See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/274383113

Socioeconomic and environmental effects of China's Conversion of Cropland

to Forest Program after 15 years: A systematic review protocol

Article in Environmental Evidence · February 2015

DOI: 10.1186/s13750-015-003

CITATIONS READS

12 388

6 authors, including:

Lucas Gutiérrez Rodríguez Wen Zhou

Southwest University of Science and Technology (SWUST) Yale University
22 PUBLICATIONS 240 CITATIONS 32 PUBLICATIONS 1,216 CITATIONS

SEE PROFILE SEE PROFILE

Louis Putzel Chen Xie

Center for International Forestry Research China national forestry economics and development research center
78 PUBLICATIONS 1,322 CITATIONS 26 PUBLICATIONS 450 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Lucas Gutiérrez Rodríguez on 20 April 2015.

The user has requested enhancement of the downloaded file.

Gutiérrez Rodríguez et al. Environmental Evidence (2015) 4:6
DOI 10.1186/s13750-015-0033-8

SYSTEMATIC REVIEW PROTOCOL Open Access

Socioeconomic and environmental effects of

China’s Conversion of Cropland to Forest Program
after 15 years: a systematic review protocol
Lucas Gutiérrez Rodríguez1*, Nick Hogarth1, Wen Zhou1, Louis Putzel1, Chen Xie2 and Kun Zhang2

Abstract
Background: Agricultural activities on sloping lands have historically led to forest loss and degradation in China
which, coupled with industrial pressures on the environment, were deemed responsible for catastrophic flooding
events in the late 1990s. After these events, China’s forest policy underwent a significant reorientation towards
ecological conservation and rural development, a process epitomized by the Conversion of Cropland to Forest
Program (CCFP). Launched in 1999, the CCFP integrates both socioeconomic and environmental objectives with the
aim of reforesting smallholder cropland on sloping lands, while compensating farmers with payments for their lost
income. Following 15 years of implementation, it is timely to conduct a comprehensive assessment of the state of
knowledge about the CCFP’s impacts on human populations and the environment.
Methods/design: The primary research question asks “What socioeconomic and environmental effects has the
Conversion of Cropland to Forest Program had on human populations and land resources during its first 15 years in
China?” We use a theory of change and a Population-Intervention-Comparator-Outcome (PICO) framework to structure
our systematic review, where populations of interest consist of both human populations and land resources targeted
by the program, while the intervention of interest is the CCFP as defined by its component activities, including
compensatory subsidies, skill-training, and enforcement with field checks. Outcomes are defined as both the
socioeconomic and environmental impacts of the program. We will conduct a search for relevant English and
Chinese language literature on Scopus, Web of Science, CAB Abstracts, AGRIS (FAO), and the China National
Knowledge Infrastructure. Search results will be screened for relevance in a two stage process (titles and abstracts,
followed by full texts) based on predefined eligibility criteria, and then further assessed for potential sources of
bias. Extraction of data from those studies that have passed full-text screening will follow a coding protocol
based on the PICO framework, and quantitative and qualitative analyses of the extracted data will be conducted
and synthesized. Finally, a narrative report will present the findings of the review, alongside a geographic map
illustrating the coverage of included studies compared with the actual implementation area of the CCFP.
Keywords: Conversion of Cropland to Forest Program, Sloping Land Conversion Program, Grain for Green,
Payment for ecosystem services, Land use change, Afforestation, Soil erosion, Flooding, Poverty alleviation and
Social equity

* Correspondence: l.grodriguez@cgiar.org
1
Center for International Forestry Research (CIFOR), Jalan CIFOR, Situ Gede,
Sindang Barang, Bogor, Barat 16115, Indonesia
Full list of author information is available at the end of the article

© 2015 Gutierrez Rodriguez et al.; licensee BioMed Central. This is an Open Access article distributed under the terms of the
Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public
Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this
article, unless otherwise stated.
Gutiérrez Rodríguez et al. Environmental Evidence (2015) 4:6 Page 2 of 11

