Read all pages carefully, selected topic, intro & must be follow format according to Sample Paper (3-3 Para each Article)       

Must be 100% Original        

I hv already attached Articles, u must be use this 3 articles for Annotated Bibliography      

Wk 3 Discussion (Due in 1 day) Urgent/..Wk 3 Discussion (Required Assignment) Due in 1 day.docx

(Must be 4 Pages wriiten, in page count/Length Not included Title 1st Page)

Must be 100% Original Work Assignment must be follow Rubric Superior Criteria.

Plz read My Note, Important tips (Wrote on 2nd Page) and also sample paper attached.

Must be use attached Three Article

NOTE: I hv attached 3 Articles & include each Article have (3 para) three paragraph
summary, Analysis
to the study.

New Selected topic: Strategies Used by Agency Leaders to Safeguard Rosewood Trade (Annotated Bibliography must be write on related this topic & Apply)

MY Notes:
(Must see sample paper)

Sample Annotated Bibliography attached so must be follow & minimum 3 pages required & three (3) peer-reviewed sources (no older than 5 years).

(4 Pages required )Must be include Abstract/Intro like in sample

Course: DDBA – Doctoral Study Mentoring

Selected topic: Strategies Used by Agency Leaders to Safeguard Rosewood Trade

Discussion 2: Annotated Bibliography

In each week of this course, you will research and select three (3) peer-reviewed, scholarly sources to develop an annotated bibliography that you can use in your Doctoral Study. You will need to take the three sources and synthesize the references into a single narrative annotated bibliography that compares/contrasts or supports your study. For example, you may develop three references that will fit into the Nature of the Study (or any other component) and then the synthesized version will help you in developing your Prospectus/Proposal. Please see this week’s Learning Resources for the Sample Annotated Bibliography Template, which you should use to complete your annotated bibliography.

By Day 3

Post your synthesized annotated bibliography narrative that includes an explanation of how these references relate to one or more components of your Doctoral Study and incorporates specific references to the Doctoral Study Rubric.

Refer to the Week 3 Discussion 2 Rubric for specific grading elements and criteria. Your Instructor will use this rubric to assess your work.

Important tips: Include each Article annotated bibliography have three paragraph summary, Analysis and applies to the study

Walden’s recommendations for formatting an AB includes three areas, typically formatted in three paragraphs: 

This first paragraph of the annotation summarizes the source. It outlines the main findings and primary methods of the study.

Summary: What did the author do? Why? What did he/she find?

This second paragraph of the annotation analyzes the source. It explains the benefits of the source but also the limitations.

Analysis: Was the author’s method sound? What information was missing? Is this a scholarly source?

This third paragraph of the annotation applies the source. It explains how the source’s ideas, research, and information can be applied to other contexts.

Application: Does this article apply to the literature? How would you be able to apply this method/study to your particular study? Is the article universal?

In general, annotated bibliographies should avoid referring to the first or second person (I, me, my, we, our, you, and us). Instead, students should aim to be objective and remove themselves from annotations. However, there may be some exceptions to this guideline. Check with your instructor if you are unsure about whether he/she will allow you to use “I” in your annotated bibliography.

Must be use Below Three Article for Annotated Bibliography & related intro & topic

Guadilla-Sáez, S., Pardo-de-Santayana, M., & Reyes-García, V. (2020). Forest Commons, traditional community ownership and ecological consequences: Insights from Spain. Forest Policy and Economics, 112, 102107.

Handavu, F., Chirwa, P. W. C., & Syampungani, S. (2019). Socio-economic factors influencing land-use and land-cover changes in the miombo woodlands of the Copperbelt Province in Zambia. Forest Policy and Economics, 100, 75–94.

Tadesse, T., Teklay, G., Mulatu, D. W., Rannestad, M. M., Meresa, T. M., & Woldelibanos, D. (2022). Forest benefits and willingness to pay for sustainable forest management. Forest Policy and Economics, 138, 102721.

(Guadilla-Sáez et al., 2020)

(Handavu et al., 2019)

(Tadesse et al., 2022)

Assignment must be follow Rubric Superior Criteria

Rubric Detail








Not Submitted

Element 1: Annotated Bibliography (post and attach document)

6.6 (30%)

Student posts and includes an attachment of his/her annotated bibliography which includes three peer-reviewed, scholarly sources that are thoroughly synthesized into a single, well-written narrative annotated bibliography that explicitly compares/contrasts or supports his/her study. A thorough and detailed explanation of how the sources relate to his/her study is evident.

6.27 (28.5%)

Student posts and includes an attachment of his/her annotated bibliography which includes three peer-reviewed, scholarly sources that are thoroughly synthesized into a single, well-written narrative annotated bibliography that explicitly compares/contrasts or supports his/her study. A detailed explanation of how the sources relate to his/her study is evident. One or two minor details are missing or lack clarity.

5.61 (25.5%)

Student posts and includes an attachment of his/her annotated bibliography which includes three peer-reviewed, scholarly sources that are synthesized into a single narrative annotated bibliography that explicitly compares/contrasts or supports his/her study. An explanation with some details of how the sources relate to his/her study is evident.

4.95 (22.5%)

Student posts and includes an attachment of his/her annotated bibliography which includes three peer-reviewed, scholarly sources that are somewhat synthesized into a single narrative annotated bibliography that compares/contrasts or supports his/her study. A cursory statement of how the sources relate to his/her study is evident.

3.3 (15%)

Does not meet minimal standards and/or is posted late.


Did not submit element.

Element 2: Follow-up Responses

8.8 (40%)

On Day 5 and on Day 7, student’s responses fully contribute to the quality of interaction by offering constructive critique, suggestions, in-depth questions, and/or additional resources related to peers’ annotated bibliography. Student demonstrates active engagement with more than one peer on at least two days in the discussion forum (or with Instructor if there are no other peers/posts).

8.36 (38%)

On Day 5 and on Day 7, student shares some constructive critique, suggestions, in-depth questions, and/or additional resources related to peers’ annotated bibliography, but more depth and/or clarity around ideas is needed. Student demonstrates active engagement with more than one peer on at least two days in the discussion forum (or with Instructor if there are no other peers/posts).

7.48 (34%)

Student did not post on Day 5 and on Day 7, but he/she did engage with at least one peer (or with Instructor if there are no other peers/posts) during the week offering constructive feedback related to peers’ annotated bibliography.

6.6 (30%)

Student posts to at least one peer (or with Instructor if there are no other peers/posts) but response is cursory and/or off topic.

4.4 (20%)

Does not meet minimal standards and/or student posted late.


Did not submit element.

Element 3: Written Delivery Style & Grammar

3.3 (15%)

Student consistently follows APA writing style and basic rules of formal English grammar and written essay style. Student communicates in a cohesive, logical style. There are no spelling or grammar errors.

3.13 (14.25%)

Student consistently follows APA writing style and basic rules of formal English grammar and written essay style. Student communicates in a cohesive, logical style. There are one or two minor errors in spelling or grammar.

2.81 (12.75%)

Student mostly follows APA writing style and basic rules of formal English grammar and written essay style. Student mostly communicates in a cohesive, logical style. There are some errors in spelling or grammar.

2.48 (11.25%)

Student does not follow APA writing style and basic rules of formal English grammar and written essay style and does not communicate in a cohesive, logical style.

1.65 (7.5%)

Does not meet minimal standards.


Did not submit element.

Element 4: Formal and Appropriate Documentation of Evidence, Attribution of Ideas (APA Citations)

3.3 (15%)

Student demonstrates full adherence to scholarly reference requirements and adheres to APA style with respect to source attribution, references, heading and subheading logic, table of contents and lists of charts, etc. There are no APA errors.

3.13 (14.25%)

Student demonstrates full adherence to scholarly reference requirements and adheres to APA style with respect to source attribution, references, heading and subheading logic, table of contents and lists of charts, etc. There are one or two minor errors in APA style or format.

2.81 (12.75%)

Student mostly adheres to scholarly reference requirements and/or mostly adheres to APA style with respect to source attribution, references, heading and subheading logic, table of contents and lists of charts, etc. Some errors in APA format and style are evident.

2.48 (11.25%)

Student demonstrates weak or inconsistent adherence scholarly reference requirements and/or weak or inconsistent adherence to APA style with respect to source attribution, references, heading and subheading logic, table of contents and lists of charts, etc. Several errors in APA format and style are evident.

1.65 (7.5%)

Does not meet minimal standards.


Did not submit element.

Wk 3 Discussion (Due in 1 day) Urgent/.Sample_Annotated_Bibliography.doc



Sample Annotated Bibliography

Student Name Here

Walden University

Sample Annotated Bibliography

research continues to grapple with activities that best serve the purpose of fostering positive interpersonal relationships for children who struggle with autism. Children have benefited from therapy sessions that provide ongoing activities to aid autistic children’s ability to engage in healthy social interactions. However, less is known about how K–12 schools might implement programs for this group of individuals to provide additional opportunities for growth, or even if and how school programs would be of assistance in the end. There is a gap, then, in understanding the possibilities of implementing such programs in schools to foster the social and thus mental health of children with autism.

Annotated Bibliography

, M. C., Dinehart, L. H., & Winick, C. B. (2016). Child-centered play therapy for children with autism spectrum disorder. In A. A. Drewes & C. E. Schaefer (Eds.), Play therapy in middle childhood (pp. 103–147). Washington, DC: American Psychological Association.

In this chapter, Kenny, Dinehart, and Winick provided a case study of the treatment of a 10-year-old boy diagnosed with autism spectrum disorder (ADS). Kenny et al. described the rationale and theory behind the use of child-centered play therapy (CCPT) in the treatment of a child with ASD. Specifically, children with ADS often have sociobehavioral problems that can be improved when they have a safe therapy space for expressing themselves emotionally through play that assists in their interpersonal development. The authors outlined the progress made by the patient in addressing the social and communicative impairments associated with ASD. Additionally, the authors explained the role that parents have in implementing CCPT in the patient’s treatment. Their research on the success of CCPT used qualitative data collected by observing the patient in multiple therapy sessions

CCPT follows research carried out by other theorists who have identified the role of play in supporting cognition and interpersonal relationships. This case study is relevant to the current conversation surrounding the emerging trend toward CCPT treatment in adolescents with ASD as it illustrates how CCPT can be successfully implemented in a therapeutic setting to improve the patient’s communication and socialization skills. However, Kenny et al. acknowledged that CCPT has limitations—children with ADS, who are not highly functioning and or are more severely emotionally underdeveloped, are likely not suited for this type of therapy

Kenny et al.’s explanation of this treatments’s implementation is useful for professionals in the psychology field who work with adolescents with ASD. This piece is also useful to parents of adolescents with ASD, as it discusses the role that parents can play in successfully implementing the treatment. However, more information is needed to determine if this program would be suitable as part of a K–12 school program focused on the needs of children with ASD

Stagmitti, K. (2016). Play therapy for school-age children with high-functioning autism. In A.A. Drewes and C. E. Schaefer (Eds.), Play therapy in middle cildhood (pp. 237–255). Washington, DC: American Psychological Association.

Stagmitti discussed how the Learn to Play program fosters the social and personal development of children who have high functioning autism. The program is designed as a series of play sessions carried out over time, each session aiming to help children with high functioning autism learn to engage in complex play activities with their therapist and on their own. The program is beneficial for children who are 1- to 8-years old if they are already communicating with others both nonverbally and verbally. Through this program, the therapist works with autistic children by initiating play activities, helping children direct their attention to the activity, eventually helping them begin to initiate play on their own by moving past the play narrative created by the therapist and adding new, logical steps in the play scenario themselves. The underlying rationale for the program is that there is a link between the ability of children with autism to create imaginary play scenarios that are increasingly more complex and the development of emotional well-being and social skills in these children. Study results from the program have shown that the program is successful: Children have developed personal and social skills of several increment levels in a short time. While Stagmitti provided evidence that the Learn to Play program was successful, she also acknowledged that more research was needed to fully understand the long-term benefits of the program.

Stagmitti offered an insightful overview of the program; however, her discussion was focused on children identified as having high-functioning autism, and, therefore, it is not clear if and how this program works for those not identified as high-functioning. Additionally, Stagmitti noted that the program is already initiated in some schools but did not provide discussion on whether there were differences or similarities in the success of this program in that setting.

Although Stagmitti’s overview of the Learn to Play program was helpful for understanding the possibility for this program to be a supplementary addition in the K–12 school system, more research is needed to understand exactly how the program might be implemented, the benefits of implementation, and the drawbacks. Without this additional information, it would be difficult for a researcher to use Stigmitti’s research as a basis for changes in other programs. However, it does provide useful context and ideas that researchers can use to develop additional research programs.

Wimpory, D. C., & Nash, S. (1999). Musical interaction therapy–Therapeutic play for children with autism. Child Language and Teaching Therapy, 15(1), 17–28. doi:10.1037/14776-014

Wimpory and Nash provided a case study for implementing music interaction therapy as part of play therapy aimed at cultivating communication skills in infants with ASD. The researchers based their argument on films taken of play-based therapy sessions that introduced music interaction therapy. To assess the success of music play, Wimpory and Nash filmed the follow-up play-based interaction between the parent and the child. The follow-up interactions revealed that 20 months after the introduction of music play, the patient developed prolonged playful interaction with both the psychologist and the parent. The follow-up films also revealed that children initiated spontaneously pretend play during these later sessions. After the introduction of music, the patient began to develop appropriate language skills.

Since the publication date for this case study is 1999, the results are dated. Although this technique is useful, emerging research in the field has undoubtedly changed in the time since the article was published. Wimpory and Nash wrote this article for a specific audience, including psychologists and researchers working with infants diagnosed with ASD. This focus also means that other researchers beyond these fields may not find the researcher’s findings applicable.

This research is useful to those looking for background information on the implementation of music into play-based therapy in infants with ASD. Wimpory and Nash presented a basis for this technique and outlined its initial development. Thus, this case study can be useful in further trials when paired with more recent research.

�The format of an annotated bibliography can change depending on the assignment and instructor preference, but the typical format for an annotated bibliography in academic writing is a list of reference entries with each entry followed by an annotation (hence the name, “annotated bibliography”).

However, APA does not have specific rules or guidelines for annotated bibliographies, so be sure to ask your instructor for any course-specific requirements that may vary from the general format.

�An introduction is a helpful addition to your annotated bibliography to tell your reader (a) your topic and focus for your research and (b) the general context of your topic.

Although your assignment instructions may not explicitly ask for an introduction, your instructor might expect you to include one. If you are not sure, be sure to ask your instructor.

�Use a Level 1 heading titled “Annotated Bibliography” or any other wording your instructor has given you to indicate to your reader that the annotations will go next and separate this section from the introduction paragraph above.

�Format your reference entries per APA, as well as follow APA style when writing your paragraphs. However, as mentioned above, this is the extent of the formatting requirements APA has for annotated bibliographies.

The content of the paragraphs and how many paragraphs you include in each annotation follows academic writing conventions, your assignment guidelines, and your instructor preferences.

�This first paragraph of the annotation summarizes the source. It outlines the main findings and primary methods of the study.

�This second paragraph of the annotation analyzes the source. It explains the benefits of the source but also the limitations.

�This third paragraph of the annotation applies the source. It explains how the source’s ideas, research, and information can be applied to other contexts.

In general, annotated bibliographies should avoid referring to the first or second person (I, me, my, we, our, you, and us). Instead, students should aim to be objective and remove themselves from annotations. However, there may be some exceptions to this guideline. Check with your instructor if you are unsure about whether he/she will allow you to use “I” in your annotated bibliography.

Wk 3 Discussion (Due in 1 day) Urgent/Forest benefits and willingness to pay.pdf

Forest Policy and Economics 138 (2022) 102721

Available online 17 March 2022
1389-9341/© 2022 Elsevier B.V. All rights reserved.

Forest benefits and willingness to pay for sustainable forest management

Tewodros Tadesse a, *, Gebreegziabher Teklay b, Dawit W. Mulatu c, d, Meley
Mekonen Rannestad e, Tigabu Molla Meresa f, Dawit Woldelibanos g

a Department of Agricultural and Resource Economics, Mekelle University, P.o.Box 231, Mekelle, Tigray, Ethiopia
b Tigray Bureau of Water Resources, P.o.Box 520, Mekelle, Tigray, Ethiopia
c Environment and Climate Research Center (ECRC), Policy Studies Institute (PSI), P.o.Box 2479, Addis Ababa, Ethiopia
d Environment, Natural Resources and the Blue Economy Global Practice, World Bank: Africa Region, Addis Ababa, Ethiopia
e Faculty of Environmental Sciences and Natural Resource Management, Norwegian University of Life Sciences (NMBU), Ås, Norway
f Department of Economics, Woldia University, P.o.Box 400, Woldia, Ethiopia.
g Relief Society of Tigray (REST), P.o.Box 361, Mekelle, Tigray, Ethiopia


Forest management
Willingness to pay
Mixed logit


The Desa’a forest in Tigray is a national forest priority area managed by the government. Government authorities
grant access to local communities as the forest is the source of income for many households. To allow the forest
provide such economic needs and ecological functions, its sustainability needs to be ensured. We studied the
effect of a demand-relevant economic attribute and two policy–relevant ecological attributes to examine pref-
erence for alternative forest management mechanisms. We find positive preference for the supply of forest
products and soil and water conservation. Marginal WTP is higher for the most sustainable levels, indicating
households’ preference to forest management mechanisms that ensure supply of forest products for longer pe-
riods and significantly reduce soil erosion and water loss. On the other hand, observed heterogeneity implies
differences in preferences and tradeoffs between economic benefits (forest products) and ecological improve-
ments (soil and water conservation). Overall, the positive preference for these important attributes emanates
both from realizations of economic benefits and ecological services from the forest, suggesting the importance of
incorporating such competing preferences (interests and needs) in forest management plans.

1. Introduction

Forest products constitute important sources of income for local
communities in many developing countries (Gordillo et al., 2019;
Kazungu et al., 2020). In this regard, different studies report forest-
based benefits both from private woodlots and public forests make up
between 12% and54% of rural household income (Tesfaye et al., 2011;
Kazungu et al., 2020; Tadesse et al., 2021), indicating the importance of
income from forest resources. Given this important role, it is argued that
the sustainable management of forests is closely linked with the direct
(Kazungu et al., 2020) and indirect benefits (Bamwesigye et al., 2020)
local communities obtain from forests. These direct and indirect benefits
in turn govern communities’ preference for sustainable forests man-
agement mechanisms, which in turn are influenced by preferences for
forest access and accessibility that define local communities’ livelihood
(Gordillo et al., 2019). However, other factors including ownership
feelings as well as inclusive and participatory decision-making related to

access and accessibility also influence preferences for alternative forest
management mechanisms.

Local communities’ sense of ownership, participation in decision-
making and benefits they obtain from forests are important to sustain-
able forest management. In many developing countries where rural
communities depend on forests for livelihood, alternative forest gover-
nance structures that consider farmers’ attitudes, sense of ownership,
participation and benefits have been implemented. Among these forest
governance structures are community-based forest management
(Takahashi and Todo, 2012), farmer-managed natural regeneration
(Haglund et al., 2011; Weston et al., 2015) and state-led participatory
forest management (Giday, 2013). All these forest governance structures
may not be practically implemented in such forestlands as the Desa’a
forest in Tigray (north-east Ethiopia), which is designated as a National
Forest Priority Area (NFPA) and owned and mainly managed by the
state. Coupled with the weak management of the Desa’a forest (Giday,
2013), effective implementation of any forest governance mechanisms

* Corresponding author at: Department of Agricultural and Resource Economics, Mekelle University, P.o.Box 231, Mekelle, Ethiopia.
E-mail address: [email protected] (T. Tadesse).

Contents lists available at ScienceDirect

Forest Policy and Economics

journal homepage:
Received 17 April 2021; Received in revised form 4 March 2022; Accepted 5 March 2022

Forest Policy and Economics 138 (2022) 102721


that do not account for local communities’ sense of ownership and
economic needs may not materialize (Gordillo et al., 2019; Woldie and
Tadesse, 2019). In such cases where the state has full control over
ownership and enforcement in the use of forest resources, policy
mechanisms that involve local communities at different stages of
decision-making can contribute not only to sustainable forest manage-
ment but also sustaining forest-based benefits to local communities
(Tesfaye et al., 2011).

Crucially therefore, whatever forest management governance struc-
tures are put in place, they need to balance direct economic needs of
local communities with ecological sustainability1 for effective and sus-
tainable forest management (Giergiczny et al., 2015). In this regard,
Tadesse et al. (2021) report that local communities would be willing to
engage in forest conservation programs if direct economic needs are
integrated into forest management plans. For integrating local com-
munities’ economic needs into forest management plans, there is a need
to capture and understand the economic value local communities place
on and preference for alternative forest management mechanisms. The
Desa’a forest in north-east Ethiopia is owned and managed by the state
and protected from encroachment (albeit weakly), with local commu-
nities participating in decision-making related to selective and restricted
access to the forest. In the process, both parties work to ensure balance
between meeting ecological conditions and economic needs. For effec-
tive and sustainable management of forests and their resources there-
fore, tradeoffs that local forest communities make among ecological

conditions and economic interests are important.
With this in the hindsight, this paper reports results on how forest-

based benefits influence preferences for (more) sustainable forest man-
agement attributes (characteristics).Based on this, we contribute to the
literature in two ways. On the one hand, we consider a demand and
policy-relevant attribute (described as forest products) that accounts for
the economic needs of local communities from the forest. Then, we
consider two other demand and policy-relevant ecological attributes
(soil and water conservation and absorption of heat wave) that reflect
ecological functions local communities put value on. Considering such
contrasting attributes that reflect competing interests(economic needs
versus ecological functions) helps explore the tradeoffs forest commu-
nities make between forest-based benefits (economic needs) and envi-
ronmental amenities (ecological services), which influence preferences
for forest management mechanisms. Based on this, choices elicited
through discrete choice experiment were analyzed using mixed logit
model to explore preferences and estimate WTP for alternative forest
management attributes.

2. Desa’a state forest, study areas and sample data

Desa’a forest, located in north-east Tigray (Ethiopia), is designated
as a National Forest Priority Area (Gebreegziabher, 1999), meaning the
state has full control over ownership and enforcement over its man-
agement. Geographically, it is situated between 13◦ 20′ and 14◦ 10′
North latitude and 39◦ 32′ and 39◦ 55′ East longitude (Fig. 1), along the
western escarpment of the Great Rift Valley facing the Afar depression.
Estimates show that the forest covers an area of about 41,000 ha (Seb-
hatleab, 2012; Giday, 2013). The forest lies on the border of two
regional states of Ethiopia: Tigray Regional State and the Afar Regional

Fig. 1. Location map of the Desa’a forest and study communities.

1 Local communities make tradeoffs between provisioning ecosystem services
(direct economic needs) such as firewood, grass, wild fruits, etc. and supporting
or regulating ecosystem services such as soil and water conservation and micro-
climate regulation (ecological benefits).

T. Tadesse et al.

Forest Policy and Economics 138 (2022) 102721


State. However, considerable part of the forestland lies within Tigray
Regional State (Gebreegziabher, 1999). Due to its role in defining live-
lihood of local communities, the forest has over a long period been
exposed to various sorts of pressures including deforestation, expansion
of agricultural land into the forest and overgrazing (Gebreegziabher,
1999; Tesfay, 2008; Giday, 2013).

For studying the benefit from the forest and preference for the sus-
tainable management attributes, local communities around the forest-
land were considered. To consider representative communities and final
set of sample households, multistage sampling procedure was used. In
the first stage, three woredas,2 namely Atsbi–Womberta and Enderta
(from Tigray Regional State) and Berahle (from Afar Regional State)
were purposively selected as the forest lies within their administrative
boundaries and are immediate stakeholders of the forest. Following this,
discussions with woreda agricultural office leaders and experts as well as
members of the local communities were made to select the communities.
Based on the discussions, seven relevant tabias were randomly selected.
Following this, roster of household population was obtained from local
administrations of each tabia, which was used as sampling frame to
select the final sample of households randomly. In the end, a sample of
240 households was selected from the tabias based on probability to
proportional size (PPS) (Krzanowski, 2007). The sample size is about
10.7% of the total household population of the study communities. The
final set of sample households was distributed across four blocks (where
each block constituted 60 sample households) for conducting the choice
experiment. We present additional description about this in subsection
3.1 and section A.1 in the appendix.

3. Preference for forest management mechanisms

3.1. Attributes of forest management

We start by hypothesizing that benefits communities obtain from
forests represent key factors in determining farmers’ preferences for the
sustainable management of forests. Our argument is for people to be
willing to commit to sustainable forest management, their needs and
priorities shall be considered. As documented in the literature (Tesfaye
et al., 2011; Kazungu et al., 2020; Tadesse et al., 2021), local commu-
nities earn significant part of their income from forests. To insure sus-
tainable forest management, such economic needs and priorities of local
communities shall be reflected in choice experiment designs. In this
regard, the most important (prioritized) need is forest products that
make up an important share of the income of local communities. The
overall benefit local communities obtain from forests could thus influ-
ence their preference for sustainable management mechanisms. For
examining preferences, an attribute–access to forest products– that cap-
tures economic needs was used in addition to two other policy and de-
mand relevant ecological attributes–soil and water conservation, and
protection from heat waves. These set of attributes were selected based
detail discussions with forest experts (scholars), stakeholder focus
groups (farmers, policy and local forest management bodies) and
reference to relevant literature.

The supply of forest products attribute was measured by the number of
years that the forest will continue to supply forest products to local
communities. In consultation with experts, community elders and
members, we needed to come up with an estimate of the number of years
that the forest will continue to supply local communities with products.
For this, we needed to know what had happened to Desa’a forest cover in
the past and what may happen in the future. According to Gebreeg-
ziabher (1999), as a result of a host of extractive activities, dating from
pre-colonial times and compounded by high population growth, the size

of the forest has been shrinking rapidly in the last century. The original
forest cover of Desa’a during demarcation was estimated to be 120,000
ha (BoANR, 1997). But because of many internal and external factors,
Desa’a forest cover has been severely diminishing in the last four de-
cades alone, and the forest cover currently is estimated to be about
41,000 ha (Sebhatleab, 2012; Giday, 2013). From this evidence, it can
be deduced that the forest cover has been diminishing at a rate of
20,000 ha per decade. This in turn would mean that access to forest
products is correlated with the rate of deforestation. Stated otherwise,
the forest would continue to supply products as long as its core integrity

Table 1
Description of attributes and attribute levels.

Attributes Description of attributes Levels

Supply of forest

Supply of forest products: This is the
number of years that the supply of
forest products is secured. Given the
current situation, the forest will
continue to provide meaningful
forest products that households need
for only 20 more years, as estimated
by Giday (2013). Depending on
alternative improved forest
management regimes, the supply of
these forest products can be secured
for longer periods. From expert
opinions and farmers perception, the
alternative length of period that the
forest will continue to supply
products meaningfully are 40 and 60
more years.

Length of period for
supply of forest
40 more years
60 more years
Status quo: 20 more

Soil and water

Soil and water conservation: This is
the amount of soil and water
conserved because of the presence of
the forest. It is known that the forest
prevents soil erosion by retaining the
soil and water through its buffering
role as wind and rainfall breaker.
However, with diminishing of forest
cover, it has been observed that
many farmers in the area experience
soil and water loss. If the current
situation persists, the current rate of
soil and water loss will continue. This
rate of soil and water loss can be
slowed down by half (50%), or
three–fourth (75%), depending on
the way the forest is managed.

Degree of reduction
in soil and water loss
50% reduction
75% reduction
Status quo: no

Heat protection/
heat absorption

Protection from heat: Protection
from heat that comes from Ertaele
molten volcanic fire (cooling effect)
is an attribute that is greatly valued
by local communities. The degree of
heat reduction through the absorbing
capacity of the forest and the
valuation local communities
however may vary. Based on
stakeholder discussion involving
local communities as well, the
alternative degrees of reduction in
heat were set at 20% and 50%.

Degree of heat
20% reduction
50% reduction
Status quo: no

Payment Payment elicitation: The estimated
payment amounts reflect local labor
market situations, where daily
laborers are paid a minimum of 140
and a maximum of 190 Birr per day.
The stakeholder discussion
(involving local bodies and
community farmers) emphasized the
subsistence-based livelihood of local
communities and a payment rate of
that amount monthly would be

140 Birr
190 Birr
Status quo: 0 (no

Note: during data collection, 1 USD was equivalent to 23 Birr (Ethiopian

2 Woreda (Plural: woredas) is an administrative boundary that is equivalent to
a district. Tabia (plural: tabias) is the smallest administrative unit in the study

T. Tadesse et al.

Forest Policy and Economics 138 (2022) 102721


remains in place (the period of which is correlated to deforestation).
Since there was no reliable data related to the length of period the forest
continues to provide local communities with products, we started with
estimates experts and local elders provided and organized a pilot survey
to ask households to rate the following 5–point Likert scale question:
given status quo, the forest will not be able to provide you and your family
with the products that you need after 20 years. 80% of the respondents
tended to agree with this statement. Given this high score, it was deemed
reasonable that an even higher proportion of people would believe that
the supply of forest products would be even less certain after a longer
period of time. Through additional consultation with forest experts and
local forest management bodies, it was decided that with current forest
management and use realities, the forest would be able to provide
meaningful resources (products) to local communities only for the
coming 20 years (i.e., status quo option value of 20 years). The other two
options were set at twice (40 years of effective supply of forest products)
and thrice (60 years of effective supply of forest products) of the length
of the status quo period.

