Forest certification as a policy option in conserving biodiversity: An empirical study of forest management in Tanzania

Forest Ecology and Management 361 (2016) 1–12

Contents lists available at ScienceDirect

Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco

Forest certification as a policy option in conserving biodiversity: An empirical study of forest management in Tanzania Severin Kusonyola Kalonga ⇑, Fred Midtgaard, Kari Klanderud Department of Ecology and Natural Resource Management, Faculty of Environmental Sciences and Technology, Norwegian University of Life Sciences (NMBU), Post Box 5003, NO-1432 Ås, Norway

a r t i c l e

i n f o

Article history: Received 15 July 2015 Received in revised form 2 October 2015 Accepted 21 October 2015

Keywords: Forest Stewardship Council Forest management Biodiversity Human forest use Green growth

a b s t r a c t Forest certification management standards aim at maintaining forest ecosystem integrity, including forest biodiversity conservation. However, studies from the Amazon and Congo basin find that forest certification may not protect forest biodiversity and ecosystems, and may therefore be unsustainable. This study evaluates the influence of forest certification on conserving biodiversity. Specifically, we (a) estimate tree (adult and seedling) species richness, diversity and density among different forest management regimes; (b) assess the relationship between environmental and human forest use variables, and species richness, diversity and density among the forest management regimes; and (c) assess the influence of forest governance of villages adjacent to the forests on tree (adult and seedling) species richness, diversity and density among the forest management regimes. This is achieved in a comparative study of Forest Stewardship Council certified community forests, non-certified open access forests, and non-certified state forest reserves in the Kilwa District in Tanzania. Our results show that forest certification standards and implementation processes are positively related to biodiversity conservation. There are significantly higher tree (adults) species richness, diversity, and density in certified community forests than in open access forests and state forest reserves. These findings suggest that forest certification may be a good policy option to conserve biodiversity. The present study is one of the first studies in tropical Africa, which contributes to the limited data on the influence of forest certification on conserving biodiversity. Our results may also serve as baseline for further research on the contribution of certified forests in conserving biodiversity at both temporal and spatial scales. Ó 2015 Published by Elsevier B.V.

1. Introduction Tropical forests provide a variety of valuable ecosystem services, such as biodiversity, carbon sequestration, water cycling and scenic beauty (Gardner et al., 2009; Sasaki et al., 2011; Sell et al., 2007). They contribute to the long-term social and economic development goals of the people who depend on them (Sebukeera et al., 2005) to achieve the vision of green growth and sustainable economy (Muthoo, 2012). Tropical forests also play an important role in addressing the Millennium Development Goals, specifically in ensuring environmental sustainability (Sebukeera et al., 2005), including forest biodiversity conservation. Unfortunately, the capacity of tropical forests to provide these ecosystem services is reduced each year by deforestation (FAO, 2010), as well as by forest ⇑ Corresponding author. E-mail addresses: [email protected] (S.K. Kalonga), fred.midtgaard@ nmbu.no (F. Midtgaard), [email protected] (K. Klanderud). http://dx.doi.org/10.1016/j.foreco.2015.10.034 0378-1127/Ó 2015 Published by Elsevier B.V.

degradation due to uncontrolled human activities such as logging and forest fires (FAO, 2006; Sasaki et al., 2011). This results in habitat degradation and fragmentation, leading to the current rampant loss of forest biodiversity (Timonen et al., 2011). Human activities have changed ecosystems more rapidly and extensively than in any comparable period of time in human history, largely to meet rapidly growing demands for food, timber, fuelwood and fibre (Levy et al., 2005; Kindt et al., 2006). During the 21st century, a substantial and ongoing loss of forest biodiversity is projected to escalate (Alkemade et al., 2009; Kim et al., 2015) in the tropics (Mwase et al., 2007; Biggs et al., 2008), including Tanzania. Forest resources in Tanzania have been managed by the state during and after the colonial eras (Burgess and Clarke, 2000). During these eras (colonial: 1880s–1961 and after independence: 1961–1990s), the state has undertaken a number of forest policy reform programmes, aiming at improving the management of natural resources (Burgess and Clarke, 2000; Zahabu et al., 2009; Petersen and Sandhövel, 2001). Most of these reforms have,

2

S.K. Kalonga et al. / Forest Ecology and Management 361 (2016) 1–12

however, not brought about the expected results (e.g. see Burgess and Clarke, 2000; Zahabu et al., 2009), as deforestation and forest degradation have escalated (Milledge et al., 2007). In combating this problem, the new Forest Policy was approved in 1998 and the Forest Act enacted in 2002. These led to the introduction of a communal forest management regime, whereby villagers have the mandate to set aside part of their village ‘general land’ forests as village land forest reserves under community-based forest management (CBFM). The CBFM aims at restoring degraded forests by controlling legal and illegal forest exploitation (URT, 1998). In spite of the institutional and legal frameworks settled, and the aim of restoring degraded forests by controlling forest exploitation, illegal exploitation of forest resources in these areas has continued (Milledge et al., 2007). In response to escalating deforestation and forest degradation, particularly in the tropics, non-governmental bodies formed the Forest Stewardship Council (FSC) in 1993 (e.g. see Auld et al., 2008; Marx and Cuypers, 2010). The FSC is an international not-for-profit multi-stakeholder organisation for promoting responsible management of the world’s forests (Karmann and Smith, 2009; FSC, 2015a), with the mission and goals of protecting forests for future generations (FSC, 2015b). The FSC forest management standards have ten principles and 70 criteria which provide details on how to manage forests responsibly (FSC, 2015a). There are about 1400 village land forest reserves under CBFM in Tanzania (MNRT, 2008), of which a total of six were FSC-certified by the end of 2012. These communities are practicing CBFM through the application of FSC management standards (see Soil Association, 2009) under the coordination of a non-governmental organisation, Mpingo Conservation and Development Initiatives (MCDI). They apply the standards to reduce pressure on forest resources by creating alternative livelihoods to communities through selective logging (i.e. sustainable harvesting) for timber production (Ball, 2009, 2010), while maintaining forest ecosystem integrity, including forest biodiversity conservation (Karmann and Smith, 2009; Sheil et al., 2010). Studies from the Amazon and the Congo Basin, employing qualitative assessments, i.e. semistructured interviews and meetings, and quantitative assessments, such as recording tree species seedlings, diversity, and logging damage, find that even limited logging affects forest biodiversity and ecosystems, and is therefore unsustainable (e.g. Ebeling and Yasué, 2009; Kukkonen et al., 2008; Poulsen and Clark, 2010; Medjibe et al., 2013). However, there is inadequate biological data on the effect of forest certification on biodiversity (see Tallis et al., 2011; Cubbage et al., 2010; Blackman et al., 2014; van Kuijk et al., 2009; Blackman and Rivera, 2010; Karmann and Smith, 2009; Sheil et al., 2010), particularly in Africa. Also, results from e.g. the Amazon and the Congo Basin forest environments may not apply to African Miombo forests. Thus, the lack of empirical evidence on the influence of certified forests on conserving forest biodiversity motivates this study, which attempts to answer the question: Is forest certification a policy option in conserving biodiversity? To discern the influence of forest management intervention on forest biodiversity conservation among management regimes, we need to explore the effects of environmental and human forest use variables on species richness, diversity and density (see Hooper et al., 2005). Generally, easily accessible forests are more affected by human activities (Sassen and Sheil, 2013) depending on tree species (Ndangalasi et al., 2007); although effective forest management planning could reverse the situation (Ball, 2011). This study examines the relationships between human forest use indicator variables and forest biodiversity indicator variables to deduce the influence of forest certification. This is achieved by comparatively assessing biodiversity in FSC-certified community forests (FSC); non-FSC-certified open access forests (OCF); and non-FSCcertified state forest reserves (FRS) in the Kilwa District in Tanzania. Specifically, the study: (a) estimates tree (adults and seedlings)

species richness, diversity and density among the forest management regimes; (b) assesses the relationship of environmental and human forest use variables with tree (adults and seedlings) species richness and diversity among the forest management regimes; and (c) assesses the influence of indicators of forest governance (e.g. rule compliance) of villages adjacent to the forests on tree (adults and seedlings) species richness, diversity and density among the forest management regimes. The study acquires ecological data, used as forest biodiversity indicators, and socioeconomic data, used as human forest use indicators, to evaluate the performance of various forest management regimes. To collect and use such data, the study applies a mixed methods research design, i.e. integrated natural and social sciences research approaches (see Creswell, 2013; Lund et al., 2014). Specifically, the study focuses on the indicators of impacts, and on how to disentangle their effects from other confounding factors that may impact on forest biodiversity, and forest governance. This is made possible by triangulation through the use of multiple data sources and methods of analysis of the observations from the study.

