Forest Policy and Economics 51 (2015) 47–56

Contents lists available at ScienceDirect

Forest Policy and Economics

journal homepage: www.elsevier.com/locate/forpol

Forest use and agriculture in Ucayali, Peru: Livelihood strategies, poverty

and wealth in an Amazon frontier
Roberto Porro a,b,⁎, Alejandro Lopez-Feldman c, Jorge W. Vela-Alvarado d
a
Embrapa Eastern Amazon, Trav. Dr. Enéas Pinheiro s/n, Belém, PA 66095-100, Brazil
b
World Agroforestry Centre, Latin-American Regional Office, c/o CIP, P.O. Box 1558, Lima 12, Peru
c
División de Economía, Centro de Investigación y Docencia Económicas, CIDE, Carretera México-Toluca 3655, Col. Lomas de Santa Fe, 01210 México, D.F., Mexico
d
Universidad Nacional de Ucayali, CFB Km 6, Pucallpa, Peru

a r t i c l e i n f o a b s t r a c t

Article history: The relevance of forests to rural well-being and poverty reduction remains a controversial issue. This paper ex-
Received 5 April 2014 amines patterns of association between household wealth, poverty, and livelihood dependency either on forest
Received in revised form 31 October 2014 extraction or agricultural activities in Ucayali, Peru. The analysis is based on survey data of 578 households
Accepted 2 December 2014
with geographical, ethnic and environmental heterogeneity. A typology of economic strategies was defined
Available online 24 December 2014
through relative income shares derived from agriculture, forest, wages and other income sources. Our results
Keywords:
show that households have multifaceted livelihood systems. While forest/environmental products provide near-
Poverty and environment ly 40% of total income, agriculture is critical to both indigenous communities and to farmers of non-Amazonian
Rural livelihoods origin. We test the hypothesis that households relying on agriculture are wealthier than forest-dependent house-
Peruvian Amazon holds. In addition, we examine the role played by ethnicity and location as interacting variables. The analysis sug-
Deforestation gests a strong role of specific ethnic and locational configurations in shaping income and asset patterns, with
some weak evidence of statistically lower poverty levels being credited to dependency on forest products.
Context-specific assessments of livelihood–environment interactions provide critical insights to development
and environmental policies and programs, which need to recognize different forms through which households
integrate forest use and agriculture.
© 2014 Published by Elsevier B.V.

1. Introduction
The globally valued ecosystem services that forests provide are
critical to ever-increasing environmental agendas seeking to mitigate
climate change through policy. The implementation of interventions
derived from such policies often affects social relations, cultural traits,
and living conditions of those who live in or near forests. Households
in forest frontier communities engage in multiple and site-specific activ-
ities that provide for their livelihoods. Therefore, livelihood strategies
adopted by these social groups should be carefully considered in pro-
grams targeted to enhance forest-related ecosystem services. Accurate
assessments of patterns, drivers, and consequences of local livelihood
systems, and particularly of synergies and trade-offs between agricul-
ture and forest use, are essential to increase the fairness and efficiency
of initiatives affecting indigenous and smallholder communities
(Kaimowitz, 2008). Moreover, comprehensive examinations of these
⁎ Corresponding author at: Embrapa Amazônia Oriental, Pavilhão de Pesquisas, Trav. Dr.
Enéas Pinheiro s/n, Belém, PA 66095-100, Brazil. Tel.: +55 91 3204 1129, +55 91 99615
6655; Fax: +55 91 3276 9845.
E-mail addresses: roberto.porro@embrapa.br, porro.roberto@gmail.com (R. Porro),
alejandro.lopez@cide.edu (A. Lopez-Feldman), jvelaunu@gmail.com (J.W. Vela-Alvarado).
livelihoods reveal the dynamic interplay of economic activities that pro-
vide them greater resilience to vulnerable social groups. Such studies
may offer relevant insights for strong safeguards and social co-benefits
to enhance the “bundle of powers” to which these resource users in
the forest margins are entitled (Chhatre et al., 2012; Larson and Ribot,
2007; Ribot and Peluso, 2003).
In this paper we discuss situations where forest products are the
main source for local livelihoods and others where there is a higher
dependency on agriculture or other activities. We draw on empirical
evidences from a case study with diverse ethnic composition at distinct
settings in the Ucayali region, Peruvian Amazon. Bridging socioeconom-
ic and environmental research domains, we examine the patterns of as-
sociation between economic indicators and livelihood dependency
either on forest extraction or agricultural activities. We use income
and wealth to assess outcomes of household decisions regarding pro-
ductive strategies. We identify household economic activities and relate
them to income. In addition, we draw on capital assets as wealth indica-
tors to attenuate the bias of assessments based only on income, which
often do not capture the realities of multidimensional rural livelihoods
and provide incomplete explanations for local patterns of behavior
(Bebbington, 1999; Brandolini et al., 2010; Reardon and Vosti, 1995).
Through the dual focus on assets and income, we show patterns of