Background Through a large-scale conversion of land use (from sloping

The Conversion of Cropland to Forest Program (CCFP), cropland into forestland) and economic reorientation
also known as the Sloping Land Conversion Program (from on-farm towards off-farm sectors) in upstream areas,
(SLCP) or ‘Grain for Green’, was initiated in a context of the CCFP is designed to provide ecosystem goods and ser-
ecological crisis and rising environmental awareness in vices, initially to upstream populations and in the long run
China [1]. In 1997 there was a severe 267 day drought in to downstream populations.
the Yellow River basin [2], followed in 1998 by massive Since the CCFP’s inception, compensating smallholders
floods that devastated both the Yangtze and Songhua for the opportunity cost of converting their sloping crop-
River basins, resulting in 3,600 deaths, 13.2 million land into forest has been the core operational mechanism
people left homeless, and widespread economic im- of the program. At the beginning of the program, compen-
pacts [3,4]. sations included a one-time payment for the purchase of
In addition to the extraordinary weather conditions saplings or seeds, an annual living allowance paid per unit
occurring between 1997–98 caused by ENSO (El Niño area of cropland enrolled, and an annual grain/cash sub-
Southern Oscillation) [3], these flooding events were as- sidy (with different amounts for households in the Yangtze
sociated with growing pressures from human activities River watershed and the Yellow River watershed regionsb)
[5,6], particularly the over-logging of natural forests and [9,13]. The payment period of this three-tiered com-
conversion of forests on steep slopes into farmland [7,8]. pensation system also depends on the type of land-use
The authorities mainly attributed this disaster to unsus- to be established, with two years of payments provided
tainable logging practices in State Forest Farms [7] and for converting cropland into grasslandsc, five years for
the conversion of forestland into cropland on steep converting cropland into forests of ‘economic trees’
slopes by smallholders throughout the catchments [4]. (trees with direct economic returns) and eight years for
In response, the central government radically reoriented converting cropland into forests of ‘ecological trees’
its forest policy by moving from a focus on timber produc- (trees with higher use restrictions). Program partici-
tion to a strategy involving conservation, restoration and pants are paid conditionally upon maintaining a tree-
livelihoods. A range of new programs related to forest survival rate higher than a minimum set at between
conservation and environmental restoration followed, 70% and 85%, depending on local criteria, which is veri-
which together are known as the Priority Forestry Pro- fied by annual site inspections [13].
grams (or the Six Key National Forestry Programs). The The nature of the CCFP’s interventions has since
first two of these programs to be introduced, and the most evolved from this three-tiered subsidy system to its
far-reaching, are the Natural Forest Protection Programa current simplified form, with a single cash payment
(NFPP) and the CCFP. now integrating the former grain compensation and
The NFPP was launched in 1998 to ban logging in the livelihood-allowance subsidies, whereas seedling subsid-
upper reaches of the Yangtze River and upper-middle ies have been removed from the CCFP intervention.
reaches of the Yellow River, and the launch of the CCFP Apart from the compensation delivered to farmers, half
soon followed, with pilot sites introduced in the Yangtze of CCFP investment has been used on complementary
and Yellow River basins in 1999. While the former pro- activities such as cropland improvement, replanting on
gram had the objective of reducing timber harvests, the CCFP land, rural energy, etc. With regard to policy en-
latter aimed to restore vegetation on sloping croplands forcement, the central government sets national-level
and lands classified as “wasteland” or “barren land” used compensation standards, while provincial governments
by smallholder farmers [9]. may make further contributions for higher farmer com-
The original intention of the CCFP was to reduce pensations. The CCFP is implemented by county-level
flooding and soil erosion; however the program was Forestry Bureaus, which were responsible for determin-
revised after a few years of operation to emphasize the ing the sloping lands eligible for conversion and later
improvement of rural livelihoods and poverty alleviation, allocating funds to those households willing to engage in
in line with the emerging focus of the national poverty the CCFP.
reduction strategy [10-12]. The CCFP can thus be con- The CCFP is currently being implemented in 25 prov-
ceptualized as an afforestation program or a large-scale inces (1,897 counties), has already afforested more than
forest Payment for Ecosystem Services (PES) scheme 25 million hectares (comprised of 9.27 million hectares
with a compensatory approach towards upstream areas of cropland and 15.8 million hectares of barren land
inhabited by economically less-advantaged populations, classified as ‘wasteland’), and provides direct subsidies to
who play a key role in providing downstream users with 32 million households (around 124 million people)
forest ecosystem services. The scheme represents an im- [14,15]. In terms of its scale and magnitude, with 298
portant monetary compensation from both central and billion CNY (~42.82 billion USD) already invested be-
local governments to these upstream smallholders. tween 1999–2013 [16], the CCFP is one of the most
Gutiérrez Rodríguez et al. Environmental Evidence (2015) 4:6 Page 3 of 11

significant forest policies implemented in the developing  Under which circumstances would/have farmers
world [9]. revert(ed) forestland back to cropland?
 Are there any unintended socioeconomic/
Objectives of the systematic review environmental outcomes?
After 15 years of implementation (1999–2014), it is
timely to conduct a systematic evaluation of the pro- Methods
gram’s impacts on both human populations and land Theory of change
resources. The objective of the systematic review is to A conceptual understanding of the CCFP using the
provide evidence from the literature that could be used methods of the theory of change can help to explain the
to actively inform the CCFP’s design and future imple- cause-and-effect interactions between CCFP interven-
mentation, while identifying research gaps and new tions and its expected socioeconomic and environmental
testable hypotheses so as to strengthen its positive im- outcomes (see Figure 1). The diagram below was de-
pacts and minimize negative ones on both human pop- signed after scoping the CCFP literature and following
ulations and land resources. The systematic review will discussions with CCFP researchers and monitoring and
contribute by reviewing and analyzing not only the evaluation specialists from the State Forestry Adminis-
English-language CCFP literature, but also the data tration at the stakeholder meeting in Kunming, held in
available within Chinese bibliographic databases. This April 2014.
systematic evaluation is also an important part of the The first step in CCFP implementation began with the
Center for International Forestry Research (CIFOR)’s selection of households and sloping lands for participa-
emerging Sloping Lands in Transition project (SLANT), tion in the program, which was undertaken by county-
which examines smallholder re/afforestation and sustain- level forestry bureaus. Under the CCFP, smallholders are
able forest management across several countries in the expected to become forest stewards on former agricul-
Asia-Pacific region. tural and barren sloping lands. The institutional regula-
Participants from CIFOR, China National Forestry tors are the central government which transfers
Economics and Development Research Center (FEDRC) economic resources to provincial governments which
of the State Forestry Administration, Beijing Forestry then, in turn, transfer the necessary funds to county-
University (BFU) and Forest Trends held together a level forestry bureaus; these bureaus are responsible for
stakeholder meeting in Kunming, China, in April 2014 CCFP implementation on the ground to provide small-
to discuss research objectives and methods of this sys- holders with compensation for converting their cropland
tematic review protocol (please refer to Additional file 1, and barren sloping lands. On these agricultural and
list of participants). Special attention was given to defin- barren sloping lands, smallholders planted ‘economic’ or
ing the populations that might have been affected by the ‘ecological’ trees, a conversion that was actively facili-
CCFP during its implementation period, and the actual tated through the delivery of a livelihood allowance, sub-
interventions along with the potential comparators and sidies for purchasing tree saplings, and skill-training to
outcomes of interest, a process that helped us to define plant the selected species.
our primary and secondary research questions. At the After the conversion, it was expected that smallholders
stakeholder meeting, we further compiled a list of con- may experience an increase in available time (freed-up
textual factors that might affect the implementation and labor), which they could use to either intensify agricul-
outcomes of the CCFP, and discussed recurrent themes tural production on their lands or pursue off-farm work
as found across a sample of articles. in urban areas; options that will be mediated by the de-
gree of social equity among households and individuals
Research questions in the CCFP implementation area (related to both intra-
The primary research question of the systematic review household and inter-household power relations across
is: age, education level, gender, income and ethnicity
What socioeconomic and environmental effects has heterogeneity factors). These options are expected to in-
the Conversion of Cropland to Forest Program had on crease household incomes, which are further supple-
human populations and land resources during its first mented by compensation for any lost agricultural
15 years in China? income from land conversion (paid following field
The secondary questions that the systematic review checks by the county-level forestry bureau). The CCFP
intends to find evidence for are as follows: thus aims for livelihood change through reduced de-
pendence on sloping agricultural lands, which will ultim-
 How effective has the CCFP been in achieving its ately lead to a generalized poverty reduction provided
own stated objectives of soil erosion control, flood that social equity and ecosystem functions are actively
prevention and poverty reduction? promoted [17]. In the medium term, the delivery of
Gutiérrez Rodríguez et al. Environmental Evidence (2015) 4:6 Page 4 of 11