The soil and water conservation attribute was measured by the percent
of soil and water conserved because of the presence of the forest. One of
the limitations regarding the selection of relevant attributes was the lack
of data on values that could be used to describe the current situation as
well as the future direction and magnitude of attributes such as soil and
water conservation. However, Tamene and Vlek (2008) estimated that
annual mean net soil loss around the study areas was 19 tons per hectare,
with water erosion accounting for the majority of the soil loss. We used
this mean net annual soil loss to help households understand the scale of
soil loss if it were to occur in their own plot (which averages to 0.53 ha).
Given this magnitude of soil loss as the current state of soil loss (status
quo indicator), households were then presented with two additional
options to choose from– 50% and 75% reductions in soil loss. In this
framework, the soil and water conservation attribute in the status quo
was framed as a ‘no change’, and the attribute levels in the other two
alternatives (option 1 and option 2) were defined by percentage re-
ductions in the amount of soil and water that could be lost. As not all
local communities in the study area face the exact same state of soil and
water erosion, this approach also carries the extra benefit of doing away
with the need to explicitly address local heterogeneity in environmental
conditions (Glenk, 2008; Hynes et al., 2011).

The protection from heat attribute was measured by the degree of
perceived heat (temperature) reduction by the absorptive capacity of the
Desa’a forest. Desa’a forest is located in the escarpment of north-east
Ethiopia in Tigray bordering the Afar Regional State where the Ertaele
molten volcano is active all year round. It is known that forest ecosys-
tems play a great role in creating suitable local micro-climate as well as
global macro-climatic conditions. Just because of its location, the Desa’a
forest serves a as buffer for heat waves from the Ertaele molten volcano.
Especially, farmers and local forest management bodies emphasized the
role the forest is playing in ensuring the continuity of habitable micro-
climate, and duly reckon the importance and want the forest to
continue its buffering role. Given this, the status quo option for this
attribute was framed as ‘no change’, reflecting existing heat wave con-
ditions. The other options involved expert-based estimated levels of heat
reductions of 20% and 50%. Table 1 summarizes the nature of these

The four attributes (including the monetary attribute) were used to
construct a full factorial design containing 1203 different combinations
of choice sets. However, given the low literacy level of sample household
heads in the local communities (where 65% of heads had zero years of
schooling), it was unrealistic to assume that respondents would
comprehend such a high number of choices and make the required
tradeoffs. As a result, D-efficient design (using the ‘dcreate’ module in

Stata 14) was used to create 16 choice sets using the D-efficiency cri-
terion. Prior values obtained from the pre-test were used to improve the
D-efficient design for the main experiment (Hynes et al., 2011; Johnston
et al., 2017). We were worried that the large number of choice sets
obtained from the D-efficient design (16 choice sets) would create sig-
nificant cognitive burden on respondents with low numerical literacy.
As a result, the 16 choice sets were randomly categorized into four
blocks of each four4 choice sets to which sample households were
randomly assigned. A sample choice set is given in table A.2 in the

3.2. Empirical approach for estimating preference for forest management

Random utility theory provides the theoretical basis for choice ex-
periments. The theory holds that the choice individuals make depends
on unobserved utility derived from their preferences for alternative
outcomes based on observable and unexplained random components
(Hanley et al., 1998; Hanley et al., 2001). To illustrate the basic utility
framework, consider a household facing choice among alternative forest
management mechanisms from a set C, which includes all the possible
forest ecosystem management alternatives. Following Lancaster’s model
of consumer choice (Lancaster, 1966), the utility function



household i and forest management alternative mechanism j can be
presented as:

Uij = V
Fj; SI

+ ε

Fj; Si


This utility function is defined by the sum of observable, V(.) and
random, ε(.) components. These components are in turn expressed as
functions of the alternative forest management mechanisms (Fj) and the
socioeconomic characteristics of the household (Si). The presence of the
random component permits us to make probabilistic statements about
households’ preferences for alternative forest management mechanisms.
Choices made between alternative forest management mechanisms will
be a function of the probability that utility associated with a particular
alternative is higher than other alternatives. In this case, household i
chooses alternative forest management mechanism j over some other
alternative mechanism k if and only if Uij > Uikfor all k ∕= j. This leads to
the expression for the probability of choice:

pij = p
Vij + εij > Vik + εik

∀ K ∈ C (2)

where k is any alternative forest management mechanism in a given
choice set (i.e. choice set C). For estimating preferences, different as-
sumptions about the distribution of the random component lead to
different models. The model in Eq. 1 can be estimated using a condi-
tional logit (CL) model on the condition that the random component of
the utility function is distributed independently and identically (IID)
with a Weibull distribution and choices are consistent with the inde-
pendence of irrelevant alternatives (IIA)5 property (Train, 2003). Often
however, such assumptions are unrealistic. This is where the mixed logit
model becomes useful as it does not exhibit the strong assumption about
the IIA property. Moreover, mixed logit model accounts for preference
heterogeneity and explicitly accounts for correlations in unobserved
utility over repeated choices by respondents (Hanley et al., 2001; Train,
2003). Based on the framework of the mixed logit model, we estimate

3 The full factorial design is computed by an(an − 1)/2 where ais number of
levels and n is the number of attributes.

4 In the pilot test, the fractional factorial design was calibrated to produce 8
choice sets, which were offered to selected households. However, we observed
that it was practically very difficult for the low-literate respondents to make
meaningful tradeoffs among 8 different choice sets. As a result, through
consultation with farmers, the number of choice sets was reduced to 4 for which
farmers were able to comprehend and make meaningful tradeoffs.

5 The IIA property states that the relative probabilities of two alternatives
being chosen are unaffected by the introduction or removal of other alternatives
(Greene, 2003).

T. Tadesse et al.

Forest Policy and Economics 138 (2022) 102721


the following empirical specification of the conditional indirect utility

Vij = β + θ

F + α′ (F × S) + ui (3)

where β is the vector of alternative specific constants (ASCs) which
capture the effect on utility of any attribute not included in the choice
specific attributes (Hanley et al., 1998). Using Eq. 3, we first estimated
the base mixed logit model without interactions (Table 3). Based on this
base specification, marginal WTP for each attribute was estimated
(Table 4). Then, we estimated the full mixed logit model with in-
teractions (Table A.4 in the appendix). In these specifications, while the
attributes were treated as random, the monetary attribute and ASCs
were assumed to be fixed. The parameters for the attributes were
assumed to be normally distributed as individuals may have higher
(positive) or lower (negative) preferences toward the attributes (Carls-
son et al., 2003; Hensher et al., 2005). The vector of coefficients that are
attached to the vector of the forest management attributes (F)are
denoted by θwhile α represents the vector of coefficients that correspond
to the interaction terms of forest management attributes and socioeco-
nomic characteristics (S)(with ubeing the random component).
Following Hanley et al. (2001), we estimated the marginal WTP using
the base specification (without interactions) in Eq. (3)6 as follows:

MWTP = −




whereβa is the coefficient of any one of the forest management attributes
and βc is the coefficient of the payment attribute. This implicit price is
useful to demonstrate the tradeoffs among attributes. Comparison of the
implicit prices reflected by household preferences on attribute–level
combination choices helps understand the relative importance and value
households place on attributes.

Based on the base specification from Eq. 3, we estimated two model
specifications. We estimated utility models based on Eq. 3 for ‘higher-
level’ and ‘most sustainable’ level of forest management attributes. The
‘higher-level’ specification is related to the utility model that estimates
preferences for higher level attributes (i.e., preference for any level that
changes the status quo management practice). This is the standard
mixed logit model whose results are presented in the first (or upper)
panels of Tables 3, 4 and 5. In this case, attributes’ levels were coded as
continuous based on the operational definition presented in Table 1.
Analysis in this case sheds light on preference for more sustainable
attribute levels as compared to the status quo. On the other hand, the
‘most sustainable’ specification estimates preferences for the most sus-
tainable forest management practice (i.e., for the highest level of each
attribute). Mixed logit model results from this specification help explore
preferences for the ‘most sustainable’ level of each attribute vis-à-vis not
only the status quo but also the ‘less sustainable’ level. For the ‘most
sustainable’ specification, we follow Hensher et al. (2015) and used
hybrid coding. According to Hensher et al. (2015), hybrid coding (0, 1
or − 1) requires the creation of new variables based on the number of
alternatives (options) minus one. For instance, if there are two alter-
natives other than the status quo (as is the case in our study), one of the
alternatives (options) is set as a base level to prevent perfect con-
founding with the overall mean (Hensher et al., 2015). In this sense, the
hybrid coding works by setting the status quo alternative (option 3)
dummy-coded as zero (0) while the lowest level (‘less sustainable’ level)
of each attribute is effects-coded as − 1 and the highest level of each
attribute is dummy-coded as one (1).

4. Results and discussion

4.1. Benefits from forest

The Desa’a forest provides different products and resources that
constitute some of the most important livelihood sources. The livelihood
sources from the Desa’a forest come in different forms (Table 2). Among
the most important forest products (resources) that households obtain
from the Desa’a forest include water (80%), firewood (62.1%), fodder
for livestock (53.8%) and honey (46.3%). Many households also re-
ported that they obtain materials to construct farm implements (29.2%)
and building materials for human and livestock housing (35.8%). Still,
the forest provides some food items (mainly wild fruits) for about28.3%
of the households. The total value of benefits households obtained from
the forest makes up only 15.2%7 (18.8% for female-headed and 12.3%
for male-headed households) of the total household income. However,
this share of income households obtained appears to be lower compared
to estimates reported in some studies such as 54% (Kazungu et al., 2020)
and 40% (Nguyen et al., 2020). The difference could be due to variations
in accounting different types of forest benefits. For instance, Kazungu
et al. (2020) add that forest income shares represent only 16.7% of
household income if we consider only unprocessed forest products such
as firewood, forest foods, and structures and fibers, etc. Else in the
literature, estimates of the share of forest income vary depending on
differences in accessibility and type of forest income considered (such as
processed an unprocessed). Studies from different parts of the globe
report different estimates. For instance, Pouliot and Treue (2013) re-
ported lower share of forest income of 5.8% in Ghana and 6.2% in
Burkina Faso (for poorest households). Belcher et al. (2015) on the other
hand reported share of forest income varying between 12 and 42% in
Indian villages. Further, Dokken and Angelsen (2015) estimated share of
forest income of 13% in Tanzania. In all the studies however, fuel wood
constituted the most important source of forest income, which is
consistent with our result.

Mean-difference tests indicate that the value of total benefits female-
headed households obtained on average was higher than that of male-
headed households. This may be due to the significant difference in
access to forest resources among female and male-headed households.
Likewise, mean-difference tests in the average number of access to the
forest per month (accessibility) indicate female-headed households on
average access the forest more frequently (6.2 time per month) than
male-headed households (with mean value of 4.2 times). This is
consistent with targeting the poor and vulnerable social groups such as
female-headed households to have higher access and accessibility to
forest resources. While targeting the poor for higher access works in

Table 2
Types of forest benefits (products and services).

Products or services from Desa’a forest Percent

Yes No

Fuelwood (firewood & charcoal) 62.1 37.9
Farm implements 29.2 70.8
Building materials 35.8 64.2
Herbs for medicine 41.3 58.8
Food/ wild fruits 28.3 71.7
Livestock fodder 53.8 46.3
Honey (honey bee farming) 46.3 53.8
Water 72.5 27.5
Leaf litter 80.4 19.6

Source: data from own survey.

6 Marginal WTP estimated using Eq. 4 was based on Eq. 3 without the
interaction terms.

7 This share of income is mainly income from unprocessed forest products.
Income from processed forest products (such as charcoal) was not collected
largely due to data unavailability and also it is legally prohibited to cut down
trees and manufacture charcoal.

T. Tadesse et al.

Forest Policy and Economics 138 (2022) 102721


some cases, other but related policies may have unintended conse-
quences. For instance, Pouliot and Treue (2013) found that conservation
policies in Western Africa tended to deprive poor people of access to
forest resources. Nevertheless, meeting economic needs of the poor are
increasingly becoming part of participatory forest management pro-
grams through targeting vulnerable social groups (Kazungu et al., 2020;
Tadesse et al., 2021).

4.2. Preference for sustainable forest management attributes

Most of the sample households (91.3%) opted for the alternative
(that change the status quo practice) forest management attributes,
suggesting positive preferences for forest management mechanisms
involving longer access to forest products, and higher reduction in soil
and water loss and heat. Only 8.7% of the sample households reported
that they prefer existing conditions and would not prefer any changes in
the management of the forest. Before estimating the mixed logit model,
Hausman–McFadden test (Hausman and McFadden, 1984) for the in-
dependence of irrelevant alternatives (IIA) under conditional logit was
performed (table A.3 in the appendix). From this test, we find results
that strongly (at 1% level) reject the null hypothesis of no systematic

differences in coefficients in response to removal of options. This in turn
suggests that the conditional logit model produces biased results. As a
result, mixed logit model results were used to explore preferences and
WTP for alternative forest management mechanisms.

One of the alternative specific constants (Option 2_ASC)was found to
be statistically significant, suggesting that there were other factors other
than the attributes that negatively affected the preference for forest
management mechanisms (Morrison et al., 2002). The negative coeffi-
cient of Option 2_ASC indicates lesser preference (disutility) for some
forest management attributes. It may also point to a status-quo bias from
the part of respondents. Moreover, the higher (in absolute terms)
negative coefficient of Option 2_ASC for sample households from Afar
Regional State suggests even lesser preference for the forest manage-
ment attributes, compared to households from Tigray Regional State.

The coefficients of two of the important attributes–supply of forest
products and soil and water conservation–were found to be statistically
significant. As the result indicates, there is positive preference for higher
levels of the supply of forest products and soil and water conservation
(Table 3, upper panel). In fact, farmers prefer to have the most sus-
tainable options as can be seen from results of the ‘most sustainable’
model specification (Table 3, lower panel). This result suggests,
compared to the status quo and ‘less sustainable’ levels, farmers prefer
the ‘most sustainable’8 levels, both for forest products (60 more years of
supply of forest products) and soil and water conservation (75%
reduction in soil and water loss). The positive coefficient of supply of
forest products suggests that households are more likely to choose al-
ternatives that ensure supply of forest products for longer periods (i.e.,
higher values of the supply of forest products attribute). In this regard,
Bamwesigye et al. (2020) and Tadesse et al. (2021) argue that the higher
value forest communities’ place on forest-based benefits (i.e., economic
needs) exert significant influence on the positive preference for such
forest management attributes. Similarly, our results indicate that
households are more likely to choose forest management attributes that
bring higher reduction in soil and water loss, suggesting households’
commitment to forest management mechanisms that also play crucial
ecological functions (Giergiczny et al., 2015; Bamwesigye et al., 2020).
Regional differences were observed in preference for the key attributes.
The results in Table 3 in this regard suggest households in Tigray tend to
have positive preference for these forest management attributes while
their utility decreases with increasing payment for forest management.
This is indicated by the positive coefficients of forest products and soil
and water conservation (SWC) attributes (higher level options). This is
further supported by the positive preference for the ‘most sustainable’
levels of both attributes, i.e., positive (higher) preference for60 more
years of supply of products and 75% reduction in soil & water loss (most
sustainable levels).As an explanation to this difference in preferences for
alternative forest management attributes, Tadesse et al. (2021) point to
the strong commitment of households in rural Tigray to soil and water
conservation (through community mobilization)activities. The strong
community mobilization programs in Tigray to rehabilitate degraded
shrub land and forestland could in part explain the strong positive
preference for the supply of forest products and SWC attributes among
households in Tigray. In this regard, Tadesse et al. (2021) emphasize the
importance of labor contribution in soil and water conservation as part
of sustainable forest management programs through wide-ranging
community mobilization to rehabilitate bush land, shrub land and
forestland in Tigray.

The payment attribute was estimated to be statistically significant

Table 3
Mixed logit model estimates.

Pooled sample Tigray Afar

Coefficients Coefficients Coefficients

Higher levels of forest management options
Choice 1_ASC − 0.099 (0.273) − 0.134 (0.163) − 0.152 (0.575)
Choice 2_ASC − 5.017 (0.000)

− 4.731(0.000)

− 8.447 (0.000)

Payment − 0.004 (0.035)

− 0.004 (0.014)

− 0.004 (0.404)

Soil & water conservation − 0.007 (0.141) 0.01 (0.029)** − 0.016 (0.365)
Supply of forest products 0.012 (0.017)

0.016 (0.002)

− 0.023 (0.153)

Heat protection − 0.005 (0.147) − 0.003 (0.382) − 0.001 (0.987)
Standard deviations
Soil & water conservation 0.043 (0.000)

0.039 (0.000)

− 0.072 (0.000)

Supply of forest products 0.021 (0.028)

0.014 (0.406) 0.006 (0.945)

Heat protection − 0.019 (0.001)

0.021 (0.000)

0.001 (0.986)

Number of observations 2880 2520 360
Log likelihood − 691.5 –591.68 − 87.804

Most sustainable forest management options
Choice 1_ASC − 0.083 (0.314) − 0.111 (0.205) 0.044 (0.863)
Choice 2_ASC − 2.56 (0.000)

− 2.93 (0.000)

− 1.90 (0.032)

Payment − 0.003 (0.038)

− 0.004 (0.025)

− 0.002 (0.736)

75% reduction in soil &
water loss

0.096 (0.029)

0.129 (0.005)

− 0.099 (0.556)

60 more years of supply of

0.111 (0.008)

0.144 (0.001)

− 0.151 (0.367)

50% reduction in heat

− 0.049 (0.222) − 0.025 (0.553) − 0.289 (0.026)

Standard deviations
75% reduction in soil &

water loss
0.304 (0.000)

0.278 (0.002)

0.492 (0.099)*

60 more years of supply of

0.128 (0.474) 0.063 (0.834) 0.332 (0.429)

50% reduction in heat

0.181 (0.087)* 0.188 (0.078)* − 0.005 (0.994)

Number of observations 2880 2520 360
Log likelihood − 736.9 –619.9 − 103.1

Note: ***, **, and * denote statistical significance at 1%, 5% and 10% levels,
Values presented in parentheses are P–values.
Source: Own estimates.

8 Two different mixed logit models were estimated. First, the standard mixed
logit model was estimated showing preference for higher attribute levels using
codes of the attributes as shown in Table 1. Then, a second mixed logit model
was estimated to explore preference for most sustainable levels not only vis-
à-vis the status quo but also with the ‘less sustainable’ level. We do this by using
hybrid coding for attribute levels following Hensher et al. (2015).

T. Tadesse et al.

Forest Policy and Economics 138 (2022) 102721


with a negative sign, consistent with expectation. The indication is that
with increasing payment, utility from households’ preferences for
higher level forest management alternatives decreases. In line with this
result, Giergiczny et al. (2015) find that respondent disutility increases
with forest management schemes that require greater intensity. This
greater intensity comes from the commitment to be willing to pay for
more(or, most) sustainable forest management programs that entail
higher costs through accessibility restrictions or reduced forest benefits
(Giergiczny et al., 2015; Tadesse et al., 2021). The significant changes in
the attributes’ effect on utility coupled with statistically significant
standard deviations indicate the presence of unobserved preference
heterogeneity, which mixed logit model helps account for (Birol et al.,

The rate at which households were willing to trade off payment for
changes in any of the forest management attributes was calculated. This
rate is calculated as the ratio of the coefficient for the attribute of in-
terest and the payment attribute, and reflects households’ WTP for the
realization of higher level of an attribute (based on Eq. 4). For the pooled
sample, marginal WTP (MWTP) for only the supply of forest products
attribute was significant in both model specifications (upper panel for
‘higher levels’ and lower panel for ‘most sustainable level’ in Table 4),
with respectively MWTP of 3.32 Birr (0.14 USD/ month) and 35.3 Birr
(1.53 USD/month). Despite the coefficients in both model specifications
are weakly significant, they nonetheless indicate the value farmers place
on the economic benefits (products and services) they get from the
forest. Interestingly, households prefer and attach even higher value on
forest management attributes that ensure sustainable forest manage-
ment as indicated by the much higher MWTP for the ‘most sustainable’
forest management level (lower panel in Table 4). As the result shows,
farmers would be willing to pay much higher (35.3 Birr per month) to
extend the forest’s provisioning capability by 60 more years, suggesting
strong preference for most sustainable forest management mechanisms.
More tellingly, results indicate that households from Tigray not only
value forest management mechanisms that maximize provisioning ser-
vices but also equally appreciate the ecological functions of the forest by
willing to pay for the most sustainable soil and water conservation level.
The estimates indicate households in Tigray would be willing to pay
higher (39.9 Birr and 35.9 Birr) for the forest management attributes
that maximize not only supply of forest products but also reduction in
soil and water loss. The negative MWTP estimates indicate that house-
holds are reluctant with higher payments despite their positive prefer-
ence for (higher levels of)the forest management attributes. Marginal
WTP estimates for forest management practices in other studies vary,
ranging from USD 1.25 per month (Bamwesigye et al., 2020) to USD
6.75 per month (Okumu and Muchapondwa, 2017). Similar studies from
the developed world such as Chaikaew et al. (2017)report even much

lower MWTP of USD 0.17 per month for mixed forest-watershed man-
agement, suggesting people’s competing interests for their discretionary
spending in relation to ecosystem services. These differences could be
attributed to differences in accessibility (Tadesse et al., 2021), benefits,
existing economic and ecological needs of communities (Belcher et al.,
2015; Bamwesigye et al., 2020).

In order to assess the sources of heterogeneity for preferences across
households, we also estimated mixed logit models with interaction
terms. The results of these models are presented in Table 5 (interaction
with payment) and table A.4 in the appendix (interaction with forest
management attributes). Table 5 contains important results related to
sources of heterogeneity in preference for (higher levels of) forest
management attributes. The results indicate positive preference for
higher levels and most sustainable level, suggesting that households
highly value both the economic contributions and ecological functions
of the forest. Moreover, age and access to natural resource management
extension were found to lead to heterogeneous preferences in payment
for the forest management attributes. Age has negative effect on WTP
due to perhaps the short planning horizon people face when they make
decisions related to choosing the most sustainable forest management
level. On the other hand, access to natural resource extension was
positively associated with WTP, which may suggest the role such
extension services play in positively influencing households’ preference
for forest management mechanisms that contribute to sustainable forest

In the model of interaction with attributes (table A.4), the two
alternative specific constants were found to be statistically significant,
indicating the presence of factors other than the attributes, which
affected tradeoffs in preferences for the attributes. The two important
attributes– supply of forest products and soil and water conservation
also carried significant coefficients, implying households’ positive
preferences for such attributes that contribute to sustainable forest
management (especially for Tigray9). While households would be
willing to pay due to the forest’s importance in providing long-term
forest products and conserving soil and water, their utility declines as
the amount of payment increases (table A.4). Moreover, the coefficients
of the interaction terms of supply of forest products and soil and water
conservation with gender and access to extension service were found to
be statistically significant and with the a priori expected signs. This
means female-headed households that are targeted for access to forest
resources were more likely to make tradeoffs for management options

Table 4
Marginal willingness to pay for forest management attributes.

Attributes and levels Pooled sample Tigray Afar

Birr USD Birr USD Birr USD

Higher levels of forest management options
Soil & water conservation − 1.862 (0.220) − 0.081 − 2.333 (0.090)* − 0.102 − 3.495 (0.563) − 0.153
Supply of forest products − 3.318

− 0.145 − 3.691 (0.038)** − 0.161 − 5.128 (0.414) − 0.224

Heat protection 1.274

0.055 0.671 (0.371) 0.029 − 3.765 (0.461) 0.165

Most sustainable forest management options
75% reduction in soil & water loss − 30.7 (0.124) 1.33 − 35.9 (0.067)* 1.56 − 58.9 (0.781) 2.56
60 more years of supply of products − 35.3 (0.084)* 1.53 − 39.9 (0.048)** 1.73 − 89.6 (0.740) 3.90
50% reduction in heat wave 15.7 (0.246) 0.68 7.03 (0.543) 0.31 − 171.2 (0.739) 7.44

Note: ** and * denote statistical significance at 5% and 10% levels, respectively.
Values presented in parentheses are P–values.
During data collection, 1USD was circa 23 Ethiopian Birr (Birr).
Source: Own estimates.

9 A separate model for Afar Regional State could not be estimated due to the
lack of sufficient degrees of freedom (emanating from limited number of ob-
servations) leading to non-convergence of the model.

T. Tadesse et al.

Forest Policy and Economics 138 (2022) 102721


that lead to more sustainable provision of forest products(than their
male-headed counterparts).Moreover, the result related to access to
natural resource management extension vis-à-vis the two important at-
tributes underlines the importance of providing natural resources
(management)extension services. As the results show, households with
access to natural resource management extension were more likely to
prefer alternative forest management mechanisms that involve extended
period of supply of forest products and significant reduction in soil and
water loss.

5. Conclusion

In this study, we considered a forestland where the governance
structure is such that the state owns the forest and enforces selective

access to rural households. In the literature, it is documented that forest-
based benefits make up an important part of household income of forest
communities. Similarly, more than 95% of the sample households
consider the forestland as an important livelihood source. Interestingly,
local communities place high stakes on the forest despite forest income
makes up less than 20% of total income (but is still the third most
important source of income). Such high stakes mean that forest man-
agement plans should integrate local communities’ economic interests
and needs for effective and sustainable forest management. In order to
understand how tradeoffs between economic needs and ecological
functions influence preference for alternative forest management attri-
butes, we considered a demand-relevant economic attribute–benefit
from forest products and two policy and demand–relevant ecological
attributes–soil and water conservation and heat absorption to examine
preferences for competing forest management alternatives.

The analysis on forest benefits on the one hand and the construct
attribute from forest benefit (which is use of supply of forest products) in
the choice experiment shows that forest communities greatly value the
benefits they obtain from forest. More importantly, the results indicate
positive preferences for the forest products and soil and water conser-
vation attributes, which represent alternative forest management
mechanisms that account for both economic needs of local communities
and ecological functions needed to ensure forest sustainability. This
preference for higher levels of the supply of forest products and soil and
water conservation further suggests that households are more likely to
choose management alternatives that continue to sustainably provide
economic (forest products) and ecological (soil and water loss reduction)

We find heterogeneous preferences across gender for the forest
management mechanisms. Results suggest that male-headed households
with less access to forest products are willing to tradeoff for less sus-
tainable forest management mechanisms. Conversely, female-headed
households that are targeted for more access to forest resources were
more likely to prefer management mechanisms that ensure continuous
supply of forest products. Moreover, the model with interaction shows
the importance of natural resource extension programs in creating
awareness and thereby inducing positive preferences for forest man-
agement mechanisms that not only ensure supply of forest products for
long periods but also significantly reduce soil and water loss. Overall,
access to forest resources is an important policy aspect in two ways. One,
it creates livelihood opportunities for targeted social groups and help
generate significant benefit. Two, it induces among targeted households
willingness to commit to forest management alternatives that not only
ensure longer period of the supply of forest products but also higher
reduction in soil and water loss. As the results imply, this preference for
higher level forest management attributes emanates both from re-
alizations of economic benefits and ecological services from the forest.
Yet, the presence of heterogeneity in preferences means that households
would make divergent tradeoffs to balance their need for economic
benefits and ecological sustainability of forests, suggesting the impor-
tance of incorporating such contrasting preferences (needs) in forest
management plans. Finally, we want to note that the presence of pref-
erence heterogeneity may inflate WTP estimates (Morrison, 2002), and
our results instead reflect regional differences, which may not be
generalized to the national level.

CRediT authorship contribution statement

Tewodros Tadesse: Conceptualization, Funding acquisition, Data
curation, Methodology, Formal analysis, Writing – original draft, Soft-
ware. Gebreegziabher Teklay: Conceptualization, Data curation,
Formal analysis, Writing – original draft. Dawit W. Mulatu: Method-
ology, Writing – review & editing, Software. Meley Mekonen Ran-
nestad: Conceptualization, Funding acquisition, Methodology, Writing
– review & editing. Tigabu Molla Meresa: Methodology, Writing – re-
view & editing. Dawit Woldelibanos: Conceptualization, Formal

Table 5
Sources of heterogeneity to preference for forest management attributes.

Versus higher level forest
management options

Versus most

Coefficient Coefficient

Choice 1_ASC − 0.093 (0.089) − 0.099

Choice 2_ASC − 5.36 (0.762)*** − 3.37

Soil and water conservation 0.008 (0.004)**
Supply of forest products 0.012 (0.005)**
Reduction in heat wave − 0.005 (0.003)
75% reduction in soil and

water loss

60 more years of supply of
forest products


50% reduction in heat wave 0.228

Payment: age –0.0001 (0.0001) − 0.0002

Payment: gender 0.003 (0.004) − 0.005

Payment: education 0.0002 (0.0007) − 0.0001

Payment: marital status –0.002 (0.007) 0.001

Payment: family size 0.0007 (0.0007) 0.0002

Payment: land size –0.001 (0.005) 0.0005

Payment: livestock wealth –0.0004 (0.0004) 0.0002

Payment: access to extension 0.014 (0.004)*** 0.016

Payment: income –4.9 × 10− 8 (1.6 × 10− 7) − 1.5 × 10− 8

(8.1 × 10− 8)

Standard deviations
Soil and water conservation − 0.036 (0.006)***
Supply of forest products 0.024 (0.013)*
Reduction in heat wave − 0.019 (0.006)***
75% reduction in soil and

water loss

60 more years of supply of
forest products


50% reduction in heat wave 0.228

Number of respondents 240 240
Number of observations 2880 2880
Log–likelihood − 680.4 − 708.6
LR χ2 69.4 13.3
Prob. > χ2 0.000 0.004

Values in parentheses are standard errors.
Statistical significance: *** for 1%, ** for 5% and * for 10% levels.
Source: Own estimates.