http://dx.doi.org/10.1016/j.forpol.2014.12.001
1389-9341/© 2014 Published by Elsevier B.V.
48 R. Porro et al. / Forest Policy and Economics 51 (2015) 47–56
association between wealth, poverty (assessed through a line of
US$1.89/day established by the Peruvian government) and the use of
forest products.
We test the hypothesis that greater dependency on forest products
is negatively correlated with income and wealth. In order to do so, we
structure a typology of livelihood strategies based on household depen-
dency on forests, agriculture, or other activities; high dependency being
considered when relative shares of agriculture or forest-derived income
are greater than two-thirds (66.6%) of total annual income. We then
clarify the observed contrasting effect of livelihood orientation on in-
come and asset levels through incorporation of ethnicity and location
as interacting variables. Our results suggest a strong role of specific eth-
nic and locational configurations in shaping income and asset patterns,
with some weak evidence of statistically lower poverty levels being
credited to dependency on forest products.
2. Background
Demand for studies on the relationship between forests and liveli-
hoods increased in the past few years with the acknowledgement
of avoided deforestation as a primary strategy for reducing carbon emis-
sions (Anderson and Bows, 2008; Kindermann et al., 2008; Ramankutty
et al., 2006). Baseline biophysical and socioeconomic assessments be-
came required protocols for sites of potential implementation of
schemes targeting reduced emissions from deforestation and forest
degradation (REDD) in developing countries. While these increased
demands require greater attention in scholarly debates on the topic,
field assessments present additional research questions and stakehold-
er perspectives that ought to be included in new research.
Scholars have convincingly presented different perspectives on
assessing poverty and well-being in forest environments (Guedes
et al., 2012; Narain et al., 2008; Nielsen et al., 2012; Sheil and Wunder,
2002). Studies have increasingly examined the extent to which multiple
dimensions of household heterogeneity in capital assets (Ellis, 2000;
Scoones, 1998) influence engagement in forest use and dependency
on forest or environmental products (e.g., Kamanga et al., 2009; Kar
and Jacobson, 2012; and see Vedeld et al. (2007), for a meta-analysis
of more than 50 cases). Yet, the links between forests and livelihoods
(the so-called forest–poverty nexus) are more often expressed through
dependency levels based on income estimations (Vedeld et al., 2004;
Wollenberg and Nawir, 1998; Wunder, 2001).
The level of poverty in Africa (Ravallion et al., 2005) has led many of
these studies to focus on sub-Saharan countries. The Amazon region is
relatively understudied in this respect, despite presenting the largest
tropical forest extension in the world, with increasing socioeconomic
inequalities and complex issues of access and use rights (Corbera
et al., 2011; Larson et al., 2008; Pacheco et al., 2012). Examinations of
Amazonian livelihood–environment links (e.g., Anderson and Ioris,
1992; Coomes and Burt, 2001; Duchelle, 2009; Pattanayak and Sills,
2001; Pinedo-Vasquez et al., 1992; Shone and Caviglia-Harris, 2006;
Stoian, 2005; Vosti et al., 2003) have more often been confined to a spe-
cific social group, either indigenous communities or long-term dwellers
with open access to resources or recently arrived colonist farmers tied
to private land use. Our approach captures differentials in economic
strategies across specific user groups. It provides greater detail on the
role of forest-derived income, adding this aspect to a prolific set of stud-
ies conducted since the 1990s on colonist farmers' economic strategies
(e.g., Browder et al., 2004; Caviglia-Harris, 2004; Marquette, 2006;
Moran et al., 2004; Murphy et al., 1997; Perz, 2005; Walker et al., 2002).
Forests and environmental resources are seen to perform three
major roles in the livelihoods of vulnerable households (Angelsen and
Wunder, 2003; Belcher, 2005; Cavendish, 2002; Fisher, 2004; Fisher
and Shively, 2005; Godoy et al., 1998; McSweeney, 2005; Paumgarten,
2005; Shackleton and Shackleton, 2004; Vedeld et al., 2007). They
support local consumption needs, provide insurance as safety nets ei-
ther through consumption or commercialization, and can be a pathway
out of poverty. Both wealthier and poorer households rely on forests
and while the wealthier tend to extract greater aggregated quantities,
the poorer households are more dependent on forest products (Byron
and Arnold, 1999; Mamo et al., 2007). Forest products may have an
equalizing effect on income inequality (Cavendish, 2000; Kamanga
et al., 2009; Mamo et al., 2007). However, forest dependency may turn
into a poverty trap, where poverty and forests are linked in a downward
spiral in which poverty causes forest loss and vice versa (Shively, 2004;
Scherr, 2000; Sunderlin et al., 2005). Poverty and impoverishment can
indeed be a major cause of environmental degradation (Angelsen and
Wunder, 2003; Fisher and Shively, 2005; Tacconi et al., 2006; Wunder,
2001), although these conditions are magnified by broader economic
inequalities and structural factors that reduce access to forest resources
and increase social vulnerability (Colchester and Lohmann, 1993;
Stonich and Dewalt, 1996).
Comparative syntheses of site-specific studies on the role of forest
products to livelihoods often address aspects such as property rights,
natural-resource governance, and modalities of market integration by
resource users. The modalities have been categorized as either subsis-
tence, diversified, or specialized strategies (Belcher, 2005; Belcher
et al., 2005; Ruíz-Pérez et al., 2004), and their correct contextualization
is critical for developing mechanisms to expand the local benefits of
trade (Scherr et al., 2003). Yet, as argued by Schmink (2004), such con-
textualization must go beyond livelihood systems to include a thorough
appreciation of the broader economic and political forces of the social
structures that impact the viability of forest management. As stated by
Belcher (2005), the neglect of such appreciation and the lack of careful
safeguards to ensure the rights of vulnerable groups may assign an
undesirable anti-poor bias to increased market integration of forest
resources.
3. Study site
With an area of 102,410 km2 (roughly 8% of the country's total)
Ucayali is the second largest of Peru's 25 administrative regions. Its
2012 population is estimated at 490,000, of which 75% reside in urban
areas and more than 60% in its capital, Pucallpa, the second most popu-
lous city in the Peruvian Amazon. Improvement in Ucayali's social con-
ditions is shown by a considerable reduction in total poverty, from
70.5% in 2001 to 20.3% in 2010 (INEI, 2011a) and the increase in its
Human Development Index (HDI), from 0.5251 in 1993 to 0.6022 in
2007 (PNUD, 2010). Substantial demographic discrepancies exist across
Ucayali's four regional provinces, with northern Coronel Portillo and
Padre Abad provinces presenting a combined demographic density
more than 10 times greater than southern Atalaya and Purus provinces,
which are predominantly rural (65%) and unconnected to paved roads
(INEI, 2009a).
Origin and cultural groups differentiate the population in Ucayali. On
the one hand, territories of nearly 300 native Amazonian communities
cover about 20% of the region, half of this area being legally titled (IBC,
2012; MINEM-GOREU, 2007). Projected to 2012, the Pano (60%) and
Arawak (40%) ethnolinguistic families have a population of about
70,000 (14% of Ucayali's total population) (IBC, 2012; INEI, 2009b;
MINEM-GOREU, 2007). On the other hand, thousands of colonists
have settled near the Federico Basadre Highway, built in 1945, or
along the banks of the Ucayali River and its tributaries, where they
joined long-term, nontribal ribereño (riverside) communities. Most of
these settlers have a non-Amazonian indigenous background (Padoch
and de Jong, 1989, p. 103). In this paper, they are designated as mestizos,
encompassing settlers of non-Amazonian background, both indigenous
and nonindigenous.
Nearly 20% of Ucayali region's GDP derives from agriculture, live-
stock, and forestry, while timber and agricultural processing companies
contribute a substantial portion of industry's 13% share of the GDP (INEI,
2011b; MINEM-GOREU, 2007). Annual cropping is mostly based on tra-
ditional short-fallow swiddens, with progressive clearing and burning
 R. Porro et al. / Forest Policy and Economics 51 (2015) 47–56
of forest or old fallows (Fujisaka, 1997; Labarta et al., 2008).
Semiperennials, usually in swidden agroforestry (Hiraoka, 1986;
Padoch and de Jong, 1989), imply longer rotations and a less frequent
need for clearing, while perennials tend to be associated with a level
of intensification that reduces the need for clearing (Gutiérrez-Vélez
et al., 2011). Ucayali pastures, on the other hand, often follow annual
crops and are managed extensively, representing low marginal cost to
those aiming to extend the use of a cleared plot with grasses adapted
to less fertile soils (Loker, 1993). The most important crops are tradi-
tional staples (rice, maize, cassava, plantains, and beans) and cash
crops that have become more relevant in the past decade, particularly
cocoa, coffee, papaya, and oil palm. Official agricultural statistics do
not mention the production of coca, a major economic driver in Ucayali
since the 1980s, mostly at the higher landscapes of the Aguaytia basin
(Salisbury and Fagan, 2011). Indeed, in 2009 the basin featured the
highest levels of coca expansion in the country (UNODC-DEVIDA, 2010).
Logging remains a major industry due to Pucallpa's road connectiv-
ity (Ramos-Delgado, 2008). Half of the nearly 8 million ha of Ucayali's
forests were declared as permanent production forests in 2002, being
exploited through management plans and concessions granted by
INRENA, the National Institute for Natural Resources, replaced in 2008
by the Ministry of Agriculture's General Directorate of Forests and Wild-
life. In the 2005–2009 period, Ucayali's annual production averaged
315,000 m3 of round wood and 193,000 m3 of lumber (INEI, 2011c).
Conservation efforts in Ucayali have resulted in two national parks,
two communal reserves, a reserved zone, and a regional conservation
area partially or entirely located in the region. Despite such efforts,
cumulative deforestation increased from 547,750 ha in 1990 to
627,064 ha in 2000 (GOREU, 2008) and to an estimated 787,000 in
2010 (Sandra Rios, personal communication), or 9% of the region's orig-
inal forested area of 8.7 million ha. Slash-and-burn farming continues to
be a major driver of tropical forest loss in the Peruvian Amazon (Alvarez
and Naughton-Treves, 2003; Smith et al., 1999). The continued expan-
sion of such system can, however, be viewed as a rational response
from farmers to secure tenure rights in a political–ecological context
of higher demand and rising prices for land due to road improvement,
hydrocarbon and mineral extraction, industrial agriculture, logging,
and even carbon speculation.
4. Methods
In this paper we analyze data gathered in Peru by the Amazon Live-
lihoods and Environment Network (RAVA), adopting the methodology
of the Poverty and Environment Network (PEN) (Angelsen et al.,
2011) for the assessment of livelihood dependency on environmental
resources.
4.1. Data collection, site selection and sampling
We collected information and data on multiple livelihood sources
through two annual and four quarterly structured surveys at the house-
hold level, as well as two village-level annual surveys. The data was col-
lected between November 2007 and February 2009. The RAVA field
research team included four site coordinators who were members of
Peruvian institutions engaged in research and/or development activities
with target communities at research sites where they performed previous
work.1 Local undergraduate students and practitioners acted as enumer-
ators trained for data collection, supervised by the respective site
1 3
Partner institutions included: (a) Pucallpa's Station of the Instituto Nacional de
Innovación Agraria (INIA), which works with farmers in nonflooded ecosystems near the
Federico Basadre Highway; (b) Instituto de Investigaciones de la Amazonía Peruana (IIAP),
which carried out research in Shipibo–Conibo communities at flooded forest sites in the
mid-Ucayali River; (c) Universidad Nacional de Ucayali (UNU), which focuses on indigenous
and mestizo communities of the Abujao basin; and (d) Associacion de Cacaocultores
Tecnificados de Padre Abad (ACATPA), which promotes extension activities with native
Cashibo–Cacataibo communities in the San Alejandro basin, Irazola District, Padre Abad.
49
coordinator. Site coordinators performed data quality control, further re-
inforced by the first author, who provided protocols and training to en-
sure comparability.
The socioeconomic and geographical context addressed by this
study encompasses the two provinces in Ucayali that have larger demo-
graphic density and greater agricultural development: Coronel Portillo
and Padre Abad. Selection of locations within these provinces consid-
ered PEN recommendations for intrasite variation of key features such
as forest integrity, distance to markets, land tenure, and social groups
(Cavendish, 2003). We integrated purposeful, clustered, and stratified
sampling seeking similar-size sample groups of indigenous Amazonian
communities and of mestizo (non-Amazonian) farmers. The final sam-
ple of 578 households2 included 296 indigenous and 282 mestizo
households, denoting sampling intensities of 5% and 2% of their respec-
tive populations in the two provinces combined.
The ethnic structure of the indigenous Amazonian population of
Coronel Portillo and Padre Abad provinces, along with predominant
landscape patterns (in order to include both upland and flooded forest
environments), guided the selection of indigenous communities in the
sample. The indigenous sample included 140 households in upland
sites—134 of Cashibo–Cacataibo and six of Ashaninka ethnicity—and
156 Shipibo–Conibo households in lowland sites. Sampling intensity
for the Cacataibo was 12% of their population in the two provinces; for
the Ashaninka it was 2%, and for the Shipibo, 3%. Liaison with local insti-
tutions' activities at each location was critical for the adoption of the
sampling procedure. Once we defined the 12 settlements of mestizo
farmers and 14 indigenous communities, we randomly selected house-
holds from residents' lists produced by the respective institutions.3
Sampling intensity at the site level averaged 25% for indigenous and
50% for mestizo households. Fig. 1 shows the location of the 26 commu-
nities where the study was conducted.
4.2. Income assessment
Contribution of multiple sources of income was based on informa-
tion obtained from household surveys. We summarized income obtain-
ed from quarterly surveys in seven categories: forest and environmental
products, fishing, agriculture, livestock and animal products, wages,
businesses, and other sources. Income data combines market sales
revenues and value of products channeled to household subsistence,
the latter by assigning “farm-gate” prices derived from local-level trans-
actions. Production costs (except household labor) are deducted from
gross values, and total income therefore refers to reported net amounts.
The survey adopted a three-month recall period for income from agri-
culture, livestock, and other sources; a one-month recall was used for
forest and environmental products, fishing, wages, and businesses,
and results scaled to the three-month period to allow computation of
annual income, calculated through the integration of the seven catego-
ries. We converted results from Peruvian nuevos soles to U.S. dollar
using the 2008 average exchange rate (1 USD = S./2.87). Analyses
and results present income adjusted to adult equivalents units (adult
equivalents)4 in order to control for differences in household demo-
graphics (Blackorby and Donaldson, 1991).
2
Of the 593 households initially surveyed, 15 that failed to respond to at least two quar-
terly surveys were excluded. Income of missing trimesters (for households who answered
two or three surveys) was imputed using Stata's impute command. Using information of
the household's existing surveys, impute runs regressions by what is known as best-
subset regression to fill in the missing values for each income category. For details on
the command, see Stata Press (2007).
Although a truly random sample was not obtained, the adopted sampling procedures
included geographic, ethnic, and environmental heterogeneity, being adequate for repre-
sentativeness of the rural population in the two Ucayali provinces, and therefore to inform
policy.
4
For greater accuracy of comparative household demographic attributes, this study
used an adult-equivalent scale with the following weights, based on the age of household
members: (0–1: 0.1 ad. eq.), (2–3: 0.2), (4–5: 0.3), (6–7: 0.4), (8–9: 0.5), (10–11: 0.6),
(12–13: 0.7), (14–15: 0.8), (16–17: 0.9), (N17: 1).
50 R. Porro et al. / Forest Policy and Economics 51 (2015) 47–56
Fig. 1. Communities in Ucayali-Peru. Map by: Laboratorio de Sistemas de Información Geográfica, CIDE.
4.3. Asset valuation
The following capital asset endowments were examined: (1) land
investments through perennial crops and pastures; (2) livestock
herds; (3) ownership of durable goods, productive equipment, and
rural construction; (4) number of fruit trees; and (5) house materials
and conditions. We estimated the use-value of the last two items and
added to the first three to compose a monetary, capital asset-based in-
dicator of wealth.
Investment in perennial crops (cocoa, oil palm, peach palm) and
pasture formation was preferred as the indicator of land-based capital
assets rather than land ownership since land ownership could only be
applied to mestizo communities; for indigenous households, land is
not a private asset. We estimated land investments through the applica-
tion of average 2008 market prices for areas under the respective land
use.5 Households also invest in the land surrounding their dwellings
by setting up home gardens and orchards of fruit and/or timber trees.
We defined an average (across species) notional value of US$10 per
fruit tree after local consultations about a conservative valuation of
fruits that a household obtains during the lifespan of a tree. The stock
of animals reported by households in the last quarterly survey supplied
herd size, with value set according to the median prices identified
for each species.6 We also based the valuation of 10 durable goods,
10 types of productive equipment, and rural construction (4 items) on
average unit prices declared by households across the sample.7 Finally,
5
Per hectare prices adopted: oil palm (US$ 2700); peach palm (US$ 1800); cocoa (US$
1250); pasture (US$ 350).
6
Median prices for livestock in Peruvian soles are as follows: bovine: $800; horse: $600;
burro: $450; swine $135; lamb: $100; turkey: $60; chicken and duck: $15; guinea-pig: $10.
7
Durable goods: car, motorcycle, refrigerator, television set, sound system, telephone,
gas stove, sewing machine, bicycle, and radio. Productive equipment: tractor, chainsaw,
boat, canoe, generator, shotgun, scale, mill, mower, and sprayer. Rural infrastructure
items: fences, rustic warehouses, rustic corrals, and fishponds.
we valued households' dwellings based on the integration of seven
parameters according to the function:
VH = 2*S * (b + c + d + e + f + g), where:
VH: estimated value of the house
S: area of the house, in square meters
b, c, d: expansion factors8 for wall, roof, and floor materials
e, f, g: expansion factors for electricity, water supply, and bathroom
type
4.4. Household typologies
Most studies on household dependency on forests compare income
quintiles of sampled households. In this study, we adopt a reverse pro-
cedure. We set up relative income thresholds to initially define house-
hold categories according to their dependence on specific activity-
sources. We thus use the level of two-thirds (66.6%) of total annual in-
come to denote high dependency on either forest or agriculture income
(Fig. 2). In our analysis, “forest income” includes forest and environ-
mental income and revenues from fishing. “Agricultural income” also
includes income from livestock and animal products (except fish).
Groups 1 and 2 comprise households respectively featuring high forest
dependency and high agriculture dependency. Group 3 features house-
holds to whom both forest and agriculture incomes account for less than
25% of their income, being referred to as wage & business dependent.
Group 4 includes households with a balanced forest–agriculture–wage
8
Expansion factors: Wall materials: 0 = no wall; 1 = reeds/straw/grass/fibers/mud/
soil; 2 = wooden boards, trunks; metal sheets, bricks, or concrete. Roof materials: 0 =
no roof; 1 = thatch; 2 = wooden boards, metal sheets, tiles. Floor materials: 0 = clay,
earth; 1 = wood; 2 = cement, tiles. Electricity: 0 = no electricity; 1 = power company,
generator, solar panel, other. Water supply: 0 = manual (river/lake); 1 = manual (well/
spring); 2 = tapped. Bathroom type: 0 = no bathroom; 1 = latrine; 2 = inside or outside
the house, with running water.

