Impact of climate and land use on plant diversity, carbon storage and
leaf area index in the Jimma Highlands, Southwest Ethiopia
Dereje Denu Rebu
Addis Ababa University
Addis Ababa, Ethiopia
June 2016
Impact of climate and land use on plant diversity, carbon storage and
leaf area index in the Jimma Highlands, southwest Ethiopia
Dereje Denu Rebu
A Dissertation Submitted to
The Department of Plant Biology and Biodiversity Management Presented
in Fulfillment of the Requirements for the Degree of Doctor of Philosophy
(Biology: Botanical Sciences)
Addis Ababa University
Addis Ababa, Ethiopia
June 2016
ADDIS ABABA UNIVERSITY
GRADUATE PROGRAMMES
This is to certify that the Dissertation prepared by Dereje Denu Rebu, entitled: Impact of
Climate and Land use on Plant Diversity, Carbon Storage and Leaf Area Index in the
Jimma Highlands, Southwest Ethiopia and submitted in fulfillment of the Requirements
for the Degree of Doctor of Philosophy (Biology: Botanical Sciences) complies with the
regulations of the University and meets the accepted standards with respect to originality
and quality.
Signed by Examining Board:
Name
Signature
Date
1. ________________________(Examiner)
____________
_______
2. ________________________(Examiner)
____________
_______
3. Ensermu Kelbessa (Advisor)
____________
_______
4. Tadesse Woldemariam (Advisor)
____________
_______
5. Rob Marchant (Advisor)
____________
_______
6. _______________________ (Chairman)
____________
_______
Abstract
Impact of Climate and Land use on Plant Diversity, Carbon Storage and Leaf Area Index
in the Jimma Highlands, Southwest Ethiopia
Dereje Denu Rebu, PhD Dissertation
Addis Ababa University, 2016
The study aimed at the impact of climate and land use on plant diversity, live carbon
storage (AGC) and leaf area index (LAI) in the Jimma Highlands of Ethiopia. Data on
woody species were collected from 155; 20 m × 20 m sample plots which were
subdivided into 2 m×2 m subplots for herbaceous species inventory. Thirty-one plots of
one ha each were randomly distributed along a study transect for -measuring diameter
at breast height for all woody species with DBH 10 cm. Upward hemispherical images
of the forest/tree canopy were taken at 12 points in the 20 m × 20 m plots established
within each one hectare plot. Two SPOT5 satellite images (path 134 / row 133) captured
simultaneously on 17th December and aerial photographs taken in October 2012 were
used for LULC mapping. The transect was classified into five major land use types from
SPOT5 images and aerial photography. Natural forest was further separated into the
natural forest with coffee shrub/tree beneath and those with no coffee under the canopy
based on field observation. Two hundred and eight-seven plant species belonging to 220
genera and 82 families were collected and identified. The highest plant species richness
per hectare was recorded from woodland and the least was from the cropland. The
highest mean abundance of tree species was recorded from the planation forest and the
least was from the pasture. Mean annual temperature and soil pH have significantly
explained the variation in herbaceous species richness; sand and clay particles
significantly explained the variation in tree species richness. Species richness,
abundance and diversity also vary along vertical stratification in semi-forest coffee
(SFC) and degraded natural forest (DNF). The highest AGC storage was recorded from
the plantation forests (152.25±24.98) followed by DNFs (82.03±32.08) and SFCs
(61.52±24.98). Land use types showed significant mean difference in AGC and LAI.
Tree species abundance and richness combined, have explained about 82% of the
variation in AGC across the land use types. There was significant linear relationship
between AGC storage and some climate variables such as mean annual temperature,
mean annual rainfall and potential evapotranspiration; between AGC and some edaphic
factors such as soil cation exchange, sand and pH. Basal area, richness of shrub, tree
and entire plant species combined have significantly explained about 82% and 81% of
the variations in LAI_true_v6 and LAI_true_v5 respectively. LAI_true_v6 explained
about 75% of the variation in AGC. Mean annual temperature and annual temperature
range significantly explained about 21% of the variation in LAI_V5. Climate change
under the current and projected scenarios affected the distribution of five plant species
across Ethiopia. In conclusion, plant richness, abundance, distribution, carbon storage
and leaf area index are affected by land use and climate variables.
Key words: Carbon storage, climate change, Jimma highlands, LAI, land use change,
plant richness
i
Acknowledgements
I would like to thank my supervisors, Prof Ensermu Kelbessa, Dr Tadesse Woldemariam
and Dr Rob Marchant for their invaluable advice and guidance from the beginning to the
end of this work. I appreciate the support of Dr Rob Marchant in finding training
opportunities and facilitating the required processes and equipping me with the required
knowledge and skills. Addis Ababa University is duly acknowledged for the training
opportunity that I got and material provision during my stay in the University. Jimma
University is also duly acknowledged for giving me a study leave for the last six years.
This work was mainly supported by the Ministry of Foreign Affairs of Finland through
the Project “Climate Change Impacts on Ecosystem Services and Food Security in East
Africa” (CHIESA), and hence is duly acknowledged.
I would also like to thank Dr Philip Platts for his support and assistance in some aspects
of data analysis when I was in York and at home. Dr Marion Pfeifer of the Imperial
College in UK is duly acknowledged for sharing her knowledge on data collection and
analysis in Leaf Area Index and above ground live carbon storage. I am also indebted to
Mr Binyam Tesfaw for his support in producing land use/land cover change map for the
study transect.
My thanks are also extended to members of the National Herbarium, Addis Ababa
University, Mr Melaku Wondafrash, Mr Wege Abebe, Mrs Shewangizew Lemma, for
ii
their support during plant identification in the National Herbarium. Gumay and Setema
districts agricultural offices, development agents and managers in Ageyo, Setema and
Difo kebeles, the community members who allowed me data collection from their farm
and coffee plots are highly acknowledge. I extend my sincere appreciation to Mr Lijalem
Takele, Mr Yohannis Takele, Mr Nasir Mohammed and Mr Awel A/Mecha for their
support and guidance in the field data collection. I also appreciate Mr Jabir Hussen who
served us not only as a driver but also as organizer of all required logistics including
food and water supply when we were in the field.
I am greateful to my wife Mrs Uwise Teka for taking responsibilities at home during my
six years stay away from home and taking care of family members. I am indebted to my
daughter Sena Dereje and family members Lijo Takele, Soressa Teka, Tarike Teka and
Gemechu Teka who shared the hard time I had, especially at the beginning of my study.
Finally, my thanks go to my mother Sirne Birri and my brothers Asrat Denu and Takele
Denu, my sister Liknesh Denu and my relatives Mrs Likie Yadessa, Mr Wondimu
Lulessa, Mr Hailu Degefe for their material and moral support during my study.
iii
List of Figures
Figure 1: Map of East Africa including Ethiopia showing location of the study area in Jimma
Highlands (designated by CHIESA Project) and landuse types .................................................24
Figure 2: Climate diagram of Jimma Highlands, southwest Ethiopia.........................................25
Figure 3: Sampling design for AGC and vegetation data collection...........................................31
Figure 4: Sampling design for LAI data collection....................................................................31
Figure 5: Land use/cover across the study transect in the Jimma Highlands for the year 2008 ...43
Figure 6: Species area curve for all land use types across the transect in the Jimma Highlands..44
Figure 7: Plant species richness per hectare in different land use types across the transect (WLD
= woodland, DNF, SFC = semi-forest coffee, PR = pasture, PF = plantation forest, CLD =
cropland) .................................................................................................................................53
Figure 8: Cluster analysis based on species presence/absence (P1–4 = DNF, P5–8 = woodland,
P9–15 = Cropland, P16–22 = SFC, P23–27 = Pastureland, P28–31 = Plantation forest) ............59
Figure 9: Group of canopy trees in the SFC in the study transect in the Jimma Highlands (I, II,
III and IV represent group 1-4 respectively) .............................................................................62
Figure 10: Box plot of species abundance in different land use types (1 = plantation forest, 2 =
DNF, 3 = SFC, 4 = woodland, 5 = cropland, 6 = pasture) .........................................................63
Figure 11: Tree species basal area in each land use type across the transect in the Jimma
Highlands (PF = Plantation Forest, DNF = Degraded natural forest, SFC = Semi-forest coffee,
WLD = Woodland, CLD = Cropland, PR = Pasture) ................................................................69
Figure 12: Tree species richness and abundance in the vertical stratification of canopy trees in
SFC (lower < 13.33m, middle = 13.33–26.67m, upper > 26.67m) ............................................72
Figure 13: Abundance of six major canopy trees in the vertical stratification of canopy trees in
SFC (lower = <13.33m, middle = 13.33–26.67m, upper = >26.67m) ........................................73
iv
Figure 14: Tree species abundance and richness in the lower, middle and upper storeys of the
canopy trees in DNFs (lower <11.67m, middle = 11.67–23.33m, upper = >23.33m) .................74
Figure 15: Abundance of six most important canopy trees in the vertical stratification of DNF
(lower <11.67m, middle = 11.67–23.33m, upper = >23.33m) ...................................................74
Figure 16: Diversity profile test in the lower, middle and upper storeys of the canopy trees in the
SFC .........................................................................................................................................80
Figure 17: Diversity profile of canopy trees in the lower, middle and upper storeys in the DNF 82
Figure 18: Boxplot for AGC storage in different land use types in Jimma transect (1 = plantation
forest, 2 = DNF, 3 = semi-managed coffee forests, 4 = woodland, 5 = pasture, 6 = cropland)....83
Figure 19: Boxplot analysis of LAI (A = True LAI_ V6, B = True LAI V5, 1 = DNF, 2 = SFC, 3
= plantation forest, 4 = woodland, 5= pasture, 6 = cropland).....................................................96
List of Tables
Table 1: Growth form distribution of plant species in DNF.......................................................45
Table 2: Growth form distribution of plant species in woodland ...............................................46
Table 3: Growth form distribution of plant species in cropland .................................................47
Table 4: Growth form distribution of plant species in SFC .......................................................48
Table 5: Growth form distribution of plant species in pasture ...................................................49
Table 6: Growth form distribution of plant species in plantation forest .....................................50
Table 7: Species rich families across the study transect and their percent composition ..............51
Table 8: X2-test for species composition in different land use types ..........................................52
Table 9: Species richness, abundance, dominance, diversity and evenness in different land use
types ........................................................................................................................................54
v
Table 10: Contribution of mean annual temperature and pH to the regression analysis of herb
richness....................................................................................................................................56
Table 11: Contribution of each explanatory variable to the regression analysis of tree species
richness....................................................................................................................................57
Table 12: Species richness, abundance, dominance, diversity and evenness in different groups of
SFC .........................................................................................................................................63
Table 13: Difference in species abundance (4th root_abundance) across different land use types
................................................................................................................................................64
Table 14: Pairwise comparison in species abundance between different land use types (LB =
lower bound, UB = upper bound) .............................................................................................65
Table 15: Homogeneous subsets among land use types in tree species abundance.....................66
Table 16: Mean abundance of tree species in different land use types .......................................66
Table 17: Contribution of each predictor variable to the model and VIF value for each
explanatory variable, (PET = Potential evapotanspiration, CEC = cation exchange capacity, BLD
= bulk density), dependent variable: abundance........................................................................67
Table 18: Basal area contribution of tree species in SFC...........................................................70
Table 19: Basal area contribution of tree species in DNF ..........................................................71
Table 20: The difference of land use types in basal area ...........................................................71
Table 21: Tree species abundance per hectare in SFC ...............................................................75
Table 22: Tree species abundance per hectare in DNF ..............................................................77
Table 23: Species abundance, richness and diversity in SFC .....................................................79
Table 24: Species richness, abundance, dominance, diversity and evenness comparison between
middle and upper storey; lower and middle storey; lower and upper storey ...............................79
Table 25: Species richness, abundance, dominance, diversity and evenness in DNF (lower =
<13.67 m, middle = 13.67–26.67 m, upper = >26.67 m) ...........................................................81
vi
Table 26: Comparison of species diversity, richness, abundance, dominance and evenness
between lower and middle; lower and upper; lower and middle storeys ....................................82
Table 27: Average AGC in six land use types across the transect..............................................87
Table 28: Analysis of variance of different land use types in AGC storage in the study transect 88
Table 29: Multiple comparison test for the differences of land use types in AGC in the Jimma
Highlands (MD = mean difference, LB = lower bound, UB =bound) ........................................89
Table 30: Homogeneity test of land use types in AGC across the study transect in the Jimma
Highlands ................................................................................................................................90
Table 31: Linear relationships between AGC and tree, herb and shrub richness and tree species
abundance................................................................................................................................91
Table 32: Variation in AGC explained by tree species richness and abundance combined.........91
Table 33: Multiple regression analysis for prediction of AGC using abundance and tree species
richness....................................................................................................................................92
Table 34: Contribution of tree species richness and abundance to the model .............................92
Table 35: Linear regression prediction of AGC by potential evapotranspiration (pet = potential
evapotranspiration....................................................................................................................93
Table 36: Linear relationships between AGC and soil factors (CEC = cation exchange capacity,
BD = bulk density)...................................................................................................................95
Table 37: Prediction of AGC by using soil pH, sand and soil cation exchange capacity ............95
Table 38: Contribution of each variable (CEC, sand and pH) to the model ................................95
Table 39: Mean True leaf area index (under CAN-EYE version 6 and 5) in six land use types
along the study transect ............................................................................................................97
Table 40: Mean±SE of true LAI under both v_6 and v_5 of CAN-EYE ....................................97
Table 41: Paired T-test showing significant statistical differences between True_LAI under
version 6 and Version_5 of CAN-EYE .....................................................................................98
vii
Table 42: Analysis of variance showing significant differences in LAI_true_v6 and v5 and
among land use types in the transect.........................................................................................99
Table 43: Multiple comparisons showing differences in LAI_true_v6 between each land use
types (SFC, DNF, MD = mean difference, LB = lower bound, UB = upper bound) ................. 100
Table 44: Multiple comparisons showing differences in LAI_true_v6 between each land use
types (SFC, DNF, LB = lower bound, UB = upper bound, MD = mean difference) ................. 101
Table 45: Linear relationships between True leaf area indices, basal area, plant species richness
and abundance (BA = basal area, abund_4th = 4th root transformed tree species abundance) .... 103
Table 46: Analysis of variance for the multiple regression of LAI_v6 with explanatory variables
.............................................................................................................................................. 103
Table 47: Contribution of basal area, shrub richness, tree species richness, plant species richness
along the entire study area and tree species abundance (BA_log = log transformed basal area,
Abund_4th = 4th root transformed tree species abundance) ...................................................... 104
Table 48: Analysis of variance test for the prediction of LAI_v5 by the explanatory variables –
Tree species abundance, basal area, richness, shrub species richness and richness across the
entire transect ........................................................................................................................ 104
Table 49: Contribution of basal area, shrub richness, tree species richness, plant species richness
along the entire study area and tree species abundance (BA_log = log transformed basal area,
Abun_4th = 4th root transformed tree species abundance) ........................................................ 105
Table 50: Linear relationships between LAI_true indices and topographic factors (elevation and
slope) and edaphic factors (SOC, CEC, silt, sand, clay and BD) ............................................. 106
Table 51: Analysis of variance test for the prediction of LAI_true_v6 by the explanatory
variables (clay, CEC and sand across the entire transect) ........................................................ 107
Table 52: Contribution of CEC, sand and clay to the model .................................................... 107
viii
Table 53: Analysis of variance test for the prediction of LAI_true_v5 by the explanatory
variables- clay, CEC and sand across the transect ................................................................... 107
Table 54: Contribution of CEC, sand and clay to the model .................................................... 108
Table 55: Linear relationships between LAI_true indices and NDVI and EVI ......................... 109
Table 56: Analysis of variance test for the prediction of LAI_true_v6 by the explanatory
variables –EVI and NDVI ...................................................................................................... 109
Table 57: Contribution of NDVI and EVI separately to the model .......................................... 109
Table 58: Analysis of variance test for the prediction of LAI_true_v5 by the explanatory
variables- EVI and NDVI....................................................................................................... 110
Table 59: Contribution of NDVI and EVI to the model........................................................... 110
Table 60: Linear relationship between LAI_true indices and AGC ......................................... 111
Table 61: Analysis of variance test for the prediction of AGC by the explanatory variable –
LAI_true_v6 .......................................................................................................................... 111
Table 62: Analysis of variance test for the prediction of AGC by the explanatory variables (bio1
= mean annual temperature and abio7 = nnual temperature range) .......................................... 113
Table 63: Contribution of mean annual temperature and annual temperature range to the model
.............................................................................................................................................. 113
Table 64: Mean of LAI under and above the coffee canopy (uc = under coffee canopy, ab =
above coffee canopy) ............................................................................................................. 115
Table 65: Shapiro-Wilk normality test for the LAI data taken above and below the coffee canopy
.............................................................................................................................................. 116
Table 66: Paired sample t-test for the true and eff_LAI taken above and below the coffee canopy
.............................................................................................................................................. 116
Table 67: Model performance under baseline (b) and projected (p) climate change scenarios for
five plant species in Ethiopia .................................................................................................. 117
ix
Table 68: Contribution of each five climate variables to the distribution of Acacia abyssinica
under the baseline climate scenario ........................................................................................ 118
Table 69: Contribution of each five climate variables to the distribution of Acacia abyssinica
under the projected climate .................................................................................................... 119
Table 70: Contribution of each five climate variables to the distribution of Cordia africana
underthe baseline climate scenarios in Ethiopia ...................................................................... 121
Table 71: Contribution of each five climate variables to the distribution of Cordia africana under
the projected climate .............................................................................................................. 122
Table 72: Contribution of each five climate variables to the distribution of Millettia ferruginea
under the baseline climate scenario ........................................................................................ 124
Table 73: Contribution of each five climate variables to the distribution of Millettia ferruginea
under the projected climate .................................................................................................... 124
Table 74: Contribution of each five climate variables to the distribution of Phytolacca
dodecandra under the baseline climate scenario ..................................................................... 127
Table 75: Contribution of each five climate variables to the distribution of Phytolacca
dodecandra under the projected climate ................................................................................. 127
Table 76: Contribution of the five climate variables to the distribution of Schefflera abyssinica
under the baseline climate scenario ........................................................................................ 129
Table 77: Contribution of the five climate variables to the distribution of Schefflera abyssinica
under the projected climate .................................................................................................... 129
x
List of Appendices
Appendix 1: Soil and Potential evapotranspiration data for the study plots in the Jimma
Highlands (CEC = cation exchange capacity, OC = organic carbon, BLD = bulk density, PET =
potential evapotranspiration) ..................................................................................................181
Appendix 2: Species list, percent and relative frequencies in the DNF ....................................182
Appendix 3: Species list, family, growth form (GF) percent and relative frequencies (%freq,
R.F) in the woodland..............................................................................................................185
Appendix 4: Species list, family, growth form (GF) percent and relative frequencies (%freq,
R.F) in the cropland ...............................................................................................................189
Appendix 5: Species list, Growth form (GF), percent and relative frequencies (%freq, R.F) of
plant species in the SFC .........................................................................................................192
Appendix 6: Species list, Growth form (GF) percent and relative frequencies (%freq, R.F) of
plant species in pastureland .................................................................................................... 196
Appendix 7: Species list, family, growth form (GF) percent and relative frequencies (%freq,
R.F) of plant species in plantation forests of Jimma Highlands ...............................................199
Appendix 8: List of plant species in all study plots along the transect in the Jimma Highlands 202
Appendix 9: Linear relationships between plant growth forms, richness and environmental
variables ................................................................................................................................210
Appendix 10: Synoptic Table for grouping canopy trees in SFC ............................................211
Appendix 11: Linear relationships between tree species abundance and environmental variables
..............................................................................................................................................212
Appendix 12: AGC in each tree species (A) and in each plant family (B) in SFC (C t ha-1) (C t
ha-1 = carbon ton per hectare) .................................................................................................213
Appendix 13: AGC in each tree species (A) and in each plant family (B) in DNF ...................215
xi
Appendix 14: AGC in each tree species (A) and in each plant family (B) in woodland ...........217
Appendix 15: AGC in each tree species (A) and in each plant family (B) in pasture ...............218
Appendix 16: AGC in each tree species (A) and in each plant family (B) in croplands............219
Appendix 17: Linear relationships between AGC and climate variables..................................220
Appendix 18: Linear relationships between true_LAI indices and climate variables ................221
Appendix 19: Habitat suitability for the distribution of five plant species in Ethiopia under
baseline and projected climate (A1, B1, C1, D1, E1 = baseline scenario, A2, B2, C2, D2, E2 =
Projected climatescenario) ..................................................................................................... 222
Appendix 20: Jackknife test (training and test data) for the distribution of five plant species (A1,
B1, C1, D1, E1 represent training gain under baseline scenario; A2, B2, C2, D2, E2 represent
test gain under the baseline scenario; A3, B3, C3, D3, E3 represent training gain under the
projected climate; A4, B4, C4, D4, E4 represent test gain under the projected climate ............225
xii
Acronyms and abrevations
abund_4th_root
fourth root transformed abundance
AGB
above ground biomass
AGC
above ground live carbon
ANOVA
analysis of variance
ASTER
Advanced Space borne Thermal Emission and Reflection Radiometer
AUC
area under curve
BA
basal area
BA_log
Log transformed basal area
bio1
mean annual temperature
bio10
mean temperature warmest quarter
bio12
mean annual rainfall
bio13
rainfall wettest month
bio14
rainfall driest month
bio15
rainfall seasonality
bio16
rainfall wettest quarter
bio17
rainfall driest quarter
bio2
mean diurnal range in temperature
bio3
isothermality
bio4
temperature seasonality
bio5
maximum temperature warmest month
bio6
minimum temperature coolest month
bio7
annual temperature range
BD
bulk density
CEC
cation exchange capacity
CEM
climate envelop models
CI
confidence interval
CSA
Central Statistical Authority
xiii
DBH
diameter at breast height
DEM
digitalelevation model
DNF
degraded natural forest
EFAP
Ethiopian Forestry Action Program
EMA
Ethiopian Mapping Agency
ETM
Enhanced Thematic Mapper
FAO
Food and Agriculture Organization of the United Nations
GBIF
Global Biodiversity Information Facility
GCP
Ground control points
GDEM
Global Digital Elevation Model
GIS
Geographic information system
GLMs
Generalized linear models
GPS
Global Positioning system
Gt
Gigatonne
ha
hectare
IPCC
International Panel on Climate Change
IUFRO
International Union of Forestry Research Organization
IVI
Importance value index
LAI
Leaf area index
LAI_eff_v5
Effective leaf area index from CAN EYE version five
LAI_eff_v6
Effective leaf area index from CAN EYE version six
LULC
Land use/land cover
Maxent
Maximum Entropy
mi
annual moisture index
mimq
Moisture index moist quarter
MSS
Multi Spectral Scanner
pet
Potential evapotranspiration
RCP
Representative concentration pathway
RF
relative frequency
ROC
Receiver operating characteristics
xiv
SDM
Species distribution model
SFC
Semi-forest coffee
SOC
Soil organic carbon
SPSS
statistical package for Social Science
Stand_Coef
Standardized coefficient
t C ha-1
tonne of carbon per hectare
TM
landsat Thematic Mapper
un
under coffee canopy
UNESCO
United Nations Education, Science and Culture Organization
UNFPA
United Nations population Fund
Unstand_Coef
unstandardized coefficient
VIF
Variance inflation factor
°C
Degree Centigrade
xv
Table of Contents
List of Figures.......................................................................................................................... iv
List of Tables ............................................................................................................................ v
List of Appendices ................................................................................................................... xi
Acronyms and abrevations ..................................................................................................... xiii
CHAPTER ONE ....................................................................................................................... 1
1. INTRODUCTION ................................................................................................................ 1
1.1. Background of the Study ................................................................................................ 1
1.2. Research Questions and Objectives................................................................................. 7
1.2.1. Research questions .................................................................................................. 7
1.2.2. General objective ..................................................................................................... 7
1.2.3. Specific objectives ................................................................................................... 8
CHAPTER TWO ...................................................................................................................... 9
2.
LITERATURE REVIEW .................................................................................................. 9
2.1. Land Use/Land Cover Change ........................................................................................ 9
2.2. Carbon Storage..............................................................................................................12
2.3. Leaf area index (LAI) ....................................................................................................17
2.4. Climate Change and Plant Distribution ..........................................................................18
CHAPTER THREE .................................................................................................................23
3. MATERIALS AND METHODS..........................................................................................23
3.1. Study Area ....................................................................................................................23
xvi
3.1.1. Location of study area.............................................................................................23
3.1.2. Climate ...................................................................................................................24
3.1.3. Human population and economy .............................................................................26
3.1.4. Land uses ...............................................................................................................27
3.2. Materials .......................................................................................................................27
3.2.1. Data acquisition ......................................................................................................27
3.2.2. Field equipment and software .................................................................................28
3.3. Methods ........................................................................................................................29
3.3.1. Study design ...........................................................................................................29
3.3.2. Data analysis ..........................................................................................................32
3.3.2.1. Land use/Land cover mapping..........................................................................32
3.3.2.2. Species area curve ............................................................................................33
3.3.2.3. Species diversity ..............................................................................................33
3.3.2.4. Basal area ........................................................................................................35
3.3.2.5. Analysis of variance .........................................................................................35
3.3.2.6. Classification and grouping study plots ............................................................35
3.3.2.7. Vertical stratification........................................................................................36
3.3.2.8. Carbon storage .................................................................................................36
3.3.2.9. Leaf area index (LAI).......................................................................................38
3.3.2.10. Species Distribution Model ............................................................................39
3.3.2.10.1. Model building ........................................................................................40
xvii
CHAPTER FOUR ...................................................................................................................42
4. Results .................................................................................................................................42
4.1. Land Use/Land Cover....................................................................................................42
4.2. Plant Species Richness and Diversity.............................................................................43
4.2.1. Species richness in each land use type .....................................................................43
4.2.1.1. Species area curve ............................................................................................43
4.2.1.2. Species richness in DNF...................................................................................44
4.2.1.3. Species richness in woodlands..........................................................................45
4.2.1.4. Species richness in cropland .............................................................................46
4.2.1.5. Species richness in SFC ...................................................................................47
4.2.1.6. Species richness in pasture ...............................................................................48
4.2.1.7. Species richness in plantation forests ................................................................49
4.2.2. Plant species across the transect ..............................................................................50
4.2.2.1. Plant species richness .......................................................................................50
4.2.2.2. Woody species richness and diversity...............................................................53
4.2.2.3. Plant growth form distribution..........................................................................54
4.2.2.4. Frequency of occurrence of species ..................................................................54
4.2.3. Climate variables and edaphic factors against species richness ................................55
4.2.3.1. Multiple regression analysis .............................................................................55
4.2.4. Classification of study plots on the bases of species presence/absence .....................57
4.2.5. Groups of canopy trees in SFC................................................................................59
xviii
4.2.6. Vegetation structure ................................................................................................63
4.2.6.1. Land use type and plant species abundance ......................................................63
4.2.6.2. Climate variables and edaphic factors against species abundance......................67
4.2.6.3. Basal area across land use types .......................................................................68
4.2.7. Vertical stratification in SFC and DNF....................................................................71
4.2.7.1. Vertical stratification and species diversity in SFC .......................................78
4.2.7.2. Vertical stratification and species diversity in DNFs .....................................80
4.3. Carbon storage ..............................................................................................................83
4.3.1. Carbon storage in SFC ............................................................................................83
4.3.2. Carbon storage in DNFs..........................................................................................84
4.3.3. Carbon storage in woodland....................................................................................85
4.3.4. Carbon storage in pasture ........................................................................................85
4.3.5. Carbon storage in cropland .....................................................................................86
4.3.6. Carbon storage in plantation forests ........................................................................86
4.3.7. Above ground live carbon storage across the transect ..............................................86
4.3.8. Carbon storage and species richness and abundance ................................................90
4.3.9. Climate variables and AGC storage.........................................................................92
4.3.10. Edaphic factors and AGC storage..........................................................................93
4.4. Leaf Area Index (LAI) ...................................................................................................96
4.4.1. LAI and Land Use Categories .................................................................................98
4.4.2. LAI and plant basal area, abundance and richness ................................................. 102
xix
4.4.3. LAI, edaphic and topographic factors .................................................................... 105
4.4.4. LAI, enhanced vegetation index and normalized difference vegetation index......... 108
4.4.5. LAI and AGC storage ........................................................................................... 110
4.4.6. LAI and climate variables ..................................................................................... 112
4.4.7. LAI above and below coffee canopies ................................................................... 114
4.4.8. Normality test ....................................................................................................... 115
4.4.9. Significance test ................................................................................................... 115
4.5. Habitat Suitability Model ............................................................................................ 117
4.5.1. Acacia abyssinica ................................................................................................. 117
4.5.1.1. Analysis of variable importance ..................................................................... 118
4.5.2. Cordia africana .................................................................................................... 119
4.5.2.1. Analysis of variable contributions .................................................................. 120
4.5.3. Millettia ferruginea............................................................................................... 122
4.5.3.1. Analysis of variable contributions .................................................................. 123
4.5.4. Phytolacca dodecandra ........................................................................................ 125
4.5.4.1. Analysis of variable importance ..................................................................... 126
4.5.5. Schefflera abyssinica ............................................................................................ 128
4.5.5.1. Analysis of variable importance ..................................................................... 128
CHAPTER FIVE ................................................................................................................... 131
5. Discussion, Conclusion and Recommendations .................................................................. 131
5.1. Discussion ................................................................................................................... 131
xx
5.1.1.
Land use /land cover change ........................................................................... 131
5.1.2.
Species richness .............................................................................................. 132
5.1.3.
Basal area ....................................................................................................... 135
5.1.4. Plant growth forms ............................................................................................... 136
5.1.5.
Above ground live carbon storage ................................................................... 141
5.1.6. Leaf area index (LAI) ........................................................................................... 145
5.1.7. Species distribution............................................................................................... 150
5.2.
Conclusion ............................................................................................................. 154
5.3.
Recommendations .................................................................................................. 156
References ............................................................................................................................. 159
Appendices ............................................................................................................................ 181
xxi
CHAPTER ONE
1. INTRODUCTION
1.1. Background of the Study
The impact of land use/land cover (LULC) change and climate variables on
biodiversity is widely understood across the world. The term biodiversity has wider
definition and includes all forms of life on earth at any level of organization as
clearly indicated at the Rio de Jeneiro Convention on Biodiversity in 1992 (UN,
1992). According to this convention, biodiversity is defined as the variability among
living organisms from all sources, including terrestrial, marine and other aquatic
ecosystems and the ecological complexes associated with them. This includes
diversity within species, between species and of ecosystems. Diversity within and
between species can also be considered genetic diversity and species diversity
respectively. According to Heywood and Watson (1995) estimation, about 300,000
species of vascular plants have been documented out of the estimated global total
volume of 400,000 species of vascular plants . Habitat loss due to anthropogenic
LULC change and global warming are threats to biodiversity across the world
(Kappelle et al., 1999) and are causes for the current extinction of species (IPCC,
2007; Pimm et al., 2014). Land use/land cover change also affects plant species
richness, diversity, abundance (Martınez et al., 2009) and biomass (Fearnside et al.,
2009; Kauffman et al., 2009).
The distribution of biodiversity across the world is not even. Some areas are
characterized by high species concentration mainly composed of high levels of
1
endemism and rapid rate of depletion (Myers, 1998). Human induced LULC change
resulted in the formation of many ecological islands within highly converted habitat
(Jenkins, 1992). This could interrupt the interactions among all biological entities on
the planet and as a result the loss of one particular species could impact the
persistence of another species. Human induced anthropogenic pressures affect the
healthy functioning of an ecosystem. Ehrlich (1981), Myers (1998), Marina (2010),
Cardinale et al. (2012) showed the impact of biodiversity loss on ecosystem
integrity, which could threaten the existence of mankind.
The increasing human population and accompanying needs has increased pressure on
the natural resources including plant cover, animal life, mineral resources and land.
According to FAO (2010), forests cover 31% of the land on the planet, of which
around 13 million is deforested each year. The annual loss of natural forest in the
tropics was estimated to be 15.2 million hectares (FAO, 2001). In Ethiopia, the land
cover change is dramatic and resulted in the decline of Ethiopian high forest cover
from about 35–40% in the 19 century (Breitenbach, 1961) to 2.3% in 1990 (EFAP,
1994). The annual loss of closed forest and the entire natural vegetation in Ethiopia is
about 10,000 ha (Landon, 1996) and 160,000–200,000 ha (Konemund et al., 2002)
respectively. Ethiopia lost 269,795.88 ha of forest cover within 13 years time (20002013) which is equivalent to about 20,754 ha per year (Hansen, et al., 2013). FAO
(2010) put the Ethiopian forest cover at about 11% (12.2 million ha) of the Ethiopian
land area. This could probably be due to the change in FAO’s forest definition.
2
The forest loss due to human induced land use change causes the loss of biodiversity
and compromises the provision of ecosystem goods and services such as pollination
services, erosion control, cycling of materials involving biotic and abiotic
components, carbon storage and climate regulation. Deforestation disrupts the
healthy functioning of an ecosystem through habitat fragmentation and formation of
landscape mosaics. The fragments (formed as a result of deforestation) are
surrounded by other land use types usually referred to as matrix. The matrix around
the fragment may have positive, negative or neutral effect on the patch (Franklin et
al., 2002), usually depending on the type of land use.
Climate change is another threat to the global biodiversity. According to IPCC
(2014), the earth’s temperature has increased by approximately 0.65°C–1.06°C over
the past 132 years. According to the report, the period from 1983–2012 was the
warmest 30 year period of the last 1400 years in the Northern Hemisphere. The two
main periods of warming were recorded between 1910 and 1945 and from 1976
onwards. The rate of warming from 1976 onwards has approximately been doubled
that of the first period and, thus, greater than at any other time during the last 1,000
years (IPCC, 2001). Of the past 10 centuries, the 20th century was the warmest of all,
and the 10 years in 1990s were the warmest of the entire period (IPCC, 2001). The
change in the climate and the occurrence of this warming pattern is an evidence for
the human induced climate change that resulted from the increased emission of
greenhouse gases (IPCC, 2001). According to IPCC (2014), the global mean surface
temperature increases by 2.6°C–4.8°C by the end of 21st century under the extreme
representative concentration pathway (RCP8.5), the scenario with the highest
3
greenhouse gas emission. The rise for Africa was predicted up to 4.5°C provided that
the current addition of greenhouse gases from fossil fuels continued (Platts et al.,
2014). Africa is also more vulnerable to the climate change impacts than the
developed nations due to its dependence on subsistence agriculture and natural
resources as the main source of livelihood.
Climate change has severe impact on the terrestrial as well as aquatic ecosystems.
Climate change also has a negative impact on the entire biodiversity and ecosystem
functioning such as seasonality/cycling of natural events, water cycles, erosion
control, pollination services, provison of fresh water and nutrient cycling. The degree
of vulnerability to climate change among the people in the world is different with
developing countries being more affected by climate change impacts. Africans are
mostly dependent on natural resources for their livelihoods including food, shelter,
medicine and the impact of climate variability on the biodiversity directly affects
natural resources and the associated livelihoods of many African nations. Most
Africans are dependent on rain-fed, hand to mouth subsistence agriculture which is
directly affected by climate variability.
Climate change has put the distribution of biodiversity including plants under
pressure (IPCC, 2014).
It affects the spatial distribution of plants (latitude,
elevation). According to Grabherr et al. (1994) and Parmesan and Yohe (2003),
plants are changing their distributional ranges both in altitude and latitude in
response to changing regional climates. Pounds et al. (1999), Still et al. (1999),
4
Thomas et al. (2004), Platts et al. (2013) also showed the range shift in species
distribution in response to climate change.
Ethiopia is endowed with different land features and topographies ranging from 125
m below sea level at Kobar sink to 4620 m above sea level (the peak of Ras Dashen).
This diversified habitat types results in the formation of different agro-climatic zones
in Ethiopia which contribute to the formation of diversified flora occurrence and
distribution in the country. There are about 6,000 species of higher plant taxa, of
which 10% are endemic, occurring in Ethiopia due to this diversified ecological and
geographical setting that results in different natural resources (Ensermu Kebessa and
Sebsebe Demissew, 2014).
Carbon storage is one of the most important ecosystem services that human beings
obtain from the natural resources. Carbon sequestration and storage in plant biomass
has been widely accepted as the most important regulators of climate change. Green
plants convert CO2 (one of the most important greenhouse gases) into organic food
during the process of photosynthesis and store in their biomass. In this regard, forests
are important sinks of carbon because they trap carbon in their biomass throughout
their life. According to Chave et al. (2014), about 50% of the above ground live plant
biomass is carbon. Globally, forests store about 289 Gt of carbon in their biomass
(FAO, 2010). Deforestation caused about 20% of greenhouse gas emissions
worldwide and about 70% of emissions in Africa (Gibbs et al., 2007).
5
Leaf is an interface between the plant canopy and the surrounding atmosphere. Leaf
area index (LAI) is very important biophysical variable which takes part in the
conversion of CO2 to organic food and insures continuity of natural interactions of
organisms in the food web (Pfefier et al., 2012).
The removal of tree cover in conversion from its natural setting to human modified
landscapes also affects the live carbon storage in the living plant biomass (Kauffman
et al., 2009). The current rate of LULC change is contributing to more carbon
emission into the atmosphere and reducing the carbon sink and enhancing the climate
change (IPCC, 2000) and loss of biodiversity (Bellard et al., 2012). The removal of
natural vegetation also affects the plant leaf area index.
The impact of land use and climate variables on plant species richness, abundance,
LAI, AGC storage in woody species biomass has not been assessed so far in the
Jimma Highlands. The impact of climate change on distribution of some plant
species such as Cordia africana (multipurpose tree), Acacia abyssinica (important
coffee shade tree), Millettia ferruginea (important coffee shade tree), Schefflera
abyssinica (important honey source) and Phytolacca dodecandra (important
medicinal plant) has not been assessed in Ethiopia so far. Therefore, there is a need
to assess the impact of land use and climate variables on plant diversity, richness,
above ground live carbon storing capacity and leaf area indices of different land use
types in the study transect. There is also a need to assess the impact of climate
change on the distribution of the selected plant species across Ethiopia.
6
1.2. Research Questions and Objectives
1.2.1. Research questions
•
Is there any difference in plant species richness, tree species abundance and
diversity across the land use types in the study transect of the Jimma
Highlands?
•
Do the same canopy trees dominate the semi-forest coffee forests across the
transect?
•
Is there any variation in above ground live carbon storage and LAI among the
land use types across the transect in the Jimma Highlands?
•
Is there any relationship between environmental variables (climate and
edaphic) and plant species richness, AGC and LAI in the Jimma Highlands?
•
Is there any change in habitat suitability for the distribution of the target plant
species under the present and future projected climates in Ethiopia?
1.2.2. General objective
The general objective of this study was to determine the difference in plant species
richness, diversity and abundance and determine the relationships of this to carbon
storage and leaf area index across different land use types along an altitudinal
transect in the Jimma Highlands; using these insights, the study then focused on
modelling the distribution of some key plant species under current and future
projected climates across Ethiopia.
7
1.2.3. Specific objectives
The specific objectives of this study include:
1. Determining the variation in plant species richness, diversity, abundance and
dominance across different land use types in the study transect of the Jimma
Highlands;
2.
Determine the dominant canopy trees in semi-forest coffee across the transect in
the Jimma Highlands;
3. Find out the difference in carbon storage among different land use types in the
study transect of the Jimma Highlands;
4.
Find out the difference in LAI among different land use types in the study
transect of the Jimma Highlands;
5. Determine the relationships between environmental variables and plant species
richness, abundance, AGC and LAI
6. Model habitat suitability for the distribution of five plant species under the
current and future climate change scenarios across Ethiopia.
8
CHAPTER TWO
2. LITERATURE REVIEW
2.1. Land Use/Land Cover Change
Anthropogenic activities are the main cause of land cover change. About 13 million
hectare of forest cover is lost annually due to deforestation (FAO, 2006). In the
tropics alone, 15.2 million hectare of forest cover is lost each year (FAO, 2001).
Anthropogenic land use change is the primary cause for the reduction of total
vegetation area in Africa (Niang et al., 2014). The annual loss of closed forest and
natural vegetation in Ethiopia is about 10,000 ha (Landon, 1996) and 160,000–
200,000 ha (Konemund et al., 2002) respectively. According to Sala et al. (2000),
land use change has probably the largest effect followed by climate change and
elevated carbon dioxide on the terrestrial biodiversity.
Ethiopia is one of the countries in Africa where the anthropogenic LULC change has
excarbated the forest loss (EFAP, 1994; Kumelachew Yeshitela, 2001; Tadesse
Woldemariam and Masresha Fetene, 2007). There has been a considerable reduction
in the Ethiopian high natural forest cover since 19 century. It was 35–40% in the 19
century (Breitenbach, 1961), 16% in the early 1950’s, 3.1% in 1982, 2.7 in 1989 and
2.3% in 1990 (EFAP, 1994). The study conducted on Shaka Forest (part of the
Afromontane rainforest in southwest Ethiopia), pointed out that the dense closed
forest declined from about 55,304 ha to 43,424 ha by 2001 and open forest decreased
from 46,594 ha to 35,077 ha during the same period (Tadesse Woldemariam and
9
Masresha Fetene, 2007). As a result the size of disturbed forests increased by 16,355
ha and the agricultural land increased from 8,620 ha to 14,672 ha.
LULC change is the most important problem that has caused the forest loss in
Ethiopia. Among the causes of LULC change contributing to the loss of the entire
forest resources and associated biodiversity are rapid human population growth,
poverty, forest clearing for cultivation, over-grazing, and exploitation of forests for
fuel wood, construction and lack of proper policy framework. The human population
growth in Ethiopia could be the most important driving force behind the loss of
forest biodiversity. According to the UNFPA (2009) report, the Ethiopian population
increased by 361% during the period of 1950–2010. UNFPA’s (2009) human
population projection also showed that the Ethiopian population will increase by
105% (reach 173.8 million) by 2050.
