phase regeneration (Busing, 1991; Shugart and Smith 1992; Urban and Shugart, 1992;. Mashapa et ...... For each tree rooted in the plot, the following were recorded; canopy depth (CD) (m) and greatest ...... Photo credit: Britney-Leosa Mashapa.
CHINHOYI UNIVERSITY OF TECHNOLOGY
HUMAN LIVELIHOODS AND SUSTAINABLE CONSERVATION: HERBIVORY AND ANTHROPOGENIC IMPACTS ON WOODY VEGETATION AND ECOSYSTEM GOODS IN SAVE VALLEY, SOUTHEASTERN LOWVELD OF ZIMBABWE
by
CLAYTON MASHAPA
A thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy in Natural Resource Management in the School of Wildlife, Ecology and Conservation at Chinhoyi University of Technology, Chinhoyi, Zimbabwe
August 2018
i
APPROVAL FORM The undersigned certify that they have read and recommended to the School of Wildlife, Ecology and Conservation, Chinhoyi University of Technology, for acceptance; a thesis titled,
“Human
Livelihoods
and
Sustainable
Conservation:
Herbivory
and
anthropogenic impacts on woody vegetation and ecosystem goods in Save Valley, southeastern lowveld, Zimbabwe’’ submitted by Clayton Mashapa in fulfilment of the requirements for the Doctor of Philosophy degree in Natural Resource Management.
Signature……………………………………………..Date………………………………...... Prof. Dr. Edson Gandiwa (Principal supervisor)
Signature……………………………………………..Date………………………………...... Prof. Dr. Never Muboko (Co-supervisor)
ii
DECLARATION I, Clayton Mashapa, do hereby declare that this thesis is the result of my own work, except to the extent indicated in the acknowledgements, references and by comments included in the body of the report, and that it has not been submitted in part or in full for any other degree to any other university.
iii
ACKNOWLEDGMENTS Praise and worship to the Almighty God for His abundant grace and mercies throughout the study duration. I am grateful to my supervisors, Prof. Dr. Edson Gandiwa and Prof. Dr. Never Muboko whose guidance significantly helped to shape this DPhil thesis. Many thanks to the vibrant team members of the School of Wildlife, Ecology and Conservation at Chinhoyi University of Technology for their technical support during our academic interactions. Prof. Dr. Victor K. Muposhi, Dr. Tongayi Mwedzi, Mr Beaven Utete and Dr. Olga Kupika all my mentors, I cannot forget your academic insights and advice all along my research study, stay blessed. Acknowledgements are expressed to S. Mashapa; K. E. Nyamwanza; M. Karinda; M. Mudhluli; T. Muhlahlubi; T. Kaunda and T. Chinho for either support or fieldwork assistance. I am indebted to the Zimbabwe Parks and Wildlife Management Authority, the Forestry Commission of Zimbabwe, the management of Save Valley Conservancy, the Ministry of Agriculture, Lands and Rural Resettlement and the traditional leadership of Mutema-Musikavanhu communities for granting me permission to conduct this study. Special thanks to Prof. Dr. Fanuel Tagwira of Africa University who inspired me to study ecology when he taught me at undergraduate studies. My thanks also go to Prof. Dr. Shakkie Kativu of the University of Zimbabwe who groomed me as a student of ecology. Finally, I am grateful to my family and to my wife Patience Mhuriro-Mashapa, for supporting me in prayers, and by way of encouragement despite my long hours without their attention.
iv
ABSTRACT The response of woody vegetation dynamics to human and herbivory disturbances across land use categories (communal lands of Mutema-Musiakavanhu, buffer zone and Save Valley Conservancy (SVC)) were assessed in Save Valley, southeastern lowveld of Zimbabwe. Disturbance regimes, such as excessive herbivory and anthropogenic activities in the study area likely influence state-and-transition dynamics in woodland ecosystems. A stratified random design was used with the study strata divided based on three defined land use categories, where data was collected from November 2016 to December 2017. The study employed mixed methods approach including household questionnaire survey (n = 400), key informant interviews (n = 20) and focus group discussions (n = 80) to collect data on anthropogenic activities based on woodland resource utilization and the impact of humanwildlife conflicts. To measure woody vegetation attributes, herbivory and human disturbances, a total of 45 plots measuring 50 m x 20 m were assessed with 15 plots randomly placed in each of the three defined study strata. Specifically, changes occurring in land use and land cover were determined based on random land classification of LANDSAT images (1990 to 2015) and ancillary data. This study further used the Markov-cellular automata model to predict the land use and land cover changes across SVC for the period 2020 to 2040. Descriptive statistics and content analysis were used to analyze quantitative and qualitative data, respectively. In communal lands, a combination of agricultural land expansion, harvesting for firewood, timber and livestock grazing have modified the status of woody vegetation. The study recorded that woodlands contributed a range of 0.04% to 12.82% to the global annual income (GAI) of about US$1600 per household based on user rights discrimination around the woodland resources. However, the majority of farmers (86%, n = 258) had incurred annual economic loss of about US$800 per household due to humanwildlife conflicts as wild animals raid crops and prey on livestock. Furthermore, the study results recorded significant differences on the status of woody vegetation across the three defined land use categories. The human and herbivory disturbances decreased in intensity within the buffer zone with maximum species diversity of woody plants, thus confirming to the intermediate disturbance hypothesis. Model simulations predicted that by the 2040s, woodland and grassland cover in SVC will decrease by 46.73% and 10.54%, respectively, with at least over 6, 000% expansion of agricultural land use and bare land cover as compared to the 1990 land use and land cover categories. It was concluded that, human and herbivory induced disturbances were the main drivers of consistent woody vegetation dynamics across the study area. The study also revealed that both formal and informal institutions in the study area do not adequately implement policy pronouncements related to sustainable woodland management as they focus more on agricultural land use management and socio-economic use of woodland resources for human livelihood. It was recommendedthat Zimbabwe should consider aligning its woodland management policies and legislation with the Zimbabwe’s National Biodiversity Strategy and Action Plan (NBSAP, 2013-2020) so as to ensure national commitment with funding for government extension service delivery in the forestry sector, effectively utilize traditional leaders and best practices to protect the environment, wildlife and equitable use of woodland resources to benefit the present and future generations. The study further recommended the need to revisit the Zimbabwe land reform policy associated with agricultural resettlement and law enforcement agents to be effective in sustainable conservation of woodlands and protected wildlife areas in Zimbabwe.
v
List of Acronyms AHP
Analytic Hierarchy Process
ANOVA
Analysis of Variance
CAMPFIRE
Communal Area Management Programme For Indigenous Resources
CBNRM
Community Based Natural Resource Management
CF
Cash flow
CITES
Convention on International Trade in Endangered Species
CV
Contingent valuation
EMA
Environmental Management Agency
FAO
Food and Agriculture Organization of the United Nations
FGD
Focus Group Discussion
FTLRP
Fast Track Land Reform Programme
GAI
Global Annual Income
GCP
Ground Control Points
GIS
Geographical Information System
GO
Gross output
GoZ
Government of Zimbabwe
GPS
Geographical Positioning System
HCA
Hierarchical cluster analysis
IDH
Intermediate Disturbance Hypothesis
IEA
International Energy Agency
IUCN
International Union for Conservation of Nature
KWH
Kruskal-Wallis H- test
MA
Millienium Assessment on Ecosystems vi
MC
Monetary costs
MOLA
Multi-objectives land allocation
NBSAP
National Biodiversity Strategy Action Plan
NDVI
Normalized Difference Vegetation Index
NGOs
Non-Governmental Organisation
OC
Operating costs
PCA
Principal Component Analysis
PQRS
Preview, Question, Read and Summarize
PRSP
Poverty Reduction Strategy Papers
RDC
Rural District Council
SADC
Southern African Development Community
SCD
Size class distribution
SDG
Sustainable Development Goal
SV
Sales value
SVC
Save Valley Conservancy
SWOT
Strengths, weakenesses, opportunities and threats
UN
United Nations
UNCBD
United Nations Convention for Conservation of Biological Diversity
UNCCD
United Nations Convention to Combat Desertification
UNEP
United Nations Environmental Programme
UNESCO
United Nations Educational, Scientific and Cultural Organisation
UNFCCC
United Nations Framework Convention on Climate Change
UNFI
United Nations Forest Instrument
UNSPF
United Nations Strategic Plan for Forests
VC
Villager Class vii
VIDCO
Village Development Committees
WADCO
Ward Development Committees
ZimAsset
Zimbabwe Agenda for Sustainable Social Economic Transformation
ZimStats
Zimbabwe National Statistical Agency
ZimVAC
Zimbabwe Vulnerability Assessment Committee
ZPWMA
Zimbabwe Parks and Wildlife Management Authority
ZRP
Zimbabwe Republic Police
viii
DEDICATION
This thesis is dedicated to:
My mother, Savie Murau-Mashapa, a subsistence farmer who toiled to fund my elementary schooling in my village.
The welfare of wild animals and the woody vegetation across the wilderness of Save Valley, southeastern lowveld of Zimbabwe.
ix
Table of contents TITLE PAGE i APPROVAL FORM ii DECLARATION iii ACKNOWLEDGMENTS iv ABSTRACT v List of Acronyms vi DEDICATION ix Table of Contents x List of Tables xiv List of Figures xv CHAPTER 1: GENERAL INTRODUCTION 1 1.1 Background to the study 1 1.2 Statement of the problem 5 1.3 Study objectives 8 1.3.1 Overall main objective 8 1.3.2 Specific objectives 9 1.4 Research questions 9 1.5 Theoretical framework 10 1.5.1 Social ecology and savanna woodland management for sustainable human livelihoods 10 1.5.2 Control systems on human activities in woodland management at the human-wildlife interface 15 1.5.3 Disturbance theories and woody vegetation dynamics in savanna ecosystem 17 1.5.4 Conceptual framework of the study 21 1.6 Justification of the study 26 1.6.1 Scientific significance of the study 26 1.6.2 Societal significance of the study 28 1.7 Thesis outline 30 CHAPTER 2: OVERVIEW OF STUDY AREA AND METHODS 31 2.1 Introduction 31 2. 1.1 Description of the study area 31 2.1.2 Geology and geomorphology 33 2.1.3 Climate 33 2.1.4 Soil type and vegetation 34 2.1.5 Wildlife 35 2.1.6 Demography and livelihood activities 36 2.2 Strategy of inquiry 37 2.3 General methods 40 2.4 Research design 41 2.5 Sampling procedure and data collection 43 2.6 Data analysis 49 2.7 Conclusion 52 CHAPTER 3: A REVIEWOF LAND USE AND SAVANNA WOODLAND MANAGEMENT IN ZIMBABWE: A CASE STUDY OF THE SAVE VALLEY, SOUTHEASTERN LOWVELD ZIMBABWE 53 3.1 Abstract 54 3.2 Introduction 55 3.3 Materials and Methods 58 x
3.3.1 Study area 58 3.3.2 Literature review process and assessment framework 59 3.3.3 Data analysis 60 3.4 Findings and Discussion 61 3.4.1 The role of local people in woodland utilization and management in the Save Valley, southeastern lowveld of Zimbabwe 61 3.4.2 Local level institutional control on woodlands management in Save Valley, southeastern lowveld of Zimbabwe 67 3.4.3 The national government and its institutional controls on woodlands management in Save Valley, southeastern lowveld of Zimbabwe 70 3.5 Conclusion 75 CHAPTER 4 : IMPACT OF ANTHROPOGENIC FACTORS AND HERBIVORY DISTURBANCES ON WOODY VEGETATION ACROSS A PROTECTION GRADIENT IN SAVE VALLEY, SOUTHEASTERN LOWVELD OF ZIMBABWE 77 4.1 Abstract 78 4.2 Introduction 79 4.3 Materials and methods 83 4.3.1 Study area 83 4.3.2 Study design and data collection 84 4.3.3 Data analysis 87 4.4 Results 88 4.4.1 Comparison of woody vegetation status in different land use category defined by a protection gradient 88 4.4.2 Association of sample plots as influenced by disturbance factors on woody vegetation in Save Valley, southeastern lowveld of Zimbabwe 91 4.5 Discussion 95 4.6 Conclusion 102 CHAPTER 5: SOCIO-ECONOMIC AND ECOLOGICAL OUTCOMES OF WOODLAND RESOURCE UTILIZATION BY PEOPLE IN MUTEMAMUSIKAVANHU COMMUNAL AREAS IN SAVE VALLEY, SOUTHEASTERN LOWVELD, ZIMBABWE 104 5.1 Abstract 105 5.2 Introduction 106 5.3 Materials and Methods 108 5.3.1 Study area 108 5.3.2 Data collection 108 5.3.2.1 Sampling procedure and sample size 108 5.3.2.2 Institutional arrangements promoted for woodland management 110 5.3.2.3 Evaluation of the socio economic of woodland utilization 110 5.3.2.4 Ecological outcomes of community based woodland management 111 5.3.3 Data analysis 111 5.3.3.1 Institutional arrangements promoted for woodland management 111 5.3.3.2 Evaluation of the socio economics of woodland utilization 112 5.3.3.3 Ecological outcomes of community based woodland management 114 5.4 Results 114 5.4.1 Institutional arrangements promoted for woodland management 114 5.4.2 Socio-economic benefits of livelihood based on woodland ecosystem services 117 5.4.3 Ecological outcomes of community based woodland management 118 5.4.3.1 Land use and land cover change 118 xi
5.4.3.2 Woody vegetation status across Mutema-Musikavanhu communal area 121 5.5 Discussion 123 5.6 Conclusion 130 CHAPTER 6: SOCIO-ECONOMIC IMPACT OF HUMAN-WILDLIFE CONFLICTS ON AGRICULTURE BASED HUMAN LIVELIHOOD IN THE PERIPHERY OF SAVE VALLEY CONSERVANCY, SOUTHERN ZIMBABWE 132 6.1 Abstract 133 6.2 Introduction 134 6.3 Materials and Methods 136 6.3.1 Study Area 136 6.3.2 Population sampling frame and sample size 137 6.3.3 Data collection 138 6.3.4 Data Analysis 139 6.4 Results 140 6.4.1 Socio-economic impact of human-wildlife conflicts on agro-based livelihoods in Mutema-Musikavanhu communities adjacent to Save Valley Conservancy, southern Zimbabwe 140 6.4.2 Drivers of human-wildlife conflicts and mitigation measures implemented by local communities to reduce human-wildlife conflicts across the study area 145 6.5 Discussion 148 6.6 Conclusion 154 6.6.1 Implication for management 155 CHAPTER 7: LAND USE AND LAND COVER CHANGES IN A HUMANWILDLIFE MEDIATED LANDSCAPE OF SAVE VALLEY CONSERVANCY, SOUTHEASTERN LOWVELD OF ZIMBABWE 156 7.1 Abstract 157 7.2 Introduction 158 7.3 Materials and Methods 160 7.3.1 Study area 160 7.3.2 Sampling design and data collection 163 7.3.3 Assessment on land use by communal settlers in SVC 164 7.3.4 Markov-cellular automata model to simulate land use and land cover change over time 165 7.3.5 Data Analysis 165 7.4 Results 168 7.4.1 Land use and land cover changes 168 7.4.2 Model validation 172 7.4.3 Predicted future land use and land cover changes for SVC 173 7.5 Discussion 176 7.6 Conclusion 179 CHAPTER 8: GENERAL DISCUSSION, CONCLUSION AND RECOMMENDATIONS 180 8.1 Introduction 180 8.2 Summary findings and discussion 181 8.3 Scientific contributions of the study 198 8.4 Societal contribution of the study 204 8.5 Management and Policy Implications 211 8.6 Study contribution related to research methods 213 8.7 Limitations of the study 216 xii
8.8 Conclusion 8.9 Recommendations 8.9.1 For research 8.9.2 For policies REFERENCES APPENDICES
217 220 222 223 224 274
xiii
LIST OF TABLES Table 3.1. Summary description of woodland ecosystems in Save Valley, southeastern lowveld of Zimbabwe, in terms of importance to woodlands and biodiversity conservation, status and underlying causes of degradation. 62 Table 4.1. Measured variables and methods used. 86 Table 4.2. Summary of statistical analyses from Kruskal-Wallis H (ANOVA) test results of study variables for 45 sample plots (median and ranges in brackets) across Save Valley, southeastern lowveld of Zimbabwe. 90 Table 5.1: Distribution and characteristics of study respondents across MutemaMusikavanhu communities in Save Valley, southeastern lowveld of Zimbabwe. 109 Table 5.2. SWOT analysis of community based woodland management system in MutemaMusikavanhu communities 116 Table 5.3a. Land use and land cover change in Mutema communal area 119 Table 5.3b. Land use and land cover change in Musikavanhu communal area 119 Table 5.4. Comparison of measured woody vegetation variables in Mutema-Musikavanhu communal areas. Data are expressed as the median and range. 122 Table 6.1. Estimated wildlife populations in Save Valley Conservancy, southern Zimbabwe 137 Table 6.2. Severity of human-wildlife conflict as driven by problem wild animals in Mutema-Musikavanhu communities in the periphery of SVC, southern Zimbabwe 142 Table 6.3. Nature of human-wildlife conflicts in Musikavanhu and Mutema communal areas, southern Zimbabwe 143 Table 6.4. Relationship between household economic and financial (US$) loss and distance from the boundary of Save Valley Conservancy to Mutema and Musikavanhu communal areas, southern Zimbabwe 144 Table 6.5. Major drivers of human-wildlife conflicts in Mutema and Musikavanhu communal areas, southern Zimbabwe 146 Table 7.1. Descriptions of land use and land cover categories for the study area. 162 Table 7.2. Biophysical and socioecological data weights derived from the Analytic Hierarchy Process (AHP) procedure. 166 Table 7.3: Area of land use andland cover and % changes over time in SVC, 1990-2015 170 Table 7.4a. Land use and land cover change of steady state and transition area matrix (in ha) for SVC for the period 1990-2000. 171 Table 7.4b. Land use and land cover change of steady state and transition area matrix (in ha) for SVC for the period 2000-2015. 171 Table 7.5. Area of land use and land cover and % changes over time in SVC, including the simulated future area of land use and land cover and % changes. 175
xiv
LIST OF FIGURES Figure 1.1. Conceptual model showing potential impacts of disturbances on woody vegetation and varying land use management on human communities and protected wildlife areas. 25 Figure 2.1. The study area of Save Valley, southeastern lowveld of Zimbabwe, showing Save Valley Conservancy and the surrounding communal areas. 32 Figure 4.1a. Principal component analysis biplot of woody vegetation variables from sample plots in Save Valley, southeastern lowveld of Zimbabwe. 92 Figure 4.1b. Principal component analysis scatter plot of 45 sample plots in Save Valley, southeastern lowveld of Zimbabwe. 93 Figure 4.2. Hierarchical cluster analysis dendrogram showing classification of sample plots into two clusters based on woody species diversity measures from 45 sample plots in Save Valley, southeastern lowveld of Zimbabwe. 95 Figure 5.1a. Land use and land cover map of Mutema communal area in Save Valley, southeastern lowveld of Zimbabwe for the period 1990 to 2015. 120 Figure 5.1b. Land use and land cover map of Musikavanhu communal area in Save Valley, southeastern lowveld of Zimbabwe for the period 1990 to 2015. 121 Figure 5.2. The size class distribution of woody species within Mutema-Musikavanhu communal areas in Save Valley, southeastern lowveld of Zimbabwe. 123 Figure 6.1. Human-wildlife conflicts prevalence across Mutema and Musikavanhu communal areas adjacent to Save Valley Conservancy, southern Zimbabwe 141 Figure 7.1. Save Valley Conservancy (SVC) and resettlement area within the protected area. 161 Figure 7.2. Land use and land cover maps of Save Valley Conservancy, southeastern lowveld of Zimbabwe, 1990-2015. 169 Figure 7.3. Simulated versus actual land use and land cover categories in 2015 for SVC, southeastern Zimbabwe 173 Figure 7.4. Simulated future land use and land cover maps of Save Valley Conservancy, southeastern Zimbabwe, 2020, 2030 and 2040. 174 Figure 8.1. Modified general conceptual framework of state and transition change in wood vegetation, highlighting prominent land use categories (large boxes), generalized woody vegetation/woodland states (small boxes), considerations for managing state change and disturbance regime shifts within a land use category (black thin arrows), and considerations for managing land-use change (bold arrows) 200
xv
CHAPTER 1
GENERAL INTRODUCTION
1.1 Background to the study Human livelihood entails the capabilities of people to access and claim resources to engage in activities required for a standard of living. A human livelihood is sustainable when it can cope with and recover from disturbances or enhance its resources to provide opportunities for sustainable livelihood approaches for both the present and future generations to realise net benefits at the local and global levels in both the immediate and long term (Chambers and Conway, 1992; Broegaard et al., 2017; Baker et al., 2018). The notion of sustainable human livelihood approach rose out of the 1992 Earth Summit held in Rio de Janeiro (Perrings, 1994; UNESCO, 2015) and its promotion of Agenda 21 (Agenda for the 21st Century) which aimed that every individual person or community must have the opportunity to earn a sustainable livelihood. The sustainable livelihood framework seeks to bring together concepts in ecology along with human, social and economic aspects of ecosystems (Conway, 1985; FAO, 2016; Turkelboom et al., 2018). These system based human livelihood approaches have practical applications and a typical practice is the evolution of new approaches to local ecological knowledge generation with woodland utilization by people for their livelihoods and conservation (Persson et al., 2016; Quintas-Soriano et al., 2018).
