This article was downloaded by: [arika virapongse]On: 26 December 2013, At: 13:01Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Forests, Trees and LivelihoodsPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/tftl20

Value chain dynamics of an emergingpalm fiber handicraft market inMaranhão, BrazilArika Virapongsea, Marianne Schminkb & Sherry Larkinca School of Natural Resources and Environment, University ofFlorida, Gainesville, FL, USAb Center for Latin American Studies, University of Florida,Gainesville, FL, USAc Food and Resource Economics Department, University of Florida,Gainesville, FL, USAPublished online: 16 Dec 2013.

To cite this article: Arika Virapongse, Marianne Schmink & Sherry Larkin , Forests, Trees andLivelihoods (2013): Value chain dynamics of an emerging palm fiber handicraft market in Maranhão,Brazil, Forests, Trees and Livelihoods, DOI: 10.1080/14728028.2013.868707

To link to this article: http://dx.doi.org/10.1080/14728028.2013.868707

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to orarising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &

http://www.tandfonline.com/loi/tftl20http://www.tandfonline.com/action/showCitFormats?doi=10.1080/14728028.2013.868707http://dx.doi.org/10.1080/14728028.2013.868707

Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

http://www.tandfonline.com/page/terms-and-conditionshttp://www.tandfonline.com/page/terms-and-conditions

Value chain dynamics of an emerging palm fiber handicraft marketin Maranhão, Brazil

Arika Virapongsea*, Marianne Schminkb and Sherry Larkinc

aSchool of Natural Resources and Environment, University of Florida, Gainesville, FL, USA;bCenter for Latin American Studies, University of Florida, Gainesville, FL, USA; cFood andResource Economics Department, University of Florida, Gainesville, FL, USA

Non-timber forest product (NTFP) markets are often used as a development tool toimprove livelihood stability of socio-economically vulnerable communities withaccess to natural resources. Interventions to encourage growth of NTFP markets,however, often succeed at reaching only subsets of target populations. To addressimpacts of changing NTFP markets on livelihoods and sustainable forestmanagement, value chain and livelihood systems analyses were used to evaluateimpacts of a buriti (Mauritia flexuosa) handicraft market on heterogeneous buritiusers in Maranhão, Brazil. Data were obtained through interviews with 97 individualswho operated as different types of buriti users. Buriti value chain diagrams wereconstructed by identifying patterns among responses. Socioeconomic characteristicsof actors were identified using statistical means comparisons. Logistic regressionanalysis was used to identify the socioeconomic factors that correlated with peoples’role in the value chain. Results showed that the new market has introduced newindividuals who interact with pre-existing buriti users. Actors differ by livelihoodstrategy, socioeconomic factors, and perceptions regarding sustainability of leafcollection. Social heterogeneity in NTFP value chains should be considered byinitiatives that seek to influence participation in NTFP markets, evaluate effects ofcommercialization on livelihoods, and effectively design and implement resourcemanagement strategies.

Keywords: Mauritia flexuosa; buriti; NTFP; Maranhão; value chain; livelihoodsystems; emerging markets

Introduction

Non-timber forest product (NTFP) markets are often used as a development tool to

improve livelihood stability of socio-economically vulnerable communities with access

to natural resources. Not all people are prepared for or interested in taking advantage of

new market opportunities, however. For many rural communities, rapidly changing forest

markets are too risky (Belcher & Schreckenberg 2007), incompatible with traditional

institutions, ecologically unsustainable, or unrealistic when conflicts exist between groups

(Schmink 2004). Market participants must also possess good organization, business and

marketing capabilities, and access to capital and resources to have successful enterprises

(Scherr et al. 2004).

Despite obstacles and risks to market participation, it is inevitable that some people in

forest areas will use NTFPs to generate income. Increasing commercialization of NTFPs

has led to concerns regarding uneven distribution of benefits (Wynberg et al. 2002; Ruiz-

Perez, Belcher, Fu, & Yang 2004), socioeconomic divisions (Kusters et al. 2006), and

This work was authored as part of the Contributor’s official duties as an Employee of the United States Government and is

therefore a work of the United States Government. In accordance with 17 U.S.C. 105, no copyright protection is available for

such works under U.S. Law

*Corresponding author. Email: arikav@ufl.edu

Forests, Trees and Livelihoods, 2013

http://dx.doi.org/10.1080/14728028.2013.868707

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

mailto:arikav@ufl.edumailto:arikav@ufl.eduhttp://dx.doi.org/10.1080/14728028.2013.868707

overharvesting. Widening social divisions due to NTFP market participation have been

well described in Brazil, for example, among Amazonian rubber harvesters (Schwartzman

1992) and babassu collectors in Maranhão (May 1986), where less privileged groups were

excluded from benefits of growing global NTFP markets. Theoretically, individuals face

markets on an equal basis, but in reality people’s relations with markets are shaped by their

social context and circumstance. A better understanding of heterogeneous actors involved

in NTFP markets and the dynamics between them can help anticipate and guide livelihood

impacts of NTFP markets (Neumann & Hirsch 2000).

Although most NTFPs are used for subsistence purposes, a household’s degree of

integration in the cash economy affects use of forest resources (Shackleton & Shackleton

2004). In a cash-based economy, households often concentrate on activities that offer

optimal financial opportunities and rewards. Peoples’ responses to NTFP markets, such as

livelihood and harvesting strategies, are shaped by socioeconomic characteristics of the

population and environmental context (Ruiz-Perez, Belcher, Achdiawan, et al. 2004a;

Kanji et al. 2005). Livelihood strategies refer to long-range goals (Sutton & Anderson

2004) in which resources are used as efficiently as possible and in different combination

depending on household constraints, goals, opportunities, and composition. Strategies are

often based on a household’s availability of assets and opportunities, current condition,

and risk management (Belcher et al. 2005; Jensen 2006), so livelihood strategies should be

conceived as one moment in an evolving long-term process.

As economic value and demand for NTFPs rise, commercialization of products often

expands outside of regions with a history of traditional use (Shackleton et al. 2009). In

comparison to traditional participants, who often have extensive exposure and knowledge

of resources (Jensen & Meilby 2008; Schmidt & Ticktin 2012), entry of new market

participants can lead to damaging harvests and uncontrolled competition for resources

(Belcher et al. 2005; Marshall et al. 2006). Emerging NTFP markets, such as the buriti

palm (Mauritia flexuosa L.f.) handicraft market in Barreirinhas, Maranhão, Brazil, provide

an opportune case study to evaluate effects of growing market demands on peoples’

changing use of forest products and development of their market roles.

In Barreirinhas, local people have traditionally exploited buriti palm derivatives as

part of a subsistence-based livelihood strategy. Fruit, mature leaves, and young leaves

were popular derivatives used for meeting household subsistence purposes. Over the

previous 15 years, buriti markets have changed from offering mostly fruit and mature

leaves to local consumers, toward favoring young leaf fiber for making handicrafts to sell

to tourists or export to a national market (Virapongse 2013). Much of this changing

consumer base has been driven by growing accessibility to and popularity of the Lenc�óis

Maranhenses National Park, which attracts increasing numbers of visitors to the region

(Lobato 2008).