2. Sites and methods 2.1. Study sites The study was conducted in the Kilwa District in the Lindi Region in Tanzania. Six forests and four villages adjacent to these forests were chosen for this study (Fig. 1 and Table 1). The sites fall on the western part of Kilwa, and the study system is characterised by miombo woodlands with some patches of coastal forests, north Zambezian undifferentiated woodlands, and wooded grassland (Lillesø et al., 2014). Miombo woodlands are dominated by woody plants, primarily trees (Chidumayo and Gumbo, 2010), with high diversity and degree of endemism (Chidumayo et al., 2011; Chidumayo and Gumbo, 2010). They are dominated by species in the genera Brachystegia, Julbernadia, and Combretum of the Caesalpinoideae subfamily (Chidumayo and Gumbo, 2010; Frost, 1996). They are the most extensive tropical savannah woodland and dry forest formations in Africa (Campbell et al., 2007; Campbell, 1996), covering about 2.7 million km2 of southern Africa including southern Tanzania (Chidumayo and Gumbo, 2010). Indicator miombo tree species, such as Acacia polyacantha Willd, Lonchocarpus capassa Rolfe, Piliostigma thonningii Schum, and Xeroderris stuhlmannii (Taub.) Mendonça & E.P. Sousa were observed in the study forests during fieldwork. Kikole and Kisangi forests are FSC-certified community forests under CBFM management regime, i.e. FSC-certified CBFM (Table 1). Likawage and Mchakama forests are village ‘general land’ forests under ‘de facto’ open access management regime without certification (OCF), i.e. nonCBFM and non-FSC-certified forests. Mitarure and Rungo are forest reserves under state management without certification (FRS), i.e. non-FSC-certified state forest reserves (Table 1). In this study, ‘open access’ regime refers to a regime which is experiencing an ineffective enforcement of laws by the appropriators, resulting in ‘de facto’ open access regime (see Milledge et al., 2007; Fennell, 2011). All of the selected forests have undergone similar historical and management processes (e.g. see Ball, 2010; Burgess and Clarke, 2000), and they have almost similar biophysical and physiographic attributes (see Burgess and Clarke, 2000), i.e., they are located in the same agro-ecological zone and similar vegetation types (miombo woodlands biome) with similar range of soils, slope, elevation, and climatic factors (rainfall, temperature, humidity). They also have several tree species of economic importance, and they have a high tree species diversity (Howell et al., 2012; Backéus et al., 2006). However, these forests are heavily influenced by

S.K. Kalonga et al. / Forest Ecology and Management 361 (2016) 1–12

3

Fig. 1. Map of Kilwa District, Lindi Region, Tanzania showing the relative location of the study sites.

humans and are still threatened as a consequence of an everincreasing demand for farmland, fuelwood and timber (Burgess and Clarke, 2000; KDC, 2008). Four villages (Kikole, Kisangi, Likawage and Mchakama) were investigated alongside with the forest sites to assess the influence

of forest governance in shaping the conditions of the adjacent forests. Before 1974, the forests adjacent to these villages were unreserved ‘general land’ forests (open access regime). Kikole and Kisangi village forests changed management from open access to village land forest reserves under CBFM in 2007, and were

4

S.K. Kalonga et al. / Forest Ecology and Management 361 (2016) 1–12

Table 1 Description of forests showing the areas, number of plots and transects, distance to nearby town, management regime, certification and legal status, forest owner and manager. Forest name

Area (ha)

Number of plots (transects)

Distance to town (km)

Current forest management regime

Forest certification

Forest legal status

Forest owner

Forest manager

Kikole

454

22 (4)

33.7

CBFM, 2007

FSC, 2009

Village

VNRC

Kisangi

1966

18 (2)

33.5

CBFM, 2007

FSC, 2009

Village

VNRC

Likawage

17,000

24 (3)

100.8

Open access

Non-FSC

State

VNRC/KFO

Mchakama

3000

20 (4)

54.4

Open access

Non-FSC

State

VNRC/KFO

Mitarure Rungo

60,484 22,586

18(3) 29 (3)

30.6 89.9

State State

Non-FSC Non-FSC

Village land forest reserve (FSC) Village land forest reserve (FSC) Village general land forest (OCF) Village general land forest (OCF) State forest reserve (FRS) State forest reserve (FRS)

State State

KFO KFO

Village Natural Resource Committee (VNRC), Kilwa District Forest Office (KFO), Number of transects shown in brackets, Community-based forest management (CBFM), Forest Stewardship Council (FSC).

FSC-certified in 2009, while Likawage and Mchakama are still ‘general land’ forests under open access regime (Table 1). Currently, the main commercial uses of the village land forest reserves are for timber production. These forests are also used for grazing and supply of subsistence and cash from forest products including charcoal, fuelwood and construction materials. Mchakama village ‘general land’ forest is less disturbed than the others (Perkin et al., 2008) most likely due to limited access (Fig. 1). Mitarure and Rungo forests were under open access regime before they were legally established as state forest reserves in 1957 and 1956, respectively (Table 1), for protection and production purposes. The reserves are also threatened by human activities such as illegal harvesting, uncontrolled fires started from farm preparations, honey extraction, hunting, charcoal burning, and recently also from livestock grazing by pastoralists moving into the area. 2.1.1. Forest certification in Kilwa The Forest Stewards Council (FSC) standards postulate that forest management shall conserve biodiversity and its associated values by minimising negative impacts through appropriate forest management planning (FSC, 2015a). These standards have principles related to: (1) compliance with laws; (2) workers’ rights and employment conditions; (3) indigenous peoples’ rights; (4) community relations; (5) benefits from the forest; (6) environmental values and impacts; (7) management planning; (8) monitoring and assessment; (9) conservation values; and (10) implementation of management activities. It is argued that these standards (e.g. Soil Association, 2009; Societe Generale de Surveillance, 2011) help forest managers to meet the social, economic and ecological needs of present and future generations by practicing sustainable forest management (SFM) (e.g. see Cerutti et al., 2014; Karmann and Smith, 2009). Specifically, the implementation of measures to protect the forests from illegal resource use and other illegal activities through managing, all activities associated with harvesting and extraction of timber and non-timber forest products so that environmental values are conserved. To implement direct measures to protect species and their habitats for their survival and viability. Such measures include effective forest management planning and social management planning, e.g. harvesting and monitoring plans, forest protection against illegal activities and fires through forest guards. The MCDI and communities in Kilwa introduced forest certification in 2004. The preparation for certification took seven years, and in this timeframe the forest managers knew about certification requirements, and adopted certification management approaches in some of these forests in March 2009. Implementing certification in these forests is a strategy for controlling legal harvesting, and for monitoring illegal harvesting, which is rampant in these areas (e.g. see Milledge et al., 2007; Smith, 2015).