Woodlands may not necessarily be owned by people for it to be an important asset to their livelihood but as long as people have access and claims to woodland resources then it will help in contributing to the way people live and manage this natural resource capital base for human livelihoods (Mashapa et al., 2014a; Duguma et al., 2018). Woody vegetation is a key 1
component of the natural resource capital which is linked to sustainable livelihood framework and wildlife management in African savanna (Perrings, 1994; FAO, 2016; Duguma et al., 2018).
Woody vegetation is comprised of plants that produces wood as its structural tissue and these are usually either trees, shrubs, or lianas (Thackway and Lesslie, 2006; Balehegn and Hintsa, 2018). African savannas are biomes that comprise ecosystems with a continuous herbaceous layer and a discontinuous woody vegetation stratum (Scholes and Archer, 1997; Sankaran et al., 2008). Woody vegetation is important for carbon sequestration as the primary producers of food which directly or indirectly provide energy across trophic levels and human livelihood is no exception (Meigs and Keeton, 2018). Woody vegetation (e.g., woodlands) constitutes a carbon sink that captures carbon dioxide from the atmosphere and stores it in biomass through the process of photosynthesis (Zeng et al., 2013; Galafassi et al., 2018). Accordingly, woody vegetation structure, abundance and composition plays important roles in the functioning of ecosystems and goods provision to human livelihoods and wild animals (Balehegn and Hintsa, 2018; Siero, 2018). The school of thought on non-equilibrium approach asserts that external disturbances like climate variability and edaphic factors are the primary determinants of large-scale vegetation patterns and the status of woody vegetation (Sankaran et al., 2005; Meigs and Keeton, 2018). Conversely, the equilibrium approach confirm that, at landscape scale woody vegetation dynamics may primarily relate to internal disturbances like human activities and excessive foraging by large herbivores (Laycock, 1991; Grime, 1997; Gowda et al., 2018). For example, the loss of woody vegetation due to elephant (Loxodonta africana) herbivory and human activities at landscape scale is a cause for concern and it has been an area of continuous research in southern Africa (e.g., Tafangenyasha, 1998; O’Connor et al., 2007; O'Kane et al., 2012). 2
In the study area of Save Valley, southeastern lowveld of Zimbabwe, climate and soil type have been uniform over a long period of time (Vincent and Thomas, 1960; Nyamapfene, 1991; Rukuni et al., 2006; Gandiwa et al., 2016). Thus, human and herbivory induced disturbances and the resultant land use changes in this human and wildlife mediated landscape is perceived to be the most influencing element of woody vegetation dynamics (Muboko et al., 2013; Kahuni et al., 2014; Zisadza-Gandiwa et al., 2014). Diebacks of woody plants and conversion of woodlands to scrublands caused by excessive elephant herbivory and human damage have been reported in many woodlands of sub-Sahara Africa and parts of Asia which affects woodland ecosystem function and services (Galiano et al., 2010). Land use change is also an important threat to the status of woody vegetation and a key factor of the global environmental change (Lambin, 1997; Balehegn and Hintsa, 2018). It is characterised by the expansion of human activities (e.g. agricultural crops, logging and livestock grazing) on the land (Mhuriro-Mashapa et al., 2018). Several woodlands in protected areas of Zimbabwe have lost parts of their natural woody vegetation because of human encroachment for communal agricultural settlement; this is the case of the Save Valley Conservancy (SVC) in southeastern lowveld of Zimbabwe (Pole, 2006; Williams et al., 2011; Gandiwa et al., 2011; Zisadza-Gandiwa et al., 2014; Fakarayi et al., 2015).
The human-wildlife interface in Mutema-Musikavanhu communal land use and the protected wildlife area of SVC was reported to experience human-wildlife conflicts as wild animals raids crops and prey on livestock (Mhuriro-Mashapa et al., 2017; Mhuriro-Mashapa et al., 2018).Humans were reported encroached for agricultural resttlement onto the protected wildlife area of SVC (Pole, 2006; Williams et al., 2011) and this posed the present study to explore how local people interact with the nearby wildlife management land use and its woodland resources. The social exchange theory posits that humans form relationships 3
keeping in mind the costs and benefits involved in them (Kelley and Thibaut, 1978; Pellegrini, 2018). Humans are rational beings, and their relationships with the natural resource capital base are based on well calculated decisions for self-benefits which are normally designed to outweigh the cost (Rusbult, 1983; Pellegrini, 2018). The social exchange theory attributes a self-motive to most actions as humans interact with the natural resource base and this can be linked to Hardin’s (1968) the tragedy of the common resource properties like woodlands in Save Valley, southeastern lowveld of Zimbabwe. The study assessed to ascertain the status of shared woody vegetation under different land use management categories of varying protection gradient and disturbance induced by humans and large herbivores.
Disturbance regimes, such as excessive herbivory by large herbivores and anthropogenic factors, would likely modify and influence state-and-transition dynamics in woodland ecosystems, thus affecting the composition, structure, abundance and regeneration potential of woody vegetation (Campbell et al., 1995; Venter et al., 2018). This, in turn, has profound effects on woodland communities and habitat status, indirectly by influencing woody growth patterns as affected by the intensity of disturbance regimes (Scholes and Walker, 1993). Woody vegetation response to these disturbances which is measured as quantitative land use and land cover changes over the years, is likely to be heterogeneously distributed on the land, and therefore, needs to be spatially monitored. Human and herbivory disturbances may influence the diversity of woody plants as promulgated by the intermediate disturbance theory which postulates that species diversity tends to be at the peak in landscape with moderate disturbances in comparison to landscapes with severe or low disturbances which result in lower levels of species diversity (Connell, 1978; Zhang et al., 2014; Muposhi et al., 4
2016). Whereas, the chaos theory of plant ecology assumes unpredictable pattern in vegetation dynamics in relation to disturbance regimes, thus, unreconcile with the intermediate disturbance hypothesis (May, 1973; Gleick, 1987; Meigs and Keeton, 2018). The relative spatio-temporal dynamic factors that drive succession in woodland ecosystems is complex and not well understood (Zhou, 2014; Archer, 2017). Successful management of habitats of natural vegetation depends to a large measure on knowledge of vegetation status, the extent to which the vegetation is utilized, and changes which take place in response to disturbance regimes (Magadza et al., 1993; Mashapa et al., 2019).
1.2 Statement of the problem A Zimbabwe human population census by ZimStats (2013) showed incremental changes in human density (i.e.,
26 people km-2) with a growth rate of 1.1 %
in Save Valley,
southeastern lowveld of Zimbabwe for the period 2002 to 2012. On the other hand, a trend analysis of results for animal population surveys from 2001 to 2014 in SVC revealed that the estimated numbers of large herbivores were expotentially increasing with elephant 1585 (± 84.9 %); cattle 15098 (± 22.2 %); and sheep and goats 5472 (± 37.8 %), along with an estimated 9617 (± 14.2 %) huts and 1754 (± 38.0 %) houses (Dunham and van der Westhuizen, 2015). These has impacts not only of human-wildlife interface, but also on land use and vegetation landscape changes as reported by other studies elsewhere (e.g., Campbell et al., 1995; Zisadza-Gandiwa et al., 2014; Mashapa et al., 2014a; Mhuriro-Mashapa et al., 2018). In the Save Valley case, a series of events have raised questions on the prevailing state of landuse and human-wildlife interface. Such events include the encroachment of communal settlement for agricultural purposes onto a wildlife protected area of SVC and land use changes related to agricultural practices in the adjacent communal areas of Mutema5
Musikavanhu communities at the periphery of SVC due to increased human population. The combination of human settlement in the SVC and presence of large herbivores is percieved to be threatening the steady-state and transition of woody vegetation and its provision of ecosystem goods to both humans and wildlife.
Furthermore, a knowledge gap exist on whether or not the current land use and land cover change provide the short-term survival rather than the sustainable management of the natural resource capital of woodlands, wildlife and its genetic diversity. There is also a dearth of knowledge on how the prevailing land tenure, local governance controls on access and how the utilization of woody vegetation is impacting on the status of woody vegetation and wildlife management in Save Valley, southeastern lowveld of Zimbabwe. It is the underlying problem of knowledge gaps on the prevailing land use, woodland status, human-wildlife interface and how best to support people’s sustainable livelihoods that influenced the present study in order to appreciate and understand the conservation status of woody vegetation, wildlife management and sustainable utilization of resources in Save Valley in Save Valley, southeastern lowveld of Zimbabwe. Based on Homans (1961) theory of social exchange, the study explored what drove communal settlers to encroach and resettle onto the southern part of a protected wildlife area of SVC given a documented human-wildlife conflicts in the adjacent communal lands (Mhuriro-Mashapa et al., 2017; Mhuriro-Mashapa et al., 2018). The study further explored the socio-economic impact of human-wildlife conflicts across the communities of Mutema-Musikavanhu communal lands on the periphery of SVC in Save Valley, southeastern lowveld of Zimbabwe (Mhuriro-Mashapa et al., 2017). It also sought to understand the status of woody vegetation under a varying protection (disturbance) gradient and contribute knowledge to woodland resource management or good governance across varying land use management categories of communal lands of Mutema-Musikavanhu 6
communities, buffer zone and a protected wildlife area of SVC in Save Valley, southeastern lowveld of Zimbabwe.
The study also focused on the sustainable livelihood theory (Conway, 1985; Meyer and Schulz, 2017) to unravel the societal needs for food security and livelihood options as well as wildlife management and how they are affected either directly or indirectly by large herbivores and human induced disturbances on woodland ecosystem. Sustainable livelihood framework has been regarded as the drive of human development which should also meet the requirements of the present generation without compromising the needs of future generations (Singh and Gilman, 1999; Pellegrini, 2018). The sustainable human livelihoods framework puts humans at the centre of development as supported through utilizing and enhancing the assets or capitals upon which livelihoods are developed, namely natural resource capital, social capital, human capital, financial capital and physical capital (Conway, 1985; Meyer and Schulz, 2017). For instance, humans, rather than the woodland resources and the wild animals they utilize or the governing structures that serve them are the priority concern. Given such overlaps and it may not be easy to see how sustainable livelihoods approach can be a framework for achieving human development at community level in the case of the study area, i.e., Save Valley, southeastern lowveld of Zimbabwe. The notion that it is humans and not woodland resources and/or wildlife whose future options need to be protected in the human sphere can be misleading as it may imply that the woody vegetation and wildlife in the case of the present study area may be considered of secondary importance; that humans can be allowed to systematically disturb the woody vegetation and wildlife if it means that they can enhance their livelihood (Rusbalt, 1983; Garrido et al., 2017).
7
The concept of human development and indeed a sustainable livelihood approach certainly does not seek to facilitate livelihood at the expense of the environment and its natural resource base of say, woody vegetation and wildlife (Brannstrom, 2014). As Hardin (1968) explained the tragedy of the commons, the pursuit of self-interest in an open-access commons of say woodland resources in Save Valley, southeastern lowveld of Zimbabwe can lead to woodland degradation and fragmentation. What Hardin (1968) termed the commons in his concept of the tragedy of the commons, can be best described as an open access commons, as there are some viable common resource properties that are managed in public and do not suffer the tragedy simply because they are subject to proper community based natural resource management. While sustainable livelihood approach is human centred, it does not advocate for the compromise on the natural resource capital base which is one of the five cornerstone of sustainable human livelihood framework (Carney, 1998; Ratner et al., 2017). It is the underlying problem of knowledge gaps on how best to support people’s sustainable livelihoods that influenced the present study outputs to contribute knowledge on the basis for sustainable utilization and conservation of woody vegetation and wildlife management in Save Valley, southeastern lowveld of Zimbabwe and elsewhere within the related African savanna ecosystem (see., Duguma et al., 2018).
1.3 Study objectives 1.3.1 Overall main objective: to assess anthropogenic influences on woody vegetation and ecosystem services in the context of human livelihoods and sustainable conservation, and contribute towards the sustainable conservation of woodland communities and associated ecosystem goods in semi-arid savanna ecosystem dominated by varying human and large herbivore activities using a case study of the Save Valley, southeastern lowveld of Zimbabwe.
8
1.3.2 Specific objectives 1. To review land use and savanna woodland management in Save Valley, southeastern lowveld of Zimbabwe. 2. To assess the impact of anthropogenic factors and herbivory disturbances on woody vegetation across a protection gradient in Save Valley, southeastern lowveld of Zimbabwe. 3. To determine the socio-economic and ecological outcomes of woodland resource utilization by people in Mutema-Musikavanhu communities in Save Valley, southeastern lowveld of Zimbabwe. 4. To analyse the socio-economic impact of human-wildlife conflicts on agriculture based human livelihood across Mutema-Musikavanhu communities in the periphery of SVC, southeastern lowveld of Zimbabwe 5. To establish historical and future land use and land cover dynamics across SVC in southeastern lowveld of Zimbabwe.
1.4 Research questions 1. What are the review outcomes of savanna woodland management in relation to anthropogenic activities and herbivory in Save Valley, southeastern lowveld of Zimbabwe? 2. How do anthropogenic activities and herbivory by large herbivores affect woody vegetation across a protection gradient in Save Valley, southeastern lowveld of Zimbabwe? 3. What are the socio-economic and ecological outcomes of woodland resource utilization by people in Mutema-Musikavanhu communities on the periphery of SVC in Save Valley, southeastern lowveld of Zimbabwe? 9
4.
What are the socio-economic outcomes of human-wildlife conflicts on agriculture based human livelihood across Mutema-Musikavanhu communities in the periphery of SVC, southeastern lowveld of Zimbabwe?
5.
What are the historical and future land use and land cover dynamics across SVC in southeastern lowveld of Zimbabwe?
1.5 Theoretical framework 1.5.1 Social ecology and savanna woodland management for sustainable human livelihoods This section presents key theories this study builds on and thereby forming the basis for the study on how earlier studies have advanced knowledge and gaps on human livelihoods and sustainable utilisation and conservation of woody vegetation and wildlife across global societies. The overall theoretical framework is based on human social ecology which is an approach to society that embraces an ecological, restoration, and communitarian view on society as people utilize natural woodland resource capital for their livelihood (Morgera and Wingard, 2009; Hellin, et al., 2018). This ideology looks to enhance and transform current outcomes on both socio-economic issues and ecological factors while promoting good governance across woodland communities under a protection gradient defined by varying land use types (Brannstrom, 2014; Piabuo et al., 2018). It looks to do away withscarcity, inequity and hierarchy in the economy in favour of a world in which human societies live in harmony with natural resources and promote its ecological integrity (Mathevet et al., 2015; Orta-Martínez et al., 2018).
10
The social ecology approach to society puts forth the idea that ecological problems are inevitably the results of social dysfunctions in human society (Little, 2007; Shrestha et al., 2018). The approach posits that problems in ecology will only be completely resolved when the underlying social issues are addressed and resolved (Leslie et al., 2015; Ratner et al., 2017). These social issues involve things such as increasing human population, agricultural expansion, excessive woody harvesting, increasing population of wild animals and herbivory, and consequently a resident class structure that designates certain portions of humanity as inferior, and a distorted view of what constitutes progress as dominated by the minority ruling class (Fischer et al., 2014; Shrestha et al., 2018). This is common in southern African communities where the traditional leaders’ family clans normally dominate as the custodians of natural resources like land and its related woodland resources (Nhira and Fortmann, 1993; Mashapa et al., 2014a). Social ecologists stress the importance of establishing more egalitarian socioecological systems in rural communities that is driven by equality and cooperation on utilization of natural resources rather than individual profits (Leslie et al., 2015; Piabuo et al., 2018). Collective action and equal concern for all aspects of life and nature are fundamental to this form of ecology.
Woody vegetation and wildlife resources in African savanna, in particular in Save Valley, southeastern lowveld of Zimbabwe are under threat from multiple stressors that include biodiversity degradation, human population growth, woodlands clearance for agricultural expansion, soil erosion, illegal harvesting of woody resources, poaching, human-wildlife conflicts, human encroachment onto protected wildlife area and retarded eco-tourism development (Williams et al., 2011; Mhuriro-Mashapa et al., 2017; Mhuriro-Mashapa et al., 2018). Woody vegetation in the Save Valley, southeastern lowveld of Zimbabwe is under 11
threat from the interaction of herbivory by large herbivores and anthropogenic disturbances such as deforestation, human population growth and human encroachment onto protected wildlife areas to expand their agricultural activities and an increase in these disturbance regimes is expected (Williams et al., 2011; Mhuriro-Mashapa et al., 2018). Excessive herbivory and anthropogenic disturbances can also exacerbate the effects of other stresses on woodlands ecosystems such as habitat fragmentation, loss and conversion, over-exploitation, invasive alien species, and pollution (Gessner et al., 2013; Meigs and Keeton, 2018).
Woodland fragmentation in a habitat can alter migratory routes of species, affect species range, population size (Case, 2006; Forrest et al., 2012; Sfair et al., 2018) and the distribution and phenology of species within terrestrial ecosystems (Lovejoy, 2008; Balehegn and Hintsa, 2018). Despite this knowledge, the extent to which current woodland conservation policies in Zimbabwe and associated international instruments that address woodland conservation and mitigation issues in the natural forest sector is mainly limited within the communal woodland management sector (Shumba, 2001). Anthropogenic disturbance of woody vegetation is also likely to affect the knowledge, innovations and practices of local communities and associated woodland ecosystem-based livelihoods (Secretariat of the Convention of Biological Diversity, 2009; Gowda et al., 2018). However, the role of integrating local traditional leadership and local people’s structures in sustaining human livelihoods is rarely considered or integrated into the design of scientific and national strategies for woodlands management in Zimbabwe and the present study area is no exception (Nhira and Frontmann, 1993; Shumba, 2001; Mashapa et al., 2014a).
12
The complex interaction between human and herbivory impacts on woodland ecosystem productive activities, societal factors such as political, economic, social, demographic and environmental processes can reshape sustainable livelihoods (Gessner et al., 2013; Leslie et al., 2015; Archer et al., 2017). Thus, there is need for an integrative approach to ecological studies that bridges the gap between the biophysical and social domains in African savanna where the majority of residents are poor rural residents who rely heavily on woodlands ecosystems as their natural resource capital base for sustainable livelihoods (Muboko et al., 2014; Mashapa et al., 2014a). Chambres and Cronway (1992) and Scoones (1998) defines livelihoods as sustainable when they can cope with, and recover from stresses and shocks, maintain or enhance its capabilities and assets, while not undermining the natural resource base in a long run. In rural southern Africa savanna, scholars have demonstrated that human and herbivory disturbance is impacting negatively on woody vegetation, which human populations rely on as the basis of their livelihoods (Mpofu et al., 2012; Muboko et al., 2014; Scoones, 2015).