Although young leaf fiber has been increasingly sought after to meet demands of a

growing market, fruit and mature leaves were still needed by local consumers for

subsistence use. Therein lay a conflict of interest, which could be understood by

considering the buriti market as interactions among several different user groups. The

course of buriti derivatives from the tree to consumer required cooperation of various

actors, including resource owners, extractors, artisans, and vendors, whose benefits from

the buriti market varied. To evaluate factors that drove peoples’ interests and participation

in Barreirinhas’ buriti market, the following research questions were addressed in this

study: What socioeconomic factors contributed to determining diverse roles in the young

leaf fiber value chain? What were the implications of value chain dynamics and

heterogeneity on sustainability of buriti resources?

2 A. Virapongse et al.

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

Methods

Analytical framework

Value chain and livelihood systems analyses were used to evaluate the structure of buriti

markets and dynamics between markets and livelihoods. Value chain analysis is a

methodological tool for identifying actors and their activities, trade routes, and attributes

of supply and demand (Kaplinsky & Morris 2001; Marshall et al. 2006; Wilsey 2008).

Value chains are comprised of different activities required to bring a product from

conception, through phases of production, and to final delivery to consumers (Kaplinsky &

Morris 2001). Actors’ roles in the value chains are affected by actors’ relationships with

each other, available livelihood assets, market proximity to end-consumers (Jensen 2009),

and changing populations, such as new consumer groups or rapidly urbanizing populations

(Williams et al. 2000; Cunningham 2001). Rather than focusing on competitiveness

among actors, value chain analysis evaluates NTFP chains as a whole (Velde et al. 2006)

by considering relationships between actors and transmission of benefits and costs along

the chain (Kanji et al. 2005).

As a complement to value chains, livelihood systems analysis provides a lens to assess

actors’ response to environmental changes. Livelihood systems analysis is used to

examine strategies and decision-making by people within a common livelihood system

(Collinson 2000). Reaching beyond assumptions that people prioritize income, livelihood

analysis considers important alternate outcomes such as food and income security, or

sustainable use of natural resources (Kanji et al. 2005). Policies, institutions, and processes

are recognized for influencing opportunities and constraints that people face while

pursuing strategies in different contexts.

Case study

Fieldwork was conducted in Barreirinhas district among 12 communities along the

Preguic�a River, ranging from the river mouth to 35 km inland (Figure 1). Barreirinhas

district is an area of 3112 km2 with 54,930 inhabitants (IBGE 2010), who are mostly

caboclos, or mixed descendants of indigenous, European, and African people. Buriti

palms grow naturally in Barreirinhas swamp forests as a dominant tree species in the

region. Buriti palms are single stem, dioecious, and arborescent palms reaching up to 25m

tall. These palms are distributed across much of South America, and hold high importance

for subsistence-based people across all areas of its occurrence (Lawrence et al. 2005).

In Maranhão state, young leaf fiber of buriti palms has been recorded as one of the most

economically valuable forest products (IBGE 2012). According to the Brazilian Institute of

Geography and Statistics (IBGE), only four districts of Maranhão harvest buriti fiber

commercially, and Barreirinhas district was the highest producer by producing 95–125

metric tons of fiber annually from 2004 to 2012. The value of buriti fiber in the district

increased from R$7178 (US$3460) per ton in 2004, adjusted for inflation to 2011 values, to

R$10,791 (US$5200) per ton in 2011 (IBGE 2012). Although accurate values of NTFP

production are notoriously difficult to obtain, these figures demonstrate recognition of an

increasingly important regional NTFP. In 2005, buriti fiber handicrafts were considered the

second most important source of income in Barreirinhas (Prefeitura Barreirinhas 2005).

Sampling strategy

The sample consisted of 97 individuals who participated in the buriti value chain. A

purposive sampling strategy was used to select individuals based on criteria developed

Value chain dynamics 3

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

during the study (Coyne 1997). Respondent-driven sampling, which is appropriate for

making estimations about hidden populations (Salganik & Heckathorn 2004), was applied

by asking community members to name individuals who participated in the buriti market

in different ways. Individuals were classified as private and communal owners of buriti

resources (n ¼ 28), extractors of buriti derivatives (n ¼ 12), and artisans (n ¼ 52) andvendors (n ¼ 19) of buriti handicrafts. Consumers and representatives of government-managed lands in value chain dynamics were not included in the sample because they were

not members of the socio-economically vulnerable communities targeted for this study.

Figure 1. Map of the study site (by Mariano González Roglich).

4 A. Virapongse et al.

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

Instead, data collected through interviews with these groups were used to build context for

the study.

Data collection and analysis

A Brazilian research visa was obtained and Institutional Review Board process completed

(protocol #2010-U-003, University of Florida) prior to beginning data collection. Data

were collected during 18 weeks from June 2009 to November 2011. Unstructured

interviews were used to collect ethnographic data from community members and

stakeholders of buriti leaf resources, such as governmental representatives and tour guides.

Semi-structured interviews were conducted with the sample group to generate data that

were analyzed qualitatively and quantitatively with Microsoft Excel and SAS 4.3

software.

Buriti value chain diagrams, which included actors and production blocks (Kaplinsky

& Morris 2001; Velde et al. 2006), were constructed based on results from qualitatively

grouping, cross checking, and identifying patterns among interview responses. These

analyses were used to identify different actors, relationships between actors, variations

among value chains of different buriti derivatives, and actors’ potential impact on buriti

forests. Prices for products as they moved through the value chain were based on actors’

reports.

Socioeconomic variables were elicited from interview responses to represent

individual and household demographics, wealth, personal history, perceptions regarding

sustainable buriti harvesting, participation in livelihood activities, and household income

sources (Table 1). The means of these variables were compared using ANOVA and

Kruskal Wallis statistical tests ( p , 0.05), and Pearson’s correlation analysis( p , 0.0001), to identify statistically significant socioeconomic factors that characterizedvalue chain actors (Table 2). These socioeconomic factors were then used to build models

for logistic regression analysis of factors associated with heterogeneity among actors.

Response variables were actor roles (owner, extractor, artisan, and vendor) and

explanatory variables were socioeconomic factors. Variables that demonstrated low

frequencies or close correlation to other explanatory variables were not included in

models. Preliminary models were tested until models with the lowest Akaike Information

Criterion, as a measurement for best-fit models, were attained (Table 3). All logistic

models were determined to have good fit of data based on the likelihood ratio ( p , 0.05)and high percent concordant value (.86). Statistical significance of factors was measuredat a 10% level or better.