Figure 1 Theory of change.

Gutiérrez Rodríguez et al. Environmental Evidence (2015) 4:6 Page 5 of 11

these annual subsidies will compensate smallholders for the range of both intended and unintended outcomes,
the opportunity cost of such livelihood changes. including studies that find forest reconversion to crop-
In terms of environmental outcomes, the core assump- land and account for its explanatory factors.
tion of the CCFP is that increased forest area and timber
volume on sloping lands will lead to a decrease in ero- PICO framework
sion and thereby a decrease in flood risk at the water- To operationalize our research questions, theoretical hy-
shed level. Thus the higher subsidies paid for planting potheses and database searches, we have further defined
‘ecological’ trees rather than ‘economic’ trees should lead a Population-Intervention-Comparator-Outcome (PICO)
to a greater incentive for reforestation on longer time- model (see Table 1).
scales. Nonetheless, both types of planted forest will
contribute to the reduction of soil erosion on sloping
Populations
lands. Moreover, skills training and monitoring provided
Our target human population consists of CCFP partici-
by county forestry bureaus are expected to lead to higher
pant households and their individual members. Our tar-
tree survival rates, as will the selection of suitable tree
get land resources population consists of CCFP enrolled
species for individual sites. The targeting of suitable
lands (cropland, wasteland, ecological trees, economic
households and the degree of farmer voluntarism in par-
trees).
ticipating in the CCFP will also affect the longevity of
land conversion and thus the achievement of its broader
environmental goals. It is expected that farmers who Interventions
have sufficient livelihood alternatives to agriculture (ie. As the CCFP is enacted through multiple activities, in-
availability of non-targeted farmland or sources of off- terventions of interest consist of CCFP subsidies paid to
farm income) and willingly choose to participate will be smallholders for land converted, skill-training for local
less likely to reconvert lands back to agriculture after farmers, and enforcement of CCFP implementation
subsidies end. On the other hand, if disadvantaged (field-based checks on compensation delivery and house-
farmers and groups are not effectively targeted [17], this hold compliance with tree-survival rates).
could also be a deterrent for achieving both the environ-
mental and socioeconomic goals of the program, i.e. to Comparators
produce a generalized socio-ecological readjustment to- These are defined as both human populations and land
wards soil conservation, flood prevention and poverty resources that have not been exposed to the CCFP inter-
reduction. Finally, the targeting of suitable lands is crit- vention (i.e. non-participant households and non-
ical for the success of the CCFP, as sloping lands that enrolled lands), with whom/which human populations/
have already experienced considerable degradation may land resources exposed to the CCFP intervention might
be difficult to rehabilitate through tree planting alone, be potentially compared. Both households and lands
and suitable sloping lands may not always be targeted if prior to receiving CCFP interventions can also be com-
they are difficult to reach (and thus monitor). pared to human populations/land resources post-CCFP
On the basis of the CCFP’s theory of change we are intervention. Other types of comparators might also in-
going to evaluate the typology, methods, geographical clude macro-level comparisons between upstream inter-
coverage (systematic map), and the extent of the socio- ventions and upstream non-interventions or upstream
economic and environmental effects brought about by socio-ecosystems prior to and following the CCFP inter-
the program during its first 15 years of implementation vention, and also comparisons between upstream and
(systematic review). Subsequently, within the systematic downstream socio-ecosystems. All these comparators
review, we will evaluate the effectiveness of the CCFP in will be used for analysis whenever there are available
achieving both its socioeconomic and environmental ob- studies that can provide these primary data (i.e. on the
jectives, as defined by soil erosion control, flood preven- actual socioeconomic and environmental effects of the
tion and poverty reduction. Moreover, we will also assess CCFP).