In a country like Ethiopia, with mostly agrarian community, the population growth
has direct relationship with the LULC change. In Ethiopia, 95% of the cultivated
land is occupied by smallholder subsistence agriculture (Shibru Tedla and Kifle
Lamma, 1998). The rising human population has increased the need for new
cultivatable land and additional energy supply. The cumulative effects of the small
holder agricultural production and increasing human population lead to loss of
natural forest and land degradation in Ethiopia. The human population of Jimma
Zone, for example, increased from 1,960,033 to 2,495,795 (CSA, 1996; CSA, 2008).
10
The rapid population growth led to expansion of agricultural land, increased
exploitation of fuel wood and construction material. These in turn led to loss of
vegetation (deforestation). According to EFAP (1994), one of the reasons for the
decline of Ethiopian forest cover is attributed to energy requirements. About 94% of
the energy requirement in Ethiopia relies on biomass alone, of which trees and
shrubs contribute the largest proportion (Haileleul Tebicke, 2002).
The new investment opportunities in southwestern Ethiopia are converting the few
remaining Afromontane rainforests into other land use systems such as coffee and tea
plantations (Taye Bekele et al., 2001). The study on Chewaka-Utto in southwest
Ethiopia (where most remnant forests of the country are present) showed the
conversion of natural forest from 85% in 1996 to 76.3% in the year 1999
(Kumelachew Yeshitela, 2001) due to clearing of the forest for tea and Eucalyptus
plantations. Another study on Shaka Forest (also in southwest Ethiopia) showed the
decline of dense forest cover from about 60% in 1973 to 20% in 2005 (Tadesse
Woldemariam and Masresha Fetene, 2007). New settlements in forests are increasing
and have resulted in the conversion of forestland into agriculture and other land use
systems. The annual deforestation rate in Kafa zone (southwest Ethiopia) where the
country has a UNESCO recognized biosphere, is approximated to be 22,500 ha.
(http://www.nabu.de/en/aktionenundprojekte/kafa/projectarea/climateprote.)
LULC change is one of the threats to biodiversity. Of the estimated 13 million
species worldwide, only 1.6 million has been described (Heywood and Watson,
1995). Biodiversity is not equally and evenly distributed throughout the world.
Mountains, for example, support about one-quarter of terrestrial biodiversity
11
worldwide and nearly half of the world’s biodiversity Hotspots (Spehn et al., 2010).
As Myers (1998) indicated there are areas with high species concentration mainly
composed of high levels of endemism and rapid rates of depletion. Kappelle et al.
(1999) in his review indicated that habitat alteration and loss, over-harvesting,
chemical pollution, invasive species and increasing human population pressure are
the major threats to the global biodiversity.
2.2. Carbon Storage
Antropogenic LULC change affects the health and wealth of an ecosystem. Carbon
dioxide is the most important green house gas. Forests could be used as both carbon
sink and carbon source. When it is managed and well conserved, it is important sink
of carbon by converting CO2 to organic food and storing in their biomass. When
deforested, they are sources of CO2 by emiting the carbon from their biomass
through burning and decomposition. Acording to FAO (2010), global forest stores
about 289 Gt of carbon in their biomass. LULC change and deforestation affected the
carbon storing capacity of ecosystems. Carbon storage is one of the ecosystem
services through which the climate could be regulated. Studies showed that LULC
change affects above ground woody carbon stock (Asner et al., 2003), carbon
emission into the atmosphere (Achard, et al., 2004, Kaplan, et al., 2010). According
to IPCC (2000) report, about 136 Gt C is emitted as a result of land-use change,
mainly from forest ecosystems, leading to an increase of CO2 by 176 Gt in the
atmosphere.
12
In the tropics, deforestation induced changes in the atmospheric circulation (Avissar
et al., 2004). About 20% of global greenhouse gas emission resulted from
deforestation. The amount of greenhouse gas emitted due to forest loss varies from
region to region. In Africa, deforestation caused about 70% of greenhouse gas
emision (Gibbs et al., 2007). The destruction of tropical forests due to human
activities contributed up to 17% of global CO2 emission (IPCC, 2007).
Anthropogenic LULC change played an important role in carbon emision world
wide. Houghton (1999) reported that LULC change contributed about 33% to total
anthropogenic carbon emison for the last 150 years. Acording to Friedlingstein et al.
(2010), the percent contribution of LULC change to anthropogenic carbon emission
was reduced to 12.5% mainly due to rise in emission from fossil fuels (during the
period 2000-2009). From 2005–2010 carbon stocks in forest biomass declined by an
estimated 0.5% Gt annually due to reduction in global forest cover (Yitebitu Moges
et al., 2010).
Agro-forestry is composed of mixed plant species. Each species of plants in the agroforestry has the capacity to sequester carbon and convert into its biomass. As
Montagnini and Nair (2004) indicated, if proper management is designed for agroforestry practices they could serve as effective carbon sinks. The average estimation
of stored carbon in agro-forestry practices indicated that 9, 21, 50, and 63 Mg C ha-1
in semi-arid, sub-humid, humid and temperate regions respectively (Montagnini and
Nair, 2004). Most countries in sub-Saharan Africa practice smallholder agriculture.
Ethiopia is one of these countries with about 85% of its population practicing
13
smallholder agriculture. In countries having agricultural communities with
smallholder agro-forestry practices in the tropics the rate of carbon sequestration
ranges from 1.3 to 3.5 Mg ha-1 yr-1 (Montagnini and Nair, 2004). In addition to its
importance in directly sequestering carbon from the atmosphere and serving as
mitigation to reduce the CO2 concentration in the atmosphere, it also serves indirectly
by reducing the human pressure on the natural forests. The soil management
practices for the improvement of agro-forestry enhances carbon storage in trees and
soils (Montagnini and Nair, 2004).
Global forest ecosystems store about 2.1 Gt of carbon (on average) annually (FAO,
2015). Studies showed that a significant portion of the absorbed carbon is returned to
the atmosphere through deforestation and forest fire. According to FAO (2003)
estimation, about 25% of the carbon emission from all human activities in the tropics
was attributed by tropical deforestation. IPCC (2000) also indicated that tropical
deforestation contributed about 25% net annual carbon emissions. The three
approaches by which forest managements help to reduce carbon in the atmosphere
are carbon sequestration, carbon conservation and carbon substitution.
The basic premise of carbon sequestration potential of land use systems including
agro-forestry revolves around the fundamental biological/ecological processes of
photosynthesis, respiration and decomposition (Nair and Nair, 2003). Essentially,
carbon sequestered is the difference between carbon gained by photosynthesis and
14
carbon lost or released by respiration of all components of the ecosystem. The
overall gain or loss of carbon is usually represented by net ecosystem productivity.
Reduction of natural forests due to human land use change reduced the amount of
carbon that would be stored in the forest. The decline in carbon storage is critical
when the forest is changed into agricultural lands for annual crops. Mixed agroforestry practices in which a variety of perennial tree species are available contribute
greatly in the amount of carbon they absorb from the atmosphere and store in their
biomass. Agro-forestry systems including shade coffee farm help in carbon storage
(Polzot, 2004; Schmitt-Harsh et al., 2012; Getachew Tadesse et al., 2014).
Traditional coffee plantation is one of the most important agro-forestry practices both
in environment conservation and in carbon sequestration. As it was indicated by
Tadesse Woldemariam and Feyera Senbeta (2008), in Ethiopia for example, there are
four different coffee producing systems: forest coffee, semi-forest coffee, shade
coffee and non-shade coffee systems. Of all these coffee agro-forestry systems, this
study addressed only the semi-forest coffee system which is commonly found in the
study transect. Multi-strata shade coffee system provides a partial compensation for
carbon loss (Van Noordwijk et al., 2002). This demonstrates the importance of
coffee shade trees in converting carbon dioxide of the atmosphere and storing in their
biomass. This can be used as one of the mitigation measures for climate change by
maintaining the normal level of carbon concentration in the atmosphere.
15
Trees are terrestrial carbon sinks. Worldwide, forest plantations were estimated to
cover 124 million hectares in 1995; 187 million hectares in 2000 (FAO, 2000) and
264 million hectares between 2000 and 2010 (FAO, 2010). According to FAO
(2000), the annual rate of planting trees was 4.5 million hectares. The annual rate of
planting trees increased to 5 million hectares from 2000–2010 (FAO, 2010). Forest
plantation showed an increasing trend in all continents from 1990 to 2010. Compared
to primary forests, plantation forests store relatively small amount of carbon
(Thornley and Cannell, 2000)
In an ecosystem, interactions between biodiversity and carbon storage have been
observed. Linear relationships with varying strengths were observed between
biodiversity and carbon stock (Midgley et al., 2010; Talbot, 2010). There are also
evidences showing the relationship between carbon storage and species richness and
abundance in an ecosystem. Strassburg et al. (2010) showed positive relationship
between terrestrial carbon stocks and biodiversity. Talemos Seta and Sebsebe
Demissew (2014) showed positive relationship between above ground carbon storage
and species richness (strong corellation) and with stem density (weak correlation).
The climate variables (temperature and moisture) have direct and indirect impacts on
the physiological activities of plants which in turn could affect the plant growth and
carbon storage. Tian et al. (1998) showed that dry weather and warmer temperatures
decrease net primary productivity. Increased precipitation leads to an increase in the
soil moisture content and this in turn affect the ecosystem productivity (Tian et al.,
16
1998). Any change in precipitation pattern could also affect the plant growth
(Myneni et al., 1997). Gentry (1982) showed a strong relationship between
precipitation and productivity of an ecosystem.
2.3. Leaf area index (LAI)
Leaves are the most important parts of a plant for photosynthesis, gas exchange,
water regulation and energy fluxes. They help as an interface between the plant
canopy and the atmosphere. The orientations of foliage in three dimensional spaces
govern the interaction between plant canopy structure and atmosphere (Fieber et al.,
2014). This derives the energy flux between the canopy and atmosphere (Koetz et
al., 2006). Leaf area index (LAI) is a very important biophysical parameter that
characterizes the interaction between the canopy and the atmosphere. Watson (1947)
defined LAI (which is applicable for broad leaves) as the total one-sided area of leaf
surface per unit ground area. Natural events and human activities could affect the
canopy leaf area index, which in turn could affect the ecosystem productivity.
Quantification of this dimensionless plant canopy trait is complex due to temporal
and spatial variability of an ecosystem. There are direct methods (Hutchison et al.,
1986; Neumann et al., 1989) and indirect ways (Norman and Campbell, 1989) to
quantify LAI.
The direct methods of acquiring LAI are difficult for large spatial extents due to its
time consuming and work intensive nature (Jonckheere, 2004; Zheng and Moskal,
2009). As an alternative to some practical limitations of the direct method of LAI
17
estimation, several indirect optical devices have been developed since 1960’s (Facchi
et al., 2010). LICOR LAI-2000 Plant Canopy Analyzer and Decagon AccuPAR-80
ceptometer are the two mostly used optical devices for LAI estimation. These optical
devics determine LAI from the radiation transmitted through the canopy. According
to Jonckheere (2004), incoming radiation, plant canopy structure and its optical
properties determine the energy intercepted by a canopy. LAI estimation from
hemispherical photos has been introduced and developed since 1980’s (Facchi et al.,
2010). RGB cameras with fine resultions solved the problem of distinguishing leaves
from the the sky (in case of upward hemi-images) or the ground (for the downward
hemi-images) (Jonckheere, 2004). CAN-EYE is one of the image processing
software packages used for estimating LAI from hemispherical images (Facchi et al.,
2010).
2.4. Climate Change and Plant Distribution
Plants respond to climate change through range shift in their spatial distribution both
in latitude and altitude (Grabherr et al., 1994; Parmesan and Yohe, 2003). Many
terrestrial and aquatic species have shifted their geographic ranges, seasonal
activities, migration patterns, abundances and species interactions in response to
climate change (IPCC, 2014). The species in the alpine region are prone to the
impact of climate change compared to those in the lowland or midlands (Thomas et
al., 2004). The current rate of plant migration in response to the changing climate is
faster than the past migrations. When the species at the lower altitude shift their
geographic ranges upward, what will happen to those at the higher elevations? In
18
response to this, Thomas et al. (2004) indicated the possible maximum of extinction
risk of the species in the alpine zone. Grabherr et al. (1994) in his study on Swiss
Alps indicated the pronounced range shift of plant species to the higher elevations.
The rate of expected upward shift is 8–10m per decade based on the mean
temperature change for the last 90 years (Grabherr et al., 1994). Pounds et al. (1999)
and Still et al. (1999) showed the loss of many cloud forest species and invasion by
the species from the lower elevations due to climate change.
There are documented range shifts in plant distribution by 6.1 km per decade towards
the pole (Parmesan and Yohe, 2003). Tree species changed their elevation or latitude
range in response to changes in Quaternary climate (Davis and Shaw, 2001). The
climate change together with change in the human land use system may disrupt the
relationship of migration and adaptation (Davis and Shaw, 2001). This clearly affects
the productivity and persistence of several species. On the basis of the current
scenarios of carbon dioxide, climate, vegetation and land use changes the scenarios
of changes in biodiversity for the year 2100 has been developed. The future climate
projections are based on the anthropogenic greenhouse gas emissions which are
summarized under four different representative concentration pathways-RCP2.6,
RCP4.5, RCP6.0 and RCP8.5) (IPCC, 2014). Of the four representative oncentration
pathways, RCP2.6 keeps global warming below 2°C preindustrial temperature and
RCP8.5 considers the extreme emissions of greenhouse gases, while the remaining
two are considered intermediate.
19
Climatic change is not only affecting the distribution of plants, but also impacts the
phenology or the seasonal activities of plants (IPCC, 2014). Plants have a fixed
period of time for flowering, for shedding their leaves and to be in foliage, to
produce fruits and seeds. These events happen at a regular cycle within the year. The
climate change may interrupt and affect this natural cycling of events in plants and
has a meaningful impact on the survival of plants.
Climate and climate factors affect the distribution of vegetation. In Ethiopia, for
example, there are different vegetation types based on climate and topographic
variations. There are two main topographic factors that govern the Ethiopian climate.
One is the location of Ethiopia in relation to the equator and the second is the relief
condition of Ethiopia. Ethiopia is located closer to the equator (the southern
boundary is at approximately 3°30' north latitude) and its relief ranges from 125
meters below sea level to 4533m above sea level.
Species distribution models (SDM) are used to show the impact of climate change in
the distribution of the species under consideration. SDM relates the spatial
distribution of organisms with environmental covariates. Species distribution models
are applied in various areas of study such as conservation work, alien species
management and in the study of evolutionary changes (Corsi et al., 1999; Peterson et
al., 1999; Guisan and Zimmermann, 2000; Kriticos and Randall, 2001; Welk et al.,
2002; Guisan and Thuiller, 2005).
20
Based on the type of biological data (occurrence data), the method of modelling and
the software applied are different. The biological data could fall in one of the
following two data sets.
1. Presence/absence data
2. Presence only data
In presence/absence data, there is a need to have both occurrence and absence
records for the species desired to model, while we need to have occurrence records
(longitude and latitude) of the species for presence only data to run the model.
According to Ponder et al. (2001), there is shortage of absence data in the tropics
where the sampling was poor and at the same time, where conservation is very
important. There are ample presence records in the tropics, while the absence data
are poorly available. General purpose statistical models such as generalized linear
models and generalized additive models address the presence/absence data types.
There are different types of climate envelop models such as Bioclim, Domain,
General additive models and Maxent (Maximum Entropy). Among the existing
Climate Envelop Models, Maximum Entropy Species distribution modelling
approach was originally designed for statistical mechanics (Jaynes, 1957) and was
applied for species habitat distribution modelling by Philips et al. (2006). It is one of
the climate envelop models applied for making extrapolations from presence only
data. When it is applied to presence only species distribution modelling, the study
area makes up the space on which the Maxent probability distribution is defined, the
points with known species occurrence records constitute the sample points, and the
features are environmental variables which could be climatic, elevational, edaphic,
vegetation or other environmental variables depending on the data at hand.
21
According to Philip et al. (2006), Maxent is advantageous in handling presence only
data
with
environmental
variables,
avoiding
model
over-fitting
through
regularization and converge to the optimum probability distributions, among others.
22
CHAPTER THREE
3. MATERIALS AND METHODS
3.1. Study Area
3.1.1. Location of study area
This study was carried out in the Jimma Highlands, southwest Ethiopia (Figure 1)
from November 2012–April 2014. The study area was designated “Jimma
Highlands” by the Climate Change Impacts on Ecosystem Services and Food
Security in Eastern Africa (CHIESA) Project. Jimma Highland is the wettest part of
Ethiopia and belongs to the Eastern Afromontane Biodiversity Hotspot (Mittermeier
et al., 2004). The landscape is a mosaic of different land uses such as semi-forest
coffee (henceforth, SFC), croplands, pastures, natural forests, woodlands and
plantation forests. The plantation forests are composed of exotic species such as
Cupressus lusitanica, Eucalyptus camaldulensis, Grevillea robusta and Pinus patula.
The study transects ranges from 1500–2200 m above sea level.
23
Figure 1: Map of East Africa including Ethiopia showing location of the study area
in Jimma Highlands (designated by CHIESA Project) and landuse types
3.1.2. Climate
Southwest Ethiopia is the wettest region of the country with eight months of heavy
rains (March-October). The precipitation and temperature data for 17 years (1990–
2007) were obtained from Ethiopian Meteorology Agency (Jimma station) for
24
constructing the climate diagram following Walter (1985) (Figure 2). As it is
indicated in Figure , Jimma (including the entire southwest Ethiopia) is characterised
by uni-modal rainfall pattern. The average monthly precipitation (1990–2007) was
1563 mm. The average monthly maximum and minimum temperatures for the period
of time indicated above are 30°C and 7.7°C respectively (Figure 2). The mean annual
maximum temperature ranges from 26.53-28.63°C, while the mean annual minimum
temperature ranges from 3.1-12.2°C. The arid period prevails in February and the
rest relatively represents the humid period. The area in black represents the perhumid period of the year getting above 100 mm of rainfall on average. This area
extends from March to October.
jimma (1760 m)
1990-07
19.7C
1563 mm
300
C
mm
50
100
40
80
30
60
20
40
10
20
30.0
7.7
0
0
J
F
M
A
M
J
J
A
S
O
N
D
Figure 2: Climate diagram of Jimma Highlands, southwest Ethiopia
25
3.1.3. Human population and economy
According to 2007 Ethiopian census report (CSA, 2008), the human population of
Setema and Gumay districts were 103,748 (males = 50999, females = 52749) and
61,333 (males = 30707, females = 30626) respectively. About 97% of the population
of Setema and 95% of Guma districts were Muslems and the remaining belong to
Christianity. Oromo is the major ethnic group in both districts. There are some
Amhara and Tigre settlers from northern Ethiopia. The majorities of the population
in both districts are rural and depend on smallholder subsistence agriculture. Coffee
(Coffea arabica) and “Chat” (Catha edulis) farming is the main agriculture activity
in both districts for generating income. Traditional honey farming is another source
of income in the districts. The conservation of forests patches for coffee farming
made both districts ideal places for honey production. Particularly the conservation
of Schefflera abyssinica (local name = Gatama) in Jimma Zone contributed to the
production of white honey which has priority in the local market compared to honey
from other sources. The common crops in both districts include cereals such as
maize, sorghum, teff, barley and wheat and pulses such as Vicia faba and Pisum
sativum. Animal farming is another important sector in supporting the rural
community in both districts. The community has been using horse back as a means of
transportation in both Setema and Gumay districts. In the past, horse was used for
transportation purposes within and out of the districts, but nowadays, vehicles have
replaced the horse back for longer journeys like from one district to another.
26
3.1.4. Land uses
According to the agricultural offices of Setema and Gumay districts the land use
types are classified into arable land, pasture, and moist montane forest and degraded
land. About 27% of the total land area of Setema District was arable, while pasture
and forests represent about 13% and 55% respectively. The remaining land (about
5%) is degraded and has no use. In Gumay District, the cultivable land is about 61%,
while pasture, forest and unusable land (swamps and mountains) represent about 8%,
5% and 20% respectively. Setema District has more forest cover compared to the
neighbouring Gumay District.
3.2. Materials
3.2.1. Data acquisition
There are two sources of data in this study.
1. Primary data
The primary data were directly collected during the field work from November
2012 - April 2014 in the Jimma Highlands.
2. Secondary data
Specific wood densities for woody species were acquired from Global Wood Density
Database (Chave et al. 2009; Zanne et al., 2009). Climate variables were taken from
down scaled AFRICLIM 1km resolution for East Africa (Platts et al., 2014) from
KITE website (http://www.york.ac.uk/environment/research/kite/). Enhanced
vegetation index (EVI) and normalised difference vegetation index (NDVI) for the
27
year 2012 were downloaded from MODDIS sattelite imagery
(https://lpdaac.usgs.gov/products/modis_products_table/mod13q1) using the Google
search engine. The hd file formats were converted to tif format in R free statistical
software version 3.0.1 (R Core Team, 2014). Extraction of EVI and NDVI values from
grid cells for the transect was caried out in QGIS version 2.2.0-Valmiera. Soil data
for 1km resolution were downloaded from World Soil Information Database (ISRIC,
2013) as additional input. Appendix 1 shows the soil and evapotranspiration data for
the study plots along the transect in the Jimma Highland.
Plant species occurrence data were obtained from herbarium specimens and from
databases. The occurrence data were collated from the herbarium specimens housed
at the National Herbarium (ETH), Addis Ababa University. The data bases of Global
Biodiversity Information Facility (GBIF) (http://www.gbif.org/) and TROPICOS
(Missouri Botanical Garden) (http://www.tropicos.org/) were also used for collecting
the occurrence data across Ethiopia.
3.2.2. Field equipment and software
Global positioning system (GPS) for recording the location of species (longitude,
latitude and altitude); hemispherical camera (Nikon D7000) with fish eye lenses for
taking upward hemispherical images which later processed to produce LAI,
clinometer for measuring slope angle and determine tree height; compass for
determination of direction and aspect of the study plots; diameter tape for taking the
diameter of tree trunk at breast height; digital range finder which measures the
28
distance from the observer and the tree trunk, and Photo camera, water balance, plant
press and tripod (camera stand) were used in the field data collection. Laptop and
desktop computers were used for collecting secondary data from different sources via
the Google search engine.
Without the application of software data organization and processing, it was
unthinkable to accomplish the objectives of the study. Therefore, the following
softwares have been utilized.
1. Quantum GIS free software version 2.2 Valmiera for image manipulation
2. R free software version 3.0.1
3. Past free software
4. CAN-EYE free software
5. PC-ORD version 5.31
3.3. Methods
3.3.1. Study design
The study design varied based on the aspects to be studied. This study addressed:
1. Plant species richness, diversity, abundance and distribution;
2. Plant canopy structure and stratification;
3. Carbon storage in different land use types across the transect and its
association with climate variables; and
29
4. Leaf area index and its association with climate variables.
The study transect with a total area of 46 km2 (23 km × 2 km) was laid from Ageyo
(a village between Toba and Dembi towns) northwest of Jimma in the upper Didessa
River Basin. The transect was placed along an altitudinal gradient (1500–2300 m
a.s.l.) by the CHIESA Project. The project incorporated a range of different aspects
such as impact of climate change on plant diversity and distribution, carbon storage
and land use/land cover changes.
Thirty-one plots of 100 m × 100 m (major plot) were established in different land use
types such as croplands, forests (SFC, degraded natural forest (henceforth, DNF),
plantation forest), woodland and pasture along the transect. Stratified random
sampling technique was used for putting the sample plots across the transect. The
main land use types for the transect were identified from 2008 SPOT5 satellite image
and was confirmed during the reconnaisence survey which was conducted before the
start of actual data collection. The major plots were placed randomly in each land use
type and their four corners were geo-referenced and the vertices have been marked to
easily trace the plot boundary for the next visit. Five subplots of 20 m × 20 m were
laid at the corners and the center of each major plot (100 m × 100 m) (Figure 3). The
20 m × 20 m plot was marked at 10 m from the corners and at 5 m into the plot from
the 10 m mark from the corners (Figure 4). Small plots of 2 m by 2 m were
established at the center and four corners of the 20 m × 20 m plot.
30
Figure 3: Sampling design for AGC and vegetation data collection
Figure 4: Sampling design for LAI data collection
Diameter at breast height (DBH) and height of all woody species with DBH
10cm
were taken from the one hectare plot for the study of carbon storage. The 20 m × 20
m plots nested within the one hectare plots were used for collecting data on
31
occurrence of woody species, upward hemispherical images for determination of leaf
area index (LAI). Thirty-five plots of 20 m × 20 m were also used for collecting
cover abundance values from SFC system. This was used for grouping SFC plots
based on the similarities in the cover abundance values of the canopy trees. The 2 m
× 2 m plots were used for collecting occurrence data on herbaceous species. Another
29 plots of size 20 m × 20 m (400 m2) were placed in SFC of size less than one
hectare, along the transect. This was designed for collecting data on LAI (both above
and below the coffee canopy).
3.3.2. Data analysis
3.3.2.1. Land use/Land cover mapping
The LULC map was produced for the year 2008 using SPOT5 satellite image. The
SPOT5 images from 2008 used for land cover mapping were two satellite images
taken on path 134 and row 133, December 17, 2008. These were utilized (i) in
panchromatic mode at 2.5 m spatial resolution with 0.48–0.71 m wavelengths, and
(ii) in multispectral mode at 10 m spatial resolution with four bands: green (0.50–
0.59 m), red (0.61–0.68 m), near infrared (0.78–0.89 m), and short-wave infrared
(1.58–1.75 m).
In addition to the satellite images for the year 2008 LULC map, true-color aerial
photographs were used and acquired in a flight campaign arranged in October 2012
using University of Helsinki’s NIKON D3X camera and EnsoMOSAIC aerial
32
imaging system. The spatial resolution of the final aerial image was 0.5 meter.
Global Digital Elevation Model with 30 m resolution was used in SPOT5 satellite
image pre-processing from ASTER. We collected ground control points and training
areas from field work in order to get the LULC map.
The classification was based on information obtained from a set of similar pixles
which is referred to as Object Based Image Analysis (OBIA) and applied in two
levels: (i) Classification of the transect into different major land use types (ii)
separation of indigenous and exotic forests from the forest cover. The Nearest
Neighbor supervised classification method was used for classifying segmented image
objects and it was based on Land Cover Classification System. The final classified
image was at 2.5 m spatial resolution.
3.3.2.2. Species area curve
Before applying different analyses such as species diversity, similarity indices,
modelling species distribution and carbon storage, species area curve analysis was
applied using PC-ORD Version 5.31 for windows (McCune and Mefford, 2006).
This helped to evaluate the sampling effort in the species data collection.
3.3.2.3. Species diversity
Species richness, abundance, diversity, dominance and diversity profile were
determined using PAST computer software for windows (Hammer et al., 2001).
Percent and relative frequency of occurrence of each species in the entire study area
33
as well as in each land use types were determined using excel spread sheet. Species
richness in each plant growth form was also determined for each land use type.
The assessment of plant diversity in the study area was also one of the targets of this
study. Shannon-Wiener diversity index (Sannon and Wiener, 1949) was used to
determine species diversity from the quantitative data on abundance of each species
obtained from the sample plots and presence/absence data. Species diversity,
richness and evenness were evaluated using Shannon-Wiener Diversity index and
Simpson evenness index.
Where,
H = Shannon and Wiener diversity index,
Pi = the ratio of a species average to the total species average
Ln = the natural logarism to base e (loge)
The species evenness is calculated using the following formula
J = Hmax/lnS
Where,
J = the species evenness
Hmax = lnS, in which S stands for the number of species
34
3.3.2.4. Basal area
The DBH of all woody species in SFC and the DNF was measured at 1.3 m above
the ground. Importance value index for each tree species was calculated from the
basal area, frequency and abundance. The basal area for the woody species was
determined from the DBH measurement. The basal area was calculated by
multiplying diameter by pie. Basal area is calculated from the following formula.
Basal area (BA) =
* (diameter/2)2
3.3.2.5. Analysis of variance
One way analysis of variance was used to determine the difference between each
land use type in plant species richness, tree species abundance and basal area across
the transect. Pearson correlation was used to evaluate the linear relationships
between species richness in herb, shrub and tree growth forms and assess the
realtionships with environmental variables such as climate, edaphic and topography.
The linear relationships between tree species abundance and environmental variables
were also determined using Pearson correlation coefficient. Multiple regression
analysis was also conducted to determine the amount of variation in species
abundance and richness explained by edaphic, topographic and climatic variables.
3.3.2.6. Classification and grouping study plots
The study plots were classified based on species presence/absence data using PCORD Version 5.31 for windows (McCune and Mefford, 2006). The plots in SFC
35
system were classified and grouped using two-way cluster analysis. Group linkage
method and Sorensen (Bray Curtis) distance measurer were used to classify canopy
trees into groups based on their similarities in cover abundance values. Modified
Braun-Blanquet approach was used for estimation of cover abundance values.
Naming of the groups of canopy trees was based on the dominant species (highest
cover abundance value).
3.3.2.7. Vertical stratification
Vertical stratification in SFC and DNF was determined following the IUFRO
classification scheme (Lamprecht, 1989). According to this scheme, trees with >2/3
height of the tallest tree represent upper storey, trees with height between 1/3 and 2/3
of the tallest tree represent the middle storey and trees with height < 1/3 of the tallest
tree represent the lower storey. The richness, abundance, diversity, dominance of
plant species in each storey was analysed. The analysis was not done for plantation
forests, woodland, cropland and pasture due to absence of strata in the repective land
use types.
3.3.2.8. Carbon storage
Carbon storage in all land use types was organized in Microsoft excel spread sheet
and the data were taken into SPSS 16.0 for windows for analysis. The above ground
live biomass was calculated using the updated, non-destructive allometric equation
(Chave et al., 2014). The above ground live biomass was, therefore, calculated from
36
the following equation, where wood specific gravity, DBH and height of the woody
spcies are used as input data.
AGB = 0.0673*( D2H)0.976
Where “AGB” stands for above ground live biomass, “ ” for wood specific gravity,
“D” for diameter at breast height and “H” for height of the woody species.
Wood specific gravity was obtained at species-level from the Global Wood Density
database (Chave et al., 2009; Zanne et al., 2009).The above ground live carbon
(AGC) was estimated at 50% of the AGB (Chave et al. 2014). Shapiro-Wilk
normality test confirmed that the data distributed normally and hence parametric test
was used.
Data exploration, descriptive statistics and analysis of variance (ANOVA) were used
to evaluate the distribution of data, carbon storage in each land use types and
comparing different land use types in terms of carbon storage. Pearson correlation
was used to determine the linear relationship between species richness (herb, tree and
shrub) and AGC; between tree species abundance and AGC. Multiple regressions
were conducted to determine the amount of variation in AGC explained by tree
species richness and abundance combined (explanatory variables). Pearson
correlation was also used to determine the relationships between AGC and climate
variables, AGC and edaphic variables, AGC and topographic variables. Multiple
37
regression analysis was also conducted to determine the amount of variation in AGC
explained by potential evapotranspiration, soil pH, CEC and sand (edaphic factors).
3.3.2.9. Leaf area index (LAI)
The upward hemispherical images taken from the plots were organized in excel
spreadsheet in the appropriate format for thresholding in MATLAB computer
software. MATLAB was used for thresholding the images and CAN-EYE computer
software was used for converting the images into digital numbers. The CAN-EYE
produces true leaf area index and effective leaf area index based on the presence and
/or absence of vegetation clumping. Both data sets were tested for normality.
Effective leaf area index was excluded from the analysis due to failure to satisfy the
assumption of normal distribution even after transformations.
LAI were produced from two versions of Can_Eye (V5.1 and V6.1) based on the
differences in the regularization term that imposes constraints for the improvement of
LAI. In Can_Eye V5.1, the regularization term uses average leaf angle which asumes
the average leaf angle close to 600, while Can_Eye V6.1 uses the retrieved plant area
index which is close to the one retrieved from the zenital angle of 570.
The linear relationships between above ground live carbon storage (AGC) and LAI,
species richness and LAI, tree species abundance and LAI were analyzed using Rfree software version 3.0.1. Analysis of variance was also used to determine the
38
difference among different land use types in LAI. Multiple regression analysis was
conducted with the explanatory variables having significant linear relationship with
the LAI. This was used to know the variation in LAI explained by the independent
variables. The difference between the LAI taken above the coffee canopy and below
the coffee canopy was tested using paired sample t-test.
3.3.2.10. Species Distribution Model
Maxent entropy (Phillips et al., 2006) was used for modelling the distributions of
five plant species encountered in the study transect. The presence records (longitude
and latitude) of species were used as input for the model. Five plant species were
selected for modelling their distribution across Ethiopia. The selection was based
mainly on: (1) occurrence in the study transect (Jimma Highlands), (2) the species
should be indigenous, (3) have economic /medicinal values among the community,
(4) their ecological role in the ecosystem. There are documented economic and
medicinal values of Cordia africana (Dawit Abebe and Ahadu Ayehu, 1993; Fichtl
and Admasu Adi, 1994; Legesse Negash, 1995; Riedl and Edwards, 2006; World
agroforestry, 2009; Sarah Tewoldeberhan et al., 2013), (Millettia ferruginea (Thulin,
1989; Fichtl and Admassu Adi, 1994; Berhanu Alemu et al., 2013), Schefflera
abyssinica (Fichtl and Admassu Adi, 1994), Phytolacca dodecandra (Aklilu Lemma
et al., 1972; Polhill, 2000; Hailu Tadeg, 2005). Acacia abyssinica, Millettia
ferruginea and Cordia africana are selected for shade provision by coffee growers
(Diriba Muleta et al., 2011). Based on the above three points, Acacia abyssinica,
Cordia africana, Millettia ferruginea, Phytolacca dodecandra and Schefflera
39
abyssinica were selected for modelling, which are decribed below. Climate data were
obtained from AFRICLIM (Platts et al., 2014).
Acacia abyssinica Hochst.ex Benth. (Fabaceae)
Cordia africana Lam. (Boraginaceae)
Millettia ferruginea (Hochst.) Bak. (Fabaceae)
Phytollaca dodecandra L’Herit. (Phytolaccaeae)
Schefflera abyssinica (Hochst. ex A. Rich.) Harms (Araliaceae)
3.3.2.10.1. Model building
The occurrence data for each of the five species were divided into training and test
data. Seventy five percent of the occurrence data were used for training; while 25%
were used for testing (referred to as random test percentage) the performance of the
model. This setting allows us to set aside 25% of the presence data to be used to
evaluate the model’s performance. In the absence of test data (25% in this case), the
model uses the training data to evaluate itself and as a result the model will be
inflated. The use of test data, therefore; avoids such inflated model outputs. Of the
three replicate run types that Maxent allows for using, the subsample option was
chosen. Ten replicates were made for each species and the average over the ten
model outputs was taken for interpretation and discussion.
40
The species have different number of occurrence localities- Acacia abyssinica (n =
92), Cordia africana (n = 108), Millettia ferruginea (n = 75), Phytolacca dodecandra
(n = 113) and Schefflera abyssinica (n = 104). Overlapping data were avoided and
only those data points which were not overlapping were used in the modelling. For
Acacia abyssinica, 40 for training and 13 for testing were used, for Cordia africana
47 for training and 15 for testing, for Millettia ferruginea, 49 for training and 16 for
testing, for Phytolacca dodecandra, 36 for training and 12 for testing, and for
Schefflera abyssinica, 33 for training and 11 for testing.
In MaxEnt, it is possible to run a model several times and then take the average of
the results from all models (Philips et al., 2006). In this study, the model was
executed to replicate 10 times and then the average was taken. This result was
combined with the result from the 25% test data for evaluating the model
performance. Executing multiple runs also provide a way to measure the amount of
variability in the model.
In the default setting, the number of iterations (convergence) was set to 500. In this
study, the number of iterations was increased to 5000 to allow adequate time for the
model to converge. This helps to avoid either over prediction or under prediction.
The application of regularization avoids or reduces model over-fit. The default value
of 1 was accepted for regularisation for this study.
41
CHAPTER FOUR
4. Results
4.1. Land Use/Land Cover
The 2008 land cover map (Figure 5) showed five main land cover categories in the
transect: these are pasture, woodland, cropland, natural and plantation forests. From
the field observation, the natural forests without coffee are found towards the end of
the transect, while the natural forests upto around 2000 m elevation were occupied
by SFC. Therefore, the natural forests were further classified into those with coffee
and those without coffee. Hence, the main LULC types of this study area were:
1. Cropland
2. Pasture
3. SFCs
4. Woodland
5. DNFs
6. Plantation forests
Plant species richness, diversity, abundance, vegetation structure, carbon storage
and leaf area indices were all affected by LULC types and hence have been
addressed separately.
42
Figure 5: Land use/cover across the study transect in the Jimma Highlands for the
year 2008
4.2. Plant Species Richness and Diversity
4.2.1. Species richness in each land use type
4.2.1.1. Species area curve
The species area curve in all land use types levelled off after some sample plots were
surveyed. This showed that the sampling effort was exhaustive (Figure 6). The
number of sample plots for plant species data collection (400 m2) was 35 for SFC, 35
for cropland, 25 for pasture, 20 for woodland, 20 for DNF and 20 for plantation
forests. Species richness and growth form distributions in each land use types are
presented below.
43
Cropland
Woodland
SFC
DNF
Pasture
Plantation
Figure 6: Species area curve for all land use types across the transect in the Jimma
Highlands
4.2.1.2. Species richness in DNF
The natural forest in the study transect is surrounded by villages and is used as
common pool for firewood, materials for building houses, timber and other nontimber forest products. As a result, it is highly degraded. This natural forest is
composed of 114 (28.25 ha-1) species of plants that are distributed among 103 genera
and 50 families (Appendix 2). The top 13 families were composed of about 56% of
44
the species in the natural forest and the remaining 37 families together contributed
about 44% of the total species composition. Asteraceae is the most species rich
family (11 species) in the DNF followed by Rubiaceae and Fabaceae each of them
with eight and seven species respectively. Twenty-five families were represented by
one species each. The number of herbaceous and tree species in the DNF is almost
the same. There are slightly more herbs than trees, while the lianas have the lowest
species diversity compared to the remaining three growth forms (Table 1).
Table 1: Growth form distribution of plant species in DNF
Growth form
Species richness Species
%Composition
richness ha-1
Herb
36
9.0
31.58
Liana
11
2.75
9.65
Shrub
32
8.0
28.07
Tree
35
8.75
30.70
4.2.1.3. Species richness in woodlands
Woodland is a stand of trees with the height of 8-20 m and a canopy cover of at least
40% of the surface (White, 1983). In this study transect, woodlands are composed of
136 (34 ha-1) plant species (Appendix 3) distributed among 105 genera and 44
families. Asteraceae and Fabaceae were equally species-rich families; each of them
containing 18 species together contributing to 26.48% of the total species
composition in the woodland. Lamiaceae and Euphorbiaceae follow with 11(8.09%)
45
and 9 (6.62%) species respectively. The families with one representative species in
the woodland are 22, together contributing about 16% of the species composition in
the woodland. Most of the plants in the woodland are herbaceous followed by shrub
species. Liana growth form is the least species rich compared to all the rest (Table 2).
Table 2: Growth form distribution of plant species in woodland
Growth form
Species richness
Species
%Composition
richness ha-1
Herb
56
14
41.18
Liana
9
2.25
6.62
Shrub
39
9.75
28.68
Tree
32
8
23.53
4.2.1.4. Species richness in cropland
Cropland is composed of 91(13 ha-1) plant species occurring in 81 genera and 37
families (Appendix 4). Species richness is the highest in the family Asteraceae
compared to other plant families in the cropland along the study transect. Fabaceae is
the second species rich family followed by Euphorbiaceae, Malvaceae and Poacaeae.
Most families are composed of one species. Out of the four plant growth forms,
herbs are the most species rich group, while liana is the least growth form in species
richness (Table 3). When woodland, degraded natural or coffee forests are converted
to plots of annual crops, the trees, shrubs and lianas are removed. The tree species
46
are scattered in farm land and their number per hectare is about three species on
average. Liana growth form is the most affected in agricultural field.
Table 3: Growth form distribution of plant species in cropland
Growth form
Species richness
Species
%Composition
richness ha-1
Herb
50
7.14
54.95
Liana
2
0.29
2.20
Shrub
17
2.43
18.68
Tree
22
3.14
24.18
4.2.1.5. Species richness in SFC
SFC ranked third in plant species richness per hectare and relatively the richest land
use type in plant family composition in the study transect. SFC is composed of 152
(52.96%) of plant species (Appendix 5) spread among 130 genera and 55 families.
Asteraceae and Fabaceae are the most species rich families. Euphorbiaceae ranked
third in species richness followed by Acanthaceae, Malvaceae and Rubiaceae each of
them with equal number of species. Twenty-four families in the SFC were
represented only by one species each.
Variation was also observed in plant growth forms. Most of the species in the SFC
are herbs, while liana is the least growth form in species richness (Table 4). Liana is
the growth form which is affected most in SFC. The most frequently occurring plant
47
species in SFC are Coffea arabica, Celts africana and Ehretia cymosa. All of these
plants were recorded from all study plots of SFC. Plant species like Achyranthes
aspera, Albizia gummifera, Cordia africana, Croton macrostachyus, Desmodium
repandum, Vepris dainellii, Vernonia amygdalina and Vernonia auriculifera were
the second most frequent species in the SFC. The most frequently occurring species
belong to the tree growth form. Of the most frequent species, Desmodium repandum
and Achyranthes aspera were herbaceous species and V. auriculifera is a shrub,
while all the remaining are trees.
Table 4: Growth form distribution of plant species in SFC
Species
Growth form
Species richness
richness ha-1
% Composition
Herb
67
9.57
44.08
Liana
3
0.34
1.97
Shrub
34
4.86
22.37
Tree
48
6.86
31.58
4.2.1.6. Species richness in pasture
There are 90 genera and 39 plant families with 113 plant species occurring in
pastures along the study transect (Appendix 6). Asteraceae is the most species rich
family in the pasture; Fabaceae and Poaceae follow. Euphorbiaceae, Lamiaceae and
Rubiaceae rank fourth species rich families with equal number of species. About half
of the families have
2 species, while the rest are represented by one species each.
