Local cost–benefit analysis for assessing the economic potential of afforestation/reforestation CDM on coca fields in the Peruvian Amazon

Background: Peru has a large extension of deforested lands converted into crops of Erythroxylum coca. Methods: A local cost–benefit analysis investigates the conversion of 500 ha of coca crops into sustainable agroforestry eligible under the Kyoto Protocol (afforestation/reforestation CDM) in the Amazon. Results: From a social and environmental perspective, CDM agroforestry generates more benefits than coca monoculture, although the cost–benefit analysis results show that illegal plantations generate higher profits. In a way, CDM makes the alternative of agroforestry to illegal coca more economically attractive than the same reforestations without carbon credit, with the advantage of attracting international investors interested in environmentally sustainable projects. Conclusion: For the permanence and success of Kyoto projects, local government and the UN continue to play a major role in avoiding leakage effects and risks of narcotraffic return.

MAP: Figure 1. Map of Peru with indications of the project area: the afforestation/reforestation CDM include two regions of central Peru.

Graph: Figure 2. Monoculture coca (Erythroxylum coca) crop. Shrubs are almost 1 m high. This is the baseline scenario of the project, indicated in the paper as H0.

Graph: Figure 3. Replacement of coca crops with selected tree species for progressive land use change, timber and nontimber productions, under the Kyoto Protocol afforestation/reforestation CDM project. This represents the shift from H0 (coca monoculture) to H1 scenarios (afforestation/reforestation CDM plantation models) of the Kyoto project (indicated in Tables 7 & 9).

Graph: Figure 4. Carbon sequestration by agroforestry systems implemented in the project area, related to a crediting period of 20 years (tCO2eq). The cumulated carbon stock corresponds to the net anthropogenic removal by sink discounted for the emissions related to harvestings and fertilizer use.

Graph: Figure 5. AR CDM alternatives and illegal coca plantation cash flow (unit: US$).

Graph: Figure 6. Sensitivity analysis: tCER price changes (US$/tCO2eq) at different coca prices (US$/kg). AR: Afforestation/reforestation; NPV: Net present value.

Graph: Figure 7. Land-use change: examples of coca fields reforested with coffee, indicated in the paper as 'project scenario' (H1 in Tables 7 & 9).

Since the 1980s, intensive deforestation has taken place in the Amazon basin (Brazil, Bolivia, Peru and Colombia), with a significant contribution to the current increase of atmospheric CO2 and associated climate changes [1]. The common practice of farmers to periodically burn the forest (slash-and-burn practice) and replace it with various cultivations (legal and illegal), have caused long-term fragmentation of the primary forest, degradation of the structure and complexity of natural ecosystems, soil erosion, and hydrogeological instability of slopes [2].

As such, illegal and subsistence crops in Central and South America are a significant cause of deforestation and CO2 emissions in tropical and biodiversity-rich areas [[3]]. The cultivation of coca (Erythroxylum coca) is mainly spread over the oriental Andean side, below 2000 m above sea level (asl), and is grows as a shrub up to 1–3 m high. Cultivation is implemented in small areas, on plots of approximately 1 ha and cultivated mostly as monoculture, although in more recent years there has been the tendency to cultivate the illegal crop under the shade of trees in order to avoid aerial detection [5]. Coca leaves can be harvested three- to five-times per year, depending on many variables: the age of the plants, soil fertility and slope gradient. When fertilizers are used, it is possible to harvest six times a year. In the 1990s, Peru was a country with major extensions of illegal cultivation of E. coca in the Amazon. Currently, Peru remains the world's second largest coca-cultivating country behind Colombia, with production of over 6200 ha in 2010. The total production of sun-dried coca leaf was estimated at 120,500 tons in 2010, and this value corresponds to a national weighted average of approximately 2100 kg/ha sun-dried coca leaf [50].

Since 1978, the national law on drugs in Peru has restricted the permission of coca cultivation for traditional uses to areas of historically verifiable pre-existing coca fields, and forbids cultivation in new areas within the national territory. The law also recognizes the National Coca Company, Peru (Empresa Nacional Coca), the monopole of the legal coca trade [6]. According to the National Institute of Statistics and Informatics, in 2006 only 8% of the coca production in Peru was destined for traditional use. The remainder was delivered via illegal narcotraffic. As a consequence, areas of coca cultivation remain in a persistent situation of social insecurity and violence [7].

Thus far, eradication activities have targeted only a few areas in the form of campaigns. In 2007, 11,056 ha of national territory experienced forced eradication and the remaining 1016 ha was voluntarily self-eradicated [8]. In recent years, the production of legal crops has increased as a result of eradication campaigns and alternative livelihood projects. Products such as palm oil and cacao, usually cultivated at these latitudes, have been successfully introduced and traded due to price increases and large international demand. Nevertheless, legal crops remain limited compared with the potentialities of Peru rural areas still under illegal coca cultivation.

