ADDIS ABABA UNIVERSITY SCHOOL of GRADUATE STUDIES Environmental Science Program

ADDIS ABABA UNIVERSITY SCHOOL OF GRADUATE STUDIES Environmental Science Program

Impact of ‘Katikala’ Production on the Degradation of Woodland Vegetation and Emission of CO and PM during Distillation in Arsi-Negele Woreda, Central Rift Valley of Ethiopia

By Nejibe Mohammed

A Thesis Submitted to the School of Graduate Studies of Addis Ababa University in Partial Fulfillment of the Requirements for the Degree of Master of Science in Environmental science

February, 2008

Environmental Impact of ‘Katikala’ Production in Arsi-Negele Woreda, Central Rift Valley of Ethiopia

A Thesis Submitted to the School Of Graduate Studies of Addis Ababa University in Partial Fulfillment of the Requirements for the Degree of Master of Science in Environmental Science

Nejibe Mohammed

Addis Ababa University

Faculty of Science, Environmental Science Program

February 2008 ACKNOWLEDGMENT

First of all I am deeply indebted to my advisors: Dr Mekuria Argaw and Dr. Seyoum Leta for their diligent and Constructive comments, guidance and flexible supervision and continuous follow up.

I wish to thank Addis Ababa University for giving me the chance to join the graduate studies of environmental science program. I need to also express my genuine thanks to horn of Africa regional environment program for providing me with financial support special thanks goes to Ato Negussu Aklilu, Forum for Environment (FfE) for his cooperation.

I would also like to thank Ato Dekebo Dale, Arsi Negele Concern for Environment and Development Association (ANCEDA) for his continuous cooperation in providing me with valuable information during the field work.

I am particularly indebted to GTZ – SUN Energy project who gave me logistic support for accomplishing this thesis. Special thanks go to Ato Anteneh Gulilat, Ato Samson Tolessa and Hiwote Teshome of GTZ – SUN Energy

Many thanks go to my family for being with me all the way right from the on set of the research work. Last but not the list I would like to thank all those who put a drop of contribution in any ways for the successful completion of my study.

i TABLE OF CONTENT

Page No.

ACKNOWLEDGMENT ...... i TABLE OF CONTENT...... ii LIST OF TABLE ...... iv LIST OF FIGURE ...... v ACRONYMS...... vi ABSTRACT...... vii 1. INTRODCTION ...... 1 1.1. GENERAL BACKGROUND ...... 1 1.2. STATEMENT OF THE PROBLEM...... 4 1.3. OBJECTIVES OF THE S.TUDY ...... 7 1.3.1. General Objective ...... 7 1.3.2. Specific Objective...... 7 2. REVIEW OF RELATED LITERATURE...... 8 2.1. FOREST RESOURE DEGRADATION...... 9 2.2. CONSEQUENCE OF WOODFUEL CONSUPTION IN THE ENVIRONMENT ...... 10 2.2.1. Impact of Woodfuel Consumption on Forest Ecosystems ...... 11 2.2.2. Fuelwood Consumption Pattern and Forest Degradation in Ethiopia ...... 12 2.3. IMPACT OF INDOOR AIR POLLUTION ...... 14 2.3.1. Indoor Air Pollution Due to Combustion and Its Health Effect in Developing Countries ...... 15 2.3.2. Major Air Pollutants Released From Biomass Combustion...... 17 2.3.3. The Gap in Pollutant Testing Methods...... 18 3. MATERIALS AND METHODS...... 21 3.1. THE STUDY AREA: ARSI-NEGELE WOREDA...... 21 3.1.1. Location and Description...... 21 3.1.2. Climate and Topography ...... 21 3.1.3. Geology and Soil ...... 22 3.1.4. Vegetation...... 23 3.1.5. Land Use and Agriculture...... 24 3.1.6. Population ...... 24 3.2.STUDY DESIGN ...... 25 3.2.1. Household survey ...... 25 3.2.2. Market survey ...... 25 3.2.3. Woody vegetation survey ...... 25 3.2.4. Indoor air pollution measuring ...... 27 3.3.DATA ANALYSIS...... 29 4. RESULTS AND DISCUSSION...... 31 4.1.FUELWOOD CONSUMPTION BY HOUSEHOLDS...... 31 4.1.1. Species Preferences by Household ...... 32 4.1.2. Most Frequently Used Species ...... 34 4.2. FUELWOOD SUPPLY AND SOURCE...... 36 4.3. IMPACT ON VEGETATION...... 39 4.3.1. Impact on different Species ...... 43 4.4.IMPACT ON INDOOR AIR QUALITY ...... 45 4.4.1. Average Pollutant Concentration...... 45

ii 4.4.2. Maximum Pollutant Concentration...... 49 4.4.3. The Relationship of Carbonmonoxide & Suspended Particulate Matter..... 53 4.4.4. Pollutant Concentration And Fuelwood Use Rate...... 55 5. CONCLSION AND RECOMMENDATION ...... 59 5.1.CONCLUSION...... 59 5.2.RECOMMENDATION...... 60