48
Lianas have fewer occurrences when compared with the remaining plant growth
forms. Herb growth form is the most species rich group (Table 5).
Table 5: Growth form distribution of plant species in pasture
Species
richness ha-1
% Composition
Growth form
Species richness
Herb
51
10.2
45.13
Liana
7
1.4
6.19
Shrub
32
6.4
28.32
Tree
23
4.6
20.35
4.2.1.7. Species richness in plantation forests
There are 79 plant species occurring in the plantation forests (Appendix 7). The
highest growth form in species richness in the plantation forests was tree and the
least was liana (Table 6). The plots of monoculture plantations include Cupressus
lusitanica, Eucalyptus camaldulensis, Grevillea robusta and Pinus patula. The
number of trees in the monoculture plantations showed the regenerating capacity of
indigenous tree species under the canopy of plantations of exotic species.
49
Table 6: Growth form distribution of plant species in plantation forest
Species
Growth form
Species richness
richness ha-1
% composition
Herb
27
6.75
34.18
Liana
3
0.75
3.80
Shrub
21
5.25
26.58
Tree
28
7
35.44
4.2.2. Plant species across the transect
4.2.2.1. Plant species richness
From the study conducted in all land use types along the established study transect,
in the Jimma Highland, 287 species of plants were collected and identified. These
287 plant species have been distributed among 220 genera and 82 families
(Appendix 8). The first 12 families contributed more to the species composition in
the study transect than the remaining 70 families (Table 7). Three families
(Cupressaceae, Proteaceae and Pinaceae) are represented by single exotic species
each.
50
Table 7: Species rich families across the study transect and their percent composition
Family
Richness
%Composition
Asteraceae
33
11.58
Fabaceae
25
8.77
Euphorbiaceae
14
4.91
Lamiaceae
14
4.91
Rubiaceae
14
4.91
Poaceae
13
4.59
Acanthaceae
10
3.51
Malvaceae
9
3.16
Solanaceae
8
2.81
Moraceae
7
2.46
Amaranthaceae
6
2.11
Ranunculaceae
6
2.11
Remaining 70 families
128
44.6
Species richness in different land use types vary. The species composition in all land
use types were compared using X2 statistics. The test showed that the species
composition is affected by land use type (Xi2 (5) = 32.258, the critical value at p =
0.05 significant level for 5° of freedom is 11.07) (Table 8). The calculated X2 is
greater than the critical value confirming that the land use types affected plant
species richness.
51
114
136
91
113
79
685
Absent
135
173
151
196
173
208
1036
Total
287
287
287
287
287
287
1722
Frequency
0.53
0.40
0.47
0.32
0.39
0.28
0.40
114.17
114.17
114.17
114.17
114.17
114.17
172.83
172.83
172.83
172.83
172.83
172.83
Ob-exp
37.83
-0.17
21.83
-23.17
-1.17
-35.17
(Ob-exp)2
1431.361 0.027778 476.6944 536.6944 1.361111 1236.694
Total
woodland
152
pasture
DNF
Present
Cropland
SFC
plantation
Table 8: X2-test for species composition in different land use types
Expected
Presence
Expected
Absence
(Ob-exp)2/exp
12.5375
0.0002
4.1754
4.7010
0.0119
10.8324 Xi2(5)=32.258
The number of plant species per hectare increased from highly modified land use
types to less modified ones (Figure 7). The land use types in decreasing order of
plant species richness per hectare are woodland, DNF, SFC, pastureland,
monoculture plantation of exotic species and cropland of annual crops (Figure 7).
52
Figure 7: Plant species richness per hectare in different land use types across the
transect (WLD = woodland, DNF, SFC = semi-forest coffee, PR = pasture, PF =
plantation forest, CLD = cropland)
4.2.2.2. Woody species richness and diversity
Woody species richness, abundance and diversity vary from land use to land use. The
highest woody species richness was recorded from SFC followed by DNF and
woodland.The least woody species with dbh 10 cm was obtained from cropland
(Table 9). The land use types also vary in woody species abundance. Plantation
forest is characterized by the highest density of trees. The number of stems per
hectare is 236 in DNF, while it is 129.7 ha-1 in SFC, whereas the least abundance was
recorded from cropland with 6.7 ha-1.
53
Table 9: Species richness, abundance, dominance, diversity and evenness in different
land use types
SFC
Richness
Cropland Woodland Pasture DNF
Plantation
44
9
27
14
32
13
Abundance
908
47
464
31
944
3193
Dominance
0.10
0.18
0.17
0.11
0.08
0.27
Shannon
2.75
1.90
2.18
2.43
2.82
1.47
Evenness
0.35
0.74
0.33
0.81
0.53
0.33
4.2.2.3. Plant growth form distribution
Plant species collected from the sample plots along the study transect were
distributed among four major plant growth forms (herb, liana, shrub and tree). The
most species rich plant growth form was herb, while the growth form with least
number of species was liana.
4.2.2.4. Frequency of occurrence of species
The 287 plant species distributed in the entire study area vary in frequency of
occurrence along the transect (Appendix 8). Nine plant species occurred in about
50% of the study plots. These are Acacia abyssinica, Achyranthes aspera,
Agerantum conyzoides, Albizia gummifera, Bidens pilosa, Cordia africana, Croton
macrostachyus, Maesa lanceolata and Vernonia auriculifera (Appendix 8). Five of
these plants were trees and three of them were herbs and there was one shrub.
54
Eight plant species occurred only once and these are Cupressus lusitanica,
Eucalyptus camaldulensis, Grevillea robusta, Kosteletzkya begoniifolia, Nuxia
congesta, Pinus patula, Schrebera alata and Sesbania sesban. Cupressus lusitanica,
Eucalyptus camaldulensis, Grevillea robusta, Pinus patula and Sesbania sesban
were exotic species and the first four are in plantations. Sesbania sesban has been
used as shade tree in home gardens and was seen in the wild escaping from the home
gardens. Kosteletzkya begoniifolia, Nuxia congesta and Schrebera alata are
indigenous species and have rare occurrence in the study area.
4.2.3. Climate variables and edaphic factors against species richness
The richness of herbaceous species showed significant linear relationships with mean
annual rainfall, mean annual temperature, maximum temperature warmest month,
mean annual temperature, annual moisture index, potential evapotranspiration and
soil pH (Appendix 9). The linear relationships of the herbaceous species richness
with rainfall wettest month, silt, soil bulk density, cation exchange capacity, sand and
clay was not significant. The linear relationship of the shrub species was only
significant with the soil clay, but its relation was not significant with other variables
(Appendix 9). Tree species richness was only significant with sand and clay
(Appendix 9). The relation with elevation and other variables was not significant.
4.2.3.1. Multiple regression analysis
The regression analysis made with the variables showed significant relationship with
herbaceous species richness. Most of the variables were excluded due to collinearity
55
and only two variables (mean annual temperature and soil pH) were taken for the
analysis. The two variables combined together have significantly explained 20.5%
(R2 = 0.205, R2adj = 0.148, SE = 7.85) of the variation in herbaceous species richness.
The contribution of each separate variable to the model was not significant (Table
10).
The two soil properties having significant relationships with the tree species richness
were tested for collinearity prior to conducting the multiple regression analysis. The
variance inflation factor for both variables was <5 (Table 11). The two variables
combined together have explained about 20% (R2 = 0.196, R2adj = 0.138, SE = 7.12)
of the variation in tree species richness. The soil clay explained about 17% of the
variation in shrub richness.
Table 10: Contribution of mean annual temperature and pH to the regression analysis
of herb richness
Unstand_Coef
Stand_Coef
B
SE
Beta
-71.572
36.082
bio1
0.391
0.353
pH
0.346
1.336
Model
Constant
Dependent variable: Herb
56
t
p
VIF
-1.984
0.057
0.375
1.106
0.278
4.04
0.088
0.259
0.798
4.04
Table 11: Contribution of each explanatory variable to the regression analysis of tree
species richness
Unstand_Coef
Stand_Coef
Model
B
SE
Beta
Constant
20. 647
72.931
Sand
0.958
1.164
Clay
-0.907
0.959
T
p
VIF
0.283
0.779
0.219
0.823
0.417
2.457
-0.251
-0.945
0.353
2.457
Dependent variable: Tree
4.2.4. Classification of study plots on the bases of species
presence/absence
The sample plots were clustered into three (with about 12.5% similarity) using the
species occurrence data (presence/absence) across all land use types in the study
transect (Figure 8). Jaccard similarity index was the measure of similarity used to
group the sample plots. The plots from different land use types were classified based
on their similarities in species composition.
Group I: Includes all forest types (DNF, SFC, plantation forest) and some plots of
woodland in the transect. The DNFs are represented by 20 sample plots, SFC by 35
plots, plantation forests by 20 plots and the woodlands by 10 plots. All plots in the
DNF were composed of indigenous plant species such as Apodytes dimidiata, Croton
macrostachyus, Galiniera saxifraga, Polyscias fulva, Prunus africana, Schefflera
57
abyssinica and Syzygium guineense, while the plantation forests are mainly
composed of Cupressus lusitanica, Eucalyptus camaldulensis, Grevillea robusta and
Pinus patula, which are all exotic species. The SFC were composed of the trees
retained on the plot by the coffee growers for the purpose of shade provision for the
coffee shrubs beneath. All the shade trees in the SFC were indigenous species. The
most important trees in the SFC were Albizia gummifera, Acacia abyssinica, Celtis
africana, Cordia africana, Croton macrostachyus and Millettia ferruginea. The
woodland plots in this group are also compsed of indigenous species such as Acacia
abyssinica, Combretum molle, Entada abyssinica and Terminalia schimperiana.
Group II: Group II includes all plots of pastureland and 10 plots from the
woodlands. All the plant species in pasture and woodlands in this group were
indigenous. The trees found doted in the pastureland include Ficus vasta, Sapium
ellipticum, Syzygium guineense, Prunus africana and Croton macrostachyus.
Group III: All plots of croplands were grouped together due to their similarity in
species composition.The tree species dotted in the croplands include Acacia
abyssinica, Cordia africana and Croton macrostachyus and all of them are
indigenous species.
58
Figure 8: Cluster analysis based on species presence/absence (P1–4 = DNF, P5–8 =
woodland, P9–15 = Cropland, P16–22 = SFC, P23–27 = Pastureland, P28–31 =
Plantation forest)
4.2.5. Groups of canopy trees in SFC
The canopy trees in the SFC were classified into four groups based on the cover
abundance values taken from 35 sample plots along the transect (Figure 9 and
Appendix 10). The group linkage method and Sorensen (Bray Curtis) distance
measurer in two ways cluster analysis was used for grouping the canopy trees into
groups. The two-way cluster analysis shows the grouping of sample plots and the
species cluster showing the plots in which the species has occurred. Based on the
calculated average value of cover abundance the coffee plots were classified into
59
four distinctive groups. Naming of the groups was based on the dominant species
(highest cover abundance value).
Croton macrostachyus and Albizia gummifera (Group I)
This group of coffee shade trees was made up of five sample plots and was named
after two dominant tree species in the group (Croton macrostachyus, relative cover
abundance = 8 and Albizia gummifera, relative cover abundance = 7.4). Cordia
africana, Ehreta cymosa, Allophylus abyssinicus, Schefflera abyssinica, Prunus
africana, Diospyros abyssinica, Ficus sur, Bersama abyssinica, Apodytes dimidiata,
Celtis africana, Galiniera saxifraga, Vernonia amygdalina and Pittosporum
viridiflorum are other species in decreasing order of average cover abundance values.
Cordia africana and Acacia abyssinica (Group II)
This group of shade trees in the SFC was composed of 15 sample plots and it was
named after two dominant canopy trees (Cordia africana, cover abundance = 6.7 and
Acacia abyssinica, cover abundance = 5.3). Other plant species in this group include
Albizia gummifera, Croton macrostachyus, Celtis africana, Ficus thonningi, Vepris
dainellii, Clausena anisata, Vernonia amygdalina, Vernonia auriculifera, Sapium
ellipticum, Ehreta cymosa, Vangueria apiculata, Maesa lanceolata, Prunus africana,
Allophylus abyssinicus, Bridelia micrantha, Ficus sur, Ficus vasta, Polyscias fulva,
Syzygium guineense, Diospyros abyssinica, Podocarpus falcatus, Schrebera alata,
Trichilia dregeana, Dracaena steudneri and Grewia ferruginea.
60
Millettia ferruginea and Acacia abyssinica (Group III)
This group was named by two coffee shade trees (Millettia ferruginea and Acacia
abyssinica) with 6.6 and 4.4 cover abundance values respectively. This group is
composed of five sample plots. Included in this group are Croton macrostachyus,
Cordia africana, Albizia gummifera, Bersama abyssinica, Prunus africana, Sapium
ellipticum, Schefflera abyssinica, Ekebergia capensis, Polyscias fulva, Maesa
lanceolata and Maytenus arbutifolia
Croton macrostahyus and Diospyros abyssinica (Group IV)
Group IV is compsed of 10 sample plots and was named by two tree species (Croton
macrostachyus and Diospyros abyssinica) with relatively high average abundance
values (6 and 5.7 respectively) than any plant species in the group. The other plant
species in this group in decreasing order of average abundance values are Millettia
ferruginea, Cordia africana, Ficus mucuso, Dracaena steudneri, Celtis africana,
Albizia gummifera, Ficus sur, Trilepisium madagascariense, Ficus vasta, Vepris
dainellii, Chionanthus mildbraedii, Ficus thonningi, Sapium ellipticum, Trichilia
dregeana, Ehreta cymosa, Rothmania urcelliformis, Bersama abyssinica, Flacourtia
indica, Prunus africana, Syzygium guineense, Vangueria apiculata, Terminalia
schimperiana, Phoenix reclinata and Maesa lanceolata.
61
II
III
IV
I
Figure 9: Group of canopy trees in the SFC in the study transect in the Jimma
Highlands (I, II, III and IV represent group 1-4 respectively)
High species richness was observed in group II and IV compared to the species
richness in gropu I and III (Table 12). Group IV was with the highest species
diversity, while group III was the least in species diversity (Table 12). Group I was
the most dominant group compared to groups.II, III and IV.
62
Table 12: Species richness, abundance, dominance, diversity and evenness in
different groups of SFC
Group_I Group_II
Group_III
Group_IV
Richness
15
27
13
26
Abundance
48.40
41.33
27.20
52.30
Dominance
0.10
0.08
0.13
0.07
Shannon_H
2.47
2.79
2.27
2.92
Evenness
0.79
0.60
0.74
0.71
4.2.6. Vegetation structure
4.2.6.1. Land use type and plant species abundance
The plant species abundance varies from land use to land use type (Figure 10). It
shows a decline in mean abundance of plant species from plantation forest to
pastureland. The mean abundance for plantation forest was the highest, while pasture
Mean abundance of Plant species
was the least in tree species abundance (Table 16).
Figure 10: Box plot of species abundance in different land use types (1 = plantation
forest, 2 = DNF, 3 = SFC, 4 = woodland, 5 = cropland, 6 = pasture)
63
Analysis of variance test showed significant difference within land use types in plant
species abundance (Table 13). A post hoc test showed significant differences
between SFC and cropland, pasture, plantation forests, but it did not show any
significant variation from DNF and woodland (Table 14). Cropland showed a strong
significant relationship with DNF, plantation forests, woodland, but did not show
significant difference from pasture (Table 14). As with the cropland, the variation
between DNF and pasture was also significant and pasture was also different from
plantation forests and with woodland. A significant statistical difference was also
shown between plantation forest and woodland wheras three homogenous groups
were also produced (Table 15).
Table 13: Difference in species abundance (4th root_abundance) across different
land use types
SS
df
MS
F
P
3.483
5
0.697
28.824
0.00
0.556
23
0.024
4.038
28
Between
Groups
Within
Groups
Total
64
Table 14: Pairwise comparison in species abundance between different land use
types (LB = lower bound, UB = upper bound)
UB
Land use 1
LB
95% CI
Land use 2
MD1 & 2
SE
P
Cropland
0.54
0.09
0.00
0.28
0.81
DNF
-0.18
0.10
0.47
-0.48
0.12
Pasture
0.59
0.09
0.00
0.30
0.87
Plantation
-0.42
0.11
0.01
-0.75
-0.09
Woodland
0.05
0.10
1.00
-0.25
0.35
DNF
-0.72
0.10
0.00
-1.03
-0.41
Pasture
0.04
0.09
1.00
-0.25
0.33
Plantation
-0.96
0.11
0.00
-1.30
-0.62
Woodland
-0.49
0.10
0.00
-0.81
-0.18
Pasture
0.77
0.10
0.00
0.44
1.09
Plantation
-0.24
0.12
0.36
-0.61
0.13
Woodland
0.23
0.11
0.33
-0.11
0.57
Plantation
-1.00
0.11
0.00
-1.36
-0.65
Pasture
Woodland
-0.54
0.10
0.00
-0.86
-0.21
Plantation
Woodland
0.47
0.12
0.01
0.10
0.84
SFC
Cropland
DNF
65
Table 15: Homogeneous subsets among land use types in tree species abundance
Tukey HSD
Subset for alpha = 0.05
land use type
N
1
2
3
Pasture
5
1.2249
Cropland
6
1.2673
Woodland
4
1.762
SFC
7
1.8116
DNF
4
1.99
Plantation
3
P
1.99
2.2294
0.998
0.279
0.233
Table 16: Mean abundance of tree species in different land use types
land use type
Mean
N
Std.
135.86
7
71.17
7.33
6
4.37
257.5
4
99.79
Pasture
6.2
5
4.49
Plantation forest
751
3
508.69
Woodland
122.25
4
94.89
Total
165.45
29
264.43
SFC
Cropland
DNF
66
4.2.6.2. Climate variables and edaphic factors against species abundance
Acorrelation was conducted to evaluate the relationship between plant species
abundance, some climate and edaphic variables and elevation (Appendix 11). Of all
the climate, edaphic and topographic variables, elevation, mean annual rainfall,,
maximum temperature warmest month, mean annual temperature, annual moisture
index, potential evapotranspiration, pH, cation exchange capacity and sand showed
significant linear relationship with the species abundance. Most of these variables
were excluded from multiple regression analysis due to collinearity effect among
themselves. Only four variables (PET, pH, CEC and sand) with variance inflation
factor (VIF) < 10 were taken into the model (Table 17). Combined together, the four
variables significantly explained about 47% of the variation in species abundance (R2
= 0.47, R2adj = 0.38, SE = 0.30, F = 5.29, P = 0.003). The separate contribution of
sand to the model was statistically significant (P = 0.01), while the remaing three
variables have no significant contribution separately (Table 17).
Table 17: Contribution of each predictor variable to the model and VIF value for
each explanatory variable, (PET = Potential evapotanspiration, CEC = cation
exchange capacity, BLD = bulk density), dependent variable: abundance
Unstand_Coef
Stand_Coef
Model
Constant
PET
PH
CEC
Sand
B
4.73
0.00
-0.05
0.06
0.107
SE
1.97
0.00
0.06
0.05
0.04
Beta
-0.46
-0.32
0.34
0.51
67
t
2.40
-1.17
-0.95
1.30
2.78
P
0.02
0.25
0.35
0.20
0.01
VIF
6.91
5.03
3.15
1.52
4.2.6.3. Basal area across land use types
Relatively, larger basal area was calculated for plantation forests followed by DNF
and SFC. The minimum basal area was recorded for pasture and cropland (Figure
11). Among the species occurring across the transect, Albizia gummifera was the top
tree species in basal area contribution in the SFC (Table 18). Croton macrostachyus,
Ficus mucuso and Cordia africana were second, third and fourth with the basal area
contribution in the SFC (Table 18). Ficus sur, Apodytes dimidiata, Schefflera
abyssinica, Syzygium guineense, Albizia gummifera and Celtis africana have
contributed >1 basal area ha-1 in the DNFs (Table 19).
One way ANOVA test showed significant mean difference in basal area among the
land use types (Table 20). Multiple comparison tests showed significant differences
between DNF and pasture, woodland, croplandand plantation forest; between SFC
and pasture, woodland, cropland and plantation forest; between pasture and
plantation forest; between woodland and plantation forest; between cropland and
plantation forest. Significant statistical difference was not observed between DNF
and SFC, between pasture and woodland, pasture and cropland and between
woodland and cropland.
68
Figure 11: Tree species basal area in each land use type across the transect in the
Jimma Highlands (PF = Plantation Forest, DNF = Degraded natural forest, SFC =
Semi-forest coffee, WLD = Woodland, CLD = Cropland, PR = Pasture)
69
Table 18: Basal area contribution of tree species in SFC
Species
BA (Total)
BA ha-1
%BA
Albizia gummifera
25.19
3.60
20.27
Croton macrostachyus
19.45
2.78
15.65
Ficus mucuso
15.98
2.28
12.85
Cordia africana
11.59
1.66
9.32
Dracaena steudneri
8.17
1.17
6.57
Acacia abyssinica
7.47
1.07
6.01
Millettia ferruginea
5.19
0.74
4.18
Ficus sur
4.35
0.62
3.50
Ficus vasta
4.28
0.61
3.44
Ehretia cymosa
3.02
0.43
2.43
Sapium ellipticum
2.72
0.39
2.19
Ficus thonningii
2.64
0.38
2.13
Celtis africana
2.61
0.37
2.10
Diospyros abyssinica
2.05
0.29
1.65
70
Table 19: Basal area contribution of tree species in DNF
DNF
BA (Total)
BA ha-1
% BA
Ficus sur
19.19
4.80
23.42
Apodytes dimidiata
12.02
3.01
14.67
Schefflera abyssinica
8.89
2.22
10.85
Syzygium guineense
7.72
1.93
9.42
Albizia gummifera
5.87
1.47
7.16
Celtis africana
4.55
1.14
5.56
Macaranga capensis
3.57
0.89
4.36
Olea welwitschii
2.33
0.58
2.84
Chionanthus mildbraedi
2.26
0.57
2.76
Millettia ferruginea
2.24
0.56
2.74
Table 20: The difference of land use types in basal area
SS
df
MS
F
P
Between Groups
6067.48
5
1213.50
29.81
0.00
Within Groups
936.30
23
40.71
Total
7003.77
28
4.2.7. Vertical stratification in SFC and DNF
Following the IUFRO classification scheme, vertical structure of trees was classified
in the SFC and DNF into three layers. Height of the tallest tree was used to decide on
the cut points for each layer. In the SFC, Albizia gummifera was the tallest tree
71
(height = 40 m), while the tallest tree in the DNF was Apodytes dimidiata (height =
35 m). The vertical stratification for the SFC and DNF was addressed one after the
other. Based on the height of Albizia gummifera, the SFC was classified into lower,
middle and upper storeys. The abundance per hectare of tree species increased from
the lower to the middle and decreased from the middle to the upper storey (Figure
12) as listed in Table 21. The six most abundant tree species are indicated in Figure
13. Some tree species such as Albizia gummifera, Croton macrostachyus and Celtis
africana have representative trees in all the three storeys.
Figure 12: Tree species richness and abundance in the vertical stratification of
canopy trees in SFC (lower < 13.33m, middle = 13.33–26.67m, upper > 26.67m)
72
Figure 13: Abundance of six major canopy trees in the vertical stratification of
canopy trees in SFC (lower = <13.33m, middle = 13.33–26.67m, upper = >26.67m)
The emergent tree species in the DNF was Apodytes dimidiata. The middle storey
was relatively with more number of species compared to the lower and upper storeys
(Figure 14). Tree species are most abundant in the middle storey compared to the
lower and upper storeys (Figure 14) and are listed in Table 23. The six most
abundant tree species in the DNF are indicated in (Figure 15). Some tree species such
as Albizia gummifera, Apodytes dimidiata, Croton macrostachyus, Millettia
ferruginea, Prunus africana, Schefflera abyssinica and Syzygium guineense were
represented in all the three storeys.
73
Figure 14: Tree species abundance and richness in the lower, middle and upper
storeys of the canopy trees in DNFs (lower <11.67m, middle = 11.67–23.33m, upper
= >23.33m)
Figure 15: Abundance of six most important canopy trees in the vertical stratification
of DNF (lower <11.67m, middle = 11.67–23.33m, upper = >23.33m)
74
Table 21: Tree species abundance per hectare in SFC
Abundance
Species name (in SFC)
Lower Middle
Upper
Sum
Croton macrostachyus
33
149
9
191
27.29
Albizia gummifera
43
49
16
108
15.43
Ehretia cymosa
69
35
0
104
14.86
Cordia africana
38
63
0
101
14.43
Millettia ferruginea
43
35
0
78
11.14
Acacia abyssinica
26
30
0
56
8
Celtis africana
6
23
1
30
4.29
Vepris dainellii
25
5
0
30
4.29
Dracaena steudneri
7
19
0
26
3.71
Ficus mucuso
0
22
0
22
3.14
Clausena anisata
14
0
0
14
2
Diospyros abyssinica
4
10
0
14
2
Vernonia amygdalina
11
1
0
12
1.71
Bersama abyssinica
10
1
0
11
1.57
Ficus sur
1
10
0
11
1.57
Allophylus abyssinicus
1
9
0
10
1.43
Ficus thonningii
2
7
0
9
1.29
Vernonia auriculifera
8
0
0
8
1.14
Chionanthus mildbraedii
7
0
0
7
1
Maesa lanceolata
7
0
0
7
1
Sapium ellipticum
1
6
0
7
1
Vangueria apiculata
7
0
0
7
1
Prunus africana
0
5
0
5
0.71
Trichilia dregeana
1
4
0
5
0.71
Ficus vasta
0
4
0
4
0.57
Schefflera abyssinica
1
3
0
4
0.57
0
4
0
4
0.57
Trilepisium
madagascariense
75
ha-1
Abundance
Species name (in SFC)
Lower Middle
Upper
Sum
Ekebergia capensis
3
0
0
3
0.43
Grewia ferruginea
3
0
0
3
0.43
Rothmania urcelliformis
1
2
0
3
0.43
Bridelia micrantha
1
1
0
2
0.29
Galiniera saxifraga
2
0
0
2
0.29
Polyscias fulva
1
1
0
2
0.29
Syzygium guineense
0
2
0
2
0.29
Apodytes dimidiata
0
1
0
1
0.14
Flacourtia indica
1
0
0
1
0.14
Maytenus arbutifolia
1
0
0
1
0.14
Phoenix reclinata
1
0
0
1
0.14
Pittosporum viridiflorum
1
0
0
1
0.14
Podocarpus falcatus
1
0
0
1
0.14
Schrebera alata
1
0
0
1
0.14
Terminalia schimperiana
1
0
0
1
0.14
76
ha-1
Table 22: Tree species abundance per hectare in DNF
Species
Lower
Middle Upper
Sum
Abundanceha-1
Albizia gummifera
2
11
11
24
6
Allophylus abyssinicus
11
26
0
37
9.25
Apodytes dimidiata
30
82
23
135
33.75
Bersama abyssinica
21
16
0
37
9.25
Brucea antidysenterica
1
0
0
1
0.25
Canthium oligocarpum
5
5
0
10
2.5
Celtis africana
0
11
18
29
7.25
Chionanthus mildbraedi
10
91
0
101
25.25
Cordia africana
0
20
0
20
5
Croton macrostachyus
3
23
5
31
7.75
Ekebergia capensis
0
0
1
1
0.25
Ficus sur
0
26
27
53
13.25
Ficus sycamoras
0
2
0
2
0.5
Galiniera saxifraga
113
15
0
128
32
Macaranga capensis
2
28
0
30
7.5
Maytenus arbutifolia
6
0
0
6
1.5
Millettia ferruginea
1
92
9
102
25.5
Nuxia congesta
3
1
0
4
1
Olea welwitschii
0
0
2
2
0.5
Oxyanthus speciosus
2
0
0
2
0.5
Phoenix reclinata
5
5
0
10
2.5
Podocarpus falcatus
0
7
0
7
1.75
Polyscias fulva
1
1
0
2
0.5
Prunus africana
2
3
3
8
2
Psychotria orophila
4
0
0
4
1
Rothmania urcelliformis
0
5
0
5
1.25
Schefflera abyssinica
3
16
5
24
6
Syzygium guineense
21
72
19
112
28
Teclea nobilis
18
16
0
34
8.5
77
Species
Lower
Middle Upper
Sum
Abundanceha-1
Trichilia dregeana
0
2
2
4
1
Vangueria apiculata
0
5
0
5
1.25
Vepris dainellii
40
20
0
60
15
4.2.7.1. Vertical stratification and species diversity in SFC
Diversity, abundance, dominance and evenness of species varied along the vertical
stratification of SFC. The individual trees that remain in the lower, those reaching
and remaining in the middle and those reaching the upper storey are composed of
different number of species (Table 23). The dominance increased from lower to
upper storey (Table 23). The Shannon diversity index also shows variation in species
diversity in the three storeys showing declines from the lower, via the middle to the
upper storey (Table 23).
The species diversities in the three storeys were compared using bootstrapping (one
of the two randomization procedures). The diversity profile test confirmed that the
three storeys showed significant difference in species diversity (Figure 16). The
bootstrapping test showed significant difference in species diversity index between
lower and middle, lower and upper and middle and upper storeys (Table 23).
78
Table 23: Species abundance, richness and diversity in SFC
Lower storey
Middle storey
Upper storey
Species
36
27
3
Abundance
383
501
26
Dominance
0.09
0.13
0.5
Shannon_H
2.77
2.48
0.79
Evenness_eH/S
0.44
0.44
0.74
Table 24: Species richness, abundance, dominance, diversity and evenness
comparison between middle and upper storey; lower and middle storey; lower and
upper storey
Layer
Species
Abundance
Dominance
Shannon (H)
Evenness eH/S
middle
27
501
0.13
2.48
0.44
upper
3
26
0.5
0.79
0.74
Boot p
0
0
0
0
0.5
Layer
Species
Abundance
Dominance
Shannon (H)
Evenness eH/S
lower
36
383
0.09
2.77
0.44
middle
27
501
0.13
2.48
0.44
Boot p
0
0
0
0
0.97
Layer
Species
Abundance
Dominance
Shannon (H)
Evenness eH/S
lower
36
383
0.09
2.77
0.44
upper
3
26
0.5
0.79
0.74
Boot p
0
0
0
0
0.7
79
Figure 16: Diversity profile test in the lower, middle and upper storeys of the canopy
trees in the SFC
4.2.7.2. Vertical stratification and species diversity in DNFs
Like in the SFC, diversity, abundance, dominance and evenness of species varied
from the lower to the upper storey in the DNF. The three storeys are different in the
number of canopy trees in the lower, middle and upper storeys (Table 25). The
middle storey is the most diverse storey compared to the lower and upper storeys
(Table 25), while the dominance was lower in the middle than in the upper and lower
storeys (Table 26). Species evenness increased from the lower to the upper storeys.
The species diversities in the three storeys of the DNF were compared using
bootstrapping. The diversity profile confirmed that there was significant difference in
species diversity between the lower and the middle; the middle and the upper storeys
and hence were comparable, while the lower and the upper storeys did not show any
significant difference in diversity and hence were not comparable (Figure 17). The
bootstrapping test showed significant difference in species diversity index between
80
lower and middle; the middle and upperstoreys, but there was no significant
difference between the lower and the upper storeys (Table 26). Species richness
showed significant difference between the lower and the upper; the middle and the
upper, but not between the lower and the middlestoreys (Table 26). The abundance
showed significant variation throughout the three storeys, while the dominance
showed significant difference between the lower and the middle; the middle and the
upperstoreys (Table 26).
Table 25: Species richness, abundance, dominance, diversity and evenness in DNF
(lower = <13.67 m, middle = 13.67–26.67 m, upper = >26.67 m)
Lower
Middle
Upper
Species
22
26
12
Abundance
304
601
125
Dominance
0.18
0.09
0.14
Shannon (H)
2.23
2.69
2.13
Evenness_eH/S
0.42
0.57
0.7
81
Table 26: Comparison of species diversity, richness, abundance, dominance and
evenness between lower and middle; lower and upper; lower and middle storeys
Layer
Species
Abundance
Dominance
Shannon (H)
Evenness eH/S
Lower
22
304
0.18
2.23
0.42
Middle
26
601
0.09
2.69
0.57
Boot p
0.28
0
0
0
0.01
Layer
Species
Abundance
Dominance
Shannon (H)
Evenness eH/S
Lower
22
304
0.18
2.23
0.42
Upper
12
125
0.14
2.13
0.7
Boot p
0
0
0.07
0.49
0
Layer
Species
Abundance
Dominance
Shannon (H)
Evenness eH/S
Middle
26
601
0.09
2.69
0.57
Upper
12
125
0.14
2.13
0.7
Boot p
0
0
0
0
0.18
Figure 17: Diversity profile of canopy trees in the lower, middle and upper storeys in
the DNF
82
4.3. Carbon storage
Carbon storage in DNF, SFC, plantation forest, pasture, woodland and cropland
(Figure 18) of Jimma Highlands has been computed from DBH and height data of
the woody species with DBH
10 cm. The above ground live carbon storage in these
AGC (ton) storage in six land use types
land use types is presented below.
Figure 18: Boxplot for AGC storage in different land use types in Jimma transect (1
= plantation forest, 2 = DNF, 3 = semi-managed coffee forests, 4 = woodland, 5 =
pasture, 6 = cropland)
4.3.1. Carbon storage in SFC
The traditional coffee management system in Ethiopia played an important role in
carbon storage. Most of the SFC are characterized by very tall trees with a large
83
diameter trunk. These trees were primarily maintained for shade provision for coffee
shrubs beneath, but now they are also important source of ecosystem services such as
climate regulation, soil and water conservation, pollination services and carbon
storage. The tree species in the SFC (Appendix 12) are important sink of carbon. The
tree species belonging to 38 genera and 26 families in the SFC stored about 62 t C
ha-1. The most importnt tree species in carbon stoarge in SFC of the transect are
Albizia gummifera (ca. 15 t C ha-1), Croton macrostachyus (ca. 10 t C ha-1), Ficus
mucuso (ca. 7 t C ha-1), Acacia abyssinica (4 t C ha-1), Dracaena steudneri (ca. 4 t C
ha-1), Cordia africana (4 t C ha-1) and Millettia ferruginea (ca.3 t C ha-1)(Appendix
12).
The five most important plant families in carbon storage in SFC of the study area
were Fabaceae (ca. 22 t C ha-1), Moraceae (ca. 12 t C ha-1), Euphorbiaceae (ca. 11 t C
ha-1), Boraginaceae (ca.5 t C ha-1) and Dracaenaceae (ca. 4 t C ha-1). The five
families with least contribution to the carbon storage were Tiliaceae (0.002 t C ha-1),
Pittosporaceae (0.002 t C ha-1), Arecaceae (0.015 t C ha-1), Combretaceae (0.0l6 t C
ha-1) and Podocarpaceae (0.019 t C ha-1) (Appendix 12).
4.3.2. Carbon storage in DNFs
The DNF in the study transect was found to be a sink for about 82 t C ha-1 which is
distributed among 32 species of trees belonging to 31 genera and 20 families
(Appendix 13). The 10 top tree species in carbon storage are Ficus sur, Apodytes
dimidiata, Syzygium guineense, Celtis africana, Albizia gummifera, Schefflera
84
abyssinica, Olea welwitschii, Millettia ferruginea, Prunus africana and Macaranga
capensis. Five tree species that contributed least to carbon storage are Brucea
antidysenterica, Oxyanthus speciosus, Psychotria orophila, Vangueria apiculata and
Maytenus arbutifolia (Appendix 13). Ten most important plant families in AGC
storage are Moraceae, Icacinaceae, Fabaceae, Myrtaceae, Ulmaceae, Araliaceae,
Oleaceae,
Euphorbiaceae,
and
Rosaceae
Melianthaceae
(Appendix
13).
Simaroubaceae, Celastraceae and Arecaceae are the families with least AGC storage
in the DNF.
4.3.3. Carbon storage in woodland
The tree species with DBH
10 cm in the woodlands across the transect were found
to be a sink for about 13 t C ha-1. This amount of carbon was distributed among 26
plant species which belong to 22 genera and 13 families (Appendix 14). The top five
plant species in woodland with high carbon storage are Acacia abyssinica, Maesa
lanceolata, Ficus sycomoras, Cordia africana and Entada abyssinica. Five families
with higher carbon storage in the woodland are Fabaceae, Moraceae, Myrsinaceae,
Combretaceae and Boraginaceae. Myrtaceae and Rubiaceae are the two families in
the woodland with the least AGC storage in the woodland.
4.3.4. Carbon storage in pasture
The scattered trees in pasture across the transect were found to be a sink for about 3 t
C ha-1. This was distributed among 13 woody species with DBH
10 cm belonging
to 13 genera and 10 families. Ficus vasta was the most important tree species in
85
AGC storage (Appendix 15). Moraceae is the most important plant family in carbon
stoarge (Appendix 15).
4.3.5. Carbon storage in cropland
Cropland is the least of all land use types in carbon storage (Appendix 16). The two
most important tree species in cropland are Cordia africana and Prunus africana.
Boraginaceae is the most important family in carbon storage (Appendix 16).
4.3.6. Carbon storage in plantation forests
Plantation forest in our study transect was found to be a sink for about 82 t C ha-1.
More carbon was stored in this land use type due to management inputs and more
density of trees in it than in any land use type in the transect. According to the
information obtained from the local community the plantation forests are in the range
of 30-40 yeas and no commercial harvesting have been reported so far.
4.3.7. Above ground live carbon storage across the transect
The highest AGC storage was recorded from plantation forests and the minimum was
recorded from cropland (Table 27). The boxplot analysis showed the highest
accumulation of carbon in the biomass of forests (plantation, DNF, SFC) the least
carbon storage in the biomass of trees in the cropland and pasture (Figure 18).
The land use types showed significant mean differences in carbon storage at 95%
confidence level (Table 28). Multiple comparisons (Table 29) showed significant
86
mean difference in carbon storage between SFC and cropland, pasture and woodland.
Significant difference was also observed between cropland and DNF, cropland and
plantation forest. The variation between DNF and pasture, DNF and woodland,
pasture and plantation, pasture and woodland and plantation and woodland were
statistically significant (Table 29). Significant statistical difference has not been
observed between SFC and DNF, SFC and plantation forest, between cropland and
pasture, cropland and woodland, between DNF and plantation forest. The
homogeneity test also showed three groups of land use types in carbon storage (Table
29). Based on the similarity in AGC storage, the land use types were categorised
under three sub-groups (Table 30).
Table 27: Average AGC in six land use types across the transect
Land use
Mean
N
Std.
SFC
61.52
7.00
24.98
Cropland
2.03
6.00
0.82
DNF
82.03
4.00
32.08
Pastureland
2.51
5.00
2.67
Plantation forest
152.25
3.00
56.81
Woodland
12.87
4.00
7.60
Total
44.54
29.00
53.33
87
Table 28: Analysis of variance of different land use types in AGC storage in the
study transect
SS
df
MS
F
P
Between Land uses
2.747
5
0.549
42.23
0.00
Within land uses
0.295
23
0.013
Total
3.041
28
88
Table 29: Multiple comparison test for the differences of land use types in AGC in
the Jimma Highlands (MD = mean difference, LB = lower bound, UB =bound)
95% CI
(I) landuse
(J) landuse
MD
SE
P
SFC
Cropland
0.58
0.06
DNF
-0.07
Pasture
Crop
DNF
Pasture
Plantation
LB
UB
0.00
0.38
0.77
0.07
0.94
-0.29
0.15
0.61
0.07
0.00
0.40
0.81
Plantation
-0.21
0.08
0.13
-0.45
0.04
Woodland
0.31
0.07
0.00
0.09
0.53
DNF
-0.64
0.07
0.00
-0.87
-0.41
Pasture
0.03
0.07
1.00
-0.18
0.25
Plantation
-0.78
0.08
0.00
-1.03
-0.53
Woodland
-0.27
0.07
0.02
-0.49
-0.04
Pasture
0.67
0.08
0.00
0.44
0.91
Plantation
-0.14
0.09
0.60
-0.41
0.13
Woodland
0.37
0.08
0.00
0.13
0.62
Plantation
-0.81
0.08
0.00
-1.07
-0.56
Woodland
-0.30
0.08
0.01
-0.54
-0.07
Woodland
0.51
0.09
0.00
0.25
0.78
89
Table 30: Homogeneity test of land use types in AGC across the study transect in the
Jimma Highlands
Subset for alpha = 0.05
land use
N
1
2
3
Pasture
5
1.049
Cropland
6
1.0824
Woodland
4
SFC
7
1.6581
DNF
4
1.7235
Plantation
3
1.8632
1.3497
P
0.998
1
0.112
4.3.8. Carbon storage and species richness and abundance
The correletion coefficient between AGC storage and plant species richness in
different growth forms was analysed (Table 31). Liana was excluded due to the
assuption of normal distribution. AGC storage satisfied the assumption of normal
distribution after 4th-root transformation. The linear relationship between herbaceous
species and carbon storage was not significant and shrub species and AGC storage
was also not significant. The analysis showed significant linear relationship between
carbon storage and tree species richness and abundance.
Multiple linear regresion model was conducted (Table 32–34) using the tree species
richness and abunadnce which showed significant relationships with the carbon
90
storage in the linear coreletion analysis. Both variables combined have explained
about 82% of the variation in AGC storage (Table 32). The contribution of tree
species abundance to the model was significant, while that of tree species richness
was not (Table 34).