In this context, afforestation/reforestation (AR) CDM project activities may represent an opportunity to generate alternative income to local communities with the trade of carbon credits, as well as timber and nontimber products, through the establishment of combined agroforestry systems that meet the needs of rural communities.

The CDM is an arrangement under the Kyoto Protocol allowing industrialized countries with GHG-reduction commitments to implement projects that reduce/remove GHG in the territory of a developing country, with the aim of contributing to the sustainable development of the host country [[9]].

AR activities under CDM may include plantations of agroforestry, commercial-grade timber species, silvi-pasture and reforestation for protection, as far as they accomplish with the national definition of forest, ensuring forest cover for the selected crediting period [[11]].

Peru fulfills the requirements needed for hosting AR CDM project activities, as defined by the Kyoto Protocol regulatory regime.

Based on a project proposal of the UN (AD/PER/99/D06), this study will assess whether the AR CDM can play a role in the efforts of illegal crop eradication in a large area of the Amazon in Peru. A carbon credit estimation is a carried out, followed by a local cost–benefit analysis (CBA) for assessing whether the economic and environmental improvements that AR CDM is supposed to provide might contribute to the disappearance of illegal coca crops. In this CBA three different scenarios are compared: business as usual (coca plantation), agroforestry systems (AR CDM) with issuance of temporary CERS (tCER), and agroforestry systems without the issuance of tCER.

This case study involves activities of reforestation on private holdings located in the central region of Peru in two subtropical areas: the 'low Amazon', Ucayali region (S 8°22'; W 7°34'; 200 m asl) and the 'high Amazon', Huánuco region (S 09°17'58"; W 67°10'57"; 600 m asl) (Figure 1). The typical ecosystem is a very humid subtropical forest: according to the data provided by the National Service of Meteorology–Hydrology of Peru, the climate of the area corresponds to a humid subtropical forest. The Ucayali and Huánuco regions are characterized by different altitude and precipitation regimes: the Ucayali region is 200 m asl, registering annual rainfalls from the lower to the upper zone between 1400 and 2500 mm, and an average temperature of 28°C. The Huanuco region is approximately 600 m asl, with annual rainfall ranging from 3000 to 4000 mm and an average temperature of 24°C [[13]].

The different ecosystems of the region can vary from forests with a high biodiversity to croplands and wetlands [15]. The primary forest is removed by slash-and-burn practices and replaced by subsistence crops such as rice, bananas, sugarcane and pastures [16]. The low productivity of subsistence agriculture has been compensated by the cultivation of illegal crops of coca, which are economically attractive albeit causing high social risk for the local communities, as well as having a negative impact to the environment (Figure 2). When the soil is completely exploited, the only vegetation types able to grow include graminoids (Paspalum virgatum, Andropon bicornis, Cyperus luzulae, Poa spp.), poor pastures (Brachearia decumbens, Bidens pilosa, Paspalum conjugatum), and few species adapted to acid soil and scarcity of nutrients such as fern (Pteridium spp.) [[17]].

To restore the fertility of soil and sustain another 3–5-year cropping phase, a long fallow period is considered necessary [20]. However, most farmers decrease the fallow period without including any rotation of crops and increase the use of fertilizers to maintain high productivities. These practices finally cause alteration of the soil quality, up to an irreversible decrease of carbon stock in the living biomass and nutrients in the soil (desertification).

The area considered for this analysis is 500 ha of deforested areas. The beneficiaries of the project are private landowners (~300 families) who voluntary expressed their interest in participating in the project by converting coca and with perennial crops (cacao and coffee), and timber plantations for carbon credit generation, assisted by the UN Office for Drug and Crime (ODC) staff (Figure 3). The private, discrete areas range from 1 to 10 ha each, thus the full project results in approximately 300 small, discrete areas belonging to private landowners. The project's plantation design is composed of the following models: forest plantations for timber production and hydrogeological protection (100 ha) to be planted on steep slopes, agroforestry with coffee (200 ha) and agroforestry with cacao (200 ha) for the production of valuable food products to be cultivated in the floodplains. All the species included in the project hypothesis are indigenous to the Amazonian basin (Table 1), while not directly belonging to the local vegetation [[21]]; some of them are hardwood trees for production of merchantable timber [24].

Theobroma cacao and Coffea arabica are perennial crops growing under the shade of timber trees in agroforestry models. The combination of mixed tree species in the plantation design is a practice well known in Central and Latin America, since the system brings advantages well suited to the livelihood strategies of rural dwellers, allowing for the sale of products at different times and income scales [[25]].