100
Share (%) of total incomefrom agriculture
2 ty to the nearest river that is navigable year-round; and distance (km)
66
HAD
4 5.1. Descriptive statistics: income sources and capital assets
BFA
25
3 1
WBD
0 25
share (%) of total income from forest
Fig. 2. Household income structure typology: (1) high forest dependency; (2) high agricul-
ture dependency; (3) wage-business dependency; (4) balanced forest–agriculture–wage
income.
dependency, when at least two of these components account for no less
than 25% of their income.
Regarding ethnicity and location, the study was conducted with
Ashaninka, Cashibo–Cacataibo, Shipibo–Conibo, and non-Amazonian
mestizo households. The Ashaninka of the sample are only in the
Abujao basin (Callería district) and the Cacataibo are only in the San
Alejandro basin in Padre Abad.9 The Shipibo and mestizos, however,
each reside in two distinct locations. Their separate assessment stems
from differences in remoteness. Mestizo farmers from the Abujao
basin have no access to permanent roads, as opposed to those in the
Irazola and Curimaná districts of Padre Abad. Conversely, the Shipibo
village of Santa Rosa, in the lower Abujao basin, is much closer to
the city of Pucallpa than the more distant Shipibo communities in the
Ucayali flooded forests. We thus examine income and assets according
to six configurations of ethnicity and location. In addition to the
Ashaninka and Cacataibo, we use the notations “R” and “NR” to desig-
nate remote and nonremote sites: mestizos from Abujao and Shipibo
from flooded forests are classified as R, while mestizos from Padre
Abad and Shipibo from Abujao are classified as NR.
4.5. Statistical and econometric analyses
In order to analyze the data we use two complementary strategies.
First we conduct analyses of variance and Bonferroni multiple-
comparison tests. This approach allows us to do in-depth comparisons
across all the different typologies. We then estimate ordinary least
squares regressions with income and wealth as dependent variables
as well as a maximum likelihood probit model with the poverty status
of the household as dependent variable. The econometric models
allow us to test the relation between ethnicity and livelihood strategies
on the one hand, and income, wealth or poverty on the other, while
controlling for household and community level characteristics.
In the econometric estimation we include the following explanatory
variables: a dummy variable for each one of the indigenous groups
(the base group are therefore the mestizos); dummy variables for
three of the four categories of households described in Section 5.2.1
(highly forest dependent households are the base group); education,
9
The Ashaninka comprised only six households of the sample and given statistical lim-
itations were not considered in further analyses based on ethnic group.
R. Porro et al. / Forest Policy and Economics 51 (2015) 47–56 51
age and gender of the household head; distance from the household
to the forest (in minutes); distance (km) from the community to the
nearest road that is open year-round; distance (km) from the communi-
from the community to the nearest market.
5. Results
Table 1, section a, presents aggregated statistics scaled to a one-year
period for the seven income categories assessed in the study. Average
annual income across the 578 households reached $4785, or $1320
when adjusted to adult equivalent units. Section b of Table 1 shows
HFD aggregated statistics for the asset categories considered in this study.
Average capital assets across the 578 households reached $4226, or
66 100 $1202 when adjusted to adult equivalent units. For a detailed descrip-
tion of the structure and composition of the seven income categories,
and of households' capital assets, see Porro, et al. (2014).
5.2. Livelihood options, poverty, wealth and interacting variables
We generated a typology of households according to prevalent live-
lihood strategies derived from their respective income structure and ex-
amined whether significant variation exists in terms of income levels
and capital assets across the resulting classes. We then investigated
the association of two other variables (ethnicity and geographical loca-
tion) with these results, initially through analysis of variation and then
with econometric models.
5.2.1. Household typology based on relative shares of income sources
The upper panel of Table 2 summarizes (in adult equivalents) total
annual income and capital asset endowments for the typology of four
groups based on household income structure. Contrasting results are
noticeable when comparing average income and asset values for groups
1 (highly forest dependent, HFD) and 2 (highly agriculture dependent,
HAD). HFD households present higher incomes when compared to
HAD but have the lowest asset values. Although the results of the
Bonferroni comparison tests indicate that higher income of HFD when
compared to HAD is not statistically significant, the econometric results
(see below) show that the differences are in fact statistically significant
when controlling for household and community level characteristics.
The difference in asset values is highly significant. For assets, however,
HAD households present higher values than all other groups, while
BFA (balanced-forest agriculture) households also present statistically
significant higher asset values than the HFD.
5.2.2. Typology based on ethnic group and location
In order to study the above discrepancy, we examined two
interacting variables: ethnicity and location. Our analysis shows that in-
comes are higher for the three ethnic groups located in Abujao (statisti-
cally significant at 99% confidence level), even when compared with
households from the same ethnicity located elsewhere (see Table 2b).
A closer examination of each income source separately reveals that
greater income in Abujao is derived from forest and environmental
products (timber is predominant), and the share of forest-derived
income is greater in Abujao than in other locations. When examining
capital assets, a much stronger association exists with ethnicity. Higher
asset endowments of mestizo farmers are clearly demonstrated when
compared to those of the Cacataibo and the Shipibo. Location in this
case plays a milder role and in the opposite direction, as average capital
asset values for Abujao mestizos are significantly lower than those for
Padre Abad.
52 R. Porro et al. / Forest Policy and Economics 51 (2015) 47–56

