.

For this report, an area of known range has been added within the boundary of the Majete Wildlife Reserve, to which elephants have been recently translocated. The shape of known range in Thuma Forest Reserve has been altered with information from the Landscan 2002 human population density data set (ORNL/GIST, 2002; see Introduction section for rationale). Based on the same data, an area of possible range to the north of Mangochi Forest Reserve has been categorized as doubtful range.

Population Data

Aerial sample counts of Kasungu National Park and Vwaza Marsh Game Reserve were conducted in 2005 (Ferreira et al., 2005). Although relatively high sampling intensities were employed in these surveys, the lack of stratification resulted in wide confidence intervals around the estimates. In addition, neither of these surveys employed a radar altimeter, which further compromises the reliability of the estimates.

The Kasungu survey, with an estimate of 58 ± 218, replaces a 1995 aerial sample count estimate of 391 ± 349. As the lower confidence interval in the 2005 survey is larger than the estimate, the number of elephants seen in the survey (25) contributes to the definite category in the summary table. A dung count of Kasungu had been conducted in 2002, giving an estimate of 117 with an asymmetric confidence interval of 96 to 142, but this estimate only covered the southern part of the park (Bhima et al., 2003).

The survey of Vwaza Marsh could not be completed due to aircraft failure (Ferreira et al., 2005), and its estimate of 270 has been categorized as an informed guess for this report, with the number of elephants seen in the survey (41) appearing under the definite category. This guess replaces a 1997 informed guess of 35 (Gibson, 1997).

The same team also conducted a survey of Nyika National Park using a point transect design (Ferreira et al., 2005). The survey returned an estimate of zero, but the sampled area excluded the north of the park, where elephants are found. For this reason, the estimate of 339 from the previous report has been retained.

As mentioned under Current Issues above, 70 elephants were translocated from Liwonde National Park to a 140 km2 fenced sanctuary within Majete Wildlife Reserve in July 2006. As the Majete population was previously extinct, a figure of 70 now appears as a new population in the table of estimates.

Prior to the Majete translocation, Liwonde National Park was said to hold in excess of 600 elephants (African Parks Foundation, 2006b), but no details could be obtained on the basis of the estimate. The difference between this figure and the 70 elephants translocated to Majete, (i.e. 530 elephants) has been entered for Liwonde in the category of other guesses. This new estimate replaces a figure of 414 from an aerial sample count conducted in 1995 (Bhima, 1996).

The estimate of 1,037 for Nkhota-Kota Wildlife Reserve (Japan International Cooperation Agency & Government of Malawi, 1997) has been degraded to the category of other guesses for the same reason, as it is now over 10 years old.

There have been substantial decreases in the estimates for the definite, probable and possible categories compared to the previous report. These decreases in estimates are mainly due to the degradation of old data to the speculative category or their replacement by low quality guesses. Estimates from recent systematic surveys only contribute marginally to these decreases. Although estimates are available for nearly 90% of Malawi's elephant range, over half of that area is covered by guesses.

Cross-border Movements

Cross-border movement between Kasungu National Park in Malawi and the North Luangwa ecosystem was documented by Jachmann and Bell (1985), but a recent survey of Zambia's Lukusuzi National Park, across the border from Kasungu, failed to find any elephants in the park (Fourie et al., 2005). Furthermore, the area between the two parks is densely settled (Bhima et al., 2003), and it is unlikely that elephant movement continues.

SUMMARY TOTALS FOR MALAWI

Data Category DEFINITE PROBABLE POSSIBLE SPECULATIVE
Direct Sample Counts and Reliable Dung Counts 74 323 323 0
Informed Guesses 111 0 309 20
Other Guesses 0 0 0 1,567
TOTALS2006 185 323 632 1,587
TOTALS 2002 647 1,569 1,649 20

INTERPRETATION OF CHANGES IN ESTIMATES FROM PREVIOUS REPORT

Cause of Change DEFINITE PROBABLE POSSIBLE SPECULATIVE
Repeat Survey -18 -308 -94 0
New Population +75 0 0 0
New Guess -439 0 +165 +530
Data Degraded -80 -938 -1,087 +1,037
TOTAL CHANGE -462 -1,246 -1,017 +1,567

AREA OF RANGE COVERED BY EACH DATA CATEGORY (km2)

Data Category Known Range Possible Range Total Range
Direct Sample Counts and Reliable Dung Counts 2,908 0 2,908
Informed Guesses 1,511 0 1,511
Other Guesses 2,316 0 2,316
Unassessed Range 0 804 804
TOTAL 6,735 804 7,538

MALAWI: ELEPHANT ESTIMATES

INPUT ZONE CAUSE OF CHANGE1 SURVEY DETAILS2 NUMBER OF ELEPHANTS SOURCE PFS3 AREA (km2) MAP LOCATION
TYPE RELIAB. YEAR ESTIMATE 95% C.L. LON. LAT.
Kasungu National Park RS' AS1 B 2005 58 218 Ferreira et al., 2005 1 2,463 33.1 E 12.9 S
Liwonde National Park NG OG3 E 2006 530 African Parks Foundation, 2006b 2 538 35.3 E 14.9 S
Majete Wildlife Reserve NP IG3 D 2006 70 African Parks Foundation, 2006b 3 140 34.7 E 16.0 S
Nkhota-Kota Wildlife Reserve DD AS2 E 1995 1,037 1,511 JICA & Government of Malawi, 1997 1 1,802 34.0 E 12.9 S
Nyika National Park AS1 B 1997 339 239 Gibson, 1997 1 3,134 33.8 E 10.6 S
Phirilongwe Forest Reserve IG3 D 1998 50 S.M. Munthali, pers. comm., 1998 2 640 35.0 E 14.6 S
Thuma Forest Reserve IG3 D 1998 30 20* S.M. Munthali, pers. comm., 1998 2 370 34.2 E 13.9 S
Vwaza Marsh Wildlife Reserve NG IG3 D 2005 270 Ferreira et al., 2005 1 976 33.4 E 11.0 S

* Range of informed guess

1 Key to Causes of Change: DA: Different Area; DD: Data Degraded; DT: Different Technique; NA: New Analysis; NG: New Guess; NP: New population; PL: Population Lost; RS: Repeat Survey (RS' denotes a repeat survey that is not statistically comparable for reasons such as different season); —: No Change

2 Key to Survey Types: AS: Aerial Sample Count; AT: Aerial Total Count; DC: Dung Count; GD: Genetic Dung Count; GS: Ground Sample Count; GT: Ground Total Count; IG: Informed Guess; IR: Individual Registration; OG: Other Guess. Survey Type is followed by an indicator of survey quality, ranked from 1 to 3 (best to worst). Survey Reliability is keyed A–E (best to worst)

3 PFS: Priority for Future Surveys, ranked from 1 to 5 (highest to lowest). Based on the precision of estimates and the proportion of national range accounted for by the site in question, PFS is a measure of the importance and urgency for future population surveys. All areas of unassessed range have a priority of 1. See Introduction for details on how the PFS is derived.

MOZAMBIQUE

General Statistics

Country area: 801,590 km2

Range area (% of country): 334,786 km2 (52%)

Protected area coverage (% of country): 7%

Protected range (% of known and possible range in protected areas): 15%

Information Quality Index (IQI): 0.48

CITES Appendix: I

Listing Year: 1989

Current Issues

A June 2005 survey in Maputo found considerable amounts of carved ivory openly for sale and export, including in the departure lounge of the international airport. It is believed that most of the ivory originates from within the country. Although the report found Mozambique's implementation of its obligations under CITES to be lacking, two weeks before the survey, the Government had initiated a campaign to encourage owners of ivory to register and regularize their stocks, and a number of ivory seizures were made thereafter (Milliken et al., 2006).

The Maputo Special Reserve is being expanded to link it to the Futi corridor and the Tembe Elephant Park in South Africa as a transfrontier conservation area. An electric fence will be erected along the western boundary of the enlarged Maputo Special Reserve, and the fence on the northern boundary of the Tembe Elephant Park will be removed (Peace Parks Foundation, 2006).

A translocation of 500 elephants from Chobe National Park (Botswana) to Gorongosa National Park (Mozambique) was planned for 2004, but did not take place.