Cacao and coffee shrubs maintain a high productivity for approximately 15–20 years, and they are then coppiced so that new sprouts can regrow. They produce fruit seeds in the quantity of 1.06 and 1.03 t ha-1yr-1 according to information given by the UNODC and Divisoria Cooperative. This means that the annual production on 400 ha of the project area destined to agroforestry is equal to 423 tonnes of seeds/year.

Timber trees taller than the coffee and cacao plants are planted according to the length of their rotation cycle: fast-growing species are harvested every 8 years; slow-growing species every 30 years. Agroforestry with coffee comprises 1497 trees/ha, of which 386 are timber trees and the remaining 1111 are cacao fruit trees.

Similarly, agroforestry with coffee includes 3719 trees/ha, of which 386 are tree species for timber production in the upper canopy layer and 3333 trees/ha of coffee fruit trees under the shade of trees. Forest plantation has a tree density of 1111 trees/ha and is comprised of six species with a scaled harvesting cycle: three fast-growing species harvestable after 8 years (capirona, bolaina and pino chuncho) and three slow-growing species harvestable after 30 years.

Over 20 years the timber exploitable from the three AR CDM plantation models is distributed over the years in order to provide deferred income to the farmers and differentiated timber products to the local and international market. Over 20 years, agroforestry with cacao produces 182 m3 ha-1 of timber, 207 m3 ha-1 with coffee, and 862 m3 ha-1 with forest plantations.

The potential carbon sequestration estimates is carried out following the 'Simplified baseline and monitoring methodologies for small-scale AR CDM project activities implemented on grasslands or croplands' (AMS0001/Version 05) [51] and the IPCC guidelines [30].

Small-scale AR CDM project activities are a Kyoto project category able to isolate a limited amount of GHG (max 16,000 tCO2eq/year) and is implemented by low-income communities (defined in 6/CMP.1) [31]. Advantages of small-scale CDM lie in restricted transaction costs and simplified standard procedures; they represent a convenient choice compared with large-scale CDM when projects are implemented in limited areas and involve many beneficiaries, such as this case study.

The AR CDM in Peru is expected to generate tCER within a crediting period of 20 years [32]. The number of credits issuable by AR CDM project activity is given by the amount of CO2 removed as a result of project activities, minus the increased emissions due to project activities within the project boundaries, minus the baseline (scenario of the study area in terms of carbon in the absence of the project activity) and leakage (increase of GHG emissions, which occurs outside the project boundary, and which is measurable and attributable to CDM project activity). The application of the selected methodology is based on the following assumptions:

• ▪ The baseline net GHG removal by sinks is assumed to be zero, because in the long term a progressive degradation of the soil due to the overexploitation of lands with a decrease of the carbon stock in the soil and in the living biomass is expected [16];
• ▪ The leakage is also assumed to be zero, since it is expected that the AR CDM agroforestry system will provide at least the same goods and services of the preproject conditions, by an increase of the productivity of highly valuable products (coffee, cacao and timber). In the grasslands, few areas are used for animal grazing, and local interviews have revealed that the number of animals owned by households involved in the project is very limited. Therefore, cattle displacement is not expected to have a significant impact;
• ▪ The carbon sequestered by the project scenario was subtracted for the biomass stock present at the start of the project (coca shrubs).

To quantify the incomes related to the plantation models, the estimation of GHG sequestration directly attributable to the growth rate of the trees, as well as the timber and fruit productions, are considered. In dealing with these issues, IPCC guidelines are used while data on biomass increment, timber and fruit volumes have been provided by UNODC and local data sources [[21], [24], [27], [52]].

In addition, an accumulation of litter and dead wood over time due to dead biomass accumulation over time is expected, but these other carbon pools are not taken into consideration for the carbon credit estimate.

CERs are calculated on the basis of the carbon contained in the living biomass (above–below ground biomass). The sum of changes in carbon stocks attributable to the implementation of the project activities within the boundary is calculated using the equation:

Graph

(Equation 1)

where N(t)= total carbon stocks in biomass at time (t) under the project scenario (tCO2); LBA(t) I = carbon stocks in the living biomass at time (t) of species (i) under the project scenario (tCO2/ha); Ai = area occupied by each (i) species i (I = total number of species); 44/12 ratio of molecular weights of CO2 and carbon; the LB(t)ifor each species was estimated using IPCC Good Practise Guidance for Land Use, Land-Use Change and Forestry 2003 equations [30]:

Graph

(Equation 2)

where GTOTAL = the average annual biomass increment above and below ground (in tons dry matter ha-1year-1); R = root–shoot ratio appropriate to increments, dimensionless (retrieved from Table 3A.1.8 of IPCC 2003 [30], data related to tropical/subtropical forest); IV = average annual net increment in volume suitable for industrial processing (in m3 ha-1 yr-1; taken from national and regional bibliographies). Where no data was found, growth data of similar species were used. The values used for the calculation are reported in Table 1; D = basic wood density (in tons dry matter m-3; data taken from UNODC in Table 1); BEF1 = biomass expansion factor for conversion of annual net increment (including bark) to above-ground tree biomass increment (dimensionless; taken from Table 3A.1.10 of IPCC 2003 [30]).