Table 1
Income sources and capital asset types of smallholder households, Ucayali, Peru, 2008.

(a) Income source n % Income (US$/adult equivalent) Income (US$/household) Income share⁎

Mean Median sd Mean Median sd Sum hhold. avg. % of total

1. Forest 552 96 478 116 1387 1857 463 5881 1,073,201 27.2 38.8
2. Fish 516 89 114 43 240 388 163 676 224,096 12.0 8.1
3. Agriculture 547 95 311 130 641 1101 465 2111 636,568 24.5 23.0
4. Livestock 489 85 135 35 362 473 145 1047 273,642 10.8 9.9
5. Wages 432 75 175 90 245 595 371 837 343,777 17.4 12.4
6. Business 168 29 84 0 450 282 0 1240 163,206 5.0 5.9
7. Other 306 53 24 1.4 77 89 7 270 51,345 3.0 1.9
Total 578 100 1320 873 1828 4785 3049 7180 2,765,836 100 100

(b) Asset category n % Assets (US$/adult equivalent) Assets (US$/household) Asset share⁎

1. Perennials 216 37 341 0 834 1244 0 2876 718,766 19.0 29.4

2. Pastures 223 39 271 0 727 952 0 2388 550,505 13.2 22.5
3. Fruit trees 515 89 114 44 267 384 190 718 222,100 17.3 9.1
4. Livestock 524 91 241 22 1506 769 89 2664 444,540 11.9 18.2
5. Goods 395 68 43 8.7 96 171 30 380 98,953 4.8 4.1
6. Equipment 234 40 39 0 189 155 0 832 89,539 5.0 3.7
7. Rural construction 141 24 36 0 117 140 0 473 80,900 2.6 3.3
8. House 563 97 117 82 148 411 320 426 237,455 25.6 9.7
Total 578 100 1202 494 2361 4226 1876 6062 2,442,758 100 100

Source: RAVA-Peru 2008 survey.

⁎ (hhold. avg.): average proportion of each income/asset type across households; (% of total): share based on total values of income/asset category.

5.2.3. Combined typology: livelihood strategy, ethnic group, and location
Table 3 reports the income structure for the observed classes of a
typology that integrates livelihood strategy, ethnic group, and loca-
tion (remoteness). This analysis clearly shows that (1) coherent
with strategies relying on multifaceted livelihoods in order to reduce
risk at the level of the social group, a balanced income structure (BFA)
is featured by nearly half of the households (frequencies of 44% to
56%) for all ethnic groups. Livelihood strategies across the entire
sample, regardless of location and ethnicity, are not overly dependent
on a single set of activities and benefit from the integration of forest ex-
traction, agriculture, and other sources; (2) livelihood orientation in
Abujao is significantly more dependent on forest products. Only 2% of
Padre Abad's mestizo (nonremote) households are highly forest depen-
dent compared to 44% for (remote) mestizos in Abujao; a similar trend,
although not as marked, occurs for the Shipibo, despite the fact that in
this case, Abujao is the nonremote location: (43% of HFD in Abujao,

Table 2
Income and capital assets for smallholder households according to classes of income structure, ethnic group and geographical location Ucayali, Peru, 2008.

Class variable n % Income (US$/ad. eq.) % of poverty Assets (US$/ad. eq.)

Mean Median sd Mean Median sd

a. Income structure, all households

1. High forest dependency (HFD) 137 23.7 1930 1158 2899 35.3 544 323 755
2. High agriculture dependency (HAD) 99 17.1 1471 933 1618 37.4 2293 1890 2444
3. Wage & business dependency (WBD) 55 9.5 1159 871 1397 38.2 997 426 1640
4. Balanced forest–agriculture (BFA) 287 49.7 1008 712 1091 48.1 1179 483 2792
Total 578 100.0 1320 873 1828 42.3 1202 494 2361
F-test from ANOVA: F = 8.58 Prob N F = 0.0000 F = 11.32 Prob N F = 0.0000

Multiple class comparisons with Bonferroni normalization: HFD HAD WBD HFD HAD BFD

HAD −460 HAD 1750***

WBD −772** −312 WBD 453 −1296***
BFA −923*** −463 −151 BFA 635** −1115*** 182

Class variable n % Income (US$/ad. eq.) % of poverty Assets (US$/ad. eq.)