2.2. Research design We used a mixed methods research design (see Creswell, 2013; Lund et al., 2014) to evaluate the influence of forest certification on conserving biodiversity, including six forests and four adjacent villages (Fig. 1). In order to minimise confounding factors, the forests and villages included were chosen based on their similarities in ecological and socioeconomic attributes, so that influence of management intervention due to management change could be assessed. The selection of each village and forest pair based on their proximity to each other. An attempt was made to maximise the distance between the certified and non-certified forests and villages, but within the same agro-ecological zone and similar vegetation types. Biases introduced by interaction effects were minimised, for instance by not selecting certified villages that shared a border with non-certified villages. This was done to strengthen the assumption of a relationship between one village and one forest per pair, so that one can assume that there was no influence between them and they could be investigated as separate case studies. Since no data were available to compare the before and after situation, differences were assessed through a crosssectional study, comparing the forest management regimes at the same point in time. To check whether or not observed differences might have happened even without forest certification, we used a retrospective quasi-experimental design (Schreckenberg and Luttrell, 2009) as a form of space-for-time substitution (Hargrove and Pickering, 1992). Similarities were maximised between the certified and non-certified groups by selecting proxy variables that helped reduce observable biases and systematic differences. Two types of counterfactuals were employed, controls in space (forests) and controls in time (recall within communities), to determine reasons for forest condition differences among the forest management regimes (see Schreckenberg and Luttrell, 2009). The state forest reserves served as a control group and FSC-certified forests and village ‘general land’ forests were used as treatment groups. In this spatial design, the FSC-certified group was compared to noncertified, so that similar values of several proxy variables could be controlled for (e.g. Strauss and Corbin, 1996). A time dimension factor (control in time, i.e. temporal design) was built into the research methods through the participatory rural appraisal (PRA) and household survey questionnaire, focusing on assessing changes related to forest management interventions. Respondents were asked about perceived changes to the biophysical and socioeconomic aspects in the present situation compared to the time before the inception of forest certification. To assess the influence of forest certification, it would be optimal to compare FSC-certified CBFM to non-FSC-certified CBFM forests (Table 1 and Fig. 1), and the inclusion of non-FSC-certified

S.K. Kalonga et al. / Forest Ecology and Management 361 (2016) 1–12

CBFM forests in the sample would have been ideal, but no nonFSC-certified CBFM forests with similar vegetation types in the same agro-ecological zone were available. Nevertheless, the preassessment of forest management before the certificate was granted to Kilwa communities, showed that the forests identified for certification suffered similar ecological problems as other forests in the area, such as for example, lack of selective harvesting to maintain biodiversity (see Soil Association, 2007). 2.2.1. Sampling procedure Biophysical aspects: Stratified systematic sampling design was used. The stratification criterion was ‘Forest Management Regimes’, which resulted in three strata: (1) FSC-certified community forests – treatment; (2) non-FSC-certified ‘de facto’ open access forests – treatment; and (3) non-FSC-certified state forest reserves – control. A reconnaissance survey, based on between plot variability and the forest area, was first done to determine the statistically required number of sample plots per forest. For all forests, the first plot was located randomly at 100 m from the boundary of the forest. Using Global Positioning System (GPS), subsequent plots were located systematically at 500 m, 500 m, 1500 m, 1000 m, 2000 m and 1500 m intervals along transects for Kikole, Kisangi, Likawage, Mchakama, Mitarure and Rungo forests, respectively. The distance between transects varied from 500 m to 5000 m depending on the size of the forest and the required number of sample plots. A total of 131 temporal concentric circular sample plots of 20 m radius (0.126 ha) with subplots of 1 m radius (0.0003 ha) were established. Socioeconomic aspects: Participants in the participatory rural appraisal (PRA) as key informants and in focus groups were identified from updated village register under the guidance of Village Councils (VCs) and Village Natural Resource Committees (VNRCs). The size of focus groups ranged from 6–12 people. To avoid dominance during the discussion and to allow the villagers to talk freely, the leaders were interviewed separately and requested to leave the PRA exercise once their interviews were completed. The PRA data were collected to complement the more quantitative variables collected during interviews, and to better qualify the strengths and possible constraints of institutions at villagers, communities and government levels. Overall, the PRA was considered to provide a reasonably accurate overview of the prevailing social, economic, ecological and institutional conditions in the study villages within a relatively short time period. 2.3. Data collection Adult tree and seedling species were recorded between July and August 2013 as part of large forest inventories for forest structure assessment in Kilwa, to collect tree species data as a proxy for forest biodiversity (e.g. see Kindt and Coe, 2005; Kindt et al., 2006). The first author, assisted with two field assistants (Forest surveyor and Botanist), performed the fieldwork. All adult trees with P3 cm diameter at breast height (dbh) were recorded within plots and all seedlings (610 cm tall) were recorded from subplots. In addition, the first and second nearest tree to the plot centre were selected for measurements of basal diameter (30 cm from ground), dbh and total tree height. Other information recorded at plot level included environmental and human forest use (disturbance) variables as proxies. Elevation (metres a.s.l) was recorded from GPS. Proxies for human forest use variables were forest management regimes, distances from the plot to access road, village, nearby town and main road, number of stumps, stumps basal diameter (65 years old), and fire (proportional area burnt). All the distances (kilometres) were recorded by GPS. Stump basal diameter was measured at 30 cm from the ground. The age of stumps was assessed by local informants who, based on the colour and degree

5

of decay of the cut surface, combined with their knowledge on historically harvesting activities in the respective forest. Proportional area burnt (%) was recorded using visual estimates. Tree species identification was done in the field by a Botanist. Unidentified species were collected, pressed and taken to the herbarium in the Department of Botany, University of Dar es Salaam for identification. The identification was done by matching with herbarium specimens by using Flora of Tropical East Africa (see Edmonds, 2012). The participatory rural appraisal (PRA) approaches (e.g. Adatho, 2011; Creswell, 2013) were used for data collection in the four villages. They generated data and information about history of forest management and resources use patterns, forest policies, laws and regulations, i.e. forest bylaws, along with forest management plans. In addition, key informants’ interviews were administered to individuals and organisations who were involved with forest management to provide information about implementation of forest governance and institutions. Data were collected to establish the differences in forest governance indicators as a proxy for rule compliance (see Kaufmann et al., 2011) in forest management practices implementation. 2.4. Data analyses Biodiversity indices including tree (adult and seedling) species richness, diversity and density were estimated as proxies for forest biodiversity indicators (see Kindt and Coe, 2005). There are often correlations between species richness, diversity and density (Hooper et al., 2005; Magurran and McGill, 2011; Rosenzweig et al., 2011). However, they can be differently affected by human activities (e.g. see Balée, 2014; Folke et al., 2004; Hooper et al., 2005; Krebs, 2001), and it is therefore relevant to include all three in the analyses. Also the relationship between the number of species observed and the number of individuals in a sample is nonlinear, which confounds direct analyses of species per individual ratios from samples of different sizes (Sheil et al., 1999). Species richness as the total number of species observed per plot (Magurran and McGill, 2011) and rarefied species richness were estimated. The rarefied species accumulation curves (with 100 permutations, based on significance level of a = 0.01) were used to compare richness among the forest management regimes. Species density (c), an environmental indicator which is useful for comparing forests of different areas was estimated using the equation, c = s/Az (see Rosenzweig et al., 2011). The s is species richness and A is the total sampled area. Plot number per forest was used as a proxy for forest area (see Rosenzweig et al., 2011). The z is a parameter which accounts for the curvature of species-arearelationships (SAR) (Magurran and McGill, 2011; Guilhaumon et al., 2010), and a value of z = 0.2 which is recommended for many tropical ecosystems was used for this study (e.g. Magurran and McGill, 2011). Species diversity was estimated using Fisher’s alpha diversity index (Magurran, 1988) as a = n(1 s)/s, where s is species richness, n is the number of individuals. The index is a measure of diversity that takes into account variability in stem number (Beck and Schwanghart, 2010) due to human intervention (Balée, 2014). To estimate biomass of removed trees (stumps biomass), dbh for the trees that were cut (stumps) were estimated from dbh-basal diameter equations developed by Kalonga et al. (2015b) for every forest. A local biomass model using dbh as independent variable developed by Mugasha et al. (2013) for Lindi miombo woodlands was used for biomass (t) estimation. Data analyses were preceded with data exploration using a protocol developed by Zuur et al. (2010) to establish appropriate statistical tools to be applied for the analyses. Cleveland dotplots were used to inspect the variables for outliers. No outliers were detected. Pairplots and variance inflation factor (VIF) of 3 were