Results

Buriti value chains

The production system, value chain actors, and estimated distribution of benefits among

actors as derivatives move through the chains are depicted in Figure 2. The figure shows

property regimes for buriti extraction and different roles of landowners, buriti extractors,

intermediaries, artisans, and vendors in three value chains for buriti fruits, mature leaves,

and young leaves. Market chains for fruit and mature leaves were directed toward

household use and sale within the community. Fruits were popularly consumed as a food

source. Mature leaves were used by residents when temporary structures were needed,

federal laws prevented use of industrial construction material (e.g., protected areas), and

ceramic roof tiles were unaffordable or inaccessible. The emerging market for young leaf

Value chain dynamics 5

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

fiber handicrafts contrasted in important ways from the other two chains. First, access to

palm resources was more difficult because most owners discouraged harvesting of young

leaves from their property. Second, artisans and intermediaries were two new actors

present only in the young fiber value chain. Artisans added value to buriti fiber by making

handicrafts to sell for higher prices than raw buriti fiber. Intermediaries played important

roles of providing artisans with access to fiber and consumers. Third, and most

importantly, the handicraft market was oriented solely to outside markets; consumers were

tourists and national and international intermediaries.

Table 1. Definitions of independent variables.

Variable name Description Range

DemographicsRegion Region where interviewee lived; Laranjeiras (0)

or Atins (1)0, 1

Age Age (years) 13–88Gender Gender; male (0) or female (1) 0,1Education Education (years) 1–13Home garden Active home garden present 0, 1Agricultural field Active agricultural field present 0, 1Household size Number of members in the household 1–10Household labor Household labor: number family members earning

income/number of people in household (%)0–1

Household members buriti Number of household members participating inburiti activities

0–5

Buriti household use Buriti leaf derivative used for household subsistence 0, 1WealthConsistent income Receives consistent income each month 0, 1Wealth index Wealth index based on presence of material items 0–6

HistoryBorn in community Born in current community of residence 0, 1Individual .10 years buriti Individuals living .10 years close to buriti 0, 1Parent born in community At least one parent born in interviewee’s current

community0, 1

Parent .10 years buriti At least one parent has lived .10 years close to buriti 0, 1Buriti learned from parent Learned current buriti trade from a parent 0, 1Planted buriti tree Has planted a buriti tree 0, 1

PerceptionsBuriti trees threatened Buriti trees threatened 0, 1Young leaf harmful Young leaf collection harmful 0, 1

ActivitiesHousework activity Housework is a main activity 0, 1Handicrafts activity Handicrafts production is a main activity 0, 1Private business activity Business owner or cooperative member is

a main activity0, 1

Agriculture activity Agriculture is a main activity 0, 1Total activities Number of livelihood activities reported 0–5

Household incomeMain income buriti Buriti provides a main income sourceHandicrafts income Handicrafts provide household income 0, 1Bolsa famila income Bolsa famila provides household income source 0, 1Retirement income Retirement provides household income source 0, 1Selling palm leaves Selling palm leaves provides household income 0, 1Total income sources Number of household income sources reported 0–6

6 A. Virapongse et al.

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

Production system

Almost all buriti forests were found in low-lying inland (.18 km from coast) areas withabundant fresh water. Most buriti derivatives were extracted from native populations

where regeneration was a natural process; trees were rarely cultivated in plantations. Local

cultivation usually consisted of discarding seeds in wet areas conducive to germination

and transplanting healthy seedlings. Buriti trees were considered to thrive naturally, so

management was usually minimal. Landowners reported that management strategies were

limited to removing mature leaves and clearing undergrowth around trees to facilitate fruit

collection and tree trunk access.

Fruit and mature leaf collection was low impact and non-intense. Community

members collected mostly fallen fruit for consumption or sale from August to December,

Table 2. Means of socioeconomic variables for value chain actors (n ¼ 97).Ownern ¼ 28

Extractorn ¼ 12

Artisann ¼ 52

Vendorn ¼ 19

DemographicsRegion 0.04 0 0.27 0.42Age 57.70 (17.0) 36.25 (12.5) 39.90 (10.9) 38.37 (14.4)Gender 0.36 0 0.90 0.89Education 2.60 (2.71) 3.67 (3.42) 4.71 (2.98) 5.88 (4.01)Home garden 0.86 0.58 0.75 0.44Agricultural field 0.96 0.67 0.44 0.50Household labor 0.38 (0.26) 0.42 (0.25) 0.42 (0.19) 0.57 (0.26)Household size* 4.68 (1.44) 5.67 (1.82) 4.98 (1.78) 4.32 (2.00)Household members buriti 0.46 (0.93) 1.75 (1.54) 1.50 (1.50) 1.05 (1.13)Buriti household use* 0.86 0.91 0.77 0.60

WealthConsistent income 0.79 0.17 0.52 0.79Wealth index 3.82 (1.54) 2.42 (1.83) 2.42 (1.56) 4.42 (1.43)

HistoryBorn in community 0.61 0.75 0.45 0.56Individual .10 years buriti 0.96 1.00 0.88 0.65Parent born in community 0.54 0.91 0.44 0.47Parent .10 years buriti 0.81 1.00 0.79 0.71Buriti learned from parent* 0.50 0.25 0.45 0.58Planted buriti tree 0.77 0.60 0.38 0.64

PerceptionsBuriti trees threatened* 0.68 0.75 0.68 0.67Young leaf harmful 0.57 0.20 0.05 0.00

ActivitiesHandicrafts activity 0.25 0.17 0.94 0.89Private business activity 0.14 0.08 0.08 0.79Total activities 2.11 (1.04) 2.75 (0.75) 2.33 (0.71) 2.53 (0.77)

Household incomeMain income buriti 0.18 0.67 0.69 0.68Handicrafts income 0.29 0.83 0.92 0.84Bolsa Famila income 0.48 0.70 0.78 0.29Retirement income 0.43 0 0.15 0.21Fishing activities income 0.15 0.33 0.37 0.16Selling palm leaves 0.04 0.50 0.02 0Total income sources 2.96 (1.37) 3.83 (0.67) 3.67 (0.97) 3.26 (1.27)

Notes: Standard deviation listed in parentheses for continuous variables. *Factors with no statistically significantdifference among actors.

Value chain dynamics 7

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

although large quantities of fruit were produced only every other year. Both mature and

young leaves were collected by extractors, who climbed trees to cut leaves during

favorable weather (no rain or wind; July–October) and according to demand. Mature

leaves were in highest demand for building temporary fishing huts before the fishing

season began each February. Collection of mature leaves was considered low impact

because leaves most valued for construction purposes were leaves that were no longer

biologically productive for the tree. Overall demand for mature leaves was decreasing due

to increased use of industrial substitutes, such as roof tiles.