Table 31: Linear relationships between AGC and tree, herb and shrub richness and
tree species abundance
Pearson
Trees
Herb
Shrub
Abundance
Correlation
GC (4th root)
R
0.601
-0.018
0.288
0.904
P
0.001
0.928
0.129
0
N
29
29
29
29
Table 32: Variation in AGC explained by tree species richness and abundance
combined
Model
1
R
R2
R2adj
SE
0.908
0.824
0.81
0.14366
Predictors: Constant, abundance, trees
91
Table 33: Multiple regression analysis for prediction of AGC using abundance and
tree species richness
Model
1
SS
df
MS
F
P
Regression 2.505
2
1.252
60.679
0.00
Residual
0.537
26
0.021
Total
3.041
28
Predictors: Constant, abundance (4th-root), trees; Dependent variable: AGC (4th-root)
Table 34: Contribution of tree species richness and abundance to the model
Unstand_Coef
Stand_Coef
B
SE
Beta
Constant
0.142
0.124
trees
0.004
0.004
0.736
0.089
Model
t
P
1.144
0.26
0.094
0.912
0.37
0.848
8.258
0.00
Abundance_4throot
Dependent variable: AGC (4th-root)
4.3.9. Climate variables and AGC storage
The linear relationships between AGC storage in woody species biomass and most
climate variables were statistically significant, while the relationships with some
variables were not significant. (Appendix 17). The variables showing significant
relationships with AGC were used as predictive variables in multiple regression
92
analysis although most of them were excluded due to collinearity effect among
themselves. They show very high variance inflation factor (VIF) when put together
in the model showing strong linear relationships among themselves. All pairs of
variables having significant correlation with the AGC showed higher VIF (>10) and
finally a single variable with relatively higher Pearson correlation value was taken as
a predictive variable in linear regression analysis of AGC. The climate variable with
relatively higher linear relationship was evapotranspiration. This variable was taken
into the model to predict the AGC. Potential evapotranspiration explained 21% (R2 =
0.21, SE = 0.298, F = 7.159, P = 0.013) of the variation in AGC along the study
transect. The ANOVA test for the linear regression showed significant variation
(Table 35).
Table 35: Linear regression prediction of AGC by potential evapotranspiration (pet =
potential evapotranspiration
Model
SS
df
MS
F
P
Regression
0.637
1
0.637
7.159
0.013
Residual
2.404
27
0.089
Total
3.041
28
Predictors: Constant, pet; Dependent variable: AGC (4th-root)
4.3.10. Edaphic factors and AGC storage
Pearson correlation depicted significant relationships between AGC and soil cation
exchange capacity, AGC and sand, and AGC and soil pH (Table 36). Soil cation
93
exchange capacity and sand showed a positive relationship with AGC, while pH
showed a negative relationship. AGC decreased with increasing pH and vice versa.
Among the soil textures, silt and clay did not show significant linear relationships
with AGC. The three edaphic factors (ECE, sand and pH), which showed significant
linear relationships with AGC were selected for carbon prediction in multiple
regression analysis (Tables 37 and 38). Before conducting the multiple regression
analysis, the three variables were tested for collinearity and all of them showed VIF<
5 (Table 38) and were taken into the model. The three variables (CEC, sand and pH)
combined, have significantly explained about 60% (R2 = 0.604, R2adj = 0.556, SE =
0.220, F = 12.685, P = 0.00) of the variation in AGC.
The unstandardized regression coefficient tells us that for every unit increase of
CEC, the AGC storage increases by 0.089 (controlling the effect of pH and sand).
For every unit increase of pH (controlling the effect of sand and CEC), the AGC
decreases by 0.145 and the same is true for sand in which the AGC increases by
0.075 for a unit increase of sand (controlling the effect of pH and CEC). The
contribution of each of the three variables to the model was statistically significant
(Table 38).
94
Table 36: Linear relationships between AGC and soil factors (CEC = cation
exchange capacity, BD = bulk density)
Silt
CEC
BD
Sand
Clay
pH
-0.08
-0.39
-0.15
0.47
-0.35
-0.64
AGC(4th_ P
0.66
0.04
0.43
0.01
0.06
0.00
Root
29
29
29
29
29
29
Cor
N
Table 37: Prediction of AGC by using soil pH, sand and soil cation exchange
capacity
Model
SS
df
MS
F
P
1.835
3
0.612
12.685
0.00
1 Residual
1.206
25
0.048
Total
3.041
28
Regression
Predictors: Constant, pH, Sand, CEC; Dependent variable: AGC (4th-root)
Table 38: Contribution of each variable (CEC, sand and pH) to the model
Model
Unstand_Coef
Stand_Coef
B
Beta
SE
4.67
1.47
Constant
CEC
0.09
0.04
0.57
Sand
0.08
0.03
0.41
pH
-0.15
0.03
-0.995
th
Dependent variable: AGC (4 -root)
95
t
P
3.18
0.004
2.51
3.05
-4.54
0.019
0.005
0.000
VIF
3.23
1.15
3.03
4.4. Leaf Area Index (LAI)
A
B
Figure 19: Boxplot analysis of LAI (A = True LAI_ V6, B = True LAI V5, 1 = DNF,
2 = SFC, 3 = plantation forest, 4 = woodland, 5= pasture, 6 = cropland)
Based on the inclusion and exclusion of vegetation clumping and different CANEYE versions, four LAI outputs were produced. LAI_true accounts for vegetation
clumping, while LAI_effective does not. The result between LAI_true and
LAI_effective are different due to the clumping effect of vegetation. The data were
checked for normality prior to testing for significance variation between the outputs
from the two different versions of Can-Eye. The normality test showed that
LAI_effective did not satisfy the assumption of normal distribution (LAI_eff_v6 (n =
29, W = 0.88, P = 0.004) LAI_eff_v5 (n = 29, W = 0.87, P = 0.002), while the square
root transformed LAI_true from both versions of CAN-EYE fulfill the assumption of
normality (LAI_true_v6 (n = 29, W = 0.94, P = 0.07), LAI_true_v5 (n = 29, W =
0.94, P = 0.13)). Therefore, LAI_eff was excluded from this analysis and the
parametric tests were conducted only for LAI_true from both versions of CAN-EYE.
96
In both LAI_true_v6 and v5, mean of LAI_true_v6 was higher than mean of
LAI_true_v5 (Table 40). Boxplot test for both LAI_true_v6 (Figure 19A) and
LAI_true_v5 (Figure 19B) showed that the LAI decreases in different land use types
along the following orders: DNF, SFC, plantation forest, woodland, cropland and
pasture. The paired sample t-test showed significant statistical difference between
LAI_true_v6 and LAI_true_v5 under two versions of Cay-Eye (Tables 39 and 40).
Table 39: Mean True leaf area index (under CAN-EYE version 6 and 5) in six land
use types along the study transect
CAN-EYE_v6
CAN-EYE_v5
Land use
Mean
N
Std.
Mean
N
Std.
SFC
1.47
7
0.43
1.39
7
0.53
Cropland
0.23
6
0.05
0.19
6
0.06
DNF
2.29
4
0.69
2.15
4
0.79
Pasture
0.08
5
0.06
0.08
5
0.07
Plantation
1.11
3
0.43
1.02
3
0.41
Woodland
0.95
4
0.59
0.88
4
0.61
Table 40: Mean±SE of true LAI under both v_6 and v_5 of CAN-EYE
Mean±SE
LAI_true_v6
0.98±0.16
N
29
Pair 1
LAI_true_v5
0.91±0.15
97
29
Table 41: Paired T-test showing significant statistical differences between True_LAI
under version 6 and Version_5 of CAN-EYE
95% CI
Mean
SEM
Lower
Upper
t
df
P
0.07
0.02
0.025
0.108
3.32
28 0.003
LAI_true_v6
LAI_true_v5
4.4.1. LAI and Land Use Categories
Land use categories were found very important determinants of LAI_true from both
versions of CAN-EYE. Analysis of variance showed significant mean difference in
LAI_true_v6 and LAI_true_v5 within all land use types. There was significant mean
difference among land use types in LAI_true_v5 and LAI_true_v6 (Table 42).
Tukey’s multiple comparison tests (Table 43) showed significant mean difference in
LAI_true_v5 between SFC and cropland, SFC and pasture; Cropland and DNF,
cropland and planation forest, cropland and woodland; between DNF and pasture,
DNF and woodland; between pasture and plantation forests, and pasture and
woodland. Significant statistical difference was not seen between SFC and DNF;
SFC and planation forest, SFC and woodland; between cropland and pasture;
between DNF and plantation forest and, and between planation forest and woodland.
There were also significant statistical differences (Table 44) in LAI_true_v6 between
SFC and cropland, SFC and pasture; between cropland and DNF, cropland and
plantation forest, cropland and woodland; between DNF and pasture, DNF and
98
woodland; between pasture and planation forest, and pasture and woodland. The
difference between SFC and DNF, SFC and plantation, SFC and woodland, DNF and
planation forest, and planation forest and woodland were not significant.
Table 42: Analysis of variance showing significant differences in LAI_true_v6 and
v5 and among land use types in the transect
SS
df
MS
F
P
5.255
5
1.051
27.919
0.00
0.866
23
0.038
6.121
28
5.144
5
1.029
21.997
0.00
1.076
23
0.047
6.219
28
Between
Groups
LAI_true_v6_sqrt
Within
Groups
Total
Between
Groups
LAI_true_v5_sqrt
Within
Groups
Total
99
Table 43: Multiple comparisons showing differences in LAI_true_v6 between each
land use types (SFC, DNF, MD = mean difference, LB = lower bound, UB = upper
bound)
Dependent Variable LAI_true_v5
Land use
MD
95% CI
I
J
I&J
SE
SFC
Crop
0.727
0.120
0.000
0.354
1.100
DNF
-0.286
0.136
0.316
-0.707
0.134
Pasture
0.935
0.127
0.000
0.542
1.328
Plantation
0.170
0.149
0.859
-0.293
0.633
Woodland 0.278
0.136
0.345
-0.142
0.699
DNF
-1.013
0.140
0.000
-1.446
-0.580
Pasture
0.208
0.131
0.614
-0.198
0.614
Plantation
-0.557
0.153
0.015
-1.031
-0.082
Woodland -0.449
0.140
0.039
-0.882
-0.015
Pasture
1.221
0.145
0.000
0.771
1.671
Plantation
0.457
0.165
0.100
-0.056
0.969
Woodland 0.565
0.153
0.013
0.090
1.039
Plantation
-0.765
0.158
0.001
-1.255
-0.275
Woodland -0.657
0.145
0.002
-1.107
-0.206
0.165
0.985
-0.405
0.621
Crop
DNF
Pasture
Plantation Woodland 0.108
100
P
LB
UB
Table 44: Multiple comparisons showing differences in LAI_true_v6 between each
land use types (SFC, DNF, LB = lower bound, UB = upper bound, MD = mean
difference)
Dependent variable: LAI_true_v6
Land use
MD
95% CI
I
J
I&J
SE
P
LB
UB
SFC
Crop
0.724
0.108
0.000
0.389
1.059
DNF
-0.296
0.122
0.185
-0.674
0.081
Pasture
0.946
0.114
0.000
0.594
1.299
Plantation
0.164
0.134
0.821
-0.252
0.579
Woodland
0.280
0.122
0.232
-0.097
0.658
DNF
-1.020
0.125
0.000
-1.409
-0.632
Pasture
0.222
0.117
0.431
-0.142
0.587
Plantation
-0.560
0.137
0.005
-0.986
-0.134
Woodland
-0.444
0.125
0.019
-0.832
-0.055
Pasture
1.243
0.130
0.000
0.839
1.647
Plantation
0.460
0.148
0.050
0.001
0.920
Woodland
0.577
0.137
0.004
0.151
1.002
Plantation
-0.782
0.142
0.000
-1.222
-0.343
Woodland
-0.666
0.130
0.000
-1.070
-0.262
Woodland
0.116
0.148
0.967
-0.343
0.576
Crop
DNF
Pasture
Plantation
101
4.4.2. LAI and plant basal area, abundance and richness
A Pearson correlation test (Table 45) showed significant linear relationships between
LAI_true_v6 and tree species basal area (BA_log), total plant species richness,
richness in shrub, tree species and woody species abundance (abund_4th_root). The
test also showed significant linear relationships between LAI_true_v5 and tree
BA_log, shrub, tree species, total plant richness and woody species abundance.
Herbaceous species richness was not significantly related to both LAI_true_v6 and
LAI_true_v5.
Multiple regressions with the explanatory variables having significant linear
relationships with both LAI_true_v6 and v5 (BA_log, richness of shrubs, trees, total
richness and abundance_4th_root transformed) was conducted (Tables 45-49). The
variables were tested for collinearity and all of them were with VIF < 10 (Table 47).
The variables combined together have significantly explained about 82% (R2 =
0.824, R2adj = 0.786, SE = 0.22) and 81% (R2 = 0.811, R2adj = 0.77, SE = 0.23) of the
variations in LAI_true_v6 and LAI_true_v5 respectively. Log transformed BA was
the only variable with the highest contribution to the model of LAI_v6 and LAI_v5.
102
Table 45: Linear relationships between True leaf area indices, basal area, plant
species richness and abundance (BA = basal area, abund_4th = 4th root transformed
tree species abundance)
Pearson BA_log Herb
LAI_v6 Cor.
Shrubs Trees
richness abund_4th
0.87
0.11
0.41
0.73
0.50
0.79
P
0.00
0.56
0.03
0.00
0.01
0.00
N
29
29
29
29
29
29
0.86
0.11
0.41
0.72
0.49
0.78
P
0.00
0.59
0.03
0.00
0.01
0.00
N
29
29
29
29
29
29
LAI_v5 Cor.
Table 46: Analysis of variance for the multiple regression of LAI_v6 with
explanatory variables
Model
SS
Regression 5.044
df
MS
F
P
5
1.009
21.553
0
0.047
1 Residual
1.077
23
Total
6.121
28
Predictors: Constant, abundance (4th-root), richness, BA (log), tree species richness
Dependent variable: LAI_v6)
103
Table 47: Contribution of basal area, shrub richness, tree species richness, plant
species richness along the entire study area and tree species abundance (BA_log =
log transformed basal area, Abund_4th = 4th root transformed tree species abundance)
Unstand_Coef
Stand_Coef
B
Beta
Model
SE
Constant
0.305
0.34
BA_log
0.404
0.13
Shrubs
-0.046
Trees
t
P
VIF
0.909
0.373
0.64
3.225
0.004
5.15
0.11
-0.088
-0.412
0.684
5.987
0.019
0.01
0.328
1.642
0.114
5.208
Richness
0.002
0.01
0.076
0.28
0.782
9.763
Abun_4th
0.045
0.25
0.036
0.177
0.861
5.478
Dependent variable: LAI_v6
Table 48: Analysis of variance test for the prediction of LAI_v5 by the explanatory
variables –Tree species abundance, basal area, richness, shrub species richness and
richness across the entire transect
Model
SS
Regression 5.042
df
MS
F
P
5
1.008
19.703
0.00
0.051
1 Residual
1.177
23
Total
6.219
28
Dependent variable: LAI_v5
104
Table 49: Contribution of basal area, shrub richness, tree species richness, plant
species richness along the entire study area and tree species abundance (BA_log =
log transformed basal area, Abun_4th = 4th root transformed tree species abundance)
Unstand_Coef
Stand_Coef
Model
B
SE
Beta
Constant
0.293
0.351
BA_log
0.416
0.131
Shrubs
-0.031
Trees
t
p
VIF
0.84
0.41
0.652
3.17
0
5.15
0.116
-0.059
-0.27
0.79
5.987
0.021
0.012
0.35
1.69
0.1
5.208
Richness
0.001
0.006
0.033
0.12
0.91
9.763
Abun_4th
0.007
0. 263
0.006
0.03
0.98
5.478
Dependent variable: LAI_v5
4.4.3. LAI, edaphic and topographic factors
Topographic (elevation and slope) and edaphic factors (soil organic carbon (SOC),
cation exchange capacity (CEC), soil texture (silt, sand, clay) and BD) were tested
for linear relationships with LAI_true_v6 and LAI_true_v5 (Table 50). All
topographic and some of the edaphic factors (SOC and BD) did not show significant
linear relationships with both LAI_true_v6 and LAI_true_v5. The relationships
between LAI_true_v6 and soil CEC, sand and clay was significant. LAI_true_v5 also
showed significant linear relationships with CEC, sand and clay.
The explanatory variables (CEC, sand, clay) which showed significant linear
relationships with both LAI _true_v6 and v5 were tested for collinearity before
105
conducting multiple regressions. All of them were found to have VIF < 10 (Tables
52 and 54). Multiple regression analysis (Tables 51–54) showed combined effect of
the three explanatory variables. Combined, the three variables explained about 45%
(R2 = 0.45, R2adj = 0.39, SE = 0.37) of the variation in LAI _true_v6 and about 42%
(R2 = 0.42, R2adj = 0.35, SE = 0.38) of the variation in LAI_true_v5. In LAI_true_v6,
sand has significant contribution to the model (Table 53), while in LAI_true_v5 all
the three variables did not show significant contribution to the model separately
(Table 54).
Table 50: Linear relationships between LAI_true indices and topographic factors
(elevation and slope) and edaphic factors (SOC, CEC, silt, sand, clay and BD)
LAI type
LAI_true_v6
LAI_true_v5
Pearson Elev SOC CEC slope silt
sand clay
Cor.
0.36 -0.03 -0.41
0.64 -0.55 -0.35
P
0.06
0.87
0.03
N
29
29
29
Cor.
0.35
0.00 -0.41
P
0.06
1.00
0.03
N
29
29
29
106
-0.05 0.00
0.79 1.00
29
29
-0.03 0.02
0.88 0.92
29
29
BD
0.00
0.00
0.06
29
29
29
0.62 -0.54 -0.35
0.00
0.00
0.07
29
29
29
Table 51: Analysis of variance test for the prediction of LAI_true_v6 by the
explanatory variables (clay, CEC and sand across the entire transect)
Model
SS
df
1 Regression
2.758
3
Residual
3.362
25
Total
6.121
28
MS
F
P
0.919 6.835 .002
0.134
Preidctors: Constant, Clay, CEC, Sand; Dependent variable: LAI_true_v6
Table 52: Contribution of CEC, sand and clay to the model
Unstand_Coef
Stand_Coef
Model
B
SE
Beta
Constant
-2.474
4.173
CEC
-0.047
0.037
Sand
0.141
Clay
-0.006
Collinearity
t
P
VIF
-0.59
0.559
-0.21
-1.27
0.214
1.22
0.064
0.548
2.204
0.037
2.809
0.058
-0.03
-0.1
0.924
2.997
Dependent variable: LAI_v6
Table 53: Analysis of variance test for the prediction of LAI_true_v5 by the
explanatory variables- clay, CEC and sand across the transect
Model
1
SS
df
MS
Regression
2.613
3
0.871
Residual
3.606
25
0.144
Total
6.219
28
F
P
6.04
0.003
Predictors: Constant, Clay, ECE, Sand; Dependent variable: LAI_v5
107
Table 54: Contribution of CEC, sand and clay to the model
Unstand_Coef
Stand_Coef
Model
B
SE
(Constant)
-1.772
4.322
CEC
-0.046
0.038
Sand
0.129
Clay
-0.014
Beta
Collinearity
t
P
-0.41
0.685
-0.21
-1.22
0.234 1.22
0.066
0.494
1.936 0.064 2.809
0.06
-0.06
-0.23
0.82
VIF
2.997
Dependent variable: LAI_v5
4.4.4. LAI, enhanced vegetation index and normalized difference
vegetation index
Enhanced vegetation index (EVI) and normalized vegetation index (NDVI)
respectively showed significant linear relationships with LAI_true_v6 and
LAI_true_v5 (Table 55). Both NDVI and EVI were used as predictor variables in
multiple regressions with LAI_v6 and LAI_v5 (Tables 56–59). NDVI and EVI
combined together have explained about 61% (R2 = 0.611, R2adj = 0.581, RSE =
0.30275, F = 20.387, P = 0.00) of the variation in LAI_true_v6 and 60% (R2 = 0.599,
R2adj = 0.568, RSE = 0.3097, F = 19. 416, P = 0.00) of the variation in LAI_true _v5.
In both LAI_true_v6 and LAI_true_v5, NDVI has significant contribution to the
model, while EVI was not singly (Tables 57 and 59). In both LAI_v6 and LAI_v5,
the analysis of variance showed significant result (Tables 56 and 58).
108
Table 55: Linear relationships between LAI_true indices and NDVI and EVI
LAI Type
Bivariate
NDVI
EVI
0.77
0.69
P
0.00
0.00
N
29
29
0.75
0.69
P
0.00
0.00
N
29
29
LAI_true_v6_sqrt Pearson Correlation
LAI_true_v5_sqrt Pearson Correlation
Table 56: Analysis of variance test for the prediction of LAI_true_v6 by the
explanatory variables –EVI and NDVI
Model
SS
1
df
MS
F
P
Regression 3.737
2
1.869
20.387
0.00
Residual
2.383
26
0.092
Total
6.121
28
Predictors: Constant, EVI, NDVI; Dependent variable: LAI_v6
Table 57: Contribution of NDVI and EVI separately to the model
Unstand_Coef Stand_Coef
Model
B
SE
Constant
-0.229
0.421
NDVI
1.612
0.527
EVI
0
0
Beta
t
p
-0.544
0.591
0.585
3.061
0.005
2.435
0.237
1.242
0.225
2.435
Dependent variable: LAI_true_v6
109
VIF
Table 58: Analysis of variance test for the prediction of LAI_true_v5 by the
explanatory variables- EVI and NDVI
Model
SS
Regression
df
MS
3.725
2
1.863
1 Residual
2.494
26
0.096
Total
6.219
28
F
P
19.416 0.00
Predictors: Constatnt, EVI, NDVI; Dependent variable: LAI_true_v5_sqrt
Table 59: Contribution of NDVI and EVI to the model
Unstand_Coef Stand_Coef
Model
B
SE
Beta
t
P
-0.82
0.418
VIF
Constant
-0.36
0.43
NDVI
1.502
0.54
0.54
2.788 0.01
2.435
EVI
0
0
0.277
1.431 0.164
2.435
Dependent variable: LAI_true_v5_sqrt
4.4.5. LAI and AGC storage
Both LAI_true_v6 and LAI_true_v5 showed strong linear relationship with AGC t
ha-1(4th-root transformed) across the land use types in the study transect (Table 60).
Due to the effect of collinearity the LAI_true_v5 was avoided from the regression
analysis and conducted with LAI_true_v6. The linear regression analysis (Table 61)
and Figure 20) showed that LAI_true_v6 explained about 75% (R2 = 0.754, R2adj =
110
0.745, RSE = 0.166, F = 82.87, P = 00) of the variation in AGC t ha-1 along the study
transect.
Table 60: Linear relationship between LAI_true indices and AGC
Peason Correlation
LAI_true_v6_sqrt
LAI_true_v5_sqrt
AGC t ha-1 (4th_Root)
R
0.87
P
0
N
29
R
0.86
P
0
N
29
Table 61: Analysis of variance test for the prediction of AGC by the explanatory
variable –LAI_true_v6
Model
1
SS
df
MS
F
P
Regression 2.294
1
2.294
82.87
0
Residual
0.747
27
0.028
Total
3.041
28
Predictors: constant, LAI_true_v6_sqrt; Dependent variable: AGCt ha-1
111
Figure 20: Prediction of above ground live carbon storage from leaf area index
4.4.6. LAI and climate variables
The linear correlation analysis (Appendix 18) showed significant linear relationships
between LAI and most climate variables, while the relationships with some other
climate variables were not statistically significant. LAI_true_v6 showed significant
linear relationships with most climate variables, while the relationships with some
other variables were not statistically significant. Almost half of the climate variables
showed significant linear relationship with LAI_true_v5, while the relationships of
LAI_true_v5 with the remaining half of the climate variables were not significant
(Appendix 18).
All the climate variables with a significant linear relationship with LAI from both
versions of the CAN-EYE showed strong collinearity in the multiple regression
analysis. The least collinearity value was calculated for mean annual temperature
(VIF = ~11) and annual temperature range (VIF = ~11) and these two variables were
used in multiple regression analysis to determine the amount of variation explained
in LAI. The two variables combined have explained about 21% (R2 = 0.207, R2 adj =
112
0.146, F = 3.401, P = 0.049) of the variation with LAI. The analysis of variance also
showed significant result (Table 62). The contribution of the two variables was not
significant separately (Table 63).
Table 62: Analysis of variance test for the prediction of AGC by the explanatory
variables (bio1 = mean annual temperature and abio7 = nnual temperature range)
Model
1
SS
df
MS
F
Regression
1.290
2
0.645
Residual
4.930
26
0.190
Total
6.219
28
P
3.401
0.049
Predictors: Constant, bio1, bio7; Dependent variable: LAI_v5
Table 63: Contribution of mean annual temperature and annual temperature range to
the model
Unstand_Coef
Stand_Coef
B
SE
Beta
T
p
0.059
VIF
Model
Constant 11.319
5.739
-0.636
1.972
1
bio7
-0.064
0.058
0.194
-1.107 0.278
10.837
bio1
0.011
0.033
0.337
10.837
Dependent variable: LAI_v5
113
0.738
4.4.7. LAI above and below coffee canopies
LAI above and below the canopy of coffee shrubs/trees was taken from 29 sample
plots of size 400 m2 in the SFC along the transect. The box plot (Figure 21) shows
the distribution of LAI data which were taken under two different conditions (above
the coffee canopy (ab) and under the coffee canopy (uc)) along the study transect.
The distribution of the data was also analysed under two varying versions of CANEYE software. In both versions of CAN-EYE, and under both presence and absence
of vegetation clumping the mean ± standard error of the LAI taken below the canopy
of coffee is higher than the mean ± standard error of the LAI taken above the coffee
canopy (Table 64). The mean of LAI_true (where the vegetation clumping was
accounted for) is higher than the mean LAI_effective (where the vegetation clumping
was not considered). Even the mean for LAI_true below the coffee canopy is higher
than the LAI_true for above coffee canopy.
Figure 21: Boxplot analysis showing more LAI value for the under coffee canopy
than those taken above the coffee canopy
114
Table 64: Mean of LAI under and above the coffee canopy (uc = under coffee
canopy, ab = above coffee canopy)
Pairs
Mean ± SE
N
LAI_eff_v6_uc vs LAI_eff_v6_ab
0.93±0.03; 0.26±0.02
29
LAI_eff_v5_uc vs LAI_eff_v5_ab
1.05±0.04; 0.29±0.02
29
LAI_true_v6_uc vs LAI_true_v6_ab
1.93±0.07; 0.51±0.04
29
LAI_true_v5_uc vs LAI_true_v5_ab
1.98±0.09; 0.53±0.05
29
4.4.8. Normality test
Before testing for significance of the variation of LAI which was taken under the
coffee canopy and above the coffee canopy, normality test was conducted, which
confirmed that except for LAI_eff_v5 (above the coffee canopy) all the rest do not
significantly deviate from the normal distribution (Table 65). Except for the two,
Shapiro-Wilk normality test did not show significant difference of the data from
normal distribution at P = 0.05 significant level.
4.4.9. Significance test
Paired sample t-test was applied (Table 66) to evaluate whether the LAI taken above
and under the coffee canopy could show significant statistical difference or not. The
t-test
showed
that
there was
statistically significant
difference between
LAI_eff_v6_un and LAI_eff_v6_ab (t28 = 20.15, P = 0.00), LAI_true_v6_uc and
115
LAI_true_v6_ab (t28 = 18.68, P = 0.00) and LAI_true_v5_uc and LAI_true_v5_ab
(t28 = 15.27, P = 0.00).
Table 65: Shapiro-Wilk normality test for the LAI data taken above and below the
coffee canopy
LAI
N
Shapiro-Wilk W
P
LAI_eff_v6_uc
29
0.95
0.17
LAI_eff_v6_ab
29
0.97
0.43
LAI_true_v6_uc
29
0.98
0.80
LAI_eff_v5_uc
29
0.96
0.28
LAI_true_v5_uc
29
0.97
0.64
LAI_true_v6_ab
29
0.96
0.31
LAI_eff_v5_ab
29
0.93
0.04
LAI_true_v5_ab
29
0.93
0.07
Table 66: Paired sample t-test for the true and eff_LAI taken above and below the
coffee canopy
95% CI
LAI under and above coffee canopies
Mean ± SE
t
df
P
lower upper
LAI_eff_v6_uc vs LAI_eff_v6_ab
0.67±0.03
0.6
0.73 20.2 28
0
LAI_eff_uc vsLAI_eff_ab
0.76±0.04
0.66
0.85 16.8 28
0
LAI_true_v6_uc vs LAI_true_v6_ab
1.41±0.08
1.26
1.57 18.7 28
0
LAI_true_v5_uc vs LAI_true_v5_ab
1.45±0.01
1.26
1.65 15.3 28
0
116
4.5. Habitat Suitability Model
The present and future distribution of Acacia abyssinica, Cordia africana, Millettia
ferruginea, Phytolacca dodecandra and Schefflera abyssinica was modelled using
Maximum Entropy modelling aross Ethiopia. The model performance under both
present and future climate change scenarios is given in Table 67 and habitat
suitability map for each of them is given in Appendix 19. The variable contribution
and jackknife test for each of them is addressed below.
4.5.1. Acacia abyssinica
The model performance was tested and the test showed good performance for both
training (AUC = 0.89) and test data (AUC = 0.86) sets. The model performance was
also evaluated under the projected climate and found that it performed well for both
training (AUC = 0.88) and test data (AUC = 0.82) sets (Table 67).
Table 67: Model performance under baseline (b) and projected (p) climate change
scenarios for five plant species in Ethiopia
Training
Test
Training
Test
AUCb
AUCb
AUCp
AUCp
Acacia abyssinica
0.89
0.86
0.88
0.82
Cordia africana
0.87
0.84
0.87
0.83
Millettia ferruginea
0.91
0.88
0.91
0.89
Phytolacca dodecandra
0.93
0.91
0.92
0.90
Schefflera abyssinica
0.91
0.90
0.91
0.87
Species
117
4.5.1.1. Analysis of variable importance
Mean annual temperature has contributed more to the model compared to all the
remaining variables under both the baseline climate change scenario (Table 68) and
projected climate (Table 69). Rainfall seasonality has least contribution to the model
under the baseline climate change scenario, while temperature seasonality
contributed the least under the projected climate.
The jackknife test showed that the mean annual temperature was with the highest
gain when used in isolation and hence has the most useful information by itself. It is
also the variable that has the information which is not present in the remaining four
variables. This is true under both present (Appendix 20A1, A3) and future (Appendix
20A2, A4) climate change scenarios. Most of the areas which are suitable for the
distribution of A.abyssinica under the current climate (Appendix 19A) will turn
unsuitable under the projected climate (Appendix 19B).
Table 68: Contribution of each five climate variables to the distribution of Acacia
abyssinica under the baseline climate scenario
Variable
% contribution
Perm. importance
Mean annual Temperature
80.5
77
Isothermality
8.6
11.8
Mean annual rainfall
5.4
4.3
Temperature seasonality
3.9
4.9
Rainfall seasonality
1.6
2.1
118
Table 69: Contribution of each five climate variables to the distribution of Acacia
abyssinica under the projected climate
Variable
% contribution
Perm. importance
Mean annual temperature
84
73.5
Isothermality
5.2
6.8
Mean annual rainfall
4
5.5
Rainfall seasonality
3.5
4.4
Temperature seasonality
3.3
9.7
4.5.2. Cordia africana
The present and future distribution of C. africana across Ethiopia was modelled
using Maximum Entropy modelling algorithm. Five climate variables were selected
to avoid the effect of collinearity. The model showed good performance under both
the current scenarios and future climate projections. The receiver operating
characteristic (ROC) curve (under the baseline scenario) showed better performance
of the model. The area under curve for training data (under the baseline scenario)
(AUC = 0.87), for test data (AUC = 0.84) which is higher than random distribution
(AUC = 0.5) (Table 67). The model also showed good performance under the future
climate change projections in which the ROC attained 0.87 for training data 0.83 for
test data (Table 67).
Under the current climate change scenario, the areas of southwest highlands of
Ethiopia (Jimma, Kaffa, Bench-Maji, Illubabbor, Shaka); central Oromia (east Shewa
119
and western part of Arsi) and southern Oromia (northern Borana, western and central
parts of Bale), Eastern Ethiopia (eastern Hararghie Highlands) are climatically the
most suitable areas for the distribution of C. africana (Appendix 19B1).The habitat
suitability decreases towards the lowlands on the northeastern, eastern and western
parts of Ethiopia.
Most of the areas in Borana and Bale zones of Oromia Region, which are currently
suitable for the distribution of C. africana will lose their suitability for the species
under the projected climate change scenario (Appendix 19B2). Areas in northern
Ethiopia which are suitable for the distribution of C. africana under the baseline
climate scenario will turn unsuitable under the projected climate.
4.5.2.1. Analysis of variable contributions
The contribution of mean annual temperature to the model under the current and
projected climate was 52.3% and 64.7% respectively followed by mean annual
rainfall with contribution of 33% and 22.6 respectively (Tables 70 and 71). Rainfall
seasonality contributed only 4% and 4.1% to the model under the current and
projected climates respectively.
The jackknife test of variable importance showed that mean annual temperature has
got the highest gain when used in isolation under both the current (Appendix 20B1 &
B3) and future climates (Appendix 20B2 & B4). This variable has the most useful
120
information by itself. The variable that reduces the gain when omitted was also mean
annual temperature in both models (present and future). Mean annual temperature is
also the most important variable of jackknife test of test data under the current and
future climates. It has the highest gain when used in isolation and the variable which
affects the model most when omitted.
Table 70: Contribution of each five climate variables to the distribution of Cordia
africana underthe baseline climate scenarios in Ethiopia
Variable
Mean annual temperature
% contribution
Perm. importance
51.3
32.5
Mean annual rainfall
33
41.8
Temperature seasonality
6.1
13.2
Isothermality
5.6
2.1
4
10
Rainfall seasonality
121
Table 71: Contribution of each five climate variables to the distribution of Cordia
africana under the projected climate
Variable
% contribution
Perm.importance
Mean annual temperature
64.7
55.7
Mean annual rainfall
22.6
24.9
Temperature seasonality
5.7
10.1
Rainfall seasonality
4.1
8.1
Isothermality
2.9
1.2
4.5.3. Millettia ferruginea
The distribution of Millettia ferruginea under the current and future climate change
scenarios was modelled using five climate variables, while other climate variables
were excluded due to collinearity. The model showed good performance under both
the current scenarios and future climate projections. The ROC curve (under the
baseline scenario) showed good performance of the model. The area under curve for
training data (under the baseline scenario) (AUC = 0.91), for test data (AUC = 0.88)
which is higher than random distribution (AUC = 0.5) (Table 67). The model also
showed good performance under the future climate change projections in which the
ROC attained 0.91 for training data 0.89 for test data (Table 67).
In general, southwest Ethiopia, particularly the highlands of Jimma, Kaffa, Shaka,
most parts of Illuababor Zone, some areas of East Wellega Zone, and eastern part of
West Wellega Zone, northern Borana Zone, Sidama and Gedeo Zones in south
122
Ethiopia, some parts of Awi and Metekel Zones are some of the most suitable areas
for the distribution of M. ferruginea under the current climate change Scenario
(Appendix 19C1).
The entire Awi Zone, most parts of Metekel, most parts of southwest Illubabor Zone
and Shaka Zone will lose their suitability for the distribution of M. ferruginea under
the future climate change scenarios (Appendix 19C2). Most highlands of Jimma and
Bench-Maji Zone will be suitable for the distribution of M. ferruginea under the
projected climate change.
Western and central highlands of Jimma, central and eastern part of Kaffa, central
and northwestern parts of Bench-Maji remain suitable areas for the distribution of M.
ferruginea. The habitat suitability declines in Shaka and western parts of Illubabor
Zone. The expansion of suitable areas under the projected climate was predicted in
the northern parts of Illubabor Zone, in the northern part of Borana and western part
of Bale, northwestern parts of Gamo Gofa and South Omo, central Arsi and eastern
Hararghe highlands (Appendix 19C2).
4.5.3.1. Analysis of variable contributions
Under the current climate, mean annual temperature is the most important variable
with the percent contribution of 48.8% followed by mean annual rainfall (44.9%)
(Table 72). The variable with the least contribution under same climate was rainfall
seasonality (0.2%). Both mean annual temperature and mean annual rainfall are more
impoartant predictors under the projected climate change scenarios (Table 73).
123
Table 72: Contribution of each five climate variables to the distribution of Millettia
ferruginea under the baseline climate scenario
Millettia_baseline
Variable
% contribution
Perm. importance
Mean annual temperature
48.8
77.4
Mean annual rainfall
44.9
16.3
Temperature seasonality
4.1
4.7
Isothermality
2
1.2
Rainfall seasonality
0.2
0
Table 73: Contribution of each five climate variables to the distribution of Millettia
ferruginea under the projected climate
Millettia _future
Variable
% contribution
Perm. importance
Mean annual temperature
52.5
76.9
Mean annual rainfall
37.9
12.1
Temperature seasonality
8.2
10.2
Isothermality
1.1
0.3
Rainfall seasonality
0.4
0.6
124
The importance of the variable was also shown by the Jackknife test (Appendix
20C1–C4). Mean annual rainfall was the most impoartant in the training data set,
under the current climate (Appendix 20C1), while mean annual temperature was the
most impoartant in test data (Appendix 20C2). Mean annual temperature is the
variable that impacts the most when it is omitted in both training and test data under
the current climate change scenarios. Under the projected climate, the variable with
the highest gain in both training and test data was the mean annual rainfall
(Appendix 20C3 & C4). The variable that decreases the gain the most when omitted
was mean annual temperature.
4.5.4. Phytolacca dodecandra
The habitat suitability for the distribution of P.dodecandra under the current climate
change scenarios and future projection was modelled using Maximum Entropy
modelling. Five climate variables (mean annual temperature, mean annual
precipitation, temperature seasonality, rainfall seasonality and isothermality) were
used. The model performance was tested and the test showed good performance for
both training (AUC = 0.93) and test data (AUC = 0.91) sets. The model performance
was also evaluated under the projected climate and found that it performed better
than it could be by random for both training (AUC = 0.92) and test data (AUC =
0.90) sets (Table 67).
Southwest highlands of Ethiopia (highlands of Kaffa, Jimma, Bench-Maji zones);
central Ethiopia (highlands of West Shewa and Guragie zones); the eastern
125
escarpments of Rift Valley (Arsi, Sidama, northern Borana and Gedeo zones, north
and western Bale) and Harargie highlands are currently the most suitable areas for
the distribution of P. dodecandra (Appendix 19D1). The suitability decreases as one
moves from the areas mentioned above towards the lowlands in all sides of Ethiopia.
The western edges (from Gambella to western Tigray), the eastern edges from
Ogaden to eastern Tigray) are the areas which are not suitable for the distribution of
the species under the current climate change scenarios.
From the northwestern highlands, the areas of southwest Ethiopia (Jimma, Kaffa);
central Ethiopia (West Shewa and Guraghe zones) will lose their suitability for the
distribution of the species under the projected climate. The areas of southeastern
highlands (most areas of northern and northwestern Bale; Sidama and Gedeo zones;
central Arsi and eastern Hararghe highlands will be suitable for the distribution of the
species under the projected climate (Appendix 19D2).
4.5.4.1. Analysis of variable importance
Temperature variables have contributed more to the model than the variables related
to precipitation (Table 74). Mean annual temperature has contributed more to the
model performance compared to all the remaining four variables under both the
current and future climate projections (Tables 74 and 75). Mean annual rainfall
contributed the least to the model under both climate change scenarios (Tables 74
and 75).
126
Table 74: Contribution of each five climate variables to the distribution of
Phytolacca dodecandra under the baseline climate scenario
Variable
% contribution
Perm. importance
Mean annualtemperature
59.9
59.1
Isothermality
24.3
13.5
Temperature seasonality
10.4
24.1
Rainfall seasonality
3.9
2.9
Mean annual rainfall
1.6
0.4
Table 75: Contribution of each five climate variables to the distribution of
Phytolacca dodecandra under the projected climate
Variable
% contribution
Perm. importance
Mean annual temperature
67.8
70.4
Isothermality
15.1
5.2
Temperature seasonality
12
18.8
Rainfall seasonality
3.7
4.8
Mean annual rainfall
1.3
0.7
The jackknife test showed that the mean annual temperature was with the highest
gain when used in isolation and hence has the most useful information by itself. It is
also the variable that has the information which is not present in the remaining four
127
variables. This is true under both present (Appendix 20D1 & D2) and future
(Appendix 20D3 & D4) climate change scenarios.
4.5.5. Schefflera abyssinica
The model performed well under both present and future climate projections. Under
the current climate change scenarios, the predicted area under curve was 0.91 for the
training data and 0.90 for the test data. Under the projected climate, the AUC was
0.91 for the training data and 0.87 for the test data (Table 67)
4.5.5.1. Analysis of variable importance
Mean annual temperature is with the highest contribution followed by mean annual
rainfall and hence it is the variable with the most impact on predicting the habitat
suitability for Schefflera abyssinica under the current climate change scenario, while
rainfall seasonality is with less impact on predicting suitable areas under the current
climate (Table 76). Almost a similar condition was obtained for the habitat suitability
of S. abyssinica under the future climate change scenarios. In the projected climate
change scenarios too, the mean annual temperature contributed more and the least
contributor was rainfall seasonality (Table 77). The second contributor to the model
was mean annual rainfall. Some areas in south, southwest Ethiopia and Arsi
highlands are suitable for the distribution of the species (Appendix 19E1). Some of
these areas lose their suitability under the projected climate (Appendix 19E2).
128
Table 76: Contribution of the five climate variables to the distribution of Schefflera
abyssinica under the baseline climate scenario
Variable
% contribution
Perm. importance
Mean annual temperature
60.3
70.2
Mean annual rainfall
23.3
19.3
Temperature seasonality
8.4
4.3
Isothermality
6.1
1.5
Rainfall seasonality
1.9
4.7
Table 77: Contribution of the five climate variables to the distribution of Schefflera
abyssinica under the projected climate
Variable
% contribution
Perm. importance
Mean annual temperature
59.7
74.5
Mean annual rainfall
22.9
16.2
Temperature seasonality
9.9
5.4
Isothermality
4.6
0.7
Rainfall seasonality
2.9
3.
The jackknife test of variable importance for both training and test data showed that
the mean annual temperature was the variable with the most useful information by
itself and at the same time it was found to be the variable that decreases the gain the
most when it was omitted (Appendix19E1 & E2). This was under the current climate
129
change scenarios. This variable remained the most important even under the
projected climate change scenarios (Appendix 19E3 & E4).