S.K. Kalonga et al. / Forest Ecology and Management 361 (2016) 1–12

3.1. Biodiversity indicators A total of 14,346 individual trees were recorded, within 188 species, 115 genera and 42 families. Dominant species were Grewia conocarpa K. Shum and Diplorhynchus condylocarpon (Müll. Arg.) Pichon, while Combretum, Grewia, and Fabaceae, Combretaceae were dominant genera and families, respectively. No introduced species was observed, i.e. all of them were endemic/core species. There were differences in adult trees and seedlings species richness, diversity and density (Fig. 2A–F and Appendix A1) among the forest management regimes (FMRs). Adult tree species (Fig. 2A) richness was higher in certified forests (FSC) and significantly different (F2,128 = 60.5, p = 0.0000) from open access forests (OCF) and state forest reserves (FRS). Diversity (Fig. 2B) was higher

p-values

OCF

FRS

FSC

FMR

OCF

30

FSC

OCF

12

F

6

8

10

Bonf, p = 1.0000

4

Seedling species density

4

6

8

Bonf, p = 1.0000

E

0

0 FSC

FRS

FMR

2

0.5

1.0

1.5

p-values FSC-CFO=0.0440 SFR-CFO=1.0000 SFR-FSC=0.0450

0.0

FRS

20

OCF

FMR

Seedling species diversity

2.0

FMR

D

FSC-OCF=0.0000 SFR-OCF=0.9306 SFR-FSC=0.0000

0

0 FSC

C

10

6

8

Adult species density

40

FSC-OCF=0.0000 SFR-OCF=0.9133 SFR-FSC=0.0000

10

12

p-values

B

4

Adult species diversity

15 10 5

Adult species richness

FSC-OCF=0.0000 SFR-OCF=0.0366 SFR-FSC=0.0000

0

FRS

Seedling species richness

3. Results

14

p-values

A

Full models were fitted initially and then simplified by backwards selection using stepwise procedure to eliminate the insignificant variables basing on Akaike’s Information Criterion (AIC) for gaussian models. Drop one procedure was used for poisson and quasipoisson models by manually removing insignificant variables basing on p-values to get the final models. All these analyses were done in R version 3.2.0 (R Core Team, 2015) employing vegan (Oksanen et al., 2013), MASS (Venables and Ripley, 2002), and visreg (Breheny et al., 2014) packages. A p-value 60.05 indicated statistical significance.

2

20

used to assess collinearity, while multi-panel scatterplots were used to visualise relationships. Furthermore, relationship between response and explanatory variables were examined using generalised linear models (GLM). The species richness, diversity and density of the different forest management regimes were first visualised using boxplots. Oneway ANOVA, followed by Post hoc test for multiple comparisons (Tukey Honestly Significant Differences/HSD) were used to test for differences between parametric variables. For non-parametric variables, Kruskal–Wallis/Bonferroni t-test was used. Due to collinearity, distances to nearby town and village variables were not included in the GLM analyses. Response variables in the models were species richness, diversity and density of adult trees and seedlings. The explanatory variables were elevation; forest management regimes, i.e. certified community forests (FSC), open access community forests (OCF) and state forest (FRS) reserves; distances to access road; main road; stump biomass and fire (i.e. area burnt), with two-way interactions of all variables. Models were fitted separately for each response variable. The GLM with quasipoisson distribution was used to model the relationship between human forest use variables and adult tree species richness and density among the forest management regimes. The GLM with gaussian distribution was used to model the relationship between human forest use variables and tree diversity. The GLM with poisson distribution was used for seedling species richness and diversity, whereas quasipoisson distribution was used for seedling density.

2

6

FRS

FSC

FMR

OCF

FRS

FSC

OCF

FMR

Fig. 2. Adult tree (A–C) and seedling (D–F) species richness, diversity and density among forest management regimes (FMR) which are state forest reserves (FRS), certified forests (FSC), open access forests (OCF). FRS stands for Mitarure and Rungo forests, FSC stands for Kikole and Kisangi forests and OCF stands for Likawage and Mchakama forests as detailed in Fig. 1and Table 1.

7

S.K. Kalonga et al. / Forest Ecology and Management 361 (2016) 1–12

in certified forests (FSC) and significantly different (F2,128 = 40.62, p = 0.0000) from open access forests and state forest reserves. Similarly, density (Fig. 2C) was also higher in certified forests (FSC) and significantly different (F2,128 = 21.26, p = 0.0000) from open access forests. Seedling species richness was high in certified forests and slightly significantly different (KW = 7.6216, df = 2, p = 0.0221, Fig. 2D) from open access forests and state forest reserves. Diversity and density did not differ among forest management regimes (Fig. 2E and F). Species richness, diversity and density were highly correlated, r = 0.9 for adults and r = 0.7 for seedlings, but were affected differently by human forest use depending on tree species. With reference to state forest reserves (FRS), adult tree species richness, diversity and density decreased with distance to main road, whereas interactions between forest management regimes and distance to main road showed that species richness, diversity and density increased with distance to main road in the open access forests (OCF) (Table 2, Fig. 3A–C). The interactions between forest management regimes and distance to access road showed that species richness, diversity and density decreased with distance to access road in the open access forests (OCF) (Table 2 and Fig. 3D–F). The interaction between forest management regimes and stumps biomass (i.e. harvesting intensity) showed that species

diversity increased with harvesting intensity in the certified forests (FSC) (Table 2 and Fig. 3G). Seedling species richness, diversity and density were generally negatively related to fires (i.e. fire decreased seedling species richness, diversity and density in all the forest management regimes), whereas species richness and diversity were positively related to management intervention (i.e. human forest use) in certified forests (FSC) and open access forests (OCF), respectively, but had no relationship with species density in both certified and open access regimes (Table 2 and Fig. 4A–C). Species diversity generally decreased with distance to access road, whereas interactions between forest management regimes and distance to access road showed that species richness and density decreased with distance to access road in certified forests (FSC) and open access forests (OCF), respectively (Table 2 and Fig. 4D–F). 3.2. Forest governance indicators Participatory rural appraisal (PRA), i.e. focus group discussions and key informants, illuminated how forest conditions have been changing over time for the different forests as a result of human forest use. Kikole and Mchakama villagers reported that illegal timber and fuelwood extraction had been the main human impact factors of the Mitarure forests, whereas in Likawage, in addition to the

Table 2 GLM regressions on the relationship between adult and seedling species richness, diversity and density against human forest use variables among the forest management regimes certified community forests (FSC), open access community forests (OCF) and state forest reserves (FRS). Coefficients

Adults

Seedlings

Estimate

Std. Error

Pr(>|t|)

Estimate

Std. Error

Pr(>|z|)

