1
CARBON DYNAMICS IN CENTRAL AFRICAN FORESTS
MANAGED FOR TIMBER
By
VINCENT DE PAUL MEDJIBE
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
2012
2
© 2012 Vincent de Paul Medjibe
3
To my Father Medjibe Paul and my Mother Deyom Agathe, you passed away during my doctoral program, this dissertation is in your memories.
4
ACKNOWLEDGMENTS
I feel privileged to have had the opportunity to work with the members of my
doctoral committee Jefferson S. “Jeff” Hall, Wendell P. Cropper, Kaoru Kitajima, and
Francis E. “Jack” Putz. I thank you all very much for your patience, infectious
enthusiasm, concern, shared knowledge, and most of all, for making me a better
scientist and researcher in the field of tropical ecology. I thank Jeff Hall for changing my
life by instilling in me a profound appreciation for tropical forestry and for supporting me
intellectually and personally throughout my graduate programs in the United States of
America. Wendell Cropper guided, advised, and helped me to develop and evaluate
simulation models, including the one I used in my dissertation research; his many
suggestions on early drafts of my manuscripts are much appreciated. I am grateful to
Dr. Kaoru Kitajima for her guidance and advice. I also acknowledge Dr. Steve
Humphrey, Dean of the School of Natural Resources and Environment (SNRE) at the
University of Florida, for his help and support throughout my doctoral program. I also
thank Dr. Claudia Romero for her friendship and many kinds of support she provided
throughout my time in Gainesville. I am grateful to the staff of SNRE, the Department of
Biology, the Center for African Studies, and Tropical Conservation and Development for
administrative support.
I also thank all my sponsors without whom my doctoral program would not have
been possible: University of Florida Alumni Fellowship, University of Florida’s School of
Natural Resources and Environment, University of Florida’s Tropical Conservation and
Development, University of Florida’s Center for African Studies, International
Foundation for Science (IFS), WWF-Kathryn Fuller Doctoral Fellowship, WWF-
5
Education for Nature (WWF-EFN), Compton Foundation, National Park Agency of
Gabon (ANPN), and Wildlife Conservation Society-Gabon Program.
I thank SEEF and CEB for permission to work in their concessions. TFF-Gabon,
WCS-Gabon, ANPN-Gabon, CEB, and SEEF all provided critical logistical support for
my fieldwork. I gratefully acknowledge the hard work of field assistants J.C. Mouandza,
Y. Mihindo, J.P. Mondjo, R. Aba’a Nseme, E. Dimoto, L.E. Mapaha, P. Feizoure, A.
Moukala, and A. Moukambi. I also thank K. Piquenot, T. Wanders, A. Moundounga, G.
Tokpa, M. Kombila, J.R. Poulsen, C.J. Clark, L.J.T. White, M.P. Starkey, G. Abitsi, H.R.
Memiaghe, A.A. Ndouna, and R. Faranga Mamadou for their support and assistance.
Of the many people who provided insightful comments on my proposals and
manuscripts, the contributions of my colleagues P. Brando, A. Sheiken, M. Slot, L.
Schreeg, A. Alencar, and G. Celis were particularly valuable and appreciated.
I would like to thank my entire family for their patience during my entire graduate
programs (from Master to Doctorate)—thank you all for putting up with my absence.
Finally, I thank my spouse Nina Flore Mapouka for being so patient and supportive; this
dissertation symbolizes your endurance.
6
TABLE OF CONTENTS page
ACKNOWLEDGMENTS .................................................................................................. 4
LIST OF FIGURES ........................................................................................................ 13
LIST OF ABBREVIATIONS AND ACRONYMS ............................................................. 15
ABSTRACT ................................................................................................................... 17
CHAPTER
1 BACKGROUND, STUDY MOTIVATION, AND OBJECTIVES ................................ 19
1.1 Background ....................................................................................................... 19
1.1.1 Congo Basin Forests ............................................................................... 20 1.1.2 Forest Management in the Congo Basin ................................................. 21
1.1.3 Timber Production in the Congo Basin .................................................... 23 1.1.4 Forest Certification in Central Africa ........................................................ 24
1.1.5 Forest Harvesting and Carbon Stocks ..................................................... 25 1.1.6 REDD+ and the Congo Basin Forests ..................................................... 26
1.2 Motivation ......................................................................................................... 27 1.2.1 Prospects for Improved Forest Management in Central Africa ................ 27
1.2.2 Impacts of Selective Logging in Central Africa ........................................ 28 1.2.3 Logging Damage ..................................................................................... 28
1.2.4 Timber Production and Carbon Emissions in Central Africa .................... 29 1.3 This Dissertation ............................................................................................... 29
1.3.1 Study Problem ......................................................................................... 29 1.3.2 Specific Study Questions ......................................................................... 30
1.3.3 Overall Goal and Outlines of the Dissertation .......................................... 30 1.3.4 Overview of Gabon .................................................................................. 34
1.3.4.1 Climate ........................................................................................... 34 1.3.4.2 Vegetation ...................................................................................... 35
1.3.4.3 Forestry sector ............................................................................... 35 1.3.4.4 Logging operations......................................................................... 36
2 IMPACTS OF SELECTIVE LOGGING ON ABOVE-GROUND FOREST BIOMASS IN THE MONTS DE CRISTAL IN GABON ............................................ 41
2.1 Introduction ....................................................................................................... 41 2.2 Methods ............................................................................................................ 43
2.2.1 Study Site ................................................................................................ 43 2.2.2 Plot-Based Measurements ...................................................................... 45
2.2.3 Damage Assessment .............................................................................. 45 2.2.4 Conversion of Above-Ground Biomass into Necromass .......................... 46
2.2.5 Data Analyses ......................................................................................... 46
7
2.2.5.1 Tree species richness .................................................................... 46 2.2.5.2 Above-ground biomass estimates .................................................. 47
2.2.5.3 Logging damage ............................................................................ 47 2.3 Results .............................................................................................................. 48
2.3.1 Trees Diameter Class Distribution ........................................................... 48 2.3.2 Tree Species Richness ............................................................................ 48
2.3.3 Above-Ground Biomass Prior to Logging ................................................ 49 2.3.4 Logging Characteristics ........................................................................... 49
2.3.5 Logging Damage ..................................................................................... 49 2.3.6 Above-Ground Biomass Converted into Necromass ............................... 51
2.4 Discussion and Concluding Remarks ............................................................... 52 2.4.1 Tree Species Richness ............................................................................ 52
2.4.2 Above-Ground Biomass .......................................................................... 52 2.4.3 Logging Damage ..................................................................................... 54
2.4.3.1 Damage due to felling .................................................................... 54 2.4.3.2 Damage due to skidding ................................................................ 55
2.4.4 Management Implications ........................................................................ 56
3 A FOREST STEWARDSHIP COUNCIL CERTIFIED LOGGING CONCESSION COMPARED WITH AN UNCERTIFIED CONCESSION IN GABON ON THE BASIS OF TREE SPECIES RICHNESS AND COMPOSITION, STAND AND SOIL DAMAGE, AND ABOVE-GROUND BIOMASS .............................................. 65
3.1 Overview ........................................................................................................... 65
3.2 Methods ............................................................................................................ 67 3.2.1 Study Site ................................................................................................ 67
3.2.1.1 FSC-Certified site (CEB-FSC) ........................................................ 67 3.2.1.2 Conventionally logged site (SEEF-CL) ........................................... 69
3.2.2 Tree Measurements ................................................................................ 70 3.2.3 Above-Ground Biomass Estimates .......................................................... 71
3.2.4 Damage Assessment .............................................................................. 71 3.2.5 Logging-Induced Conversion of Above-Ground Biomass into
Necromass .................................................................................................... 73 3.2.6 Data Analysis .......................................................................................... 73
3.2.6.1 Tree species richness .................................................................... 73 3.2.6.2 Statistical analyses and modes of data presentation ..................... 74
3.3 Results .............................................................................................................. 75 3.3.1 Forest Structure and Tree Species Richness .......................................... 75
3.3.1.1 Forest structure .............................................................................. 75 3.3.1.2 Changes in tree species richness and composition ....................... 75
3.3.2 Above-Ground Biomass .......................................................................... 77 3.3.3 Logging Intensities................................................................................... 77
3.3.4 Logging Damage ..................................................................................... 78 3.3.4.1 Felling damage .............................................................................. 79
3.3.4.2 Skidding damage ........................................................................... 79 3.3.4.3 Felling gaps and ground area disturbed ......................................... 80
3.3.4.4 Above-ground biomass converted into necromass ........................ 80
8
3.3.4.5 Damage due to road construction and maintenance ...................... 81 3.4 Discussion and Conclusion ............................................................................... 81
3.4.1 Impacts of Certified or Conventional Logging on Forest Structure and Composition .................................................................................................. 81
3.4.2 Above-Ground Biomass Impacts and Conversion into Necromass ......... 82 3.4.3 Logging Damage ..................................................................................... 83
3.4.3.1 Damage due to felling .................................................................... 83 3.4.3.2 Damage due to skidding ................................................................ 83
3.4.3.3 Damage due to road construction and maintenance ...................... 84 3.4.4 Impacts of Forest Management Certification on Biodiversity and
Carbon Retention .......................................................................................... 85
4 COST COMPARISONS OF REDUCED-IMPACT AND CONVENTIONAL LOGGING IN THE TROPICS ............................................................................... 102
4.1 Background Information .................................................................................. 102
4.2 Methods .......................................................................................................... 105 4.2.1 Descriptions of the Case Studies........................................................... 105
4.2.1.1 Fazenda Cauaxi, Paragominas, Brazil ......................................... 105 4.2.1.2 Pibiri, Guyana .............................................................................. 106
4.2.1.3 Fazenda Agrosete, Brazil ............................................................. 107 4.2.1.4 Sabah, Malaysia........................................................................... 108
4.2.1.5 Malinau, East Kalimantan, Indonesia ........................................... 109 4.2.1.6 Belém, Brazil ................................................................................ 110
4.2.1.7 Sarawak, Malaysia ....................................................................... 110 4.2.1.8 Dungun, Terengganu, Malaysia ................................................... 111
4.2.1.9 Tapajós National Forest, Brazil .................................................... 111 4.2.1.10 Terengganu, Malaysia ................................................................ 111
4.2.1.11 Kelantan, Malaysia ..................................................................... 112 4.2.2 Study Site: Monts de Cristal, Gabon...................................................... 113
4.2.3 Logging Costs ....................................................................................... 113 4.2.3.1 Labor costs .................................................................................. 114
4.2.3.2 Costs of equipment and materials ................................................ 114 4.2.3.3 Training costs in the tropics ......................................................... 115
4.2.3.4 Analysis of logging profits ............................................................ 115 4.3 Results ............................................................................................................ 116
4.3.1 Costs of Logging on Monts de Cristal, Gabon ....................................... 116 4.3.1.1 Pre-harvest operations ................................................................. 116
4.3.1.2 Harvest operation ......................................................................... 117 4.3.2 Costs and Profits of Logging in the Tropics ........................................... 118
4.3.2.1 Cost of logging ............................................................................. 118 4.3.2.1.1 Pre-harvest activities ................................................................. 119
4.3.2.1.2 Harvest planning ....................................................................... 120 4.3.2.1.3 Infrastructure costs.................................................................... 121
4.3.2.1.4 Harvest operations .................................................................... 121 4.3.3.2 Profits for CL and RIL ................................................................... 122
4.4 Concluding Remarks ....................................................................................... 123
9
5 POST-HARVEST STAND RECOVERY OF TREE ABUNDANCE AND ABOVE-GROUND BIOMASS AFTER SELECTIVE LOGGING IN GABON ....................... 137
5.1 Introductory Remarks ...................................................................................... 137 5.2 Methods .......................................................................................................... 139
5.2.1 Study Site .............................................................................................. 139 5.2.2 Tree Growth .......................................................................................... 139
5.2.3 Tree Mortality and Recruitment ............................................................. 140 5.2.4 Functional Groups ................................................................................. 140
5.2.5 Model Structure ..................................................................................... 141 5.2.6 Transition Probabilities .......................................................................... 142
5.2.7 Model Runs ........................................................................................... 142 5.2.8 Sensitivity Analyses ............................................................................... 143
5.3 Results ............................................................................................................ 143 5.3.1 Tree Abundance .................................................................................... 143
5.3.2 Tree Growth and Recruitment ............................................................... 144 5.3.3 Tree Mortality ........................................................................................ 144
5.3.4 Above-Ground Biomass Increments ...................................................... 145 5.3.5 Above-Ground Biomass Loss ................................................................ 146
5.3.6 Evaluation of the Simulation Model ....................................................... 146 5.3.6.1 Simulated tree abundance ........................................................... 146
5.3.6.2 Simulated Above-Ground Biomass .............................................. 148 5.3.7 Variability in Tree Abundance and Above-Ground Biomass .................. 148
5.3.7.1 Tree abundance and above-ground biomass ............................... 148 5.3.7.2 Tree Mortality ............................................................................... 149
5.4 Discussion ...................................................................................................... 150 5.4.1 Tree Abundance .................................................................................... 150
5.4.2 Above-Ground Biomass ........................................................................ 151 5.5 Conclusion ...................................................................................................... 152
APPENDIX
A CHARACTERISTICS OF TREE SPECIES IN THE 10 1-HA PERMANENT PLOTS IN THE RIL ZONE OF THE SEEF CONCESSION ON MONTS DE CRISTAL, GABON ................................................................................................ 167
B CHARACTERISTICS OF TREE FAMILIES IN THE 10 1-HA PERMANENT PLOTS IN THE RIL ZONE OF THE SEEF CONCESSION ON THE MONTS DE CRISTAL, GABON ................................................................................................ 177
C CHARACTERISTICS OF TREE SPECIES IN THE 20 1-HA PERMANENT PLOTS IN THE CEB-FSC CONCESSION, GABON ............................................. 179
D CHARACTERISTICS OF TREE FAMILIES IN THE 20 1-HA PERMANENT PLOTS IN THE CEB-FSC CONCESSION, GABON ............................................. 193
10
E CHARACTERISTICS OF TREE SPECIES IN THE 12 1-HA PERMANENT PLOTS IN THE SEEF-CL CONCESSION, GABON ............................................. 195
F CHARACTERISTICS OF TREE FAMILIES IN THE 12 1-HA PERMANENT PLOTS IN THE SEEF-CL CONCESSION, GABON ............................................. 205
G LOGGING ACTIVITIES REPORTEDLY CONDUCTED UNDER CONVENTIONAL (CL) AND REDUCED-IMPACT LOGGING (RIL) SYSTEMS. .. 207
H TRANSITION PROBABILITIES AND OTHER PARAMETERS USED IN THE MATRIX MODEL................................................................................................... 209
I CHANGES IN NUMBER OF TREES PER STEMS DIAMETER CLASS IN 10 1-HA PERMANENT PLOTS FOR EACH SPECIES FUNCTIONAL GROUP TWO AFTER REDUCED-IMPACT LOGGING ON MONTS DE CRISTAL IN GABON. . 212
J CHANGES IN ABOVE-GROUND BIOMASS PER STEMS DIAMETER CLASS IN 10 1-HA PERMANENT PLOTS FOR EACH SPECIES FUNCTIONAL GROUP TWO YEARS AFTER REDUCED-IMPACT LOGGING ON MONTS DE CRISTAL IN GABON. ........................................................................................... 214
LIST OF REFERENCES ............................................................................................. 216
BIOGRAPHICAL SKETCH .......................................................................................... 234
11
LIST OF TABLES Table page 1-1 Overview of forests by country in the Central Africa Region ............................... 38
1-2 Natural forest management in Central Africa ...................................................... 38
1-3 Distribution of forest certification in Gabon ......................................................... 38
2-1 Characteristics of the ten 1-ha plots before logging on Monts de Cristal, Gabon. ................................................................................................................ 58
2-2 Timber trees harvested in the 104-ha RIL area on Monts de Cristal, Gabon ...... 59
2-3 Number of damaged tree per 1-ha plot and associated above-ground biomass losses on the RIL study area on Monts de Cristal, Gabon. .................. 60
2-4 Number and surface area of skid trails in the 50 ha RIL study area on Monts de Cristal, Gabon. ............................................................................................... 60
2-5 Number of damaged trees ≥10 cm dbh and associated AGB loss along 3450 m of skid trails in 50 ha and felling damage in 5 1-ha plots in the study area on Monts de Cristal, Gabon. ............................................................................... 61
2-6 Above-ground biomass loss due to selective logging in the tropics .................... 61
3-1 Biophysical conditions and pre-logging tree community characteristics in the CEB-FSC and SEEF-CL logging concessions in Gabon. ................................... 86
3-3 Number and volume of timber trees harvested in the 508 ha CEB-FSC and 200 ha SEEF-CL study areas in Gabon. ............................................................ 89
3-4 Logging impacts in the entire 508-ha CEB-FSC and 200-ha SEEF-CL study areas in Gabon. .................................................................................................. 90
3-6 Ground area disturbed and logging waste in the 508-ha CEB-FSC and 200-ha SEEF-CL study areas in Gabon.. .................................................................. 93
3-7 Roads in 508 ha in the CEB-FSC concession and 200 ha in the SEEF-CL (200 ha) concession in southeastern Gabon compared with FAO standards. .... 94
4-1 Comparison of reduced-impact (RIL) and conventional logging (CL) logging intensities and harvesting costs per cubic meter in the tropics based on published data.. ................................................................................................ 128
4-2 Harvesting costs per cubic meter of timber extracted and per logged hectare divided into pre-harvest activities, harvest planning, infrastructure construction, and harvest operations.. .............................................................. 130
12
4-3 Costs of conventional (CL) and reduced-impact logging (RIL) on Monts de Cristal, Gabon. .................................................................................................. 132
4-4 Profits from conventional logging (CL) and reduced-impact logging (RIL) in the tropics based on values published in the case studies ............................... 133
5-1 Transition probabilities and other parameters used in the matrix model. .......... 153
5-2 Changes in number of trees in the 10 1-ha permanent plots two years after reduced-impact logging on Monts de Cristal in Gabon. .................................... 154
5-3 Changes in number of trees per stems diameter class in the 10 1-ha permanent plots two years after reduced-impact logging on Monts de Cristal in Gabon. .......................................................................................................... 155
5-4 Changes in above-ground biomass in the 10 1-ha permanent plots two years after reduced-impact logging on Monts de Cristal in Gabon. ............................ 156
5-5 Changes in above-ground biomass per stems diameter class in 10 1-ha permanent plots two years after reduced-impact logging on Monts de Cristal in Gabon. .......................................................................................................... 157
13
LIST OF FIGURES
Figure page 1-1 Land affectation in the Congo Basin... ................................................................ 39
1-2 Location of the study sites in logging concessions in Gabon. ............................. 40
2-1 Stem diameter class distribution in the 10 1-ha plots on Monts de Cristal. .. ...... 62
2-2 Percentage of damaged trees per damage category from felling and skidding during selective logging on Monts de Cristal, Gabon .......................................... 63
2-3 Relationships between stem diameter (dbh) of harvested trees on Monts de Cristal, Gabon. ................................................................................................... 64
3-1 Locations of study sites in the CEB-FSC and SEEF-CL logging concessions in Gabon. ............................................................................................................ 95
3-1A Study area and plot locations in the CEB-FSC logging concession. ................... 95
3-1B Study area and plot locations in the SEEF-CL logging concession. ................... 96
3-2 Pre-logging stem densities by diameter (dbh) class in the twenty 1-ha plots in the CEB-FSC and twelve 1-ha plots in the SEEF-CL concessions. .................... 96
3-3 Location of plots based on the number of trees per species before and after logging in the CEB-FSC and SEEF-CL study sites... .......................................... 97
3-4 Percentage of trees by damage category resulting from felling and skidding in the foresst concessions in Gabon.. ................................................................. 98
3-5 Relationships between harvested tree stem diameter (dbh) in the CEB-FSC (508 ha) and SEEF-CL (200 ha) concessions in Gabon. ................................... 99
3-6 Maps of skid trails and felled trees in the CEB-FSC and SEEF-CL concessions at two scales. ............................................................................... 100
3-7 Logging roads in the CEB-FSC and SEEF-CL concessions. ............................ 101
4-1 Costs of RIL compared to CL in relation to harvest volume per hectare. .......... 134
4-2 Relative costs of RIL and CL in relation to the relative timber volume harvested per hectare. ...................................................................................... 135
4-3 Relative profits of RIL and CL. .......................................................................... 136
5-1 Variability in stems density per dbh class in 2009 and 2011. . ......................... 158
14
5-2 Post-logging stem diameter and above-ground biomass growth . .................... 159
5-3 Variability in above-ground biomass per dbh class in 2009 and 2011.. ............ 160
5-4 Post-logging tree abundance over timer using different logging intensities. ..... 161
5-5 Post-logging tree abundance over time using the observed logging intensity and the government mandated minimum cutting cycle per species functional group. ............................................................................................................... 162
5-6 Post-logging above-ground biomass growth over timer using different logging intensities.......................................................................................................... 163
5-7 Post-logging above-ground biomass over time using the observed logging intensity and the government mandated minimum cutting cycle per species functional group. ............................................................................................... 164
5-8 Variability of tree density associated with hypethetical changes 5% in post-harvest tree growth and mortality using observed logging intensity. ................. 164
5-9 Variability of above-ground biomass associated with hypothetical changes by 5% in post-harvest tree growth and mortality using observed logging intensity. ........................................................................................................... 165
5-10 Trends of the post-harvest tree mortality... ....................................................... 166
15
LIST OF ABBREVIATIONS AND ACRONYMS
AFD Agence Française de Developpement
AGB Above-Ground Biomass
ANCOVA Analysis of Covariance
ANOVA Analysis of Variance
ANPN Agence Nationale des Parcs Nationaux du Gabon
ATIBT Association Technique Internationale des Bois Tropicaux
ATO African Timber Organization
CAR Central African Republic
CARPE Central Africa Regional Program for Environment
CEB Compagnie Equatoriale de Bois
CIFOR Center for International Forestry Research
CL Conventional Logging
CCA Canonical Correspondence Analysis
COMIFAC Central African Forest Commission
COP Conference of the Parties
DBH Diameter at Breast Height
DTL Demerara Timber Ltd
EIA Environmental Impacts Assessment
FAO United Nations Food and Agriculture Organization
FCPF Forest Carbon Partnership Facility
FFT Fundação Floresta Tropical
FORM International International Forestry Consultancy
FSC Forest Stewardship Council
IFS International Foundation for Science
16
IISD International Institute for Sustainable Development
ITTO International Tropical Timber Organization
MCI Malaysian Criteria, Indicators, Activities, and Management Specification
OLB Origine Légale du Bois
PARPAF Projet d’Appui à la Réalisation des Plans d’Aménagement Forestier
REDD+ Reduced Emissions from Deforestation and forest Degradation, conservation and enhancing forest carbon stocks
RIL Reduced-Impact Logging
ROC Republic of Congo
R-PIN REDD Proposal Idea Notes
SEEF Société Equatoriale d’Exploitation Forestière
SFM Sustainable Forest Management
TEREA Terre Environnement Aménagement
TFF Tropical Forest Foundation
USAID United States Agency for International Development
WWF World Wildlife Fund for Nature
17
Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
CARBON DYNAMICS IN CENTRAL AFRICAN FORESTS MANAGED FOR TIMBER
By
Vincent de Paul Medjibe
May 2012
Chair: Francis E. Putz Major: Interdisciplinary Ecology
As globally valued storehouses of biodiversity and carbon, the Congo Basin
forests are threatened by human activities. Given the importance of extractive industries
and the role of forests in climate change mitigation and maintenance of biodiversity, it is
crucial to investigate the extent to which improvements in forest management practices
can serve to reduce carbon emissions, maintain biodiversity, and increase future timber
yields.
I assessed the impacts of logging in three different sites in two forest concessions
in Gabon: a forest logged with reduced-impact logging (RIL) methods conducted by
Tropical Forest Foundation (TFF-RIL); a Forest Stewardship Council-certified forest
concession (FSC); and, a conventionally logged forest concession (CL). Logging
impacts were evaluated in 42 randomly located 1-ha permanent plots in which all trees
≥ 10 cm dbh were measured, identified to species, and marked prior to harvesting.
Impact comparisons were made between the FSC and CL sites. After logging, tree
damage and biomass losses were recorded as being due to felling, skidding, and road-
related construction and maintenance. Given the financial importance of logging, I
18
assessed the costs and benefits of RIL and CL using new data from Gabon and
published studies from elsewhere in the tropics.
Logging intensities were light and differed among the three sites, but no short-term
effects of logging were detected on tree species richness. At the TFF-RIL site, the initial
biomass declined by 4.1% with committed emissions of 17 Mg ha-1. At the FSC site, the
biomass declined by 2.9% with committed emissions of 11.2 Mg ha-1. At the CL site,
biomass declined by 6.3% with committed emissions of 24.6 Mg ha-1. Based on a matrix
model and with the observed logging intensity at the TFF-RIL site, stand volume is
predicted to recover about 21 years after logging. Post-harvest tree abundance and
biomass increment were sensitive to tree mortality. A meta-analysis of logging costs
revealed substantial variation but RIL was more costly to implement than CL when
these costs were calculated either per timber volume harvested or per hectare logged.
In conclusion, selective logging causes carbon emissions that can be reduced
through use of improved forest management.
19
CHAPTER 1 BACKGROUND, STUDY MOTIVATION, AND OBJECTIVES
1.1 Background
Tropical forests cover 7-10% of the global land area, store 40-50% of the carbon in
terrestrial vegetation (FAO, 2010), and provide a variety of important economic,
environmental, and socio-cultural benefits to people across the globe (e.g., Canadell
and Raupach, 2008; FAO, 2011). Forests of the Congo Basin, the second large tropical
forest biome after the Amazon Basin have high conservation value and are crucial for
national development (i.e., source of revenue for the central African countries through
timber production) and the livelihoods of millions of people (de Wasseige et al., 2009).
These forests continue also to absorb and store huge quantities of carbon (Lewis et al.,
2009) and support over one thousand plant species and populations of endangered
animal.
The production forests that are so critical to both conservation and development
are often degraded due to poor logging practices (see Putz et al., 2000) that continue,
and constitute a fundamental threat to these forests (Malcolm and Ray, 2000; Hall et al.,
2003). Unplanned logging by untrained and poorly supervised crews also results in
carbon emissions to the atmosphere that otherwise are avoidable, reducing future
timber yields, increasing the risk of forest fires, and decreasing future economic values
of the forest (Putz and Nasi, 2009). Such poor logging is underway in many places in
the tropics, but less is known about the consequences of these practices in Africa than
elsewhere. With global attention on the carbon stored in African forests and efforts to
reduce carbon emission from forest due to logging (e.g., Baccini et al., 2008; Lewis et
20
al., 2009; Houghton, 2009), there is increased interest in avoiding degradation of forests
and promoting improved forest management in the Congo Basin region.
1.1.1 Congo Basin Forests
Congo Basin forests, which cover 529 million ha, expand through the Central
Africa Region (Figure 1-1; Table 1-1; FAO, 2011). They are the most “intact” remaining
tropical forest worldwide and provide shelter and food to thousands of people who
depend on them for daily subsistence (de Wasseige et al., 2009; FAO, 2010). The
forests expand across Cameroon around the Gulf of Guinea through Gabon and then
along the central Congo River basin (Richards, 1996).
These forests are generally seasonal and wet with precipitation greater than 2 000
mm per year with the highest precipitation occurring along the Atlantic coast (e.g., West
coast of Gabon received more than 3 000 mm of rain per year) followed by the central
basin and areas to the East (Ashton and Hall, 2011). Soils of the Congo Basin forest are
predominantly oxisols and are developed on a Pre-cambrian continental shield that
comprises sedimentary shale and sandstone (Richards, 1996). Fertile soils are found in
volcanic areas in Cameroon and on the floodplains of large rivers (see Ashton and Hall,
2011).
Congo Basin forests can be divided by on the basis of moisture regimes into
coastal belt and the core wet evergreen region of the upper Congo Basin
Gilbertiodendron dewevrei (Caesalpiniaceae) forms monodominant stands in some
places (Hart, 2001), but the forest is mostly mixed with a variety of timber species (e.g.,
Entandrophragma spp, Millitia excelsa, Pterocarpus soyauxii, Triplochyton excleroxylon,
Aukoumea klaineana). The semi-evergreen forests of the northern Congo Basin are
characterized by scattered Meliaceae (Hall, 2002) and Burseraceae, among which
21
Aucoumea klaineana Pierre has sustainedtimber industries in Gabon for years (Doucet,
2003; White and Abernathy, 2007).
These forests provide critical ecosystem goods and services, house a large variety
of plants and animals species, and store large quantities of carbon. They also constitute
a major source of revenue from timber extraction (Table 1-1; de Wasseige et al., 2009)
and employ many people in timber extraction, biodiversity conservation, and
ecotourism. Finally, these forests are important for research and education on forest
resource management and biodiversity conservation. In the Congo Basin, only two
major land uses are officially recognized: logging concessions and protected areas
(Figure 1-1). Community forests are rare except in Cameroon (Karsenty and Assembe,
2011).
1.1.2 Forest Management in the Congo Basin
Congo Basin forests have been the subject of some research on forest
management. Early studies pointed out deficiencies in natural regeneration of the main
harvested timber species and recommended growing these species in plantations or
increasing their abundance in managed natural forest through enrichment planting
(MacGregor, 1934; Lancaster, 1960; Nwoboshi, 1987). The method implemented was
based on the modified selection system employed in Ghana that resembles the group
shelterwood system described by Nwoboshi (1987). The recommended method is
similar to the liberation thinning technique described by Dupuy et al. (1993) and Petrucci
and Tandeau de Marsac (1994) that was applied in Central African Republic (CAR).
Other studies conducted in the region by CIRAD-Forêt under the FORAFRI Project
(Petrucci and Tandeau de Marsac, 1994; Bedel et al., 1998) also focused on the
dynamics of the main harvested timber species whereas more recent research focuses
22
on the abiotic and biotic influences on natural regeneration and growth rates. For
example, Hall et al. (2003) studied four species of Entandrophragma (African
mahogany) and suggested that the growth and recruitment of these species in natural
forest is favored by the creation of small to medium-sized canopy gaps. Similarly,
Makana and Thomas (2003 and 2005) reported high seedlings survival and growth of
Khaya anthotheca and Entandrophragma spp. (Meliaceae) in gaps in the DRC. Doucet
(2003) also reported that A. klaineana, the most important timber species in Gabon,
performs better under high light conditions. In keeping with these findings, Hall (2008)
recommended enrichment planting of seedlings of commercial timber species in forest
gaps. Medjibe et al. (in progress) reported that seedlings of seven of the timber species
in the ROC survived and recruited better under low light conditions and small canopy
gaps.
Few studies have been conducted on how logging operations in Central Africa
affect natural regeneration or other aspects of forest management. In particular, the
effects of logging on forest structure and carbon stocks are under-studied. Although the
surge of forest certification (Table 1-2) gives hope for sustainable forest management in
the Congo Basin, little is known about the extent to which switching from conventional
logging (CL) to reduced-impact logging (RIL) promotes sustainable timber production
and preserves forest ecosystem integrity. Overall, implementation of improved forest
management practices in Central Africa still faces major challenges (see Ruiz Perez et
al., 2005; Cerutti et al., 2008). Recent revisions of forestry laws in the region may help
to promote these practices. For example, the forestry code in CAR was first developed
in 1990 then revised in 2008, in Cameroon it was developed in 1994 and is currently
23
under revision, and in Congo, Gabon and DRC the codes were developed in 2000,
2001, and 2002, respectively (Nasi et al., 2012). Although forest management plans
developed on the bais of these codes share a common set of activities, there are still
differences among concessions and countries (ATIBT, 2007; Bayol and Borie, 2004).
1.1.3 Timber Production in the Congo Basin
Timber production is of substantial long-term socio-economic importance in the
Congo Basin (Minnemeyer et al., 2002; Karsenty and Gourlet-Fleury, 2006; Karsenty et
al., 2008, de Wasseige et al., 2009). Starting after World War II and accelerating in the
early 1990s, African tropical timbers have drawn substantial attention from international
markets (Ashton and Hall, 2011). Logging in the region can be distinguished by three
periods: the pioneer phase (1945-1970) when logging was limited to coastal regions;
expansion into the hinterlands (1970-1990) under a pure timber production standard but
with increased interest in silviculture research; and, the last two decades characterized
by the opening of the forestry sector and the importance of forest management under
sustainable development (Nasi et al., 2006; de Wasseige et al., 2009).
In Central Africa, Entandrophragma spp. and A. klaineana supply most of the
highly valued timber. These species have important medicinal, food, and social uses for
local people. For example Entandrophragma cylindricum hosts a caterpillar species that
is rich in protein and a delicacy for local people, A. klaineana produces resin used to
light fires and as candle. Exploitation of the timbers of these species will deplete
populations, which could have negative consequences for local people. Clearly, timber
harvests can be sustained only if forests are harvested without excessive damage to the
remaining stand and tree growth is promoted with silvicultural treatments (e.g., liberation
thinning). The practice of silviculture in rainforests is to insure sustained production of
24
timber. But even with careful management, sustaining timber yields over multiple cutting
cycles is difficult to achieve because government-specified minimum cutting cycles in
Central Africa average only 25-30 years (25 years in Gabon) and average logging
intensities are 1-4 trees per hectare (Durrieu de Madron et al., 2000). These conditions
make it important to understand the dynamics of Congo Basin forests, especially in
regards to the most harvested timber species.
In Central Africa, logging industries contribute up to 7% of national economics
(Table 1-1) and are the second largest employer after the government. About 41% of
the region’s forests are allocated to timber production (Laporte et al., 2004), with an
average annual timber production of 2.7-3.3x106 m3. For example, of the forests of
Gabon, which cover 85% (22x106 ha) of the nation, about 59% (13x106 ha) are
considered production forests (FCPF R-PIN Gabon, 2008) with logging roads
constituting >60% of the length of all roads (Laporte et al., 2007). Because of the
importance of timber production in the region, implementation of improved forest
management is imperative. Despite this recognized need, logging practices in the
region still range from conventional logging practices to reduced-impact logging with
and without forest certification.
1.1.4 Forest Certification in Central Africa
In Central Africa, some of the timber companies that have their forest
management plans approved and are working towards forest certification. The region
has witnessed a recent surge in forest certification by the Forest Stewardship Council
(FSC), the Origine Légale du Bois (OLB), and the Pan African Forest Certification
(PAFC; Table 1-2). Forest certification has demonstrated considerable progress in the
region from no certified forests in 1995 to about 4.8 million hectares certified in 2010
25
(Nasi et al., 2012). Overall, 15% of the Congo Basin forests are in logging concessions,
of which 13% of the total area isFSC certified. A total 293 concessions were officially
recognized in 2010 with 90 operating under approved management plans and 95 in the
process of developing their plans (Nasi et al., 2012). In most of the countries in the
region, forestry administrations face some challenges to develop and monitor forest
management plans due in part to a lack of technical capacity. Forest management plans
are typically developed by independent consulting groups (for example PARPAF; Projet
d‘Appui à la Réalisation des Plans d’Aménagement Forestier in CAR; TEREA; Terre
Environnement Aménagement in Gabon) hired by logging companies; the consultants
work closely with governments, scientific institutions, non-governmental organizations,
and local communities (Nasi et al., 2006). Theforest management guidelines developed
for the region foster the transition from conventional to planned logging.
In 2003, the African Timber Organization (ATO) edited its principles, criteria and
indicators of sustainable forest management to include specific reference to certification
(ATO&ITTO, 2003). The Association Technique Internationale des Bois Tropicaux
(ATIBT) published a set of practical guidelines for concessioners and forest managers
of Central Africa (ATIBT, 2007). Also in 2003, FAO and Tropical Forest Foundation
(TFF) published reduced-impact logging guidelines for the Central African forest (FAO,
2003). These documents indicate that progress is being made towards a sustainable
management of the Congo Basin forests despites the persistence of many challenges
(Ezzine de Blas and Ruiz Perez, 2008).
1.1.5 Forest Harvesting and Carbon Stocks
Forests in Central Africa are degraded or lost as a direct or indirect consequence
of logging, road construction, mining, subsistence agriculture, agro-industries, and fires
26
(Laporte et al., 2007, de Wasseige et al., 2009; FAO, 2011). Emissions from land use-
change in Africa were estimated to be 240 Tg C y−1, of which 11% was from industrial
wood harvesting (Canadell et al., 2009). In 2010 the African share of the global
emissions from land-use change was 17% (Canadell et al., 2009). Despite the
importance of timber industries in the Congo Basin, forest carbon stocks are
jeopardized by logging especially where it is carried out by untrained and unsupervised
laborers working without the aid of adequate management plans (Putz et al., 2008a;
Angelsen et al., 2009).
1.1.6 REDD+ and the Congo Basin Forests
Many efforts to reduce carbon emissions from deforestation and forest
degradation (REDD+) and other REDD-readiness activities are underway in Central
Africa. To reduce emissions and to promote economic development, the Central African
Forests Commission (COMIFAC) aims to provide orientation and coordination of efforts
to promote conservation and sustainable management of forests (IISD, 2006).
COMIFAC countries (Cameroon, Central African Republic, Congo, Democratic Republic
of Congo, Equatorial Guinea, and Gabon) also proposed the use of an adjustment
development factor (Karsenty, 2008; Parker et al., 2009) so that development is not
unduly curtailed in the interest of REDD+. Since the Montreal Conference of the Parties
(COP-11) in 2005, COMIFAC countries have chosen to work together to develop
concerted and common position and to make their presence felt in climate negotiations.
They requested the explicit inclusion of conservation and sustainable forest
management in REDD and underscored their requirements to strengthen their technical
capacity for monitoring forest cover and carbon stocks (Tadoum et al., 2010).
Fortunately, both REDD and development goals are served by adoption of improved
27
forest management practices (Cerutti et al., 2008; Ezzine de Blas et al., 2008). The
implementation of REDD+ mechanism is facing some challenges in the region since
most of the land is owned by the States (i.e., national governments; Karsenty and
Assembe, 2011). Fair implementation of REDD+ requires the adaptation or
transformation of the land tenure systems to prevent the increase of the value of forests
while local people lack real land rights and risk dispossession in favor of carbon
investors (Cotula and Mayers, 2009). In contrast, REDD+ can provide a lever to make
the needed changes in land tenure, but some countries may be reluctant to undertake
land-tenure reforms in the face of vested interests and concerns of potential conflicts.
Social movements and civil society are divided over REDD+, seeing it either as a threat
to the rights of the communities, or as an opportunity to fight against poverty (Karsenty
and Assembe, 2011). Despite these difficulties, some countries are in the phase of
developing their REDD-Project Idea Notes (R-PIN) while others are working on R-
Project Proposals (R-PP). It is still unclear whether REDD+ will be successful in Central
Africa.
1.2 Motivation
1.2.1 Prospects for Improved Forest Management in Central Africa
Employment of improved forest management techniques such as reduced-impact
logging (RIL) elsewhere in the tropics is known to result in substantial reductions in
carbon emissions and increases in biodiversity retention (see van Rheenen et al., 2004;
van Kuijk et al., 2009). Several studies have shown that RIL generates competitive or
superior financial returns compared to conventional logging (CL) due to gains in harvest
efficiency (Barreto et al., 1998; Boltz et al., 2001; Holmes et al., 2002; Applegate et al.,
2004; Houghton et al., 2009). Unfortunately, results of cost comparisons of RIL and CL
28
conducted elsewhere cannot confidently be generalized to Central Africa due to the
heterogeneity among tropical forests (Ghazoul and Sheil, 2010). It is clear that
implementation of improved forest management technique requires up-front investment
for training in logging operations and compliance with best management practice
guidelines.
1.2.2 Impacts of Selective Logging in Central Africa
Due to high tree species diversity in Central Africa, logging is and will remain
selective. Even where RIL practices are employed, selective logging unavoidably
causes carbon emissions, destroys some wildlife habitat, and has a variety of direct and
indirect impacts on biodiversity, other ecosystem values, and human welfare. Where
good timber management practices are implemented, these impacts are substantially
reduced while still providing financial benefits to loggers and landowners, be they
private individuals, communities, or governments. Furthermore, given that logging is
among the world's most dangerous professions (ILO, 1990; Putz and Nasi, 2009), the
safety benefits of RIL training of forest workers should not be disregarded.
1.2.3 Logging Damage
Logging activities always damage forests, but the extent of this damage varies
with the intensity and methods of harvesting (Table 1-3). This conclusion is supported
by studies conducted in Cameroon (Forni et al., 1994; Durrieu de Madron et al., 1998;
Jonkers et al., 2000), Central Africa Republic (Durrieu de Madron et al., 2000), Republic
of Congo (Brown et al., 2005), Gabon (White, 1994; Medjibe et al., 2011), and a meta-
analysis for the Central African region (Durrieu de Madron et al., 2011; Picard et al.,
2012). What is apparently missing are comparative studies on the carbon impacts of CL
29
and RIL operations in Central Africa based on field measurements rather than
secondary data (e.g., Durrieu de Madron et al. , 2011).
1.2.4 Timber Production and Carbon Emissions in Central Africa
When properly conducted, timber harvesting can be a component of sustainable
forest management (SFM; van Kuijk et al., 2009). RIL is also the main component of
SFM shown to have benefits for carbon retention (Pinard and Putz, 1996) and
biodiversity conservation (Pearce et al., 2003; see van Rheenen et al., 2004). The rarity
of RIL in the tropics in general and most likely Central Africa as well is apparently due to
the perceptions of loggers and concessionaires that it is too expensive, that there is
nothing wrong with CL, and that markets do not demand that RIL practices be
implemented (Putz et al., 2000; Applegate et al., 2004). Nevertheless, given the
potential carbon, biodiversity, and future timber yield benefits of RIL, Central African
countries where timber production is of great importance need to promote its adoption.
1.3 This Dissertation
This dissertation focuses on the impacts of different forest harvesting techniques
on forest structure and composition, forest carbon dynamics, and tree species diversity
and composition in Central Africa.
1.3.1 Study Problem
In this dissertation I address the question of: “To what extent do improved forest
management practices in Central Africa reduce carbon emissions, preserve tree
species diversity, and increase future timber yields compared to traditional conventional
logging?” I therefore compare different logging practices (i.e., RIL and CL) in regards to
reduction and recovery of forest carbon and timber stocks in Gabon. I also investigate
30
the impacts of these practices on tree species diversity and on financial harvesting
costs.
1.3.2 Specific Study Questions
The study investigated:
• How do the impacts of improved forest harvesting techniques (i.e., RIL) compare
with conventional logging (CL) in regards to carbon stocks?
• How does the switch from CL to RIL affect the maintenance of tree species
diversity and above-ground biomass?
• What are the differences in financial costs and logging profits of forest harvesting
operations with CL and RIL practices?
• How does the forest respond to RIL in terms of carbon dynamics and tree
species composition?
1.3.3 Overall Goal and Outlines of the Dissertation
The overarching goal of the dissertation is to contribute to the information base
needed for developing forest management approaches that effectively and
simultaneously reduce carbon emissions, assist in the adaptation to global climate
change, guarantee the sustained yield of timber, and preserve tree species in forests of
the Congo Basin. Ultimately, I hope the results of this study will help inform
governments, communities, private land owners, forest industries, and conservation
managers of the benefits of implementing improved methods of timber extraction. The
resulting data should also be useful to policy-makers and implementers of REDD+
initiatives (reduced emissions from deforestation and forest degradation with
enhancement of forest carbon stocks) in the Congo Basin. The study was conducted in
31
three different sites in the forests of two logging companies: SEEF Monts de Cristal
(TFF-RIL), CEB Milolé (FSC certified), and SEEF Milolé (CL; Conventional Logging).
Chapter 1: Background and overview. This introductory chapter provides
background information with some literature reviews on the history of forest
management in Central Africa as well as studies on the regeneration
ecology of the main harvested timber species. The motivation for
conducting this study resulted from that review, which showed a clear lack
of information on the subject for the region. This chapter also discusses the
main goals and objectives of this dissertation.
Chapter 2: Impacts of timber extraction on forest above-ground biomass on
the Monts de Cristal, Gabon (Published in Forest Ecology and Management
Vol.262, Issue 9, 2011). This chapter focuses on Monts de Cristal in
northwestern Gabon, a landscape with high diversity of plant and animal
species. Timber exploitation in the study site is carried out by the Société
Equatoriale d’Exploitation Forestière (SEEF), a Gabonese-owned company.
In collaboration with the Tropical Forest Foundation (TFF), SEEF allocated
a portion of its concession for demonstration and research projects on
improved forest management using RIL guidelines. Advantage was taken of
the collaboration with TFF to conduct a before-and-after logging impacts
study on forest structure and above-ground biomass (AGB). The main
purpose was to evaluate the impacts of felling and skidding using RIL
guidelines on the abundance of tree species and AGB.
32
Chapter 3: A Forest Stewardship Council Certified Logging Concession
Compared with an Adjacent Uncertified Concession in Gabon on the basis
of Tree Species Richness and Composition, Stand and Soil Damage, and
Changes in Above-Ground Biomass (In review for Environmental
Management). This chapter encompasses studies conducted in two logging
concessions (i.e., sites) with different management practices. It compares
the short-term impacts of selective logging in a FSC certified logging
concession of the Compagnie Equatoriale de Bois (CEB), member of the
Precious-Woods group in Milolé and a matched uncertified SEEF
concession in Milolé in Gabon. SEEF was still implementing conventional
logging techniques in which trees are not mapped and annual allowable
cuts were not designated, but the concession was divided into annual
cutting blocks. I focus on the impacts on damage to trees and soils as well
as logging-induced changes in tree species composition and above-ground
carbon stocks. CEB is implementing RIL guidelines with logging operations
planned in advance and forest workers trained and the concession is
managed based on FSC standards. The study sites are very similar in
biophysical conditions. The main objective of the study was to compare the
effects of improved forest management practices (i.e., FSC certified), on
forest structure and AGB relative to unplanned CL operations. Logging
damage was assessed for felling, skidding, log construction, and road-
related construction and maintenance in each site and then compared.
Impacts on AGB, tree species richness, and other environmental
33
characteristics were also evaluated and compared between sites. Logging
impacts varied substantially between the CL and the FSC sites. Certification
was related to damage reduction and reduced carbon emissions.
Chapter 4: Cost comparisons of reduced-impact and conventional logging
in the tropics (Accepted in the Journal of Forest Economics). This chapter
encompasses a case study of logging cost data from TFF-Gabon and a
broad review of published data on the costs of RIL and CL in the tropics.
Although RIL operations have been promoted as having a financial
advantage over CL, a review of 12 studies conducted in Africa, Asia, and
South America revealed that logging costs vary with practices and the
activities included in the cost estimates. Given the financial and ecological
importance of forest management, detailed information about the costs and
benefits of different improved practices are needed to make informed
decisions that will determine the fates of forests and forest industries. In
addition, the lack of adequate information about logging costs should not
continue to preclude sound decision-making about adoption of improved
forest management techniques. There is a need for accurate cost estimates
to inform decision-making about forest management and to secure
incentives to reduce carbon emissions due to logging.
Chapter 5: Post-harvest stand recovery of tree abundance and above-
ground biomass after RIL harvest in Gabon. In this chapter I investigate
forest recovery and carbon dynamics after logging using data from the
Monts de Cristal site where plots were established before logging in 2009
34
and remeasured 2 years later to derive growth and mortality data. A stand
simulation model, developed based on a matrix population model, was used
to simulate changes in AGB and tree abundance over time for different
scenarios of logging intensities and cutting cycle durations. The simulated
tree abundance and AGB are predicted reach the initial level before the end
of the recommended cutting cycle. Both tree abundance and AGB are
sensitive to tree mortality and logging intensity.
1.3.4 Overview of Gabon
The dissertation research was conducted in Gabon in three sites within two
logging concession with different logging practices (Figure 1-2). Gabon straddles the
equator, between 2°30' latitude north and has a land area of 267,667 km2. The country
is bordered on the north by Equatorial Guinea and Cameroon, on the east by the
Republic of Congo, and on the west by the Atlantic Ocean. Gabon is the least densely
populated country in Central Africa with 1.5 million inhabitants or an average density of
4 inhabitants per km2. 73% of the population is urban, with more than 50% living in
Libreville, the capital.