Table 1 PICO elements of the systematic review

Population(s) Intervention(s) Comparator(s) Outcome(s)
CCFP households and their CCFP (subsidies , skill-training, Non-participant households, Socioeconomic outcomes (changes in
individual members and enforcement with field households prior to CCFP households’ income structure, migration, etc.)
checks implementation
CCFP enrolled lands (cropland/ CCFP (subsidies , skill-training, Non-enrolled sloping lands, Environmental outcomes (changes in water
wasteland/ecological trees/ and enforcement with field lands prior to CCFP discharge, soil erosion, flood risk, local
economic trees) checks implementation biodiversity, etc.)
Gutiérrez Rodríguez et al. Environmental Evidence (2015) 4:6 Page 6 of 11

Outcomes comparator and outcome categories were combined

Socioeconomic outcomes of the CCFP include impacts using the Boolean command OR, then combined in a
on household production (income, labor allocation, em- comprehensive search string using the Boolean command
ployment), household consumption, land tenure, food AND. Our searches will be adjusted for the specific
security and nutrition, social equity, farmers’ autonomy requirements/features of each database and their specific
in decision-making, power relations (including between truncation and/or wildcard symbols. For instance, as
income groups, ethnic groups, gender), and rural out- Google Scholar does not allow the use of complex
migration and remittances. search strings, the following intervention terms will
Environmental outcomes include impacts on water- be used: Conversion of Cropland to Forest Program
sheds (floods, discharge rates, filtration), soil (erosion, (CCFP), Sloping Land Conversion Program (SLCP),
nutrients), changes in forest cover and standing volume, Grain for Green, Upland Conversion Program, China.
tree-survival rate, changes in tree biomass and carbon The Chinese search strategy was reviewed by a subject
storage, changes in biodiversity, changes in energy specialist librarian at the University of Michigan (see
sources (biomass, coal, hydro, solar), and other land-use Additional file 2, for a detailed account of the search
and cover changes (LUCC). strings that have been employed in this protocol).
Other socioeconomic and environmental outcomes re-
ported in the literature will be noted. These potential in- Estimating the comprehensiveness of the searches
teractions and hypotheses have been explained in detail Our scoping searches confirmed our previous expecta-
within the ‘theory of change’ section and, whenever there tions that Chinese databases would retrieve a far
are available studies, socioeconomic-environmental in- higher number of results than English databases, and
teractions will also be reported. that this difference is striking in quantitative terms.
Initial search results showed Chinese results to be over
Search strategy a thousand orders of magnitude greater than English
Our search strategy has been structured according to the results. After refining our search strings, around 900
Collaboration for Environmental Evidence’s guidelines [18] results were identified from English databases (486 hits
and a PICO framework to consider the CCFP’s impacts on in Web of Science, 253 hits in Scopus, 144 hits in CAB
both human populations and land resources. abstracts and 21 hits in AGRIS, or 879 unique hits
after duplicate removal) compared to around 3,500 hits
Searches (titles) from Chinese databases.
During our initial literature scoping, we assessed the
breadth of the CCFP’s current bibliography and determined Publication databases We aim to identify CCFP peer-
that Chinese research databases contain an enormous body reviewed articles, CCFP doctoral theses and CCFP mas-
of potentially relevant literature. Around 500,000 hits were ter theses, and other study types through the following
initially identified within the Chinese China National sources:
Knowledge Infrastructure (CNKI) databases when
employing the phrase 退耕还林 (CCFP) as a search  Web of Science, Scopus, CAB Abstracts, AGRIS
term. This huge number would later be reduced to (FAO).
about 3,500 hits by making use of population, interven-  CNKI or 中国知网, which includes China Academic
tion plus comparator and outcome search terms within Journals Full-text Database, China Doctoral
the frame of our research strategy. Dissertations Full-text Database, China Masters’
Languages: Searches will be conducted in English and Theses Full-text Database, China Core Newspapers
Chinese. Spanish was also used in our scoping search Full-text Database, China Proceedings of Conference
searches, but as there were no meaningful results (no Full-text Database.
published literature), Spanish has been removed from
the final search strategy. Internet searches We will use Google Scholar to conduct
Time frame: Searching will be limited to studies pub- internet searches. In developing these methods, the first
lished or produced in and after 1999, the first year of CCFP 200 studies listed were retrieved for screening as a test of
implementation the search strategy. The comprehensiveness of the data-
Search terms: See Table 2 below for a comprehensive bases versus the internet as a source of articles determined
list of search terms as organized by their Population, to be relevant through screening will be reported in the
Intervention, Comparator, and Outcome categories. review.

Search strings and/or combinations of searches Search Specialist searching for grey literature, contacts and
terms from each of the population, intervention, organizations: During the searching process we will
Gutiérrez Rodríguez et al. Environmental Evidence (2015) 4:6 Page 7 of 11

