PHE- CONSORTIUM

A Resource Base and Climate Change Risk Maps for

Tezera Chernet ERCAND Consult

2015,

Financed by the SCIP Fund: the SCIP Fund is financed by the Governments of the United Kingdom (UK), Norway and Denmark. Table of Contents

Table of Contents

ACRONYMS ...... VI ACKNOWLEDGMENTS ...... VII EXECUTIVE SUMMARY ...... VIII 1. OVERVIEW OF AWASH NATIONAL PARK ...... 1

1.1 INTRODUCTION ...... 1 1.2 OBJECTIVE...... 2 1.3. THE STUDY AREAS ...... 2 1.3.1 Location ...... 2 1.3.2 Topography ...... 3 1.3.3 Soil ...... 4 1.3.4 Water Resources and drainage ...... 4 1.3.5 Brief History of the Awash National park ...... 4 2. LAND COVER CLASSIFICATION ...... 6

2.1 METHODOLOGY ...... 6 2.1.1 Data set ...... 6 2.1.2 Preprocessing of Images ...... 6 2.1.3 Image Classification for land use and land cover mapping ...... 6 2.1.5 Land cover change detection ...... 7 2.1.6 Identification of change drivers ...... 7 2.1.7 Materials and tools ...... 7 2.2. RESULTS AND DISCUSSION ...... 8 2.2.1 Present Vegetation types ...... 8 2.2.2 Description of present land cover ...... 9 2.2.3 Land use / cover changes ...... 12 2.2.4 Drivers of Land cover Change ...... 12 3. WILD ANIMALS HABITAT MAPPING ...... 13

3.1 METHODOLOGY ...... 13 3.1.1 Data Set...... 13 3.1.2 Habitat assessment and mapping ...... 13 3.2 RESULTS AND DISCUSSIONS ...... 14 3.2.1 Bio-Diversity of Awash National Park ...... 14 3.2.2 Trends in Number of Wild-animal Population ...... 16 3.2.3 Habitat Fragmentation ...... 18 3.2.3.1 Deforestation and Degradation ...... 18 3.2.3.1.1 Expansion of Cultivated Land ...... 18 3.2.3.1.2 Competition from Grazing by Domestic Animals ...... 19 3.2.3.1.3 Collection of Fuel and construction wood ...... 19 3.2.3.2 Infrastructure development ...... 19 3.2.3.3 Impacts of Fragmentation ...... 20 3.2.4 Other Pressures ...... 21 3.2.4.1 Road Kill ...... 21 3.2.4.2 Alien Invasive Species ...... 22 4. CLIMATE CHANGE RISK PROFILES ...... 23

ERCAND Consult Page ii Table of Contents

4.1 OBJECTIVE...... 23 4.2 METHODOLOGY ...... 23 4.2.1 Data Set...... 23 4.2.2 Hazard Assessment ...... 23 4.2.2.1 Fire hazard Assessment ...... 24 4.2.2.2 Drought hazard Assessment ...... 26 4.2.2.3 Flood hazard Assessment ...... 27 4.2.3 Risk Assessment ...... 28 4.3 RESULTS AND DISCUSSIONS ...... 28 4.3.1 Climate ...... 28 4.3.1.1 The Present Climate of ANP ...... 28 4.3.1.2 The Future Climate ...... 30 4.3.2 Hazards Assessment ...... 30 4.3.2.1 Fire Hazard ...... 30 4.3.2.1.1 Fire Frequency ...... 30 4.3.2.1.2. Fire Prone Areas ...... 34 4.3.2.2 Drought Hazard ...... 34 4.3.2.3 Flood Hazard ...... 35 4.3.2.3.1 Flood Prone Areas ...... 35 4.3.2.3.2 Flood Frequency ...... 36 4.3.2.3.3 Beseka Lake Level ...... 38 4.3.3 Vulnerability Assessment ...... 40 4.3.4 Risk Assessment ...... 40 4.3.4.1 Fire Risk ...... 40 4.3.4.2 Drought Risk ...... 41 4.3.4.3 Flood Risk ...... 42 5. CONCLUSIONS AND RECOMMENDATIONS ...... 44

5.1. CONCLUSIONS ...... 44 5.2. RECOMMENDATIONS ...... 45 6. REFERENCES ...... 47

ERCAND Consult Page iii Table of Contents

List of Tables

TABLE 1: AREA SIZES OF LAND COVER TYPES ...... 9 TABLE 2: POPULATION SUPPORT CAPACITY ANALYSIS FOR WOREDA (SOURCE: WBISPP) ...... 18 TABLE 3: VARIABLES AND WEIGHT FACTOR USED IN FIRE HAZARD MAPPING ...... 26 TABLE 4: FACTORS USED IN FLOOD HAZARD ASSESSMENT ...... 27

List of Figures

FIGURE 1: LOCATION MAP OF THE STUDY AREAS ...... 3 FIGURE 2: PROFILE GRAPH: BESEKA LAKE - TOP OF FENTALE MOUNTAIN – HOT SPRING AT ANP ...... 3 FIGURE 3: PRESENT LAND USE / COVER MAP OF ANP ...... 8 FIGURE 4: LAND USE / LAND COVER CLASS CHANGE BETWEEN 1973 AND 2014 AT ANP ...... 12 FIGURE 5: CHANGE IN NUMBER OF SPECIES FOR SOME WILD ANIMALS OCCURRING IN ANP BETWEEN 2008 AND 2010 ...... 16 FIGURE 6: MAP DEPICTING THE PRESENT DISTRIBUTIONS OF SOME WILD ANIMALS AT ANP...... 17 FIGURE 7: FREQUENT ROAD KILL AREAS BY SPECIES TYPE AT ANP ...... 21 FIGURE 8: CROSS SECTION VIEW OF ROAD CROSSING ANP ...... 21 FIGURE 9: MAPS OF WEIGHTED VARIABLES ...... 25 FIGURE 10: MEAN MONTHLY RAINFALL: METHARA...... 28 FIGURE 11: MEAN MONTHLY TEMPERATURE: METHARA...... 29 FIGURE 12: THERMAL ZONES ...... 29 FIGURE 13: NUMBER OF FIRES OBSERVED BY YEAR AT AWASH NATIONAL PARK ...... 30 FIGURE 14: NUMBER OF FIRES OBSERVED BY MONTH AT AWASH NATIONAL PARK...... 31 FIGURE 15: NUMBER OF FIRES OBSERVED AT AWASH NATIONAL PARK AND SURROUNDING AREAS BY MONTH ...... 31 FIGURE 16: DISTRIBUTION OF FIRE POINTS IN THE KAS SURROUNDING ANP ...... 32 FIGURE 17: NUMBER OF FIRES OBSERVED PER YEAR AT AWASH NATIONAL PARK AND SURROUNDING AREAS ...... 32 FIGURE 18: LOCATION OF FIRE PIXELS IN ANP AND THE SURROUNDING AREAS ...... 33 FIGURE 19: ANP - FIRE HAZARD MAP ...... 34 FIGURE 20: COMPARISON OF NDVI IMAGES OVER ANP (NORTH EASTERN PART OF IMAGE) AND SURROUNDING AREAS FOR 2003 (A) AND 2008 (B) ...... 35 FIGURE 21: FLOOD PRONE AREAS ...... 36 FIGURE 22: MEAN ANNUAL FLOW OF AT METHARA ...... 37 FIGURE 23: MONTHLY FLOW OF AWASH RIVER AT METHARA GAUGING STATION (1982-2009) ...... 37 FIGURE 24: LOG PEARSON TYPE III ANALYSIS (USING AVERAGE DAILY MAXIMUM FLOW VALUES) FOR AWASH RIVER AT METHARA STATION...... 38 FIGURE 25: WATER LEVEL OF BESEKA LAKE BETWEEN 2009 AND 2013 ...... 38 FIGURE 26: FIRE RISK MAP OF ANP ...... 41 FIGURE 27: FLOOD RISK MAP OF ANP ...... 42

List of Plates

PLATE 1: THICKET AT ANP ...... 9 PLATE 2: BUSH LAND (A) AND (B) AT ANP ...... 10 PLATE 3: GRASSLAND WITH TREES AT ANP ...... 10