1.3.4.1 Climate
The climate of Gabon is hot and humid with annual temperature ranging from 22°
to 32°C, humidity of 85%, and precipitation of 2000 - 3800 mm year-1(Christy et al.,
2003). The climate is driven by the trade winds from the south Atlantic. The country is
generally characterized by four seasons: two dry seasons (June-August and December-
January) and two rainy seasons (September-December and February-May; Medzegue,
2007).
35
The country is divided into three climate zones: 1) the equatorial climate in the
northeastern area characterized by two distinct dry seasons; 2) a transitional tropical
climate in the central region, characterized by a three-month dry season; 3) the tropical
climate in the southeast, characterized by a five-month dry season (Leonard and
Richard, 1993).
1.3.4.2 Vegetation
The vegetation of Gabon is principally Guinean-Congolese rainforest (85% of the
land area) with 15% of the national area consists of mosaics forest and savanna,
swamps, and mangroves (Nasi et al., 2006). Gabonese forests are classified as
mangrove, flooded swamps, non-flooded forest of the coastal region, montane forest,
forest of the inner plateau, and forest lacking Aucoumea klaineana Pierre, Burseraceae
(okoumé) on the northeast plateau (Christy et al., 2003). For logging purposes, the
vegetation is divided into a forest zone characterized by the abundance of okoumé that
covers 30% of the forest domain and a forest zone with diversified habitats that covers
70% of the forest domain (WRI, 2009).
1.3.4.3 Forestry sector
The forestry sector is the second largest employer in Gabon (28% of the active
population) after the government (WRI, 2009). Timber production is economically very
important for the country, although the contribution of the forest sector to the gross
domestic product is relatively low (6%). To improve activities in the forestry sector, a
forestry code was adopted in 2001 (Law 016/01 from 31 December 2001) based on a
review of Law 01/82 of 22 July 1982. The forestry code focuses on three main points:
sustainable management of forest resources and industrialization of the timber sector;
36
conservation and protection of ecosystems; and, contribution to the efforts of poverty
alleviation.
1.3.4.4 Logging operations
Logging operations in Gabon are supposed to be conducted in compliance with
the 2001 Forestry Code. This code specifies that logging permits be reviewed, and
describes the types of forest concession (see details below) and the granted modalities.
For example, it obligates companies to associate forest management with local timber
processing through installation of processing facilities with a goal of 75% local
production by 2011. In 2009, Gabon had reached only 23% local processing, far below
the targeted threshold, which resulted in a presidential decision in January 2010 to ban
log export.
Currently there are three types of forest concessions in Gabon (WRI, 2009):
1. Forest concessions under sustainable management (CFAD) for which the area
should not exceed 600,000 ha. A temporary convention of management,
exploitation, and timber processing should be signed for a period of three years to
provide time for the concessionaire to develop forest management and
industrialization plans.
2. Associated forest permit (PFA) is intended for nationals only. Several
concessionaires of PFA titles can be grouped to constitute one CFAD. For that,
each forest concession cannot exceed 15 000 ha when integrated into a CFAD.
However, the forest concession area can be 50 000 ha if it constituted a managed
unit by itself but a unique management plan is required.
37
3. Agreed-upon mutual consent permits (permis de gré à gré; PGG) granted for local
wood processing of up to 50 trees to nationals in the rural forest domain. The
permit is not subjected to special management rules.
Gabonese forestry policies aim to promote progress towards the sustainable
management of production forests. Logging companies are required to provide
management plans and some forestry operators are committed to third-party
certification (Table 1-3).
38
Table 1-1. Overview of forests by country in the Central Africa Region
Countries Forest area (ha)
Forest cover (%)
Carbon stock (x106
Mg)
Concession (x106 ha)
Timber produced (x106 m3)*
Contribution to the GDP
(%)
Cameroon 21,245 45.6 5,043 8.84 2.29 6.0
CAR 22,755 36.5 5,460 3.6 0.54 6.3
Congo 22,471 65.8 4,219 11.4 1.33 5.6
DRC 133,610 58.9 27,258 20-45 0.31 1.0
Eq. Guinea 16,320 58.2 445 1.4 0.52 0.2
Gabon 21,775 84.5 4,383 13.4 3.35 4.3
* Figures from 2007 Source: FAO, 2005; de Wasseige et al., 2009 Table 1-2. Natural forest management in Central Africa* Countries Concession
area (x106
ha)
# of granted concessions
Concession with approved management
plans
# Certified concessions
Area certified
(ha)
Type of certification
Cameroon 8.84 103 65 8 899,822 FSC
CAR 3.6 11 8 1 195,500 OLB
Congo 11.4 52 3 2 834,302 FSC
DRC 20-45 65 0 1 - OLB
Equatorial Guinea
1.4 6 - - - -
Gabon 13.4 44 11 3 1,626,177 FSC
* These figures do not include plantations Source: FORAF, 2008 in de Wasseige et al., 2009 ; Nasi et al., 2012
Table 1-3. Distribution of forest certification in Gabon Type of certificates Number of concessions Area (ha)
Keurhout 2 688,262 FSC 3 1,626,177 OLB 2 622,399 ISO 14001 1 549,327
Source: Chevalier, 2009; Nasi et al., 2012
39
Figure 1-1. Land affectation in the Congo Basin. A) Protected areas (green). B) Logging concessions (brown). (From WRI, 2009).
40
Figure 1-2. Location of the study sites in logging concessions in Gabon.
41
CHAPTER 2 IMPACTS OF SELECTIVE LOGGING ON ABOVE-GROUND FOREST BIOMASS IN
THE MONTS DE CRISTAL IN GABON
2.1 Introduction
The 2 million km² of Congo Basin forest is high in conservation value and crucial
for both national development and the livelihoods of about 100 million people (de
Wasseige et al., 2009; FAO, 2011). These forests are among the world's most intact
and provide a variety of ecosystem services (e.g., protection of water and soil
resources), support thousands of plant species, and harbor populations of many
endangered animal species (Blake et al., 2007; Clark et al., 2009). Although these
forests are actively used by people, they are needed to maintain viable populations of
many wide-ranging species (Putz et al., 2001; Sanderson, 2006). In addition to
supporting biodiversity, these forests also store and absorb substantial but seldom
measured quantities of carbon (Lewis et al., 2009).
Logging by untrained and unsupervised laborers working without the aid of
adequate management plans is taking a great toll on Congo Basin forests (Hall et al.,
2003; Ruiz-Perez et al., 2005; Cerutti et al., 2008 and 2011; Ezzine de Blas and Ruiz
Perez, 2008; Angelsen et al., 2009; Poulsen et al., 2009). In Gabon, for example, of the
12.4 x 106 ha of “forêt productive enregistrée” in 2008, management plans were
prepared for only about a quarter of the area being logged (WRI, 2009).
Based on studies conducted in tropical forests but not in Africa, employment of
reduced-impact logging (RIL) practices reduces collateral forest damage, which results
in both substantial reductions in carbon emissions and increased biodiversity retention
(see Johns et al., 1996; Pinard and Putz, 1996; van Rheenen et al., 2004; Putz et al.,
2008a; van Kuijk et al., 2009; Putz and Nasi, 2009). RIL is a suite of techniques based
42
on scientific and engineering principles that, in combination with worker education,
training, and supervision, improves the efficiency of application of labor and equipment
in the harvesting of industrial timber while reducing damage to residual stands (Dykstra,
2002). While definitions of RIL are imprecise and not all of these techniques are used in
every site claiming to be applying RIL, studies of areas logged using some or all of
these techniques in South America and Asia have shown increased forest carbon
retention both immediately after logging and for decades afterwards (Putz et al., 2008b).
Unfortunately, the implementation of RIL requires up-front capital investment in timber
inventories, staff training, and sometimes machinery, along with substantial
modifications in working practices. Logging company managers and owners are unlikely
to make these investments without clear indications of their benefits. Evidence for these
benefits is scarce in Central Africa where little is known about the extent to which
employment of RIL could serve to preserve carbon and tree species diversity.
If poorly implemented, selective logging causes substantial damage to residual
stands. Logging directly and indirectly affects all components of biodiversity, from genes
to landscapes (Putz et al., 2001; van Kuijk et al., 2009). Even light selective logging
affects tree species composition, densities, and size-class frequency distributions, but
the deleterious environmental impacts of logging can be substantially reduced if
appropriate techniques are used (Bertault and Sist, 1997; Durrieu de Madron, 1998; van
Kuijk et al., 2009). In a selectively logged forest in Indonesian Borneo, for example,
Cannon et al. (1998) found that 8 years after logging the density of trees ≥ 20 cm dbh
was lower than in unlogged plots but detected no difference in tree species richness.
Similarly, in Southwestern Central African Republic, Hall et al. (2003) found that forest
43
sampled 18 years post-logging had lower tree densities than either unlogged stands or
stands sampled 6 months post-harvest, but did not detect difference in tree species
richness.
Tree harvesting unavoidably opens canopy gaps, reduces overall canopy cover,
and disturbs soil surface. In Bolivia, for example, Jackson et al. (2002) found that
planned harvesting of 4.3 trees ha-1 (12.1 m3 ha-1) opened the canopy over 25% of their
study area. In eastern Brazil, Johns et al. (1996) found canopy reduction of 10% after
planned logging of 4.5 trees ha-1 (37 m3 ha-1). In the semi-deciduous forest of Mbaïki,
Central African Republic, Durrieu de Madron et al. (2000) reported 14 - 22% of area
disturbed due to the harvest of 2.6 - 4.0 trees ha-1 (mean = 118 cm dbh). These results
show that the harvest of even small numbers of large trees results in substantial canopy
opening.
Selective timber harvesting degrades forests in the sense that it results in
reductions in carbon stocks that will need to be accounted for if a reduced-emissions
from deforestation and forest degradation (REDD; Angelson, 2009) program is
implemented. If Congo Basin countries are to benefit from REDD and REDD-like
programs, data will be needed on the carbon consequences of logging. To fill some of
the gaps in knowledge about the effects of logging on forest structure, composition, and
above-ground biomass in Africa, we measured the damage resulting from felling and
skidding due to low-intensity RIL in Gabon.
2.2 Methods
2.2.1 Study Site
This study was conducted on the Monts de Cristal region of northwestern Gabon
(0°20’ N; 10°20’ E; Figure 1-2) in the 477,033 ha logging concession of SEEF. The
44
natural vegetation of this region is dense humid evergreen rainforest (Fuhr et al., 1998;
Sunderland et al., 2004). The long-lived pioneer A. klaineana is the most common tree
species. The soils are mostly oxisols, the climate is tropical with a long dry season
(July-September), annual rainfall is 2000-2400 mm (Leonard and Richard, 1993), and
average temperatures are 24- 26o C (Sunderland et al., 2004).
Logging in what is now the SEEF concession started in the 1950s but due mostly
to the undulating topography, elevation (689 m above sea level), and inaccessibility,
harvesting was extremely light and spatially patchy until SEEF started more thorough
exploitation in 2000. Although we lack maps or other information about the history of the
study site, we found no evidence of earlier episodes of harvesting in our study plots.
Timber extraction in SEEF is selective with A. klaineana alone making up about
60% of the total volume. At the time of this study the concessionaire was preparing
management plans for the entire concession with the intention of attaining Forest
Stewardship Council (FSC) certification (T. Ricordeau 2010, pers. comm.). With this
goal in mind, SEEF allocated 250 ha to the Tropical Forest Foundation (TFF) and
FORM International for training, demonstration, and research on RIL as part of an
International Tropical Timber Organization project. In this area, harvestable trees of 27
commercial species were tagged, mapped, measured (dbh; diameter at breast height of
1.4 m or above buttresses), and identified to species. Prior to harvesting a total of 104
ha in July 2009 and January-February 2010, skid trails were planned by TFF on the
basis of topographic and stock maps prepared by an inventory crew. Each feller
received one week of training in directional felling techniques conducted by a
professional trainer contracted by TFF. Trees were felled using chainsaws (Stihl MS
45
880) and yarded by a trained worker with a tracked skidder (Caterpillar D527); both
operations were coordinated and supervised by TFF. Logging intensity in the 104 ha
(0.82 trees ha-1 and 8.11 m3 ha-1) was within the typical range (0.7-4 trees ha-1) for
Central Africa (Durrieu de Madron et al., 1998; Ruiz-Perez et al., 2005).
2.2.2 Plot-Based Measurements
Prior to harvesting, we established ten permanent 200 x 50 m (1 ha) plots at
random locations in 50 ha of the area to be subjected to RIL to capture variability in
logging impacts. All trees ≥10 cm dbh in each plot were measured, tagged, mapped,
classified according to stem quality and crown position (suppressed, side-lighted, sub-
dominant, co-dominant, dominant; see Hall et al., 2003), and assessed for the presence
or absence of lianas. Trees were identified to the species level where possible based on
vegetative characteristics. Voucher specimens were collected for species that could not
be identified in the field and then identified at the National Herbarium in Libreville.
2.2.3 Damage Assessment
Logging damage was assessed in the 1-ha sample plots using methods well-
established in the literature (e.g., Johns et al., 1996; Whitman et al., 1997; reviewed by
Putz et al., 2008b). Damage to roots, boles, and crowns were ranked on a scale from
minor to very severe. Crown damage was recorded as severe (> 66% crown loss),
moderate (33-66% crown loss), or minor (< 33% crown). Bole damage was recorded as
severe (broken bole), moderate (> 100 cm2 of bark removed), or minor (< 100 cm2 of
bark removed). Uprooted trees were recorded as such. Root damage was recorded as
major (>10% of surface roots injured) or minor (< 10% of surface root injured). Crown,
bole, and root damage were attributed to felling and/or skidding.
46
The soil surface area disturbed by skid trails and skidder activities in logging gaps
was measured in the entire 50 ha study area. Skid trails were assigned to one of three
categories based on the number of logs skidded (one log per pass): primary > 10;
secondary 2-10; and, tertiary 1. Skid trail widths were measured every 10 m.
Felling gaps were measured from a central point to the gap edge in the eight
cardinal and inter-cardinal directions based on Brokaw (1982) definition (gaps = forest
canopy openings that are > 20 m2 and extend down through all foliage levels to a height
of < 2 m above ground). A total of 31 single-tree felling gaps were measured in the 50
ha block.
2.2.4 Conversion of Above-Ground Biomass into Necromass
To estimate the amount of aboveground biomass (AGB) converted into necromass
during logging (= “committed emissions” defined here as the estimated equivalent of
carbon dioxide to be emitted by killed trees), we estimated the mass of each tree that
was harvested or destroyed using one of Chave et al. (2005) allometric equations (see
below) with published wood density estimates for trees harvested in Africa (Zanne et al.,
2009). When more than one wood density value was available, we used the arithmetic
mean; in the absence of species-specific wood density data we used the mean value for
the genus or family; and, when none of these values were available, we used the plot
mean (see Section 2.2.5.2).
2.2.5 Data Analyses
2.2.5.1 Tree species richness
Species richness values in the 1 ha plots before and after logging were compared
using EstimateS 8.2.0 (Colwell, 2006) with sample-based rarefaction to compute
expected species accumulation curves. A randomization test without replacement was
47
run to compute richness estimators and diversity indexes (e.g., Fisher’s alpha) based on
the sample size. Randomizations were based on 50, 75, and 100 iterations to produce
relatively smooth diversity index curves. We used the outputs for 100 randomizations
and Fisher’s alpha, defined by the relation S=α ln〖(1+N/α)〗 where S represents the
number of species, N the number of individuals, and α the diversity index (Leigh, 2008).
2.2.5.2 Above-ground biomass estimates
Total above-ground biomass of each tree in the plots was estimated using Chave
et al. (2005) allometric equation:
ln (AGB) = - 1.576 + 2.179* ln (D) + 0.198*(ln(D))2 - 0.0272*(ln(D))3 + 1.036*ln(ρ)
where D= dbh (5-156 cm range) and ρ=wood density (df = 1501). For unidentified
species, we applied the mean wood density for each plot weighted by the number of
trees from each species (see Lewis et al., 2009). For trees larger than 156 cm dbh,
AGB was extrapolated.
2.2.5.3 Logging damage
Analysis of variance (ANOVA) was used to test for differences among skid trail
orders in width and numbers of trees damaged per unit length. For these tests, the units
of replication are skid trails in each category (n = 5 for primary, n = 16 for secondary,
and n = 18 for tertiary). After an overall difference was detected, Tukey’s Honest
Difference tests were employed for pairwise comparisons of means. Regression
analyses were used to determine the relationships between the dbh of the felled trees
and felling gap area and the number of trees damaged.
48
2.3 Results
2.3.1 Trees Diameter Class Distribution
A total of 4527 trees ≥ 10 cm dbh were inventoried in the 10 1-ha plots. Individuals
in the smallest diameter class (10-20 cm) contributed in large proportion to the number
of trees (59%), while trees ≥ 60 cm dbh represented only 5% of the total trees (Figure 2-
1a). Of the inventoried trees, commercial species represented only 25.5%, most of the
individuals found in the 10-20 cm diameter class (52%). Trees above the minimum
cutting diameter (60+ cm) represented 12% of the total commercial species (Figure 2-
1b). Among the commercial species, A. klaineana represented only 2% of the total trees
with variability in the diameter class distribution. Individuals above the minimum cutting
diameter of this tree species (80 cm) represented 43% of the total number of A.
klaineana trees (n = 21) inventoried in the 10 1-ha plots (Figure 2-1c).
2.3.2 Tree Species Richness
The 4527 trees ≥10 cm dbh inventoried represented 45 families, 188 genera, and
214 species were encountered in the 10 1-ha plots (Table 2-1). Of the inventoried trees,
A. klaineana, the main harvested timber species in Gabon, represented 0.46% of the
trees in the plots. Other important timber species such as Bikinia durandii
(Caesalpiniaceae) represented 1.3% of the identified trees, Dacryodes igaganga
(Burseraceae) 4%, and Dialium angolense (Caesalpiniaceae) 4.5%; Dacryodes
buettneri (Burseraceae), a species that produces commercial timber but is banned from
harvesting due to its high biodiversity value, represented 1.2% of the trees. Coula edulis
(Olacaceae), an important non-timber forest product (NTFP) species, constituted 2.9%
of the trees (Appendix A). Caesalpiniaceae is the dominant family in the area, with
49
28.9% of the trees, followed by Olacaceae (13.4%), Burseraceae (12.7%), and
Euphorbiaceae (8.1%; Appendix B).
In the 10 1-ha plots before and after logging, respectively, average species
richness of trees >10 cm dbh were 214 and 206, Fisher’s alphas were 47.01 and 46.56.
None of these descriptors of species richness and diversity were affected by logging (all
comparisons based on t-tests with df = 18; species richness t = 0.33, p = 0.75; Fisher’s
alpha t = 0. 26, p = 0.79).
2.3.3 Above-Ground Biomass Prior to Logging
Prior to harvesting, the total AGB estimates for the 10 1-ha plots ranged 293.4 -
511.1 Mg ha-1 with an average of 420.4 ± 92.95 Mg ha-1 (Table 2-2). Trees of the main
commercial species, A. klaineana, represented 6.9% of the total AGB. Other timber
species that contributed substantially to the total AGB included D. angolense (7.2%),
Bikinia le-testui (6.3%), B. durandii (3.9%), and D. buettneri (3.2%), and C. edulis
(11.9%; Appendix A).
2.3.4 Logging Characteristics
A total of 85 timber trees were harvested in the 104-ha RIL area and the size
ranged 60 - 142 cm dbh with a mean of 85 ± 18.31 cm; harvest intensity was light with
0.82 trees ha-1 (8.11 m3 ha-1) extracted. A. klaineana was the dominant tree species
harvested representing 50.4% of the total volume harvested and 41.3% of the total
biomass. Another important timber species Bikinia durrandii constituted 42.3% of the
total harvested biomass and 34.2% of the total volume (Table 2-2).
2.3.5 Logging Damage
Of the ten intensively monitored 1-ha plots, five yielded one harvested tree and
three other plots in which there was no felling contained trees damaged due to skidding;
50
the other two plots suffered no damage from either felling or skidding (Table 2-3). Based
on data from the 10 1-ha plots, the felling of one tree resulted in damage to an average
of 11.0 other trees ≥ 10 cm dbh while skidding to the log landing resulted in damage to
an additional 15.6 trees ≥ 10 cm. One plot was severely damaged due to both felling
and skidding with 103 trees damaged (22.2% of trees >10 cm dbh) and 17 trees
completely destroyed (i.e., uprooted) and converted into 32.9 Mg of necromass (Table
2-3).
The 39 skid trails in the 50 ha study area covered 2.8% of the ground surface
(Table 2-4). Machine maneuvering resulted in an additional 2.6% of the ground surface
being disturbed. Thus, a total of 5.4% of the ground was disturbed by log yarding
activities. Of the trees damaged by skidding, about 26% were uprooted, 6% suffered
severe bole damage, and 1% suffered severe root damage. Therefore, skidding was a
major cause of tree uprooting and to injuries to tree boles and roots (Figure 2-2). No
differences were detected among the numbers of trees damaged per length of primary
(0.11 trees m-1), secondary (0.12 trees m-1), and tertiary (0.10 trees m-1) skid trails (F =
0.68; df = 36; p = 0.88). The widths of the three orders of skid trails also did not differ (F
= 0.70; df = 342; p = 0.50; Table 2-4).
Skidding operations caused damage to trees in the 10 1-ha permanent plots and
the interior of the 50-ha study area. The number of trees damaged differed among the
skid trail categories with the secondary skid trail damaging more trees than the tertiary
and the primary skid trails. On average 15.6 trees ≥10 cm dbh in plots and 7.9 trees ≥10
cm dbh outside the plots were damaged during the skidding operation resulting in an
51
average biomass of damaged trees of 9.3 Mg ha-1 for the plots and 4.6 Mg ha-1 outside
the plots. Therefore, on average 13.9 Mg ha-1 were damaged due skidding (Table 2-5).
Felling activities that occurred in five of the ten 1-ha plots damaged a total of 110
trees ≥ 10 cm representing an average of 11.0 trees ± 5.93 damaged per felled tree
(Table 2-3). Of the trees damaged during felling, 56% suffered severe crown damage,
7% suffered moderate crown damage, and 15% suffered minor crown damage. Some
trees were uprooted during felling operations (10%) and a few suffered root damage
(9%; Figure 2-2). Therefore, felling activities constituted a major source of tree crown
loss and bole damage, but also resulted in a complete destruction of trees (i.e.,
uprooted) and severe root injuries (Figure 2-2).
Felling gap area increased exponentially with dbh of the felled tree (Figure 2-3a)
with a mean gap area per felled tree of 320.3 ± 44.29 m2 (mean ± S.E). The number of
damaged trees also increased with the size the felled tree (Figure 2-3b).
2.3.6 Above-Ground Biomass Converted into Necromass (i.e., Committed Emissions)
Trees destroyed during felling and skidding (25.6 Mg ha-1) plus the AGB in the
harvested trees (8.6 Mg ha-1) resulted in an average biomass loss of 34.2 Mg ha-1. Due
to extreme heterogeneity in logging intensity, AGB of the trees that were completely
destroyed per hectare ranged from 0 to 32.9 Mg ha-1 with an average of 8.4 Mg ha-1.
Overall, the total committed emissions due to harvesting of 0.82 trees ha-1 (8.11 m3 ha-
1) from the 10 monitored 1-ha plots averaged 17.0 Mg ha-1 or 4.05% of the pre-logging
total (Table 2-3).
52
2.4 Discussion and Concluding Remarks
2.4.1 Tree Species Richness
The modest impacts of low-intensity selective logging using improved techniques
(RIL) on species richness and other measures of tree diversity on Monts de Cristal in
Gabon is similar to results reported for forests elsewhere in the tropics. For example,
Hall et al. (2003) found very similar Shannon diversity indices for trees in an unlogged
forest stand and in nearby stands 6 months and 18 years post-harvest in the Central
African Republic. While comforting from a biodiversity conservation perspective, it is
important to note that such measures cannot indicate changes in species composition,
and do not address the longer-term impacts of repeated harvesting (Kariuki et al.,
2006).
2.4.2 Above-Ground Biomass
Immediately after the harvest of 0.82 tree ha-1 (8.11 m3 ha-1), aboveground
biomass declined from the pre-logging average of 420.4 Mg ha-1 (Table 2-2) to 386.2
Mg ha-1 (Table 2-3) with an additional biomass of 17.2 Mg ha-1 in trees that suffered
minor to moderate damage. Compared with other studies on logging damage, the
impacts on Monts de Cristal were small most probably due to both the use of RIL
techniques and the low logging intensity (Table 2-5).
The likely benefits of RIL impacts varied with logging practices and intensities in
the tropics. For example under RIL practices in Brazil, Mazzei et al. (2010) reported
AGB loss of 94.5 Mg ha-1 due to the harvest of 6 trees ha-1 (21.3 m3 ha-1). Under RIL in
the Republic of Congo, with a logging intensity even lower (0.53 trees ha-1) than in our
study site (0.82 trees ha-1), Brown et al. (2005) reported AGB loss of 20.4 Mg ha-1, not
accounting for AGB of extracted logs, slightly greater than our value (17.0 Mg ha-1;
53
Table 2-5). These results suggest that residual stand damage reduction and biomass
retention vary with both the intensity of logging and the harvesting practices employed.
Although large trees (i.e., trees ≥200 cm dbh) occurred at low densities in our
study area (0.2% of all trees >10 cm dbh; 1 individual/2.5 ha), they contained
substantial AGB. Given the difficulty of making accurate measurements of buttressed
trees and the presence of many extremely large trees of commercial species in our
study area, this minor infraction is understandable and perhaps pardonable. In light of
the exponential increase in residual stand damage with increasing dbh of harvested
trees (Figure 2-3b), such upper bounds on harvestable tree size are justifiable from a
carbon-balance perspective. On the other hand, given that the conversion efficiency of
round wood into veneer and saw timber and hence the commercial value of logs
increases with their diameter, the opportunity costs of foregoing the harvest of very
large trees will remain a controversial issue in managed natural forests.
While we used what we deemed the most robust allometric equation available for
estimating AGB from dbh and wood density (Chave et al., 2005), no African trees were
included in the dataset on which the equation was based. Nevertheless, similar results
were obtained when we used the equations of Brown et al. (2005) and Djomo et al.
(2010) that include African data but do not include wood density as an independent
variable. A bigger problem is that some trees in our study area were substantially larger
than any used to develop any of these equations (156 cm dbh). Extrapolation beyond
the range of modeled data is questionable practice but truncation at the AGB of the
largest tree used to generate the allometric equation results in obvious underestimate
and thus does not seem like an appropriate alternative (e.g., Mazzei et al., 2010).
54
2.4.3 Logging Damage
2.4.3.1 Damage due to felling
The severity of felling damage increased substantially with the size of trees
harvested (Figure 2-3a). Mean gap size per felled tree in our study was 302.3 ± 44 m2
(mean ± S.E) with the mean dbh of harvested trees of 85 cm dbh. In contrast, for trees
that average 81 cm dbh, Jackson et al. (2002) reported a mean felling gap per tree
extracted of using RIL techniques in Bolivia of 591 ± 92 m2 (mean ± S.E). Using RIL in
Cameroon, Jonkers (2000) found a mean felling gap size of 720 m2 per tree extracted.
Most of the trees harvested in our study site were okoumé (A. klaineana) that do not
have large crowns relative to other tree species such as Bikinia durrandii
(Caesalpiniaceae; Table 2-2). This suggests that gap size likely varies with crown size,
the spatial distribution of the felled trees, and the physiognomy of individual tree
species.
On average 11 trees suffered minor to severe damage per tree felled in our study
site (Figure 2-2b). In Bolivia, Jackson et al. (2002) found 14.6 trees were damaged per
tree felled. In contrast, Johns et al. (1996) found that even in a planned operation in
Brazil, 20.5 trees were damaged per harvested tree. At the other extreme, in a RIL
study in the Republic of Congo, Brown et al. (2005) reported only 7.3 trees damaged
per felled tree with a mean dbh of 123 cm. Similarly, in Southern Cameroon, Jonkers
(2000) found 19.3 trees damaged per felled tree with mean dbh of 91 cm. Even when
carried out with care, felling still results in stand damage, but the number of trees
damaged likely varies with stem density, the abundance of lianas connecting tree
crowns, crown dimensions, and terrain. For example, the lower damage per felled tree
reported by Brown et al. (2005) for their study in the Republic of Congo may result from
55
the relatively flat terrain in their study area compared to the rolling terrain in our study
site in Gabon.
2.4.3.2 Damage due to skidding
Although skid trails were planned in advance, skidding still caused residual stand
damage. Most trees damaged during skidding were uprooted but some suffered severe
bole damage (Figure 2-2). Overall, skid trails and machine maneuvering disturbed 2.8%
and 2.6% of the ground surface, respectively, during the harvest of 0.82 trees ha-1.
During a planned harvest of 4.5 trees ha-1 in Brazil, Johns et al. (1996) reported 5.4% of
the logging area in skid trails and 0.4% affected by machine maneuvering. Similarly, in a
RIL harvest of 6 trees ha-1 in Eastern Amazon, Brazil, Sist and Ferreira (2007) reported
that skid trails occupied 7% of the surface area. In a study on RIL harvest in Cameroon,
Durrieu de Madron (1998) reported that skid trails occupied 3.0% of the surface area
due to the harvest of 0.5-1 trees ha-1 (5 to 15 m3 ha-1); in the same country, Jonkers
(2000) reported 3.9% of area disturbed resulting from the planned harvest of 1.4 trees
ha-1. In Southern Central African Republic, skid trails covered 7.4% of the surface area
due to the harvest 3.7 trees ha-1 (Durrieu de Madron et al., 1998). While these values
vary with logging intensity, they suggest that well-planned skidding operations result in
relatively little damage to the ground surface.
Despite efforts to minimize residual stand damage, collateral damage to trees >10
cm dbh during RIL of 0.82 tree ha-1 (8.11 m3 ha-1) in Gabon resulted in an average AGB
impact of 34.2 Mg ha-1 of which 17.0 Mg ha-1 (4.05% of the pre-harvest AGB; Table 2-3)
was converted into necromass. In a study conducted in Eastern Amazon, Mazzei et al.
(2010) reported that with RIL of 6.0 trees ha-1 (21.3 m3 ha-1), 94.5 Mg ha-1 (23% of the
pre-harvest AGB; includes harvested and destroyed trees) were converted into
56
necromass, which is much greater than the value reported here. It is important to note
that an average of 8.6 Mg ha-1 was in the trees harvested from our plots, of which a
portion will be converted into end forest products with long carbon storage, but we do
not account for this retained carbon.
2.4.4 Management Implications
Future timber yields from selectively logged tropical forests will vary with the ways
forests are harvested. Yield sustainability is crucial in Central Africa where timber
production is of great economic importance. Forest industries contribute up to 7% to the
economies of Congo Basin countries, and, in Gabon, they are the second largest
employer after the government (Minnemeyer, 2002; de Wasseige et al., 2009). If
logging is wasteful, timber stocks will decline rapidly, thereby compromising the ability of
the forest to support future extractive economic activities.
While the focus of this study was on logging damage and carbon losses when RIL
practices are used in selectively harvested forest, sustainability of management was the
underlying motivation. One concern is that by implementing RIL, regeneration of
disturbance-dependent species will be disfavored. In particular, managing for the
sustained yield of A. klaineana, the main timber species in Gabon, is problematic
because it is light-demanding and unlikely to regenerate under the small canopy gaps
that are the objective of RIL and that favor carbon retention. To retain this and other
light-demanding species, silviculural treatments such as liberation thinning or even
scarification of the surface soil in felling gaps may be required (Fredericksen and Putz,
2003; Sist and Ferreira, 2007; Peña-Claros et al., 2008). Clearly research is needed on
forest management strategies that mitigate carbon emissions while ensuring the
maintenance of timber stocks. Given the many environmental and social benefits of RIL
57
and the fact that adoption of good timber harvesting practices in the Congo Basin is still
facing challenges (Ezzine de Blas and Ruiz Perez, 2008; Cerutti et al., 2011), perhaps
concerns about species-specific sustainability of volume yields need to be considred in
light of the carbon and biodiversity benefits of sound harvesting practices. Thus, with
economic development as part of the “development adjustment factors” set by the
Central African Commission for Forests (COMIFAC) countries and efforts to reduce
emissions from deforestation and forest degradation and enhancing forest carbon
stocks (REDD+), the implementation of better forest harvesting techniques is timely for
the sub-region.
58
Table 2-1. Characteristics of the ten 1-ha plots before logging on Monts de Cristal, Gabon (trees ≥10 cm dbh).
Plot Total number of
trees
Number of species a
Maximum tree size
(cm)
Total basal area (m2 ha-1)
Total AGB (Mg)
RIL-P1 463 89 (12) 198b 31.6 431.1
RIL-P2 440 95 (5) 113 26.8 315.8
RIL-P3 432 90 (8) 195 b 35.5 508.2
RIL-P4 445 91 (7) 116 26.3 293.4
RIL-P5 520 99 (6) 152 36.6 508.2
RIL-P6 471 90 (4) 144 33.4 468.3
RIL-P7 478 86 (2) 150 27.3 357.3 RIL-P8 462 88 (1) 122 24.7 305.0
RIL-P9 391 81 (2) 146 34.3 506.1
RIL-P10 425 86 (2) 169 b 34.2 511.1
Mean c 452.7±34.83 89.5±4.97 150.5±29.91 31.1±4.37 420.4±92.95 a Number of unidentified species noted in parentheses. b Includes AGB of trees larger than those used to generate the allometric equation with which AGB was estimated (i.e., extrapolated estimates). c Values are mean ± 1 standard deviation.
59
Table 2-2. Timber trees harvested in the 104-ha RIL area on Monts de Cristal, Gabon
Species No. harvested
trees
Mean DBH (cm ± s.d)
Max DBH (cm)
Basal Area (m2)
Harvested volume (m3)
AGB harvested (Mg)
Afzelia bipindensis 2 54.5±2.12 56.0 0.47 6.90 7.56
Aukoumea klaineana 43 84.8±15.03 140.0 25.03 425.04 252.20
Bikinia durrandii 22 100.8±16.24 142.5 17.99 284.58 258.63
Guibourtia ehie 5 57.5±0.94 58.5 1.30 27.95 20.63
Guibourtia tessmannii
1 85.50 85.5 0.57 8.76 11.42
Mitragyna ciliata 10 75.1±5.65 84.0 4.45 81.53 49.91
Pterocarpus soyauxii 2 66.3±0.35 66.5 0.69 8.68 10.77
Total 85 85.0±18.31 142.5 50.49 843.44 611.12
60
Table 2-3. Number of damaged trees (≥10 cm dbh) per 1-ha plot and associated above-ground biomass (AGB) losses on the RIL study area on Monts de Cristal, Gabon.
Plot # of trees
harvested
AGB of trees
harvested (Mg)
# of trees
damaged by
felling
AGB damage
d by felling (Mg)
# of trees
damaged by
skidding
AGB damage
d by skidding
(Mg)
# of destroyed trees
AGB destroyed trees
AGB converted
into necromas
s (Mg)a
AGB damage & trees
harvested (Mg)b
Total AGB after
harvest (Mg)
RIL-P1 1 48.3 58 93.9 45 32.4 17 32.9 81.2 174.6 256.5
RIL-P2 1 12.7 27 35.8 - - 4 15.5 28.2 48.5 267.3
RIL-P3 - - - - 17 5.6 9 1.0 1.0 5.6 502.6
RIL-P4 - - - - - - - - - - 293.4
RIL-P5 1 6.7 2 7.1 - - - 6.9 13.6 13.8 494.4
RIL-P6 1 10.5 14 23.6 10 4.8 7 18.3 28.8 38.9 429.4
RIL-P7 - - - - 22 13.9 4 0.3 0.3 13.9 343.4
RIL-P8 1 8.1 9 2.2 30 20.1 8 1.8 9.9 30.4 274.6
RIL-P9 - - - - - - - - - - 506.1
RIL-P10 - - - - 32 16.6 2 7.3 7.3 16.6 494.5
Mean 8.6 11.0 16.3 15.6 9.3 5.1 8.4 17.0 34.2 386.2
a The AGB converted into necromass (i.e., the committed emissions) includes the entire above-ground biomass of harvested trees plus the AGB of trees that were completely destroyed. b The AGB damage & trees harvested includes the entire above-ground biomass of harvested trees plus AGB of trees damaged by both felling and skidding. Table 2-4. Number and surface area of skid trails in the 50 ha RIL study area on Monts de Cristal, Gabon.
Category Number Total length Mean width (±sd) Ground area covered
Primary 5 350 m 3.9 m (±1.09) 1365 m2 (0.3 %)
Secondary 16 1630 m 4.1 m (±1.45) 6683 m2 (1.3 %)
Tertiary 18 1470 m 4.2 m (±1.43) 6174 m2 (1.2 %)
61
Table 2-5. Number of damaged trees ≥10 cm dbh and associated AGB loss along 3450 m of skid trails in 50 ha and felling damage in 5 1-ha plots in the study area on Monts de Cristal, Gabon.
In plots In the 50-ha study area
Logging operation # of damaged
trees
AGB loss (Mg)
AGB density loss (Mg ha-1)
# of damaged
trees
AGB loss (Mg)
AGB density loss (Mg ha-1)
Total AGB loss (Mg ha-1)
Felling 110 162.6 16.3 - - - 16.3
Primary skid trail - - - 40 16.4 0.3 0.3
Secondary skid trail 104 75.7 7.6 200 105.6 2.1 9.7
Tertiary skid trail 52 17.7 1.8 154 105.7 2.1 3.9
Total skidding 156 93.4 9.3 394 227.7 4.6 13.9
Table 2-6. Above-ground biomass loss due to selective logging in the tropics
Logging Method
AGB before logging (Mg ha-1)
AGB loss from logging (Mg ha-1)
Percent loss (%)
Trees harvested (trees ha-1)
Volume extracted (m3 ha-1)
Study site Source
CL 331.4 159.4 48.1 12.9 154 Sabah, Malaysia
Pinard and Putz 1996
RIL 327.0 90.0 27.5 8.7 104 Sabah, Malaysia
Pinard and Putz 1996
RIL 410.0 94.5 23.0 6.0 21.3 Paragominas, Brazil
Mazzei et al. 2010
CL 423.5 170.0 40.1 10.4 32.5 French Guiana Blanc et al. 2009; Rutishauser et al. 2010
RIL 554.0 20.4 3.7 0.53 11.0 Sangha, Rep. of Congo
Brown et al. 2005
RIL 420.4 17.0 4.1 0.82 8.11 Monts de Cristal, Gabon
Medjibe et al. 2011
62
Figure 2-1. Stem diameter class distribution in the 10 1-ha plots on Monts de Cristal. A)
All tree species. B) All timber trees. C) Okoumé (A. klaineana).
63
Figure 2-2. Percentage of damaged trees per damage category from felling and skidding during selective logging on Monts de Cristal, Gabon
64
Figure 2-3. Relationships between stem diameters (dbh) of harvested trees on Monts de Cristal, Gabon. A) Felling gap size. B) Number of damaged trees
65
CHAPTER 3 A FOREST STEWARDSHIP COUNCIL CERTIFIED LOGGING CONCESSION
COMPARED WITH AN UNCERTIFIED CONCESSION IN GABON ON THE BASIS OF TREE SPECIES RICHNESS AND COMPOSITION, STAND AND SOIL DAMAGE, AND
ABOVE-GROUND BIOMASS
3.1 Overview
Forest certification emerged in the 1990s as a market-based mechanism to
promote sustainable management of forest resources (e.g., Auld et al., 2008). While
there is a great deal of mostly indirect evidence indicating certification’s impacts on
improving management practices in production forests, direct field measures of impacts
are scarce (Blackman and Rivera, 2010). This study addresses this deficiency by
providing data on the environmental impacts of selective logging of a certified and an
adjacent uncertified forest concession in Gabon.
The principles on which responsible forest management are judged in certification
programs, including the Forest Stewardship Council (FSC), include measures to reduce
the deleterious environmental impacts of logging (Burger and Eschborn, 2005). The
recommended harvesting practices, generally referred to as reduced-impact logging
(RIL), include worker training and supervision, detailed harvest plans that include
demarcated extraction paths (i.e., skid trails), directional felling to reduce collateral
stand damage, and protection of riparian areas. While FSC principles address other
environmental as well as social and economic factors, in this study the focus is on the
direct, short-term, biophysical impacts.
Like many studies of the impacts of forest management, this one is pseudo-
replicated (Helme et al., 2011), but the two concessions compared are well-matched on
the basis of many criteria (Table 3-1), which argues for their appropriateness as
counterfactuals (Andam et al., 2008; Pattanayak et al., 2010).
66
Selective logging, no matter how carefully conducted or low in intensity, changes
forests. For example, timber removal reduces standing stocks of carbon in forests that
continue to satisfy the definition of forest; from a carbon emissions perspective, this
constitutes degradation (FAO, 2005). Similarly, logging unavoidably changes forest
structure and species composition, but the reported magnitudes of these impacts vary
among research protocols (e.g., spatial and temporal aspects of sampling), taxa, and
logging intensities; they presumably also vary with logging techniques, but these
difference are less well studied (Putz et al., 2012). The adoption of sound management
practices, as indicated by voluntary third-party certification by organizations like the
FSC, is presumed to reduce these impacts to acceptable levels, which some authors
refer to as sustainable forest management (SFM; Write et al., 2002; Sheil et al., 2004; ).
Although credible certification is widely touted for its high environmental, economic, and
social standards (Tikina and Innes, 2008; Eba’a Atyi et al., 2009; van Kuijk, Putz, and
Zagt 2009; Sheil, Putz, and Zagt, 2010), the assumed benefits need to be evaluated
with field measurements (Blackman and Rivera, 2010).
FSC Principle 6 states that “Forest management shall conserve biological diversity
and its associated values, water resources, soils, and unique and fragile ecosystems
and landscapes, and, by so doing, maintain the ecological function and the integrity of
the forest” (FSC, 2004). It is widely agreed that whether this principle is being upheld
needs to be tested with field data (Sheil et al., 2004; van Kuijk et al., 2009; Ebeling and
Yasué, 2009), but there are few data available from ground-based studies on forest
carbon stocks and biodiversity in certified logging areas in the tropics in general or the
Congo Basin in particular (but see Poulsen and Clark, 2009; Wanders, 2010).
67
All credibly certified forest operations include measures for biodiversity
conservation and environmental protection in their plans. Examples include the
designation and protection of riparian zones, water catchments, and high conservation
value forests. Typically, certified forest concessions are logged with compliance to RIL
guidelines that encompass careful planning of the road infrastructure, operational
logging plans, directional felling, and other elements (Eba’a Atyi, 2002). These logging
practices haves been demonstrated to result in substantial reductions in carbon
emissions and increases biodiversity retention in logged forests (see Pinard and Putz,
1996; van Rheenen et al., 2004; van Kuijk et al., 2009).
In Gabon where this study was conducted, forest certification appears to have
helped convince forest policy-makers to define new technical and legal standards for
the management of forest resources (Eba’a Atyi, 2004). In early 2012, 3 of 44 granted
concessions (9.7% of the area of the production forests) were FSC certified, 2
concessions (8.2% of the concession area) were Keurhout certified, and 2 were OLB
certified (de Wasseige et al., 2009; WRI, 2009; Nasi et al., 2012). FSC certification was
granted by Bureau Veritas and the certified forests covered 61% of the total certified
concessions (http://info.fsc.org). To determine whether forest certification serves to
sustain timber yields, reduce carbon emissions, and preserve tree species diversity,
field data are needed.
3.2 Methods
3.2.1 Study Site
3.2.1.1 FSC-Certified site (CEB-FSC)
This 508-ha research site is located in the FSC certified logging concession of
Compagnie Equatoriale de Bois (CEB), a member of the Switzerland-based Precious-
68
Woods Group (www.preciouswoods.com). The study was conducted in the 2008 cutting
area in Milolé, Province of Ogooué-Ivindo, southeastern Gabon (0°15’S; 12°45’E; Figure
3-1A). The concession is bordered on the northwest by Ivindo National Park, on the
north and southeast by the CWG (Corà Wood Gabon) concession, and on the west by
the SEEF (Société Equatoriale d’Exploitation Forestière) concession described below
(Figure 3-1). This mixed forest is characterized by the presence of Aucoumea klaineana
(Burseraceae, okoumé), the principal source of commercial timber, along with
Julbernardia pellegriniana (Ceasalpiniaceae, béli), and Scyphocephalium mannii
(Myristicaceae, sorro). The rich fauna includes large mammals such as forest elephants
(Loxodonta africana cyclotis), gorillas (Gorilla gorilla), and chimpanzes (Pan
troglodytes). Soils are mostly oxisols developped on megalithic mid-Precambrian rocks
(Martin et al., 1981; Leonard and Richard, 1993). The climate is equatorial with a long
rainy season (February-May), mean annual precipitation of 1300-1500 mm, and mean
annual temperatures of 21-28oC (Doucet, 2003).
CEB started commercial logging in the region after obtaining 616,700 ha
concession in 1987; the first management plan was developed in 2000. Forest classified
as High Conservation Value Forests (HCVFs) are demarcated in the concession but
their areas vary with the principal value considered (i.e., biodiversity, ecological
services, traditional uses, or cultural identity, etc.; CEB, 2007). Of this area, 15,737 ha
are allocated solely for conservation objectives and 18,588 for non-timber forest
products and other services (http://info.fsc.org). CEB operates a sawmill in the forest
concession with a monthly capacity of 1,000 m3. The concession was FSC certified by
Bureau Veritas in 2008 after starting to implement RIL practices in 2007 when its
69
management plan was approved by the Gabonese Ministry of Waters and Forests. CEB
harvests a total of 41 species, but A. klaineana, constitutes about 80% of the annual
average production of 200,000 m3 (CEB 2007). 4 tree species, among withi 3 timber
species, are protected for their biodiversity values and were flagged prior to harvest.
Until the governmental log export ban in 2010, most of the logs were exported; since
then, annual harvests were reduced to 160 000 m3 and the roundwood is processed
locally (G. Tokpa, 2010 pers. comm.).
We established permanent plots (see below for details) prior to logging in April-
August 2010 in a previously unlogged ha area separated from Ivindo National Park with
a 5-km width buffer zone (Figure 3-1A). None of CEB’s HCVFs or other areas set aside
from logging are located within the study area, which was logged in October-December
2010 by trained fellers using Stihl 780 chainsaws with log yarding along pre-planned
skid trails with a Caterpillar 535 crawler-tractor. Logging damage was assessed in
January-February 2011.
3.2.1.2 Conventionally logged site (SEEF-CL)
The conventionally logged study site was a previously unlogged 200-ha block in
the 48,000 ha concession of SEEF in Milolé, southeastern Gabon (Figure 3-1B). The
concession is bordered on the west by CEB and on the south by CWG. As in the CEB-
FSC site, the forest is dominated by A. klaineana, J. pellegriniana and S. mannii and the
fauna, soils, and climate are likewise similar (Martin et al., 1981).
The logging system used by SEEF-CL since 1998 employs tree finders who
search harvestable trees, principally of A. klaineana (90% of the timber species
harvested from the study area). Once found, the trees are measured, assessed for
quality, and numbered but none of the protected timber species was flagged before
70
harvest. Tree diameter, estimated height to the first branch, stem quality, and
approximate location are recorded by the team leader working without the aid of
surveying equipment or a reference map. Access routes to the trees to be harvested are
marked by tree finders with sticks notched to indicate the number of trees to be found
and slashed with machetes. SEEF possesses a plywood mill that processes only logs of
A. klaineana. The 2010 ban in logs export apparently did not affect SEEF’s harvests.
Trees were felled using chainsaws (Stihl 070 and MS 880) and yarded with crawler-
tractors (CAT 525) in January-February 2011. Logging damage was assessed in March
2011. Although CL practices are generally followed, the SEEF-CL fellers were trained in
directional felling techniques by a feller from CEB, which means there was some
potential for spillage of the certification treatment. Also, while the field study was
underway, SEEF was reportedly in the process of developing a forest management plan
as a step towards forest certification (T. Ricordeau 2009, pers. comm.), but that plan
had not yet been implemented.
