Hunting safaris are traditionally sold as ‘packages’ based on dangerous and charismatic key species (lion, elephant, leopard, buffalo Syncerus caffer and rare antelope species) that demand higher prices and longer hunts. Income is accrued through daily rates, which are paid by clients regardless of whether hunts are successful, and trophy fees. We used data from standardized hunt reports (n = 267) submitted by clients (www.thehuntingreport.com, accessed June 2011) to establish typical hunt packaging for the five main lion hunting countries (Table 1), and obtained mean prices (daily rates, trophy fees, and the minimum duration of hunts) for packages by surveying operator websites in 2005 (n = 114) and 2011 (n = 165). Operators were randomly selected from lists of those presenting at US and European hunting conventions and the websites of a minimum of 10–15 from each country sampled (where it was possible to find that number of sites). We compared the prices of key species hunts and assessed changes in prices from 2005–2011. We used the compound US inflation rate to convert 2011 hunt prices into 2005 US dollars, and compared these prices with actual 2005 hunt prices to determine the real increase or decrease in hunt prices during the period.

Estimates for income earned from trophy hunting in each country were obtained by collecting data on hunting quota and off-takes from as many different hunting areas as possible. In Tanzania, hunting quotas were available from all 143 of the hunting blocks in the country from 2007 (since when, some blocks have been subdivided, taking the total number of blocks in the country to 176, V. Booth unpublished data). Mean percentage utilization of quotas for each species was obtained from the Tanzanian Wildlife Division and applied to quotas in each block to provide an estimate of typical off-takes. In Zimbabwe, data on percentage utilization of quotas were provided by hunting operators for 23 hunting blocks. Mean quota utilization of each species was then applied to the 2011 quota data for all (state owned) safari and forestry areas in Zimbabwe (provided by the Parks and Wildlife Management Authority), and the two largest private conservancies in the country. Community and privately owned hunting blocks (except large conservancies) were excluded from this analysis in Zimbabwe (except for cases where actual off-take data were available) because the status of wildlife in such areas is highly variable so we did not feel confident applying mean quota utilization data to them. For Zambia, 2007 quota data were obtained for the Game Management Areas, excluding a small number of private ranches on which hunting occurs. Mean percentage off-takes (derived from 13 areas, R. Martin unpublished data), were then applied to quotas from Zambian game management areas. In Namibia, data were only available for community conservancies, and so state and privately owned hunting areas were excluded from the analysis (though lions are not hunted in state concessions, and the species only occurs on 8.2% of Namibian farmlands so are rarely hunted, [14]. For Namibian communal conservancies, we only had quota data, and no information was available on the percentage utilization of quotas. Consequently, we used mean quota utilization for each species from the other countries in the analysis. For species that were unique to Namibia, we applied the percentage utilization value from the most ecologically similar species (e.g. for Hartmann's mountain zebra Equus zebra, we used the percentage utilization value for plains zebra Equus burchelli). In Mozambique, hunting quota data were available for all hunting blocks in the country (from the Ministry of Tourism), but data were excluded for all areas except those for which quota utilization data were available because many hunting areas in that country are severely depleted and quotas bear little resemblance to actual off-takes. Off-take data in Mozambique were limited to the Niassa Reserve (9 blocks, V. Booth unpublished data) and the Coutada 9 and 13 hunting blocks (data provided by operators).

We estimated income (sum of daily rates and trophy fees) accrued per block using off-take data and hunt package prices obtained from the 2011 online survey. We assumed that animals would always be hunted in the most lucrative package available, with excess non-key species hunted in 7-day specialized ‘plains-game’ (primarily antelope) hunts (with the exception of Tanzania and Zambia where plains game hunts are rarely sold). To estimate the total number of packages sold, we multiplied off-take of key species by the mean success rates of hunts (calculated from the hunt return data). This ensured we accounted for daily rates earned from unsuccessful hunts where safaris for key species are paid for by clients, but the target animals are not successfully hunted. To estimate the financial value of each species to a given hunting block, we estimated earnings from trophy fees and daily rates for key species (i.e. those used to sell hunt packages) and non-key species, as follows:

Financial contribution of key species = (% of total trophy fee income comprised by that species*daily rate income from all key species hunts)+(trophy fees from that species*off-take of that species)

Financial contribution of non-key species = (% of total trophy fee income comprised by that species*daily rate income from all plains-game hunts)+(trophy fees from that species*off-take of that species).