Table 2 PICO search terms

Population English Chinese
search terms
Human household, farmer, family peasant 农户, 农民
population
Land resources sloping land, cropland, wasteland, economic forest/tree, 坡地, 耕地, 荒地, 经济林, 生态林, 土地使用, 土壤, 水,
ecological forest/tree, land use, soil water, basin 流域
Intervention English Chinese
search terms
Conversion of Cropland to Forest, Sloping Land Conversion 退耕还林
Grain for Green Upland Conversion
Comparator English Chinese
search terms
Human Participant, non-participant intra-household, upstream user 参与者, 非参与者, 家庭成员/农户成员, 上游使用者, 下游
comparators downstream user, uphill resident, lowland resident cross- 使用者, 山地居民, 低地(平原)居民 横截面(数据), 比较,
sectional, comparison comparative, longitudinal space, time, 纵向(数据), 空间, 时间, 面板数据
panel data
Land resources Enrolled, non-enrolled upstream, downstream uphill, lowland 退耕地, 未退耕地, 上游, 下游, 山地, 低地(平原), 横截面
comparators cross-sectional, comparison comparative, longitudinal time series, (数据), 比较, 纵向(数据), 空间,时间, 面板数据
space, time panel data.
Outcome English Chinese
search terms
Socioeconomic household production household consumption food security, 农户生产/家庭生产, 农户消费/家庭消费, 粮食安全/粮食保
outcomes nutrition livelihood equity, power relations, equality, Gini, gender, 障, 营养 民生/生计, 公平, 权力关系,平等, 基尼系数, 性
intra-household, ethnic decision making, governance voluntary 别平等, 家庭内关系, 少数民族, 政府决策, 治理, 自愿程
migration, remittances. 度/志愿程度, 迁移, 汇款.
Environmental watershed, floods, discharge rate, soil, filtration, erosion, soil 流域, 洪灾/水灾/洪水灾难, 流量, 土壤, 土壤渗滤, 水土流
outcomes nutrient deforestation, forest degradation, afforestation, 失/土壤流失/土壤侵蚀, 土壤养分, 森林砍伐 /毁林, 滥砍
reforestation forest cover, survival rate biodiversity, biomass, 滥伐, 森林退化, 造林, 再造林, 森林覆盖率, 林木成活率/
carbon energy, land use and land use change 林木保存率, 生物多样性, 生物量, 碳汇量, 能源, 土地利用
和, 土地利用变化

identify key institutions/organizations that could poten- Relevant interventions: These include CCFP compen-
tially be involved in conducting research studies linked sation subsidies, skill training for local farmers, and en-
to the CCFP. Afterwards, we will search for additional forcement work with field checks. When possible, we
reports delivered by these institutions/organizations’ will retrieve all information on other types of subsidies
websites. Moreover, we have also issued a call for grey that might have an impact on household livelihoods and
literature, with both English and Chinese language bro- the environment. Broadly speaking, the Natural Forest
chures being circulated online and hard copies distrib- Protection Program (NFPP) does not overlap with the
uted at relevant meetings and conferences [19]. In the CCFP, as the former is related to state forestland
meantime, an advisory group formed by the people who whereas the latter mainly occurs over collective forest-
took part in the stakeholder meeting in Kunming will land. Therefore, the NFPP is not included in our analysis
also provide key inputs in recommending relevant re- (although it is taken into account as a contextual factor).
ports/datasets that may be unpublished or not found in Relevant comparators: We are interested in assessing
our searching. the existing evidence comparing the effects of the CCFP
between participating and non-participating CCFP house-
Inclusion criteria holds. This systematic review will simultaneously consider
Relevant subjects: Both human populations and land re- the available evidence about CCFP land resources’ com-
sources are to be included as relevant populations, in- parators such as both enrolled and non-enrolled lands
cluding: CCFP participant households, their individual (under the management by both types of households
members and their CCFP enrolled lands (cropland, dwelling upstream). This systematic review will also use
wasteland, ecological trees, economic trees). Grasslands the available empirical data to track those ‘before-and-
are excluded from our analysis since they no longer form after’ comparators in both human populations (i.e. the
part of the CCFP, they are under the administration of socioeconomic status of both participant and non-
the Ministry of Agriculture, and because they contribute participant households before and after the CCFP inter-
to significantly different environmental outcomes as ventions) and land resources (i.e. the environmental status
compared with forests. of both enrolled and non-enrolled lands before and after
Gutiérrez Rodríguez et al. Environmental Evidence (2015) 4:6 Page 8 of 11