3.2.2 Tree Measurements
Prior to logging we established twenty and twelve permanent 200 x 50 m (1 ha)
plots at random locations in the CEB-FSC and SEEF-CL areas, respectively. Although
as noted earlier these are technically sub-plots and not true treatment replicates (i.e.,
they are pseudo-replicates; Hurlbert, 1984, 2004), the size and heterogeneity of the
areas at least partially mitigates this statistical design deficiency (e.g., Heffner et al.,
1996; Oksanen, 2001; Paillet et al., 2010). Within each plot, all trees ≥ 10 cm dbh were
measured, tagged, mapped, classified according to stem quality and crown position
(suppressed, side-lighted, sub-dominant, co-dominant, dominant; see Hall et al., 2003),
and assessed for the presence or absence of lianas. Trees were identified to species
71
based on vegetative characteristics. Voucher specimens were collected from trees that
could not be identified in the field for species determination at the herbarium in
Libreville.
3.2.3 Above-Ground Biomass Estimates
Total AGB of trees in each plot was estimated using Chave et al.’s (2005)
allometric equation that is based on the destructive harvest of 1501 tropical trees 5-156
cm dbh:
ln (AGB) = -1.576 + 2.179* ln (D) + 0.198*(ln(D))2-0.0272*(ln(D))3 + 1.036*ln(ρ)
where D = dbh and ρ = wood density. For trees > 156 cm dbh, AGB was
extrapolated using this same equation. Species-specific wood density estimates were
used when available (Zanne et al., 2009). When there were multiple wood density
values for the same taxon, we used the arithmetic mean; in the absence of species-
specific wood density data, we used generic or familial means; and, if none of these
values were available, we used the plot mean weighted by the number of trees of each
species.
3.2.4 Damage Assessment
Logging damage was assessed in the 1-ha sample plots using methods well-
established in the literature (e.g., Johns et al. 1996; Whitman et al., 1997; reviewed by
Putz et al., 2008b). Crown damage was recorded as severe (> 66% crown loss),
moderate (33-66%), or minor (< 33%). Bole damage was recorded as severe (broken
bole), moderate (> 100 cm2 of bark removed), or minor (< 100 cm2 of bark removed).
Uprooted trees were recorded as such. Root damage was recorded as major (>10% of
surface roots injured) and minor (< 10% of surface roots injured). Crown, bole, and root
72
damage were attributed to felling and/or skidding, log deck construction, or road
construction.
Canopy gaps in the 1-ha plots that were opened by felling were measured in the
cardinal and sub-cardinal directions from a central point based on Brokaw’s (1982)
definition (gaps = forest canopy openings that are >20 m2 and extend down through all
foliage levels to an average height of < 2 m above ground). Log landings and machine
maneuvering zones in felling gaps as well as the ground surface disturbed on skid trails
and other skidder activities were also mapped and measured but in the entire 508 ha
(FSC) and 200 ha (CL) study areas. Skid trails were classified on the basis of the
number of passes by the skidders (which always yarded single logs): primary >10;
secondary 2-10; and, tertiary 1. Skid trail widths were measured every 10 m. The
number of skid trails crossing streams was determined using maps created based on
field measurements and existed hydrologic map of each study site. Avoidable stream
crossing was determined through the analysis of the processed map based on the
proximity of tree stumps to streams and relative to the skid trail network. Maps were
made using ArcGIS Desktop10.
Damage to trees >10 cm dbh and soils that resulted from road construction and
maintenance (primary = main road used all year; secondary = used only during the 4-
month harvest season) was assessed in 5 x 100 m (0.05 ha) plots at 100 m intervals on
each side of the roads in the 508 ha (FSC) and 200 ha (CL) study areas. Road widths
were measured at 20 m intervals.
73
3.2.5 Logging-Induced Conversion of Above-Ground Biomass (AGB) into Necromass
To estimate the amount of AGB converted into necromass (and therefore
committed as emissions) during logging operations, we estimated the mass of each
harvested and destroyed (i.e., uprooted or bole snapped) tree as described above.
Biomass left in the forest in branches and leaves (i.e., tree crowns) was estimated from
the difference between total estimated biomass of the entire trees and the portions in
stumps and logs that were extracted or abandoned. Stump volumes were measured
and included in necromass estimates using the same species-specific wood density
expansion factors.
We estimate the committed emissions of carbon due to logging in two ways, but
neither considers decomposition rates and instead assumes that the emissions are
immediate and both assume that the harvested wood did not count for the retained
carbon. For all cases, emissions are discussed as if they were instantaneous rather
than by accounting for differential rates of decomposition. Unfortunately, data on post-
logging mortality rates of damaged trees that could support or refute this assumption
are scarce in the literature and apparently absent for Congo Basin forests. In
recognition of the fact that some damaged trees will recover completely while others
develop heartrots and stem hollows, our estimates of committed emissions that include
both destroyed and harvested trees should be interpreted with caution.
3.2.6 Data Analysis
3.2.6.1 Tree species richness
Tree species richness in the 1 ha plots before and after logging were compared
using EstimateS 8.2.0 (Colwell, 2006) with sample-based rarefaction to compute
74
expected species accumulation curves. A randomization without replacement was run
to compute the richness estimators and diversity indexes (e.g., Fisher’s alpha). We
compared species richness before and after logging at each site with paired t-tests of
the outputs from 100 randomized iterations.
3.2.6.2 Statistical analyses and modes of data presentation
Analysis of variance (ANOVA) was used to test for differences in numbers of trees
damaged per unit length of skid trail and road by order. In these tests, the units of
replication are skid trails and roads. If overall differences were detected, Tukey’s Honest
Difference tests were employed to make pairwise comparisons among means.
Further analyses were conducted to compare logging-induced changes in species
density, stem density, basal area, and AGB in the certified and uncertified concessions.
Analysis of covariance (ANCOVA) was used to determine the relationships between the
dbh of the felled trees and felling gap area and the number of damaged trees in both
sites. To detect changes in tree species abundance as determinant of shifting in plot
location before and after logging at each site, we used a canonical correspondence
analysis (CCA). CCA is a multivariate method that elucidates the relationships between
biological assemblages of species and their environment (Ter Braak, 1986; Borcard,
Gillet, and Legendre, 2011). The solution of the CCA was displayed in an ordination
diagram with plots represented by points. We used vegan package that contains
common ordination methods of principal component analysis and correspondence
analysis. The vegan package provides tools for descriptive community ecology and it
has most basic functions of diversity analysis (Oksanen et al.,2011). All analyses were
conducted with version 2.10.1 R (R Development Core Team, 2010).
75
Due to a slight difference in logging intensity between the two concessions, results
are presented on a per hectare basis as well as by the number of trees and volumes of
timber harvested.
3.3 Results
3.3.1 Forest Structure and Tree Species Richness
3.3.1.1 Forest structure
Prior to logging a total of 6083 trees ≥ 10 cm dbh were inventoried in the twenty 1-
ha FSC plots and 4649 trees ≥ 10 cm dbh in the twelve 1-ha CL plots (Table 3-1).
Stem density (Figure 3-2), basal area, and above-ground biomass were all lower
at the FSC site (Table 3-2).
Liana-infested trees were relatively uncommon in both sites, especially among the
large trees; only 3.8 and 1.9% of trees > 60 cm dbh were liana infested in the FSC and
CL plots, respectively (Table 3-2).
3.3.1.2 Changes in tree species richness and composition
Prior to logging the FSC site, the 20 1-ha plots supported 298 species of trees ≥
10 cm dbh representing 169 genera and 45 families (40 unidentified morphospecies
representing 13% of inventoried trees are included). In the 12 1-ha plots at the CL site,
inventoried trees represented 47 families, 140 genera, and 201 species including 22
morphospecies (i.e., 11% of the total inventoried trees; Table 3-2). Of the tree species
encountered in the FSC plots, Erypetalum tessmannii (Caesalpiniaceae) represented
11.7% of the total trees inventoried. Neochevalierodendron stephanii (Caesalpiniaceae)
7.8%, Dialium angolense (Caesalpiniaceae) 4.7%, Scorodolphloeus zenkeri
(Caesalpiniaceae) 4.5%, and Santiria trimera (Burseraceae) 4.5%; A. klaineana
represented only 0.2% of the total trees (Appendix C). Timber species protected from
76
logging due to their high biodiversity values, namely Coula edulis (Olacaceae),
Dacryodes buettneri (Burseraceae), and Baillonella toxisperma (Sapotaceae),
represented 1.8%, 0.1%, and 0.3% of the total number trees, respectively, all
representing a density of 9.4 stems ha-1 (Table 3-1). Caesalpiniaceae dominated at the
family level with 51.7% of the trees followed by Burseraceae 6.6%, Ebenaceae 5.2%,
and Euphorbiaceae 4.4% (Appendix D).
In the CL plots, S. trimera represented 13.5% of the total trees, S. zenkeri 13.3%,
Polyalthia suaveolens (Annonaceae) 4.7%, and Pausynisthalia macroceras (Rubiaceae)
4.0%.; A. klaineana represented only 0.9% of the trees (Appendix E). Of the protected
timber species only D. buettneri was found in the CL site and represented 1.1% of the
total trees with a density of 4.4 stems ha-1 (Table 3-1). At the family level,
Caesalpiniaceae represented 27.5% of the trees, Burseraceae 18.9%, Myristicaceae
5.3%, and Mimosaceae 2.7% (Appendix F). During logging, similar proportions of
protected trees were damaged in the FSC plots (9 of 131) and in the CL plots (4 of 53;
Chi-Squared = 0.03, p = 0.89).
Logging had no detectable short-term impact on any of the measures we used for
tree diversity in either CEB-FSC or SEEF-CL. The canonical correspondence analysis
(CCA) used to detect short-term changes in species composition due to logging based
on tree abundance showed only very slight shifts in the FSC (Figure 3-3a) and CL
(Figure 3-3b) sites. As expected, based on the scores of the first component of the
CCA, plots from which no trees were harvested and that suffered no damage from
skidding did not undergo any shifts in the FSC and CL sites. Nevertheless, the mean
Euclidian distances between plots prior to and after logging was greater for the CL site
77
(Figure 3-3d; mean = 0.07, s.d. = 0.04, n = 12) than for the FSC site (Figure 3-3c; mean
= 0.04, s.d. = 0.03, n = 20; t = 2.25, p = 0.03).
3.3.2 Above-Ground Biomass
Pre-logging total AGB estimates for the 20 1-ha FSC plots were 215.2–636.8 Mg
ha-1 with a mean of 380.0 ± 86.78 Mg ha-1, while for the 12 1-ha CL plots, they were
306.7–547.6 Mg ha-1 with a mean of 387.3 ± 81.01 Mg ha-1 (Table 3-1). Tree species
contributed to the AGB in somewhat different proportions in the two concessions. The
FSC plots were dominated by Caesalpiniaceae including Julbernardia. pellegriniana,
which represented 11.7% of the total AGB, E. tessmannii (7.1%), Hymenostegia
pellegrinii (6.7%), and Gilletiodendron pierreanum (6.2%); Scyphocephalium mannii
(Myristicaceae) represented 5.2%. Among the commercial timber-producing species, A.
klaineana represented 1.7% of the total AGB, Pterocarpus soyauxii (Fabaceae) 0.9%,
Pycnanthus angolensis (Myristicaceae) 0.5%, Staudtia gabonensis (Myristicaceae)
0.4%, Canarium schweinfurthii (Burseraceae) 0.3%, Nesogordonia papaverifera
(Malvaceae) 0.3%, and Nauclea diderrichii (Rubiaceae) 0.2% (Appendix C). For
commercial timbers in the CL plots, S. mannii represented 12.6% of the total AGB, A.
klaineana 9.5%, S. zenkeri 9.3%, J. pellegriniana 7.0%, and S. trimera 4.3% (Appendix
E). Overall, Caesalpiniaceae contributed 55.6% and 31.0% of the total AGB in the FSC
and CL sites, respectively, while Mimosaceae represented 7.1% and 6.6%,
Myristicaceae 6.6% and 15.4%, and Burseraceae 5.5% and 19.2% (Appendices D and
F).
3.3.3 Logging Intensities
A total of 198 trees ≥ 60 cm dbh (range: 60–148 cm dbh; 93.7 ± 19.57 (mean ±
s.d)) were harvested at an intensity of 0.39 trees ha-1 (5.67 m3 ha-1) from the 508-ha
78
FSC site and 153 trees ≥ 58 cm dbh (range: 58–148 cm dbh; 100.3 ± 21.32) with an
intensity of 0.76 trees ha-1 (11.38 m3 ha-1) from the 200-ha CL site (Table 3-3). The
mean AGB in logs harvested from the FSC and CL sites were, respectively, 5.97 Mg ha-
1 and 6.93 Mg ha-1. The lower logging intensity at the FSC site was reportedly related to
the log export ban (G. Topka 2010 pers. comm.). Although High Conservation Value
Forests were demarcated and protected in the FSC concession, none were in the 508
study area in which there were also no streamside buffer zones in which logging was
prohibited. Mean stump heights of harvested trees in the FSC and CL sites did not differ
(Table 3-4). When expressed per cubic meter, biomass in tree stumps at the FSC site
was slightly higher but in contrast, biomass left in tree crowns at the FSC site was
substantially lower than at the CL site. Therefore, AGB converted to necromass from
felled trees was lower at the FSC site than the CL site expressed either on a per
hectare basis or in relation to the timber volume harvested (Table 3-4).
3.3.4 Logging Damage
Of the 20 1-ha FSC plots, the 6 that were harvested yielded a total of 10 trees,
while 16 trees were harvested from 9 of the 12 1-ha CL plots. At the FSC site, damage
to trees in plots from which none were harvested resulted from felling of nearby trees (1
plot), skidding (6 plots), felling and skidding (2 plots), skidding and road-related activities
(1 plot), and felling, skidding, and road-related activities (1 plot); 3 plots were untouched
by logging. Among the 3 CL plots from which no timber was extracted, one was
damaged by both the felling of trees outside the plots and skidding, another suffered
only skidding damage, and one was untouched by logging.
79
3.3.4.1 Felling damage
At the FSC site, the felling that occurred in 6 of the 20 1-ha plots damaged on
average 9.1 trees ≥ 10 cm dbh per felled tree. In contrast, 20.9 trees ≥ 10 cm were
damaged per felled tree at the CL site (Table 3-5). Of the damaged trees, 19% in the
FSC site and 25% in the CL site suffered severe crown damage, while some suffered
severe bole damage (i.e., broken, FSC = 13%, CL = 8%) and other were uprooted (FSC
= 20%; CL = 8%; Figure 3-4a and 3-4b). The impact of felling on AGB was lower at the
FSC site than the CL site but somewhat greater when expressed per timber volume
extracted (Table 3-5). An ANCOVA revealed that while the number of trees damaged
due to felling increased with the diameter (dbh) of the felled trees at both sites, the
number of damaged trees per felled tree was greater at the CL site (Figure 3-5a).
3.3.4.2 Skidding damage
Skid trails covered 1.6% of the 508-ha FSC site and 4.5% of the 200-ha of the CL
site (Table 3-6; Figure 3-6). The number of skid trails that crossed streams was also
lower in the FSC than CL sites, with 1.2 and 2 crossings per 100 ha that seemed
avoidable, respectively. Skidding operations damaged fewer trees per logged hectare in
the 20 1-ha FSC plots than in the 12 1-ha CL plots but trees damaged per timber
volume extracted did not differ (Table 3-5). Of the damaged trees at the FSC site, 31%
were destroyed whereas 23% were destroyed in the CL site. Root damage was suffered
by 4% and 7% of the trees at the FSC and CL sites, respectively (Figure 3-4a and 3-
4b).
Skid trails of all orders were consistently narrower in the FSC than CL concession
(primary: 4.2 vs. 5.8; secondary: 3.6 vs. 5.1 m; and, tertiary: 3.4 vs. 5.1 m; Table 3-6). In
the FSC concession the number of damaged trees per length of skid trail differed
80
between primary (0.08 trees m-1), secondary (0.08 trees m-1), and tertiary (0.05 trees m-
1) skid trails (F = 4.31; p = 0.013), but damage per length of primary and secondary skid
trails were equal. At the CL site, 0.18 trees m-1 were damaged per meter of both primary
and secondary skid trails, but only 0.15 trees m-1 were damaged along tertiary trails.
The impact of skidding on AGB was 6.2 Mg ha-1 lower at the FSC site than the CL site
but did not differ when expressed on the basis of timber volume harvested (Table 3-5).
3.3.4.3 Felling gaps and ground area disturbed
In the 20 1-ha FSC plots, the 12 felling gaps were 270–1794 m2 with a mean gap
size per felled tree of 787.4 m2. In the 12 1-ha CL plots, the 13 felling gaps were 293–
1565 m2 with a mean gap per felled tree of 751.1 m2 (Table 3-5). Felling gaps increased
with the size if the felled tree in both areas (Figure 3-5a), but were relatively larger in the
FSC than in the CL sites (Figure 3-5b).
Log landings covered 0.12% of the ground surface in the 508-ha FSC site and
0.13% of the 200-ha CL site, while machine maneuvering disturbed 0.31% and 0.25%
of the ground area in the FSC and CL sites, respectively (Table 3-6).
3.3.4.4 Above-ground biomass converted into necromass
AGB of destroyed trees per logged hectare did not differ but was 0.17 Mg m-3
higher in the 20 1-ha FCS plots than in the 12 1-ha CL plots when expressed per timber
volume extracted (Table 3-4). In contrast, AGB of damaged trees was 15 Mg ha-1 lower
at the FSC site than at the CL site but did not differ when expressed per timber
extracted. Considering all together, the AGB of damaged, destroyed, and harvested
trees averaged 28.8 Mg ha-1 (5.1 Mg m-3) at the FSC site and 54.7 Mg ha-1 (4.7 Mg m-3)
at the CL site (Table 3-5). Overall, AGB converted into necromass (i.e., AGB of
harvested plus completely destroyed trees) was less in the FSC site (11.2 Mg ha-1; 1.99
81
Mg m-3 or 2.9% of the pre-logging total biomass) than in the CL site (24.6 Mg ha-1; 2.3
Mg m-3 or 6.3% of the pre-logging total converted into necromass; Table 3-5).
3.3.4.5 Damage due to road construction and maintenance
The unique measured primary road covered 0.8% (75.7 m2 ha-1) and two
secondary roads covered 0.7% (33.5 m2 ha-1) of the ground surface at the FSC site
(Table 3-6). In the CL study area there were no primary roads but two secondary roads
covered 5.4% (539.7 m2 ha-1) of the ground surface and none crossed permanent plots
(Table 3-6). Secondary roads in the CL site were substantially wider (66.6 m) than in the
FSC site (15.1 m; Table 3-6 and 3-7; Figure 7). In the 20 1-ha FSC plots, road
construction and maintenance damaged and destroyed a total of 30 trees ≥ 10 cm dbh
in (Table 3-5) of which, 9% suffered severe crown damage and 55% were uprooted.
None of the 12 1-ha CL plots suffered damage from road construction and
maintenance. Whether expressed per tree or per cubic meter of timber harvested, roads
in the CL area, which were wider, covered nearly an order of magnitude more ground
than the FSC area; the number of trees damaged per length of road differed among the
primary (0.16 trees m-1) and two secondary roads (0.18 and 0.14 trees m-1; F = 8.7; df =
610; p <0.001). At the CL site, none of the 12 1-ha plots suffered road-related damage
but in the entire 200-ha study area, the two secondary road construction damaged and
destroyed an average of 0.49 trees m-1 (Table 3-6).
3.4 Discussion and Conclusion
3.4.1 Impacts of Certified or Conventional Logging on Forest Structure and Composition
Based on the FSC Principle # 6, forest management shall conserve biological
diversity and its associated values by minimizing the deleterious impacts of logging. In
82
both our CL and FSC study sites, which share many environmental characteristics,
logging intensities were light and had no apparent short-term effects on tree species
richness or composition. We expect that changes are forthcoming, especially in the
most severely damaged plots, but this will require continued monitoring.
3.4.2 Above-Ground Biomass Impacts and Conversion into Necromass
Due to the low harvest intensities, the immediate losses of AGB were only, 2.9%
and 6.3% at the FSC and CL sites, respectively. Logging intensity at the uncertified site
was nearly twice that at the uncertified site, but the total impacts on AGB were also
higher at the uncertified when expressed per cubic meter harvested. Lower logging
intensity in the FSC-certified concession was apparently related to the manager’s
response to the 2010 log export ban and not due to implementation of RIL techniques or
compliance with certification standards. In particular, in the 508 ha study area in the
certified concession, there was no foregone timber due to harvesting restrictions on
steep slopes, in stream buffer zones, or in demarcated conservation areas.
The harvest intensity from the CL concession was higher than from the FSC
concession even though more species were harvested from the latter. From the CL site
90% of the volume harvested was from A. klaineana, all of which was processed in the
concession’s plywood mill and was thus not affected by the log export ban. The higher
logging intensity was thus due mostly related to the presence of higher standing stocks
of A. klaineana timber in the CL area. As a consequence, AGB converted into
necromass (i.e., committed emissions) was lower at the FSC than the CL site when
expressed per logged hectare, but the difference persisted when expressed per cubic
meter of timber harvested. Note that our necromass estimates include the biomass in
83
the extracted logs, a portion will be converted into end forest products that store carbon
for long periods.
3.4.3 Logging Damage
3.4.3.1 Damage due to felling
Felling caused more collateral damage to residual trees at the CL site than the
FSC site for a number of reasons. First to all, prior to the harvest, stem densities were
higher at the CL site, particularly among small trees. Not surprising, therefore, more
trees suffered severe crown and bole damage than at the FSC site. Determination of
the fates of damaged trees awaits future plot measurements, but if the responses
resemble those reported in other tropical forests (e.g., Durrieu de Madron et al., 2000
and 2011; Jonkers, 2000; Sist and Ferreira, 2007; Blanc et al., 2009; Mazzei et al.,
2010; Hawthorne et al., 2012), then carbon emissions from the CL site will be much
greater than from the FSC site where fewer trees were damaged.
The small difference between the FSC and CL sites in biomass impacts of felling
may be a result of the training in directional felling received by workers in both
concessions. This might also explain the lack of a difference in stump heights but do not
explain why so much more ostensibly marketable timber was wasted in the CL site due
to poor bucking.
3.4.3.2 Damage due to skidding
The observed reductions in tree and soil damage due to skidding in the FSC site
as compared to the CL are due mostly to skid trails being narrowed, presumably due to
pre-felling skid trail planning, training, and supervision of skidder drivers at the FSC site
(Table 3-5; Figure 3-6). While there was no difference between the CL and FSC
concessions in skid trail length per harvested tree, because skid trails in the CL area
84
were 40% wider, they covered 38 and 36% more ground per tree and per cubic meter
harvested, respectively. The lack of a difference in the number of avoidable stream
crossings per unit area in the two concessions (Table 3-6; Figure 6) may have resulted
from the failure of the certified concession to demarcate and protect streamside buffer
zones, as called for in the 2008 and 2011 audit reports by Bureau Veritas
(http://info.fsc.org).
It is also important to note that although skidder drivers at the CL site received no
formal training in RIL and had no maps of planned skid trails to follow, the tree finders
that preceded them slashed paths to the harvestable trees that avoided steep slopes
and large trees and were otherwise selected to facilitate efficient log yarding.
Consequently, there was no difference in biomass of trees damaged and destroyed by
skidding when expressed on the basis of the volume of timber harvested.
3.4.3.3 Damage due to road construction and maintenance
Compared to the FSC concession, roads in the CL concession were wider, longer,
and therefore caused more damage whether expressed per unit area, number of trees
harvested, or harvested timber volumes (Table 3-7). The roadside clearings in the CL
site that resulted in so much collateral stand damage were opened to facilitate the
drying of the road surface (Figure 3-7). Roads in the certified concession in contrast,
trafficability of the narrow logging, which were designed and constructed as
recommended for the Congo Basin by the FAO (FAO, 2003), was maintained by
assuring proper drainage and widespread surfacing with laterite gravel.
85
3.4.4 Impacts of Forest Management Certification on Biodiversity and Carbon Retention
Although based on a comparison of only one certified and one un-certified
concession, the results of this study support the contention that forest management
certification, leads to reduced stand damage and reduced carbon emissions. The lack of
any short-term impact of logging on tree species richness and composition in either
concession is apparently due to the relatively low logging intensities in both. These
impacts may become apparent with longer-term studies and investigations of other taxa,
but it seems likely that if there are impacts. Further studies are needed to explore the
likely variability in the impacts of certified logging on AGB, biodiversity, and other
ecosystem services. It will be also necessary to investigate the natural regeneration
response of the most harvested timber species, A. klaineana, in both FSC and CL
concessions.
Despite its shortcomings, this study constitutes one of the first examples of a
counterfactual study comparing certified forest with uncertified forest in the tropics as a
whole and the Congo Basin in particular. As such, it provides some of the data needed
to inform decision-making about investments in forest management and certification.
86
Table 3-1. Biophysical conditions and pre-logging tree community characteristics in the CEB-FSC and SEEF-CL logging concessions in Gabon.
Variables CEB-FSC SEEF-CL Statistical Contrasts
Concession area 616,700 ha 477,033 ha Ownership French/Swiss Gabonese Administrative status/Year of acquisition
CFAD/2000 a PFA/2004 b
Management plan approved/date Yes/2000 No/In progress Certification/date Yes/2008; Bureau Veritas No Workers 530 About 200 Vegetation type Mixed with A. klaineana
dominance Mixed with A. klaineana dominance
Slopes in logged areas (%) 8-15 10-15 Mean altitude 300–500 m 300–600 m Soil type Oxisols on megalithic rock Oxisols on megalithic
rock
Rainfall 1300–1500 mm 1300–1500 mm Area study site (ha) 508 200 Buffer zone Yes/but no in study area No High Conservation Value Forest Yes/ but no in study area No Road surface Laterite gravel, hump well
drained Laterite on slippery area
a CFAD = Concession Forestière d’Aménagement Durable (Forest Concession Under Sustainable Management) b PFA = Permis Forestier Associé (Associated Forest Permit).
87
Table 3-1 continued
Variables CEB-FSC SEEF-CL Statistical Contrasts
Number of 1-ha permanents plots 20 12 Total trees ≥10 cm dbh 6083 4649 Number of harvestable timber species 41 21 Stem density (stems ha-1; mean ±1s.d) 304.2 ± 30.3 387.4 ± 36.8 t = 6.6; p <0.001 Stem density of A. klaineana (stems ha-1) 0.55 ± 0.18 3.33 ± 0.94 t = 2.4; p = 0.03 Stem density of A. klaineana ≥ 80 cm dbh (stems ha-1)
0.40 ± 0.14 2.67 ± 0.62 t = 2.7; p = 0.01
Density protected tree species (stem ha-1) 6.55 ± 3.40a 4.4 ± 2.71b t = 1.8; p = 0.03 AGB protected trees (Mg ha-1) 16.5 ± 6.1 11.4 ± 3.3 t = 1.02; p = 0.04 Mean dbh 27.0 ± 2.1 cm 25.9 ± 2.0 cm t = 1.4; p = 0.17 Maximum dbh 196 cm 175 cm t = 0.1; p = 0.92 Basal area (m2 ha-1) 26.7 ± 3.8 31.3 ± 6.2 t = 2.3; p = 0.03 Above-ground biomass (Mg ha-1) 380.0 ± 86.8 387.3 ± 81.0 t = 0.2; p = 0.81 Species richness 298 201 t = 3.8; p <0.001 Rarefied species (Fisher’s alpha) 57.8 42.8 t = 7.8; p <0.001
a Baillonella toxisperma, Sapotaceae (n = 4) ; Coula edulis, Olacaceae (n = 107) ; Dacryodes buettneri, Burseraceae (n = 20). b Dacryodes buettneri, Burseraceae (n = 53).
88
Table 3-2. Pre-logging tree community structures in the CEB-FSC (508 ha) and SEEF-CL (200 ha) concessions in Gabon. Crown Classes (%)
Site DBH (cm)
# Spp.
# Families
Density (Ind./ha)
BA (m2/ha)
AGB (Mg/ha)
Understory Midstory Canopy Emergent Liana loads (%) a
CEB-FSC 10 207 38 151.8 2.4 17.0 42.9 7.1 - - 24.4
20 148 35 60.9 2.9 28.5 14.6 5.4 - - 12.5
30 130 33 37.5 3.5 42.6 0.1 12.3 - - 6.7
40 92 29 20.7 3.2 45.3 - 6.8 - - 4.0
50 76 24 12.8 3.0 45.7 - 1.1 2.8 0.3 3.1
60 64 21 7.8 2.5 41.5 - 0.6 1.7 0.2 1.7
70 41 15 4.9 2.1 35.5 - 0.5 1.0 0.1 0.7
80 32 12 3.1 1.7 29.7 - 0.3 0.6 0.1 0.4
90 15 9 1.4 0.9 17.0 - 0.2 0.3 0.1 0.3
100 11 5 1.1 0.9 15.2 - 0.1 - 0.2 0.3
110+ 15 7 2.2 3.5 62.1 - 0.1 - 0.6 0.4
Total
298 45 304.2 26.7 380.0
SEEF-CL 10 147 40 202.2 3.3 23.8 49.8 2.4 - - 28.4
20 108 34 81.4 3.9 37.0 10.3 10.7 - - 10.5
30 90 32 43.8 4.1 47.7 1.4 9.9 - - 4.8
40 79 28 21.8 3.4 44.3 0.1 5.5 - - 2.2
50 56 23 15.3 3.6 48.8 0.1 1.7 2.2 - 1.6
60 41 19 8.0 2.6 38.2 0.04 0.5 1.5 0.02 0.7
70 22 9 6.2 2.7 38.9 0.02 0.3 1.3 0.02 0.5
80 19 13 3.8 2.2 33.1 0.02 0.1 0.9 - 0.3
90 10 5 1.5 1.1 16.8 - 0.04 0.3 - 0.1
100 5 3 0.7 0.6 7.5 - - 0.2 - 0.04
110+ 8 5 2.8 3.8 51.0 - 0.02 0.7 0.02 0.3
Total 201 47 387.4 31.3 387.3
a Percent of trees infested by liana in each diameter class
89
Table 3-3. Number and volume of timber trees harvested in the 508 ha CEB-FSC and 200 ha SEEF-CL study areas in
Gabon. CEB-FSC SEEF-CL
Species # Trees Harvested
Total Volume (m3)
Volume Extracted (m3 ha-1)
# Trees Harvested
Total Volume (m3)
Volume Extracted (m3 ha-1)
Aucoumea klaineana 153 2151.2 4.23 137 2136.9 10.68
Cylicodiscus gabunensis 8 176.8 0.35 13 109.9 0.55
Distemonanthus benthamianus 2 15.3 0.03 - - -
Entandrophragma cylindricum 3 85.2 0.17 1 10.1 0.05
Entandrophragma utile 1 39.9 0.08 - - -
Erythophyllum ivorensis 1 8.6 0.02 - - -
Guibourtia tessmannii 3 45.5 0.09 - - -
Khaya anthotheca 1 15.17 0.03 - - -
Oxystigma oxyphyllum 2 57.7 0.11 - - -
Pterocarpus soyauxii 21 257.6 0.51 2 18.6 0.09
Swartzia fistuloides 3 24.8 0.05 - - -
Total 198 2877.7 5.66 153 2275.5 11.38
90
Table 3-4. Logging impacts in the entire 508-ha CEB-FSC and 200-ha SEEF-CL study areas in Gabon.
Variables CEB-FSC SEEF-CL Statistical Constrast
Number of trees harvested 198 153 Number of felled and abandoned trees 1 a 4 b Maximum dbh of harvested trees (cm) 148 148 Logging intensity (trees ha-1) 0.39 0.76 Mean dbh of harvested trees (cm; mean ± 1 s.d) 93.7 ± 19.6 100.3 ± 21.3 t = 2.9; p <0.01 Total volume harvested (m3 ha-1) 5.7 11.4 t = 4.6; p <0.001 Total AGB of harvested trees (Mg ha-1) 6.0 6.9 t = 9.1; p <0.001 Total AGB extracted in logs (Mg ha-1) 5.5 4.8 t = 12.2; p <0.001 Total AGB due to waste (Mg ha-1) 0.5 2.1 t = 7.3; p <0.001 Stump height (m; mean ± 1 s.d) 0.89 ± 0.2 0.92 ± 0.2 t = 1.3; p = 0.21 Total AGB of stumps (Mg ha-1) 0.11 0.15 t = 5.9; p <0.001 Total AGB of stumps (Mg m-3) 0.02 0.01 t = 5.9; p <0.001 Total AGB of crowns of felled trees (Mg ha-1) 0.39 1.99 t = 7.8; p <0.001 Total AGB of crowns of felled trees (Mg m-3) 0.07 0.18 t = 8.3; p <0.001
AGB to necromass from felled trees (Mg ha-1) c 1.00 4.29 t = 7.3; p <0.001
AGB to necromass from felled trees (Mg m-3) 0.19 0.40 t = 7.7; p <0.001 a Hollow tree b Two trees were hollow and two broke upon felling c Calculated as the sum of AGB in stumps, crowns, and waste (i.e., the difference between AGB of harvested trees and extracted in logs); not counting for AGB in logs.
91
Table 3-5. Logging impacts in permanent plots in the 508-ha CEB-FSC and 200-ha SEEF-CL study areas in Gabon
Variables CEB-FSC SEEF-CL Statistical Contrasts
Number of 1-ha plots 20 12 Number of trees harvested in plots 10 16 Maximum dbh of harvested trees (cm) 148 147 Mean dbh of harvested trees (cm; mean ± 1 s.d) 113.9 ± 29.72 113.4 ± 20.85 t = 0.04; p = 0.96 Number of felling gaps 12 a 13 b Felling gap area per tree harvested (m2) 787.4 751.1 t = 0.5; p = 0.63 Trees damaged & destroyed per tree harvested (trees tree-1) 9.1 20.9 t = 2.2; p = 0.02 Trees damaged & destroyed by skidding (trees ha-1) 9.4 (2.3%) c 19.8 (4.1%) c t = 2.1; p = 0.03 Trees damaged & destroyed by skidding (trees m-3) 1.6 (2.3%) c 1.7 (4.1%) c t = 0.2; p = 0.87 Total trees damaged & destroyed by road-related 30 - Total damaged & destroyed protected tree species (trees m-3) 0.05 0.04 AGB trees damaged & destroyed by felling (Mg ha-1) 13.1 24.2 t = 1.6; p = 0.06 AGB trees damaged & destroyed by felling (Mg m-3) 2.3 2.1 t = 0.2; p = 0.41 AGB trees damaged & destroyed by skidding (Mg ha-1) 5.9 12.1 t = 1.5; p = 0.07 AGB trees damaged & destroyed by skidding (Mg m-3) 1.05 1.06 t = 1.2; p = 0.49 AGB trees damaged & destroyed by road-related (Mg ha-1) 1.5 - AGB trees damaged & destroyed by road-related (Mg m-3) 0.3 - AGB damaged & destroyed protected timber species (Mg m-3) 0.3 0.1 t = 0.9; p = 0.19
a Two canopy gaps were opened by trees harvested outside the plots b Resulted from multiple-tree gaps (i.e., overlapping gaps) c As percent of total trees damaged and destroyed by skidding
92
Table 3-5. continued
Variables CEB-FSC SEEF-CL Statistical Contrasts
AGB destroyed trees (Mg ha-1) 1.9 1.8 t = 0.2; p = 0.43
AGB destroyed trees (Mg m-3) 0.32 0.15 t = 2.1; p = 0.05
AGB total trees damaged (Mg ha-1) 18.7 33.7 t = 1.9; p = 0.04 AGB total trees damaged (Mg m-3) 3.3 2.9 t = 0.4; p = 0.73 AGB trees destroyed & damaged (Mg ha-1) 20.5 35.5 t = 1.8; p = 0.04 AGB trees destroyed & damaged (Mg m-3) 3.6 3.1 t = 0.5; p = 0.31 AGB destroyed & harvested trees (Mg ha-1) 10.2 20.3 t = 1.6; p = 0.06 AGB destroyed & harvested trees (Mg m-3) 1.8 1.9 t = 0.1; p = 0.49 AGB damaged & harvested trees (Mg ha-1) 27.0 52.3 t = 2.0; p = 0.03 AGB damaged & harvested trees (Mg m-3) 4.7 4.6 t = 0.1; p = 0.92 AGB damaged, destroyed, & harvested trees (Mg ha-1) 28.8 54.0 t = 1.9; p = 0.03 AGB damaged, destroyed, & harvested trees (Mg m-3) 5.1 4.7 t = 0.2; p = 0.41 Total AGB converted into necromass (Mg ha-1) d 11.2 24.6 t = 4.9; p <0.001 Total AGB converted into necromass (Mg m-3) d 1.99 2.30 t = 1.9; p = 0.04 Post-harvest AGB (Mg ha-1) 353.0 ± 72.6 335.0 ± 78.0 t = 0.7; p = 0.52 AGB percent reduction (%) 2.9 6.3 d Determined as the sum of necromass from harvested (c in Table 3-4) and destroyed trees; including tree crowns and stumps and not trees that were damaged and may survive.
93
Table 3-6. Ground area disturbed and logging waste in the 508-ha CEB-FSC and 200-ha SEEF-CL study areas in Gabon. Numbers in parentheses represent the proportion of the total of the total area.
Variables CEB-FSC SEEF-CL Statistical Contrasts
Area in felling gaps 0.9 ha (0.2%) 1.0 ha (0.5%) t = 1.5; p = 0.06 Skid trail length (m ha-1) 15.2 28.7 t = 4.4; p = 0.04 Skid trail area (m2 ha-1) 55.1 (1.6%) 149.9 (4.5%) t = 4.5; p = 0.04 Skid trail length per harvested tree (m tree-1) 39.0 37.5 t = 0.4; p = 0.71 Skid trail length per harvested volume (m m-3) 2.7 2.7 t = 0.6; p = 0.59 Width skid trail (m; mean ± 1 s.d) 3.8 ± 0.9 5.3 ± 1.4 t = 38.7; p <0.001 Skid trail area per harvested tree (m2 tree-1) 141.4 195.9 t = 3.2; p = 0.04 Skid trail area per harvested volume (m2 m-3) 9.7 13.2 t = 3.1; p = 0.04 Skidding damage intensity (trees m-1) 0.27 0.19 t = 1.3; p = 0.31 Primary road length (m ha-1) 3.3 - Width primary road (mean; m ± 1 s.d) 22.6 ± 5.89 - Primary road area (m2 ha-1) 75.7 (0.8%) - Primary road length per harvested tree (m tree-1) 8.5 - Primary road length per harvested volume (m m-3) 0.6 - Primary road area per harvested tree (m2 tree-1) 194.1 - Primary road area per harvested volume (m2 m-3) 13.3 - Secondary road length (m ha-1) 2.2 8.1 t = 10.9; p = 0.03 Width secondary road (mean; m ±1 s.d) 15.1 ± 4.26 66.6 ± 23.27 t = 21.9; p <0.001 Secondary road area (m2 ha-1) 33.5 (0.7%) 539.7 (5.4%) t = 5.5; p = 0.05 Secondary road length per harvested tree (m tree-1) 5.8 10.6 t = 12.5; p = 0.02 Secondary road length per harvested volume (m m-3) 0.4 0.7 t = 10.7; p = 0.02 Secondary road area per harvested tree (m2 tree-1) 85.8 705.5 t = 5.6; p = 0.05 Secondary road area per harvested volume (m2 m-3) 5.9 47.4 t = 5.6; p = 0.05 Primary road damage intensity (trees m-1) 0.16 - Secondary road damage intensity (trees m-1) 0.35 0.49 t = 2.9; p = 0.21 Area covered by log landings 0.12% 0.13% t = 0.6; p = 0.59 Skid trails crossing streams per 100 ha 2.8 6.0 Number of avoidable stream crossings per 100 ha 1.2 2.0
94
Table 3-7. Roads in 508 ha in the CEB-FSC concession and 200 ha in the SEEF-CL (200 ha) concession in southeastern Gabon compared with FAO standards (FAO 2003).
Category Side Cast (±s.d) Road Bed (±s.d)
Matahari Clearing (±s.d)
Total Road Width (±s.d)
RIL Guidelines, FAO 2003 15–30 m 7–12 m 5–30 m 30–40 m
CEB-FSC Primary 20.4 m (±5.46) 6.8 m (±1.14) 7.9 m (±2.88) 22.6 m (±5.89)
Secondary 11.8 m (±3.51) 4.1 m (±0.73) 5.5 m (±2.20) 15.1 m (±6.26)
SEEF-CL Secondary 21.3 m (±5.92) 7.9 m (±2.12) 29.3 m (±11.56) 66.6 m (±23.26)
95
Figure 3-1. Locations of study sites in the CEB-FSC and SEEF-CL logging concessions in Gabon.
Figure 3-1A. Study area and plot locations in the CEB-FSC logging concession.
96
Figure 3-1B. Study area and plot locations in the SEEF-CL logging concession.
Figure 3-2. Pre-logging stem densities by diameter (dbh) class in the twenty 1-ha plots
in the CEB-FSC and twelve 1-ha plots in the SEEF-CL concessions.
97
Figure 3-3. Location of plots based on the number of trees per species before and after logging in the CEB-FSC and
SEEF-CL study sites. A&B) First axis derived from a Canonical Correspondence Analysis. C&D) Euclidean distances between plots. Results showed differences between CEB-FSC and SEEF-CL (t = 2.25; df = 16; p = 0.003).
98
Figure 3-4. Percentage of trees by damage category resulting from felling and skidding in the forest concessions in Gabon. A) CEB-FSC. B) SEEF-CL.
99
Figure 3-5. Relationships between harvested tree stem diameter (dbh) in the CEB-FSC (508 ha) and SEEF-CL (200 ha) concessions in Gabon. A) Number of damaged trees. B) Felling gap area. Results are based on linear model of an ANCOVA: a) CEB: Number of Damaged Trees = 0.09 x DBH - 0.05 (n = 10; R2 = 0.21; p = 0.181); SEEF: Number of Damaged Trees = 0.09 x DBH + 1.6 (n = 16; R2 = 0.05; p = 0.38); b) SEEF: Felling Gap Area = 10.9 x DBH – 388.1 (n = 13; R2 = 0.30; p = 0.05); CEB: Felling Gap Area = 10.1 x DBH + 205.3 (n = 12; R2 = 0.59; p = 0.008).
100
Figure 3-6. Maps of skid trails and felled trees in the CEB-FSC (left) and SEEF-CL (right) concessions at two scales.
101
7
Figure 3-7. Logging roads in the CEB-FSC (left) and SEEF-CL (right) concessions.
102
CHAPTER 4 COST COMPARISONS OF REDUCED-IMPACT AND CONVENTIONAL LOGGING IN
THE TROPICS
4.1 Background Information
Selective logging of tropical forest is an important contributor to economic
development, but when carried out by untrained and poorly supervised workers
operating without the aid of management plans or reliable stand maps, it is a major
cause of forest degradation. Conventional logging (CL), as shown by many studies
conducted around the tropics (e.g., Pinard and Putz, 1996; Sist et al., 1998; van der
Hout, 1999; Barreto et al., 1998; Healey et al., 2000; Boltz et al., 2001 and 2003; Tay et
al., 2002; Holmes et al., 2002), resulted in about twice as much damage to the residual
stand and soils as when “reduced-impact logging” (RIL) practices are employed. RIL
refers to planned harvest operations based on sound scientific and engineering
principles that, in combination with worker education, training, and supervision, serve to
improve the efficiency of labor and equipment in the harvesting of timber while reducing
damage to residual stands (Dykstra, 2002). To foster the adoption of RIL, many sets of
national and regional RIL guidelines have been developed around the tropics (e.g.,
Elias et al., 2001; FAO, 2006), most of which are heavily based on the FAO Model
Code for Forest Harvesting Practice (Dykstra and Heinrich, 1996).
Despite substantial and prolonged efforts at reforming logging practices in the
tropics and more than 300 published studies on RIL (Schwab et al., 2001), loggers have
been slow to adopt the improved harvesting practices at least partially due to their
perception that RIL is more costly to implement than CL (Putz et al., 2000 and 2008a;
Dykstra, 2002; Pearce et al., 2003; Ezzine de Blas and Ruiz Perez, 2008; Cerutti et al.,
2008). This perception contrasts markedly with the often repeated claim by non-
103
governmental organizations and others that RIL is actually more cost effective (e.g.,
Tropical Forest Foundation, CIFOR, Rainforest) and should therefore be adopted
spontaneously by loggers out of enlightened self-interest. Unfortunately, the
implementation of RIL requires up-front capital investment in timber inventories, staff
training, and sometimes machinery, along with substantial modifications in working
practices (Applegate et al., 2004). Logging company managers and owners are unlikely
to make these investments without clear indications of their benefits. Using all the
published studies we could find that report on the costs of CL and RIL in the tropics, we
attempt to shed light on this controversy.
Protecting tropical forest from degradation by destructive logging is critical in many
tropical countries where timber production provides an important source of income for
governments and workers along the forest product market chain. Furthermore, with
increased interest in the climate change mitigation potential of improved forest
management (Putz et al., 2008b), additional attention should be paid to the benefits of
RIL. But if payments are to be made for the additional carbon retained in forest when
RIL practices are employed, the costs associated with the adoption of these practices,
these costs need to be known. As long as there are discrepancies in estimates of the
operational costs of CL and RIL, carbon-based ecosystem service payments for
example though REDD+ (reduced emissions from deforestation and forest degradation)
are unlikely. This analysis of studies reporting the financial costs of RIL and CL will
hopefully help settle this dilemma.
Despite the economic importance of logging natural tropical forests, data on costs
and productivity of CL and RIL are difficult to interpret because they are affected by
104
many factors including the heterogeneity and diversity of tropical forests (Ghazoul and
Sheil, 2010), the quantity and quality of available labor, wages, forest history, the
species and size of trees harvested, terrain, climate, accessibility, transportation
distance, and differences in harvesting equipment, methods, efficiency, and intensity
(Sundberg and Silversides, 1988). Also, few studies consider all activities and
associated costs from harvest planning and worker training to the costs of the logging
operations themselves. Furthermore, whether harvesting costs are calculated up to the
log landing, forest border, or mill gate affects the findings. Complicating comparison
further, logging costs are expressed either per volume or per area, and if the latter, then
either on the basis of the area actually logged or the entire area designated for logging.
Comparison of cost effectiveness of CL and RIL are also affected by the spatial
and temporal scales of the research. For example, small plots that are entirely suited for
harvesting but that include no logging roads are likely to differ from studies of large
heterogeneous areas with logging roads, stream crossing, and areas from which log
extraction is prohibited. Similarly, if following RIL guidelines requires shutdowns of
logging operations due to wet weather and other temporary restrictions, log delivery
schedules will be disrupted, which has some associated (but difficult-to-determine)
financial costs; such costs were not evaluated in any of the studies surveyed here.
Finally, if the long-term effects of RIL and CL on timber yields and the full range of
environmental and social impacts are monetized and included, then the results will
likewise be affected; these effects are not studied here.
Comparisons of published values for the costs of RIL and CL are further
complicated by inconsistencies in the activities and payments included and how the
105
costs are calculated and expressed. In the first phase of our analysis we mostly
disregard these inconsistencies and compare published costs within studies. We then
standardize the data to allow comparisons among studies by including only forest
management activities terminating with the loading of logs onto trucks. Given this
restrictive frame-of-reference, road building and hauling costs, taxes and fees, and the
many long-term and indirect effects of logging are disregarded. This study included a
case comparing costs of CL and RIL in Gabon and a review of published costs of CL
and RIL in the tropics. We aimed to understand why loggers have been slow to adopt
improved forest management practices and to also to investigate if studies have over-
or under-estimate the costs of CL vs. RIL.
4.2 Methods
4.2.1 Descriptions of the Case Studies
The analyses are based on the 13 cases studies, including this study, briefly
described in this section. The studies are generally summarized, except that studies in
which the distinction between RIL and CL was not clear and those that employed
radically different log yarding equipment are at the end.