Species richness: (Intercept) Access road Main road Fires FSC OCF Access road:FSC Access road:OCF Main road:FSC Main road:OCF

2.3512 0.0177 0.0062 – 0.2110 0.2030 0.0213 0.0600 0.0083 0.0107

0.1052 0.0152 0.0018 – 0.4746 0.1303 0.0237 0.0191 0.0244 0.0024

2e 16*** 0.2490 0.0008*** – 0.6574 0.1218 0.3699 0.0021** 0.7348 1.82e 05***

1.0051 0.1203 – 30.467 1.4022 0.5924 0.3688 0.1521 – –

0.4986 0.0748 – 8.5078 0.6101 0.6148 0.1654 0.0926 – –

0.0438* 0.1080 – 0.0003*** 0.0215* 0.3353 0.0258* 0.1006 – –

Species diversity: (Intercept) Access road Main road Fires StumpsB FSC OCF Access road:FSC Access road:OCF Main road:FSC Main road:OCF StumpsB:FSC StumpsB:OCF

5.4218 0.0370 0.0286 – 0.7927 1.3660 2.0335 0.0079 0.2669 0.0258 0.0555 7.7341 0.8236

0.7635 0.0987 0.0122 – 0.5164 4.2574 0.9441 0.1888 0.1293 0.2208 0.0171 2.5100 1.0545

9.74e 11*** 0.7086 0.0209* – 0.1274 0.7489 0.0333* 0.9669 0.0413* 0.9073 0.0015** 0.0026** 0.4363

0.2014 0.0970 17.352 – 0.3084 0.6597 – – – – – –

0.2967 0.0395 – 4.5662 – 0.3270 0.3196 – – – – – –

0.4972 0.0139* – 0.0001*** – 0.3457 0.0390* – – – – – –

Species density: (Intercept) Access road Main road Fires FSC OCF Access road:FSC Access road:OCF Main road:FSC Main road:OCF

3.1954 0.0142 0.0064 – 0.8558 0.2482 0.0336 0.0641 0.0419 0.0098

0.1390 0.0202 0.0024 – 0.6682 0.1742 0.0326 0.0264 0.0348 0.0033

2e 16*** 0.4840 0.0085* – 0.2027 0.1567 0.3043 0.0167* 0.2310 0.0035**

1.0200 0.1398 – 32.293 0.7183 0.6462 0.2662 0.1746 – –

0.4307 0.0631 – 7.8075 0.5746 0.5314 0.1551 0.0793 – –

Pr(>|t|) 0.0194* 0.0286* – 6.46e 05** 0.2136 0.2263 0.0887 0.0296* – –

The intercepts include state forest reserves. StumpsB stands for stumps biomass. *** Signif. at a = 0.001. ** Signif. at a = 0.01. * Signif. at a = 0.05.

–

8

S.K. Kalonga et al. / Forest Ecology and Management 361 (2016) 1–12

Fig. 3. The relationship between human forest use variables: distance to main road with adult species richness, diversity and density (A–C); distance to access road (D–F); and stumps biomass (G) among the forest management regimes. Colours denote forest management regimes: Red for non-certified state forest reserves (FRS) as reference, Green for certified forests (FSC) and Blue for non-certified open access community forests (OCF). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

above, uncontrolled forest fires had been the main human impact factors of Likawage and Rungo forest resources decline. Kikole and Kisangi villagers reported that charcoal production, establishment of settlements and uncontrolled forest fires had been the main cause of Kikole and Kisangi forest resources decline. The illegal timber extraction targeted specific tree species regardless of stock and allowed harvestable size in non-certified forests, while legal timber extraction in certified community forests considered tree species stock, distribution and allowed harvestable size. Certified forests (FSC) in Kikole and Kisangi had forest management and harvesting plans in place, while open access forests (OCF) in Likawage and Mchakama, and state forest reserves (FRS) in Mitarure and Rungo had none. In addition, FSC villages had forest bylaws which were operational and effective into some extent to prevent illegal harvesting as compared to non-FSC villages. Generally, from forest inventory (stumps count), harvesting intensity for the past five years was higher in non-certified (OCF: 22% and FRS: 68%) than certified forests (FSC: 9.5%). 4. Discussion Our results show that biodiversity indices are higher in certified than in non-certified forests studied in Tanzania. This could be due

to effective harvesting plans and assessment of timber stocks, as supported by certification audit reports saying that harvesting is selective based on a rational of removing target species and trees of a minimum defined size and felling girth (Soil Association, 2013). A study by Medjibe et al. (2013) in the Congo Basin also showed that certified forests exhibit higher biodiversity indicators than non-certified forests because of effective implementation of forest management plans. On the contrary, another study from Tanzania reported high intensity of human disturbance in state forest reserves as a result of inadequate forest management practices (Sawe et al., 2014). Results from our study further show that forest access, fires and forest management are significantly related to biodiversity indicators in certified forests, and that harvesting intensity is lower in the certified forests than in the non-certified forests. This is possibly due to effective implementation of forest management planning and practices in certified forests. Human activities are regulated compared to non-certified forests, which lack forest management plans. This is consistent with certification audit reports that harvesting is based on selective felling regulated by quotas (Soil Association, 2013). The lower harvesting intensity is also due to closer local protection of the forest by the villagers through forest guards (Soil Association, 2013) as a result of the economic benefits

S.K. Kalonga et al. / Forest Ecology and Management 361 (2016) 1–12

9

Fig. 4. The relationship between human forest use variables: fires with seedling species richness, diversity and density (A–C); and distance to access road (D–F) among the forest management regimes. Colours denote forest management regimes: Red for non-certified state forest reserves (FRS) as reference, Green for certified forests (FSC) and Blue for non-certified open access community forests (OCF). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

they gain from the forests under CBFM and FSC (e.g. Kalonga et al., 2015a). Tree adult and seedling species density as well as adult species richness and diversity decrease with distance from the forests to access roads in open access forests (OCF) in our study area, suggesting that illegal cutting is more likely inside the forest than along edges. People encroach forests with easy access from adjacent villages, possibly due to ineffective forest management, which is as a result of, among others, inadequate economic benefits villagers receive from forest products, e.g. harvesting licenses (see Kalonga et al., 2015a). This implies that forest management is not effective in conserving biodiversity in open access forests (OCF) compared to certified forests (FSC). It has also been assumed that the closer the forest is to main roads, the higher is the encroachment in forests with inadequate management (e.g. Chidumayo, 2002). The increase in adult tree species richness, diversity and density with distance to main road in open access forests (OCF), could be attributed to physiographic factors. The Mavuji River makes Mchakama forest inaccessible, and also Likawage being far away from the main road. This corroborates Sassen and Sheil (2013) who found that part of the forests which are far away from reach of people, have higher tree species richness and diversity. Human activities strongly influence ecological processes from local to global scales (Folke et al., 2004; Hooper et al., 2005; Krebs, 2001). Intensive human use of certain highly valued species may have a greater impact on diversity than less intense forest use (e.g. Ndangalasi et al., 2007). The higher tree species diversity in the certified forests (FSC), as shown by high biomass of harvested trees, implies that forest use is less intense due to effective harvesting plans, corroborating certification audit reports which indicate gradual biodiversity maintenance and/or enhancement in the area (e.g. Soil Association, 2013). Studies show that communitybased forest management may be an effective conservation tool by itself (e.g. see Berkes, 2006, 2007; Bromley, 1992; McKean, 1992; Ostrom, 1990). However, Rametsteiner (2002) argues that

