This paper contains supplementary material that can be found online at http://journals.cambridge.org
Introduction
Wildlife management in southern Africa has evolved through three stages (Child, Reference Child, Child, Suich and Spencely2009). With white settlement wildlife populations were decimated by unregulated hunting and habitat fragmentation (MacKenzie, Reference MacKenzie1988). Declines were exacerbated by outbreaks of bovine pleuropneumonia and rinderpest (Bond et al., Reference Bond, Child, de la Harpe, Jones, Barnes, Anderson and Child2004). A protectionist phase followed in which colonial administrations established legislation that centralized control over wildlife and limited commercial use, making wildlife on farmlands a burden for landowners (MacKenzie, Reference MacKenzie1988; Murombedzi, Reference Murombedzi2003). Wildlife populations continued to wane because of illegal hunting, persecution by landowners, state-sponsored hunting to remove tsetse fly Glossina spp. hosts, and construction of veterinary fences (Child & Riney, Reference Child and Riney1987; Taylor & Martin, Reference Taylor and Martin1987; Bond et al., Reference Bond, Child, de la Harpe, Jones, Barnes, Anderson and Child2004). Negative wildlife population trends improved following legislative changes during the 1960s and 1970s that enabled landowners to utilize wildlife on their land (Bond et al., Reference Bond, Child, de la Harpe, Jones, Barnes, Anderson and Child2004).
During the 1980s there was a rising demand for tourism and safari hunting, providing incentives for landowners to begin wildlife ranching (Bond et al., Reference Bond, Child, de la Harpe, Jones, Barnes, Anderson and Child2004). Recurrent droughts, declining range productivity because of overstocking with livestock and declining state subsidies for livestock production hastened the shift to wildlife ranching (Jansen et al., Reference Jansen, Bond and Child1992; Child, Reference Child, Prins, Grootenhuis and Dolan2000; Carruthers, Reference Carruthers2008). Predictions that wildlife could produce more meat than livestock (Dasmann & Mossman, Reference Dasmann and Mossman1961) were not borne out because of the costs of harvesting wildlife, veterinary restrictions and lack of support infrastructure. Rather, the comparative advantage of wildlife lay in multiple values from ecotourism, safari hunting, meat and hides (Child, Reference Child, Prins, Grootenhuis and Dolan2000; Carruthers, Reference Carruthers2008). Wildlife ranching spread rapidly across semi-arid private lands in southern Africa. There are now at least 9,000 wildlife ranches in South Africa, covering c. 205,000 km2 (Falkena, Reference Falkena2003; NAMC, 2006). In Zimbabwe there were 27,000 km2 of wildlife ranches prior to so-called land reform (Bond et al., Reference Bond, Child, de la Harpe, Jones, Barnes, Anderson and Child2004) and in Namibia there were 400 registered hunting farms by 2001 (Krug, Reference Krug2001).
A growing body of evidence suggests that wildlife-based land uses confer several ecological and socio-economic benefits compared to livestock farming in semi-arid areas (Price-Waterhouse, 1994; Langholz & Kerley, Reference Langholz and Kerley2006). For example, wildlife-based land use has resulted in the restoration of degraded rangelands in some areas and stimulated recoveries of wildlife populations, including threatened species (Barnes & de Jager, Reference Barnes and de Jager1996; Bothma et al., Reference Bothma, Suich, Spenceley, Child, Such and Spenceley2009; Child, Reference Child, Child, Suich and Spencely2009). In semi-arid areas wildlife-based land use is commonly more profitable than livestock, generates more foreign currency, and is less susceptible to drought (Price-Waterhouse, 1994; Sims-Castley et al., Reference Sims-Castley, Kerley, Geach and Langholz2005). Wildlife-based land uses contribute to food security through employment, foreign currency and venison (Price-Waterhouse, 1994; Langholz & Kerley, Reference Langholz and Kerley2006).
Few studies have examined the economic, social and conservation impacts of wildlife ranching on private land in southern Africa, despite the scale of the industry. Lack of data and a perception that the primary beneficiaries of wildlife-based land uses are white landowners have meant that wildlife ranching is not always fully supported by governments (Duffy, Reference Duffy2000). Some politicians believe that wildlife ranches threaten food security (du Toit, Reference du Toit2004) and others perceive wildlife ranching as an attempt by white landowners to avoid land reform (Gibson, Reference Gibson1999). Current policies artificially inflate the profitability of livestock farming and suppress that of wildlife ranching through veterinary restrictions on wildlife reintroductions (Scoones & Wolmer, Reference Scoones and Wolmer2008; Albertson, Reference Albertson, Ferguson and Hanks2010). Without intervention to raise awareness among politicians of the benefits of wildlife-based land uses there is a risk that land reform will cause a reversion to livestock in areas best suited for wildlife. Southern Africa inherited skewed land ownership from colonial governments and the transfer of land from white to black farmers is a political imperative. There is therefore a need for research into the scale and impacts of wildlife ranching in southern Africa to guide land-use planning, veterinary policies and land reform. Here we provide an assessment of wildlife ranching in Namibia.