4.2.1.1 Fazenda Cauaxi, Paragominas, Brazil (Holmes et al., 2002; Boltz et al., 2001; both used data from Holmes et al., 2001)
This research was conducted in Fazenda Cauaxi, a privately owned forest in
Paragominas, Pará, Brazil. The terrain is undulating and the soils are oxisols with an
argilic horizon. The mean annual rainfall is 1750 mm and mean temperature 28oC. In
1995, Fundação Floresta Tropical (FFT) established three 100 ha plots, one as an un-
harvested control, one for CL, and one for RIL. The RIL plot was logged by trained
operators working for FFT whereas the CL plot was logged by an industrial cooperator
106
whose workers were not trained in RIL methods. Logging intensity was 4.25 trees ha-1
(26.09 m3 ha-1) for CL and 3.31 trees ha-1 (24.95 m3 ha-1) for RIL. A Stihl AV 51
chainsaw was used for felling and bucking in both CL and RIL plots. Logs were yarded
with a Caterpillar D6 bulldozer (=crawler tractor) in the CL plot and a Caterpillar 525
rubber-tired skidder in the RIL plot.
Holmes et al. (2002) used different techniques to estimate the costs and
productivity of RIL and CL. For RIL they estimated cost per cubic meter as the sum of
average fixed costs (pre-harvest, harvest planning, and infrastructure), average variable
costs (activities associated with felling, bucking, skidding, and log deck operations),
average costs due to waste (felled logs not found, volume lost due to poor felling,
volume left unused on the log deck, volume lost due to high stumps or poor bucking),
average stumpage cost (harvesting rights), and average training costs. For CL they
estimated productivity and costs of CL from survey data collected from seven loggers in
the Paragominas timbershed. The authors calculated training costs by amortizing the
cost per worker over 5-years assuming that at the end of that period a worker would
either need retraining or have changed occupations. They then divided the amortized
cost by the estimated volume harvested per worker over this period.
4.2.1.2 Pibiri, Guyana (van der Hout, 1999)
This research was conducted in the Demerara Timber Ltd (DTL) concession in
Central Guyana on an undulating sedimentary plain at 50-100 m a.s.l. with mixed
ultisols and arenosols. The evergreen forest in the study area has canopy heights of 30-
40 m with patches of the valuable commercial timber greenheart (Chlorocardium rodiei,
Lauraceae) concentrated on upper slopes. Mean annual rainfall is approximately 2700
mm and mean temperature is 26oC. DTL had allocated approximately 1000 ha to the
107
Tropenbos-Guyana Program and the Guyana Forestry Commission for research on RIL
and impacts of logging intensity on the residual stand. RIL harvesting was implemented
by trained Tropenbos/DTL teams, while CL was conducted by untrained DTL workers
employed by the same company. Harvesting operations focused on the extraction of
greenheart, which comprised 53% and 91% of total volume harvested for RIL and CL,
respectively. Felling was carried out with a Stihl AV 066 chainsaw for RIL and a Stihl
070 for CL. A Caterpillar 528 wheeled skidder equipped with a choker hook on the
winch cable was used for yarding logs from the RIL area, whereas the same skidder
with a bull-hook was used in the CL area. Logging intensity for RIL (69 ha; 12 plots of
5.76-ha each with harvest intensities of 4, 8, and 16 trees ha-1) was 9.7 trees ha-1 (29.1
m3 ha-1) while the CL intensity (3 plots of 11-ha plot each) was 8.3 trees ha-1 (24.1 m3
ha-1). Logging costs were calculated from hourly costs and production rates for felling
and skidding. Costs of pre-harvest planning, harvest preparation, harvest operation,
transport to mills, and overhead were included.
4.2.1.3 Fazenda Agrosete, Brazil (Barreto et al., 1998)
This study was conducted in Fazenda Agrosete, Pará, Brazil in evergreen forest
with canopy heights of 25-40 m, mean annual rainfall of 1700 mm, mean annual
temperature is 28oC, and soils that are predominantly latisols. RIL by trained workers
was carried out in a 105 ha plot whereas CL operations were conducted in an adjacent
75 ha plot by personnel with no formal training in logging. Logging intensity in the RIL
plot was 4.5 trees ha-1 (38.6 m3 ha-1), while in the CL it was 5.6 trees ha-1 (29.7 m3 ha-1).
For the RIL operations, two machines were used for yarding logs: a Caterpillar D5E
bulldozer equipped with a winch and tower (Hyster W5B) and a Caterpillar 518C rubber-
tired skidder equipped with a winch, tower, and grapple. Logs from the CL area were
108
skidded with a Caterpillar D5B bulldozer. Data were collected for each logging activity
on the productivity of machines and workers as well as labor costs including effective
work hours. Barreto et al. (1998) assumed that 0.25 m3 were lost during CL operations
due to high stumps and improper bucking and thus compared costs per 1 m3 for RIL
and per 0.75 m3 for CL. RIL cost estimates included pre-harvest operations, detailed
harvest planning, and harvest operations whereas CL costs included only basic harvest
planning (roads and log landings) and harvest operations. Cost estimates for both
logging systems included log transport to mills.
4.2.1.4 Sabah, Malaysia (Tay, 1999; Healey et al., 2000; Tay et al., 2002)
This study in Sabah, Malaysia focused on 230 ha allocated for harvesting by
trained and closely supervised crews using RIL techniques and 176 ha harvested by
untrained crews using CL methods in the Sabah Foundation concession in Ulu Segama
Forest Reserve (Pinard and Putz, 1996; Pinard and Cropper, 2000). Trees in the
Dipterocarpaceae comprised 49% of the total basal area for trees ≥10 cm dbh and
constituted 70-80% of total harvested timber. The terrain is hilly at 100-1,200 m a.s.l.,
the soils are oxisols and ultisols, the mean annual rainfall during the study period (1993-
1994) was 2739 mm, and the mean temperature was 26.7oC. In the RIL areas, only 129
ha of the 230 ha allocated for logging (56%) were actually logged due mostly to the RIL-
project requirement to follow state guidelines about harvesting on steep slopes (> 35
degrees) and in streamside buffer zones. In contrast, only 1 ha of the 176 ha (<1%) was
not harvested in the CL area. For this study, there were four independent blocks of RIL
and of CL in matched pairs. The average timber yield per logged hectare was 136 m3
ha-1 in the CL blocks and 106 m3 ha-1 in the RIL blocks. All felling was with Stihl 070
chainsaws. Skidding was carried out with a Komatsu D60 bulldozer in the RIL area and
109
a Caterpillar D7F bulldozer in the CL area. Logging costs were estimated by combining
operating costs, labor costs, and production rates. RIL costs included those for pre-
harvest, harvest, and post-harvest operations as well as non-operational costs (i.e.,
general expenses), while CL practices did not included post-harvest operations (i.e.,
culvert removal, log landing amelioration, and damage assessment) therefore, there are
no associated costs.
4.2.1.5 Malinau, East Kalimantan, Indonesia (Dwiprabowo et al., 2002)
The study was conducted in the 50,000 ha INHUTANI II forest concession in
Malinau, East Kalimantan, Indonesia. The forest is dominated by Dipterocarpaceae but
with the noteworthy presence of Agathis borneensis, Araucariaceae, another valuable
timber species. Mean annual rainfall is 4000 mm and mean annual temperature is
23.5oC with relative humidity of 75%-98%. The topography is hilly with a dense network
of steep ridges and valleys. In the logging site at 100-300 m above sea level, the 244 ha
of CL yielded 5.9 trees ha-1 (52.8 m3 ha-1) whereas the 138 ha of RIL yielded 6.9 trees
ha-1 (60.9 m3 ha-1). Logs were yarded to roadside landings with two CAT 527 rubber-
tired skidders in the RIL area and a Caterpillar D7G bulldozer in the CL area. Costs of
logging operations were estimated from data on the productivity of each activity together
with machine and labor costs. Both CL and RIL cost estimates included some pre-
harvest and harvest operations, but RIL costs also included those of training. The
authors estimated training costs based on an assumption that the trained worker would
be employed for 5 years and divided the estimated initial cost of training by the
estimated harvested volume over that period.
110
4.2.1.6 Belém, Brazil (Verissimo et al., 1992)
Research on CL was carried out during the June-December dry season in three
concessions separated by 100-110 km along a 340 km stretch of the Belém-Brasilia
Highway (3oS, 50oW) in evergreen lowland rainforest with canopy heights of 25-40 m.
The terrain is undulating, soils are oxisols and ultisols, and rainfall is seasonal with
about 2000 mm year-1. Harvested areas and harvesting intensities in the three areas
were 115 ha at 3 trees ha-1 (18 m3 ha-1), 37 ha at 7 trees ha-1 (35 m3 ha-1), and 16 ha at
16 trees ha-1 (62 m3 ha-1). Logging cost and profit estimates of both logging and wood
processing were based on interviews with five forest extraction crews and 33 mill
owners. The authors combined data on production rates and labor costs to estimate CL
costs including log transport to mills.
4.2.1.7 Sarawak, Malaysia (Dagang et al., 2002)
This research was conducted in mixed hill dipterocarp forest in the FOMISS-
Samling Pilot Area near Ulu Baram, Sarawak. The study area is underlain by
sedimentary rocks and has moderately sloping topography with regosols/acrisols and a
mean annual rainfall of 4600-5000 mm. Logging intensity was 64.5 m3 ha-1 in the CL
area (90 ha) and 27.8 m3 ha-1 in the RIL area (70 ha). Trees were felled using Stihl 07
chainsaws by workers trained in directional felling techniques in the RIL area and
untrained workers in the CL area. A Caterpillar D6 DLS bulldozer with a cable winch
was used for skid trail and log landing preparation and for log yarding in both RIL and
CL areas. One stated difference in harvesting operations was that winching distances
were longer in the RIL area, but no data on distances were provided. CL costs included
marking of harvestable trees and harvest operations whereas RIL costs included worker
111
training, harvest planning, harvesting, and post-harvest operations (i.e., damage
assessment and RIL compliance assessment).
4.2.1.8 Dungun, Terengganu, Malaysia (Ahmad et al., 2009)
RIL operations were conducted in Jengai Forest Reserve (197 ha) at an intensity
of 3.2 trees ha-1 (65 m3 ha-1), while CL operations were carried out in Tembat Forest
Reserve (135 ha) at an intensity of 3.3 tree ha-1 (92 m3 ha-1). The terrain of both forest
reserves is hilly with some rocks and slopes of 5-22o. Logs were yarded from both sites
with a Caterpillar D6C bulldozer. Logging costs were estimated using a combination of
time-study and survey-questionnaire methods to obtain information on fixed and
operating costs to estimate productivity and costs of machines and labor.
4.2.1.9 Tapajós National Forest, Brazil (Bacha et al., 2007)
This study was conducted during 1999-2003 in 3,130 ha of forest in Tapajós
National Forest, Parà, Brazil that was harvested using RIL practices at an average
intensity of 3.2 tree ha-1 (20.25 m3 ha-1). Cost of RIL operations were estimated using
data on the average productivity of labor and machines for harvesting operations
provided by the logging company. Included were costs of pre-harvest and harvest
operations and general expenses (i.e., support, logistics, supervision, lodging, etc.),
taxes, fees, and freight to the sawmill. No additional information was provided about the
characteristics of the study site or the equipment used. Workers were trained in RIL
practices by TFF.
4.2.1.10 Terengganu, Malaysia (Rahim Nik et al., 2002; data also used by Fisher et al., 2011)
Research was carried out in Jengai Forest Reserve in the 128,720 ha Kumpulan
Pengurusan Kayan-Kayan Terengganu Sdh. Bhd forest concession in Peninsular
112
Malaysia where dipterocarps accounted for 30-40% the total volume of all trees ≥50 cm
dbh. Harvesting was carried out according to the Malaysian Criteria, Indicator, Activities
and Management Specifications (MC&I) in a 43-ha research plot and what the authors
referred to as “conventional practices” in a nearby 364 ha area. In keeping with most
RIL guidelines, MC&I compliance requires worker training and pre-harvest activities
including an environmental impact assessment (EIA), pre-felling inventory, tree marking
and mapping, compartment boundary demarcation, marking and delimitation of buffer
zones, and planning of road alignments. In contrast, the conventional practices, which
we refer to as +/-RIL, included all the impact-reduction measures specified by MC&I
except the EIA and tree mapping. Logs were extracted from the +/-RIL area with a
bulldozer, and from the MC&I area with a modified excavator with a 300 m yarding
radius. On average 32.4 m3 ha-1 was harvested with the excavator and 33.8 m3 ha-1 with
bulldozers. Logging cost data were obtained from the State Forestry Department, the
District Forest Office, the concessionaire, and the contractors. Costs of excavator and
bulldozer logging included post-harvest assessments of stocking and residual stand
damage to determine the silvicultural treatment needed.
Compliance with MC&I rules resulted in 28% of the area harvested with the
excavator (12 of 43 ha) being set aside for soil and water conservation. In contrast, only
1.65% of the bulldozer-yarded area (6 of 364 ha) was not logged. Based on pre-logging
stock inventories in MC&I and the +/-RIL areas, of 202.6 m3 (16.9 m3 ha-1) and 392.9 m3
(65.5 m3 ha-1) were not harvested from the buffer zones, respectively.
4.2.1.11 Kelantan, Malaysia (Rahim Abdul et al., 2009)
Data for this case study on logging with an excavator modified for log skidding
(“LogfisherTM”; see Gan et al., 2006 for a description) were collected from a
113
concessionaire in Sungai Betis Forest Reserve in Kelantan, Peninsular Malaysia. Data
on logging with bulldozers were collected in a nearby concession in Gua Musang.
Excavator-yarding (+/-RIL-ex) and bulldozer-yarding (+/-RIL) were conducted in 90 and
100-ha plots, respectively, with corresponding logging intensities of 50.7 m3 ha-1 and
54.3 m3 ha-1. Because the excavator winched logs up to with 300 m, no skid trails were
needed. In contrast, skid trail densities in the crawler-tractor plots averaged 300 m ha-1.
Published costs included payments for premiums, royalties and levy charges,
administration, and machines.
4.2.2 Study Site: Monts de Cristal, Gabon
This study was conducted on Monts de Cristal in the SEEF concession (see
section 2.2.1 for description). SEEF allocated approximately 500 ha to the Tropical
Forest Foundation and FORM International for training, demonstration, and research on
RIL as part of an International Tropical Timber Organization project. 104 ha were logged
under RIL at an intensity of 0.82 trees ha-1 (8.1 m3 ha-1), while an adjacent 54 ha were
logged under CL with an intensity of 0.97 trees ha-1 (7.7 m3 ha-1). Prior to harvesting the
RIL area, each feller received one week of training in directional felling and skid trails
and log landings were planned. Trees in the RIL area were felled using chainsaws (Stihl
MS 880) and yarded with a track-skidder with a centrally mounted winch (Caterpillar
527). Trees in the CL area were felled by a contracted untrained independent feller
using a Stihl 070 chainsaw and yarded with the same machine as in the RIL plot.
4.2.3 Logging Costs
Rigorous comparisons of the costs of RIL and CL should be based on the costs of
each activity (e.g., stock mapping, felling, skidding), daily and effective hourly labor
costs, operating costs and depreciation of equipment, and materials used. Because the
114
studies reviewed vary in the data provided and in how costs were calculated and
expressed, the published values presented in Table 4-1 are not directly comparable
(Appendix G). We standardized the data by restricting our analysis to activities leading
up to log loading, as described above, and present those estimates in Table 4-2.
4.2.3.1 Labor costs
The total wages received by workers can include basic salaries, yield-based
bonuses, and other benefits paid by the contractor, concessionaire, or forest owner.
Daily labor costs were generally calculated based on the monthly incomes due to
variability in base salaries among workers and in their monthly production. Labor costs
are expressed as either a daily or an hourly basis, depending on the activity.
4.2.3.2 Costs of equipment and materials
Equipment costs include operating costs (i.e., variable cost of operations) and
effective daily machine time (van der Hout, 1999; Holmes et al., 2002) plus the fixed
costs of ownership (depreciation, insurance, and interest) and the variable costs of
operation (fuel, lubricants and filters, maintenance and repairs, special wear items, etc.;
van der Hout, 1999). Annual depreciation (D) was determined as D = (P-S)/N where P is
the equipment purchase price, S is the salvage value (i.e., the amount for which the
equipment can be sold at the time of its disposal), and N the economic life in years (i.e.,
the period over which the equipment can operate with acceptable operating costs and
productivity; Miyata, 1980). Costs of fuel and lubricants were estimated as a
combination of the hourly operating cost, hourly consumption, and the price per liter of
fuel or lubricant (based on prices in petrol stations in Libreville, Gabon in mid-2010).
Effective machine-use time (hr/yr) varies with the type of equipment and with the
115
effective crew hours. For this study, values of effective machine-use time from Holmes
et al. (2002) were used.
4.2.3.3 Training costs in the tropics
To estimate the training costs per cubic meter of timber or per hectare of forest
harvested we amortized the training costs over the productive life of the trained worker
(Applegate et al., 2004). Most published studies on logging costs, even those with a RIL
focus, often fail to report training costs, do not adequately explain how they were
calculated, or assign to the small experimentally harvested areas all the costs of
workers that will presumably continue to employ what they learned elsewhere. To
standardize estimates of the training costs for workers that conduct RIL, we started our
calculations with the average of training cost per worker in Malaysia (US$596; Tay,
1999), Gabon (US$710; Wanders pers.com.), and at TFF training facilities in Brazil
(US$750; Schulze et al., 2009), Guyana ($US633; G. Marshall, 2011, pers. com.), and
Indonesia (US$706 assuming 6 fellers; Dwiprabowo et al., 2002). Following Holmes et
al. (2002) and Dwiprabono et al. (2002), we amortized the average cost of training over
a 5-year working life of the average worker. For fellers, we then assumed an average
daily productivity of 5 trees, 20 m3, and 1 ha for 200 d/yr over a 5-year working life. On
this basis the estimated cost of providing a worker with the training needed to apply RIL
techniques at US$0.68 per tree, US$0.17 m-3, and US$1.44 ha-1.
4.2.3.4 Analysis of logging profits
Logging profits (π) were estimated as: π = (p*h)-C; where p is the market price of
logs per cubic meter, h is the per-hectare harvested volume, and C is the per hectare
cost of logging. We used the log prices reported in each case study when available. For
studies reporting log prices by category, of high, medium, and low values, we used each
116
price category with the assigned timber volume accordingly. Where those values were
not reported, we employed the average log price for the study country.
4.3 Results
4.3.1 Costs of Logging on Monts de Cristal, Gabon
4.3.1.1 Pre-harvest operations
Boundary demarcation: The 104 ha RIL block was demarcated by four workers
and one supervisor with a production rate of 2 000 meters per day. To calculate the cost
of block demarcation, we divided the daily production by the total linear distance to get
the number of days and then multiplied that value by daily labor costs (US$142.34). We
expressed this in terms of cubic meters of harvested timber by dividing the cost by the
total volume harvested and that resulted in a cost of US$0.84 m-3 (Table 4-3).
Pre-felling inventory: The pre-felling inventory of all trees of commercial species
≥70 cm dbh was conducted by three workers (including one tree identifier) and one
supervisor. Cost was determined by combining a productivity (12.5 ha/day) with daily
labor costs (US$142.34) resulting in a cost of US$1.40 m-3 (Table 4-3).
Tree hunting: Hunting for trees suitable for felling and ≥70 cm dbh was conducted
in the 54-ha area by four laborers with a productivity of 10 ha/day and daily cost of
US$98.16. The cost of this operation was estimated at US$0.51 m-3 (Table 4-3).
Training: Training costs were determined by amortizing the cost per feller
(US$710) over the expected 5-year work life of each feller divided by the total timber
volume expected to be harvested by the feller during that period. We estimated training
cost for this study at US$0.14 m-3 (Table 4-3).
Tree marking and mapping: Tree marking was conducted as described for the
pre-felling inventory by four workers with a production rate of 12.5 ha per day for RIL
117
operations and daily cost of US$98.16. The cost of this activity was estimated at
US$0.43 m-3 (Table 4-3).
4.3.1.2 Harvest operation
Felling and bucking: For CL, felling was conducted by 1 untrained feller and 1
helper with a productivity of 83.12 m3 per day. A Stihl 070 chainsaw was used with an
estimated operating cost of US$4.97/hr and daily cost of US$12.08 per day for a total
labor cost of US$44.17 per day. The combination of these costs and the productivity
resulted in felling and bucking cost using CL of US$0.68 m-3. For RIL, felling was
conducted by a trained feller with a helper with production rates of 39.6 m3 per day. A
Stihl MS 880 chainsaw was used with an operating cost of US$8.65/hr and daily cost of
US$21.02 per day for a total labor cost of US$44.17 per day. The cost of this activity for
RIL was estimated at US$1.65 m-3 (Table 4-3).
Skidding: Log yarding under both CL and RIL was with a tracked skidder
Caterpillar D527 bulldozer with one operator and an assistant. For CL, the machine
operating cost was US$18.25/hr, daily production cost was US$199.99, labor cost was
US$53.99/day, for a production of 134.7 m3 per day and that resulted in cost of
US$1.48 m-3. For RIL, the machine operating cost was US$18.35/hr for a daily cost of
US$146.80. The labor cost was similar to CL and the production was 122.95 m3 per day
leading to a cost of US$1.63 m-3 (Table 4-3).
Log deck operations: In both CL and RIL, the operations were conducted using a
Komatsu WA 470 loader with an operating cost was US$18.09/hr and a daily cost of
US$144.72. For CL, the operation was conducted by 3 workers (1 loader operator, 1
sawyer, and 1 laborer) with a total labor cost of US$73.62 per day. Chainsaw operating
cost was US$8.24 per day and the production rate was 129.4 m3 per day. The cost of
118
this activity was estimated at US$1.75 m-3. The same operation under RIL involved four
workers (a loader operator, one sawyer, one laborer, and a supervisor). Labor cost was
US$93.25 per day and the production was 114.9 m3 per day. Chainsaw operating cost
was US$13.94 per day and the cost this operation was estimated at US$2.19 m-3 (Table
4-3).
4.3.2 Costs and Profits of Logging in the Tropics
4.3.2.1 Cost of logging
Among the studies comparing RIL and CL, three reported low relative costs for
RIL, one showed identical costs with CL, and six reported high RIL cost (Figure 4-1a).
Analysis of the relative costs (CL cost / RIL costs) with the relative timber volume
extracted (Volume CL/ Volume RIL), six of the ten studies comparing RIL and CL
showed CL relatively cheaper than RIL, while only one study showed similar cost
between the two logging types (Figure 4-2). This suggested that RIL costs, despite of
many of its components, could be lower with higher timber volume harvested. The low
cost of CL however, resulted from higher timber volume harvested for this logging type.
The difference in costs on the volume basis of both logging types is strongly influenced
by the volume of timber harvested. When the costs of log hauling and other expenses
(i.e., taxes, overheads, etc.) were excluded, the total cost of pre-harvest activities,
harvest planning, skid trail layout, log deck construction and operations, and harvest
operations varied among the case studies and between logging methods. On this basis,
when expressed as the cost per volume of timber harvested (Figure 4-1b), two studies
showed lower costs for RIL, two reported equal costs for RIL and CL, and six reported
that RIL cost more. When expressed per logged hectare, four studies reported lower
119
costs for RIL, two reported almost identical costs for RIL and CL, and four reported
higher RIL cost (Figure 4-1c).
Reported costs of CL had a range of US$13.54-80.02 m-3 and for RIL US$13.39-
68.26 m-3 (Table 4-2). The highest cost of CL (US$80.02 m-3) was reported for
Kalimantan, Indonesia (Ruslandi et al., 2011) whereas the highest cost for RIL
(US$68.26 m-3) was reported in Kelantan, Malaysia where timber harvesting used a
modified excavator to yard logs (Rahim Abdul et al., 2009). Note that some of the cost
estimates (Table 4-2) include taxes, overheads, and the cost of log transport to the mill
gate, which means that comparisons among studies should be made with caution. CL
and RIL costs were almost identical in Guyana (US$28.28 m-3 and US$28.23 m-3,
respectively; van der Hout, 1999) and in Gabon where CL cost US$17.66 m-3 and RIL
cost US$20.90 m-3.
4.3.2.1.1 Pre-harvest activities
Cost of pre-harvest activities (i.e., tree marking, block layout, inventory, vine
cutting, data processing, and map making) for studies comparing CL and RIL had a
range of US$0.13–0.51 m-3 for CL and US$0.32–2.81 m-3 for RIL (Table 4-2). In
Malaysia, the pre-harvest costs for five studies ranged US$0.13–0.29 m-3 for CL and
US$0.32–2.25 m-3 for RIL. The cost of these activities for studies using only CL was
US$1.54 m-3 in Malaysia (Fisher et al., 2011; Ruslandi et al., 2011), US$1.62 m-3 in
Indonesia (Ruslandi et al., 2011), and US$1.84 m-3 in Brazil (Verissimo et al., 1992).
From four studies in Brazil, the cost of pre-harvest activities ranged US$0.14–1.84 m-3
for CL and US$0.67–1.38 m-3 for RIL. In Indonesia, two studies on CL reported that
these activities cost US$0.29 m-3 and US$ 0.70 m-3 whereas one study on RIL reported
a cost of US$0.65 m-3 (Bacha et al., 2007). In a comparative study on CL and RIL in
120
Guyana, pre-harvest activities cost US$0.18 m-3 and US$0.50 m-3, respectively (van der
Hout, 1999). In Gabon these activities cost US$0.51 m-3 using CL and US$2.81m-3
using RIL (Table 4-1).
Published pre-harvest costs varied with whether the costs of vine cutting were
included; only 7 of 13 case studies included this cost (Table 4-2). The reported costs of
vine cutting ranged from US$0.13 m-3 in studies in Guyana and Indonesia to US$0.25
m-3 in one study in Malaysia and four in Brazil. One source of variation in these costs is
whether the vine cutting was a blanket vine cutting treatment, as in Malaysia (Pinard et
al., 1995), or what we presumed to be a treatment involving only vines on trees to be
harvested and perhaps those infesting some future crop trees. Also, it would depend
upon whether vines were cut prior to harvest by survey crew or immediately before
harvest by fellers.
Variation in training costs also contributed to inter-study differences in pre-harvest
costs. The average reported cost of training ranged from US$0.60 m-3 in Malaysia down
to US$0.14 m-3 in Gabon. This variation results from differences in the duration of
training and whether the trainers were locals or expatriates.
4.3.2.1.2 Harvest planning
Costs of harvest planning (i.e., road, skid trail, and log deck planning) ranged
US$0.09-0.14 m-3 for CL and US$0.02–0.55 m-3 for RIL (Table 4-2). In Brazil, harvest
planning costs incurred under CL was US$0.14 m-3 (Holmes et al., 2002), whereas in
the four RIL studies they had a range of US$0.16–0.55 m-3. In Malaysia, the harvest
planning costs for three RIL studies were US$0.06 m-3, US$0.09 m-3, and US$0.49 m-3
(Mattsson-marn et al., 1981; Tay et al., 2002; Dagang et al., 2002, respectively).
Overall, for RIL, road planning costs were not reported in three of the studies whereas
121
costs of skid trail and log deck planning were not reported in ten studies; it is not clear
whether these expected RIL activities were not carried out or simply not included among
the reported costs.
4.3.2.1.3 Infrastructure costs
With the termination of our cost analysis with the loading of logs on trucks,
infrastructure costs included only log deck construction and operations and skid trail
construction and maintenance. Overall, these costs had a range of US$0.14–1.47 m-3
for CL and US$0.07–1.89 m-3 for RIL (Table 4-2). The highest costs were reported in
Malaysia for both CL and RIL (Tay et al., 2002) and that was only for skid trails; no cost
was reported for log deck construction for CL and log landings were not used in the RIL
area (logs were loaded from roadsides; Pinard et al., 1995). In Brazil the cost of log
deck and skid trail construction had a range of US$0.07–0.61 m-3 for RIL and US$0.17–
0.29 m-3 for CL. For three studies in Brazil, the average cost for RIL was US$0.41 m-3
and for CL US$0.23 m-3. One study in Brazil reported only the cost of log deck
construction for both RIL (US$0.07 m-3) and CL (US$0.17 m-3; Barreto et al., 1998). In
Guyana, the cost for RIL infrastructure was US$0.20 m-3, which included the costs of
skid trail (US$0.08 m-3) and log deck construction (US$0.12 m-3); for CL, the cost was
US$0.13 m-3 and consisted of the cost of log deck construction but the difference
resulted from higher log volume recovery in RIL (van der Hout, 1999).
4.3.2.1.4 Harvest operations
Reported harvesting operations (i.e., felling, bucking, skidding, and log deck
activities) cost US$3.15–14.81 m-3 for CL and US$2.34–16.38 m-3 for RIL. The highest
CL harvesting costs with tractor yarding were reported in Indonesia (Ruslandi et al.,
2011) and for RIL in Malaysia (Tay, 1999). In Malaysia, harvesting operations for five
122
studies on CL cost US$3.15–9.15 m-3, which included felling and bucking (US$2.56 m-
3), skidding (US$5.94 m-3), and log deck activities (US$1.07 m-3). For RIL, these
operations cost US$3.15–16.38 m-3. In four studies in Brazil, costs of harvesting
operation using CL had a range of US$4.36–10.04 m-3 and for RIL ranged US$3.14–
6.28 m-3. The CL estimates included the costs of felling and bucking, skidding, and log
deck activities. RIL costs included those for felling and bucking, skidding, and log deck
activities. In Guyana, CL harvest operations cost US$5.23 m-3 and RIL cost US$5.59 m-
3, both costs included felling and bucking, skidding, and log deck activities (van der
Hout, 1999).
Costs for CL harvest operations in two studies in Indonesia averaged US$8.89 m-3
whereas for one RIL study the cost was US$2.34 m-3. Using RIL, costs of felling and
bucking were US$0.04 m-3, skidding US$0.20 m-3, and log deck activities US$2.10 m-3
(Dwiprabowo et al., 2002). In Gabon, the cost of these operations was US$3.91 m-3 for
CL and US$5.47 m-3 for RIL (Table 4-3).
4.3.3.2 Profits for CL and RIL
Based on reported data on logging costs and timber sale prices, logging profits
varied substantially between logging systems and intensities when expressed either per
logged hectare or per harvested timber volume (Table 4-4; Figure 4-3).
Among the ten studies comparing RIL and CL, four studies showed higher relative
(CL/RIL) logging profit per cubic meter of timber harvested (i.e., profit per logged
hectare divided by the per hectare timber volume harvested) for CL, two reported
almost identical profits, and four showed higher profits for RIL (Figure 4-3a). When
expressed per logged hectare, six studies showed lower relative profits for RIL and four
showed lower profit for CL (Figure 4-3b). The results suggested that timber volume
123
extracted and the logged area are the overarching factors influencing RIL profit. These
results that are based on published values from the first cut showed that in both cases,
RIL is relatively less financially profitable than CL. But this varied among the reviewed
studies with a variability in logging intensity (i.e., timber volume harvested) and logged
area between CL and RIL.
4.4 Concluding Remarks
Costs of conventional and reduced-impact selective logging (CL and RIL,
respectively) in tropical forests varied substantially among the reviewed studies but
without clear relationships with region of by differences in research and reporting
methods. These challenges notwithstanding, of the 13 published studies that compared
CL and RIL, two found RIL to cost less than CL on cubic meter basis, three reported
almost identical costs, and eight that found RIL to cost more than CL. Of the ten studies
that reported per cubic meter profits from logging, six found CL to be more profitable
than RIL, two reported almost equal profits, and two found higher profits for RIL. Some
of the variation among studies derived from differences in what costs were included and
from differences in the yarding equipment used in logging operations.
Restricting the analysis of logging costs up to and including log deck operations
allowed meaningful comparisons to be made among the studies. But it may be
potentially a serious problem to disregard the major costs of road construction and
maintenance as well as the considerable costs of hauling logs to markets or mills
(Dykstra and Heinrich, 1996). The principal justification for this disregard is that only 6 of
the 13 studies in our analysis provided the necessary data and none reported costs for
bridges and culverts. Given that inclusion of road-related expenses can double the
costs of logging (Rahim Nik et al., 2002; Rahim Abdul et al., 2009; Fisher et al., 2011;
124
Ruslandi et al., 2011), they clearly deserve more attention. It is also important to know
by how much and at what cost employment of RIL road, skid trail, and log landing
construction and maintenance techniques reduces erosion and minimizes other
deleterious environmental impacts such as impoundments, landslides, fire, carbon
emissions, poaching, forest colonization by people and invasive species, and population
fragmentation. These impacts to ecosystem services should be valued and included in
the cost-benefit analyses of logging but it is beyond the scope of this study. At the
minimum, studies on logging costs that include road-related expenses should consider
the increased trafficability of well-made roads and the decreased need for maintenance
and vehicular repair, even if the environmental benefits are disregarded.
The studies reviewed varied in whether they included the cost of wood waste in
the forest and how waste was measured. Only one study reviewed provided detailed
data on wood waste due to improper felling (i.e., high stumps and broken logs), as well
from inefficiencies in volume recovery due to poor bucking and lost and abandoned logs
(Holmes et al., 2002). In that study in Brazil, Holmes et al. (2002) reported that
merchantable wood wasted represented 23.9% of the harvestable volume for CL but,
only 7.6% for RIL. If future yields are to be considered in financial analyses of logging
(see Healey et al., 2000), then the 50 to 200% reductions in residual stand damage that
result from the implementation of RIL also need to be considered (Mattsson-Marn and
Jonkers, 1981; Johns et al., 1996; Berthault and Sist, 1997). Despite most RIL
guidelines call for future crop trees to be mapped and flagged prior to logging, none of
the studies reviewed included the cost of this activity.
125
Interpretation of the results of some of the published comparisons of RIL and CL is
complicated by the introduction of new technology. In most studies used in our analysis,
logs were yarded to roadside log landings with bulldozers or rubber-tired skidders, but
Rahim Abdul et al. (2009) and Rahim Nik et al. (2002) used a modified excavator
(LogfisherTM) with a 300 m cable to yard logs directly to roadsides. While the costs of
logging with the excavator were higher either per unit volume or per unit area,
environmental damage was presumably much reduced but not accounted for in their
financial assessments.
Compliance with governmental regulations, forest certification principles, and RIL
guidelines often requires demarcation and protection of buffer zones along water
courses and restrictions on harvesting on steep slopes. Where such areas are typically
harvested (i.e., under CL), RIL-related restrictions result in foregone timber that some
loggers consider as a cost. With so much deforestation of areas on level terrain with
well-drained soils, logging is increasingly being relegated to adverse sites where the
foregone timber issue is likely to increase in prominence. Already in the studies we
reviewed that were conducted at the scale of logging blocks and not sample plots, there
were sometimes substantial differences between the area designated for logging (i.e.,
the gross area) and the area from which timber was actually harvested (i.e., the net
area) when RIL guidelines were followed. For example, in Peninsular Malaysia,
compliance with the Malaysian Criteria and Indicator rules resulted in 28% of the area
(12 of 43 ha) being set aside for soil and water conservation in Terengganu (Rahim Nik
et al., 2002) and 2% in Kelantan (Rahim Abdul et al., 2009). Such differences in net and
126
gross areas will obviously affect logging costs, especially if those costs are expressed
on the basis of the entire area allocated for logging.
In the studies we reviewed, logging costs were expressed per unit of harvested
volume or per unit area, metrics that do not fully consider the financial importance of
time from the perspective of log buyers. In studies that tracked the timing of harvesting
activities (e.g., Holmes et al., 2002), there can be accounting of the costs to logging
firms from temporary shutdowns of logging operations. During such times, workers
typically continue to be paid their base salaries and many fixed costs continue to accrue
(e.g., camp running, administration fees). For example, in Malaysia Tay 1999 and Tay
et al. (2002) reported much greater number of unproductive rainy days with RIL that
was a major component in the greater costs of RIL than in their study. Logging is an
outdoor activity that, even in CL, sometimes ceases due to inclement weather or road
closures. To the extent that following RIL guidelines causes more frequent or longer
duration interruptions in the flow of logs out of the forest, then at least from the
perspective of the buyers of logs, it is important to account for this variable in cost
comparisons with CL.
The analysis of published data revealed substantial variability in logging costs
among the case studies. Some of this variability may reflect the spatial and temporal
scales at which CL and RIL are compared. For example, comparisons based on small
and uniformly harvestable experimental plots are likely to find cost savings with RIL. In
contrast, in the large areas where industrial logging is carried out, RIL may not seem so
favorable due to foregone timber in protected areas within logging units. Logging
restriction due to compliances to RIL guidelines constitutes a factor that dictates the
127
cost of RIL either per logged area or timber volume harvested. But those costs can be
minimized if RIL was coupled with silvicultural treatment such as liberation thinning (i.e.,
RIL+) through efficiency of logging practices. Overall, given the financial and ecological
importance of forest management, detailed information about the costs and benefits of
different improved practices are needed to make informed decisions that will determine
the fates of forests and forest industries.
Lack of adequate information about logging costs should not continue to preclude
sound decision-making about adoption of improved forest management techniques.
Where adoption of those improvements will reduce logging profits, financial subsidies or
incentives might be needed. Fortunately, since carbon emissions from disturbed soils
and damaged residuals stands are reduced by the application of RIL techniques,
REDD+ and other carbon conservation programs should consider a payment scheme
that account for the benefit associated with RIL.
128
Table 4-1. Comparison of reduced-impact (RIL) and conventional logging (CL) logging intensities and harvesting costs per cubic meter in the tropics based on published data. These cost estimates cannot be directly compared among studies due to differences in which logging operations were included.
Source Locations Logging Method
Area (ha) Trees harvested (# ha-1)
Volume extracted (m3 ha-1)
Costs (US$ m-3)
Holmes et al., 2002
Fazenda Cauaxi, Brazil
CL 100 4.26 26.1 15.66
RIL 100 3.31 24.9 13.84
Boltz et al., 2001 Para, Brazil CL 100 - 25.8 13.50
RIL 100 - 19.7 16.34
van der Hout, 1999
Pibiri, Guyana CL 32 8.7 28.5 28.28
RIL 69 8.9 31 28.23
Barreto et al., 1998
Fazenda Agrosete, Brazil
CL 75 5.6 29.7 24.95
RIL 105 4.5 38.6 26.48
Tay, 1999 and Tay et al., 2002
Sabah, Malaysia CL 175 13 136 35.75
RIL 129 9 106 42.25
Healey et al., 2000 a
Sabah, Malaysia CL 175 13 136 35.75
RIL 129 9 106 42.25
Dwiprabowo et al., 2002
East Kalimantan, Indonesia
CL 11 x 1 7.6 83 20.03
RIL 12 x 1 7.5 60 20.38
Verissimo et al., 1992
Belém, Brazil CL 242 6.4 38 22.42
Dagang et al., 2002
Sarawak, Malaysia
CL 90 - 64.5 7.35
RIL 70 - 27.8 11.39
This study Monts de Cristal, Gabon
CL 54 0.97 7.7 17.66
RIL 104 0.82 8.1 20.9
129
Table 4-1. continued
Source Locations Logging Method
Area (ha) Trees harvested (# ha-1)
Volume extracted (m3 ha-1)
Costs (US$ m-3)
Ahmad et al., 1999
Dungun, Terengganu, Malaysia
CL 135 3.2 92 13.93
RIL 197 3.3 65 13.39
Bacha et al., 2007 Tapajós, Brazil RIL 3,130 - 20.3 19.22
Mattsson-Marn et al., 1981
Sarawak, Malaysia
CL 122 - 53 31.71
RIL 122 - 55 29.88
Rahim Nik et al., 2002
Terengganu, Malaysia
±RIL b 364 - 33.2 41.09
RIL (MC&I-ex) c
31 - 32 52.25
(Fisher et al., 2011 d
Sabah, Malaysia ±RIL b >220,000 - 152 60.36)
Ruslandi et al., 2011
Kalimantan, Indonesia
CL - - 40-60 80.02
Ruslandi et al., 2011
Sabah, Malaysia CL - - - 60.36
Rahim Abdul et al., 2009
Kelantan, Malaysia
±RIL b 100 - 54.3 44.8
±RIL-ex c
90 - 50.7 68.26
a Data from Tay, 1996. b Some RIL techniques applied, yarding with a bulldozer. c Yarding with a modified excavator. d Logging cost data from Rahim Nik et al., 2002.
130
Table 4-2. Harvesting costs per cubic meter of timber extracted and per logged hectare divided into pre-harvest activities, harvest planning, infrastructure construction, and harvest operations (felling, bucking, skidding, and log landing operations). Sums of the cost data in this table do not equal those in Table 1 because the latter include a variety of other costs (e.g., hauling, taxes, etc.).
Source Locations Logging method
Pre-harvest Harvest planning Infrastructure a Harvest operation Total costs
US$/m3 US$/ha US$/m
3 US$/ha US$/m
3 US$/ha US$/m
3 US$/ha US$/m
3 US$/ha
Holmes et al., 2002
Fazenda Cauaxii, Brazil
CL - - 0.14 3.61 0.29 (0.28)
7.57 (7.31)
4.49 117.19 4.92 128.41
RIL 1.18b 29.38 0.16 3.98 0.43
(0.16) 10.71 (3.98)
3.14 78.19 4.91 122.26
Boltz et al., 2001
Paragominas, Brazil
CL 0.14 3.61 - - 0.28 (0.22)
7.22 (5.68)
4.49 115.89 4.91 126.73
RIL 1.87b 36.91 0.55 10.84 0.61
(0.28) 12.02 (5.53)
3.21 63.37 5.71 112.71
van der Hout, 1999
Pibiri, Guyana
CL 0.19 5.42 0.09 2.56 0.14 (2.30)
3.99 (65.55)
5.23 149.06 5.65 161.02
RIL 0.50b 15.50 0.09 2.79 0.20
(2.00) 6.20
(62.00) 5.59 173.29 6.38 197.78
Barreto et al., 1998
Fazenda Agrosete, Brazil
CL - - - - 0.17 (0.23)
5.05 (6.83)
4.36 129.49 4.53 134.54
RIL 0.99b 38.21 0.35 13.51 0.07
(0.22) 2.70
(8.49) 4.15 160.19 5.56 214.62
(Tay, 1996 Sabah, Malaysia
CL 0.13 17.68 - - 1.47 (6.03)
199.92 (820.08)
9.04 1229.44 10.64 1447.04)
RIL 1.33b 140.98 0.09 9.54 1.89
(7.74) 200.34
(820.44) 16.38 1736.28 19.69 2087.14)
Dwiprabowo et al., 2002
East Kalimantan, Indenosia
CL 0.29 24.07 - - - - 3.15 261.45 3.44 285.52
RIL 0.66b 39.60 0.02 1.20 - - 2.34 140.40 3.02 181.20
Verissimo et al., 1992
Belém, Brazil CL 1.84 69.92 - - - - 10.04 381.52 11.88 451.44
Dagang et al., 2002
Sarawak, Malaysia
CL 0.13 8.39 - - - - 7.22 465.69 7.35 474.08
RIL 2.25 62.55 0.49 13.62 - - 8.34 231.85 11.08 308.02
131
Table 4-2. continued Source Locations Logging
method Pre-harvest Harvest planning Infrastructure
a Harvest operation Total costs
US$/m3 US$/ha US$/m
3 US$/ha US$/m
3 US$/ha US$/m
3 US$/ha US$/m
3 US$/ha
This study
Monts de Cristal, Gabon
CL 0.51 3.93 - - - - 3.91 30.11 4.42 34.04
RIL 2.81 22.76 - - - - 5.47 44.31 8.28 67.07
Bacha et al., 2007
Tapajós, Brazil
RIL 1.49b 30.25 0.20 4.06 -
(2.67) -
(54.20) 6.28 127.48 7.97 161.79
Mattsson-Marn et al., 1981
Sarawak, Malaysia
CL 0.14 7.42 - - - (2.66)
- (140.98)
9.15 484.95 9.29 492.37
RIL 0.32 17.60 0.06 3.30 - (2.66)
- (146.30)
7.57 416.35 7.95 437.25
Rahim Nik et al., 2002
Terengganu, Malaysia
±RIL c 1.54 51.13 - - -
(1.36) -
(45.15) 9.92 329.34 11.64 380.47
RIL (MC&I-ex)
d
4.73 151.36 - - - (8.83)
- (282.56)
12.49 399.68 17.22 551.04
(Fisher et al., 2011
Sabah, Malaysia
±RIL c 1.54 234.08 - - - - 9.92 1507.84 11.64 1741.92)
Ruslandi et al., 2011
Kalimantan, Indonesia
CL 1.62 81.00 - - - (0.92)
- (46.00)
14.81 740.50 16.43 821.50
Rahim Abdul et al., 2009
Kelantan, Malaysia
±RIL c 2.78 150.95 - - -
(6.64) -
(360.55) 8.78 476.75 11.56 627.71
±RIL-ex d
3 152.10 - - - (0.31)
- (15.72)
22.23 1127.06 25.23 1279.16
Ahmad et al., 1999
Dungun, Terengganu, Malaysia
CL - - - - - - 4.79 440.68 4.79 440.68
RIL - - - - - - 4.79 311.35 4.79 311.35
a Infrastructure costs included only skid trail layout and log deck construction which were only reported in three studies on RIL (van der Hout, 1999; Boltz et al., 2001; Holmes et al., 2002), one study reported only the cost of log deck construction (Barreto et al., 1998), and one reported the cost of skid trail construction (Tay,1999). Costs of road construction and maintenance are noted in parentheses when available. b Studies included cost of vine cutting c Some RIL techniques applied, yarding with a bulldozer. d Yarding with a modified excavator.
132
Table 4-3. Costs of conventional (CL) and reduced-impact logging (RIL) on Monts de Cristal, Gabon.
Logging phase Activity CL (US$ m-3) RIL (US$ m-3)
Pre-Harvest Boundary demarcation - 0.84 Pre-felling inventory - 1.40 Tree hunting 0.51 - Training - 0.14
Tree marking and mapping
- 0.43
Harvest Felling and bucking 0.68 1.65 Skidding 1.48 1.63 Log deck operation 1.75 2.19 Sub-total 4.42 8.28 Others Royalty a 10.80 10.80 Area tax b 1.36 0.65 Administration fees c 1.08 1.17 Sub-total 13.24 12.62
Grand total 17.66 20.90
a Royalty: Log royalties were calculated as a percent of the log export value with an adjustment based on the location of the concession. b Area tax: According to the Code Forestier du Gabon (2001), the area tax for concessions with an approved management plan reduced by half. We used the full area tax for the CL area, and half that rate for the RIL area. To determine the cost, we divided the per-hectare fee by the per-hectare timber volume harvested using CL and RIL. c Administration fees: Charges paid for activities expected to be conducted by technicians from the Ministry of Water and Forests: stump hammering (US$2.16 per tree); surveying (US$5.39 per ha); line clearing (US$5.39 per km); and, monitoring of felling (US$1.73 per m3; Code Forestier du Gabon, 2001). All costs were calculated using the 2009-2010 US dollar currency conversion rate of US$1.00=463.05Fcfa.
133
Table 4-4. Profits from conventional logging (CL) and reduced-impact logging (RIL) in the tropics based on values published in the case studies
Logging profit (US$)
Source Locations Logging Method
Per m3 a Per logged hectare b
Holmes et al., 2002 Fazenda Cauaxi, Brazil CL 9.84 256.82 RIL 11.66 290.33
Boltz et al., 2001 Paragominas, Brazil CL 12.00 309.60 RIL 9.16 180.45
van der Hout, 1999 Pibiri, Guyana CL 183.72 5236.02
RIL 120.77 3743.87
Barreto et al., 1998 Fazenda Agrosete, Brazil
CL 13.30 395.01
RIL 14.32 552.75
Tay et al., 2002 Sabah, Malaysia CL 14.25 1938.00
RIL 6.75 728.00
Healey et al., 2000 Sabah, Malaysia CL 14.25 1938.00
RIL 6.75 724.50
Dwiprabowo et al., 2002
East Kalimantan, Indonesia
CL 101.97 8463.51
RL 101.62 6097.20
Dagang et al., 2002 Sarawak, Malaysia CL 75.41 5338.02
RIL 79.66 2531.19
This study Monts de Cristal, Gabon CL 92.34 711.02
RIL 89.10 721.71
Mattsson-Marn et al., 1981
Sarawak, Malaysia CL 70.13 3716.89
RIL 71.96 3957.80
Ahmad et al., 1999 Dungun, Terengganu, Malaysia
CL 144.49 13293.08
RIL 145.03 9426.95
Rahim Nik et al., 2002 c
Terengganu, Malaysia CL 117.33 3895.36
+/-RIL 106.17 3397.44
Rahim Abdul et al., 2009 c
Kelantan, Malaysia CL 93.22 5061.85
+/-RIL 69.76 3536.83
Bacha et al., 2007 Tapajós, Brazil RIL 18.12 367.83
Verissimo et al., 1992 c
Belém, Brazil CL 5.08 193.04
Fisher et al., 2011 c Sabah, Malaysia CL 77.66 11804.32
Ruslandi et al., 2011 c
Kalimantan, Indonesia CL 41.98 2099.00
a Calculated by dividing the net revenue (US$/ha) by the timber volume extracted (m3/ha) and expressed in US$/m3. b Calculated as the difference between gross revenue and total costs (US$/ha) c Studies not included in figures 1 and 2.