The sustainable livelihoods framework has been successfully used to understand and promote rural development (Chambers, 2011; Baker et al., 2018). According to Acharya (2006) rural households can acquire livelihoods in four ways namely production-based livelihood, labourbased livelihood, exchange or market based livelihood and transfer based entitlements. The sustainable livelihoods framework is therefore applied in this study to understand communal and wildlife land use categories on how they influence woody vegetation dynamics and the emerging livelihoods strategies in Save Valley, southestern lowveld of Zimbabwe. Recent literature demonstrates that rural communities in Save Valley, southeastern lowveld of Zimbabwe are diversifying livelihoods from consumptive woodland resource utilization and 13
agriculture based livelihoods strategies in order to cope with perturbations resulting from the declining economy of Zimbabwe (Mashapa et al., 2014a; Mhuriro-Mashapa et al., 2017; Mhuriro-Mashapa et al., 2018). In that regard, the sustainable livelihood approach was used to analyse peoples’ current livelihood in Save Valley, southern lowveld of Zimbabwe and what is needed for restoration ecology in avoiding the inappropriate utilatisation of woody vegetation in communal and wildlife land use categories and critiqued based on selected ecological concepts and theories.
There are five capitals of sustainable livelihoods that are classified as human, social, physical, natural resource and financial which constitute as collective drivers of a standard way of living (Singh and Gilman, 1999). The present study explored the natural resource capital of woody vegetation across the study area in terms of its vulnerability to shocks and the local institutional context within which this capital exist and how it shape the livelihoods of local people in the study area. Once this is understood then recommended interventions were highlighted to enhance livelihoods and their sustainability, perhaps by reducing vulnerability of woody vegetation as it is utilized by humans and wild animals in Sava Valley, southeastern lowveld of Zimbabwe.
Thus, the study was also about understanding the current woody vegetation dynamic, its conservation strategies and develop suggestions for improvement based upon that understanding. The sustainable livelihood approach is meant to avoid a situation where intervention is unguided giving little positive impact or is at worst a detrimental means of livelihood (Chambers and Conway, 1992; Baker et al., 2018). Instead it might involve making the current means of woodlands resource utilisation and agriculture based livelihood 14
less susceptible to environmental, social or economic stresses in Save Valley, southeastern lowveld of Zimbabwe. The framework of sustainable livelihood approach helped aid an appreciation of the natural resource capital of woody vegetation which are available to the study area and the vulnerability of woodlands and the involvement of local institutions which resulted in recommendations that local people themselves and the wildlife management authority of SVC may be able to put into practice rather than be dependent upon the actions of outsiders.
1.5.2 Control systems on human activities in woodland management at the humanwildlife interface Human interactions with nature are based on a mix of socio-ecological norms and expectations
(Holling
and
Meffe,
2002).
Social
exchange
theory
posits
that
human relationships are formed by the use of a subjective cost-benefit analysis and the comparison of alternatives (Rusbalt, 1983). However, control is a deeply entrenched aspect of contemporary human societies, i.e., human behavior is controlled through laws, incentives, threats, contracts, and agreements (Holling and Meffe, 2002). Save Valley, southeastern lowveld of Zimbabwe was reported to experience human-wildlife conflicts as wild animals from SVC, mainly large herbivores were recorded to raid crops and large carnivores prey on livestock in Mutema-Musikavanhu communities (Mhuriro-Mashapa et al., 2017; MhuriroMashapa et al., 2018). On the other hand, communal settlers encroached and resettled onto the protected wildlife area of SVC for agricultural settlement (Pole, 2006; Williams et al., 2011).
15
The study applied the social exchange theory in exploring to understand and incorporate into how local people experiencing human-wildlife conflicts came to decide to encroach for resettlement within this nearby protected wildlife area of SVC in Save Valley, southeastern lowveld of Zimbabwe. The social exchange theory has excellent applications for real-world situations, such as to understand business and interpersonal relationships of human and nature (Rusbalt, 1983; Mashapa et al., 2019). The theory of social exchange views human interactions and exchanges as a kind of results and consequences driven social behavior (Leslie et al., 2015). The fundamental concept of the theory of social exchange is cost and rewards and means that cost and reward comparisons drive human decisions and behavior (Rusbalt, 1983; Mashapa et al., 2019). Related to the cost and benefit concept of the social exchange theory is Hardins (1968) concept of the “tragedy of the commons” which posits that common resource properties like communal woodlands in Mutema-Musikavanhu communities are a natural resource capital shared by many individuals as they utilize woody resources for their livelihood. In the context of the Mutema-Musikavanhu communal lands study area stratum that would mean that each individual would not have a claim to any part of the woodland resources, but rather, to the use of a portion of it for own benefit. The tragedy would be that, in the absence of clear and proper regulations on woodland resource utilization, each individual would have a tendency to exploit the common woodland resources to own advantage without limit and control systems and under this state of affairs, the woody vegetation would be eventually degraded or lost.
Socio-ecology and control systems in the case of Save Valley, southeastern lowveld of Zimbabwe can be defined as the formal and informal rules, regulations, by-laws, social values and belief systems that influence human behavior including land clearance practices, 16
hunting, grazing and using fire. In the protected wildlife areas like SVC, national policy instruments are designed and expected to change human behavior in order to minimize its impacts on natural resources and stocking rates of wild animals (Rowcliffe et al., 2004). This suggests that laws and regulations enforcement can be crucial in combating excessive and illegal exploitation of woody vegetation in Save Valley, southeastern lowveld of Zimbabwe where human and herbivory induced disturbances were perceived to influence vegetation dynamics (Pole, 2006; Williams et al., 2011). Compulsory human behavior arising from coercion (top-down enforcement) that is derived from laws and regulations or constraints caused by restrictive provisions could be applied to sustainably manage woodlands in Save Valley, southeastern lowveld of Zimbabwe (Rowcliffe et al., 2004; Fischer et al., 2011).
1.5.3 Disturbance theories and woody vegetation dynamics in savanna ecosystem State-and-transition models hold great potential to aid in understanding communal woodlands and rangeland ecosystems’ response to land use management-induced disturbances by providing a framework for organizing current understanding of potential ecosystem dynamics. Varying land use influence vegetation cover (Bouriaud et al., 2005; Nacoulma et al., 2011; Meigs and Keeton, 2018), especially in regions exposed to herbivory and human population growth like Save Valley, southeastern lowveld of Zimbabwe (ZimStats, 2013; Dunham and Westhauizen, 2015). Land use change within the region of Save Valley, southeastern lowveld of Zimbabwe is related to agriculture through cropland expansion (Kahuni et al., 2014; Mhuriro-Mashapa et al., 2017; Mhuriro-Mashapa et al., 2018) which was perceived to threaten natural woody vegetation. Woody vegetation dynamics are likely to respond to these human and herbivory induced disturbances (Gessner et al., 2013; Siero, 2018). However, there was still a knowledge gap in understanding woody vegetation 17
response to human and large herbivores induced disturbances, particularly in Save Valley, southeastern lowveld of Zimbabwe where vegetation dynamics is poorly understood and needs further investigation given the onset of the Zimbabwe Land Reform Programme in 2000 which triggered agricultural expansion associated with woodland clearances as well as human encroachment onto protected wildlife area of SVC (Rukuni et al., 2006; Mashapa et al., 2014a; Mhuriro-Mashapa et al., 2018).
Varying land use categories of Mutema-Musikavanhu communal lands, the buffer zone and the protected wildlife area of SVC presented an opportunity to assess the influence of equilibrium forces at landscape scale to determine the status of woody vegetation. The present study highlighted that despite considerable progress in the theory of disturbances, for different authors the notion of disturbance has different meanings (Evans and Barkham, 1992; Meigs and Keeton, 2018). The differences follow from an alternative understanding of the term, assumption of the holistic or reductionistic concept of the study or different approaches to equilibrium and non-equilibrium in nature (Westoby, 1980; Manning et al., 2018). Different theoretical and methodological assumptions taken when analysing vegetation changes under the effect of disturbances allow relative arbitrariness in the interpretation of results (Westoby et al., 1989; Sfair et al., 2018). However, not all biotic interactions and ecological processes taking place in plant communities can be identified as disturbances (Veblen and Lorenz, 1986; Meigs and Keeton, 2018).
The effects of disturbances should be described on the basis of an objective assessment of what is normal for the functioning and stabilisation of a community as well as what disturbs and disrupts this equilibrium (Westoby et al., 1989; Sfair et al., 2018). Moreover, assuming 18
the equivalent treatment of ecological systems, the interaction of the disturbing factors should be related to a specific level of plant life organisation (an individual, population or community), which would permit unambiguous discernment of the disturbance effects and finally generalise its effects on different levels of the hierarchical structure of vegetation (Veblen and Lorenz, 1986; Rusch, 1988; Evans and Barkham, 1992; Meigs and Keeton, 2018).
Much interest has been devoted to the analysis of disturbances on many levels of the hierarchical structure of vegetation and its evolutionary changes (Veblen and Lorenz, 1986; Sfair et al., 2018). They have been analysed in the context of the effect of natural and anthropogenic factors (Rusch, 1988; Evans and Barkham, 1992; Gowda et al., 2018) or of variable space-time relationships (Foster and Reiners, 1986; Coffin and Lauenroth, 1989; Sfair et al., 2018). The concept of disturbances also includes other natural processes taking place in communities, mainly renewal of woody plant-stands in gaps formed as a result of woody plant falls (Veblen, 1992; Sfair et al., 2018). Gaps appearing in woody plant-stands are analysed in different aspects related to woody vegetation dynamics: gap models, gapphase regeneration (Busing, 1991; Shugart and Smith 1992; Urban and Shugart, 1992; Mashapa et al., 2019) or in view of the effect they exert on the final shape of the renewing woody plant-stands depending on their origin, structure or dynamics (Canham, 1988; Lorimer et al., 1988; Worrall and Harrington, 1988; Mashapa et al., 2019).
Following Wiens (1984), DeAngelis and Waterhouse (1987), Ellis and Swift (1988) and Sfair et al. (2018) it is suggested that in non-equilibrium savanna ecosystem biomes, plant-humanherbivore interactions are loosely coupled and herbivore populations are controlled by 19
density-independent factors, whereas, plant biomass is abiotically controlled by climate and soil type where competition among plant species is not an important force in structuring vegetation communities. In contrast, equilibrium savanna ecosystems at landscape level like the case of the present study area of Save Valley, southeastern lowveld of Zimbabwe were characterized by tightly coupled plant-herbivore-human interactions with density dependent controls on herbivore populations, where herbivores and human density control plant biomass with competition as possible important structuring force in plant communities. The study adopted the intermediate disturbance hypothesis theory to explain the woody vegetation dynamics and its ecological outcomes across a protection gradient of varying land use categories (communal, buffer zone and protected wildlife area) in Save Valley, southeastern lowveld of Zimbabwe.
In the study area, several studies have focused on socio-economic outcomes of agricultural activities and its contribution to sustainable livelihoods of people in Save Valley (Mashapa et al., 2014b; Mashapa et al., 2014c; Mhuriro-Mashapa et al., 2017) and in agricultural based systems in Zimbabwe (Jiri et al., 2016; Mapfumo, 2016; Murungweni et al., 2014). However, no study have assessed the influence of human and herbivory on woody vegetation dynamics and how woodlands resources contribute towards sustainable livelihoods of people in Save Valley, southeastern lowveld of Zimbabwe. This present study, therefore, evaluated regulations and local institutional structures for woodland utilisation in the context of socioeconomic implications related to both communal woodland management for rural livelihoods and wildlife management in Save Valley, southeastern lowveld of Zimbabwe in southern African savanna ecosystem.
20
The study combined the sustainable livelihoods framework (Conway, 1985; Baker et al., 2018), social exchange theory (Kelley and Thaibaut, 1978; Archer et al., 2017), tragedy of the commons (Hardin, 1968; Shrestha et al., 2018) and the intermediate disturbance hypothesis (Connell, 1978; Sfair et al., 2018) theories within the social ecological systems framework (Ostrom, 2007; Gowda et al., 2018) to explore human livelihood (Chambers and Conway, 1992; Piabuo et al., 2018) and sustainable conservation of woody vegetation and wildlife within the legal and institutional frameworks for land use management of communal lands, buffer zone and protected wildlife area in Save Valley, southeastern lowveld of Zimbabwe, which constitute the northern part of the globally renowned Greater Limpopo Transfrontier Conservation Area of southern Africa. The study sought to develop a socioecological system model for assessment and simulation of land use and land cover changes and woody vegetation dynamics within a southern Africa savanna ecosystem.
1.5.4 Conceptual framework of the study Figure 1.1 represents a set of conceptual framework that is expected to explain the impact of human and large herbivores induced disturbances on woody vegetation in relation to human livelihood and sustainable utilisation and conservation of woodlands under varying land use management categories. The conceptual framework shows the linkages (bold dotted arrows) and the consequences/opportunities of the threats (bold normal) on woody vegetation and livelihoods as they relate to disturbance regimes of herbivory and anthropogenic activities. The conceptual framework shows how the independent variables (rise in population of people and large herbivores in land use categories of Mutema-Muskivanhu communal lands and the protected wildlife area of SVC, respectively, as well as social ecology and power dynamics of local instituitions on woodland management across the defined land use categories) and relate these to the influence of the dependent variable (status of woody 21
vegetation/woodlands dynamics) which shape the land use and land cover of Save Valley, southeastern lowveld of Zimbabwe. Climate and soil type are the intervening variables that might affect the relationship of the dependent and independent variables but it is difficult to measure or to see the nature of their influence given that the study area was at landscape level of uniform habitat in the case of these intervening variables of non-equilibrium factors (see, Yeh, 2000).
Complex bottom up and top down mechanisms could affect the status of biophysical resources consequently the abundance, composition and structure of woody vegetation within the study area (Gowda et al., 2018). Bottom up mechanisms refers to the influence of climate and soil type on woody vegetation dynamics (Westoby, 1980; Gandiwa et al., 2013). However, the climate and soil type of the study area at the defined landscape level has been considered uniform over time and considered negligible to influence the status of woody vegetation (Vincent and Thomas, 1960; Nyamapfene, 1991; Gandiwa et al., 2016; MhuriroMashapa et al., 2017). Top-down control mechanisms are based on the premise that management and/or predators can regulate and control the numbers of herbivore populations thereby preventing them from overexploiting vegetation (Terborgh, 2001; Boulanger, 2018). Top-down control mechanisms are driven by consumers such as predators and herbivores and are believed to control producer woody biomass through regulation of primary production (Meserve et al., 2003; Boulanger, 2018).
There are also moderating variables and these behaves like the independent variables in that they have a significant contributory or contingent effect on the relationship between the dependent and the independent variable (in this case the social ecology and livelihoods component e.g., the local people/communal settlers, key informants, legislations/rules and 22
regulations, local instistution structures, community traditional leaders and management of SVC/Zimbabwe Parks and Wildlife Management Authority (ZPWMA) as they influence land use management and woodlands utilisation across specific and respective defined land use categories). In this study, there was a focus on the role of local social ecological knowledge in bottom-up autonomous adaptation of local people and their local institutions in promoting sustainable livelihoods as people engage in community based natural resource management concept of utilising woodland resources for productive activities and agriculture for their socio-economic capital generation. The study also explored on how local people behave and respond to the reported human-wildlife conflicts (Mhuriro-Mashapa et al., 2017) in MutemaMusikavanhu communities adjacent to SVC. Bottom-up adaptation refers to the use of informal local values and perceptions from local communities (Rodima-Taylor, 2012; Gandiwa et al., 2013).
The study was based on the proposition that there exists a disturbance continuum/disturbance gradient across the defined three land use categories with the communal lands being at the high extreme end of frequent disturbances on woody vegetation, SVC as a protected area with lower levels of disturbance, whereas, the buffer zone of low protection with moderate disturbance on woody vegetation. The propositions were that; (i) the communal lands study stratum was proposed to have high frequency of anthropogenic disturbances and predicted to have a low woody species diversity dominated by pioneer species with low woody plant abundance and small basal area, (ii) the buffer zone stratum was proposed to have intermediate disturbances and predicted to have high woody species diversity with high woody plant abundance and large basal area, whereas, (iii) SVC was proposed as a study stratum with infrequent disturbances and predicted to have moderate abundance, structure and diversity of woody plants. The study modelled and predicted the land use and land cover 23
changes across the study area. It was assumed that there was no significant difference in the past, current and future land use and land cover over for the period 1990 to 2040. The research propositions were based on the intermediate disturbance hypothesis that stipulate that maximum diversity is at intermediate level of disturbance and infrequent disturbance can trigger low diversity due to competitive exclusion by dominant species, whereas, frequent disturbance cause low diversity with community dominated by pioneer species (Meigs and Keeton, 2018).
Armesto et al. (1986) hypothesized that frequent, large-scale disturbances should create random spatial patterns, whereas individuals of woody species should appear clumped in woodlands where canopy gaps are the major cause of structural change. Hence, the frequently disturbed woodlands should exhibit more random dispersions among individuals of woody species, and the infrequent disturbed habitat should exhibit more clumped dispersions. If the disturbance has been sufficiently large, it may be expected to see succession where a distinctly different woody species composition may be emerging in the intermediate disturbed buffer zone. Species experiencing successful regeneration in the disturbed woodlands should have size distributions more skewed towards saplings and smaller trees than the same species within the undisturbed woodlands (Manning et al., 2018). Unsuccessfully regenerating species should either be unusually rare or lack these smaller size classes of trees. The conceptual framework of this study link issues of woodland utilisation, management and conservation across a protection gradient of varying land use categories under an African savanna vegetation.
24
Figure 1.1: Conceptual model showing potential impacts of disturbances on woody vegetation and varying land use management on human communities and protected wildlife areas. Adapted from Bellard et al. (2012).
25
1.6 Justification of the study 1.6.1 Scientific significance of the study Relationship between species diversity and disturbance remain contentious in community ecology (Veblen, 1992 ; Meigs and Keeton, 2018) as highlighted by conflicting theories, namely; the intermediate disturbance hypothesis (Connel, 1978 ; Sfair et al., 2018), chaos theory of plant ecology (Strogatz, 2000) and the invisibility theory of plant ecology (Davies et al., 2000). High plant diversity has been linked to high primary productivity in natural communities (Tilman, 1999; Meigs and Keeton, 2018). The intermediate disturbance theory posits that species diversity is high at intermediate disturbance, while chaos theory do not show consistent effect of disturbance on diversity (Huston, 2000). Thus, diversity-disturbance relationship remains largely unresolved and the present study presented appropriate findings where a disturbance gradient across woodland ecosystem on varying land use management conformed to the intermediate disturbance hypothesis with maximum diversity of woody species within the buffer zone of moderate disturbance as compared to the communal lands.