Emissions due to the removal of the pre-existing biomass are taken into consideration in the first year of project activities, on the basis of carbon storage in coca shrubs. Samples of coca biomass in monoculture were collected in 2007, weighed and measured to calculate average above-ground biomass stocks of coca per ha, which were finally converted into carbon stock. The carbon content of the coca shrubs was also measured on three plots of 1 m2 each. The carbon stored in 1 ha of coca crop resulted in 3.08 ± 0.13 tCO2eq (the carbon content measured in the wood was 42.08 ± 0.47%). The pre-existing biomass is considered a net emission due to the CDM project, although the replacement of coca with agroforestry systems in the project scenario does not foresee the complete removal of pre-existing shrubs in most cases. Nevertheless, it is applied as a conservative approach by considering complete eradication of pre-existing vegetation. The related CO2 emissions are estimated to be equal to 1541 tCO2eq emitted in the first year of project implementation (Figure 2).

The ex-ante carbon sequestration of each plantation system is measured as a variation of carbon stock during a crediting period of 20 years, taking into account the living biomass only (above-ground and below-ground pools). The data of mean annual increments of the tree species have been converted into biomass and finally into carbon stock. The cumulative carbon sink is then estimated for 20 years.

The estimated biomass increment is 9.15 m3 ha-1 yr-1 for agroforestry with cacao. The agroforestry scheme with coffee shows similar performances: tree species show a mean biomass increment of 10.38 m3 ha-1 yr-1, while forest plantation has a mean increment of 43.1 m3 ha-1 yr-1, thanks to some species with a high productivity such as pino chuncho (Schizolobium amazonicum), as documented by other studies [52].

In order to quantify the impact of the AR CDM, a CBA was carried out by comparing, in monetary units, the changes from current land use (coca monoculture) to AR CDM. The economic analysis performed here follows the principle of potential compensation formulated from Hicks and Kaldor [[33]], and is adopted by the EC [[35]], for which "an action is more efficient if those that are made better off could potentially compensate those that are made worse off and lead to a Pareto optimal outcome". Despite the large use of CBA analysis [[37]], there are some criticisms about its use for environmental issues [[39]]; in the present study, a simplified CBA is used and considered as an instrument for supporting decision-makers.

The principle of potential compensation can be expressed as:

Graph

(Equation 3)

where H0 is representative of the current land use and H1 is the proposed AR CDM land use. When benefits and costs are distributed over time, from '0' to 'n', decisions are based on the comparison of the present benefit value (B) and cost (K), discounted at time '0' using the appropriate discounted rate. Formally:

Graph

(Equation 4)

Table 2 reports the main indicators commonly used to quantify the investment performance in the evaluation process: the net present value (NPV) and benefit/cost ratio (B/CR).

The CBA is carried out taking into consideration the indications of the Organization for Economic Cooperation and Development, which suggests the adoption of simpler approaches when the assessment of projects has no significant impact on the general economic system (input and output impacts on national and regional markets) while providing environmental effects on small-scale activities [35]. The evaluation compares the AR CDM with illegal coca cultivation, assessing whether agroforestry represents a worthy alternative by taking into consideration the benefits with carbon credit revenues (AR CDM) and without carbon credit revenues (no CDM agroforestry activities). This approach reflects the expectation of an entrepreneur who bases its decisions in terms of maximizing profits and evaluates its cost and revenue using the market price.

The analysis is articulated in two steps. In the first step, the absolute value for each agroforestry option and for coca plantation is quantified. In the second step, the comparative value is obtained comparing the AR CDM project with current land use (illegal coca). To assess the impact of the various parameters used in the CBA, a sensitivity analysis is also performed where a stability test of monetary analysis is carried out by modifying the magnitude of the major critical variables that can affect the results, such as: discount rates, carbon credit price, exclusion of benefits guarantee from Kyoto AR CDM and carbon market and reduction of the illegal coca price.

In the CBA the following costs and benefits are considered: timber and nontimber products, coca leaves, carbon credits (Table 3), initial investment from industrialized countries interested in carbon credit purchases, agroforestry implementation costs, and transaction costs linked to the CDM project cycle (Table 4).

Information regarding expenditure and revenue inputs (workers, machines as well as other equipment) is obtained by local market survey. The economic value of commercial wood used in the calculation was defined for each species according to local market prices related to the year 2007 (UNODC data). The market price for cacao and coffee seeds estimated for the year 2007 is equal to US$1.5 kg-1. The average farm-gate price of sun-dried coca leaf is $3–4 kg-1 dry weight [8].