Legislative basis for wildlife ranching in Namibia
Several forms of consumptive utilization of wildlife are allowed on Namibian farmlands with appropriate permits, including: shoot-and-sell (shooting of animals for meat to sell), safari hunting (sale of guided hunts mainly to foreign hunters), management hunts (sale of guided hunts targeting non-trophy animals), biltong hunting (mainly local hunters sold the right to shoot animals for meat), wildlife harvesting (wildlife is culled by specialized teams to produce venison), shooting for own use, and capture and sale of live wildlife (Gödde, Reference Gödde2008). Consumptive wildlife use is governed by the Ministry of Environment and Tourism via Nature Conservation Ordinance No. 4 of 1975. This legislation was amended with the Nature Conservation Amendment Act of 1996, which conferred similar user rights to residents of communal land conservancies (Barnett & Patterson, Reference Barnett and Patterson2006). A Parks and Wildlife Management Bill is being drafted but is not yet operational (Laubscher et al., Reference Laubscher, Jooste, Mbai and Idsardi2007). Permits for consumptive use are allocated by the Ministry of Environment and Tourism following submission of management plans by farmers and field inspections/wildlife counts (Gödde, Reference Gödde2008). Twenty-five conservancies have developed in which multiple landowners manage wildlife cooperatively (comprising 1,008 farms and c. 43,250 km2; Ministry of Environment and Tourism, pers. comm., 2010).
Methods
A structured, pre-tested questionnaire was used to gather quantitative data on land-use, wildlife, employment and venison production (Appendix). Sixty of the 81 member associations of the Namibian Agricultural Union (NAU) were randomly selected. From each association four farmers were randomly sampled and interviewed in person. If respondents were not reachable, alternatives were randomly selected. Interviews were conducted in English, Afrikaans or Herero by four interviewers. Farmers were informed that the survey was part of a university study on wildlife-based land uses in Namibia and that the results would be anonymous. The interviewers were provided with training in survey techniques and observed multiple pre-tests of the survey and conducted several supervised practice surveys before commencing data collection.
Two hundred and fifty farmers were interviewed (sample distribution is depicted in Fig. 1). Because of multiple farm-ownership/lease-holding the sample covered 412 farms (28,038 km2). There are 3,500 commercial farms in Namibia (Giel Schoombee, NAU, pers. comm., 2010). The margin of error with this sample size is 4.9% (i.e. 95% confidence interval). Refusal rate was 4.8%, which is unlikely to introduce non-response bias (Lindner, Reference Lindner2002).
Estimates of wildlife populations and wildlife biomass on freehold farms were made by multiplying mean values per km2 from our sample in each region by the area of farmland in each region. Wildlife biomass was estimated by multiplying the mean mass of individuals of a species (0.75 of standard female mass; Hayward et al., Reference Hayward, Henschel, O'Brien, Hofmeyr, Balme and Kerley2006) by respondents’ estimates of populations of those species on their properties. When estimating venison production mean dressing percentages were multiplied by the number of individuals reported utilized each year (Bothma & du Toit, Reference Bothma and Du Toit2009). All animals utilized were assumed to be adults (sex ratios of harvests were provided by farmers), except for culling, for which 0.75 of standard female mass was used.
To estimate meat production on a national scale two methods were used. (1) Mean percentage offtake of populations of each species in each region were calculated and multiplied by population estimates for each region (after Barnes et al., Reference Barnes, Lange, Nhuleipo, Muteyauli, Katoma and Amupolo2009). These values were then multiplied by the mean meat yield from an individual of each species via each form of utilization. For small antelopes for which population estimates were not available the percentage of total meat production from the sample that they comprised was calculated, and the national meat production estimate adjusted upwards by the same proportion. (2) Meat production was also calculated based on available land, by multiplying the mean meat production per km2 from all forms of use in each region by the area of farmlands in each region.
Estimates of venison produced per km2 in each region were multiplied by mean prices (from the survey) to calculate earnings from meat. Warthog Phacochoerus africanus meat was assumed not to be sold, except where the species was shot under shoot-and-sell permits, as the sale of meat from the species is restricted (F. Joubert, Directorate of Veterinary Services, pers. comm., 2010).
Survey data were analysed using multiple logistic regressions, χ2 tests and analyses of variance (JMPIN, 2000). When commencing with multiple logistic regressions or analyses of variance all variables expected to influence the dependent variable were included in the models and removed following a backwards stepwise procedure until all remaining variables were statistically significant. To analyse percentage income from different land uses we categorized income data as 0–25% income, 26–50% and > 50%.