The study answered research questions and worked within multi-disciplines of; ecology, land use (agriculture and wildlife interface), spatial patterning (land use planning/geography) and biology (structure, abundance and diversity of woody plants). The study also employed research methods cutting across disciplines like remote sensing, geographical information system and ground based measurement of vegetation attributes, herbivory and human activities which gave scope in vegetation assessment and monitiring methods based on: disturbance analysis; biodiversity; ecosystem services and livelihood impacts. The significance of woody vegetation response to disturbance perturbations was linked to its dynamics persistence. Woody vegetation persistence across the study strata indicated a vegetation tendency to persist in increasing or decreasing (Simoniello et al., 2008), and it 26
proved related to the concept of resilience, that is, the ability of woody vegetation to recover from disturbances (Simoniello et al., 2008; Gowda et al., 2018). In that regard, the present thesis findings contributed to fill the gaps on scientific information in the research methods to monitor and assess savanna woodland dynamics of Zimbabwe and elsewhere with related tropical vegetation. The present thesis demonstrated the feasibility of assessing the savanna woodland ecosystem by using the Markov-cellular automata model which proved to be a robust approach in spatial and temporal dynamic modelling of land use and land cover changes because geographic information system (GIS) and remote sensing data which was efficiently incorporated with biophysical and socioecological data (Li and Reynolds, 1997; Silverton et al., 1992).
This study focused on current approach of remote sensing data such as the LANDSAT images to better assess vegetation dynamics regarding land use and land cover change and vegetation trends (Aduah and Aabeyir, 2012). Moreover, this research provides information on deforestation, determines the vegetation change drivers and predicted future land use and land cover change by the 2040s for SVC which is important for policy makers to develop vegetation monitoring strategies to combat woodland degradation and to ensure better environmental management. This research contributes to the United Nations programmes (e.g., Reducing Emissions from Deforestation and Forest Degradation, and United Nations Convention to Combat Desertification) to help maintaining or restoring the environmental resilience in order to secure the availability of productive land to satisfy present and future generations.
Disturbance regimes on woody vegetation exacerbates changes in woody species composition, abundance and structure through; extirpations, extinctions, changes in species’ 27
ranges, mismatches in their phenology (timing of pollination, flowering, etc.), and population declines (Lambin et al., 2003; Bowd et al., 2018). The sustainability and productivity of woody vegetation in Save Valley, southeastern lowveld of Zimbabwe is under threat from disturbances and rising resource demands associated with increasing population of human and large herbivores (Dunham et al., 2002; Williams et al., 2011; Dunham and van der Westhuizen, 2015). The complex geospatial interaction of wildlife, human and shifting environmental conditions have had significant impact on woody vegetation, its productivity and its provision of ecosystem services (Gandiwa et al., 2013). Vegetation dynamics are likely to respond to human and herbivory disturbances across the study area (Dunham and van der Westhauizen, 2015). However, there is still knowledge gap in understanding vegetation response in SVC and the periphery communal areas of Mutema-Musikavanhu communities in Save Valley under savanna ecosystem where vegetation dynamics is poorly understood and needs further investigation after the onset of Zimbabwe’s fast track land reform programme in 2000. This programme triggered expansion of agricultural activities (Rukuni et al., 2006) as well as human encroachment onto protected wildlife area of Save Valley Conservancy (Williams et al., 2011). This necessitates the gathering of empirical and theoretical evidence to understand the status of the socio-ecological systems of the woodland ecosystem across Save Valley, southeastern lowveld Zimbabwe.
1.6.2 Societal significance of the study Study sought to contribute knowledge on socio-ecological challenges related to woodland utilisation in communal settlement within Save Valley, southeastern lowveld of Zimbabwe, specifically on: livestock grazing and human access to woodland resources. Woodland resources are the foundation of many developing countries’ economies and a source of livelihoods for millions of people (Duguma et al., 2018). Assessing woody vegetation 28
dynamics provided an objective means of assessing equall and equitable woodland resource utilization levels thereby facilitating decisions on improving community based natural resource management (CBNRM) approaches (Piabuo et al., 2018). These facilitate designing of appropriate management plans (e.g., optimal stocking and removal rates of livestock and large herbivores) that promotes persistence and coexistence of woodland resources and human. Although woodlands contribute significantly towards the diversification of livelihoods of communities in Save Valley, southeastern lowveld of Zimbabwe, inadequate and unequal power dynamics of community involvement in the management and governance of communal woodland resources, has been identified as an area of research in regards to the escalation of woodland ecosystem destruction. Failure to recognise for the different classes of people as woodland resource users could lead to patterns of woodland degradation and losses with many detrimental environmental consequences in the study area and there is need for devolution of power and authority on community based woodland management (OrtaMartínez et al., 2018).
The study further explored the socio-ecology and livelihood implications of communal woodland utilisation by people. The socio-economic and ecological outcomes of woodland management in situations where land tenure are conferred to communal people have been less investigated in Zimbabwe (Campbell et al., 1996; Mashapa et al., 2013). In this study, the linkages between communal woodland property rights and institutions at work on one hand in Zimbabwe is assessed and woodland utilisation and its contributions to the livelihoods of local communities and communal woodland conservation on the other hand in Save Valley, southeastern lowveld of Zimbabwe was explored. Assessing woodland dynamics provide an objective means of assessing woody vegetation utilization levels thereby facilitating decisions regarding optimal livestock-wildlife stocking and removal rates 29
(Mutanga and Rugege, 2006). These facilitate designing of appropriate management plans that promote persistence and coexistence of woody plants and human-wildlife and the ecological integrity of the woodland ecosystem. The study contributed to knowledge on managing socio-ecological challenges like human-wildlife conflicts related to human-wildlife interface by recommending compensatory mechanisms for victims of human-wildlife conflicts as well as documenting indegenous mitigation measures to reduces human-wildlife conflicts.
1.7 Thesis outline This DPhil thesis has eight Chapters. Chapter 1 focuses on background to the study, statement of the problem, main and specific research objectives, research questions, the theoretical and conceptual framework of the study, justification of the study. Chapter 2 presents the description of the study area and methods employed in this research are addressed. Chapter 3 devoted to the first study objective presents a review and analysis of the status of savanna woodland in relation to anthropogenic activities and herbivory in Save Valley, southeastern lowveld of Zimbabwe. Chapters 4, 5, 6 and 7 are devoted to the second, third, fourth and fifth objectives of the study, respectively. They present in detail the introductions, methods used, the results achieved, their discussion and conclusions, respectively. Chapter 8 presents the general discussion, conclusion and recommendations highlighting the overall contribution of the study as well as recommendations for application of research findings and for further research.
30
CHAPTER 2
OVERVIEW OF STUDY AREA AND METHODS 2.1 Introduction This Chapter is a description of the study area, followed by an outline of the general methodology and inquiry including the sampling design, data collection procedures and data analysis. Additional and detailed materials and methods for respective objectives are discussed in the respective Chapters.
2.1.1 Description of the study area The research site is a protected wildlife area SVC and the bordering communities of MutemaMusikavanhu communal lands in Save Valley, southeastern lowveld of Zimbabwe (Figure 2.1). The study area lies within the globally renowned Greater Limpopo Transfrontier Conservation Areas of southern Africa (Muboko, 2017). SVC (area of 348, 450.53ha, location at latitude 19°50′S-21°00′S and longitude 31°30 ′E-32°30′E) is a cooperative wildlife management area located 385km south of Harare and it is about 120 km from Mutare City, Zimbabwe. SVC is located along Save River stretching from Birchenough Bridge in Chipinge District to Chiredzi District in Save Valley, southeastern lowveld of Zimbabwe. SVC was partially resettled by communal settler farmers since 2000 as part of the Zimbabwe Land Reform programme of agricultural resettlement (Pole, 2000). The perennial Save River which tends to attract wild animals to trespass towards human communities in MutemaMusikavanhu communal lands separates the protected wildlife area of SVC and the local communities (Mhuriro-Mashapa et al., 2018).
31
Figure 2.1. The study area of Save Valley, southeastern lowveld of Zimbabwe, showing Save Valley Conservancy and the surrounding communal areas. NB: Ward 3 is Mutema communal area and Ward 16 is Musikavanhu communal area.
32
2.1.2 Geology and geomorphology The geology of the study area can be classified into one formation. This is the Precambrian D which is essentially a compound of undifferentiated granitic and gneissic rocks (Nyamapfene, 1991). The geology of Save Valley, southeastern lowveld of Zimbabwe consists of the granophyre complex, the basaltic intrusions and Cretaceous sedimentary series of the cave sandstone type (Purves and Fullstone, 1975). The sandstone gives rise to deep, highly permeable soils. The topography is gently undulating, with gneisse, paragneisse and granite outcrops rising up to 250 m above ground and an elevation of 480-620 m above sea level (Pole et al., 2004).
2.1.3 Climate Monthly total rainfall data for the period 1990 to 2015, from two weather stations within the study areas, namely, SVC (two rain stations on northern and southern part of the conservancy and in less than 20 km proximity to Mutema-Musikavanhu communal area) were collected (Gandiwa et al., 2016). Thus, between 1990 and 2015, the study area recorded one wet year (2000, which experienced a cyclone eline) and three drought years (1992, 1994 and 2005). The recorded number of wet and drought years in the study area was similar between the the two rain stations on each far end of the study area. Overall, there was a strong synchrony in the occurrence of wet and drought periods across the two stations, with only the year 2000 being the wettest period whereas 1992, 1994 and 2005 were the common drought recorded across the two rain stations.
Mean annual rainfall for the period from 1990 to 2015 did not differ significantly across the two rain stations far apart within the study area. Thus, the rainfall pattern of the study was regarded uniform.Mean annual precipitation for Save Valley, southeastern lowveld of 33
Zimbabwe (1990-2015) is +-450 mm, and has varied between 92 mm in 1992 and 1114 mm in 2000 (Mhuriro-Mashapa et al., 2017). The wet season is between November and March and a dry season between April and October. Mean monthly maximum temperature ranged from 25.9 oC in July to 36 oC in January, whilst mean monthly minimum temperature ranges from 9 oC in June to 24 oC in January (1990-2015 SVC temperature records). The high summer temperatures (with peaks in the January-February period) and clear skies induce high evapo-transpiration rates. Effective rainfall is thus generally lower than the recorded values (Pole, 2006).
2.1.4 Soil type and vegetation Just like the common soils of savanna, soils on the study area are mainly developed from acidic metamorphic rocks, hence they have low kaolinitic clay, but have coarse texture in the topsoil and have low organic matter content (Nyamapfene, 1991). In common, the soils have little supply of nitrogen, phosphorus and sometimes micronutrients and they are very susceptible to erosion and compaction (Rukuni et al., 2006). Based on the geomorphology data collected from ZPWMA at a scale of 1:500,000, the soils of the study area are made up of siallitic (70%), regosol (15%), sodic groups (10%) lithosol (3%) and vertisol (2%). According to Nyamapfene, (1991), the largest soil substrate on the study area is the malvernia cretaceous sandstones which covers most of southeastern lowveld of Zimbabwe with siallitic soil group which dominate the study strata. This is an unusual formation for Zimbabwe (Nyamapfene, 1991). In depressions, the soils tend to be sodic because of relatively high amounts of exchangeable sodium (Purves and Fullstone, 1975). Alluvium/colluvium sediments are found in slight depressions (inland valleys) along the river such as Save River and its tributaries.
34
The study area is characterized by a savannah vegetation of the Zambezian phytogeographic zone which is semi-arid deciduous African savanna with the the dominating taxonomical families of woody plants being Leguminosae-Caesalpinioidae-Combretaceae woodlands being the common woody vegetation types (Pole, 2006). Accordingly, the most occurring woody tree species are Colophospermum mopane, Acacia nigrescens, Acacia weiwitchii, Combretum apiculatum, Kirkia acuminata, Commiphora mollis and Adansonia digitata with understorey woody shrubs including the Markhamia acuminata Cassia abbreviata, Sclerocarya caffra, Pterocarpus angolensis and Commiphora pyracanthoides, Gardenia resiniflua and Monodora jonodi (Shumba, 2001; Pole, 2006).
2.1.5 Wildlife Dunham and van der Westhauizen (2015) reported aerial survey results of population estimates of a wide variety of wild large herbivore species in SVC and these include 3112 (± 103 %) buffalo (Syncerus caffer), 1585 (± 84.9 %) elephant (Loxodonta africana), 314 (± 69.0 %) kudu (Tragelaphus strepsiceros) 661 (± 64.1 %) giraffe (Giraffa camelopardalis), 7050 (± 41.8%) impala (Aepyceros melampus), 2895 (± 39.7 %) zebra (Equus quagga), 3326 (± 46.1 %) wildebeest (Connochaetes taurinus), 690 (± 50.0%) eland (Taurotragus oryx), This survey revealed that, since 2001 to 2015, the estimated numbers of kudu, eland, impala and warthog (Phacochoerus africanus) in SVC had declined, while the numbers of elephants, buffalo and giraffe significantly increased. There is an over-abundance of elephants at a density of 2 elephants per km-2 (Dunham and van der Westhuizen, 2015). Human-wildlife conflicts have been reported across the study area with people engaging in poaching from SVC (Lindsey et al., 2009) whereas, large herbivores were reporting damaging crops while large carnivores prey on livestock (Mhuriro-Mashapa et al., 2017; Mhuriro-Mashapa et al., 2018). 35
2.1.6 Demography and livelihood activities The human population of the study area is dynamic like in the case of the whole country of Zimbabwe which has experienced high population growth rate in recent years due to high fertility rates (ZimStats, 2013). In Mutema-Musikavanhu commuties, the population size increased to 29 163 people (7 054 households) at a growth rate of 2.4 for the period 1992 to 2012 (ZimStat, 2013; Mhuriro-Mashapa et al., 2017). In rural settlements, such a population growth combined with subsistence farming increased the need of more agricultural land leading to excessive pressure on vegetation and the depletion of natural resource stock in general (Rukuni et al., 2006; Kahuni et al., 2014). The human population density varied from 26 to 81 people per km-2 (Williams et al., 2011; Zimstats, 2013). The population of the study area is essentially rural and made up of the Mutema-Musikavanhu communal area and southern part of SVC where communal settlers resettled on this protected wildlife area (Williams et al., 2011). Agriculture which included crop farming and livestock rearing is the principal source of livelihood for the rural people of the study area as well as for the entire country where 78% of the workforce is primarily devoted to agriculture (Rukuni et al., 2006). The main crops produced are maize (Zea mays), sorghum (Sorghum bicolor), tomatoes (Solanum lycopersicum) bananas (Musa sapientum), and vegetables (Mhuriro-Mashapa et al., 2018). The cultivation of maize and sorghum is considered important for the provision of staple food to the local people, whereas, livestock keeping is practiced often as a livelihood strategy in the face of crop failure and income losses (Mhuriro-Mashapa et al., 2018).
In 2000 and 2001, an area of approximately 960 km² on the southern part of SVC was resettled under the Zimbabwe Land Reform programme of agricultural resettlement reducing the effective wildlife area (Williams et al., 2011). Agricultural crop lands are normally located relatively near to homesteads, which included buildings constructed of wooden poles, 36
clay or brick, often with roofs of thatching grass or occasionally corrugated metal. The human population density in the resettlement area within SVC was reported 36 people per km-2 with an average rate of natural increase at 4.3 % against a Zimbabwe national average of 33 people per km-2 at an average rate of natural increase of 2.3 % over the period 2002 and 2012 (ZimStat, 2013). Aerial survey data by Dunham and van der Westhauizen (2015) reported that communal settlement in SVC significantly increased from the year 2001 to 2015 currently at 15098 (± 22.2 %) cattle (Bos taurus) and 5472 (± 37.8 %) and goats (Capra aegagrus). These domestic animals were in the southern half of SVC, along with an estimated 9617 (± 14.2 %) huts and 1754 (±38.0 %) houses.
2.2 Strategy of inquiry Firstly, Chinhoyi University of Technology Senate Research Committee granted clearance to carry out the research. Since SVC and part of the buffer zone fall under the jurisdiction of the ZPWMA and SVC management, permissions were sought and granted from the director of the ZPWMA and another permit from SVC management to carry out research in the protected wildlife area. Consequently, research permit was provided for 2017. The researcher also sought consent and authority from the local authority for the communal lands of Mutema-Musikavanhu communities which was granted by the office of the Manicaland Provincial Governor for the Ministry of Local Government, Public Works and National Housing. Permission was also sought and granted at local level with the Chipinge Administrators office and the traditional leadership of the chieftainship of MutemaMusikavanhu communities.
With regards the consent of households heads, the traditional leadership (from the chief to the village heads), political leadership, i.e., ward councilors were also consulted prior to data 37
collection. The research was therefore granted authority, access to and cooperation from the stakeholders in Save Valley including parks officials, Ministry of Agriculture and Rural Resettlement, other government departments, Non-Governmental Organisations (NGOs), traditional leadership and local experts. Once approval was granted by the Rural District Councils (RDCs) the research team got to the study area, reached up to targeted key traditional authorities for the study and accordingly, their individual consent was solicited and granted during the pilot study, which was carried out in the third week of June 2016. This allowed them to inform their respective local community leadership and prior to the household survey. During the data collection, prior informed consent was sought from the respondents through verbal means. Where possible, recording of the interviews was also done prior to obtaining verbal consent from the interviewees. Each household’s homestead was captured using a Garmin Geographical Position System (GPS) Map unit for further reference. However, specific names of respondents were not disclosed serve for the professional position of the individual in the community.
The study used an integrated or mixed methods approach of both qualitative and quantitative (Terrell, 2012; Quintas-Soriano et al., 2018). One of the merits of using the integrated methods approach is the ability to use various data sets to get a consolidated output of a phenomenon (Johnson et al., 2007; Angeler et al., 2018). However, with various sources of data, this entails that the researcher has to learn various data analysis skills (White et al., 2005; Creswell and Clark, 2011). Qualitative research methods used have their roots in social sciences such as social ecology and this approach seeks to understand people's subjective reality and social exchange which determine behaviour in terms of their knowledge, attitudes, perceptions, beliefs and fears (Creswell and Clark, 2010; Garrido et al., 2017). The qualitative methodology is rooted within the interpretive paradigm which emphasizes 38
analyzing and describing phenomena without necessarily relying on quantitative measurements and statistics (Creswell, 2012; Squires and Wiber, 2018). The interpretive paradigm (Leedy and Ormrod, 2005 in Muboko, 2011:104) views values as an integral part of social ecology such that it is based on subjective social ecological knowledge, which is context dependent. The disadvantages of qualitative research is that data collection methods and analysis are so cumbersome such that in most cases, only small numbers of subjects can be studied (White et al., 2005; Boylorn and Orbe, 2014). Other scholars have also criticized the method based on the fact that it involves difficulties in rigorous analysis of data, lack of reproducibility of the findings (i.e., results may not be applicable to other related settings), all these makes the qualitative methodology to be subject to researcher bias (Creswell, 2007; Hennink et al., 2010). However, proponents of qualitative research argue that there are strategies available to the qualitative researcher to protect against these potential biases and to enhance the rigour of the findings (Creswell, 2013). Such strategies include triangulation of qualitative data with quantitative data which was employed in the present study.
On the other hand, quantitative research methods involves counting and measuring parameters or variables to produce estimates of averages and differences between and within the groups or specific units of measurement (Creswell, 2013). However, quantitative research approaches partially contribute towards social problems and are unbalanced in the sense that results from the study informs us about inputs and outputs concerning variables or parameters but does not describe how such phenomenon or that unit of measure is locally constituted (Silverman, 2006). The quantitative research methodology is rooted within the positivist paradigm, which focuses on the need to discover laws that are generalizable and govern the universe. It places value on natural science based on objective knowledge derived from precise observations and measurements that are verifiable (Becker, 2008; Creswell, 2013). 39
The mixed and integrated method strategy allowed triangulation of results from different research methods (White et al., 2005) used in this study. Triangulation is whereby one performs concurrent but separate collection and analysis of quantitative and qualitative data to understand the research problem (Creswell and Clarke, 2010). The study also employed the convergent model which is a variant of the triangulation model whereby the results from the study were then converged by comparing and contrasting the different results during interpretation. The model allows for comparison, confirmation or corroboration of the qualitative results with the quantitative results (Clark and Creswell, 2008; Denzin and Lincoln, 2011). Between-method triangulation allowed us to confirm the findings generated through one particular method by another (Dey, 1993). For example, in this data from a household questionnaire about perceived land use and land cover changes across Save Valley, southeastern lowveld of Zimbabwe was triangulated with related data on land use and land cover changes derived from remote sensing for the same research site to confirm the validity of the former.