134
0 20 40 60 80 100 120 140
-20
24
6
Loggin
g c
osts
(R
IL-C
L; U
S$/m
3) a) Costs per cubic meter using published values
This study-AF
Tay+Healey-AS
Boltz-SA
Holmes-SA
van der Hout-SA
Barreto-SA
Mattsson-Marn-AS
Dagang-AS
Dwiprabowo-AS
Ahmad-AS
0 20 40 60 80 100 120 140
-50
510
Loggin
g c
osts
(R
IL-C
L; U
S$/m
3)
This study-AF
Tay+Healey-AS
Boltz-SA
Holmes-SA
van der Hout-SA
Barreto-SA
Mattsson-Marn-AS
Dagang-AS
Dwiprabowo-AS
Ahmad-AS
b) Costs per cubic meter of the same set of logging activities for all studies
0 20 40 60 80 100 120 140
-200
0200
400
600
Volume extracted (m3/ha)
Loggin
g c
osts
(R
IL-C
L; U
S$/h
a)
This study-AF
Tay+Healey-AS
Boltz-SA
Holmes-SA van der Hout-SA
Barreto-SA
Mattsson-Marn-AS
Dagang-AS
Dwiprabowo-AS
Ahmad-AS
c) Costs per hectare of the same set of logging activities for all studies
Figure 4-1. Costs of RIL compared to CL in relation to harvest volume per hectare: a)
published costs per cubic meter of harvested timber (Table 4-1); b) costs per cubic meter for pre-harvest activities, harvest planning, skid trail layout, log deck construction, and harvest operations ending with the loading of log trucks; c) same as b but expressed as cost per hectare. Values above the dashed line indicate that RIL was more costly to implement than CL. AF = Africa; AS = Asia; SA = South America; surnames of the senior authors are used to further identify the studies.
RIL Cheaper
CL Cheaper
135
1.0 1.5 2.0 2.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
Relative Volume (m3 ha
1; CL/RIL)
Re
lative
Co
sts
(U
S$
m3; C
L/R
IL)
Holmes-SA
Boltz-SA
van der Hout-SA
Barreto-SA
Healey+Tay-AS
Dwiprabowo-AS
Dagang-AS
This study-AF
Ahmad-AS
Mattsson-Marn-AS
Figure 4-2. Relative costs of RIL and CL in relation to the relative timber volume
harvested per hectare. Values above the dashed line indicate that RIL was relatively cheaper to implement than CL. AF = Africa; AS = Asia; SA = South America; surnames of the senior authors are used to further identify the studies.
RIL Cheaper
CL Cheaper
136
1.0 1.5 2.0
1.0
1.5
2.0
Re
lative
Pro
fit (C
L/R
IL; U
S$
m3)
This study-AF
van der Hout-SA
Barreto-SA
Boltz-SA
Holmes-SA
Tay+Healey-AS
Dagang-AS
Dwiprabowo-AS
Mattsson-Marn-AS Ahmad-AS
a) Relative profit per cubic meter using published values
1.0 1.5 2.0
1.0
1.5
2.0
2.5
Relative Timber Volume Extracted (CL/RIL; m3 ha
1)
Re
lative
Pro
fit (C
L/R
IL; U
S$
ha
1)
This study-AF
van der Hout-SA
Barreto-SA
Boltz-SA
Holmes-SA
Tay+Healey-AS
Dagang-AS
Dwiprabowo-AS
Mattsson-Marn-AS
Ahmad-AS
b) Relative profit per hectare using published values
Figure 4-3. Relative profits of RIL and CL: a) published values for profits per cubic meter of harvested timber; b) same as a) but expressed as profit per hectare. Values above the dashed line indicate that RIL was less profitable than CL. AF = Africa; AS = Asia; SA = South America; surnames of the senior authors are used to further identify the studies.
CL Profitable
RIL Profitable
137
CHAPTER 5 POST-HARVEST STAND RECOVERY OF TREE ABUNDANCE AND ABOVE-
GROUND BIOMASS AFTER SELECTIVE LOGGING IN GABON
5.1 Introductory Remarks
In tropical forests managed for timber, there is a great variation in the extent to
which selective logging alters tree population dynamics immediately and during the
subsequent stand recovery period (e.g., Ruslandi et al., 2012). Given the general lack of
post-logging silvicultural treatments in the tropics, the sustained yield of timber mostly
hinges on the logging impacts. In regards to the next harvest, the most critical
population parameters are the survival and subsequent growth of future crop trees
(FCTs; i.e., trees of commercial species but lower than the minimum size for
harvesting). In tropical countries where timber-based industries contribute substantially
to economic development, sustainability of yields is critical for the long-term economic
welfare. While some of the impacts of logging on tree species composition, forest
structure, and above-ground biomass (AGB) are fairly well known (e.g., Pinard and
Putz, 1996; Pinard and Cropper, 2000; Brown et al., 2005; Blanc et al., 2009; Mazzei et
al., 2010; Medjibe et al., 2011), there is a need for better information on post-logging
recovery, especially in the Congo Basin. Here we provide information about post-
logging dynamics of tree populations and recovery of AGB based on data collected for
two years after reduced-impact logging (RIL) in Gabon. We then use these field data to
project future stand conditions using a matrix model.
Although logging is a major source of anthropogenic disturbance to tropical forests
(Asner et al., 2011), its impacts vary with the intensity of harvesting and the care with
which it is carried out (Picard et al., 2012). For example, where logging results in the
creation of large gaps that are suitable for the growth and recruitment of pioneer
138
species, rates of timber volume recovery can dampened or accelerated, depending on
the silvics of the commercial species (see Denslow, 1987; Jennings et al., 1999; Hall et
al., 2003; Wright et al., 2003; Makana and Thomas, 2005) while hampering the
regeneration of other species (i.e., shade tolerant species). But it can also result in
considerable residual stand damage (White, 1994; Johns et al., 1996; Whitman et al.,
1997; Jackson et al., 2002, Medjibe et al., 2011). In any case, management can be
improved with the data-based projections of the long-term impacts of logging on the
dynamics of tree populations and forest biomass; the need for such information is
particularly critical in the Congo Basin.
A wide variety of stand models exits to simulate the long-term dynamics of tree
populations (Gourlet-Fleury et al., 2005). Some models are spatially explicit single-tree
distance-dependent models (Kohler and Huth, 1998; Chave, 1999; Gourlet-Fleury and
Houillet, 2000; Kammesheidt et al., 2001; Kohler et al., 2001; Phillips et al., 2004).
Others approaches include gap models (Shugart et al., 1980; Pastor and Post, 1986;
Bossel and Krieger, 1991; Acevedo et al., 1995) and matrix models that describe forest
structure (Gourlet-Fleury et al., 2005; Fortini and Zarin, 2011). Matrix models have been
widely used to examine population structure and to simulate the effects of logging on
forest structure and composition (Karsenty and Gourlet-Fleury, 2006). Furthermore,
matrix models have been used in tropical forest to deal with the economy of timber
production (Boscolo and Vincent, 2000), carbon sequestration (Boscolo et al., 1998),
and forest management (Macpherson et al., 2010; Boscolo and Vincent, 2000; Gourlet-
Fleury et al., 2005; Chikumbo and Steward, 2007; Fortini and Zarin, 2011). Although
matrix models suffer some limitations when used to predict the recovery of a species
139
population especially over several cutting cycles (Gourlet-Fleury et al., 2005), they are
useful in making projections of tree diameter class distributions that may be used to
estimate the time needed for stand recovery after a single logging entry (Caswell,
2001).
This study uses a matrix model to simulate post-logging recovery of tree
abundance and AGB under different scenarios of logging intensity and cutting cycles.
The model that we apply to a selectively logged forest in Gabon was calibrated using
data from unlogged permanent plots in Central African Republic. Sensitivity analyses
were also conducted to examine changes in tree abundance and forest biomass with
changes in rates of tree growth and mortality.
5.2 Methods
5.2.1 Study Site
The study site is located in the SEEF (Société Equatoriale d’Exploitation
Forestière) concession on the Monts de Cristal in Gabon. The research was conducted
in the 10 1-ha permanent plots in the RIL area (see Chapter 2, section 2.2.1 for
description of the study site and section 2.2.2 for plot measurements).
5.2.2 Tree Growth
Trees were measured 2 months prior to RIL operations in 2009 and approximately
two years later. Data from stems with characteristics that made them difficult to
measure with accuracy (e.g., prominent swellings or fissures) as well as those with
aberrant growth rates (i.e., extreme growth or shrinkage) were not included in the
growth rates calculation (e.g., 171 trees,, 4% of total; see Condit et al., 1993; van der
Hout and Zagt, 2005).
140
5.2.3 Tree Mortality and Recruitment
Trees were aggregated into ten 10-cm DBH classes. Tree mortality was
calculated as:
Mortality = 1- ((No-Nm)/No)(1⁄θ) where No is the total numbers of stems at the
beginning of time interval θ, and Nm is the number of trees that died during the interval
(Sheil et al., 1995). New recruits were identified to species, tagged, and mapped. The
bisection method was use to estimate recruits necessary to reach the "carrying
capacity" based on observed population size. The bisection method has been used in
studies associated with root-finding to show that a continuous function on a closed
interval achieves its maximum (Wood, 1992).
5.2.4 Functional Groups
To facilitate the understanding of the general trends in tree population dynamics,
tree species were classified into functional groups based on their environmental
requirements for recruitment and growth. Following Hawthorne (1995), we used three
functional groups: pioneer; non-pioneer light-demanding (NPLD); and, shade tolerant
(ST) species.
• Pioneer species require light for recruitment and regenerate only in gaps. They are characterized by small seeds, early reproduction, rapid colonization of canopy openings, rapid growth, and low wood density.
• NPLD species can regenerate in the shade but require light for growth and survival at later stages of their life cycle and are capable of rapid growth under high light conditions.
• ST species regenerate and perform well under shaded conditions. They typically have large seeds, seedlings capable of prolonged survival in the shaded forest understory, high density wood, and low mean and maximum growth rates.
141
5.2.5 Model Structure
The basic structure of the model, which is a density dependent model, was based
on studies by Boscolo and Vincent (2000) and Macpherson et al. (2010). The model
simulated an equilibrium number of trees (E) that is reached when the number dying is
equal to the number of recruits. The total mortality increases as a function of population
size. The equilibrium population size is an input to the model. Recruitment numbers,
which are difficult to measure in the field, can be set to a constant rate that balances
mortality at E. Bisection can be used to estimate the recruitment necessary to reach the
selected value of either E or a set population size at a particular time following
disturbance. The growth model was based upon the characterization of forest structure
and composition using tree diameter distribution. The model is structured as:
Yt+θ =G (Yt - ht - dt) + Rt where θ is the growth interval. The pre-harvest stand state
was given by the vector Yt= [Yijt] where Yijt is the number of trees of group i (=1,.…,m)
and diameter size class j (=1,….,n) alive at time t. In this application, the size classes
range from > 10 cm dbh to >100 cm. The harvest of trees of group i and size class j at
time t is given by ht= [hijt], a vector of the same dimension. Residual stand damage is
assumed to be function of the logging intensity (i.e., harvested stems ha-1) and is
represented by dt = ( hijt) x D x Yt where D is a matrix of logging damage for each
tree in group i and diameter class j with the diagonal containing the proportion of trees
damaged. Rt is a vector representing the recruitment parameters, as a function of stand
density, by diameter class at time t. G is a transition matrix of trees in each diameter
class:
142
where Si represents the stasis probability, the probability that a tree in species group i
and diameter size class j remain alive and in size class j during the time interval θ; Gi
the probability that a tree in group i and size class j remains alive and grows to next
diameter size class during the same time interval.
5.2.6 Transition Probabilities
Transition probabilities were calculated based on Crouse et al. (1987):
Gi = prob(survival)i × prob(growth)i where Gi is the probability of growing to next
class; Stasis = prob(survival)i ×(1-prob(growth)i) where Stasis is the probability of
remaining in the same class. Transition probabilities were calculated for tree abundance
and above-ground biomass over the time interval θ for all species (Table 5-1) and
species functional groups (Appendix H).
5.2.7 Model Runs
Because logging intensity predicts residual stand damage, the growth model was
run for tree abundance and AGB two years after logging using the following scenarios:
1) no changes to the recorded logging intensity (0.82 trees ha-1) and the official cutting
cycle of 25 years; 2) an increased in logging intensity slightly beyond the range of the
intensity in Central Africa (Ruiz-Perez et al., 2005) to 2 and 5 trees ha-1 using the same
cutting cycle; and 3) an extension of the cutting cycle to 30 years using the previous
logging intensities. The model was run for 50 years assuming that this is a reasonable
143
timeframe to detect changes in tree species abundances and biomass recovery. In the
model, tree mortality was assumed constant throughout the cutting cycle. Simulation
analyses were conducted with Python 2.6.6 Tk version 8.5 (www.python.org). Additional
analyses were conducted with the R program, version 2.10.1 (R Development Core
Team, 2010).
5.2.8 Sensitivity Analyses
Alternative assumptions were incorporated into the model to evaluate the relative
influence of different parameters on tree abundance and forest biomass over a 50-year
simulation period. Assuming that very large or small growth and mortality rates relative
to the field estimates are less plausible, the tests were conducted by increasing and
reducing growth and mortality rates by 5%. Matrix population models are often
evaluated in this way, using elasticity of the matrix elements with respect to the
dominant eigenvalue (Caswell, 2001).
5.3 Results
5.3.1 Tree Abundance
Two years after logging, the diameter class structure of tree population in the ten
1-ha permanent plots was different than before logging (Table 5-2). Overall, total tree
abundance declined by 5% (Figure 5-1a) and differed among functional groups (F =
63.8; df = 783; p <0.001). The abundance of NPLD species declined by 3% (Figure 5-
1b), while pioneer species declined by 15% (Figure 5-1c), and ST species declined by
4% (Figure 5-1d). Trees in the small dbh classes (i.e., ≤ 30 cm) had higher stem
densities for all species groups, while the larger dbh classes contained fewer individuals
(Table 5-3; Figure 5-1). Trees ≥ 70 cm dbh represented only 3% of the total trees while
NPLDs include 6% of the total trees in contrast to pioneers (3%) and STs (2%; Figure 5-
144
1). The abundance of NPLD species in the larger dbh classes could be linked to the
growth and survival capacity of these species under high light conditions. ST species
(90% of the total species of this functional group) dominated the small dbh classes (i.e.,
≤ 30 cm) compared to NPLDs (78%) and pioneers (86%) as a consequence of the
persistence of these species in the forest understory (Figure 5-1).
5.3.2 Tree Growth and Recruitment
Stem diameter growth varied among plots (Table 5-2) and diameter classes (Table
5-3) averaging 0.71 ± 0.02 cm yr-1 (mean ± S.E), with higher growth of pioneer species
(0.96 ± 0.12 cm yr-1) relative to NPLDs (0.76 ± 0.06 cm yr-1) and STs (0.68 ± 0.03 cm yr-
1). Overall, stem diameter growth differed among functional groups (F = 4.0; df = 4166;
p = 0.02). Pioneer species grew 29% faster than STs and 21% than NPLDs (t = 2.3, p =
0.019; Figure 5-2a). In contrast, there was considerable variability in tree growth per
stem dbh class. Individuals in the 50 cm dbh class grew slowly while those in 20-40 cm
class grew relatively quickly (Table 5-3). For NPLDs, individuals in all dbh classes
showed fast growth except those in the 50 and ≥ 100 cm dbh. For pioneer species, slow
growth was observed for trees in the 10 and ≥ 100 cm dbh classes in contrast to the
STs that showed considerable variability in tree growth with only individuals in the 90
cm dbh class having fast growth (Appendix I).
Tree recruitment also varied among plots. Two plots had large number of recruits
while in two others recruitment was scarce (Table 2). All the recruits were restrictied to
the 10 cm stem dbhclass (Table 3; Appendix I).
5.3.3 Tree Mortality
Logging, on average, resulted in the death of 9.2 trees ha-1, which is substantially
higher than mortality from natural causes with the highest mortality recorded in plot 1
145
(Table 5-2). Overall, annual mortality of trees by logging (2.0 ± 0.0%) was higher than
mortality of undamaged trees (1.0 ± 0.0 %; Tables 5-2 and 5-3). Mortality also varied
among stem diameter classes with high natural mortality recorded for trees in the 90 cm
dbh (Table 5-3). In general, mortality induced by logging was substantially higher for
trees ≥ 80 cm dbh due in part to the harvest of trees in this size class (Table 5-3). STs
had the highest tree mortality, mostly of trees ≤ 30 cm dbh, followed by NPLDs and
pioneer species (Appendix I). These differences might be explained by the higher stem
density of STs relative to NPLDs and pioneer species (Figure 5-1).
5.3.4 Above-Ground Biomass Increments
Of the ten 1-ha plots, two had biomass growth more than one standard deviation
above the mean and only one had growth less than half of any other plot (Table 5-4).
Overall, there was an average AGB accumulation of 3.8 ± 1.07 Mg ha-1 yr-1, with major
contribution from growth (3.2 ± 1.31 Mg ha-1 yr-1) and recruitment into the 10 cm dbh
class (0.5 ± 0.19 Mg ha-1 yr-1; Table 5-4; Figure 5-2b). Species functional group
explained much of the variation in AGB growth (F = 2.51; df = 4169; p = 0.017). After
logging, pioneer species had 41% higher biomass growth than STs and 34% higher
than NPLDs (t = 1.99; p = 0.04), while biomass growth of NPLDs was only 11% higher
than STs. However, there was substantial variability in the AGB growth distribution
within diameter size classes and functional groups (F = 65.5; df = 783; p <0.001). For all
species combined, trees ≥ 100 cm contributed by only 0.10 Mg ha-1 yr-1 to the biomass
growth (Table 5-5) even though they contributed much of the total biomass (Table 5-5;
Figure 5-3a), similarly to NPLDs (Figure 5-3b) and pioneer species (Figure 5-3c). ST
species, in contrast, showed a higher proportion of the biomass growth in trees ≤ 60 cm
146
dbh (Figure 5-3d), which is not surprising since trees ≥ 70 cm represented only 2% of
the total individuals of this functional group (Figure 5-1d).
Despite the high biomass of large trees, trees ≤ 40 cm had higher biomass growth
than trees ≥ 90 cm (Table 5-5). The large contribution to the biomass gain came from
the growth of trees ≤ 40 cm dbh and in-growth of trees in the 10 cm dbh classes for all
tree species combined (Table 5-5) and by species functional group (Appendix J).
5.3.5 Above-Ground Biomass Loss
Logging operations severely damaged one of the plots (ref. Chapter 2, Table 2-3),
which resulted in large net negative biomass balance (Table 5-4). Biomass loss induced
by logging averaged 7.8 ± 9.75 Mg ha-1 yr-1 and from natural mortality of trees averaged
2.6 ± 3.83 Mg ha-1 yr-1 (Table 5-4). Therefore, there was an average net AGB balance
(i.e., emission factor) of -6.6 ± 10.78 Mg ha-1 yr-1. Two years after logging, 95% of the
pre-harvest AGB was retained in the forest. The harvest of trees ≥ 70 cm dbh
contributed to the high emissions while growth of trees ≤ 40 cm dbh and ingrowth
helped reduce the emissions (Table 5-5, Appendix J). Losses from natural tree mortality
however, varied within stem diameter class where natural loss from trees ≥ 60 cm dbh
was larger than from trees ≤ 40 cm dbh (Table 5-5). The loss from logging was
substantially higher for both NPLDs and STs trees ≤ 40 cm dbh than from natural
mortality (Appendix J).
5.3.6 Evaluation of the Simulation Model
5.3.6.1 Simulated tree abundance
Over a 50 year time span, simulated tree abundance fluctuated between 414.5
and 454.5 stems ha-1 after logging based on the recorded logging intensity (i.e., 0.82
trees ha-1) at a 25-year cutting cycle. Under this scenario, tree abundance reached the
147
pre-harvest stem density 23 years after logging (Figure 5-4a), as it was when the cutting
cycle was extended by 5 years (i.e., 30 years; Figure 5-4b). Considerable variation in
simulated tree abundance was observed among functional groups. NPLDs and STs
showed a quick decline in abundance after harvest and a steady increase a few years
later. Abundance fluctuated between 101 and 108 stems ha-1 for NPLDs and 274 and
300 stems ha-1 for STs, both reaching an asymptote at the end of the 25-year cutting
cycle (Figure 5-5). In contrast, pioneer species showed a peak in post-harvest
abundance that fluctuated between 38 and 35 stems ha-1 over the cutting cycle (Figure
5-5). This decline in tree abundance might result from the physiological characteristics
of species in this functional group.
Increased logging intensity drastically affected simulated tree abundance after
harvest as tree abundance did not reach the initial density after intensities of 2 or 5
trees ha-1. Under the normal cutting cycle, post-harvest tree abundance fluctuated
between 383 and 437 for the extraction of 2 trees ha-1 and between 306 and 393 stems
ha-1 for extracting 5 trees ha-1. When extending the normal cutting cycle to 30 years and
with an extraction rate of 2 trees ha-1, tree abundance fluctuated between 383.9 and
444.7 stems ha-1 and did not reach the initial density at the end of the cutting cycle
(Figure 5-4b). Under this scenario, tree abundance could probably be at the nominal
pre-harvest level 33 years after logging. These scenarios showed that under either
cutting cycle with increased in extraction rates, tree abundance is not projected to reach
the prior logging stems density (Figures 5-4a and 5-4b). Logging practices that exceed
these levels should be treated with caution and monitored carefully to assess recovery
potential.
148
5.3.6.2 Simulated Above-Ground Biomass
Post-harvest AGB recovery varied with logging intensity and felling cycle (Figures
5-6a and 5-6b). Using the normal cutting cycle and the recorded logging intensity, the
simulated post-logging AGB fluctuated between 386 and 422 Mg ha-1 and recovered the
initial biomass (420 Mg ha-1) 24 years after logging. Contrary to the simulated tree
abundance, simulated AGB of the three species functional groups showed a decline
after harvest and then an increase over time. With the observed logging intensity (i.e.,
0.82 trees ha-1), the simulated AGB of NPLDs and pioneer species reached the initial
value right at the end of the 25-year cutting cycle, while STs reached the initial AGB 21
years after logging (Figure 5-7).
For all species combined, post-logging AGB fluctuated between 359 and 407 Mg
ha-1 and is predicted to recover the initial biomass after 33 years with a logging intensity
of 2 trees ha-1. For an intensity of 5 trees ha-1, the simulated post-logging biomass
fluctuated between 291 and 370 Mg ha-1 and is not expected to reach the initial biomass
(Figure 5-6a). Similar to the normal felling cycle, scenarios with the felling cycle
extended to 30 years result in post-harvest forest biomass recovery after 24 years
(Figure 5-6b). If 2 trees ha-1 were extracted, post-harvest AGB could barely make the
initial value (417.3 vs. 420.5 Mg ha-1) at the end of the cutting cycle. The simulated AGB
also showed that under high logging intensity (i.e., ≥ 5 trees ha-1), post-harvest AGB
would take longer to reach the prior-logging AGB in this study site (Figure 5-6b).
5.3.7 Variability in Tree Abundance and Above-Ground Biomass
5.3.7.1 Tree abundance and above-ground biomass
Predictions of post-harvest tree abundance over 50 years are most sensitive to
changes in tree mortality. Increasing tree mortality by 5% reduced stem density by 5%
149
throughout the cutting cycle, whereas reducing tree mortality by 5% increased stem
density by 2% (Figure 5-8). If simulated tree mortality was reduced by 5%, tree
abundance could achieve the initial stem density by 19 years after logging. However,
tree abundance was relatively stable with respect to the 5% changes in tree growth.
Increasing the stem growth by 5% resulted in only a 3% increase in tree abundance. A
similar reduction in stem growth resulted in a 3% decrease, but tree abundance could
attain the initial stem density one year earlier with a 5% increase in tree growth (Figure
5-8).
Post-logging AGB recovery was sensitive to changes in tree growth and mortality.
When tree morality is reduced by 5%, biomass recovered 17 years after logging,
whereas if tree mortality is increased by 5%, biomass recovered after 21 years. With an
increased in tree growth of 5%, biomass reached an asymptote of 441 Mg ha-1 at the
end of the cutting cycle (i.e., 25 years; Figure 5-9). However, if tree growth is reduced
by 5%, biomass reached an asymptote at 440 Mg ha-1, 1.3% lower than with the 5%
reduction in tree mortality. The sensitivity analyses showed that tree mortality
constitutes a fundamental factor influencing tree abundance and forest biomass after
logging.
5.3.7.2 Tree Mortality
In this study, tree mortality induced by logging was higher than natural tree
mortality (t = 2.19; df = 9; p = 0.03; Table 5-2; Appendix I). The model predicting tree
mortality rates showed a decrease over time, but natural death of trees depends on the
environmental conditions that govern the survivorship of individual tree (Figure 5-10a).
Similar patterns were also observed among species functional groups with faster
declines in tree mortality of NPLDs and pioneer species over time (Figure 5-10b).
150
5.4 Discussion
5.4.1 Tree Abundance
Considerable variability in tree abundance was found among plots and stem
diameter classes. As expected, tree abundance in plots that suffered severe damage
from felling and skidding declined as a result of the death of damaged trees. Most of the
damaged trees were in the smaller diameter classes that are dominated by ST species
followed by NPLDs. Not surprising is the dominance of ST species in these size
classes, which explained the capacity of species of this functional group to persist for
long periods in the forest understory. In contrast, few individuals of ST species were
found in the larger stem size classes that were dominated by NPLD and pioneer
species.
Despite the decline in tree abundance (i.e., 3%), logging apparently had positive
effects on the subsequent growth and recruitment of tree species. Two years after
logging we found that residual trees grew in stem diameter at an average rate of 0.71
cm yr-1, which is within the range reported in Gabon (0.7-1.52 cm yr-1; Furh et al., 1998)
and helps explain the rapid stand recovery. These results were supported by the
simulation analysis that predicted tree abundance to reach the initial stem density 2
years earlier before the end of the normal cutting cycle (i.e., 25 years) even through the
patterns varied among functional groups and was dependent on the low logging
intensity. As reported in Cameroon, low intensity logged site recovered within 14 years
(van Gemerden et al., 2003). We did find that tree abundance is very sensitive to
mortality, which due to logging was very high relative to trees that died from natural
causes. Natural tree mortality is indeed difficult to assess since it depends on
environmental conditions that govern growth and survival of tree species. In contrast,
151
mortality due to logging is avoidable and controllable through better harvest planning
and operations. More importantly, our model showed that any increase in logging
intensity should be accompanied by increased care in the harvest operations if the
forest management objective is to reach the initial stem density at the end of each
cutting period. For example in Ghana, Hawthorne et al. (2012) reported that annual tree
mortality rates return to normal rates of less than 2% after 15 years in the unlogged
forest and 22 years in the logged forest. We expect that faster decline in tree morality,
therefore reduction in collateral logging damage, would favor rapid recovery in tree
abundance.
5.4.2 Above-Ground Biomass
Two years after logging, AGB losses were still greater than gains. 75% of the AGB
loss was from trees that were harvested or destroyed during logging. Undamaged trees
that died from natural causes contributed 25% to the total ABG loss, but we expect a
rapid AGB balance with an average gain rate of 3.8 Mg ha-1 yr-1 from growth and
recruitment. However, contributions to AGB growth differed among functional groups.
Pioneer species showed faster AGB growth but they were less abundant than shade
tolerant species. In contrast, shade tolerant species had substantially higher AGB in all
dbh classes compared to pioneer and NPLD species. Such variability in AGB density
among functional groups explained the capacity of individual trees species to sequester
and store carbon.
The simulation showed that with the current accumulation rate and at low logging
intensity, AGB will recover before the end of the recommended cutting cycle.
Additionally, any increase in logging intensity will result in longer AGB recovery time
(see e.g., Blanc et al., 2009; Mazzei et al., 2010). Therefore, AGB cannot attain the
152
initial value with any increase in logging intensity within the recommended current
cutting cycle using current harvesting practices.
5.5 Conclusion
Logging alters the biophysical structure of forests but can create conditions that
are favorable for some tree species and functional groups. Despite the limitations on a
projection of tree abundance and AGB growth based on only 2-years of post-logging
tree data, it is clear that logging intensity, collateral damage during logging and
subsequent tree mortality substantially influence residual tree abundance and forest
biomass recovery. Therefore, timber extraction operations should be carried out with
caution and monitored carefully to favor post-logging recovery.
153
Table 5-1. Transition probabilities and other parameters used in the matrix model. Gi is the probability of growing to the next class; Stasis in the probability of remaining in the same class.
DBH Class (cm)
Gi Stasis Tree Mortality (%)*
Proportion of damage trees
Recruitment (stems ha-1 yr-1)
Tree abundance 10-20 0.011 0.960 3.1 0.06 12.5 20-30 0.018 0.964 1.9 0.06 - 30-40 0.019 0.969 1.2 0.06 - 40-50 0.018 0.966 1.7 0.06 - 50-60 0.009 0.991 0.0 0.06 - 60-70 0.018 0.972 1.1 0.04 - 70-80 0.010 0.946 4.5 0.07 - 80-90 0.016 0.952 3.4 0.07 - 90-100 0.018 0.903 8.2 0.16 - > 100 0.009 0.947 4.7 0.00 - Above-ground biomass Loss rate (%) Mg ha-1 yr-1 10-20 0.064 0.909 2.7 0.06 0.657 20-30 0.072 0.911 1.7 0.06 0.732 30-40 0.061 0.929 1.0 0.06 0.620 40-50 0.045 0.940 1.6 0.07 0.455 50-60 0.016 0.984 0.0 0.08 0.159 60-70 0.029 0.933 3.8 0.04 0.298 70-80 0.012 0.971 1.7 0.07 0.125 80-90 0.009 0.954 3.8 0.06 0.091 90-100 0.004 0.808 1.9 0.12 0.046 > 100 0.009 0.991 0.0 0.00 0.092
* Annual mortality of damaged and undamaged trees.
154
Table 5-2. Changes in number of trees in the 10 1-ha permanent plots two years after reduced-impact logging on Monts de Cristal in Gabon.
2009 Trees (stems ha-1)
Gain (stems ha-1)
2009-2011 Mortality (Stems ha-1) Trees (stems ha-1)
Growth (cm yr-1)
Tree Mortality (%)
Plot Before logging
Ingrowth # Tree harvested
Damaged trees
Undamaged trees
Total 2011 2009-2011
Natural Logging
RIL-P1 463 9 1 50 8 59 413 0.79 1.0 6.0 RIL-P2 440 7 1 24 7 32 415 0.81 1.0 3.0
RIL-P3 432 17 - 13 11 24 425 0.65 1.0 2.0 RIL-P4 445 5 2 13 9 24 426 0.67 1.0 2.0 RIL-P5 520 3 1 1 2 4 519 0.83 - -
RIL-P6 471 8 1 19 3 23 456 0.69 - 2.0 RIL-P7 478 11 - 8 2 10 479 0.70 - 1.0 RIL-P8 462 14 1 16 14 31 445 0.99 1.0 2.0
RIL-P9 391 3 - 1 4 5 389 0.51 1.0 - RIL-P10 425 10 - 10 3 13 422 0.45 - 1.0 Mean±1s.d 452.7±34.8 8.7±4.6 0.7 15.5±14.1 6.3±4.2 22.5±16.3 438.9±37.6 0.71±0.16 1.0±0.0 2.0±0.0
155
Table 5-3. Changes in number of trees per stems diameter class in the 10 1-ha permanent plots two years after reduced-impact logging on Monts de Cristal in Gabon.
2009 Trees (Stems ha-1)
Gain (stems ha-1)
2009-2011 Mortality (stems ha-1) Trees (Stems ha-1)
Growth (cm yr-1)
Mortality rate (%)
DBH Class (cm)
Before logging
Ingrowth # Tree harvested
Damaged trees
Undamaged trees
Total 2011 2009-2011
Natural Logging
10 266.8 87 - 11.9 4.2 16.1 254.9 0.58 0. 8 2.3
20 86.4 - - 2.3 0.9 3.2 85.5 0.90 0. 5 1.3
30 42.0 - - 0.6 0.4 0.1 40.8 0.97 0. 5 0. 7
40 23.9 - - 0.5 0.3 0.8 24.9 0.93 0. 6 1.1
50 11.4 - - - - - 11.5 0.46 - -
60 9.5 - - - 0.2 0.2 9.4 0.80 1.1 -
70 4.5 - - 0.3 0.1 0.4 4.5 0.54 1.1 3.4
80 3.0 - 0.1 0.1 0.1 0.2 2.9 0.82 1.7 1.7
90 1.9 - 0.2 0.2 0.1 0.3 1.5 0.67 2.7 5.4
100+ 3.3 - 0.4 0.3 - 0.3 3.0 0.48 - 4.7
Mean 452.7 8.7 0.7 1.6 0.6 2.2 438.9 0.71 1.0 2.0
156
Table 5-4. Changes in above-ground biomass in the 10 1-ha permanent plots two years after reduced-impact logging on Monts de Cristal in Gabon.
Total AGB (Mg ha-1)
2009-2011 Gain (Mg ha-1 yr-1) 2009-2011 Mortality (Mg ha-1 yr-1)
Net Balance (Mg ha-1 yr-1)
Plot Before logging
Growth Ingrowth Total Logging Natural Total
RIL-P1 431.1 2.6 0.3 2.9 -30.3 -3.5 -33.8 -30.9
RIL-P2 315.8 3.6 0.3 3.9 -10.1 -3.4 -13.5 -9.6
RIL-P3 508.2 1.0 1.7 2.7 -0.8 -12.9 -13.7 -11.0
RIL-P4 293.4 2.3 0.3 2.6 -6.9 -1.9 -8.9 -6.3
RIL-P5 508.2 5.1 0.2 5.3 -3.5 -0.2 -3.7 1.6
RIL-P6 468.3 2.4 1.7 4.1 -19.1 -1.4 -20.4 -16.3
RIL-P7 357.3 5.1 0.3 5.4 -3.9 -0.2 -4.0 1.4
RIL-P8 305 3.8 0.4 4.2 -0.7 -1.6 -2.3 1.9
RIL-P9 506.1 2.4 0.1 2.5 - -0.5 -0.5 2.0
RIL-P10 511.1 3.9 0.3 4.2 -2.9 -0.5 -3.4 0.8
Mean ± 1 s.d 420.5±92.96 3.2±1.31 0.5±0.19 3.8±1.07 -7.8±9.75 -2.6±3.83 -10.4±10.37 -6.6±10.78
157
Table 5-5. Changes in above-ground biomass per stems diameter class in 10 1-ha permanent plots two years after reduced-impact logging on Monts de Cristal in Gabon.
DBH Class (cm)
2009 AGB (Mg ha-1) Gain (Mg ha-1 yr-1) Loss (Mg ha-1 yr-1) Net Balance (Mg ha-1 yr-1)
Before logging Growth Ingrowth Total Logging Natural Total
10 295.5 0.64 0.49 1.13 -0.57 -0.21 -0.79 0.34 20 388.2 0.70 - 0.70 -0.43 -0.22 -0.65 0.05 30 453.9 0.61 - 0.61 -0.30 -0.15 -0.45 0.16 40 505.4 0.45 - 0.45 -0.54 -0.25 -0.79 -0.34 50 400.2 0.16 - 0.16 - - - 0.16 60 506.5 0.30 - 0.30 - -0.59 -0.59 -0.29 70 337.1 0.13 - 0.13 -1.31 -0.30 -1.61 -1.48 80 297.8 0.08 - 0.08 -0.26 -0.42 -0.69 -0.61 90 246.5 0.05 - 0.05 -0.67 -0.47 -1.14 -1.09 100+ 773.3 0.10 - 0.10 -3.73 - -3.73 -3.63 Mean 420 0.32 0.05 0.38 -0.78 -0.26 -1.04 -0.66
158
10 20 30 40 50 60 70 80 90 100
log10 (
Tre
e A
bundance;
Ste
ms h
a1)
0.0
00.0
50.1
00.1
50.2
00.2
5a) All Species 2009
2011
10 20 30 40 50 60 70 80 90 100
0.0
00.0
50.1
00.1
50.2
0
b) N.P.L.D Species
10 20 30 40 50 60 70 80 90 100
DBH Class (cm)
log10 (
Tre
e A
bundance;
Ste
ms h
a1)
0.0
00.0
40.0
80.1
2
c) Pioneer Species
10 20 30 40 50 60 70 80 90 100
DBH Class (cm)
0.0
00.0
50.1
00.1
50.2
00.2
5
d) Shade Tolerant Species
Figure 5-1. Variability in stems density per dbh class in 2009 and 2011. A) All species,
B) Non-pioneer light demander species. C) Pioneer species. D) Shade tolerant species
159
Ste
m D
iam
ete
r G
row
th (
cm
yr
1)
0.0
0.2
0.4
0.6
0.8
1.0
All NPLD Pioneer ST
a)
Above-G
round B
iom
ass G
row
th (
Mg h
a1 y
r1)
01
23
45
6
All NPLD Pioneer ST
b)
Figure 5-2. Post-logging stem diameter and above-ground biomass growth for all species (n = 4172) and by functional group: NPLD (n = 1035), Pioneer (n = 288), and ST (n = 2849).
160
10 20 30 40 50 60 70 80 90 100
Above-G
round B
iom
ass (
Mg h
a1)
02
46
810
a) All Species
10 20 30 40 50 60 70 80 90 100
01
23
45
67
20092011
b) N.P.L.D Species
10 20 30 40 50 60 70 80 90 100
DBH Class (cm)
Above-G
round B
iom
ass (
Mg h
a1)
02
46
810
c) Pioneer Species
10 20 30 40 50 60 70 80 90 100
DBH Class (cm)
01
23
4
d) Shade Tolerant Species
Figure 5-3. Variability in above-ground biomass per dbh class in 2009 and 2011. A) All species. B) Non-pioneer light demander species. C) Pioneer species. D) Shade tolerant species.
161
Figure 5-4. Post-logging tree abundance over timer using different logging intensities. A) The government-mandated minimum cutting cycle. B) A scenario with a 5-year extension of the cutting cycle.
162
Figure 5-5. Post-logging tree abundance over time using the observed logging intensity and the government mandated minimum cutting cycle per species functional group.
163
Figure 5-6. Post-logging above-ground biomass growth over timer using different logging intensities. A) The government-mandated minimum cutting cycle. B) a scenario of a 5-year extension of the cutting cycle.
164
Figure 5-7. Post-logging above-ground biomass over time using the observed logging intensity and the government mandated minimum cutting cycle per species functional group.
Figure 5-8. Variability of tree density associated with hypethetical changes 5% in post-harvest tree growth and mortality using observed logging intensity.
165
Figure 5-9. Variability of above-ground biomass associated with hypothetical changes by 5% in post-harvest tree growth and mortality using observed logging intensity.
166
Figure 5-10. Trends of the post-harvest tree mortality. A) Natural and logging-induced tree mortality. B) Species functional group. Red colors represented mortality 2 years after logging and the following dark colors are tree mortality from the model.