In 2006, the United States Government denied applications for CITES import permits for a number of elephant trophies, originating from areas other than the Niassa Game Reserve, on the grounds that there was insufficient information on elephant populations in such areas to set rational trophy quotas (Jackson, 2006). In 2005 Mozambique increased its CITES export quota from 20 to 80 tusks (40 animals) (UNEPWCMC, 2006).

Range Data

Reliable knowledge on the distribution of elephants in Mozambique is scanty in many areas. The country's most important population is in the far north, in and around the Niassa Game Reserve. Smaller known populations occur in the west and far south.

Two new areas of known range have been added to the north and to the east of Limpopo National Park, where elephants are known to be moving to (Anderson, quest. reply, 2005). Based on information provided by the same source, the area of range in Banhine National Park has been categorized as doubtful.

A total of 20 areas across the country have been categorized as doubtful range based on data from the Landscan ambient human population data set (ORNL/GIST, 2002). All areas re-categorized are estimated to have a human population density of at least 15 persons per km2, which makes elephant presence unlikely (see Introduction section for details). These changes reflect better information rather than recent real changes in elephant distribution. The Gilé Game Reserve, incorrectly shown as non-range in the previous report, is now depicted as known range.

SUMMARY TOTALS FOR MOZAMBIQUE

Data Category DEFINITE PROBABLE POSSIBLE SPECULATIVE
Aerial or Ground Total Counts 630 0 0 0
Direct Sample Counts and Reliable Dung Counts 13,427 2,396 2,396 0
Informed Guesses 22 0 237 68
Other Guesses 0 0 0 6,912
TOTALS2006 14,079 2,396 2,633 6,980
TOTALS 2002 11,647 2,786 3,073 6,902

INTERPRETATION OF CHANGES IN ESTIMATES FROM PREVIOUS REPORT

Cause of Change DEFINITE PROBABLE POSSIBLE SPECULATIVE
Repeat Survey +428 +749 +1,032 0
Different Technique +730 0 +97 -4
Different Area +1,254 -1,139 -1,569 0
New Guess +19 0 0 +82
TOTAL CHANGE +2,432 -390 -440 +78

AREA OF RANGE COVERED BY EACH DATA CATEGORY (km2)

Data Category Known Range Possible Range Total Range
Aerial or Ground Total Counts 10,786 0 10,786
Direct Sample Counts and Reliable Dung Counts 48,309 0 48,309
Informed Guesses 4,831 2,592 7,423
Other Guesses 123,197 66,794 189,991
Unassessed Range 33,853 44,423 78,276
TOTAL 220,977 113,809 334,786

MOZAMBIQUE: ELEPHANT ESTIMATES

INPUT ZONE CAUSE OF CHANGE1 SURVEY DETAILS2 NUMBER OF ELEPHANTS SOURCE PFS3 AREA (km2) MAP LOCATION
TYPE RELIAB. YEAR ESTIMATE 95% C.L. LON. LAT.
Banhine National Park NG IG3 D 2005 0 Anderson, quest. reply, 2005 6,510 32.9 E 22.8 S
Cabo Delgado Province OG3 E 1998 670 DNFFB, 1999 1 43,780 39.2 E 12.4 S
Cabora Bassa North DA AS2 B 2003 1,718 807 Dunham, 2004a 3 3,708 30.7 E 15.4 S
Gilé Game Reserve IG3 D 2002 15 18* C.P. Ntumi, pers. comm., 2003 3 2,100 38.4 E 16.6 S
Gorongosa National Park NG IG3 D 2004 22 Dunham, 2004b 3 3,689 34.3 E 18.8 S
Inhambane Province OG3 E 1998 260 DNFFB, 1999 3 2,235 34.0 E 22.7 S
Limpopo National Park DT AT2 A 2006 630 Whyte, 2006 2 10,000 31.9 E 23.3 S
Magoe District RS AS2 B 2003 1,628 794 Dunham, 2004a 3 2,621 30.7 E 15.9 S
Manica Province OG3 E 1998 260 DNFFB, 1999 1 36,441 33.5 E 19.1 S
Maputo GR & Futi Corridor IG3 D 2002 200 50* R. Morley, pers. comm., 2002 3 900 32.7 E 26.6 S
Mecuburi Forest Reserve OG3 E 2002 5 5* Anderson, quest. reply, 2002 4 195 39.0 E 14.3 S
Moribane-Chimanimani Forest Reserve IG3 D 2002 22 C.P. Ntumi, pers. comm., 2003 4 185 33.4 E 19.5 S
Niassa Game Reserve & Buffer Zone RS AS2 B 2004 12,477 2,111 Craig & Gibson, 2004 2 42,612 37.2 E 12.1 S
Quirimbas National Park NG OG3 E 2005 2,000 Cumming & Jones, 2005 2 7,845 40.0 E 12.5 S
Sofala Province OG3 E 1999 800 200* DNFFB, 1999 2 33,138 34.5 E 18.8 S
Tete Province OG3 E 1998 2,260 1650* DNFFB, 1999 1 63,581 32.5 E 15.5 S
Zambezia Province OG3 E 1998 657 DNFFB, 1999 1 01,289 35.9 E 17.8 S

* Range of informed guess

1 Key to Causes of Change: DA: Different Area; DD: Data Degraded; DT: Different Technique; NA: New Analysis; NG: New Guess; NP: New population; PL: Population Lost; RS: Repeat Survey (RS' denotes a repeat survey that is not statistically comparable for reasons such as different season); —: No Change

2 Key to Survey Types: AS: Aerial Sample Count; AT: Aerial Total Count; DC: Dung Count; GD: Genetic Dung Count; GS: Ground Sample Count; GT: Ground Total Count; IG: Informed Guess; IR: Individual Registration; OG: Other Guess. Survey Type is followed by an indicator of survey quality, ranked from 1 to 3 (best to worst). Survey Reliability is keyed A–E (best to worst)

3 PFS: Priority for Future Surveys, ranked from 1 to 5 (highest to lowest). Based on the precision of estimates and the proportion of national range accounted for by the site in question, PFS is a measure of the importance and urgency for future population surveys. All areas of unassessed range have a priority of 1. See Introduction for details on how the PFS is derived.

Population Data

An aerial sample count of Niassa Game Reserve and Buffer Zone (Craig & Gibson, 2004) estimated 12,477 ± 2,111 elephants. Although this is lower than the estimate of 13,061 ± 2,433 (Craig & Gibson, 2002) which it replaces, the difference between the two estimates is not statistically significant, and the low carcass counts in both surveys is indicative of a secure population.

Much of the Cabora Bassa North area was surveyed for the first time in 2003 as part of a cross-border aerial sample count covering the Zambezi Valley between Lakes Kariba and Cabora Bassa (Dunham, 2004a). The same survey also covered the Magoe South District, and the estimate for this area replaces a 2002 estimate of 1,264 ± 1,359 from a similar survey (Mackie, 2001). These areas were surveyed again in 2005, but a survey report had not been produced at the time of writing.

The Limpopo National Park was systematically surveyed for the first time in 2006, as part of the regular aerial counts that cover Kruger National Park in South Africa. The estimate of 630 from this survey (Whyte, 2006) replaces an informed guess of 150 to 200 (Anderson, quest. reply, 2002).

An aerial sample count of Gorongosa National Park conducted in 2004 failed to detect any elephants within the search strips, but a herd of 20–22 and a single bull were seen when flying back to camp (Dunham, 2004b). These sightings have been entered as an informed guess in the table of estimates.

The previous estimate of 8 for Banhine National Park (Anderson, quest. reply, 2002) has been replaced by a new guess of zero, as elephants do not currently occur there (Anderson, quest. reply, 2005). All other estimates have been retained from the previous report.

The number of elephants in the definite category has increased by nearly 2,450, largely due to the inclusion of the considerably larger area covered by the Cabora Bassa (Magoe) North survey and new survey estimates for Limpopo. An overall increase in precision is responsible for the decreases of around 400 in the probable and possible categories. Less than a fifth of Mozambique's elephant range is covered by good quality estimates, nearly 60% is covered by guesses, and the remaining quarter remains unassessed.

Cross-border Movements

Elephants in the Niassa Game Reserve are part of a transboundary population whose range includes the Selous Game Reserve in Tanzania (Mpanduji et al., 2002), and which constitutes one of the most important populations on the continent.