REFERENCE...... 61 ANNEX ...... 69 ANNEX 1: HOUSE HOLD SURVEY QUESTIONNAIRE...... 69 ANNEX 2: MARKET QUESTIONNAIRE ...... 72 ANNEX 3: MARKET SUPPLY COOUNTING FORMAT ...... 74 ANNEX 4: VEGETATION DATA COLLECTION FORMAT...... 75 ANNEX 5: HOBO CO SAMPLING DATA FORM...... 76 ANNEX 6: UCB PARTICLE MONITOR SAMPLING DATA FORM ...... 77 ANNEX 7: INDOOR AIR POLLUTIN: POST-MONITORING QUESTIONAIRE ... 78

iii LIST OF TABLE

TABLE 4.1 DAILY CONSUMPTION RATEKG/DAY ...... 31 TABLE 4.2 FUELWOOD SPECIES PREFERENCE OF CONSUMERS...... 32 TABLE 4.3 MOST FREQUENTLY USED FUELWOOD SPECIES BY HOUSEHOLDS ...... 35 TABLE 4.4 SOURSES OF FUELWOOD TO ARSI-NEGLE TOWN FROM DIFFERENT LOCATION ...... 37 TABLE 4.5 DAILY SUPPLY OF FUELWOOD TO ARSI-NEGELE TOWN (MEAN (+ SD...... 38 TABLE 4.6 DENSITY OF TREES (SS= SEEDLINGS+SAPLING, M= MATURE TREES) PER HECTARE WITH IN VILLAGES ...... 39 TABLE 4.7 SIZE CLASS DISTRIBUTION OF TREES COMPARISON OF SPECIES B/N THE PLOTS IN THE KEBELE AND THE PARK...... 40 TABLE 4.8 SIZE CLASS DISTRIBUTIONS OF TREES WITHIN SPECIES COMPARISON BETWEEN THE KEBELES AND PARK...... 42 TABLE 4.9 NUMBER, NUMBER PER HECTARE AND FREQUENCY OF TREES CUT IN THE VILLAGES...... 42 TABLE 4.10: WOOD FUEL USE RATE & ENVIRONMENTAL FACTORS ...... 45 TABLES 4.11 MEAN POLLUTANT CONCENTRATIONS DURING DISTILLATION PROCESS...... 46 TABLE 4.12 AVERAGE CO CONCENTRATION (PPM): MULTIPLE REGRESSION ESTIMATES...... 48 TABLE 4.13 AVERAGE PM CONCENTRATION (MG/M3): MULTIPLE REGRESSION ESTIMATES...... 48 TABLE 4.14 THE MAXIMUM 15-MINUTE AVERAGE POLLUTANT CONCENTRATIONS ...... 50 TABLE 4.15 MINUTE MAXIMUM CO CONCENTRATION (PPM): MULTIPLE REGRESSION ESTIMATES...... 51 TABLE 4.16 15-MINUTE MAXIMUM PM CONCENTRATION (MG/M3): MULTIPLE REGRESSION ESTIMATES ...... 52 TABLE 4.17 FUEL USE IMPACT ON INDOOR CO CONCENTRATIONS : BIVARIATE REGRESSION ESTIMATES ...... 56 TABLE 4.18 FUEL USE IMPACT ON INDOOR PM CONCENTRATIONS : BIVARIATE REGRESSION ESTIMATES ...... 57 TABLE 4.19 WOOD USE RATE: MULTIPLE REGRESSION ESTIMATES ...... 57

iv LIST OF FIGURE

FIGURE 2.1 LAND USE / LAND COVER CATEGORIES OF DIFFERENT YEARS OF ARSI-NEGELE WOREDA...... 20 FIGURE 3.1 LOCATION MAP OF ARSI NEGEL WEREDA ...... 22 FIGURE 3.2 FIELD VEGETATION SURVEY SAMPLE POINT...... 26 FIGURE 4.1 HISTOGRAM OF MEAN CONCENTRATION OF CO (PPM) DURING DISTILLATION PROCESS ...... 46 FIGURE 4.2 HISTOGRAM OF MEAN CONCENTRATION OF PM (MG/M3) DURING DISTILLATION PROCESS...... 47 FIGURE 4.3 HISTOGRAM OF MAXIMUM 15-MINUTE AVERAGE CO (PPM) CONCENTRATIONS...... 50 FIGURE 4.4 MAXIMUM 15-MINUTE AVERAGE PM (MG/M3) CONCENTRATIONS ...... 50 FIGURE 4.5 THE RELATIONSHIP BETWEEN CO (PPM) AND PM (MG/M3 ...... 54 FIGURE 4.6 CONCENTRATION OF CO AND SUSPENDED PM FOR A SINGLE DAY FOR A REPRESENTATIVE HOUSEHOLD ...... 55