T. Tadesse et al.

Forest Policy and Economics 138 (2022) 102721


analysis, Writing – review & editing.

Declaration of Competing Interest



The authors would like to thank enumerators and facilitators in
different villages who worked hard to conduct the choice experiments.
We also would like to thank farmers who sat for long periods to provide
data. This work was supported by the “Steps toward Sustainable Forest
Management with Local Communities in Tigray, northern Ethiopia
[with grant number: ETH-13-0018]” project financed by the Norwegian
Support for Higher Education and Development (NORHED/NORAD),
Norway. We thank NORAD for the funding. Any remaining errors are
solely the responsibilities of authors.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.


Bamwesigye, D., Hlavackova, P., Sujova, A., Fialova, J., Kupec, P., 2020. Willingness to
pay for forest existence value and sustainability. Sustainability 12, 891. https://doi.

Belcher, B., Achdiawan, R., Dewi, S., 2015. Forest-based livelihoods strategies
conditioned by market remoteness and Forest proximity in Jharkhand, India. World
Dev. 66, 269–279.

Birol, E., Karousakis, K., Koundouri, P., 2006. Using a choice experiment to account for
preference heterogeneity in wetland attributes: the case of Cheimaditida wetland in
Greece. Ecol. Econ. 60, 145–156.

BoANR (Bureau of Agriculture and Natural Resources), 1997. Desa’a State Forest
Management Plan. Bureau of Agriculture and Natural Resource Development,
Mekelle, Tigray.

Carlsson, F., Frykblomb, P., Liljenstolpec, C., 2003. Valuing wetland attributes: an
application of choice experiments. Ecolog. Econ. 47, 95–103.

Chaikaew, P., Hodges, A.W., Grunwald, S., 2017. Estimating the value of ecosystem
services in a mixed-use watershed: a choice experiment approach. Ecosystem
Services 23, 228–237.

Dokken, T., Angelsen, A., 2015. Forest reliance across poverty groups in Tanzania. Ecol.
Econ. 117, 203–211.

Gebreegziabher, Z., 1999. An Assessment of Human Impact, Evolutionary Pattern and
Options for Sustainable Management. UNESCO-MAB (Man and the Biosphere)
Programme, Mekelle, Ethiopia.

Giday, K., 2013. Management Interventions to Assist Restoration of Degraded Dry
Afromontane Forests in North Ethiopia. PhD Thesis. KU Leuven, Heverlee, Belgium.

Giergiczny, M., Czajkowski, M., Żylicz, T., Angelstam, A., 2015. Choice experiment
assessment of public preferences for forest structural attributes. Ecol. Econ. 119,

Glenk, K., 2008. Unveiling Regional Preferences for Biological Diversity, Economic
Valuation of Biological Diversity Exploring Non-market Perspectives in the Vicinity
of the Lore-Lindu National Park in Indonesia’s Central Sulawesi Region. PhD Thesis.
University of Gottingen, Gottingen, Germany.

Gordillo, F., Elsasser, P., Günter, S., 2019. Willingness to pay for forest conservation in
Ecuador: results from a nationwide contingent valuation survey in a combined
“referendum” – “consequential open-ended” design. Forest Policy Econ. 105, 28–39.

Greene, W.H., 2003. Econometric Analysis, fifth ed. Prentice Hall, New York, US.
Haglund, E., Ndjeunga, J., Snook, L., Pasternak, D., 2011. Dry land tree management for

improved household livelihoods: farmer managed natural regeneration in Niger.
J. Environ. Manag. 92 (7), 1696–1705.

Hanley, N., Wright, R., Adamowicz, W., 1998. Using choice experiments to value the
environment. Environ. Resour. Econ. 11 (3), 413–428.

Hanley, N., Mourato, S., Wright, R.E., 2001. Choice modeling approaches: a superior
alternative for environmental valuation? J. Econ. Surv. 15 (3), 435–462. https://doi.

Hausman, J.A., McFadden, D., 1984. Specification tests for the multinomial logit model.
Econometrica 52, 1219–1240.

Hensher, D., Ho, C., Mulley, C., 2015. Identifying preferences for public transport
investments under a constrained budget. Transport. Res. A Policy Pract. 72, 27–46.

Hensher, D., Shore, N., Train, K., 2005. Households’ willingness to pay for water service
attributes. Environ. Resour. Econ. 32, 509–531.

Hynes, S., Campbell, D., Howley, P., 2011. A holistic vs. an attribute-based approach to
Agri-environmental policy valuation: do welfare estimates differ? J. Agric. Econ. 62,

Johnston, R.J., Boyle, K.J., Adamowicz, W., Bennett, J., Brouwer, R., Cameron, T.A.,
Hanemann, W.M., Hanley, N., Ryan, M., Scarpa, R., Tourangeau, R., Vossler, C.A.,
2017. Contemporary guidance for stated preference studies. J. Associat. Environ.
Res. Econ. 4, 319–405.

Kazungu, M., Zhunusova, E., Yang, A.L., Kabwe, G., Gumbod, D.J., Günter, S., 2020.
Forest use strategies and their determinants among rural households in the Miombo
woodlands of the Copperbelt Province, Zambia. Forest Policy Econ. 111, 102078

Krzanowski, W.J., 2007. Statistical Principles and Techniques in Scientific and Social
Investigations. Oxford University Press, New York, USA.

Lancaster, K.J., 1966. A new approach to consumer theory. The J. Polit. Econ. 74 (2),

Morrison, M., 2002. Aggregation biases in stated preference studies. Aust. Econ. Pap. 39
(2), 215–230.

Morrison, M., Bennett, J., Blamey, R., Louviere, J., 2002. Choice modeling and tests of
benefit transfer. Am. J. Agric. Econ. 84, 161–170.

Nguyen, H.T., Ariyawardana, A., Ratnasiri, S., 2020. Forest plantation owners’
willingness to pay for hybrid nursery stock: the case of Acacia hybrids in Central
Vietnam. Forest Policy Econ. 116, 102184

Okumu, B., Muchapondwa, E., 2017. Economic valuation of Forest ecosystem Services in
Kenya: implication for design of PES schemes and participatory Forest management.
In: ERSA working paper 693, economic research southern Africa (ERSA), Cape

Pouliot, M., Treue, T., 2013. Rural People’s reliance on forests and the non-Forest
environment in West Africa: evidence from Ghana and Burkina Faso. World Dev. 43,

Sebhatleab, M., 2012. Land Use Land Cover Change Detection and Deforestation
Susceptibility Analysis of Desa’a Forest. MSc Thesis. Bahir Dar University, Bahir Dar,

Tadesse, T., Berhane, T., Mulatu, D.W., Rannestad, M.M., 2021. Willingness to accept
compensation for Afromontane forest ecosystems conservation. Land Use Policy 105,

Takahashi, R., Todo, Y., 2012. Impact of community-based Forest management on Forest
protection: evidence from an aid-funded project in Ethiopia. Environ. Manag. 50,

Tamene, L., Vlek, P.L.G., 2008. Soil erosion studies in northern Ethiopia. In: Braimoh, A.
K., Vlek, P.L.G. (Eds.), Land Use and Soil Resources. Springer, Stockholm,
pp. 73–100.

Tesfay, A., 2008. MSc Thesis, Ethno Botanical Study of Desa’a Forest, North-Eastern
Escarpment of Ethiopia, with Emphasis on Use and Management of Forest Resources
by the Local People. Addis Ababa University, Addis Ababa, Ethiopia.

Tesfaye, Y., Roos, A., Campbell, B.M., Bohlin, F., 2011. Livelihood strategies and the role
of forest income in participatory-managed forests of Dodola area in the bale
highlands, southern Ethiopia. Forest Policy Econ. 13, 258–265.

Train, K.E., 2003. Discrete Choice Methods with Simulation. Cambridge University Press,
Cambridge, UK.

Weston, P., Hong, R., Kaboré, C., Kull, C., 2015. Farmer-managed natural regeneration
enhances rural livelihoods in dryland West Africa. Environ. Manag. 55, 1402–1417.

Woldie, B.A., Tadesse, S.A., 2019. Views and attitudes of local people towards
community versus state forest governance in Tehulederi District, South Wollo,
Ethiopia. Ecol. Process. 8, 4.

T. Tadesse et al.

  • Forest benefits and willingness to pay for sustainable forest management
    • 1 Introduction
    • 2 Desa’a state forest, study areas and sample data
    • 3 Preference for forest management mechanisms
      • 3.1 Attributes of forest management
      • 3.2 Empirical approach for estimating preference for forest management mechanisms
    • 4 Results and discussion
      • 4.1 Benefits from forest
      • 4.2 Preference for sustainable forest management attributes
    • 5 Conclusion
    • CRediT authorship contribution statement
    • Declaration of Competing Interest
    • Acknowledgments
    • Appendix A Supplementary data
    • References

Wk 3 Discussion (Due in 1 day) Urgent/Forest commons, traditional community.pdf

Contents lists available at ScienceDirect

Forest Policy and Economics

journal homepage:

Forest commons, traditional community ownership and ecological
consequences: Insights from Spain
Sara Guadilla-Sáeza,⁎, Manuel Pardo-de-Santayanab,c, Victoria Reyes-Garcíaa,d
a Institut de Ciència i Tecnologia Ambientals, Universitat Autònoma de Barcelona, Barcelona, Spain
b Departamento de Biología, Universidad Autónoma de Madrid, Madrid, Spain
c Centro de Investigación en Biodiversidad y Cambio Global (CIBC-UAM), Madrid, Spain
d Institució Catalana de Recerca i Estudis Avançats (ICREA), Barcelona, Spain


Biodiversity conservation
Community forests
Forest history
Historical review


With a terrestrial surface increasingly dominated by human activities, conservation scholars nowadays seek to
reconcile extractive land uses, such as low-intensity agriculture, forestry or agroforestry, with biodiversity
conservation. This approach has been widely adopted by the international forestry community, which advocates
for implementing management strategies both favourable to forest biodiversity and economically profitable.
Along these lines, considerable attention is being given to the potential of traditional community management
for guaranteeing long-term forest-related resources conservation. Here, we extend this line of research to explore
whether certain local forms of use and governance of traditional community forests contribute to the con-
servation of biodiversity-rich habitats by examining the historical evolution of collective property regimes in
Spain. The establishment of a political and economic framework by the late eighteen century that did not
recognize community ownership as a form of property, largely disrupted the traditional management systems of
Spanish community forests, offering a unique context to analyse the ecological consequences of replacing tra-
ditional forms of forest use by other management systems. Results of our historical analysis illustrate that the
abolition of traditional uses had negative ecological consequences. In the short term, the privatization of forest
commons resulted in a decline of forest cover due to the cut of the woodlots acquired by the new owners, causing
flooding and soil erosion. In the long term, the limitation of traditional land uses due to State interventionism of
the forest commons not privatized seems to have favoured the decline of biodiversity-rich semi-natural habitats
dependent on human practices and the simplification of the rural landscape mosaic. These findings further
support the idea that traditional community management can provide useful insights for designing forest
management strategies reconciling economic benefits and forest biodiversity conservation. Additionally, the
historical evolution detailed in this manuscript helps to understand the multiple legacies of community-own-
ership forests recognized in Spanish present-day legal code.

1. Introduction

The term ‘biodiversity’ refers to the great variety of life forms and
the high diversity of interactions and processes that occur at the many
levels of biological organization (McElhinny et al., 2005). Despite its
relatively recent introduction in the scientific literature –Walter Rosen
was the first to use the word in 1986–, the term has been widely
adopted by the general public. Over the last decades, biodiversity
conservation has become a target for many international organizations,
as reflected in the signature of diverse international agreements for
biodiversity conservation (Brockington et al., 2008) and the creation of

the Intergovernmental Science-Policy Platform on Biodiversity and
Ecosystem Services (Schmeller and Bridgewater, 2016). Moreover, un-
derstanding what are the many threats to biodiversity (e.g., growing
human population, land use change, overuse of natural resources, en-
vironmental degradation) and finding strategies to mitigate them are
major sources of social concern (McElhinny et al., 2005).

To date, the backbone official instrument for biodiversity con-
servation has been the establishment and maintenance of a system of
legally recognized protected areas (Gaston et al., 2008; Gray et al.,
2016). However, the efficacy –in terms of biodiversity maintenance– of
the different protection categories is much debated (Hayes, 2006;
Received 31 January 2019; Received in revised form 10 January 2020; Accepted 24 January 2020

⁎ Corresponding author at: Institut de Ciència i Tecnologia Ambientals, Carrer de les Columnes s/n, Edificio Z, Campus UAB, 08193 Bellaterra (Cerdanyola del
Vallès), Barcelona, Spain.

E-mail addresses: [email protected] (S. Guadilla-Sáez), [email protected] (M. Pardo-de-Santayana), [email protected] (V. Reyes-García).

Forest Policy and Economics 112 (2020) 102107

Available online 28 January 2020
1389-9341/ © 2020 Elsevier B.V. All rights reserved.


Geldmann et al., 2013; Coetzee et al., 2014). Several studies suggest
that, overall, protected areas do help protect biodiversity, although
researchers have also noted that granting an area the ‘protected’ status
does not necessarily leads to biodiversity protection, as regulations
designed to protect biodiversity are not always effective or not suffi-
ciently enforced (Bruner et al., 2001; Dudley et al., 2004; Dudley et al.,
2005). While some conservationists argue that the solution to that lies
in ensuring compliance with regulations, others posit that efforts should
also be directed to maintain biodiversity outside the physical bound-
aries of protected areas. Following this logic, approaches oriented to
conserve biodiversity beyond the network of designated areas are in-
creasingly being adopted worldwide (Poiani et al., 2000; Mathur and
Sinha, 2008; Guadilla-Sáez et al., 2019).

Of particular interest are approaches aiming to reconcile extractive
land uses, such as agriculture or forestry, with biodiversity conserva-
tion, as these approaches could complement protected areas (Kremen,
2015). Indeed, this perspective seems to have been widely adopted by
the international forestry community, which increasingly advocates for
implementing management systems that combine economic benefits
and sustainable use of forest-related resources (Hernando et al., 2010).
Accordingly, several authors have analysed whether multipurpose
forest management strategies provide profitable forest-related uses
without actually compromising biodiversity conservation. Although it
has already been established that owners’ willingness to combine
nature-oriented and economic uses of forests is determinant for the
establishment of conservative management objectives (Nielsen et al.,
2017; Bergstén et al., 2018; Pynnönen et al., 2018; Weiss et al., 2018),
many of these studies do not include considerations on the effects of
land ownership type on biodiversity. Moreover, the few studies that
include proprietorship in the analyses mainly focus on comparing
public and private forest ownership forms, despite the literature stres-
sing the potential of traditional community ownership for guaranteeing
long-term forest resources conservation (Agrawal and Gibson, 1999;
Ostrom, 1999).

Traditional community forest, i.e., community forests or forest
commons, refers to woodlands collectively managed by local commu-
nities, considering community as a social group living in a small spatial
unit, with a homogeneous social structure, frequent interactions, and
shared interests and norms (Agrawal and Gibson, 1999). Understanding
how local communities manage their forests is important to global
biodiversity, as traditional community-based management comprise
nearly 18% of worldwide forest biodiversity-rich areas (Chhatre and
Agrawal, 2008). Here, we contribute to the analysis of the ecological
outcomes of forest commons ownership by examining the historical
evolution of collective property regimes in Spain, a country in Western
Europe with long history of forest community-ownership. During the
eighteenth and nineteenth centuries, the transition from the medieval
period to modernity in Europe brought the establishment of a political
and economic framework that introduced the concept of property law
to previous feudal land tenure regimes (Izquierdo, 2007). Because
traditional community ownership does not fit well in the private vs.
public property dichotomy, forests commons had to be classified as
private, public, semi-private or semi-public, depending on the regional
context (see Weiss et al., 2018). The heterogeneous land tenure change
occurring in Western Europe at the time offers a unique context to
analyse the woodland landscape dynamic resulting from the replace-
ment of forest commons’ traditional management systems by other land
ownership forms.

In particular, in this work we examine the different woodland
landscape dynamics resulting from the heterogeneous dismantling
process of forest commons occurred in Spain. The analysis should help
us to better understand the ecological consequences of replacing tra-
ditional community ownership by other forms of land ownership. In the
last section of the article, we provide a succinct description of the
multiple legacies of community-ownership forests recognized in the
contemporary Spanish legal code and the nature conservation interest

of these categories.

2. Historical evolution of forest commons in Spain

In Spain, there is a tradition of obtaining natural resources from
forest commons since the Germanic tribes’ invasion to the northwest
areas of the Iberian Peninsula in the fifth century (Aranda, 1996). From
the eighth to the fifteenth century, during the Christian Reconquest, the
communal regime was generalized to other parts of the peninsula, be-
coming the most habitual type of tenure regime of Spanish woodlands.
The importance of forest-related resources to Spanish peasant com-
munities resulted in formal –e.g., local ordinances– and informal –e.g.,
cultural practices– norms and rules to manage forest commons and
prevent them from overuse (Ezquerra and Gil, 2004). However, the
political and economic framework established by the late eighteenth
century in the country did not recognize community ownership
(Caballero, 2015), initiating a process of privatization and replacement
of forest commons by public management forms that negatively af-
fected forest biodiversity conservation in the long term. This process of
abolition of forest commons use, detailed below, is necessary to un-
derstand current management of Spanish forest commons (Montiel,

2.1. Initial records of the commons

Woodlands collectively managed in Spain date back, at least, to the
Middle Ages, when the territories that now constitute Spain experi-
enced a process of human resettlement and land use redistribution as-
sociated to the Christian Reconquest (Pardo and Gil, 2005; Montiel,
2007). From the eighth to the fifteenth century, medieval kings granted
land privileges to the Christian settlers who displaced Muslim popula-
tions from the newly gained territories. Such strategy created a special
type of common tenure, in which settlers, organized in village councils,
or concejos, collectively managed land concessions consisting of mea-
dows, woods, and streams. Similarly, the collective use and manage-
ment of natural resources and institutional formal structures created by
the commoners (i.e., users of the commons) spread across Western
Europe countryside from the eleventh and fifteenth centuries, becoming
the most salient feature of the economic life of the peasantry until the
nineteenth century (Laborda-Pemán and Moor, 2012).

As the management of such common lands and resources was not
officially regulated, with time, commoners developed a set of informal
rules adapted to local social-ecological conditions that became widely
accepted by community members (Mangas, 2013; Blanco, 2014). Such
rules were mainly orally transmitted and rarely written down until the
thirteenth century, when las Siete Partidas, or the Seven Divisions, a
legal code compiled by Alfonso X the Learned of Castile, referred to the
management of the commons stating that ‘mountains and pastures and
all places similar (…) belong to the common. Every man who is a re-
sident can make use of them’ (Law IX, Title XXVII, Third Partida) and
‘Cities and towns can own fields and other lands (…) although these are
property of all inhabitants, nevertheless, each one of them cannot se-
parately and individually make use of them’ (Law X, Title XXVII, Third
Partida) (Burns, 2012). From the thirteenth century onwards, the con-
cejos issued local ordinances to define the condition of resident and to
guide the long-term conservation of resources in common use by all
residents (Moreno, 1998; Arango, 2009). For example, in Eastern Spain,
local regulations dating from 1271 were issued to avoid the entrance of
non-resident livestock herds and the ploughing of forest commons
(Piqueras, 2002).

Later, in the fifteenth century, the Catholic Monarchs issued a de-
cree aiming to regulate the use of forest commons. According to Wing
(2015), the new decree shows crown’s intention in regulating forest use
at the same time that recognized local municipalities, handicraftsmen,
and shepherds needs of forest resources, as well as farmers’ interests to
extend their arable lands by ploughing local woodlands and pastures. In

S. Guadilla-Sáez, et al. Forest Policy and Economics 112 (2020) 102107


an attempt to preserve tree canopy layer while fulfilling local com-
munities’ livelihood needs, the royal decree promoted a sustainable use
of forest-related resources. For instance, one regulation prevented ve-
teran trees from excessive cutting, so while it allowed cutting branches
for firewood and carving, it regulated that this could only be done to
the extent that it did not impede new growth.

From the sixteenth century onwards, crown’s regulation of woods
management intensified, resulting on the promulgation of several royal
ordinances limiting local population’s use of woodlands. In 1518, for
instance, a royal decree on ‘Formation of new forest plantations and
ordinances to conserve old and new forests’ called for the designation of
local guards to defend against the cutting of trees (Wing, 2015). This
protective approach to forest management, which continued during the
seventeenth and eighteenth centuries (Ramos, 2007), is considered by
modern historians as an exercise of power in behalf of naval ship-
building interests against peasants’ use of forest-related resources to
meet the Spanish navy’s high requirements of timber (Rey, 2004; Wing,
2015). Indeed, the promulgation of 1748 Forest Ordinances, which
forbade cutting trees marked by and for the navy and authorized the
expropriation of lands suitable for forest nurseries, is considered the
first attempt of appropriation of forests by Spanish authorities
(Valbuena-Carabaña et al., 2010).

Thus, between the sixteenth and the eighteenth centuries, crown’s
forest regulations shifted from defending the use of forest commons to
increasingly restricting it, initiating the abolishment of traditional
communal forms of forest use in Spain.

2.2. Dismantling the commons

The nineteenth century largely resulted in the breakdown of the
traditional communal system in Spain. Following the liberal movement
spread through Europe at the end of the Old Regime, two major reforms
were enacted during that period, both with important effects on com-
munal lands. First, Spanish earliest written Constitution was issued in
1812. This new political and economic framework distinguished only
between public and private ownership forms. Despite the long historical
tradition of forest commons in Spain, the new legal code did not include
community ownership as a form of property and common lands were
considered as public property and concejos were replaced by larger and
hierarchically dependent municipalities, the Town Councils (Izquierdo,
2007; Serrano, 2014; Caballero, 2015). In other words, ancient ordi-
nances were not recognized by the new legal framework, which implied
that, from then forth the different laws governing the commons –and
specifically those related to woodlands– only recognized Town Councils
as valid intermediaries between villagers and public administrations.
Consequently, local residents, represented by the concejos, were not
authorized to profit from their woodlands except through Town
Councils (Serrano, 2014). However, given that the 1812 Constitution
was repealed only two years after its promulgation, common lands
survived this first political attempt of abolishing them.

The second reform, this one with real important effects on common
lands, was the disentailment policy (Desamortización) issued by the
Minister of Finances Mendizábal in 1836–1837 and which continued
until the twentieth century. Originally, this policy aimed to increase the
number of rural small landowners by releasing to the market land
properties that were, in liberal terms, lying stagnant (Arango, 2009).
The process also aimed at decreasing the social influence of the Catholic
Church, forcing the sale of ecclesiastical properties. However, the
sought improvement of land distribution was not achieved, as vast
quantities of property were acquired by an increasingly dominant
bourgeoisie (Arango, 2009). In 1855, new disentailing policies affected
public lands, many of them held in common, which imposed munici-
palities the sale of their lands through public auctions (Beltrán, 2015).
In such context, and particularly in mountain areas of central and
northern Spain where rural communities feared to loose local resources
that were essential for their everyday life, inhabitants organized

themselves and pooled capital to collectively bid in the auctions and
acquire disentailed forest commons for themselves (Medrano et al.,
2013). However, in many occasions, the same fear resulted in in-
dividual appropriation of forest commons, typically by enclosing (i.e.,
delimiting common lands within a surface demarcated, for example, on
a cadastral map) and then ploughing (i.e., transforming forestland into
crop fields) forest that had belong to the common.

Disentailment policies resulted in the individual and State appro-
priation of common lands (Caballero, 2015). The process had a great
impact in forest commons, as this type of property was widespread in
the countryside. For instance, in the northern regions of Spain, such as
La Rioja or Leon, community property at that time represented more
than three quarters of the mountain areas (Moreno, 1998; Rey, 2004).
Before disentailment, there were ten million hectares of public moun-
tain areas1 in Spain, mainly integrated by municipal properties that
included forest commons; and the amount of public woodlands sold to
particulars during 1855–1924 is estimated to be five million hectares
(Laso and Bauer, 1964; Pérez-Soba, 2013).

2.3. State interventionism and people’s resistance

Given the Spanish political and social instability, enclosing and
ploughing of forest commons and other illegal practices flourished
during the nineteenth century, largely by fear to disentailment (Aedo
et al., 1990; GEPC, 2004). Other drivers of deforestation, such as the
high demand of timber by naval shipbuilding, charcoal and cellulose
industries, development of communications infrastructure (e.g., rail-
roads ties), and population increase in rural areas, also contributed to
the depletion of forest resources (Liaño and García, 2003; Arango,
2009; Beltrán, 2015; Iriarte-Goñi, 2017). The loss of forest cover ne-
gatively affected the environment through soil erosion, reservoir sedi-
mentation, successive floods, and changes in watercourses (Jiménez-
Blanco, 2002).

Deforestation raised authorities’ awareness, who reacted to prevent
further deterioration, but aimed at doing so with a top-down scientific-
based forestry approach (Liaño and García, 2003; Parrotta and Trosper,
2012). In line with this aim, the Spanish School of Forest Engineers was
founded in 1846. In this School, professionals were trained using the
theories developed by the German Forestry Faculty of Tharandt, who
promoted the idea that States should assume the management of
woodland areas through a bureau of forest technicians for imposing on
‘disorderly nature the neatly arranged constructs of science’ (Linares,
2000). Scientific foresters would apply mathematical utilitarianism and
rationalism to promote the sustainable yield of natural resources
(Johann et al., 2012). Following this logic, the Spanish Corps of Forest
Engineers was created in 1853. Interestingly, the first request made to
this corps was to produce a list that included all country’s public
woodlands to be exempted from disentailment. The selection criteria of
the public woodlands to be exempted from the sales was established by
1863 Forestry Act and required minimum of 100ha of forest cover with
pine, oak or beech as the dominant tree species. The classification
generated a ‘Catalogue of Public Woodland exempted from disentail-
ment’ that dates from 1864 and which included 4,365,083ha of Spanish
public woodlands (Sala, 2000).

After several corrections to this Catalogue, in 1896 the approach
applied to classify public woodlands shifted the emphasis towards the
ecological importance, introducing a novel type of woodland property:
Public Utility Woodland or Monte de Utilidad Pública. According to
Royal Decree of 20 September 1896, forest stands and forestlands
should be considered as public utility woodland when their soil con-
ditions and land area requires preservation or reforestation to safeguard

1 Note that since 1864, when Spanish first classification of woodland areas
was carried out, forest commons were considered to belong to municipalities,
that is, they were considered public properties.

S. Guadilla-Sáez, et al. Forest Policy and Economics 112 (2020) 102107


public health, water regime, and soil productivity. This new criteria for
the declaration of public woodlands resulted in a ‘Catalogue of forests
and other forest lands exempted from disentailment for reason of public
utility’ that dates from 1901, and which included 5,051,112ha of
public woodlands (Pérez-Soba, 2013). Once the Catalogue classified a
municipal forest as Public Utility Woodland, its monitoring was trans-
ferred to the State Forestry Administration (Sieira, 1956; GEHR, 1999).
The inclusion implied that, from then forth, villagers had to ask for the
approval of the Forest Administration to obtain goods from forest
commons. Thus, with this catalogues, the Liberal State initiated a
process of privatization of forest commons and increased its influence in
the management of those forest commons exempted from privatization
(GEHR, 1994; Beltrán, 2015).

The great monitoring exerted over forest commons by the State gave
rise to tensions between forest authorities and villagers (Cobo et al.,
1992; Linares, 2000). Still, until the nineteenth century, rural com-
munities often contested State actions with protests like illegally felling
trees or enclosing common lands. Popular resistance processes have
been also documented in other European countries associated to the
transition from the communal system to the institutionalization of
forestry (Hölzl, 2011; Radkau, 2012). During the nineteenth century,
however, local strategies to recover traditional rights lost because of the
disentailment policies were more complex, including a combination of
individual illegal actions such as ploughing or fires, and organized legal
actions like the collective purchase of communal lands in public auc-
tions (Gavira, 1998; Linares, 2000; Piqueras and Sanz, 2007; Valbuena-
Carabaña et al., 2010).

The opposition to the abolition of the historical communal property
and use rights provoked numerous conflicts during the nineteenth and
twentieth centuries. Still, as Soto et al. (2007) remark, it is convenient
to distinguish between conflicts generated by the loss of ownership
rights and conflicts generated by the loss of traditional community use
rights. On the one side, the Spanish State did not recognize community
ownership; moreover, a royal decree issued in 1848 denied any possi-
bility of community ownership, and forest commons’ ownership was
considered as public property and their ownership rights were trans-
ferred to municipalities (Izquierdo, 2007; Caballero, 2015). Several
instances of conflicts related to the loss of traditional community
property rights have been documented particularly in Northwest Spain,
where forest commons have a private ownership origin (Cuadrado,
1980; Caballero, 2015).