Although most of the fence separating Limpopo National Park and Kruger National Park in South Africa has yet to be removed, elephants are reported to be moving out of Kruger and into Limpopo and north of it out of their own volition (Anderson, quest. reply, 2005; Marshall, 2005; Peace Parks Foundation, 2006).

It is expected that elephant movement between the Maputo Special Reserve and Tembe Elephant Park in South Africa will resume once the electric fence along Tembe's northern boundary is removed (Peace Parks Foundation, 2006).

NAMIBIA

General Statistics

Country area: 825,418 km2

Range area (% of country): 146,921 km2 (18%)

Protected area coverage (% of country): 13%

Protected range (% of known and possible range in protected areas): 23%

Information Quality Index (IQI): 0.46

CITES Appendix: II

Listing Year: 1997

Current Issues

Namibia has developed a national management plan for elephants. The plan foresees the devolution of authority to landowners, both communal and private, to manage elephants on their land as a pre-requisite for the adoption of successful adaptive co-management. It also calls for range expansion and the removal of veterinary fences that restrict elephant movement, but does not rule out management intervention to reduce populations in the future (Ministry of Environment and Tourism, 2005c).

New conservancies are being created north of Etosha National Park in areas into which elephant range has been expanding. The expansion of elephant range in the northeast, due largely to immigration from Botswana is causing levels of human-elephant conflict to escalate (Martin, 2005b; Ministry of Environment and Tourism, 2005a).

Namibia's proposal to be allowed to export individually marked and certified traditional ivory amulets, known as ekipas, for non-commercial purposes was approved at the 13th Meeting of the Conference of the Parties to CITES, held in 2004 in Bangkok, Thailand (CITES, 2004). In January 2007 Namibia and Botswana submitted a joint proposal to CITES to maintain their elephant populations, as well as those of South Africa and Zimbabwe, in Appendix II, and to establish annual export quotas for these four countries to trade in raw ivory to approved trading partners (Government of Botswana & Government of Namibia, 2007).

Range Data

Elephants are only found in the north of the country. While large populations are found in Etosha National Park, Khaudom Game Reserve and the Caprivi region, Namibian elephants range widely in search of water depending on annual rainfall patterns, and have some of the largest home ranges recorded anywhere on the continent (Lindeque, 1995).

Few changes have been made to the range map for this report. Some areas of possible range in the Caprivi strip have been re-categorized as known range (Kolberg, 2004; Martin, 2005b), and known range has been slightly extended to the northwest of Etosha (C.R. Thouless, pers. comm., 2006) and to the west of Khaudom Game Park (Kolberg, 2004). Five elephants were seen in a survey of the N#a-Jaqna Conservancy, to the southwest of Khaudom, and this sighting is shown as a cross on the map (Kolberg, 2004).

Population Data

Aerial sample counts of all major elephant populations in Namibia were conducted between 2004 and 2005, as part of Namibia's regular survey programme. A 2004 survey of Etosha National Park estimated 2,057 ± 598 elephants in the park (Kilian & Kolberg, 2004). This replaces a 2002 aerial sample count estimate of 2,417 ± 663 (Kilian, 2003). The difference between the two estimates, however, is not statistically significant.

The estimate for Caprivi, on the other hand, has nearly doubled from the previous estimate of 4,576 ± 1,223 (Ministry of Environment and Tourism, 1998) to 8,725 ± 2,206 (Kolberg, 2004). This significant difference cannot be due to natural population increase alone and is most likely to have been influenced by elephant immigration from Botswana.

The estimate of 1,966 ± 973 for Khaudom/Nyae-Nyae featured in the previous report, which originated from an aerial sample count conducted in 2000 (Craig, 2003), has been replaced by the results of two simultaneous surveys conducted in Khaudom/Kavango and Nyae-Nyae by different survey teams in 2004 (Kolberg, 2004; Stander, 2004). While the Nyae Nyae estimate of 967 ± 481 from the 2004 survey is comparable in magnitude to the estimate of 755 ± 554 from the 2000 survey (Craig, 2003), the 2004 Khaudom/Kavango estimate is nearly three times higher than corresponding estimate from the 2000 survey. This difference cannot be accounted for by natural population increase alone. The area surveyed was larger in 2004 than in 2000, but only by 10%. Thus, the difference in area is unlikely to be a major contributor to the difference in the estimates. Elephants are reported to be moving from Botswana into Khaudom in increasing numbers (Martin, 2005b), and this could account for much of the difference.

An estimate of 210 ± 157 from a 2005 survey in Kunene (Ministry of Environment and Tourism, 2005b) replaces an estimate of 663 ± 790 from an aerial sample count conducted in 2000 (Craig, 2003). Although the area covered in the earlier survey was nearly three times larger, the core elephant range was covered in its entirety in the 2005 survey, and the difference between the two estimates is not statistically significant.

An informed guess of 20 for Mangetti Game Reserve (Martin, 2005b) replaces an aerial total count estimate of 19 (Lindeque et al., 1995), which would have otherwise been degraded to the category of other guesses, as the estimate is over 10 years old.

The increase in the definite category is largely the result of new estimates from methodologically comparable surveys, while an overall decrease in precision has resulted in increases in the probable and possible categories. It is worth stressing that all these changes are likely to have been influenced by elephant immigration from Botswana.

Nearly 45% of Namibia's elephant range remains unsurveyed. Although the much of the unsurveyed area is potentially important for elephant movements, only small numbers of elephants are likely to be found there at any one time.

Cross-border Movements

The Caprivi Strip in Namibia is an extension of the northern Botswana population, and provides an important corridor for the movement of elephants between Zimbabwe, Botswana, Zambia and Angola (Craig, 1996a). There may also be movement between other areas in northern Namibia and southern Angola. As elephant range expands in northern Botswana, elephants are moving in increasing numbers into Khaudom/Nyae-Nyae in Namibia (Martin, 2005b).

SUMMARY TOTALS FOR NAMIBIA

Data Category DEFINITE PROBABLE POSSIBLE SPECULATIVE
Direct Sample Counts and Reliable Dung Counts 12,531 3,276 3,276 0
Informed Guesses 0 0 20 0
TOTALS2006 12,531 3,276 3,296 0
TOTALS 2002 7,769 1,872 1,872 0

INTERPRETATION OF CHANGES IN ESTIMATES FROM PREVIOUS REPORT

Cause of Change DEFINITE PROBABLE POSSIBLE SPECULATIVE
Repeat Survey +4,760 +2,579 +3,157 0
New Population +4 +159 +402 0
Different Area +13 -1,334 -2,205 0
New Guess -16 0 +70 0
TOTAL CHANGE +4,762 +1,404 +1,424 0

AREA OF RANGE COVERED BY EACH DATA CATEGORY (km2)

Data Category Known Range Possible Range Total Range
Direct Sample Counts and Reliable Dung Counts 73,567 6,964 80,531
Informed Guesses 359 0 359
Unassessed Range 15,082 50,949 66,031
TOTAL 89,008 57,913 146,921

NAMIBIA: ELEPHANT ESTIMATES

INPUT ZONE CAUSE OF CHANGE1 SURVEY DETAILS2 NUMBER OF ELEPHANTS SOURCE PFS3 AREA (km2) MAP LOCATION
TYPE RELIAB. YEAR ESTIMATE 95% C.L. LON. LAT.
Caprivi Region RS AS2 B 2004 8,725 2,206 Kolberg, 2004 2 17,943 23.5 E 17.9 S
Etosha National Park RS AS1 B 2004 2,057 598 Kilian & Kolberg, 2004 2 18,551 15.9 E 19.0 S
Khaudom -Kavango RS AS2 B 2004 3,787 2,289 Kolberg, 2004 2 10,485 20.8 E 18.6 S
Kunene DA AS1 B 2005 210 157 MET, 2005b 1 31,144 13.7 E 19.3 S
Mangetti Game Reserve NG IG3 D 2005 20 MET, 2005 3 762 19.1 E 18.7 S
N#a-Jaqna Conservancy NP AS2 B 2004 61 115 Kolberg, 2004 2 9,143 19.6 E 19.6 S
Nyae Nyae Conservancy RS AS1 B 2004 967 481 Stander, 2004 2 2,957 20.5 E 19.6 S

1 Key to Causes of Change: DA: Different Area; DD: Data Degraded; DT: Different Technique; NA: New Analysis; NG: New Guess; NP: New population; PL: Population Lost; RS: Repeat Survey (RS' denotes a repeat survey that is not statistically comparable for reasons such as different season); —: No Change

2 Key to Survey Types: AS: Aerial Sample Count; AT: Aerial Total Count; DC: Dung Count; GD: Genetic Dung Count; GS: Ground Sample Count; GT: Ground Total Count; IG: Informed Guess; IR: Individual Registration; OG: Other Guess. Survey Type is followed by an indicator of survey quality, ranked from 1 to 3 (best to worst). Survey Reliability is keyed A–E (best to worst)

3 PFS: Priority for Future Surveys, ranked from 1 to 5 (highest to lowest). Based on the precision of estimates and the proportion of national range accounted for by the site in question, PFS is a measure of the importance and urgency for future population surveys. All areas of unassessed range have a priority of 1. See Introduction for details on how the PFS is derived.