130
CHAPTER FIVE
5. Discussion, Conclusion and Recommendations
5.1. Discussion
5.1.1. Land use /land cover change
The 2008 LULC map revealed that the transect was covered by cropland, natural
forest, plantation forest, woodland and pasture. The information obtained from the
Agricultural Office of Setema District and some elder people in the transect, the
ubiquitous and extensive conversion of natural vegetation (grassland and forest) to
agriculture and plantation forests started since 1975 in the lower part of the transect.
These sources confirmed that agriculture was expanded to the upper part of the
transect from 1985 onwards. This coincides with the time when people from norther
Ethiopia were brought to resettle in the study area (one of the fertile parts of
southwest Ethiopia selected at that time) as the remedy to combat the 1984 famine in
northern Ethiopia. Since then, due to the wide spread of agricultural expansion,
larger areas of the forest and grassland has been converted to croplands. This agrees
with several reports that addressed the decline of forest cover in Ethiopia
(Breitenbach, 1961; EFAP, 1994; Hylander et al., 2013) and expansion of agriculture
lands (Tadesse Woldemariam and Masresha Fetene, 2007). Due to human activities,
the transect has been converted to a mosaic of different land use types such as
cropland, pasture, plantation forest, SFC and DNF and woodland. This agrees with
Landon (1996) and Konemund et al. (2002) who respectively addressed the annual
loss of closed forests and natural vegetation in Ethiopia respectively. This study also
agrees with a study on closed forest decline in Shaka Zone of Ethiopia by Tadesse
131
Woldemariam and Masresha Fetene (2007) and with net forest cover decline in
Bonga and Goma-Gera area (Hylander et al., 2013).
5.1.2. Species richness
The species area curve became flat after five plots in DNF, ten plots in woodland,
five plots in cropland, ten plots in SFC, ten plots in pasture and four plots in
planation forests showing that the sampling effort to incorporate all species occurring
in different land use types was exhaustive.
The result of the study showed that the transect was rich in plant species richness and
diversity. About 287 species belonging to 220 genera and 82 families were
documented. As part of the Eastern Afromontane Biodiversity Hotspot area
(Mittermeier et al., 2004), the study transect in the Jimma Highlands has been
endowed with plant species richness. This agrees with Coetzee (1978) who
conducted a study on plant diversity and richness in East African Mountains. Among
the 82 families recorded from the entire transect, Asteraceae was the most species
rich family with total number of species (n = 33) followed by Fabaceae (n = 25).
Asteraceae is the most species rich family in almost all land use types along the
transect. Most floristic studies conducted in Ethiopia showed high number of plant
species belonging to family Asteraceae (Dereje Denu, 2007; Dereje Denu and
Tamene Belude, 2012; Ermias Lulekal, 2014) on Bibita Forest (Guraferda), sacred
landscapes in Bedele District and Dense Forest in Ankober, respectively. Three of
132
the forty one families were represented by introduced species (Cupressaceae,
Pinaceae and Proteaceae). According to the Setema District agricultural office, most
of the exotic species were planted in the transect in late 1970’s.
The six land use types of the transect vary in plant species richness, abundance and
diversity. From this study it is apparent that the plant species richness decreases from
woodland to the cropland. The chi-square test conducted indicated the impact of land
use on the plant species richness. The plant species richness per hectare decreased
from the relatively less modified (natural vegetation) to the highly modified
landscapes (cropland, monoculture plantation and pasture). This agrees with most
studies on the impact of land use change on plant species richness and diversity
(Bobo et al., 2005; Bremer and Farley, 2010; Getachew Tadesse et al., 2014). More
species per hectare was recorded from the woodland and followed by the DNF. The
SFC was in the third place in terms of plant species richness per hectare. The SFC
are high in plant species richness per hectare compared to the monoculture
plantations, pasture and cropland. Plant species richness declined from the degraded
natural vegetation to the monoculture plantation forests, to the cropland and pasture.
Different land use types have different number of plant species richness per hectare.
Natural vegetation (forest, woodland and grassland) have been converted to
cropland, pasture and manmade monoculture plantation forests by compromising the
plant species richness and abundance. In this study, the cropland was relatively
species poor per hectare compared to all other land use types. This agrees with Bobo
133
et al. (2005) in which the conversion of forest to cropland affected the plant species
diversity and richness. Following the cropland, least species richness was recorded
from the manmade monoculture plantations of exotic species. The site occupied by
the manmade monoculture plantations of exotic species today, were occupied by
natural vegetation in the past. The manmade plantations were introduced in the area
in late 1970s by clearing the existing dense natural forests of the area. This was
revealed from the land use map and personal communications with the local elders
who have good knowledge about the vegetation change in the area and from the
District agricultural office. The finding agrees with Bremer and Farley (2010).
According to these authers plantation forests help in conservation of biodiversity
when applied on degraded land, but not when they replace the natural vegetation.
Land use change affected not only the plant species richness, but also abundance of
the tree species. The highest abundance was recorded from the plantation forests
(798.25 ha-1) followed by DNF (236 ha-1) and SFC (129.7 ha-1). The motive behind
the plantation forests was commercial benefit and as a result the trees have been
planted at regular intervals and have been protected until they mature for the planned
purpose. The plantation forests in the study transect was owned by the state that is
protecting the forests from exploitation by the local community. The abundance in
the plantation forests was attributed by the management inputs and the protection
provided to maximize the income up on selling the trees for timber.
134
The natural forests in the transect have been used as a common pool for different
purposes such as poles and vines for construction purposes, fire wood for fuel and
other ecosystem goods and services. There are frequent illegal felling of trees for
logging, fuel and house contrustruction; this impact is compounded by cattle grazing
due to shortage of pastureland for the communities around the natural forest. The
combined effect of illegal felling and grazing by animals has greatly affected the
germination and recruitment of plant species to the adult stage that has highly
affected the abundance of trees and shrubs in the natural forest.
SFC is modified natural forest with wild coffee beneath the canopy. In the
modifications of natural forests to the SFC, some individual trees have been
removed, while others are retained for shade provision for the coffee shrubs/trees
beneath. Less abundance of trees in the SFC compared to the DNFs was attributed by
purposive removal of some individual trees as to maximize the growth of and yield
obtained from coffee. This agrees with Demel Teketay (1999b; Kitessa Hundera et
al. (2013). Woodland occupied the forth place in the tree species abundance whereas
the cropland and pasture were least abundant in woody species due to the clearing of
natural vegetation for the expansion of agriculture and livestock farming. Clearing of
natural vegetation affected the diversity and abundance of plant species.
5.1.3. Basal area
Compared to all land use types in the transect, the plantation forests were the most
important land use types in terms of basal area per hectare and this was attributed
mainly by management inputs. The basal area for both DNF and SFC in the study
135
area are less than the normal basal area value for virgin tropical forests in Africa
(Lamprecht, 1989). Compared to other studies conducted in different natural forests
of Ethiopia (Tamrat Bekele, 1993; Tamrat Bekele, 1994; Abate Ayalew, 2003;
Kitessa Hundera, 2003; Kumelachew Yeshitela and Taye Bekele, 2003; Simon
Shibru and Girma Balcha, 2004; Ermias Lulekal, 2005; Genene Bekele, 2005; Dereje
Denu, 2007), the basal areas for DNF and SFC in the transect are the least. This is
due to degradation of the natural forest and selective removal of some trees during
the conversion of natural forest to the SFC systems.
The most important tree species contributing higher basal area in the SFC are Albizia
gummifera, Croton macrostachyus, Ficus mucuso and Cordia africana. Croton
macrostachyus and Albizia gummifera have also been reported as important canopy
trees in SFC (Driba Mulleta et al., 2007; Aerts et al., 2011). Driba Mulleta et al.
(2007) also reported C. africana as important coffee shade tree. The most important
tree species contributing higher basal area in the DNFs were Ficus sur, Apodytes
dimidiata, Schefflera abyssinica, Syzygium guineense, Albizia gummifera and Celtis
africana. As it was indicated in Dereje Denu (2007), Ficus sur was also one of the
top ten tree species with high basal area in Bibita Forest.
5.1.4. Plant growth forms
Of the four plant growth forms herb was the richest in species composition compared
to shrub, tree and liana across the transect and this agrees with Tadesse
Woldemariam (2003) in Yayu forest, Schmitt (2006) in Bonga Forest, Dereje Denu
136
(2007) in Bibita Forest– all in southwest Ethiopia. Trees ranked second in species
richness across the transect. This disagrees with Dereje Denu (2007) and Dereje
Denu and Tamene Belude (2012) in which the tree species ranked third, while the
shrub ranked second in Bibita Forest and had the same rank with herbaceous species
in sacred forests of Bedele District, but it agrees with Schmitt (2006).
The herbs dominated all land use types in the study transect except the plantation
forests where its richness follows the tree growth form. The herbaceous species
distribution in the DNFs (n = 36, ~32%), woodlands (n = 56, ~41.18%), cropland (n
= 50, ~55%), SFC (n = 64, ~44%), pasture (n = 51, ~45%) and in plantation forests
(n = 27, ~34%). The number of herbaceous species per hectare was the least in DNF,
while it was the highest in cropland. Compared to all land use types, except in the
DNF, the number of herbaceous species per hectare was the least in plantation
forests. The number of herbaceous species is less in plantation forests than in pasture
or woodland and higher than in the natural forests. This disagrees with Kamo et al.
(2002), which could be due to the degradation of the natural forest by anthropogenic
activities.
The richness of liana species across the transect was lowest compared to herbs,
shrubs and tree species richness. This agrees with Tadesse Woldemariam (2003) in
Yayu Forest, Getaneh Belachew (2006) in Beshilo and Abay riverine vegetation,
Dereje Denu (2007) in Bibita Forest. The least number of liana species richness per
hectare was recorded from the cropland followed by the SFC with 0.29 and 0.34 ha-1
137
respectively. The lianas were highly compromised in the conversion of natural
vegetation to the cropland and SFC. When the forest is converted to cropland, the
lianas are also removed with the trees and shrubs. In the conversion of forest to SFC,
where the canopy trees are recruited, lianas have no chance to be recruited for shade
provision by the coffee growers. Lianas are removed during the thinning activities
due to their negative impact on the growth of coffee shrubs/trees and blocking access
during harvesting of the ripe coffee berries. The highest record of lianas was in the
DNF where it was highly successful than in any land use types along the transect.
The impact of human induced disturbances on liana success was also shown
somewhere by other authors (Schnitzer et al., 2004; Addo-Fordjour, 2009;
Rutishauser, 2011). The illegal felling favors the expansion of lianas which
completely or partially covers the passage of light to the forest floor and the open
sites are affected by the grazing animals. In the pasturelands, the number of
herbaceous species was higher, while the woody species including the trees are rare.
A study on pasturelands (Tracy and Sanderson, 2000) in America also showed a
similar result with the finding of this study.
Of the 287 plant species recorded from the transect, nine were the most frequent ones
occurring in about 50% of the study plots. Five of them were trees, three herbs and
one shrub. Among the trees, Albizia gummifera and Acacia abyssinica are protected
by the community for their shade provision for the coffee shrubs; Cordia africana
has been protected for its various uses such as raw material for making household
furniture and as coffee shade tree as it was indicated in Diriba Muleta et al. (2011).
Species with rare occurrence in the study transect were Eucalyptus camaldulensis,
138
Grevillea robusta, Kosteletzkya begoniifolia, Nuxia congesta, Pinus patula,
Schrebera alata and Sesbania sesban. Eucalyptus camaldulensis, Grevillea robusta,
Pinus patula and Sesbania sesban are exotic species and they all are in manmade
plantations except Sesbania sesban which has escaped from the home gardens and
naturalized in the wild. Kosteletzkya begoniifolia, Nuxia congesta and Schrebera
alata are indigenous species and have rare occurrence in the study area.
The SFC was classified in DNFs into lower storey, middle storey and upper storey
using the classification scheme of IUFRO (Lamprecht, 1989). The number of tree
species with the canopy remaining in the lower storey is greater than the number of
tree species reaching the middle and upper storeys in the SFC. This disagrees with
Getaneh Belachew (2006) and Dereje Denu (2007). The coffee growers retain some
tree species in the coffee plot not only for shade provision, but also for other
purposes such as house construction, building fences around their home garden and
source of fire wood. They maintain the trees until they reach the stage that could help
them for the mentioned purposes. In the DNF, where the canopy is highly dominated
by lianas and the passage of light to the forest floor is blocked, the tree species
richness is lower in the lower storey than in the middle. The forest is degraded due to
uncontrolled human activities, which also facilitated the domination of the lianas as a
canopy cover. In such liana dominated canopies, the germination and recruitment of
the seedlings into saplings and then into trees is compromised. It allows only the
germination and growth of shade loving species, while those light seeking species are
unfavored. The tree species abundance in the middle storey is greater than that in the
lower and upper storeys both in SFC and in DNF. This is in agreement with Getaneh
139
Belachew (2006), even though, the comparison is not a direct one due to spatial and
temporal variations and also variability of other underlying environmental factors; it
is to give some indication about the similarity and differences with other studies.
Among the most dominant canopy tree species in SFC, Albizia gummifera and
Croton macrostachyus are at forefront and this agrees with Aerts et al. (2011).
Few species dominated the upper storey compared to the middle and lower storey.
The dominance is high when few species dominate the canopy and the dominance is
low when several species are evenly distributed. The storey with high species
richness is less in dominance, while the upper storey with less diversity is more
dominant. This is because; few species dominated the upper story, while the lower
storey is with several species with relatively even distribution. In general, in SFC, the
diversity increased from the lower to the upper story via the middle storey. In DNF,
the diversity is higher in the middle which is followed by the lower storey.
The upper storey in DNF is relatively with more diversity (12 species) compared to
the diversity (3 species) in upper storey of SFC. The cluster analysis using Jaccard
similarity index as distance measurer grouped the study plots into three. All plots of
forests (SFC, DNF and planation forests) were grouped together because they share
more common traits with each other than with the other remaining land use types.
140
5.1.5. Above ground live carbon storage
LULC changes are among the anthropogenic contributors to the global carbon
emissions (Friedlingstein et al., 2010; Houghton, 1999). All land use types are not
equally important in AGC storage. In this study, AGC storage (AGC t ha-1) varies
from land use to land use across the study transect in the Jimma Highlands. The
highest AGC was recorded in the plantation forest followed by DNF and the least
was in the cropland and pasture.
The management input which increased the tree species density in the plantation
forests also contributed to the relatively higher AGC storage. The conversion of
natural forests to DNF along the transect affected the amount of AGC in the above
ground tree biomass. The AGC in DNF in this study was lower than the amount of
AGC reported by Tadesse et al. (2014); Yohannes et al. (2015); Brown (1997) and
WBISPP (2005). This is mainly due to the anthropogenic activities exerted on the
DNF from the surrounding villages.
The human land use change affected forests which are important in the global carbon
balance. As it was indicated in FAO (2010) Global forest resources store about 289
Gt of carbon. This very important global resource is affected by human land use
change. The woodlands followed SFC in the amount of AGC stored in the above
ground woody species biomass. Compared to the carbon storage in woodlands
reported by WBISPP (2005) in Ethiopia, the amount of AGC calculated in the
woodlands of this study area was lower. This difference actually emanates from the
141
methodologies and tools applied and the level of anthropogenic activities exerted on
the woodlands. As indicated by Asner et al. (2003) land cover change is the most
important factor that impacts on the AGC storage in woody vegetation.
Compared to croplands and pasture, SFCs are very important in their AGC storage.
In this system, the natural forests have been modified through thinning by coffee
growers for better coffee yield. It is one of the coffee management systems in
Ethiopia (Demel Teketay, 1999b; Wiersum et al., 2008; Feyera Senebeta et al., 2009;
Schmitt et al., 2010; Kitessa Hundera et al., 2013). More AGC t ha-1 storage was
calculated for SFC than in croplands and pastures. This agrees with WBISPP (2005).
The amount of carbon stored in SFC varies depending on management intensity,
which in turn varies by region and tradition. For example, compared to carbon
storage in nearby natural forests, coffee agro-forests have been reported to retain
42% of AGC carbon in Panama (Kirby and Potvin, 2007), 49% in Indonesia (Kessler
et al., 2012) and 50-62% in Yeki and Decha of Ethiopia (Getachew Tadesse et al.,
2014).
The most important tree species in AGC storarge in the SFC are Acacia abyssinica,
Albizia gummifera, Cordia africana, Croton macrostachyus, Dracaena steudneri,
Ficus mucuso and Millettia ferruginea. Among these tree species, Acacia abyssinica,
Albizia gummifera, Cordia africana and Millettia ferruginea were reported by Diriba
Mulleta et al. (2011) as important coffee shade trees in southwest Ethiopia. These
species are also the first four tree species in farmers’ preference as coffee shade trees
142
in this study. Cordia africana is also retained on farm for its good quality timber.
The coffee farmers look after these tree species in their coffee farm for many years
and as a result the trees are relatively with high DBH and stem density. The higher
AGC storage was attributed by the management inputs in the conservation of the
selected coffee shade trees. Albizia gummifera and Croton macrostachyus were also
reported by Aerts et al. (2010) as important coffee shade trees forming the dominant
canopies in some areas of southwest Ethiopia. Ficus mucuso was not among the top
tree species of choice by the farmers for shade provision. This species dominated the
coffee forest along Didessa River and was not recoded from other plots. The species
is characterised by its large trunk diameter which has contributed to the highest
carbon storage in its above ground biomass.
The five most important plant families in AGC in the SFC of the study area were
Fabaceae, Moraceae, Euphorbiaceae, Boraginaceae and Dracaenaceae. These are the
families to which the above most important coffee shade trees belong. In the
conversion of natural vegetation to agriculture, most trees, shrubs and lianas are
cleared, while some trees are retained on farm for different purposes. The least AGC
in tree biomass was calculated for the croplands covered by annual crops and
pasturelands. The AGC calculated for cropland in this study is within the range of
AGC calculated for croplands (1.78–2.47 t ha-1) in Ethiopia (WBISPP, 2005). The
most important tree species in the cropland was Cordia africana which the farmers
purposely retained on farm for its good timber. The species was also ecologically
important in improving soil fertility (Abebe Yadessa et al., 2009). Pasture and
143
croplands are characterised by small number of individual trees sparsely scattered
and retained for different purposes.
Experience accumulated over decades living and working in southwest Ethiopia tells
us that the profitability of traditional coffee farming is finely balanced: when the
market price of coffee drops, there often follows a wave of conversion from SFC to
cropland. If such livelihood pressures were to cause the coffee growers along our
study transect to similarly convert their land, then we estimate that 59.5 t ha-1
(conversion to cropland) or 59.0 t ha-1 (pasture) would be released as greenhouse gas
emissions into the atmosphere. This is in agreement with Achard et al. (2004);
Houghton (1999); Friedlingstein et al. (2010); Kaplan, et al. (2010) in which the
impact of LULC change in carbon emission was addressed.
AGC storage has significant linear relationships with tree species richness and
abundance. These two varibales (richness and abundance) explained about 82% of
the variation in carbon storage. AGC storage was calculated from woody species
with DBH
10 cm. Most of the woody species in the transect with this DBH are
trees. That is why tree species abundance and richness were highly correlated with
AGC storage. This agrees with Strassburg et al. (2010).
AGC storage also showed significant linear relationships with some climate variables
such as mean temperature warmest quarter, mean temperature coolest quarter, mean
144
annual rainfall, rainfall driest quarter, potential evapotranspiration, moisture index
moist quarter, annual moisture index, mean diurnal range in temperature, mean
annual temperature, min temperature coolest month, annual temperature range and
maximum temperature warmest month. Potential evapotranspiration explained about
21% of the variation in carbon storage when all other variables were excluded due to
collinearity. Net primary productivity (the basis for carbon storage in the woody
species biomass) decreases by dry weather and warmer temperatures (Tian, et al.,
1998).
Among the edaphic variables, CEC, sand and pH showed significant linear
relationship with AGC. Cation exchange capacity and pH negatively correlated to
AGC, while sand showed positive relationship with AGC. AGC decreases with
increasing CEC and soil pH, but increases with increasing percentage of sand
particle. In the regression analysis, combined together, soil pH, CEC and sand
significantly explained the variation in AGC t ha-1 across different land use types
along the transect.
5.1.6. Leaf area index (LAI)
LAI under both Version 5 and 6 of CAN-EYE were influenced by human land use
change. The decreasing order of land use types in LAI values includes DNF, SFC,
plantation forests, woodland, cropland and pasture. This agrees with Kozlowski et al.
(1991), Mass et al. (1995) who addressed the variation in LAI among ecosystems
and within ecosystems respectively. The three forest types (SFC, DNFs and
145
plantation forests) and woodlands were not statistically different in LAI_true_v5 and
v6. In the same way cropland and pasture were not statistically different in LAI.
Almost all canopy trees have been removed in the conversion of forests and
woodlands to croplands and pasture in the study transect. Few scattered trees have
been retained in the croplands and pastures. This has contributed to relatively lower.
LAI values for the canopy trees in these land use types. There were also slight
differences in LAI between cropland and pasture. Trees with more canopies were
found in the croplands than in the pasture contributing to more LAI in the cropland.
The three forest types (DNF, SFC and plantation) are different in LAI. The DNFs
were relatively higher in LAI compared to SFC and plantation forests. SFC is the
result of modification of natural forest through thinning activities in the conversion
to SCFs. Lianas and shrubs are totally removed in the modification of natural forests
to the SFC. In addition to lianas and shrubs, the individual trees are removed as to
allow enough light to the coffee shrubs. Therefore; though it is degraded, the natural
forest in the study transect has relatively more tree density and lianas than the SFC
and all other land use types in the transect. This has contributed to more LAI in the
DNF than in the SFC. The DNF is composed of broad leaved indigenous tree
species, while the plantation forests are all with exotic species and are mostly needle
leaved. The LAI in DNF is higher than the LAI in the plantation forests most
probably attributed by the variation in the leaf morphology. In the DNF, the lianas
climbing to the canopy of trees also have contribution to the LAI, while lianas as
canopy component are absent from the plantation forests. The woodlands are
composed of short trees, shrubs and grasses. The canopies are not closed as it has
been observed in the DNF and are also relatively less dense. This has contributed to
146
lower LAI value compared to the three forest types. Though there were variations in
the three forest types and woodlands in terms of LAI, the variations were not
statistically significant, most probably due to the similarity in the canopy cover. In
the woodlands, shrubs, taller herbaceous species contributed to the canopy.
LAI_true was strongly correlated with tree species basal area, richness and
abundance. Leaf area is a very important biophysical factor that plays an important
role in photosynthesis and primary productivity. The biomass accumulated in the
trees and other woody species is the result of this primary productivity. The growth
in basal area is the result of biomass accumulation as a result of primary productivity.
The strong linear relationship between LAI and woody species basal area, most
probably, emanates from the physiological relationships they have. When the
abundance of tree species is compromised due to the conversion of natural forest- as
in the SFC and croplands, the LAI is also compromised. This indicates the
importance of density in its contribution to the LAI.
Herbaceous species richness was not significantly related to both LAI_true_v6 and
LAI_true_v5. In the DNF where the LAI was relatively higher, the herbaceous
species richness was poor. The growth of light loving herbaceous species is
compromised under the canopies of natural and plantation forests. The management
inputs in SFC and plantation forests also impacted the herbaceous species growth.
147
Topographic factors such as elevation and slope have no strong relationship with the
LAI. In the study transect, land use was found more important than topographic
factors. The natural vegetation has been changed to different land use types
irrespective of elevation and slope. The areas where more species richness and
abundance are expected may have less number of species abundance due to the
human land use change. At the same time, at elevations where less number of species
and abundance are expected, you may come up with more species than at the
elevation with more theoretical species richness and abundance.
Among the edaphic factors, soil cation exchange capacity, sand and clay have
significant relationship with the LAI. CEC and clay were negatively related with
LAI, while the sand was positively related. These three edaphic factors have also
significantly explained the variation in LAI. EVI and NDVI are significantly related
to the LAI. This agrees with Goswami et al. (2015) that showed strong correlation
between LAI and NDVI. In the multiple regression analysis, the two variables
combined, significantly explained the variation in LAI. NDVI has significant
contribution to the model, while the contribution of EVI was not significant singly.
LAI and AGC storage: Both LAI_true_v6 and LAI_true_v5 have strong linear
relationship with the above ground carbon storage across the land use types in the
study transect. LAI_true_v6 has explained about 75% of the variation in AGC t ha-1.
Land use types have significant variation in AGC t ha-1. The LAI which is influenced
148
by human land use change has significantly explained the variation in AGC, which is
also influenced by human land use change.
LAI and climate variables: As very important biophysical element, leaf acts as an
interface between the plant canopy and the atmosphere. It is a very important site for
absorption of energy from the sun, for gas exchange and regulation of water loss.
Leaf area has a direct role to play in this important biological process in green plants.
The significant linear relationships between LAI and climate variables are attributed
to these natural interactions. LAI showed significant linear relationships with climate
variables such as mean temperature warmest quarter, mean annual rainfall, maximum
temperature warmest month, mean annual temperature, annual temperature range,
mean temperature coolest quarter, moisture index moist quarter,mean diurnal range
in
temp,
rainfall
driest
quarter,
annual
moisture
index
and
potential
evapotranspiration. This agrees with Jin and Zhang (2001), Xavier and Vettorazzi
(2003) and Luo et al. (2004) also showed the existence of linear relationship between
LAI and precipitation. All the above climate variables showed strong collinearity in
the multiple regression analysis. The two climate variables with relatively low
collinearity value were mean annual temperature and annual temperature range.
These two variables combined, have significantly explained the variation in LAI
across the land use types in the study transect.
The LAI data taken below and above coffee canopy (taken above the canopy of
coffee shrubs/trees and below the canopy of coffee shrubs/trees) are important to
149
determine the contribution of coffee canopy to the LAI across the coffee agroforestry in the transect. In the LAI data taken below the coffee canopy, it is obvious
that the contribution of coffee canopy to the LAI was included, while, in the LAI
data taken above the coffee canopy, the contribution of coffee canopy to the LAI was
excluded. The LAI taken below the coffee canopy was higher than the LAI taken
above the coffee canopy due to the inclusion and exclusion of the coffee canopy
respectively. The significant variation between the LAI taken above the coffee
canopy and LAI taken below the coffee canopy showed a significant contribution of
coffee canopy to the LAI. Therefore; it is possible to deduce that the coffee canopy,
in addition to the canopy of shade trees, has contribution in regulating the
microclimates under the canopy. Hardwick et al. (2015) also showed the importance
of LAI in regulating the microclimate beneath the canopy. According to him, the air
under the canopies having high LAI is cooler and has high relative humidity during
the day. The canopy below the coffee canopy could have cooler climate than the
canopy above the coffee canopy.
5.1.7. Species distribution
The model output showed that the lowlands of Ethiopia are not suitable for the
distribution of Cordia africana. This agrees with the description of elevational
distribution of the species by Riedl and Edwards (2006). The areas which are
climatically most suitable for the distribution of C.africana are parts of southwest
highlands such as Jimma, Kaffa, Bench-Maji, Illubabbor, Shaka; central and
southern Oromia and eastern Hararghe Highlands. These areas are within the range
150
of elevations reported by Riedl and Edwards (2006) for the distribution of the
species. Most areas of Borana and Bale zones in Oromia, which are climatically
suitable under the current climate change scenarios for the distribution of C.
africana, will lose their suitability for the distribution of the species in the future.
The suitable areas in northern Ethiopia for the distribution of C. africana under the
baseline climate scenario will turn unsuitable under the projected climate.
The two climate variables contributing more to the model are mean annual
temperature and mean annual rainfall. The lowlands of Ethiopia which are not
suitable for the distribution of the species are the areas where the temperature is high
and rainfall is low (Daniel Gemechu, 1977). The southwest Ethiopian highlands
which are climatically suitable for the distribution of C. africana are characterised by
relatively high rainfall and low temperature (Daniel Gemechu, 1977). This shows
how the temperature and rainfall influence the distribution of C. africana. The areas
which are suitable for the distribution of the species under the current climate change
scenario will lose their suitability under the projected climate. This may be due to the
rise in temperature. As it was indicated by Platts et al. (2014) the temperature
increases in sub-Sahara Africa under both IPCC concentration pathways by 2100.
The mean annual temperature for the occurrence locations of C. africana will
increase by 4.3 to 5.1°C by late century (data from Platts et al., 2014).
Millettia ferruginea is one of the endemic species of Ethiopia with least concern in
IUCN Red list category (Vivero et al., 2005). The species is distributed in upland
forests and rainforests covering the elevational range from 1000–2500m above sea
151
level. The lowlands around Ethiopian highlands are not suitable for the distribution
of M. ferruginea under the baseline and projected climate scenarios. Southwest
highlands of Ethiopia, particularly the highlands of Jimma, Kaffa, Shaka, most parts
of Illuababor zones; some areas of East Wellega zone, and eastern part of West
Wellega and northern Borana zones, Sidama and Gedeo zones in south Ethiopia, and
some parts of Awi and Metekel zones are some of the most suitable areas for the
distribution of Millettia ferruginea under the baseline scenario. Under the projected
climate change scenarios, the entire Awi Zone, most parts of Metekel, most parts of
southwest Illubabor and Shaka zones will lose their suitability for the distribution of
Millettia ferruginea under the future climate change scenarios.
Western and central highlands of Jimma, central and eastern parts of Kaffa, central
and northwestern parts of Bench-Maji, south and southeastern parts of Illubabor
remain suitable areas for the distribution of Millettia ferruginea. These are the areas
with high rainfall and low temperature compared to the surrounding areas (Daniel
Gemechu, 1977; Platts et al., 2014). They are characterized by low average annual
water deficit, low temperature and high rainfall (Daniel Gemechu, 1977; Platts et al.,
2014). They are the wettest regions of the country (Daniel Gemechu, 1977).
Similarly, mean annual temperature and mean annual rainfall are the variables with
the highest contribution to the model. The rise in temperature for the points of
occurrence of M. ferruginea ranges from 4.4–5.0°C by the year 2100 (Platts et al.,
2014). Under the baseline scenario, the mean annual rainfall for the occurrence
localities of M. ferruginea ranges from 714–1868 mm. This shifts to 724–1862 mm
under the projected climate by the year 2100 (Platts et al., 2014). From this one can
152
deduce that change in temperature is more important than the change in rainfall for
the distribution of M. ferruginea.
Phytolacca dodecandra is distributed in an elevation range of 1500–3000 m above
sea level (Polhill, 2000). Elevation and temperature have inverse relationship where
the temperature decreases with increasing altitude. Mean annual temperature for the
occurrence localities of the species ranges from 10.8–21.8°C under the current
climate change scenarios (Platts et al., 2014). This temperature range is suitable for
the distribution of P. dodecandra under the baseline scenario. The mean annual
temperature ranges from 15.4–26°C under the projected climate for the occurrence
localities of P. dodecandra with rise in temperature by 4.6°C and 4.2°C for the
higher and lower elevations respectively. The areas with mean annual temperature
higher than 21.8°C are not suitable for the distribution of the species. The lower
mean annual temperature for occurrence localities of P .dodecandra shifts from
10.8°C (under baseline) to 15.4°C (under the projected climate) for the higher
elevations, while the mean annual temperature for the lower elevations shifts from
21.8–26°C (Platts et al., 2014). Based on the temperature and elevation relationship,
it is possible to deduce the elevational shift in spatial distribution of P. dodecandra if
the current climate change scenario continues. This agrees with Parmesa and Yohe
(2003).
Highlands of Kaffa, Jimma, Bench-Maji (parts of southwest highlands of Ethiopia);
west Shewa and Guragie zones (central Ethiopia); Arsi, Sidama, Borena and Gedeo
highlands, north and western Bale;and Harargie highlands are currently the most
suitable areas for the distribution of P.dodecandra. The areas with the mean annual
153
temperature above 21.8°C lost their suitability for the distribution of the species.
From the northwestern highlands, the areas of southwest Ethiopia (Jimma, Kaffa);
central Ethiopia (west Shewa and Guraghe zones) will lose their suitability for the
distribution of the species under the projected climate due to the shift of mean annual
temperature above 21.8°C. The areas of southeastern highlands (most areas of
northern and northwestern Bale; Sidama and Gedeo zones; Central Arsi and eastern
Hararghe highlands will be suitable for the distribution of the species under the
projected climate
Mean annual temperature for occurrence locations of Schefflera abyssinica under the
baseline climate change scenarios ranges from 13.8–21.5°C. This shifts to 18.4–
26.3°C under the projected climate change scenarios with a minimum change of
4.6°C and a maximum of 4.8°C.
5.2. Conclusion
1. Land use/land cover change affected plat species richness, abundance and
diversity across the study transect. The richness decreases from less modified
to highly modified land use types in the trasect.
2. Plant spcies richness and abundance showed significant linear relationships
with some climate variables across the transect. Therefore, climate has
impacts on the richness and abundance of plant species.
3. In all land use types, except in planation forests, herbaceous species dominate
the species composition.
154
4. Land use/land cover change affected the above ground live carbon storage
with the maximum in the planation forests followed by DNF and SFC and the
minimum in cropland.
5. The linear relationship between above ground live carbon storage and climate
variables shows that climate affects its storage.
6. Tree species richness and abundance vary along the vertical stratification in
SFC and DNF. In both cases, the middle storey is with the highest species
abundance followed by the lower storey.
7. As in plant richness and carbon storage, LAI also varies with land use types
and the differences were statistically significant. This shows how the land use
types influence the distribution of LAI. The three forests (SFC, DNF and
plantation forests) and woodlands were not significantly differed in LAI.
8. Leaf area index highly influenced the above ground live carbon storage. It
explains about 75% of the variation in carbon storage across the land use
types.
9. Leaf area index has linear relationships with climate variables such as mean
annual rainfall, max temp warmest month, mean annual temperature and
potential evapotranspiration. These climate variables could affect LAI
directly or indirectly.
10. SFC is significantly different from pasture and cropland in LAI, while the
difference with DNF, plantation forest and woodland was not significant.
11. Topographic variables did not significantly explain the variation in LAI
across the study transect, while land use types did well.
155
12. Significant statistical variation was obtained between the LAI taken above the
coffee canopy and the LAI taken below the coffee canopy. This shows the
importance of coffee canopy in the contribution to the total leaf area index
when taken below the coffee canopy.
13. The distribution model of five plant species across Ethiopia showd how the
climate change affects their distribution both under the present and projected
scenarios.
14. Most of the areas which are suitable for the distribution of these species under
the current climate will turn unsuitable under the most extreme representative
concentration pathways (RCP8.5).
5.3. Recommendations
1. In the transect, most of the grasslands and some areas of forests have
been converted to farm lands of annual crops by the small holder
subsistence agriculture. Each year, new areas have been added to
agricultural land. This is to get more yields to satisfy the increasing
family members. To combat this problem, new agricultural
technologies by which the farmers could get sufficient yield from
small areas of land should be introduced in the area.
2. Most of the areas in the transect are covered by SFC, while the DNF
is confined to smaller areas located above 2000 m elevation. Farmers
tend to convert coffee forests to alternative land use types like
croplands during yield loss and failure in coffee price. This
compromises the plant species richness, diversity, carbon storage, leaf
156
area index and other ecosystem services from which the community
could benefit. To make the SFC sustainable in provision of ecosystem
goods and services, there should be a mechanism by which the
farmers could be supported during the yield loss and failure in market
price.
3. SFC is in the third place in carbon storage in the biomass of the coffee
shade trees. They are important sinks of carbon because the coffee
shade trees are looked after by the coffee farmers and stay longer in
the coffee plot providing shade for the coffee trees/shrubs beneath.
The SFC forests satisfy the forest definition for REDD+ mechanism.
Therefore, SFC should be considered in any climate debates and the
coffee growers should be benefited from the carbon funds.
4. The DNF is confined to the upper part of the transect and is used as a
common pool for provison of materials for construction, timber and
non-timber forest products and other ecosystem goods and services
for the surrounding community. There is also a pressure from the
community to convert the forest to croplands which strongly
compromises the species richness, diversity, carbon storage, leaf area
index and other ecosystem goods and services. Therefore, there
should be a mechanism by which the community around the forest
could get sustainable benefit from the forest around them. Non-timber
forest products like honey production could sustainably benefit the
community and also conserve the forest.
157
5. Climate change affects the distribution of plant species. Most of the
areas which are currently suitable for the distribution of the species
may become unsuitable in the future if the current climate change
scenario continues. Climate has no boundary and the measure to be
taken needs integrated effort across the world. Therefore, humanity
across the world should agree to fight the climate change through
reduction in greenhouse gas emissions.
158
References
Abate Ayalew (2003). Floristic composition and structural analysis of the Denkoro
forest. M.Sc Thesis. Addis Ababa University.
Abebe Yadessa, Fisseha Itanna, Olsson, M. (2009). Scattered trees as modifiers of
agricultural landscapes: the role of waddeessa (Cordia africana Lam.) trees in
Bako area, Oromia, Ethiopia. Afri.J.Eco. 47: 78–83.
Achard, F., Eva, H. D., Mayaux, P., Stibig, H-J., Belward, A. (2004). Improved
estimates of net carbon emissions from land cover change in the tropics for
the 1990s. Glob. Biogeochem.Cyc. 18(2): doi: 10.1029/2003GB002142
Addo-Fordjour, P., Anning, A. K., Larbi, J. A., and Akyeampong, S. (2009). “Liana
species richness, abundance and relationship with trees in the Bobiri forest
reserve, Ghana: impact of management systems.” For.Eco.and Manag.257,
no.8: 1822–1828.
Aerts, R., Kitessa Hundera; Gezahegn Berecha; Gijbels, P., Baeten, M., Van
Mechelen, M., Hermy, M., Muys, B., Honnay, O. (2011). Semi-forest coffee
cultivation and the conservation of Ethiopian Afromontane rainforest
fragments. For.Ecol. and Manag. 261, no.6: 1034–1041.
Aklilu Lemma, Brody, G., Newell, G.W., Parkhurst, R. M., and Skinner, W.
A.(1972). Studies on the Molluscicidal Properties of Endod (Phytolacca
dodecandra): I. Increased Potency with Butanol Extraction. The J. of Parasit.
58, No.1: 104-107
Anderson, R.P., Peterson, A.T., Gòmez-Laverde, M. (2002). Using niche-based GIS
modelling to test geographic predictions of competitive exclusion and
competitive release in South American pocket mice. Oikos 98: 3–16.
159
Asner, G. P., Archer, S., Hughes, R. F., Ansley, R. J. and Wessman, C. (2003). Glob.
Chang. Biol. 9(3): 316–335.
Avissar, R., da Silva, R.R. and Werth (2004). Implications of tropical deforestation
for regional and global hydroclimate.Ecosyst. Land use change. 153: 73–83.
Bellard, C., Bertelsmeier, C., Leadley, P., Thuiller, W. and Courchamp, F. (2012).
Impacts of climate change on the future of biodiversity. Ecology Letters,
(2012) doi: 10.1111/j.1461-0248.2011.01736.x
Berhanu Alemu, Getachew Animut and Adugna Tolera (2013). Millettia ferruginea:
An endemic legume tree as forage for ruminants in southern and northwestern Ethiopia. Livestock Rese for Rural Development, 25(3), Article #44.
Retrieved May 23, 2015.
Bobo, K.S., Waltert, M., Sainge, N.M., Njokagbor, J., Fermon, Heleen, and
Mu¨Hlenberg, M. (2005). From forest to farmland: species richness patterns
of trees and understorey plants along a gradient of forest conversion in
Southwestern Cameroon. Biodiv. and Con.15: 4097–4117. doi
10.1007/s10531-005-3368-6.
Breitenbach, V. F. (1961). Forests and woodlands of Ethiopia- a geobotanical
contribution to the knowledge of the principal plant communities of Ethiopia
with special regard to forestry. Ethiop. For. Rev. 1: 5–16.
Bremer, L.L. and Farley, K.A. (2010). Does plantation forestry restore biodiversity
or create green deserts? A synthesis of the effects of land-use transitions on
plant species richness. Biodiv.Cons. 19: 3893–3915.
160
Brown, S. (1997). Estimating Biomass and Biomass Change of Tropical Forests: A
Primer. FAO For. Paper 134. FAO, Rome.
Cardinale, B., Duffy, J.E., Gonzalez, A., Hooper, D.U., Perrings, C., Venail, P.,
Narwani, A., Mace, G.M., Tilman, D., Wardle, D.A., Kinzig, A.P., Daily,
G.C., Loreau, M., Grace, J.B., Larigauderie, A., Srivastava, D.S. & Naeem, S.
(2012). Biodiversity loss and its impact on humanity. Nat. 486: 59-67.
Chave, J., Coomes, D.A., Jansen, S., Lewis, S.L., Swenson, N.G. & Zanne, A.E.
(2009) Towards a worldwide wood economics spectrum. Eco.Lett. 12: 351–
366.
Chave, J., Réjou-Méchain, M., Búrquez, A., Chidumayo, E., Colgan, M.S., Delitti,
W.B.C., Duque, A., Eid, T., Fearnside, P.M., Goodman, R.C., Henry, M.,
Martínez-Yrízar, A., Mugasha, W.A, Muller-Landau, H.C., Mencuccini, M.,
Nelson, B.W., Ngomanda, A., Nogueira, E.M., Ortiz-Malavassi, E., Pélissier,
R., Ploton, P., Ryan, C.M., Saldarriaga, J.G. & Vieilledent, G. (2014)
Improved allometric models to estimate the above ground biomass of tropical
trees. Glob. Chan. Biol. 10: 3177–3190.
Coetzee, J.A. (1978). Phytogeographical aspects of the montane forests of the chains
of mountains on the eastern side Africa. Erdwissenschaft Forschung, 11:482–
494.
Corsi, F., Dupr`e, E., Boitani, L., (1999). A large-scale model of wolf distribution in
Italy for conservation planning. Cons. Biol.13: 150–159.
CSA (1996). The 1994 population and housing census of Ethiopia, results for
Ormoia Region. Federal Democratic Republic of Ethiopia Office of
Population and Housing Census Commission, Ethiopia.