forest certification changes the forestry sector more intensely than many governmental initiatives. The FSC standards, which are more stringent and more detailed than national laws, may serve as an incentive to comply with the bylaws or even to help enforce the laws. They are also regularly updated, controlled and verified through third party auditors. The national laws do not uphold as high a standard and are weakly implemented and verified (Petersen and Sandhövel, 2001). Harvesting plans employing selective logging based on tree species stocking and allowable cut suggest that forest certification conserves diversity in certified forests compared to non-certified forests. The PRA information on historical forest management and use reveal that open access forests (OCF) and state forest reserves (FRS) are more affected by human use, with more illegal activities for timber production targeting large trees. Ecologically, the artificial reduction of large trees is a concern, as large trees affect micro-climate, and are important as habitat for other flora and fauna as well as for forest regeneration (Clark and Clark, 1996; Sagar and Singh, 2005). Furthermore, selective logging of preferred tree species without considering species stocking and size may result in less seed production. A main factor affecting seedlings survival, among others, is forest fire (e.g. see Gambiza et al., 2000; Mapaure and Moe, 2009; Ryan and Williams, 2011). Fire affects trees differently depending on species and age; young trees been affected the most (Ryan and Williams, 2011). The decrease in seedling species richness and diversity in the forests in our study could be attributed to fire effects. This is supported in the present study as none of the management regimes had not registered fire incidences, corroborating a study by Miteva et al. (2015) in Southeast Asia that forest certification had no positive effect on fire management. Certification audit reports in Kilwa show that the forest managers have not been effective in controlling forest fires from outside the forest management units causing a threat to the certified forests (Soil Association, 2013). However, certified forests in Kilwa had registered less fire incidences (FSC: 9%) than non-certified forests

10

S.K. Kalonga et al. / Forest Ecology and Management 361 (2016) 1–12

Fig. A1. Rarefied species accumulation curves per forest management regimes with standard deviation. Colours denote forest management regimes: Red for non-certified state forest reserves (FRS), Green for certified forests (FSC) and Blue for non-certified open access community forests (OCF). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

(OCF: 36% and FRS: 45%), suggesting that the use of forest guards in certified forests helps to suppress forest fires (see Soil Association, 2013). This implies that certified forests may enhance seedlings recruitment and survival, which in turn result in high adults’ species density compared to non-certified forests. Species density provides indicators of environmental conditions that are easy to comprehend and use, and serves as an environmental indicator of the effect of human forest use (see Rosenzweig et al., 2011). The different levels of tree species richness, diversity and density among the forest management regimes could also be attributed to previous over-harvesting, or to relatively high present harvesting levels in these forests, or both. Thus, the cause and effect of human forest use on forest biodiversity in the certified forests in the present study are difficult to determine. Although the preparation for forest certification took a while, certification has been operational in the area for a short time (four years) making it hard to fully attribute the findings to certification effects alone. However, there is a positive and significant relationship between biodiversity indicators and human forest use practices in certified forests compared to non-certified forests. The findings of this study suggest that forest certification as a policy option may help to conserve biodiversity. The study is one of the first in Africa that contributes to the limited empirical evidence on the influence of forest certification on conserving biodiversity. Although, the causal relationships in this study are limited, the findings may serve as a baseline for further research on the contribution of forest certification in conserving biodiversity at both temporal and spatial scales in the tropics. 5. Conclusion Human forest use practices related to forest certification appear to conserve biodiversity. There are higher biodiversity indicators (species richness, density and diversity) in FSC-certified community forests than in open access community forests (OCF) and state forest reserves (FRS). These findings suggest that forest certification may be a policy option to help conserve biodiversity. This implies that the adoption and incorporation of forest certification standards in the forest policy and legal framework, could be one of the sustainable forest management tools in conserving biodiversity in Tanzania and in the tropics as a whole. The certification experiences from certified community forests could then be extended to landscape level mosaics of forests under other uses, and to the various products they provide. Because of the sample size in this study, and the short duration of time the certification has been operational, it is hard to identify the precise effects of the certification. Nevertheless, our findings may serve as a baseline for future research at temporal and spatial scales on building up empirical evidence on the contribution of certified forests on conserving biodiversity.

Acknowledgements This study was funded by the Climate Change Impacts, Adaptation and Mitigation (CCIAM) Programme in Tanzania. We are grateful to the field assistants Mr. Rashid Matanda (forest inventory), Mr. Selemani Haji (tree identification), and Mr. Crawford Ndanshau (driver). The suggestions made by two anonymous referees and the editor are also gratefully acknowledged.

Appendix A See Fig. A1.

References Adatho, M., 2011. Combining Quantitative and Qualitative Methods for Program Monitoring and Evaluation: Why are Mixed-Method Designs Best? World Bank, USA. Alkemade, R., van Oorschot, M., Miles, L., Nellemann, C., Bakkenes, M., ten Brink, B., 2009. GLOBIO3: a framework to investigate options for reducing global terrestrial biodiversity loss. Ecosystems 12, 374–390. Auld, G., Gulbrandsen, L.H., McDermott, C.L., 2008. Certification schemes and the impacts on forests and forestry. Annu. Rev. Environ. Resour. 33, 187–211. Backéus, I., Pettersson, B., Strömquist, L., Ruffo, C., 2006. Tree communities and structural dynamics in miombo (Brachystegia–Julbernardia) woodland, Tanzania. For. Ecol. Manage. 230, 171–178. Balée, W. (Ed.), 2014. Historical Ecology and the Explanation of Diversity: Amazonian Case Studies. Springer. BALL, S., 2010. Biodiversity and certified community forests in Tanzania. In: Sheil, D., Francis, E., Roderick, J. (Eds.) Biodiversity conservation in certified forests. Ball, S., 2011. Putting the Forestry into Participatory Forest Managmenet – Simple Inventory Protocols for Sustainable Logging. In: Geldenhuys, C., Ham, C., Ham, H. (Eds.), Sustainable Forest Management in Africa. Stellenbosch. BALL, S.M.J. Putting Forestry into PFM – Simple inventory protocol for sustainable logging. In: Nshubemuki, L., Chamshama, S.A.O., Mbwambo, L.A., Balama, C., (Eds.), Proceedings of the First PFM Research Workshop: PFM for Improved Forest Quality. Livelhoods and Governance, 2009 Morogoro, Tanzania: 15–16 June, 2009. Tanzania Forest Research Institute (TAFORI). Beck, J., Schwanghart, W., 2010. Comparing measures of species diversity from incomplete inventories: an update. Meth. Ecol. Evol. 1, 38–44. Berkes, F., 2006. From community-based resource management to complex systems: the scale issue and marine commons. Ecol. Soc. 11, 45. Berkes, F., 2007. Community-based conservation in a globalized world. Proc. National Acad. Sci. 104, 15188–15193. Biggs, R., Simons, H., Bakkenes, M., Scholes, R.J., Eickhout, B., van Vuuren, D., Alkemade, R., 2008. Scenarios of biodiversity loss in southern Africa in the 21st century. Glob. Environ. Change 18, 296–309. Blackman, A., Raimondi, A., Cubbage, F., 2014. Does Forest Certification in Developing Countries Have Environmental Benefits? Insights from Mexican Corrective Action Requests. Insights from Mexican Corrective Action Requests (March 31, 2014). Resources for the Future Discussion Paper. Blackman, A,. Rivera, J., 2010. The evidence base for environmental and socioeconomic impacts of ‘‘sustainable” certification. Discussion paper, Resources for the future RFF DP 1017. Washington DC, USA. p. 34. Breheny, P., Burchett, W., Breheny, M.P., 2014. Package ‘visreg’: Visualization of regression models. Bromley, D.W. (Ed.), 1992. Making the Commons Work: theory, Practice and Policy. Institute for Contemporary Studies Press, San Francisco. Burgess, N., Clarke, G. (Eds.), 2000. Coastal forests of eastern Africa: IUCN.