SOUTH AFRICA

General Statistics

Country area: 1,219,912 km2

Range area (% of country): 30,455 km2 (2%)

Protected area coverage (% of country): 4%

Protected range (% of known and possible range in protected areas): 85%

Information Quality Index (IQI): 0.96

CITES Appendix: II

Listing Year: 2000

Current Issues

As elephant populations in South Africa continue to grow, arguments between those in favour of the resumption of culling and those against it have become increasingly heated (Cumming & Jones, 2005). A consultative process, convened by South African National Parks (SANParks) in 2004, reviewed the opinions of scientists and stakeholders on both sides of the debate (South African National Parks, 2004a, 2005). In 2005, SANParks put forward a recommendation to the Minister of Environmental Affairs and Tourism that culling be resumed (Mabunda, 2005), as envisaged by the Policy for Elephant Management in Kruger National Park (South African National Parks, 2004b). A decision was postponed pending further consultation with an international panel of elephant experts. In 2006, the panel issued a statement of scientific consensus which, rather than supporting a particular management action, recommended the establishment of a long-term research programme to better understand the consequences of any management action that may be taken.

Range Data

South Africa's elephants are confined to protected areas and private reserves, largely in the north and east of the country, although some populations remain – and others are being newly established – in the far south. The largest portion of elephant range falls within and around Kruger National Park, from where most of the elephants in populations elsewhere in the country have been translocated.

Some new areas of known range have been added to the AED as a result of recent translocations and new information. These include the Kapama, Kwandwe, Kariega, Shamwari, Lalibela, Mthetomusha and Great Fish River Game Reserves.

Population Data

A complete update of all populations in South Africa has been obtained from SANParks (Whyte, 2006) and the Elephant Management and Owners Association (Elephant Management and Owners Association, 2005). Most surveys in South Africa are aerial total counts conducted from helicopters, and therefore appear under the definite category in the summary table. The estimates for some reserves for which detailed census methods could not be obtained have been classified as informed guesses. At the request of their owners, private reserves holding 50 or fewer elephants are grouped under one overall estimate and are not shown on the map.

The elephant population in Kruger National Park and its environs has continued to grow in recent years. At the time of the 2005 census, the population in the park stood at 12,467 elephants (I.J. Whyte, pers. comm., 2005), up by nearly 2,000 from the figure of 10,459 featured in the previous report. The slightly lower estimate of 12,427 from the 2006 survey (Whyte, 2006), featured in this report, is likely to have resulted from net emigration to neighbouring private reserves and the Limpopo National Park in Mozambique. The population in the entire Kruger ecosystem, including the surrounding private reserves, increased from 12,439 in 2002, to 14,735 in 2005 (I.J. Whyte, pers. comm., 2005), and then to 15,387 in 2006 (Whyte, 2006).

The population at Addo Elephant National Park has also continued to increase, from 337 in 2002 to 459 in 2005. As a result of higher estimates from these and other populations, the number of elephants in the definite category has increased by nearly 3,800 compared to the previous report.

Cross-border Movements

Movement of elephants out of Kruger National Park into Mozambique is reported to have increased recently (Anderson, quest. reply, 2005), facilitated by the removal of an additional 30 km of the fence that separates Kruger from Limpopo National Park in Mozambique.

Elephants from Botswana and possibly Zimbabwe have moved into the Mapungubwe National Park, raising some concern about their potential impact on woodlands (South African National Parks, 2006).

SUMMARY TOTALS FOR SOUTH AFRICA

Data Category DEFINITE PROBABLE POSSIBLE SPECULATIVE
Aerial or Ground Total Counts 17,501 0 0 0
Informed Guesses 346 0 638 22
TOTALS2006 17,847 0 638 22
TOTALS 2002 14,071 0 855 0

INTERPRETATION OF CHANGES IN ESTIMATES FROM PREVIOUS REPORT

Cause of Change DEFINITE PROBABLE POSSIBLE SPECULATIVE
Repeat Survey +3,489 0 0 0
New Population +180 0 0 0
Different Technique +71 0 -100 0
New Guess +36 0 -51 +22
Population Lost 0 0 -66 0
TOTAL CHANGE +3,776 0 -217 +22

AREA OF RANGE COVERED BY EACH DATA CATEGORY (km2)

Data Category Known Range Total Range
Aerial or Ground Total Counts 29,204 29,204
Informed Guesses 1,251 1,251
TOTAL 30,455 30,455

SOUTH AFRICA: ELEPHANT ESTIMATES

INPUT ZONE CAUSE OF CHANGE1 SURVEY DETAILS2 NUMBER OF ELEPHANTS SOURCE PFS3 AREA (km2) MAP LOCATION
TYPE RELIAB. YEAR ESTIMATE 95% C.L. LON. LAT.
Addo Elephant National Park RS IR1 A 2005 459 EMOA, 2005 2 1,250 25.5 E 33.3 S
Atherstone Nature Reserve RS AT3 A 2005 60 EMOA, 2005 3 136 26.8 E 24.5 S
Balule Nature Reserve RS AT3 A 2006 457 Whyte, 2006 3 400 31.0 E 24.2 S
Great Fish River Reserve Complex NP IR1 A 2005 2 EMOA, 2005 3 440 26.8 E 33.1 S
Greater St Lucia Wetland Park (Managed Nature Reserve) RS AT3 A 2005 45 EMOA, 2005 3 539 32.5 E 27.9 S
Hluhluwe-Imfolozi Game Reserve NG IG3 D 2004 346 22* EMOA, 2005 2 965 31.9 E 28.3 S
Itala Nature Reserve RS GT1 A 2005 84 EMOA, 2005 3 297 31.3 E 27.5 S
Kapama Private Game Reserve NP IR1 A 2005 36 EMOA, 2005 3 246 31.1 E 24.4 S
Kariega Private Game Reserve NP IR1 A 2005 11 EMOA, 2005 3 190 26.7 E 33.5 S
Klaserie Private Nature Reserve RS AT2 A 2006 569 Whyte, 2006 2 628 31.2 E 24.2 S
Knysna Forest Reserve RS IR1 A 2005 4 EMOA, 2005 3 126 23.0 E 34.0 S
Kruger National Park RS AT2 A 2006 12,427 Whyte, 2006 1 19,624 31.5 E 24.0 S
Kwandwe Private Game Reserve NP IR1 A 2005 29 EMOA, 2005 3 158 26.6 E 33.1 S
Lalibela Private Game Reserve NP IR1 A 2005 11 EMOA, 2005 3 75 26.2 E 33.5 S
Letaba Game Ranch PL GT1 A 2006 0 Whyte, 2006 420 31.1 E 23.7 S
Madikwe Nature Reserve RS AT3 A 2005 455 EMOA, 2005 2 700 26.3 E 24.8 S
Makalali Private Game Reserve RS GT1 A 2005 72 EMOA, 2005 3 140 30.7 E 24.2 S
Makuya National Park RS AT2 A 2006 54 Whyte, 2006 3 165 30.9 E 22.6 S
Manyeleti Game Reserve DT AT2 A 2006 71 Whyte, 2006 3 228 31.5 E 24.6 S
Marakele National Park RS IR1 A 2005 110 EMOA, 2005 3 380 27.6 E 24.4 S
Mkuzi Game Reserve RS IR1 A 2005 37 EMOA, 2005 3 380 32.3 E 27.7 S
Mthetomusha Nature Reserve NP IR1 A 2005 30 EMOA, 2005 3 80 31.3 E 25.5 S
Phalaborwa Mining Company RS AT2 A 2006 77 Whyte, 2006 4 41 31.2 E 24.0 S
Pilanesberg National Park RS AT3 A 2005 140 EMOA, 2005 3 553 27.1 E 25.2 S
Pongola Nature Reserve RS IR1 A 2005 55 EMOA, 2005 3 119 32.0 E 27.4 S
Private Reserves NG IG3 D 2005 578 EMOA, 2005 1 4,000 Not Shown
Sabie Sands Game Reserve RS AT2 A 2006 857 Whyte, 2006 3 572 31.5 E 24.8 S
Selati Game Reserve RS GT1 A 2005 85 EMOA, 2005 3 300 30.8 E 24.0 S
Shamwari Game Reserve NP IR1 A 2005 61 EMOA, 2005 3 200 26.1 E 33.4 S
Songimvelo Game Reserve NG IG3 D 2005 60 EMOA, 2005 2 490 31.0 E 26.0 S
Tembe Elephant Park RS IR1 A 2005 167 Morley, 2005 3 300 32.5 E 26.9 S
Timbavati Private Nature Reserve RS AT2 A 2006 712 Whyte, 2006 3 494 31.3 E 24.4 S
Umbabat Private Nature Reserve RS AT2 A 2006 163 Whyte, 2006 3 144 31.4 E 24.1 S
Venetia-Limpopo National Park RS AT3 A 2005 61 EMOA, 2005 3 91 29.3 E 22.2 S
Welgevonden Private Game Reserve RS AT3 A 2005 100 EMOA, 2005 3 330 27.8 E 24.3 S