Young leaves were in demand throughout most of the year and particularly for the

tourist season that was highest during June and July. Young leaves most valued by artisans

were collected as leaf spikes .2m long and from mature trees with trunk height .3m.According to interviewees, approximately one young leaf per tree was produced each

month, and leaf harvesting was sustainable if at least two to three leaves remained on the

tree and two subsequent young leaves were never harvested from the same tree.

Overharvested trees along riverbanks, however, provided evidence that extractors did not

always follow their own local harvesting rules. In comparison to fruit and mature leaf

value chains, increasing demands, production, and value of fiber handicrafts could have

negative impacts on buriti resources.

Table 3. Results from logistic regression models for value chain actors reporting coefficients, oddsratio, and p-value in parentheses.

Variables Owner Extractor Artisan Vendor

Model statistics:Observations

54 54 65 75

Model evaluation:Likelihood ratio

16.62 (0.0343) 21.93 (0.0155) 23.20 (0.0031) 43.24 (,0.0001)

Percent concordant 85.1 89.5 82.8 93.0AgeAGE

– 20.1130.893 (0.048)*

20.0480.953 (0.094)**

0.0081.01 (0.880)

GenderGENDER

21.8290.161 (0.066)**

2 2 0.6751.97 (0.583)

EducationEDUCAT

20.0081.00 (0.957)

20.6470.524 (0.033)*

2 0.4571.58 (0.065)**

Home gardenHOMEGA

– 21.6630.189 (0.186)

2 20.2540.776 (0.812)

Household members buritiMFAMBA

20.6670.513 (0.117)

0.4751.61 (0.254)

0.6281.88 (0.021)*

20.6070.545 (0.158)

Planted buriti treePLABUR

2.2229.22 (0.029)*

0.9662.63 (0.401)

– –

Buriti trees threatenedTHREAT

– – 20.2850.752 (0.681)

–

Wealth indexWEALTH

0.5031.65 (0.040)*

20.3510.704 (0.320)

20.5940.552 (0.007)*

1.6845.39 (0.001)*

Born in communityBORCOM

0.0911.10 (0.923)

3.12922.84 (0.027)*

20.9660.381 (0.164)

–

Buriti learned from parentLEAPAR

0.6201.86 (0.468)

21.0170.362 (0.406)

0.8962.45 (0.186)

2.40311.05 (0.052)**

Total activitiesATOTAL

20.1240.884 (0.798)

0.5711.77 (0.419)

20.8730.418 (0.019)*

1.7215.59 (0.031)*

Main income buritiMINCBU

– 20.6270.534 (0.604)

20.4200.657 (0.534)

2.2769.73 (0.089)**

Notes: *p , 0.05, **p , 0.10; dash signifies that the factor was not included in the specific model.

8 A. Virapongse et al.

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

Principal actors

Of the sample group, most individuals participated in the buriti market (n ¼ 81).Individuals who did not participate in the buriti market were owners (n ¼ 16) who did notsell buriti derivatives. Individuals could fill multiple roles in the value chain (24% of

sample group), although they usually specialized in one role. Owning buriti resources was

often a secondary role because land was inherited or purchased for other uses. Artisans and

vendors commonly overlapped (13% of sample group) because some artisans transitioned

to vendors as markets became more accessible; 32% of vendors were artisans prior to

becoming vendors. Owners and extractors lived in Laranjeiras area communities, where

buriti forests were located because abundant freshwater was available. About 27% of

artisans and 42% of vendors were from coastal Atins area communities. With exception of

some handicraft artisans and vendors, most actors engaged in the buriti market as part of a

diversified livelihood strategy.

Owners were defined in the study as the household individual who made decisions

regarding land use. Most owners were men who valued buriti trees for their current or

potential economic value, so they rarely cut or removed trees. Although 59% of owners

sampled did not participate in buriti markets, they were considered part of the value chain

because they managed buriti resources that they gave away upon request or were taken

by extractors without permission. Most owners actively managed their land, but others

became absent managers (18% of owners sampled) because they inherited land or

became too elderly. Owners participating in the market often optimized for fruit

production. As reported by owners, fruit could be dried and sold for R$5/kg (US$3) in the

Figure 2. Schematic model of the buriti handicraft value chain showing different options andscenarios for property regimes, value chain actors, and the distribution of benefits earned by eachactor as the derivatives moves through the chain.

Value chain dynamics 9

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

Barreirinhas market. Mature leaves were harvested once every 2–3 years for R$80/100

leaves (US$46). Hiring a leaf extractor cost R$25/100 leaves (US$14). No owners sold

young leaves, mostly because they believed that intense young leaf harvesting negatively

impacted trees. In fact, about half of owners reported preferring to collect young leaves

from unmanaged buriti forests (private land with absent owners) in order to conserve

their own trees. There were also potential use conflicts between actors. For example,

owners often managed for buriti fruit by conserving leaves on productive female trees. In

contrast, artisans favored young leaf fiber collected from female trees, which could have

negative impacts on fruit production.

Extractors were often young men, who were athletically fit to climb tall buriti trees.

Leaf collection was considered a risky activity that did not pay well; extractors often

discouraged their sons from participating. Upon request from a buyer, extractors

collected mature leaves from privately owned land. Extractor reported paying owners

R$25/100 leaves (US$14) or collecting leaves from unmanaged open access land at no

cost. Although collecting from unmanaged land ensured greater profit, access to

resources was physically more challenging and had more potential for land rights

conflicts. Young leaves were collected mostly from unmanaged land. Extractors reported

earning R$95/100 mature leaves (US$54) and R$0.50–2.00/young leaf (US$0.29–1.00).

Extractors of the sample group used two different strategies to gain benefits from leaves:

half of extractors sold buriti leaves to other artisans for direct income, while other

extractors collected leaves for a household-based artisan to make and sell handicrafts.

Because handicrafts were sold for higher prices than leaves, extractors could indirectly

earn more household income by collecting leaves for a family artisan rather than selling

leaves.

Artisans were usually young women with good eyesight and dexterity to make

handicrafts. Almost all handicrafts were made for sale rather than household use. Women

prepared young leaf fiber by stripping the bottom layer of the leaf blade into fibers, boiling,

dyeing, and sun drying the fiber, and then painstakingly knotting fibers together into a

single fiber. Women made complex products such as hats, bags, and tablecloths via

crochet, macramé, and weaving techniques. Most male artisans made simple fiber

handicrafts, such as cords, or assisted their artisan wives. Common bottlenecks in the

handicraft production process were having ability to strip leaves into fiber and market

access. Women lacking skills (15% of artisans) or time to process fiber depended on the

help of other women or purchased fiber for R$25–30/kg (US$14–17). Artisans in the

Atins area depended on intermediaries to transport fiber from the Laranjeiras area;

unprocessed young leaves could not be transported between regions because of their

delicate nature and regional laws that prohibit exportation (Barreirinhas municipal law no.

161/1975). These intermediaries were local men who traditionally transported resources,

such as fish and manioc flour, between regions.