Data on the start up and running costs of trophy hunting operations were obtained through a randomized survey of hunting operators at US hunting conventions (Dallas and Houston Safari Clubs, Atlanta Africa Hunting Show), using a structured questionnaire survey following methods outlined in [12] (Table 2). At the shows, an attempt was made to survey every African operator present that sells lion hunts, resulting in coverage of 73.8% of the operators present who offer lion hunts and a sample of n = 111 operators. Operators were asked to determine the length of lease of their hunting block(s), and provide an estimate of the total start-up and annual running costs (split into fixed and variable) associated with their hunting operation.

When assessing the viability of hunting operations under different scenarios of lion hunting, we assumed a capital structure of 50% equity and 50% debt for initial investment. Start up costs were calculated by multiplying the concession size by the mean start up costs/km², within bounds set as the minimum and maximum estimates made by operators for start up costs in each country. We split start up costs into lease acquisition costs (60%), camp (10%) and vehicles and equipment (30%). Projected income from trophy hunting was used to calculate income per km² in each hunting area, followed by a mean for each country. We calculated Net Profit Before Tax (NPBT) by subtracting depreciation, interest, and running costs from Revenue. Depreciation on capital investments was calculated by dividing the cost of the investment by the term of the lease, except for vehicles and equipment, which were depreciated over a 5-year period. Interest was calculated at 4.25%, (the US prime interest rate plus one percent, which is the mean rate for commercial loans of medium risk in the US (http://www.federalreserve.gov/Releases/E2/Current/default.htm, accessed July 2011). A standard rate was used for all countries, as hunting operators are often not from the country in which they operate and are thus likely to source funds outside of those countries. Running costs were split into fixed (61%) and variable (39%) based on the mean of estimates from the operator survey, and were converted to a ‘cost per client day’ based on estimated client days in each block (estimated as the number of hunts of each type sold multiplied by the number of days for which such hunt packages are sold). Net Profit After Tax (NPAT) was calculated by reducing NPBT by the country-specific corporate tax rate (assuming NPBT was positive).

For each block, we divided NPAT by the start up costs to calculate Return on Investment (ROI). ROI was compared with a hurdle rate to evaluate financial viability of operations. For this hurdle rate, we used the Weighted Average Cost of Capital (WACC) of a major tourism company, which it uses to evaluate new projects in sub-Saharan Africa, adjusted to accommodate our assumed capital structure and cost of debt, which resulted in a WACC of 6.96%. The adjust WACC of 6.96% is used as the hurdle rate to evaluate financial viability of the hunting operations. We acknowledge that, in addition to financial considerations, there is a “lifestyle” element to the decision to invest in a hunting operation, but we have ignored this for the purposes of our analysis, as it is impossible to quantify. We calculated ROI under three scenarios: i) current lion off-takes, ii) off-takes reduced to 0.5 lions/1000 km² (the recommended sustainable off-take for lions in Tanzania, excluding the Selous Game Reserve; [3], and iii) universal discontinuation of lion hunting. The recommended harvest rates are more conservative than those advised by other authors [15], [16] and would thus likely be safe to apply in other countries.

Key weaknesses in our analyses were estimates of start-up and running costs of hunting operations, where we applied mean values across hunting blocks within each country. In reality, costs vary among blocks due to the varying prices and remoteness of concessions. This weakness affected the accuracy of predictions of profitability of individual blocks, and the proportion of blocks that are estimated to be profitable. However, variation in these parameters did not affect the key conclusion: that the presence/absence of lions on quota affects the proportion of hunting blocks across which trophy hunting is viable.