ERCAND Consult Page iv Table of Contents

PLATE 4: GRASSLAND WITH SHRUBS AT ANP ...... 10 PLATE 5: RIVERINE FOREST AT ANP ...... 11 PLATE 6: DOUM PALM AT ANP ...... 11 PLATE 7: DIK-DIK AT ANP ...... 14 PLATE 8: SOEMMERING’S GAZELLE AT ANP ...... 14 PLATE 10: BIRDS AT ANP ...... 15 PLATE 9: HAMADRYAS BABOONS AT ANP ...... 15 PLATE 11: GOAT AND SHEEP HERDS NEARBY ILALA-SALA PLAIN AT ANP ...... 19 PLATE 12: PART OF ADDIS- RAIL-WAY UNDER CONSTRUCTION AT ANP ...... 20 PLATE 13: COMPARISON OF IMAGES OVER BESEKA LAKE (FROM LEFT TO RIGHT: 1973, 1986, 1998 AND 2014) ...... 39 PLATE 14: NORTHERN PART OF LAKE BESEKA (AT THE BACKGROUND IS FENTALE MOUNTAIN) ...... 39

List of Annexes

ANNEX 1: LIST OF WILD ANIMALS AT AWASH NATIONAL PARK LISTED IN THE IUCN RED LIST ...... 49 ANNEX 2: NUMBER OF DOMESTIC ANIMALS AT KAS SURROUNDING AWASH NATIONAL PARK ...... 50 ANNEX 3: COMPARISON OF NO. OF LIVESTOCK IN FENTALE WOREDA BETWEEN 2008/09 AND 2013/14K ...... 51

ERCAND Consult Page v Acronyms

Acronyms

ANP Awash National Park ASTER Advanced Space-borne Thermal Emission and Reflection Radiometer CSA Central Statistics Agency DEM Digital Elevation Model EMA Ethiopian Mapping Agency EWCA Ethiopian Wildlife Conservation Authority FDRE Federal Democratic Republic of Ethiopia FIRMS Fire Information for Resource Management System GCP Ground Control Point GIS Geographical Information System GPS Global Positioning System HH House hold KA Kebele Administration LUPRD Land-use Planning and Regulatory Department m.a.s.l. meters above sea level MoARD Ministry of Agriculture and Rural Development MODIS Moderate Resolution Imaging Spectro-radiometer MSS Multi-Spectral Scanner NMSA National Meteorological Service Agency PA Protected Area PHE Population, Health and Environment SCIP Strategic Climate Institutions Program SDPASE Sustainable development of the protected area system of Ethiopia TM Thematic Mapper UTM Universal Transverse Mercator WBISPP Woody Biomass Inventory and Strategic Planning Project WGS World Geodetic System

ERCAND Consult Page vi Acknowledgment

Acknowledgments

This study was financed by the SCIP Project and the author would like to thank PHE-Ethiopia Consortium management and staff for providing the opportunities and facilitating the logistic support to conduct the study. The author appreciates and thanks Ato Tadesse Hailu, project coordinator and Mr. Adrian Young, Environmental Program officer at PHE-Ethiopia, for their valuable advice to improve the document. Thanks are also due to EWCA for permission to work and support in acquisition of data on the management plan of the park. Awash National Park Office and the management were also very supportive in facilitating the field work within the Protected Area; Ato Mamo Alemu and Ato Zerihun Ketema, wildlife experts at Awash National Park have also helped in the acquisition of ground data, provision of valuable description of habitats and wild animals frequenting it and arranging consultative meetings with sector Offices.

Due appreciation also goes to the Department of Hydrology and GIS Section under the Ministry of Water and Energy for their prompt provision of data on discharge and water level and spatial data produced under the Awash basin studie. We acknowledge the use of FIRMS data and imagery from the Land, Atmosphere Near real-time Capability for EOS (LANCE) system operated by the NASA/GSFC/Earth Science Data and Information System (ESDIS) with funding provided by NASA/HQ. Gratitude is also due to Ato Arega Mekonen, SDPASE-GIZ national project coordinator, for provision of spatial data on Awash National Park.

The author is very grateful to Ato Kidane Mengistu, General Manager, ERCAND Consult, Dr. Shiferaw Alem and Yehuala S. for support and providing comments on the draft document. Thanks are also due to Ato Temsgen Yohannes in the vegetation survey and Ato Hailu for assistance during the field work. Special thanks also goes to park scouts for assistance in vegetation survey and providing valuable information on habitats, staff members of Woreda Agricultural Offices in the administration of household level questionnaire and farm households at the KAs who have spent their precious time for providing responses.

ERCAND Consult Page vii Executive Summary

Executive Summary

This study presents the resource bases at Awash National Park (ANP) and examines future exposure to identified climate change impacts. The methodology used includes analysis of Landsat images to delineate the vegetation units and describe time-series land cover changes. Locations of wild animals inhabiting the park were also mapped based on field observation, documented data and discussions with key informants. Further, using basic climatic, historical river flow and lake level data, Digital Elevation Model (DEM), Normalized Difference Vegetation Index (NDVI) and FIRMS hot spots data analysis were made to evaluate exposure to hazards.

Comparison of time series Landsat image over a period of 40 years on Awash national park showed that there is a reduction in size of the grassland units with most of the gain for grassland with shrub. A major driver of change at ANP was competition for grazing by domestic animals.

Fire, proneness to drought, river flooding and lake level rise were the types of hazards faced by the protected area. An occurrence of fire hazard within the protected area was of anthropogenic origin. Most fire prone areas were also located in or adjoining areas where such activities are permitted.

Fire risk assessment shows that 64.28% of ANP area was found in high fire risk areas and 30.60% under moderate fire risk areas. An estimated 90% of land cover units whose major component is the grassland, namely, the grassland, grass with trees and grass with shrub, fall under high fire risk area. Likewise, about 24% and 73% of thicket and bush lands are found in areas categorized as high and moderate fire risk areas respectively. Analysis of the flood risk reveals that, of the 14.6 percent of the park area susceptible to flooding, 3.27 percent had a very high and 17.17% a high risk of flooding. 3.25 percent of land cover classes in which the major component is the grassland, and a proportionate percent of thicket and bush lands fall under very high flood risk area. Findings will help strengthen efforts in appraisal of natural resources toward meeting challenges of climate change.

ERCAND Consult Page viii Section 1: Overview

1. Overview of Awash National Park

1.1 Introduction Toward meeting the multiple needs of communities while building resilience to future development and climate change, the core interventions of Population, Health and Environment Ethiopia Consortium (PHEEC) programs focus in situation where demographic trends put pressure on the environment and where natural resources degraded threaten human health and livelihoods (PHEEC, 2012).

Ethiopia is known to possess some 284 wild mammal, 861 bird, 201 reptile, 63 amphibian, 188 fish and 1225 arthropod species with about 10, 2, 5, 54, 0.6 and 21 per cent of endemism, respectively (IBC, 2009). Protected areas have significant value in the conservation and development of these resources. The Ethiopian government, in recognition of these, has put in place conservation mechanisms in a bid to protect wild animals from various disasters (FDRE, 2005). The wildlife conservation areas of the country are estimated to cover 158,200 km2 or about 14% of the country size.

Despite the efforts made to conserve and develop the protected areas, there had been different pressures and mostly were of human origin. Furthermore, these are now exacerbated by climate change; the extent and degree in which climate change affects local level Protected Areas (PAs) are also little known. Awash National Park (ANP) was one of the first two national park gazetted in the country. In the past forty years since establishment, however, there are still a number of constraints that hinder the expected progress.

Accordingly, understanding the nature and characteristics of the existing resources, trends in the past forty years, and the pressures now faced along with its exposure to climate change related stresses, shall help in developing mechanisms to protect and develop the resources.

ERCAND Consult Page 1 Section 1: Overview

The first chapter of this paper provides the background and objective of the study and describes the study area. Chapter two explains the present land cover and examines the historical trends of land use and cover change. Chapter three discusses the habitat type and distribution within the park including habitat fragmentation and make reviews of pressures faced. Chapter four focuses on hazards related with climate change and is followed by the final chapter, which presents the conclusion and recommendations.