On the other side, rural inhabitants were concerned by the loss of
use rights in common lands (Cobo et al., 1992; Soto et al., 2007). As
mentioned, forests hold resources that were critical to rural livelihood,
notably for the poorest peasants. Disentailment policies, along with
State monitoring of forest commons classified as Public Utility Wood-
lands, resulted in the decrease of the forest area that local communities
were able to use (Cobo et al., 1992). In addition, the forest management
system adopted by State forest technicians –based on the assumption
that some traditional uses, such as grazing or prescribed burns, were
incompatible with long-term conservation of forest cover– limited tra-
ditional forest-related practices (Sala, 2000; Serrano, 2005). However,
partly due to the key role of these traditional practices on local liveli-
hood (Cobo et al., 1992; Balboa, 1999), but also as a means of protest
(Piqueras, 2002; GEHR, 1999), these uses continued to be carried out
by rural communities. Moreover, as the new legal framework dismissed
concejos’ authority to sanction illegal uses, activities such as the en-
closing and ploughing of forest commons proliferated (Serrano, 2005).
Therefore, the penalization of traditional uses, rather than resulting in
forest cover preservation, seems to have had the opposite effect
(Campos et al., 2013).

Overall, the disentailing process negatively affected forest con-
servation in the short term for two reasons. First, privatized forests were
logged, as private owners were inclined to compensate the cash value of
their purchase (Ezquerra and Gil, 2008). The decline of tree canopy
soon thereafter resulted in natural damages caused by flooding and soil

erosion (Laso and Bauer, 1964). Second, local opposition to the cessa-
tion of forest commons’ historical uses resulted in the proliferation of
illegal practices that were not longer monitored by the concejos
(Serrano, 2005).

2.4. Replacement of traditional uses by modern forestry approaches

Spanish rural landscapes entered the twentieth century drastically
deforested (Valbuena-Carabaña et al., 2010). During the twentieth
century, the State Forest Administration focused its efforts on reversing
the degradation trend and restoring the vegetation cover of public
woodlands through afforestation policies (GEHR, 1999). During the
first third of the century (1901–1939), afforestation focused on pro-
tective outcomes, such as to prevent periodic flooding or fixing coastal
sand dunes, for which fast growing tree species, like Pinus species, were
used (Valbuena-Carabaña et al., 2010; Vadell et al., 2016). Later, from
1940 to 1986, afforestation shifted to an intensive silvicultural treat-
ment towards wood production in which non-native fast growing tree
species that could be harvested in less than ten years, such as Populus
and Eucalyptus species, were favoured over traditional ones (GEPC,
2004; Ramos, 2007; Valbuena-Carabaña et al., 2010; Vadell et al.,

Another measure taken by Spanish public administration to avoid
further degradation of forested landscapes was to adopt a more con-
servationist interventionism in those forests classified as Public Utility
Woodlands. The management of forest commons, monitored by the
State Forest Administration since 1863, aimed to prevent traditional
uses in catalogued woodlands. This intervention was done under the
argument that some practices, such as logging, firewood collection or
small ruminant livestock grazing, were incompatible with the long-term
maintenance of forest cover (Cobo et al., 1992; Linares, 2000; Linares,
2007; Montiel, 2007). To that end, from the beginning of the nineteenth
century, the access to and use of public woodlands became regulated
through forest management plans (GEHR, 1999), which –in an attempt
to reduce peasants’ use of forest resources– were often overly restrictive
(Cobo et al., 1992; Balboa, 1999).

Both interventions –afforestation policies and traditional forest uses
control–, along with the dismantling of forest commons, resulted in a
decline in the use of woodlands by rural communities. The intensive
afforestation created very specific ecological systems that were not
connected to local productive systems, impeding the multiple-use of
forest resources (GEPC, 2004). Livelihood activities such as pastoralism,
small-scale agriculture, or fuelwood gathering were forbidden in re-
forested plantations (Jiménez-Blanco, 2002). Additionally, and partly
due to restrictions in the use of forest resources, local communities
increasingly abandoned forest-based activities shifting to other eco-
nomic activities in response to national market demands (Ramos,

The disentailment process –which continued until the first decades
of the twentieth century– also encouraged the entrance of a market-
based economy in agricultural production, as the enclosing and
ploughing of common lands allowed farmers to enlarge their productive
capacity. However, farmers’ illegal appropriations of common lands
was a heterogeneous phenomenon in the peninsula, because in those
areas where farming practices competed with other land uses, such as
extensive livestock grazing, conflicts among community members lim-
ited the enclosing process (Beltrán, 2015). Thus, in semi-arid Medi-
terranean areas of Spain, where environmental conditions are favour-
able for agriculture, cropland was favoured at the expense of the
dehesas, a woodland-pasture managed in common in the past (Linares,
2000; Campos et al., 2013). In contrast, in Mediterranean continental
areas, with environmental conditions less favourable for farming cul-
tivation, summer pastures were favoured instead, as a mean of guar-
anteeing the basement for traditional stockbreeding, frequently man-
aged through systems of agrarian collectivism (Montiel, 2007).
Similarly, northern areas located in the Atlantic ecosystem did not

S. Guadilla-Sáez, et al. Forest Policy and Economics 112 (2020) 102107


experience the enlargement of arable land at the expense of forestlands.
In contrast to Mediterranean climate territories covering the rest of
Spain, where dry weather conditions limited the increase agricultural
production to extensive agriculture, the higher production capacity of
Atlantic weather systems due to their humid conditions resulted in the
intensification of agricultural productivity of these lands without re-
sorting to the expansion of crops. Additionally, a sharp relief and de-
ficient communications also contributed to reduce the ploughing of
northern forestlands (GEHR, 1994; GEPC, 2004).

2.5. Rural landscape dynamics resulting from the abandonment of
traditional forest management

From mid-twentieth century onwards, the enclosing process de-
clined. The rural crisis associated with depopulation, agricultural and
livestock intensification and mechanization, and the abandonment of
traditional activities, took out pressure from the arable land supply
leading to a progressive natural vegetation succession of abandoned
lands (Rotherham, 2013; Viedma et al., 2015). These changes led to the
densification and homogenization of the traditional rural landscape
mosaic, which in turn resulted in an impoverishment of forest biodi-
versity because of the transformation of woodland to shrubland and in
an increasing risk of wildfires due to a higher fuel load and continuity
(Loepfe et al., 2010).

In Spain, as in other areas of mainland Europe, the abandonment of
traditional forest management systems seems to have had long-term
negative ecological outcomes to species dependent on features of old
forests and old agricultural landscapes such as dead wood, old trees, fire
and open grazed areas (Bengtsson et al., 2000). Although not all cus-
tomary practices might necessarily had been positive for biodiversity
conservation, local communities contributed to preserve landscape
patchiness and favoured the occurrence of stress-tolerant species
through their small and intermediate-scale disturbances and a set of
site-specific norms to manage resources sustainably (Potee and Ostrom,
2004; Babai et al., 2015). As previously mentioned, the importance of
forest commons’ resources to the livelihood strategies of Spanish rural
communities, particularly in mountain areas, often resulted in local
forms of multiple use of forest-related resources (e.g. cultural practices)
(Ezquerra and Gil, 2004; Piqueras and Sanz, 2007; Serrano, 2014).

In the absence of these historical forms of management, over the last
decades many Spanish biodiversity-rich semi-natural habitats have
declined or been lost. For instance, the progressive decline of chestnut
groves (Castanea sativa Mill.) –a forest habitat of high species diversity
that is included in the European Habitats and Species Directive (92/43/
CEE)– has been attributed to the abandonment of human management
practices on these stands, including grazing, regular pruning, or peri-
odical understory burns (Gondard et al., 2006; Guitián et al., 2012).
Similarly, the cessation of livestock grazing and periodic hay cutting
seems to have resulted in the progressive ecological succession in
mountain areas from shrubland of Erica ciliaris and Erica tetralix, a
priority natural habitat type of interest according to the European Di-
rective 97/62/EC (Corbelle-Rico et al., 2015), to areas encroached by

The much-managed dehesa landscape, which is also included in the
European Habitats Directive for its positive effects for biodiversity
conservation (Campos et al., 2013), currently suffers from abandon-
ment of traditional uses, particularly traditional pruning. Traditional
pruning of Quercus spp. makes compatible farming and herding with the
persistence of a canopy layer, at the same time that allows the persis-
tence of veteran trees, which are key for saproxylic fauna and flora and
as a habitat niche for cavity-nesting birds and wood-inhabiting fungi
(Olea and San Miguel-Ayanz, 2006; Siitonen and Ranius, 2015). The
cessation of Quercus spp. traditional management is linked to the woody
encroachment of these habitats. Moreover, traditional management is
being substituted by more intensive systems, such as commercial con-
ifer forestlands, with the consequent loss of flora and fauna associated

to Quercus forests (Taboada et al., 2006). Another traditional practice
worth mentioning for its positive ecological outcomes is the transhu-
mance, an ancestral pastoral practice consisting of seasonal moving of
livestock to graze on higher pastures in summer, which arguably con-
tributes to species biodiversity by increasing landscape complexity
through the creation of grassland-woodland habitat mosaics (Oteros-
Rozas et al., 2012; Orlandi et al., 2016).

Thus, overall, the State Forest Administration initial assumption
that decreasing traditional forest-related practices would result in forest
conservation seems to have had the opposite effect in the long-term. On
the one side, the abandonment of traditional uses led to the en-
croachment of forest habitats and the simplification and homogeniza-
tion of rural landscape mosaic shaped by traditional management, ne-
gatively affecting biodiversity conservation and increasing fire hazard
risks. On the other side, reforestation processes with monospecific even-
aged stands, despite being effective for preventing soil erosion, criti-
cally impoverished shrub and ground flora across the country and in-
creased the occurrence of natural hazards such as pests and diseases
(Jiménez-Blanco, 2002; Iriarte-Goñi, 2017).

3. Multiple legacies in current legal framework of Spanish forest

As the historical account suggest, the persistence of forest collective
ownership has been closely related to the environmental conditions of
the different geographical areas of Spain (Beltrán, 2015). Thus, in
southern areas where agricultural lands were favoured at the expense of
forest commons, enclosing rates were higher than in mountainous re-
gions of north Spain, where forest-related goods were a major support
to the rural livelihood. In addition, Spanish northern rural communities
fought for legal recognition of their traditional community forests’
rights, which were restored by the 1957 Forestry Act that recognized a
particular type of collective woodlands known as neighbour woodlands
or montes vecinales en mano común (Cuadrado, 1980; Balboa, 1999;
GEPC, 2004; IDEGA, 2013; Mangas, 2013). Also in some regions of
northern and central Spain, where local communities’ collectively
purchased forest commons during disentailment, this legacy of com-
munity ownership has been recently legally recognized, being catalo-
gued as pro indiviso forests in an additional disposition to Spanish 2003
Forestry Act.

The most recent Spanish Forestry Act (Law No. 21/2015 of Spanish
Government) distinguishes between three different categories of com-
munity-ownership forests: (1) Forest commons, (2) Partners’ wood-
lands, and (3) Neighbour woodlands.

3.1. Forest commons

Forest commons, or montes comunales in Spanish, are conformed by
former forest commons that survived the privatization wave or –in
other words– that were considered as exempted from disentailment
during the nineteenth century classification carried out by the Corps of
Forest Engineers. Forest commons typically belong to municipalities,
but their use corresponds to local communities (Sieira, 1956; Pérez-
Soba and Solá, 2004). The management of forest commons is ruled by
ordinances approved by residents, with the forestry administration
exerting its influence by monitoring the commoners’ access to grazing,
firewood, and other forest-related goods (Balboa, 1999).

Forest commons are the most abundant community forests in Spain,
with presence in all regions of the country. They used to feature a
regulated spatial planning in which commoners organized themselves
to carry out traditional practices (Montiel, 2003; Couto and Gutiérrez,
2012). Examples of these practices include the vecería, a communal
pastoral activity consisting in shifting turns among community mem-
bers to move a common herd to graze in forest commons, which was
habitual in Spanish northern regions until recent times (González,
2001). According to Vázquez (2016), apart from reducing agricultural

S. Guadilla-Sáez, et al. Forest Policy and Economics 112 (2020) 102107


workload, the vecería system also strengthened social relations in rural
communities. The same author posited that, although the persistence of
this collective organization system was undesirable by liberals, local
requests to preserve it were frequently accepted, as commoners argued
that individual pastoral systems are not viable for the poorest peasants,
likely resulting in cattle mismanagement and uncontrolled grazing.

A remarkable example of forest common with positive ecological
impact is the Urbión Forest in Castile and Leon, north Spain. In this
region, thirteenth century local ordinances enforcing local commu-
nities’ right to make use of forest goods have been legally endorsed until
the present, for which traditional community management was not
affected by exclusionary policies. Interestingly, the Urbión Forest
nowadays is the most extensive continuous wooden area on the pe-
ninsula, with Pinus spp., Quercus spp., Fagus sylvatica L. and Juniperus
thurifera L. stands (Segur et al., 2014). These high nature value habitats
overlap now with Cañón del Río Lobos Natural Park and several Natura
2000 sites.

3.2. Partners’ woodlands

Partners’ woodlands, or montes de socios, are a second type of
community-ownership woodlands that survived the disentailing po-
licies. Partners’ woodlands were created during the nineteenth century
through the association of neighbours that pooled economic resources
to buy the forests that they had traditionally managed in common
(Montiel, 2005; Lana and Iriarte-Goñi, 2015). This category constitutes
a type of ownership form in which the forest is private property, owners
being a group of people who collectively bought it, resulting in a great
number of co-owners (Mangas, 2013; Medrano et al., 2013).

Most of partners’ woodlands originated from the collective response
of rural communities in forested provinces of inner Spain such as
Burgos and Soria, where a strong local opposition to forest commons’
usurpation took place (Piqueras, 2002; Montiel, 2007). Partners’
woodlands can be also found in other north and inner provinces of
Spain, referred by a large variety of names –e.g., montes del común
(woodlands of the common), sociedad del monte (society’s woodland),
monte de la sociedad de vecinos (neighbours’ society woodland)–, a di-
versity that actually reflects their past abundance and importance
(Medrano et al., 2013). Nowadays, more than 1,500,000ha of forest-
land have the status of partners’ woodlands, although a legal framework
for the management of this type of property was not issued until the
2015 Spanish Forestry Act. The new legal code included Partners’
Woodlands category to respond to the management problems resulted
from not having updated the co-ownership property details related to
those forests acquired by groups of local residents during public auc-
tions. This legal endorsement allows the identified commoners to
manage the woodlands, in order to avoid the mismanagement –and
consequently deterioration– of these ecosystems.

From an ecological perspective, partners’ woodlands can also pro-
vide important insights for designing sustainable conservation strate-
gies. For instance, the economic profit obtained from the traditional
practice of quotas or suertes, in which woodlands are divided in cut-
blocks among residents for timber harvesting, is known to increase
commoners’ interest in forest conservation, reflected in the internal
regulations issued to prevent illegal uses such as logging or burning
(Gogeascoechea, 1999). Nowadays, several partner’s woodlands
overlap with Natura 2000 areas, such as Sierra de Cabrejas in Castile
and Leon that constitutes the largest Juniperus thurifera forest in Europe
(Pecurul-Botines et al., 2014).

3.3. Neighbour woodlands

Neighbour woodlands, or montes vecinales en mano común (hereafter
MVMC, for their acronym in Spanish), originated during the second half
of the twentieth century from the social resistance of Galician peasants,
in Northwest Spain, to disentailment policies (Couto and Gutiérrez,

2012). MVMC are a type of ownership in which the forest is private
property, owned by all neighbours of a particular local community. In
other words, the status of neighbour is required to obtain forest own-
ership and use rights (Caballero, 2015).

Noteworthy, MVMC have a very different historical evolution than
forest commons and partners’ woodlands (Balboa, 1999). MVMC have a
private ownership origin, in which ownership collectively belong the
local community members (Cuadrado, 1980; Caballero, 2015). This
particular private ownership form was not recognized by the liberal
legislation, which in 1848 denied any possibility of community-own-
ership form in Spain and transferred neighbour woodlands’ ownership
rights to municipalities. However, in Galicia, more than in other re-
gions, the livelihood dependence of forest resources led to a strong,
long-term rural resistance to the appropriation process (Piqueras, 2002;
GEPC, 2004). The combination of both factors –an original private
ownership form and strong local opposition to the dismantling process–
resulted in the early legal recognition of the community-ownership
form in Galicia through the inclusion of MVMC in the 1957 Spanish
Forestry Act, and through the promulgation in 1968 of a specific For-
estry Act returning MVMC ownership rights to Galician local commu-
nities. In 1975, the legal framework of MVMC was extended to the
neighbouring northwest provinces of Zamora, León, Asturias and Can-
tabria (Cuadrado, 1980; Fernández, 1992; Blanco, 2014).

Nowadays MVMC represent one-third of the total surface of Galician
forests, covering approximately 673,000ha, with more than 2800
community-owners ruling the management of MVMC based on tradi-
tional norms (Arango, 2009; IDEGA, 2013; Caballero, 2015). Due to the
1968 legal endorsement, neighbours have been able to manage forest
resources for many decades, which has resulted in many high valued
ecosystems that currently overlap with different protection categories
(Table 1).

4. Conclusion

This paper provides a historical examination of community-owner-
ship tenure regimes in Spanish woodlands and their potential impacts
on forest biodiversity maintenance. It highlights how local communities
created norms and institutions regulating and monitoring the multiple
uses of forest commons to prevent resources depletion back in the
Middle Ages. The replacement of traditional community governance

Table 1
Examples of Protected Areas overlapping with Neighbour woodlands. Source:
Biodiversity database, Spanish Ministry of Agriculture, Fisheries and Food.

Category of Protected Area Name of the site

Biosphere Reserve Área de Allariz
Coruñesas e Terras do Mandeo
Geres-Xures (Transboundary Biosphere Reserve)
Os Ancares Lucenses y Montes de Cervantes, Navia
y Becerrea
Río Eo, Oscos y Ferras de Buron
Terras do Miño

Natural Monument Serra de Pena Corneira
Souto da Retorta
Pregamento xeolóxico de Campodola-Leixazós

Natural Park O Invernadeiro
Serra da Enciña da Lastra
Fragas do Eume
Monte Aloia
Baixa Limia-Serra do Xurés
Complexo Dunar de Corrubedo e Lagoas de
Carregal e Vixán

Natural Protected Area Sobreiras do Faro
Protected Landscape Val do río Navea
Protected Wetland Complexo intermareal Umia-O Grove, A Lanzada,

punta Carreirón e lagoa Bodeira
Complexo das Praias, Duna e lagoa de Corrubedo

Site of national interest A Curotiña

S. Guadilla-Sáez, et al. Forest Policy and Economics 112 (2020) 102107


systems occurring during the process of privatization and state inter-
ventionism of communal lands in the nineteenth and twentieth cen-
turies had negative consequences for forest cover maintenance in Spain.
On the one side, forests acquired by private owners were cut to com-
pensate the cash value of their purchase. On the other, forest considered
as public suffered from illegal uses no longer sanctioned by local in-

In the long-term, the abandonment of traditional uses lead to a
simplification and homogenization of the rural landscape mosaic, as-
sociated to a decrease of biodiversity and an increase of fire hazard risk.
Interestingly, in geographical areas showing a stronger opposition to
forest commons’ dismantling policies and where traditional community
ownership rights were restored earlier, we find several instances of
overlap between forests commons and conservation areas. These results
provide support to the idea that community-based management can
hold useful insights for the maintenance of diverse, high ecological
valued ecosystems, while allowing local communities the use of natural
resources. Further research should aim to identify which particular
management practices traditionally applied by local communities in
forest ecosystems of Spain had been favourable to biodiversity. Ideally,
future research might address how present legacies of Spanish com-
munity forest can complement forestry science in the design and im-
plementation of conservation policy agenda.

Declaration of Competing Interest

The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.


This research was supported by the Agència de Gestió d’Ajuts
Universitaris i de Recerca AGAUR of the Government of Catalonia
(2015FI_B00333) and by the Spanish Ministry of Economy and
Competitiveness through the research project ‘Citizen Science and tra-
ditional agroecological knowledge: How to increase citizen’s partici-
pation in the Spanish inventory of traditional knowledge related to
biodiversity?’ (CSO2014-59704-P). We thank Pablo Domínguez, Mar
Grau and Petra Benyei for revising and improving earlier versions of the
manuscript. This work contributes to Environmental Sciences and
Technology Institute ICTA-UAB ‘Unit of Excellence’ (MinECo,


Aedo, C., Diego, C., García, J.C., Moreno, G., 1990. El bosque en Cantabria. Universidad
de Cantabria, Santander.

Agrawal, A., Gibson, C.C., 1999. Enchantment and disenchantment: the role of commu-
nity in natural resource conservation. World Dev. 27 (4), 629–649.

Aranda, G., 1996. La selvicultura en España hasta el siglo XIX. Ecología 10, 173–184.
Arango, J., 2009. Montes comunales en Asturias y otras cuestiones agrarias. KRK

Ediciones, Oviedo.
Babai, D., Tóth, A., Szentirmai, I., Biró, M., Máté, A., Demeter, L., Szépligeti, M., Varga,

A., Molnár, A., Kun, R., Molnár, Z., 2015. Do conservation and agri-environmental
regulations effectively support traditional small-scale farming in east-central
European cultural landscapes? Biodivers. Conserv. 24, 3305–3327.

Balboa, X., 1999. La historia de los montes públicos españoles (1812-1936): Un balance y
algunas propuestas. Historia Agraria 18, 95–128.

Beltrán, F.J., 2015. Social and environmental filters to market incentives: the persistence
of common land in nineteenth-century Spain. J. Agrar. Chang. 15 (2), 239–260.

Bengtsson, J., Nilsson, S.G., Franc, A., Menozzi, P., 2000. Biodiversity, disturbances,
ecosystem function and management of European forests. For. Ecol. Manag. 132,

Bergstén, S., Stjernström, O., Pettersson, Ö., 2018. Experiences and emotions among
private forest owners versus public interests: why ownership matters. Land Use Policy
79, 801–811.

Blanco, A.L., 2014. La imprescriptibilidad de los bienes comunales: su trascendencia para
la conservación de la propiedad forestal pública. Revista jurídica de Castilla y León

32, 1–37.
Brockington, D., Duffy, R., Igoe, J., 2008. Nature Unbound: Conservation, Capitalism and

the Future of Protected Areas. Earthscan, London.
Bruner, A.G., Gullison, R.E., Rice, R., da Fonseca, G.A., 2001. Effectiveness of parks in

protecting tropical biodiversity. Science 291, 125–128.

Burns, R.I., 2012. The medieval world of law: lawyers and their work (Partida III). In: Las
Siete Partidas, volume 3. University of Pennsylvania Press, Philadelphia.

Caballero, G., 2015. Community-based forest management institutions in the Galician
communal forests: a new institutional approach. For. Policy Econ. 50, 347–356.

Campos, P., Huntsinger, L., Oviedo, J.L., Starrs, P.F., Díaz, M., Standiford, R.B., Montero,
G., 2013. Mediterranean Oak Woodland Working Landscapes. Springer, Netherlands.

Chhatre, A., Agrawal, A., 2008. Forest commons and local enforcement. PNAS 105 (36),

Cobo, F., Cruz, S., González de Molina, M., 1992. Privatización del monte y protesta
campesina en Andalucía Oriental (1936-1920). Agric. Soc. 65, 253–302.

Coetzee, B.W.T., Gaston, K.J., Chown, S.L., 2014. Local scale comparisons of biodiversity
as a test for global protected area ecological performance: a meta-analysis. PLoS One
9 (8), e105824.

Corbelle-Rico, E., Butsic, V., Enríquez-García, M.J., Radeloff, V.C., 2015. Technology or
policy? Drivers of land cover change in northwestern Spain before and after the ac-
cession to European economic community. Land Use Policy 45, 18–25.

Couto, S., Gutiérrez, J.E., 2012. Recognition and support of ICCAs in Spain. In: Kothari,
A., Corrigan, C., Jonas, H., Neumann, A., Shrumm, H. (Eds.), Recognising and
Supporting Territories and Areas Conserved by Indigenous Peoples and Local
Communities: Global Overview and National Case Studies. Secretariat of the
Convention on Biological Diversity, ICCA Consortium, Kalpavriksh, and Natural
Justice, Montreal.

Cuadrado, M., 1980. Aprovechamiento en común de pastos y leñas. Servicio de
Publicaciones Agrarias, Madrid.

Dudley, N., Belokurov, A., Borodin, O., Higgins-Zogib, L., Hockings, M., Lacerda, L.,
Stolton, S., 2004. Are Protected Areas Working? An Analysis of Forest Protected
Areas by WWF. WWF, Gland.

Dudley, N., Baldock, D., Nasi, R., Stolton, S., 2005. Measuring biodiversity and sustain-
able management in forest and agricultural landscapes. Philos. Trans. R. Soc. B 360,

Ezquerra, F.J., Gil, L., 2004. La transformación histórica del paisaje forestal en la co-
munidad de Cantabria. Tercer Inventario Forestal Nacional, Ministerio de Medio
Ambiente, Madrid.

Ezquerra, F.J., Gil, L., 2008. La transformación histórica del paisaje forestal en
Extremadura. Ministerio de Medio Ambiente, Madrid.

Fernández, M., 1992. Los aprovechamientos comunales del monte en el cuadrante nor-
occidental de España. Anales del Museo del Pueblo Español IV 49–83.

Gaston, K.J., Jackson, S.F., Cantú-Salazar, L., Cruz-Piñón, G., 2008. The ecological per-
formance of protected areas. Annu. Rev. Ecol. Evol. Syst. 39, 93–113. https://doi.

Gavira, C., 1998. Administración y jurisdicción en el valle de Soba a fines del siglo XVIII.
Altamira 56, 141–165.

GEHR, 1994. Más allá de la “propiedad perfecta”. El proceso de privatización de los
montes públicos españoles (1859-1926). Historia Agraria 8, 99–152.

GEHR, 1999. Diversidad dentro de un orden. Privatización, producción forestal y
represión en los montes públicos españoles, 1859-1926. Historia Agraria 18,

Geldmann, J., Barnes, M., Coad, L., Craigie, I.D., Hockings, M., Burgess, N.D., 2013.
Effectiveness of terrestrial protected areas in reducing habitat loss and population
declines. Biol. Conserv. 161, 230–238.

GEPC, 2004. La devolución de la propiedad vecinal en Galicia (1960-1985). Modos de uso
y conflicto de propiedad. Historia Agraria 33, 105–130.

Gogeascoechea, A., 1999. Los montes proindivisos en Vizcaya. Lurralde: Investigación y
espacio 22, 299–322.

Gondard, H., Romane, F., Santa Regina, I., Leonardi, S., 2006. Forest management and
plant species diversity in chestnut stands of three Mediterranean areas. Biodivers.
Conserv. 15, 1129–1142.

González, M., 2001. Sociología y ruralidades. La construcción social del desarrollo rural
en el valle de Liébana. Ministerio de Agricultura, Pesca y Alimentación, Madrid.

Gray, C.L., Hill, S.L.L., Newbold, T., Hudson, L.N., Börger, L., Contu, S., Hoskins, A.J.,
Ferrier, S., Purvis, A., Scharlemann, J.P.W., 2016. Local biodiversity is higher inside
than outside terrestrial protected areas worldwide. Nat. Commun. 7, 12306. https://

Guadilla-Sáez, S., Pardo-de-Santayana, M., Reyes-García, V., Svenning, J.-C., 2019.
Biodiversity conservation effectiveness provided by a protection status in temperate
forest commons of North Spain. For. Ecol. Manag. 433, 656–666.

Guitián, J., Guitián, P., Munilla, I., Guitián, J., Garrido, J., Penín, L., Domínguez, P.,
Guitián, L., 2012. Biodiversity in chestnut woodlots: management regimen vs woo-
dlot size. Open J. For. 2 (4), 200–206.

Hayes, T.M., 2006. Parks, people, and forest protection: an institutional assessment of the
effectiveness of protected areas. World Dev. 34 (12), 2064–2075.

Hernando, A., Tejera, R., Velázquez, J., Núñez, M.V., 2010. Quantitatively defining the
conservation status of Natura 2000 forest habitats and improving management op-
tions for enhancing biodiversity. Biodivers. Conserv. 19, 2221–2233.

Hölzl, R., 2011. Forests in conflict: Rural populations and the advent of modern forestry

S. Guadilla-Sáez, et al. Forest Policy and Economics 112 (2020) 102107


in pre-industrial Germany, 1760-1860. In: Massard-Guilbaud, G., Mosley, S. (Eds.),
Common Ground: Integrating the Social and Environmental in History. Cambridge
Scholars Publishing, UK, pp. 198–223.

IDEGA, 2013. La política forestal gallega en los montes vecinales en mano común.
Ambienta 104, 114–125.

Iriarte-Goñi, I., 2017. El contexto socio-económico de las repoblaciones en España (1939-
c.1980). In: Pemán, J., Iriarte-Goñi, I., Lario, F.J. (Eds.), La restauración forestal de
España. Ministerio de Agricultura y Pesca. Alimentación y Medio Ambiente, Madrid,
pp. 23–41.

Izquierdo, J., 2007. En nombre de la comunidad. Antropología de la propiedad en el
Antiguo Régimen. In: Congost, R., Lana, J.M. (Eds.), Campos cerrados, debates
abiertos. Análisis histórico y propiedad de la tierra en Europa (siglos XVI-XIX).
Universidad Pública de Navarra, Pamplona, pp. 53–73.

Jiménez-Blanco, J.I., 2002. El monte: una atalaya de la Historia. Historia Agraria 26,

Johann, E., Agnoletti, M., Bölöni, J., Yurdakul Erol, S.Y., Holl, K., Kusmin, J., García
Latorre, J., García Latorre, J., Molnár, Z., Rochel, X., Rotherham, I.D., Saratsi, E.,
Smith, M., Tarang, L., van Benthem, M., van Laar, J., 2012. Europe. In: Parrotta, J.A.,
Trosper, R.L. (Eds.), Traditional Forest-Related Knowledge. Springer, New York, pp.