Data on vegetation, mean annual rainfall, human densities, distances from towns/national roads of each farm in the sample, and estimates of the area of freehold farms in each region were calculated or derived from NACSO (2010), using ArcInfo v. 9.3 (ESRI, Redlands, USA). Land was categorized as falling in the ‘small-stock’ or ‘large-stock’ farming areas, following Erb (Reference Erb2004).
Results
Interviewees
Eighty-seven percent (87.1%) of respondents interviewed were white, of which 54.2% were Afrikaans-speaking, 42.1% were German-speaking and the remainder English-speaking. Thirteen percent of farmers were black (of which 74.0% were Herero and 9.2% Damara). Of farmers in commercial conservancies, only 0.86% were black.
Land use
Livestock production was the most common land use (92.3% of respondents) and generated the largest mean proportion of respondents’ income (66.9%; Figs 1 & 2). Cattle were the most widespread livestock (93.4% of respondents; mean density where kept 5.1 ± SE 0.36 km−2), followed by sheep (72.7%, 13.6 ± SE 2.3 km−2), and goats (61.6%, 2.20 ± SE 0.19 km−2). Percentage income from livestock was influenced by region (highest in Kunene, 79.5%, and Otjozondjupa, 67.1%, and lowest in Erongo, 54.2%) and by age of respondent (higher among older farmers; F Ratio = 3.69, df = 8, P < 0.001; JMPIN, 2000).
Fig. 1 Spatial patterns in primary land use (i.e. that accounting for the majority of farmers’ income) on freehold land in Namibia (categorized as either consumptive wildlife utilization, ecotourism or livestock), and the line denoting the boundary between the small-stock area (to the south) and large-stock area (to the north).
Fig. 2 Percentage of farmers interviewed who were engaged in various land-use forms (wildlife cropping refers to the large-scale culling of wildlife to produce meat for sale, the shooting often done at night, and differs from shoot-and-sell which typically involves the more selected removal of one individual at a time).
Seventy-five percent of respondents practised commercial wildlife-based land uses (Fig. 2). Wildlife-based land uses are practised over c. 287,000 km2 and exclusively over c. 32,000 km2 (Table 1). Whether or not wildlife-based land uses are practised was related to conservancy membership (94.0% of respondents in conservancies cf. 69.4% outside), and wildlife diversity (9.0 ± SE 0.32 wild ungulate species where wildlife-based land use is practised cf. 5.2 ± SE 0.26). Percentage of income from wildlife-based land uses was higher among conservancy members (35.3% cf. 19.1% among non-members) and was higher among whites than blacks (29.6% cf. 6.6%).
Table 1 Total area, percentage (and area), and mean % income generated from each land use, of farms practising safari hunting, ecotourism, any wildlife-based land uses (i.e. safari hunting, ecotourism, shoot-and-sell, biltong hunting, management hunts, cropping, live sales), wildlife only and livestock only, in 10 regions of Namibia.
The commonest forms of wildlife-based land uses were shoot-and-sell, safari hunting and ecotourism (Fig. 2). Safari hunting (9.2%), ecotourism (6.8%) shoot-and-sell (2.7%), live sales (1.8%) and biltong hunting (1.3%) generated most income from wildlife-based land uses (Table 1). Percentage of income from safari hunting was greater among younger ranchers and among conservancy members (22.6% cf. 7.5%; F Ratio = 11.5, df = 2, P < 0.001). Percentage income from ecotourism was higher in the small-stock than large-stock area (7.7% cf. 6.4%).
Twenty-one percent (21.4%) of farmers would consider removing all livestock and practising only wildlife-based land uses in the future. Willingness of respondents to make such a change was influenced by race (30.8% of blacks were willing cf. 20% of whites), proportion of income from ecotourism (willing respondents derived 8% of income from ecotourism cf. 4% among unwilling respondents), proportion of income from safari hunting (willing respondents derived 22% of income from safari hunting cf. 11% among unwilling respondents), and respondents’ distance from a town (willing respondents were nearer towns 142 ± SE 20 km cf. 192 ± SE 11.4 km; χ2 = 152, df = 158, P < 0.001). Most farms have stock-proof fencing, even in conservancies (Table 2). Game-proof fencing is relatively uncommon (Table 2).
Table 2 Percentage occurrence of various forms of fencing on Namibian commercial farmlands.