* Range of informed guess

1 Key to Causes of Change: DA: Different Area; DD: Data Degraded; DT: Different Technique; NA: New Analysis; NG: New Guess; NP: New population; PL: Population Lost; RS: Repeat Survey (RS' denotes a repeat survey that is not statistically comparable for reasons such as different season); —: No Change

2 Key to Survey Types: AS: Aerial Sample Count; AT: Aerial Total Count; DC: Dung Count; GD: Genetic Dung Count; GS: Ground Sample Count; GT: Ground Total Count; IG: Informed Guess; IR: Individual Registration; OG: Other Guess. Survey Type is followed by an indicator of survey quality, ranked from 1 to 3 (best to worst). Survey Reliability is keyed A–E (best to worst)

3 PFS: Priority for Future Surveys, ranked from 1 to 5 (highest to lowest). Based on the precision of estimates and the proportion of national range accounted for by the site in question, PFS is a measure of the importance and urgency for future population surveys. All areas of unassessed range have a priority of 1. See Introduction for details on how the PFS is derived.

SWAZILAND

General Statistics

Country area: 17,360 km2

Range area (% of country): 50 km2 (1%)

Protected area coverage (% of country): 5%

Protected range (% of known and possible range in protected areas): 81%

Information Quality Index (IQI): 1.00

CITES Appendix: I

Listing Year: 1989

Current Issues

Conservation priorities in the Hlane and Mkhaya parks concentrate on the black rhino population and treenesting vultures, and the management aims to minimize impact on these species by controlling elephant numbers (Reilly, quest. reply, 2005).

Three elephants in Malolotja Nature Reserve, originally from the adjacent Songimvelo Nature Reserve in South Africa, are reported to be coming into conflict with communities to the east of the reserve (Reilly, quest. reply, 2005) and to be causing some impact on the woodland within Malolotja (Mtui & Owen-Smith, 2006).

There is ongoing controversy between the Swaziland National Trust Commission (SNTC), Yonge Nawe – a local environmental NGO – and Big Game Parks of Swaziland (BGP), which manages the Hlane and Mkhaya parks. SNTC and Yonge Nawe have challenged the status of BGP as Swaziland's delegated CITES national management and scientific authority and have questioned plans for the enlargement of Mkhaya (Douglas Consulting & LKM, 2004; Reilly, quest. reply, 2005).

Range Data

Elephant distribution is well understood in Swaziland, being mainly restricted to fenced enclosures within Hlane Royal National Park and Mkhaya Game Reserve. The enclosures only occupy a fraction of the reserves (6% and 19% respectively).

Three elephants from Songimvelo Game Reserve in South Africa use the Komati Valley in the Malolotja Nature Reserve as part of their range (Reilly, quest. reply, 2005). This area has been added to the map as known range.

Population Data

Individual registration of all elephants is maintained for the Hlane and Mkhaya populations. Eleven elephants from these two reserves were exported to zoos in the United States in 2003 (Reilly, quest. reply, 2005). The consequent reduction in the population figures from the 39 reported in the AESR 2002 is reflected in the summary table. An estimate of three elephants from Malolotja Nature Reserve has been added to the table of estimates (Reilly, quest. reply, 2005).

Cross-border Movements

In January 2005 an elephant from the Pongola Nature Reserve in South Africa crossed into Swaziland near Lavumisa and moved north past Maloma to Kubuta, where it turned back and returned to Pongola after being radio-collared in Swaziland (Reilly, quest. reply, 2005).

SUMMARY TOTALS FOR SWAZILAND

Data Category DEFINITE PROBABLE POSSIBLE SPECULATIVE
Aerial or Ground Total Counts 28 0 0 0
Informed Guesses 3 0 0 0
TOTALS2006 31 0 0 0
TOTALS 2002 39 0 0 0

INTERPRETATION OF CHANGES IN ESTIMATES FROM PREVIOUS REPORT

Cause of Change DEFINITE PROBABLE POSSIBLE SPECULATIVE
Repeat Survey -11 0 0 0
New Population +3 0 0 0
TOTAL CHANGE -8 0 0 0

AREA OF RANGE COVERED BY EACH DATA CATEGORY (km2)

Data Category Known Range Total Range
Aerial or Ground Total Counts 22 22
Informed Guesses 28 28
TOTAL 50 50

SWAZILAND: ELEPHANT ESTIMATES

INPUT ZONE CAUSE OF CHANGE1 SURVEY DETAILS2 NUMBER OF ELEPHANTS SOURCE PFS3 AREA (km2) MAP LOCATION
TYPE RELIAB. YEAR ESTIMATE 95% C.L. LON. LAT.
Hlane Royal National Park RS IR1 A 2005 13 Reilly, quest. reply, 2005 1 142 31.9 E 26.3 S
Malolotja Nature Reserve NP IG3 D 2005 3 Reilly, quest. reply, 2005 1 28 31.1 E 26.0 S
Mkhaya Nature Reserve RS IR1 A 2005 15 1 65 31.7 E 26.6 S

1 Key to Causes of Change: DA: Different Area; DD: Data Degraded; DT: Different Technique; NA: New Analysis; NG: New Guess; NP: New population; PL: Population Lost; RS: Repeat Survey (RS' denotes a repeat survey that is not statistically comparable for reasons such as different season); —: No Change

2 Key to Survey Types: AS: Aerial Sample Count; AT: Aerial Total Count; DC: Dung Count; GD: Genetic Dung Count; GS: Ground Sample Count; GT: Ground Total Count; IG: Informed Guess; IR: Individual Registration; OG: Other Guess. Survey Type is followed by an indicator of survey quality, ranked from 1 to 3 (best to worst). Survey Reliability is keyed A–E (best to worst)

3 PFS: Priority for Future Surveys, ranked from 1 to 5 (highest to lowest). Based on the precision of estimates and the proportion of national range accounted for by the site in question, PFS is a measure of the importance and urgency for future population surveys. All areas of unassessed range have a priority of 1. See Introduction for details on how the PFS is derived.

ZAMBIA

General Statistics

Country area: 752,610 km2

Range area (% of country): 201,247 km2 (28%)

Protected area coverage (% of country): 31%

Protected range (% of known and possible range in protected areas): 77%

Information Quality Index (IQI): 0.47

CITES Appendix: I

Listing Year: 1989

Current Issues

In 2003 Zambia developed a national Elephant Policy and Action Plan (Ministry of Tourism, Environment and Natural Resources, 2003). The policy identifies human-elephant conflict and poaching as the major threats to elephant populations in Zambia, and proposes improved land use planning, decentralized decision-making processes and revenue sharing from the sustainable use of elephants as the principal means to mitigate these problems.