In the evaluation of AR CDM options, transaction costs are also accounted for, including all costs undertaken to cover prefeasibility, feasibility and operational expenditure related to the AR CDM project cycle. Some of them are fixed costs and others are variable costs, the latter being a function of the project area extension (monitoring and verification) or of the volume of carbon sequestered by AR CDM (issuing fees). For the calculation of tCERs, a price of 5.95 $/tCER, is applied according to the market price of tCER used in the AR CDM Project Design Document and reported by the Institute for Global Environmental Strategies database [53].

The analysis was conducted for the total project area (500 ha) and the 20 years of the project period. The results are provided per unit area (1 ha). The discount rate used is the long bond yield of the 20-year Peruvian obligation (5.22%) that has been adopted as the benchmark for the AR CDM, since no benchmark for agro-forestry investments is available (Table 4). Monetary performances obtained from the CBA are evaluated using the following criteria:

• ▪ If (H1 - H0) > 0: the AR CDM project should be accepted;
• ▪ If (H1 - H0) < 0: the AR CDM project should be rejected;
• ▪ If (H1 - H0) = 0: either leave current land use as is or introduce the AR CDM project.

On the other hand, the public approach is not considered, since an illegal activity such as drug cultivation represents an evident social cost at local and international levels, which is not explicitly quantified here.

The agroforestry systems used for the project activity involve high plant diversity and multistrata plantations. The preparation of soil is limited to manual clearance of invasive weeds; no mechanization is applied. The use of fertilizers is limited to the first 3 years, the time required for perennial crops and tree seedlings to rapidly grow and overcome the competition of weeds. These activities generate an initial CO2 emission at the implementation phase of the project (see negative values in Figure 4). Based on the selected SSC-AR CDM methodology, starting from the mean annual increment of the tree species and the harvesting cycles, the estimated carbon sequestration of the mixed forest plantation is: 7.34 tC ha-1 yr-1 according to the growth rate of the species and the differentiated harvesting cycles. The perennial agroforestry models have a potential carbon sequestration capacity of between 2.12 and 4.72 tC ha-1 yr-1 (agroforestry with cacao and coffee, respectively). These results are in the range of other studies. Lapereye et al. reports for the same region in Peru a carbon sink of 3.15 tC ha-1 yr-1 for agroforestry with cacao 15 years old, and 4.82 tC ha-1yr-1 for agroforestry with coffee plantations 4 years old [41]. A study by Pandey provides an estimate of carbon sequestration by agroforestry systems distinguished between complex agroforestry models (2–4 tC ha-1yr-1) and simple agroforestry, with only one dominant species (7–9 tC ha-1 yr-1) [42]. In the same article, Pandey reports other results in which agroforestry has a potential carbon sequestration of between 0.2 and 3.1 tC ha-1yr-1. Takimoto et al. analyzed the carbon stock and sequestration potential of different agroforestry systems in arid and semiarid regions: 8–9-year-old agroforestry systems show a carbon sequestration rate of 2.8–6.75 tC ha-1 yr-1 in the living biomass, without considering the carbon stored in the soil [43].

Figure 4 shows the net gain of carbon that will be achieved during a crediting period of 20 years, calculated by a conservative approach. At the time of plantation establishment, the system results in a small source of GHG due to the loss of carbon from the pre-existing vegetation and the use of nitrogen fertilizers. The plantation then shows fast growth of newly established trees and a rising carbon sink capacity; at maturation phase, the carbon is finally stabilized in the above-ground and below-ground biomass stocks.

The net CO2 sink is 82,670 tCO2eq at the end of 20 years within 500 ha of the project area (Table 5). Carbon sequestration is a dynamic process: its trend shows a series of contractions starting from year 6 onward, which corresponds to the harvesting cycles; however, these inflections are temporary and followed by new carbon storage due to tree-planting activities, which add up with the growth of those species with longer rotation cycles.

The carbon stock in the soil is also expected to rise in the 20-year project lifetime. As reported in some studies, although tropical pastures in the humid tropics have been shown to maintain high carbon stocks in the soil, the persistence of slash-and-burn practices would result in a gradual decline of the total carbon stock [[44]]. The proposed agroforestry systems avoid the return of this trend in the next 20 years, at least.

Economic analysis can be properly expressed in terms of the cash flow of each productive alternative. The cash flow of AR CDM and coca plantations are reported in Figure 5. Illegal coca plantations register costs and revenues on an annual basis. On the other hand, AR CDM generates cacao and coffee seeds on an annual basis, while timber and carbon credit revenues are periodic.