Mean Median sd Mean Median sd

b. Ethnic group and location

1. NR-Mestizo, Padre Abad 214 37.0 1261 898 1250 38.3 2432 1617 3451
2. R-Mestizo, Abujao 68 11.8 2852 1584 3892 16.2 919 454 1179
3. R-Shipibo, Ucayali flooded forests 114 19.7 875 624 789 51.8 256 175 369
4. NR-Shipibo, Abujao 42 7.3 1892 1349 1942 14.3 299 249 225
5. Cacataibo, San Alejandro basin 134 23.2 758 512 815 62.7 505 338 513
6. R-Ashaninka, Abujao 6 1.0 3066 2884 1336 0 365 318 181
F-test from ANOVA: F = 17.62 Prob N F = 0.0000 F = 22.81 Prob N F = 0.0000

Multiple class comparisons with Bonferroni normalization: 1.nr-Me 2.r-Me 3.r-Sh 4.nr-Sh 5.Cac 1.nr-Me 2.r-Me 3.r-Sh 4.nr-Sh 5.Cac

2. R-Me 1591*** −1513***

3. R-Sh −386 −1976*** −2177*** −663
4. NR-Sh 631 −960* 1017** −2133*** −620 44
5. Cacat. −503 −2094*** −117 −1134*** −1927*** −414 249 206
6. Asha. 1805 214 2191** 1174 2308** −2067 −554 109 66 −140

* p b .10, ** p b .05, *** p b .01. Source: RAVA-Peru 2008 survey.

 R. Porro et al. / Forest Policy and Economics 51 (2015) 47–56 53

Table 3 Table 4
Distribution of households according to income structure, ethnic group and location. Econometric estimations of income, assets and poverty.

Income % by category of location-ethnicity Income Assets Poverty+

structure NR-Mestizo R-Mestizo NR-Shipibo R-Shipibo Cacataibo Shipibo −892.698*** −1994.200*** 0.101
(n by group) (214) (68) (42) (114) (134) [312.701] [397.193] [0.085]
Cacataibo −1176.764*** −1789.027*** 0.375***
HFD 1.87% 44.12% 42.86% 18.42% 44.78%
[228.274] [289.954] [0.056]
HAD 37.38% 4.41% – 7.02% 5.97%
HAD −854.104*** 814.699** 0.131**
WBD 10.75% 7.35% 2.38% 18.42% 3.73%
[260.582] [330.991] [0.06]
BFA 50.00% 44.12% 54.76% 56.14% 45.52%
WBD −1151.107*** 106.958 0.178**
Source: RAVA-Peru 2008 survey. [294.249] [373.755] [0.075]
BFA −1160.876*** 220.943 0.211***
[191.894] [243.744] [0.047]
Education 65.821*** 35.986 −0.020***
18% at remote sites). A possible explanation for these findings is
[22.913] [29.105] [0.006]
presented at the end of Section 5.3; (3) significant frequency of high Age −0.059 −0.820 0.004***
agriculture dependency is featured only by mestizos in nonremote [5.479] [6.960] [0.001]
locations (37%), where current availability of high-value forest product Male −373.926 396.879 0.068
[244.442] [310.490] [0.064]
is lower; (4) comparing the three indigenous groups, dependency on
Dist_forest −0.742 0.441 0.001
forest products is lower at the more remote site: 45% of the Cacataibo [1.386] [1.760] [0.001]
and 43% of the (nonremote) Shipibo households in Abujao are HFD, Dis_road −5.985 21.329*** 0.002
contrasting with 18% for the (remote) Shipibo at flooded forest sites. [6.495] [8.250] [0.002]
Critical to these results is the approach of measuring dependency ac- Dis_river −5.094 −11.104** 0.002**
[3.741] [4.751] [0.001]
cording to shares of forest and agricultural income. Income generated
Dis_market 7.517 −23.806*** −0.001
from forests will largely depend on high value products and accessible [6.293] [7.993] [0.002]
points of sale. Constant 2524.681*** 1855.596***
[454.386] [577.161]
5.2.4. Poverty assessment R2 0.129 0.169
In order to strengthen our analysis based on income and wealth Pseudo R2 0.109
indicators, we also examined the percentage of households posi- N 572
tioned below the poverty line across the sample. We use the line of * p b 0.10, ** p b 0.05, *** p b 0.01.
US$1.89/day proposed by the Peruvian government for rural loca- Standard errors shown in brackets.
tions in the Amazon region (INEI, 2011b). By contrasting such threshold + The results shown are marginal effects after a probit estimation.