Kremen, C., 2015. Reframing the land-sparing/land-sharing debate for biodiversity con-
servation. Ann. N. Y. Acad. Sci. 1355, 52–76.

Laborda-Pemán, M., De Moor, T., 2012. A Tale of Two Commons: Some Preliminary
Hypotheses on the Long-Term Development of the Commons in Western and Eastern
Europe, 1000-1900. Papers 0031. Utrecht University, Centre for Global Economic

Lana, J.M., Iriarte-Goñi, I., 2015. Commons and the legacy of the past. Regulation and
uses of common lands in twentieth century Spain. Int. J. Commons 9, 510–532.

Laso, M.P., Bauer, E., 1964. La propiedad forestal en España. Revista de Estudios
Agrosociales 49, 7–53.

Liaño, C.D., García, J.C., 2003. La Corona y los pueblos en la explotación de los montes de
Cantabria: deforestación y gestión del bosque en la segunda mitad del siglo XVIII.
Cuad. Soc. Esp. Cien. For. 16, 215–220.

Linares, A., 2000. Economic Adaptation and Social Resilience to State Intervention in the
Spanish Common Forests (1855–1925). In: Proceedings of the VIIIth IASCP
Conference, Bloomington.

Linares, A.M., 2007. Forest planning and traditional knowledge in collective woodlands
of Spain: the Dehesa system. For. Ecol. Manag. 249, 71–79.

Loepfe, L., Martinez-Vilalta, J., Oliveres, J., Piñol, J., Lloret, F., 2010. Feedbacks between
fuel reduction and landscape homogenisation determine regimes in three
Mediterranean areas. For. Ecol. Manag. 256, 2366–2374.

Mangas, J.M., 2013. Génesis y evolución de los patrimonios territoriales públicos y co-
munitarios. Ambienta 104, 22–53.

Mathur, P.K., Sinha, P.R., 2008. Looking beyond protected area networks: a paradigm
shift in approach for biodiversity conservation. Int. For. Rev. 10 (2), 305–314.

McElhinny, C., Gibbons, P., Brack, C., Bauhus, J., 2005. Forest and woodland stand
structural complexity: its definition and measurement. For. Ecol. Manag. 218, 1–24.

Medrano, P.A., Marín, A., Gracia, P., 2013. Montes de socios: Un ejemplo de gestión
forestal al servicio del desarrollo rural. Ambienta 104, 102–113.

Montiel, C., 2003. Origen y evolución de la propiedad forestal colectiva en España. Cuad.
Soc. Esp. Cien. For. 16, 285–290.

Montiel, C., 2005. Los montes de socios: un problema territorial de difícil gestión. Boletín
de la A.G.E. 40, 181–200.

Montiel, C., 2007. Cultural heritage, sustainable forest management and property in in-
land Spain. For. Ecol. Manag. 249, 80–90.

Moreno, J.R., 1998. El régimen comunal y la reproducción de la comunidad campesina en
las sierras de La Rioja (siglos XVIII-XIX). Historia Agraria 15, 75–111.

Nielsen, A.S.E., Strange, N., Bruun, H.H., Jacobsen, J.B., 2017. Effects of preference
heterogeneity among landowners on spatial conservation prioritization. Conserv.
Biol. 31 (3), 675–685.

Olea, L., San Miguel-Ayanz, A., 2006. The Spanish dehesa. A traditional Mediterranean
silvopastoral system linking production and nature conservation. In: Lloveras, J.,
González-Rodríguez, A., Vázquez-Yáñez, O., Piñeiro, J., Sanatamaría, O., Olea, L.,
Poblaciones, M.J. (Eds.), Sustainable Grassland Productivity. SEEP, Madrid, pp. 3–13.

Orlandi, S., Probo, M., Sitzia, T., Trentanovi, G., Garbarino, M., Lombardi, G., Lonati, M.,
2016. Environmental and land use determinants of grassland patch diversity in the
western and eastern Alps under agro-pastoral abandonment. Biodivers. Conserv. 25,

Ostrom, E., 1999. Self-governance and forest resources. In: CIFOR. Bogor, Indonesia.

Oteros-Rozas, E., González, J.A., Martín-López, B., López, C.A., Montes, C., 2012.
Ecosystem services and social–ecological resilience in transhumance cultural land-
scapes: Learning from the past, looking for a future. In: Plieninger, T., Bieling, C.
(Eds.), Resilience and the Cultural Landscape in Human-Shaped Environments.
Cambridge University Press, Cambridge, pp. 242–260.

Pardo, F., Gil, L., 2005. The impact of traditional land use on woodlands: a case study in

the Spanish central system. J. Hist. Geogr. 31, 390–408.

Parrotta, J.A., Trosper, R.L., 2012. Traditional forest-related knowledge. In: Sustaining
Communities, Ecosystems and Biocultural Diversity. Springer, New York. https://doi.

Pecurul-Botines, M., Di Gregorio, M., Paavola, J., 2014. Discourses of conflict and col-
laboration and institutional context in the implementation of forest conservation
policies in Soria, Spain. Biodivers. Conserv. 23, 3483–3499.

Pérez-Soba, I., 2013. Los montes de utilidad pública: un patrimonio con mucho pasado y
mucho futuro. Ambienta 104, 54–72.

Pérez-Soba, I., Solá, M.A., 2004. La tragedia de los comunales: Legalidad y realidad de los
montes comunales en España. Estudios Agrosociales y Pesqueros 203, 187–232.

Piqueras, J.A., 2002. Bienes comunales: propiedad, arraigo y apropiación. In: Ministerio
de Agricultura. Alimentación, Madrid, Pesca y.

Piqueras, J.A., Sanz, V., 2007. A social History of Spanish Labour: New Perspectives on
Class, Politics and Gender. Berghahn Books, New York.

Poiani, K.A., Richter, B.D., Anderson, M.G., Richter, H.E., 2000. Biodiversity conservation
at multiple scales: functional sites, landscapes, and networks. Bioscience 50 (2),

Potee, A.R., Ostrom, E., 2004. Heterogeneity, group size and collective action: the role of
institutions in forest management. Dev. Chang. 35 (3), 435–461.

Pynnönen, S., Paloniemi, R., Hujala, T., 2018. Recognizing the interest of forest owners to
combine nature-oriented and economic uses of forests. Small-scale For. 17, 443–470.

Radkau, J., 2012. Wood: A History. Polity Press, Cambridge.
Ramos, J.M., 2007. Aprovechamientos forestales, ordenanzas de montes y conflictividad

social en el norte de Palencia en la Edad Moderna (siglos XVII-XVIII). Polígonos 17,

Rey, O., 2004. Montes, bosques y zonas comunes: aprovechamientos agrícola-ganaderos,
forestales y cinegéticos. In: Aranda, F.J. (Ed.), El Mundo Rural en la España Moderna.
Universidad de Castilla-La Mancha, Cuenca, pp. 907–964.

Rotherham, I.D., 2013. Cultural Severance and the Environment. Springer, Dordrecht.
Sala, P., 2000. Modern Forestry and Enclosure: Elitist State Science against Communal

Management and Unrestricted Privatisation in Spain. Environ. Hist. Camb. 6,

Schmeller, D.S., Bridgewater, P., 2016. The intergovernmental platform on biodiversity
and ecosystem services (IPBES): progress and next steps. Biodivers. Conserv. 25,

Segur, M., Martel, S., Picardo, Á., Medrano, P., Lucas, J.A., Martínez, A., Valbuena, P.,
Alonso, J., Gil, P., Galindo, E., 2014. Old solutions for today’s problems in the Urbión
model forest. In: Katila, P., Galloway, G., de Jong, W., Pacheco, P., Mery, G. (Eds.),
Forests under pressure: Local responses to global issues. International Union of Forest
Research Organizations, Vienna, pp. 399–410.

Serrano, J.A., 2005. La defensa del comunal y de los usos colectivos, León (1800-1936):
¿una «economía moral»? Historia Agraria 37, 431–463.

Serrano, J.A., 2014. When the enemy is the state: common lands management in north-
west Spain (1850–1936). Int. J. Commons 8 (1), 107–133.

Sieira, C., 1956. Distinta titularidad de los montes incluidos en el Catálogo de Montes de
Utilidad Pública. Su naturaleza jurídica. Revista de estudios de la vida local 86,

Siitonen, J., Ranius, T., 2015. The importance of veteran trees for Saproxylic insects. In:
Kirby, J.K., Watkins, C. (Eds.), Europe’s Changing Woods and Forests. CABI, UK, pp.

Soto, D., Herrera, A., González de Molina, M., Ortega-Santos, A., 2007. La protesta
campesina como protesta ambiental, siglos XVIII-XX. Historia Agraria 42, 277–301.

Taboada, A., Kotze, D.J., Tárrega, R., Salgado, J.M., 2006. Traditional forest manage-
ment: do carabid beetles respond to human-created vegetation structures in an oak
mosaic landscape? For. Ecol. Manag. 237, 436–449.

Vadell, E., de-Miguel, S., Pemán, J., 2016. Large-scale reforestation and afforestation
policy in Spain: ahistorical review of its underlying ecological, socioeconomic and
political dynamics. Land Use Policy 55, 37–48.

Valbuena-Carabaña, M., López, U., Fuentes-Utrilla, P., González-Doncel, I., Gil, L., 2010.
Historical and recent changes in the Spanish forests: a socio-economic process. Rev.
Palaeobot. Palynol. 162, 492–506.

Vázquez, I., 2016. Aprendizaje histórico en gestión de bienes comunales: Los pastos en
Cantabria (España). University of Cantabria, Ph.D. dissertation, pp. 192–199.

Viedma, O., Moity, N., Moreno, J.M., 2015. Changes in landscape fire-hazard during the
second half of the 20th century: agriculture abandonment and the changing role of
driving factors. Agric. Ecosyst. Environ. 207, 126–140.

Weiss, G., Lawrence, A., Hujala, T., Lidestav, G., Nichiforel, L., Nybakk, E., Quiroga, S.,
Sarvašová, Z., Suarez, C., Živojinović, I., 2018. Forest ownership changes in Europe:
state of knowledge and conceptual foundations. For. Policy Econ. 99, 9–20. https://

Wing, J.T., 2015. Roots of Empire. Forests and State Power in Early Modern Spain,
c.1500–1750. Boston, Brill, Leiden.

S. Guadilla-Sáez, et al. Forest Policy and Economics 112 (2020) 102107


  • Forest commons, traditional community ownership and ecological consequences: Insights from Spain
    • Introduction
    • Historical evolution of forest commons in Spain
      • Initial records of the commons
      • Dismantling the commons
      • State interventionism and people’s resistance
      • Replacement of traditional uses by modern forestry approaches
      • Rural landscape dynamics resulting from the abandonment of traditional forest management
    • Multiple legacies in current legal framework of Spanish forest commons
      • Forest commons
      • Partners’ woodlands
      • Neighbour woodlands
    • Conclusion
    • mk:H1_13
    • Acknowledgments
    • References

Wk 3 Discussion (Due in 1 day) Urgent/Socio-economic factors influencing land-use.pdf

Contents lists available at ScienceDirect

Forest Policy and Economics

journal homepage:

Socio-economic factors influencing land-use and land-cover changes in the
miombo woodlands of the Copperbelt province in Zambia
Ferdinand Handavua,b,⁎, Paxie W.C. Chirwaa, Stephen Syampunganic
a Postgraduate Forest Science Programme, Department of Plant and Soil Sciences, University of Pretoria, Pretoria, South Africa
b Zambia Forestry College, Kitwe, Zambia
c Department of Plant and Environmental Sciences, Copperbelt University, Kitwe, Zambia


Miombo woodland
Socio-economic factor


This study examined socio-economic factors that influence land-use and land-cover dynamics in the Copperbelt
miombo woodlands of Zambia. Data were collected through household surveys and focus group discussions. All
households that have lived in the area for 5 years and above were considered eligible to participate in the survey.
A total of 372 households and 30 discussants within a 5-km buffer zone of the forest reserves were selected for
interview. Pearson’s Chi-square tests were used to test association between independent variables (gender, age,
education, wealth status, and household size) and use of forest products. Furthermore, binary logistic regression
models were developed to examine determinants of forest products use and land-use and land-cover change. The
chi-square results revealed a strong association among the following; charcoal production with gender, age and
wealth; use of construction poles with household size; firewood collection with wealth; wild fruits collection
with gender and household size; caterpillar collection with education; honey harvesting with gender and wealth;
wild vegetable collection and use with education; use of thatching grass with wealth and household size; live-
stock fodder use with wealth and level of education; collection and use of bush meat with age and residence
status; collection of material for brooms with age and wealth respectively. The logistic regression model results
revealed that charcoal, construction poles, wild fruits and animal fodder were statistically significant at 1% level
while, honey, thatching grass and bush meat were significant at 5%. Gender, age, education, wealth status,
household size and residence status were significant determinants in the use of various forest products.
Furthermore, the regression model showed that agriculture expansion (p < .031) and population growth
(p < .032) were significant determinants of changes in forest cover. The study concludes that there is high level
of dependence on forest products by local communities and hence any attempts to avert deforestation should
consider addressing social and economic problems faced by local communities. We further conclude that de-
velopment of sustainable forest management policies and strategies that provide for inclusion of local ecological
knowledge and various utilization practices such as charcoal production into sustainable forest management.

1. Introduction

Land-use and land-cover (LULC) change is a complex socio-eco-
nomic and environmental issue that requires a comprehensive under-
standing of the interaction and relationship between human induced
activities and the environment (Brown et al., 2013). It is arguably the
most pervasive socio-economic force driving changes and degradation
of ecosystems. LULC change have become a central component in
current strategies for managing natural resource and monitoring en-
vironmental changes (Tiwari and Saxena, 2011). Studies undertaken on
land and forests indicate that fluctuating human demographic patterns
particularly population growth, its density and distribution over longer

time scales greatly influence land-use (Lambin et al., 2003), quality and
extent of forests (Ashraf et al., 2017), with immigration being con-
sidered the most important demographic component driving LULC
change (Lambin et al., 2003; Verbist et al., 2005; Yohannes et al.,
2018). It is argued that in order to improve our understanding of land-
use and land-cover relationships, there is need to link LULC change to
human actions (Turner II et al., 1995; Nurrochmat et al., 2017) and
with specific attention to the economic drivers behind these actions
(Martinez and Mollicone, 2012).

Several studies have been carried out to assess the contribution of
forest products to household income provision (Campbell et al., 1991;
Sunderlin et al., 2004; Kalaba et al., 2013; Kamwi et al., 2015) and the
Received 5 June 2017; Received in revised form 11 October 2018; Accepted 24 October 2018

⁎ Corresponding author at: Zambia Forestry College, P/B 1, Mwekera, Kitwe 10101, Zambia.
E-mail address: [email protected] (F. Handavu).

Forest Policy and Economics 100 (2019) 75–94

Available online 29 November 2018
1389-9341/ © 2018 Elsevier B.V. All rights reserved.


results showed that forest products are vital in sustaining rural house-
hold needs. Forests remain an important source of products and services
that are critical to household livelihood support and emergency safe-
guards (Angelsen et al., 2014). In line with the above, Wunder et al.
(2014) referred the contribution of forests to income and diet as the
supermarket of the wild. A wide range of studies have indicated an
important role non-timber forest products (NTFPs) play in supporting
rural household income base (Rasmussen et al., 2017). Furthermore,
Hickey et al. (2016) noted the role of wild foods in supporting food
security and nutrition. For example, it was noted that people living
around forests have greater access to forest foods such as wild fruits,
leafy greens and bush meat (Ickowitz et al., 2014; Powell et al., 2013,
2015). In addition to food, other products that are of substantial con-
tribution to household subsistence include fuelwood, medicinal plants,
construction materials, fodder and other non-cash material goods
(Sunderlin et al., 2005; Rasmussen et al., 2017). Access to these forest
products is associated with individual characteristics such as gender,
age of the household head, household size, education level and total
household income, among others (Coulibaly-Lingani et al., 2009).

Socio-economic factors have been reported to influence LULC
change. For example, Schwartz and Caro (2003) noted that socio-eco-
nomic factors alter or deplete forest cover and also alter forest structure
and species composition. Among the socio-economic factors, agri-
culture expansion (Defries and Pandey, 2010; Kamwi et al., 2015; Vu
et al., 2014., Sunderland et al., 2017), population growth (Giliba et al.,
2011; Kamwi et al., 2015; Ariti et al., 2015; Yohannes et al., 2018),
daily livelihood needs (Giliba et al., 2011), oil palm plantation estab-
lishment (Austin et al., 2017; Susanti and Maryudi, 2016) and policy
shifts and regime change (Ariti et al., 2015; Maryudi and Sahide, 2017;
Yohannes et al., 2018; Rahman et al., 2018) are reported to be among
the most critical factors driving LULC change. Furthermore, other stu-
dies noted that household size, education, period of residence, distance
to forest reserve and farmland size (Mitinje et al., 2007; Giliba et al.,
2011) influence deforestation and degradation of forest resources.
Haule (2014) argues that age structure and gender composition have
important implications on influencing the pace and/or extent of de-
forestation. On the other hand, Babulo et al. (2008) noted that factors
that condition a house’s economic reliance on forest environmental
resources may vary depending on resource endowment of the house-
hold, household’s demographic and economic characteristics. Given this
scenario, understanding demographic and economic characteristics of
the communities in a given area would provide for contribution to-
wards understanding changes in forest cover. Furthermore, the dual
role of human beings in contributing to the cause and experiencing the
cause of LULC change processes emphasises the need for better un-
derstanding of the interactions and linkages between humans and the
terrestrial environment (Berkes and Folke, 1998; Ungirwalu et al.,
2017., Nurrochmat et al., 2017, Pudyatmoko et al., 2017). Currently,
the major environmental challenge faced in Zambia is deforestation and
forest degradation. Based on the Integrated Land-use Assessment
(ILUA) 2000–2014 report (ILUA, 2017), the official annual deforesta-
tion rate for Zambia is estimated at approximately 276,021 ha per
annum, a figure poised to be one of Africa’s highest deforestation rates
per annum.

Although there have been some studies on socio-economic aspect of
forest resource use and its influence on LULC change in sub-Saharan
Africa (Tugume et al., 2015; Alelign et al., 2011; Giliba et al., 2011;
Nzunda et al., 2013), variations in ethnicity and cultural orientation
influence use of forest resources differently. Furthermore, differences in
demand-led variability in biophysical and socio-economic factors have
been reported to lead to countries experiencing different rates and
trajectories of land use (see Ashraf et al., 2017). Little is known about
the influence of socio-economic factors on LULC change in Zambia and
hence a deeper understanding of the human dimension of this phe-
nomenon is desirable. Few studies (e.g. Chileshe, 2005; Umar, 2014)
have been undertaken to analyse land tenure and rural livelihoods.

However, these studies did not critically look at socio-economic factors
that influence LULC change. Therefore, developing an understanding of
the influence of socio-economic factors on forest products use and land-
use and land-cover dynamics will contribute to developing sustainable
land-use and resource management policies. The objective of the study
was to address a number of key questions namely: (i) What are the
influences of the demographic characteristics on land-use and land-
cover change in the study area? (ii) What type of land-uses are in the
study area? (iii) What influence does a household’s socio-economic
status have on resource use? (iv) What are the major socio-economic
drivers of land-use and land-cover change?

2. Materials and methods

2.1. Description of the study sites

The study was undertaken in Miengwe and Katanino catchment
areas, both situated in Masaiti District of Zambia and located approxi-
mately 90 km, south of Ndola town along the Ndola – Kapiri-Mposhi
road. Miengwe Forest reserve lies between latitude 130 24′ 05“S and
longitude 280 49’ 00”E (Fig. 1), with a gross area of 8, 094 ha. The
forest reserve is shared by three chiefdoms namely: Mushili, Chiwala
and Nkambo. Katanino forest reserve lies between 130 25′ 00“S and 130

45’ 00”S and 280 25′ 00“E and 280 40’ 00”E at an altitude of 1200 m.
The forest has a gross area of 4, 532 ha and falls within four (4)
chiefdoms namely; Mushili, Malembeka, Nkole and Nkambo.

The forests occur at almost the boundary of Copperbelt and Central
Provinces. The forest areas occur on Katanga rock system and fall in one
agro-ecological zone (Zone III) and receive rainfall averaging 1200 mm
per annum and experience three distinct seasons based on rainfall and
temperature namely; hot dry season (September – November), hot wet
season (December – April) and the cold dry season (May – August)
(Syampungani, 2008). The rainfall is unimodal and occurs between
November and March.

2.2. Socioeconomic profile

The inhabitants of the two study areas are traditionally Lamba
speaking people, though since the 1990s, the area has seen an influx of
people from other parts of Zambia. The local communities’ main eco-
nomic activities are heavily dependent on rain-fed slash and burn
subsistence farming, charcoal production and trading (Njovu et al.,
2004). Other livelihood systems include fishing, hunting, and fruit
trade, bee keeping (Kalaba, 2005) and gardening.

2.3. Sampling design and data collection

2.3.1. Site selection
The two study sites were purposefully selected based on a set of

criteria used to represent various socio-economic conditions. First, the
study sites have experienced massive changes in land-use/land-cover
over the last thirty one years. The catchment areas have been ex-
tensively used for income sources which have influenced patterns of
forest resource use and dependence. One other reason for selecting the
areas with similar characteristics was to provide generalizable results
from close replicates to help build a cumulative knowledge base. On the
other hand, Katanino forest reserve has previously been jointly man-
aged by local communities and Government under Joint Forest
Management (JFM) scheme. The selection of the sites also provided an
opportunity to compare between human dominated landscapes and the
protected forest reserves falling under formalised management regimes.
It is our view that this selection criteria covers key components that
depict interaction between the environment (natural forest), human
capital (demographic pressure), and economic pressure (presence of
local small farmers) (see Boissiere et al., 2014).

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94


2.3.2. Sampling design
The first step involved identifying all the clusters falling within 5-

km radius from the forest reserve boundaries. A 5-km buffer was ar-
rived at to specifically capture information from households who really
depend on the forests and assessing the connectivity of communities to
the forest reserves and associated patterns of resource use. This is in line
with Obua et al. (1998) who observed that people within 5-Km range
from the forest reserve tend to interact more frequent and intense with
the forest reserve. With the help of traditional leaders, we identified 78
(40 in Miengwe and 38 in Katanino) clusters falling within the 5-km
buffer from which individual households were randomly selected for
the study. A total of 1, 488 households (776 and 712 households in
Miengwe and Katanino, respectively) were recorded from all the
identified clusters. Thereafter, random sampling using lottery method
(Singh and Masuku, 2014) was employed to select participants from
each of the cluster list at 25% sampling intensity, which is higher than
the 20% recommended by other studies (see Adhikari et al., 2004) as
the minimum size considered to be a representative sample for the
population. We conducted a household survey using a semi-structured
questionnaire in 372 selected households (194 and 178 households in
Miengwe and Katanino, respectively) (Fig. 1). All farm owners/house-
holds that have been resident in the area for 5 years and above were
eligible to participate in the survey. We interviewed household heads or
the next of kin in each household. If we did not find any eligible person
or in an event that the household refused to participate, another ran-
domly selected household was used to replace that household (see Bong
et al., 2016). The sampling unit was a household while the unit of
observation was the head of the house or next of kin. A household was

defined as a group of people living together, making common ar-
rangements for food and other essentials for survival (FAO, 2010). Prior
to undertaking the survey, the questionnaire was pre-tested on ten
households within the study area (these were exempted from the main
survey) to check for errors and ambiguity and hence improve the va-
lidity of the survey tools (see Babbie and Mouton, 2014; Barribeau
et al., 2015).

We used focus group meetings to assess a number of issues related to
community resource use, household income generating activities,
changes in forest cover and factors that drive the changes and also to
develop a community based wealth ranking criteria. According to
Morgan (1996), focus group interview is the purposeful use of inter-
action designated to generate qualitative data. He identifies three major
components of focus group discussion as (i) method devoted to data
collection, (ii) interaction as a source of data and (iii) the active role of
the researcher in creating group discussion for data collection. In the
study, we purposively identified traditional leaders, who have knowl-
edge of forest cover changes and their drivers, forest resource use and
local members’ socio-economic status. Some of the Focus Group Dis-
cussion (FGD) participants were suggested by traditional leaders during
our initial visit. In the case of Katanino study site, three of the FGD
participants had been key members of the Joint Forest Management
(JFM) committee. We desired to identify and select FGD members that
were especially knowledgeable or experienced with rural household
incomes needs, land use practices and were willing to participate and
able to communicate experiences and provide opinions in a reflective
and articulate manner. A total of 30 discussants (16 for Miengwe and 14
for Katanino, respectively) participated.

Fig. 1. Map showing the location of the study areas and participating households in each study area.

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94


2.3.3. Data collection
The two data collection methods employed was focus group dis-

cussions and administration of semi-structured questionnaire. The
questionnaire was prepared in English and individual questions were
orally translated into the local language (Bemba/Lamba) while ad-
ministering to respondents. Six trained research assistants who were
good at both English and the local language participated in the exercise.
The questionnaire included several sections covering the following
areas; general household socio-economic data (demographics, assets
and land-use) (see also Kamwi et al., 2015), level of dependence on
forest products and income generating activities. Local community
members who were familiar with the study sites were engaged to help
with locating the selected farms/households. Upon location of the
households and prior to administering the questionnaire, coordinates
were taken using a hand-held Global Positioning System (GPS) equip-
ment to ensure that only households within the 5-km buffer were

The sampled households were stratified by wealth classes generated
during focus group discussion. Proportion of wealth categories were
divided into four classes, namely: Very poor (n = 78), Poor (n = 189),
Rich (n = 74) and Very rich (n = 31) (see Appendix A). Wealth ranking
based on Participatory Rural Appraisal (PRA) technique of household
mapping as provided during Focus Group Discussion was undertaken.
The discussants were guided to align their household wealth ranking
process to include socio-economic attributes such as amount of land
cultivated, size and style of houses owned, quality of household assets,
livestock ownership, income from off-farm agricultural activities (see
Richards et al., 1999). The card game method was used for the wealth
ranking exercise to identify local criteria used to differentiate house-
holds based on wealth and well-being (see Mukherjee, 1992).

Furthermore, four (4) focus group discussion meetings (two per
study site) were carried out in both Miengwe and Katanino catchment
areas. These meetings consisted of 6–9 discussants per group (e.g.
Appendix B). The information gathered from focus group discussions
was used through triangulation technique (see also Kamwi et al., 2015)
to validate data obtained from questionnaires with focus group inter-
view data to provide in-depth understanding of community activities.
In particular, issues such as development of criteria for wealth ranking,
migration of households into study sites, and community resource use
were discussed. The meetings lasted approximately 90 min and were all
moderated by the researcher and one assistant. A voice recorder was
used to capture the sessions.

2.4. Data analysis

Quantitative data were analysed using Statistical Package for Social
Sciences (IBM SPSS) software version 23 (IBM Corp. Released, 2015.
IBM SPSS Statistics for Windows, Version 23.0. Armonk, NY: IBM Corp).
The main statistical analysis methods used to analyse the quantitative
data were both descriptive and inferential statistical methods (Giliba
et al., 2011; Pallant, 2014). Descriptive statistics such as mean values,
standard error of the mean, frequency counts and percentages of ob-
served attributes (Giliba et al., 2011) were used to explain demographic
characteristics of communities within the 5-km buffer. We examined
respondents’ perceived roles of household members in collection of
forest products. The Pearson’s Chi-square test of association was used to
show the levels of association between independent variables and fre-
quency of use of forest products. In-depth data was analysed through
content analysis. Furthermore, in order to assess the socio-economic
variables that were determinants of communities’ use of forest products
and perceived land-cover loss, a series of binary logistic regression
models were developed (see Coulibaly-Lingani et al., 2009; Kamwi
et al., 2015). The logistic regression model is a suitable statistical tool
when determining the influence of explanatory variables on the di-
chotomous dependent variables (i.e. with only two categories or va-
lues), when the former are continuous, categorical or dummy variables

(Peng et al., 2002; Coulibaly-Lingani et al., 2009). The model contained
eight (8) explanatory variables which included the following: gender,
age of respondent, education level, wealth, household size and re-
sidence status, which were introduced simultaneously to assess factors
influencing the use of forest products.

The logit is the natural logarithm (ln) of odds of Y, and odds are
ratios of probabilities (π) of Y happening to probabilities (1 − π) of Y
not happening. The logistic model is specified as:

= = + + + …+logit Y In X X X( )
1 i i k ki0 1 1 2 2

where β0 is the intercept and β1, β2 … βk are the coefficients of the
independent variables X1, X2 … Xk.

The response variables for the logistic regression model applied in
this study were various forest products that communities collect. Use of
forest products was a binary choice variable (1 = yes and 0 = other-
wise) that helped to establish whether a household collects any pro-
ducts from the forest or not. Furthermore, the response variables for the
logistic regression model on causes of land-cover change were: charcoal
production, agriculture expansion, population growth and settlements,
which were also defined as binary variables with a value 1 for re-
spondents agreeing to either of the perceived causes or 0 for otherwise.
Other causes such as lack of conservation knowledge, firewood har-
vesting and honey harvesting were not included in the logistic regres-
sion model because they were mentioned by few respondents, thereby
making them insufficient for inclusion in the model (see Kamwi et al.,
2015). Qualitative data from focus group discussions was analysed
through content analysis, whereby the discussions were objectively and
subjectively analysed (Hsieh and Shannon, 2005; Kamwi et al., 2015).