Wildlife populations
Wildlife populations on freehold land may be larger than previously recognized (Barnes et al., Reference Barnes, Lange, Nhuleipo, Muteyauli, Katoma and Amupolo2009; Table 3). Wild ungulate diversity was higher in conservancies (10.1 ± SE 0.39 species per farm cf. 6.72 ± SE 0.36), positively related to income from safari hunting (ranchers obtaining with 0–25% of their income from safari hunting had 5.4 ± SE 0.2 species, those earning 26–50% had 10.8 ± SE 4.4 species, and those deriving > 50% had 12.4 ± SE 0.8 species), negatively related to income from livestock (ranchers obtaining 0–25% of income from livestock had 10.3 ± SE 1.2 species, those earning 26–50% had 10.8 ± SE 0.65 species, and those deriving > 50% had 7.0 ± SE 0.3 species), negatively related to farmer age, and influenced by vegetation (F Ratio = 18.9, df = 15, P < 0.001). Wildlife diversity was highest in thorn-bush shrub-land (11.6 ± SE 0.9 species) and southern Kalahari (10.3 ± SE 1.3), and lowest in Karas dwarf shrub-land (4.4 ± SE 0.5). Four of the so-called big five (buffalo Syncerus caffer, lion Panthera leo, elephant Loxodonta africana and rhinoceros Diceros bicornis) were rare on farms whereas the leopard Panthera pardus was not (Fig. 3). Springbok Antidorcas marsupialis, oryx Oryx gazella, kudu Tragelaphus strepsiceros and warthog were the most abundant species on farmlands (Table 3).
Fig. 3 Percentage occurrence of large wild mammals on Namibian farmlands within and outside of conservancies.
Table 3 Estimates of wildlife populations on freehold land, by region and overall, based on mean densities of each species derived from farmers’ estimates of population sizes, and the estimates of Barnes et al. (Reference Barnes, Lange, Nhuleipo, Muteyauli, Katoma and Amupolo2009), ordered by total population.
1 Assuming that wildlife densities in Oshikoto, Oshana and Omusati equal those in Kunene, the nearest region for which density estimates are available
2 Assuming an area of 356,533 km2 of freehold land (Mendelsohn, 2006)
Livestock biomass (mean 2,251 ± SE 140 kg km−2) was higher than wild ungulate (and ostrich Struthio camelus) biomass (936 ± SE 84.1 kg km−2; F Ratio = 64.0, df = 1, P < 0.001). Livestock production on freehold land contributed NAD 1.97 billion (USD 235 million at mean 2009 rates) to gross national income (GNI) in 2009, compared to at least USD 166 million from wildlife and tourism (Barnes et al., Reference Barnes, Alberts and MacGregor2010). Wildlife biomass is thus more efficient at generating revenue than livestock. Wildlife, which comprises 29.4% of mammalian biomass, generates 41.5% of the revenue from livestock, wildlife and tourism combined, or 1.41% revenue per 1% biomass, whereas livestock generates 0.83% revenue per 1% biomass.
Wildlife biomass was negatively related to income from livestock (ranchers earning 0–25% of income from livestock had 2,712 ± SE 900 kg of wildlife biomass km−2, those deriving 26–50% had 1,516 ± SE 137 kg km−2, and those deriving >50% had 911 ± SE 136 kg km−2), positively related to income from safari hunting (ranchers deriving 0–25% of income from safari hunting had 973 ± SE 256 kg of wildlife biomass km−2, those deriving 26–50% had 1,369 ± SE 108 kg km−2, and those deriving >50% 2,179 ± 258 kg km−2), and positively related to income from ecotourism (ranchers deriving 0–25% from ecotourism had 1,129 ± 123 kg of wildlife biomass km−2, those deriving 26–50% had 1,137 ± SE 166 kg km−2, and those deriving >50% had 2,849 ± 1,324 kg km−2), and was positively related to wildlife diversity (F Ratio = 69.9, df = 8, P < 0.001). Some ranchers have significant wildlife populations and yet generate little or no income from wildlife-based land uses, suggesting that the resource is underutilized in some areas.
Fifty-eight percent (57.6%) of respondents thought wildlife populations were increasing on their land, 23.7% thought they were stable, and 18.6% thought they were declining. Percentage of income from safari hunting was higher on properties with stable or increasing wildlife populations (18.7 ± SE 2.8%) than where wildlife was declining (1.50 ± SE 0.76%; χ2 = 12.1, df = 4, P < 0.001). Wildlife was more commonly stable or increasing inside (87.9%) than outside conservancies (75.0%; χ2 = 3.3, df = 1, P = 0.068). Explanations for increasing wildlife populations included favourable rainfall (35.3%; rainfall was generally above average during 2000–2009; Namibian Ministry of Works and Transport, 2011), good management (26.4%), conservative harvests (19.1%, Table 4), artificial water-points (10.3%), and incentives for conservation through safari hunting (8.8%). Explanations for declining wildlife populations among the 18.6% of ranches reporting such trends included excessive utilization (50.0%), drought (13.6%), poaching (9.1%), and persecution by livestock farmers (4.5%).
Table 4 Game meat production on Namibian freehold farms, the percentage of meat produced in each region, offtake as a proportion of populations and intrinsic rates of increase for each species by comparison, ordered by estimate of meat produced.