1.2 Objective The general objective of this study was to assess the biological resource potential and values of the Awash and develop resource as well as climate risk maps.

The specific objectives of this study were: a. to assess and develop baseline data for the terrestrial biological resources of Awash National park; b. to assess the biological resources present in the national park and their location in order to enable them to be conserved appropriately; c. to provide information for decision makers and justify the conservation and protection of the park ecosystems as a mechanism to eradicate poverty and mitigate and adapt to climate change; and d. to create climate change risk profiles that cover all the major hazards (floods, droughts, forest fires, landslides etc.) prevailing in and around Awash National Park.

1.3. The study areas

1.3.1 Location Awash National park, is found some 220 Km East of Addis Ababa in the Awash basin, part of the Great Rift, between 8o40' and 9o9' North and 39o50' to 40o10' East (Figure 1). Administratively, it is located between the Afar and Regional states. Two districts from the former in the east and north east and four districts from the latter in the North- west, west and south bound the national park.

ERCAND Consult Page 2 Section 1: Overview

Figure 1: Location map of the study areas

1.3.2 Topography Altitudinal variations in ANP range between 740 near the hot spring to 1820 m.a.s.l at top of Mt. Fentale. lies to the southwest of the park; some 3.6 kms from the park boundary towards north and some 10 Kms towards east (Figure 2).

Figure 2: Profile graph: Beseka lake - top of Fentale Mountain – Hot spring at ANP

ERCAND Consult Page 3 Section 1: Overview

80% of the Awash national park area has slopes lower than 5 degrees, while the remaining area, particularly at Mt. Fentale and a nearby hill, has varying slopes; up to 70 degrees in certain places.

1.3.3 Soil

Based on the study made by FAO/LUPRD (1982), the dominant soil types found in the ANP are Lithic leptosols (36%), Eutric cambisols (35%) and Vertic cambisols (28%) with small occurrence of Eutric leptosols (1%).

1.3.4 Water Resources and drainage

Awash river passes in the south and south east part of the national park and forms part of the boundary between East Hararghe and Arsi zones of Oromia Region. The river makes a water fall, south of Gotu camp, before entering the gorges. Hot spring is also found at the far northern part of the park. Pastoralists and agro-pastoralists found around the park use the water to drink their animals. Two other man-made watering points are located within the park designed in the early days of the park to drink wild animals.

1.3.5 Brief History of the Awash National park Awash National park, was gazetted in 1969. With respect to its area size, the ANP previously had an area of 760 km2, but since 2002 it was reduced to 598 Km2.

It is believed that ANP is a haven for more than 78 mammals, including the Near threatened species like Beisa Oryx (Oryx beisa) and (Tragelaphus imberbis), and of more than 472 bird species, including six endemic (Yirmed and Girum, 2013).

An effort had been made to introduce some species such as Ostrich (Struthio camelus) and Swayne’s (Alcelaphus buselaphus swaynei) to the national park in the mid 1970's. In the recent past, the park was Ostrich’s natural habitat, however, when the number declined and was on the verge of diminishing, reintroduction, including ostrich eggs from Alegedhi, had been made. Despite these, the population declined and the only remaining ostrich was killed by vehicle collision in 1995. At the same time, some 60 Swayne’s had also been introduced to the park from Senkele sanctuary; however, the effort had been unsuccessful as it survived only to the late 1990s. Similarly,

ERCAND Consult Page 4 Section 1: Overview wild animals such as Bush duiker (Sylvicapra grimmia), African hunting dog (Lycaoii pictiis), buffalo (Syncerus caffer), Grevy’s (Equus grevyi), black (Diceros bicornis) and African (Loxodonta Africana) had once inhabited this national park.

Tourists that have visited ANP in 2013/14 were 11,843. During the same time period, the tourism industry has generated 1.2 million ETB (of which 64,000 ETB is from penalties paid by trespassers).

ERCAND Consult Page 5 Section 2: Land Cover Classifications

2. Land Cover Classification

2.1 Methodology

2.1.1 Data set Data that characterize the protected areas and influence the fauna and flora existing within it are acquired and/or generated to form the individual layers for use in GIS analysis. These include raster and vector data, and other available historical survey and map data from local institutions and on-line databases.

Landsat images acquired between mid 1970’s and 2014 are used to estimate the geographic extent, distribution and types of land cover changes in the two national park and their immediate surroundings. Overall eight cloud free scenes, from sensors including L8, TM and MSS, and taken between December and early January were used in the study.

2.1.2 Preprocessing of Images The image scenes received were geo-referenced using GCP and the required atmospheric corrections were applied. Further, images were stacked, mosaicking and sub-setting to the area of interest done for proper investigation of land cover changes.

2.1.3 Image Classification for land use and land cover mapping A preliminary land use land cover map (through unsupervised classification) was generated to help guide the field work. Data collection formats were also developed for vegetation survey and the collection of information on wild animals.

The sampling design employed was stratified sampling. Adequate and representative ground data were collected from selected observation plots in December, 2014. At each plot, the species composition, percent cover of the different cover types, life-forms, and other relevant characteristic and information were recorded on prepaed formats along with GPS location data.

Maximum Likelihood classification technique, which assumes that the statistics for each class in each band are normally distributed, was employed to classify the satellite images.

ERCAND Consult Page 6 Section 2: Land Cover Classifications

Post-classification change detection method was employed to determine magnitude of change. To validate classification, reference data were collected separately and the associated statistics were generated.

2.1.5 Land cover change detection

Landsat images, acquired between early 1970’s and 2014, at an interval of 12 years, were analyzed to estimate the geographic extent, distribution and types of land cover changes in the protected areas. The time span selected between series of layers was determined based on availability of cloud free data and preferred season.

Historical maps and data, depicting vegetation types and distribution, were used to identify cover units. These include studies made by LUPRD, WBISPP, Awash Basin Study and aerial photographs.

2.1.6 Identification of change drivers Local level data on proximate factors was collected from sampled households of selected Kebele administrations that adjoin the national park, through discussions with key informants and on site observations. Questionnaires were administered at household level by individuals who have a good knowledge of the locality and its culture. A total of 31 HH were selected randomly from four KAs to conduct the survey. Furthermore, basic information such as population census, stock density and other available relevant documents and reports were collected from CSA and adjoining woreda administration and reviewed.

2.1.7 Materials and tools ENVI 4.3 was used in the processing, analysis and display of raster data. Further more, ArcGIS had also been used for manipulation, editing and viewing of geographic and tabular data. SPSS was used for the analysis, conversion and presentation of collected HH data.

Garmin 12 GPS and silva-compass were used for identification of waypoints (locate ground control points), registration of geographical coordinates in the description of land cover units and marking of training areas for signature definition, during the field survey. Furthermore, binoculars were used in habitat assessments, desktop and notebook computers were also used for entry, processing and analysis of different input data.

ERCAND Consult Page 7 Section 2: Land Cover Classifications

2.2. Results and Discussion

2.2.1 Present Vegetation types

Based on classification of Landsat images, eight different land use / cover classes were identified in ANP (Fig. 3). Grassland with scattered shrubs (27.42%), Grassland with scattered trees (23.67%), Thicket (22.86%) and Bushland (14.26%) constitute the major cover types at Awash national park (table 1).