3. Results

3.1. Household demographics characteristics

Table 1 shows the household demographic characteristics of the
study areas. Results from the 372 respondents showed that 65.3%
(n = 243) were males and 34.7% (n = 129) were females. Of the in-
terviewed respondents, 56.7% (n = 211) attended grade 1–7 and 30.4%
(n = 113) grade 8–12, while only 1.6% (n = 6) attended tertiary edu-
cation. Those that never went to school accounted for 11.3% (n = 42).
With respect to their wealth status, the sampled households consisted of
71.8% poor and very poor families and only 28.2% of the households
belonged to the rich and very rich categories.

Results on ethnicity of participating household heads revealed a
total number of 24 ethnic groups. Of these, the 5 most dominant tribes
included the following; Lamba (37.4%), Bemba (16.4%), Lala (15.1%),
Tonga (8.6%) and Lenje (4.8%). The rest in the minority included the
following: Namwanga, Chewa, Ngoni, Swaka, Ushi, Tumbuka, Kaonde,
Ndebele, Lozi, Luvale, Nsenga, Bisa, Shona, Zezulu, Lunda, Soli,
Karanga, Kunda, and Lungu respectively.

Table 1
Demographic characteristics of participating households.

Demographic characteristic

Gender of household respondents Male headed 65.3%
Female headed 34.7%

Average household size 8.33 ± 0.225
Av. Land holding size 14.35 ± 18.511 ha
Av. Cultivated land size 3.66 ± 3.676 ha
Household status Indigenous 40.5%

Migrant 59.5%
Wealth categories
Very rich 8.3% (n = 31)
Rich 19.9% (n = 74)
Poor 50.8% (n = 189)
Very poor 21% (n = 78)

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94


We investigated farm sizes owned by individual land owners. The
mean farm size in the study areas was 14.35 ± 18.511 ha, with ap-
proximately 40% of the households falling in the 0–5-ha category. The
mean cultivated land size for two study sites was 3.66 ± 3.676 ha per

3.2. Factors affecting land-use change

3.2.1. Population distribution
Population distribution is influenced by migration, a complex pro-

cess driven by many factors. The study revealed that majority of the
respondents was migrants (59.5%) while local inhabitants accounted
for 40.5% of the respondents (Table 1). Forces of migration were in-
vestigated during focus group discussions. Discussants indicated that
attraction into a new area was mainly stimulated by factors such as:
good soils, availability of productive land and abundance of rainfall.
Similarly, scarcity of farmland, decreasing agricultural productivity,
erratic rainfall, limited income opportunities was reported as causes of
people leaving an area. Furthermore, it was revealed that majority of
those that migrated were informed by relatives or close associates. For
example, when asked why people have been flocking to the area, one
male discussant in Katanino site recalled the drought of the 1996/1997
farming season by saying;

“The drought of 1997 left us with shame to an extent where our
pride of being a maize belt province was no longer on our lips. We
survived entirely on a government initiative of “food for work” just
to have at least a meal for the day. The year that followed, I had no
option but relocate to this area where rainfall is still abundant.”

3.3. Farming practices and production trends

Results on the main farming practices showed that local commu-
nities engage in three (3) major practices namely: shifting cultivation
(32%, n = 119), crop rotation (29%, n = 108) and intensive mono-
cropping (26%, n = 95) (Fig. 2). The crops mainly grown in the study
areas are maize (Zea mays) (staple food), sweet potatoes (Ipomea

batatas) and groundnuts (Arachis hypogaea). In terms of production
trends, majority of respondents indicated low productivity (69.2%) for
maize, while response on sweet potatoes showed a slightly higher trend
(56.5%). Respondents attributed the low productivity for maize mainly
to loss of soil fertility, lack of farming inputs, insufficient rainfall and
non-availability of extension services. As for sweet potatoes, the lack of
impressive yield was attributed to insufficient rainfall and attack on
tubers by moles.

3.4. Sources of income and community forest resource utilization

3.4.1. Household income sources
Fig. 3 shows that sale of farm produce (58.2%) was ranked to be the

most important source of income in the study areas. Furthermore,

Fig. 2. Farming practices prominently engaged by households in the study area.







0.3 1.3 0.5

















Income sources

Fig. 3. Household income sources in the study areas.

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94


charcoal production and sale (52.5%) and wild food harvesting (45%)
were ranked to be the second and third sources of income in the study
sites. Income from family business and piecework accounted for 34.9%
and 29% respectively.

Furthermore, other sources of income included livestock production
with chickens (33.3%), chicken + goats (16.4%) and goats (7.5%)
being the most sold. However, focus group discussion results indicated
that charcoal production and sale was a major source of income in the
area. For example, some of the discussants around Katanino forest re-
serve had this to say;

“farming inputs are very expensive to afford so we treat this forest as
our reserve bank from which we draw a living through manu-
facturing and sale of charcoal”.

Discussants indicated that charcoal bags are packaged in various bag
sizes according to market points (e.g. Fig. 4). Very few households
engaged in specialised skills such as the sale of hand crafts and garden
chair production displayed (Appendix C).

Overall, pooled results on level of dependence showed that 98.1% of
the respondents indicated that they highly depend on forest for various
products. Rural household’s dependency on fuel wood for energy was
investigated and results indicated that 96% of the communities use
firewood, while only 4% use charcoal as the main household energy
source. In addition, majority of the firewood users rely on charcoal for
household supplementary energy needs.

Fig. 5 shows a total of 16 forest products frequently obtained from
the two study sites. Results showed that the highest proportion of
households engaged in mushroom (90.9%), firewood (90.9%), wild
fruits (88.7%), thatching grass (83.6%), construction poles (75.3%),
wood fibre (74.5%), medicine (72%), charcoal (68.3%), honey (59.9%)
and wild vegetables (55.4%), which constitute the ten (10) most fre-
quently extracted forest product in the study areas.

In addition, association between predictor variables (gender, age,
education, wealth status, and household size) and forest products use
was investigated. The results showed significant level of association
between some explanatory variables and use of some of the forest
products. The study revealed statistical evidence of association among
the following; gender, age and wealth with charcoal production
(X2 = 9.145, p < .002; X2 = 17.928, p < .006; X2 = 8.841,
p < .031), household size with use of construction poles (X2 = 22.831,
p < .000), wealth with firewood collection (X2 = 10.193, p < .017),
gender and household size with collection and use of wild fruits
(X2 = 10.839, p < .001; X2 = 15.966, p < .007), education with ca-
terpillar collection and use (X2 = 7.993, p < .046), gender & wealth
with wild honey harvesting (X2 = 6.345, p < .012; X2 = 9.036,
p < .029), education with wild vegetable collection and use
(X2 = 10.327, p < .016), wealth and household size with use of
thatching grass (X2 = 10.167, p < .017; X2 = 16.777, p < .005),
wealth and education with livestock fodder (X2 = 54.209, p < .000;
X2 = 21.272, p < .000), age and residence status with use of bush
meat (X2 = 21.581.928, p < .000; X2 = 4.940, p < .026); age and
wealth with collection of materials for making brooms (X2 = 35.199,
p < .000; X2 = 15.980, p < .001) (Table 2). However, no significant
association was observed between explanatory variables and use of
forest products such as medicine, roots/tubers, wood fibre and mush-

3.4.2. Determinants of forest products utilization
Several reasons exist why households with varying socio-economic

and demographic characteristics depend on forest products differently.
This may among others relate to consumption motives and response to
varying challenges households come across. In order to establish the
likelihood that socio-economic factors influence households’ use of
forest products, six independent variables (gender, age, education,
wealth status, household size and residence status) were entered in the
logistic regression model. Overall assessment of the logistic regression
model for the use of various forest products revealed positive results.
For the Hosmer-Lemeshow Goodness of Fit Test poor fit is designated by
a significance value < 0.05, so to support our models we require va-
lues > 0.05. In our study, the chi-square values showed that p-values
were larger than 0.05, implying adequate fit to the data and hence
support for the models (Appendix D).

The full models containing predictors that were statistically sig-
nificant using the Omnibus Test of Model Coefficients were seven (7).
Coefficients on charcoal [X2(16, 372) = 37.756, p < .002]; construction
poles [X2(16, 372) = 37.756, p < .001]; wild fruits [X

(16, 372) = 37.756,

p < .005]; and animal fodder [X2(16, 372) = 37.756, p < .001] were
statistically significant at 1%, while honey [X2(16, 372) = 37.756,
p < .026]; thatching grass [X2(16, 372) = 37.756, p < .023]; and bush
meat [X2(16, 372) = 37.756, p < .024] were statistically significant at
5% level (Table 3).

Fig. 4. Photos showing charcoal bags read for transportation to Ndola (a) and Lusaka (b).











Forest products

Fig. 5. Proportion (%) of households collecting various forest products.

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94


Furthermore, in terms of influence of socio-economic determinants
on use of forest products, the output of the logistic regression model
revealed that though with varying levels of dominance, gender, age of
household head, education, wealth and residence status were sig-
nificant determinants of use of specific forest products. For example,
results revealed that collection and use of timber was found to be sig-
nificantly influenced by residence status (p < .013), gender and age
had significant influence on charcoal production (p < .003, p < .002),
age class (36–55 years) had influence on use of construction poles
(p < .037), gender and household size (11–15 class) had influence on
collection and use of wild fruits (p < .001, p < .035), age had influ-
ence on collection of roots and tubers (p < .030), age groups 36–55 &
56–75 years influenced the collection and use of wood fibre (p = .040,
p < .030), secondary education level (p < .046), and wealth status of
households (p < .049) influenced the gathering of caterpillars, honey
gathering and use were influenced by gender of household head
(p < .011) and age groups 36–55 & 56–75 (p < .015), wild vegetable
gathering and use was statistically influenced by level of education
(p < .018), collection and use of thatching grass was influenced by age
groups (16–35, p < .041; 36–55, p < .031; 56–75, p < .048),

collection and use of livestock fodder was influenced by age group
(16–35, p < .038); education (p < .005) and wealth status
(p < .001), and finally household use of bush meat was influenced by
age (p < .037) and residence status (p < .018) (Appendix E).

As part of the understanding of the level of abundance of forest
products, rural household heads were asked to rank the forest products.
Table 4 shows that in Miengwe catchment area, respondents indicated
eight (8) most abundant forest products as follows; thatching grass
(86.1%), firewood (85.1%), mushroom (83.1%), wild fruits and con-
struction poles (73.7% each), wood fibre (70%) medicinal plants (67%),
charcoal (54.1%) and wild vegetables. In Katanino study site, results
revealed the following order of abundance; firewood (77%), medicinal
plants (66.3%), wild fruits (64.6%), thatching grass (60.7%), mush-
room (59.6%), charcoal (54.5%) and wood fibre (50%). Furthermore, it
was observed that there was a high level of community organisation
during mushroom gathering in Miengwe study site than in Katanino
area. This was evidenced by existence of well organised market outlets
where mushroom was sold. Furthermore, in Miengwe, communities
formed small groups where each group had specific days within the
week when they collected mushroom in Miengwe forest reserve. The
mushroom is sold in baskets, commonly known as “imiseke” (Fig. 6a) at
wholesale order price to locals and people (locally known as “Ba Kun-
gula”) from urban areas.

3.4.3. Local Ecological Knowledge (LEK) and forest conservation
We investigated existence of local ecological knowledge systems

(norms and cultural values) that help in conservation of the forest re-
sources. The pooled results showed that 82% of the respondents were
aware of local ecological practices that help to conserve the forest re-
sources. Key LEK practices established during the survey were as fol-
lows; protection of trees in grave sites (Fig. 7), no cutting of trees along
riverbanks, no cutting of fruit trees. Results from this study revealed
that protection of burial sites is one of the highly-respected practices
among communities.

3.4.4. Land clearance and perceived causes of forest cover loss
Results indicate that 79% of the interviewed participants cleared

land in the last 10 years in the study sites and the major reasons ad-
vanced were; increasing agricultural productivity and loss of soil fer-
tility. Average land cleared per household per annum was
2.1 ± 3.528 ha and the mean cultivated farmland size was
3.66 ± 3.676 ha.

Table 2
Explanatory variables and associated use of forest products.

Product Statistical evidence of levels of association

Gender Age Wealth Education Household size Residence status

2 p-value 2 p-value 2 p-value 2 p-value 2 p-value 2 p-value

Timber 1.505 0.220 4.670 0.587 2.968 0.397 3.862 0.277 4.790 0.442 7.398 0.007⁎
Charcoal 9.145 0.002⁎ 17.928 0.006⁎ 8.841 0.031⁎ 2.779 0.427 3.208 0.668 0.164 0.686
Construction poles 0.077 0.782 11.934 0.063 0.371 0.946 6.834 0.077 22.831 0.000⁎ 0.119 0.731
Firewood 1.113 0.292 8.481 0.205 10.193 0.017⁎ 4.051 0.256 4.585 0.469 0.002 0.969
Wild fruits 10.839 0.001⁎ 9.152 0.165 4.430 0.219 0.912 0.870 15.966 0.007⁎ 0.078 0.780
Medicine 1.511 0.219 6.596 0.360 6.748 0.080 2.708 0.439 8.870 0.114 2.330 0.127
Roots/tubers 1.698 0.193 10.670 0.099 4.712 0.194 0.967 0.809 1.566 0.905 1.506 0.220
Wood fibre 0.971 0.325 7.676 0.263 7.519 0.057 1.432 0.698 5.104 0.403 0.039 0.843
Caterpillars 1.492 0.222 6.940 0.326 4.100 0.251 7.993 0.046⁎ 2.998 0.700 0.306 0.580
Honey 6.345 0.012⁎ 11.218 0.082 9.036 0.029⁎ 0.631 0.889 5.326 0.377 0.001 0.980
Wild vegetables 1.001 0.317 4.838 0.565 3.115 0.374 10.327 0.016⁎ 6.169 0.290 0.001 0.971
Thatching grass 0.401 0.526 11.812 0.066 10.167 0.017⁎ 0.221 0.974 16.777 0.005⁎ 0.069 0.792
Livestock fodder 0.000 0.993 5.676 0.460 54.209 0.000⁎ 21.272 0.000⁎ 6.645 0.248 0.144 0.705
Bush meat 0.036 0.850 21.581 0.001⁎ 1.084 0.781 4.573 0.206 7.104 0.213 4.940 0.026⁎
Mushroom 0.458 0.499 8.480 0.205 1.623 0.654 1.660 0.646 4.608 0.466 2.030 0.154
Brooms 2.109 0.348 35.199 0.000⁎ 15.980 0.001⁎ 4.548 0.603 4.010 0.947 0.703 0.704

⁎ Significant at 0.05 level of confidence.

Table 3
Table showing model performance using the Omnibus Test of Model

Product Chi-Square df Sig

Timber 25.331 16 0.064
Charcoal 37.756 16 0.002⁎⁎

Construction poles 38.188 16 0.001⁎⁎

Firewood 16.040 16 0.450
Wild fruits 34.393 16 0.005⁎⁎

Medicine 21.282 16 0.168
Roots/Tubers 17.161 16 0.375
Wood fibre 21.251 16 0.169
Caterpillars 24.144 16 0.086
Honey 28.728 16 0.026⁎

Wild vegetables 25.973 16 0.054
Thatching grass 29.175 16 0.023⁎

Animal fodder 90.297 16 0.000⁎⁎

Bush meat 28.958 16 0.024⁎

Mushroom 19.426 16 0.247

⁎ Statistically significant at 5% level of significance.
⁎⁎ Indicates significance at 1% level.

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94


Communities’ perceptions on the causes of forest cover loss were
investigated in the study. Of the interviewed, 87.1% of the people noted
that there has been a decrease in forest cover over the last 10 years
(Fig. 8).

Table 5 shows that there is a general understanding that charcoal
production (77.2%) is the main cause of change in land-use and land-
cover in the study area. Agriculture expansion was ranked as the second
(25.6%) most important driver of land-use and land-cover change,
while population growth (18.3%) ranked third, respectively. Drivers

such as settlements and firewood harvesting were considered of little
influence (6.2 and 0.5% respectively).

3.5. Determinants of forest cover loss

Four logistic regression models were developed to explore the key
determinants of the likelihood that charcoal production, agricultural
expansion, settlements and population growth could influence forest
cover loss. The model contained all the six independent variables

Table 4
Percentage response on levels of abundance of forest products per study site.

Product Level of abundance (%) Miengwe Level of abundance (%) Katanino

Abundant Not abundant Not found No idea Abundant Not abundant Not found No idea

Timber 5.7 6.2 82.5 6.7 6.7 7.3 84.8 1.1
Charcoal 54.1 18.6 0.5 26.8 54.5 9.0 2.8 33.7
Construction Poles 73.7 14.4 0.0 11.9 46.6 11.2 1.1 41.0
Firewood 85.1 8.8 0.0 6.2 77.0 6.7 0.6 15.7
Wild fruits 73.7 18.0 0.5 7.7 64.6 16.3 0.6 18.5
Medicinal plants 67.0 5.7 0.0 27.3 66.3 2.8 1.1 29.8
Wild tubers/roots 35.4 14.9 1.0 49.5 21.3 10.7 18.5 49.4
Wood fibre 70.1 9.8 0.0 20.1 50.0 15.2 1.1 33.7
Caterpillars 49.0 13.4 1.5 36.1 15.7 24.2 17.4 42.7
Honey 49.0 13.4 1.5 36.1 36.0 21.9 5.1 37.1
Wild vegetables 52.1 10.3 0.0 37.6 38.8 6.2 3.9 51.1
Thatch grass 86.1 6.7 0.0 7.2 60.7 10.1 2.8 26.4
Livestock fodder 22.1 1.0 0.5 76.3 13.5 1.1 3.9 81.5
Bush meat 13.9 16.5 22.2 47.4 5.6 4.5 81.5 8.4
Mushroom 83.5 10.3 0.0 6.2 59.6 27.0 1.7 11.8
Brooms 1.0 0.0 0.0 99.0 1.1 0.0 3.4 95.5

Fig. 6. Forest products community members obtain from forests around study sites.

Fig. 7. Preserved vegetation at grave sites.

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94


(gender, age, education, wealth, household size and residence status).
Overall assessment using the Omnibus Test of Model Coefficients re-
vealed that agricultural expansion (X2(16, 372) = 28.077, p < .031) and
population growth (X2(16, 372) = 27.985, p < .032) were statistically
significant in influencing perceived changes in land-use and land-cover.
The models explained between 4.8% (Cox & Snell R2) & 7.2%
(NagelkerkeR2) and 7.2% (Cox & Snell R2) & 11.7% (NagelkerkeR2) of
the variance in land-cover status resulting from agricultural expansion
and population growth respectively (Table 6). The two models correctly
classified 75.8% and 81.5% of the cases each.

Table 7 shows that secondary education had a significant influence
(p < .020; p < .021) when engaging in charcoal production,

expanding fields and also fostering population growth. In addition,
household size was found to significantly influence agricultural ex-
pansion (p < .003) and population growth (p < .034).

4. Discussion

4.1. Demographic characteristics

Our results indicate a higher average household size than most
studies in the region. The larger the family size, the greater the po-
tential for overall population growth of the area, growing number of
people and higher population density. The high household size in our
study areas may be attributed to a number of demographic factors such
as: (a) belief in extended family systems; (b) traditional values and
belief systems that consider large family as a source of labour force for
cultivation as most families still use hand-hoeing farming methods and;
(c) in-migration at village level. A study by Bongaart (2001) revealed
that overall average household size in rural sub-Saharan Africa is 5.3
persons. Similarly, Alelign et al. (2011), Teshome et al. (2015) observed
the average household size ranging from 5.6 to 7 in some parts of
Ethiopia, while Giliba et al. (2011), Kalaba et al. (2013) and Kamwi
et al. (2015) observed the average family size of 5.0 and 6.0 in the
Bereku Forest Area of Tanzania, some parts of Copperbelt province of
Zambia and the Zambezi region of Namibia, respectively.

Household size is an important indicator of the population’s po-
tential to impact on the environment. Our study revealed that house-
hold size had influence on collection and use of wild fruits and livestock
fodder. Furthermore, the results showed significant association between
household size and collection of construction poles, wild fruits and
thatching grass. This is reinforced by existing research findings in
Ethiopia (Mamo et al., 2007), Burkina Faso (Coulibaly-Lingani et al.,
2009) and Uganda (Tugume et al., 2015) who observed a positive
correlation between dependence on forest products and household size.
This suggests that households with large families, especially those with
limited income opportunities, are more dependent on forest resources
to fulfil their basic needs (see Bhandari and Jianhua, 2017). A study by

Fig. 8. Participants’ perception on extent and state of forest cover in the past 10 years.

Table 5
Participants’ response on causes of land use and land cover change.

Perceived causes of forest cover loss Participant response

Frequency Percentage (%)

Charcoal production 287 77.2
Agricultural expansion 94 25.6
Settlements 23 6.2
Population increase 70 18.3
Firewood harvesting 2 0.5

Table 6
Factors influencing land-use and land cover change.

Dependent variable Chi-square

df Sig Cox & Snell


Charcoal production 18.159 16 0.315 0.048 0.072
Agricultural expansion 28.077 16 0.031⁎ 0.073 0.107
Settlements 11.018 16 0.808 0.029 0.079
Population growth 27.985 16 0.032⁎ 0.072 0.117

⁎ Significant at 0.05 level of confidence.

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94


Ashraf et al. (2017) indicated that demographic changes particularly
population growth, its density and distribution greatly influence the
quality of forests. Ashraf et al. further noted that a high share of rural
populations living in poverty rural communities is significantly detri-
mental to forest cover and forest condition. Coulibaly-Lingani et al.
(2009) asserted that individuals from large families may have diffi-
culties in accessing alternative sources of subsistence and thus become
inclined to use of forest resources.

Although large family size constitute a social burden in terms of
basic livelihood needs, households with many members of productive
and economically active age, use their available labour inputs to an
advantage in farming and forest products exploitation. Logistic re-
gression results indicated that household size was found to significantly
influence agriculture expansion and population growth. These results
reinforce existing research findings that as household size increases,
demand for new agricultural land grows (Pan et al., 2007). This sce-
nario creates greater potential for migration, land shortages and also an
increase in rates of deforestation. Furthermore, the results show that as
many of the individual household members reach adulthood, more
demand for resources, income and decisions for more land for sub-
sistence crop production arise, and in return affect LULC through forest
clearing. Similarly, Haule (2014) also noted that with more members of
the household within the age group of 20–45 years, the higher the
likelihood of being involved in activities that cause or accelerate

deforestation. Furthermore, a number of studies in other parts of the
developing world namely Rwanda (Nduwamungu, 2001) and Tanzania
(Madulu, 1996) observed a strong relationship between size of house-
hold and environmental degradation.

4.2. Education level

Education level of rural communities can influence households’
dependence on forest products use (Timko et al., 2010; Bhandari and
Jianhua, 2017). Analysis of educational levels in the two study sites
revealed that a significant proportion of household members ended at
primary and junior secondary school levels only. As such, levels of low
education in the areas are high. This may underline the local commu-
nities’ over dependence on forest products for livelihood due to limited
capacity to seek employment opportunities in the formal sector. Higher
education levels are associated with low dependence on forests for li-
velihoods and this is mainly because education provides a wider range
of employment opportunities in other sectors of the economy (see
Adhikari et al., 2004; Mamo et al., 2007; Tugume et al., 2015) and
generally wider asset base (Timko et al., 2010). Furthermore,
Coulibaly-Lingani et al. (2009) observed that education substantially
influenced individuals’ access to Non-Timber Forest Products (NTFPs).
Our results indicate that there were significant levels of association
between education level and use of products such as caterpillars, wild

Table 7
Socio-economic variables influencing land use and land cover change.

Dependent Independent variable B Std. Error Wald df Sig. Exp(B) 95% Confidence Interval for Exp(B)

Lower Upper

Charcoal Gender 0.513 0.282 3.315 1 0.069 1.671 0.961 2.904
Age class 0.005 0.165 0.001 1 0.976 1.005 0.727 1.388
Education – Primary 0.625 0.384 2.653 1 0.103 1.868 0.881 3.964
Education – Secondary 1.014 0.436 5.408 1 0.020⁎ 2.757 1.173 6.481
Education – Tertiary 1.094 1.180 0.860 1 0.354 2.985 0.296 30.128
Wealth category – Very rich 0.571 0.577 0.982 1 0.322 1.771 0.572 5.482
Wealth category – Rich 0.098 0.380 0.067 1 0.796 1.103 0.524 2.325
Wealth category – Poor 0.419 0.316 1.758 1 0.185 1.520 0.818 2.824
Household size class −0.014 0.144 0.010 1 0.920 0.986 0.743 1.307
Residence status −0.009 0.257 0.001 1 0.972 0.991 0.599 1.641

Agriculture expansion Gender 0.180 0.262 0.471 1 0.493 1.197 0.716 2.002
Age class 0.264 0.168 2.482 1 0.115 1.303 0.937 1.810
Education – Primary 1.006 0.520 3.741 1 0.053 2.736 0.987 7.586
Education – Secondary 1.258 0.546 5.307 1 0.021⁎ 3.519 1.207 10.264
Education – Tertiary −18.586 0.000 . 1 . 8.477E-9 8.477E-9 8.477E-9
Wealth category – Very rich −0.094 0.513 0.034 1 0.854 0.910 0.333 2.489
Wealth category – Rich −0.338 0.400 0.712 1 0.399 0.714 0.326 1.563
Wealth category – Poor −0.179 0.322 0.308 1 0.579 0.836 0.445 1.573
Household size class 0.404 0.134 9.033 1 0.003⁎ 1.498 1.151 1.950
Residence status 0.133 0.251 0.279 1 0.597 1.142 0.698 1.868

Settlements Gender −0.645 0.530 1.482 1 0.223 0.525 0.186 1.482
Age class −0.272 0.294 0.859 1 0.354 0.762 0.428 1.355
Education – Primary 0.871 1.077 0.654 1 0.419 2.390 0.289 19.739
Education – Secondary 1.160 1.104 1.104 1 0.293 3.189 0.367 27.748
Education – Tertiary −16.616 0.000 . 1 . 6.080E-8 6.080E-8 6.080E-8
Wealth category – Very rich −1.383 1.119 1.527 1 0.217 0.251 0.028 2.248
Wealth category – Rich −1.017 0.725 1.969 1 0.161 0.361 0.087 1.497
Wealth category – Poor −0.438 0.506 0.751 1 0.386 0.645 0.239 1.738
Household size class 0.142 0.249 0.326 1 0.568 1.153 0.708 1.876
Residence status 0.076 0.442 0.030 1 0.863 1.079 0.454 2.566

Population growth Gender 0.036 0.292 0.015 1 0.902 1.037 0.585 1.838
Age class 0.178 0.183 0.947 1 0.331 1.195 0.835 1.710
[Education – Primary 0.675 0.574 1.383 1 0.240 1.963 0.638 6.043
[Education – Secondary 1.294 0.594 4.739 1 0.029⁎ 3.646 1.138 11.686
[Education – Tertiary 1.148 1.258 0.833 1 0.361 3.153 0.268 37.099
Wealth category – Very rich −0.572 0.660 0.751 1 0.386 0.564 0.155 2.058
[Wealth category – Rich −0.107 0.475 0.051 1 0.821 0.898 0.354 2.280
[Wealth category – Poor 0.554 0.377 2.156 1 0.142 1.740 0.831 3.644
Household size class 0.313 0.148 4.512 1 0.034⁎ 1.368 1.025 1.827
Residence status −0.055 0.279 0.039 1 0.843 0.946 0.548 1.634

⁎ Significant at 0.05 level of confidence

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94


vegetables and collection of livestock fodder. Furthermore, our results
revealed that education level was an important determinant capable of
substantially influencing households’ patterns of use of land and also
population growth.

4.3. Population distribution

Ethnic diversity in rural areas is often strongly influenced by the
medium and long-term demographic consequences of immigration. Our
study revealed that even though the two areas were originally inhabited
by Lamba speaking people, over the years, there has been a steady in-
flux of other tribes mainly dominated by Bemba, Lala, Tonga and Lenje
speaking groups respectively. Thus the majority of the interviewed
household members had migrated into the study sites. The resulting
consequences of such migration into a new area are high population
densities and pressure on land. Our findings are in agreement with a
study by Ashraf et al. (2017), who noted that demographic changes,
particularly population growth, density and distribution have sig-
nificant influence on quality and extent of forests. Furthermore, the
increase in tribal groupings entail a rich cultural diversity and also
varying land-use practices which in the long run influence LULC
changes in the catchment areas. These findings are consistent with a
study by Lacuna-Richman (2003) who observed that ethnic mix brings
about a diversity of abilities, experiences, cultures which may be pro-
ductive and may lead to innovation and creativity. This scenario pro-
vides an insight of the potential of migration forces in fostering popu-
lation growth and influence of socio-cultural characteristics on LULC

Two forms of migration were noticed and these are rural-rural mi-
gration – mostly motivated by the rural people’s search for new farm-
land, and urban-rural migration, mainly facilitated by tough living
conditions necessitated by loss of jobs or retrenchments. Majority of
participants that migrated cited loss of employments due to Structural
Adjustment Programme (SAP) which the Zambian Government im-
plemented between 1990 and 1995. Lall et al. (2006) and Halue (2010),
in their study of population dynamics noted that the choice to migrate
involves circumstantial factors, i.e. “push factors” – that constitute the
problems, limitations or adverse situations which force people to move
out of a specific locality and also “pull factors” which attract migrants
to areas of preference.