As envisaged by the policy, and after a ban of 21 years, sport hunting of elephants was reopened in August 2005, and Zambia notified the CITES Secretariat that it would maintain an annual export quota of 40 tusks (20 animals) as hunting trophies (UNEP-WCMC, 2006).

Range Data

Elephants in Zambia are distributed in four major populations, namely in the Luangwa Valley, the Kafue ecosystem and nearby West Lunga, the Lower Zambezi Valley and Sioma Ngwezi and its environs. A number of smaller populations are scattered along the country's borders.

The shape of the range map for Zambia has not changed substantially since the last report, but a number of areas have been categorized as doubtful range. These include an area along the shores of Lake Kariba (G.C. Craig, pers. comm., 2006; Ministry of Tourism, Environment and Natural Resources, 2003), a strip adjacent to the Mukungule Game Management Area (E. van der Westhuizen, pers. comm., 2006), and two areas on the eastern side of the Luangwa Valley, where human population density is estimated to exceed 15 persons per km2 (ORNL/GIST, 2002; see Introduction section for details on rationale).

Population Data

Most of the Luangwa Valley has been systematically surveyed in the last five years. Two aerial sample counts were conducted in North Luangwa National Park and surrounding areas, in 2003 (van der Westhuizen, 2003) and 2005. The result from the latter survey, however, is believed to be a considerable overestimate, perhaps caused by the use of a different survey crew (E. van der Westhuizen, pers. comm., 2006). For this reason, the results of the 2003 survey have been used in this report, replacing an aerial total count conducted in 2000 (Aucamp, 2000). The elephant population in North Luangwa is currently believed to be stable or increasing (E. van der Westhuizen, pers. comm., 2006).

A 2004 aerial sample count covered several game management areas in the Luangwa Valley (Simwanza, 2004b). This survey excluded the Luambe National Park, last surveyed by Jachmann (1999a), when it was counted as a single block together with the southern sector of Lumimba Game Management Area. This southern sector was included in the 2004 survey and appears on the table as Mwanya hunting block. Despite occasional sightings, elephants have been largely absent from Luambe National Park for many years. For this reason, the 1999 estimate for Luambe has been removed from the table of estimates. There are nevertheless reports that elephants are beginning to return to Luambe, and it would be important for this park to be systematically surveyed as part of an ecosystem-wide survey of the Luangwa Valley.

SUMMARY TOTALS FOR ZAMBIA

Data Category DEFINITE PROBABLE POSSIBLE SPECULATIVE
Direct Sample Counts and Reliable Dung Counts 16,229 5,899 5,899 0
Other Dung Counts 27 49 9 0
Informed Guesses 306 0 0 0
Other Guesses 0 0 0 813
TOTALS2006 16,562 5,948 5,908 813
TOTALS 2002 12,457 6,961 7,631 235

INTERPRETATION OF CHANGES IN ESTIMATES FROM PREVIOUS REPORT

Cause of Change DEFINITE PROBABLE POSSIBLE SPECULATIVE
Repeat Survey +1,953 +37 -45 0
New Population +36 -85 +179 0
Different Technique -41 +1,100 -1,632 0
Different Area +1,806 -2,029 -231 0
New Guess +326 0 0 +119
New Analysis +29 -37 +5 +477
Data Degraded -3 0 0 -19
TOTAL CHANGE +4,105 -1,013 -1,723 +578

AREA OF RANGE COVERED BY EACH DATA CATEGORY (km2)

Data Category Known Range Possible Range Total Range
Direct Sample Counts and Reliable Dung Counts 109,944 5,143 115,087
Other Dung Counts 482 0 482
Informed Guesses 75 0 75
Other Guesses 6,851 734 7,585
Unassessed Range 1,204 76,814 78,018
TOTAL 118,556 82,692 201,247

ZAMBIA: ELEPHANT ESTIMATES

INPUT ZONE CAUSE OF CHANGE1 SURVEY DETAILS2 NUMBER OF ELEPHANTS SOURCE PFS3 AREA (km2) MAP LOCATION
TYPE RELIAB. YEAR ESTIMATE 95% C.L. LON. LAT.
Chanjuzi Hunting Block RS AS3 B 2004 65 81 Simwanza, 2004b 3 600 32.6 E 12.3 S
Chiawa Game Management Area RS AS2 B 2003 45 53 Dunham, 2004a 3 900 28.9 E 15.8 S
Chisomo & Sandwe Game Management Area AS2 B 1999 128 155 Jachmann & Phiri, 1999a 3 750 30.9 E 13.8 S
Isangano National Park DD GT1 E 1993 3 Tembo, quest. reply, 1993 3 840 30.6 E 11.2 S
Kafinde Game Management Area IG3 E 1991 50 Tembo, quest. reply, 1993 2 3,860 30.1 E 12.4 S
Kafue National Park DA AS2 B 2004 6,306 5,227 Simwanza, 2004a 2 22,400 25.9 E 15.2 S
Kasanka National Park NA DC2 C 1999 76 9 Jachmann & Phiri, 1999b 3 390 30.2 E 12.6 S
Kasonso-Busanga Game Management Area RS AS2 B 2004 401 378 Simwanza, 2004a 2 7,780 25.6 E 14.1 S
Katokota Game Ranch AT3 E 1991 19 Tembo, quest. reply, 1993 5 15 28.0 E 16.8 S
Lavushi Manda National Park IG3 E 1991 15 Tembo, quest. reply, 1993 3 1,500 30.8 E 12.4 S
Lower Zambezi National Park RS AS2 B 2003 1,477 744 Dunham, 2004a 2 4,084 29.7 E 15.5 S
Luano Game Management Area IG3 E 1996 150 Jachmann, 1996 2 8,930 29.6 E 14.8 S
Luawata Hunting Block RS AS2 B 2003 968 652 van der Westhuizen, 2003 3 1,092 31.9 E 12.3 S
Lukusuzi National Park RS AS1 B 2005 0 0 Fourie et al., 2005 3,200 32.6 E 12.8 S
Lunga-Luswishi Game Management Area RS AS2 B 2004 195 169 Simwanza, 2004a 2 13,340 26.9 E 13.8 S
Lupande Game Management Area AS2 B 2002 975 586 Dunham & Simwanza, 2002 2 4,959 32.0 E 13.3 S
Mosi-oa-Tunya National Park NG IG3 D 2006 306 Chase, 2006 4 66 25.8 E 17.9 S
Mukungule Game Management Area RS AS2 B 2003 156 119 van der Westhuizen, 2003 3 788 32.0 E 11.7 S
Mulobezi Game Management Area RS AS2 B 2004 55 96 Simwanza, 2004a 2 3,420 25.4 E 16.5 S
Mumbwa Game Management Area DT AS2 B 2004 181 208 Simwanza, 2004a 2 3,370 26.4 E 15.1 S
Musalangu Game Management Area DA AS2 B 2004 1,011 898 Simwanza, 2004b 3 2,190 32.8 E 11.2 S
Mwanya Hunting Block DA AS3 B 2004 503 237 Simwanza, 2004b 3 860 32.3 E 12.7 S
Mweru wa Ntipa Ecosystem NP AS2 B 2003 0 Simwanza, 2003 7,274 29.8 E 8.7 S
Namwala Game Management Area DT AS2 B 2004 127 134 Simwanza, 2004a 2 3,600 26.3 E 15.5 S
Nchete Island Wildlife Sanctuary AT3 E 1991 49 Tembo, quest. reply, 1993 4 25 27.6 E 17.4 S
Nkala Game Management Area NP AS2 B 2004 210 306 Simwanza, 2004a 4 194 26.0 E 16.0 S
North Luangwa National Park RS AS3 B 2003 3,235 695 van der Westhuizen, 2003 2 4,688 32.2 E 11.9 S
Nsumbu National Park DT AS2 B 2003 65 92 Simwanza, 2003 3 2,063 30.4 E 8.8 S
Nyampala Game Management Area DA AS2 B 2004 284 133 Simwanza, 2004b 4 330 31.7 E 12.5 S
Rufunsa Game Management Area RS AS2 B 2003 0 Dunham, 2004a 3,128 30.0 E 15.2 S
Sekula Island Wildlife Sanctuary AT3 E 1991 7 Tembo, quest. reply, 1993 5 10 27.5 E 17.4 S
Sichifulo Game Management Area RS AS2 B 2004 0 Simwanza, 2004a 3,600 25.7 E 16.8 S
Sioma Ngwezi National Park RS AS2 B 2005 385 371 Chase & Griffin, 2005b 2 4,377 23.4 E 17.3 S
South Luangwa National Park AS2 B 2002 4,459 1,519 Dunham & Simwanza, 2002 2 8,448 31.6 E 13.1 S
West Lunga National Park AS3 E 1996 520 C.M. Phiri, pers. comm., 1998 3 1,684 24.8 E 12.8 S
West Petauke Game Management Area AS2 B 1999 897 1,399 Jachmann & Phiri, 1999a 3 905 30.3 E 14.3 S

South Luangwa National Park has not been surveyed since 2002, and the estimate shown on the table has been retained from the previous report. A 2005 aerial survey of Lukusuzi National Park failed to find any elephants in the park (Fourie et al., 2005), and an estimate of zero appears on the table.