the CCFP intervention). At a more aggregated level, we With regards to qualitative evidence, we will consider the
will synthetize the available evidence on the effects of the following design/methods: participant and non-participant
CCFP on watershed with-intervention upstream and observations, structured, semi-structured, and unstructured
without-intervention upstream socio-ecosystems (i.e. both interviews, focus group discussions, and qualitative data
between intervention and non-intervention watershed re- from surveys and questionnaires. On the other hand, with
gions, and between pre-CCFP and post-CCFP watershed regard to quantitative evidence, we will consider the follow-
regions), and also between upstream and downstream ing design/methods: direct measurements of observed phe-
socio-ecosystems. nomena, including use of geo-spatial technologies (GIS and
Relevant outcomes: From the stakeholder workshop remote sensing) as well as the use of polls, questionnaires,
and initial literature scoping, we have identified a num- and surveys where answers are restricted to given choices.
ber of relevant socioeconomic outcomes; including Finally, studies that will not be considered for data
CCFP impacts on: household production and consump- extraction include reviews, meta-analyses, summary stud-
tion; changes in household land tenure; changes in food ies, theoretical and methodological framework studies,
security and nutrition; social equity (between and within and editorials and commentaries, although these will be
households), farmers’ autonomy in decision-making, considered in our background and discussion.
power relations (income groups, intra-household and
gender levels, ethnic groups); migration and remittances. Study screening Given the big volume of references ex-
With regards to environmental outcomes of interest, we pected from the Chinese literature database, we will first
have identified the following: floods and watershed dis- perform title screening, then abstract screening, and
charge rates; soil filtration, erosion and nutrient cycling; finally full-text screening of retrieved search results
deforestation and forest degradation on slopes (should according to the inclusion criteria stated above. For the
farmers revert converted land back to cropland); forest English literature database, we will first conduct title-
cover, afforestation and reforestation; tree-survival rates, and-abstract screening and later full-text screening. At
biodiversity, biomass and carbon storage; changes in the beginning of each stage of screening, four reviewers
household energy sources (biomass, coal, hydro, solar, etc.); for English (NH, WZ, LP and LGR) and three reviewers
and land-use cover change (LUCC) dynamics. Studies for Chinese (CX, KZ and LGR) will review a sample of
assessing potential or future outcomes of CCFP, including 50 studies to conduct kappa analysis on their screening
model projections or other predictions of program impact, decisions. Should the kappa statistic fall below 0.6, the
will not be included as this review only seeks to assess the reviewers will discuss points of disagreement and conduct
actual impacts of CCFP implementation (those which have a second round of screening. This will take place for both
already taken place). Socioeconomic-environmental interac- the Chinese and English language literature, with LGR
tions will be reported, whenever there are available studies coordinating pilot screening in both languages. Once an
on this issue. acceptable level of agreement is reached, the remainder of
Relevant types of study design: Primary studies using studies will be screened in each stage. Members of the
quantitative and qualitative methods will be considered; advisory group will also randomly review the screening,
these can include experimental and quasi-experimental and for a 25% selection of the scoped files, tests will be
designs, case–control experiments and broad sample-size conducted first on the abstracts and then on the full-texts
surveys of participant and non-participant populations to ensure screening decisions remain consistent. Then
(cross-sectional analyses), surveys of populations prior to full-text reading and extraction of qualitative and quantita-
and following CCFP implementation (longitudinal ana- tive information will proceed into several categories (see
lyses), and individual case studies of populations that have next subsection, study quality assessment).
been targeted for CCFP interventions. Studies must use
primary data to present actual impacts that have already Potential effect modifiers and reasons for heterogeneity
happened, and are causally linked or correlated to the During the stakeholder workshop in Kunming (April
CCFP interventions. Primary studies concerning farmers’ 2014), a set of independent socioeconomic variables with
perceptions of CCFP impacts will also be included, pro- potential influence on CCFP outcomes was defined, in-
vided that a robust and reliable methodology was used, as cluding: household members’ age, gender, education level,
these perceptions can be used as a proxy for measuring income group and ethnicity. These variables will be useful
certain socioeconomic impacts. Modeling exercises that to assess cross-household and intra-household heterogen-
use primary data to calculate actual impacts shall be in- eity of CCFP impacts. For instance, we can assess if im-
cluded for further analysis, whereas models that project pacts of the CCFP equally affect households with different
potential or future impacts will not be included (although income levels, or whether different members of the same
they will be collected in a separate folder for future household experience different socioeconomic impacts.
analysis). At the same time, a set of independent environmental
Gutiérrez Rodríguez et al. Environmental Evidence (2015) 4:6 Page 9 of 11

variables with potential effects on CCFP outcomes at the and using the set of factors of interest that were raised
targeted land plots were also defined, including: orientation, in discussions at the stakeholder meeting in Kunming.
slope, size, distance to home, weather/climate, altitude and The data extraction categories are as follows:
latitude. Socioeconomic-environmental interactions among
these factors can be especially relevant for the CCFP imple- Study metadata and methodology:
mentation process.
 Bibliographic information: author, year, title,
Study quality assessment (critical appraisal) institution of the lead author.
Those studies that meet our inclusion criteria through  Type of study: quantitative/qualitative study, or both
the full-text screening will be assessed based on the (mixed methods).
following five quality criteria:  Comparative methods: cross-sectional, longitudinal,
or both.
1. Data collection methods are thoroughly explained  Geographic location (county and GPS coordinates
and clear and replicable. whenever available)
2. Qualitative or quantitative analysis methods are  Time-span covered by the study.
thoroughly explained and clear and replicable; key
terms and variables are well defined. Population:
3. Sample size is well explained and representative of
the population.  Type of population: human population, land
4. Results/conclusions are logically derived and resources population, or both.
supported by presented evidence.  Unit of comparative analysis (scale): household/
5. Confounding factors are considered and well individual, village/community, county, provincial or
explained. national levels.
 Sample size and land area: number of households
We will document individual study quality based on covered by study or land area covered by the study.
each of these five criteria, and report on the overall qual-
ity of the evidence base in our systematic review. For Intervention:
each study, we will also record yes/no answers for each
criteria, where “yes” is equal to a score of one and “no”  Type and duration of intervention: compensation
equal to a score of zero. Each study will thus have a subsidies plus tree-sapling provision, skill-training,
quality assessment score of 0 to 5, where scores of 3 to 5 enforcement with field checks (one or multiple
will be considered acceptable while studies with scores intervention types can be present).
of 0 to 2 will be considered low quality. For our system-
atic review, we will consider and compare the outcomes Outcomes:
from both sets of studies to determine whether the low
quality studies demonstrate significantly different results  Socioeconomic outcome categories: changes in
from those of acceptable quality studies, and whether upstream household production and production
their inclusion in our final analysis leads to any change structure (as measured by income, labor,
in our overall assessment. Assessment of the studies employment); changes in household consumption
against these criteria is a strong indicator of the presence and household income structure; changes in
or absence of most types of potential bias. In addition, household land tenure; changes in social equality
we will check for patterns of correspondence between (Gini coefficient) and intra-household equality, both
authors’ affiliations and specific findings so as to identify of them across income levels, gender, age groups and
additional sources of potential bias. We will also deter- educational levels; changes in household migration
mine whether there are discernable biases in results that and remittances; enforcement (voluntary/compulsory
correspond to the study designs used (i.e. control/coun- and degree of tree species selection).
terfactual, longitudinal study, or case study).  Environmental outcome categories: changes in
upstream forest cover and standing volume, tree
Data extraction strategy survival rates, changes in measures of biodiversity
For all studies that have met our critical appraisal (species richness, composition, and abundance), tree
criteria after full-text screening, we will proceed with biomass and carbon storage; changes in upstream
extracting both quantitative and qualitative data for soil erosion and soil nutrient content; changes in
both socioeconomic and environmental outcomes, fol- upstream household energy use and energy
lowing the general structure of our PICO framework structure; changes in upstream land-use and
Gutiérrez Rodríguez et al. Environmental Evidence (2015) 4:6 Page 10 of 11