We used mean percentage utilization of quotas to estimate off-takes, which may have introduced error in estimates of revenue in some blocks. For Namibia, we lacked data on actual off-takes and had to rely on estimates of the proportions of quotas utilized from other countries. For Tanzania and Zambia, the most recent quota data was from 2007 (though the pricing data for all countries was from 2011), and quotas may have changed since then (for example, lion quotas in Zambia were cut in 2011, although the effect on estimated profits is presumably small as percentage utilization would likely increase when quotas are lowered).

The subset of hunting blocks for which we had data may not be entirely representative in some cases: for Namibia and Zimbabwe, private land was largely excluded from the analyses, and in Mozambique we had a relatively small sample of (albeit large) blocks in our analysis. Finally, hunt reports were submitted voluntarily by clients and it is difficult to gauge how representative the sample was of the total number of hunts conducted (for most species, they represented <10% of hunts undertaken annually). However, these were the only standard data available (reporting of hunt composition, duration and success is notoriously poor among statutory authorities) and we confirmed the accuracy of packaging and relative success of hunts with multiple operators from each of the countries assessed. Despite these shortcomings, the data presented provide novel and robust insights into the financial significance of lions to the viability of trophy hunting.

With the exception of rhinoceroses (Ceratotherium simum and Diceros bicornis) in Namibia and South Africa and exceptionally large elephant trophies, lions generate the highest revenue per hunt of any species in Africa. Prices for lion hunts are particularly high in Tanzania and were also costly in Botswana prior to the moratorium (up to $140,000/hunt, G. Rann, Rann Safaris, pers comm.), presumably because of the renowned trophy quality of Kalahari lions (www.scirecordbook.org; accessed June 2011). The price of lion hunts has increased faster than most species, and this will likely continue as the supply of wild lion trophies declines (3). Lions have the potential to suffer from an anthropogenic Allee effect, where consumers place disproportionate value on rare animals, driving a cycle that could theoretically lead to a species' extinction [17]. However, contrary to that suggestion, the cheapest lion hunts offered are from West Africa where the species is considered Regionally Endangered [18]. The future for major increases in the price of lion hunts is also likely to be undermined by the rapid increase in availability of cheap, high quality trophies from captive-bred lions in South Africa (Lindsey et al. unpublished data) which will assumedly undermine the financial value of wild lions.

If lion hunting was restricted there would be scope in some places for compensating for lost income through more strategic packaging of quotas, and by increasing quotas and off-takes of other key species. In Tanzania and Zambia, the government imposes tight restrictions on the way in which trophy hunts can be packaged and sold, through dictation of hunt lengths and species compositions of certain packages. If operators were free to market hunts as they wished and in response to market forces, quotas could be sold more efficiently and profitably [29]. Elephant off-takes in Mozambique (32.8/year from a population of 14,079, Blanc et al. 2007), Tanzania (58.8 from 108,816), and Zambia (3.6 from 16,562) are less than the 0.5% that is considered to be sustainable for trophy off-takes and could potentially be increased in some areas [30], [31], [32]. In other cases, trade restrictions on key species imposed by hunting import countries such as the US could be removed to elevate profitability of trophy hunting. However, there is a limit to the extent to which other species can be used to compensate for income from lion hunting as homogenization of the trophy product among countries may compromise the viability in less popular and accessible countries [12].

The trophy hunting industry is not dependent on lions for viability in most areas, and other species (notably elephant, buffalo and leopard) are more important in financial terms. However, in a significant minority of hunting areas lions are of key importance, and if hunting of the species was discontinued, hunting operations comprising approximately 59,538 km² could potentially become unviable in the countries assessed, posing a risk that those areas could be lost as lion habitat. This represents 11.5% of the 516,738 km² where lions are currently hunted in the countries included in the analysis (Balme et al. unpublished data), and at least 3.6% of total lion range (1,674,664 km²; [33]). Furthermore, lions are hunted across ∼94,000 km² in Central African Republic, ∼7,000 km² in Burkina Faso and ∼4,000 km² in Benin (Balme et al. unpublished data) and inclusion of those countries in the analysis would have likely significantly increased the size of the area across which viability of trophy hunting would be lost if lion hunting was banned (particularly given the low numbers and diversity of other key species hunted in those countries). Even where viability is predicted to be retained, restrictions on lion hunting would affect the overall profitability of trophy hunting and thus reduce the competitiveness of wildlife-based land uses relative to alternatives such as livestock production. Net returns from livestock in semi-arid African rangelands ($10–$30/km²/year in areas with 400–800 ml of annual rainfall [34]) are similar to those from trophy hunting in some areas (mean $–24 to $164/km²); hence, maximizing returns from hunting is key to ensuring competitiveness of wildlife-based land uses.