167
APPENDIX A CHARACTERISTICS OF TREE SPECIES IN THE 10 1-HA PERMANENT PLOTS IN THE RIL ZONE OF THE SEEF
CONCESSION ON MONTS DE CRISTAL, GABON
Species name Average dbh (cm)
Maximum tree size
(cm)
Basal area (m2
ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density
(%)
Relative dominance
(%)
Relative Importance
(%)
Acioa pallenscens 13.2 16.2 0.03 0.2 0.2 0.04 0.01 0.29
Acioa sp. 19.3 27.6 0.16 0.5 1.6 0.11 0.05 0.39
Afrostyrax lepidophyllus 20.8 40.8 0.21 0.5 2.0 0.11 0.07 0.53
Albizia andianthifolia 59.6 59.6 0.28 0.1 4.4 0.02 0.09 0.23
Allanblackia sp. 38.3 51.6 0.26 0.2 3.4 0.04 0.08 0.24
Amanoa sp. 17.8 32.8 0.15 0.5 1.0 0.11 0.05 0.51
Amphimas ferrugineus 12.8 12.8 0.01 0.1 0.1 0.02 0.00 0.14
Angylocalyx pynaertii 14.0 20.7 0.18 1.1 1.3 0.24 0.06 0.77
Angylocalyx sp. 12.8 18 0.08 0.6 0.5 0.13 0.03 0.62
Anisophyllea myriostica 34.1 74.1 6.23 5.3 82.0 1.17 2.00 3.76
Anisophyllea polyneura 16.1 43.3 1.75 7.3 14.8 1.61 0.56 3.22
Anisophyllea purperensis 13.7 30.9 0.79 4.9 5.5 1.08 0.25 1.92
Anisophyllea sp. 19.7 56.8 1.32 3.3 13.7 0.73 0.42 1.85
Anthonotha fragrans 29.4 36.9 0.21 0.3 1.8 0.07 0.07 0.37
Anthonotha sp. 25.8 59.7 0.61 0.9 6.0 0.20 0.20 0.75
Aphanocalyx margininervatus
37.4 41.7 0.34 0.3 2.9 0.07 0.11 0.29
Aphanocalyx microphyllus
27.8 47 0.45 0.6 3.8 0.13 0.15 0.51
Araliopsis sp. 11.9 16.2 0.12 1.1 0.7 0.24 0.04 0.75
Aucoumea klaineana 72.9 124.4 10.22 2.1 101.3 0.46 3.29 4.57
Baikieae insignis 12.2 12.2 0.01 0.1 0.1 0.02 0.00 0.14
Baphia sp. 16.0 31 0.10 0.4 1.0 0.09 0.03 0.59
168
Species name Average dbh (cm)
Maximum tree size
(cm)
Basal area (m2
ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density
(%)
Relative dominance
(%)
Relative Importance
(%)
Barteria fustulosa 14.5 22.6 0.11 0.6 0.8 0.13 0.03 0.86
Beilschmiedia congolana 19.0 64.2 1.48 3.6 12.8 0.80 0.48 2.20
Beilschmiedia fulva 36.6 62.3 0.31 0.2 3.6 0.04 0.10 0.38
Beilschmiedia sp. 14.2 20.1 0.12 0.7 0.6 0.15 0.04 0.54
Berlinia congolensis 20.8 24.4 0.07 0.2 0.6 0.04 0.02 0.18
Berlinia sp. 15.5 15.5 0.02 0.1 0.1 0.02 0.01 0.14
Bridelia sp. 22.9 38.1 0.66 1.4 5.3 0.31 0.21 0.99
Calpocalyx dinklagei 21.8 41.2 0.41 0.9 4.9 0.20 0.13 0.68
Canarium schweinfurthii 43.0 52.1 0.45 0.3 4.4 0.07 0.14 0.44
Canthium arnoldiana 17.8 17.8 0.02 0.1 0.2 0.02 0.01 0.15
Canthium sp. 44.5 67 0.39 0.3 5.8 0.07 0.13 0.42
Carapa procera 30.6 46.5 0.86 1.1 8.9 0.24 0.28 1.22
Carapa sp. 29.1 64.6 0.47 0.5 5.9 0.11 0.15 0.38
Cassipourea sp. 19.8 38.7 0.41 1.1 3.9 0.24 0.13 0.49
Centroplacus glaucinus 13.6 26.5 0.75 4.9 5.0 1.08 0.24 2.14
Centroplacus sp. 13.2 19 0.34 2.4 2.1 0.53 0.11 1.10
Chrysobalanus sp. 10.0 10 0.01 0.1 0.1 0.02 0.00 0.14
Chrysophyllum sp. 36.2 39.4 0.21 0.2 2.8 0.04 0.07 0.34
Chytranthus sp. 12.4 21.9 0.11 0.8 0.8 0.18 0.03 0.79
Cleistanthus africana 17.3 28.9 0.09 0.3 1.1 0.07 0.03 0.21
Cleistanthus gabonii 19.7 64.8 1.78 4 28.3 0.88 0.57 2.27
Cleistanthus sp. 17.2 48.4 1.91 6.9 23.6 1.52 0.62 2.95
Coelocaryon klanei 46.4 46.4 0.17 0.1 1.8 0.02 0.05 0.19
Coelocaryon preussii 44.0 70.9 5.54 3.3 65.8 0.73 1.78 3.32
Coffea sp. 14.4 21 0.05 0.3 0.4 0.07 0.02 0.43
169
Species name Average dbh (cm)
Maximum tree size
(cm)
Basal area (m2
ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density
(%)
Relative dominance
(%)
Relative Importance
(%)
Cola flovovolquina 10.8 11.6 0.02 0.2 0.1 0.04 0.01 0.17
Cola lizae 16.8 28.8 0.32 1.3 2.3 0.29 0.10 0.85
Cola sp. 18.5 50.2 0.84 2.7 6.9 0.60 0.27 1.21
Copaifera mildbraedii 34.4 49.2 0.44 0.4 4.6 0.09 0.14 0.46
Corynanthe sp. 10.5 10.9 0.02 0.2 0.1 0.04 0.01 0.17
Coula edulis 39.2 120.4 20.42 12.9 413.0 2.85 6.57 10.58
Dacryodes buettneri 41.9 151.8 10.61 5.4 134.9 1.19 3.42 5.77
Dacryodes edulis 25.9 25.9 0.05 0.1 0.4 0.02 0.02 0.16
Dacryodes igaganga 20.0 80.9 7.65 18.2 68.6 4.02 2.46 7.65
Dacryodes klaineana 33.2 33.2 0.09 0.1 1.2 0.02 0.03 0.17
Dacryodes macrophylla 43.5 58.3 1.47 0.9 17.0 0.20 0.47 1.49
Dacryodes normandii 31.8 48.6 0.66 0.7 6.4 0.15 0.21 0.83
Dacryodes sp. 14.8 16.5 0.03 0.2 0.2 0.04 0.01 0.17
Daniellia soyauxii 67.8 67.8 0.36 0.1 4.0 0.02 0.12 0.25
Desbordesia glaucescens
45.0 119.7 7.96 3.3 187.8 0.73 2.56 4.45
Dialium bipendense 29.6 98.1 4.99 4.9 98.4 1.08 1.61 3.74
Dialium dinklagei 45.4 67.9 0.71 0.4 11.1 0.09 0.23 0.55
Dialium pachyphyllum 25.5 93.1 16.18 20.5 316.2 4.53 5.21 10.90
Dialium soyauxii 12.8 14.9 0.04 0.3 0.3 0.07 0.01 0.20
Dialium sp. 21.3 59.5 1.12 2.3 15.6 0.51 0.36 1.91
Dichaetanthera africana 42.4 42.4 0.14 0.1 2.3 0.02 0.05 0.18
Dichostemma glaucescens
13.5 24.7 1.99 13.2 9.2 2.92 0.64 4.72
Didelotia sp. 12.0 12 0.01 0.1 0.1 0.02 0.00 0.14
170
Species name Average dbh (cm)
Maximum tree size
(cm)
Basal area (m2
ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density
(%)
Relative dominance
(%)
Relative Importance
(%)
Diogoa zenkeri 25.1 89.2 6.75 10.2 87.7 2.25 2.17 5.59
Diospyros abyssinica 13.0 16 0.04 0.3 0.3 0.07 0.01 0.20
Diospyros crassifolia 12.5 14.5 0.03 0.2 0.2 0.04 0.01 0.28
Diospyros melocarpa 58.3 58.3 0.27 0.1 5.0 0.02 0.09 0.22
Diospyros punctata 10.5 10.5 0.01 0.1 0.1 0.02 0.00 0.14
Diospyros sp. 14.3 27.8 1.02 5.8 9.6 1.28 0.33 2.77
Diospyros viridicans 19.5 19.5 0.03 0.1 0.3 0.02 0.01 0.15
Diospyros zenkeri 12.4 14.3 0.09 0.7 0.7 0.15 0.03 0.65
Discoglypremna caloneura
36.8 36.8 0.11 0.1 0.6 0.02 0.03 0.17
Drypetes gossweileri 34.3 53.2 0.24 0.2 3.4 0.04 0.08 0.35
Drypetes sp. 25.7 49.6 1.59 2.5 17.7 0.55 0.51 2.11
Duboscia macrocarpa 14.3 14.3 0.02 0.1 0.1 0.02 0.01 0.14
Duguetia confinis 29.3 56.8 0.44 0.5 5.4 0.11 0.14 0.83
Duguetia staudtii 12.8 15 0.03 0.2 0.2 0.04 0.01 0.29
Duvigneaudia inopinata 27.6 36.9 1.01 1.6 7.3 0.35 0.33 1.38
Enantia chlorantha 19.4 33.7 0.60 1.8 3.5 0.40 0.19 1.52
Engomegoma gordonii 58.2 105.9 1.36 0.4 29.0 0.09 0.44 0.99
Entandrophragma candollei
35.3 35.3 0.10 0.1 1.0 0.02 0.03 0.17
Eriocoelum oblongum 12.2 12.2 0.01 0.1 0.1 0.02 0.00 0.14
Eriocoelum sp. 17.7 31.3 0.65 2.4 4.1 0.53 0.21 1.67
Erismadelphus exsul 38.9 107.1 6.11 3.3 90.3 0.73 1.97 3.86
Essewa sp. 12.7 12.7 0.01 0.1 0.1 0.02 0.00 0.14
Filaeopsis discophora 40.5 94.5 2.25 1.1 27.8 0.24 0.73 1.67
Garcinia epunctata 15.6 27.4 1.30 6.5 12.2 1.44 0.42 2.90
171
Species name Average dbh (cm)
Maximum tree size
(cm)
Basal area (m2
ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density
(%)
Relative dominance
(%)
Relative Importance
(%)
Garcinia kola 20.6 20.6 0.03 0.1 0.3 0.02 0.01 0.15
Garcinia lucida 11.9 11.9 0.01 0.1 0.1 0.02 0.00 0.14
Garcinia sp. 14.5 26.7 1.20 6.9 10.3 1.52 0.39 2.96
Gilbertiodendron dewevrei
17.7 76.1 3.25 10.4 35.7 2.30 1.05 4.04
Gilbertiodendron ogoouense
26.0 93.5 3.18 3.4 47.0 0.75 1.02 2.59
Gilbertiodendron sp. 17.6 49.8 1.58 5.3 15.7 1.17 0.51 2.73
Grewia coriacea 25.4 34 0.37 0.7 2.5 0.15 0.12 0.97
Grewia polyasles 20.1 20.1 0.03 0.1 0.2 0.02 0.01 0.15
Guarea cedrata 88.5 88.5 0.62 0.1 8.8 0.02 0.20 0.34
Guarea sp. 66.9 66.9 0.35 0.1 4.5 0.02 0.11 0.25
Guibourtia ehie 27.9 79.7 1.98 2.3 28.2 0.51 0.64 2.08
Guibourtia tessmannii 146.1 146.1 1.68 0.1 37.7 0.02 0.54 0.68
Heisteria parvifolia 29.7 76 1.15 1.3 18.8 0.29 0.37 1.35
Homalium sarcopetelum 21.1 21.1 0.03 0.1 0.3 0.02 0.01 0.15
Hymenostegia klainei 29.8 45 0.45 0.6 6.6 0.13 0.15 0.86
Hymenostegia ngouniensis
22.2 74.7 16.27 33.5 215.0 7.40 5.24 13.80
Irvingia excelsa 39.2 80 3.07 2.2 47.3 0.49 0.99 2.40
Irvingia gabonensis 23.3 53 1.40 2.6 18.7 0.57 0.45 2.07
Irvingia grandifolia 37.1 66.9 0.43 0.3 7.7 0.07 0.14 0.55
Irvingia robur 44.4 60.3 0.35 0.2 6.0 0.04 0.11 0.27
Irvingia sp. 43.4 43.4 0.15 0.1 2.3 0.02 0.05 0.19
Klaineanthus gabonii 14.7 27.5 1.35 7.4 6.8 1.63 0.43 3.12
Klainedoxa gabonensis 23.5 56.3 0.87 1.4 15.6 0.31 0.28 1.29
172
Species name Average dbh (cm)
Maximum tree size
(cm)
Basal area (m2
ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density
(%)
Relative dominance
(%)
Relative Importance
(%)
Klainedoxa microphylla 53.3 98.6 2.04 0.7 47.8 0.15 0.66 1.16
Klainedoxa trinesi 117.7 117.7 1.09 0.1 29.2 0.02 0.35 0.49
Macaranga monundra 16.8 16.8 0.02 0.1 0.1 0.02 0.01 0.15
Macaranga sp. 17.8 18.5 0.05 0.2 0.2 0.04 0.02 0.18
Macaranga staudtii 39.9 52.8 0.66 0.5 5.2 0.11 0.21 0.55
Maesobotrya sp. 13.5 20.3 0.14 0.9 0.6 0.20 0.04 0.48
Mammea africana 18.8 41.4 0.31 0.9 3.0 0.20 0.10 1.00
Manilkara fouilloyana 56.6 64.4 0.51 0.2 10.1 0.04 0.16 0.33
Manilkara sp. 25.9 45.2 0.26 0.4 4.1 0.09 0.08 0.64
Maprounea membranacea
19.5 23.6 0.06 0.2 0.4 0.04 0.02 0.18
Maranthes chrysophylla 23.9 23.9 0.04 0.1 0.5 0.02 0.01 0.15
Maranthes glabra 21.6 33.7 0.29 0.7 3.8 0.15 0.09 0.71
Mareya micrantha 16.3 20.7 0.07 0.3 0.3 0.07 0.02 0.32
Mareyopsis sp. 10.5 10.5 0.01 0.1 0.1 0.02 0.00 0.14
Marquesia excelsa 34.5 65.5 0.95 0.8 15.1 0.18 0.30 0.95
Melocarpodium sp. 13.7 14.6 0.03 0.2 0.1 0.04 0.01 0.17
Monodora myristica 11.7 11.7 0.01 0.1 0.0 0.02 0.00 0.14
Monopetalanthus dibata 35.5 35.5 0.10 0.1 0.9 0.02 0.03 0.17
Monopetalanthus durandii
33.9 198 11.93 5.9 167.9 1.30 3.84 6.19
Monopetalanthus letestu 57.7 194.7 13.77 2.6 217.4 0.57 4.43 5.59
Monopetalanthus microphyllus
42.0 72.9 0.43 0.2 5.2 0.04 0.14 0.41
Monopetalanthus pelligrini
16.4 28 0.08 0.3 0.5 0.07 0.03 0.44
173
Species name Average dbh (cm)
Maximum tree size
(cm)
Basal area (m2
ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density
(%)
Relative dominance
(%)
Relative Importance
(%)
Monopetalanthus sp. 20.1 33.7 0.30 0.8 2.3 0.18 0.10 0.85
Mytragyna ciliata 33.8 113.1 2.09 1.4 26.6 0.31 0.67 1.21
Nauclea didderichii 37.3 49.7 0.24 0.2 3.3 0.04 0.08 0.35
Newtonia duparquetiana 12.0 12 0.01 0.1 0.1 0.02 0.00 0.14
Newtonia leucocarpa 90.9 90.9 0.65 0.1 10.4 0.02 0.21 0.35
Newtonia sp. 48.9 73.3 0.47 0.2 6.4 0.04 0.15 0.43
Ochna sp. 16.6 16.6 0.02 0.1 0.2 0.02 0.01 0.15
Ochthocosmus africanus 43.7 43.7 0.15 0.1 2.3 0.02 0.05 0.19
Odyendyea gabonensis 31.1 64.3 0.90 0.9 5.8 0.20 0.29 1.19
Ongokea gore 57.9 90.2 1.59 0.5 29.9 0.11 0.51 1.09
Pancovia sp. 16.6 22.3 0.23 1 1.9 0.22 0.07 0.53
Parkia bicolor 63.2 78.6 0.97 0.3 10.3 0.07 0.31 0.73
Pausinystalia johimbe 29.0 53 0.66 0.8 7.4 0.18 0.21 0.97
Pausinystalia macroceras
24.5 49.1 0.65 1.1 6.7 0.24 0.21 1.27
Pausinystalia sp. 10.4 10.4 0.01 0.1 0.0 0.02 0.00 0.14
Pentaclethra macrophylla 49.3 60.2 0.40 0.2 5.5 0.04 0.13 0.29
Petersianthus macrocarpus
51.2 81.7 0.56 0.2 9.6 0.04 0.18 0.34
Phyllocosmus africana 17.3 17.3 0.02 0.1 0.2 0.02 0.01 0.15
Piptadeniastrum africanum
96.3 104.9 2.20 0.3 36.8 0.07 0.71 1.12
Plagiostyles africana 33.7 60.8 5.48 5.6 76.0 1.24 1.76 4.16
Poga oleosa 18.8 26 0.13 0.4 0.7 0.09 0.04 0.48
Polyalthia suaveolens 14.7 32.8 1.72 9.3 13.7 2.05 0.55 3.77
Psychotria sp. 34.5 70.8 0.52 0.4 6.9 0.09 0.17 0.49
174
Species name Average dbh (cm)
Maximum tree size
(cm)
Basal area (m2
ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density
(%)
Relative dominance
(%)
Relative Importance
(%)
Psychotria vogellei 37.9 80.9 0.56 0.3 8.3 0.07 0.18 0.59
Pteleopsis discofera 41.6 59.6 0.32 0.2 4.0 0.04 0.10 0.26
Pterocarpus soyauxii 33.5 38.8 0.18 0.2 2.1 0.04 0.06 0.33
Pycnanthus angolensis 38.1 51.5 0.61 0.5 4.9 0.11 0.20 0.77
Rabdophyllum sp. 27.5 66.3 1.59 2 21.6 0.44 0.51 1.89
Rauvolfia vomitoria 15.4 20.1 0.27 1.4 1.4 0.31 0.09 0.63
Rothmannia sp. 17.8 17.8 0.02 0.1 0.2 0.02 0.01 0.15
Santiria trimera 21.1 58.5 11.51 28.3 94.2 6.25 3.70 11.12
Scaphopetalum sp. 13.0 15.1 0.04 0.3 0.2 0.07 0.01 0.31
Schaumaniophyton magnificum
11.8 11.8 0.01 0.1 0.1 0.02 0.00 0.14
Scorodophloeus zenkeri 27.7 69.7 9.89 13.5 127.1 2.98 3.18 7.33
Scyphocephalium ochocoa
53.4 91 4.74 1.9 52.5 0.42 1.52 2.87
Scytopetalum klaineanum
31.2 69 3.32 3.4 41.3 0.75 1.07 2.98
Scytopetalum sp. 45.9 45.9 0.17 0.1 2.1 0.02 0.05 0.19
Sindoropsis letestui 34.2 124.1 5.24 3.6 81.0 0.80 1.69 3.65
Sorindeia sp. 13.4 17.5 0.15 1.1 0.7 0.24 0.05 0.99
Stachyothyrsus staudtii 28.7 53.5 0.73 0.9 8.1 0.20 0.23 1.13
Staudtia gabonensis 37.9 91.1 2.24 1.5 37.9 0.33 0.72 2.10
Staudtia kamerounensis 52.2 52.2 0.21 0.1 3.7 0.02 0.07 0.21
Sterculia tragacantha 35.1 52.3 0.33 0.3 3.3 0.07 0.11 0.29
Strephonema sp. 14.4 25.4 0.11 0.6 0.7 0.13 0.04 0.40
Strombosia grandifolia 18.4 47.3 2.39 7.6 28.5 1.68 0.77 3.61
Strombosia pustulata 18.9 44.5 2.81 8.4 34.3 1.86 0.91 3.92
175
Species name Average dbh (cm)
Maximum tree size
(cm)
Basal area (m2
ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density
(%)
Relative dominance
(%)
Relative Importance
(%)
Strombosiopsis tetrandra 18.2 38.4 5.07 17.3 45.2 3.82 1.63 6.62
Strychnos sp. 66.4 66.4 0.35 0.1 5.9 0.02 0.11 0.25
Symphonia globulifera 39.6 48.3 0.89 0.7 10.0 0.15 0.29 1.02
Synsephalium sp. 19.6 43.8 0.36 0.7 4.5 0.15 0.12 0.85
Syzygium sp. 19.3 26.3 0.12 0.7 1.0 0.15 0.04 0.43
Syzygium staudtii 24.9 85 1.13 1.6 14.5 0.35 0.36 1.07
Terminalia sp. 28.9 28.9 0.07 0.1 0.4 0.02 0.02 0.16
Tessmannia africana 18.9 23 0.06 0.2 0.6 0.04 0.02 0.30
Tessmannia anomala 33.0 45.9 0.38 0.4 5.8 0.09 0.12 0.68
Tessmannia lescrauwaetii
14.3 14.3 0.02 0.1 0.1 0.02 0.01 0.14
Tessmannia sp. 24.1 40.5 0.17 0.3 2.3 0.07 0.05 0.47
Tetraberlinia bifoliolata 20.7 79.1 2.81 5.6 28.6 1.24 0.91 3.19
Tieghemella africana 108.9 108.9 0.93 0.1 17.1 0.02 0.30 0.44
Treculia africana 17.1 17.1 0.02 0.1 0.1 0.02 0.01 0.15
Trichilia sp. 23.0 34.7 0.19 0.4 1.5 0.09 0.06 0.50
Trichoscypha abut 18.2 31.1 0.19 0.6 1.8 0.13 0.06 0.66
Trichoscypha accuminata 14.7 29.2 0.86 4.7 6.3 1.04 0.28 2.36
Trichoscypha engong 62.0 133.9 1.80 0.4 31.7 0.09 0.58 1.13
Trichoscypha sp. 17.5 56.1 0.72 2.3 7.2 0.51 0.23 1.55
Uapaca paludosa 29.0 46.4 0.30 0.4 3.4 0.09 0.10 0.53
Uapaca sp. 29.6 80 3.04 3.5 37.6 0.77 0.98 2.57
Unknown 18.6 73.1 3.63 10.1 34.9 2.23 1.17 4.56
Vitex doniana 44.4 44.4 0.15 0.1 2.3 0.02 0.05 0.19
Warneckea sp. 14.9 32.8 1.08 5.4 10.8 1.19 0.35 2.70
Xylopia aethiopica 34.3 57.4 1.60 1.6 13.5 0.35 0.52 1.57
176
Species name Average dbh (cm)
Maximum tree size
(cm)
Basal area (m2
ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density
(%)
Relative dominance
(%)
Relative Importance
(%)
Xylopia pynaertii 29.1 53.4 0.49 0.5 4.6 0.11 0.16 0.62
Xylopia quintasii 20.2 35.5 0.61 1.6 7.0 0.35 0.20 1.25
Xylopia rubasens 43.0 46.8 0.44 0.3 5.2 0.07 0.14 0.32
Xylopia sp. 20.2 48.3 1.22 3 11.6 0.66 0.39 2.10
Xylopia staudtii 25.8 47.5 1.52 2.5 10.1 0.55 0.49 1.27
177
APPENDIX B CHARACTERISTICS OF TREE FAMILIES IN THE 10 1-HA PERMANENT PLOTS IN THE RIL ZONE OF THE SEEF
CONCESSION ON THE MONTS DE CRISTAL, GABON
Name of family Average dbh (cm)
Maximum tree size
(cm)
Basal area (m2 ha-1)
Density (stem ha-1)
AGB (Mg)
Relative density
(%)
Relative dominance
(%)
Relative Importance
(%)
Anacardiaceae 17.4 133.9 3.75 9.3 47.7 2.05 1.21 3.26
Anisophylleaceae 20.7 74.1 10.09 20.8 116.0 4.59 3.25 7.84
Annonaceae 20.0 57.4 8.72 21.7 74.9 4.79 2.81 7.60
Apocynaceae 15.4 20.1 0.27 1.4 1.4 0.31 0.09 0.39
Burseraceae 25.3 151.8 42.74 56.3 428.6 12.44 13.76 26.19
Caesalpiniaceae 24.6 198 100.94 128.9 1506.5 28.47 32.48 60.96
Chrysobalanaceae 18.9 33.7 0.54 1.7 6.3 0.38 0.17 0.55
Combretaceae 22.0 59.6 0.50 0.9 5.1 0.20 0.16 0.36
Dipterocarpaceae 34.5 65.5 0.95 0.8 15.1 0.18 0.30 0.48
Ebenaceae 14.6 58.3 1.48 7.3 16.1 1.61 0.47 2.09
Euphorbiaceae 20.2 80 15.09 36 154.0 7.95 4.86 12.81
Guttifae 16.6 51.6 3.99 15.3 39.2 3.38 1.29 4.67
Huaceae 20.8 40.8 0.21 0.5 2.0 0.11 0.07 0.18
Irvingiaceae 36.8 119.7 17.36 10.9 362.2 2.41 5.59 7.99
Ixonanthaceae 43.7 43.7 0.15 0.1 2.3 0.02 0.05 0.07
Lauraceae 19.0 64.2 1.91 4.5 17.0 0.99 0.61 1.61
Lecythidaceae 51.2 81.7 0.56 0.2 9.6 0.04 0.18 0.22
Linaceae 17.3 17.3 0.02 0.1 0.2 0.02 0.01 0.03
Loganiaceae 66.4 66.4 0.35 0.1 5.9 0.02 0.11 0.13
Malvaceae 19.6 52.3 1.92 5.4 15.3 1.19 0.62 1.81
Melastomataceae 15.4 42.4 1.22 5.5 13.0 1.21 0.39 1.61
Meliaceae 33.3 88.5 2.59 2.3 30.5 0.51 0.83 1.34
178
Name of family Average dbh (cm)
Maximum tree size
(cm)
Basal area (m2 ha-1)
Density (stem ha-1)
AGB (Mg)
Relative density
(%)
Relative dominance
(%)
Relative Importance
(%)
Mimosaceae 44.3 104.9 7.37 3.2 102.1 0.71 2.37 3.08
Moraceae 17.1 17.1 0.02 0.1 0.1 0.02 0.01 0.03
Myristicaceae 45.0 91.1 13.51 7.4 166.5 1.63 4.35 5.98
Myrtaceae 23.8 85 1.26 2 15.5 0.44 0.40 0.85
Ochnaceae 16.6 16.6 0.02 0.1 0.2 0.02 0.01 0.03
Olacaceae 25.0 120.4 41.53 58.6 686.4 12.94 13.37 26.31
Papilionaceae 17.3 59.6 0.84 2.5 9.4 0.55 0.27 0.82
Passifloraceae 14.5 22.6 0.11 0.6 0.8 0.13 0.03 0.17
Phyllanthaceae 18.1 64.8 3.77 11.2 52.9 2.47 1.21 3.69
Putranjavaceae 26.3 53.2 1.83 2.7 21.1 0.60 0.59 1.19
Rhizophoraceae 19.6 38.7 0.53 1.5 4.6 0.33 0.17 0.50
Rubiaceae 28.5 113.1 5.25 5.3 65.9 1.17 1.69 2.86
Rutaceae 11.9 16.2 0.12 1.1 0.7 0.24 0.04 0.28
Salicaceae 21.1 21.1 0.03 0.1 0.3 0.02 0.01 0.03
Sapindaceae 19.9 66.3 2.59 6.3 28.4 1.39 0.83 2.22
Sapotaceae 30.4 108.9 2.29 2 38.7 0.44 0.74 1.18
Scytopetalaceae 31.6 69 3.49 3.5 43.4 0.77 1.12 1.90
Simabouraceae 20.7 22 0.07 0.2 0.3 0.04 0.02 0.07
Simaroubaceae 34.1 64.3 0.83 0.7 5.5 0.15 0.27 0.42
Unknown 18.6 73.1 3.63 10.2 34.9 2.25 1.17 3.42
Verbenaceae 44.4 44.4 0.15 0.1 2.3 0.02 0.05 0.07
Vochysiaceae 38.9 107.1 6.11 3.3 90.3 0.73 1.97 2.69
179
APPENDIX C CHARACTERISTICS OF TREE SPECIES IN THE 20 1-HA PERMANENT PLOTS IN THE CEB-FSC CONCESSION,
GABON
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Acioa sp. 13.0 15.8 0.07 0.25 0.5 0.08 0.01 0.30
Afrosersalisa afzelii 75.0 85 0.90 0.1 14.9 0.03 0.17 0.41
Afrostyrax lepidophyllus
15.0 29.5 0.31 0.8 2.1 0.26 0.06 1.01
Afzelia bella 12.2 16.4 0.05 0.2 0.3 0.07 0.01 0.28
Afzelia bipendensis 39.2 56.5 0.29 0.1 4.5 0.03 0.05 0.22
Albizia ferruginea 57.4 68.1 0.53 0.1 6.0 0.03 0.10 0.27
Allanblackia cf. gabonensis
14.9 14.9 0.02 0.05 0.1 0.02 0.00 0.09
Alstonia boonei 54.0 80 2.63 0.5 19.5 0.16 0.49 1.00
Amanoa strobilacea 24.6 34.7 0.16 0.15 1.1 0.05 0.03 0.15
Amphinas ferrugineus 43.9 56.1 0.65 0.2 9.1 0.07 0.12 0.46
Angylocalyx sp. 15.3 24 0.32 0.8 2.4 0.26 0.06 0.94
Anonidium mannii 48.4 70 1.35 0.35 8.2 0.12 0.25 0.71
Anopyxis klaineana 20.3 24.6 0.07 0.1 0.7 0.03 0.01 0.18
Anthocleista vogelii 13.5 15.5 0.09 0.3 0.4 0.10 0.02 0.25
Anthonotha fragrans 33.9 59.9 1.52 0.7 15.8 0.23 0.28 1.27
Anthonotha sp. 29.9 81.2 2.19 1 35.3 0.33 0.41 1.56
Anthonotha wijmacampensi
25.7 68.5 4.33 3.5 54.1 1.15 0.81 2.44
Antidesma cf vogelianum
13.9 13.9 0.02 0.05 0.1 0.02 0.00 0.09
Antrocaryon klaineanum
55.0 55 0.24 0.05 2.8 0.02 0.04 0.13
180
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Aphanocalyx margininervatus
18.9 35.3 0.84 1.3 5.4 0.43 0.16 1.13
Aphanocalyx microphyllus
19.5 59.8 0.55 0.6 4.7 0.20 0.10 0.64
Aucoumea klaineana 108.1 180 11.72 0.55 126.5 0.18 2.20 2.93
Baikiaea insignis 32.9 66.4 0.62 0.3 8.7 0.10 0.12 0.35
Baillonella toxisperma 120.2 160 5.15 0.2 109.7 0.07 0.96 1.24
Baphia cf. buettneri 18.1 38 1.82 3.1 17.4 1.02 0.34 1.77
Baphia cf. pubescens 34.7 34.7 0.09 0.05 1.2 0.02 0.02 0.10
Baphia pubescens 22.0 22 0.04 0.05 0.4 0.02 0.01 0.09
Baphia sp. 15.7 48.6 0.87 1.8 8.5 0.59 0.16 1.03
Barteria fustulosa 15.3 24.4 0.18 0.45 1.4 0.15 0.03 0.66
Beilschmiedia calicitranthera
10.3 10.3 0.01 0.05 0.0 0.02 0.00 0.09
Beilschmiedia congolana
35.4 47.7 0.22 0.1 2.6 0.03 0.04 0.14
Beilschmiedia fulva 15.7 15.7 0.02 0.05 0.1 0.02 0.00 0.09
Beilschmiedia pierreanum
10.9 10.9 0.01 0.05 0.0 0.02 0.00 0.09
Beilschmiedia sp. 20.3 75 1.05 1.05 12.3 0.35 0.20 1.36
Beilschmiedia sp1 13.7 16.5 0.06 0.2 0.4 0.07 0.01 0.28
Berlinia auriculata 11.9 11.9 0.01 0.05 0.1 0.02 0.00 0.09
Berlinia bracteosa 55.6 80 4.38 0.8 62.3 0.26 0.82 1.29
Berlinia confusa 26.0 47.2 0.43 0.3 4.8 0.10 0.08 0.52
Berlinia congolensis 13.3 15.7 0.10 0.35 0.6 0.12 0.02 0.34
Berlinia sp. 20.2 78.3 1.02 1 12.5 0.33 0.19 1.00
Bertiera sp. 11.4 11.6 0.02 0.1 0.1 0.03 0.00 0.17
181
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Bikinia dibata 23.5 23.5 0.04 0.05 0.3 0.02 0.01 0.09
Bikinia evrardii 20.9 32.8 0.30 0.4 2.2 0.13 0.06 0.46
Bikinia grisea 35.1 124.4 6.59 2 84.1 0.66 1.24 2.65
Bikinia le-testui 28.3 100 1.38 0.65 17.3 0.21 0.26 0.81
Bikinia pellegrinii 51.9 71.2 0.48 0.1 5.5 0.03 0.09 0.19
Bikinia sp. 11.5 11.5 0.01 0.05 0.0 0.02 0.00 0.09
Blighia welwitschii 25.0 29.5 0.10 0.1 1.2 0.03 0.02 0.19
Bombax buonopozense
36.6 36.6 0.11 0.05 0.6 0.02 0.02 0.10
Brenania breyi 42.6 42.6 0.14 0.05 1.7 0.02 0.03 0.11
Calpocalyx dinklagei 15.0 33.9 1.82 4.75 15.2 1.56 0.34 3.14
Canarium schweinfurthii
67.7 84.8 2.31 0.3 23.4 0.10 0.43 0.87
Canthium sp. 15.2 15.2 0.02 0.05 0.1 0.02 0.00 0.09
Carapa cf. parvifolia 26.4 44.3 1.06 0.9 10.1 0.30 0.20 0.90
Carapa procera 28.5 37.9 0.59 0.45 5.8 0.15 0.11 0.60
Celtis tessmannii 53.2 78.6 3.95 0.85 58.7 0.28 0.74 1.43
Centroplacus glaucinus 14.5 28.8 1.29 3.65 7.9 1.20 0.24 2.74
Chrysophyllum africanum
47.8 75 0.68 0.15 9.8 0.05 0.13 0.25
Chrysophyllum sp. 47.5 47.5 0.18 0.05 2.2 0.02 0.03 0.12
Chytranthus sp. 12.6 16.1 0.14 0.55 0.9 0.18 0.03 0.69
Cleistanthus cf. racemosus
19.7 19.7 0.03 0.05 0.1 0.02 0.01 0.09
Cleistanthus gabonensis
20.0 40 0.27 0.35 1.6 0.12 0.05 0.44
182
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Cleistanthus sp. 18.5 62.9 2.10 2.95 12.6 0.97 0.39 2.25
Cleistanthus sp1. 13.8 15.1 0.05 0.15 0.2 0.05 0.01 0.13
Coelocaryon klainei 59.9 69.5 1.44 0.25 17.7 0.08 0.27 0.63
Coelocaryon preussii 46.2 54 0.34 0.1 3.5 0.03 0.06 0.23
Coffea sp. 14.0 16.9 0.05 0.15 0.3 0.05 0.01 0.20
Cola altissima 48.5 54 0.56 0.15 6.8 0.05 0.10 0.22
Cola cf. altissima 58.9 65 0.83 0.15 11.0 0.05 0.16 0.34
Cola cf. griseiflora 11.2 11.2 0.01 0.05 0.0 0.02 0.00 0.09
Cola flavo-velutina 18.9 28.8 0.16 0.25 1.3 0.08 0.03 0.32
Cola lizea 26.4 58.7 0.99 0.75 10.1 0.25 0.19 0.64
Cola sp. 19.8 60.8 2.25 2.6 22.8 0.85 0.42 2.30
Cola sp1 26.7 55.2 2.12 1.55 22.0 0.51 0.40 1.25
Cola sp2 17.8 26.2 0.08 0.15 0.6 0.05 0.02 0.20
Cola sp3 13.1 13.1 0.01 0.05 0.1 0.02 0.00 0.09
Copaifera mildbraedii 16.5 16.5 0.02 0.05 0.2 0.02 0.00 0.09
Coula edulis 28.2 98.8 9.53 5.35 176.6 1.76 1.79 4.30
Crudia cf ledermannii 32.5 55 0.31 0.15 4.0 0.05 0.06 0.24
Cylicodiscus gabunensis
124.4 143 5.05 0.2 117.4 0.07 0.95 1.22
Dacryodes buettneri 39.3 130.5 3.55 1 43.1 0.33 0.66 1.54
Dacryodes cf. buettneri 46.9 185.5 3.74 0.55 52.0 0.18 0.70 1.29
Dacryodes igaganga 26.2 50.8 2.70 2.15 24.7 0.71 0.51 1.97
Dacryodes klaineana 28.0 44.9 0.28 0.2 3.6 0.07 0.05 0.32
Dacryodes macrophylla 15.2 15.2 0.02 0.05 0.1 0.02 0.00 0.09
Dacryodes normandii 28.3 86.6 2.18 1.2 23.7 0.39 0.41 1.69
183
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Dacryodes sp. 27.2 54.8 0.70 0.5 6.8 0.16 0.13 0.77
Daniellia soyauxii 50.3 73.4 0.88 0.2 9.1 0.07 0.16 0.44
Desbordesia glaucescens
63.2 63.2 0.31 0.05 6.9 0.02 0.06 0.14
Detarium macrocarpum 34.7 97.2 0.96 0.3 16.8 0.10 0.18 0.55
Dialium angolense 25.0 83.3 19.06 14.25 298.9 4.69 3.57 9.63
Dialium bipindense 24.7 98.5 4.83 3.45 93.6 1.13 0.90 3.34
Dialium dinklagei 28.4 62.3 0.66 0.35 10.1 0.12 0.12 0.58
Dialium lopense 21.2 25.7 0.15 0.2 1.6 0.07 0.03 0.30
Dialium sp. 18.0 75 2.12 3.05 29.5 1.00 0.40 2.43
Dialium tessmannii 18.5 48.7 1.15 1.8 13.8 0.59 0.22 1.77
Didelotia africana 30.5 44.2 0.26 0.15 3.1 0.05 0.05 0.30
Didelotia sp. 12.9 12.9 0.01 0.05 0.1 0.02 0.00 0.09
Diogoa zenkeri 26.8 70.2 2.32 1.45 32.1 0.48 0.43 1.32
Diospyros canaliculata 15.5 15.5 0.02 0.05 0.2 0.02 0.00 0.09
Diospyros cf. epunctata 16.1 22.7 0.20 0.45 1.9 0.15 0.04 0.53
Diospyros cf. piscatoria 23.3 23.3 0.04 0.05 0.5 0.02 0.01 0.09
Diospyros dendo 13.0 19.8 0.28 1 2.6 0.33 0.05 1.07
Diospyros epunctata 16.5 23.6 0.09 0.2 0.9 0.07 0.02 0.29
Diospyros flavolitina 15.4 15.4 0.02 0.05 0.2 0.02 0.00 0.09
Diospyros fragrans 15.5 33.4 0.86 2.05 8.5 0.67 0.16 1.38
Diospyros gabunensis 13.0 20.9 0.48 1.75 4.0 0.58 0.09 1.49
Diospyros hoyleana 12.7 18.6 0.67 2.6 5.4 0.85 0.13 1.80
Diospyros iturensis 11.4 11.4 0.01 0.05 0.1 0.02 0.00 0.09
Diospyros mannii 16.4 26.9 0.12 0.25 1.2 0.08 0.02 0.38
Diospyros melocarpa 15.1 33.6 1.48 3.75 14.5 1.23 0.28 2.54
184
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Diospyros obliquifolia 13.1 13.3 0.03 0.1 0.2 0.03 0.01 0.17
Diospyros sanza-minika
18.4 39.5 0.48 0.75 5.0 0.25 0.09 0.88
Diospyros sp. 16.4 43.9 1.26 2.45 14.5 0.81 0.24 2.21
Diospyros zenkeri 17.6 17.7 0.05 0.1 0.5 0.03 0.01 0.18
Discoglypremna caloneura
16.9 22.5 0.12 0.25 0.5 0.08 0.02 0.38
Distemonanthus benthamianus
61.8 89.8 2.29 0.35 34.1 0.12 0.43 0.89
Drypetes gossweileri 29.6 59.1 1.09 0.65 13.9 0.21 0.20 1.10
Drypetes sp. 24.5 54.2 1.52 1.2 19.7 0.39 0.29 1.64
Drypetes sp1 12.0 13.9 0.02 0.1 0.2 0.03 0.00 0.17
Drypetes sp2 18.2 18.2 0.03 0.05 0.2 0.02 0.00 0.09
Drypetes sp3 14.0 14 0.02 0.05 0.1 0.02 0.00 0.09
Duboscia macrocarpa 37.2 60 0.30 0.1 3.4 0.03 0.06 0.23
Duguetia confinis 37.0 63.8 0.41 0.15 5.6 0.05 0.08 0.33
Duguetia staudtii 17.2 17.2 0.02 0.05 0.2 0.02 0.00 0.09
Duvigneaudia inopinata 38.0 38 0.11 0.05 0.9 0.02 0.02 0.11
Endodesmia cf. calophylloides
12.1 12.1 0.01 0.05 0.1 0.02 0.00 0.09
Entandophragma cylindricum
20.1 20.1 0.03 0.05 0.2 0.02 0.01 0.09
Entandrophragma angolense
50.0 50 0.20 0.05 2.0 0.02 0.04 0.12
Eriocoelum cf. macrocarpum
15.9 24 0.17 0.4 1.1 0.13 0.03 0.51
185
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Eriocoelum macrocarpum
27.0 27 0.06 0.05 0.4 0.02 0.01 0.10
Eriocoelum oblongum 33.6 33.6 0.09 0.05 0.8 0.02 0.02 0.10
Eriocoelum sp. 23.2 49.1 1.25 1.2 10.8 0.39 0.23 1.31
Erypetalum tessmannii 24.5 88.4 45.23 35.45 539.3 11.66 8.47 21.22
Erythrophleum ivorense
65.5 98.5 1.81 0.25 34.7 0.08 0.34 0.69
Fillaeopsis discophora 43.3 196 10.50 1.95 138.1 0.64 1.97 3.43
Garcinia cf. epunctata 15.7 25.4 0.22 0.55 2.2 0.18 0.04 0.77
Garcinia courauana 14.6 19.6 0.09 0.25 0.8 0.08 0.02 0.44
Garcinia epunctata 13.5 24.3 0.11 0.35 1.1 0.12 0.02 0.41
Garcinia kola 29.6 29.6 0.07 0.05 0.8 0.02 0.01 0.10
Garcinia ovalifolia 13.3 18.1 0.04 0.15 0.4 0.05 0.01 0.26
Garcinia smeathmannii 10.8 12.9 0.11 0.6 0.7 0.20 0.02 0.70
Garcinia sp. 12.3 15.2 0.33 1.35 2.4 0.44 0.06 1.26
Gilbertiodendron dewevrei
23.8 75.9 3.98 3.05 55.6 1.00 0.75 2.09
Gilbertiodendron ogoouense
18.9 38.5 2.15 3.3 20.6 1.08 0.40 2.10
Gilbertiodendron sp. 25.1 60.8 0.46 0.3 6.4 0.10 0.09 0.46
Gilletiodendron insignis 26.6 34 0.23 0.2 3.3 0.07 0.04 0.18
Gilletiodendron pierreanum
36.6 105 22.64 8 467.2 2.63 4.24 8.24
Gossweilrodendron balsamiferum
32.1 32.1 0.08 0.05 0.5 0.02 0.02 0.10
Grewia coriacea 20.3 38.9 1.11 1.5 7.0 0.49 0.21 1.66
Guarea cedrata 20.3 28.3 0.07 0.1 0.6 0.03 0.01 0.18
186
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Guarea thompsonii 30.9 30.9 0.07 0.05 0.7 0.02 0.01 0.10
Heisteria parvifolia 40.0 72.6 3.74 1.4 61.4 0.46 0.70 1.91
Homalium letestu 47.6 64 0.40 0.1 6.3 0.03 0.07 0.24
Hylodendron gabunensis
44.2 55 1.18 0.35 19.7 0.12 0.22 0.68
Hymenostegia klainei 23.6 30.2 0.09 0.1 1.2 0.03 0.02 0.19
Hymenostegia pellegrinii
32.8 92.7 28.12 12.3 507.7 4.04 5.27 10.61
Hypodaphnis zenkeri 29.8 29.8 0.07 0.05 0.7 0.02 0.01 0.10
Unknown 14.3 37.2 0.79 2 6.3 0.66 0.15 1.90
Irvingia excelsa 45.8 45.8 0.16 0.05 2.6 0.02 0.03 0.12
Irvingia gabunensis 32.7 110 3.62 1.45 63.7 0.48 0.68 2.25
Irvingia grandifolia 54.9 70 0.51 0.1 9.5 0.03 0.10 0.27
Irvingia sp. 20.4 20.4 0.03 0.05 0.3 0.02 0.01 0.09
Isomacrolobium hallei 11.8 11.8 0.01 0.05 0.1 0.02 0.00 0.09
Isomacrolobium sp. 12.7 13.6 0.03 0.1 0.2 0.03 0.00 0.11
Julbernardia cf. seretii 45.4 50.8 0.33 0.1 4.6 0.03 0.06 0.16
Julbernardia pellegriniana
79.1 210 47.97 3.85 889.3 1.27 8.99 11.42
Klaineanthus gabonii 20.6 52.7 2.76 3.65 18.0 1.20 0.52 3.02
Klainedoxa gabonensis 44.3 115 1.94 0.45 44.8 0.15 0.36 1.06
Klainedoxa trillesii 40.9 63.5 0.46 0.15 9.3 0.05 0.09 0.34
Lasianthera africana 10.0 10 0.01 0.05 0.0 0.02 0.00 0.09
Letestua durissima 60.3 60.3 0.29 0.05 6.6 0.02 0.05 0.14
Lindackeria dentata 16.8 18.1 0.04 0.1 0.3 0.03 0.01 0.18
Macaranga barteri 56.0 56 0.25 0.05 2.1 0.02 0.05 0.13
187
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Macaranga monandra 28.4 48.3 0.45 0.3 3.1 0.10 0.08 0.46
Mammea africana 18.7 27.6 0.09 0.15 0.8 0.05 0.02 0.27
Manilkara fouilloyana 25.3 25.3 0.05 0.05 0.6 0.02 0.01 0.09
Manilkara sp. 41.9 41.9 0.14 0.05 2.4 0.02 0.03 0.11
Maprounea membranacea
28.4 28.4 0.06 0.05 0.4 0.02 0.01 0.10
Maranthes gabonensis 61.2 93 1.33 0.2 28.2 0.07 0.25 0.59
Maranthes glabra 37.0 81.9 0.56 0.15 12.4 0.05 0.11 0.36
Mareya cf. micrantha 11.1 12.7 0.04 0.2 0.2 0.07 0.01 0.21
Mareya micrantha 14.4 16.9 0.18 0.55 0.8 0.18 0.03 0.49
Mareyopsis longifolia 10.9 12.6 0.14 0.75 0.5 0.25 0.03 0.75
Marquesia excelsa 38.3 105 12.06 3.8 210.9 1.25 2.26 4.74
Macaranga monandra 28.4 48.3 0.45 0.3 3.1 0.10 0.08 0.46
Massularia acuminata 11.6 12.8 0.08 0.4 0.5 0.13 0.02 0.56
Memecylon sp. 11.6 11.6 0.01 0.05 0.1 0.02 0.00 0.09
Millettia grifoniana 15.7 19.8 0.10 0.25 0.9 0.08 0.02 0.17
Mitragyna ciliata 25.2 62.7 0.55 0.4 5.6 0.13 0.10 0.37
Monodora sp. 16.4 21.6 0.07 0.15 0.4 0.05 0.01 0.13
Myrianthus arboreus 52.7 52.7 0.22 0.05 2.0 0.02 0.04 0.13
Myrianthus serratus 26.9 38.9 0.41 0.35 2.8 0.12 0.08 0.40
Nauclea cf. latifolia 27.9 47.5 0.35 0.25 4.1 0.08 0.07 0.28
Nauclea diderrichii 43.6 81.6 1.14 0.3 17.9 0.10 0.21 0.72
Neochevalierodendron stephanii
23.6 64.9 25.52 23.65 276.1 7.78 4.78 13.65
Nesogordonia papaverifera
45.6 75 1.63 0.45 23.0 0.15 0.30 0.73
188
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Newtonia duparquetiana
29.5 29.5 0.07 0.05 0.6 0.02 0.01 0.10
Newtonia glandulifera 39.0 39.5 0.24 0.1 3.2 0.03 0.04 0.15
Newtonia sp. 22.2 33.7 0.10 0.1 0.9 0.03 0.02 0.19
Ochna sp. 11.0 11 0.01 0.05 0.1 0.02 0.00 0.09
Ochthocosmus 12.5 12.5 0.01 0.05 0.1 0.02 0.00 0.09
Oddoniodendron micranthum
20.2 82.1 6.90 8.3 106.2 2.73 1.29 4.84
Odyendyea gabonensis
46.5 80 1.07 0.25 8.1 0.08 0.20 0.49
Omphalocarpum elatum
75.0 75 0.44 0.05 6.1 0.02 0.08 0.17
Ongokea gore 80.4 80.4 0.51 0.05 9.9 0.02 0.10 0.18
Oubanguia africana 22.7 41.1 0.32 0.3 4.0 0.10 0.06 0.36
Oxystigma oxyphyllum 45.8 59.9 0.56 0.15 5.4 0.05 0.11 0.22
Pancovia cf floribunda 12.9 21.6 0.42 1.5 3.0 0.49 0.08 1.46
Panda oleosa 42.6 69.4 5.82 1.9 67.1 0.62 1.09 2.33
Parinari sp. 53.9 53.9 0.23 0.05 3.6 0.02 0.04 0.13
Pausinystalia johimbe 17.6 43.2 1.58 2.8 12.9 0.92 0.30 2.18
Pausinystalia macroceras
18.0 46.5 1.79 2.95 15.1 0.97 0.34 2.33
Pentaclethra eetveldeana
44.1 64.8 4.06 1.2 57.1 0.39 0.76 1.91
Pentaclethra macrophylla
55.1 118.3 4.09 0.7 85.8 0.23 0.77 1.61
Petersianthus macrocarpus
47.0 56.9 0.71 0.2 10.1 0.07 0.13 0.47
189
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Phyllocosmus africanus
23.4 46.9 0.31 0.25 4.5 0.08 0.06 0.41
Picralima nitida 20.8 22.1 0.07 0.1 0.7 0.03 0.01 0.18
Pinacopodium congolense
38.6 56.3 0.75 0.3 12.7 0.10 0.14 0.45
Piptadeniastrum africanum
90.9 150 6.06 0.4 103.5 0.13 1.13 1.68
Piptostigma sp. 12.0 15.4 0.05 0.2 0.2 0.07 0.01 0.21
Placodiscus sp. 18.6 18.6 0.03 0.05 0.3 0.02 0.01 0.09
Plagiosiphon gabonensis
26.9 59.9 0.49 0.3 6.3 0.10 0.09 0.33
Plagiosiphon sp. 26.3 49.4 1.04 0.75 12.2 0.25 0.19 0.65
Plagiostyles africana 34.4 61.3 7.14 3.65 96.9 1.20 1.34 3.84
Polyalthia gabunensis 16.4 38.9 0.68 1.35 6.3 0.44 0.13 1.26
Polyalthia suaveolens 24.3 64.2 3.33 3 38.9 0.99 0.62 2.64
Porterandia cladanta 12.7 14.8 0.05 0.2 0.3 0.07 0.01 0.21
Prioria balsamifera 102.6 102.6 0.83 0.05 9.3 0.02 0.15 0.24
Prioria joveri 31.7 31.7 0.08 0.05 0.6 0.02 0.01 0.10