2.3 General methods The research method is divided into four components based on the five study objectives with objectives 3 and 4 compounded together. The first objective target is Chapter 3 where a case study reviewed and analysed land use and savanna woodland management across the study area. The second objective which made up Chapter 4 quantified the impact of human and large herbivores induced disturbances and its influence on woody vegetation dynamics based on varying land use of communal lands, buffer zone and a protected wildlife area in SVC in Save Valley, southeastern lowveld of Zimbabwe. This emphasized on the assessment of woody vegetation dynamics in response to human and large herbivores induced disturbances in the study area. The third component consists of objective 3 and 4 which are presented in 40
Chapters 5 and 6, respectively, and these are based on the assessessment of the socioeconomic and ecological outcomes of human utilization of communal woodland resources as well as the socioeconomic impact of human-wildlife conflicts on agriculture based livelihoods in Mutema-Musikavanhu communities in the periphery of SVC. The fourth part is objective 5 which make up Chapter 7 which was on classifying the study area into the main land use and land cover classes Finally, the study simulated and predicted land use and land cover change by the year 2020, 2030 and 2040 across SVC in Save Valley, southeastern lowveld of Zimbabwe.
2.4 Research design The research design refers to the overall strategy that one chooses to integrate the different components of a study such as the outline for the collection, measurement, and analysis of data in a coherent and logical way, in order to address the research problem (De Vaus, 2006). Data collection instruments or research instruments are the tools which are used to collect data to accomplish the goals of the study (Harrison et al., 2013). The main methods of collecting data in qualitative research include the following: household surveys, measurement of woody vegetation attributes and variables of herbivory and anthropogenic factors, direct observations, diary method, simulation and case study (White et al., 2005; Creswell, 2013).
This study in Chapter 3 employed a case study approach to review and analyze land use and savanna woodland management in Mutema-Musikavanhu communities and the protected wildlife area of SVC in Save Valley, southeastern lowveld of Zimbabwe. According to Creswell (2012), a case study can be defined as an empirical inquiry whose focus is on delimitated systems or contemporary phenomenon within its real-life context. The study strata of the case study is a communal agriculture-wildlife land use interface. Although some 41
scholars advance that case studies has no basis for scientific generalization, Yin (1994) highlighted that unlike scientific studies, the purpose of a case study is to highlight the theoretical propositions in research. One advantage of the case study design is that it accounts for background effects and information on behaviors that other forms of research may attempt to isolate (Flyvbjerg, 2006). In addition, results from a case study tend to be generalizable to the extent that readers can compare and contrast reported cases with their own (Yin, 1994). Flyvbjerg (2006) a case study is based on specific phenomenon hence it can generate context-dependent expert knowledge. In this study, the choice of the case study approach was because the researcher wanted to solicit novel local ecological knowledge information concerning savanna woodland management within a communal land use category and a protected wildlife land use management in Zimbabwe. According to Codjoe et al. (2014), local ecological knowledge is time-specific; is derived from people occupying a specific local community within a specific culture; and is a result of handing over to successive generations their accumulated experiences with their natural environment. Such information can be compared with case studies from other African biomes or it can be used to generate a model for adaptive management of other woodlands under related land use categories.
The status of woody vegetation under defined varying land use categories (communal, buffer zone and protected wildlife area) was assessed. Human disturbance on woody vegetation under communal land use was compared with large herbivores induced disturbance on woodland stands in varying land use category. The study area was stratified according to three study strata land use management systems with a varying protection gradient based on land use categories, namely; (i) SVC a strict conservation and protected wildlife area for consumptive use restricted to hunting of big game and was proposed as a study stratum of 42
high level of protection and lower level of disturbance on woody vegetation perceived to be driven by herbivory intensity from large herbivores, (ii) Mutema-Musikavanhu communal area of low level of protection and unrestricted access to woodland resources by local people was proposed as a study stratum of high level of disturbance on woody vegetation mainly driven by human activities and, (iii) the buffer zone which is also a less strict conservation area of low level of protection was proposed to the study stratum of moderate disturbance on woody vegetation due to both herbivory and human activities at the convergence zone of human-wildlife interface. For Chapter 4, a field survey was conducted in October to December 2017 to measure or record selected variables of woody vegetation and to record local population practices that influence woody vegetation dynamics that cause changes in land use and land cover changes. For Chapters 5 and 6, surveys through administration of household questionnaires were done in Mutema-Musikavanhu communities and across the communal settlers who resettled onto the southern part of the protected wildlife area of SVC. Whereas, for Chapter 7 remote sensing technique was employed to assess land use and land cover changes across the study stratum of SVC.
2.5 Sampling procedure and data collection For Chapter 3, secondary data were obtained from literature and other secondary sources. The data collection process covered the period November 2016-December 2017 and followed processes as outlined in the next sections. Secondary data involved the collection of published and unpublished information from relevant academic and non-academic sources. Secondary data aimed at documenting empirical research on woodland management issues in both communal woodlands and woodlands within protected wildlife area not only to supplement the study but also to identify gaps in the existing information. In Chapter 3, a review of existing literature was conducted (1990-2015) using peer-reviewed journal research 43
articles, books, edited book chapters, electronic academic theses, and technical reports from Google Scholar, Scopus, and Web of Science covering issues on woodland utilization by communal people as well as woodland dynamics for wildlife management in Save Valley, southeastern lowveld of Zimbabwe. In line with the case study approach, the review of empirical literature focused on case studies related to the impacts of woodland ecosystem productive activities and adaptation strategies in selected protected areas in close proximity to communal land use in Zimbabwe.
Empirical case studies were derived from the largest protected areas in Zimbabwe that is Gonarezhou National Park, Chiredzi River Conservancy, Hwange National Park and Mana Pools National Park. Thus in Chapter 3, literature review was done to obtain the bigger picture in terms of land use and savanna woodland management in Mutema-Musikavanhu communities and the protected wildlife area of SVC in Save Valley, southeastern lowveld of Zimbabwe. The international, regional, national and local biodiversity policy documents were downloaded and retrieved from official websites such as the United Nations Convention on Biological Diversity (CBD). Other policy documents included National Adaptation Programme of Action documents, National Communications and National Biodiversity Strategic Action Plans. Data were also obtained from unpublished internal scientific reports and management plans from the Zimbabwe Parks and Wildlife Management Authority (ZPWMA), technical reports and public publications from key international environmental organizations such as the United Nations Environmental Programme (UNEP).
In Chapter 3, overall, the goal of the study was to establish whether biodiversity related policies mainstream issues on sustainable woodland utilization and management and the role of local governance institutions and related rules and regulations. Part of this study aimed at 44
identifying knowledge gaps in terms of legal and institutional frameworks for woodland management in the context of communal lands and protected areas. Care International (2009) noted that decisions made by central governments, with particular focus on natural resource policies can have a profound effect on the ability of communities to adapt to woodland management. For instance, natural resource management policies in sectors such as wildlife, forests, water, agriculture, health, infrastructure, and economic development can facilitate or constrain adaptation and mitigation. Thus, integration of woodland considerations into these natural resource management policies can ensure that they contribute to building adaptive capacity from national to local levels. In some cases, existing policies or practices can provide opportunities or deter progress to address climate change. However, this depends on whether the available resources and political will in place to support implementation. Thus, in this study, key institutions in SVC and the periphery communal lands of MutemaMusikavanhu communities in Save Valley, southeastern lowveld of Zimbabwe and their role in addressing threats to woodland resources and human-wildlife interface under a varying protection gradient were identified.
For Chapter, 4 data collection was done based on the study strata which was categorized according to the defined contrasting land use (communal, buffer zone and protected wildlife area of SVC). Sampling study sites, as plots, were randomly selected using random number tables based on Save Valley, southeastern lowveld of Zimbabwe topographical map grid square intercept system, in relation to the stratified study sites. For each of the three study stratum, 15 replicate sample plots were pegged for data collection. Thus, a total of 45 plots were sampled for data collection of Chapter 4. A standard sample plot 50 m length by width of 20 m wide was used at each study site, in accordance with methods by Mapaure (2001) and Anderson and Walker (1974). 45
To describe the cumulative and interactive determinacy of harbivory intensity and human activities and its impact on the status of woody vegetation across the study area. The following were measured or recorded: woody species, plant height, basal stem circumference at 1.3m height, number of stems per plant, canopy volume, herbivory damage on woody plants, human damage on woody plants, plant status (alive or dead), stem density and dung counts. The measured variables follow the methods of Mpofu et al. (2012) and Gandiwa et al. (2012). Large herbivores and cow dung counts according to Gandiwa et al. (2011) were used to indicate a measure of large herbivores and cattle occupancy and utilization of a study stratum. Total counts of dung were recorded in each sample plot. The sum of dung counts represent an accumulation of large herbivore/cow occupancy and utilization of a study stratum. Dung counts have the advantage that they give data, not only on numbers, but also on distribution and differential habitat use (Gandiwa et al., 2011). More details are presented in Chapter 4.
A sampling produre of household survey was employed for Chapters 5 and 6. A base map was obtained from the Chipinge and Chiredzi Rural District Councils to identify the study sites. In the first stage, purposive sampling of Mutema-Musikavanhu communities was done based on their close proximity to SVC, same as well with the sampling of communal settlers who resettled onto the southern part of SVC. The assumption was that proximity to the protected wildlife area rendered the villages more vulnerable to impacts of woodlands disturbances due to mediating effects of human-wildlife conflict, whereas, the communal settlers who resettled onto the protected wildlife area of SVC were perceived to be influencing woody vegetation dynamics as they clear land for settlement and agricultural expansion. Due to the communal settlement model of Zimbabwe, households in communal lands form a cluster hence are located very close to each other. A distinct cluster forms a 46
village and is headed by a village head that is also accorded the status of being under the traditional leadership of a village head. A random systematic sampling technique was employed to select every third households within the village.
The sampling unit in this study stratum of communal land use was the household (MhuriroMashapa et al., 2018). A household here was taken as the unit of analysis because it is where all decisions are primarily taken (Mhuriro-Mashapa et al., 2017). The official statistics for the ward and village population were obtained from the ward councilors and village registers respectively. The sampling size for the households’ survey was determined by using probability proportional to sampling technique (Midzuno, 1951; Mhuriro-Mashapa et al., 2018). Consequently, a total of 400 sample households were selected and interviewed: 300 from Mutema-Musikavanhu communities, and 100 from the communal settlers who resettled onto the southern part of the protected wildlife area of SVC. The SVC was divided into two administrative units in terms of authorities who oversee the designate areas, i.e., the northern part managed from Chishakwe offices and the southern part managed from Sango/Humani offices and there is a buffer zone partially managed by the management of SVC and ZPWMA, the Ministry of Local Government, Public Works and Provincial Affairs and the Forestry Commission of Zimbabwe.
For Chapters 5 and 6, the study focused on local population survey which was conducted to determine the socioeconomic and ecological outcomes of communal woodland management across the study area in Mutema-Muskivanhu communities and communal settlers who resettled in southern part SVC. Local residents were considered for the survey that focused on 400 household heads selected randomly. The questionnaire survey which form part of a broad study on human effects on woody vegetation involved a sample of local people drawn 47
from the villages occurring in Mutema-Musikavanhu communities adjacent to the two protected areas as well as communal settlers who resettled on the southern part of SVC. Questionnaires are particularly suitable tools for approaching studies of local knowledge and perceptions of ecological processes (White et al., 2005). Current village registers kept by traditional leaders of the study villages formed the sampling pool and households were randomly selected. The household heads or another permanently resident adult (≥ 18 years) were targeted as the respondents and took part in the interviews in the study respondents’ residence. Respondents were interviewed using a semi-structured questionnaire. Questions were constructed to gather information on community based woodland management and the changing status of woody vegetation across the study area. More details are presented in Chapters 5 and 6, respectively.
For Chapter 7, the study focused on land use and land cover classification which entailed the use of LANDSAT TM (Thematic Mapper) images from 1990 to 2015, and environmental data (Elevation, slope and geomorphological units). The reference data for validating the different land use and land cover maps were collected from high-resolution images (Quickbird, Rapid Eye, Google Earth and aerial photo) and field data. All the classifications were performed with the random forest algorithm which is a machine learning classifier. This classifier gives the importance of each variable in the classification using the mean decrease accuracy based on which the importance of satellite image bands and ancillary data were assessed. The land use and land cover map statistics were based on adjusted error matrix (Olofsson et al., 2013) that enables computation of classification, area estimates and area uncertainty. This method adjusts the ordinary pixel count error matrix (Congalton, and Green, 2009) with the area of each land use and land cover categories on the classified maps. Furthermore, the simulated and predicted trends of land use and land cover changes were 48
modelled with Markov-cellular automata model using classification algorithm biophysical and anthropogenic variables as predictors. A detailed description of the method applied for land use and land cover classification and land use and land cover simulation is presented in Chapter 7.
2.6 Data analysis In Chapter 3, data analysis started with simple frequency counts of savanna woodlands management related terminologies in the reviewed biodiversity policy documents. The counts were done using the advanced search function for whole words in PDF documents whilst the search function was used for documents in Microsoft Word. For each official document from the websites, first, sections which contained information on woodland management and human-herbivores induced disturbance and biodiversity issues were retrieved. Thematic analysis was used to analyse the qualitative data. Each theme captured important aspects about the gathered data in relation to the research objectives (Braun and Clarke, 2006). An inductive qualitative data analysis approach (Strauss and Corbin, 1990) was applied where themes related to woodland management and sustainable livelihoods were derived. The issues were then presented in a text box and tables.
For Chapter 4, descriptive statistics were used to summarize measured study attributes data and study variables. The Shannon Wiener Index (H) formula was used to calculate woody species diversity and basal area was calculated from stem circumference. Canopy volume were calculated from canopy parameters, whereas, plant density was calculated based on number of woody plant per sample plot area. The Kruskal-Wallis H, Analysis of Variance (ANOVA) test (KWH test), was used to test for significant differences across the three defined land use categories. Post-hoc analyses using the z-test were carried out where 49
significant differences across study strata had been recorded. Using STATISTICA version 10 for Windows, The Principal Component Analysis (PCA) was performed to highlight the driving factors of variation in the woody vegetation status across the study strata. Hierarchical cluster analysis (HCA) using Ward’s method was performed using a matrix of 45 plots and 266 woody species using species diversity data to group sample plots on the based on their woody composition similarity.
Data collected through the household questionnaire survey (Chapters 5 and 6) were coded to facilitate entry using the Microsoft Office Excel Package 2007. Coding involved organization of data into categories and where each response was assigned a numerical code as described by Strauss and Corbin (2008). Descriptive statistics such as percentages of responses, frequencies and means were obtained and cross tabulations were made. Data from the household interviews were summarized into frequencies of responses. Coordinates for sampled households were entered in a Geographical Positioning System for future reference. Field observations were analyzed using deductive/theoretical thematic analysis (Strauss and Corbin, 2008). According to Braun and Clarke (2006), the theoretical thematic analysis is a qualitative data analysis approach which is based on themes which are predetermined by the researcher's analytic interest in the study area and is more explicitly driven. Thus, in the present study, themes were related to socio-economic values derived from woodland ecosystem productive activities, indicators of sustainable livelihoods, coping and adaptation strategies towards land use and land cover changes as well as socio-economic impact of human-wildlife on agriculture based livelihood.
Qualitative data from household interviews were analyzed using content analysis where the key issues were grouped into themes (see, Mutanga et al., 2017). Systematic analysis of text 50
institutions and their structures involved in communal woodland management was done (Allendorf, 2010). Focus group discussant responses were sorted into different influencing determinants of promotion of institutions of communal woodland management. The strengths, weakenesses, opportunities and threats (SWOT) tool analysis was used to group the high frequently mentioned internal and external drivers of promoting local level institutional arrangement for community woodland management (Mashapa et al., 2014a). Following the methods of Gbedomon et al. (2016) household economic benefits derived from woodlands were analyzed. Annual incomes of all woodland ecosystem productive activities were aggregated to get the global annual income (GAI) of individual household study respondent.
For Chapter 7, land use and land cover percentage change was cross-tabulated and descriptive analysis of change was performed over the period 1990 to 2015. Using the IDRISI standardization algorithm, the biophysical/socioecological data was transformed into comparable transition potential values that represent the probabilities of land use and land cover category changes (Eastman, 2003). Then, an Analytic Hierarchy Process (Saaty, 1977) was used to compute biophysical/socioecological factor weights based on preference factors derived from study respondents of the household survey of communal settlers resettled on southern part of SVC. Computation of steady state and transition probabilities of the landscape of SVC was done using Markov chain analysis. Following Pastor et al. (1993), the Markov chain analysis was used to compute transition probabilities based on the LANDSAT derived land use and land cover maps for 1990, 2000, 2005, 2010 and 2015. Spatial allocation of the simulated land use and land cover probabilities was processed mainly from the three datasets of SVC maps, namely, (1) the preceding initial land use and land cover base map of SVC derived from the LANDSAT, (2) the transition potential maps of SVC, and (3) the periodic transition area matrix, which were integrated using multi-objectives land 51
allocation (MOLA) and cellular automata spatial filter in order to simulate the subsequent land use and land cover map. The Markov-cellular automata model was run through a MOLA procedure and a cellular automata spatial filter was done in order to simulate future maps (2020, 2030 and 2040) as prediction of land use and land cover of SVC in Save Valley, southeastern lowveld of Zimbabwe.
2.7 Conclusion Chapter 2 highlighted a variety of research methods design employed to address the set study objectives. Justification of the materials and methods which were adopted in this study are described and presented in this Chapter. A critique of the quantitative and qualitative approaches and the justification for their use in this study is also highlighted. The data collection instruments, which were used to collect the qualitative and quantitative data, which addressed the objectives of the study, are presented in detail. The sampling procedures and the ethical considerations, which were considered during each stage in data collection procedure, are explained. The steps, which were followed in data capture, sorting, transcription and final analysis and presentation, are described in the end. However, in-depth details on materials and methods for each Chapters 3, 4, 5, 6 and 7 are presented specifically in each Chapter, respectively.
52
CHAPTER 3
A REVIEW OF LAND USE AND SAVANNA WOODLAND MANAGEMENT IN ZIMBABWE: A CASE STUDY OF THE SAVE VALLEY, SOUTHEASTERN LOWVELD ZIMBABWE
This Chapter is based on a slightly modified version and submitted for publication: Mashapa, C., Gandiwa, E. and Muboko, N. (Submitted). An analysis of land use and savanna woodland management in Zimbabwe: A case study of the Save Valley, southeastern lowveld Zimbabwe.
53
3.1 Abstract This study explored the role of local people in Save Valley, southeastern lowveld of Zimbabwe and their respective local level institutions play in woodlands utilization and management. The study conducted in April 2017 reviewed existing literature from published documents and national reports on the subject under review. Exploitation of timber from indigenous woodlands of Mutema-Musikavanhu communities in Save Valley, southeastern lowveld Zimbabwe, began in the 1890s to supply the mines and construction of the railway infrastructure. This exploitation of timber has continued to date albeit changed use patterns, where timber is now used mostly for fuelwood for brickmaking, building construction and furniture. Local traditional institutions of woodland management are rather diffuse and need to be supported by government institutions in Save Valley, southeastern lowveld Zimbabwe. Furthermore, poor coordination of policy implementation and enforcement among state forestry and agriculture sector agencies cause conflicts in land use and woodland management. There is need to examine whether under this disturbance gradient of land use, the woodland ecosystem can be able to self-reorganise in a steady state and transition to ensure persistence and co-existence of wildlife, human and woody vegetation for sustainable ecosystem services and functions across the study area.