S.K. Kalonga et al. / Forest Ecology and Management 361 (2016) 1–12 Campbell, B., Angelsen, A., Cunningham, A., Katerere, Y., Sitoe, A., Wunder, S., 2007. Miombo Woodlands–Opportunities and Barriers to Sustainable Forest Management. CIFOR, Bogor, Indonesia. Campbell, B.M., 1996. The Miombo in Transition: Woodlands and Welfare in Africa. CIFOR, Indonesia. Cerutti, P., Lescuyer, G., Tsanga, R., Kassa, S., Mapangou, P., Mendoula, E., MissambaLola, A., Nasi, R., Eckebil, P., Yembe, R., 2014. Social Impacts of the Forest Stewardship Council Certification: An Assessment in the Congo Basin. Center for International Forestry Research (CIFOR). Chidumayo, E., 2002. Changes in miombo woodland structure under different land tenure and use systems in central Zambia. J. Biogeogr. 29, 1619–1626. Chidumayo, E., Okali, D., Kowero, G., Larwanou, M., 2011. Climate Change and African Forest and Wildlife Resources African Forest Forum. Nairobi, Kenya. Chidumayo, E.N., Gumbo, D.J., 2010. The Dry Forests and Woodlands of Africa: Managing for Products and Services. Earthscan. Clark, D.B., Clark, D.A., 1996. Abundance, growth and mortality of very large trees in neotropical lowland rain forest. For. Ecol. Manage. 80, 235–244. Creswell, J.W., 2013. Research Design: Qualitative, Quantitative, and Mixed Methods Approaches. Sage. Cubbage, F., Diaz, D., Yapura, P., Dube, F., 2010. Impacts of forest management certification in Argentina and Chile. For. Policy Econo. 12, 497–504. Ebeling, J., Yasué, M., 2009. The effectiveness of market-based conservation in the tropics: forest certification in Ecuador and Bolivia. J. Environ. Manage. 90, 1145– 1153. Edmonds, J.M., 2012. Flora of Tropical East Africa – Series. Royal Botanic Gardens. FAO, 2006. Global Forest Resources Assessment 2005: Progress Towards Sustainable Forest Management. Food and Agriculture Organization of the United Nations, Rome. FAO, 2010. Global Forest Resources Assessment 2010. FAO, Rome. Fennell, L., 2011. Ostrom’s law: property rights in the commons. Int. J. Commons 5, 9–27. Folke, C., Carpenter, S., Walker, B., Scheffer, M., Elmqvist, T., Gunderson, L., Holling, C., 2004. Regime shifts, resilience, and biodiversity in ecosystem management. Annu. Rev. Ecol., Evol., Syst., 557–581 Frost, P., 1996. The Ecology of Miombo Woodlands. In: Campbell, B. (Ed.) The Miombo in Transition: Woodlands and Welfare in Africa. CIFOR, Bogor. FSC, 2015a. Forest Stewardship Council Principles and Criteria for Forest Stewardship. FSC-STD-01-001 V5-2 [Online]. Bonn/Germany. . FSC, 2015b. Forest Stewardship Council: Mission and Vision. https://us.fsc.org/ mission-and-vision.187.htm [Online]. Bonn/Germany: FSC. Available: (accessed 3 September 2015). Gambiza, J., Bond, W., Frost, P., Higgins, S., 2000. A simulation model of miombo woodland dynamics under different management regimes. Ecol. Econo. 33, 353–368. Gardner, T.A., Barlow, J., Chazdon, R., Ewers, R.M., Harvey, C.A., Peres, C.A., Sodhi, N. S., 2009. Prospects for tropical forest biodiversity in a human-modified world. Ecol. Lett. 12, 561–582. Guilhaumon, F., Mouillot, D., Gimenez, O., 2010. MmSAR: an R-package for multimodel species–area relationship inference. Ecography 33, 420–424. Hargrove, W.W., Pickering, J., 1992. Pseudoreplication: a sine qua non for regional ecology. Landsc. Ecol. 6, 251–258. Hooper, D., Chapin III, F., Ewel, J., Hector, A., Inchausti, P., Lavorel, S., Lawton, J., Lodge, D., Loreau, M., Naeem, S., 2005. Effects of biodiversity on ecosystem functioning: a consensus of current knowledge. Ecol. Monogr. 75, 3–35. Howell, K.M., Msuya, C.A., Mligo, C., Werema, C., Kihaule, P., Honorati, M.K., Suleiman, H.O., 2012. Biodiversity Surveys of Poorly Known Coastal Forests of Southeastern Tanzania and Zanzibar. WWF Tanzania, Dar es Salaam. Kalonga, S., Kulindwa, K., Mshale, B., 2015a. Equity in distribution of proceeds from forest products from certified community-based forest management in Kilwa District, Tanzania. Small-scale For. 14, 73–89. Kalonga, S.K., Midtgaard, F.M., EID, T.H., 2015b. Does forest certification enhance forest structure? Empirical evidence from certified community-based forest management in Kilwa District, Tanzania. J. Int. For. Rev. 17, 182–194. Karmann, M., Smith, A., 2009. FSC reflected in scientific and professional literature. Literature Study on the Outcomes and Impacts of FSC Certification. FSC Policy Series, Bonn/Germany. Kaufmann, D., Kraay, A., Mastruzzi, M., 2011. The worldwide governance indicators: methodology and analytical issues. Hague J. Rule Law 3, 220–246. KDC, 2008. Socio-economic profile of Kilwa district. Kilwa, Lindi: Lindi District Council. Kim, D.H., Sexton, J.O., Townshend, J.R., 2015. Accelerated deforestation in the humid tropics from the 1990s to the 2000s. Geophys. Res. Lett. Kindt, R., Coe, R., 2005. Tree Diversity Analysis: A Manual and Software for Common Statistical Methods for Ecological and Biodiversity Studies. World Agroforestry Centre Eastern and Central Africa Program. Kindt, R., van Damme, P., Simons, A., 2006. Patterns of species richness at varying scales in western Kenya: planning for agroecosystem diversification. Biodiver. Conserv. 15, 3235–3249. Krebs, C.J., 2001. Ecology: The Experimental Analysis of Distribution and Abundance, California, USA., Benjamin Cummings, San Francisco, CA. Kukkonen, M., Rita, H., Hohnwald, S., Nygren, A., 2008. Treefall gaps of certified, conventionally managed and natural forests as regeneration sites for Neotropical timber trees in northern Honduras. For. Ecol. Manage. 255, 2163– 2176.