land-cover change (NDVI and leaf-area indices); be also presented in a searchable database, provided
changes in downstream discharge rates and floods; along with a geographic map that identifies and locates
frequency of natural disasters; trade-offs among the available evidence from studies across China. GIS
ecosystem services. and Remote Sensing techniques will be used to document
 Socioeconomic-environmental interactions (among the locations of included studies within China to compare
the aforementioned outcomes), whenever there are the coverage of CCFP evaluations with the actual area of
available studies on this issue. its implementation.
As for the secondary research questions, the existing re-
Potential effect modifiers and reasons for sults on the program’s effectiveness will be first synthesized
heterogeneity: and accordingly presented in terms of its achievements in
Socioeconomic factors: household members’ age, gender, poverty reduction, soil erosion control and flood preven-
education, income group and ethnicity: tion. Second, conclusions will be made in terms of condi-
tions that may have led to the reconversion of sloping
– Age: average and percentage distribution across lands to agriculture. Third, CCFP’s unintended socioeco-
several ranges (over 20, between 20–40, between nomic and environmental effects will be synthesized and
40–60, and over 60). presented in order to guide future program implementa-
– Gender: percentage of women/men tion and research, and also to uncover possible knowledge
– Education: average and percentage distribution gaps and new hypotheses on the program’s impacts.
across several ranges (primary school or less, middle
school, high school or above). Endnotes
– Income group: locally-defined low, middle and high a
The NFPP was approved in 1998 so as to stop natural
income groups. forest loss and degradation [15]. The introduction of this
– Ethnicity: Han/non-Han ‘logging ban’ policy meant the re-structuring of state-
– Environmental factors: land orientation, slope, size, owned forestry enterprises, into which government sub-
distance to home, and elevation of land plots sidies have been channeled to compensate laid-off workers
– Other: Voluntarism of participation in CCFP and alleviate the economic crisis faced by these companies
in the late 1990s.
b
Since 2004, grain transfers were completely replaced
Data synthesis and presentation by cash.
c
One narrative synthesis report, i.e. a systematic review, Although the CCFP initially included the conversion of
will be produced relying on both qualitative and descrip- cropland into grassland, this land-use transformation no
tive statistics so as to assess the available evidence on longer forms part of the program, and so has become a
the CCFP’s socioeconomic and environmental outcomes. different program which is currently under the Ministry of
Using descriptive statistics, we will first present the re- Agriculture.
sults of each screening stage (title, abstract, and full text
screening) and will also use statistics to show the results Additional files
of the quality assessment (high quality vs. low quality
studies). Secondly, we will organize and synthesize the Additional file 1: Participants in stakeholder meeting, Kunming,
data according to the types of research conducted on the April 2015.
Conversion of Cropland to Forest Program, through: 1) Additional file 2: Search String Combinations.
categorizing the empirical evidence as socioeconomic
and/or environmental; 2) classifying its research methods Abbreviations
BFU: Beijing Forestry University; CCFP: Conversion of Cropland to Forest
in typologies; and 3) presenting the geographical distribu- Program; CIFOR: Center for International Forestry Research; CNKI: China
tion of studies throughout the country. Qualitative data National Knowledge Infrastructure; ENSO: El Niño Southern Oscillation;
on both socioeconomic and environmental outcomes will FEDRC: Forestry Economics and Development Research Center (China State
Forestry Administration); NFPP: Natural Forest Protection Program;
be grouped by the individual measures they address (i.e. PES: Payments for Ecosystem Services; SLCP: Sloping Land Conversion
income change, forest cover change, etc.) and synthesized Program.
narratively alongside quantitative data on the same
Competing interests
measures. In so far as we have sufficient data to perform
The authors declare that they have no competing interests.
meta-analysis of quantitative data of outcome measures
(and particularly their correlation with human and envir- Authors’ contributions
onmental population characteristics), we will do so. All authors took part in the stakeholder meeting held in Kunming in April
2014. NH carried out initial database searches of English studies, and
With regard to the main research question, the socio- provided key research inputs to this protocol. LP and CX provided key
economic and environmental outcomes of the CCFP will research guidance for the definition of primary and secondary questions. LP,
 Gutiérrez Rodríguez et al. Environmental Evidence (2015) 4:6 Page 11 of 11