161
CSA (2008). Summary and Statistical Report of the 2007 Population and Housing
Census, population size by age and sex. Federal Democratic Republic of
Ethiopia Population Census Commission, Ethiopia.
Daniel Gemechu (1977). Aspects of climate and water budget in Ethiopia. Addis
Ababa University Press. Addis Ababa, Ethiopia.70pp.
Davis, M., and Shaw, R.G. (2001). Range shifts and adaptive responses to
Quaternary climate change: Sci. 292: 673–679.
Dawit Abebe and Ahadu Ayehu (1993). Medicinal plants and enigmatic health
practices of northern Ethiopia. B.S.P.E., Addis Ababa.
Demel Teketay (1999b). History, Botany and Ecological Requirements of
Coffee.Walia, No.20: 28–50.
Dereje Denu (2007). Floristic composition and Ecological study of Bibita Forest
(Gura Ferda), southwest Ethiopia. M.Sc Thesis, Addis Ababa University,
Addis Ababa, Ethiopia.
Dereje Denu and Tamene Belude (2012). Floristic Composition of Traditional Sacred
Landscapes in Bedelle Woreda, Illubabor Zone, Oromia Regional State,
Ethiopia. Ethiop. J.Educ. and Sci. 8, No 1: 75–91.
Diriba Muleta, Fassil Assefa, SileshiNemomissa and Granhall, U. (2011).
Socioeconomic benefits ofshade trees in coffee production systems in Bonga
and Yayu Hurumu Districts, southwest Ethiopia: Farmers’ perceptions.
Ethiop. J. Educ.and Sci.1: 39–56.
162
EFAP (1994).The challenge for development.Ethiopian forestry action program,
EFAP, Addis Ababa.
Ehrlich, P.R., Ehrlich, A.H. (1981). Extinction: The Causes and Consequences of the
Disappearance of Species. New York: Random House.
Ermias Lulekal (2005). Ethnobotanical study of medicinal plants and floristic
compostion of the Menna–Angetu moist montane forest in Menna-Angetu
District, Bale Ethiopia. M.Sc. Thesis, Addis Ababa University.
Ermias Lulekal (2014). Plant diversity and Ethnobotanical study of medicinal plants
in Ankober Distyrict, north Shewa zone of Amhara region, Ethiopia. PhD
Dissertation, ADDIS Ababa University, Ethiopia.
Facchi, A., Baroni, G., Boschetti, M., Gandolfi, C. (2010). Comparing optical and
direct methods for leaf area index determination in a maiz crop. J.of Agr.
Eng. 1: 33-40
FAO (2000).The global outlook for future wood supplies from forest plantations.
Brwon, C., FAO Working Paper GFPOS/WP/03. FAO, Rome, Italy. 129pp.
FAO (2001). Global Forest Resources Assessment 2000. FAO Forestry Paper 140,
Rome. 511pp.
FAO (2003). State of the world’s Forests 2003. Food and Agreculture Organization
of the United Nations. Rome, Italy 126pp.
FAO (2006). Global Forest resources assessment 2005. Progress towards sustainable
forest management. FAO Forestry Paper 147, Rome. 350pp.
FAO (2010) Global Forest Resources Assessment 2010 ‐ Country Report Ethiopia.
Food and Agriculture Organisation (FAO), Rome, Italy (www.
fao.org/forestry/ fra/fra2010/en/, accessed on September 15, 2015).
163
FAO (2015). FAO assessment of forests and carbon stocks, 1990–2015. Reduced
overall emissions, but increased degradation. http://www.fao.org/3/ai4470e.pdf . Accessed on January 8, 2016.
Fearnside, P.M., Righi, C.A., Lima de Alencastro Graca, P.M., Keizer, E.W.H.,
Cerri, C.C., Nogueira, E.M., Barbosa, R.I. (2009). Biomass and greenhousegas emissions from land-use change in Brazil’s Amazonian ‘‘arc of
deforestation’’: The states of Mato Grosso and Rondonia. For.Ecol. and
Manag. 258: 1968–1978.
Pfeifer, M., Alemu Gonsamo, Disney, M., Pellikka, P., Rob Marchant, R. (2012).
Leaf area index for biomes of the Eastern Arc Mountains: Landsat and SPOT
observations along precipitation and altitude gradients. Remote Sens. of
Environ. 118:103–115.
Fichtl, R. and Admasu Adi (1994). Honeybee Flora of Ethiopia, Margraf Verlag,
Addis Ababa, Ethiopia. 511pp.
Fieber, K. D., Davenport, I. J., Tanase, M. A., Ferryman, J. M., Gurney, R. J.,
Walker, J. P. and Hacker, J. M. (2014) Effective LAI and CHP of a single
tree from small-footprint full-waveform LiDAR. IEEE Geosci. Remote Sens.
Lett. 11 (9): 1634 -1638.
Franklin, A.B., Noon, B.R., George, L.T. (2002). What is habitat fragmentation?
Stud. in Avian Biol. 25: 20–29.
Friedlingstein, P., Houghton, R. A., Marland, G., Hackler, J., Boden,T. A., Conway,
T. J., Canadell, J. G., Raupach, M. R., Ciais, P. and Le Qu´er´e, C.
(2010).Update on CO2 emissions, Nat. Geosci. 3: 811–812.
164
Genene Bekele (2005). Floristic composition and structure of the vegetation of
Magada forest, Borana zone, Oromia National Regional State. M.Sc. Thesis,
Addis Ababa University.
Gentry, A.H. (1982).Neotropical Floristic Diversity: Phytogeographical Connections
Between Central and South America, Pleistocene Climatic Fluctuations, or an
Accident of the Andean Orogeny? Annals of the Miss.Bot. Gard. 69, No. 3:
557–593.
Getachew Tadesse, Zavaleta, E. and Shennan, C. (2014). Effects of land-use changes
on woody species distribution and above ground carbon storage of forest-coffee
systems. Agri.Ecosy.and Env.197: 21–30.
Getaneh Belachew (2006). Floristic composition and structure in Beschillo and Abay
(Blue Nile) Riverine Vegetation.M.Sc Thesis, Addis Ababa University, Addis
Ababa, Ethiopia.
Gibbs, H.K., Brown, S., Niles, J.O. and Foley, J.A. (2007). Monitoring and
estimating tropical forest carbon stocks: making REDD a reality. Env. Res
.Lett. No2. doi:10.1088/1748-9326/2/4/045023.
Global Biodiversity Information Facility (GBIF) (http://www.gbif.org/). Accessed on
March 20, 2014.
Goswami, S., Gamon, J.A., Vargas, S., Tweedie, C.E. (2015). Relationships of
NDVI, Biomass and Leaf area index (LAI) for six key plant species in
Barrow, Alaska. https://dx.doi.org/10.7287/peerj.preprints.913v1.
Grabherr, G., Gottfried, M. and Pauli. H. (1994). Climate effects on mountain plants.
Nat.369: 448.
165
Guisan A., Thuiller, W. (2005). Predicting species distribution: offering more than
simple habitat models. Ecol. Lett. 8: 993–1009.
Guisan A., Zimmermann, N.E. (2000). Predictive habitat distribution models in
ecology. Ecol. Model. 135: 147–186.
Haileleul Tebicke (2002). A scan of sustainable energy, Environment and
development in Ethiopia. In: Tibebua Hectett (ed). Ethiopia civil society.
Berhanena Selam Printing Enterprise, Addis Ababa, Ethiopia.
Hailu Tadeg, Endris Mohammed, Kaleab Asres, Tsige Gebre-Mariam (2005).
Antimicrobial activities of some selected traditional Ethiopian medicinal
plants used in the treatment of skin disorders. J. of Ethnopharmac. 100: 168–
175.
Hamere Yohannes, Teshome Soromessa, Mekuria Argaw (2015). Carbon Stock
Analysis Along Altitudinal Gradient in Gedo Forest: Implications for Forest
Management and Climate Change Mitigation, Amer. J.Env. Prot. 4(5): 237–
244.
Hammer, Ø., Harper, D.A.T., and Ryan, P. D. (2001). PAST: Paleontological
Statistics Software Package for Education and Data Analysis. Palaeontologia
Electronica 4(1): 9pp.
Hansen, M. C., P. V. Potapov, R. Moore, M. Hancher, S. A. Turubanova, A.
Tyukavina, D. Thau, S. V. Stehman, S. J. Goetz, T. R. Loveland, A.
Kommareddy, A. Egorov, L. Chini, C. O. Justice, and J. R. G. Townshend.
2013. “High-Resolution Global Maps of 21st-Century Forest Cover Change.”
166
Science 342 (15 November): 850–53. Data available on-line from:
http://earthenginepartners.ap
pspot.com/science-2013-global-forest.
Hardwick, S.R., Toumi, R., Pfeifer, M., Turner, E.C., Nilus, R., Robert M. Ewers,
R.M. (2015). The relationship between leaf area index and microclimate in
tropical forest and oil palm plantation: Forest disturbance drives changes in
microclimate. Agri.and For.Meteo. 201: 187–195.
Heywood, V.H. and Watson, R.T. (1995). Global Biodiversity Assessment.
Cambridge University Press, Cambridge.1152pp.
Houghton, R. A. (1999). The annual net flux of carbon to the atmosphere from
changes in land use 1850–1990.Tellus B. 51: 298–313.
http://chiesa-gis.geography.helsinki.fi:8080/geonetwork/srv/en/main.home.
https://lpdaac.usgs.gov/products/modis_products_table/mod13q1.
http://www.nabu.de/en/aktionenundprojekte/kafa/projectarea/climateprote.
Hutchison, B.A., Matt, D.R., McMillen, R.T., Gross, L.J., Tajchman, S.J. and
Norman, J.M. (1986). The architecture of a deciduous forest canopy in
eastern Tennessee, USA. J.Ecol. 74: 635-646.
Hylander, K., Nemomissa, S., Delrue, J. & Enkosa, W. (2013). Effects of Coffee
Management on Deforestation Rates and Forest Integrity. Conservation
Biology Available at
http://www.academia.edu/5367898/Effects_of_Coffee_Management_on_Def
orestation_Rates_and_Forest_Integrity
167
IPCC (2000). Special Report on Land Use, Land Use Change and Forestry. Summary
for Policy Makers. Geneva, Switzerland.
IPCC (2001).Third Assessment Report of the Intergovernmental Panel on Climate
Change. IPCC (WG I & II) Cambridge Univ. Press, Cambridge.
IPCC (2007). Contribution of working group I to the fourth assessment reportof the
intergovernmentalpanel on climate change. Accessed on 10 January, 2012.
IPCC (2014). Climate Change 2014, Synthesis Report Summary for Policymakers,
www.ipcc.ch. Accessed on 20 November 2014.
ISRIC (2013). Soil property maps of Africa at 1 km. World Soil Information,
www.isric.org (Accessed on 28 January 2014).
Jaynes, E.T. (1957). Information theory and statistical mechanics. Phys. Rev.106:
620–630.
Jenkins, M. (1992). Species Extinction. In: Groombridge B (ed) Global Biodiversity:
Status of the Earth’s Living Resources. Compiled by the World Conservation
Monitoring Centre (WCMC), Chapman and Hall, London.
Jin, M. and Zhang, D.L. (2001). Observed variations of leaf area index and its
relationship with surface temperatures during warm seasons. Meteorol.
Atmos. Phys. 80: 117-129.
Jonckheere, I., Fleck, S., Nackaerts, K., Muys, B., Coppin, P.,Weiss, M., Baret, F.
(2004). Review of methods for in-situ leaf area index determination Part I.
Theories, sensors and hemispherical photography. Agri. For. Meteo. 121: 1935.
Kamo, K., Vacharangkura, T., Tiyanon, S., Viriyabuncha, C., Nimpila, S. and
Doangsrisen, B. (2002). Plant Species Diversity in Tropical Planted Forests
168
and Implication for Restoration of Forest Ecosystems in Sakaerat,
Northeastern Thailand. JARQ 36 (2): 111–118.
Kaplan, J. O., Krumhardt, K. M., Ellis, E. C., Ruddiman, W. F., Lemmen, C.,
Goldewijk, K. K (2010). Holocene carbon emissions as a result of
anthropogenic land cover change.The Holocene. 21(5): 775–791.
Kappelle, M., Van Vuuren, M. M.I. and Baas Pieter (1999). Effects of climate
change on biodiversity: a review and identification of key research issues.
Biodiv. and Cons. 8: 1383–1397.
Kauffman, J.B., Hughes, R.F., Heider, C. (2009). Carbon pool and biomass dynamics
associated with deforestation, land use, and agricultural abandonment in the
neotropics. Ecol. Appl.19 (5): 1211-1222.
Kessler, M., Hertel, D., Jungkunst, H.F., Kluge, J., Abrahamczyk, S., Bos, M.,
Buchori, D., Gerold, G., Gradstein, S.R., Köhler, S., Leuschner, C., Moser,
G., Pitopang, R., Saleh, S., Schulze, C.H., Sporn, S.G., Steffan-Dewenter, I.,
Tjitrosoedirdjo, S.S. & Tscharntke, T. (2012). Can joint carbon and
biodiversity management in tropical agro-forestry landscapes be optimized?
PLoS ONE 7: e47192.
Kirby, K.R. and Potvin, C. (2007). Variation in carbon storage among tree species:
implications for the management of a small-scale carbon sink project.
For.Ecol. and Manag. 246: 208–221.
Kitessa Hundera (2003). Floristic composition and structure of the Dodolla forest,
Bale zone, Oromia Regional State. M.Sc thesis, Addis Ababa University
169
Kitessa Hundera., Aerts, R., Fontaine, A., Van Mechelen, M., Gijbels, P., Honnay,
O. & Muys, B. (2013) Effects of coffee management intensity on
composition, structure, and regeneration status of Ethiopian moist evergreen
Afromontane forests. Env.Manag. 51: 801–809.
Koetz, B., Morsdorf, F., Sun, G., Ranson, K. J., Itten, K. and Allgöwer, B. (2006).
Inversion of a LiDAR waveform model for forest biophysical parameter
estimation. IEEE Geosci. Remote Sens. Lett. 3, No.1: 49–53.
Konemund, T., Hiwete Teshome and Samson Tolessa (2002). What alternatives do
we have to bridge the gap between fuel-wood demand and supply in
Ethiopia? In: Tibebura Hectett (ed.), Ethiopian Civil Society Preparation for
tiotio.
Kozlowski, T.T., Kramer, P.J. and Pallardy, S.G. (1991). The Physiological Ecology
of Woody Plants. Academic Press, New York, 657 pp.
Kriticos, D.J., Randall, R.P. (2001). A comparison of systems to analyse potential
weed distributions.In: Groves RH, Panetta FD, Virtue JG, (eds). Weed Risk
Assessment Melbourne: CSIRO Publishing.
Kumelachew Yeshitela (2001). Loss of forest biodiversity associated with changes in
land use the case of Chewaka-Utto tea plantation. Proceedings of a Workshop organized by Biological Society of Ethiopia Faculty of Science, Addis
Ababa University.
Kumelachew Yeshitela and Taye Bekele (2003). The woody species composition and
structure of Masha Anderacha forest, Southwestern Ethiopia. J.Biol. Soci.
Ethiop. 2 (1): 31– 48.
170
Lamprecht, H. (1989). Silviculture on the Tropics. Tropical forest ecosystems and
their tree species – Possibilitiesand methods for their long term utilizations.
Landon, J.R. (1996). Booker tropical soil manual. A hand book for soil survey and
agriculture and evolution in the tropics and sub-tropics. Longman.
Legesse Negash (1995). Indigenous trees of Ethiopia: Biology, uses and Propagation
Techniques. SLU Reprocentralen, Umea, Sweden. 285pp.
Martinez, M. L., Perez-Maqueo, O., Vazquez, G., Castillo-Campos, G., Garcı´aFranco, J., Mehltreter, K., Equihua, M., Landgrave, R. (2009). Effects of land
use change on biodiversity and ecosystem services in tropical montane cloud
forests of Mexico. For.Ecol. and Manag. 258: 1856–1863.
Mass, J.M., Vose, J.M., Swank, W.T., Martinez-Yrizar, A. (1995). Seasonal changes
of leaf area index (LAI) in a tropical deciduous forest in west Mexico. For.
Ecol. and Manag.74: 171–180.
McCune, B. and Mefford, M. J. (2006). PC-ORD. Multivariate Analysis of
Ecological Data.Version 5.31.MjM Software, Gleneden Beach, Oregon,
U.S.A.
Midgley, G.F., Bond, W.J., Kapos, V., Ravilious, C., Scharlemann, J.P., Woodward,
F.I. (2010). Terrestrial carbon stocks and biodiversity: key knowledge gaps
and some policy implications. Cur. Opin. in Environ. Sust. 2: 1-7.
Luo, Y., Su, B., Currie, W.S., Dukes, J.S., Finz, A., Hartwig, U., Hungate,
B., Mcmurtrie, R.E., Oren, R., Parton, W.J., Pataki, D.E., Shaw, M.R., ZAK,
D.R., and Field, C.B. (2004). Progressive Nitrogen Limitation of Ecosystem
Responses to Rising Atmospheric Carbon Dioxide. BioSci. 54 No.8: 264-270.
171
Mittermeier, R.A., Robles, G.P., Hoffman, M., Pilgrim, J., Brooks, T., Mittermeier,
C.G., Lamoreux, J. and da Fonseca, G.A.B. (2004). Hotspots Revisited.
Mexico City, Mexico: CEMEX.
Montagnini, F. and Nair, P.K.R. (2004). Carbon sequestration: An underexploited
environmental benefit of agroforestry systems. Agrofor. Syst. 61: 281–295.
Myers, N. (1998). Threatened biotas: hotspots in tropical forests. The
Environmentalist, 8: 178–208.
Myneni, R.B., Keellng, C.D., Tucker, C.J., Asrar, G. and Nemani, R.R. (1997).
Increased plant growth in the northern high latitudes from 1981 to 1991. Nat.
386: 698–702.
Nair, P. K. R., Nair, V. D. (2003). Carbon storage in North American agro-forestry
systems.In: Kimble, J., Heath, L.S., Birdsey, R. A., Lal, R. (eds): The
Potential of U.S. Forest Soils to Sequester Carbon and Mitigate the
Greenhouse Effect. CRC Press, Boca Raton, FL, USA, 333–346pp.
Neumann, H.H., Hartog, G.D. and Shaw, R.H. (1989). Leaf area measurements
based on hemispheric photographs and leaf-litter collection in a deciduous
forest during autumn leaf fall. Agr. For. Meteo. 45: 325-345.
Norman, J.M. and Campbell, G.S. (1989). Canopy structure. In: R.W. Pearcy, H.A.
Mooney, J.R. Ehleringer and P.W. Rundel (Eds), Physiological Plant
Ecology: Field Methods and Instrumentation. Chapman and Hall, New York
Marina, A. (2010). Maintaining ecological integrity and sustaining ecosystem
function in urban areas. Cur. Opin. in Env. Sust. 2: 178–184.
Newton, A.C. (2007). Forest Ecology and Conservation, A Hand book of techniques.
Oxford University Press.
172
Niang, I., O.C. Ruppel, M.A. Abdrabo, A. Essel, C. Lennard, J. Padgham, and P.
Urquhart (2014). Africa: Impacts, Adaptation, and Vulnerability. Part B:
Regional Aspects. Contribution of Working Group II to the Fifth Assessment
Report of the Intergovernmental Panel on Climate Change. In: Barros, V.R.,
C.B. Field, D.J. Dokken, M.D. Mastrandrea, K.J. Mach, T.E. Bilir, M.
Chatterjee, K.L. Ebi, Y.O. Estrada, R.C. Genova, B. Girma, E.S. Kissel, A.N.
Levy, S. MacCracken, P.R. Mastrandrea, and L.L. White (eds). Cambridge
University Press, Cambridge, United Kingdom and New York, NY, USA.
Parmesan, C. and Yohe, G. (2003).A global coherent fingerprint of climate change
impacts across natural systems. Nat. 421: 37–42.
Peterson, A.T., Cohoon, K.P. (1999). Sensitivity of distributional prediction
algorithms to geographic data completeness. Ecol.Model. 117: 154–164.
Phillips, S.J., Anderson, R.P., Schapired, R.E. (2006). Maximum entropy modelling
of species geographic distributions. Ecol. Model. 190: 231–259.
Pimm, S.L., Jenkins, C. N., Abell, R., Brooks, T. M., Gittleman, J. L., Joppa, L. N.,
Raven, P. H., Roberts, C. M., Sexton, J. O. (2014). The biodiversity of
species and their rates of extinction, distribution, and protection. Sci. 344,
Issue 6187. DOI: 10.1126/science.1246752
Platts, J.P., Omeny, P.A. and Marchant, R. (2014). AFRICLIM: high-resolution
climate projections for ecological applications in Africa. Afr.J. Ecol. 1-6.
Platts, J.P., Gereau, R.E., Burgess, N.D., and Marchant, R. (2013). Spatial
heterogeneity of climate change in an Afromontane centre of endemism.
Ecogra. 36: 518–530.
173
Polhill, R.M. (2000). Phytolaccaceae.In: Edwards, S., Mesfin Tadesse, Sebsebe
Demissew and Hedberg, I. (eds). Flora of Ethiopia and Eritrea, Magnoliaceae
to Flacourtiaceae, Vol.2 part 1.Addis Ababa, Ethiopia, Uppsala, Sweden
Polzot, C.L. (2004). Carbon Storage in Coffee Agro-ecosystems of Southern Costa
Rica: Potential applications for the Clean Development Mechanism. Master’s
thesis, pp. 1-162. York University, Toronto, Ontario, Canada.
Ponder, W. F., Carter, G. A., Flemons, P., and Chapman, R. R. (2001). Evaluation of
museum collection data for use in biodiversity assessment. Cons. Biol.15:
648–657.
Pounds, J.A., Fogden, M.P.L. and Campbell, J.H. (1999). Biological responses to
climate change on a tropical mountain. Nat. 398: 611–615.
R Core Team (2014). R: A language and environment for statistical computing. R
Foundation for Statistical Computing, Vienna, Austria.http://www.Rproject.org/.
Riedl and Edwards (2006). Boraginaceae. In: Hedberg, I., Ensermu Kelbessa,
Edwards, S., Sebsebe Demissew and Persson, E. (eds). Flora of Ethiopia and
Eritrea, Gentianaceae to Cyclocheilaceae, Vol.5, Addis Ababa, Ethiopia,
Uppsala, Sweden.
Rutishauser, S. E. (2006). Increasing liana abundance and biomass in tropical forests:
testing mechanistic explanations M.Sc. thesis, University of Wisconsin,
Milwaukee, Wis, USA.
Sala, O.E., Chapin, F.S., Armesto, J.J., Berlow, E., Bloomfield, J., Dirzo, R., HuberSanwald, E.,Huenneke ,L.F., Jackson, R.B.,Kinzig, A., Leemans, R., Lodge,
174
D.M., Mooney, H.A., Oesterheld, M.,Poff, N.L.,Sykes, M.T., Walker, B.H.,
Walker, M., Wall, D.M. (2000). Global Biodiversity Scenarios for the year
2100. Sci. 287: 1770–1774.
Sarah Tewolde_Berhan; Fagertun, R.S., AbegazKebede, Judith, N., Abay Fetien and
Trude, W. (2013).Ferric reducing antioxidant power and total phenols in
Cordia africana fruit. Afr.J.Biochem.Res., 7(111): 215–2030.
Schmitt, C.B., Senbeta, F., Denich, M., Preisinger, H. and Boehmer, H.J. (2010).
Wild coffee management and plant diversity in the montane rainforest of
southwestern Ethiopia. Afr. J. Ecol. 48: 78–86.
Schmitt, C.B. (2006). Montane rainforest with wild Coffea arabica in the Bonga
region (SW Ethiopia): plant diversity, wild coffee management and
implications for conservation. Ecology and Development Series No. 47
Schmitt-Harsh, M., Evans, T.P., Castellanos, E. & Randolph, J.C. (2012) Carbon
stocks in coffee agro-forests and mixed dry tropical forests in the western
highlands of Guatemala. Agrofor. Syst. 86: 141–157.
Schnitzer, S. A., Parren, M. P. E., and Bongers, F. (2004). Recruitment of lianas into
logging gaps and the effects of pre-harvest climber cutting in a lowland forest
in Cameroon. For.Ecol.and Manag. 190, no. 1: 87–98.
Shannon, C.E. and Wiener, W. (1949). The Mathematical Theory of Communication.
University of Illinois, Chicago, USA.
Shibru Tedla and Kifle Lemma (1998). Environmental Management in Ethiopia:
Have the National Conservation Plans Worked? Environmental Forum
Publications Series No. 1.
175
Simon Shibru and Girma Balcha (2004). Composition, structure and regeneration
status of woody species in Dindin Natural Forest, southeast Ethiopia: An
implication for conservation. Ethiop. Biol. Sci. (1) 3: 15–35.
Spehn, E.M., Rudmann-Maurer, K., Körner, C., Maselli, D. (2010). Mountain
Biodiversity and Global Change. GMBA-DIVERSITAS, Basel Global
Mountain Biodiversity Assessment (GMBA), Institute of Botany, University
of Basel, Switzerland.
Still, C. J., Foster, P. N. and Schneider, S. H. (1999). Simulating the effects of
climate change on tropical montane cloud forests. Nat. 398: 608–610.
Strassburg, B.B.N., Kelly, A., Balmford Davies, R.G., Gibbs, H.K., Lovett, A.,
Miles, L., Orme, C.D.L., Price, J., Turner, R.K., Rodrigues, A.S.L. (2010).
Global congruence of carbon storage and biodiversity in terrestrial
ecosystems. Cons. Lett. 3: 98–105.
Tadesse Woldemariam (2003). Vegetation of the Yayu Forest in SW Ethiopia:
impacts of human use and implications for in situ conservation of wild Coffea
Arabica L. populations. Ecology and Development SeriesNo. 10, 2003,
Cuvillier Verlag Göttingen.
Tadesse Woldemariam and Feyera Senbeta (2008). Sustainable Management and
Promotion of Forest Coffee in Bale, Ethiopia. Submitted to: Bale Eco-Region
Sustainable Management Programme, SOS Sahel/FARM-Africa August
2008, Addis Ababa.
Tadesse Woldemariam and Masresha Fetene (2007). Forests of Shaka: Ecological,
social, legal and economic dimensions of recent land use/land cover change176
overview and synthesis. In: Masresha Fetene (ed). Forests of Shaka
multidisciplinary case studies on impactsof land use/land cover changes,
southwest Ethiopia, Melca Mahber (Movement for Ecological Learning and
Community Action), Addis Ababa.
Talbot, J.D. (2010). Carbon and biodiversity relationships in tropical forests.Multiple
Benefits Series 4. Report prepared on behalf of the UN-REDD
Programme.School of Geography, University of Leeds, Leeds / UNEP World
Conservation Monitoring Centre. Cambridge, UK.
Talemos Seta and Sebsebe Demissew (2014). Diversity and standing carbon stocks
of native agro-forestry trees in Wenago District, Ethiopia. J. of Emerg. Trends
in Engin. and Appl. Sci. (JETEAS) 5(7): 125-132.
Tamrat Bekele (1993). Vegetation ecology of remnant afromontane forests on the
Central plateau of Shewa, Ethiopia. Acta phytogeogr. Suec. 79: 1–59.
Tamrat Bekele (1994). Phytosociology and ecology of a humid Afromontane forest
in the central plateau of Ethiopia. J. Veg. Sci. 5: 87–98.
Taye Bekele, Getachew Berhan, EliasTaye, Matheos Ersado, and Kumlachew
Yeshitela (2001). Regeneration status of moist Montane Forests of Ethiopia:
Consideration for conservation, part I: Boginda, Bonga, Masha - Anderacha
and Yayu Forests. Walya 22: 45–60.
Thomas, C.D., Cameron, A., Green, R.E., Bakkenes, M., Beaumont, L.J.,
Collingham, Y.C., Erasmus, B.F.,De Siquerira, M.F., Grainger, A., Hannah,
L., Hoghes, L., Huntley, B., Van Jaarsveld, A.S., Midgley, G.F., Miles, L.,
Ortega-Huerta, M.A., Peterson, A.T.,Phillips, O.L., Williams, S.E. (2004).
Extinction risk from climate change. Nat.427: 145–148.
177
Thornley, J. H. M. and Cannell, M. G. R. (2000). Managing forests for wood yield
and carbon storage: a theoretical study. Tree Physio. 20: 477-484.
Thulin, M. (1989). Fabaceae.In: Hedberg, I. and Edwards.S. Flora of Ethiopia,
Pittosporaceae to Araliaceae. Vol.3. Addis Ababa and Asmara, Ethiopia,
Uppsala, Sweden.
Tian, H., Melillo, J.M., Kicklighter, D.W., McGuire, A.D., Helfrich, J.V.K., Berrien
Moore, B. and Vörösmarty, C.J. (1998). Effect of interannual climate
variability on carbon storage in Amazonian ecosystems. Nat. 396: 664–667.
Tracy, B.F. and Sanderson, M.A. (2000).Patterns of plant species richness in pasture
lands of the northeast United States. Plant Ecol. 149: 169–180.
TROPICOS (Missouri Botanical Garden) (http://www.tropicos.org/). Accessed on
March 25, 2014.
UN (1992). Convention on Biological Diversity. https://www.cbd.int/doc/legal/cbden.pdf. Accessed on February 12, 2015.
UNFPA (2009). State of World Population 2009. Facing a changing World: Women,
Population and Climate.
USITC (2005). Export Opportunities and Barriers in African Growth and
Opportunity Act-Eligible Countries, pp. 347-384. Investigation No. 332-464,
USITC Publication 3785. Washington DC, USA: United States International
Trade Commission.
Van Noordwijk, M., Rahayu, S., Hairiah, K., Wulan, Y.C., Farida, A., Verbist, B.
(2002). Carbon stockassessment for a forest-to-coffee conversion landscape
178
in Sumber-Jaya (Lampung, Indonesia): fromallometric equations to land use
change analysis. Sci. in China, 45: 76–86.
Vivero, J. L., Ensermu Kelbessa and Sebsebe Demissew (2005). The Red List of
Endemic Trees & Shrubs of Ethiopia and Eritrea.Fauna and Flora
International, Cambridge, UK.23 pp.
Walter, H. (1985). Vegetation of the earth and an ecological system of the
geobiosphere, third ed.Berlin, Heidelberg, New York. 318 pp.
Watson, D.J. (1947). Comparative physiological studies in the growth of field crops:
I. Variation in net assimilation rate and leaf area between species and
varieties, and within and between years. Ann. Bot. 11: 41-76.
WBISPP (2005) A national strategy plan for the biomass sector. Addis Ababa,
Ethiopia.
Welk, E., Schubert, K., Hoffmann, M.H., 2002. Present and potential distribution of
invasive mustard (Alliara petiolata) in North America. Divers. Distributions
8: 219–233.
Wiersum, K.F., Tadesse Woldemariam Gole, Gatzweiler, F.W, Volkman, J.,
Bognetteau, E. and Wirtu, O. (2008). Certification of wild coffee in Ethiopia:
Experiences and Challenges. Forests, Trees and Liveli. 18: 9–21.
World Agro-forestry (2009).
http://www.worldagroforestry.org/treedb/AFTPDFS/Cordia_africana.PDF
Xavier, A.C. and Vettorazzi, C.A. (2003). Leaf area index of ground covers in a
subtropical watershed. Sci. Agricola, 60 (3): 425–431.
179
Yitebitu Moges, Zewdu Eshetuand Sisay Nune (2010). Ethiopian Forestry
Resources: Current status and future management options in view of access to
carbon finance. Literature review prepared for the Ethiopian climate research
and networking and the United Nations Development Program (UNDP),
Addis Ababa
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. and Chave, J. (2009). Data
from: Towards a worldwide wood economics spectrum. Dryad Digital
Repository, http://dx.doi.org/10.5061/dryad.234.
Zheng G., Moskal L.M (2009). Retrieving Leaf Area Index (LAI) Using Remote
Sensing: Theories, Methods and Sensors. Sensors, 9(4):2719-2745.
180
Appendices
Appendix 1: Soil and Potential evapotranspiration data for the study plots in the
Jimma Highlands (CEC = cation exchange capacity, OC = organic carbon, BLD =
bulk density, PET = potential evapotranspiration)
Land Use
SFC 1
SFC 2
SFC 3
SFC 4
SFC 5
SFC 6
SFC 7
DNF1
DNF 2
DNF 3
DNF 4
Pasture 1
Pasture 2
Pasture 3
Pasture 4
Pasture 5
Woodland 1
Woodland 2
Woodland 3
Woodland 4
Cropland 1
Cropland 2
Cropland 3
Cropland 4
Cropland 4
Cropland 6
Cropland 7
Plantation 1
Plantation 2
Plantation 3
Plantation 4
Sand
33.500
36.167
35.000
35.500
34.833
32.167
32.333
32.167
34.167
34.333
33.500
30.167
28.333
31.500
30.833
32.000
36.167
33.000
32.333
33.500
32.167
32.333
33.667
30.833
30.667
33.010
32.510
33.000
32.167
31.167
32.500
Clay
37.667
37.333
37.833
40.333
38.833
43.667
41.167
43.667
37.667
38.167
41.333
42.333
43.333
42.500
43.500
42.167
37.333
39.833
41.000
39.500
42.333
41.167
41.167
43.500
41.333
39.833
43.657
42.167
42.333
42.167
43.667
Silt
28.667
26.833
27.000
23.833
26.500
24.167
26.667
24.167
27.667
27.667
25.333
27.167
27.833
25.833
25.667
25.833
26.833
26.833
26.667
26.667
25.333
26.667
25.333
25.667
27.833
26.833
24.000
24.667
25.333
26.667
24.000
181
PH
5.067
5.617
5.467
5.517
5.567
5.267
5.433
5.317
4.983
5.017
4.983
5.567
5.483
5.517
5.367
5.683
5.283
5.533
5.567
5.467
5.067
5.433
5.567
5.633
5.383
5.533
5.257
5.017
5.067
5.266
5.267
CEC
(ds/m)
0.202
0.222
0.225
0.233
0.235
0.253
0.233
0.253
0.192
0.198
0.193
0.242
0.263
0.223
0.253
0.243
0.222
0.243
0.210
0.238
0.200
0.233
0.232
0.253
0.203
0.243
0.261
0.203
0.200
0.220
0.262
OC (kg)
18.667
16.833
16.167
18.833
17.500
21.667
18.667
21.667
17.833
18.833
18.667
22.000
22.000
17.833
18.667
22.500
16.833
17.167
20.167
18.833
17.333
18.667
19.333
18.667
20.167
17.167
23.001
18.833
17.333
22.167
23.000
BLD (kg)
1138.500
1133.333
1127.167
1134.167
1108.500
1158.500
1195.833
1158.500
1164.167
1124.167
1147.667
1161.833
1152.000
1157.833
1186.833
1159.833
1133.333
1084.667
1226.833
1095.833
1128.167
1195.833
1136.333
1186.833
1178.333
1084.667
1119.657
1137.833
1128.167
1249.833
1119.667
PET
(mm)
1674
1759
1811
1818
1726
1814
1715
1755
1655
1650
1655
1748
1785
1826
1823
1785
1663
1674
1718
1715
1674
1807
1795
1826
1795
1718
1748
1759
1816
1814
1795
Appendix 2: Species list, percent and relative frequencies in the DNF
S.N
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Species name (in DNF)
Acanthus eminens
Achyranthes aspera
Adiantum poiretii
Ageratum conyzoides
Albizia gummifera
Albizia schimperiana
Allophylus abyssinicus
Allophylus macrobotrys
Apodytes dimidiata
Arthraxon micans
Asparagus racemosus
Asplenium aethiopicum
Asplenium formosum
Bersama abyssinica
Bidens pilosa
Brucea antidysenterica
Calpurnia aurea
Canthium oligocarpum
Carissa spinarum
Cassipourea malosana
Celtis africana
Chionanthus mildbraedii
Cissus petiolata
Clausena anisata
Coffea arabica
Combretum paniculatum
Conyza bonariensis
Cordia africana
Crassocephalum rubens
Croton macrostachyus
Cynodon aethiopicus
Cyphostemma cyphopetalum
Dalbergia lactea
Desmodium repandum
Dichrocephala integrifolia
Doryopteris concolor
Dracaena afromontana
Dracaena steudneri
Family
Acanthaceae
Amaranthaceae
Adiantaceae
Asteraceae
Fabaceae
Fabaceae
Sapindaceae
Sapindaceae
Icacinaceae
Poaceae
Asparagaceae
Aspleniaceae
Aspleniaceae
Melianthaceae
Asteraceae
Simaroubaceae
Fabaceae
Rubiaceae
Apocynaceae
Rhizophoraceae
Ulmaceae
Oleaceae
Vitaceae
Rutaceae
Rubiaceae
Combretaceae
Asteraceae
Boraginaceae
Asteraceae
Euphorbiaceae
Poaceae
Vitaceae
Fabaceae
Fabaceae
Asteraceae
Sinopteridaceae
Dracaenaceae
Dracaenaceae
182
Growth
form
Freq %Freq R.F
75
0.011
S
3
75
0.011
H
3
100
0.014
H
4
50
0.007
H
2
T
4
100
0.014
T
2
50
0.007
T
4
100
0.014
S
2
50
0.007
T
3
75
0.011
50
0.007
H
2
75
0.011
S
3
75
0.011
H
3
75
0.011
H
3
100
0.014
T
4
50
0.007
H
2
50
0.007
S
2
25
0.004
T
1
50
0.007
T
2
S
3
75
0.011
T
1
25
0.004
100
0.014
T
4
S
2
50
0.007
L
2
50
0.007
50
0.007
S
2
50
0.007
S
2
50
0.007
L
2
50
0.007
H
2
50
0.007
T
2
50
0.007
H
2
100
0.014
T
4
50
0.007
H
2
50
0.007
H
2
50
0.007
S
2
100
0.014
H
4
75
0.011
H
3
H
2
50
0.007
S
2
50
0.007
T
2
50
0.007
S.N
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
Species name (in DNF)
Droguetia iners
Ehretia cymosa
Ekebergia capensis
Embelia schimperi
Eremomastax speciosa
Erythrococca trichogyne
Ficus sur
Flacourtia indica
Galiniera saxifraga
Geranium aculeolatum
Gouania longispicata
Grewia ferruginea
Hippocratea goetzei
Hypoestes forskaullii
Impatiens aethiopica
Isoglossa somalensis
Jasminum abyssinicum
Jasminum repandum
Laggera crispata
Landolphia buchananii
Loxogramme abyssinica
Macaranga capensis
Maesa lanceolata
Maytenus arbutifolia
Maytenus gracilipes
Maytenus undata
Microsorium scolopendria
Mikaniopsis clematoides
Millettia ferruginea
Myrsine africana
Nuxia congesta
Ocimum urticifolium
Olea welwitschii
Oplismenus compositus
Oplismenus hirtellus
Oxyanthus speciosus
Passiflora edulis
Paullinia pinnata
Peperomia abyssinica
Family
Urticaceae
Boraginaceae
Meliaceae
Myrsinaceae
Acanthaceae
Euphorbiaceae
Moraceae
Flacourtiaceae
Rubiaceae
Geraniaceae
Rhamnaceae
Proteaceae
Celastraceae
Acanthaceae
Balsaminaceae
Acanthaceae
Oleaceae
Oleaceae
Asteraceae
Apocynaceae
Polypodiaceae
Euphorbiaceae
Myrsinaceae
Celastraceae
Celastraceae
Celastraceae
Polypodiaceae
Asteraceae
Fabaceae
Myrsinaceae
Loganiaceae
Lamiaceae
Oleaceae
Poaceae
Poaceae
Rubiaceae
Passifolraceae
Sapindaceae
Piperaceae
183
Growth
form
Freq %Freq R.F
75
0.011
H
3
T
2
50
0.007
50
0.007
T
2
50
0.007
S
2
50
0.007
H
2
50
0.007
S
2
75
0.011
T
3
50
0.007
T
2
75
0.011
T
3
50
0.007
H
2
75
0.011
L
3
50
0.007
S
2
50
0.007
L
2
50
0.007
H
2
H
2
50
0.007
S
2
50
0.007
L
4
100
0.014
L
3
75
0.011
50
0.007
H
2
100
0.014
L
4
75
0.011
H
3
50
0.007
T
2
50
0.007
T
2
75
0.011
T
3
50
0.007
S
2
50
0.007
S
2
50
0.007
H
2
50
0.007
H
2
50
0.007
T
2
S
2
50
0.007
25
0.004
T
1
S
2
50
0.007
T
2
50
0.007
H
3
75
0.011
50
0.007
H
2
50
0.007
S
2
50
0.007
L
2
50
0.007
L
2
100
0.014
H
4
S.N
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
Species name (in DNF)
Peperomia tetraphyla
Pergularia daemia
Periploca linearifolia
Phoenix reclinata
Phyllanthus mooneyi
Phytolacca dodecandra
Piper capense
Podocarpus falcatus
Polyscias fulva
Pouzolzia mixta
Prunus africana
Psychotria orophila
Rhamnus prinoides
Rothmannia urcelliformis
Rubus steudneri
Rumex natalensis
Rytigynia neglecta
Sanicula elata
Schefflera abyssinica
Senna didymobotrya
Setaria megaphylla
Setaria verticillata
Solanecio gigas
Solanecio mannii
Solanum anguivi
Solanum giganteum
Stephania abyssinica
Syzygium guineense
Teclea nobilis
Tectaria gemmifera
Tragia cinerea
Trichilia dregeana
Urera hypselodendron
Vangueria apiculata
Vepris dainellii
Vernonia auriculifera
Vernonia biafrae
Family
Piperaceae
Asclepiadaceae
Asclepiadaceae
Arecaceae
Euphorbiaceae
Phytolacaeae
Pinaceae
Lamiaceae
Podocarpaceae
Araliaceae
Lamiaceae
Myrtaceae
Ranunculaceae
Capparidaceae
Rosaceae
Rosaceae
Polygonaceae
Rubiaceae
Lamiaceae
Oleaceae
Fabaceae
Poaceae
Malvaceae
Asteraceae
Asteraceae
Solanaceae
Caryophyllaceae
Bignoniaceae
Asteraceae
Rutaceae
Ranunculaceae
Euphorbiaceae
Tiliaceae
Urticaceae
Rubiaceae
Asteraceae
Asteraceae
184
Growth
form
Freq %Freq R.F
100
0.014
H
4
H
2
50
0.007
50
0.007
L
2
75
0.011
T
3
50
0.007
H
2
50
0.007
S
2
75
0.011
S
3
100
0.014
T
4
50
0.007
T
2
50
0.007
S
2
50
0.007
T
2
50
0.007
T
2
75
0.011
S
3
50
0.007
T
2
S
3
75
0.011
H
2
50
0.007
S
3
75
0.011
H
3
75
0.011
50
0.007
T
2
50
0.007
S
2
75
0.011
H
3
50
0.007
H
2
75
0.011
S
3
75
0.011
T
3
50
0.007
S
2
50
0.007
S
2
50
0.007
H
2
100
0.014
T
4
50
0.007
T
2
H
2
50
0.007
50
0.007
H
2
T
2
50
0.007
L
2
50
0.007
T
2
50
0.007
75
0.011
T
3
50
0.007
S
2
50
0.007
S
2
Appendix 3: Species list, family, growth form (GF) percent and relative frequencies
(%freq, R.F) in the woodland
S.N
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Species name (in woodland)
Acacia abyssinica
Acacia lahai
Acalypha racemosa
Acanthus pubescens
Achyranthes aspera
Aeschynomene schimperi
Ageratum conyzoides
Ajuga integrifolia
Albizia gummifera
Amphicarpa africana
Arthropteris monocarpa
Aspilia mossambicensis
Asplenium formosum
Berkheya spekeana
Bidens pilosa
Bidens prestinaria
Bridelia micrantha
Buchnera hispida
Caesalpinia decapetala
Calpurnia aurea
Cardiospermum halicacabum
Carissa spinarum
Cissus petiolata
Clausena anisata
Clematis cadatus
Clematis longicauda
Clematis simensis
Clerodendron myricoides
Coelorhachis afraurita
Coffea arabica
Combretum collinum
Combretum molle
Combretum paniculatum
Commelina diffusa
Cordia africana
Croton macrostachyus
Cuscuta campestris
Family
Fabaceae
Fabaceae
Euphorbiaceae
Acanthaceae
Amaranthaceae
Fabaceae
Asteraceae
Lamiaceae
Fabaceae
Fabaceae
Oleandraceae
Asteraceae
Aspleniaceae
Asteraceae
Asteraceae
Asteraceae
Euphorbiaceae
Scrophulariaceae
Fabaceae
Fabaceae
Sapindaceae
Apocynaceae
Vitaceae
Rutaceae
Ranunculaceae
Ranunculaceae
Ranunculaceae
Lamiaceae
Poaceae
Rubiaceae
Combretaceae
Combretaceae
Combretaceae
Commelinaceae
Boraginaceae
Euphorbiaceae
Cuscutaceae
185
GF Freq %Freq R.F
100 0.35
T
4
50 0.17
T
2
50 0.17
H
2
S
4
100 0.35
H
3
75 0.26
50 0.17
H
2
H
2
50 0.17
H
2
50 0.17
50 0.17
T
2
50 0.17
H
2
50 0.17
H
2
50 0.17
S
2
50 0.17
H
2
50 0.17
H
2
100 0.35
H
4
50 0.17
H
2
50 0.17
T
2
50 0.17
H
2
25 0.09
S
1
50 0.17
T
2
H
2
50 0.17
S
2
50 0.17
50 0.17
L
2
50 0.17
S
2
50 0.17
L
2
50 0.17
L
2
50 0.17
L
2
50 0.17
S
2
50 0.17
H
2
25 0.09
S
1
50 0.17
T
2
50 0.17
T
2
50 0.17
L
2
50 0.17
H
2
T
3
75 0.26
T
3
75 0.26
H
2
50 0.17
S.N
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
Species name (in woodland)
Cyathula uncinulatA
Cycnium herzfeldianum
Cynodon aethiopicus
Cyperus welwitschii
Cyphostemma cyphopetalum
Cyprus triceps
Desmodium repandum
Desmodium salisifolium
Dichondra repens
Dicliptera laxata
Dioscorea bulbifera
Entada abyssinica
Erythrina brucei
Euphorbia cyparissioides
Euphorbia tirucalli
Ficus mucuso
Ficus sp.