An aerial sample count was conducted in the Kafue ecosystem in 2004 (Simwanza, 2004a). Estimates from this survey replace a number of ground and aerial sample counts conducted between 1997 and 2001 (Fairall & Kampamba, 2001; Jachmann, 2000; Zyambo, 1997). Another aerial sample count covering much of Kafue National Park, conducted in two stages between September and November 2004, gave a combined estimate of 1510 ± 61 (van Aarde & Guldemond, 2004; van Aarde et al., 2004). Elephant density for the extreme northern sector of the park, which was not covered in the survey, was extrapolated from the rest of the survey area to yield an overall estimate of 1,738 ± 355. This survey, however, suffered from technical and design limitations, and has not been used for this report.

The Lower Zambezi National Park was surveyed, together with its surrounding Game Management Areas, as part of a 2003 transboundary survey that also covered adjacent areas in Mozambique and Zimbabwe (Dunham, 2004a). The results of this survey replace estimates from methodologically comparable surveys conducted by Phiri (1996). These areas were surveyed again in 2005, but a survey report had not been produced at the time of writing.

Three aerial sample counts were conducted in Sioma Ngwezi National Park since the last report: one in January 2004, which returned an estimate of 1,212 ± 920 (Chase et al., 2004), another in August 2004 (899 ± 755) and a third in November 2005 (385 ± 389) (Chase & Griffin, 2005b). Despite being lower than the previous two, the estimate from the most recent (2005) survey is shown in the table of estimates. The use of this result, which replaces an estimate of 250 (Mwiya, 1996) is justified by the fact that none of the differences between the estimates from the last three surveys are statistically significant. This lack of significance is due to the wide confidence limits in all three estimates, which may have been brought about by elephant distribution clustering around the centre of the park (Chase et al., 2004).

The Mweru wa Ntipa ecosystem was surveyed in its entirety for the first time in 2003 (Simwanza, 2003). The survey also covered the Nsumbu National Park, the population of which had last been estimated at 45 in 1998 (L. Saiwana, pers. comm., 1998).

The number of elephants in the definite category has increased by over 4,100 from the previous report, whereas the probable and possible categories have declined by about 1,000 and 1,700 respectively. These changes arise from methodologically comparable surveys, as well as from surveys conducted over different areas or using different techniques. While nearly 60% of Zambia's estimated elephant range is currently covered by good quality estimates, no figures are available for most of the remaining range. It is likely, however, that a more detailed knowledge of elephant presence around protected areas will cause the proportion of unassessed range to decline.

Cross-border Movements

Elephant range in southwestern Zambia is contiguous with range in Angola's Luiana Reserve (Chase & Griffin, 2005b) and with northern Botswana through the Caprivi Strip in Namibia. Elephant movement also occurs between Zambia, Zimbabwe and Mozambique in the Zambezi Valley. It is not known whether movement continues between the North Luangwa ecosystem and Kasungu National Park in Malawi.

ZIMBABWE

General Statistics

Country area: 390,580 km2

Range area (% of country): 76,931 km2 (29%)

Protected area coverage (% of country): 13%

Protected range (% of known and possible range in protected areas): 58%

Information Quality Index (IQI): 0.91

CITES Appendix: II

Listing Year: 1997

Current Issues

The Department of National Parks and Wildlife Management completed its transition into the present National Parks and Wildlife Management Authority (NPWMA) in 2004. The new authority is no longer funded from the national budget and must raise its own revenue. Major cuts in NPWMA's operating budget have been necessitated by a decline in tourist revenues in recent years, and this has resulted in a decreased capacity to manage elephant populations.

Three consecutive droughts between 2002 and 2005 caused a number of elephants in Hwange and other areas to die of starvation (Dunham et al., 2006b). This, together with reports of an increasing incidence in human-elephant conflict, kindled the debate on whether elephant numbers should be reduced through culling.

Poaching in the Sebungwe region is reported to have increased in recent years. A recent survey of the area recorded a 2.5-fold increase in the number of dead elephants since 2001, suggesting a notable increase in elephant mortality in the last five years. A high incidence in illegal activity was also recorded in parts of survey zone, suggesting that illegal killing is at least partly responsible for the estimated increase in elephant mortality (Dunham et al., 2006a).

Range Data

The majority of Zimbabwe's elephants are found in and around protected areas along the borders with neighbouring countries. There are four major populations, namely Northwest Matabeleland, Sebungwe, the Zambezi Valley and Gonarezhou. Most of the central highveld and eastern highlands are extensively settled and farmed, and have long been devoid of elephants.

All areas previously depicted as possible range have been categorized as doubtful based on Landscan 2002 human population density data (ORNL/GIST, 2002; see Introduction section for rationale) and information provided by C. Craig (2006). The area corresponding to Matibi II communal lands was incorrectly depicted as range in previous reports, but now appears as doubtful range. No other changes have been made to the range map.

Population Data

A survey of Northwest Matabeleland in 2006 could not be completed due to technical difficulties, and only covered 65% of the area surveyed in previous counts (Dunham et al., 2006b). The estimate for the areas covered, which excluded the northern sectors of Hwange National Park and the Matetsi complex, was 25,087 ± 5,301. This is not significantly different from the estimate of 26,602 ± 4,155 for the exact same areas obtained in the previous survey, conducted in 2001 (Dunham, 2002). In view of this, estimates from the 2001 survey have been retained from the previous report. Nevertheless, it is worth noting that the 2006 survey recorded a considerable increase in the carcass ratio (i.e., the proportion of dead to dead plus live elephants), from 3.2% in 2001 to 5.6% in 2006. A tenth of the estimated number of dead elephants in the 2006 survey were found to have their tusks intact, suggesting that they may have died of natural causes, possibly drought-related (Dunham et al., 2006b).

An aerial sample count of the Sebungwe region conducted in late 2006 (Dunham et al., 2006a) gave an overall estimate of 15,024 ± 2,133. Results from this survey replace estimates from a methodologically comparable survey conducted in 2001 (Mackie, 2002b). Although considerable increases in elephant mortality and illegal activities were recorded, with a carcass ratio of 15.6% for the study area (see Current Issues above), the estimate of live elephants in 2006 did not differ significantly from that in 2001 (Dunham et al., 2006a).

The Zambezi Valley area was surveyed as part of a 2003 aerial sample count that also included adjacent populations in Mozambique and Zambia. This survey returned an estimate of 19,981 ± 2,392 for the Zimbabwe portion (Dunham, 2004a), and the estimate replaces a 2001 aerial sample count estimate of 19,227 ± 2,493. Another survey of the same area was conducted in 2005, but no report had yet been produced at the time of writing.

An aerial survey of Gonarezhou was planned for 2006, but had to be postponed until 2007 due to technical difficulties. An aerial sample count of the Save Valley Conservancy, conducted in 2003, gave an estimate of 527 ± 310 (Dunham, 2003), and this replaces the previous aerial sample count estimate of 535 ± 318.

Both the AED 1998 and the AESR 2002 showed an incorrect estimate of 33 elephants for Matibi II communal lands. The correct estimate of zero (Davies et al., 1996) is now shown in the table of estimates.

Estimates from surveys conducted since the previous report, all of which are comparable in design and intensity to previous counts, have resulted in an increase of about 2,850 in the definite category. The marginal decline in the probable and possible categories is brought about by a minor increase in precision and the correction of the Matibi II estimate.