NH, WZ, and KZ participated in pilot Kappa tests for study screening. LGR 13. Bennett M, Xie C, Hogarth N, Peng DL, Putzel L. China’s Conversion of
carried out literature scoping in English, Chinese and Spanish databases, and Cropland to Forest Program for Household Delivery of Ecosystem Services:
led the writing of this draft protocol. All authors reviewed the final version of How Important is a Local Implementation Regime to Survival Rate
this protocol before submission. All authors read and approved the final Outcomes? Forests. 2014;5(9):2345–76.
manuscript. 14. Song C, Zhang YL, Mei Y, Liu H, Zhang ZQ, Zhang QF, et al. Sustainability of
forests created by China’s sloping land conversion program: a comparison
Acknowledgments among three sites in Anhui, Hubei and Shanxi. Forest Policy Economics.
Special thanks to Peng Daoli (Beijing Forestry University), Michael Bennett 2014;38:161–7.
(Forest Trends), Wang Jianan (FEDRC), Wang Jiang (BFU), Christine Padoch 15. Delang CO, Wang W. Chinese Forest Policy Reforms After 1998: The Case of
(CIFOR), Kiran Asher (CIFOR) and Yustina Artati (CIFOR), who participated and the Natural Forest Protection Program and the Slope Land Conversion
contributed to the stakeholder meeting held in Kunming (April 12th 2014). Program. Int For Rev. 2013;15(3):290–304.
We also want to express our gratitude to Liangyu Fu (University of 16. Zhai DL, Xu JC, Dai ZC, Cannon CH, Grumbine RE. Increasing tree cover
Michigan), who reviewed our Chinese search strategy. The review team while losing diverse natural forests in tropical Hainan, China. Reg Environ
would like to thank CIFOR’s Evidence Based Forestry Initiative and the UK Change. 2014;14:611–21.
Department for International Development (DfID) for financing this research 17. Pascual U, Phelps J, Garmendia E, Brown K, Corbera E, Martin A, et al. Social
through its KNOWFOR program grant. equity matters in payments for ecosystem services. Bioscience. 2014;64
(11):1027–36.
Author details 18. Collaboration for Environmental Evidence. Guidelines for Systematic Review
1
Center for International Forestry Research (CIFOR), Jalan CIFOR, Situ Gede, and Evidence Synthesis in Environmental Management. Version 4.2. In:
Sindang Barang, Bogor, Barat 16115, Indonesia. 2China National Forestry Environmental Evidence. 2013. http://environmentalevidence.org/wp-
Economics and Development Research Center, State Forestry Administration, content/uploads/2014/06/Review-guidelines-version-4.2-final.pdf.
Hepingli Dongjie No. 18, Beijing 100714, China. 19. Evidence-Based Forestry. http://www1.cifor.org/ebf/reviews/current-reviews/
systematic-reviews/chinas-conversion-of-cropland-to-forest-program-ccfp.html.
Received: 11 June 2014 Accepted: 8 February 2015

References
1. Xi XL, Fan LH, Deng XM. The situation of environmental awareness in China.
Analysis of results from a survey of Chinese citizens (In Chinese). Zhongguo
gongzhong huanjing yishi zhuangkuang - gongzhong diaocha jieguo pouxi
Zhongguo ruan kexue-gongzhong juece-huanjing baohu ji qi chanye
zhuanti yantao. 1998;9:24–30.
2. Gao YC. Analysis on reasons for the Yellow River’s dry-up and its
eco-environmental impacts. J Environ Sci. 1998;10(3):357–64.
3. Qian Y, Glantz M. The 1998 Yangtze Floods: the use of short-term forecasts
in the context of seasonal to interannual water resource management.
Mitig Adapt Strat Glob Chang. 2005;10:159–82.
4. Yin RS, Xu JT, Li Z, Liu C. China’s Ecological Rehabilitation: The
Unprecedented Efforts and Dramatic Impacts of Reforestation and Slope
Protection in Western China. China Environment Series. 2005;7:17–32.
5. Li WH. Yangtze’s floods and ecological construction (In Chinese). Changjiang
hongshui yu shengtai jianshe Ziran ziyuan xuebao. 1999;14(1):1–8.
6. Shi DM. Analysis on the relationship between soil erosion and flood
disasters in the Yangtze basin (In Chinese). Changjiang liuyu shuitu liushi
yu honglao zaihai guanxi pouxi Turang qinshi yu shuitu baochi xuebao.
1999;5(1):1–7.
7. Liu C. An economic and environmental evaluation of the Natural Forest
Protection Program. China National Forest Economics and Development
Research Center (FEDRC), State Forestry Administration (SFA) 2002
(working paper).
8. Xu ZG, Xu JT, Deng XZ, Huang JK, Uchida E, Rozelle S. Grain for Green
versus Grain: Conflict between Food Security and Conservation Set-Aside in
China. World Dev. 2006;34(1):130–48.
9. Uchida E, Xu JT, Rozelle S. Grain for Green: Cost-Effectiveness and
Sustainability of China’s Conservation Set-Aside Program. Land Econ.
2005;81(2):247–64.
10. Gauvin C, Uchida E, Rozelle S, Xu JT, Zhan JY. Cost-Effectiveness of Payments
for Ecosystem Services with Dual Goals of Environment and Poverty Submit your next manuscript to BioMed Central
Alleviation. Environ Manage. 2010;45:488–501. and take full advantage of:
11. China Green Times. 10 years of monitoring, sending our country ‘a postcard
of data’. Looking back 10 years of monitoring the socioeconomic benefits
• Convenient online submission
of the key forestry programs (In Chinese). Shi nian jiance, jigei zuguo de
shuju mingxinpian——linye zhongdian gongcheng shehui jingji xiaoyi • Thorough peer review
jiance shinian huigu. Zhongguo lvse shibao 2014-January-2nd. Retrieved • No space constraints or color figure charges
from: http://www.greentimes.com/green/news/zhuanti/lyzdgc/content/
• Immediate publication on acceptance
2014-01/02/content_242505.htm.
12. CCICED. Implementing the Natural Forest Protection Program and the • Inclusion in PubMed, CAS, Scopus and Google Scholar
Sloping Land Conversion Program: Lessons and policy recommendations. • Research which is freely available for redistribution
In: Xu J, Katsigris E, White TA, editors. China Council for International
Cooperation on Environment and Development; Western China Forests and
Grassland Task Force. 2002. Submit your manuscript at
www.biomedcentral.com/submit

View publication stats