1 Extrapolating from population estimates made by Barnes et al. (Reference Barnes, Lange, Nhuleipo, Muteyauli, Katoma and Amupolo2009); this is conservative as our estimates of wildlife populations are considerably higher
2 Calculated as the total number of animals of each species harvested on the ranches surveyed as a percentage of the populations of those species estimated by the ranchers
3 Black wildebeest, nyala Tragelaphus angasi, tsessebe, white rhinoceros Ceratotherium simum, klipspringer Oreotragus oreotragus, dik dik, grey duiker Sylvicapra grimmia, blesbok Damaliscus pygargus, ostrich
Meat production
An annual mean of 67.7 ± SE 6.8 kg of venison was produced per km2 on farmland. Safari hunting generated the highest quantity of venison (21.9 ± SE 3.9 kg km−2), followed by shooting for own use (21.1 ± SE 3.0 kg km−2), shoot-and-sell (13.9 ± SE 2.6 kg km−2) and biltong hunting (6.5 ± SE 1.5 kg km−2). Most venison was from oryx, kudu and springbok (Table 4). Typical harvests of wildlife on Namibian farms were well within intrinsic rates of increase for those species (Table 4). Venison production per km2 was related positively to wildlife biomass, wildlife diversity and livestock biomass (F Ratio = 48.9, df = 3, P < 0.001).
Between 15,917 t (extrapolated from mean utilization of available wildlife populations) and 24,952 t (extrapolated from mean production per km2 to available land area) of venison are produced on freehold farms per year (Tables 4 & 5). In contrast, 93,045 t of meat from domestic stock are produced in Namibia (including communal land) annually, of which 86.9% is exported (W. Schutz, Namibian Meat Board, pers. comm., 2010). Approximately 805 t of venison are exported from Namibia each year (including 85 t to Europe, 160 t exported legally to South Africa, and a tentative estimate of 720 t smuggled to South Africa; Laubscher, Reference Laubscher, Jooste, Mbai and Idsardi2007; D. Museler, pers. comm., 2010). These exports correspond to 3.0–5.0% of venison produced on freehold land, so more venison than meat from livestock on freehold land remains in Namibia (15,200–22,200 t cf. 12,100 t).
Table 5 Estimated amount of game meat produced on commercial farmlands in Namibia, by region, from various forms of wildlife utilization, and overall, based on mean meat production per km2 for various forms of wildlife utilization, ordered by total.
1 Based on an estimate of the total area of freehold land (which excludes resettlement farms) and using the proportional breakdown of farms in each region (Mendelsohn, 2006)
2 Assuming that meat production values in these regions equals those in Kunene, the nearest region with available data
Venison is typically sold to butcheries (37.0%), used for workers’ rations (23.5%), or personal consumption (13.7%). Sixty-five percent (64.6%) is sold as whole carcasses, 22.5% as unselected cuts, 6.8% as processed meat and 5.4% as selected cuts. Prices obtained by farmers for unprocessed venison increased from c. USD 1.42 kg−1 in 2006/2007, to USD 2.07 kg−1 in 2009 but remains lower than the beef (USD 2.44) and sheep price (USD 2.50; mean 2009 Namibian Meat Board values). Prices of meat from eland Tragelaphus oryx and springbok were 14.1 and 9.8% higher than other wildlife species. Farmers obtained higher prices for selected cuts (USD 3.71 kg−1) and processed venison (USD 9.47 kg−1).
Annual earnings from venison sales were USD 12.4–116.0 km2 depending on the region; extrapolating from this USD 23.8 million was generated annually from meat sales on freehold land. Including meat obtained from harvesting/culling and shoot-and-sell from eland, hartebeest Alcelaphus buselaphus, impala, oryx, kudu, springbok and Hartmann's mountain zebra Equus zebra (species likely to be most marketable) c. 4,100 t of venison could be exported annually from farmland, which could generate a potential annual return of USD 34.6 million, assuming a price of USD 9.47 kg−1 and that a market exists for that quantity of venison.
Farm workers receive more venison as rations (3.82 ± SE 0.34 kg week−1) than meat from livestock (2.11 ± SE 0.42 kg week−1; F Ratio = 8.1, df = 1, P = 0.005). There are c. 22,855 workers on commercial farmland in Namibia (Giel Schoombee, pers. comm., 2010) and, extrapolating from our sample, c. 4,500 t of meat are used to feed workers annually compared to c. 2,500 t of meat from livestock. Venison rations probably benefit >33,000 workers and their dependants on freehold farms.
Employment
Respondents employed 9.91 ± SE 0.94 workers per management unit (farm or multiple adjacent farms managed by one person or company), or 0.22 ± SE 0.08 workers km−2. Farmers housed an additional 1.94 ± SE 0.11 family members per worker, or a total of 26.4 ± SE 1.9 people per management unit (0.41 ± SE 0.09 people km−2). Employment was positively related to income from ecotourism (farmers earning 0–25% of income from ecotourism employed 0.10 ± SE 0.01 people km−2, those deriving 25–50% employed 0.09 ± SE 0.01 km−2 and those deriving >50% employed 0.31 ± SE 0.09 km−2), and negatively related to income from livestock (farmers deriving 0–25% of income from livestock employed 0.24 ± SE 0.03 people km−2, those deriving 25–50% employed 0.14 ± SE 0.02 km−2 and those deriving >50% employed 0.08 ± SE 0.01 km−2; F Ratio = 12.3, df = 3, P < 0.001).