11

Levy, M., Rhoe, V., Balmford, A., 2005. Ecosystems and Human Well-Being. vol. 1: Current State and Trends: Ecosystem Change and Human Wellbeing. Lillesø, J.-P.B., van Breugel, P., Kindt, R., Mbago, F., Moshi, H.N., Ndangalasi, H.J., Uronu, L.O.N., Jamnadass, R., Graudal, L., 2014. Potential Natural Vegetation of Eastern Africa. Atlas and Tree Species Composition for Tanzania, vol. 10. University of Copenhagen: Department of Geosciences and Natural Resource Management. Lund, J., Burgess, N., Chamshama, S., Dons, K., Isango, J., Kajembe, G., Meilby, H., Moyo, F., Ngaga, Y., Ngowi, S., Njana, M., Mwakalukwa, E., Skeie, K., Theilade, I., Treue, T., 2014. Mixed method approaches to evaluate conservation impact: evidence from decentralized forest management in Tanzania. Environ. Conserv., 9 Magurran, A.E., 1988. Ecological Diversity and Its Measurement. Springer. Magurran, A.E., Mcgill, B.J. (Eds.), 2011. Biological Diversity: Frontiers in Measurement and Assessment. Oxford University Press, Oxford. Mapaure, I., Moe, S.R., 2009. Changes in the structure and composition of miombo woodlands mediated by elephants (Loxodonta africana) and fire over a 26-year period in north-western Zimbabwe. Afric. J. Ecol. 47, 175–183. Marx, A., Cuypers, D., 2010. Forest certification as a global environmental governance tool: what is the macro-effectiveness of the Forest Stewardship Council? Regul. Gover. 4, 408–434. Mckean, M.A., 1992. Management of traditional common lands (Iriaichi) in Japan. In: Bromley, D.W. (Ed.), Making the Commons Work: Theory, Practice, and Policy. ICS Press, San Francisco. Medjibe, V., Putz, F.E., Romero, C., 2013. Certified and uncertified logging concessions compared in gabon: changes in stand structure, tree species, and biomass. Environ. Manage. 51, 524–540. Milledge, S., Gelvas, I., Ahrends, A., 2007. Forestry, Governance and National Development: Lessons Learned from a Logging Boom in Southern Tanzania. TRAFFIC East/Southern Africa, Tanzania Development Partners Group, Tanzania Ministry of Natural Resources and Tourism, Dar es Salaam. Miteva, D.A., Loucks, C.J., Pattanayak, S.K., 2015. Social and environmental impacts of forest management certification in Indonesia. PloS one 10, e0129675. MNRT, 2008. Ministry of Natural Resources and Tourism: Participatory forest management in Tanzania. Facts and figures. Dar es Salaam: Government Printer. Mugasha, W.A., Eid, T., Bollandsås, O.M., Malimbwi, R.E., Chamshama, S.A.O., Zahabu, E., Katani, J.Z., 2013. Allometric models for prediction of above-and belowground biomass of trees in the miombo woodlands of Tanzania. For. Ecol. Manage. 310, 87–101. Muthoo, M., 2012. Forest certification and the green economy. Unasylva (FAO) 63, 17–23. Mwase, W.F., Bjørnstad, Å., Bokosi, J.M., Kwapata, M.B., Stedje, B., 2007. The role of land tenure in conservation of tree and shrub species diversity in miombo woodlands of southern Malawi. New For. 33, 297–307. Ndangalasi, H.J., Bitariho, R., Dovie, D.B., 2007. Harvesting of non-timber forest products and implications for conservation in two montane forests of East Africa. Biol. Conserv. 134, 242–250. Oksanen, J., Blanchet, F.G., Kindt, R., Oksanen, M.J., Suggests, M., 2013. Package ‘vegan’. Community Ecology Package Version. Ostrom, E., 1990. Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press. Perkin, A., Leonard, C., Doggart, N., 2008. Landscape Profile: Kilwa. Tanzania Forest Conservation Group, Dar es Salaam. Petersen, L., Sandhövel, A., 2001. Forestry policy reform and the role of incentives in Tanzania. For. Policy Econo. 2, 39–55. Poulsen, J.R., Clark, C.J. (Eds.), 2010. Congo Basin timber certification and biodiversity conservation. Rametsteiner, E., 2002. The role of governments in forest certification—a normative analysis based on new institutional economics theories. For. Policy Econo. 4, 163–173. Rosenzweig, M., Donoghue, J., Li, Y., Yuan, C. (Eds.), 2011. Estimating Species Density. Oxford University Press, New York. Ryan, C.M., Williams, M., 2011. How does fire intensity and frequency affect miombo woodland tree populations and biomass? Ecol. Appl. 21, 48–60. SA, 2007. Soil Association Woodmark: Summary of strengths and weaknesses identified in the pre-assessment audit of Mpingo Conservation Proiect by Woodmark Octobet 2007-10-20. [Online]. South Plaza, uk: SoiAl ssociation Certification LImited. SA, 2009. Soil Association Woodmark: Forest Management Standards – Woodmark Generic Standard and Checklist adapted for Tanzania. [Online]. SA, 2013. Soil Association Woodmark: MCDI SLIMF Certification Public Report. [Online]. South Plaza, UK: Soil Association Certification Ltd. Sagar, R., Singh, J., 2005. Structure, diversity, and regeneration of tropical dry deciduous forest of northern India. Biodivers. Conserv. 14, 935–959. Sasaki, N., Asner, G.P., Knorr, W., Durst, P.B., Priyadi, H.R., Putz, F.E., 2011. Approaches to classifying and restoring degraded tropical forests for the anticipated REDD+ climate change mitigation mechanism. iFor.-Biogeosci. For. 4, 1. Sassen, M., Sheil, D., 2013. Human impacts on forest structure and species richness on the edges of a protected mountain forest in Uganda. For. Ecol. Manage. 307, 206–218.

12

S.K. Kalonga et al. / Forest Ecology and Management 361 (2016) 1–12

Sawe, T.C., Munishi, P.K., Maliondo, S.M., 2014. Woodlands degradation in the Southern Highlands, Miombo of Tanzania: Implications on conservation and carbon stocks. Int. J. Biodivers. Conserv. 6, 230–237. Schreckenberg, K., Luttrell, C., 2009. Participatory forest management: a route to poverty reduction? Int. For. Rev. 11, 221–238. Sebukeera, C., Muramira, E., Mumokama, C., Elkholei, A., Elbagouri, I., Masumbuko, B., Rabesahala, 2005. Forests and Woodlands: Africa Environment Outlook 2. Our Environment. Our Wealth. UNEP. Sell, J., Koellner, T., Weber, O., Proctor, W., Pedroni, L., Scholz, R.W., 2007. Ecosystem services from tropical forestry projects–the choice of international market actors. For. Policy Econo. 9, 496–515. SGS, 2011. Société General de Surveillance: SGS QUALIFOR Forest Management Certification Report. . Sheil, D., Putz, F.E., Zagt, R.J., 2010. Biodiversity conservation in certified forests, Tropenbos International. Sheil, D., Sayer, J.A., O’Brien, T., 1999. Tree species diversity in logged rainforests. Science 284, 1587. Smith, D., 2015. Environmental sustainability. Tanzania: illegal logging threatens tree species with extinction. The Guardian. Strauss, A., Corbin, J., 1996. Basics of Qualitative Research: Techniques and Procedures for Developing Grounded Theory. Sage publications.

Tallis, H., Ricketts, T., Guerry, A., Nelson, E., Ennaanay, D., Wolny, S., Olwero, N., Vigerstol, K., Pennington, D., Mendoza, G., 2011. InVEST 2.0 beta User’s Guide. The Natural Capital Project. Stanford. Team, R.C., 2015. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL . Timonen, J., Gustafsson, L., Kotiaho, J.S., Mönkkönen, M., 2011. Hotspots in cold climate: conservation value of woodland key habitats in boreal forests. Biol. Conserv. 144, 2061–2067. URT, 1998. The Forest Plocy. In: Division, F.A.B. (Ed.), Dar es Salaam: Government Printer. Van kuijk, M., Putz, F., Zagt, R., 2009. Effects of forest certification on biodiversity, Wageningen, Tropenbos International. Venables, W.N., Ripley, B.D., 2002. Modern Applied Statistics with S, fourth ed. Springer, New York. Zahabu, E., Eid, T., Kajembe, G., Mbwambo, L., Mongo, C., Sangeda, A., Malimbwi, R., Katani, J., Kashaigili, J., Luoga, E., 2009. Forestland tenure systems in Tanzania: an overview of policy changes in relation to forest management [Eiendomsforholdene for skogarealer i Tanzania: en oversikt over policyendringer i skogforvaltningen]. INA-fagrapport No. 14. Zuur, A.F., Ieno, E.N., Elphick, C.S., 2010. A protocol for data exploration to avoid common statistical problems. Meth. Ecol. Evol. 1, 3–14.