with adult equivalent income, we find that 42% of the sample house-
holds are below the poverty line. Table 2 includes in its intermediate
column the percentage of households for each group that do not reach
the minimum income and are considered poor. Coherent with the as-
sessment based on income, the data indicates that: (1) small differences
in poverty level exist on the basis of household livelihood orientation
alone; (2) poverty is significantly greater for the Cacataibo and the
(remote) Shipibo from Ucayali flooded forests, and lower for the three
groups settled in the Abujao basin.
5.3. Econometric analysis
After examining correlations and testing the statistical significance
of differences between household typologies, we employ regression
analyses to further understand whether ethnicity and livelihood strate-
gies exert a determinant role on income, asset levels and poverty. By
including household level characteristics (e.g., age and education)
we can control for the effect that heterogeneity in those observable
variables can have on income, wealth and poverty. We also included a
set of distance variables that try to capture access to natural resources
as well as access to markets. Results of the analyses, presented in
Table 4, are summarized below.
The income regression shows that the Shipibo and Cacataibo have
lower incomes than the Mestizos. HAD, WBD and BFA households
have lower income than HFD households, but the average per-adult
equivalent income across these three categories is not statistically dif-
ferent from each other. Among the control variables the only statistical-
ly significant relationship is between education and income; everything
else constant an additional year of education implies a higher income.
The asset levels of the Shipibo and the Cacataibo are lower than
those of the mestizos, while the difference between Cacataibo and
Shipibo assets' is not statistically significant. HAD households have
more assets than HFD households. There are no statistically significant
differences between HFD, WBD and BFA in terms of asset holdings.
The finding that, controlling for household and community level
characteristics, assets in HAD households are higher than the assets of
the other three categories is not surprising because the definition of as-
sets includes many that are needed for agricultural activities. Three of
the distance variables are statistically significant: households that are
further away from roads have more assets, while those living in com-
munities further away from rivers and local markets have less assets.
The latter can be interpreted as evidence that being disconnected
from markets has a negative effect on wealth while the former is related
to the fact that in many communities that are remote or hard to reach by
land there is a relatively high prevalence of illegal activities (i.e., coca
plantations and illegal logging).
The last column of Table 4 shows the results of a probit model where
the dependent variable is a dummy that takes the value of one if the
household is below the poverty line. Cacataibo households are more
likely to be poor than a mestizo or a Shipibo household. The probability
of a household to be below the poverty line is always higher if the
household is HAD, WBD or BFA than if it is HFD. Nevertheless, this neg-
ative relationship between poverty and dependency on forest products
is relatively weak as it vanishes when we include in the probit model a
dummy variable to identify households located in the Abujao region.10
There is anecdotal evidence pointing out to a situation where house-
holds from Abujao use money from illegal activities to finance logging.
That can explain the fact that when we include the Abujao dummy it
becomes significant (and negative) while the dummies for livelihood
categories stop being significant. The education of the household head
decreases the likelihood of the household being poor while age
increases it. Finally, households that are further away from a river are
more likely to be poor.
10
Results from this regression are not shown but no other significant changes occur
when we include the Abujao dummy. We ran the regressions for income and assets with
the Abujao dummy as well and the results are qualitatively the same as those presented in
Table 4.
54 R. Porro et al. / Forest Policy and Economics 51 (2015) 47–56
6. Discussion
Our analysis provides mixed insights to the examination of links be-
tween forest dependency, poverty and wealth. The differences in asset
ownership identified through the study offer unequivocal support
to the hypothesis that households relying predominantly on agriculture
are wealthier than forest-dependent households. We have shown
that wealthier households happen to be mestizo farmers of non-
Amazonian origin. Households highly dependent on forests present
lower asset levels in the broader typology based on livelihood orienta-
tion. Yet, this difference is not confirmed when the analysis is disaggre-
gated: aside from nonremote mestizos, no statistically significant
difference exists in asset levels across classes of livelihood strategies
for all other ethnic/location configurations.
A more complex picture is portrayed when assessing income levels,
or poverty. Livelihood orientation alone cannot provide sufficient
evidence for income variation across households. A clear effect of geo-
graphical location is manifested through the higher incomes of mestizo
residents of the Abujao basin, as a much greater share of these house-
holds rely on forest products than on agriculture. The economic dynam-
ic of forest resource extraction in locations still abundant in natural
capital is likely to explain such contrasting results when compared to
asset endowments of these households.
Our study suggests that despite accessibility constraints, existent mar-
ket channels (including considerable illegal trade mechanisms) allow
mestizo households to effectively harness the current availability of
timber and transform it into financial capital. Moreover, such commercial
operations are coupled with logging-derived businesses and wage labor
opportunities observed in the area, confirmed by the incomes from
WBD households.
In terms of capital assets, however, contrasting results clearly
confirm greater endowments only for mestizo households in more
consolidated land, and particularly those with stronger agricultural
orientation or balanced income structure. For all other ethnic-location
configurations (except the Cacataibo), households with the BFA income
structure present a higher average value of assets (although not statisti-
cally significant at the 90% level).
Our study provides evidence that livelihoods and type of dependen-
cy on forest products are quite distinct according to the degree of mar-
ket integration. Subsistence use and limited market channels at more
remote indigenous communities, such as the Shipibo in Ucayali flooded
forests, contrast with active trading, particularly for timber, at remote
sites exploited by mestizo farmers in Abujao. Nonetheless, in areas
where resource users opted for greater market integration, such condi-
tions have not yet supported sustainable local economies through value
chains for timber and nontimber products. Rather, as unregulated ex-
traction prevails, little benefit remains locally other than low payments
for unprocessed timber. Institutions collaborating with indigenous and
smallholder communities in the area have worked toward greater
benefit sharing. But these efforts are still shy of reaching the target
and require the true commitment and innovative partnerships from
national and regional governments, the private sector, and research
professionals to establish open dialog with the communities in order
to progressively harness this unexploited potential in a sustainable
manner. This is particularly critical in the current context marked
by the primacy of global debates on climate-change mitigation. If not
offering concrete livelihood alternatives that provide cultural continuity
to traditional practices, a policy framework heavily relying on carbon-
market transactions will affect the land market, increase social stratifi-
cation, and counteract efforts to enhance social co-benefits of rights-
based environmental policy approaches.
Relevant policy implications can be derived from these findings,
particularly to stakeholder groups with survival strategies characterized
by a strong interaction of swidden agriculture and forest use. Indeed,
despite the significance of forest-related income, our analysis shows
that livelihood strategies in Ucayali are not overly dependent solely on
forests or on any other single set of activities. Rather, resource users
benefit from the integration of forest extraction, fishing, agriculture,
livestock, and other sources, including wages and businesses, with nu-
ances of greater relevance for specific components expressed through
ethnic/locational configurations. Interventions and management options
that have the concurrent objectives of environmental conservation, social
equity, and economic development can only be successful when strength-
ening the multiple components of these livelihood systems.
7. Conclusions
Our study of 578 households from 26 communities with diverse
ethnic composition at distinct environmental settings in the Peruvian
region of Ucayali shows that, on average, nearly 40% of the annual in-
come derives from forests and environmental products (including
fish), followed by agriculture (25%), wages (17%), and livestock and
animal products (11%). Compared to similar assessments in other trop-
ical forest locations, these figures are a strong indicator of the criticality
of multiple uses of forest products for a large share of the local popula-
tion, both indigenous communities and non-Amazonian settlers.
Our empirical observations in the area, supported by the statistical