Table 6 shows the net present value of costs and revenues over the 20 years of the project's lifetime for both alternatives. While illegal coca plantation has a revenue of $85,715/ha, AR CDM projects generate $25,334/ha, including carbon credit revenues ($2081/ha) and AR CDM initial external investment ($770). Production costs are equal to $6656/ha for planting and managing agroforestry activities, while $523/ha are transaction costs. Illegal coca plantation is particularly demanding in terms of inputs ($17,732/ha), although this is compensated by high revenues, with a present value that is 3.75-times more than the AR CDM project.

The results of the CBA are reported in Table 7. The absolute value of the two alternatives investigated has a NPV > 0 and B/CR > 1, which means in both cases that revenue is always more than expenditure. On the other hand, the comparative analysis shows a great advantage of coca plantation ($49,829/ha) and more efficiency in terms of revenue for unit of expenditure (4.83 vs 3.53%). The AR CDM project is an investment that is less convenient than current land use, despite initial investment from industrialized countries and carbon credit revenues.

For a more comprehensive evaluation, the AR CDM scenario and the same agroforestry/forestry project scenario without the CDM benefits (carbon credit revenues, funds from CDM investors) and costs (transaction costs of CDM procedures) are compared in order to evaluate the convenience of entering into the CDM project cycle. Table 8 shows that NPV and B/CR are positive for the AR CDM project and for the same agroforestry/forestry project implemented without CDM benefits. Nevertheless, the AR CDM project ensures greater NPV ($18,155) than the same project without Kyoto investments ($15,826). The indicators show a difference of $2328 for NPV and 0.15 for B/CR.

The results of this comparison indicate that the AR CDM produces much better revenues compared with project implementation without carbon credit sale.

The success of introducing carbon sequestration benefits into smallholder agroforestry activities in developing countries also come from the fact that any additional incomes for subsistence farmers who have limited alternative employment generate a positive value to the introduced practices, although the profit from carbon sales do not represent the major quota of the farmer's income, which has been demonstrated in other studies [46]. In addition, the contribution to the socioeconomic development of low-income rural communities has become an opportunity for private and public enterprises to provide 'environmental friendly' and 'social consciousness' images to multinational investors that look for carbon compensation project activities.

Sensitivity analysis provides an evaluation of robustness of the results, when some parameter values such as productivity, prices and discount rates vary. The variables investigated in the sensitivity analysis include: discount rate, AR CDM investment from industrialized countries, the area occupied by each of the three AR CDM schemes indicated in Table 1, carbon credit price and illegal coca price. The results are reported in Table 9.

Generally, the discount rate is the most sensitive factor of an investment.

The benchmark might decrease or increase, but the results do not change significantly. In both cases, for a private owner there is no convenience in introducing AR CDM. In particular, when the discount rate decreases, illegal coca plantation convenience becomes greater. Regarding the effects of transaction costs, CDM investment and carbon credit revenue as a whole lead to a reduction of NPV: the proposed agroforestry project without the carbon benefits (and costs) becomes even less acceptable for landowners compared with illegal coca plantation.

In the hypothesis where we implement only one of three AR CDM schemes, the results do not show significant changes, except for one aspect: the CDM forest plantation scheme (H1) becomes more convenient than H0 (6.57 vs 4.83%; Table 9) in case landowners select the forestry model as a unique option to be implemented in the 500 ha of the project area: for each US$ of expenditure, they could obtain $6.57.

Another applicable hypothesis concerns the changes in the market prices system (Figure 6). When the price of coca leaves is $5/kg [8] and the price of tCER is $5/tCO2eq, a significant difference between the two scenarios is detected: H1 - H0 = $-35,105. When the coca price decreases to $2/kg, while the tCER price is stable, a very small advantage would be achieved by the AR CDM (H1 - H0 = $277.39). On the contrary, if the coca price remains stable ($5/kg), the price of tCERs rise to more than $95/tCO2eq to achieve the break-even point of the NPV.

A local CBA compares two alternative land-use systems on 500 of deforested areas in Peru: illegal coca cultivation (monoculture) versus AR CDM mixed agroforestry, by comparing the monetary return of each alternative. In addition, a comparison between AR CDM convenience versus the same agroforestry project without the carbon credit revenues is carried out.

The CBA results indicate that illegal coca cultivation generates high revenue over time, and enters into the production phase after the first year of plantation. AR CDM agroforestry does not provide the same yields. Cacao and coffee can guarantee annual revenues starting from the third to fourth year onward, and forest plantations have periodic productions (Figure 7).

Regardless of the positive absolute value (NPV > 0; B/CR > 1), comparative analysis indicates that AR CDM is less profitable than coca crops. On the other hand, AR CDM has a positive effect on wealth distribution because producers can act directly in the market in a more competitive way compared with the market monopoly of coca. Assuming that the plantations will be preserved for 20 years, the CO2 sequestered by AR CDM activities generates $1293/ha carbon credits; in other words, funding contributions from industrialized countries for AR CDM activities improves the performance by carbon credit trade compared with the same project without credit generation, although not reversing the results.