Discussion
The veracity of our findings is dependent on the reliability of the answers provided by respondents. Because of the care taken when explaining the purpose of the study to respondents, linguistic matching of respondents and interviewers, and the anonymous and non-contentious nature of the survey, respondents were willing to participate and we believe the data provided are reliable. Wildlife-based land use is practised by 75% of Namibian farmers (according to our data), and is increasing in prevalence (Barnes & Jones, Reference Barnes, Jones, Child, Suich and Spenceley2009). Safari hunting is a more commonly practised form of wildlife-based land use on freehold land than ecotourism (and generates a higher mean percentage of farmers’ earnings), contrasting with the findings of Barnes et al. (Reference Barnes, Lange, Nhuleipo, Muteyauli, Katoma and Amupolo2009). Our study may have underestimated the contribution of ecotourism: farms practising large-scale ecotourism can generate high revenues but are probably clustered spatially and may be underrepresented in our survey (J. Barnes, pers. comm., 2010).
Livestock farming is the most widespread land-use and generates the majority of income for most farmers. However, livestock numbers have declined on freehold land in recent years because of range degradation (including bush encroachment) caused by overgrazing and the rise of wildlife-based land uses (de Klerk, Reference de Klerk2004; Barnes & Jones, Reference Barnes, Jones, Child, Suich and Spenceley2009), although improved herd management has maintained output (Erb, Reference Erb2004). Wildlife production is probably less affected by bush encroachment (many species are browsers) and, as long as stocking rates are not excessive, replacement of livestock with wildlife should stimulate gradual rangeland recovery (Child, Reference Child, Child, Suich and Spencely2009).
Economic role of wildlife-based land uses
Wildlife and tourism on freehold land contributed USD 166 million to GNI in Namibia in 2009 (or USD 213 million, if all natural resources are taken into account), compared to USD 235 million from livestock (Barnes et al., Reference Barnes, Alberts and MacGregor2010). These estimates are conservative, as the economic value of venison (USD 23.8 million per year, excluding export earnings) is higher than previously thought (USD 532,544; Barnes et al., Reference Barnes, Lange, Nhuleipo, Muteyauli, Katoma and Amupolo2009). The economic contribution of wildlife and tourism on freehold land may already exceed that of livestock despite policies and subsidies favouring the latter. With continued growth in tourist and hunter arrivals likely, the economic contribution of wildlife will probably increase further. The trophy hunting industry increased in value from USD 28.5 to 44.8 million during 2004–2007 (Lamprechts, Reference Lamprechts2009) and international tourist arrivals in Namibia are predicted to increase by 5.7% per annum over the next 10 years (WTTC, 2012). Wildlife-based land uses are popular among younger farmers and earnings from wildlife are projected to be 60% less affected by climate change than those from livestock (Barnes et al., Reference Barnes, Alberts and MacGregor2010).
Social benefits of wildlife-based land uses
Employment on Namibian farmlands is related positively to income from ecotourism but negatively to income from livestock, in keeping with findings from Zimbabwe (Price-Waterhouse, 1994) and South Africa (Langholz & Kerley, Reference Langholz and Kerley2006). In the Eastern Cape the switch to wildlife-based land uses increased employment by 4.5 times, wage bills by 32 times and conferred improved working conditions for employees (Langholz & Kerley, Reference Langholz and Kerley2006). Such improvements are crucial as farm workers earn among the lowest wages (LEAD, 2005). Wildlife-based land uses also confer social benefits through protein provision. More venison is produced on Namibian farms than previously recognized (16,000–23,000 cf. 4,300 t, Laubscher et al., Reference Laubscher, Jooste, Mbai and Idsardi2007) and acts as a key food source for workers and their families.
Ecological significance of wildlife-based land uses
The area of farmland used for wildlife-based land uses is more than twice as large as the protected area network (c. 287,000 cf. 114,079 km2, Cumming, Reference Cumming and Child2004). Although the primary objective of wildlife ranches is typically profit they nonetheless confer biodiversity gains. For example, 82,000 Hartmann's mountain zebras and 13,500 black-faced impalas Aepyceros melampus petersi live on Namibian farmlands, and cheetahs Acinonyx jubatus are present on 71.2% of farms (our data). Populations of most wildlife species are increasing on farmlands and the proportion of mammalian biomass comprised by wildlife increased from 8% in 1972 to 29% in 2009 (Barnes & de Jager, Reference Barnes and de Jager1996). Wildlife numbers on commercial farms (1.8–2.8 million) exceed those in protected areas (c. 121,000) and community conservancies (150,000–200,000; Barnes et al., Reference Barnes, Lange, Nhuleipo, Muteyauli, Katoma and Amupolo2009; C. Weaver, WWF–Namibia, pers. comm.). Wildlife abundance on freehold land may also be higher than previous estimates. Our extrapolations of wildlife numbers require caution as they rely on farmers’ estimates. However, the Ministry of Environment and Tourism (and conservancies) conduct regular wildlife counts and most farmers probably have a reasonable impression of their wildlife populations.