analysis of forest-derived income indicate that a substantial portion of
such income is obtained from the sale of timber from natural and
successional forests by mestizos of non-Amazonian origin and from
the consumption and sale of bush meat. This denotes potential deple-
tion of natural capital and threats to biodiversity through rather unsus-
tainable practices, particularly for remote mestizo settlements. While
prompted by the very political–economic configuration of a system
that is unable to stop illegal logging and drug smuggling, the conse-
quences of this unregulated extraction resemble the boom-to-bust pat-
tern so common in Amazonian locations and suggest somber livelihood
prospects related to depletion of natural stocks.
Contrasting results about income (poverty) and capital assets
(wealth) for households in remote areas demonstrate contexts of abun-
dant natural capital at earlier stages of accumulation, with higher
incomes derived from the availability of marketable products. Instead
of associating dependency on forest use with reduced wealth and
increased poverty, our analysis confirms the critical role of ethnicity in
combination with the agrarian dynamics involved in forest-resource
extraction in determining households' poverty levels (measured
through income). Relative dependency on forest products is shown to
be lower at the more remote indigenous sites, inhabited by the Shipibo
in Ucayali's flooded forests, suggesting that indigenous communities at
these remote sites are less engaged in logging and extract timber with
lower intensity. Ethnic/locational combinations also prove to be rele-
vant in the assessment of wealth (measured through capital assets).
These results point to rather different outcomes, indicating that mestizo
households at nonremote sites relying predominantly on agriculture
are wealthier. Whether these households should also be considered
better off in terms of well-being is however a subjective call since
well-being is experienced differently according to each cultural context
and related individual perceptions. Our findings confirm that aggregat-
ed statistics often mask important distinctions even within relatively
limited geographical areas. Moreover, they support the argument for in-
clusive approaches and multiple analytical dimensions for the proper
understanding of social–ecological interactions and the assessment of
even only the tangible, material aspects of well-being in the forest
frontier.
Acknowledgments
We sincerely thank the families of the 26 communities who allowed
our multiple visits and shared with us invaluable information on
their livelihoods, their knowledge and perceptions on well-being. This
research was possible with the financial support of the World Bank In-
stitutional Development Fund (project grant TF090577). Patrica Seijas,
 R. Porro et al. / Forest Policy and Economics 51 (2015) 47–56
Miguel Vasquez, Clemente Salazar, Vladimir Nuñez and Jefferson
Cardenas deserve our recognition as team leaders during fieldwork,
and collected the data that allowed this analysis. Lourdes Quiñones pro-
vided efficient support with data checking and verification. Alexander
Mahr and Leroy Mwanza assisted with data management at various
stages of the research. We gratefully acknowledge Jonathan Cornelius,
Julio Ugarte, Mary Menton and two anonymous reviewers for their
valuable comments and suggestions to improve the manuscript. We
take full responsibility for any remaining imprecisions or omissions.
References
Alvarez, N., Naughton-Treves, L., 2003. Linking national agrarian policy to deforestation in
the Peruvian Amazon: a case study of Tambopata, 1986–1997. Ambio 23, 269–274.
Anderson, K., Bows, A., 2008. Reframing the climate change challenge in light of post-
2000 emission trends. Philos. Trans. R. Soc. A Math. Phys. Eng. Sci. 366 (1882),
3863–3882.
Anderson, A.B., Ioris, E.M., 1992. Valuing the rain forest: economic strategies by small-
scale forest extractivists in the Amazon estuary. Hum. Ecol. 20 (3), 337–369.
Angelsen, A., Wunder, S., 2003. Exploring the Forest–Poverty Link. CIFOR Occasional Paper
40. CIFOR, Bogor, Indonesia, pp. 1–20.
Angelsen, A., Larsen, H.O., Lund, J.F., Smith-Hall, C., Wunder, S. (Eds.), 2011. Measuring
Livelihoods and Environmental Dependence. Earthscan, Edinburgh, UK.
Bebbington, A., 1999. Capitals and capabilities: a framework for analyzing peasant viabil-
ity, rural livelihoods and poverty. World Dev. 27 (12), 2021–2044.
Belcher, B.M., 2005. Forest product markets, forests and poverty reduction. Int. For. Rev. 7
(2), 82–89.
Belcher, B., Ruíz-Pérez, M., Achdiawan, R., 2005. Global patterns and trends in the use and
management of commercial NTFPs: implications for livelihoods and conservation.
World Dev. 33 (9), 1435–1452.
Blackorby, C., Donaldson, D., 1991. Adult-equivalence scales, interpersonal comparisons of
well-being, and applied welfare economics. In: Elster, J., Roemer, J. (Eds.), Interper-
sonal Comparisons and Distributive Justice. Cambridge University Press, Cambridge,
pp. 164–199.
Brandolini, A., Magri, S., Smeeding, T.M., 2010. Asset‐based measurement of poverty.
J. Policy Anal. Manag. 29 (2), 267–284.
Browder, J.O., Pedlowski, M.A., Summers, P.M., 2004. Land use patterns in the Brazilian
Amazon: comparative farm-level evidence from Rondônia. Hum. Ecol. 32 (2),
197–224.
Byron, N., Arnold, M., 1999. What futures for the people of the tropical forests? World
Dev. 27 (5), 789–805.
Cavendish, W., 2000. Empirical regularities in the poverty–environment relationship of
rural households: evidence from Zimbabwe. World Dev. 28 (11), 1979–2003.
Cavendish, W., 2002. Quantitative methods for estimating the economic value of resource
use to rural households. In: Campbell, B., Luckert, K.M. (Eds.), Uncovering the Hidden
Harvest: Valuation Methods for Woodland and Forest Resources. Earthscan, London,
U.K., pp. 17–65.
Cavendish, W., 2003. How do Forests Support, Insure and Improve the Livelihoods of the
Rural Poor? A Research Note. Unpublished paper, CIFOR, available at http://www.
cifor.cgiar.org/pen
Caviglia-Harris, J.L., 2004. Household production and forest clearing: the role of farming in
the development of the Amazon. Environ. Dev. Econ. 9 (2), 181–202.
Chhatre, A., Lakhanpal, S., Larson, A.M., Nelson, F., Ojha, H., Rao, J., 2012. Social safeguards
and co-benefits in REDD+: a review of the adjacent possible. Curr. Opin. Environ.
Sustain. 4 (6), 654–660.
Colchester, M., Lohmann, L. (Eds.), 1993. The Struggle for Land and the Fate of the Forests.
World Rainforest Movement, Penang, Malaysia.
Coomes, O.T., Burt, G.J., 2001. Peasant charcoal production in the Peruvian Amazon:
rainforest use and economic reliance. For. Ecol. Manag. 140 (1), 39–50.
Corbera, E., Estrada, M., May, P., Navarro, G., Pacheco, P., 2011. Rights to land, forests and
carbon in REDD+: insights from Mexico, Brazil and Costa Rica. Forests 2 (1), 301–342.
Duchelle, A.E., 2009. Conservation and livelihood development in Brazil nut-producing
communities in a tri-national Amazonian frontier. (Doctoral Dissertation). University
of Florida.
Ellis, F., 2000. Rural Livelihoods and Diversity in Developing Countries. Oxford University
Press, Oxford.
Fisher, M., 2004. Household welfare and forest dependence in Southern Malawi. Environ.
Dev. Econ. 9 (2), 135–154.
Fisher, M., Shively, G., 2005. Can income programs reduce tropical forest pressure?
Income shocks and forest use in Malawi. World Dev. 33 (7), 1115–1128.
Fujisaka, S., 1997. Land use strategies in Pucallpa. In: Murray, T., Gallopin, G. (Eds.),
Proceedings of the First International Workshop: Integrated Conceptual Framework
for Tropical Agroecosystem Research Based on Complex Systems Theories. Interna-
tional Center for Tropical Agriculture, Cali, Colombia, pp. 37–46.
Gobierno Regional de Ucayali—GOREU, 2008. Diagnóstico Socio-económico del
Departamento Ucayali, 2007. GOREU, Pucallpa, Peru (Accessed May 2, 2012 at
http://www.regionucayali.gob.pe/grpp/images/planes/diag.dep.uca-indice.pdf).
Godoy, R., Jacobson, M., Wilkie, D., 1998. Strategies of rain-forest dwellers against misfor-
tunes: the Tsimane' Indians of Bolivia. Ethnology 37 (1), 55–69.
Guedes, G., Brondízio, E., Barbieri, A., Resende, A., Penna-Firme, R., D'Antona, A., 2012.
Poverty and inequality in the rural Brazilian amazon: a multidimensional approach.
Hum. Ecol. 40 (1), 41–57.
55
Gutiérrez-Vélez, V.H., DeFries, R., Pinedo-Vásquez, M., Uriarte, M., Padoch, C., Baethgen,
W., Fernandes, K., Lim, Y., 2011. High-yield oil palm expansion spares land at the ex-
pense of forests in the Peruvian Amazon. Environ. Res. Lett. 6 (4) (Accessed Septem-
ber 11, 2012 at http://iopscience.iop.org/1748-9326/6/4/044029/pdf/1748-9326_6_
4_044029.pdf).
Hiraoka, M., 1986. Zonation of mestizo riverine farming systems in northeast Peru. Natl.
Geogr. Res. 2 (3), 354–371.
Instituto del Bien Común—IBC, 2012. Sistema de Información sobre Comunidades Nativas
de la Amazonía Peruana–SICNA. Accessed May 3, 2012 at. http://www.ibcperu.org/
sicnabd/.
Instituto Nacional de Estadística e Informática—INEI, 2009a. Perfil Sociodemográfico del
Departamento de Ucayali. Censos Nacionales, 2007: XI de Población y VI de Vivienda.
INEI, Lima, Peru.
Instituto Nacional de Estadística e Informática—INEI, 2009b. Resultados Definitivos
de Comunidades Indígenas. Tomo 1. Censos Nacionales, 2007: XI de Población y VI de
Vivienda. INEI, Lima, Peru.
Instituto Nacional de Estadística e Informática—INEI, 2011a. Perú: Perfil de la Pobreza por
Departamentos, 2001–2010. INEI, Lima, Peru.
Instituto Nacional de Estadística e Informática—INEI, 2011b. Producto Bruto Interno
por Departamentos, 2001–2010. INEI, Lima, Peru (Accessed May 2, 2012 at
http://www.inei.gob.pe/biblioineipub/bancopub/Est/Lib0995/Libro.pdf).
Instituto Nacional de Estadística e Informática—INEI, 2011c. Ucayali: Compendio
Estadístico Departamental, 2010. INEI. Oficina Departamental de Estadística e