Key words: communal area, livelihood, savanna woodlands, protected area, Zimbabwe
54
3.2 Introduction Savanna woodlands are one of the world's most extensive biomes (Williams et al., 1996). Woody vegetation constitute the most extensive tropical woodland and dry forest formation in Africa (perhaps even globally), covering an estimated 2.7 million km2 in regions receiving more than 600 mm mean annual rainfall on nutrient-poor soils (Campbell, 1996). Woodland is defined as property claims to categories of trees and shrubs by people on land under various kinds of tenure (Fortmann and Nhira, 1992). Fortmann and Nhira (1992) defined woodland tenurial niches covering substantial portions of southern Africa: Angola, Malawi, Mozambique, Tanzania, Zambia and Zimbabwe. Savanna woodlands comprise systems with a continuous herbaceous layer and a discontinuous woody stratum. Accordingly, woody vegetation plays important roles in the functioning of woodland ecosystems and service provision to both wildlife and livelihoods (Tesfaye, 2017). Nevertheless, the impact of human activities was recognized much earlier for natural forest and woodlands than other ecosystems (Clement, 1916). The first references can be traced back to Plato (ca 400 BC), who suggested that soil erosion and the drying up of springs could be due to deforestation (Daily, 1997). Human activities, such as land use for agriculture, wildlife ranching and woodland exploitation has an impact on almost woody vegetation over recent centuries (Dezecache et al., 2017).
Arnold et al. (2006) pointed to the continuing importance of woodlands in Africa, citing the prediction of the International Energy Agency (IEA) (2002) that biomass energy will still account for an estimated three quarters of total domestic energy in Africa by 2030, and that the absolute number of people using wood fuel will rise by more than 40% during 2000-30 to about 700 million people. Global consumption of fuel wood remains the dominant source of domestic fuel in some developing countries like Zimbabwe, where alternative energy supply 55
are limited and expensive for rural people (Muboko et al., 2013; Muboko et al., 2014). The International Union for Conservation of Nature (IUCN) in a General Statement on its African Special Project which began in 1960, stated that the accelerated rate of destruction of wild flora and habitat in Africa, is the most urgent international conservation problem of the present time (Worthington and Treichel, 1961). It was suggested that the peoples of Africa and their administrations should be continuously induced to look favorably upon their unique inheritance of faunal and floral resources. This highlighted an old special concern over the limited number of African systems and institutional structures engaged in wildlife and woodlands conservation, and Zimbabwe is no exception (IUCN, 2000).
The disturbances on woody vegetation due to herbivory, fires, drought, frost, diseases and human activities is a cause for concern in Save Valley, south eastern lowveld of Zimbabwe and it has been an area of continuous research focus (e.g., Gandiwa and Kativu, 2009; Dunham, 2002; Mashapa et al., 2014a; Zisadza-Gandiwa et al., 2014; Mhuriro-Mashapa et al., 2017; Mhuriro-Mashapa et al., 2018). With the onset of the Zimbabwe’s ‘fast track’ resettlement programme of 2000, villagers moved into parts of protected areas in Save Valley, southeastern lowveld of Zimbabwe, opening up the woodlands for agriculture (Rukuni et al., 2006). The land resettlement process of the Zimbabwe land reform programme took insufficient account of some of the fundamental requirements for wildlife and woodlands conservation (Fakarayi et al., 2014). Societal needs for food security and basic livelihood options depended on woody vegetation are therefore perceived affected, either directly or indirectly by wildlife and human induced disturbances on woodland ecosystem (Mashapa et al., 2014b). The vulnerability and resilience of such socio-ecological systems need to be established in order to enhance the adaptive capacity of woodland ecosystems (Langat et al., 2016). 56
This article reviews land use and management of woodlands across the protected Save Valley Conservancy and the bordering communities of Mutema-Musikavanhu communal lands in Save Valley, southeastern lowveld of Zimbabwe. The attempt is to highlight and inform how human activities and protected area management influence woodlands conservation under varying protection gradient (communal land and protected wildlife area). Successful management of large areas of woody vegetation depends to a large measure on knowledge on how the woodland resources are utilised and the policy environment to conserve woodlands (Magadza et al., 1993; Tesfaye, 2017). For the purpose of this review, a distinction was made that involves the land use controls that relate to the use of wood resources from woodlands. This review is in two sections which attempts to first: i) understand the role of local people on woodlands utilization and management in Save Valley, southeastern lowveld of Zimbabwe, and secondly, ii) to analyze factors influencing the effectiveness of local community and state institutions on land and woodlands management within the socioecological system in Save Valley, southeastern lowveld of Zimbabwe. We argue that contests over the control of woody resources within the communal areas, protected areas and unbalanced power relations between the community and state agencies, explain the ineffectiveness of community and national institutions to collaboratively regulate woodland management (Rita et al., 2017). We discussed critically at these local and national level processes within the prevailing policy environment of land use/land tenure and woodlands management in Zimbabwe.
57
3.3 Materials and Methods 3.3.1 Study area For the purpose of this review study, Save Valley, southeastern lowveld Zimbabwe is defined as constituted of the protected area namely, Save Valley Conservancy and the bordering communities of Mutema-Musikavanhu communal areas (see, Figure 2.1). This study area lies in the globally renowned northern Greater Limpopo Transfrontier Conservation Area of southern Africa (Muboko, 2017). Originally, Save Valley Conservancy was a co-operative structure comprising twenty-four private, individual cattle ranches with no marked buffer zone to boarder the adjacent rural settlement of Mutema-Musikavanhu communal areas. These connected cattle ranches were then converted into a single co-operatively managed commercial wildlife reserve conservancy in 1993 (Wels, 2003). However, given that the land use of Save Valley Conservancy was initially planned for cattle ranching which is compatible with close proximity of human settlement without much need for large buffer zones, thus, the conversion into a protected wildlife conservancy in the early 1990s pose socio-economic challenges to the local communities and protected area in regards to human-wildlife conflicts as human and wildlife share woodlands resources in close proximity (Mhuriro-Mashapa et al., 2017; Mhuriro-Mashapa et al., 2018).
In this drought prone semi-arid area, irrigation agriculture, rain-fed dry land conservation agriculture and livestock production are ideal and common livelihoods activities across Mutema-Musikavanhu communal areas, with most people relying on small-scale commercial farming (Mhuriro-Mashapa et al., 2018). The study area is located in the semi-arid climate environment with deciduous woodland savanna of low and variable rainfall and poor-quality soils (Poshiwa et al., 2013). The Save Valley Conservancy, southeastern lowveld of Zimbabwe has a diverse vertebrate fauna that consists of 89 species of mammals, 400 species 58
of birds, 76 species of reptiles, 28 species of amphibians and 50 species of fish (Dunham, 2002; Dunham, 2012). For more details on the description of study area, see Chapter 2.
3.3.2 Literature review process and assessment framework We approached this review from a historical perspective (Gandiwa et al., 2014) to allow for the tracking and evaluation of cultural, legal institutional framework for land use and woodland conservation and governance in communal and protected areas. The historical perspective allowed us to trace the developments in the mainstreaming of land use and woodlands management and governance in Save Valley, southeastern Lowveld of Zimbabwe and how it is related to the Zimbabwean scenario. In this review article, we primarily noted and dealt with two of the woodland tenure niches i.e. gazetted woodlands controlled by the state, in this case is Save Valley Conservancy and the communal woodland in human settlement of Mutema-Musikavanhu communal areas.
Land use and savanna woodland management in Save Valley, south eastern lowveld of Zimbabwe was reviewed mainly based on case studies and research articles drawn from Zimbabwe, where the authors are more familiar with. The researchers are familiar with Save Valley, southeastern lowveld of Zimbabwe from previous related socio-ecological research (e.g. Gandiwa and Kativu, 2009; Mashapa et al., 2014ab; Zisadza-Gandiwa et al., 2014; Muboko, 2017; Mhuriro-Mashapa et al., 2017; Mhuriro-Mashapa et al., 2018) and the study area appeared appropriate because it had experienced human-wildlife conflicts, spontaneous encroachment of human settlement onto protected areas (see, ZPWMA, 2011; Kahuni et al., 2014; Zisadza-Gandiwa et al., 2014; Mhuriro-Mashapa et al., 2017; Mhuriro-Mashapa et al., 2018).
59
Following the methods by Spellerberg (1998) and Muboko (2017), the review concentrated mostly on reports of original research articles [see reference list]. To access and collect information, peer-reviewed and published research articles, obtained through internet sources, books and reports were examined. Following Muboko (2017), internet scholarly search engines were used with inquiry guided by inputting the following words or phrases, “land tenure,” “communal woodland utilization and management,” “woodland management in protected areas” “woodland utilization,” “state actors in woodland management,” “non-state actors in woodland management” “indigenous peoples”. To get more relevant data from the internet search engines and to ensure that the relevant documents were obtained, each document, especially the abstract, was first gleaned following a method proposed by Cohen (1990) referred to as the preview, question, read and summarize (PQRS) system outputs for review analysis. At the preview stage of PQRS, articles whose abstracts were deemed relevant to the objectives of the review study were selected. The selection was based on the presence of at least two of the key words, phrases or similar wordings which necessitated the downloading of a document, resulting in the review analysis of 73 documents.
3.3.3 Data analysis After the preview stage of PQRS, a meta-synthesis technique was used to evaluate, analyze and interpret the findings from the literature review. The downloaded 73 documents were read through to make an assessment and identification of common core elements and emerging themes following Cronin et al. (2008). According to Polit and Beck (2006), a meta-synthesis involves analyzing and synthesizing key elements in each study and transforming the findings into new conceptualizations and interpretations. To assist in classifying and grouping documents by type of content, content was grouped 60
into components for analysis based on the two objectives of the review study, i) indigenous people, land tenure-land use systems (communal and protected) and woodland resource use, woodland management, human settlement and encroachment onto protected areas, local level institutional (traditional and state actors) framework and its influence on land and woodlands
management, and, ii) undertaking a more
systematic and critical analysis and review of the content, and then making a summary of the findings (Muboko, 2017). Using this analysis, findings were discussed based on the two defined study objectives.
3.4 Findings and Discussion 3.4.1 The role of local people in woodland utilization and management in the Save Valley, southeastern lowveld of Zimbabwe Woodland ecosystems in Zimbabwe exist under different land tenure systems (Shumba, 2001). The protected area estates like Save Valley Conservancy, falls under the state agent namely Zimbabwe Parks and Wildlife Management Authority and the Forestry Commission of Zimbabwe. Protected areas make up to 28% of the land area of Zimbabwe (GoZ, 2010). The communal woodland of Mutema-Musikavanhu communities falls under the auspices of Chipinge Rural District Council of the Zimbabwe Ministry of Local Government. The role of local people and threats to woodlands and biodiversity in Save Valley, southeastern lowveld of Zimbabwe were recorded in year 2000 during the development process of the National Biodiversity Strategy Action Plan (NBSAP) of Zimbabwe (Shumba, 2001; GoZ, 2010). Critical threats to woodlands and biodiversity conservation and the underlying causes of degradation as recorded by NBSAP in 2000, are presented in Table 3.1, highlighting the role and impact of local people in shaping woodland status in Save Valley, southeastern lowveld of Zimbabwe. 61
Table 3.1. Summary description of woodland ecosystems in Save Valley, southeastern lowveld of Zimbabwe, in terms of importance to woodlands and biodiversity conservation, status and underlying causes of degradation. Source: Shumba (2001) and Muboko (2017) Woodland ecosystem and importance to biodiversity conservation southern Africa bushveld;
Threat woodlands
to Underlying causes of Impact woodland degradation
-conversion agricultural land
to - agriculture expansion -habitat fragmentation converting woodlands and woodland into cultivation lands degradation
-pollution; Protected area provides wildlife and woodlands protection in Save Valley Conservancy; forms part of the Greater Limpopo and Limpopo Shashe Transfrontier Conservation Areas
- illegal settlements -unsustainable harvesting of timber
-reduced woodland ecosystem services and - human encroachment declining human wellinto protected areas being
-bush encroachment by - deforestation invasive species like tickberry -increased frequency (Lantana camara) and intensity of forest/veld fires -excessive herbivory resulting in loss of due to vegetation cover and overpopulation of biodiversity herbivores -veld fires -increased reliance on natural resources (trees -climate change and and forests) for Communal high frequencies of livelihoods woodlands of drought Mutema-growing market for Musikavanhu - harvesting of non- firewood in urban communal timber forest centres due to limited areas in products grid electricity southeastern accessibility and other lowveld of -woodland loss economic challenges Zimbabwe;-dominated by -unfavorable agriculture mega outputs and market herbivores prices, resulting in more -dominated by people unsustainably woodlands harvesting wood -high species resources as an richness and alternative income diversity of source. woody plants 62
-biodiversity loss -increased conflict between humans and wildlife at interface -extinction or extirpation of threatened species -increased vulnerability for species, with low productivity and population numbers -restricted habitats -limited ranges
and
patchy
ecosystem
-Woodlands converted into scrublands
For rural communities in Save Valley, southeastern lowveld of Zimbabwe, woodlands are the primary energy source in the form of wood fuel or charcoal. Fuel wood demand in the region was reported on the increase due to a variety of factors which include escalating national energy cost, rapid population growth, persistent poverty and lack of realistic energy alternatives (Mashapa et al., 2014b). Shumba (2001) state that 31 percent of Zimbabwe’s total energy consumption is wood fuel, with 80 percent of the energy demands of communal areas being met by wood fuel. Therefore, preferred woody species are being rapidly destroyed for firewood and other benefits that the local people derived from woodland resources. In Save Valley, southeastern lowveld of Zimbabwe, natural forests and woodlands have been the primary energy sources in the form of wood fuel/charcoal and a source of commercial timber, with hardwood species such as Kiaat (Pterocarpus angolensis), Chamfuta (Afzelia quanzensis), African blackwood (Dalbergia melanoxylon), Leadwood (Combretum imberbe), African ebony (Diospyros mespiliformis), African teak (Pericopsis elata) and Red Mahogany (Khaya anthotheca) being reportedly targeted with excess harvesting (McCullum, 2000; Pole, 2006; Mashapa et al., 2014). These important woody species are now mainly confined to protected areas of Save Valley Conservancy (e.g., Gandiwa, 2011).
It was noted that during the pre-colonial periods of the 18th century, wood poaching in protected areas was minimal in Save Valley, southeastern lowveld of Zimbabwe, if ever in existence (Mamimine and Mandiregerei, 2001). However, there has been a drastic rise in wood poaching and woodlands clearance for agriculture and related wood resources driven industries in colonial periods from the 1890s onwards, as triggered by industrialisation in the country (Shumba, 2001; Rukuni et al., 2006). It is projected few of Zimbabwe’s protected natural forests and woodlands will survive to the end of the 21st century in their current pristine state (Magadza et al., 1993; Shumba, 2001), thus the communal woodlands with open 63
access and its natural exposure to the “tragedy of the common” are at high risk of degradation in Save Valley, southeastern lowveld of Zimbabwe (Hardin, 1968; Gandiwa, 2011; Mashapa et al., 2014b).
The savanna woodland cover an area of about 2 million ha in southeastern lowveld of Zimbabwe (Shumba, 2001), whereas, of this area, 638,000 ha is in communal areas; 343,000 ha is on private lands (especially the large scale commercial farms); 568,000 ha is in Zimbabwe National Parks and Wildlife areas (where commercial logging is prohibited); and 439,000ha is on gazetted natural forest areas (GoZ, 2010). Exploitation of timber from these indigenous woodlands which were not gazetted by then, began in the 1890s of colonial period to supply the mines and rails of the railway lines established by the white settlers who mapped the Pioneer Column route of colonizing the present country Zimbabwe (Wels, 2003; Scoones et al., 2010; GoZ, 2010). However, this operation became so extensive that regulation became necessary, leading to the enactment of the Forest Act in 1949 (McGregor, 1991). Despite a lack of adequate data, it is reported that these woodlands have been overexploited and by 1977 timber demand had already exceeded supply in southeastern lowveld of Zimbabwe (Wels, 2003). Apart from commercial timber which is limited to eastern highlands of Zimbabwe and the teak indigenous woodlands limited to western Zimbabwe, the woodlands of Save Valley, southeastern lowveld of Zimbabwe is a source of a wide range of products of a variety of woodland based livelihoods activities which include commercial hardwood timber, fuel wood, small artisan crafts, fodder, wild fruits, honey, mushrooms, insects, bark for rope, medicines, leaf litter and gum.
Harvesting of wild vegetables has increased as interest in traditional foods as an alternative and healthier lifestyle has grown (Shumba, 2001; Mashapa et al., 2014b). An estimated gross 64
value from harvesting of woodland resource products in Zimbabwe was estimated US$110 million a year (Campbell, 1996). Woodland resource harvesting is a significant component of rural livelihood strategies in both communal and resettled areas in Save Valley, southeastern lowveld of Zimbabwe given the high prices of agricultural products, unreliable rainfall and unemployment (ZimStats, 2013; Mashapa et al., 2014b; Mhuriro-Mashapa et al., 2017).
Save Valley, southeastern lowveld of Zimbabwe is an area of global significance for the characteristic mega-fauna and flora of southern Africa (ZPWMA, 2011). However, over the past three decades, rapid land use changes have also substantially reduced wildlife habitats and woodlands within this region (Pole, 2006; Kahuni et al., 2014; Zisadza-Gandiwa et al., 2014). Land use has changed from wildlife protected area to small scale farming characterized by subsistence mixed crop and livestock production in many parts of Save Valley, southeastern lowveld of Zimbabwe, following the Zimbabwe government’s ‘fast track’ resettlement programme post 2000 (Rukuni et al., 2006; Scoones et al., 2010; MhuriroMashapa et al., 2017). With the onset of the resettlement programme in Zimbabwe, villagers moved and resettled into parts of the protected Save Valley Conservancy clearing woodland habitats for agriculture and human settlement (Kahuni et al., 2014; Zisadza-Gandiwa et al., 2014). Large areas of woodland were cleared for subsistence farming within Save Valley Conservancy. Due to the poor nutrient content of the soils, resettled farmers practice slash and burn agriculture to boost yields, resulting in increasing woodland clearance (Kahuni et al., 2014).
There is human resettlement of Chigwete and Masapas game ranches in the centre of the protected Save Valley Conservancy, there was clearance of vast woodlands and threatens to prevent wildlife movement between the northern and southern halves of the conservancy. 65
There ia also human occupation in the south-eastern part of Save Valley Conservancy (Mkwasine and parts of Senuko and Levanga ranches) which jeopardize connectivity between Save Valley Conservancy and the rest part of Greater Limpopo Transfrontier Conservation Area (Scoones et al., 2010). Thus, this evidence of human settlement encroaching onto protected areas exhibit habitat fragmentation which can threatens protected woodlands and biodiversity in Save Valley, southeastern lowveld of Zimbabwe. The encroaching human settlement onto protected areas further decrease the home range of large herbivores thereby triggering over-browsing in protected areas. With reported over population of elephants in Save Valley Conservancy (Dunham, 2012), woodland degradation is likely in habitat of small home ranges. Elephants are known to cause top-kill and die-back of trees and shrubs due to excessive herbivory which can degrade woodlands and convert them into scrublands (ZPWMA, 2011, Mashapa et al., 2013).
As human population continues to increase in Save Valley, southeastern lowveld of Zimbabwe (ZimStats, 2013) more woodlands are likely to be opened up for settlement and cultivation. Land use and land cover changes within Save Valley Conservancy are related to herbivory disturbances and anthropogenic factors (ZPWMA, 2011). The conservation frontier status of Save Valley Conservancy has several implications. First, local people might have had intentions of finding enough space to create an idealized model of their areas of origin (Chibemene, 2006; Kahuni et al., 2014; Zisadza-Gandiwa et al., 2014) as they unilaterally encroached for human settlement on the protected area of Save Valley Conservancy.