In addition to the potential loss of habitat, restrictions on lion hunting could potentially reduce the tolerance of communities in some areas, such as on private land or in Namibian conservancies where land holders are the effective owners of the wildlife resource [25]. Restrictions on lion hunting may also reduce the funds available for management activities such as anti-poaching and community outreach. State budgets for most African parks are below that required to protect them effectively and there is typically little state funding for hunting blocks [35]. In some cases, investments from hunting operators in anti-poaching activities are notable. For example, trophy hunting generates $380–400,000/annum for Niassa National Reserve, almost 20% of the total funds required to maintain the 42,000 km² protected area [36]. Niassa is the focus of lion distribution in Mozambique and the large population (730–1,000 individuals) is believed to be stable or even increasing [37]. Hunting operators in Savé Valley Conservancy (SVC) in Zimbabwe (who removed livestock and reintroduced wildlife, including lions) invest $546,000/annum on anti-poaching and employ 186 permanent scouts, enabling the lion population to increase [38] (Ox Hacking, SVC CEO pers. comm.). Similarly, operators in Coutada 9, Mozambique invest $60,000/annum on anti-poaching, have removed 5,000 gin traps, and have reintroduced lions [20]. However, returns from trophy hunting in most concessions are low, reducing available funds for anti-poaching, regardless of whether lions are hunted. Recent estimates suggest that as much as $1,000/km² may be required to maintain lion populations at a density of at least 50% of their potential carrying capacity (C. Packer, unpublished data) suggesting that hunting may generate a fraction of the funding needed to protect lions effectively in the long term. Similarly, in some countries (notably Tanzania and Zambia), leases of hunting concessions are short, undermining incentives for operators to invest in protecting wildlife. In countries where earnings from hunting are centralized (notably Tanzania and Zambia), wildlife is likely to disappear from hunting blocks in the absence of reform to make communities the primary beneficiaries of trophy hunting (in areas where hunting occurs on community lands) (22, 23, 39).

While trophy hunting could survive without lion hunting in most areas, the species is an important financial component of an industry which is marginal in some areas and vulnerable to reductions in profitability. Blanket trade restrictions would unfairly punish countries where lion hunting is well managed, and could be negative for lions by undermining the competitiveness of wildlife-based land uses and by undermining tolerance for lions which are typically a high-cost species due to their tendency to kill livestock. A preferable alternative would be the introduction of recommended quotas (0.5 lions/1000 km²) as such an intervention would allow lion hunting to be sustainable, while retaining conservation-incentives from trophy hunting. Sustainability would be enhanced further if age-based regulations were implemented (e.g. as in Niassa National Reserve) [37] and if governance of the industry was improved to provide communities with greater stakeholdings. Temporary moratoria on lion hunting could be used to allow recoveries in areas where hunting is implicated in negative lion population trends. Lion populations recover quickly when the pressure of excessive harvests is removed. Consequently, over-hunting is likely to pose little threat to the long term persistence of lions so long as interventions are made to address excessive quotas where they occur. Conversely, if lion hunting was banned, and wildlife-based land uses were replaced by alternatives in some areas, the long term prospects for lion conservation in those areas would be poor and reversing negative trends would be unlikely. Precluding lion hunting may therefore be a greater long term risk to lions than over-hunting. That said, urgent efforts are needed by range states to reform lion hunting management, and temporary moratoria could be considered for use as levers to promote such changes.