Cross-border Movements

All of Zimbabwe's major elephant populations are located along the border with neighbouring countries, and movements can be expected to take place across all of them, except across Lake Kariba (Cumming & Jones, 2005; Dunham et al., 2006a). The population in Hwange National Park and surrounding areas is part of a much larger population that spans the borders of Zimbabwe, Botswana, Namibia (Cumming & Jones, 2005) and perhaps also Angola. Elephants are also known to move between Zimbabwe, Zambia and Mozambique along the Zambezi Valley in the north, and to Botswana and South Africa in the south (Selier et al., 2002).

SUMMARY TOTALS FOR ZIMBABWE

Data Category DEFINITE PROBABLE POSSIBLE SPECULATIVE
Aerial or Ground Total Counts 236 0 0 0
Direct Sample Counts and Reliable Dung Counts 83,991 7,033 7,033 0
Informed Guesses 189 0 334 91
Other Guesses 0 0 0 200
TOTALS2006 84,416 7,033 7,367 291
TOTALS 2002 81,555 7,039 7,373 291

INTERPRETATION OF CHANGES IN ESTIMATES FROM PREVIOUS REPORT

Cause of Change DEFINITE PROBABLE POSSIBLE SPECULATIVE
Repeat Survey +2,864 -4 -4 0
New Analysis -3 -2 -2 0
TOTAL CHANGE +2,861 -6 -6 0

AREA OF RANGE COVERED BY EACH DATA CATEGORY (km2)

Data Category Known Range Possible Range Total Range
Aerial or Ground Total Counts 2,998 0 2,998
Direct Sample Counts and Reliable Dung Counts 63,671 0 63,671
Informed Guesses 6,746 0 6,746
Other Guesses 2,423 0 2,423
Unassessed Range 717 375 1,092
TOTAL 76,555 375 76,931

ZIMBABWE: ELEPHANT ESTIMATES

INPUT ZONE CAUSE OF CHANGE1 SURVEY DETAILS2 NUMBER OF ELEPHANTS SOURCE PFS3 AREA (km2) MAP LOCATION
TYPE RELIAB. YEAR ESTIMATE 95% C.L. LON. LAT.
Binga Communal Lands RS AS2 B 2006 431 373 Dunham et al., 2006a 2 2,217 27.9 E 17.4 S
Bubi Valley Conservancy IG3 D 2001 53 Dunham & Mackie, 2002 2 2,895 30.1 E 21.5 S
Bubiana Conservancy IG3 D 2001 50 50* Dunham & Mackie, 2002 2 1,772 29.8 E 21.1 S
Chete Safari Area RS AS2 B 2006 971 310 Dunham et al., 2006a 3 1,260 27.8 E 17.4 S
Chewore IV AS1 B 2001 580 335 Mackie, 2002a 3 610 29.9 E 16.2 S
Chiredzi River Conservancy GT1 A 2001 28 Dunham & Mackie, 2002 3 895 31.6 E 20.8 S
Chirisa Safari Area RS AS2 B 2006 4,231 1,260 Dunham et al., 2006a 2 1,529 28.3 E 17.9 S
Chizarira National Park RS AS1 B 2006 3,071 1,117 Dunham et al., 2006a 2 2,084 27.9 E 17.8 S
Doma Safari Area AS2 B 2001 336 383 Mackie, 2002a 3 975 30.2 E 16.4 S
Gonarezhou National Park AS2 B 2001 4,987 1,577 Dunham & Mackie, 2002 2 4,987 31.9 E 21.6 S
Hartley Safari Area IG3 D 2001 100 20* Dunham & Mackie, 2002 3 445 29.6 E 17.9 S
Home Farm & Greystone Ranches IG3 D 2001 3 1* Dunham & Mackie, 2002 4 60 27.9 E 20.8 S
Hwange National Park & Safari Area AS2 B 2001 44,492 5,770 Dunham & Mackie, 2002 2 12,900 26.6 E 19.1 S
Kariba Communal Lands RS AS2 B 2006 3,715 1,033 Dunham et al., 2006a 2 3,224 28.4 E 17.1 S
Kavira Forest Land IG3 D 2001 100 Dunham & Mackie, 2002 3 287 27.0 E 18.1 S
Lusulu AS2 B 2001 33 63 Mackie, 2002b 3 543 27.8 E 18.0 S
Mahenye Ward AS2 B 2001 0 Dunham & Mackie, 2002 221 32.4 E 21.2 S
Malilangwe Conservancy AT3 A 2001 116 Dunham & Mackie, 2002 3 425 31.9 E 21.1 S
Malipati Safari Area AS2 B 2001 5 9 Dunham & Mackie, 2002 3 175 31.4 E 21.9 S
Mambali Communal Lands AT3 A 2001 10 Dunham & Mackie, 2002 3 327 28.4 E 21.5 S
Maramani Communal Lands AT3 A 2001 0 Dunham & Mackie, 2002 367 29.4 E 22.1 S
Matabeleland Communal Lands AS2 B 2001 64 79 Dunham & Mackie, 2002 2 3,110 27.1 E 19.6 S
Matetsi Safari Complex AS2 B 2001 4,201 1,670 Dunham & Mackie, 2002 2 4,399 25.7 E 18.2 S
Matibi II Communal Lands NA AS2 E 1996 0 Davies et al., 1996 400 31.7 E 21.5 S
Matusadona National Park RS AS2 B 2006 1,925 443 Dunham et al., 2006a 2 1,413 28.6 E 17.0 S
Mavuradonha Wilderness Area AS2 B 2001 13 26 Dunham & Mackie, 2002 3 617 30.9 E 16.5 S
Mukwiche Area AS1 B 2001 228 296 Mackie, 2002a 3 337 29.9 E 16.4 S
Ngamo & Sikumi State Forests AS2 B 2001 553 496 Dunham & Mackie, 2002 2 2,344 27.3 E 18.8 S
North Gokwe Communal Lands RS AS2 B 2006 192 172 Dunham et al., 2006a 2 3,082 28.5 E 17.5 S
Nyatana Wildlife Management Area IG3 D 2001 150 Dunham & Mackie, 2002 3 651 32.5 E 16.7 S
Protea Farm IG3 D 2001 7 Dunham & Mackie, 2002 5 14 29.6 E 16.5 S
Save Valley Conservancy RS AS1 B 2003 527 310 Dunham, 2003 2 3,047 32.1 E 20.4 S
Sengwe Communal Land OG3 E 2001 200 Dunham & Mackie, 2002 2 2,422 31.3 E 22.1 S
Sentinel & Nottingham Ranches AT3 A 2001 82 Dunham & Mackie, 2002 3 568 29.6 E 22.1 S
Shangani Ranch IG3 D 2001 60 20* Dunham & Mackie, 2002 3 628 29.3 E 19.6 S
Sijarira Forest Area RS AS2 B 2006 488 333 Dunham et al., 2006a 3 270 27.5 E 17.6 S
Tuli Circle Safari Area AT3 A 2001 0 Dunham & Mackie, 2002 416 29.1 E 22.0 S
Zambezi Valley RS AS2 B 2003 19,981 2,392 Dunham, 2004a 1 16,476 29.7 E 16.1 S

* Range of informed guess

1 Key to Causes of Change: DA: Different Area; DD: Data Degraded; DT: Different Technique; NA: New Analysis; NG: New Guess; NP: New population; PL: Population Lost; RS: Repeat Survey (RS' denotes a repeat survey that is not statistically comparable for reasons such as different season); —: No Change

2 Key to Survey Types: AS: Aerial Sample Count; AT: Aerial Total Count; DC: Dung Count; GD: Genetic Dung Count; GS: Ground Sample Count; GT: Ground Total Count; IG: Informed Guess; IR: Individual Registration; OG: Other Guess. Survey Type is followed by an indicator of survey quality, ranked from 1 to 3 (best to worst). Survey Reliability is keyed A–E (best to worst)

3 PFS: Priority for Future Surveys, ranked from 1 to 5 (highest to lowest). Based on the precision of estimates and the proportion of national range accounted for by the site in question, PFS is a measure of the importance and urgency for future population surveys. All areas of unassessed range have a priority of 1. See Introduction for details on how the PFS is derived.

< previous section  < index >  next section >