Local people may have assumed that land and woodland resources were abundant, which determined their behaviour, even though there was evidence to the contrary (Shumba, 2001; Rusike et al., 2006). Thus, there are problems of human encroachment on protected areas in 66
Save Valley, southeastern lowveld of Zimbabwe, which led to the clearance of woodlands to establish agricultural fields and homesteads (Scoones et al., 2010; Zisadza-Gandiwa et al., 2014). Furthermore, neighbouring communal land residents poach animal and wood resources from Save Valley Conservancy, the claims which may be based on prior usage rights as indigenous people who were earlier relocated by the white settlers during the colonial era to pave way for establishment of the same protected area (Chibemene, 2006; Kahuni et al., 2014). Moreover there were reports of wild animals which raid crops as well as high incidences of livestock predation depicting a record of human-wildlife conflicts within Mutema-Musikavanhu communities surrounding Save Valley Conservancy in southeastern lowveld of Zimbabwe (Mhuriro-Mashapa et al., 2017; Mhuriro-Mashapa et al., 2018).
With high incidents of such human-wildlife conflict local people tend to retaliate by poaching wild animals and woodland resources from protected areas. In that regard, a recognized and locally appropriate response to human-wildlife conflicts, probably in the form of Communal Area Management Programme For Indigenous Resources (CAMPFIRE) (Gandiwa et al., 2013) is suggested to mitigate local people’s perceptions of conservation as a conflict, this approach aim for mutual benefits for wildlife, woodland management and agro-based livelihood thereby promoting co-existence of local people and the protected Save Valley Conservancy (Mhuriro-Mashapa et al., 2018).
3.4.2 Local level institutional control on woodlands management in Save Valley, southeastern lowveld of Zimbabwe Traditional leadership are custodians of communal land with the powers vested in them by the Traditional Leaders Act of Zimbabwe and they play prominent roles in decision making concerning communal woodland governance. With regard to regulation (rules establishment), 67
access to woodland resources, monitoring/ control and punishment of offenders, they have full privileges of making decisions and are also in charge for their enforcement. For instance, special hard woody species like African mahogany (Khaya anthotheca), were only cut after seeking permission form the traditional leaders, even though the consent was reported tainted with corruption or nepotism (Mashapa et al., 2014b). So traditional access to woodland resources appears deeply etched into African culture. Mutema-Musikavanhu communities in Save Valley, southeastern lowveld of Zimbabwe are characterized by the presence of sacred controls enforced by customary or traditional institutions and practices as defined by Chibemene (2006) and Nhira and Fortmann (1993). Sacred controls are defined as norms of wood resource use and protection that are based in folklore-traditional religious beliefs or indigenous ecological knowledge systems that are enforced by individual internalization of the norms, community sanction or by religious and/or traditional leaders (Kopytoff, 1987).
In Mutema-Musikavanhu communities some households did not use certain tree species because
of
their
ritualistic
significance,
e.g.,
Kudu
berry
(Pseudolachnostylis
maprouneifolia). These woody species could still be cut, e.g., when clearing fields so the fact that they could not be used after being cut did not preclude their being cut. However, it was noted that local people only cut trees or shrubs when they intended to use them and so would not normally cut a tree or shrub for which there is no use except when clearing fields for cultivation.
The idea that a tree is only cut for a purpose has been linked to the fact that people of Mutema-Musikavanhu communities show that they are stewards of woodlands (Nhira and Fortmann, 1993). This cognitive set also applied to indigenous wild fruit trees such as Bird plum (Berchemia discolor), Baobab (Adansonia digitata) and African ebony (Diospyros 68
mespiliformis) which occur in large numbers in Save Valley, southeastern lowveld of Zimbabwe, are not cut except in the process of clearing fields for cultivation (Mashapa et al., 2014). Live trees on sacred places like graveyards are not cut (C. Mashapa, pers. observ. 2017). Such controls arise from a combination of common sense and preference as part of a cognitive set embedded in the local traditional culture (or more crudely cultural baggage) of woodland resource conservation. In this sense a frontier can be capitalized as a force for cultural-historical continuity of indigenous knowledge and ecological woodland conservatism (Kopytoff, 1987). The existence of the cultural controls suggests that local people of Mutema-Musikavanhu have indigenous ecological knowledge systems which they can establish for land use and woodland conservation. However, traditional authority is unrecognized in protected area of Save Valley Conservancy and large scale commercial farms in southeastern lowveld of Zimbabwe.
There are a number of local level state institutions responsible for monitoring woodland resource usage and, more generally, natural forest resource use in Save Valley, southeastern lowveld of Zimbabwe. A common characteristic to all the state institutions is that they are imposed on the community by government bodies and that they all depend on the selected Village Development Committees (VIDCOs) and traditional leaders as Village heads to execute the functions (Murombedzi, 1992). The Environmental Management Agency (EMA) in collaboration with the Chipinge Rural District Council work with the local VIDCOs and traditional leaders with a view to exhort local people to preserve their environment with each VIDCO having two or three natural resource overseers nominated by local people based on trust. The duties of natural resource overseers of the VIDCO entails enforcing people on the need to conserve woodlands, soil, water and wildlife; control of problem animals; assessing crop damages; control of snaring, hunting and fishing as communicated to local Mutema69
Musikavanhu communal residents (Mhuriro-Mashapa et al., 2017). However, the relationship between natural resource overseers of the local VIDCOs and game rangers of Save Valley Conservancy lacks integration and there is no collaboration between these actors of the woodland conservation value chain (Mhuriro-Mashapa et al., 2018). Save Valley Conservancy game rangers embrace conservation, management and exploitative/consumption responsibilities within the protected area game ranch unlike the VIDCOs natural resource overseers who are conservative within the communal areas domain of communal woodland management (Murphree and Metcalfe, 1997).
3.4.3 The national government and its institutional controls on woodlands management in Save Valley, southeastern lowveld of Zimbabwe The review highlights the preoccupations of government agencies which did not concern themselves much with natural forest and woodland management yet much of the livelihood components in Zimbabwe had implications for the status of woodlands. National state power of the government of Zimbabwe is represented by government extension personnel, district administration, the rural district council, and the various institutions emanating from the President's Office Directive on Local Government through the Provincial Administration and Provincial Governorships of 1984 (Murombedzi, 1992). In theory community development plans are channeled up the hierarchy of development committees where choices would be made about what should be implemented and in turn channeled down the hierarchy for their implementation. In practice the lowest structure of VIDCOs did not have budgets nor are they legal persona, therefore their natural resource/woodland management plans did not usually see implementation despite that the plans could be the priority of the local people. VIDCOs are more appropriately seen as agents of the government because they had little autonomous action beyond serving as conduits for ideas emanating from the local government hierarchy 70
although they are comprised of local residents (Murombedzi, 1992). The political ties of VIDCOs to the state government had left them dependent. These issues have been the subject of other studies e.g., Scoones and Matose (1993). See, Murombedzi (1992) for an extended treatment of these structures. Community leaders and VIDCOs did react to state maneuvers, just as ordinary community members did, by manipulating state power for their own ends of excessive benefits from woodlands even if it means engaging woodland based livelihoods activities causing woodlands degradation.
Recorded is the Zimbabwe fast-track land resettlement program, which was adopted post 2000, aimed at redressing the inequitable distribution of land (Rukuni et al., 2006). It entailed acquiring land from the large scale commercial sector and redistributing it to small scale farmers from the communal lands, as part of the overall rural resettlement program (Moyo, 2000; Rukuni et al., 2006). The program was designed to redress the colonial imbalances of land holdings between the communal and large scale commercial sectors (Fakarai et al., 2015). However, some people moved onto the commercial farms and Save Valley Conservancy was not spared, allegedly as new settlers but some people with the intention of looting resources from the farms (Rukuni et al., 2006; Pole, 2006; Kahuni et al., 2014; Fakarayi et al., 2014) including wood resources, giving rise to conflicts over woodland tenure and control of woodland resources (Mashapa et al., 2014a). In cases of disputes over encroachments and threats of excessive damage on woody vegetation, the Zimbabwe office of the District Administration is normally called in to arbitrate if government extension workers fail to mediate on illegal woodland utilization. District Administration claim to act on behalf of the Rural District Council which is the local authority with legal jurisdiction over the communal area and its natural resources, as per the provisions of the Communal Lands Act 1982 and the District Councils Act 1980 (and as updated). 71
In most scenarios, VIDCO members are perceived to stay out of the more intractable land or woodland resource use disputes as they appear to favour disguising the disruptiveness of land or woodland resource use disputes while deploying what Rose (1992) has termed harmony ideologies. On many occasions and as noted, the disputants involved are concerned with protecting and promoting their land or woodland resource use interests, while the state government functionaries are concerned with maintaining their relative political power and control over the land with populist decisions at the expense of sustainable land use and woodlands conservation in Save Valley, southeastern lowveld of Zimbabwe. Surprisingly, a regime of rigid control over land settlement is coupled with a laissez faire attitude to the use of wood resources and woodlands management even though the Rural District Council could invoke the Natural Resources Act, the Communal Forest Produce Act, the Communal Land (Model) (Land Use and Conservation) By-Laws of 1985 and associated legislation to effectively manage and conserve woodlands in Zimbabwe.
The laissez faire attitude taken by the local government authorities and institutions with regard to land use and woodlands management in the study area could provide some limited space for the nascent local traditional institutions to attempt to manage natural forests and woodlands. The land-use planning and land reform programme in Save Valley, southeastern lowveld of Zimbabwe was intended to focus on both land redistribution to the landless people and natural resources management in protected areas but this element of woodland management tended to be neglected in the execution of the plan. Regrettably, communal woodland and wildlife management was not tightly integrated into the land-use planning process of land reform and rural resettlement in Zimbabwe (Kahuni et al., 2014; ZisadzaGandiwa et al., 2014), yet land use planning has profound implications on the availability of woodland resources (McGregor, 1991). The major concern on the land distribution aspects of 72
Zimbabwe land reform program was how land and woodlands could have been sustainably exploited in Save Valley, southeastern lowveld of Zimbabwe. This is an important consideration given the fact that inappropriate communal farming systems and woodlands degradation in the communal lands of the Save River catchment in Save Valley, southeastern lowveld of Zimbabwe were reported to have caused soil loss ranging from 40 to 100 tons per hectare annually (du Toit, 1985; du Toit, 1992).
The provision of government extension services tended to favour the agricultural sector to the exclusion of woodland management and wildlife sectors (Shumba, 2001; Rukuni et al., 2006). The content of community extension messages neither emphasize management of natural resources nor its woodlands conservation a priority in the farmers' own perceptions. Extension activities also tend to support activities that rely on the exploitation of woodlands e.g. tobacco and grain growing which requires the building of bam, granary from wood, respectively. No effort is made to control such exploitation of woodlands by farmers, leading to other farmers being unwilling to live by the rules imposed by the VIDCOs and traditional leaders enforcing woodlands conservation. Thus the community extension mode followed also lead to segmentation of the communities living little room for community effort towards natural forest and woodlands management in Save Valley, southeastern lowveld of Zimbabwe (Zwart, 1990; Mashapa et al., 2013).
There also exist policy disjunctions, where e.g. the Zimbabwe Ministry of Agriculture, Lands and Rural Resettlement has a long-term planning framework spanning 20 years, in contrast with all the other sectoral ministries and cross-sectorial plans for the forestry sector that have five-year planning time frames (GoZ, 2010; Rukuni et al., 2006). Such a short planning phase does not allow for an adequate assessment of the impacts of these plans and polices on 73
woodland ecosystems and human well-being in the long term. The Zimbabwe Comprehensive Agricultural Policy Framework 2012-2032 recognizes the need for compliance with intellectual property rights requirements and international and local sanitary and phytosanitary standards. The agricultural policy recognizes the value of agro-ecological zones and recommends their re-assessment in response to climate change impacts (Rukuni et al., 2006). This is all evidence of a well-planned long term agriculture development strategy for Zimbabwe with a less pronounced plan for natural forestry and woodland management and conservation.
As for the natural resources and forestry sector in Zimbabwe, intra-sectoral government agency coordination on natural forests and woodlands conservation issues is largely weak (Shumba, 2001). There is no coordination of conservation issues at national level as functions are split between the Biodiversity Office in the Ministry of Environment, Water and Climate and Environmental Management Agency (Shumba 2001; Goz, 2010). This is also reflected in the lack of harmonized reporting and monitoring on multilateral environmental agreements to leverage resources, especially with the United Nations (UN) Convention to Combat Desertification (UNCCD), the Convention on Wetlands of International Importance (also known as the Ramsar Convention), the UN Framework Convention on Climate Change (UNFCCC), the UN Convention on Biological Diversity (UNCBD) and the Convention on International Trade in Endangered Species (CITES) (Dzingirai and Matuture, 2008). Natural forest, woodlands and Biodiversity conservation has not been mainstreamed into the Ministry of Environment, Water and Climate as the Biodiversity Office in Zimbabwe is considered a project that is externally funded (GoZ, 2010). This has led in part to data and other information on woodlands and biodiversity becoming outdated, unavailable or scattered across various institutions (Shumba, 2001). 74
Zimbabwe is party to the United Nations Convention on Biological Diversity (UNCBD) and accordingly has obligations to implement the provisions of the convention (Shumba, 2001). The convention encouraged Zimbabwe to prepare a national biodiversity strategy and to ensure that it is mainstreamed into the planning and activities of all sectors that have an impact on biodiversity and woodlands management is no exception. In line with this provision, Zimbabwe developed its first national biodiversity strategy and action plan (NBSAP) in 1998, which covered the period 2000-2010 (Sumba, 2001). In 2013, Zimbabwe embarked on a process of reviewing the NBSAP and aligning it with the UNCBD Strategic Plan 2011-2020 which emphasizes the communication, education and public awareness and the ecosystems approach, including the value of ecosystems, in the development of NBSAPs (GoZ, 2010). The NBSAP, through its strategic objectives, was designed to contribute to the national development targets in the economic blueprint for the period 2013 to 2018, of the Zimbabwe Agenda for Sustainable Social Economic Transformation (ZimAsset) (GoZ, 2013), but little had been achieved by the time of the review study, mainly due to limited national funding of ZimAsset (ZimStats, 2013). There is need for Zimbabwe’s national commitment with funding for government extension service delivery in the forestry sector to utilize traditional indigenous ecological knowledge, research, technology, innovations and best practices to protect the environment, conserve and sustainably use biodiversity and woodland ecosystems to benefit present and future generations.
3.5 Conclusion The study reviewed the role of local people in woodland utilization and management with the range of local level institutional controls that are shown in the context of communal and protected areas in Save Valley, southeastern lowveld of Zimbabwe. The role of people is 75
characterized by their intensive utilization of woodland resources to meet their increased demand for fuel wood, timber, land for agricultural expansion and settlements. Woodland ecosystem support people with tangible and intangible goods and services but are strongly influenced by anthropogenic disturbances that occur regularly in woodland ecosystems originating from people overharvesting wood resources, clearing woodland for cultivation and causing woodland degradation in Save Valley, southeastern lowveld Zimbabwe (Campbell, 1996; Casey and Mathew, 2011).
The drivers of woodland degradation are perceived to lie in the failure of the government institutions to adequately involve local traditional institutions and other stakeholders in the management and governance of the woodland resources in Save Valley, southeastern lowveld of Zimbabwe. Agencies associated with the local government are pre-occupied with controlling agricultural extension service delivery while showing little intervention for woodland management or the implications of land use for the status of woodlands. The state government extension activities in collaboration with local level traditional institutions should include woodlands management as an integral form of land use management. We recommend strengthening of the national forest policy that is mainstreamed into all sectors and incorporated into the national accounting and reporting system. This calls for the development and implementation of a comprehensive communication, education and public awareness strategy on the conservation and sustainable use of land and natural forests. Furthermore, we recommend a study of woodlands and livelihood in the context of sustainable management to understand the links and interactions between the woodland resource, users, and institutions that mediate between them.
76
CHAPTER 4
IMPACT OF ANTHROPOGENIC FACTORS AND HERBIVORY DISTURBANCES ON WOODY VEGETATION ACROSS A PROTECTION GRADIENT IN SAVE VALLEY, SOUTHEASTERN LOWVELD OF ZIMBABWE
This Chapter is based on a slightly modified version and submitted for publication: Mashapa, C., Gandiwa, E., Muboko, N and Chinho, T. (Submitted). The influence of anthropogenic and herbivory induced disturbance on woody vegetation across a protection gradient in Save Valley, southeastern lowveld of Zimbabwe.
77
4.1 Abstract This study evaluated the influence of humans and large herbivores utilization on woody vegetation status across three land use categories; a protected area of Save Valley Conservancy, the buffer zone and the adjacent communal lands of Mutema-Musikavanhu communities in Save Valley, southeastern lowveld of Zimbabwe. A total of 45 plots (size, 50 m x 20 m) were assessed with 15 plots randomly placed in each of the three defined study stratum where vegetation attributes, herbivory and anthropogenic disturbance levels data were collected in November 2017. Significant differences were recorded on tree, shrub and sapling density, dung counts density, stump density, tree and grass height, tree canopy cover, basal area, herbivory damage level on woody plants, human damage level on woody plants, grass, tree and shrub species diversity across the three study strata. There were no significant differences on fire damage levels on woody plants, number of stems per tree, alien woody plant density and coppiced woody plant density across the study strata. It was evident that anthropogenic activities and herbivory induced disturbances influence the woody vegetation status in a distinctive pattern across the study area. Human damage on woody plants were the most frequent and pronounced disturbances occurring in communal lands and herbivory damage on woody plants were most frequent and pronounced disturbances in the protected area. These human-herbivory disturbances decreased in intensity within the buffer zone, thus confirming to the intermediate disturbance hypothesis. This study presents the first quantification of significant changes in woody vegetation status due to human utilization in communal areas, herbivory in a protected area and intermediate level of woodland utilization in the buffer zone in Save Valley, southeastern lowveld of Zimbabwe. We recommend the need to revisit the management concept and policy measures associated with woodland ecosystem services and goods provision for protected areas and the surrounding rural communities. 78
Key words: buffer zone, communal woodlands, Intermediate disturbance hypothesis, protected area management, southern Africa, tropical savannah
4.2 Introduction Explanations of spatial and temporal patterns of vegetation structure, abundance and composition in nature have been a pervading ecological research theme in the realm of vegetation succession and conservation science (Clements, 1916; Walker, 1976). Resilience of savanna woodlands ecosystem to disturbance and direction of progression or retrogression would likely vary with disturbance regime severity and the prevailing environmental factors during the intervening period (Westoby, 1980; Chinuwo et al., 2010; Muposhi et al., 2016). Vegetation succession patterns of late have been explained by contradicting theories, in support of or against the conventional Clementsian mono-climax theory which postulates deterministic one climax vegetation community in a climatic or geographical region (Clements, 1916), with most of the later theories having tradeoffs but remain unreconciled (Chinuwo et al., 2010), notably the deterministic intermediate disturbance theory (Connell, 1978) and the stoichastic chaos theory of plant ecology (May, 1973; Gleick, 1987).
Local community species assembly during succession may be driven by deterministic or stochastic processes (e.g., Hidding et al., 2013; Raevel et al., 2013; Bhaskar et al., 2014). The relative spatio-temporal dynamic factors that drive succession in ecosystems is complex and not well understood (Chase and Myers, 2011; Zhou, 2014). This conundrum has its origin in the early work on ecological succession where Clements (1916) describes succession as a largely deterministic process of predictable species composition, structure, abundance, niche, and vegetation replacements, whereas Gleason (1926) and Elliot (2014) focus more on ecological succession as an outcome of stochastic processes and individual chaotic responses 79
of species to their environment. All these theories of plant ecology clearly give conflicting models for predicting/forecasting habitat changes under various disturbance regimes, thus, understanding natural ecosystem dynamics, and the disturbance cycles that govern them, will allow for more informed management decisions to be taken.