Ficus sur
Ficus sycamoras
Ficus thonningii
Ficus vasta
Flacourtia indica
Galinsoga parviflora
Gardenia volkensii
Girardinia diversifolia
Gnidia glauca
Grewia ferruginea
Guizotia schimperi
Helinus mystacinus
Helychrysum forskaulii
Hibiscus berberidifolius
Hibiscus dongolensis
Hippocratea goetzei
Hyparrhenia rufa
Hypericum peplidifolium
Hypericum revolutum
Hypolepis glandulifera
Indigofera spicata
Justicia ladanoides
Keetia guiinzii
Family
Amaranthaceae
Scrophulariaceae
Poaceae
Cyperaceae
Vitaceae
Cyperaceae
Fabaceae
Fabaceae
Convolvulaceae
Acanthaceae
Dioscoreaceae
Fabaceae
Fabaceae
Euphorbiaceae
Euphorbiaceae
Moraceae
Moraceae
Moraceae
Moraceae
Moraceae
Moraceae
Flacourtiaceae
Asteraceae
Rubiaceae
Urticaceae
Thymelaeaceae
Proteaceae
Asteraceae
Rhamnaceae
Asteraceae
Malvaceae
Malvaceae
Celastraceae
Poaceae
Hypericaceae
Hypericaceae
Dennstaedtiaceae
Fabaceae
Acanthaceae
Rubiaceae
186
GF Freq %Freq R.F
H
2
50 0.17
50 0.17
H
2
50 0.17
H
2
50 0.17
H
2
75 0.26
H
3
50 0.17
H
2
50 0.17
H
2
50 0.17
H
2
75 0.26
H
3
50 0.17
H
2
50 0.17
H
2
100 0.35
T
4
25 0.09
T
1
H
2
50 0.17
T
2
50 0.17
T
2
50 0.17
25 0.09
T
1
50 0.17
T
2
50 0.17
T
2
50 0.17
T
2
50 0.17
T
2
50 0.17
T
2
50 0.17
H
2
75 0.26
S
3
50 0.17
H
2
50 0.17
S
2
50 0.17
S
2
50 0.17
H
2
L
2
50 0.17
H
1
25 0.09
50 0.17
S
2
S
2
50 0.17
50 0.17
L
2
75 0.26
H
3
50 0.17
H
2
50 0.17
S
2
H
2
50 0.17
50 0.17
S
2
50 0.17
H
2
S
2
50 0.17
S.N
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
Species name (in woodland)
Laggera crispata
Lantana trifolium
Leucas martinicensis
Lippia adoensis
Maesa lanceolata
Mikaniopsis clematoides
Millettia ferruginea
Momordica foetida
Ocimum lamiifolium
Ocimum urticifolium
Oplismenus compositus
Otostegia tomentosa
Passiflora edulis
Pavonia urens
Pennisetum sphacelatum
Phoenix reclinata
Phyllanthus ovalifolius
Premna schimperi
Pseudarthria hookeri
Psidium guajava
Pychnostachys emini
Pycnostachys abyssinica
Pycreus nitida
Rhamnus prinoides
Rhoicissus tridentata
Rhus natalensis
Ricinus communis
Rothmannia urcelliformis
Rubus steudneri
Rumex natalensis
Sapium ellipticum
Satureja paradoxa
Senna didymobotrya
Senna occidentalis
Senna petersiana
Sesbania sesban
Sicyos polyacanthus
Sida schimperiana
Sida tenuicarpa
Sida ternata
Family
Asteraceae
Verbenaceae
Lamiaceae
Verbenaceae
Myrsinaceae
Asteraceae
Fabaceae
Cucurbitaceae
Lamiaceae
Lamiaceae
Poaceae
Lamiaceae
Passifolraceae
Malvaceae
Poaceae
Arecaceae
Euphorbiaceae
Urticaceae
Rosaceae
Fabaceae
Fabaceae
Lamiaceae
Lamiaceae
Ranunculaceae
Rhamnaceae
Vitaceae
Anacardiaceae
Capparidaceae
Rosaceae
Rosaceae
Apiaceae
Euphorbiaceae
Oleaceae
Fabaceae
Fabaceae
Malvaeae
Poaceae
Cucurbitaceae
Malvaceae
Malvaceae
187
GF Freq %Freq R.F
H
2
50 0.17
50 0.17
S
2
50 0.17
H
2
75 0.26
S
3
50 0.17
T
2
50 0.17
H
2
50 0.17
T
2
50 0.17
H
2
50 0.17
S
2
50 0.17
S
2
50 0.17
H
2
50 0.17
S
2
25 0.09
L
1
S
3
75 0.26
H
2
50 0.17
T
2
50 0.17
50 0.17
S
2
50 0.17
S
2
50 0.17
H
2
50 0.17
T
2
50 0.17
H
2
50 0.17
H
2
50 0.17
H
2
50 0.17
S
2
50 0.17
L
2
50 0.17
S
2
50 0.17
H
2
25 0.09
T
1
S
2
50 0.17
H
2
50 0.17
50 0.17
T
2
H
2
50 0.17
50 0.17
S
2
50 0.17
H
2
75 0.26
T
3
25 0.09
T
1
25 0.09
H
1
50 0.17
S
2
50 0.17
S
2
H
2
50 0.17
S.N
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
Species name (in woodland)
Solanum capsicoides
Solanum giganteum
Solanum incanum
Stachys albigena
Stereospermum kunthianum
Syzygium guineense
Terminalia schimperiana
Tragia cinerea
Triumfetta pilosa
Vangueria apiculata
Vernonia adoensis
Vernonia amygdalina
Vernonia auriculifera
Vernonia hochstetteri
Vernonia ischnophylla
Vernonia karaguensis
Vernonia theophrastifolia
Vernonia thomsoniana
Veronica abyssinica
Family
Solanaceae
Solanaceae
Solanaceae
Poaceae
Menispermaceae
Bignoniaceae
Tectariaceae
Ranunculaceae
Moraceae
Urticaceae
Rutaceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
188
GF Freq %Freq R.F
S
2
50 0.17
50 0.17
S
2
75 0.26
S
3
50 0.17
H
2
75 0.26
T
3
50 0.17
T
2
100 0.35
T
4
50 0.17
H
2
50 0.17
H
2
50 0.17
T
2
50 0.17
S
2
50 0.17
T
2
100 0.35
S
4
S
2
50 0.17
S
3
75 0.26
S
2
50 0.17
50 0.17
S
2
50 0.17
S
2
50 0.17
H
2
Appendix 4: Species list, family, growth form (GF) percent and relative frequencies
(%freq, R.F) in the cropland
S.N
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Species name (in cropland)
Acacia abyssinica
Acanthus pubescens
Achyranthes aspera
Ageratum conyzoides
Ajuga integrifolia
Albizia gummifera
Alchemila pedata
Alectra sessiliflora
Amaranthus hybridus
Amaranthus sparganiocephalus
Arthraxon micans
Bersama abyssinica
Bidens pilosa
Bidens prestinaria
Brassica carinata
Brucea antidysenterica
Calpurnia aurea
Caylusea abyssinica
Celtis africana
Chenopodium ambrosioides
Cirsium dender
Cissampelos mucronata
Clematis cadatus
Clematis simensis
Coelorhachis afraurita
Combretum collinum
Commelina diffusa
Commelina imberbis
Conyza bonariensis
Cordia africana
Crassocephalum macropappum
Croton macrostachyus
Cuscuta campestris
Cyathula uncinulatA
Cycnium herzfeldianum
Cynodon aethiopicus
Dalbergia lactea
Family
Fabaceae
Acanthaceae
Amaranthaceae
Asteraceae
Lamiaceae
Fabaceae
Rosaceae
Scrophulariaceae
Amaranthaceae
Amaranthaceae
Poaceae
Melianthaceae
Asteraceae
Asteraceae
Brassicaceae
Simaroubaceae
Fabaceae
Resedaceae
Ulmaceae
Chenopodiaceae
Asteraceae
Menispermaceae
Ranunculaceae
Ranunculaceae
Poaceae
Combretaceae
Commelinaceae
Commelinaceae
Asteraceae
Boraginaceae
Asteraceae
Euphorbiaceae
Cuscutaceae
Amaranthaceae
Scrophulariaceae
Poaceae
Fabaceae
189
GF
T
S
H
H
H
T
H
H
H
H
H
T
H
H
H
S
T
H
T
H
H
H
L
L
H
T
H
H
H
T
H
T
H
H
H
H
S
Freq %Freq R.F
71 0.022
5
29 0.009
2
43 0.013
3
7
100 0.030
2
29 0.009
57 0.017
4
3
43 0.013
2
29 0.009
29 0.009
2
29 0.009
2
29 0.009
2
43 0.013
3
43 0.013
3
57 0.017
4
43 0.013
3
43 0.013
3
43 0.013
3
43 0.013
3
29 0.009
2
43 0.013
3
2
29 0.009
2
29 0.009
29 0.009
2
29 0.009
2
29 0.009
2
29 0.009
2
43 0.013
3
29 0.009
2
29 0.009
2
71 0.022
5
29 0.009
2
29 0.009
2
43 0.013
3
71 0.022
5
2
29 0.009
5
71 0.022
2
29 0.009
S.N
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
Species name (in cropland)
Datura stramonium
Dichondra repens
Dioscorea bulbifera
Echium plantagineum
Ehretia cymosa
Erythrina brucei
Euphorbia tirucalli
Ficus mucuso
Ficus sp.
Ficus sur
Galinsoga parviflora
Glycine wightii
Guizotia schimperi
Hibiscus dongolensis
Hygrophila asteracanthoide
Hyparrhenia rufa
Indigofera spicata
Justicia ladanoides
Laggera crispata
Leucas martinicensis
Lippia adoensis
Maesa lanceolata
Maytenus gracilipes
Millettia ferruginea
Momordica foetida
Nicandra physaloides
Ocimum urticifolium
Pavonia glechomifolia
Pavonia urens
Pennisetum nubicum
Phyllanthus mooneyi
Physalis peruviana
Phytolacca dodecandra
Plantago lanceolata
Ritchiea albersii
Rumex natalensis
Sapium ellipticum
Satureja paradoxa
Schefflera abyssinica
Senna didymobotrya
Family
Solanaceae
Convolvulaceae
Dioscoreaceae
Boraginaceae
Boraginaceae
Fabaceae
Euphorbiaceae
Moraceae
Moraceae
Moraceae
Asteraceae
Fabaceae
Asteraceae
Malvaceae
Acanthaceae
Poaceae
Fabaceae
Acanthaceae
Asteraceae
Lamiaceae
Verbenaceae
Myrsinaceae
Celastraceae
Fabaceae
Cucurbitaceae
Solanaceae
Lamiaceae
Malvaceae
Malvaceae
Poaceae
Euphorbiaceae
Solanaceae
Phytolacaeae
Pittosporaceae
Euphorbiaceae
Rosaceae
Apiaceae
Euphorbiaceae
Lamiaceae
Oleaceae
190
GF
H
H
H
H
T
T
T
T
T
T
H
H
H
S
H
H
S
H
H
H
S
T
S
T
H
H
S
H
S
H
H
H
S
H
T
H
T
H
T
S
Freq %Freq R.F
3
43 0.013
29 0.009
2
29 0.009
2
29 0.009
2
29 0.009
2
14 0.004
1
29 0.009
2
29 0.009
2
14 0.004
1
29 0.009
2
43 0.013
3
29 0.009
2
71 0.022
5
2
29 0.009
2
29 0.009
2
29 0.009
29 0.009
2
29 0.009
2
29 0.009
2
29 0.009
2
29 0.009
2
29 0.009
2
43 0.013
3
29 0.009
2
57 0.017
4
29 0.009
2
29 0.009
2
29 0.009
2
2
29 0.009
2
29 0.009
29 0.009
2
2
29 0.009
57 0.017
4
57 0.017
4
14 0.004
1
43 0.013
3
29 0.009
2
29 0.009
2
29 0.009
2
3
43 0.013
S.N
78
79
80
81
82
83
84
85
86
87
88
89
90
91
Species name (in cropland)
Sida schimperiana
Sida tenuicarpa
Solanecio gigas
Solanum incanum
Soncus asper
Stereospermum kunthianum
Tagetes minuta
Terminalia schimperiana
Tragia cinerea
Triumfetta rhomboidea
Vernonia amygdalina
Vernonia auriculifera
Vernonia ischnophylla
Veronica abyssinica
Family
Cucurbitaceae
Malvaceae
Malvaceae
Solanaceae
Solanaceae
Menispermaceae
Myrtaceae
Tectariaceae
Ranunculaceae
Tiliaceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
191
GF
S
S
S
S
H
T
H
T
H
H
T
S
S
H
Freq %Freq R.F
2
29 0.009
43 0.013
3
29 0.009
2
43 0.013
3
43 0.013
3
29 0.009
2
29 0.009
2
29 0.009
2
29 0.009
2
29 0.009
2
29 0.009
2
57 0.017
4
29 0.009
2
2
29 0.009
Appendix 5: Species list, Growth form (GF), percent and relative frequencies (%freq,
R.F) of plant species in the SFC
S.N
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Species name (in SFC)
Celtis africana
Coffea arabica
Ehretia cymosa
Achyranthes aspera
Albizia gummifera
Cordia africana
Croton macrostachyus
Desmodium repandum
Vepris dainellii
Vernonia amygdalina
Vernonia auriculifera
Clausena anisata
Maesa lanceolata
Oplismenus compositus
Vangueria apiculata
Acacia abyssinica
Acanthus eminens
Ageratum conyzoides
Allophylus abyssinicus
Bidens pilosa
Cyathula uncinulatA
Erythrococca trichogyne
Girardinia diversifolia
Millettia ferruginea
Pentas lanceolata
Peperomia abyssinica
Rytigynia neglecta
Sapium ellipticum
Allophylus macrobotrys
Bersama abyssinica
Bidens prestinaria
Calpurnia aurea
Clematis hirsuta
Diospyros abyssinica
Dracaena steudneri
Ficus thonningii
Ficus vasta
Family
Ulmaceae
Rubiaceae
Boraginaceae
Amaranthaceae
Fabaceae
Boraginaceae
Euphorbiaceae
Fabaceae
Rubiaceae
Asteraceae
Asteraceae
Rutaceae
Myrsinaceae
Poaceae
Urticaceae
Fabaceae
Acanthaceae
Asteraceae
Sapindaceae
Asteraceae
Amaranthaceae
Euphorbiaceae
Urticaceae
Fabaceae
Rubiaceae
Piperaceae
Polygonaceae
Apiaceae
Sapindaceae
Melianthaceae
Asteraceae
Fabaceae
Ranunculaceae
Ebenaceae
Dracaenaceae
Moraceae
Moraceae
192
GF Freq %Freq R.F
100 0.017
T
7
100 0.017
S
7
100 0.017
T
7
H
6
86 0.015
T
6
86 0.015
86 0.015
T
6
T
6
86 0.015
H
6
86 0.015
86 0.015
T
6
86 0.015
T
6
86 0.015
S
6
71 0.012
S
5
71 0.012
T
5
71 0.012
H
5
71 0.012
T
5
57 0.010
T
4
57 0.010
S
4
57 0.010
H
4
57 0.010
T
4
57 0.010
H
4
H
4
57 0.010
S
4
57 0.010
57 0.010
H
4
57 0.010
T
4
57 0.010
S
4
57 0.010
H
4
57 0.010
S
4
57 0.010
T
4
43 0.007
S
3
43 0.007
T
3
43 0.007
H
3
43 0.007
T
3
43 0.007
L
3
43 0.007
T
3
T
3
43 0.007
T
3
43 0.007
T
3
43 0.007
S.N
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
Species name (in SFC)
Flacourtia indica
Hypoestes aristata
Impatiens aethiopica
Justicia schimperiana
Maytenus arbutifolia
Maytenus gracilipes
Ocimum lamiifolium
Pavonia urens
Pergularia daemia
Phyllanthus ovalifolius
Phytolacca dodecandra
Pittosporum viridiflorum
Prunus africana
Ritchiea albersii
Rubus steudneri
Sicyos polyacanthus
Sida tenuicarpa
Stephania abyssinica
Tragia cinerea
Acalypha racemosa
Achyrospermum schimperi
Adenostemma perottettii
Albizia schimperiana
Ampelocissus bombycina
Arthropteris monocarpa
Asparagus racemosus
Aspilia mossambicensis
Asplenium aethiopicum
Asplenium formosum
Bridelia micrantha
Caesalpinia decapetala
Celosia anthelminthica
Celosia trigyna
Ceropegia racemosa
Chenopodium ambrosioides
Chionanthus mildbraedii
Cissampelos mucronata
Commelina diffusa
Crassocephalum macropappum
Crotalaria emarginella
Family
Flacourtiaceae
Acanthaceae
Balsaminaceae
Acanthaceae
Celastraceae
Celastraceae
Lamiaceae
Malvaceae
Asclepiadaceae
Euphorbiaceae
Phytolacaeae
Piperaceae
Lamiaceae
Euphorbiaceae
Rosaceae
Poaceae
Malvaceae
Caryophyllaceae
Ranunculaceae
Euphorbiaceae
Lamiaceae
Asteraceae
Fabaceae
Vitaceae
Oleandraceae
Asparagaceae
Asteraceae
Aspleniaceae
Aspleniaceae
Euphorbiaceae
Fabaceae
Amaranthaceae
Amaranthaceae
Asclepiadaceae
Chenopodiaceae
Oleaceae
Menispermaceae
Commelinaceae
Asteraceae
Fabaceae
193
GF Freq %Freq R.F
T
3
43 0.007
43 0.007
H
3
43 0.007
H
3
43 0.007
S
3
43 0.007
T
3
43 0.007
S
3
43 0.007
S
3
43 0.007
S
3
43 0.007
H
3
43 0.007
S
3
43 0.007
S
3
43 0.007
T
3
43 0.007
T
3
T
3
43 0.007
S
3
43 0.007
H
3
43 0.007
43 0.007
S
3
43 0.007
H
3
43 0.007
H
3
29 0.005
H
2
29 0.005
H
2
29 0.005
H
2
29 0.005
T
2
29 0.005
H
2
29 0.005
H
2
29 0.005
S
2
29 0.005
S
2
29 0.005
H
2
H
2
29 0.005
T
2
29 0.005
29 0.005
S
2
H
2
29 0.005
29 0.005
H
2
29 0.005
H
2
29 0.005
H
2
29 0.005
S
2
29 0.005
H
2
29 0.005
H
2
29 0.005
H
2
H
2
29 0.005
S.N
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
Species name (in SFC)
Cyphostemma cyphopetalum
Dalbergia lactea
Desmodium dichotomum
Dicliptera laxata
Doryopteris concolor
Dracaena afromontana
Droguetia iners
Ekebergia capensis
Ensete ventricosum
Eremomastax speciosa
Euphorbia candelabrum
Ficus sp.
Ficus sur
Galiniera saxifraga
Galinsoga parviflora
Geranium aculeolatum
Grewia ferruginea
Guizotia schimperi
Hibiscus berberidifolius
Hippocratea goetzei
Indigofera spicata
Isoglossa somalensis
Justicia ladanoides
Laggera crispata
Leucas martinicensis
Loxogramme abyssinica
Microsorium scolopendria
Momordica foetida
Ocimum urticifolium
Pavonia glechomifolia
Pennisetum nubicum
Peperomia tetraphyla
Phoenix reclinata
Physalis peruviana
Podocarpus falcatus
Polyscias fulva
Premna schimperi
Pseudarthria hookeri
Psydrax schimperiana
Pteris pteridioides
Family
Vitaceae
Fabaceae
Fabaceae
Acanthaceae
Sinopteridaceae
Dracaenaceae
Urticaceae
Meliaceae
Musaceae
Acanthaceae
Euphorbiaceae
Moraceae
Moraceae
Rubiaceae
Asteraceae
Geraniaceae
Proteaceae
Asteraceae
Malvaceae
Celastraceae
Fabaceae
Acanthaceae
Acanthaceae
Asteraceae
Lamiaceae
Polypodiaceae
Polypodiaceae
Cucurbitaceae
Lamiaceae
Malvaceae
Poaceae
Piperaceae
Arecaceae
Solanaceae
Lamiaceae
Podocarpaceae
Urticaceae
Rosaceae
Rubiaceae
Rubiaceae
194
GF Freq %Freq R.F
H
2
29 0.005
29 0.005
S
2
29 0.005
H
2
29 0.005
H
2
29 0.005
H
2
29 0.005
S
2
29 0.005
H
2
29 0.005
T
2
29 0.005
H
2
29 0.005
H
2
29 0.005
T
2
29 0.005
T
2
29 0.005
T
2
T
2
29 0.005
H
2
29 0.005
H
2
29 0.005
29 0.005
S
2
29 0.005
H
2
29 0.005
S
2
29 0.005
L
2
29 0.005
S
2
29 0.005
S
2
29 0.005
H
2
29 0.005
H
2
29 0.005
H
2
29 0.005
H
2
29 0.005
H
2
29 0.005
H
2
S
2
29 0.005
H
2
29 0.005
29 0.005
H
2
H
2
29 0.005
29 0.005
T
2
29 0.005
H
2
29 0.005
T
2
29 0.005
T
2
29 0.005
S
2
29 0.005
H
2
29 0.005
T
2
H
2
29 0.005
S.N
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
Species name (in SFC)
Pterolobium stellatum
Pycnostachys abyssinica
Ranunculus multifidus
Ricinus communis
Rothmannia urcelliformis
Sanicula elata
Schefflera abyssinica
Senna occidentalis
Senna petersiana
Senra incana
Setaria verticillata
Sida schimperiana
Solanecio gigas
Solanum anguivi
Solanum incanum
Soncus asper
Stellaria mannii
Syzygium guineense
Teclea nobilis
Tectaria gemmifera
Thalictrum rhynchocarpum
Trichilia dregeana
Trilepisium madagascariense
Triumfetta pilosa
Triumfetta rhomboidea
Urera hypselodendron
Vernonia karaguensis
Apodytes dimidiata
Cassipourea malosana
Ficus mucuso
Gouania longispicata
Kosteletzkya begoniifolia
Olea welwitschii
Psidium guajava
Schrebera alata
Family
Pteridaceae
Lamiaceae
Pyperaceae
Anacardiaceae
Capparidaceae
Rubiaceae
Lamiaceae
Fabaceae
Fabaceae
Fabaceae
Poaceae
Cucurbitaceae
Malvaceae
Asteraceae
Solanaceae
Solanaceae
Lamiaceae
Bignoniaceae
Asteraceae
Rutaceae
Combretaceae
Euphorbiaceae
Meliaceae
Moraceae
Tiliaceae
Tiliaceae
Asteraceae
Icacinaceae
Rhizophoraceae
Moraceae
Rhamnaceae
Malvaceae
Oleaceae
Fabaceae
Araliaceae
195
GF Freq %Freq R.F
S
2
29 0.005
29 0.005
H
2
29 0.005
H
2
29 0.005
H
2
29 0.005
T
2
29 0.005
H
2
29 0.005
T
2
29 0.005
H
2
29 0.005
T
2
29 0.005
H
2
29 0.005
H
2
29 0.005
S
2
29 0.005
S
2
S
2
29 0.005
S
2
29 0.005
H
2
29 0.005
29 0.005
H
2
29 0.005
T
2
29 0.005
T
2
29 0.005
H
2
29 0.005
H
2
29 0.005
T
2
29 0.005
T
2
29 0.005
H
2
29 0.005
H
2
29 0.005
L
2
29 0.005
S
2
14 0.002
T
1
T
1
14 0.002
T
1
14 0.002
14 0.002
L
1
H
1
14 0.002
14 0.002
T
1
14 0.002
T
1
14 0.002
T
1
Appendix 6: Species list, Growth form (GF) percent and relative frequencies (%freq,
R.F) of plant species in pastureland
S.N
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Species name (in pasture land)
Acacia abyssinica
Acanthus pubescens
Achyranthes aspera
Aeschynomene schimperi
Ageratum conyzoides
Albizia gummifera
Allophylus macrobotrys
Amaranthus hybridus
Amaranthus sparganiocephalus
Apodytes dimidiata
Asplenium formosum
Bauhinia tomentosa
Becium verticillifolium
Berkheya spekeana
Bersama abyssinica
Bidens pilosa
Bridelia micrantha
Buchnera hispida
Cardiospermum halicacabum
Cirsium dender
Clausena anisata
Clematis hirsuta
Clematis longicauda
Clematis simensis
Combretum collinum
Combretum paniculatum
Commelina imberbis
Conyza bonariensis
Crossopteryx febrifuga
Croton macrostachyus
Cyathula uncinulatA
Cynodon aethiopicus
Cyperus digitatus
Cyprus triceps
Desmodium dichotomum
Desmodium repandum
Dichondra repens
Family
Fabaceae
Acanthaceae
Amaranthaceae
Fabaceae
Asteraceae
Fabaceae
Sapindaceae
Amaranthaceae
Amaranthaceae
Icacinaceae
Aspleniaceae
Fabaceae
Lamiaceae
Asteraceae
Melianthaceae
Asteraceae
Euphorbiaceae
Scrophulariaceae
Sapindaceae
Asteraceae
Rutaceae
Ranunculaceae
Ranunculaceae
Ranunculaceae
Combretaceae
Combretaceae
Commelinaceae
Asteraceae
Rubiaceae
Euphorbiaceae
Amaranthaceae
Poaceae
Cyperaceae
Cyperaceae
Fabaceae
Fabaceae
Convolvulaceae
196
GF Freq %Freq R.F
T
2
40 0.163
S
2
40 0.163
H
3
60 0.244
H
2
40 0.163
H
5
100 0.407
T
2
40 0.163
S
2
40 0.163
H
2
40 0.163
H
2
40 0.163
T
1
20 0.081
H
2
40 0.163
T
2
40 0.163
S
2
40 0.163
H
2
40 0.163
T
2
40 0.163
H
2
40 0.163
T
3
60 0.244
H
2
40 0.163
H
2
40 0.163
H
2
40 0.163
S
2
40 0.163
L
2
40 0.163
L
2
40 0.163
L
2
40 0.163
T
3
60 0.244
L
2
40 0.163
H
2
40 0.163
H
2
40 0.163
T
2
40 0.163
T
2
40 0.163
H
2
40 0.163
H
2
40 0.163
H
3
60 0.244
H
3
60 0.244
H
2
40 0.163
H
2
40 0.163
H
4
80 0.325
S.N
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
Species name (in pasture land)
Digitaria abyssinica
Digitaria ternata
Doryopteris concolor
Ehretia cymosa
Entada abyssinica
Euphorbia cyparissioides
Euphorbia schimperiana
Ficus thonningii
Ficus vasta
Flacourtia indica
Galinsoga parviflora
Gardenia volkensii
Glycine wightii
Gnidia glauca
Grewia ferruginea
Guizotia schimperi
Helinus mystacinus
Helychrysum forskaulii
Hygrophila asteracanthoide
Hyparrhenia rufa
Hypericum peplidifolium
Justicia ladanoides
Keetia guiinzii
Keetia zanzibarica
Laggera alata
Laggera crispata
Lantana trifolium
Lippia adoensis
Maesa lanceolata
Maytenus arbutifolia
Maytenus senegalensis
Micractis bojeri
Nephrolepis undulata
Ocimum urticifolium
Oplismenus compositus
Oplismenus hirtellus
Otostegia tomentosa
Paullinia pinnata
Pavonia urens
Pennisetum sphacelatum
Family
Poaceae
Poaceae
Sinopteridaceae
Boraginaceae
Fabaceae
Euphorbiaceae
Euphorbiaceae
Moraceae
Moraceae
Flacourtiaceae
Asteraceae
Rubiaceae
Fabaceae
Thymelaeaceae
Proteaceae
Asteraceae
Rhamnaceae
Asteraceae
Acanthaceae
Poaceae
Hypericaceae
Acanthaceae
Rubiaceae
Rubiaceae
Asteraceae
Asteraceae
Verbenaceae
Verbenaceae
Myrsinaceae
Celastraceae
Celastraceae
Asteraceae
Nephrolepidaceae
Lamiaceae
Poaceae
Poaceae
Lamiaceae
Sapindaceae
Malvaceae
Poaceae
197
GF Freq %Freq R.F
H
2
40 0.163
H
2
40 0.163
H
2
40 0.163
T
2
40 0.163
T
2
40 0.163
H
2
40 0.163
H
2
40 0.163
T
2
40 0.163
T
3
60 0.244
T
3
60 0.244
H
2
40 0.163
S
3
60 0.244
H
2
40 0.163
S
2
40 0.163
S
2
40 0.163
H
2
40 0.163
L
2
40 0.163
H
2
40 0.163
H
3
60 0.244
H
2
40 0.163
H
2
40 0.163
H
2
40 0.163
S
1
20 0.081
S
2
40 0.163
H
2
40 0.163
H
2
40 0.163
S
2
40 0.163
S
2
40 0.163
T
3
60 0.244
T
2
40 0.163
T
2
40 0.163
H
2
40 0.163
H
2
40 0.163
S
2
40 0.163
H
2
40 0.163
H
2
40 0.163
S
2
40 0.163
L
2
40 0.163
S
3
60 0.244
H
2
40 0.163
S.N
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
Species name (in pasture land)
Pentas lanceolata
Persicaria setosula
Phoenix reclinata
Phyllanthus mooneyi
Phyllanthus ovalifolius
Plectranthus punctatus
Premna schimperi
Prunus africana
Pterolobium stellatum
Pychnostachys emini
Pycreus nitida
Ranunculus multifidus
Rhamnus prinoides
Rhoicissus tridentata
Rhus natalensis
Rubus apetalus
Rubus steudneri
Rytigynia neglecta
Sapium ellipticum
Satureja paradoxa
Senna petersiana
Sida schimperiana
Sida ternata
Solanum anguivi
Solanum dasyphyllum
Solanum incanum
Sporobolus africanus
Syzygium guineense
Vangueria apiculata
Vernonia adoensis
Vernonia auriculifera
Vernonia hochstetteri
Vernonia ischnophylla
Vernonia ituriensis
Vernonia theophrastifolia
Xanthium strumanium
Family
Rubiaceae
Polygonaceae
Arecaceae
Euphorbiaceae
Euphorbiaceae
Plantaginaceae
Urticaceae
Lamiaceae
Pteridaceae
Fabaceae
Lamiaceae
Pyperaceae
Ranunculaceae
Rhamnaceae
Vitaceae
Rubiaceae
Rosaceae
Polygonaceae
Apiaceae
Euphorbiaceae
Fabaceae
Cucurbitaceae
Malvaceae
Asteraceae
Solanaceae
Solanaceae
Asteraceae
Bignoniaceae
Urticaceae
Rutaceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
Scrophulariaceae
198
GF Freq %Freq R.F
S
2
40 0.163
H
2
40 0.163
T
2
40 0.163
H
2
40 0.163
S
2
40 0.163
H
2
40 0.163
S
2
40 0.163
T
1
20 0.081
S
2
40 0.163
H
2
40 0.163
H
2
40 0.163
H
2
40 0.163
S
2
40 0.163
L
2
40 0.163
S
3
60 0.244
S
2
40 0.163
S
2
40 0.163
S
2
40 0.163
T
3
60 0.244
H
4
80 0.325
T
2
40 0.163
S
4
80 0.325
H
2
40 0.163
S
2
40 0.163
H
2
40 0.163
S
2
40 0.163
H
2
40 0.163
T
2
40 0.163
T
2
40 0.163
S
2
40 0.163
S
3
60 0.244
S
2
40 0.163
S
2
40 0.163
S
2
40 0.163
S
2
40 0.163
H
2
40 0.163
Appendix 7: Species list, family, growth form (GF) percent and relative frequencies
(%freq, R.F) of plant species in plantation forests of Jimma Highlands
S.N
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Species name (in plantation
forest)
Acacia abyssinica
Achyranthes aspera
Ageratum conyzoides
Albizia gummifera
Apodytes dimidiata
Bersama abyssinica
Bidens pilosa
Brucea antidysenterica
Calpurnia aurea
Celtis africana
Cirsium dender
Clausena anisata
Clutia lanceolata Forssk.