Figure 3: Present Land use / cover map of ANP

ERCAND Consult Page 8 Section 2: Land Cover Classifications

Table 1: Area sizes of land cover types Cover type / use Symbol Area (ha) Percentage Thicket F1 13479.00 22.86 Bush land F2 8409.00 14.26 Grassland F3 5685.67 9.64 Grassland with scattered shrubs F4 16168.67 27.42 Grassland with scattered trees F5 13957.67 23.67 Palm Forest F7 375.00 0.64 Exposed surface N 305.00 0.52 Lava flow L 587.00 1.00 TOTAL 58967.00 100.00

2.2.2 Description of present land cover

Thicket (F1) had a composition of trees, shrub, and grass between 5-20%, 10-15% and 25- 55%, respectively (Plate 1). Exposed surface within the unit occupy 10-25%. Characteristics tree species found in the unit include Acacia senegal, Acacia tortolis, and Dobera glabra. A. mellifera is also dominant among the shrub layer while Balanites aegyptica and others are also present among others. The unit covers an area of 13479 hectare

(22.8%). Plate 1: Thicket at ANP

Bushland (F2) covers an area of 8409 hectare (14.3%). It is characterized by trees (<10%), shrubs (25-55 %) and grasses (20-70%). Most of the identified Bush land areas were found dispersed along the eastern side, south western and central part of the park. Balanites spp., A. senegal, A. tortolis, Celtis spp., Acacia oerfota, Ziziphys spp., A. nilotica, Dobera glabra, A. mellifera and Grewia spp were some of the species found in the unit (see Plate 2A).

ERCAND Consult Page 9 Section 2: Land Cover Classifications

Plate 2: Bush land (A) and Grassland (B) at ANP

Grassland (F3) covers an area of 5685 hectare (9.6%) and is characterized by trees (0-1 %), shrubs (0-2%), grasses (80-90%) and bare lands (<15%). Scattered trees found on the grassland include Acacia nilotica and saplings of B. aegyptica, A. Senegal, A. tortolis and A.

mellifera (see Plate 2B).

Grassland with scattered Trees (F4): the unit is mostly found in the lower central part of the park and covers an area of 16168 hectare (Plate 3). It is characterized by trees (<5 %), shrubs (2-5%), grasses (70-90%) and bare lands (up to 10%). The shrub layer includes species such as Dichristachy cineria, A.nilotica, A.senegal,

and Commiphora spp. Plate 3: Grassland with Trees at ANP

Grassland with scattered Shrubs (F5) covers an area of 13957 hectare (23.67%). The unit (Plate 4) is composed of herbs (65%), trees (< 2%), shrubs (< 15%) and bare land (< 20%),.

Plate 4: Grassland with shrubs at ANP

ERCAND Consult Page 10 Section 2: Land Cover Classifications

Riverine forest (F6) is located along the Awash river and is not mapped due to the spatial resolution of the classified image. Tree species found in the unit include Acacia etbaica, Celtis africana, Ficus sur, Grewia bicolor, Tamarindus indica, Ziziphus muro, and Z.

mucronata. Saplings encountered Plate 5: Riverine forest at ANP include Z. mulenona, B. discolor, Z.

mucronata, Acacia senegal and Ficus sur. Seedlings of and T. indica were also observed (Plate 5).

Palm trees (F7) are located in the northern and north eastern part of the national park, around the hotspring (Plate 6). The unit covers an area of 375 ha and the dominant species was Hyphaene coriacea.

Plate 6: Doum Palm at ANP

Marsh land (W): the unit is not mapped. It is located adjoining the hot springs and has a 96- 100% real cover; water surface cover <4% of the unit. Papyrus was the dominating grass

type.

Lava flow (L) is almost all located at the top most and parts of the upper slope of Fentale mountain. The unit cover an area of 587 hectares.

ERCAND Consult Page 11 Section 2: Land Cover Classifications

2.2.3 Land use / cover changes

Class change analysis at ANP (Figure 4) shows that the most persistent and considerable change, in all measured time period, had occurred in the grassland unit. Between the end of 1990’s and 2014 significant part of this cover type had gone to Grassland with shrub (34.6%) and an almost equivalent amount (29.14%) was changed to grassland with trees. Similarly, between the early 1970’s and mid 1980’s, two-third of the loss in grassland was a gain to Grassland with shrub. Since the early 1970s there were also regular, but significantly declining, changes in thicket.

Figure 4: land use / land cover class change between 1973 and 2014 at ANP

2.2.4 Drivers of Land cover Change The vegetation types at ANP had undergone changes in the studied period mainly due to the influence of over-grazing by domestic animals which came from different directions and destruction of trees.

Majority of surveyed HH (70%) at ANP had also expressed their opinion that high number of animal population coupled with scarcity of grazing land and animal feed, especially during the dry season, had led to degradation of the vegetation. Other mentioned causes were also an increase in human population and occurrence of drought.

ERCAND Consult Page 12 Section 3: Wild animals Habitat Mapping

3. Wild Animals Habitat Mapping

3.1 Methodology

3.1.1 Data Set Input data on environmental variables were obtained from different sources and some are derived from existing data sets. Surface data including DEM from data centers of NASA and USGS and data on infrastructure were collected from local governmental institutions in geo- tiff and shape formats. Further, historical data on recorded wild animal density and other available information were collected from the two Park. Information on habitat requirements for selected species of the area was collected from different documents and in consultation with wild animal experts.

3.1.2 Habitat assessment and mapping

Field work was made between end of November and December 2014. Based on the preliminary land cover map and in consultation with wildlife experts and other key informants, data and information on wild animal habitats within the park were collected. These include data on habitat use, seasonality, and routes frequented. Further, location of animal watering places, frequent wild animal-vehicle collision areas and man-made features were registered. On areas which lacks vegetation information, survey of vegetation classes was also carried out in four transects and a total of 22 plots at ANP. Information gathered and observations made were registered in formats developed and in reference with registered points (GPS location). Along with the trend analysis, sampled HH were also interviewed in the historic type and abundance of wild-animals along with their personal views for the incline / decline observed. Habitat maps were produced through the analysis of generated and derived data sets in GIS and with the integration of ancillary data and information.

ArcGIS 9.2 was used to manage and manipulate geographic data including DEM images. The software was also used to build and run habitat model. Garmin 12 GPS was used to register geographical location during inventory of existing land units and collection of other spatial information. Binoculars and digital cameras were used in the spotting and taking of pictures of wild-animals.

ERCAND Consult Page 13 Section 3: Wild animals Habitat Mapping

3.2 Results and Discussions

3.2.1 Bio-Diversity of Awash National Park Wild animals like, Klipspringer (Oreotragus oreotragus), Salt’s dik dik (Madoqua saltiana), mountain Reed buck (Redunca fulvorufula), Jackal, and Crested Porcupine (Hystrix cristata) inhabit the bush land and thickets (Plate 7). Caracal (Felis caracal) favors the thicket as well the grassland area, especially toward mid northern part of the park. In the wet season species like Lesser kudu (Tragelaphus imberbis) were mainly localized to areas where the units adjoin water bodies, such

as west of Gotu camp. The geographical Plate 7: Dik-Dik at ANP distribution of these two unit ranges from

the side and foot slope of Fentale mountain to the undulating and flat lands at the center and far eastern parts of the park. Tree species such as Acacia senegal, A. tortolis, A. mellifera, A. nilotica, Dobera glabra, Grewia spp, Balanites spp., Celtis spp., Ziziphys spp., and Grewia spp are found in these units.

The grass land, the grassland with trees, and grassland with shrubs (Plate 8) are the main habitat of Beisa Oryx (Oryx beisa), Sommerring’s gazelle (Gazella soemmerringi), and warthog (Phacochoerus aethiopicus). These species, during the dry season, visit the thickets for food (browsing). Swayne’s hartebeest (Alcelaphus buselaphus

swaynei) and Ostrich (Struthio camelus) Plate 8: Soemmering’s gazelle at ANP once also exclusively inhabit these units. Warthog (Phacochoerus aethiopicus), is

ERCAND Consult Page 14 Section 3: Wild animals Habitat Mapping

mostly confined to peripheries of the unit adjoining the water areas during the dry season, however, during the wet season didn’t show particular affinity towards any.

The grass land, marshy area and the adjoining cliff nearby the hot spring provide habitat for Rock hyrax (Procavia capensis) and Hamadryas baboons (Papio hamadryas); range for the latter extends as far to south-eastern parts of the park (Plate 9). Warthog (Phacochoerus aethiopicus) and

Waterbuck (Kobus ellipsiprymnus) are Plate 9: Hamadryas Baboons at ANP also found in the dry season. (Panthera leo) and (Panthera pardus) are mostly found in the mountainous areas on the central part of the park and visits almost all identified units.

Many bird species are found distributed in the park land (Plate 10). The riverine forest and the marshy areas around the hot spring were some of the places where numerous bird species are observed.