There are, however, a number of conservation problems on Namibian farmlands, including continued intolerance towards predators (Marker et al., Reference Marker, Mills and Macdonald2003). Lions and wild dogs Lycaon pictus occur on < 10% of farms, suggesting that lethal control is preventing them from recovering. Ranchers may persecute predators to protect their investment in valuable extralimital wildlife species. In addition, the increasing prevalence of game-proof fencing can interrupt natural processes such as migration, reduce the ability of ungulates to utilize patchy primary productivity (Fryxell & Sinclair, Reference Fryxell and Sinclair1988) and increase the risk of localized overstocking (Lindsey et al., Reference Lindsey, Romañach and Davies-Mostert2009). Finally, although wild ungulate populations are thriving in most areas, there are negative trends in some groups of farms. Such trends are possibly because of excessive harvesting related to high venison prices and are most common outside conservancies, where harvests are not coordinated.
Lack of development of wildlife ranching in Namibia
Despite expansion of wildlife-based land uses in Namibia it has not yet been embraced as fully by farmers as in South Africa or, as formerly, in Zimbabwe. Most Namibian farmers (>90%) retain livestock whereas by 2001 > 50% of ranchers in several semi-arid parts of South Africa and Zimbabwe had removed all livestock (Lindsey et al., Reference Lindsey, Romañach and Davies-Mostert2009). Six factors in particular undermine the development and value of wildlife-based land uses in Namibia.
Inadequate devolution of user rights over wildlife
In Namibia user-rights over wildlife were not devolved as far to landowners as in Zimbabwe and South Africa (NNF, 2010). Landowners in Namibia are required to apply for permits to hunt wildlife, reducing management flexibility and profitability, increasing transaction time and costs, and impinging on farmers’ autonomy, thus creating disincentives for wildlife-based land uses (NNF, 2010). The permit system and seasonal restrictions on hunting also limit venison exports (Gödde, Reference Gödde2008).
Veterinary restrictions
A veterinary cordon across northern Namibia controls the spread of foot-and-mouth disease to retain access to export markets for beef. Most freehold farms occur south of the cordon in the foot-and-mouth disease free zone, where the reintroduction of buffalo is prohibited (including individuals free of foot-and-mouth disease; DVS, 2007). The buffalo is a key species for safari hunting because it commands high trophy fees (USD 6,400 cf. <USD 1,000 for most antelopes) and is used to sell hunting packages (buffaloes generate c. USD 14,000 in daily rates per hunt cf. c. USD 4,000 for antelope hunts; P. Lindsey, unpubl. data). Buffaloes generate 4.1–49.0% of income from safari hunting depending on the country (Lindsey et al., Reference Lindsey, Balme, Booth and Midlane2012). Historically, the buffalo occurred in most areas with > 250 mm of rainfall, including much of what is now farmlands (Martin, Reference Martin and Osofsky2004). Costs of veterinary restrictions are borne by the state but the benefits are enjoyed by individual farmers, artificially inflating the profitability of livestock (Scoones & Wolmer, Reference Scoones and Wolmer2008). Nonetheless, the profitability of commercial livestock production is low across much of southern Africa (Jansen et al., Reference Jansen, Bond and Child1992; McLaughlin, Reference McLaughlin2010) and is projected to decline (Barnes et al., Reference Barnes, Lange, Nhuleipo, Muteyauli, Katoma and Amupolo2009). Long-term access to European markets for beef is not guaranteed, the costs of maintaining veterinary restrictions are increasing and the efficacy of control measures for foot-and-mouth disease is declining (Scoones & Wolmer, Reference Scoones and Wolmer2008; Thomson, Reference Thomson2008). The wisdom of continued subsidization of the livestock industry at the expense of wildlife-based land uses is thus questionable. At the very least provision should be made for the reintroduction of certified disease-free buffalo on wildlife ranches in the freehold farming area. Alternatively, several different approaches to veterinary control could be considered to allow for the unfettered development of wildlife-based land uses in certain areas. For example, foot-and-mouth disease-infected zones could be expanded, or veterinary disease control could be compartmentalized, to allow for the creation of wildlife production zones in areas of particularly suitable habitat. Lastly, commodity-based trade could be considered (Scoones & Wolmer, Reference Scoones and Wolmer2008). Through commodity-based trading meat processed in a manner proven to provide minimal risk of transmitting foot-and-mouth disease would be acceptable for export (Thomson, Reference Thomson2008). If accepted by the International Organization for Animal Health and the EU, commodity-based trading would provide scope for reintroduction of buffalo on freehold land while permitting continued export of beef (Cumming, Reference Cumming, Ferguson and Hanks2010).