Protomegabaria sp. 28.6 28.6 0.06 0.05 0.6 0.02 0.01 0.10
Pseudospondias cf microcarpa
45.1 46.5 0.32 0.1 3.2 0.03 0.06 0.23
Pseudospondias microcarpa
43.5 64.8 0.80 0.25 8.3 0.08 0.15 0.51
Pseudospondias sp. 31.6 44.1 0.18 0.1 1.7 0.03 0.03 0.14
Psychotria vogeliana 23.6 24.1 0.09 0.1 0.7 0.03 0.02 0.19
Pterocarpus cf tessmannii
25.6 46.4 0.62 0.5 7.0 0.16 0.12 0.69
190
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Pterocarpus soyauxii 81.5 106.9 3.77 0.35 65.5 0.12 0.71 1.30
Pterocarpus tessmannii 66.0 118.3 1.81 0.2 32.0 0.07 0.34 0.61
Pterygota bequaertii 65.4 65.4 0.34 0.05 4.2 0.02 0.06 0.15
Pycnanthus angolensis 79.8 117.5 3.66 0.35 38.7 0.12 0.69 1.14
Rhabdophyllum cf. afine
12.6 17.1 0.04 0.15 0.3 0.05 0.01 0.13
Rhabdophyllum sp. 30.4 35.5 0.15 0.1 2.1 0.03 0.03 0.20
Rinorea dentata 13.8 23.7 0.30 0.95 2.1 0.31 0.06 0.64
Santiria trimera 22.7 71.3 13.15 13.55 117.5 4.46 2.46 8.29
Sarcocephalus pobeguinii
29.7 98 1.13 0.5 15.6 0.16 0.21 0.51
Scorodolphloeus zenkeri
29.7 100 24.00 13.65 337.9 4.49 4.50 10.29
Scottellia coriacea 21.4 45.7 1.57 1.85 13.9 0.61 0.29 1.45
Scottellia klaineana 21.8 32.7 0.09 0.1 0.9 0.03 0.02 0.19
Scyphocephalium mannii
70.2 165.9 29.99 3.4 394.2 1.12 5.62 8.04
Scytopetalum klaineanum
37.6 37.6 0.11 0.05 1.3 0.02 0.02 0.11
Shirakiopsis elliptica 59.0 59 0.27 0.05 3.2 0.02 0.05 0.14
Sorindeia africana 23.4 23.4 0.04 0.05 0.3 0.02 0.01 0.09
Sorindeia cf gabonensis
10.2 10.2 0.01 0.05 0.0 0.02 0.00 0.09
Sorindeia gabonensis 18.9 30.4 0.26 0.4 1.8 0.13 0.05 0.52
Sorindeia sp. 17.9 32 0.36 0.65 2.4 0.21 0.07 0.83
Stachyothyrsus staudtii 21.8 55.7 1.42 1.5 15.4 0.49 0.27 1.03
191
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Staudtia gabonensis 41.8 74.5 1.87 0.6 29.9 0.20 0.35 1.16
Stemonocoleus micranthus
82.1 87.5 1.06 0.1 16.0 0.03 0.20 0.37
Sterculia tragacantha 35.4 35.4 0.10 0.05 1.1 0.02 0.02 0.10
Strephonema cf sericeum
10.0 10 0.01 0.05 0.0 0.02 0.00 0.09
Strephonema mannii 30.9 61.5 0.41 0.2 5.4 0.07 0.08 0.35
Strephonema sp. 18.5 30.4 0.21 0.35 1.8 0.12 0.04 0.43
Strombosia pustulata 15.8 43.3 0.99 2.05 11.4 0.67 0.18 1.68
Strombosiopsis tetrandra
39.3 53.8 0.89 0.35 11.6 0.12 0.17 0.62
Swartzia fistuloides 28.2 37.9 0.14 0.1 2.1 0.03 0.03 0.20
Symphonia globulifera 27.3 34.3 0.43 0.35 4.0 0.12 0.08 0.54
Synsepalum cf. longecuneatum
11.6 11.8 0.02 0.1 0.2 0.03 0.00 0.17
Synsepalum longecuneatum
14.7 24.1 0.09 0.25 0.9 0.08 0.02 0.37
Syzygium cf. congolense
26.9 26.9 0.06 0.05 0.5 0.02 0.01 0.10
Syzygium staudtii 37.1 72 1.06 0.4 14.1 0.13 0.20 0.67
Tabernaemontana crassa
18.4 18.4 0.03 0.05 0.2 0.02 0.00 0.09
Tessmannia africana 29.7 84 2.03 1 36.0 0.33 0.38 1.46
Tessmannia anomala 39.9 85.5 1.30 0.35 25.1 0.12 0.24 0.77
Tessmannia lescrauwaetii
63.0 87.6 1.17 0.15 23.1 0.05 0.22 0.47
Tessmannia sp. 12.3 12.3 0.01 0.05 0.1 0.02 0.00 0.09
192
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Tetraberlinia bifoliolata 22.8 81.4 3.06 2.35 31.6 0.77 0.57 2.10
Tetrapleura tetraptera 57.7 70 0.83 0.15 10.3 0.05 0.16 0.41
Treculia africana 30.2 37.2 0.32 0.2 3.0 0.07 0.06 0.33
Treculia obovoidea 13.0 14.2 0.03 0.1 0.1 0.03 0.01 0.17
Treculia sp. 11.0 11 0.01 0.05 0.0 0.02 0.00 0.09
Tricalysia anomala 11.6 13.3 0.03 0.15 0.2 0.05 0.01 0.26
Trichoscypha abut 20.7 30.8 0.27 0.35 2.6 0.12 0.05 0.37
Trichoscypha acuminata
13.8 24.1 0.40 1.25 2.7 0.41 0.07 1.51
Trichoscypha arborea 46.7 46.7 0.17 0.05 2.2 0.02 0.03 0.12
Trichoscypha sp. 15.3 18.7 0.08 0.2 0.5 0.07 0.01 0.35
Uapaca sp. 50.0 50 0.20 0.05 2.6 0.02 0.04 0.12
Uapaca vanhoutei 34.2 34.2 0.09 0.05 1.0 0.02 0.02 0.10
Uvariastrum sp. 10.5 10.6 0.02 0.1 0.1 0.03 0.00 0.17
Uvariastrum sp1. 17.4 17.4 0.02 0.05 0.1 0.02 0.00 0.09
Uvariastrum sp2. 15.5 15.5 0.02 0.05 0.1 0.02 0.00 0.09
Vangueriopsis cf. rubiginosa
29.2 50 0.25 0.15 2.9 0.05 0.05 0.30
Warnekea sp. 11.4 11.4 0.01 0.05 0.1 0.02 0.00 0.09
Xylopia aethiopica 37.4 65.1 0.90 0.3 8.7 0.10 0.17 0.68
Xylopia phoiodora 27.4 27.4 0.06 0.05 0.6 0.02 0.01 0.10
Xylopia pynaertii 26.4 35.4 0.46 0.4 4.6 0.13 0.09 0.56
Xylopia sp. 25.0 48.6 0.35 0.3 3.8 0.10 0.07 0.51
Xylopia staudtii 31.7 45.6 0.19 0.1 1.4 0.03 0.04 0.21
Zanthoxylum tessmannii
33.2 33.2 0.09 0.05 0.9 0.02 0.02 0.10
Zeyherella sp. 45.0 50 0.32 0.1 4.5 0.03 0.06 0.16
193
APPENDIX D CHARACTERISTICS OF TREE FAMILIES IN THE 20 1-HA PERMANENT PLOTS IN THE CEB-FSC CONCESSION,
GABON
Family name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Anarcadiaceae 20.6 64.8 3.13 3.5 28.6 1.15 0.59 2.89
Annonaceae 24.2 70.0 7.93 6.6 79.1 2.17 1.49 5.83
Apocynaceae 46.2 80.0 2.73 0.7 20.4 0.21 0.51 0.94
Bombacaceae 36.6 36.6 0.11 0.1 0.6 0.02 0.02 0.05
Burseraceae 28.1 185.5 40.3 20.0 421.3 6.58 7.56 20.71
Caesalpiniaceae 27.5 210.0 278.1 157.2 4225.6 51.67 52.10 155.44
Cannabaceae 53.2 78.6 3.95 0.85 58.7 0.28 0.74 1.30
Centroplaceae 14.5 28.8 1.29 3.65 7.9 1.20 0.24 2.64
Chrysobalanaceae 36.5 93.0 2.18 0.65 44.7 0.21 0.41 0.84
Combretaceae 21.9 61.5 0.63 0.6 7.2 0.20 0.12 0.51
Dipterocarpaceae 38.3 105.0 12.1 3.8 210.9 1.25 2.26 4.76
Ebenaceae 14.8 43.9 6.1 15.7 60.2 5.15 1.14 11.43
Euphorbiaceae 23.5 62.9 14.5 13.5 146.7 4.42 2.72 11.56
Fabaceae 39.6 55.0 1.27 0.45 20.9 0.15 0.24 0.53
Guttiferae 14.7 34.3 1.52 3.9 13.3 1.28 0.29 2.85
Huaceae 15.0 29.5 0.31 0.8 2.1 0.26 0.06 0.58
Icacinaceae 10.0 10.0 0.01 0.05 0.0 0.02 0.00 0.03
Unkown 14.7 37.2 0.58 1.45 5.0 0.48 0.11 1.06
Irvingiaceae 37.2 115.0 7.04 2.3 137.2 0.76 1.32 2.83
Ixonanthaceae 23.4 46.9 0.31 0.25 4.5 0.08 0.06 0.22
Lauraceae 20.0 75.0 1.44 1.55 16.2 0.51 0.27 1.29
Lecythidaceae 47.0 56.9 0.71 0.2 10.1 0.07 0.13 0.26
194
Family name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Linaceae 38.6 56.3 0.75 0.3 12.7 0.10 0.14 0.34
Loganiaceae 13.5 15.5 0.09 0.3 0.4 0.10 0.02 0.21
Malvaceae 25.1 75.0 10.5 7.85 113.4 2.58 1.96 7.13
Melastomataceae 11.5 11.6 0.02 0.1 0.1 0.03 0.00 0.07
Meliaceae 27.3 50.0 2.03 1.6 19.3 0.53 0.38 1.43
Mimosaceae 34.1 196.0 33.3 9.7 538.0 3.19 6.25 12.62
Moraceae 26.6 52.7 0.99 0.75 7.9 0.25 0.19 0.68
Myristicaceae 66.6 165.9 38.1 4.75 499.8 1.56 7.14 10.26
Myrtaceae 35.9 72.0 1.11 0.45 14.7 0.15 0.21 0.50
Ochnaceae 18.3 35.5 0.20 0.3 2.5 0.10 0.04 0.23
Olocaceae 27.8 98.8 18.0 10.7 303.0 3.50 3.37 10.37
Pandaceae 42.6 69.4 5.82 1.9 67.1 0.62 1.09 2.34
Papilionaceae 24.1 118.3 11.7 7.45 169.5 2.45 2.20 7.10
Passifloraceae 15.3 24.4 0.18 0.45 1.4 0.15 0.03 0.33
Putranjuvaceae 25.1 59.1 2.68 2.05 34.0 0.67 0.50 1.85
Rhizophoraceae 20.3 24.6 0.07 0.1 0.7 0.03 0.01 0.08
Rubiaceae 19.5 98.0 7.39 8.95 78.8 2.94 1.38 7.27
Rutaceae 33.2 33.2 0.09 0.05 0.9 0.02 0.02 0.05
Salicaceae 22.2 64.0 2.12 2.2 21.5 0.72 0.40 1.84
Sapindaceae 17.1 49.1 2.27 3.9 18.6 1.30 0.43 3.02
Sapotaceae 52.6 160.0 8.26 1.2 157.8 0.38 1.55 2.30
Scytopetalaceae 24.8 41.1 0.43 0.4 5.3 0.12 0.08 0.31
Simabouraceae 42.8 80.0 1.11 0.3 8.3 0.10 0.21 0.41
Violaceae 13.8 23.7 0.30 0.9 2.1 0.31 0.06 0.68
195
APPENDIX E CHARACTERISTICS OF TREE SPECIES IN THE 12 1-HA PERMANENT PLOTS IN THE SEEF-CL CONCESSION,
GABON
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Afrostyrax lepidophyllus
17.1 46.7 1.41 4.4 10.9 1.14 0.37 2.78
Afzelia bipindense 11.4 12.5 0.02 0.2 0.1 0.04 0.01 0.28
Allanblackia cf. klainei 39.8 39.8 0.12 0.1 1.5 0.02 0.03 0.17
Amphimas ferrugineus 32.5 37 0.17 0.2 1.9 0.04 0.04 0.32
Andosmia sp. 11.2 11.2 0.01 0.1 0.1 0.02 0.00 0.14
Angylocalyx sp. 21.2 32.4 0.28 0.6 2.6 0.15 0.07 0.92
Anisophyllea myriostica
29.6 29.6 0.07 0.1 0.8 0.02 0.02 0.15
Anonidium mannii 35.4 60 1.40 1.0 7.9 0.26 0.37 1.09
Anopyxis klaineana 52.4 52.4 0.22 0.1 3.8 0.02 0.06 0.19
Anthonotha cf. wijmacampensis
42.4 42.4 0.14 0.1 2.0 0.02 0.04 0.17
Anthonotha sp. 16.0 23.6 0.07 0.3 0.6 0.06 0.02 0.43
Aucoumea klaineana 110.1 171.9 41.33 3.3 442.9 0.86 11.02 13.15
Baikiea sp. 25.4 25.4 0.05 0.1 0.5 0.02 0.01 0.15
Baphia cf. buettneri 18.1 43.8 2.24 6.1 22.6 1.57 0.60 2.74
Barteria fistulosa 14.1 17.5 0.11 0.6 0.8 0.15 0.03 0.87
Beilschmiedia cf. klainei
14.2 14.2 0.02 0.2 0.1 0.04 0.00 0.28
Beilschmiedia fulva 34.1 40 0.28 0.3 3.0 0.06 0.07 0.37
Beilschmiedia gdes feuilles
36.6 36.6 0.11 0.1 1.2 0.02 0.03 0.16
Beilschmiedia klainei 13.4 13.4 0.01 0.1 0.1 0.02 0.00 0.14
196
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Beilschmiedia ptes feuilles
13.5 20.6 0.06 0.3 0.4 0.09 0.02 0.22
Beilschmiedia sp. 15.0 40.5 1.05 4.3 8.1 1.10 0.28 2.64
Berlinia auriculata 12.6 14.3 0.08 0.5 0.5 0.13 0.02 0.49
Berlinia cf. confusa 16.7 28.2 0.20 0.7 1.5 0.17 0.05 0.80
Berlinia cf. congolensis 50.2 50.2 0.20 0.1 2.6 0.02 0.05 0.19
Bikinia grisea 43.6 134.6 8.74 3.2 113.7 0.82 2.33 3.72
Blighia welwitschii 26.5 42.6 0.26 0.3 3.5 0.09 0.07 0.62
Brenania brieyi 52.2 52.2 0.21 0.1 2.8 0.02 0.06 0.19
Calpocalyx dinklagei 16.5 24.2 0.27 1.0 2.3 0.26 0.07 1.02
Canarium schweinfurthii
74.5 74.5 0.44 0.1 4.4 0.02 0.12 0.25
Carapa procera 22.8 29.5 0.34 0.7 2.8 0.17 0.09 0.84
Casearia barteri 35.1 35.1 0.10 0.1 1.1 0.02 0.03 0.16
Celtis tessmannii 60.0 95.6 8.84 2.4 140.3 0.62 2.36 4.02
Centroplacus glaucinus
14.9 25 1.71 7.7 10.9 1.98 0.46 3.82
Chrysophyllum lacourtianum
47.2 84.6 1.81 0.7 27.2 0.17 0.48 1.12
Chrysophyllum sp. 12.8 12.8 0.01 0.1 0.1 0.02 0.00 0.14
Chytranthus sp. 13.2 18.1 0.15 0.9 1.1 0.24 0.04 1.20
Cleistanthus itsogensis 35.0 65.5 2.98 2.2 22.3 0.56 0.79 1.93
Cleistanthus racemosa 27.7 39.8 0.20 0.3 1.2 0.06 0.05 0.35
Coelocaryon preussii 42.7 71.9 5.17 2.8 51.5 0.73 1.38 3.26
Cola lizea 14.6 20 0.09 0.4 0.6 0.11 0.02 0.25
Cola sp. 15.1 18.8 0.13 0.6 0.8 0.15 0.04 0.76
197
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Copaifera mildbraedii 30.1 46 0.50 0.5 6.0 0.13 0.13 0.72
Cylicodiscus gabunensis
98.9 116.1 2.34 0.3 51.5 0.06 0.62 1.03
Dacryodes buettneri 46.6 114.2 11.46 4.4 137.1 1.14 3.06 5.58
Dacryodes edulis 15.6 17.7 0.08 0.3 0.4 0.09 0.02 0.45
Dacryodes igaganga 19.2 43.3 1.66 3.9 13.4 1.01 0.44 2.83
Dacryodes klaineana 23.0 31.7 0.18 0.3 1.9 0.09 0.05 0.59
Dacryodes macrophylla
45.1 57.1 0.66 0.3 7.4 0.09 0.18 0.72
Dacryodes normandii 26.4 65.2 1.11 1.3 11.1 0.32 0.30 1.42
Dacryodes sp. 52.3 55.3 0.43 0.2 5.1 0.04 0.11 0.39
Daniellia klainei 65.1 96.4 0.82 0.2 9.3 0.04 0.22 0.38
Daniellia soyauxii 28.9 32 0.13 0.2 1.0 0.04 0.04 0.31
Desbordesia glaucescens
37.3 62.9 0.97 0.7 18.2 0.17 0.26 0.89
Detarium macrocarpum
52.1 63.6 0.45 0.2 7.1 0.04 0.12 0.39
Dialium angolense 23.0 97.1 9.06 12.1 151.4 3.12 2.42 6.92
Dialium bipindense 22.9 59.9 3.07 4.8 52.1 1.25 0.82 3.45
Dialium dinklagei 23.4 42.2 0.50 0.8 5.9 0.22 0.13 0.81
Dialium sp. 19.5 40 1.05 2.5 12.7 0.65 0.28 1.96
Dialium tessmannii 16.6 28.6 0.48 1.7 4.9 0.43 0.13 1.48
Dichostemma glaucescens
13.8 20.4 0.37 2.0 1.7 0.52 0.10 0.96
Diogoa zenkeri 32.9 69.8 4.39 3.7 58.8 0.95 1.17 2.69
Diospyros cf. crassiflora
65.9 65.9 0.34 0.1 6.8 0.02 0.09 0.23
198
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Diospyros fragrans 23.7 32.7 0.15 0.3 2.0 0.06 0.04 0.34
Diospyros hoyleana 14.5 27.8 0.41 1.8 3.9 0.47 0.11 1.39
Diospyros macroceras 12.7 12.7 0.01 0.1 0.1 0.02 0.00 0.14
Diospyros melocarpa 17.0 44.7 3.68 12.0 39.6 3.10 0.98 5.46
Diospyros sp. 13.2 15.5 0.10 0.6 0.8 0.15 0.03 0.87
Discoglypremna caloneura
47.5 47.5 0.18 0.1 1.2 0.02 0.05 0.18
Distemonanthus benthamianus
39.8 59.1 0.70 0.4 8.7 0.11 0.19 0.75
Dracaena cf. arborea 14.3 14.3 0.02 0.1 0.1 0.02 0.00 0.14
Drypetes gossweileri 24.5 24.5 0.05 0.1 0.5 0.02 0.01 0.15
Drypetes sp. 25.6 63.7 1.16 1.3 16.6 0.32 0.31 1.78
Duguetia confinis 24.4 30.6 0.10 0.2 0.9 0.04 0.03 0.18
Duvigneaudia inopinata
35.7 41.3 0.51 0.4 4.1 0.11 0.14 0.82
Engomangoma gordonii
43.8 46.7 0.30 0.2 5.0 0.04 0.08 0.24
Entandrophragma candollei
50.4 87.1 0.61 0.2 9.1 0.04 0.16 0.44
Entandrophragma congoense
27.1 31.9 0.18 0.3 1.4 0.06 0.05 0.34
Eriocoelum macrocarpum
20.1 38.9 0.30 0.7 2.4 0.17 0.08 0.83
Eriocoelum sp. 19.1 26.6 0.09 0.3 0.7 0.06 0.03 0.20
Erismadelphus exsul 22.5 33.3 0.23 0.4 2.1 0.11 0.06 0.51
Erythropleum ivorense 77.1 77.1 0.47 0.1 9.2 0.02 0.12 0.26
Erythroxylum mannii 18.3 18.3 0.03 0.1 0.2 0.02 0.01 0.14
199
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Fillaeopsis discophora 37.2 118.7 7.05 3.8 97.7 0.97 1.88 4.11
Funtumia africana 31.1 31.1 0.08 0.1 0.5 0.02 0.02 0.16
Garcinia epunctata 13.6 24.3 1.74 9.5 15.4 2.45 0.47 3.95
Garcinia smeathmannii 10.9 10.9 0.01 0.1 0.1 0.02 0.00 0.14
Garcinia sp. 12.9 14.8 0.11 0.7 0.8 0.17 0.03 0.89
Gilbertiodendron dewevrei
17.5 28.4 0.32 1.0 2.9 0.26 0.09 0.46
Gilbertiodendron ogoouense
29.7 78.9 0.95 0.8 14.0 0.19 0.25 0.79
Gilbertiodendron sp. 16.4 33.9 0.16 0.5 1.5 0.13 0.04 0.40
Gilletiodendron pierreanum
37.5 89.5 5.79 3.2 122.0 0.82 1.54 2.94
Grewia coriacea 19.3 40.6 1.12 2.7 7.1 0.69 0.30 1.91
Guarea sp. 25.4 30.9 0.11 0.2 0.9 0.04 0.03 0.30
Guarea thompsonii 28.4 28.4 0.06 0.1 0.5 0.02 0.02 0.15
Guibourtia tessmannii 57.6 77.4 0.83 0.3 10.7 0.06 0.22 0.63
Heisteria parvifolia 34.3 62.6 4.09 3.3 64.1 0.86 1.09 3.22
Heisteria sp. 16.1 17.4 0.08 0.3 0.7 0.09 0.02 0.34
Hodoniodendron micrantha
23.0 23 0.04 0.1 0.4 0.02 0.01 0.15
Hymenostegia pellegrinii
23.4 64.2 3.70 5.6 54.9 1.44 0.99 3.23
Irvingia excelsa 57.9 67.1 0.80 0.3 14.5 0.06 0.21 0.51
Irvingia gabonensis 23.6 42.4 1.32 2.3 15.9 0.58 0.35 2.20
Irvingia grandifolia 34.0 82.9 0.91 0.6 16.5 0.15 0.24 0.85
Irvingia robur 15.4 15.4 0.02 0.1 0.2 0.02 0.00 0.14
200
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Julbernardia pellegriniana
58.4 175 18.68 4.4 324.6 1.14 4.98 6.93
Julbernardia seretii 30.2 42.4 0.40 0.4 4.9 0.11 0.11 0.33
Klaineanthus gabonii 20.5 33.5 1.46 3.3 9.1 0.86 0.39 2.63
Klainedoxa gabonensis 31.3 66.7 1.66 1.3 32.1 0.34 0.44 1.71
Klainedoxa trillesii 40.5 40.5 0.13 0.1 2.3 0.02 0.03 0.17
Lindackeria dentata 11.9 11.9 0.01 0.1 0.1 0.02 0.00 0.14
Lovoa trichilioides 52.5 53.1 0.43 0.2 4.2 0.04 0.12 0.39
Macaranga monandra 19.1 35.8 0.21 0.5 1.2 0.13 0.06 0.76
Mammea africana 40.0 52.3 0.40 0.3 4.9 0.06 0.11 0.52
Manilkera fouilloyana 18.6 18.6 0.03 0.1 0.3 0.02 0.01 0.14
Maranthes chrysophylla
53.7 70.5 1.01 0.3 20.4 0.09 0.27 0.82
Maranthes gabunensis 49.9 49.9 0.20 0.1 3.6 0.02 0.05 0.19
Maranthes glabra 71.7 122.9 2.52 0.4 59.6 0.11 0.67 1.12
Mareyopsis longifolia 11.2 13.6 0.11 0.9 0.4 0.24 0.03 0.61
Marquesia excelsa 33.4 78 2.99 1.8 52.6 0.47 0.80 2.31
Memecylon sp. 31.2 45 0.26 0.3 3.9 0.06 0.07 0.36
Millettia cf. grifoniana 11.9 11.9 0.01 0.1 0.1 0.02 0.00 0.14
Millettia laurentii 19.8 57.1 0.93 1.8 12.0 0.47 0.25 1.87
Mitragyna ciliata 42.8 105.6 2.28 1.0 28.2 0.26 0.61 1.33
Monodora sp. 13.5 15.2 0.03 0.2 0.2 0.04 0.01 0.28
Nauclea diderrichii 40.5 46.5 0.39 0.3 5.1 0.06 0.10 0.40
Neochevolierodendron staphanii
25.1 53.6 3.38 4.7 37.5 1.20 0.90 2.34
Ochna sp. 16.8 23.4 0.09 0.3 0.9 0.09 0.02 0.46
201
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Odyendyea gabonensis
34.2 80 0.86 0.5 6.4 0.13 0.23 0.70
Oncoba glauca 17.6 21.1 0.13 0.4 1.0 0.11 0.03 0.60
Ongokea gore 56.2 82.9 1.43 0.4 26.2 0.11 0.38 0.95
Oxystigma oxyphyllum 41.8 78.5 1.35 0.6 13.0 0.15 0.36 0.86
Pancovia cf. floribunda 14.8 24.5 0.28 1.3 2.2 0.32 0.07 1.43
Panda oleosa 32.8 54.5 1.00 0.8 10.8 0.22 0.27 1.06
Parkia bicolor 57.4 88.3 1.40 0.4 14.8 0.11 0.37 0.83
Pausinystalia johimbe 23.9 40.1 1.12 1.8 11.0 0.45 0.30 1.44
Pausinystalia macroceras
19.4 59.9 6.38 15.3 55.2 3.96 1.70 7.04
Pentaclethra eetveldeana
32.3 65 4.15 3.5 53.4 0.90 1.11 3.28
Pentaclethra macrophylla
39.7 67.4 1.53 0.9 26.5 0.24 0.41 1.10
Petersianthus macrocarpus
48.0 80 1.83 0.8 27.9 0.19 0.49 1.14
Phyllocosmus africanus
26.0 59.1 0.48 0.5 7.6 0.13 0.13 0.83
Picralima nitida 12.6 16.3 0.04 0.3 0.3 0.06 0.01 0.31
Pinacopodium congolense
31.4 56.1 2.07 1.9 33.4 0.49 0.55 2.20
Piptadeniastrum africanum
82.7 96 3.28 0.5 52.3 0.13 0.87 1.58
Plagiosiphon gabonensis
19.2 41.1 0.39 0.9 3.7 0.24 0.10 0.46
Plagiostyles africana 29.2 49.8 5.57 6.7 69.0 1.72 1.49 4.59
202
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Polyalthia gabonensis 13.6 19.8 0.17 0.9 1.2 0.24 0.04 0.74
Polyalthia suaveolens 18.2 46.6 6.55 18.3 62.8 4.71 1.75 7.84
Porterandia cladantha 17.4 30.5 0.30 0.9 2.4 0.24 0.08 1.12
Pouteria altissima 43.5 43.5 0.15 0.1 1.3 0.02 0.04 0.18
Pseudospondias cf. microcarpa
16.8 18.6 0.04 0.2 0.3 0.04 0.01 0.29
Pseudospondias macrocarpum
23.4 23.4 0.04 0.1 0.3 0.02 0.01 0.15
Pseudospondias sp. 23.2 33.8 0.27 0.5 2.0 0.13 0.07 0.55
Pteleopsis hylodendron
66.9 66.9 0.35 0.1 4.2 0.02 0.09 0.23
Pterocarpus soyauxii 66.3 89.8 1.12 0.3 18.5 0.06 0.30 0.59
Pterocarpus tessmannii
35.1 60.9 0.61 0.4 8.4 0.11 0.16 0.50
Pycnanthus angolensis 68.3 91.3 2.68 0.6 26.9 0.15 0.72 1.44
Rinorea dentata 17.6 32.8 0.88 2.8 7.3 0.71 0.23 1.75
Rothmannia sp. 21.5 31.6 0.16 0.3 1.4 0.09 0.04 0.36
Rovolfia cf. macrophylla
15.6 15.6 0.02 0.1 0.1 0.02 0.01 0.14
Santiria trimera 19.9 63.7 23.74 52.3 199.2 13.51 6.33 21.22
Scorodophloeus zenkeri
23.5 64.8 34.14 51.6 430.1 13.31 9.10 23.80
Scottellia coriacea 14.1 14.1 0.02 0.1 0.1 0.02 0.00 0.14
Scyphocephalium mannii
57.5 113.2 48.77 14.3 586.8 3.70 13.00 18.08
Scytopetalum klaineanum
43.6 49.2 0.30 0.2 3.8 0.04 0.08 0.35
203
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Sorindeia gabonensis 20.1 46.8 0.54 1.2 4.2 0.30 0.14 1.14
Sorindeia sp. 13.8 21.3 0.11 0.6 0.6 0.15 0.03 0.76
Stachyothyrsus staudtii 10.3 10.3 0.01 0.1 0.1 0.02 0.00 0.14
Staudtia gabonensis 33.4 69.6 3.33 2.8 49.0 0.71 0.89 2.75
Strombosia pustulata 17.2 49.1 2.44 7.3 28.8 1.87 0.65 3.79
Strombosiopsis tetrandra
32.5 60.3 3.17 2.9 38.3 0.75 0.85 2.63
Symphonia globulifera 31.0 45.8 0.19 0.2 2.2 0.04 0.05 0.32
Synsepalium cf. longecuneatum
15.5 31.1 0.38 1.5 3.7 0.39 0.10 1.41
Syzygium oworiense 24.6 56.5 0.82 1.1 13.3 0.28 0.22 1.07
Syzygium staudtii 50.1 57.6 0.60 0.3 8.0 0.06 0.16 0.57
Tabernaemontana crassa
15.9 16.7 0.06 0.3 0.4 0.06 0.02 0.31
Tessmannia africana 20.1 36.7 0.52 1.2 6.3 0.30 0.14 1.01
Tessmannia anomala 14.7 19.2 0.05 0.3 0.5 0.06 0.01 0.31
Tessmannia cf. anomala
19.3 19.3 0.03 0.1 0.3 0.02 0.01 0.14
Tessmannia lescrauwaetii
23.0 54.3 0.31 0.4 4.6 0.11 0.08 0.54
Tetraberlinia bifoliolata 26.2 61.6 1.05 1.3 10.0 0.32 0.28 0.95
Tetrapleura tetraptera 55.2 67.9 0.50 0.2 6.1 0.04 0.13 0.41
Thomandersia cf. congolana
11.6 15.8 0.33 2.6 2.1 0.67 0.09 1.33
Treculia africana 53.0 53 0.22 0.1 2.5 0.02 0.06 0.20
Treculia obovoidea 23.9 39.9 3.42 5.9 27.2 1.53 0.91 3.24
Tricalysia cf. anomala 12.2 12.2 0.01 0.1 0.1 0.02 0.00 0.14
204
Species name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Trichilia sp. 10.5 10.5 0.01 0.1 0.1 0.02 0.00 0.14
Trichoscypha abut 16.8 28.4 0.50 1.8 3.9 0.47 0.13 1.53
Trichoscypha acuminata
14.9 29.4 0.71 3.2 5.2 0.82 0.19 2.39
Trichoscypha arborea 42.9 42.9 0.14 0.1 1.8 0.02 0.04 0.18
Trichoscypha cf. lucens
14.6 22.9 0.14 0.7 1.0 0.17 0.04 0.56
Uapaca paludosa 11.0 11 0.01 0.1 0.1 0.02 0.00 0.14
Uvariastrum sp. 14.5 18 0.03 0.2 0.2 0.04 0.01 0.28
Unknown 24.3 121.2 2.55 2.5 12.1 0.65 0.68 2.59
Vangueriopsis cf. rubiginosa
18.4 18.4 0.03 0.1 0.2 0.02 0.01 0.14
Vangueriopsis rubiginosa
28.3 28.3 0.06 0.1 0.6 0.02 0.02 0.15
Xylopia aethiopica 35.7 49.8 0.57 0.4 4.8 0.11 0.15 0.72
Xylopia hypolampra 50.9 50.9 0.20 0.1 2.8 0.02 0.05 0.19
Xylopia pynaertii 16.7 21 0.07 0.3 0.5 0.06 0.02 0.31
Xylopia quintassii 23.8 27.7 0.09 0.2 1.0 0.04 0.02 0.18
Xylopia sp. 18.2 29.7 0.29 0.8 2.4 0.22 0.08 0.98
Xylopia staudtii 18.1 31.4 0.32 0.9 1.7 0.24 0.08 1.01
Zanthoxylum heitzii 66.9 66.9 0.35 0.1 3.8 0.02 0.09 0.23
205
APPENDIX F CHARACTERISTICS OF TREE FAMILIES IN THE 12 1-HA PERMANENT PLOTS IN THE SEEF-CL CONCESSION,
GABON
Family name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Acanthaceae 11.6 15.8 0.33 2.6 2.1 0.67 0.09 1.42
Agavaceae 14.3 14.3 0.02 0.1 0.1 0.02 0.00 0.05
Anacardiaceae 16.9 46.8 2.51 8.2 19.3 2.11 0.67 4.89
Anisophylleaceae 29.6 29.6 0.07 0.1 0.8 0.02 0.02 0.06
Annonaceae 19.2 60 9.82 23.3 86.42 6.02 2.62 14.67
Apocynaceae 16.5 31.1 0.19 0.7 1.3 0.17 0.05 0.40
Burseraceae 26.8 171.9 86.71 73.3 892.1 18.91 23.12 60.97
Caesalpiniaceae 26.0 175 99.64 106.3 1440.2 27.45 26.56 81.45
Cannabaceae 60.0 95.6 8.84 2.4 140.3 0.62 2.36 3.60
Centroplacaceae 15.0 25 1.75 7.8 10.9 2.00 0.47 4.47
Chrysobalanaceae 62.3 122.9 3.73 0.8 83.5 0.22 0.99 1.42
Combretaceae 66.9 66.9 0.35 0.1 4.2 0.02 0.09 0.14
Dipterocarpaceae 33.4 78 2.99 1.8 52.6 0.47 0.80 1.74
Ebenaceae 16.9 65.9 4.69 14.8 53.2 3.83 1.25 8.91
Erythroxylaceae 18.3 18.3 0.03 0.1 0.2 0.02 0.01 0.05
Euphorbiaceae 22.4 65.5 6.02 9.8 41.4 2.52 1.60 6.64
Guttiferae 14.7 52.3 2.59 10.8 24.9 2.80 0.69 6.28
Huaceae 17.1 46.7 1.41 4.4 10.9 1.14 0.37 2.66
Hypericaceae 11.2 11.2 0.01 0.1 0.1 0.02 0.00 0.05
Inconnu 18.5 43.1 0.50 1.2 5.2 0.30 0.13 0.74
Irvingiaceae 30.0 82.9 5.83 5.3 99.7 1.38 1.55 4.31
Lauraceae 16.2 40.5 1.53 5.1 12.8 1.31 0.41 3.03
Lecythidaceae 48.0 80 1.83 0.8 27.9 0.19 0.49 0.87
Leguminoseae 21.9 21.9 0.04 0.1 0.3 0.02 0.01 0.05
206
Family name Average dbh (cm)
Maximum tree size (cm)
Basal area (m2 ha-1)
Density (stems ha-1)
AGB (Mg)
Relative density (%)
Relative dominance (%)
Relative importance (%)
Linaceae 30.3 59.1 2.55 2.4 40.9 0.62 0.68 1.93
Malvaceae 18.1 40.6 1.34 3.7 8.5 0.95 0.36 2.25
Melastomataceae 31.2 45 0.26 0.3 3.9 0.06 0.07 0.20
Meliaceae 29.4 87.1 1.74 1.6 18.9 0.41 0.46 1.28
Mimosaceae 39.2 121.2 21.67 10.6 304.6 2.73 5.78 11.24
Moraceae 24.3 53 3.64 6.0 29.7 1.55 0.97 4.07
Myristicaceae 52.4 113.2 60.00 20.6 714.2 5.31 16.00 26.58
Myrtaceae 29.4 57.6 1.42 1.3 21.4 0.34 0.38 1.07
Ochnaceae 16.8 23.4 0.09 0.3 0.9 0.09 0.02 0.20
Olacaceae 27.1 82.9 15.91 18.1 221.9 4.67 4.24 13.56
Pandaceae 32.8 54.5 1.00 0.8 10.8 0.22 0.27 0.70
Papilionaceae 20.6 89.8 5.20 9.3 64.3 2.41 1.39 6.21
Passifloraceae 14.1 17.5 0.11 0.6 0.8 0.15 0.03 0.33
Putranjuvaceae 25.6 63.7 1.21 1.3 17.1 0.34 0.32 1.01
Rhizophoraceae 52.4 52.4 0.22 0.1 3.8 0.02 0.06 0.10
Rubiaceae 21.2 105.6 10.97 20.1 107.0 5.18 2.92 13.30
Rutaceae 66.9 66.9 0.35 0.1 3.8 0.02 0.09 0.14
Salicaceae 18.6 35.1 0.25 0.7 2.2 0.17 0.07 0.41
Sapindaceae 16.7 42.6 1.10 3.5 9.9 0.90 0.29 2.10
Sapotaceae 25.2 84.6 2.38 2.4 32.6 0.62 0.63 1.88
Scytopetalaceae 43.6 49.2 0.30 0.2 3.8 0.04 0.08 0.17
Simaroubaceae 34.2 80 0.86 0.5 6.4 0.13 0.23 0.49
Violaceae 17.6 32.8 0.88 2.8 7.3 0.71 0.23 1.65
Vochysiaceae 22.5 33.3 0.23 0.4 2.1 0.11 0.06 0.28
207
APPENDIX G LOGGING ACTIVITIES REPORTEDLY CONDUCTED UNDER CONVENTIONAL (CL)
AND REDUCED-IMPACT LOGGING (RIL) SYSTEMS.
Logging phase Logging activity CL (n=12) RIL (n=11)a
Pre-harvest
Management plan 2 2
Process and licensee fees 4 2
Pre-felling inventory 1 10
Environmental impact assessment
1 2
Boundary demarcation 5 9
Data processing 2 6
Tree hunting 3 0
Harvest planning
Tree marking & mapping1 3 11
Vine cutting 0 7 Road planning2 3 8
Skid trail planning 0 4
Log deck planning 0 3
Training 0 7
Infrastructure
Road construction & maintenance
10 8
Skid trails layout 1 4
Log deck construction 4 4
Harvest
Log deck operations3 11 10
Felling & bucking 12 11
Skidding 11 9
Waste adjustment 1 1
Log transport4 7 5
Post-harvest Monitoring/supervision 4 7
Research fees5 2 3
General expenses
Other expenditures6 8 7
Taxation 4 2
Closing reports 1 1
Premiums & royalties 6 5
Administration7 4 3
Heavy equipment investment8 1 3
Building investment & maintenance9
2 2
a Studies using the Malaysian Criteria and Indicator (MC&I; Rahim Nik et al., 2002) and the modified excavator “Logfisher” (Rahim Abdul et al., 2009) were considered as RIL studies. 1 This comprises a GPS plan and tree tagging (Rahim Abdul et al., 2009), tree marking, and tree mapping. 2 This includes road planning and proposed road alignment (Rahim Abdul et al., 2009).
208
3 Log deck operations include log landing operations (Bacha et al., 2007) and log loading. 4 Included are log hauling to mills, log loading, short distance haulage (Rahim Abdul et al., 2009), and log rafting to mills. 5 These include rehabilitation treatments (Rahim Abdul et al., 2009), silvicultural obligation, research & development fee, and social payments. 6 This includes support and logistics (Bacha et al., 2007) and overhead. 7 Included are log grading & administration and log yard administration (Rahim Abdul et al., 2009). 8 This includes “Logfisher” (Rahim Abdul et al., 2009), depreciation of tractors and other vehicles (Bacha et al., 2007). 9 It includes base camp, lodging (Bacha et al., 2007).
209
APPENDIX H TRANSITION PROBABILITIES AND OTHER PARAMETERS USED IN THE MATRIX MODEL. Gi IS THE PROBABILITY
OF GROWING TO THE NEXT CLASS; STASIS IN THE PROBABILITY OF REMAINING IN THE SAME CLASS.
Non-Pioneer Light Demander (NPLD) species.
DBH Class (cm)
Gi Stasis Mortality rate (%)
Proportion of damage trees
Recruitment (stems ha-1 yr-1)
Tree abundance 10-20 0.011 0.964 2.5 0.05 2.3 20-30 0.018 0.966 1.6 0.05 - 30-40 0.022 0.970 0. 8 0.07 - 40-50 0.020 0.968 1.2 0.06 - 50-60 0.009 0.991 0.0 0.14 - 60-70 0.017 0.970 1.3 0.03 - 70-80 0.011 0.937 5.1 0.05 - 80-90 0.029 0.971 0.0 0.06 - 90-100 0.012 0.945 4.3 0.08 - > 100 0.009 0.967 2.4 0.00 - Above-ground biomass Loss rate (%) Mg ha-1 yr-1
10-20 0.146 0.834 2.0 0.05 0.149 20-30 0.161 0.820 1.9 0.06 0.164 30-40 0.117 0.876 0.7 0.06 0.118 40-50 0.136 0.851 1.3 0.06 0.138 50-60 0.064 0.936 0.0 0.17 0.064 60-70 0.143 0.843 1.5 0.02 0.145 70-80 0.039 0.904 5.6 0.04 0.042 80-90 0.000 1.000 0.0 0.06 0.000 90-100 0.003 0.969 2.8 0.09 0.003 > 100 0.009 0.980 1.1 0.00 0.009
210
Pioneer species.
DBH Class (cm)
Gi Stasis Mortality rate (%)
Proportion of damage trees
Recruitment (stems ha-1 yr-1)
Tree abundance 10-20 0.013 0.962 2.5 0.05 0.5 20-30 0.026 0.974 0.0 0.02 - 30-40 0.032 0.948 2.0 0.04 - 40-50 0.051 0.949 0.0 0.11 - 50-60 0.024 0.977 0.0 0.00 - 60-70 0.043 0.957 0.0 0.00 - 70-80 0.015 0.985 0.0 0.00 - 80-90 0.032 0.834 1.3 0.25 - 90-100 0.000 0.707 2.9 1.00 - > 100 0.010 0.990 0.0 0.00 - Above-ground biomass Loss rate (%) Mg ha-1 yr-1
10-20 0.037 0.938 2.4 0.04 0.038 20-30 0.053 0.947 0.0 0.02 0.053 30-40 0.063 0.914 2.2 0.05 0.065 40-50 0.048 0.952 0.0 0.08 0.048 50-60 0.002 0.998 0.0 0.00 0.002 60-70 0.009 0.991 0.0 0.00 0.009 70-80 0.027 0.973 0.0 0.00 0.027 80-90 0.054 0.842 1.1 0.20 0.060 90-100 0.028 0.687 2.8 1.00 0.039 > 100 0.009 0.992 0.0 0.00 0.008
211
Shade Tolerant species.
DBH Class (cm)
Gi Stasis Mortality rate (%)
Proportion of damage trees
Recruitment (stems ha-1 yr-1)
Abundance 10-20 0.011 0.958 3.1 0.05 8.2 20-30 0.017 0.965 1.8 0.07 - 30-40 0.018 0.971 1.1 0.06 - 40-50 0.014 0.971 1.5 0.06 - 50-60 0.007 0.993 0.0 0.02 - 60-70 0.015 0.975 1.0 0.06 - 70-80 0.009 0.945 4.7 0.09 - 80-90 0.008 0.935 5.7 0.00 - 90-100 0.048 0.952 0.0 0.00 - > 100 0.011 0.989 0.0 0.00 - Above-ground biomass Loss rate (%) Mg ha-1 yr-1
10-20 0.444 0.528 2.8 0.06 0.467 20-30 0.477 0.506 1.7 0.07 0.485 30-40 0.425 0.564 1.1 0.05 0.430 40-50 0.262 0.719 1.9 0.07 0.267 50-60 0.090 0.910 0.0 0.02 0.090 60-70 0.147 0.843 1.0 0.05 0.148 70-80 0.054 0.899 4.7 0.10 0.057 80-90 0.016 0.933 5.1 0.00 0.017 90-100 0.009 0.991 0.0 0.00 0.008 > 100 0.086 0.914 0.0 0.00 0.059
212
APPENDIX I CHANGES IN NUMBER OF TREES PER STEMS DIAMETER CLASS IN 10 1-HA PERMANENT PLOTS FOR EACH
SPECIES FUNCTIONAL GROUP TWO AFTER REDUCED-IMPACT LOGGING ON MONTS DE CRISTAL IN GABON.
Non-Pioneer Light Demander Species 2009 Trees
(Stems ha-
1)
Gain (stems ha-1)
2009-2011 Mortality (stems ha-1) Trees (Stems ha-1)
Growth (cm yr-1)
Tree Mortality rate (%)
DBH Class (cm)
Before logging
Ingrowth # Tree harvested
Damaged trees
Undamaged trees
Total 2011 2009-2011
Natural Logging
10 54.1 10 - 2.1 0.6 2.7 51.3 0.11 0.6 2.0
20 18.7 - - 0.4 0.2 0.6 17.7 0.19 0.5 1.1 30 12.3 - - 0.1 0.1 0.2 13.3 0.22 0. 4 0. 4
40 8.4 - - 0.1 0.1 0.2 8.7 0.20 0. 6 0. 6 50 4.3 - - - - - 4.4 0.09 - -
60 4.0 - - - 0.1 0.1 4.1 0.17 0. 13 - 70 2.0 - - 0.1 0.1 0.2 1.8 0.12 0. 25 2.5
80 1.8 - - - - - 1.7 0.29 - - 90 1.2 - - - 0.1 0.1 1.2 0.13 0. 43 -
100+ 2.1 - 2 0.1 - 0.1 1.9 0.09 - 2.4
213
Pioneer Species 2009 Trees
(Stems ha-1) Gain (stems ha-1)
2009-2011 Mortality (stems ha-1) Trees (Stems ha-
1)
Growth (cm yr-1)
Tree Mortality rate
DBH Class (cm)
Before logging
Ingrowth # Tree harvested
Damaged trees
Undamaged trees
Total 2011 2009-2011
Natural Logging
10 22.1 5 - 0.8 0.3 0.11 18.9 0.13 0. 7 1.8
20 4.6 - - - - - 4.3 0.26 - - 30 2.5 - - 0.1 - 0.1 1.7 0.33 - 2.0
40 1.8 - - - - - 1.5 0.51 - -
50 1.2 - - - - - 1.0 0.24 - - 60 0.5 - - - - - 0.6 0.43 - -
70 0.3 - - - - - 0.3 0.15 - - 80 0.4 - - 0.1 - 0.1 0.3 0.38 - 1.34
90 0.2 - 3 0.1 - 0.1 - - - 2.93 100+ 0.5 - 1 - - - 0.4 0.10 - -
Shade Tolerant Species 2009 Trees
(Stems ha-
1)
Gain (stems ha-1)
2009-2011 Mortality (stems ha-1) Trees (Stems ha-
1)
Growth (cm yr-1)
Tree Mortality rate (%)
DBH Class (cm)
Before logging
Ingrowth # Tree harvested
Damaged trees
Undamaged trees
Total 2011 2009-2011
Natural Logging
10 183.8 35 - 8.2 3.0 11.2 170.1 0.12 0. 8 2.3
20 61.1 - - 1.5 0.7 2.2 61.7 0.18 0. 6 1.2
30 26.3 - - 0.4 0.2 0.6 25.1 0.18 0. 4 0. 8
40 13.6 - - 0.3 0.1 0.4 14.6 0.14 0. 4 1.1
50 5.8 - - - - - 6.0 0.07 - -
60 4.9 - - - 0.1 0.1 4.6 0.15 1.0 -
70 2.2 - - 0.2 - 0.2 2.4 0.09 - 0.4.7
80 0.9 - - - 0.1 0.1 0.9 0.08 0. 57 -
90 0.4 - - - - - 0.3 0.48 - -
100+ 0.7 - 1 - - - 0.7 0.11 - -
214
APPENDIX J CHANGES IN ABOVE-GROUND BIOMASS PER STEMS DIAMETER CLASS IN 10 1-HA PERMANENT PLOTS FOR
EACH SPECIES FUNCTIONAL GROUP TWO YEARS AFTER REDUCED-IMPACT LOGGING ON MONTS DE CRISTAL IN GABON.
Non-Pioneer Light Demander Species
2009 AGB (Mg ha-1)
Gain (Mg ha-1 yr-1) Loss (Mg ha-1 yr-1) Net Balance (Mg ha-1 yr-1)
DBH Class (cm)
Before logging Growth Ingrowth Total Logging Natural Total
10 62.1 0.15 3.4 3.55 0.017 0.003 0.020 3.53 20 86.2 0.16 - 0.16 0.012 0.007 0.019 0.14 30 145.0 0.12 - 0.12 0.004 0.003 0.007 0.11 40 182.8 0.14 - 0.14 0.007 0.006 0.013 0.13 50 150.0 0.06 - 0.06 - - - 0.06 60 209.9 0.14 - 0.14 - 0.015 0.015 0.13 70 149.8 0.04 - 0.04 0.035 0.021 0.056 -0.02 80 198.1 - - - - - - - 90 170.8 - - - - 0.028 0.028 -0.03 100+ 538.8 0.01 - 0.01 0.011 - 0.011 -
215
Pioneer Species
2009 AGB (Mg ha-1)
Gain (Mg ha-1 yr-1) Loss (Mg ha-1 yr-1) Net Balance (Mg ha-1 yr-1)
DBH Class (cm)
Before logging Growth Ingrowth Total Logging Natural Total
10 16.7 0.038 0.3 0.338 0.018 0.006 0.024 0.31 20 15.0 0.053 - 0.053 - - - 0.05 30 20.4 0.065 - 0.065 0.022 - 0.022 0.04 40 26.7 0.048 - 0.048 - - - 0.05 50 30.6 0.002 - 0.002 - - - - 60 17.4 0.009 - 0.009 - - - 0.01 70 16.9 0.027 - 0.027 - - - 0.03 80 26.2 0.060 - 0.060 0.105 - 0.105 -0.05 90 13.5 0.039 - 0.039 0.285 - 0.285 -0.25 100+ 69.7 0.008 - 0.008 - - - 0.01
Shade Tolerant Species
2009 AGB (Mg ha-1)
Gain (Mg ha-1 yr-1) Loss (Mg ha-1 yr-1) Net Balance (Mg ha-1 yr-1)
DBH Class (cm)
Before logging Growth Ingrowth Total Logging Natural Total
10 215.2 0.457 3.30 3.757 0.019 0.008 0.027 3.73 20 284.2 0.485 - 0.485 0.011 0.006 0.017 0.47 30 286.1 0.430 - 0.430 0.007 0.004 0.011 0.42 40 298.2 0.267 - 0.267 0.014 0.004 0.018 0.25 50 219.4 0.090 - 0.090 - - - 0.09 60 279.3 0.148 - 0.148 - 0.010 0.010 0.14 70 170.6 0.057 - 0.057 0.047 - 0.047 0.01 80 85.5 0.017 - 0.017 - 0.051 0.051 -0.03 90 45.1 0.009 - 0.009 - - - 0.01 100+ 164.9 0.086 - 0.086 - - - 0.09
216
LIST OF REFERENCES
Acevedo, M., Urban, D., Ablan, M., 1995. Transition and gap models of forest dynamics. Ecol. Appl. 5, 1040-1055.