Plate 10: Birds at ANP

ERCAND Consult Page 15 Section 3: Wild animals Habitat Mapping

The riverine forest is mainly composed of tree and shrub species like Acacia etbaica, Celtis spp, Ficus species, Grewia Bicolor, Tamarindus indica, Trichilia spp., B. discolor Acacia Senegal, Ziziphus muro, and Z. mucronata. The unit provides habitat for Colobus monkey (Colobus guereza), Grivet monkey (Chlorocebus aethiops) and Anubis baboons (Papio anubis), among others.

The adjoining water-bodies are also habitats for species like crocodile and (Hippopotamus amphibius). Waterbuck (Kobus ellipsiprymnus) was also found along awash river but at present mainly near marshy lands at hot spring area; during the wet season, when human and domestic animals interference is at its low, it can spent most of its time in the grassland, otherwise feed during the night and remain in the adjoining thicket during the day.

3.2.2 Trends in Number of Wild-animal Population

Available census data indicate that the number of Beisa Oryx (Oryx beisa) and Soemmering’s gazelle (Gazella soemmerringi) population has declined between 2008 and 2010 at ANP (Figure 5). The number of Salt’s dik dik (Madoqua saltiana), (Phacochoerus aethiopicus) and spotted hyena (Crocuta crocuta), however, had shown an increment of approximately 200 each between 2008 and 2010. The present distribution of some wild animals in the national park is shown in Figure 6.

Figure 5: Change in number of species for some wild animals occurring in ANP between 2008 and 2010

ERCAND Consult Page 16 Section 3: Wild animals Habitat Mapping

Figure 6: Map depicting the present distributions of some wild animals at ANP

ERCAND Consult Page 17 Section 3: Wild animals Habitat Mapping

3.2.3 Habitat Fragmentation IPCC (2002) had emphasized that changes in ecosystem and attendant loss of biodiversity are caused by land use cover changes and habitat fragmentation, among others. Further, Fischlin et. al. (2007) have stated that the present and future land-use change and associated landscape fragmentation are very likely to impede species’ migration and thus impair natural adaptation via geographical range shifts. Studies also show that (Barbara et.al, 2009) fragmentation leads some species to become more abundant while make others decline to local or regional extinction.

In the past 40 years, the vegetation units in Awash National Park have undergone changes in area size and distribution patterns of vegetation unit. The major identified anthropogenic causes of fragmentation are described below.

3.2.3.1 Deforestation and Degradation Overgrazing by domestic animals, collection of wood and expansion of farmlands were the main factors causing deforestation and degradation in the national park.

3.2.3.1.1 Expansion of Cultivated Land In 2002, WBISPP had also carried out an indicative study on the population support capacity of rain-fed cropland in order to determine the capacity of the land to support the current and projected populations, that is, the degree of population pressure. Based on this study the potential limit for Fentale woreda has been much exceeded and was categorized as critical (Table 2).

Table 2: population support capacity analysis for Fentale woreda (source: WBISPP)

current/ yr 10/ yr 20/ Total rural Potential total ZONE wereda potential potential potential Category population population (%) (%) (%)

East Shewa Fantale 56,768 6,506 873% 1206% 1666% Critical

After the re-demarcation of Awash National Park boundary in 2012, parts of the left out areas was converted to agricultural plots.

ERCAND Consult Page 18 Section 3: Wild animals Habitat Mapping

3.2.3.1.2 Competition from Grazing by Domestic Animals

During the rainy season domestic animals graze around homesteads. However, starting from the onset of the dry periods, between mid-October and end of May, herders, from the surrounding areas of ANP most often send their cattle, goats and sheep to other areas, including the park, for grazing (Plate 11). Most frequented areas of grazing in the national park include Mt. Fentale and its surrounding, the mid eastern, and southern (including Ilala Sala plane) parts of the national park. Plate 11: Goat and sheep herds nearby Ilala-Sala plain at ANP

3.2.3.1.3 Collection of Fuel and construction wood

No data on the amount, season and intensity of collection of fuel and construction wood was available. However, considering that 100% of the surveyed HHs use biomass as the main source of cooking and heating and noting that in the Ethiopian context the average estimated annual consumption is 560kg, it can be inferred that there is high pressure on the national park.

3.2.3.2 Infrastructure development

Addis-Djibouti Rail way

Since the early 1900’s, the rail way linking Addis and Djibouti town passes between the mid- eastern and south-western part of the Awash national park for approximately 26 Km, and was not functional for the past decade or so. Nevertheless, a new railway is under construction (Plate 12), and unlike the previous old railway, it shall support modern and faster ones. The route shall also probably be fenced throughout its passage with mesh-wire to avoid collision. In line with this, park authorities have asked for the construction of underpasses / overpasses at different points along the route to allow movement of wild animals.

ERCAND Consult Page 19 Section 3: Wild animals Habitat Mapping

Plate 12: Part of Addis-Djibouti Rail-way under construction at ANP

One big concern is related with time of adaptation of some species with the proposed over and/or underpasses. The habitat of Oryx, for example, are found in both sides of the railway; it frequents the southern side most often during the dry season but also moves to the northern side at the onset of the rainy season and when the number of domestic animals decrease. Oryx stays there mostly till the end of the wet season, by which time availability of water reduced and number of domestic animals rise, then returning to the southern parts.

Settlements

Awash town was expanding at a faster rate towards the boundary lines of ANP. Associated with it, human interference, and as witnessed in most other towns, disposal of urban waste would be one of the major problems that would be faced in the near future. Further, intensity of fuel wood collection and intrusion into the park area would be more aggravated.

3.2.3.3 Impacts of Fragmentation

Based on collected HH data, the numbers of baboons in the national park and adjoining KAs have increased while that of most other animals had decreased, especially in the past 5 to 15 years. One of the main habitats for species such as baboons is peripheries of wood and bush lands.

ERCAND Consult Page 20 Section 3: Wild animals Habitat Mapping

3.2.4 Other Pressures

3.2.4.1 Road Kill Available records show that there were a number of road kill at ANP in the past years. A number of wild animals, such as Oryx beissa, Lesser kudu (Tragelaphus imberbis), Hamadryas baboon (Papio hamadryas), Hyena (Crocuta crocuta), Civet (Vivera civetta), as well as many baboons and big birds were killed on the road (Figure 7).

In 2013/14 some eight Oryx and three lesser kudu were killed among others. The figures, however, were much underestimated, as it was based only on data registered at the traffic office. Further, it was believed that the killing of a single wild animal can bring more collision, especially

carnivores. Figure 7: Frequent road kill areas by species type at ANP

Some of the Oryx were killed on the flatlands while most others at a junction towards Awash town and where the inclination starts. Other animals were killed on areas which have gentle to moderate slope (Figure 8). Vehicles speed and fragmentation of habitats could be the major causes for the disaster. Figure 8: Cross section view of road crossing ANP During the first week of the field work at ANP, humps were constructed in some six places on part of the Addis – highway that crosses the park, to lower the speed of vehicles.

ERCAND Consult Page 21 Section 3: Wild animals Habitat Mapping

3.2.4.2 Alien Invasive Species

The invasive tree species, Prosopis julliflora, which is widely spread in the adjoining Afar plains, was slowly penetrating to the ANP from the NE and Eastern part. The species is known for its aggressive expansion and invading characteristics. Different studies show that and interviewed HH are evidence for, one way in which the introduction and spread of the species is associated with the movements of animal herds from one area to another.

ERCAND Consult Page 22 Section 4: Climate Change Risk Profiles

4. Climate Change Risk Profiles

4.1 Objective Specific objective of this study was to create climate change risk profiles that cover all the major hazards prevailing in and around the Awash National Park.

4.2 Methodology

4.2.1 Data Set Ten years climatic data from meteorological stations at Methara (ANP) was collected from the National Meteorological Service Agency (NMSA). The basic parameters that were considered in the analysis are monthly rainfall and monthly temperature.

Surface and atmospheric data files including archive fire from FIRMS1, surface topography such as digital elevation model (DEM) and vegetation indices are acquired in geo-tiff and shape formats from data centers of NASA and USGS. Climatic data were also acquired from FAO. Classified land use and land cover map and other derived data are also used in the analysis.