Commelina diffusa
Cordia africana
Croton macrostachyus
Cupressus lucitanica
Cyathula uncinulatA
Dalbergia lactea
Dichondra repens
Ehretia cymosa
Ekebergia capensis
Erythrococca trichogyne
Eucalyptus camaldulensis
Euphorbia schimperiana
Ficus sur
Ficus thonningii
Flacourtia indica
Galinsoga parviflora
Girardinia diversifolia
Gouania longispicata
Grevillea robusta
Guizotia schimperi
Hippocratea goetzei
Hypoestes forskaullii
Jasminum abyssinicum
Family
Fabaceae
Amaranthaceae
Asteraceae
Fabaceae
Icacinaceae
Melianthaceae
Asteraceae
Simaroubaceae
Fabaceae
Ulmaceae
Asteraceae
Rutaceae
Euphorbiaceae
Commelinaceae
Boraginaceae
Euphorbiaceae
Cupressaceae
Amaranthaceae
Fabaceae
Convolvulaceae
Boraginaceae
Meliaceae
Euphorbiaceae
Myrtaceae
Euphorbiaceae
Moraceae
Moraceae
Flacourtiaceae
Asteraceae
Urticaceae
Rhamnaceae
Tiliaceae
Asteraceae
Celastraceae
Acanthaceae
Oleaceae
199
GF
T
H
H
T
T
T
H
S
T
T
H
S
S
H
T
T
T
H
S
H
T
T
S
T
H
T
T
T
H
H
L
T
H
L
H
L
Freq %Freq R.F
25 0.006
1
75 0.018
3
75 0.018
3
75 0.018
3
25 0.006
1
75 0.018
3
50 0.012
2
50 0.012
2
4
100 0.024
50 0.012
2
50 0.012
2
75 0.018
3
50 0.012
2
50 0.012
2
50 0.012
2
75 0.018
3
25 0.006
1
50 0.012
2
50 0.012
2
50 0.012
2
50 0.012
2
2
50 0.012
2
50 0.012
25 0.006
1
50 0.012
2
50 0.012
2
50 0.012
2
50 0.012
2
50 0.012
2
50 0.012
2
50 0.012
2
25 0.006
1
50 0.012
2
50 0.012
2
75 0.018
3
2
50 0.012
S.N
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
Species name (in plantation
forest)
Justicia ladanoides
Kalanchoe petitiana
Laggera alata
Laggera crispata
Lantana trifolium
Leucas martinicensis
Macaranga capensis
Maesa lanceolata
Maytenus gracilipes
Myrsine africana
Ocimum lamiifolium
Ocimum urticifolium
Olea welwitschii
Oplismenus compositus
Pentas lanceolata
Peperomia abyssinica
Peperomia tetraphyla
Pergularia daemia
Phoenix reclinata
Phyllanthus mooneyi
Phyllanthus ovalifolius
Pinus patula
Pittosporum viridiflorum
Plantago lanceolata
Premna schimperi
Pterolobium stellatum
Rothmannia urcelliformis
Rubus steudneri
Rytigynia neglecta
Sapium ellipticum
Satureja paradoxa
Senna didymobotrya
Sida schimperiana
Solanecio mannii
Solanum incanum
Syzygium guineense
Tagetes minuta
Tectaria gemmifera
Thalictrum rhynchocarpum
Family
Acanthaceae
Crassulaceae
Asteraceae
Asteraceae
Verbenaceae
Lamiaceae
Euphorbiaceae
Myrsinaceae
Celastraceae
Myrsinaceae
Lamiaceae
Lamiaceae
Oleaceae
Poaceae
Rubiaceae
Piperaceae
Piperaceae
Asclepiadaceae
Arecaceae
Euphorbiaceae
Euphorbiaceae
Asteraceae
Piperaceae
Pittosporaceae
Urticaceae
Pteridaceae
Capparidaceae
Rosaceae
Polygonaceae
Apiaceae
Euphorbiaceae
Oleaceae
Cucurbitaceae
Asteraceae
Solanaceae
Bignoniaceae
Myrtaceae
Rutaceae
Combretaceae
200
GF
H
H
H
H
S
H
T
T
S
S
S
S
T
H
S
H
H
H
T
H
S
T
T
H
S
S
T
S
S
T
H
S
S
T
S
T
H
H
H
Freq %Freq R.F
75 0.018
3
3
75 0.018
100 0.024
4
50 0.012
2
25 0.006
1
50 0.012
2
50 0.012
2
100 0.024
4
50 0.012
2
50 0.012
2
50 0.012
2
50 0.012
2
25 0.006
1
75 0.018
3
3
75 0.018
3
75 0.018
2
50 0.012
2
50 0.012
50 0.012
2
75 0.018
3
50 0.012
2
25 0.006
1
25 0.006
1
50 0.012
2
50 0.012
2
50 0.012
2
50 0.012
2
75 0.018
3
75 0.018
3
2
50 0.012
75 0.018
3
2
50 0.012
2
50 0.012
1
25 0.006
50 0.012
2
25 0.006
1
50 0.012
2
50 0.012
2
50 0.012
2
S.N
76
77
78
79
Species name (in plantation
forest)
Vangueria apiculata
Vepris dainellii
Vernonia auriculifera
Vernonia ituriensis
Family
Urticaceae
Rubiaceae
Asteraceae
Asteraceae
201
GF
T
T
S
S
Freq %Freq R.F
50 0.012
2
1
25 0.006
100 0.024
4
50 0.012
2
Appendix 8: List of plant species in all study plots along the transect in the Jimma
Highlands
S.N
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Species name (All
Plots)
Acacia abyssinica
Acacia lahai
Acalypha racemosa
Acanthus pubescens
Acanthus eminens
Achyranthes aspera
Achyrospermum
schimperi
Adenostemma perottettii
Adiantum poiretii
Aeschynomene
schimperi
Ageratum conyzoides
Ajuga integrifolia
Albizia gummifera
Albizia schimperiana
Alchemila pedata
Alectra sessiliflora
Allophylus abyssinicus
Allophylus macrobotrys
Amaranthus hybridus
Amaranthus
sparganiocephalus
Ampelocissus
bombycina
Amphicarpa africana
Apodytes dimidiata
Arthraxon micans
Arthropteris monocarpa
Asparagus racemosus
Aspilia mossambicensis
Asplenium aethiopicum
Asplenium formosum
Bauhinia tomentosa
Becium verticillifolium
Berkheya spekeana
Bersama abyssinica
Bidens pilosa
Fabaceae
Fabaceae
Euphorbiaceae
Acanthaceae
Acanthaceae
Amaranthaceae
Growth
form
T
T
H
S
S
H
DD1
DD2
DD3
DD4
DD5
DD6
Lamiaceae
H
DD7
2 6. 45
0. 004
Asteraceae
Adiantaceae
H
H
DD8
DD9
2 6. 45
4 12. 90
0. 004
0. 008
Fabaceae
H
DD10
4 12. 90
0. 008
Asteraceae
Lamiaceae
Fabaceae
Fabaceae
Rosaceae
Scrophulariaceae
Sapindaceae
Sapindaceae
Amaranthaceae
H
H
T
T
H
H
T
S
H
DD11
DD12
DD13
DD14
DD15
DD16
DD17
DD18
DD19
74. 19
12. 90
67. 74
12. 90
9. 68
6. 45
25. 81
22. 58
12. 90
0. 046
0. 008
0. 042
0. 008
0. 006
0. 004
0. 016
0. 014
0. 008
Amaranthaceae
H
DD20
4 12. 90
0. 008
Vitaceae
H
DD21
2 6. 45
0. 004
Fabaceae
Icacinaceae
Poaceae
Oleandraceae
Asparagaceae
Asteraceae
Aspleniaceae
Aspleniaceae
Fabaceae
Lamiaceae
Asteraceae
Melianthaceae
Asteraceae
H
T
H
H
S
S
H
H
T
S
H
T
H
DD22
DD23
DD24
DD25
DD26
DD27
DD28
DD29
DD30
DD31
DD32
DD33
DD34
Family
202
Col.No. Freq
16
2
4
8
7
21
23
4
21
4
3
2
8
7
4
2
6
4
4
5
4
5
9
2
2
4
15
17
%
Freq
51. 61
6. 45
12. 90
25. 81
22. 58
67. 74
6. 45
19. 35
12. 90
12. 90
16. 13
12. 90
16. 13
29. 03
6. 45
6. 45
12. 90
48. 39
54. 84
R. F
0. 032
0. 004
0. 008
0. 016
0. 014
0. 042
0. 004
0. 012
0. 008
0. 008
0. 010
0. 008
0. 010
0. 018
0. 004
0. 004
0. 008
0. 030
0. 034
S.N
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
Species name (All
Plots)
Bidens prestinaria
Brassica carinata
Bridelia micrantha
Brucea antidysenterica
Buchnera hispida
Caesalpinia decapetala
Calpurnia aurea
Canthium oligocarpum
Cardiospermum
halicacabum
Carissa spinarum
Cassipourea malosana
Caylusea abyssinica
Celosia anthelminthica
Celosia trigyna
Celtis africana
Ceropegia racemosa
Chenopodium
ambrosioides
Chionanthus mildbraedii
Cirsium dender
Cissampelos mucronata
Cissus petiolata
Clausena anisata
Clematis cadatus
Clematis hirsuta
Clematis longicauda
Clematis simensis
Clerodendron
myricoides
Clutia lanceolata
Coelorhachis afraurita
Coffea arabica
Combretum collinum
Combretum molle
Combretum paniculatum
Commelina diffusa
Commelina imberbis
Conyza bonariensis
Cordia africana
Asteraceae
Brassicaceae
Euphorbiaceae
Simaroubaceae
Scrophulariaceae
Fabaceae
Fabaceae
Rubiaceae
Growth
form
H
H
T
S
H
S
T
T
DD35
DD36
DD37
DD38
DD39
DD40
DD41
DD42
Sapindaceae
H
DD43
Apocynaceae
Rhizophoraceae
Resedaceae
Amaranthaceae
Amaranthaceae
Ulmaceae
Asclepiadaceae
S
T
H
H
H
T
H
DD44
DD45
DD46
DD47
DD48
DD49
DD50
Chenopodiaceae
H
DD51
Oleaceae
Asteraceae
Menispermaceae
Vitaceae
Rutaceae
Ranunculaceae
Ranunculaceae
Ranunculaceae
Ranunculaceae
S
H
H
L
S
L
L
L
L
DD52
DD53
DD54
DD55
DD56
DD57
DD58
DD59
DD60
Lamiaceae
S
DD61
Euphorbiaceae
Poaceae
Rubiaceae
Combretaceae
Combretaceae
Combretaceae
Commelinaceae
Commelinaceae
Asteraceae
Boraginaceae
S
H
S
T
T
l
H
H
H
T
DD62
DD63
DD64
DD65
DD66
DD67
DD68
DD69
DD70
DD71
Family
203
Col.No. Freq
9
3
7
7
4
3
13
2
%
Freq
29. 03
9. 68
22. 58
22. 58
12. 90
9. 68
41. 94
6. 45
R. F
0. 018
0. 006
0. 014
0. 014
0. 008
0. 006
0. 026
0. 004
4 12. 90
0. 008
16. 13
6. 45
9. 68
6. 45
6. 45
48. 39
6. 45
0. 010
0. 004
0. 006
0. 004
0. 004
0. 030
0. 004
5 16. 13
0. 010
12. 90
19. 35
12. 90
12. 90
45. 16
12. 90
16. 13
12. 90
19. 35
0. 008
0. 012
0. 008
0. 008
0. 028
0. 008
0. 010
0. 008
0. 012
5
2
3
2
2
15
2
4
6
4
4
14
4
5
4
6
2 6. 45
2
4
10
7
2
6
9
4
6
18
6. 45
12. 90
32. 26
22. 58
6. 45
19. 35
29. 03
12. 90
19. 35
58. 06
0. 004
0. 004
0. 008
0. 020
0. 014
0. 004
0. 012
0. 018
0. 008
0. 012
0. 036
S.N
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
Family
%
Growth
Col.No. Freq
Freq
form
Asteraceae
H
DD72
Asteraceae
Rubiaceae
Fabaceae
Euphorbiaceae
Cupressaceae
Cuscutaceae
Amaranthaceae
Scrophulariaceae
Poaceae
Cyperaceae
Cyperaceae
H
T
H
T
T
H
H
H
H
H
H
DD73
DD74
DD75
DD76
DD77
DD78
DD79
DD80
DD81
DD82
DD83
Vitaceae
H
Cyprus triceps
Cyperaceae
Dalbergia lactea
Datura stramonium
Desmodium dichotomum
Desmodium repandum
Desmodium salisifolium
Dichondra repens
Dichrocephala
integrifolia
Dicliptera laxata
Digitaria abyssinica
Digitaria ternata
Dioscorea bulbifera
Diospyros abyssinica
Doryopteris concolor
Dracaena afromontana
Dracaena steudneri
Droguetia iners
Echium plantagineum
Ehretia cymosa
Ekebergia capensis
Embelia schimperi
Ensete ventricosum
Entada abyssinica
Eremomastax speciosa
Species name (All
Plots)
Crassocephalum
macropappum
Crassocephalum rubens
Crossopteryx febrifuga
Crotalaria emarginella
Croton macrostachyus
Cupressus lusitanica
Cuscuta campestris
Cyathula uncinulatA
Cycnium herzfeldianum
Cynodon aethiopicus
Cyperus digitatus
Cyperus welwitschii
Cyphostemma
cyphopetalum
R. F
4 12. 90
0. 008
6. 45
6. 45
6. 45
64. 52
3. 23
16. 13
48. 39
12. 90
35. 48
9. 68
6. 45
0. 004
0. 004
0. 004
0. 040
0. 002
0. 010
0. 030
0. 008
0. 022
0. 006
0. 004
DD84
7 22. 58
0. 014
H
DD85
5 16. 13
0.
010
Fabaceae
Solanaceae
Fabaceae
Fabaceae
Fabaceae
Convolvulaceae
S
H
H
H
H
H
DD86
DD87
DD88
DD89
DD90
DD91
Asteraceae
H
DD92
Acanthaceae
Poaceae
Poaceae
Dioscoreaceae
Ebenaceae
Sinopteridaceae
Dracaenaceae
Dracaenaceae
Urticaceae
Boraginaceae
Boraginaceae
Meliaceae
Myrsinaceae
Musaceae
Fabaceae
Acanthaceae
H
H
H
H
T
H
S
T
H
H
T
T
S
H
T
H
DD93
DD94
DD95
DD96
DD97
DD98
DD99
DD100
DD101
DD102
DD103
DD104
DD105
DD106
DD107
DD108
204
2
2
2
20
1
5
15
4
11
3
2
8
3
4
14
2
11
25. 81
9. 68
12. 90
45. 16
6. 45
35. 48
3 9. 68
4
2
2
4
3
6
4
5
5
2
15
6
2
2
6
4
12. 90
6. 45
6. 45
12. 90
9. 68
19. 35
12. 90
16. 13
16. 13
6. 45
48. 39
19. 35
6. 45
6. 45
19. 35
12. 90
0. 016
0. 006
0. 008
0. 028
0. 004
0. 022
0. 006
0. 008
0. 004
0. 004
0. 008
0. 006
0. 012
0. 008
0. 010
0. 010
0. 004
0. 030
0. 012
0. 004
0. 004
0. 012
0. 008
S.N
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
Species name (All
Plots)
Erythrina brucei
Erythrococca trichogyne
Eucalyptus
camaldulensis
Euphorbia candelabrum
Euphorbia
cyparissioides
Euphorbia schimperiana
Euphorbia tirucalli
Ficus mucuso
Ficus sp.
Ficus sur
Ficus sycamoras
Ficus thonningii
Ficus vasta
Flacourtia indica
Galiniera saxifraga
Galinsoga parviflora
Gardenia volkensii
Geranium aculeolatum
Girardinia diversifolia
Glycine wightii
Gnidia glauca
Gouania longispicata
Grevillea robusta
Grewia ferruginea
Guizotia schimperi
Helinus mystacinus
Helychrysum forskaulii
Hibiscus berberidifolius
Hibiscus dongolensis
Hippocratea goetzei
Hygrophila
asteracanthoide
Hyparrhenia rufa
Hypericum peplidifolium
Hypericum revolutum
Hypoestes aristata
Hypoestes forskaullii
Hypolepis glandulifera
Fabaceae
Euphorbiaceae
%
Growth
Col.No. Freq
Freq
form
T
DD109
2 6. 45
S
DD110
8 25. 81
Myrtaceae
T
DD111
1 3. 23
0. 002
Euphorbiaceae
T
DD112
2 6. 45
0. 004
Euphorbiaceae
H
DD113
4 12. 90
0. 008
Euphorbiaceae
Euphorbiaceae
Moraceae
Moraceae
Moraceae
Moraceae
Moraceae
Moraceae
Flacourtiaceae
Rubiaceae
Asteraceae
Rubiaceae
Geraniaceae
Urticaceae
Fabaceae
Thymelaeaceae
Rhamnaceae
Proteaceae
Tiliaceae
Asteraceae
Rhamnaceae
Asteraceae
Malvaceae
Malvaceae
Celastraceae
H
T
T
T
T
T
T
T
T
T
H
S
H
H
H
S
L
T
S
H
L
H
S
S
L
DD114
DD115
DD116
DD117
DD118
DD119
DD120
DD121
DD122
DD123
DD124
DD125
DD126
DD127
DD128
DD129
DD130
DD131
DD132
DD133
DD134
DD135
DD136
DD137
DD138
12. 90
12. 90
16. 13
12. 90
35. 48
6. 45
29. 03
25. 81
38. 71
16. 13
35. 48
19. 35
12. 90
25. 81
12. 90
12. 90
19. 35
3. 23
25. 81
41. 94
12. 90
9. 68
12. 90
12. 90
25. 81
0. 008
0. 008
0. 010
0. 008
0. 022
0. 004
0. 018
0. 016
0. 024
0. 010
0. 022
0. 012
0. 008
0. 016
0. 008
0. 008
0. 012
0. 002
0. 016
0. 026
0. 008
0. 006
0. 008
0. 008
0. 016
Acanthaceae
H
DD139
5 16. 13
0. 010
Poaceae
Hypericaceae
Hypericaceae
Acanthaceae
Acanthaceae
Dennstaedtiaceae
H
H
S
H
H
H
DD140
DD141
DD142
DD143
DD144
DD145
7
4
2
3
5
2
0. 014
0. 008
0. 004
0. 006
0. 010
0. 004
Family
205
4
4
5
4
11
2
9
8
12
5
11
6
4
8
4
4
6
1
8
13
4
3
4
4
8
22. 58
12. 90
6. 45
9. 68
16. 13
6. 45
R. F
0. 004
0. 016
S.N
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
Species name (All
Plots)
Impatiens aethiopica
Indigofera spicata
Isoglossa somalensis
Jasminum abyssinicum
Jasminum repandum
Justicia ladanoides
Justicia schimperiana
Kalanchoe petitiana
Keetia guiinzii
Keetia zanzibarica
Kosteletzkya
begoniifolia
Laggera alata
Laggera crispata
Landolphia buchananii
Lantana trifolium
Leucas martinicensis
Lippia adoensis
Loxogramme abyssinica
Macaranga capensis
Maesa lanceolata
Maytenus arbutifolia
Maytenus gracilipes
Maytenus senegalensis
Maytenus undata
Micractis bojeri
Microsorium
scolopendria
Mikaniopsis clematoides
Millettia ferruginea
Momordica foetida
Myrsine africana
Nephrolepis undulata
Nicandra physaloides
Nuxia congesta
Ocimum lamiifolium
Ocimum urticifolium
Olea welwitschii
Oplismenus compositus
Oplismenus hirtellus
Balsaminaceae
Fabaceae
Acanthaceae
Oleaceae
Oleaceae
Acanthaceae
Acanthaceae
Crassulaceae
Rubiaceae
Rubiaceae
Growth
form
H
S
S
L
L
H
S
H
S
S
DD146
DD147
DD148
DD149
DD150
DD151
DD152
DD153
DD154
DD155
Malvaceae
H
DD156
Asteraceae
Asteraceae
Apocynaceae
Verbenaceae
Lamiaceae
Verbenaceae
Polypodiaceae
Euphorbiaceae
Myrsinaceae
Celastraceae
Celastraceae
Celastraceae
Celastraceae
Asteraceae
H
H
L
S
H
S
H
T
T
T
S
T
S
H
DD157
DD158
DD159
DD160
DD161
DD162
DD163
DD164
DD165
DD166
DD167
DD168
DD169
DD170
Polypodiaceae
H
DD171
Asteraceae
Fabaceae
Cucurbitaceae
Myrsinaceae
Nephrolepidaceae
Solanaceae
Loganiaceae
Lamiaceae
Lamiaceae
Oleaceae
Poaceae
Poaceae
H
T
H
S
H
H
T
S
S
T
H
H
DD172
DD173
DD174
DD175
DD176
DD177
DD178
DD179
DD180
DD181
DD182
DD183
Family
206
Col.No. Freq
5
6
4
6
3
11
3
3
3
2
%
Freq
16. 13
19. 35
12. 90
19. 35
9. 68
35. 48
9. 68
9. 68
9. 68
6. 45
1 3. 23
6
12
4
5
8
7
5
4
18
8
10
2
2
2
R. F
0. 010
0. 012
0. 008
0. 012
0. 006
0. 022
0. 006
0. 006
0. 006
0. 004
0. 002
19. 35
38. 71
12. 90
16. 13
25. 81
22. 58
16. 13
12. 90
58. 06
25. 81
32. 26
6. 45
6. 45
6. 45
0. 012
0. 024
0. 008
0. 010
0. 016
0. 014
0. 010
0. 008
0. 036
0. 016
0. 020
0. 004
0. 004
0. 004
4 12. 90
0. 008
4
10
8
4
2
2
1
7
12
4
15
4
12. 90
32. 26
25. 81
12. 90
6. 45
6. 45
3. 23
22. 58
38. 71
12. 90
48. 39
12. 90
0. 008
0. 020
0. 016
0. 008
0. 004
0. 004
0. 002
0. 014
0. 024
0. 008
0. 030
0. 008
S.N
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
Species name (All
Plots)
Otostegia tomentosa
Oxyanthus speciosus
Passiflora edulis
Paullinia pinnata
Pavonia glechomifolia
Pavonia urens
Pennisetum nubicum
Pennisetum sphacelatum
Pentas lanceolata
Peperomia abyssinica
Peperomia tetraphyla
Pergularia daemia
Periploca linearifolia
Persicaria setosula
Phoenix reclinata
Phyllanthus mooneyi
Phyllanthus ovalifolius
Physalis peruviana
Phytolacca dodecandra
Pinus patula
Piper capense
Pittosporum viridiflorum
Plantago lanceolata
Plectranthus punctatus
Podocarpus falcatus
Polyscias fulva
Pouzolzia mixta
Premna schimperi
Prunus africana
Pseudarthria hookeri
Psidium guajava
Psychotria orophila
Psydrax schimperiana
Pteris pteridioides
Pterolobium stellatum
Pychnostachys emini
Pycnostachys abyssinica
Pycreus nitida
Ranunculus multifidus
Family
Lamiaceae
Rubiaceae
Passifolraceae
Sapindaceae
Malvaceae
Malvaceae
Poaceae
Poaceae
Rubiaceae
Piperaceae
Piperaceae
Asclepiadaceae
Asclepiadaceae
Polygonaceae
Arecaceae
Euphorbiaceae
Euphorbiaceae
Solanaceae
Phytolacaeae
Pinaceae
Piperaceae
Pittosporaceae
Plantaginaceae
Lamiaceae
Podocarpaceae
Araliaceae
Urticaceae
Lamiaceae
Rosaceae
Fabaceae
Myrtaceae
Rubiaceae
Rubiaceae
Pteridaceae
Fabaceae
Lamiaceae
Lamiaceae
Cyperaceae
Ranunculaceae
207
Growth
form
S
S
L
L
H
S
H
H
S
H
H
H
L
H
T
H
S
H
S
T
S
T
H
H
T
T
S
S
T
H
T
T
T
H
S
H
H
H
H
Col.No. Freq
DD184
DD185
DD186
DD187
DD188
DD189
DD190
DD191
DD192
DD193
DD194
DD195
DD196
DD197
DD198
DD199
DD200
DD201
DD202
DD203
DD204
DD205
DD206
DD207
DD208
DD209
DD210
DD211
DD212
DD213
DD214
DD215
DD216
DD217
DD218
DD219
DD220
DD221
DD222
4
2
3
4
4
11
4
4
9
11
8
7
2
2
11
9
9
4
9
1
3
4
6
2
6
4
2
8
6
4
3
2
2
2
6
4
4
4
4
%
Freq
12. 90
6. 45
9. 68
12. 90
12. 90
35. 48
12. 90
12. 90
29. 03
35. 48
25. 81
22. 58
6. 45
6. 45
35. 48
29. 03
29. 03
12. 90
29. 03
3. 23
9. 68
12. 90
19. 35
6. 45
19. 35
12. 90
6. 45
25. 81
19. 35
12. 90
9. 68
6. 45
6. 45
6. 45
19. 35
12. 90
12. 90
12. 90
12. 90
R. F
0. 008
0. 004
0. 006
0. 008
0. 008
0. 022
0. 008
0. 008
0. 018
0. 022
0. 016
0. 014
0. 004
0. 004
0. 022
0. 018
0. 018
0. 008
0. 018
0. 002
0. 006
0. 008
0. 012
0. 004
0. 012
0. 008
0. 004
0. 016
0. 012
0. 008
0. 006
0. 004
0. 004
0. 004
0. 012
0. 008
0. 008
0. 008
0. 008
S.N
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
Species name (All
Plots)
Rhamnus prinoides
Rhoicissus tridentata
Rhus natalensis
Ricinus communis
Ritchiea albersii
Rothmannia
urcelliformis
Rubus apetalus
Rubus steudneri
Rumex natalensis
Rytigynia neglecta
Sanicula elata
Sapium ellipticum
Satureja paradoxa
Schefflera abyssinica
Schrebera alata
Senna didymobotrya
Senna occidentalis
Senna petersiana
Senra incana
Sesbania sesban
Setaria megaphylla
Setaria verticillata
Sicyos polyacanthus
Sida schimperiana
Sida tenuicarpa
Sida ternata
Solanecio gigas
Solanecio mannii
Solanum anguivi
Solanum capsicoides
Solanum dasyphyllum
Solanum giganteum
Solanum incanum
Soncus asper
Sporobolus africanus
Stachys albigena
Stellaria mannii
Stephania abyssinica
Stereospermum
%
Freq
22. 58
12. 90
16. 13
12. 90
12. 90
Rhamnaceae
Vitaceae
Anacardiaceae
Euphorbiaceae
Capparidaceae
Growth
form
S
L
S
H
T
DD223
DD224
DD225
DD226
DD227
7
4
5
4
4
Rubiaceae
T
DD228
7 22. 58
0. 014
Rosaceae
Rosaceae
Polygonaceae
Rubiaceae
Apiaceae
Euphorbiaceae
Lamiaceae
Araliaceae
Oleaceae
Fabaceae
Fabaceae
Fabaceae
Malvaeae
Fabaceae
Poaceae
Poaceae
Cucurbitaceae
Malvaceae
Malvaceae
Malvaceae
Asteraceae
Asteraceae
Solanaceae
Solanaceae
Solanaceae
Solanaceae
Solanaceae
Asteraceae
Poaceae
Lamiaceae
Caryophyllaceae
Menispermaceae
Bignoniaceae
S
S
H
S
H
T
H
T
T
S
H
T
H
T
H
H
H
S
S
H
S
T
S
S
H
S
S
H
H
H
H
H
T
DD229
DD230
DD231
DD232
DD233
DD234
DD235
DD236
DD237
DD238
DD239
DD240
DD241
DD242
DD243
DD244
DD245
DD246
DD247
DD248
DD249
DD250
DD251
DD252
DD253
DD254
DD255
DD256
DD257
DD258
DD259
DD260
DD261
6. 45
41. 94
22. 58
38. 71
16. 13
41. 94
35. 48
19. 35
3. 23
29. 03
12. 90
22. 58
6. 45
3. 23
9. 68
12. 90
12. 90
38. 71
25. 81
12. 90
22. 58
12. 90
19. 35
6. 45
6. 45
12. 90
38. 71
16. 13
6. 45
6. 45
6. 45
16. 13
16. 13
0. 004
0. 026
0. 014
0. 024
0. 010
0. 026
0. 022
0. 012
0. 002
0. 018
0. 008
0. 014
0. 004
0. 002
0. 006
0. 008
0. 008
0. 024
0. 016
0. 008
0. 014
0. 008
0. 012
0. 004
0. 004
0. 008
0. 024
0. 010
0. 004
0. 004
0. 004
0. 010
0. 010
Family
208
Col.No. Freq
2
13
7
12
5
13
11
6
1
9
4
7
2
1
3
4
4
12
8
4
7
4
6
2
2
4
12
5
2
2
2
5
5
R. F
0. 014
0. 008
0. 010
0. 008
0. 008
S.N
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
Species name (All
Plots)
kunthianum
Syzygium guineense
Tagetes minuta
Teclea nobilis
Tectaria gemmifera
Terminalia
schimperiana
Thalictrum
rhynchocarpum
Tragia cinerea
Trichilia dregeana
Trilepisium
madagascariense
Triumfetta pilosa
Triumfetta rhomboidea
Urera hypselodendron
Vangueriaapiculata
Vepris dainellii
Vernonia adoensis
Vernonia amygdalina
Vernonia auriculifera
Vernonia biafrae
Vernonia hochstetteri
Vernonia ischnophylla
Vernonia ituriensis
Vernonia karaguensis
Vernonia
theophrastifolia
Vernonia thomsoniana
Veronica abyssinica
Xanthium strumanium
Family
%
Growth
Col.No. Freq
Freq
form
Myrtaceae
Asteraceae
Rutaceae
Tectariaceae
T
H
T
H
DD262
DD263
DD264
DD265
Combretaceae
T
Ranunculaceae
R. F
35. 48
16. 13
12. 90
19. 35
0. 022
0. 010
0. 008
0. 012
DD266
6 19. 35
0. 012
H
DD267
4 12. 90
0. 008
Euphorbiaceae
Meliaceae
H
T
DD268
DD269
9 29. 03
4 12. 90
0. 018
0. 008
Moraceae
T
DD270
2 6. 45
0. 004
Tiliaceae
Tiliaceae
Urticaceae
Rubiaceae
Rutaceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
H
H
L
T
T
S
T
S
S
S
S
S
S
DD271
DD272
DD273
DD274
DD275
DD276
DD277
DD278
DD279
DD280
DD281
DD282
DD283
Asteraceae
S
Asteraceae
Scrophulariaceae
Asteraceae
S
H
H
209
11
5
4
6
12. 90
12. 90
12. 90
41. 94
32. 26
12. 90
32. 26
74. 19
6. 45
12. 90
22. 58
12. 90
12. 90
0. 008
0. 008
0. 008
0. 026
0. 020
0. 008
0. 020
0. 046
0. 004
0. 008
0. 014
0. 008
0. 008
DD284
4 12. 90
0. 008
DD285
DD286
DD287
2 6. 45
4 12. 90
2 6. 45
0. 004
0. 008
0. 004
4
4
4
13
10
4
10
23
2
4
7
4
4
-0.22 clay
0.25
sand
0.20
0.27
31
-0.42
31
-0.21
-0.18
31
0.29
-0.22 BD
0.23
CEC
0.23
0.22
pH
0.41
31
-0.05
31
0.31
0.18
31
0.02
silt
0.73
0.07
bio4
0.05
0.35
pet
0.48
31
0.01
31
0.00
0.01
31
0.01
-0.45 mi
0.01
bio1
0.01
0.45
bio5
0.45
31
0.00
31
0.01
31
0.01
-0.42 bio12
0.02
-0.01 bio13
0.94
bio3
0.35
0.05
31
0.33
Cor
0.09
0.80
0.34
0.26
0.14
0.02
N
31
31
31
31
31
31
31
31
31
31
31
31
31
31
31
N
31
31
31
31
31
31
31
31
31
31
31
31
31
31
(Elev = elevation, bio3 = isothermality, bio13 = rainfall wettest month, bio12 = mean annual rainfall, bio5 = maximum temperature
warmest month, bio1 = mean annual temperature, mi = annual moisture index, pet = potential evapotranspiration, bio4 = rainfall
seasonality, CEC = cation exchange capacity, BD = bulk density)
210
0.02
0.41
0.02
0.23
0.24
0.10
0.15
0.42
0.88
0.20
0.25
0.18
0.17
0.19
0.26
0.15
0.70
0.83
0.58
P
-0.42
0.35
-0.22
0.94
-0.22
0.99
-0.30
0.97
-0.03
0.99
-0.24
0.97
-0.25
0.94
-0.24
0.07
0.07
0.64
0.04
P
Cor
Tree
31
0.01
31
-0.09
N
Shrub
-0.57 Elev
P
0.11
Herb
Cor
0.00
Pearson
Appendix 9: Linear relationships between plant growth forms, richness and environmental variables
31
Appendix 10: Synoptic Table for grouping canopy trees in SFC
Binomial
Acacia abyssinica
Albizia gummifera
Allophylus abyssinica
Apodytes dimidiata
Bersama abyssinica
Bridelia micrantha
Celtis africana
Chionanthus mildbraedii
Clausena anisata
Cordia africana
Croton macrostachyus
Diospyros abyssinica
Dracaena steudneri
Ehreta cymosa
Ekebergia capensis
Ficus mucuso
Ficus sur
Ficus thonningi
Ficus vasta
Flacourtia indica
Galiniera saxifraga
Grewia ferruginea
Maesa lanceolata
Maytenus arbutifolia
Millettia ferruginea
phoenix reclinata
Pittosporum viridiflorum
Podocarpus falcatus
Polyscias fulva
Prunus africana
Rothmania urcelliformis
Sapium ellipticum
Schefflera abyssinica
Schrebera alata
Syzygium guineense
Terminalia schimperiana
Trichilia dregeana
Trilepisium madagascariense
Vangueria apiculata
Vepris dainellii
Vernonia amygdalina
Vernonia auriculifera
Group I
Group II
0.00
7.40
4.20
1.40
2.00
0.00
1.20
0.00
0.00
4.80
8.00
2.60
0.00
4.80
0.00
0.00
2.40
0.00
0.00
0.00
1.00
0.00
0.00
0.00
0.00
0.00
0.60
0.00
0.00
3.00
0.00
0.00
4.20
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.80
0.00
5.33
4.93
0.47
0.00
0.00
0.47
3.33
0.00
1.60
6.73
3.53
0.33
0.27
1.07
0.00
0.00
0.47
2.40
0.47
0.00
0.00
0.13
0.53
0.00
0.00
0.00
0.00
0.33
0.47
0.53
0.00
1.27
0.00
0.33
0.47
0.00
0.33
0.00
0.67
2.20
1.33
1.33
211
Group III
4.40
2.00
0.00
0.00
1.60
0.00
0.00
0.00
0.00
2.80
3.20
0.00
0.00
0.00
1.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.40
0.40
6.60
0.00
0.00
0.00
0.60
1.60
0.00
1.40
1.20
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Group IV
0.00
2.20
0.00
0.00
0.60
0.00
3.40
1.30
0.00
4.60
6.00
5.70
3.50
1.00
0.00
4.10
2.00
1.30
1.80
0.60
0.00
0.00
0.20
0.00
4.80
0.30
0.00
0.00
0.00
0.60
0.80
1.30
0.00
0.00
0.50
0.40
1.20
1.90
0.50
1.70
0.00
0.00
Appendix 11: Linear relationships between tree species abundance and environmental variables
bio1
mi
pet
bio4
silt pH cec bld sand clay
Elev bio3 bio13 bio12 bio5
Abundance Cor 0.41 0.22 0.04 0.45
-0.46 -0.47 0.46 -0.48 -0.27 0.15 0.54 0.40 0.06 0.42 0.27
P
0.03 0.26 0.86 0.01
0.01
0.01 0.01
0.01
0.16 0.45 0.00 0.03 0.76 0.02 0.16
N
29
29
29
29
29
29
29
29
29
29
29
29
29
29
29
(Elev = elevation, bio13 = rainfall wettest month, bio3 = isothermality, bio1 = mean annual temperature, bio4 = rainfall seasonality, silt =
soil silt, pH = soil pH, CEC = cation exchange capacity, bld = bulk density, bio12 = mean annual rainfall, PET = potential
evapotranspiration, MI = annual moisture index, bio5 = maximum temperature warmest month, abund = tree species abundance)
212
Appendix 12: AGC in each tree species (A) and in each plant family (B) in SFC (C t ha-1)
(C t ha-1 = carbon ton per hectare)
A (in SFC)
Species name
Albizia gummifera
Croton macrostachyus
Ficus mucuso
Acacia abyssinica
Dracaena steudneri
Cordia africana
Millettia ferruginea
Ficus vasta
Ficus sur
Sapium ellipticum
Celtis africana
Diospyros abyssinica
Prunus africana
Ficus thonningii
Ehretia cymosa
Schefflera abyssinica
Vepris dainellii
Syzygium guineense
Trilepisium madagascariense
Apodytes dimidiata
Trichilia dregeana
Allophylus abyssinicus
Maytenus arbutifolia
B (in SFC)
C t ha-1
Family
14.618 Fabaceae
9.682 Moraceae
C t ha-1
22.017
12.073
7.307 Euphorbiaceae
4.293 Boraginaceae
11.438
4.153 Dracaenaceae
4.105 Ulmaceae
4.153
4.976
1.604
3.093 Ebenaceae
1.773 Rosaceae
1.321
1.746 Araliaceae
1.714 Rutaceae
0.755
1.604 Myrtaceae
1.321 Iccacinaceae
0.294
1.214 Meliaceae
0.97 Sapindacea
0.224
0.873 Asteraceae
0.665 Celastraceae
0.116
1.214
0.423
0.270
0.193
0.110
0.394 Melianthaceae
0.294 Oleaceae
0.075
0.277 Rubiaceae
0.27 Flacourtiaceae
0.054
0.071
0.047
0.214 Myrsinaceae
0.193 Podocarpaceae
0.041
0.11 Combretaceae
0.016
213
0.019
A (in SFC)
Species name
B (in SFC)
C t ha-1
Family
C t ha-1
0.015
Polyscias fulva
0.093 Arecaceae
0.091 Pitosporaceae
Bersama abyssinica
0.075 Tiliaceae
0.002
Maesa lanceolata
Flacourtia indica
Bridelia micrantha
Schrebera alata
Chionanthus mildbraedii
Clausena anisata
Rothmania urcelliformis
Vernonia auriculifera
Vangueria apiculata
Podocarpus falcatus
Terminalia schimperiana
Phoenix reclinata
Ekebergia capensis
Galiniera saxifraga
Pittosporum viridiflorum
Grewia ferruginea
Total
0.051
0.047
0.043
0.037
0.034
0.029
0.027
0.023
0.023
0.019
0.016
0.015
0.011
0.004
0.002
0.002
61.5
Vernonia amygdalina
214
0.002
Appendix 13: AGC in each tree species (A) and in each plant family (B) in DNF
A (in DNF)
Species name
B (in DNF)
C t ha-1
Family
C t ha-1
Ficus sur
19.752 Moraceae
19.978
Apodytes dimidiata
14.152 Icacinaceae
14.152
Syzygium guineense
8.085 Fabaceae
9.525
Celtis africana
6.609 Myrtaceae
8.085
Albizia gummifera
6.485 Ulmaceae
6.609
Schefflera abyssinica
5.842 Araliaceae
6.102
Olea welwitschii
3.203 Oleaceae
4.863
Millettia ferruginea
3.040 Euphorbiaceae
3.935
Prunus africana
2.464 Rosaceae
2.464
Macaranga capensis
2.307 Meliaceae
1.549
Chionanthus mildbraedi
1.660 Rutaceae
1.321
Croton macrostachyus
1.628 Melianthaceae
0.915
Ekebergia capensis
1.156 Rubiaceae
0.842
Bersama abyssinica
0.915 Sapindaceae
0.528
Galiniera saxifraga
0.892 Podocarpaceae
0.383
Vepris dainellii
0.723 Boraginaceae
0.358
Allophylus abyssinicus
0.528 Loganiaceae
0.212
Trichilia dregeana
0.393 Arecaceae
0.168
Podocarpus falcatus
0.383 Celastraceae
0.038
Cordia africana
0.358 Simaroubaceae
0.002
Teclea noblis
0.310
Polysciasfulva
0.261
Ficus sycamoras
0.226
Nuxia congesta
0.212
215
A (in DNF)
Species name
B (in DNF)
C t ha-1
Phoenix reclinata
0.168
Canthium oligocarpum
0.121
Rothmania ulceriformis
0.071
Maytenus arbutifolia
0.038
Vangueria apiculata
0.027
Psychotria orophila
0.013
Oxyanthus speciosus
0.007
Brucea antidysenterica
0.002
Total
Family
82.029
216
C t ha-1
Appendix 14: AGC in each tree species (A) and in each plant family (B) in woodland
A (in woodland)
B (in woodland
Species name
C t ha-1
Family
C t ha-1
Acacia abyssinica
4.406
Anacardiaceae
0.016
Acacia lahai
0.094
Arecaceae
0.034
Albizia gummifera
0.116
Asteraceae
0.051
Bridelia micrantha
0.005
Bignoniaceae
0.555
Combretum collinum
0.586
Boraginaceae
0.834
Combretum molle
0.019
Combretaceae
1.374
Cordia africana
0.834
Euphorbiaceae
0.129
Croton macrostachyus
0.069
Fabaceae
5.532
Entada abyssinica
0.896
Flacourtiaceae
0.112
Euphorbia tirucalli
0.003
Moraceae
2.510
Ficus sur
0.512
Myrsinaceae
1.703
Ficus sycamoras
1.190
Myrtaceae
0.007
Ficus vasta
0.808
Rubiaceae
0.008
Flacourtia indica
0.112
Gardenis volkensii
0.008
Maesa lanceolata
1.703
Millettia ferruginea
0.005
Phoenix reclinata
0.034
Psidium guajava
0.007
Rhus natalensis
0.016
Sapium ellipticum
0.051
Senna petersiana
0.009
Sesbania sesban
0.007
Stereospermum kunthianum
0.555
Terminalia schimperiana
0.769
Vernonia amygdalina
0.051
Total
12.865
217
Appendix 15: AGC in each tree species (A) and in each plant family (B) in pasture
A
Species
B
C t ha-1
Family
C t ha-1
Acacia abyssinica
0.073
Bignoniaceae
0.042
Bauhinia tomentosa
0.019
Boraginaceae
0.003
Combretum collinum
0.124
Combretaceae
0.124
Ehretia cymosa
0.003
Euphorbiaceae
0.243
Entada abyssinica
0.022
Fabaceae
0.113
Ficus vasta
1.932
Flacourtiaceae
0.008
Flacourtia indica
0.008
Moraceae
1.932
Gardenia volkensii
0.012
Myrsinaceae
0.002
Keetia zanzibarica
0.003
Myrtaceae
0.023
Maesa lanceolata
0.002
Rubiaceae
0.015
Sapium ellipticum
0.243
Stereospermum kunthianum
0.042
Syzygium guineense
0.023
Total
2.507
218
Appendix 16: AGC in each tree species (A) and in each plant family (B) in croplands
A
Species
B
C t ha-1
Families
C t ha-1
Acacia abyssinica
0.191 Araliaceae
0.125
Albizia gummifera
0.002 Asteraceae
0.103
Combretum molle
0.123 Boraginaceae
1.171
Cordia africana
0.794 Combretaceae
0.122
Ficus vasta
0.114 Fabaceae
0.161
Ficus sycamoras
0.299 Moraceae
0.344
Prunus africana
0.611
Scheffleria abyssinica
0.150
Terminalia schimperiana
0.024
Vernonia amygdalina
0.123
Total
2.432
219
29 0.02 -0.44 bio5
29 0.10 -0.31 bio4
29 0.02 -0.44 bio7
29 0.02 -0.43 bio6
29 0.01 -0.45 bio1
29 0.42 -0.16 bio3
mi
29 0.02 0.44
29 0.01 -0.45 bio2
mimq
bio17
29 0.01 0.45
29 0.02 0.44
bio16
29 0.30 0.20
29 0.02 -0.46 pet
bio15
29 0.12 0.30
29 0.18 -0.25 bio14
29 0.97 -0.01 bio13
bio12
29 0.02 0.42
P
29 0.02 -0.44 bio11
Cor
29 0.02 -0.44 bio10
AGC _4th root
Appendix 17: Linear relationships between AGC and climate variables
N
bio10 = mean temperature warmest quarter, bio11 = mean temperature coolest quarter, bio12 = mean annual rainfall, bio13 =
rainfall wettest month, bio14 = Rainfall driest month, bio15 = Rainfall seasonality, bio16 = Rainfall wettest quarter, bio17
=Rainfall driest quarter, pet = potential evapotranspiration, mimq = Moisture index moist quarter, mi = Annual moisture
index, 2 = Mean diurnal range in temperature, bio3 = Isothermality, bio1 = mean annual temperature, bio6 = Min temp coolest
month, bio7 = Annual temperature range, bio4 =Temperature seasonality, bio5 = Max temp warmest month
220
LAI_v6
Cor
.
P
N
Cor
.
P
N
Pearson
bio3
29
0.785
0.053
29
0.823
0.043
29
0.029
-0.406
29
0.024
-0.418
29
0.515
-0.126
29
0.532
-0.121
29
0.053
0.363
29
0.045
0.375
29
0.029
-0.405
29
0.025
-0.415
29
0.051
0.365
29
0.044
0.376
29
0.026
-0.412
29
0.022
-0.423
29
0.665
0.084
29
0.633
0.093
29
0.014
-0.452
29
0.012
-0.46
29
0.503
-0.13
29
0.496
-0.132
29
0.057
-0.357
29
0.05
-0.368
29
0.032
-0.4
29
0.027
-0.41
29
0.026
-0.414
29
0.022
-0.425
29
0.022
0.423
29
0.017
0.44
29
0.035
0.394
29
0.029
0.405
29
0.269
0.212
29
0.252
0.22
29
0.027
-0.41
29
0.024
-0.419
29
0.182
-0.255
29
0.167
-0.264
bio10
bio13
bio12
bio5
mimq
bio1
bio16
221
bio7
bio14
bio6
bio11
bio2
bio17
mi
bio15
pet
bio4
Appendix 18: Linear relationships between true_LAI indices and climate variables
LAI Type
LAI_v5
Appendix 19: Habitat suitability for the distribution of five plant species in Ethiopia under baseline and projected climate (A1,
B1, C1, D1, E1 = baseline scenario, A2, B2, C2, D2, E2 = Projected climatescenario)
A1
A2
222
B2
B1
C1
C2
223
D1
D2
E1
E2
224
Appendix 20: Jackknife test (training and test data) for the distribution of five plant species (A1, B1, C1, D1, E1 represent training gain
under baseline scenario; A2, B2, C2, D2, E2 represent test gain under the baseline scenario; A3, B3, C3, D3, E3 represent training gain
under the projected climate; A4, B4, C4, D4, E4 represent test gain under the projected climate
Jackknife test for the distribution of five species
A2
A1
A4
A3
B2
225
B1
B3
B4
C1
C2
C3
C4
D1
D2
226
D3
D4
E2
E1
E3
E4
A1-4 = Acacia abyssinica, B1-4 = Cordia africana, C1-4 = Millettia ferruginea, D1-4= Phytolacca dodecandra, E1-4= Schefflera
abyssinica)
227
I, the undersigned, hereby declare that this thesis is my original work and that all
sources of materials used for the thesis have been acknowledged.
Name: Dereje Denu Rebu
Signature: ____________________
Date of submission: May 06, 2016
228