Most artisans believed that they should earn higher returns for their handicrafts. For

example, a tablecloth that required 20 h and two young leaves (0.5 kg of fiber) was sold to

vendors for R$25–30 (US$14–17). A very productive artisan could make 5 tablecloths/

month, which would earn them R$125/month (US$63). Although this value was lower

than federal minimum wage for an equivalent working week (R$270/month, US$135),

artisans could multitask in their home as they made handicrafts. In any case, there were

few jobs available in the region that paid minimum wage. Some artisans managed for lack

of market access by selling to different vendors to maintain a diversity of market outlets.

Artisans who made higher quality handicrafts could demand better prices from vendors

who competed to purchase their wares. It was relatively easy for a vendor to be an artisan,

10 A. Virapongse et al.

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

while an artisan had to overcome socioeconomic limitations to transition into a vendor.

Many vendors owned businesses or shops where they sold handicrafts to consumers, but

few artisans owned shops. Not all artisans were willing to move closer to the consumer end

of the chain, however. Some artisans enjoyed making handicrafts and identified culturally

as artisans. Artisans also associated greater proximity to the market end of the chain with

more stress and responsibility.

Vendors were considered a type of intermediary that provided market access for

artisans by regularly buying and reselling handicrafts to other vendors or consumers for at

least 25% premium. Although 53% of vendors were skilled artisans, they often preferred

to purchase products from other artisans because handicraft production was time

consuming. To make unique products or save time and money, some vendors purchased

uncompleted handicrafts to finish the product personally. Vendors took on risks by

investing in handicrafts that could potentially be left unsold, and earning delayed returns

from their investments. Overall, most vendors were financially successful; many of their

shops had year-round market access.

Socioeconomic factors affecting value chain actors

Socioeconomic characteristics of actors

Means of socioeconomic factors among actors were compared to understand heterogeneity

among actors. All socioeconomic factors, except for four factors, demonstrated

statistically significant differences (Table 2). Results from pair-wise correlation analysis

are qualitatively discussed.

Actors most closely associated with buriti resources (owners and extractors) had less

education and wealth, and fewer livelihood opportunities in comparison to actors more

closely associated with the emergent buriti handicraft market (artisans and vendors).

Progressively higher education was apparent from owner to vendor in the value chain,

which suggested that education helped prepare or encourage people to work in markets.

Age and education were negatively correlated. Older people, such as owners, had low

education partly because widespread secondary education was established only within the

previous decade.

Wealth was measured using an index based on presence of household goods (e.g., tile

roof, inside bathroom, well-made floor and walls, water plumbing, and vehicle ownership),

access to consistent income, availability of household labor, andBolsaFamilia, whichwas a

governmental subsidy awarded to women of poor households with school aged children.

The wealth index and having a consistent source of income were positively correlated.

Extractors were the poorest actors, as they had a low wealth index, few consistent sources

of income, and lacked access to credit, which was often dependent on formal property

ownership. Among all actors, extractors also had the most diverse livelihood strategy

because they participated in a high number of household income sources and livelihood

activities. In contrast, vendors were the wealthiest group, as they had consistent income and

high integration in the market economy. Owners were also relatively wealthy because they

had consistent income, few income sources, and their households were most likely to

receive retirement payments, which were considered a lucrative income source within

communities. Vendors and owners both demonstrated high livelihood stability. Bolsa

Familia was received mostly by extractor and artisan households, and considerably less by

owner and vendor households. Household labor was calculated based on the number of

household members earning income compared to the number of household members.

Although the number of household members was not significantly different among actors,

Value chain dynamics 11

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

household labor was highest among actors on the market end of the value chain, such as

vendors, who had few dependents in their household.

Buriti activities provided a main income source for extractors, artisans, and vendors. In

comparison, owners earned little household income from handicraft production or other

buriti-related activities. Although household use of buriti was not statistically different

between actors, household use of buriti was highest among actors closest to the forest.

Actors closest to the forest were associated with having home gardens and agricultural

plots, which represented subsistence level activities. In contrast, few vendors had home

gardens and agricultural fields, and few artisans had agricultural fields. Vendors were

particularly reliant on purchased food.

Facing similar livelihood pressures, extractors and artisans often cooperated to meet

livelihood goals. Both extractors and artisans had high numbers of household income

sources and earned household income from similar sources, such as Bolsa Familia and

fishing activities. In comparison to other actors, artisans and extractors had more

household members who worked with buriti derivatives, which helped increase efficiency

for completing the labor-intensive handicraft process. In comparison to artisan

households, however, extractor households depended more on household income from

selling palm leaves. Extractors had the largest household sizes of all actors. Palm leaf sale

formed part of their diversified livelihood strategy to meet income demands of a large

young family with low labor availability.

To evaluate the impact of history and tradition regarding buriti on actors’ participation

in the buriti market, actors’ personal history and parental ties with buriti were analyzed.

Statistically, actors had similar rates of being born in their current community and having a

parent with extensive exposure to buriti. All interviewees had greater parental ties to buriti

resources than to their current community of residence. In comparison to other actors,

actors closest to the forest end of the value chain were more likely to have had extensive

exposure to buriti trees. Extractors, of all actors, had the strongest historical ties to their

current community of residence, and exposure and history with buriti. They were also

likely to be born in their current community; all extractors and at least one of their parents

had spent over 10 years close to buriti. Although there was no statistical difference among

actors in terms of reporting that they had learned buriti skills from their parents, extractors

had the lowest rate among actors. Instead, extractors reported being self-taught; they relied

on their extensive exposure to buriti to learn to climb trees.

Perceptions about sustainability of buriti use were measured based on individuals’

perceptions that buriti trees were threatened and collection of young leaves was harmful.

There was little variability among actors with regard to their general perception that buriti

forests were threatened (not statistically different), but actors closer to the forest end of the

value chain were more likely to believe that young leaf collection was harmful to buriti

trees. In contrast, few artisans and, interestingly, no vendors believed that collecting young

leaves was harmful. Reflecting familiarity with the green market discourse, vendors often

defended the sustainability of their craft to tourists, or researchers, by asserting their long

tradition of buriti use and contribution to buriti sustainability (64% of vendors reported

having planted a buriti tree).