Historical data on Awash river discharge at different locations and Beseka water level data were obtained from the Hydrology department of Ministry of Water and Energy. Perennial and intermittent river features and surface water bodies found within the study areas were extracted and digitized from topographic maps.

4.2.2 Hazard Assessment Data and information on hazards that have occurred in and around the national park and its impact were collected from archived records obtained from Woreda Bureau of Agriculture, in discussion with pertinent officials and key informants, household level survey and direct observation.

1 MCD14ML: This data set was produced by the University of Maryland and provided by NASA FIRMS operated by NASA/GSFC/ESDIS with funding provided by NASA/HQ Available on-line https://earthdata.nasa.gov/active-fire-data#tab- content-6

ERCAND Consult Page 23 Section 4: Climate Change Risk Profile

4.2.2.1 Fire hazard Assessment

Distribution of archived fire data were analyzed and described in terms of spatial distribution and frequency. FIRMS Fire hot-spots are acquired by the MODIS instruments, on board NASA’s Terra and Aqua Earth Observing System (EOS) satellites. A MODIS active fire detection represents the center of an approximately 1km pixel flagged as containing one or more actively burning hotspots/fires. Fire detection is performed using a contextual algorithm that exploits the strong emission of mid-infrared radiation from fires. The algorithm examines each pixel of the MODIS swath, and ultimately assigns to each one of the following classes: missing data, cloud, water, non-fire, fire, or unknown (EOS-DIS, 2009).

Input data were used to estimate the fire prone areas at the study sites. The present land use and land cover map, as part of it include the vegetation cover, data on slope, aspect, elevation, proximity to road and to settlements were the variables selected as an input to the model.

Initially, each selected variable were reclassified and categorized between classes. Class categorization of each variable was made between the ranges of very high to very low (Figure 9).

ERCAND Consult Page 24 Section 4: Climate Change Risk Profile

Figure 9: Maps of weighted variables

ERCAND Consult Page 25 Section 4: Climate Change Risk Profile

Further, environmental and anthropogenic factors that strongly influence the ignition and spread of fire in the PA are selected and parameterized with the relative degree of influence each has on the fire danger (Table 3). Variables weighted include the present land cover classification, slope, aspect, distance from roads, distance from settlements and elevations.

Table 3: Variables and weight factor used in fire hazard mapping Variable Assigned weight Land cover 66 Slope 20 Aspect 7 Distance from Road 3 Distance from Settlements and towns 3 Elevation 1 Total 100

A model widely used in the determination of fire hazard was used to delineate areas susceptible to fire.

4.2.2.2 Drought hazard Assessment In order to determine the impact of drought on wild animals, assessment of NDVI values between a known drought (2003) and non-drought period (1995) was made to determine the vegetation unit most affected by the drought.

Vegetation indices were generated from Landsat TM images. The Normalized Difference Vegetation Index (NDVI) was computed using:

NIR RED NDVI = NIR + RED − Where NIR is reflectance in the near infrared (band 4 of landsat TM) and Red is reflectance in the red wavelengths (band 3 of landsat TM).

ERCAND Consult Page 26 Section 4: Climate Change Risk Profile

4.2.2.3 Flood hazard Assessment Derived and acquired data sets including slope (derived from digital elevation model), drainage density, soil type, present land use and land cover and elevation data were used as input in flood hazard assessment. The main flooding factors were reclassified into groups and ranked based on significance / influence on the hazard (Table 4).

Table 4: Factors used in Flood Hazard Assessment Factor Sub-factor Rank Description 27.109-79.120 5 Very high 16.496-27.109 4 High Slope 8.966-16.496 3 Moderate 4.033-8.966 2 Low 0-4.03 1 Very low

0.365-0.568 5 Very high 0.249-0.365 4 High Drainage density 0.149-0.249 3 Moderate 0.047-0.149 2 Low 0-0.047 1 Very low

Vertic cambisols 5 Very high Lithic leptosols 4 High Soil type Eutric fluvisols 3 Moderate Eutric cambisols 2 Low Eutric leptisols 1 Very low

Exposed surface 4 V. High Cultivated land 3 High Land use and land cover type Grassland 2 Moderate Thicket/bush land 1 Low

760-920 5 Very high 920-1080 4 High Elevation 1080-1320 3 Moderate 1320-1600 2 Low 1600-1960 1 Very low

The standardized map layers were then assigned weights based on their relative importance and normalized by the sum of weights. Accordingly, weights of 33.33, 26.67, 20.00, 13.33

ERCAND Consult Page 27 Section 4: Climate Change Risk Profile

and 6.67 were assigned for slope, drainage density, soil type, land use / cover type and elevation respectively. GIS based weighted linear combination was used to derive flood hazard areas.

Furthermore, Awash river discharge and Beseka lake level data were processed using Log Pearson Type III analysis.

4.2.3 Risk Assessment

Based on integration of derived data on hazard and information on the elements at risk, assessment of risk was made using weighted overlay. Ranking was made based on relative importance attached and categorized from very high to very low. Available data on previous impacts were also reviewed.

4.3 Results and Discussions

4.3.1 Climate

4.3.1.1 The Present Climate of ANP

The total mean annual rainfall for Methara, nearby ANP, is about 496 mm. The area receives 60% of the total rain fall between July and September and some 25% between March and May (Figure 10).

Figure 10: Mean Monthly Rainfall: Methara.

ERCAND Consult Page 28 Section 4: Climate Change Risk Profile

The mean maximum and mean minimum temperature over Methara, 944 m.a.s.l, is 34.0 and 17.8OC, respectively. In December, the mean maximum and mean minimum temperatures are at the lowest and in June at their highest (Figure 11).

.

Figure 11: Mean Monthly Temperature: Methara.

The spatial pattern of temperature in ANP (Figure 12) shows that, except the northern and mid western (Fentale Mountain) parts, most other parts have temperature in between 20 and 22.5oC.

Figure 12: Thermal zones

ERCAND Consult Page 29 Section 4: Climate Change Risk Profile

At national level, a study (NMSA, 2005) showed that there has been a warming trend in the annual minimum temperature over the past 50 years and an increase by about 0.370C in every ten years. However, even though there were fluctuations, rainfall for the same period had remained more or less constant when averaged over the whole country.

4.3.1.2 The Future Climate

A study by NMSA (2005) also showed that, for the IPCC mid-range (A1B) emission scenario, the mean annual temperature will increase in the range of 0.9 -1.1°C by 2030, in the range of 1.7 - 2.1°C by 2050 and in the range of 2.7-3.4°C by 2080 over Ethiopia compared to the 1961-1990 normal. A small increase in annual precipitation is also expected over the country.

4.3.2 Hazards Assessment

4.3.2.1 Fire Hazard

4.3.2.1.1 Fire Frequency

In between 2001 and 2013 there had been 26 fire incidences at Awash national park. Of these, seven each had been recorded in year 2012 and 2013. Furthermore, 54% of the hot spots observed between 2001 and 2013 had occurred in just two years, 2012 and 2013 (Figure 13).

Figure 13: Number of fires observed by year at Awash National Park

ERCAND Consult Page 30 Section 4: Climate Change Risk Profile

The seasonal distribution of fire incidence at ANP (Figure 14) shows that most fire incidences had occurred between early October and end of December, after which it shows sharp decline in the month of January. Afterwards, it starts to rise again, rather slowly, and reach its second peak in the month of May.

Figure 14: Number of fires observed by month at Awash National Park

Including the surrounding areas, February, May and December are the months in which fire occurrences were most observed (Fig. 15). This implies that most fires had occurred at the end of the dry season and a month or two before the onset of the rainy season.

Figure 15: Number of fires observed at Awash National Park and surrounding areas by month

Considering the adjoining KAs, the number of fires recorded is proportionally very high at the vicinity of Merti sugar Factory (Figure 16), where frequent firing of sugar cane fields

ERCAND Consult Page 31 Section 4: Climate Change Risk Profile

were practiced at different times annually. Merti-sugar, adjoins the national park in the south western side, in between, there is Awash river.