4.4. Farming practices and production trends

The main farming practices among farmers in the study area were
crop rotation and shifting cultivation. However, it was observed that
shifting cultivation was more predominant and this could mainly be
attributed to local peoples’ customs and tradition. By virtue of its
dominance and taking short-term period of cultivation before clearing
new areas, shifting cultivation has contributed to change in land-cover
and overall forest ecosystems. This is corroborated by other studies (e.g.
Luoga, 2000; Mwampamba, 2009) who noted that shifting cultivation
alone contributes > 50% of deforestation in Tanzania. Furthermore,
effects of shifting cultivation on forest ecosystems are exuberated by
high population densities.

The fact that production trends (in an area predominantly under
traditional cultivation systems) for the main crop have shown low
productivity signifies that communities have to look for sustenance
from off-farm activities such as charcoal production, which causes
changes in land-use and land-cover of forest ecosystems. Therefore,
better methods of farming need to be explored. Notwithstanding, the
cultivated land size for majority of households was too small to ade-
quately satisfy basic household needs, especially with the low tech-
nology agriculture production in the study areas. The small size of
cultivated land per household may be attributed to inadequate capital
to purchase farm implements and agricultural inputs (see Central
Statistical Office (CSO), 2010). Mostly, households depend on hand

tools such as hoes and axes to prepare and till their land. According to
Arega et al. (2013) the use of such outdated production technologies
characterizes most rural agricultural activities in rural Africa and makes
it difficult to do much at household level. Most interviewed households
indicated low yields for the main crop (i.e. maize) hence placing most
of households in a relatively vulnerable situation. They attributed low
yield to the following; lack agricultural inputs, labour and limited
agricultural extension services. This may explain why most of the in-
terviewed households indicated the production trends show low pro-
ductivity across the two study areas. When agricultural activities fail to
make households earn sufficient living from farm production alone,
they look for supplementary non/off-farm income generating activities
(Yizegaw et al., 2015). This may be the reason why the community in
the study areas have developed alternative livelihood strategies such as
charcoal production, collection of wild fruits and mushroom for sale as
a way of supplementing their failing agricultural activities.

4.5. Household income and community forest resource utilization

Our study revealed that rural communities living around Miengwe
and Katanino forests have a strong attachment to forest products and
services. The long-term contribution of forest resources to household
income and subsistence use of products has been widely acknowledged
as noteworthy (Levang et al., 2005; Sunderlin et al., 2005; Rasmussen
et al., 2017). For example, Shimizu (2006); Vedeld et al. (2007);
Wunder et al. (2014) noted the safety-net role of forest resources in the
lives of rural poor in improving livelihoods and quality of life. Simi-
larly, Hickey et al. (2016) and Broegaard et al. (2017) indicate that in
addition to agricultural crops, large proportions of rural populations
continue to rely on forests and other habitats in order to secure ade-
quate food and nutritionally balanced diets for their families. In our
study, respondents noted that mushrooms, wild fruits and firewood
were the most important and used subsistence forest products. Rural
households’ continued dependence on forest products for income gen-
eration implies that forest products form part of the income sources for
the rural communities. Likewise, in Nigeria, Adetola and Adetoro
(2014) show that local communities encroach into conservation areas
to obtain forest products for sustenance and income generating activ-
ities. In this study, we observed that the choice of forest products is
influenced by the product’s potential as a source of income. The ma-
jority of the discussants indicated that charcoal production is the most
reliable off-farm income-generating activity among the rural poor. For
example, one female discussant had this to say:

“Charcoal sale in the city of Lusaka currently is a lucrative business
venture as such it doesn’t take long for it to be bought”

The major implication of this finding is that it demonstrates that
charcoal production and sale is central to local communities’ income
base in the miombo eco-region (see CHAPOPSA, 2002). Therefore, the
realisation of the importance of miombo woodlands to well-being of
rural communities is key to integrating household needs and environ-
mental security into sustainable forest management planning that is
cognisant of incorporated disturbances such as charcoal production (see
Syampungani, 2008).

The age of household head plays a significant role in determining
income generating activities. Our study revealed significant levels of
association between age and charcoal production, bush meat hunting
and collection of brooms. Similar findings were noted by Kalaba et al.
(2013) who reported that labour-demanding activities such as charcoal
production were more commonly practiced by young men. Wealth
differentiation in the use of forest products was observed in the study
areas. Households belonging to the “wealthier” income groups, espe-
cially the Tonga migrants, who are culturally cattle keepers extract and
use the forests as a source of fodder for livestock. Similar studies by
Kalaba et al. (2013), observed a significantly greater proportion of
wealthy households using the forest for animal fodder. This association

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94


is likely due to the households’ ability and capacity to purchase and rear
animals mainly cattle.

4.6. Local ecological knowledge and implications for forest conservation

Religious beliefs, traditional beliefs, cultural values and practices
have been known to play a crucial role for the successful conservation
of the environment and specific organisms especially in the developing
countries (Lingard et al., 2003). Considering the future scenarios that
global warming presents, traditional knowledge systems, through cul-
tural and religious belief system can significantly contribute to ecolo-
gical recovery and hence a build-up of C stocks. Our study revealed that
protection of burial sites was one of the highly-respected practices
among communities. The communities never extract or collect forest
products from burial sites and therefore, such areas are mostly intact
(see Fig. 9). Traditional knowledge and practices have been observed to
contribute positively to protection of forests not only in the Miombo
woodlands (Luoga et al., 2000a) but many other African vegetation
formations (Lingard et al., 2003; WWF (World-Wide Fund for Nature),
2006). This respect and subsequent preservation of certain species in-
dicates that some indigenous practices have a positive impact on forest
conservation. Results of this study show that traditional ecological
knowledge plays a critical role in resource management and hence the
need for inclusion in long-term planning. According to Ngara and
Mangizyo (2013), indigenous and local knowledge system is embedded
in a context of values and social conventions, ethical principles, re-
ligious beliefs, ritual taboos, customs and other cultural practices
among local community members, which if abrogated may result in bad
omen. For example, Msuya and Kideghesho (2009) reported the posi-
tive impact of traditional practice on protection of medicinal plant
species in burial sites. These social values and cultural norms regarding
sacred places command high respect among many African societies (Saj
et al., 2006) and hence the existing potential for contributing towards
reduced forest degradation and mitigating effects of climate change.

4.7. Community perception on cause of forest cover loss

It has been noted that local community’s perception is that forests
cover has decreased in the study areas in the past 10 years. Overall,
respondents ranked charcoal production as the most predominant cause
of forest cover loss. This was followed by agriculture expansion and
population increase. A number of factors such as ready market and
poverty may have contributed to the dominance of charcoal production
among other forest use practices. The influence of ready market on
charcoal production has also been reported in Tanzania (see Luoga
et al., 2000b). The absence of other alternative sustainable energy
sources is a major driver of charcoal production in the Southern African
savanna (Gumbo et al., 2013). Additionally, land-use and land-cover
are aggravated by the fact that rural households have to cope with both
poverty and variability in agriculture and forestry sectors which are
impacted by the changes in climate (Kamwi et al., 2015).

On the other hand, results from logistic regression analysis revealed
that agriculture expansion and population growth were key determi-
nants of perceived changes in land-use and land-cover. These findings
are consistent with studies by Defries et al. (2010) and Kamwi et al.
(2015), who noted that agriculture is the main causes of dwindling
forest cover, mainly due to its high responsiveness to realities of life
such as communities’ everyday need for food and the desire to pursue
this need against all odds. In this study, we observed that the real cause
of forest cover loss is poverty, which permeates many rural households
in the two study areas and is deeply rooted in the daily needs of the
rural communities. Similar observations were made by Aung et al.
(2012) and Miyamoto et al. (2014) who, reported that poverty levels

are the principle variable influencing forest product extraction and
cover change.

5. Conclusions and recommendations

Our research has contributed to the wider body of knowledge on the
role of forests in rural household income provision from sale of products
and livelihood support through subsistence use of forest products. Our
analysis therefore confirm that forests play a critical role in rural live-
lihoods. The results showed that use of forest products is associated
with individual household characteristics. For example, among the
eight variables examined, gender, age, education, wealth status,
household size and residence status were found to be significant de-
terminants in the use of various forest products. These findings un-
derscore the need to critically understand the relationship between
household socio-economic factors and local forest utilization attributes
for better forest management, policy and decision-making processes.
The major policy implication of our finding is that miombo woodland is
an importance resource in supporting household needs of rural com-
munities and that practices such as charcoal production cannot be
stopped or excluded. On the other hand, logistic regression model
showed that agriculture expansion and population growth were found
to be significant determinants of changes in land-use and land-cover.
Therefore, forest policies should promote the integration of these uti-
lization practices in forest or woodland management. Additionally, any
attempts to avert deforestation should consider addressing social and
economic problems faced by local communities. The study also revealed
that local ecological knowledge systems (cultural and religious belief
systems) can significantly contribute to ecological recovery of forest

From this study, we make two key recommendations namely: (i) the
interrelationships between socio-economic factors and local forest uti-
lization attributes should be contextualised critically by defensible and
quantitative data, if policy strategies on forest conservation are to
provide for sustainable land-use and land-cover management; (ii) in-
tegrating household needs and environmental security into policy fra-
meworks that provide for inclusion of various utilization practices such
as charcoal production into sustainable forest management; (iii) in-
tegrating local ecological knowledge into strategic land use planning
and policy formulation processes.

Conflict of interest

The authors declare no conflict of interest.


The authors are grateful to the African Forest Forum, South African
Forestry Company Limited (SAFCOL) and the University of Pretoria for
funding the first author to undertake this study. Additionally, National
Science and Technology Council, through Zambia-Mozambique bi-
lateral Agreement has also funded fieldwork. Special thanks go to the
following people: Nalukui Matakala, Mulako Muimui, Cassius Mweele,
Joseph Chikasha, Dimus Change, Lutangu Kalubi, Alice Mukuka,
Chrispin Masanduko and Louis Sikalumbi for assisting in data collec-
tion. Gratitude also go to Benson Kabungo for his assistance with map
development and other GIS/Remote Sensing related works. We owe
much gratitude to Chief Nkambo (Late) and the farmers surrounding
Miengwe and Katanino forest reserves for their hospitality and agreeing
to the interviews. We thank the special issue editors and the two
anonymous reviewers for their constructive critique and suggestion on
the manuscript.

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94


Appendix A. Appendix

Appendix A
: Wealth ranking by communities in the Katanino-Miengwe study area.

Wealth Indicator Wealth category

Very rich Rich household Poor household Very poor house-

Land ownership Owns land Owns Land Owns land Owns land
Quality of house Burnt bricks, cement floor & iron roofing

Unburnt bricks, Cement floor & iron sheets House – unburnt bricks

with grass thatch
Walls- poles/mud &
grass thatched

Water sources Borehole/protected well protected well Open shallow well Open shallow well,

Cultivation Large field – above 5 ha Large field – 2 to 5 ha Field – 1-1.5 Small fields 1 lima-
1 ha

Means of cultivation Uses mechanical power – Tractor & animal
drawn ploughs, Hired labour

Drought power as farm implements, temporal workers
as hired labour

Uses Oxen or Hand hoes,

Only uses hand

Productivity Above 100 bags 50–100 × 50 kg bags 10–50 bags (50 kg) 0–10 bags (50 kg)
Livestock Cattle (5& above), Goats, Pigs, Sheep,

Chicken, Other fowls
Cattle (1–5), Goats, Pigs, Chicken (30 and above) Goats, chicken (10−20) Chicken (1−10)

Source of livelihood Farm products, Farm products, Off-farm activities & other business

Farm products & charcoal


Source of income Farm produce, off-farm activities & other

Sell of farm produce,business from hire of farming
equipment, Vehicle, Hammer mills

Sell of farm produce, gar-


Assets Vehicle, Tractor, Genset Hammer mill, Ox-cat, Motor cycle, Bicycle, Solar,
Generator, cell phone

Bicycle, radio, Television,
Solar, cell phone

May/not have a bi-
cycle, Radio

Family Education Support children’s education up to University Support children up to col-

No support

Appendix B. Focus group members undertaking mapping of households prior to wealth ranking

Appendix C. Rare income generating skills providing income sources for some of the locals

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94


D: Table showing results of the Hosmer-Lemeshow Goodness of Fit Test & model summary of Cox & Snell R Square and Nagelkerke R Square.

Hosmer and Lemeshow Test Cox & Snell R square Nagelkerke R square

Product Chi-Square df Sig

Timber 3.704 8 0.883 0.066 0.130
Charcoal 5.270 8 0.728 0.097 0.135
Construction poles 4.013 8 0.856 0.098 0.145
Firewood 8.712 8 0.367 0.042 0.092
Wild fruits 9.423 8 0.308 0.088 0.175
Medicine 8.350 8 0.400 0.056 0.080
Roots/Tubers 12.047 8 0.149 0.045 0.061
Wood fibre 7.437 8 0.490 0.056 0.082
Caterpillars 4.164 8 0.842 0.063 0.087
Honey 8.736 8 0.365 0.074 0.100
Wild vegetables 9.732 8 0.284 0.067 0.090
Thatching grass 8.994 8 0.343 0.075 0.128
Animal fodder 5.927 8 0.655 0.216 0.344
Bushmeat 9.155 8 0.329 0.075 0.117
Mushroom 10.576 8 0.227 0.051 0.111

E: Socio-economic determinants influencing households’ use of forest products.

Dependable variable Independent variables β S.E Wald df Sig Exp (β) 95% C.I.for EXP(B)

Lower Upper

Timber Gender 0.371 0.388 0.915 1 0.339 1.449 0.678 3.100
Age 3.803 3 0.283
16–35 0.671 0.671 1.002 1 0.317 1.956 0.526 7.281
36–55 1.174 0.675 3.030 1 0.082 3.236 0.863 12.140
56–75 1.024 0.904 1.282 1 0.258 2.784 0.473 16.384
Education 2.821 3 0.420
Primary −0.104 0.377 0.075 1 0.784 0.902 0.430 1.889
Secondary −0.088 1.238 0.005 1 0.943 0.916 0.081 10.360
Tertiary −1.767 1.054 2.809 1 0.094 0.171 0.022 1.349
Wealth status 2.729 3 0.435
Very rich −0.534 0.634 0.711 1 0.399 0.586 0.169 2.030
Rich −0.405 0.577 0.493 1 0.483 0.667 0.215 2.066
Poor −1.114 0.713 2.437 1 0.119 0.328 0.081 1.329
Household size 2.291 5 0.808
≤5 −0.421 0.435 0.935 1 0.334 0.657 0.280 1.540
6–10 0.178 0.507 0.124 1 0.725 1.195 0.443 3.226
11–15 0.160 0.678 0.055 1 0.814 1.173 0.310 4.433
16–20 −19.772 19,775.386 0.000 1 0.999 0.000 0.000 .
Residence status 0.998 0.404 6.109 1 0.013⁎ 2.714 1.230 5.990

Charcoal Gender −0.808 0.268 9.097 1 0.003⁎ 0.446 0.264 0.754
Age 14.907 3 0.002⁎
16-35 −0.916 0.435 4.432 1 0.035⁎ 0.400 0.171 0.939
36–55 −1.438 0.449 10.276 1 0.001⁎ 0.237 0.099 0.572
56–75 −1.986 0.605 10.777 1 0.001⁎ 0.137 0.042 0.449
Education 1.623 3 0.654
Primary 0.075 0.272 0.076 1 0.783 1.078 0.633 1.835
Secondary −0.168 0.899 0.035 1 0.851 0.845 0.145 4.926
Tertiary −0.440 0.386 1.300 1 0.254 0.644 0.302 1.372
Wealth status 5.549 3 0.136
Very rich 0.265 0.471 0.316 1 0.574 1.303 0.518 3.277
Rich 0.764 0.442 2.996 1 0.083 2.147 0.904 5.102
Poor 0.918 0.495 3.444 1 0.063 2.505 0.950 6.609
Household size 1.303 5 0.935
≤5 0.123 0.295 0.173 1 0.677 1.131 0.634 2.017
6–10 0.418 0.400 1.094 1 0.296 1.519 0.694 3.324
11–15 0.180 0.515 0.122 1 0.727 1.197 0.436 3.288
16–20 0.677 1.242 0.297 1 0.586 1.968 0.172 22.474
Residence status 0.130 0.243 0.287 1 0.592 1.139 0.708 1.833

Construction poles Gender 0.136 0.274 0.247 1 0.620 1.146 0.670 1.961
Age 5.734 3 0.125
16–35 −0.412 0.445 0.858 1 0.354 0.662 0.277 1.584
36–55 −0.945 0.453 4.355 1 0.037⁎ 0.389 0.160 0.944
56–75 −0.675 0.673 1.004 1 0.316 0.509 0.136 1.906
Education 5.598 3 0.133

(continued on next page)

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94


Appendix (continued)

Dependable variable Independent variables β S.E Wald df Sig Exp (β) 95% C.I.for EXP(B)

Lower Upper

Primary −0.488 0.280 3.023 1 0.082 0.614 0.354 1.064
Secondary 19.743 16,175.5 0.000 1 0.999 375,312,938.8 0.000 .
Tertiary 0.615 0.493 1.562 1 0.211 1.851 0.705 4.859
Wealth status 1.581 3 0.664
Very rich 0.312 0.546 0.327 1 0.567 1.366 0.469 3.982
Rich 0.052 0.494 0.011 1 0.916 1.053 0.400 2.773
Poor 0.438 0.560 0.612 1 0.434 1.550 0.517 4.645
Household size 6.701 5 0.244
≤5 0.369 0.317 1.350 1 0.245 1.446 0.776 2.693
6–10 −0.524 0.390 1.806 1 0.179 0.592 0.275 1.272
11–15 0.101 0.564 0.032 1 0.858 1.106 0.366 3.341
16–20 −21.918 20,073.513 0.000 1 0.999 0.000 0.000 .
21–25 −22.572 40,192.970 0.000 1 1.000 0.000 0.000 .
Residence status 0.202 0.262 0.595 1 0.441 1.224 0.732 2.047

Firewood Gender −0.555 0.427 1.693 1 0.193 0.574 0.249 1.325
Age 3.353 3 0.340
16–35 −0.767 0.789 0.944 1 0.331 0.465 0.099 2.181
36–55 −1.106 0.794 1.939 1 0.164 0.331 0.070 1.569
56–75 0.289 1.305 0.049 1 0.825 1.335 0.103 17.235
Education 3.412 3 0.332
Primary 0.271 0.461 0.345 1 0.557 1.311 0.531 3.237
Secondary 18.598 15,981.009 0.000 1 0.999 119,342,309.8 0.000 .
Tertiary −0.822 0.526 2.447 1 0.118 0.439 0.157 1.231
Wealth status 0.964 3 0.810
Very rich 0.452 0.815 0.307 1 0.580 1.571 0.318 7.762
Rich 0.243 0.729 0.111 1 0.739 1.275 0.306 5.325
Poor −0.555 0.427 1.693 1 0.193 0.574 0.249 1.325
Household size −0.092 0.776 0.014 1 0.905 0.912 0.199 4.176
≤5 3.545 5 0.617
6–10 0.248 0.478 0.270 1 0.603 1.282 0.503 3.269
11–15 −0.338 0.578 0.343 1 0.558 0.713 0.230 2.212
16–20 −0.719 0.708 1.031 1 0.310 0.487 0.121 1.953
21–25 −1.133 1.254 0.816 1 0.366 0.322 0.028 3.763
Residence status 18.782 40,192.97 0.000 1 1.000 143,593,670.7 0.000 .

Wild fruits Gender −1.771 0.553 10.266 1 0.001⁎ 0.170 0.058 0.503
Age 4.661 3 0.198
16–35 −0.215 0.685 0.098 1 0.754 0.807 0.211 3.086
36–55 −0.872 0.678 1.654 1 0.198 0.418 0.111 1.579
56–75 −1.226 0.853 2.063 1 0.151 0.294 0.055 1.563
Education 1.073 3 0.784
Primary −0.206 0.400 0.265 1 0.606 0.814 0.371 1.783
Secondary −0.684 1.258 0.296 1 0.587 0.505 0.043 5.937
Tertiary −0.512 0.568 0.812 1 0.368 0.599 0.197 1.826
Wealth status 0.387 3 0.943
Very rich 0.071 0.748 0.009 1 0.924 1.074 0.248 4.650
Rich −0.219 0.690 0.101 1 0.751 0.803 0.208 3.108
Poor −0.174 0.751 0.054 1 0.816 0.840 0.193 3.657
Household size 7.203 5 0.206
≤5 −0.124 0.473 0.069 1 0.793 0.883 0.349 2.233
6–10 −0.818 0.561 2.125 1 0.145 0.441 0.147 1.325
11–15 −1.399 0.664 4.436 1 0.035⁎ 0.247 0.067 0.907
16–20 −1.443 1.313 1.207 1 0.272 0.236 0.018 3.099
21–25 −25.395 40,192.97 0.000 1 0.999 0.000 0.000 .
Residence status 0.088 0.360 0.059 1 0.807 1.092 0.539 2.213

Medicine Gender −0.258 0.263 0.964 1 0.326 0.773 0.462 1.293
Age 2.429 3 0.488
16–35 −0.163 0.395 0.171 1 0.679 0.849 0.392 1.842
36–55 −0.519 0.404 1.651 1 0.199 0.595 0.269 1.314
56–75 −0.417 0.587 0.503 1 0.478 0.659 0.209 2.083
Education 2.153 3 0.541
Primary −0.380 0.271 1.965 1 0.161 0.684 0.402 1.163
Secondary −0.495 0.954 0.269 1 0.604 0.609 0.094 3.954
Tertiary −0.251 0.398 0.399 1 0.528 0.778 0.357 1.696
Wealth status 3.155 3 0.368
Very rich 0.738 0.501 2.166 1 0.141 2.092 0.783 5.588
Rich 0.230 0.450 0.261 1 0.610 1.258 0.521 3.039
Poor 0.133 0.492 0.073 1 0.786 1.143 0.436 2.996
Household size 6.240 5 0.284
≤5 0.404 0.292 1.914 1 0.167 1.498 0.845 2.654
6–10 0.247 0.394 0.392 1 0.531 1.280 0.591 2.773
11–15 −0.626 0.499 1.573 1 0.210 0.535 0.201 1.422
16–20 −0.645 1.061 0.369 1 0.543 0.525 0.066 4.198
21–25 19.957 40,192.97 0.000 1 1.000 464,681,699.5 0.000 .

(continued on next page)

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94


Appendix (continued)

Dependable variable Independent variables β S.E Wald df Sig Exp (β) 95% C.I.for EXP(B)

Lower Upper

Residence status −0.292 0.251 1.357 1 0.244 0.747 0.457 1.221
Root tubers Gender −0.306 0.231 1.752 1 0.186 0.737 0.469 1.158

Age 8.924 3 0.030⁎
16-35 −0.762 0.332 5.272 1 0.022⁎ 0.467 0.244 0.894
36–55 −0.930 0.354 6.884 1 0.009⁎ 0.395 0.197 0.790
56–75 −1.345 0.562 5.730 1 0.017⁎ 0.261 0.087 0.784
Education 1.558 3 0.669
Primary 0.053 0.249 0.046 1 0.830 1.055 0.648 1.717
Secondary 0.443 0.886 0.250 1 0.617 1.558 0.274 8.844
Tertiary 0.425 0.362 1.382 1 0.240 1.530 0.753 3.107
Wealth status 0.873 3 0.832
Very rich −0.112 0.455 0.061 1 0.805 0.894 0.367 2.179
Rich −0.311 0.426 0.535 1 0.465 0.732 0.318 1.688
Poor −0.307 0.467 0.434 1 0.510 0.735 0.295 1.836
Household size 0.606 5 0.988
≤5 0.060 0.269 0.050 1 0.823 1.062 0.626 1.801
6–10 −0.003 0.363 0.000 1 0.993 0.997 0.490 2.030
11–15 0.268 0.489 0.299 1 0.584 1.307 0.501 3.408
16–20 −0.559 1.193 0.220 1 0.639 0.572 0.055 5.928
21–25 −20.713 40,192.97 0.000 1 1.000 0.000 0.000 .
Residence status 0.270 0.224 1.453 1 0.228 1.310 0.844 2.032

Wood fibre Gender −0.283 0.270 1.100 1 0.294 0.754 0.444 1.278
Age 5.577 3 0.134
16–35 −0.842 0.462 3.326 1 0.068 0.431 0.174 1.065
36–55 −0.971 0.472 4.221 1 0.040⁎ 0.379 0.150 0.956
56–75 −1.351 0.624 4.695 1 0.030⁎ 0.259 0.076 0.879
Education 3.345 3 0.341
Primary −0.430 0.277 2.402 1 0.121 0.651 0.378 1.120
Secondary −0.927 0.957 0.938 1 0.333 0.396 0.061 2.583
Tertiary 0.099 0.422 0.055 1 0.814 1.104 0.483 2.527
Wealth status 5.469 3 0.141
Very rich −0.509 0.603 0.711 1 0.399 0.601 0.184 1.962
Rich −1.096 0.565 3.762 1 0.052 0.334 0.110 1.012
Poor −0.870 0.606 2.062 1 0.151 0.419 0.128 1.374
Household size 1.880 5 0.866
≤5 0.277 0.306 0.825 1 0.364 1.320 0.725 2.402
6–10 −0.054 0.393 0.019 1 0.890 0.947 0.438 2.047
11–15 −0.236 0.539 0.192 1 0.661 0.790 0.275 2.270
16–20 0.247 1.201 0.042 1 0.837 1.280 0.122 13.477
21–25 −22.516 40,192.97 0.000 1 1.000 0.000 0.000 .
Residence status −0.061 0.255 0.058 1 0.810 0.940 0.570 1.550

Caterpillar Gender −0.243 0.240 1.027 1 0.311 0.784 0.490 1.255
Age 1.440 3 0.696
16–35 0.204 0.355 0.329 1 0.566 1.226 0.611 2.459
36–55 −0.117 0.387 0.091 1 0.762 0.890 0.417 1.898
56–75 −0.020 0.565 0.001 1 0.971 0.980 0.323 2.968
Education 6.153 3 0.104
Primary −0.221 0.265 0.693 1 0.405 0.802 0.477 1.349
Secondary 2.329 1.165 3.995 1 0.046⁎ 10.273 1.046 100.852
Tertiary 0.363 0.375 0.938 1 0.333 1.438 0.689 3.000
Wealth status 7.881 3 0.049⁎
Very rich −0.019 0.477 0.002 1 0.969 0.981 0.386 2.498
Rich −0.045 0.445 0.010 1 0.920 0.956 0.400 2.287
Poor −0.956 0.515 3.445 1 0.063 0.384 0.140 1.055
Household size 1.630 5 0.898
≤5 −0.112 0.287 0.152 1 0.697 0.894 0.510 1.569
6–10 0.066 0.375 0.031 1 0.859 1.069 0.512 2.229
11–15 0.450 0.495 0.828 1 0.363 1.569 0.595 4.135
16–20 −0.231 1.210 0.037 1 0.848 0.793 0.074 8.503
21–25 −19.992 40,192.97 0.000 1 1.000 0.000 0.000 .
Residence status 0.140 0.235 0.358 1 0.550 1.151 0.726 1.823

Honey Gender 0.601 0.235 6.518 1 0.011⁎ 1.824 1.150 2.893
Age 10.472 3 0.015⁎
16-35 −0.676 0.366 3.414 1 0.065 0.509 0.248 1.042
36–55 −0.846 0.382 4.916 1 0.027⁎ 0.429 0.203 0.906
56–75 −1.793 0.574 9.765 1 0.002⁎ 0.167 0.054 0.513
Education 1.057 3 0.788
Primary −0.197 0.255 0.597 1 0.440 0.821 0.498 1.353
Secondary 0.190 1.027 0.034 1 0.853 1.209 0.162 9.051
Tertiary 0.167 0.369 0.206 1 0.650 1.182 0.574 2.436
Wealth status 4.974 3 0.174
Very rich −0.135 0.498 0.073 1 0.787 0.874 0.329 2.322
Rich −0.418 0.464 0.814 1 0.367 0.658 0.265 1.633

(continued on next page)

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94


Appendix (continued)

Dependable variable Independent variables β S.E Wald df Sig Exp (β) 95% C.I.for EXP(B)

Lower Upper

Poor −0.849 0.500 2.880 1 0.090 0.428 0.160 1.141
Household size 1.575 5 0.904
≤5 0.334 0.273 1.496 1 0.221 1.396 0.818 2.384
6–10 0.254 0.365 0.484 1 0.487 1.289 0.630 2.638
11–15 0.126 0.501 0.063 1 0.802 1.134 0.425 3.027
16–20 0.059 1.078 0.003 1 0.957 1.060 0.128 8.771
21–25 −20.639 40,192.970 0.000 1 1.000 0.000 0.000 .
Residence status 0.033 0.228 0.021 1 0.884 1.034 0.661 1.617