Predicting actors

Socioeconomic factors identified in the previous section were used to build best-fit models

for each actor, in order to identify the probability that certain socioeconomic factors were

correlated with different roles in the value chain (Table 3). Logistic regression models

12 A. Virapongse et al.

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

were as follows:

Predicted logit of ðownerÞ¼22:479þð21:829Þ*GENDERþð20:008Þ*EDUCATþð20:667Þ*MFAMBAþð2:222Þ*PLABURþð0:503Þ*WEALTHþð0:091Þ*BORCOMþð0:620Þ*LEAPARþð20:124Þ*ATOTAL;

Predicted logit of ðextractorÞ¼ 3:443þð0:113Þ*AGEþð20:647Þ*EDUCATþð21:663Þ*HOMEGAþð0:475Þ*MFAMBAþð0:966Þ*PLABURþð20:351Þ*WEALTHþð3:129Þ*BORCOMþð21:017Þ*LEAPARþð0:571Þ*ATOTALþð20:627Þ*MINCBU;

Predicted logit of ðartisanÞ ¼ 6:200þ ð20:048Þ*AGEþ ð0:628Þ*MFAMBAþ ð20:594Þ*WEALTHþ ð20:966Þ*BORCOMþ ð0:896Þ*LEAPARþ ð20:873Þ*ATOTALþ ð20:420Þ*MINCBU;

Predicted logit of ðvendorÞ ¼ 216:509þ ð0:008Þ*AGEþ ð0:675Þ*GENDERþ ð0:457Þ*EDUCATþ ð20:254Þ*HOMEGAþ ð20:607Þ*MFAMBAþ ð1:684Þ*WEALTHþ ð2:403Þ*LEAPARþ ð1:721Þ*ATOTALþ ð2:276Þ*MINCBU:

Results showed that the probability of being an owner was lower for men, and higher

for those that planted buriti trees or had a higher wealth index. The probability of being an

extractor was lower for older respondents and for those with high education, but there was

higher probability for individuals born in the community. The probability of being an

artisan was, like for extractors, lower for older respondents and for those with a higher

wealth index and who participated in more activities. If the household had more family

members working with buriti, they were more likely to be artisans. The probability for

being a vendor was higher if an individual was more educated and wealthier, had learned

about their current buriti trade from a parent, reported more activities, and earned a main

income from buriti.

In summary, older individuals were less likely to be extractors or artisans. Although

most owners were men, some owners were female. Men often made land use decisions, but

women often inherited land of which they maintained ownership. Those who were more

highly educated were less likely to be extractors, and more likely to be vendors. Handicraft

production was labor intensive, so young family members often assisted in the process,

which led to higher probabilities of artisans. Individuals who planted a buriti tree were

more likely to be an owner, which was expected because they were land managers.

Wealthier households were more likely to be owners or vendors, but less likely to be an

artisan. Individuals born in the same community in which they currently resided in were

more likely to be extractors. Those who learned their buriti trade from a parent were more

likely to be a vendor. Households reporting more livelihood activities were more likely to

be vendors and less likely to be artisans. Lastly, households whose main income was from

buriti are more likely to be vendors.

Discussion

Although NTFP markets are often based on existing knowledge and use of forest

resources, emerging global markets for NTFPs can introduce socioeconomic and

ecological challenges. Global export markets, of which buriti fiber handicraft markets are

Value chain dynamics 13

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

classified, can be socially and geographically foreign to NTFP users (Philip 2002). As

suggested in other studies (Shillington 2002), a value chain perspective helped reveal the

roles of more hidden groups in the buriti market, such as independent artisans, tree owners,

and extractors. Forest-based users are often erroneously considered as a homogeneous

group, which has implications for the effectiveness of conservation and development

interventions.

Privileged groups, such as actors at the market end of the value chain, tend to dominate

over other users in global export NTFP markets (Belcher & Schreckenberg 2007;

Shackleton et al. 2007). They are also common recipients of external assistance

(McSweeney 2004). In Barreirinhas, long-term capacity-building programs by SEBRAE

(Brazilian micro and small businesses support service) helped women artisans gain new

skills and better access to existing and new markets (e.g., higher paying clients), and

organize handicraft cooperatives (SEBRAE 2007). Through this assistance, many women

used their market advantage to take on roles as vendors to earn higher income.

As markets grow to favor higher quality products, market differentiation can

reconfigure value chains by excluding people who cannot meet product standards (Kanji

et al. 2005; Velde et al. 2006). These dynamics, however, can also create new roles in the

value chain. Less-skilled buriti handicraft artisans found a market for selling products to

other artisans, who reassembled handicrafts for resale. Some extraordinarily skilled

artisans could overcome the social and market advantage that other artisans gained from

external assistance, and could strike out independently in their own niche markets.

Additionally, not all artisans were interested in joining handicraft cooperatives, which

often required women to devote all of their time to producing handicrafts. A household

also needed the appropriate socioeconomic circumstances to support a specialized

livelihood strategy that optimized for one activity.

Strategies used to overcome limitations in the buriti market also had implications for

resource sustainability. Extractors that collected young leaves for artisans in the

household, rather than selling leaves for direct income, participated in value-added

dynamics. This practice could lead to more sustainable use of buriti resources because a

household invested more time into fewer products (less NTFP resources) to generate

higher financial return. Investing into more profitable, higher quality products can reduce

destructive impacts of NTFP harvesting (Varghese & Ticktin 2008) and increase market

advantage (Belcher & Schreckenberg 2007).

Artisans overcame a limited access to buriti resources by relying on social networks,

such as intermediaries or family relations, to ensure a supply of young leaves. Some

independent artisans also maintained market links by selling handicrafts to regular buyers

for low prices. Unfortunately, low prices for handicrafts often drove artisans to optimize

for quantity, instead of quality, which led to greater exploitation of young leaves. By

obtaining greater education, income, and skills, female artisans found an avenue of upward

mobility through the market chain by working as handicraft vendors. Although NTFP

trade often allows women to earn income with little interference by men (Schreckenberg

et al. 2006), more men entering the market as vendors could present future problems as

men often dominate lucrative forest market opportunities (Ruiz Perez et al. 2002).

Although buriti users had found some solutions for overcoming market challenges,

sustainability of buriti resources remained vulnerable. With increased commercialization

of NTFPs, people often intensify extraction activities to take advantage of market

opportunities (Godoy et al. 1993). In comparison to fruit and mature leaves, young buriti

leaf collection had greater potential for overharvesting. Fruit and mature leaves have

higher harvest limits than young leaves (Ticktin 2004). Secure property rights for NTFP

14 A. Virapongse et al.

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

users are considered to be one of the first steps for achieving poverty reduction and

sustainable resource management through commercialization of NTFPs (Ros-Tonen &

Kusters 2011). Indeed, private buriti tree owners resisted young leaf collection to prevent

overharvesting of buriti resources on their property. A negative effect, however, was that

young leaf users were driven to rely on resources collected furtively or from unmanaged

lands. Buriti users often treated unmanaged land, such as the large swaths of buriti forests

in the region with absentee owners, as open access land to be overexploited.

Insecure land rights also meant that extractors were often unwilling to discuss their

extraction activities, which could hinder collection of harvesting information that is vital

for formulating resource management strategies. Extractors often harvested because they

lacked other income-earning options. Extractors’ social and economic vulnerability can

have dire consequences for sustainability of buriti leaf resources because they are more

likely influenced by market pressure to engage in overharvesting practices. They were also

likely to change livelihood activities as their socioeconomic situation improved, leaving

less skilled extractors to fill their gaps in the value chain. Experienced extractors are more

likely than inexperienced extractors to leave forest populations intact by carefully

selecting resources for extraction (Jensen & Meilby 2008; Schmidt & Ticktin 2012).