Figure 16: Distribution of fire points in the KAs surrounding ANP

Still, year 2012, 2013 and 2014 were the years in which most fire occurrences had been observed (Figure 17)

Figure 17: Number of fires observed per year at Awash National Park and surrounding areas

The spatial distribution of fires (Figure 18) shows that there were accumulations of fire points at three areas; in far north, mid-eastern and central parts.

ERCAND Consult Page 32 Section 4: Climate Change Risk Profile

Figure 18: Location of Fire pixels in ANP and the surrounding areas

Within the national park, the main causes of fire are believed to be careless throwing of cigarette butts from passing vehicles and passenger trains as well as from car exhaust. Overlaying the road network, the argument can easily be justified, as some of the fire pixels are observed near it (Figure 18) and probably explaining the causes for the central part. In the far northern part, unlike Merti-sugar, at southwest, there is no major river that can potential hamper the propagation of fire from kesem irrigated areas.

Based on HH interviews conducted in adjoining KAs under , the most probable causes of fire in the mid-eastern part of the park could be charcoal production.

The main road is now made much wider (since 1994 / 95) and the old rail way has stopped functioning (since some 5 to 8 years) and a new one is under construction.

ERCAND Consult Page 33 Section 4: Climate Change Risk Profile

4.3.2.1.2. Fire Prone Areas

Fire hazard areas identified at ANP (Figure 19) shows that most of the high fire hazard areas are located at and near Fentale Mountain. The high fire hazard area covers 17.20% of ANP. Almost half of ANP area is categorized under moderate (47.08%), 30.60% as low and some 5.00% as very low to none fire hazard areas.

Figure 19: ANP - Fire Hazard Map

4.3.2.2 Drought Hazard Comparison of NDVI values between a known drought (2003) and non-drought period (2008) show that there were changes in the NDVI values over the national park.

ERCAND Consult Page 34 Section 4: Climate Change Risk Profile

The mean NDVI values for year 2003 images (drought year) was -0.064733 with a standard deviation of 0.253738 while that of year 2008 was -0.055139 with a standard deviation of 0.238072 (Figure 20)

Figure 20: Comparison of NDVI images over ANP (North eastern part of image) and surrounding areas for 2003 (A) and 2008 (B)

4.3.2.3 Flood Hazard

4.3.2.3.1 Flood Prone Areas

3.91 percent of ANP area was categorized as very high flood prone area. 20.34 , 33.56, 29.47 and 12.72 percent of the PA were also assigned as high, moderate, low and very low hazard areas respectively. Very high hazard areas are located in the north and north western part of the national park (Figure 21).

ERCAND Consult Page 35 Section 4: Climate Change Risk Profile

Figure 21: Flood Prone Areas

4.3.2.3.2 Flood Frequency

Based on a 28 years instantaneous daily flow data for Methara (gauging station 032003) estimates of discharges associated with different return periods have been calculated to infer the probability of future Awash river water level.

The mean annual flow of Awash river, at Methara gauging station, in 1983, 1993, 1999 and 2007 had exceeded 40m3/s (Figure 22).

ERCAND Consult Page 36 Section 4: Climate Change Risk Profile

Figure 22: Mean Annual Flow of Awash River at Methara

It should be noted that, most of the flow from the upper awash basin are regulated at Koka reservoir, located some 110 km south-west of the Methara gauging station, with proper notification on the release. However, there exist a proportional area of catchment between the reservoir and the gauging station at Methara that contribute to the observed flow.

Figure 23: Monthly Flow of Awash River at Methara gauging station (1982-2009)

The monthly flow of Awash river shows high variation between August and September (Figure 23). Accordingly, high water flow that can potentially overtop the natural levee of the river could be expected during this time period. Within ANP, river flooding occurs in the SW part of the river path, before cascading into the gorge, and affects the immediate surroundings of its path.

ERCAND Consult Page 37 Section 4: Climate Change Risk Profile

Flood frequency, using Log Pearson Type III analysis for Awash river at Methara station is shown in Figure 24.

Figure 24: Log Pearson Type III Analysis (using average daily maximum flow values) for Awash River at Methara Station

4.3.2.3.3 Beseka Lake Level

Ten years (1997- 2009) daily water level data records for Beseka lake, located some 11 kilometres south east of the ANP, shows that the water level is continuously on the rise (Fig 25). Between March, 2009 and March 2013, Beseka lake has shown a 0.85 meters rise.

Figure 25: Water level of Beseka lake between 2009 and 2013

ERCAND Consult Page 38 Section 4: Climate Change Risk Profile

The expansion of the lake, based on satellite images acquired (Plate 13) shows a substantial increase in area size between early 1970’s and 2014.

Plate 13: Comparison of images over Beseka Lake (from left to right: 1973, 1986, 1998 and 2014)

Studies (Girma et.al., 2005) show that major causes of degradation were accelerated soil erosion by wind and water, increasing salinization of soils and near-surface groundwater supplies and a reduction in soil moisture retention. Same study further note that, in turn these has led to an increase in surface runoff and stream flow variability, a reduction in species diversity and plant biomass, and a reduction in the overall productivity in dry land ecosystems with an attendant impoverishment of the communities dependent on these ecosystems.

Plate 14: Northern part of Lake Beseka (at the Background is Fentale Mountain)

ERCAND Consult Page 39 Section 4: Climate Change Risk Profile

4.3.3 Vulnerability Assessment

Hazard affects wild animals mainly through habitat modification (Kennedy and Fontaine, 2009). Even though, the extent, size, life form and combination of vegetation types required by wild animals differ from one species to another, in general, habitat provide shelter, food and water, and space in which the survival and continuity of wild animals depend (Yarrow G., 2009).

Awash national park is also surrounded by pastoralist and agro-pastoralist communities, whose livelihood depend on the range and agricultural lands. Number of human and livestock in the adjoining KAs were given in Annex 2.

4.3.4 Risk Assessment An occurrence of hazard within the park might potentially affect the size and species composition of habitat and the fauna that depend on it. Further, it will also affect the livelihood of local community who mainly graze their livestock on the range lands. Disruption of tourism may also decrease revenue and income derived by business operators in the area.

4.3.4.1 Fire Risk 64.28% of ANP area was found in high fire risk areas and 30.60% under moderate fire risk areas, together these constitutes almost 95%. The remaining 5% fall under low and very low fire risk areas. An estimated 90% of land cover units whose major component is the grassland, namely, the grassland, grass with trees and grass with shrub, fall under high fire risk area. Likewise, about 24% and 73% of thicket and bush lands are found in areas categorized as high and moderate fire risk areas respectively (Figure 26).

ERCAND Consult Page 40 Section 4: Climate Change Risk Profile

Figure 26: Fire Risk Map of ANP

4.3.4.2 Drought Risk There are no sufficient census data to determine the impact of drought on wild animals, nevertheless, most farmers, based on frequency of visit at their locality, are of the opinion that the 2003 drought has a strong influence especially on Warthog.

In year 2002/2003, about 9,372 households holding 56,313 family members residing in Fentale district were affected with drought; translated to 81.21% of the total district population. With this drought, it was estimated that about 62,234 cattle, 19,466 sheep and 11,908 goats were dead (BoFED, 2011).

ERCAND Consult Page 41 Section 4: Climate Change Risk Profile

4.3.4.3 Flood Risk Of the estimated 598 km2 total area of ANP, 14.6% is susceptible to flooding. Of these, 3.27 percent is categorized as having a very high risk of flooding and 17.17% as a high flood risk area. The remaining 33.61%, 30.76% and 15.18% are delineated under moderate, low and very low flood risk areas (Figure 27).

3.25 percent of land cover classes in which the major component is the grassland, namely the grassland, grassland with trees and grassland with shrubs units, fall under very high flood risk areas, while 17.26, 29.10, 35.56 and 14.84 percent were under high, moderate, low and very low flood risk areas respectively. Similarly, 3.46 % of the thicket and bush lands fall under very high flood risk area whereas 16.13%, 38.95%, 25.35% and 16.12% of these units are found in the high, moderate, low and very low flood risk areas respectively.