Informática de Ucayali, Pucallpa, Peru.
Kaimowitz, D., 2008. The prospects for reduced emissions from deforestation and degra-
dation (REDD) in mesoAmerica. Int. For. Rev. 10 (3), 485–495.
Kamanga, P., Vedeld, P., Sjaastad, E., 2009. Forest incomes and rural livelihoods in
Chiradzulu District, Malawi. Ecol. Econ. 68 (3), 613–624.
Kar, S.P., Jacobson, M.G., 2012. NTFP income contribution to household economy and related
socio-economic factors: lessons from Bangladesh. Forest Policy Econ. 14 (1), 136–142.
Kindermann, G., Obersteiner, M., Sohngen, B., Sathaye, J., Andrasko, K., Rametsteiner, E.,
Beach, R., 2008. Global cost estimates of reducing carbon emissions through avoided
deforestation. Proc. Natl. Acad. Sci. 105 (30), 10302–10307.
Labarta, R.A., White, D.S., Swinton, S.M., 2008. Does charcoal production slow agricultural
expansion into the Peruvian Amazon Rainforest? World Dev. 36 (3), 527–540.
Larson, A.M., Ribot, J.C., 2007. The poverty of forestry policy: double standards on an
uneven playing field. Sustain. Sci. 2 (2), 189–204.
Larson, A.M., Cronkleton, P., Barry, D., Pacheco, P., 2008. Tenure Rights and Beyond:
Community Access to Forest Resources in Latin America. CIFOR, Bogor, Indonesia.
Loker, W.M., 1993. The human ecology of cattle raising in the Peruvian Amazon: the view
from the farm. Hum. Organ. 52 (1), 14–24.
Mamo, G., Sjaastad, E., Vedeld, P., 2007. Economic dependence on forest resources: a case
from Dendi District, Ethiopia. Forest Policy Econ. 9 (8), 916–927.
Marquette, C.M., 2006. Settler welfare on tropical forest frontiers in Latin America. Popul.
Environ. 27 (5), 397–444.
McSweeney, K., 2005. Natural insurance, forest access, and compounded misfortune:
forest resources in smallholder coping strategies before and after Hurricane Mitch,
northeastern Honduras. World Dev. 33 (9), 1453–1471.
Ministerio de Energía y Minas—MINEM, Gobierno Regional de Ucayali—GOREU, 2007.
Caracterización del Departamento de Ucayali, con Fines de Ordenamiento Territorial.
MINEM-GOREU, Pucallpa, Peru (Accessed May 2, 2012 at http://www.ibcperu.org/
doc/isis/12377.pdf).
Moran, E.F., Siqueira, A., Brondizio, E., 2004. Household demographic structure and its
relationship to deforestation in the Amazon Basin. In: Fox, J., Rindfuss, R., Walsh,
S.S., Mishra, V. (Eds.), People and the Environment: Approaches for Linking
Household and Community Surveys to Remote Sensing and GIS. Kluwer Academic
Publishers, pp. 61–89.
Murphy, L., Bilsborrow, R., Pichon, F., 1997. Poverty and prosperity among migrant
settlers in the Amazon rainforest frontier of Ecuador. J. Dev. Stud. 34 (2), 35–65.
Narain, U., Gupta, S., Van't Veld, K., 2008. Poverty and resource dependence in rural India.
Ecol. Econ. 66 (1), 161–176.
Nielsen, M.R., Pouliot, M., Bakkegaard, R.K., 2012. Combining income and assets measures
to include the transitory nature of poverty in assessments of forest dependence:
evidence from the Democratic Republic of Congo. Ecol. Econ. 78, 37–46.
Pacheco, P., Barry, D., Cronkleton, P., Larson, A., 2012. The recognition of forest rights in
Latin America: progress and shortcomings of forest tenure reforms. Soc. Nat. Resour.
25 (6), 556–571.
Padoch, C., de Jong, W., 1989. Production and profit in agroforestry: an example from the
Peruvian Amazon. In: Browder, J.G. (Ed.), Fragile Lands of Latin America: Strategies
for Sustainable Development. Westview Press, Boulder, Colorado, pp. 102–113.
Pattanayak, S.K., Sills, E.O., 2001. Do tropical forests provide natural insurance? The micro-
economics of non-timber forest product collection in the Brazilian Amazon. Land
Econ. 77 (4), 595–612.
Paumgarten, F., 2005. The role of non-timber forest products as safety-nets: a review of
evidence with a focus on South Africa. GeoJournal 64 (3), 189–197.
Perz, S.G., 2005. The importance of household asset diversity for livelihood diversity and
welfare among small farm colonists in the Amazon. J. Dev. Stud. 41 (7), 1193–1220.
Pinedo-Vasquez, M., Zarin, D., Jipp, P., 1992. Economic returns from forest conversion in
the Peruvian Amazon. Ecol. Econ. 6 (2), 163–173.
Programa de las Naciones Unidas para el Desarrollo—PNUD, 2010. Informe
sobre Desarrollo Humano Perú, 2009. Parte 2. Anexos. Accessed May 2, 2012 at.
http://www.pnud.org.pe/data/publicacion/idh2009vol2-09anexos.pdf.
Porro, R., Lopez-Feldman, A., Vela-Alvarado, J.W., Quiñones-Ruiz, L., Seijas-Cardenas, Z.P.,
Vásquez-Macedo, M., Salazar-Arista, C., Nuñez-Paredes, V.I., Cardenas-Ruiz, J., 2014.
Forest Use and Agriculture in Ucayali, Peruvian Amazon: Interactions Among Liveli-
hood Strategies, Income and Environmental Outcomes. Tropics 23, 47–62.
56 R. Porro et al. / Forest Policy and Economics 51 (2015) 47–56
Ramankutty, N., Gibbs, H.K., Achard, F., Defries, R., Foley, J.A., Houghton, R.A., 2006.
Challenges to estimating carbon emissions from tropical deforestation. Glob. Chang.
Biol. 13 (1), 51–66.
Ramos-Delgado, N.R., 2008. Impacto de la producción forestal maderable en la economía
de la Región Ucayali, Perú. (M. Sc. Thesis). Universidad Nacional Agraria, Lima, Peru.
Ravallion, M., Datt, G., Walle, D., 2005. Quantifying absolute poverty in the developing
world. Rev. Income Wealth 37 (4), 345–361.
Reardon, T., Vosti, S.A., 1995. Links between rural poverty and the environment in develop-
ing countries: asset categories and investment poverty. World Dev. 23 (9), 1495–1506.
Ribot, J.C., Peluso, N.L., 2003. A theory of access. Rural. Sociol. 68 (2), 153–181.
Ruíz-Pérez, M., Belcher, B., Achdiawan, R., Alexiades, M., Aubertin, C., Caballero, J.,
Sunderland, T.J., 2004. Markets drive the specialization strategies of forest peoples.
Ecol. Soc. 9 (2), 4 (online http://www.ecologyandsociety.org/vol9/iss2/art4).
Salisbury, D.S., Fagan, C., 2011. Coca and conservation: cultivation, eradication, and traf-
ficking in the Amazon borderlands. GeoJournal 11, 1–20.
Scherr, S.J., 2000. A downward spiral? Research evidence on the relationship between
poverty and natural resource degradation. Food Policy 25 (4), 479–498.
Scherr, S.J., White, A., Kaimowitz, D., 2003. A new agenda for forest conservation and pov-
erty reduction: Making forest markets work for low-income producers. Forest
Trends, CIFOR, and IUCN.
Schmink, M., 2004. Communities, forests, markets, and conservation. In: Zarin, D.,
Alavalapati, J.R.R., Putz, F.E., Schmink, M. (Eds.), Working Forests in the Tropics—
Conservation through Sustainable Management. Columbia University Press,
New York, pp. 119–129.
Scoones, I., 1998. Sustainable rural livelihoods: a framework for analysis. IDS. Working
Paper. Institute of Development Studies, Brighton, UK (Accessed June 8, 2012 at
http://www.ntd.co.uk/idsbookshop/details.asp?id=419).
Shackleton, C.M., Shackleton, S.E., 2004. The importance of non-timber forest products
in rural livelihood security and as safety nets: a review of evidence from South
Africa. S. Afr. J. Sci. 100 (11 & 12), 658–664.
Sheil, D., Wunder, S., 2002. The value of tropical forest to local communities: complica-
tions, caveats, and cautions. Conserv. Ecol. 6 (2) (online URL: http://www.consecol.
org/vol6/iss2/art9).
Shively, G.E., 2004. Poverty and forest degradation: introduction to the special issue.
Environ. Dev. Econ. 9 (02), 131–134.
Shone, B.M., Caviglia-Harris, J.L., 2006. Quantifying and comparing the value of non-
timber forest products in the Amazon. Ecol. Econ. 58 (2), 249–267.
Smith, J., van de Kop, P., Reategui, K., Lombardi, I., Sabogal, C., Diaz, A., 1999. Dynamics of
secondary forests in slash-and-burn farming: interactions among land use types in
the Peruvian Amazon. Agric. Ecosyst. Environ. 76 (2), 85–98.
Stata Press, 2007. Stata Data Management. Reference Manual. Release 10. Stata Press,
College Station, TX.
Stoian, D., 2005. Making the best of two worlds: rural and peri-urban livelihood options
sustained by nontimber forest products from the Bolivian Amazon. World Dev. 33
(9), 1473–1490.
Stonich, S.C., Dewalt, B.R., 1996. The political ecology of deforestation in Honduras. In:
Sponsel, L.E., Headland, T.N., Bailey, R.C. (Eds.), Tropical Deforestation, The Human
Dimension. Columbia University Press, New York, pp. 187–225.
Sunderlin, W.D., Angelsen, A., Belcher, B., Burgers, P., Nasi, R., Santoso, L., Wunder, S., 2005.
Livelihoods, forests, and conservation in developing countries: an overview. World
Dev. 33 (9), 1383–1402.
Tacconi, L., Siagian, Y., Syam, R., 2006. On the Theory of decentralization, forests and live-
lihoods. Environmental Management and Development Occasional Papers, 09-2006.
Asia Pacific School of Economics and Governments. Australian National University,
Canberra.
United Nations Office on Drugs and Crime—UNODC, Comisión Nacional para el Desarrollo
y Vida sin Drogas—DEVIDA, 2010. Perú. Monitoreo de Cultivos de Coca, 2009.
www.unodc.org/unodc/en/crop_monitoring/index.html.
Vedeld, P., Angelsen, A., Sjaastad, E., Berg, G.K., 2004. Counting on the Environment: Forest
Incomes and the Rural Poor. World Bank Environment Department, Washington, DC.
Vedeld, P., Angelsen, A., Bojö, J., Sjaastad, E., Berg, G.K., 2007. Forest environmental
incomes and the rural poor. Forest Policy Econ. 9, 869–879.
Vosti, S.A., Braz, E.M., Carpentier, C.L., d'Oliveira, M.V., Witcover, J., 2003. Rights to forest
products, deforestation and smallholder income: evidence from the western Brazilian
Amazon. World Dev. 31 (11), 1889–1901.
Walker, R., Perz, S., Caldas, M., Silva, L.G.T., 2002. Land use and land cover change in forest
frontiers: the role of household life cycles. Int. Reg. Sci. Rev. 25 (2), 169–199.
Wollenberg, E., Nawir, A.S., 1998. Estimating the incomes of people who depend on
forests. In: Wollenberg, E., Ingles, A. (Eds.), Incomes from the Forest: Methods for
the Development and Conservation of Forest Products for Local Communities.
CIFOR, Bogor, Indonesia, pp. 157–187.
Wunder, S., 2001. Poverty alleviation and tropical forests—what scope for synergies?
World Dev. 29 (11), 1817–1833.