Anthropogenic and herbivory disturbances may influence woody species diversity as promulgated by the intermediate disturbance theory (Connell, 1978). The intermediate disturbance hypothesis postulates that species diversity tends to be at the peak in landscape with moderate disturbances in comparison to landscapes with severe and less disturbances which result in lower levels of species diversity (Rawal et al., 2012; Zhang et al., 2014). However, the stochastic chaos theory of plant ecology assumes unpredictable pattern in vegetation structure, abundance and composition in relation to disturbance regimes. Disturbance of woodland ecosystems by people and large herbivores are at the centre of socio-ecology, wildlife and conservation critiques, yet woody vegetation, large herbivores and people have co-existed for millennia, and are ecologically linked in influencing biodiversity and woodland ecosystem services (Ickowitz, 2006; O’Connor et al., 2007).
Woody vegetation status in tropical savanna is influenced by dynamics of the disturbance regimes in the form of herbivory (O’Connor et al., 2007), climate (Lindenmayer et al., 2006), fire (Gandiwa and Kativu, 2009) and human (Ndangalasi et al., 2007; Muboko et al., 2014). While these disturbance regimes are needed to maintain a high level of heterogeneity conducive to high levels of woody species diversity (Perrings and Lovett, 1999), the dominance of one or several of these elements can have detrimental effects (Western and Maitumo, 2004), whereas insufficient disturbance also reduces diversity of woody vegetation through homogenization (Lindenmayer and Noss, 2006). 80
Rural people in southern Africa and the effect of high levels of anthropogenic activities may lead to general woody vegetation change and fragmentation of woodland ecosystems, or even extirpation of woody species in communal woodlands (Schmidt-Soltau, 2003; Ndangalasi et al., 2007). Although protected areas are usually the object of conservation measures, the effects of excessive herbivory in these confined areas can be detrimental to woody vegetation in ways that anthropogenic disturbances can trigger woodland degradation in communal areas (Western and Maitumo, 2004). Several studies elsewhere have demonstrated that large herbivores can alter the structure of vegetation through their browsing across their home range, (e.g., Ben-Shahar, 1993; Moolman and Cowling, 1994; Cumming et al., 1997; Kerley and Landman, 2006; Mpofu et al., 2012). Similarly, anthropogenic disturbances have been reported to influence the status of woody vegetation in tropical savanna ecosystems (Bond and Keeley, 2005; Pricope and Binford, 2012; Muboko et al., 2013). Currently, anthropogenic activity is reported the driver of planetary change with the present era suggested to be referred as the Anthropocene (Smith and Zeder, 2013; Malhi et al., 2014).
There is a school of thought which focus on the causes of vegetation changes in nature, that is explained by the non-equilibrium approach (Westoby et al., 1989) that is external disturbances such as annual rainfall and its variability and on the other hand the equilibrium approach (Breman and de Wit, 1983) that favour anthropogenic and herbivory disturbances as the cause of vegetation changes. However, in the Save Valley, southeastern lowveld of Zimbabwe where we carried out the study, climate and soil type have been uniform over a long period of time (Vincent and Thomas, 1960; Nyamapfene, 1991; Moyo and O’Keefe, 1993; Pole, 2006; Mashapa et al., 2019). However, an increase in human population in Mutema-Musikavanhu communal areas adjacent to Save Valley Conservancy and an increase 81
in density of large herbivores in this protected area occurred simultaneously throughout the last 25 years and therefore the equilibrium approach is better placed to explain the perceived vegetation change of this semi-arid tropical savanna ecosystem in Save Valley, southeast lowveld of Zimbabwe (Dunham and van der Westhuizen, 2015; ZimStats, 2013).
Large herbivores like elephants are bulk feeders (O’Connor et al., 2007) and may negatively impact on the vegetation especially when confined to certain geographic boundaries. Elephant numbers have subsequently increased in Save Valley Conservancy, where there is now concern that the vegetation of this protected area may be compromised (Dunham and van der Westhuizen, 2015; Cumming et al., 1997). The loss of woody vegetation due to herbivory in protected areas, fires and human induced disturbances on communal woodlands is perceived a threat to woodland ecosystem in Save Valley, southeastern lowveld of Zimbabwe, (e.g., Ben-Shahar, 1998; Tafangenyasha; 1998; Holdo, 2007; Alam and Agbeja, 2011; O’Kane et al., 2012).
There has been a realization that anthropogenic disturbances in communal woodlands bordering protected areas, herbivory disturbances in within protected areas and the resultant edge effect within the buffer zone are seldom quantified or assessed across the protection gradient and is a cause for concern in wildlife, socio-ecology and conservation management realm in southeastern lowveld of Zimbabwe which constitute the northern Greater Limpopo Transfrontier Conservation Area in southern Africa (Magadza et al., 1993; Balme et al., 2010; Muposhi et al., 2016). The aim of this study was to determine the status of woody vegetation at three sites within a similar environment but under different disturbance regimes and to relate the differences in the vegetation to the impacts of the disturbance agents
82
operating at the three sites (communal, buffer zone and protected area) in Save Valley, southeastern lowveld of Zimbabwe.
It is important to firstly establish the quality of woodland ecosystems and secondly to quantify differences at both landscape/land use and local level, between protected area, buffer zone and communal lands (Lindenmayer et al., 2006). Study objectives were two fold: (i) to assess woody vegetation structure, abundance and composition across different land use categories of varying protection gradient and, (ii) to determine disturbance factors that influence the structure, abundance and composition of woody vegetation in Save Valley, southeastern lowveld of Zimbabwe. Proper management of habitat areas under a varying protection gradient depends on knowledge of ecological processes and variation in species composition, abundance and diversity along anthropogenic and herbivory disturbance regimes (Nesheim et al., 2010).
4.3 Materials and methods 4.3.1 Study area The study area within Save Valley, southeastern lowveld of Zimbabwe is defined as constituted of three land use categories, namely; theprotected wildlife area of Save Valley Conservancy, a strip stretch of buffer zone land use area along the eastern river bank of Save River and the bordering communal lands of Mutema-Musikavanhu communities. For detailed description of the research site and study area map, see Chapter 2. The buffer zone constitutes a strip of land to the east of Save River, defined by a radius of about 4 km from Saver River into the communal lands. The Mutema-Musikavanhu communal lands are characterized by high human density estimated at 48 people per km2 as compared to the national average of 33 people per km2. The land of Save Valley, southeastern lowveld of Zimbabwe is threatened by 83
the regional human population growth and the pressure it exerts on natural resources, especially wood as a source of fuel or building material (Pole, 2006; ZimStats, 2013; Mashapa et al., 2014a). Although, at present, human population growth and its effects appear limited to areas along the buffer zone, it is anticipated to lead to the use of increasingly vast sections of land, thus endangering natural ecological processes in the region (ZPWMA, 2011; ZimStats, 2013).
4.3.2 Study design and data collection This study was based on a stratified random design (Mueller-Dombois and Ellenberg, 1974) based on land use management. The position of the plots followed stratification according to three land use categories, namely, protected area of Save Valley Conservancy, buffer zone and Mutema-Musikavanhu communal area. For each of the three land use categories, 15 plots (size, 50 m x 20 m) were randomly placed in each study stratum across the defined protection gradient from a protected area, then buffer zone to a communal land use area. Sample plots in protected area and communal lands within less 2 km radius from the buffer zone boundary were discarded to avoid the edge effect of the intermediate zone of disturbance (Muposhi et al., 2016). From the vegetation map, sampling plots areas were randomly selected using random number tables. Sampling plots within the study area were located by generating random points (Geographical positioning System (GPS) coordinates) in the defined study strata of the study area map (Giri and Jenkins, 2005). Guided by Save Valley, southeastern lowveld of Zimbabwe study area vegetation map, GPS handsets were used to identify sample plot location across the study strata. However, inaccessible areas and were discarded and the next sampling point was considered.
84
Mutema-Musikavanhu communities live adjacent to a protected area of Save Valley Conservancy with elephants, where elephants have been present for 25 years prior to the study and were at increasing densities (between 0.5-2 elephants per km2) in southeastern lowveld of Zimbabwe. We considered the absence of large herbivores on communal lands of Mutema-Musakivanhu area adjacent to a protected area with large herbivores like elephants as a regional control for the study. The strip buffer zone area between the communal lands and the protected area was considered the zone of intermediate disturbances from both anthropogenic and herbivory. Consequently, 15 sample plots (50 m x 20 m) within a belt transect (15 km x 3 km) from communal lands, through the buffer zone to the protected area. This standard sample plot size was used according to Walker (1976) and included at least 1520 trees. The sample plots were located adjacent to each of the study stratum. The non-large herbivores sites of communal areas and the buffer zone were selected based on their linear proximity to each large herbivore site of protected area sample plot so as to mirror all other conditions (i.e., vegetation, rainfall, geology, aspect, and the presence/absence of other ungulates) as closely as possible and to account for the range of inherent spatial and temporal variability across non-large herbivore sites.
Sampling started by locating the position of the plot and the GPS position was taken as the center of the plot. A 100 m tape measure and 4 metal pegs were used to peg the plot on the ground and laid around the plot perimeter. Study variable measurements were done following the methods described in Table 4.1. For each woody plant within the sample plot, species name, number of saplings, number of trees, number of shrubs, height, stem circumference, canopy diameter, canopy depth, number of stems per plant, fire damage, evidence of browse and evidence of human damage on woody plants were recorded. Dung counts of large
85
herbivores were recorded for each plot. Data collection was done and recorded onto a data a collection tools (see Apendix 4a,b) and this was done from October to December 2017.
Table 4.1. Measured variables and methods used. Variable Species composition Tree height
Assessment methods Woody plant species were identified using a field guide by Coates-Pelgrave (1997) and with the aid of an experienced research technician. Tree height was measured by placing a calibrated 12m pole against a tree. For trees >12m, height was observed at a distance and visual estimated.
Stem Stem circumference was measured at breast height (1.3m) using a flexible 10m tape circumference measure. Density Numbers of plants were physically counted and a multi-stemmed plant was considered when the stem starts from underground. Evidence of herbivory damage on woody plants
Evidence of fire damage on woody plants Woody plant status
Canopy volume
Every woody component in the plot was assessed for herbivory by large herbivores. Herbivory damage was measured in terms of herbivory damage levels categorized as, (1) harvested leaves and twigs, (2) broken branches, (3) bark stripping and scaring, (4) broken stems, and (5) uprooted or felled down trees (Mpofu et al., 2012). The herbivory disturbances were qualitatively assessed in each 20m x 50m plot following methods adapted from (Mpofu et al., 2012). using a six-point scale (0-5) to represent the percentage of the plot disturbed as follows: 0 = no disturbance; 1 = 0-20% of the plot moderately disturbed; 2 = 21-40% of the plot severely disturbed; 3 = 41-60% of the plot excessively disturbed; 4 = 61-80% of the plot acutely disturbed; and 5 = 81-100% of the plot disturbed and completely destroyed. All woody plants occurring within plots were assessed for fire damage. Fire damage was measured in terms of the number of burnt and affected plants within a plot (Ben-Shahar, 1998). Indicators of fire damage were scorch marks on branches, fire scars and characteristic new growth from the woody plants (Gandiwa and Kativu, 2009). All woody components were assessed whether dead or alive. For dead plants, the cause of death was attributed to the one of the four factors that is (1) herbivory (2) drought (3) fire (4) human action and (5) unknown. Any dead woody plant with intact branches and standing upright, cause of death was attributed to drought. Where a plant is dead with its intact branches, stems and leaves standing upright, with no noticeable signs of browsing, fire or human action, the cause of death was unknown. For each tree rooted in the plot, the following were recorded; canopy depth (CD) (m) and greatest canopy diameters (D1 and D2) at 900. For each shrub rooted in the plot, height (m) and greatest canopy diameters (D1 and D2) were measured. Canopy diameters were measured to the nearest 0.1 m using a flexible 20 m tape.
Number stems
of Stems were physically counted for each woody plant. For each plot, total number of stems was divided by number of plants to give an average number of stems per plant. This was recorded specifically for trees, shrubs and saplings. Dung counts Dung counts of large herbivores were done according to and that was used to indicate a measure of large herbivore occupancy and utilization of a habitat. Total counts of old 86
Number of saplings, shrubs and trees
Evidence of human damage on woody plants
dung and fresh dung of large herbivores were recorded in each plot. The sum of old and fresh dung represented an accumulation of large herbivore occupancy and utilization of a study strata. Dung counts have the advantage that they give data, not only on numbers, but also on distribution and differential habitat use (Gandiwa et al., 2011). Trees, shrubs and samplings were counted and recorded based on based on the defined categories, e.g., trees were defined as rooted, woody, self-supporting plants ≥3 m in height with a basal stem diameter ≥0.06m, shrubs were defined as rooted, selfsupporting stemmed plants >1 m, but 3 to 7 km and >7 km. Each study site zone’s households were then randomly sampled as an independent sub-population. Hundred (100) questionnaires were administered to each distance zone, thus, the cumulative total sample size for the present study was 300 household heads.
6.3.3 Data collection Data on opportunity and direct costs of wildlife raided crop and livestock predation was collected during the 2014/16 farming season spanning from October 2014 to October 2016. Specific interview questions addressed the encountered incidences of crop raids and livestock predation on a monthly and seasonal basis and mitigation measures adopted by farmers. Economic value of crop/livestock loss was estimated based on market value. Furthermore, the following data collection tools were used to collect data; 20 key informants were purposively interviewed and these included the locals, namely; 2-Ward Councilors, 2-Traditional leaders, 2-Agriculture Officials, 2-Zimbabwe Republic Police (ZRP) Officers, 2-Parks Senior Wildlife Officers, 1-District Administrator. 2-Save Valley Conservancy Senior Rangers, 1Member of Parliament, 4-Farmer Association Committee members and 2-Shareholders (Operators) in Save Valley Conservancy. Extensive compilation of literature review and consolidation of the data from Save Valley Conservancy and the Zimbabwe Parks and Wildlife Management Authority on local Problem Animal Control reports and other scientific documents/reports were also done as desk review of available literature and information triangulation. Data was collected using a set of questionnires (see Apendix 5).
138
6.3.4 Data Analysis Contingent valuation (CV) analysis was adopted for determination of household economic cost value of crop and livestock loss due to human-wildlife conflicts in Mutema and Musikavanhu communities. Univariate analysis of variance was used to determine if household total economic loss varied with distance from the park boundary of Save Valley Conservancy to the farming communities. Direct costs at net present value were given by making comparison of spatial dimensions of perceived and actual crop and/or livestock losses. These direct costs were derived from area dimensions in respect to crop type grown or livestock reared. At the household base level, the spatial dimensions of agriculture damage were associated with the amount of crop and/or livestock expected to be harvested if not impacted upon by natural hazards as suggested by Bell and McShane-Caluzi (1986) and O’connell-Rodwell et al. (2000). Based on knowledge of local smallholder farmers, perception on magnitude of damage incurred was obtained (Mashapa et al., 2014b; MhuriroMashapa et al., 2017). Determination of costs assumed homogeneity of crops, livestock and market prices in Mutema and Musikavanhu communal areas as there was uniformity in crops/livestock and their associated prices across the study area (Mashapa et al., 2014b; Mhuriro-Mashapa et al., 2017). Moreover, study responses on recorded drivers of humanwildlife conflicts and mitigation measures were summarized, then content and descriptive analyses were performed. Wilcoxon’s (rank sums) test was performed, comparing incidences of human-wildlife conflicts between wet and dry season. Fisher’s exact test analysis was performed to compare percentage damage of all crops and livestock types across the two study stratum. All statistical analyses of data were done using the Statistical Package for Social Sciences (SPSS version 17.0, Chicago, USA.).
139
6.4 Results 6.4.1 Socio-economic impact of human-wildlife conflicts on agro-based livelihoods in Mutema-Musikavanhu communities adjacent to Save Valley Conservancy, southern Zimbabwe In terms of livelihood activities 97.9% (n = 294) of the study respondents households reported that they were agriculture farmers producing crops and livestock for a living. The other 2.1% (n = 6) households stated that they were dependent on wild food, remittances from employed relatives and selling of bush meat amongst others. Across the two study strata over 80% (n = 240) of study respondents reported incidences of crop and/or livestock damage by wild animals (Figure 6.1).
140
Figure 6.1. Human-wildlife conflicts prevalence across Mutema and Musikavanhu communal areas adjacent to Save Valley Conservancy, southern Zimbabwe
Common agriculture enterprises reported to be highly prone to wildlife damage in MutemaMusikavanhu communities are shown in table 6.2. High severity of human-wildlife conflict was recorded in all of the two study strata, mainly driven by selected large herbivores, namely; elephants and buffalo raiding crops like banana (Musa sapientum) and maize (Zea mays), respectively. Whereas, selected carnivores like lion (Panthera leo) and African wild dog (Lycaon pictus) mainly driving livestock predation, killing cattle (Bos taurus) and goats (Capra aegagrus).
141
Table 6.2. Severity of human-wildlife conflict as driven by problem wild animals in MutemaMusikavanhu communities in the periphery of Save Valley Conservancy, southern Zimbabwe Agriculture produce affected by wild animals
Main problem wild animals
Cattle (Bos taurus), Goats (Capra aegagrus),
Lion African wild dogs and Spotted hyenas Elephant and buffalo Elephant Elephant and Warthog Elephant and Warthog Elephant and buffalo
Maize (Zea mays), Bananas (Musa sapientum) Sugar beans (Phaseolus vulgaris), Pumpkins (Cucurbita sp.) Sorghum (Sorghum bicolour), Wheat (Triticum aestivum)
Sugar cane (Saccharum officinarum). Tomatoes (Solanum lycopersicum)
Severity of human-wildlife conflict in Musikavanhu study stratum High Moderate
Comments on severity of human-wildlife in Mutema study stratum High Moderate
High High Moderate
High High High
Low
High
Low
High
Elephant, warthog and hippopotamus (Hippopotamus amphibious) Elephant
Moderate
Moderate
Low
Low
Elephant
Moderate
High
The nature of human-wildlife conflicts reported in the study area is shown in table 6.3. High incidences of both crop raid and livestock predation were reported across the study strata. Records from literature review and 80% (n = 16) of key informants also indicated that the total number of people killed by elephants, buffalos, hippopotamus, and Nile Crocodile (Crocodylus niloticus) per annum was increasing in Mutema-Musikavanhu communal areas where 7 people were reported killed and 16 people wounded by wild animals for the period 2010 to 2016.
142
Table 6.3. Nature of human-wildlife conflicts in Musikavanhu and Mutema communal areas, southern Zimbabwe Nature
of
human-wildlife
Musikavanhu (n = 150)
conflicts across the study area
%
Mutema (n = 150) %
Crop raids by large herbivores
68
43
Livestock predation by carnivores
40
74
People killed by wild animals
4
3
Majority of study respondent households across the study strata (96%, n = 288) had incurred household economic losses. Total household economic losses were significantly different across the distances between the Save Valley Conservancy boundary and human settlement (Table 6.4). The highest households mean economic loss was within a radius of 3 km between Save Valley Conservancy boundary and the human settlement, and the lowest was over 7 km from the Save Valley Conservancy boundary into the human settlement. About 88% (n = 264) of the study respondent households within the 0-3 km radius zone from the Save Valley Conservancy boundary indicated that they had been negatively affected by wild animals, 63% (n = 189) and 53% (n = 159) reported that they were negatively affected by wild animals within the > 3 to 7 km and > 7 km radius zones across the study area, respectively. There was a correlation between the distance from the Save Valley Conservancy boundary and the number of households affected by wild animals (R2=0.986; p 3 to 7 km
703.96 ±113.35a
668.16 ±103.23a
686.06 ± 128.92a
> 7 km
617.80 ±177.97a
724.20 ±17.98a
671.00 ± 182.32a
boundary
Means within the same column with the same superscript were not significantly different at p>0.05. Means within the same column with different superscript were significantly different at p