Figure 27: Flood Risk map of ANP

ERCAND Consult Page 42 Section 4: Climate Change Risk Profile

Flooding affects the local community surrounding the PA. Flooding from Awash River retreats within few hours and deposit sediments. Areas mostly affected by flooding are localized on the plane areas along the course of the river. Considering the nutrient level of the sediment deposited and effect on productivity, unless otherwise it occurred untimely and devastates cropped lands, farmer’s perception towards Awash river flooding is not negative as that for Beseka lake level rise. The latter, which is saline and showed no retreat in months have rendered farmlands productivity low and forced abandonment of most nearby agricultural lands and residential quarters.

ERCAND Consult Page 43 Section 5: Conclusions and Recommendations

5. Conclusions and Recommendations 5.1. Conclusions

Identified land cover classes at ANP were subjected to changes of different scale in the past forty years. Among these, the most significant one was the decline in the area size of grassland. The major causes for the decline in area size differ in the time period studied. Significant ones include the deforestation and degradation of the vegetation through extraction of wood, overgrazing and expansion of cultivation.

Time-series comparison of census data revealed that there was a decline in the number of Oryx beisa. Changes in the observed land use and cover classes might have significant impact on the composition and abundance of wild animals in the national park.

Studies also reveal that the national park was exposed to climatic variations, especially related with temperature. Moreover, risk analysis shows that the national park and surrounding areas was exposed to fire, drought and flood hazard of different degrees. Some 64.28% of ANP were exposed to high fire risks and some 3.27 percent to very high risk of flooding.

The national park possesses significant number of wild animals, including mammals and birds that are endemic to Ethiopia. Beyond the ecosystem goods and services, the park has the potential to generate a substantial amount of economic and social benefit from the tourism industry such as livelihood diversification and job creation. These will help contribute to the government plan for poverty reduction.

ERCAND Consult Page 44 Section 5: Conclusions and Recommendations

5.2. Recommendations

In order to support the ongoing efforts toward the conservation and overall development of the protected area in general and resilience to climate change in particular the followings are recommended:

• Developing monitoring system, deployment of the required material and training of personnel is required to minimize risk and, on the occurrence of one, for an effective response to any identified hazard. • There is a need to collate the existing documents on traits of wild animals found in the country, if not available, to conduct brainstorming workshops in which wild animal experts would provide the scientific basis that help determine the interaction in face of climate change. • There probably exists a wide range of spatial and non-spatial data related to the bio- physical and environmental characteristics of wild animals, park, etc. in the country. However, such data, at least, are not available at park levels. It is therefore, very important and timely to develop a geo-data base and migrate all available maps into this for a better retrieval, storage and analysis (can be linked with the existing web page of EWCA, access if necessary, can be with permission). • Studies (Girma et.al.) show that the quality of Awash river had underwent considerable changes in the past; some parts of it are also not frequented by herders; it is thus timely to test the water quality of Awash River at different location and salinity level of near-surface water to consider alternatives based on results. • Sustainable development can be achieved when local communities are in direct benefit from the ANP. Therefore, as suggested by NOPA (1992), there is a need to establish park revenue sharing schemes that can, based on prioritized needs of local community, be directed toward animal health and feed, supplementing current efforts underway by the Fentale woreda office of Pastoralist and Agro-pastoralists in nearby KAs of Oromia region and improved water supply. • It is imperative to conduct carrying capacity assessment of the grassland, composition of the herbaceous layer; including seasonal differences in dry mass.

ERCAND Consult Page 45 Section 5: Conclusions and Recommendations

• The location, size and probably length of under and over passes, currently under construction should be reconciled in consideration of the biological traits of wild animals found in the park and that can potentially be re-introduced into the park. • Whatever possible interventions made to stop the encroachment and overgrazing of domestic animals shall minimize the risk of invasive species. • Installment of road signs, at proper places nearby the newly constructed humps, is required to avoid breakdown of passing vehicles. • Anthropogenic impacts poses difficulties in the management of the wild animals found at the national park. Nevertheless, data to accurately quantify such activities are meager. It would be advisable to collect data on factors such as wild animal dispersal areas, number and distribution of domestic animals and season of use, as well as fuel collections and use of adjoining kebeles and production and sale of charcoal.

ERCAND Consult Page 46 References

6. References

Barbara Wilson, Leonie Valentine, Janine Kinloch, Tracy Sonneman, Marnie Swinburn, 2009. Habitat Loss and Fragmentation IN: Biodiversity values and threatening processes of the Gnangara; Editors: Barbara A. Wilson and Leonie E. Valentine BoFED, 2011. Socio-economic profile of Fentale woreda. Chuvieco, E. and Congalton, R. 1989. Application of Remote Sensing and Geographic Information Systems to Forest Fire Hazard Mapping. Remote Sensing and Environment, 29: 147-159. Engstrom, R.T. 2010. First-order fire effects on animals: review and recommendations. Fire Ecology 6(1): 115-130. doi: 10.4996/fireecology.0601115 EOS-DIS, 2009. Active Fire Data. Retrieved on March 06, 2015 from EOSDIS Earth Data website. FAO/LUPRD, 1982. Soil association map of Ethiopia, Technical document 6. FDRE, 2005. wildlife strategy. Addis Abeba Fischlin, A., G.F. Midgley, J.T. Price, R. Leemans, B. Gopal, C. Turley, M.D.A. Rounsevell, O.P. Dube, J. Tarazona, A.A. Velichko, 2007: Ecosystems, their properties, goods, and services. Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, M.L. Parry, O.F. Canziani, J.P. Palutikof, P.J. van der Linden and C.E. Hanson, Eds., Cambridge University Press, Cambridge, 211-272. Girma Taddese, Kai Sonder and Don Peden, 2005. The Water of the Awash River Basin A Future Challenge to Ethiopia, ILRI. Greg Yarrow, 2009. Habit Requirements of Wildlife: Food, Water, Cover and Space. Cooperation Extension’s Forestry and Natural Resources Team, Clemson University. Institute of Biodiversity Conservation, 2009. Convention on Biological Diversity, Ethiopia’s 4th Country Report, Addis Ababa. IPCC, 2002. Climate change and biodiversity. Technical paper IV. IUCN SSC Antelope Specialist Group 2008. Oryx beisa. The IUCN Red List of Threatened Species. Version 2014.3. . Downloaded on 26 January 2015. MoARD, 2005. Wildlife Development, Protection and Utilization Policy and Strategy, Addis Ababa.

ERCAND Consult Page 47 References

NMSA, 2007. Climate Change National Adaptation Programme of Action (NAPA) of Ethiopia, Editor: Abebe Tadege "Preparation of National Adaptation Programme of Action for Ethiopia”, GEF / UNDP. Addis Ababa, Ethiopia. NOPA, 1992. Pastoralists at a cross road: Survival and development issues in African Pastoralism. Kennedy P. L. and Fontaine B. J., 2009. Synthesis of Knowledge on the Effects of Fire and Fire Surrogates on Wildlife in U.S. Dry Forests, Special Report 1096. Extension and Experiment Station Communications, Oregon State University. PHE, 2012. Three years Strategic plan of PHE Ethiopia Consortium, 2011 – 2013. Addis Ababa. WBISPP, 2002. A strategic plan for the sustainable development, conservation, and management of the woody biomass resources, Volume 5, Methodology: Land Use Systems Analysis. Yirmed and Girum, 2013. Birds of Awash National Park, Revised Check list.

ERCAND Consult Page 48 Annex

Annex 1: List of wild animals at Awash National Park listed in the IUCN Red list Estimated Common name Scientific name IUCN status Total no. Park

(2010) ANP 1 Oryx Beisa oryx Near-threatened 410

2 Lesser kudu Tragelaphus imberbis Near-threatened 1208

3 Tragelaphus strepsiceros Least concern 35

ERCAND Consult Page 49 Annex

Annex 2: Number of domestic animals at KAs surrounding Awash National Park

ERCAND Consult Page 50 Annex

Annex 3: Comparison of No. of Livestock in Fentale Woreda between 2008/09 and 2013/14k

ERCAND Consult Page 51