Buriti tree owners and extractors, who were closer to the forest end value chain, were

generally overlooked by external efforts to improve distribution of benefits and

sustainability of resources. Owners and extractors, however, had extensive exposure and

contact with natural resources, and provided links between artisans and buriti resources.

Having great influence over harvesting practices and investment in protecting resources,

tree owners and extractors would be good candidates as subjects for conservation and

development initiatives.

Conclusions

This study has analyzed the impacts of an emerging buriti handicrafts market on the

heterogeneity of actors in value chains in Maranhão, Brazil. The young buriti leaf

handicraft market has introduced new actors, such as artisans and vendors, and resource

demands that compete with pre-existing local and subsistence uses of buriti fruits and

mature leaves. This study demonstrates that there is much heterogeneity among users of

the same resources, which can affect efforts to improve equal distribution of financial

returns among actors. In addition, financial benefit from buriti products was not always

actors’ main objective. More studies are needed to understand how actors manage the

trade-offs and transitions between subsistence use and commercialization of NTFPs.

Based on this study, some recommendations are made to address socioeconomic and

ecological challenges introduced by an emerging buriti handicraft market:

. Secure property rights for buriti users. New market demand poses a threat tosustainability of buriti harvesting partly because young leaf extraction takes place

primarily on unmanaged public or open access land.

. Prioritize livelihood stability in development interventions. Local people relied onburiti resources for both subsistence and market purposes, depending on their

livelihood strategy and socioeconomic limitations. Protecting against over-

harvesting of young leaf buriti resources would help ensure that all groups can

benefit from the buriti palm.

. Target owners and extractors for training programs aimed to improve sustainableharvest of young buriti leaves.

Value chain dynamics 15

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

. Train a broader group of artisans. By including people who do not optimize for highproduction of handicrafts, all artisans have equal opportunity to produce higher

quality handicrafts, with more awareness of harvesting pressures on buriti

resources.

This analysis of buriti fiber value chains contributes to our understanding of the

complex relations between growing markets and natural resources, which can be used to

inform and guide policy and development interventions that seek to influence livelihoods

and sustainable resource management through participation in NTFP markets. Managing

harvesting and use strategies that are affected by value chain dynamics and heterogeneity

can help to alleviate the negative impacts of increasing and changing NTFP use.

Acknowledgements

The authors are grateful for the generous cooperation of the community members of Barreirinhasthat made this research possible. We thank our research collaborator, Noemi Miyasaka Porro ofUniversity Federal de Pará, for her valuable in-country assistance. We also appreciate the helpfulsuggestions made by anonymous reviewers.

Funding

This research was supported by a doctoral dissertation improvement grant from the US NationalScience Foundation (award ID1032034); a Botany in Action fellowship from Phipps Conservatory,Pittsburg, PA; and a Brazilian Initiation Scholarship from the Brazilian Studies Association.

References

Belcher B, Ruı́z-Pérez M, Achdiawan R. 2005. Global patterns and trends in the use andmanagement of commercial NTFPs: implications for livelihoods and conservation. World Dev.33:1435–1452.

Belcher B, Schreckenberg K. 2007. Commercialisation of non-timber forest products: a realitycheck. Dev Policy Rev. 25:355–377.

Collinson M. 2000. A history of farming systems research. Oxon (UK): FAO/CABI.Coyne IT. 1997. Sampling in qualitative research. Purposeful and theoretical sampling: merging or

clear boundaries? J Adv Nurs. 26:623–630.Cunningham AB. 2001. Applied ethnobotany people, wild plant use and conservation. London:

Earthscan.Godoy R, Lubowski R, Markandya A. 1993. A method for the economic valuation of non-timber

tropical forest products. Econ Bot. 47:220–233.IBGE: Brazilian Institute of Geography and Statistics. 2010. Barreirinhas-MA, Summary of

information [Internet]. [cited 2012 Dec 9]. Available from: http://www.ibge.gov.br/cidadesat/link.php?uf¼ma

IBGE: Brazilian Institute of Geography and Statistics. 2012. Production of vegetal extraction andsilvaculture 2004–2011 [Internet]. IBGE; [cited 2012 Jul 12]. Available from: http://www.ibge.gov.br/estadosat/

Jensen A. 2006. The scramble for cash: access and access mechanisms to income from a commercialNTFP. Scand Soc Forest Econ. 41:115–120.

Jensen A. 2009. Valuation of non-timber forest products value chains. Forest Policy Econ.11:34–41.

Jensen A, Meilby H. 2008. Does commercialization of a non-timber forest product reduce ecologicalimpact? A case study of the critically endangered aquilaria crassna in Lao PDR. Oryx.42:214–221.

Kanji N, MacGregor J, Tacoli C. 2005. Understanding market-based livelihoods in a globalisingworld: combining approaches and methods. London: International Institute for Environment andDevelopment (IIED).

Kaplinsky R, Morris M. 2001. A handbook for value chain research. Sussex (UK): IDRC.

16 A. Virapongse et al.

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

http://www.ibge.gov.br/cidadesat/link.php?uf=mahttp://www.ibge.gov.br/cidadesat/link.php?uf=mahttp://www.ibge.gov.br/estadosat/http://www.ibge.gov.br/estadosat/

Kusters K, Achdiawan R, Belcher B, Pérez M. 2006. Balancing development and conservation? Anassessment of livelihood and environmental outcomes of nontimber forest product trade in Asia,Africa, and Latin America. Ecol Soc. 11:20. Available from: http://www.ecologyandsociety.org/vol11/ iss2/art20/

Lawrence A, Phillips O, Ismodes A, Lopez M, Rose S, Wood D, Fardan A. 2005. Local values forharvested forest plants in Madre de Dios, Peru: towards a more contextualized interpretation ofquantitative ethnobotanical data. Biodivers Conserv. 14:45–79.

Lobato FM. 2008. Diagnóstico Turı́stico do Municı́pio de Barreirinhas – MA [Tourism diagnostic ofBarreirinhas municipality – MA]. São Luı́s: Agência Espanhola de Cooperac�ão Internacional(AECI), SEBRAE-LEGAL.

Marshall E, Schreckenberg K, Newton A. 2006. Commercialization of non-timber forest products:factors influencing success: lessons learned fromMexico and Bolivia and policy implications fordecision-makers. Cambridge (UK): UNEP World Conservation Monitoring Centre.

May PH. 1986. A modern tragedy of the non-commons: agro-industrial change and equity in Brazil’sbabassu palm zone. New York: Cornell University.

McSweeney K. 2004. Forest product sale as natural insurance: the effects of householdcharacteristics and the nature of shock in eastern Honduras. Soc Nat Resour. 17:39–56.