Ahmad, S., Brodie, J.D., Sessions, J., 1999. Analysis of two alternative harvesting systems in Peninsular Malaysia: sensitivity analysis of costs, logging damage and buffers. Jour. Trop. For. Sci. 11, 809-821.
Andam, K. S., P. J. Ferraro, A. Pfaff, G. A. Sanchez-Azofeifa, and J. A. Robalino. 2008. Measuring the effectiveness of protected area networks in reducing deforestation. Proceedings of the National Academy of Sciences of the United States of America 105:16089-16094.
Angelsen, A. Brockhaus, M., Kanninen, M., Sills, E., Sunderlin, W. D. and Wertz-Kanounnikoff, S. (eds) 2009. Realising REDD+: National strategy and policy options. CIFOR, Bogor, Indonesia.
Applegate, G., Putz, F.E., Snook, L.K., 2004. Who pays for and who benefits from improved timber harvesting practices in the tropics? Lessons learned and information gaps. CIFOR, Bangor, Indonesia.
Asner, G.P., Knapp, D.E., Broadbent, E.N., Oliveira, P.J.C., Keller, M., Silva, J.N., 2005. Selective logging in the Brazilian Amazon. Science 310, 480-482.
Asner, G.P., Broadbent, E.N., Oliveira, P.J.C., Keller, M., Knapp, D.E., Silva, J.N., 2006. Condition and fate of logged forests in the Brazilian Amazon. PNAS 103, 12947-12950.
Asner, G.P., Mascaro, J., Muller-Landau, H.C., Vieilledent, G., Vaudry, R., Rasamoelina, M., Hall, J.S., van Breugel, M., 2011. A universal airborne LiDAR approach for tropical forest carbon mapping. DOI 10.1007/s00442-011-2165-z.
Ashton, M.S., Hall, J.S., 2011. The ecology, silviculture, and use of tropical wet forests with special emphasis on timber rich types. In S. Gunter et al. (eds) Silviculture in the Tropics, Tropical Forestry 8. DOI 10.10079/978-3-642-19986-8_12, Springer-Verlag Berlin Heidelberg 2011.
ATIBT, 2007. The requirement of a practical forest development plan for natural tropical African production forests. Application in the case of Central Africa. Volume I: Production Forests. Association Technique Internationale des Bois Tropicaux, Saint Mandé, France.
ATO-ITTO, 2003. Principles, Criteria and Indicators for the sustainable management of African natural tropical forests. African Timber Organization and International Tropical Timber Organization, ITTO Policy Development Series No. 14, Yokohama, Japan.
217
Auld, G., Gulbrandsen, L. H., McDermott, C. L., 2008. Certification schemes and the impacts on forests and forestry. Ann Rev Environ Res 33, 187-211.
Baccini, A., Laporte, N., Goetz, S.J., Sun, M., Dong, H., 2008. A first map of tropical Africa’s above-ground biomass derived from satellite imagery. Environmental Research Letters, doi:10.1088/1748-9326/3/4/045011.
Bacha, C.J.C., Rodriguez, L.C.E., 2007. Economic and social impacts of reduced impact logging in the Tapajós National Forest, Brazil: A case study. Ecol. Econo. 63, 70-77.
Barreto, P., Amaral, P., Vidal, E., Uhl, C., 1998. Costs and benefits of forest management for timber production in eastern Amazonia. Forest Ecol. Manage. 108, 9-26.
Bayol, N., Borie, J.M., 2004. Itineraires techniques d’amenagement des forets de productions en Afrique centrale. Bois et Forêts des Tropiques 281, 35-48.
Bedel, F., Durrieu de Madron, L., Dupuy, B., Favrichon, V., Maitre, H. F., Bar Hen, A., Narboni. P., 1998. Dynamique de croissance dans des peuplements exploités et éclaircis de forêt dense africaine. Le dispositif de M’Baïki en République Centrafricaine (1982 - 1995). CIRAD-Forêt, Montpellier, France.
Berry, N.J., Phillips, O.L., Lewis, S.L., Hill, J.K., Edwards, D.P., Tawatao, N. B., Ahmed, N., Magintan, D., Khen, C.V, Maryati, M., Ong, R.C, Hamer, K.C., 2010. The high value of logged tropical forests: lessons from northern Borneo. Bio. Cons. 19,985-997.
Blackman, A., Rivera, J., 2010. The evidence base for environmental and socioeconomic impacts of “sustainable” certification. Discussion paper, Resources for the future RFF DP 10-17. Washington DC, USA.
Blake, S., Strindberg, S., Boudjan, P., Makombo, C., Bila-Isia, I., Ilambu, O., Grossmann, F., Bene-Bene, L., de Semboli, B., Mbenzo, V., S’hwa, D., Bayogo, R., Williamson, L., Fay, M., Hart, J., Maisels, F., 2007. Forest elephant crisis in the Congo Basin. Public Library of Science Biology 5,1-9.
Blanc, L., Echard, M., Herault, B., Bonal, D., Marcon, E., Chave, J., Baraloto, C., 2009. Dynamics of aboveground carbon stocks in a selectively logged tropical forest. Ecol. App. 19, 1397-1404.
Boltz, F., Carter. D.R., Holmes, T.P., R. Pereira, Jr., 2001. Financial returns under uncertainty for conventional and reduced-impact logging in permanent production forests of the Brazilian Amazon. Ecol. Econo. 39, 387-398.
218
Boltz, F., Holmes, T.P., Carter, D.R., 2003. Economic and environmental impacts of conventional and reduced-impact logging in Tropical South America: a comparative review. Forest Policy and Economics 5, 69-81.
Borcard, D., Gillet, F., Legendre, P., 2011. Numerical ecology with R. DOI 10.1007/978-1-4419-7976-6 Springer New York Dordrecht London Heidelberg.
Boscolo, M, Vincent, J.R, 1998. Promoting better logging practices in tropical forests: A simulation analysis of alternative regulations. Policy Research Working Paper 1971. The World Bank Development Research Group.
Boscolo, M., Vincent, J., 2000. Promoting better logging practices in tropical forests: a simulation analysis of alternative regulations. Land Economics, pp.1-14.
Bossel, H., Krieger, H., 1991. Simulation model of natural tropical forest dynamics. Ecol. Model. 59, 37-71.
Broadbent, E.N, Asner, G.P., Peña-Claros, M., Palace, M., Soriano, M., Spatial partitioning of biomass and diversity in a lowland Bolivian forest: Linking field and remote sensing measurements. For. Ecol. Manage. 255, 2602-2616.
Brokaw, N.V.L., 1982. The definition of tree-fall gap and its effect on measures of forest dynamics. Biotropica 14, 158-160.
Brown, S., 1997. Estimating biomass and biomass change of tropical forests: a primer. FAO Forestry paper 134. Rome, Italy.
Brown, S. Pearson T., Moore N., Parveen A., Ambagis S., and Shoch D., 2005. Impact of selective logging on the carbon stock of tropical forests: Republic of Congo as a case study. Winrock International, Virginia. 21pp.
Burger, D., Hess, J., Lang, B., 2005. Forest Certification: An innovative instrument in the service of sustainable development? Deutsche Gesellschaft für Technische Zusammenarbeit. Eschborn 2005.
Cannon, C. H., Peart, D. R., Leighton, M., 1998. Tree species diversity in commercially logged Bornean rainforest. Science 281, 1366-1368.
Cashore, B., Gale, F., Meidinger, E., Newsom, D., 2006. Confronting sustainability: Forest certification in developing and transitioning countries. Yale F&ES Publication Series, Report Number 8. New Haven, Connecticut USA.
Caswell, H., 2001. Matrix population models, Construction, Analysis and Interpretation. Sinauer Associates, Inc., Sunderland, Mass, 722 pp.
CEB (Compagnie Equatoriale de Bois), 2007. Résumé publique du plan d’aménagement. Precious Woods and TERRA. 31 pages.
219
Cerutti, P.O., Nasi, R., Tacconi, L., 2008. Sustainable forest management in Cameroon needs more than approved forest management plans. Ecol. Society 13(2): 36. [online] RUL: http://www.ecologyandsociety.org/vol13/iss2/art36/
Cerutti, P.O., Assembe-Mvondo, S., German, L., Putzel, L., 2011. Is China unique? Exploring the behaviour of Chinese and European firms in the Cameroonian logging sector. Intern. For. Rev. 13, 23-34.
Chave, J., 1999. Study of structural, successional and spatial patterns in tropical forests using TROLL, a spatially explicit forest model. Ecol. Model. 124, 233-254.
Chave, J., Andalo, C., Brown, S., Cairns, M.A., Chambers, J.Q., Eamus, D., Fölster, H., Fromard, F., Higuchi, N., Kira, T., Lescure, J.P., Nelson, B.W., Ogawa, H., Puig, H., Riéra, B., Yamakura, T., 2005. Tree allometry and improved estimation of carbon stocks and balance in tropical forests. Oecologia 145, 87-99.
Chikumbo, O., Steward, G., 2007. A stand basal area model for plantation grown New Zealand kauri. Ecol. Model. 209, 367-376.
Clark, D. A., Clark. D. B., 1992. Life history diversity of canopy and emergent trees in a Neotropical rain forest. Ecological Monographs 62, 315-344.
Clark, J.C., Poulsen, J.R., Malonga, R., P. W. Elkan Jr., 2009. Logging concessions can extend the conservation estate for Central African tropical forests. Cons. Bio. 23, 1281-1293.
Clark, M. R., Kozar, J.S., 2011. Comparing Sustainable Forest Management Certifications Standards: A Meta-analysis. Ecology and Society, 16(1):3.
Colchester, M. 2006. Reflection on the social dimension of verification in FLEGT processes: Issues, risks and challenges. VERIFOR Viewpoint no.1, ODI, London (www.verifor.org).
Cole, T.G, Ewel, J.J., 2006. Allometric equations for four valuable tropical tree species. Forest Ecol. Manage. 299, 351-360.
Colwell, R. K., 2006. EstimateS: Statistical estimation of species richness and shared species from samples. Version 8. Persistent URL <purl.ocloc.org/estimates>.
Compagnie Equatoriale de Bois. Precious Wood Gabon [cited 23 September 2010]. Available from http://www.preciouswoods.com/.
Cotula, L., Mayers, M. 2009. Tenure in REDD: Start-point or Afterthought? London, International Institute for Environment and Development. 67 p.
Cropper, Jr. W.P. and D. DiResta. 1999. Simulation of a Biscayne Bay, Florida commercial sponge population: Effects of harvesting following Hurricane Andrew. Ecol. Model. 118, 1-15.
220
Crouse, D.T., Crowder, L.B., Caswell, H., 1987. A stage-based population model for loggerhead sea-turtles and implications for conservation. Ecology 68, 1412-1423.
Dagang, A.A., Richter, F., Hahn-Schilling., Manggil, P., 2002. Financial and economic analyzes of conventional and reduced impact harvesting systems in Sarawak. In: Enters, T., Durst, P.B., Applegate, G.B, Kho, Peter C.S., Man, G., 2002. Applying reduced impact logging to advance sustainable forest management. FAO 2002 International Conference Proceedings, Bangkok, Thailand, 526 p.
De Chatelperron, G., Commercon, R., 1996. Mise en exploitation du dispositif de recherche en foret naturelle dans les forets de Boukoko et La Lole en Republique Centafricaine Projet F.A.C.A.R.R.F.C.T.F.T. In : Durrieu de Madron, L., Forni, E., Mekok, M., 1998. Les techniques d’exploitation à faible impact en forêt dense humide camerounaise. Document No. 17 Série FORAFRI. CIRAD-Forêt, Montpellier, France.
de Wasseige, C., Devers, D., de Marcken, P., Eba’a Atyi, R., Nasi, R., Mayaux, Ph., 2009. The Forests of the Congo Basin - State of the Forest 2008, Luxembourg: Publications Office of the European Union, doi: 10.2788/32259.
Denslow, J. S., 1987. Tropical rainforest gaps and tree species diversity. Annual Review of Ecology and Systematics 18, 431-451.
Djomo, A.N., A. Ibrahima, J. Saborowski, and G. Gravenhorst., 2010. Allometric equations for biomass estimations in Cameroon and pan moist tropical equations including biomass data from Africa. Forest Ecol. Manage. 260, 1874-1885.
Doucet, J.L. 2003. L’alliance délicate de la gestion forestière et de la biodiversité dans les forêts du centre du Gabon. Thèse de doctorat, Faculté Universitaire des Sciences Agronomiques, B-5030 Gembloux, 323 p.
Dupuy, B., Brevet, R., Doumbia, F., Diahuissié, A., 1993. Sylviculture et productivité de la forêt dense humide en Côte d’Ivoire: les périmètres expérimentaux d’Irobo et de Mopri. Typescript report, CIRAD Forêt, Abidjan, Côte d’Ivoire.
Durrieu de Madron, L., Forni, E., Mekok, M., 1998. Les techniques d’exploitation à faible impact en forêt dense humide camerounaise. Document No. 17 Série FORAFRI. CIRAD-Forêt, Montpellier, France.
Durrieu de Madron, L., Fontez, B., Dipapoundji, B., 2000. Dégâts d’exploitation et de débardage en fonction de l’intensité d’exploitation en forêt dense humide d’Afrique Centrale. Bois et Forêts des Tropiques 264, 57-60.
Durrieu de Madron, L., Bauwens, S., Giraud, A., Hubert, D., Billand, A., 2011. Estimation de l’impact de differents modes d’exploitation forestiere sur les stocks de carbone en Afrique centrale. Bois et Forêts des Tropiques 308, 75-86.
221
Dykstra, D.P., 2002. Reduced impact logging: concepts and issues. in: Enters, T., Durst, P.B., Applegate, G.B., Kho, P.C.S., Man, G., 2002. Applying reduced impact logging to advance sustainable forest management. FAO 2002 International Conference Proceedings, Bangkok, Thailand.
Dwiprabowo, H., Grulois, S., Sist, S., Kartawinata, K., 2002. Cost-benefit analysis of reduced-impact logging in a lowland Dipterocarp forest of Malinau, East Kalimantan. In Forest, Science and Sustainability: The Bulungan model forest. ITTO Technical report.
Dykstra, D.P., Heinrich, R., 1996. FAO Model code of forest harvesting practice. FAO, Rome.
Dykstra, D.P., 2002. Reduced impact logging: concepts and issues. In: Enters, T., Durst, P.B., Applegate, G.B., Kho, P.C.S., Man, G., 2002. Applying reduced impact logging to advance sustainable forest management. FAO 2002 International Conference Proceedings, Bangkok, Thailand, 526 p.
Eba’a Atyi. R. 2006. Forest certification in Gabon. In Cashore, B., Gale, F., Meidinger, E., Newsom, D., 2006. Confronting sustainability: Forest certification in developing and transitioning countries. Yale F&ES Publication Series, Report Number 8. New Haven, Connecticut USA.
Ebeling, J., Yasué, M., 2009. The effectiveness of market-based conservation in the tropics: forest certification in Ecuador and Bolivia. J. Env. Manage. 90, 1145-1153
Elias, Applegate, G., Kartawinata, K., Machfudh, Klassen, A., 2001. Reduced-impact logging guidelines for Indonesia. CIFOR, Jakarta, Indonesia.
Ezzine de Blas, D., Ruiz Perez, M., 2008. Prospect for reduced impact logging in Central African logging concessions. Forest Ecol. Manage. 256, 1509-1516.
FAO, 2003. Regional Code of Practice for Reduced-Impact Harvesting in Tropical Moist Forests of West and Central Africa. Food and Agriculture Organization of the United Nations, Rome, Italy.
FAO, 2005. Global forest resource assessment 2005. Progress towards sustainable forest management. FAO forest paper 174. Rome, Italy.
FAO, 2006. Global Forest Resources Assessment 2005. Progress towards sustainable forest management. FAO Forestry Paper No. 147. Rome.
FAO, 2010. Global forest resources assessments 2010. Main report. FAO Forestry Paper 163. Rome, Italy.
FAO, 2011. State of the world’s forest 2011. FAO, Rome, Italy.
222
FCPF (Forest Carbon Partnership Facility) R-PIN (Readiness Plan Idea Note). 2008. Gabon. http://www.forestcarbonpartnership.org/fcp/GA (accessed 26 August 2010).
Fisher, B., Edwards, D.P., Giam, X., Wilcove, D.S., 2011. The high costs of conserving Southeast Asia’s lowland ranforests. Front Ecol. Environ 2011; doi: 10.1890/100079.
Forest Certification System. Certification public report. Forest Management Certification. [Reported 17 December 2008; cited 16 Feburary 2012]. Available from http://info.fsc.org/servlet/.
Forest Stewardship Council (FSC) 2004. FSC Principles and Criteria for forest stewardship. Reference Code FSC-STD-01-001. Forest Stewardship Council A.C.
Forest Stewardship Council (FSC 2011). www.fsc-watch.org/archives/2011/06/20/CIB___FSC_s_great_Af.
Forni, E., 1994. Etude de l’exploitation – Bilan de l’exploitation de la vente de coupe 1112. Rapport technique A.P.I Dimako, 12p. In : Durrieu de Madron, L., Forni, E., Mekok, M., 1998. Les techniques d’exploitation à faible impact en forêt dense humide camerounaise. Document No. 17 Série FORAFRI. CIRAD-Forêt, Montpellier, France.
Forshed, O., Udarbe, T., Karlson, A., Falck, J., 2006. Initial impact of supervised logging and pre-logging climber cutting compared with conventional logging in a dipterocarp rainforest in Sabah, Malaysia. Forest Ecol. Manage. 221, 233-240.
Fortini, L.B., Zarin, D.J., 2011. Population dynamics and management of Amazon tidal floodplain forests: links to the past, present and future. Forest Ecology and Management 261:551-561.
Fredericksen, T.S., Putz, F.E., 2003. Silvicultural intensification for tropical forest conservation. Bio. Cons.12, 1445-1453.
Fuhr, M., Delègue, M.A., Nasi, R., Minkoué, J.M., 1998. Dynamique et croissance de l’Okoumé en zone côtière du Gabon. Document No.16 Série FORAFRI. CIRAD-Forêt, Montpellier, France.
Gan, B.K, Amir Abdul Nasir, S., Zulkifli, A., 2006. The Logfisher - Its development and application in a New Ground-Based Reduced-Impact Logging System in Peninsular Malaysia. Proceeding of the ITTO – MoF Regional Workshop on RIL Implementation in Indonesia with reference to Asia-Pacific Region: Review and Experience, Bongor, Indonesia, 137-245.
Gerwing, J.J., Johns, J.S., Vidal, E., 1996. Reducing waste during logging and log processing: forest conservation in eastern Amazonia. Unasylva 187, 17-25.
223
Ghazoul, J., Sheil, G., 2010. Tropical rain forest ecology, diversity, and conservation. Oxford University Press Inc., New York.
Gordon, E., Ham, C., Eba’a Atyi, R., Mwima, P.M., Biryahwaho, B., Njovu, F.C., Eilu, G., Cashore, B., Gombya-Ssembajjwe, W., 2006. Regional overview: Forest certification in Sub-Saharan Africa. In Cashore, B., Gale, F., Meidinger, E., Newsom, D., 2006. Confronting sustainability: Forest certification in developing and transitioning countries. Yale F&ES Publication Series, Report Number 8. New Haven, Connecticut USA.
Gourlet-Fleury, S., Houillier, F., 2000. Modelling diameter increment in a lowland evergreen rainforest in French Guuiana. Forest Ecol. Manage. 131, 269-289.
Gourlet-Fleury, S., Cornu, G., Jésel, S., Dessard, H., Jourget, J.G., Blanc, L., Picard, N., 2005. Using models to predict recovery and assess tree species vulnerability in logged tropical forests: a case study from French Guiana. Forest Ecol. Manage. 209, 69-85.
Gullison, R. E., Frumhoff, P. C., Canadell, J. G., Field, C. B., Nepstad, D. C., Hayhoe, K., Avissar, R., Curran, L. M., Friedlingstein, P., Jones, C. D., Nobre, C., 2007. Tropical Forests and Climate Policy. Science 316, 985-986.
Hall, J.S., 2002. The role of habitat specialization in the regeneration of Entandrophragma spp: implicatiosn for the maintenance of species diversity and forest management in Central Africa. Ph.D Dissertation,Yale University, New Haven, Connecticut, USA.
Hall, J.S., Medjibe, V., Berlyn, G.P., Ashton, P.M.S., 2003. Seedling growth of three co-occurring Entandrophragma species (Meliaceae) under simulated light environments: implications for forest management in central Africa. Forest Ecol. Manage. 179, 135-144.
Hall, J.S., Harris, D.J., Medjibe, V., Ashton, P. Mark. S., 2003. The effects of selective logging on forest structure and tree species composition in a Central African forest: implications for management of conservation areas. Forest Ecol. Manage. 183, 249-264.
Hall, J.S., 2008. Seed and seedling survival of African mahogany (Entandrophragma spp.) in the Central African Republic: Implications for forest management. Forest Ecol. Manage. 255, 292-299.
Hall, J.S., 2011. Natural forest silviculture for Central African Meliaceae. In S. Gunter et al. (eds) Silviculture in the Tropics, Tropical Forestry 8. DOI 10.10079/978-3-642-19986-8_12, Springer-Verlag Berlin Heidelberg 2011.
Halme, P.T., Toivanen, T., Honkanen, M., Kotiaho, J.S., Monkkonen, M., Timonen, J., 2010. Flawed meta-analysis of biodiversity effects of forest management. Cons. Bio. Doi: 10.1111/j.1523-1739.2010.01542.x.
224
Hart, T.B., 2001. Forest dynamics in the Ituri basin (DRC Congo). In Weber, W., White, A., Vedder, L.J.T, Naughton-Irenes, L. (eds) African rainforest ecology and conservation: An interdisciplinary perspective. Yale University Press, New Haven, CT, pp 154-164.
Hawthorne, W. D. 1995. Ecological profiles of Ghanaian forest trees. Tropical Forestry Papers 29. Oxford Forestry Institute, Oxford.
Hawthorne, W. D., Sheil, D., Agyeman, V.K., Abu Juam, M., Marshall, C.A.M., 2012. Logging scars in Ghanian high forest: towards improved models for sustainble production. Forest Ecol. Manage. 271, 27-36.
Healey, J.R., Price, C., Ray, J., 2000. The cost of carbon retention by reduced impact logging. Forest Ecol. Manage. 139, 237-255.
Heffner, R. A., Butler, M. J., Reilly, C. K. 1996. Pseudoreplication revisited. Ecology 77, 255-2562.
Henry, M., Besnard. A., Asante W.A., Eshun, J., Abu-Bredu, S., Valentini, R., Bernoux, M., Saint-Andre, L., Woody density, phytomass variations within and among trees, and allometric equations in tropical rainforest of Africa. Forest Ecol. Manage. 260, 1375-1388.
Holmes, T.P, Blates, G.M., Zweede F.C., Perrera, R., Barretto. P., Boltz .F., Bauch, R., 2002. Financial and ecological indicators of reduced-impact logging performance in eastern Amazon. Forest Ecol. Manage. 163, 93-110.
Houghton, R. A., Hall, F., Goetz, S. J., 2009. Importance of biomass in the global carbon cycle, Jour. Geophy. Res. 114, G00E03, doi:10.1029/2009JG000935.
Hulbert, S.H., 1984. Pseudoreplication and the design of ecological field experiments. Ecol. Mono. 54, 187-211.
Hurlbert, S.H., 2004. On misinterpretations of pseudoreplication and related matters: a reply to Oksanen. Oikos 104, 591-597.
Jackson, S.M., Fredericksen, T.S., Malcolm, J.R., 2002. Area disturbed and residual stand damage following logging in a Bolivian tropical forest. Forest Ecol. Manage. 166, 271-283.
Jennings, S.B., Brown, N.D., Sheil, D., 1999. Assessing forest canopies and understory illumination: canopy closure, canopy cover and other measures. Forestry 72, 59-73
Johns, J. S., Barreto, P., Uhl, C., 1996. Logging damage during planned and unplanned logging operations in the eastern Amazon. Forest Ecol. Manage. 89, 59-77.
225
Jonkers, W.B.J., 2000. Logging, damage and efficiency: a study of the feasibility of the reduced impact logging in Cameroon. Tropenbos Cameroon Report 00-3, Tropenbos-Cameroon Programme, Kribi, Cameroon.
Kammesheidt, L., Kholer, P., Huth, A., 2001. Sustainable timber harvesting in Venezuela: a modelling approach. J. Appl. Ecol. 38, 756-770.
Kant, S. S., 2004. Economics of sustainable forest management. Forest Pol. Econo. 6, 197-203.
Kariuki, M., Kooyman, R.M, Smith, R.G.B., Wardell-Johnson, Vanclay, J.K., 2006. Regeneration changes in tree species abundance, diversity and structure in logged and unlogged subtropical rainforest over a 36-year period. Forest Ecol. Manage. 236,162-176.
Karsenty, A., Gourlet-Fleury, S., 2006. Assessing sustainability of logging practices in the Congo Basin’s managed forests: the issue of commercial species recovery. Ecology and Society 11: 26. [online] URL: http://www.ecologyandsociety.org/vol11/iss1/art26/
Karsenty, A. 2008. The architecture of proposed REDD schemes after Bali: facing critical choices. International forestry review 10:1-25.
Karsenty, A., Assembe, S., 2011, Les régimes fonciers et la mise en œuvre de la REDD+ en Afrique Centrale. Revues des questions foncières 2, 105-130.
Kohler, P., Huth, A., 1998. The effects of tree species grouping in tropical rainforest modelling: Simulations with the individual-based model FORMIND. Ecol. Model. 109, 301-321.
Kohler, P., Ditzer, T., Ong, R.C., Huth, A., 2001. Comparison of measured and modelled growth on permanent plots in Sabah's rain forest. Forest Ecol. Manage. 144, 101-111.
Lancaster, P., 1960. Investigations into methods of obtaining and encouraging the growth of natural regeneration. Nigerian Forest Information Bulletin 8.
Laporte, N.T., Stabach, J.A., Grosch, R., Lin, T.S., Goetz, S.J., 2007. Expansion of industrial logging in Central Africa. Science 316:1451.
Leigh, E.G.Jr., 2008. Tropical forest ecology: sterile or virgin for theoreticians? In Carson, W.P., and Schnitzer S.A. 2008. Tropical Forest Community Ecology. John Wiley & Sons Ltd, West Sussex, UK.
Leonard, G., Richard, A., 1993. Le Gabon. Institut Pedogogique National, 287p. In: Drouineau, S., Nasi, R., Legault, F., Cazet, M., 1999: L’aménagement forestier au Gabon: historique, bilan, perspectives. Document No.19 Série FORAFRI. CIRAD-Forêt, Montpellier, France.
226
Leslie, A.D, 2004. The impacts and mechanics of certification. The International Forestry Review, 6, 30-39.
Lewis, S. L., Lopez-Gonzalez, G., Sonké, B., Affum-Baffoe, K., Baker, T. R., Ojo, L. O., Phillips, O. L., Reitsma, J. M., White, L., Comiskey, J. A., Djuikouo, M. N. K., Ewango, C. E. N., Feldpausch, T.R., Hamilton, A. C., Gloor, M., Hart, T., Hladik, A., Lloyd, J., Lovett, J.C., Makana, J. R., Malhi, Y., Mbago, F. M., Ndangalasi, H. J., Peacock, J., Peh, K. S.-H., Sheil, D., Sunderland, T., Swaine, M.D., Taplin, J., Taylor, D., Thomas, S.C., Votere, R., Wöll, H., 2009. Increasing carbon storage in intact African tropical forests. Nature 457, 1003-1007.
Luckert, M.K., Williamson, T., 2005. Should sustained yield be part of sustainable forest management? Canadian Journal Forest Research. 35, 356-364.
MacGregor, W. D., 1934. Silviculture of the mixed deciduous forests of Nigeria with special references to the south-western provinces. Oxford Forest Memoirs 18,1-108.
Macpherson, A. et al., 2010. A Model for comparing reduced impact logging with conventional logging for an Eastern Amazonian Forest. Forest Ecol. Manage. 260, 2002-2011.
Makana, J. R., Thomas, S. C., 2005. Effects of lights gaps and litter remoal on the seedling performance of six African timber species. Biotropica 37, 227-237.
Malcolm, J. R., Ray J. C., 2000. Influence of timber extraction routes on central African small-mammal communities, forest structure, and tree diversity. Cons. Bio.14, 1623-1638.
Martin, D., Chatelin Y., Collinet, J., Guichard E., Sala, G., 1981. Les sols du Gabon : pédogenèse, répartition et aptitudes : cartes à 1:2 000 000 - Paris (FRA) : ORSTOM, Office de la recherche scientifique et technique outre mer. 67 p. : 2 cartes 1/2 000 000 - (Notice Explicative, No 92).
Mattsson-Marn, H., Vel, E., de Jongh, O., Hui, D.C.K., 1981. Planning and cost studies in harvesting in the mixed dipterocarp forest of Sarawak: Part I. Based on maps derived from ground survey. Food and Agriculture Organization of the United Nations, FO: MAL/76/008, Field Document No. 7. 76pp. In: Schwab, O., Pulkki, R., Bull, G.Q., 2001. Reduced impact logging in tropical forests: Literature synthesis, analysis and prototype statistical framework. Forest production division, working paper series FOP/08. FAO, Rome, Italy.
Mattsson-Marn, H.G., Jonkers, W., 1981. Logging damage in tropical high forest. Food and Agriculture Organization of the United Nations, Project FO: MAL/76/008, Working Paper No. 5. 15pp.
227
Mazzei, L., Sist, P., Ruschel, A., Putz, F.E., Marco, P., Peña, W., Ferreira, J.E.R., 2010. Above-ground biomass dynamics after reduced-impact logging in the Eastern Amazon. Forest. Ecol. Manage. 259, 367-373.
Medjibe V.P., Putz, F.E., Starkey, P.M, Ndouna, A. A., Memiaghé, H. R., 2011a. Impacts of selective logging on above-ground forest biomass in the Monts de Cristal in Gabon. Forest Ecol. Manage. 262, 1799-1806.
Medjibe, V.P., Putz, F.E., 2011b. In review, Cost comparison of reduced-impact and conventional logging in the tropics.
Minnemeyer, S.T., Walker, T., Collomb, J.G, Cotton, L., Bryant, D., 2002. An analysis of access to central Africa’s rainforests. World Resources Institute Report, Washington DC, USA.
Miyata, E.S., 1980. Determining fixed and operating costs of logging equipment. North Central Forest Experiment Station Forest Service, U.S. Department of Agriculture, St Paul, Minnesota, USA.
Nasi, R., Cassagne, B., Billand, A., 2006. Forest management in Central Africa : where are we? Int. For. Rev. 8, 14-20.
Nasi, R., Mayaux, P., Devers, D., Bayol, N., Eba’a Atyi, R., Mugnier, A., Cassagne, B., Billand, A., Sonwa, D., 2008. A first look at carbon stocks and their variation in Congo Basin forests. In de Wasseige, C., Devers, D., de Marcken, P., Eba’a Atyi, R., Nasi, R., Mayaux, Ph., 2009. The Forests of the Congo Basin - State of the Forest 2008, Luxembourg: Publications Office of the European Union, doi: 10.2788/32259.
Nasi, R., Billand, A., Vanvliet, N., 2012. Managing for timber and biodiversity in the Congo Basin. Forest Ecol. Manage. 268, 103-111.
Nwoboshi, L. C., 1987. Regeneration success of natural management, enrichment planting, and plantations of native species in West Africa. In F. Mergen and J. R. Vincent, eds. Natural Management of Tropical Moist Forests: Silviculture and Management Prospects of Sustained Utilization. Yale University Press, New Haven.
Oksanen, L., 2001. Logic of experiment in ecology: is pseudoreplication a pseudoissue? Oikos 94, 27-38.
Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, P., Minchin, P.R., O’Hara, R.B., Simpson,G.L., Solymos, P., Stevens,M.H.H, Wagner, H., 2011. Package “vegan”. Community ecology package Version 2.0-0. http://cran.r-project.org, http://vegan.r-forge.r-project.org/
228
Paillet, Y., Bergès, L., Hjältén, J., et al., 2010. Biodiversity Differences between Managed and Unmanaged Forests: Meta-Analysis of Species Richness in Europe. Cons. Bio. 24, 101-112.
Pattanayak, S. K., S. Wunder, and P. J. Ferraro. 2010. Show Me the Money: Do Payments Supply Environmental Services in Developing Countries? 2010. Review of Environmental Economics and Policy 4: 254–274.
Parker, C., Mitchell, A., Trivedi, M., Mardas, N., 2009. The Little REDD+ Book. Global Canopy Programme, www.globalcanopy.org.
Pastor, J., Post, W.M., 1985. Development of a linked forest productivity-soil process model. Environmental sciences division. Carbon dioxide research division, office of research, Publication No. 2455.
Pearce, D., Putz, F.E., Vanclay, J.K., 2003. Sustainable forestry in the tropics: panacea or folly? Forest Ecol. Manage. 172, 229-247.
Peña-Claros, M., Fredericksen, T.S., Alarcón, Blate, G.M., Choque, U., Leaño, C., Licona, J.C., Mostacedo, B., Pariona, W., Villegas, Z., Putz, F.E., 2008. Beyond reduced-impact logging: Silvicultural treatments to increase growth rates of tropical trees. Forest Ecol. Manage. 256, 1457-1468.
Petrucci, Y., Tandeau de Marsac, G., 1994. Evolution du peuplement adulte et de la régénération acquise après interventions sylvicoles. Ministère des Eaux, Forêts, Chasses, Pêches, Tourisme, et de l’Environement, République Centrafricaine. 56 pp.
Phillips, P.D., de Azevedo, C.P., Degen, B., Thompson, I.S., Silva, J.N.M., van Gardingen, P.R., 2004. An individual-based spatially explicit simulation model for strategic forest management planning in the eastern Amazon. Ecol. Model. 173, 335-354.
Picard, N., Gourlet-Fleury, S., Forni, E., 2012. Estimating damage from selective logging and implications for tropical forest management. Can. J. For. Res. 42, 1-9.
Pinard, M.A., Putz, F.E., Tay, J., Sullivan, T., 1995. Creating timber harvest guidelines for a reduced impact logging project in Malaysia. Jour. For. 93, 41-45.
Pinard, M.A., Putz, F.E., 1996. Retaining forest biomass by reducing logging damage. Biotropica 28, 278-295.
Pinard, M.A., Cropper, W.P., 2000. Simulated effects of logging on carbon storage in dipterocarp forest. Jour. App. Ecol. 37,267-283.
Post, W.M., Pastor, J., 1996. Linkages-an individual-based forest ecosystem model. Climatic Change 34, 253-261
229
Poulsen, J.R., Clark, C.J., Mavah, G., Elkan, P.W., 2009. Bushmeat supply and consumption in a tropical logging concession in northern Congo. Cons. Biol. 23, 1597-1608.
Putz, F.E., Viana, V., 1996. Biological challenges for certification of tropical timber. Biotropica 28, 323-330.
Putz, F. E., Dykstra, D.P., Heinrich, R., 2000. Why poor logging practices persist in the tropics. Cons. Bio. 14, 951-956.
Putz, F. E., Blate, G.M., Redford, K.H., Fimbel, R., Robinson, J., 2001. Tropical forest management and conservation of biodiversity: an overview. Cons. Biol. 15, 7-20.
Putz, F. E., Sist, P., Fredericksen, T., Dykstra, D., 2008a Reduced-impact logging: challenges and opportunities. Forest Ecol. Manage. 256, 1427-1433.
Putz, F. E., Zuidema, P. A., Pinard, M. A., Boot, R. G.A., Sayer, J. A., Sheil, D., Sist, P., Vanclay, E. J. K., 2008b.Tropical forest management for carbon retention. PLoS Biology 6, 1368-1369.
Putz, F. E., Redford, K. H., 2009 Dangers of carbon-based conservation. Global Environnemental Change 19, 400-401.
Putz, F.E., Nasi, R., 2009. Carbon benefits from avoiding and repairing forest degradation. In Angelsen, A. with Brockhaus, M., Kanninen, M., Sills, E., Sunderlin, W. D. and Wertz-Kanounnikoff, S. (eds) 2009. Realising REDD+: National strategy and policy options. CIFOR, Bogor, Indonesia.
Putz, F.E., Zuidema, P.A., Synnott, T., Pena-Claros, M., Pinard, M.A., Sheil, D., Vanclay, J.K., Sist, P., Gourlet-Fleury, S., Griscom, B., Palmer, J., Zagt, R., 2012. In press. Sustaining conservation values in selectively logged tropical forests: The attained and the attainable. Conservation Letters, Policy Perspective.
Python Software Foundation 2001-2010. Available from http://www.python.org.
R Development Core Team, 2010. R: a language and environment for statistical computing. Version 2.11.1. R Foundation for Statistical Computing, Vienna, Austria. Available from http:// www.R-project.org.
Rahim Abdul, N., 2002. A model project for cost analysis to achieve sustainable forest management. Vol. II Main Report. Forest Research Institute Malaysia and International Tropical Timber Organization. p288.
Rahim Abdul, A.S., Mohd Shahwahid, H.O, Zariyawati, M.A., 2009. A comparison analysis of logging cost between conventional and reduced-impact logging practices. Int. Jour. Econo. Manage. 3, 354-366.
230
Richards, P.W., 1996. The tropical rain forest: an ecology study. Cambridge University Press, Cambridge.
Rockwell, C. A. 2007. Future crop tree damage in a certified community forest in southwestern Amazonia. Forest Ecol. Manage. 242, 108-118.
Ruiz Perez M., Ezzine de Blas, D., Nasi, R., Sayer, J. A., Sassen, M., Angoue, C., Gami, N., Ndoye, O., Ngono, G., Nguinguiri, J. C., Nzala, D., Toirambe, B., Yalibanda, Y., 2005. Logging in the Congo Basin: A multi-country characterization of timber companies. Forest Ecol. Manage. 241, 231-236.
Ruslandi, Venter, O., Putz, F.E., 2011. Over-estimating the costs of conservation in Southeast Asia. Front. Ecol. Environ. 9, 542-544.
Ruslandi, Halperin, J., Putz, F.E., 2012. Effects of felling gap proximity on residual tree mortality and growth in a dipterocarp forest in East Kalimantan, Indonesia. J. Trop. Sc. 24 (1).
Rutishauser, E., Wagner, F., Herault, B., Nicolini, E.A., Blanc, L., 2010. Contrasting above-ground biomass balance in a Neotropical rain forest. J.Veg. Science. 21, 672-682.
Sanderson, E.W., 2006. How many animals do we want to save? The many ways of setting population target levels for conservation. BioScience 56, 911-922.
Schulze, M., Lentini, M., Zweede, J.C., 2009. Training needs for RIL and improved forest management. In Angelsen, A., Brockhaus, M., Kanninen, M., Sills, E., Sunderlin, W. D., Wertz-Kanounnikoff, S., (Eds) 2009. Realising REDD+: National strategy and policy options. CIFOR, Bogor, Indonesia.
Schwab, O., Pulkki, R., Bull, G.Q., 2001. Reduced impact logging in tropical forests: Literature synthesis, analysis and prototype statistical framework. Forest production division, working paper series FOP/08. FAO, Rome, Italy.
Sheil, D., Burselem, D.F.R.P, Alder, D., 1995. The interpretation and misinterpretation of mortality rate measures. J. Ecol. 83, 331-333.
Sheil, D., Nasi, R., Johnson, B., 2004. Ecological criteria and indicators for tropical forest landscapes: challenges in the search for progress. Ecol. Soc. 9(1):7. [online] URL: http://www.ecologyandsociety.org/vol9/iss1/art7
Sheil, D., Meijaard, E., 2010, Purity and prejudice: deluding ourselves about biodiversity conservation. Biotropica 42, 566–568.
Sist, P., Nolan, T., Bertault, J.G., Dykstra, D., 1998. Harvesting intensity versus sustainability in Indonesia. Forest Ecol. Manage. 108, 251-260.
231
Sist, P., 2000. Reduced impact logging in the tropics: objectives, principles and impacts. Int. Forest. Rev. 2, 3-10.
Sist, P., Ferreira, F.N., 2007. Sustainability of reduced-impact logging in the Eastern Amazon. Forest Ecol. Manage. 243, 199-209.
Slik, J.W. F, Shin-Ichiro Aiba, Brearley, F. Q., Cannon, C. H., Forshed, O., Kitayama, K., Nagamasu, H., Nilus, R., Payne, J., Paoli, G., Poulsen, A. D., Raes, N., Sheil, D., Sidiyasa, K., Suzuki, E., Johan L.C.H., van Valkenburg., 2009. Environmental correlates of tree biomass, basal area, wood specific gravity and stem density gradients in Borneo’s tropical forests. Global Ecology and Biogeography DOI: 10.1111/j.1466-8238.2009.00489.x www.blackwellpublishing.com/geb.
Sundberg, U., Silversides, C.R., 1988. Operational efficiency in forestry. Volume 1: Analysis. Kluwer Academic Publishers, Dordrecht, the Netherlands. Quoted in: van der Hout, P., 1999. Reduced impact logging in the tropical rain forest of Guyana: ecological, economic and silvicultural consequences. Tropenbos-Guyana Series 6. Tropenbos-Guyana Programme, Georgetown, Guyana.
Sunderland, T., Walters, G., Issembe, Y., 2004. A preliminary vegetation assessment of the Mbe National Park, Monts de Cristal, Gabon. Smithsonian Institution Report. p51.
Système de Certification FSC. Rapport publique de certification. Certification de gestion forestière. [Reported 29 December 2011; cited 16 February 2012]. Available from http://info.fsc.org/servlet/servlet/.
Tadoum, M., Seya Makonga, V.K., Boundzanga, G.C., Bouyer, O., Hamel, O., Creighton, G.K. 2010. International negociations on the future climate regime beyond 2012: achievements from Copenhagen to Cancun and benefits to the forests of the Congo Basin. In de Wasseige, C., de Marcken, P., Bayol, N., Hiol Hiol, F., Mayaux, Ph., Desclee, B., Nasi, R., Billand, A., Defourny, P., Eba’a Atyi, R., 2010. The Forests of the Congo Basin - State of the Forest 2010, Luxembourg: Publications Office of the European Union, doi: 10.2788/48830.
Tay, 1999. Economic assessment of reduced-impact logging in Sabah, Malaysia. Ph.D. Dissertation, University of Wales, UK.
Tay, J., Healey, J., Price, C., 2002. Fianacial assessment of reduced-impact logging techniques in Sabah, Malaysia. Enters, T., Durst, P.B., Applegate, G.B., Kho, P.C.S., Man, G., 2002. Applying reduced impact logging to advance sustainable forest management. FAO 2002 International Conference Proceedings, Bangkok, Thailand.
Ter Braak, C.J.F, 1986. Canonical correspondence analysis: a new eigenvector technique for multivariate direct gradient analysis. Ecology 67, 1167-1179.
232
Tikina, A.T.V., Innes, J.I.L., 2008. A framework for assessing the effectiveness of forest certification. Canadian Journal of Forest Research 38,1357-1365.
van der Hout, P., 1999. Reduced impact logging in the tropical rain forest of Guyana: ecological, economic and silvicultural consequences. Tropenbos-Guyana Series 6. Tropenbos-Guyana Programme, Georgetown, Guyana.
van der Hout, P., Zagt, R.J., 2005. Responses of tree populations and forest composition to selective logging in Guyana. in Arets, E.J.M.M., 2005. Long-term responses of populations and communities of trees to selective logging in tropical rain forests in Guyana. Tropenbos-Guyana Series 13. Tropenbos-Guyana Programme, Georgetown, Guyana.
van Gemerden, B.S., 2003. The pristine rainforest: remnants of historical human impacts on current tree species composition and diversity. J. Biogeo. 30, 1381–1390
van Kuijk, M., Putz, F.E., Zagt, R.J., 2009. Effects of Forest Certification on Biodiversity. Tropenbos International, Wageningen, the Netherlands.
van Rheenen (Jacaranda), H.M.P.J.B., Boot, R.G.A., Werger, M.J.A., Ulloa Ulloa, M., 2004. Regeneration of timber trees in a logged tropical forest in North Bolivia. Forest Ecol. Manage. 200, 39-48.
Veríssimo, A., Barreto , P., Mattos, M., Tarifa, R., Uhl, C,. 1992. Logging impacts and prospects for sustainable forest management in an old Amazonian frontier: the case of Paragominas. Forest Ecol. Manage. 55, 169-199.
Vidal, E., Jones, J. Grewing, J.J. Barreto, P. Uhl, C. 1997. Vine management for reduced-impact logging in eastern Amazonia. Forest Ecol. Manage. 98, 117-129.
White, L.J.T. 1994. The effects of commercial mechanized selective logging on a transect in lowland rainforest in the Lope Reserve, Gabon. Journal of Tropical Ecology 10, 313-322.
White, L., Abernethy, K., 1997. A guide to the vegetation of the Lope Reserve, Gabon Wildlife Conservation Society. New York, USA.
Whitman, A.A, Brokaw, N.V.L., Hagan, J.M., 1997. Forest damage caused by selection logging of mahogany (Swietenia macrophylla) in northern Belize. Forest Ecol. Manage. 92, 87-96.
Wood, G.R., 1992. The bisection method in higher dimensions. Mathematical Programming 55, 391-337.
WRI (World Resources Institute). 2009. Atlas forestier interactif du Gabon (version pilote): document de synthèse. WRI : Washington, DC (Etats Unis d’Amérique), 58p.
233
Wright, S.J., Muller-Landau, H.C., Condit, R., Hubbell, S. P., 2003. Gap-dependent recruitment, realized vital rates, and size distribution of tropical trees. Ecology 84, 3174-3185
Write, P. A., Gregory, A., Hoekstra, T., Tegler, B., Turner, M., 2002. Monitoring for forest management unit scale sustainability. Final Report of the Local Unit Criteria and Indicators Development (LUCID) Test: Executive Summary. USDA Forest Service IMI (Inventory and Monitoring Institute) Report 4, Fort Collins, Colorado, USA.
Zanne, A.E., Lopez-Gonzalez, G., Coomes, D.A., Ilic, J., Jansen, S., Lewis, S.L., Miller, R.B., Swenson, N.G., Wiemann, M.C., Chave, J., 2009. Global wood density database. Dryad. Retrieved from: http://hdl.handle.net/10255/dryad.235.
234
BIOGRAPHICAL SKETCH
Vincent de Paul Medjibe, son of the late Medjibe Paul and Deyom Agathe, was
born in 1974 in northern Central African Republic (CAR). He has two sisters and three
brothers. When he was 5 years old, his parents sent him to Bangui to live with his uncle
so that he could go to school. He earned a baccalorate in sciences and went to the
University of Bangui where he graduated with a Master’s degree in plant biology. He
had the opportunity to do fieldwork in the forests of southwestern CAR as part of his
Master’s research.
From October 2001 through June 2002, he worked as a research assistant on
forest management for the Dzanga-Sangha Project, working in a protected area and an
adjacent logging concession in CAR. Through this work, he learned a great deal about
the ecology of tropical rainforests but wanted to improve his knowledge of forest
management, so he sought an additional degree at Yale University.
In August 2002, he started working on a master’s degree in forestry at Yale’s
School of Forestry and Environmental Studies. Following completion of this degree he
was employed by the Wildlife Conservation Society -Congo for two years to examine
forest structure, floristic diversity, and regeneration of timber species in the northern
Republic of Congo.
After working in the Republic of Congo, he applied to pursue a doctoral degree at
the University of Florida (UF), where he started the program in the fall of 2007. His
dissertation research concerns the impacts of logging on the forests of Central Africa,
precisely in Gabon. While a student at UF he had the opportunity to conduct research
on leaf toughness and carbon baseline in a secondary succession stand and pasture in
the Republic of Panama for 3 months, taught a field course on research methods in
235
Guyana, supervised and guided students from the school of water and forests in Gabon,
and participated in several volunteering works conducted by the Alachua Conservation
Trust in Gainesville, Florida. He received his Ph.D from the University of Florida in
spring of 2012. After completion of this doctoral program, he will work for a few months
for the Gabonese Carbon Council before starting in a post-doctoral program at Duke
University in North Carolina. His work for the Gabonese government as well as his post-
doctoral research will build on his dissertation research in Gabon.
Top Related