Failure to reintroduce other high-value species
Because of the shortage of so-called big game on freehold land most farmers offer similar, low-value hunting/tourism products involving antelopes. In South Africa ranchers with the big five charge more than double for ecotourism than individuals lacking these species (Lindsey et al., Reference Lindsey, Romañach and Davies-Mostert2009). Namibia and Botswana generate similar revenues from safari hunting even though Namibia attracts 4,000–6,000 annually compared to the 500 that visit Botswana (Martin, Reference Martin2008; NAPHA, pers. comm.), because of the shortage of high-value species on Namibian farms (Humavindu & Barnes, Reference Humavindu and Barnes2003).
Failure to develop fully integrated conservancies
A key reason for the absence of the largest species on farmlands is the failure of landowners to cooperate to form fully integrated conservancies. In Zimbabwe and South Africa large conservancies have developed in which all livestock and internal fencing has been removed, and all indigenous mammal species reintroduced (Lindsey et al., Reference Lindsey, Romañach and Davies-Mostert2009). By contrast, Namibian conservancies lack key species, are fractured because not all farms within their boundaries are members, and typically retain livestock and internal fencing. Fully integrated conservancies would facilitate higher-end ecotourism and safari hunting and would confer a variety of social and ecological benefits (Lindsey et al., Reference Lindsey, Romañach and Davies-Mostert2009). At present, however, the Namibian government does not formally recognize private conservancies and the permit system discourages their formation. Landowners with properties surrounded by game fencing are granted longer hunting seasons and more complete user rights over wildlife than those without fencing (including within conservancies; Gödde, Reference Gödde2008). This situation should be reversed.
Failure to integrate development of wildlife-based land uses with land reform
As currently practised in Namibia land reform may cause a shift from wildlife-based land uses to livestock because of a lack of the necessary experience, expertise and start-up capital among many emerging farmers, and inadequate efforts by government to promote their integration into wildlife ranching. Government could identify suitable farms as wildlife ranches and purchase them for allocation to interested emerging farmers, whom our data suggest may be numerous. Promoting the development of fully integrated conservancies could also assist land reform. The economies of scale and centralized management in conservancies would remove key barriers for entry into wildlife-based land uses for emerging farmers. Conservancies could be structured as corporate entities to allow investment by emerging farmers, creating alternative avenues for achieving land reform that would allow for the retention of existing capital and capacity. Proactive efforts by commercial conservancies to facilitate the integration of black farmers may improve prospects of being granted a favourable legislative environment.
Failure to exploit export markets for venison
The economic value of wildlife-based land uses has been limited by failure to exploit potential export markets for venison, because of inconsistent meat supplies, lack of facilities to store venison, a shortage of EU-approved abattoirs, and lack of awareness among target markets of the health qualities of venison (Gödde, Reference Gödde2008).
Similar constraints limit the value of wildlife ranching elsewhere in southern Africa and our recommendations have regional applicability. Wildlife is outcompeting livestock throughout semi-arid areas of southern Africa, despite policies favouring the latter. A more level legislative environment would allow the full potential of wildlife-based land uses to be harnessed and could generate significant economic, social and conservation benefits.
Acknowledgements
We thank TRAFFIC East/Southern Africa and Tom Milliken for instigating this project and the German Federal Ministry for Economic Cooperation and Development and African Wildlife Conservation Fund for funding, the Ministry of Environment and Tourism, Namibian Agricultural Union, Namibian Professional Hunters Association, Conservancies Association of Namibia, WWF–Namibia, Annetjie du Preez, Harrald Marggraff, Almut Kronsbein, Danica Shaw, Laurie Marker, Chris Weaver, Jon Barnes, Uapii Kazahe, Sakkie Van Der Merwe and Wilfried Pack and all the farmers interviewed.
Appendix
The appendix for this article is available online at http://journals.cambridge.org
Biographical sketches
Peter Lindsey works throughout Southern Africa on wildlife-based land uses, the bushmeat trade and predator conservation. Carl Havemann is studying the roan antelope in Botswana. Robin Lines has undertaken applied research on large carnivore conservation in Southern Africa since 2002. Aaron Price works on livestock water quality compliance. Tarryn Retief is studying the effects of biological gradients on biodiversity in Botswana. Tiemen Rhebergen works as a consultant in geographical information systems and agriculture. Cornelis Van der Waal's research interests include applied rangeland ecology in savannah and desert systems, and mine restoration. Stephanie S. Romañach investigates wildlife responses to climate change and ecosystem restoration in the Everglades. She has worked widely in Southern and Eastern Africa and helps run the African Wildlife Conservation Fund.