Neumann RP, Hirsch E. 2000. Commercialisation of non-timber forest products: review and analysisof research. Bogor (Indonesia): Center for International Forestry Research.

Philip K. 2002. The quest for rural enterprise support strategies that work the case of mineworkers’development agency. Small Enterprise Dev. 13:13–25.

Prefeitura Barreirinhas. 2005. Código municipal do meio ambiente de Barreirinhas [Municipal codeof Barreirinhas environment]. Barreirinhas, MA: Prefeitura Barrerinhas.

Ros-Tonen MAF, Kusters K. 2011. Pro-poor governance of non-timber forest products: the need forsecure tenure, the rule of law, market access and partnerships. In: Shackleton S, Shackleton C,Shanley P, editors. Non-timber forest products in the global context. Berlin: Springer.p. 189–207.

Ruiz-Perez M, Belcher B, Achdiawan R, Alexiades M, Aubertin C, Caballero J, Campbell B,Clement C, Cunningham T, Fantini A. 2004. Markets drive the specialization strategies of forestpeoples. Ecol Soc. 9:4. Available from: http://www.ecologyandsociety.org/vol9/iss2/art4/

Ruiz Perez M, Ndoye O, Eyebe A, Ngono DL. 2002. A gender analysis of forest product markets inCameroon. Afr. Today. 49:97–126.

Ruiz-Perez MR, Belcher B, Fu M, Yang X. 2004. Looking through the bamboo curtain: an analysisof the changing role of forest and farm income in rural livelihoods in China. Int Forestry Rev.6:306–316.

Salganik MJ, Heckathorn DD. 2004. Sampling and estimation in hidden populations usingrespondent-driven sampling. Sociol Methodol. 34:193–240.

Scherr S, White A, Kaimowitz D, Trends F. 2004. A new agenda for forest conservation and povertyreduction: making forest markets work for low-income producers. Washington (DC): ForestTrends.

Schmidt IB, Ticktin T. 2012. When lessons from population models and local ecological knowledgecoincide: effects of flower stalk harvesting in the Brazilian savanna. Biol Conserv.152:187–195.

Schmink M. 2004. Communities, forests, markets, and conservation. Working forests in the tropics:conservation through sustainable management? New York: Columbia University Press.p. 119–129.

Schreckenberg K, Rushton J, Newton A, Marshall E, Velde DWt. 2006. Research methodology. In:Marshall E, Schreckenberg K, Newton A, editors. Commercialization of non-timber forestproducts: factors influencing success: lessons learned from Mexico and Bolivia and policyimplications for decision-makers. Cambridge (UK): UNEP World Conservation MonitoringCentre.

Schwartzman S. 1992. Social movements and natural resource conservation in the BrazilianAmazon. Sustainable strategies for saving tropical forests. London (UK): Friends of the Earth.p. 207–212.

SEBRAE. 2007. Projeto: Artesanato nos Lenc�óis [Project: handicrafts of Lenc�óis [Internet].Maranhão: SEBRAE. Available from: http://sebrae-legal.com.br/arquivos/Projeto_Artesanato_Lencois.htm (accessed 11 November 2012).

Value chain dynamics 17

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

http://www.ecologyandsociety.org/vol11/ iss2/art20/http://www.ecologyandsociety.org/vol11/ iss2/art20/http://www.ecologyandsociety.org/vol9/iss2/art4/http://sebrae-legal.com.br/arquivos/Projeto_Artesanato_Lencois.htmhttp://sebrae-legal.com.br/arquivos/Projeto_Artesanato_Lencois.htm

Shackleton C, Shackleton S. 2004. The importance of non-timber forest products in rural livelihoodsecurity and as safety nets: a review of evidence from South Africa. South African J Sci.100:658–664.

Shackleton S, Shanley P, Ndoye O. 2007. Invisible but viable: recognising local markets for non-timber forest products. Int Forestry Rev. 9:697–712.

Shackleton SE, Shackleton CM, Wynberg R, Sullivan CA, Leakey RRB, Mander M, McHardy T,Den Adel S, Botelle A, Du Plesis P, et al. 2009. Livelihood trade-offs in the commercializationof multiple-use NTFP: lessons from marula (Sclerocarya birrea subsp. caffra) in SouthernAfrica. In: Shaanker RU, Hiremath A, Joseph G, Rai ND, editors. Non-timber forest products:conservation, management and policy in the tropics. Bangalore: Ashoka Trust for Research inEcology and the Environment (ATREE). p. 139–173.

Shillington LJ. 2002. Non-timber forest products, gender, and households in Nicaragua: acommodity chain analysis. Blacksburg, Virginia: Virginia Polytechnic Institute and StateUniversity.

Sutton MQ, Anderson EN. 2004. Introduction to cultural ecology. Walnut Creek, CA: AltaMiraPress.

Ticktin T. 2004. The ecological implications of harvesting non-timber forest products. J Appl Ecol.41:11–21.

Varghese A, Ticktin T. 2008. Regional variation in non-timber forest product harvest strategies,trade, and ecological impacts: the case of black dammar (Canarium strictum Roxb.) use andconservation in the Nilgiri biosphere reserve, India. Ecol Soc. 13:11. Available from: http://www.ecologyand society.org/vol13/iss2/art11/

Velde DW, Rushton J, Schreckenberg K, Marshall E, Edouard F, Newton A, Arancibia E. 2006.Entrepreneurship in value chains of non-timber forest products. Forest Policy Econ. 8:725–741.

Virapongse A. 2013. Forest products for subsistence and markets: livelihood systems and valuechains of buriti (Mauritia flexuosa) in Brazil [dissertation]. Gainesville: School of NaturalResources and Environment, University of Florida.

Williams V, Balkwill K, Witkowski E. 2000. Unraveling the commercial market for medicinal plantsand plant parts on the Witwatersrand, South Africa. Econ Bot. 54:310–327.

Wilsey D. 2008. Integrating conservation and development objectives through non-timber forestproduct certification: the case of Chamaedorea palms [dissertation]. Gainesville: School ofNatural Resources and Environment, University of Florida.

Wynberg R, Laird S, Botha J, Den Adel S, McHardy T. 2002. The management, use andcommercialisation of marula: policy issues. London (UK): DFID/FRP Winners and Losers inForest Product Commercialisation.

18 A. Virapongse et al.

Dow

nloa

ded

by [

arik

a vi

rapo

ngse

] at

13:

01 2

6 D

ecem

ber

2013

http://www.ecologyand society.org/vol13/iss2/art11/http://www.ecologyand society.org/vol13/iss2/art11/

AbstractIntroductionMethodsAnalytical frameworkCase studySampling strategyData collection and analysis

ResultsBuriti value chainsProduction systemPrincipal actors

Socioeconomic factors affecting value chain actorsSocioeconomic characteristics of actorsPredicting actors

DiscussionConclusionsAcknowledgementsReferences