On the other hand, uncontroversial benefits of agroforestry CDM have to be taken into consideration from a social and environmental point of view. The CBA carried out at a local scale does not take into account the indirect global benefits of Amazon ecosystems, such as the values of watershed services, biodiversity, pharmaceutical and food reserve, conservation of genetic resources of endemic species, and ecosystem preservation, such as other studies do [47]. A social analysis might fulfill a comprehensive evaluation by also taking into account the public costs related to the illegal consumption of coca.

The AR CDM is relevant in the fight against climate change; however, its greatest contribution comes from the opportunity to attract potential investors from industrialized countries, as well as promote sustainable and fair development of legal goods that enter into the international market with good profits for rural communities, finally providing global benefits from economic, environmental and social perspectives. The AR CDM contributes to create conditions that make the shift from illegal crops into agroforestry activities more attractive for landowners who voluntarily accept participation in Kyoto initiatives. This result nevertheless needs to be constantly supported by a leading responsibility of local and international institutions to efficiently contrast the return of narcotraffic, avoid leakage effects outside the project area and ensure public safety [48].

For all these reasons, it is strategic that an improved cooperation between the UN, the Peruvian Government and local communities continues in order to make the permanence of alternative activities such as the AR CDM more feasible, as well as increase the disadvantages and penalties to farmers who persist with coca cultivations by implementing adequate policies [48].

The dynamic fluctuations of coca cultivation is a phenomenon well known by the UNODC. A constant remote monitoring activity by the UNODC shows a variable increase and decrease in illegal crop development according to a number of variables linked to the illegal global market of drugs, as well as the policy dynamics of South America and other countries. North America, South America, and western and central Europe continue to be the world's largest cocaine markets.

On the other hand, eradication efforts by local governments rose from 2006 onward: in Peru eradicated lands rose from approximately 7500 ha in 2005 to more than 12,000 ha in 2011. The same trends have occurred in Colombia and Bolivia.

This means that in areas where an active and persistent intervention of public security and policy is carried out against illegal cultivations, there are increasing possibilities in future years to spread the conversion of degraded illegal crops into sustainable agroforestry systems under AR CDM, enabling rural communities to conduct highly profitable and legal activities, and establish safer living conditions.

The orientation of future climate policy still renews and incentivizes the development of CDM during the second Kyoto period (2013–2020), as was decided during the COP of Durban (COP 17; South Africa) in December 2011.

Table 1. Biomass increment of trees in the afforestation/reforestation CDM plantation schemes.

1 †Data of wood density and tree growth rate have been provided by the UN Drug and Crime office, local sources.

Table 2. Indicators of cost–benefit analysis.

Graph

Table 2. Indicators of cost–benefit analysis.

Graph

Table 2. Indicators of cost–benefit analysis.

Graph

Table 2. Indicators of cost–benefit analysis.

Graph

Table 2. Indicators of cost–benefit analysis.

2 B/CR: Benefit/cost ratio; NPV: Net present value. Formulas refer to [35].

Table 3. Economic variables used in the cost–benefits analysis.

3 †Data of wood density and tree growth rate have been provided by UN Drug and Crime office, local sources.

Table 4. Small-scale afforestation/reforestation CDM transaction costs.

4 AR: Afforestation/reforestation. Data taken from [49].

Table 5. Volume of harvested timber and carbon sequestration over 20 years.

5 †Authors' own elaboration starting from UNODC data. ‡Authors' data processing from Wood biomass to carbon dioxide (CO2eq).

Table 6. Present value of revenues and costs of afforestation/reforestation CDM.

6 Benchmark: 5.22%; unit: US$/ha. †1 ha of AR CDM has 40% agroforestry with cacao, 40% agroforestry with coffee and 20% forest plantation. AR: Afforestation/reforestation.

Table 7. Investment performances in 20 years period (benchmark: 5.22%.

7 †1 ha of AR CDM has 40% agroforestry with cacao, 40% agroforestry with coffee and 20% forest plantation. AR: Afforestation/reforestation; B/CR: Benefit/cost ratio; NPV: Net present value.

Table 8. Comparative analysis of agroforestry activities with and without carbon credits.

8 †1 ha of AR CDM has 40% agroforestry with cacao, 40% agroforestry with coffee and 20% forest plantation. Data are referred to 1 ha (benchmark = 5.22%). AR: Afforestation/reforestation; B/CR: Benefit/cost ratio; NPV: Net present value.

Table 9. Sensitivity analysis.