Wild vegetables Gender −0.152 0.234 0.421 1 0.516 0.859 0.543 1.360
Age 3.668 3 0.300
16–35 −0.425 0.345 1.520 1 0.218 0.654 0.332 1.285
36–55 −0.605 0.362 2.794 1 0.095 0.546 0.269 1.110
56–75 −0.874 0.552 2.506 1 0.113 0.417 0.141 1.231
Education 10.068 3 0.018⁎
Primary −0.083 0.245 0.116 1 0.733 0.920 0.569 1.486
Secondary −0.155 0.890 0.030 1 0.862 0.856 0.150 4.904
Tertiary 1.276 0.421 9.205 1 0.002⁎ 3.583 1.571 8.170
Wealth status 3.659 3 0.301
Very rich 0.338 0.460 0.541 1 0.462 1.402 0.569 3.455
Rich −0.216 0.424 0.258 1 0.611 0.806 0.351 1.852
Poor −0.221 0.466 0.225 1 0.635 0.802 0.322 1.999
Household size 3.728 5 0.589
≤5 0.006 0.271 0.001 1 0.982 1.006 0.592 1.711
6–10 −0.508 0.361 1.973 1 0.160 0.602 0.297 1.222
11–15 0.046 0.492 0.009 1 0.926 1.047 0.399 2.747
16–20 −1.185 1.194 0.986 1 0.321 0.306 0.029 3.172
21–25 −21.385 40,192.97 0.000 1 1.000 0.000 0.000 .
Residence status 0.039 0.224 0.030 1 0.863 1.039 0.670 1.612

Thatching grass Gender −0.304 0.324 0.881 1 0.348 0.738 0.390 1.393
Age 4.936 3 0.177
16–35 −1.568 0.768 4.165 1 0.041⁎ 0.209 0.046 0.940
36–55 −1.672 0.773 4.677 1 0.031⁎ 0.188 0.041 0.855
56–75 −1.803 0.914 3.894 1 0.048⁎ 0.165 0.027 0.988
Education 0.094 3 0.993
Primary 0.094 0.343 0.076 1 0.783 1.099 0.561 2.155
Secondary 0.092 1.239 0.006 1 0.941 1.097 0.097 12.428
Tertiary 0.090 0.490 0.034 1 0.854 1.094 0.419 2.858
Wealth status 4.529 3 0.210
Very rich 1.239 0.650 3.630 1 0.057 3.453 0.965 12.353
Rich 0.367 0.535 0.471 1 0.493 1.444 0.506 4.124
Poor 0.247 0.589 0.176 1 0.675 1.280 0.403 4.065
Household size 9.306 5 0.097
≤5 0.622 0.377 2.727 1 0.099 1.863 0.890 3.900
6–10 −0.350 0.436 0.643 1 0.422 0.705 0.300 1.657
11–15 −0.260 0.578 0.202 1 0.653 0.771 0.248 2.394
16–20 −1.457 1.076 1.833 1 0.176 0.233 0.028 1.920
21–25 −22.596 40,192.970 0.000 1 1.000 0.000 0.000 .
Residence status −0.013 0.306 0.002 1 0.967 0.988 0.542 1.799

Fodder Gender −0.289 0.331 0.763 1 0.382 0.749 0.391 1.433
Age 5.105 3 0.164
16–35 −0.921 0.445 4.284 1 0.038⁎ 0.398 0.167 0.952
36–55 −0.520 0.487 1.139 1 0.286 0.595 0.229 1.544
56–75 −0.099 0.680 0.021 1 0.884 0.906 0.239 3.431
Education 12.866 3 0.005⁎
Primary 0.887 0.347 6.537 1 0.011⁎ 2.427 1.230 4.790
Secondary 1.774 1.013 3.069 1 0.080 5.894 0.810 42.886
Tertiary 1.427 0.490 8.475 1 0.004⁎ 4.168 1.594 10.895
Wealth status 49.297 3 0.000⁎
Very rich −0.808 0.485 2.777 1 0.096 0.446 0.172 1.153
Rich −2.726 0.504 29.226 1 0.000⁎ 0.065 0.024 0.176
Poor −3.385 0.662 26.132 1 0.000⁎ 0.034 0.009 0.124
Household size 0.609 5 0.988
≤5 0.232 0.403 0.330 1 0.566 1.261 0.572 2.780
6–10 0.126 0.542 0.054 1 0.816 1.134 0.392 3.281
11–15 0.448 0.632 0.502 1 0.479 1.565 0.453 5.400
16–20 −19.160 19,170.513 0.000 1 0.999 0.000 0.000 .
21–25 −17.529 40,192.970 0.000 1 1.000 0.000 0.000 .
Residence status −0.114 0.319 0.129 1 0.720 0.892 0.477 1.667

Bush Meat Gender 0.053 0.293 0.033 1 0.856 1.055 0.594 1.873
Age 8.472 3 0.037⁎
16-35 −0.713 0.365 3.809 1 0.051 0.490 0.240 1.003
36–55 −1.140 0.411 7.688 1 0.006⁎ 0.320 0.143 0.716
56–75 −1.207 0.672 3.225 1 0.073 0.299 0.080 1.117

(continued on next page)

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94


Appendix (continued)

Dependable variable Independent variables β S.E Wald df Sig Exp (β) 95% C.I.for EXP(B)

Lower Upper

Education 4.423 3 0.219
Primary 0.385 0.308 1.568 1 0.211 1.470 0.804 2.687
Secondary 1.278 0.964 1.760 1 0.185 3.591 0.543 23.734
Tertiary 0.678 0.421 2.586 1 0.108 1.969 0.862 4.498
Wealth status 2.392 3 0.495
Very rich 0.332 0.601 0.306 1 0.580 1.394 0.430 4.524
Rich 0.057 0.574 0.010 1 0.920 1.059 0.344 3.264
Poor 0.551 0.606 0.827 1 0.363 1.735 0.529 5.693
Household size 3.129 5 0.680
≤5 −0.410 0.316 1.687 1 0.194 0.663 0.357 1.232
6–10 −0.475 0.465 1.041 1 0.308 0.622 0.250 1.548
11–15 −1.167 0.806 2.097 1 0.148 0.311 0.064 1.511
16–20 −19.615 19,788.691 0.000 1 0.999 0.000 0.000 .
21–25 −20.636 40,192.970 0.000 1 1.000 0.000 0.000 .
Residence status −0.645 0.273 5.556 1 0.018⁎ 0.525 0.307 0.897

Mushroom Gender −0.298 0.413 0.521 1 0.470 0.742 0.330 1.667
Age 2.622 3 0.454
16–35 −0.169 0.623 0.073 1 0.787 0.845 0.249 2.865
36–55 −0.753 0.621 1.468 1 0.226 0.471 0.139 1.592
56–75 18.781 8094.504 0.000 1 0.998 143,360,529.3 0.000 .
Education 0.697 3 0.874
Primary −0.341 0.410 0.691 1 0.406 0.711 0.318 1.589
Secondary 18.048 15,104.120 0.000 1 0.999 68,860,035.3 0.000 .
Tertiary −0.069 0.683 0.010 1 0.920 0.934 0.245 3.563
Wealth status 3.767 3 0.288
Very rich −0.577 0.869 0.441 1 0.507 0.562 0.102 3.083
Rich −0.640 0.821 0.607 1 0.436 0.528 0.106 2.635
Poor 0.435 0.951 0.210 1 0.647 1.546 0.240 9.973
Household size 3.330 5 0.649
≤5 0.760 0.451 2.835 1 0.092 2.138 0.883 5.177
6–10 0.267 0.550 0.236 1 0.627 1.306 0.444 3.843
11–15 0.338 0.864 0.153 1 0.695 1.402 0.258 7.623
16–20 −0.382 1.243 0.095 1 0.758 0.682 0.060 7.795
21–25 17.685 40,192.970 0.000 1 1.000 47,937,255.2 0.000 .
Residence status −0.522 0.405 1.663 1 0.197 0.593 0.269 1.312

⁎ Significant at 0.05 level of confidence


Adetola, B.O., Adetoro, A.C., 2014. Threats to biodiversity conservation in cross River
National Park. Nigeria. Int. J. Conserv. Sci. 5 (4), 547–552.

Adhikari, B., Di Falco, S.D., Lovett, J.C., 2004. Household characteristics and forest de-
pendency: evidence from common property forest management in Nepal. Ecol. Econ.
48 (2), 245–257.

Alelign, A., Yemshaw, Y., Teketay, D., Edwards, S., 2011. Socio-economic factors af-
fecting sustainable utilization of woody species in Zegie Peninsula, North-Western
Ethiopia. Trop. Ecol. 52 (1), 13–24.

Angelsen, A., Jagger, P., Babigumira, R., Belcher, B., Hogarth, N.J., Bauch, S., Borner, J.,
Smith-Hall, C., Wunder, S., 2014. Environmental Income and Rural Livelihoods: a
Global-Comparative Analysis. World Dev. 64 (S1), S12–S28.

Arega, B., Woldeamlak, B., Melanie, N., 2013. Rural households’ livelihood assets, stra-
tegies and outcomes in drought-prone areas of the Amhara Region, Ethiopia: Case
Study in Lay Gaint District. Afr. J. Agric. Res. 8 (46), 5716–5727.

Ariti, A.T., van Vliet, J., Verburg, P.H., 2015. Land-use and land-cover changes in the
Central Rift Valley of Ethiopia: Assessment of perception and adaptation of stake-
holders. Appl. Geogr. 65, 28–37.

Ashraf, J., Pandey, R., de Jong, W., 2017. Assessment of biophysical, social and economic
drivers for forest transition in Asia-Pacific region. For. Policy Econ. 76, 35–44.

Aung, P.S., Pretzsch, J., Peter, R., 2012. Understanding forest dependency and resource
extraction of local communities living around the protected area in Myanmar,
Unpublished MSc Thesis. In: Dresden University of Technology. Dresden, German.

Austin, K.G., Mosnier, A., Pirker, J., McCallum, I., Fritz, S., Kasibhatla, P.S., 2017. Shifting
patterns of oil palm driven deforestation in Indonesia and impolcations for zero-de-
forestation commitments. Land Use Policy 69, 41–48.

Babbie, E., Mouton, J., 2014. The Practice of Social Research. South African edition.
Oxford University Press, Southern Africa.

Babulo, B., Muys, B., Nega, F., Tollens, E., Nyssen, J., Deckers, J., Mathijs, E., 2008.
Household livelihood strategies and forest dependence in the highlands of Tigray,
Northern Ethiopia. Agr. Syst. Vo. 98 (2), 147–155.

Barribeau, P., Butler, B., Corney, J., Doney, M., Gault, J., Gordon, J., Fetzer, R., Klein, A.,
Rogers, C.A., Stein, I.F., Steiner, C., Urschel, H., Wagoner, T., Palmquist, M., 2015.
Survey research. Colorado State University, USA.

guides/pdfs/guide68.pdf (Accessed on 19/3/2017).
Berkes, F., Folke, C. (Eds.), 1998. Linking social and ecological systems: management

practices and social mechanisms for building resilience. Cambridge University Press,
Cambridge, UK.

Bhandari, D., Jianhua, Z., 2017. Household dependency on buffer zone community forest
and its implication for management of Chitwan National Park. Nepal. Int. J. Sci. 6 (3).

Boissiere, M., Beaudoin, G., Hofstee, C., Rafanoharana, S., 2014. Participating in REDD+
Measurement, Reporting, and Verification (PMRV): Opportunities for local people?
Forests 5, 1855–1878.

Bong, I.W., Felker, M.E., Maryudi, A., 2016. How are local people driving and Affected by
Forest Cover Change? Opportunities for Local Participation in REDD+ Measurement,
Reporting and Verification. PLoS One 11 (11), e0145330.

Bongaart, J., 2001. Household size and composition in the developing world. (Policy
Research Division working paper # 144).

Broegaard, R.B., Rasmussen, L.V., Dawson, N., Mertz, O., Vongvisouk, T., Grogan, K.,
2017. Wild food collection and nutrition under commercial agriculture expansionin
agriculture-forest landscapes. For. Policy Econ. 84, 92–101.

Brown, D.G., Verburg, P.H., Pontius Jr., R.G., Lange, M.D., 2013. Opportunities to im-
prove impact, integration, and evaluation of land change models. Curr. Opin. Env.
Sust. 5, 452–457.

Campbell, B.M., Vermeulen, S.J., Lynam, T.J.P., 1991. Value of trees in the small-farming
sector of Zimbabwe. International Development Research Centre, Ottawa.

CHAPOPSA. 2002, 2002. Charcoal Potential in Southern Africa. INCO_DEV: International
Cooperation with Developing Countries (1998-2002).
publications/charcoal-potential-southern-africa/ Contract number:

Chileshe, R.A., 2005. Land Tenure and Rural Livelihoods in Zambia: Case studies of
Kamena and St. Joseph. University of Western Cape, South Africa PhD Thesis.

Coulibaly-Lingani, P., Tigabu, M., Savadogo, P., Oden, P.C., Quadba, J.M., 2009.
Determinants of access to forest products in southern Burkina Faso. For. Policy Econ.
11, 516–524.

Central Statistical Office (CSO), 2010. Zambia 2010 Census of Population and housing:
Report on characteristics of households and housing. CSO, Lusaka, Zambia.

Defries, R., Pandey, D., 2010. Urbanisation, the energy ladder and forest transitions in
India’s emerging economy. Land Use Policy 27 (2), 130–138.

Defries, R.S., Rudel, T., Uriarte, M., Hansen, M., 2010. Deforestation driven by urban

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94


population growth and agricultural trade in the twenty-first century. Nat. Geosci. 3,

FAO, 2010. A system of integrated agricultural censuses and surveys. Italy: World
Programme for the Census of Agriculture.

Giliba, R.A., Boon, E.K., Kayombo, C.J., Chirenje, L.I., Musamba, E.B., 2011. The influ-
ence of socio-economic factors on deforestation: a case study of the Bereku Forest
Reserve in Tanzania. J. Biodiv. 2 (1), 31–39.

Gumbo, D.J., Moombe, K.B., Kandulu, M.M., Kabwe, G., Ojanen, M., Ndhlovu, E.,
Sunderland, T.C.H., 2013. Dynamics of the charcoal and indigenous timber trade in
Zambia: A scoping study in Eastern, Northern and North-western provinces. Occasional
Paper. CIFOR, Bogor, Indonesia, pp. 86.

Halue, M.J.M., 2010. Population Dynamics and Sustainable Forest Conservation: A Case
Study of the West Matogoro catchment area in Songea. University of KwaZulu-Natal,
Tanzania (Unpublished PhD thesis).

Haule, J.M., 2014. Population, Development and deforestation in Songea District.
Tanzania. Nat. Resourc. 5, 15–24.

Hickey, G.M., Poiliot, M., Smith-Hall, C., Wunder, S., Nielsen, M.R., 2016. Quantifying
the economic contribution of wild food harvests to rural livelihoods: a global-com-
parative analysis. Food Policy 62, 122–132.

Hsieh, H., Shannon, S.E., 2005. Three approaches to qualitative content analysis. Qual.
Health Res. 15 (9), 1277–1288.

IBM Corp, 2015. Released. IBM SPSS Statistics for Windows, Version 23.0. IBM Corp,
Armonk, NY.

Ickowitz, A., Powell, B., Salim, M.A., Sunderland, T.C.H., 2014. Dietary quality and tree
cover in Africa. Glob. Environ. Chang. 24, 287–294.

Integrated Land Use Assessment (ILUA), 2017. Integrated Land Use Assessment Phase II –
Report for Zambia. The Food and Agriculture Organization of the United Nations and
the Forestry Department. Ministry of Lands and Natural Resources, Lusaka, Zambia.

Kalaba, F.K., 2005. The role of indigenous fruit trees in the rural livelihood: A case of
Mwekera, Copperbelt Province. University of Stellenbosch, Zambia MSc. Thesis.

Kalaba, F.K., Quinn, C.H., Dougill, A.J., 2013. Contribution of forest provisioning eco-
system services to rural livelihoods in the Miombo woodlands of Zambia. Popul.
Environ. 35, 159–182.

Kamwi, J.M., Chirwa, P.W.C., Manda, S.O.M., Graz, P.F., Katsch, C., 2015. Livelihoods,
land-use and land-cover change in the Zambezi region, Namibia. Popul. Environ.

Lacuna-Richman, C., 2003. Ethnicity and the utilization of non-wood forest products: fi
ndings from three Philippine villages. Silva Fennica 37 (1), 129–148.

Lall, S.V., Selod, H., Shalizi, Z., 2006. Rural-Urban migration in developing countries: A
survey of theoretical predictions and empirical findings. In: World Bank Policy
Research paper 3915, (Washington DC, 20433, USA.

Lambin, E.F., Geist, H.J., Lepers, E., 2003. Dynamics of Land-Use and Land-Cover Change
in Tropical regions. Annu. Rev. Environ. Resour. 28, 205–241.

Levang, P., Dounias, E., Sitorus, S., 2005. Out of the forest, out of poverty? For. Trees and
Liveli. 15, 211–235.

Lingard, M., Raharison, N., Rabakonandrianina, E., Rakotoarisoa, J.A., Elmqvist, T.,
2003. The role of local taboos in conservation and management of species: the ra-
diated tortoise in southern Madagascar. Conserv. Soc. 1, 223–246.

Luoga, E.J., 2000. Subsistence use of wood products and shifting cultivation within
miombo woodland of eastern Tanzania, with special note on commercial uses. S. Afr.
J. Bot. 66, 72–85.

Luoga, E.J., Witkowski, E.T.F., Balkwill, K., 2000a. Differential utilization and ethnobo-
tany of trees in Kitulanghalo forest reserve and surrounding communal lands, eastern
Tanzania. Econ. Bot. 54 (3), 328–343.

Luoga, E.J., Witkowski, E.T.F., Balkwill, K., 2000b. Economics of charcoal production in
miombo woodlands of eastern Tanzania: some hidden costs associated with com-
mercialisation of the resources. Ecol. Econ. 35, 243–257.

Madulu, N.F., 1996. Environmental degradation and marital fertility changes in Tanzania: The
Case of Kondoa District, Unpublished PhD Thesis. University of Dar es Salaam.

Mamo, G., Sjaastad, E., Vedeld, P., 2007. Economic dependence on forest resources: a
case from Dendi District, Ethiopia. For. Policy Econ. 9, 916–927.

Martinez, S., Mollicone, D., 2012. From land-cover to land-use: a methodology to assess
land-use from remote sensing data. Remote Sens_Basel 4, 1024–1045.

Maryudi, A., Sahide, M.A.K., 2017. Research trends: Power analysis in polycentric and
multi-level forest governance. Forest Policy Econ. 81, 65–68.

Mitinje, E., Kessy, J.F., Mombo, F., 2007. Socio-economic factors influencing deforesta-
tion on the Uluguru Mountains. Discov. Innovat. 19, 139.

Miyamoto, M., Parid, M.M., Aini, Z.N., Michinaka, T., 2014. Proximate and underlying
causes of forest cover change in Peninsular Malaysia. For. Policy Econ. 44, 18–25.

Morgan, D.L., 1996. Focus groups. Annu. Rev. Soc. 22, 129–152.
Msuya, T.S., Kideghesho, J.R., 2009. The role of traditional management practices in

enhancing sustainable use and conservation of medicinal plants in West Usambara
Mountains. Tanzania. Trop. Conserv. Sci. Vol. 2 (1), 88–105(Accessed from: www. [Accessed no: 8/3/2017]).

Mukherjee, N., 1992. Villagers’ perception of rural poverty through the mapping methods
of PRA. In: Rapid Rural Appraisal Notes Number 15. Special Issue on Applications of
Wealth Ranking. London International Institute for Environment and Development.

Mwampamba, H.T., 2009. Forest recovery and carbon sequestration under shifting cul-
tication in the eastern Arc Mountain. In: Tanzania: Landscape and land-use effects
Unpublished desertation submitted in partial satisfaction of the requirements for the
degree of Doctor of Philosophy in Ecology in the office of graduate studies of the
University of California, (183p).

Nduwamungu, J., 2001. Dynamics of deforestation in Miombo woodlands: The case of Kilosa
District. In: Tanzania. Unpublished PhD thesis. Sokoine University of Agriculture,

Ngara, R., Mangizyo, R.V., 2013. Indigenous knowledge systems and the conservation of

natural resources in the shangwe community in Gokwe District. Zimbabwe. Int. J.
Asian Social Sci. 3 (1), 20–28.

Njovu, P., Chuzu, P., Nguvulu, C.Z., 2004. Key threats and opportunities for conservation
of biological and natural resources in the Copperbelt areas of biological significance.
Analytical report, WWF-SARPO.

Nurrochmat, D.R., Nugroho, I.A., Hardjanto Purwadianto, A., Maryudi, A., Erbaugh, J.T.,
2017. Shifting contestation into cooperation: Strategy to incorporate different in-
terest of actors in medicinal plants in Meru Betiri National Park, Indonesia. For.
Policy Econ. 83, 162–168.

Nzunda, N.G., Munishi, P.K.T., Soka, G.E., Monjare, J.F., 2013. Influence of socio-eco-
nomic factors on land-use and vegetation cover changes inand around Kagoma Forest
Reserve in Tanzania. Ethoip. J. Env. Studies Manag. 6 (5), 480–488.

Obua, J., Banana, A.Y., Turyahabwe, N., 1998. Attitudes of local communities towards
forest management practices in Uganda: the case of Budongo forest reserve.
Commonw Forest Rev. 77 (2), 113–118.

Pallant, J., 2014. SPSS Survival Manual: A Step by Step Guide to Data Analysis Using SPSS
for Windows (Version 10). Open University press, Buckingham, Philadelphia.

Pan, W., Carr, D., Barbierri, A., Bilsborrow, R., Suchindran, C., 2007. Forest Clearing in
the Ecuadorian Amazon: a Study of patterns over Space and Time. Popul. Res. Policy
Rev. 26, 635–659.

Peng, C.Y.J., Lee, K.L., Ingersoll, G.M., 2002. An introduction to logistic regression ana-
lysis and reporting. J. Educ. Res. 96 (1), 3–14.

Powell, B., Ickowitz, A., Mcmullin, S., Jamnadass, R., Padoch, C., Pinedo-Vasque, M.,
Sunderland, T., 2013. The role of forests, trees and wild biodiversity for nutrition-
sensitive food systems and landscapes. In: Expert Background Paper for the
International Conference on Nutrition (ICN 2). FAO, Rome.

Powell, B., Thilsted, S., Ickowitz, A., Termote, C., Sunderland, T., Herforth, A., 2015.
Improving diets with wild and cultivated biodiversity from across the landscape.
Food Sec. 7 (3), 535–554.

Rahman, M.S., Sarker, P.K., Sadath, M.N., Giessen, L., 2018. Policy changes resulting in
power changes? Quantitative evidence from 25 years of forest policy development in
Bangladesh. Land Use Policy 70, 419–431.

Rasmussen, L.V., Watkins, C., Agrawal, A., 2017. Forest contribution to livelihoods in
changing agriculture-forest landscapes. For. Policy Econ. 84 (1–8).

Richards, M., Kanel, K., Maharjan, M., Davies, J., 1999. Towards participatory economic
analysis by forest user groups in Nepal. Overseas Development Institute in colla-
boration with the Nepal-UK Community Forestry Project, London. (Accessed from:
6139.pdf [Accessed on 6/3/2017]).

Saj, T.L., Mather, C., Sicotte, P., 2006. Traditional taboos in biological conservation: the
case of Colobus vellerosus at the Boabeng-Fiema Monkey Sanctuary, Central Ghana.
Soc. Sci. Inform. 45 (2), 285–310.

Schwartz, M.W., Caro, T.M., 2003. Effect of selective logging on trees and understory
regeneration in miombo woodlands. Afr. J. Ecol. 41, 75–82.

Shimizu, T., 2006. Assessing the access to forest resources for improving livelihoods in
West and Central Asia countries. Livelihoods Support Programme. In: LSP Working
Paper. Vol. 33. Food and Agriculture Organisation of the United Nations, FAO,
Rome, pp. 42.

Singh, A.S., Masuku, M.B., 2014. Sampling techniques and determination of sample size
in applied statistics research: an overview. Int. J. Econ. Comm. Manag. II, 11.

Sunderland, T., Abdoulaye, R., Ahammad, R., Asaha, S., Baudron, F., Deakin, E., Duriaux,
J.Y., Eddy, I., Foli, S., Gumbo, D., Khatun, K., Kondwani, M., Kshatriya, M., Leonald,
L., Rowland, D., Stacey, N., Tomscha, S., Yang, K., Gergel, S., Vianen, J.V., 2017. A
methodological approach for assessing cross-site landscape change: Understanding
socio-ecological systems. For. Policy Econ. 84, 83–91.

Sunderlin, W.D., Angelsen, A., Wunder, S., 2004. Forests and Poverty alleviation. Centre
for International Forestry Research (CIFOR).

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.

Susanti, A., Maryudi, A., 2016. Development narratives, notions of forest crisis, and boom
of oil palm plantations in Indonesia. For. Policy Econ. 73, 130–139.

Syampungani, S., 2008. Vegetation Change Analysis and Ecological Recovery of the
CopperbeltMiombo woodlands of Zambia. PhD Thesis. University of Stellenbosch,

Teshome, B., Kassa, H., Mohammed, Z., Padoch, C., 2015. Contribution of dry forest
products to household income and determinants of forest income levels in the
Northwestern and Southern Lowlands of Ethiopia. Natural Resources 6, 331–338.

Timko, J.A., Waeber, P.O., Kozak, R.A., 2010. The socio-economic contribution of non-
timber forest products to rural livelihoods in Sub-Saharan Africa: knowledge gaps and
new directions. Int. Forest. Rev. 12 (3), 284–289.

Tiwari, M.K., Saxena, A., 2011. Change detection of land-use/landcover pattern in and
around Mandideep and Obedullaganj area using remote sensing and GIS. Int. J.
Techn. Eng. Syst. 2 (3).

Tugume, P., Buyinza, M., Namaalwa, J., Kakudidi, E.K., Mucunguzi, P., Kalema, J.,
Kamatenesi, M., 2015. Socio-economic predictors of dependence on non-timber forest
products: lessons from Mabira Central Forest Researve communities. J. Agr. Environ.
Sci. 4 (2), 195–214.

Turner II, B.L., Skole, D., Sanderson, S., Fischer, G., Fresco, L., Leemans, R., 1995. Land-
Use and Land-Cover Change Science/Research Plan; IGBP Global Change Report No. 35
and HDPReport No. 7; International Geosphere-Biosphere Programme and the
Human Dimensions of Glob. Environ. Chang. Programme: Stockholm, Sweden and
Geneva, Switzerland. 3.

Umar, B.B., 2014. A critical review and re-assessment of theories of smallholder decision-
making: a case of conservation agriculture households, Zambia. In: Renewable
Agriculture and Food Systems. 29. Cambridge University Press, pp. 277–290.

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94

Ungirwalu, A., Awang, S.A., Suryanto, P., Maryudi, A., 2017. The ethno-techno-con-

servation approach in the utilization of Black Fruit (Haplolobus sp.) by the
Wandamen people of Papua. Indonesia. Biodiversitas 18, 1336–1343.

Vedeld, P., Angelsen, A., Bojo, J., Sjaastad, F., Berg, G., 2007. Forest environmental in-
come and the rural poor. For. Policy Econ. 9 (7), 869–879.

Verbist, B., Putra, A.E.D., Budidarsono, S., 2005. Factors driving Land-use Change: Effects
on watershed functions in a coffee Agroforestry System in Lampung, Sumatra. Agric.
Syst. 85, 254–270.

Vu, Q.M., Le, Q.B., Frossard, E., Vlek, P.L.G., 2014. Socio-economic and biophysical de-
terminants of land degradation in Vietnam: an integrated causal analysis at the

national level. Land Use Policy 36, 605–617.
Wunder, S., Borner, J., Shively, G., Wyman, M., 2014. Safety nets, gap filling and forests:

a global-comparative perspective. World Dev. 64, S29–S42.
WWF (World-Wide Fund for Nature). 2006. WWF – Environmental & conservation news:

Africa (Accessed on).
Yizegaw, Y.S., Okoyo, E.N., Beyene, F., 2015. Determinants of livelihood diversification

strategies: the case of smallholder rural farm households in Debre Elias Woreda, East
Gojjam Zone. Ethiopea. Afr. J. Agric. Res. Vol. 10 (19), 1998–2013.

Yohannes, W.A., Cotter, M., Kelboro, G., Dessalegn, W., 2018. Land Use and Land Cover
changes and their effects on the Landscape of Abaya-Chamo Basin, Southern Ethiopia.
Land 7, 2.

F. Handavu et al. Forest Policy and Economics 100 (2019) 75–94


  • Socio-economic factors influencing land-use and land-cover changes in the miombo woodlands of the Copperbelt province in Zambia
    • Introduction
    • Materials and methods
      • Description of the study sites
      • Socioeconomic profile
      • Sampling design and data collection
        • Site selection
        • Sampling design
        • Data collection
      • Data analysis
    • Results
      • Household demographics characteristics
      • Factors affecting land-use change
        • Population distribution
      • Farming practices and production trends
      • Sources of income and community forest resource utilization
        • Household income sources
        • Determinants of forest products utilization
        • Local Ecological Knowledge (LEK) and forest conservation
        • Land clearance and perceived causes of forest cover loss
      • Determinants of forest cover loss
    • Discussion
      • Demographic characteristics
      • Education level
      • Population distribution
      • Farming practices and production trends
      • Household income and community forest resource utilization
      • Local ecological knowledge and implications for forest conservation
      • Community perception on cause of forest cover loss
    • Conclusions and recommendations
    • Conflict of interest
    • Acknowledgement
    • Appendix
    • Focus group members undertaking mapping of households prior to wealth ranking
    • Rare income generating skills providing income sources for some of the locals
    • References