9 †1 ha of AR CDM has 40% agroforestry with cacao, 40% agroforestry with coffee and 20% forest plantation. AR: Afforestation/reforestation.

The permanent destruction/harvesting of forests in order to make the land available for other uses. Deforestation is considered to be one of the main contributing factors to global climate change, since trees absorb CO2, produce oxygen and perpetuate the water cycle. The Kyoto Protocol formally and economically recognizes this unique role the forest has to mitigate and contrast global warming effects.

The geographical area of the Amazon river that belongs to many countries: Brazil, Bolivia, Peru and Colombia. In the Amazon the most biodiverse tropical forests are currently threatened by illegal harvesting and forest burning.

Erythroxylum coca (coca) is a high-altitude South American shrub whose leaves are the source of cocaine. This plant has a long history of human use, but the illegal narcotraffic made this species under control of cultivation in the South America.

One of the mechanisms recognized by the Kyoto Protocol for achieving carbon credits, allowing industrialized countries with a GHG reduction commitment to implement projects that reduce/remove GHGs in the territory of a developing country.

Also called 'carbon offset', this is a financial instrument that represents a tonne of CO2 removed or reduced from the atmosphere, achieved by a dedicated project, which can be used by governments, industry or private individuals to offset damaging carbon emissions that they are generating. Afforestation and reforestation activities are a key means by which existing emissions can be removed from the atmosphere.

The combination of annual and/or perennial crops mixed with forest species, compliyng with the ecological and socio-economic needs of the rural communities. Agroforestry systems allow the production of food and timber at different times and income scales.

An internationally recognized method used for assessing the monetary value of different scenarios in which costs and revenues of each component are taken into consideration for economic assessment as a support for the decision-making process.

• ▪ Illegal and subsistence crops in Central and South America are a significant cause of deforestation in tropical and biodiversity-rich areas.

• ▪ The afforestation/reforestation (AR) CDM project activities may represent an opportunity to generate alternative income for local communities with the trade of carbon credits and timber/nontimber products by agroforestry systems that meet the needs of rural communities.

• ▪ At the time of AR CDM plantation establishment, the AR CDM system has resulted in a source of GHGs due to the loss of carbon from pre-existing vegetation. The project scenario then quickly shifts to a carbon sink because of the fast growth of trees. The carbon is stored in the above-ground and below-ground biomass of agroforestry: the net sequestration of carbon dioxide is 82.670 tCO2eq at the end of 20 years within 500 ha of the project area.

• ▪ Timber is another commodity, with an annual aggregate production of 16.43 m3/ha/year for a production of 8215.23 m3/year on 500 ha. In addition, an annual production of nontimber products, such as cacao and coffee seeds (423.00 tons/year), is generated within the project area.

• ▪ Using a cost–benefit analysis, the net present value of costs and revenues over the 20 years of the project's lifetime was calculated for both alternatives (illegal vs legal crops). Illegal coca plantation showed a revenue of US$85,715.31, with a present value 3.75-times more than the AR CDM project; the AR CDM project generated $25,333.86, including carbon credit revenues. According to the monetary analysis, the hypothesis to introduce a Kyoto project such as the AR CDM should in theory be rejected, but there is evidence that coca plantations generate an expensive cost for environmental and social security.

• ▪ The profits achievable by the AR CDM come from timber and nontimber products, and carbon credits (CERs). This latter output has the advantage to attract potential investors interested in sustainable actions in favor of low-income communities.

• ▪ The AR CDM still has relevance in the fight against climate change; however, its greatest contribution comes from the ability to promote the sustainable and fair development of the local area. AR CDM monetary contributions reduce the disadvantages of landowners when replacing current land use into alternative CDM agroforestry.

• ▪ The cost–benefit analysis results might provide useful information to local institutions and policy makers for future strategic decisions. Indeed, the AR CDM is based on a participatory approach and local farmers can voluntarily decide to change their current illegal crops into alternative agroforestry eligible under the Kyoto Protocol, but this is possible only where the local government and UN show a high level of presence and responsibility to counteract the action of narcotraffic return over time.

The authors would like to thank the following Cooperativa Divisoria members and United Nations Office of Drug and Crime of Lima (UNODC) staff for helping in the in the field, as well as with data processing: T Herrera, C Torres, R Calderon. The authors also thank the National University de la Selva, Peru, for contribution to laboratory analysis, supported by students RV Alegría and DB Peña.

This study was undertaken as part of the bilateral Cooperation between the Italian Ministry of Environment Land and Sea and the UNDOC. Afforestation/reforestation CDM activities described refers to the project of alternative development AD/PER/99/D06 of United Nation of Drug and Crime 2004–2007: 'Recover of degraded ecosystem by illegal crops in the Aguaytia and Tulumayo watershed', led in 2007 by HJ Wiese of the UNODC office. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.