v ACRONYMS

ADAA Africa Development Aid Association ANOVA Analysis of Variance ºC Degrees Celsius cm cent-meter CO Carbon Monoxide CO Carbon dioxide 2 CSA Central Statistical Agency EARO Ethiopia Agricultural Research Organization FAO Food and Agricultural Organization GPS Global Positioning System IAP Indoor Air Pollution IAQ Indoor Air Quality IPCC Inter-governmental Panel on Climate Change kg kilogram m meter m3 cubic meter mg milligram N number in sample ORS Oromia Region State ppm parts per million RVLB Rift Valley Lake Basin SPM Suspended Particulate Matter SPSS Statistical Package for Social Sciences μg/m3 microgram per cubic meter mg/m3 milligram per cubic meter UNDP United Nations Development Program UNEP United Nations Environmental Prpgramme SD Standard Deviation USEPA United States Environmental Protection Agency’s WHO World Health Organization of the United Nations WRI World Resources Institute

ABSTRACT

Arsi-Negele is located 225 km south of Addis Ababa, the capital of Ethiopia. Most of the residence of the town is engaged by production of ‘Katikala’ (alcoholic beverage produced by distillation process). The process involves series of tedious steps which requires more fuelwood energy and all of which are done by women. This system seems to work well for most families but is facing problems because of environmental concerns. The utilization of woodfuel coupled with use of low efficiency traditional stove has been contributing to forest resource degradation and prevalence of health problems due to indoor air pollution. In this study, impact on the surroundings vegetation and indoor air pollution due to katikala production was assessed.

Household and fuelwood market information was gathered by personal interviewing through structured questionnaire. Vegetation sample was taken at the Acacia-Balanites woodland to crosscheck the information gathered through the questionnaire with the impact on the vegetation around the town. Indoor air pollutants were measured during distillation process. The major pollutants quantified in this study are CO and PM by using HOBO CO data logger and UCB particle monitor, respectively.

It was found that 87.3 % of the respondents were engaged in ’katikala’ production. Woodfuel consumption rate in the town was equivalent to 76.98 kg /day for distillers and 17.82 kg /day for non- distillers. By most of the distillers A. etbaica was found to be the most preferred plant species followed by R. natalenis, where as dry wood materials and branches of tree was preferred by most non-distillers for woodfuel use. However, most frequently used woodfuel species by distillers did not coincide with what they preferred because of the scarcity of this preferred species. The shift of woodfuel use from the more preferred to the less preferred one suggests that the most preferred species are already threatened. The first two preferred species by distillers, A. etbaica and R. natalenis, are going to extinct in the study area. Average concentration of CO and PM: during distillation process as 68.81 ppm and 3.11 mg/m3; 8-hour concentration as 75.25 ppm and 5.57 mg/m3; the 15-minute average maximum as 115.31 ppm and 12.31 mg/m3, were measured, respectively. The indoor air pollutants level during the process exceeded air quality guidelines set by the World Health Organization standards.

Keywords: Arsi-Negele, Katikala, Distillers, Non-Distillers, Fuelwood, CO, PM

vii 1. INTRODUCTION

1.1 General Background

We have been living in the natural environment ever since we were born. The natural environment provides us with life support and material and cultural services. Without the natural environment, there will be no human civilization. With the increase of population and the improvement of living standard, the material demand of mankind will gradually approach the limits of the tolerance of the natural environment and an irreversible degradation will occur to the natural environment if we stick to the traditional mode of development. In fact, mankind already has the ability powerful enough to destroy the globe, which has become increasingly incapable of satisfying the demand of mankind (China Development Gateway, 2006). One of the most challenging current issues, taking into account the finite capacity of the earth's land and water resources, is the urgent need to increase world food production, to keep pace with the growing global population. According to UNEP (2000) report, many regions in the world already suffer from severe symptoms of overexploitation, degradation, and reduced productivity caused by unbalanced management of natural agro-ecosystems

What we eat and drink affects not only our health and well-being but also the environment and natural resources. Although it is commonly accepted and it has been widely demonstrated that the farming phase accounts for most of the environmental impact in the food production cycle, manufacturing and household consumption also play a key role due to their high use of resource and their high levels of by product generation. According to Giampietro (1994), in less developed countries, the value of energy spent for 1 nutritional kcal is about 35 kcal (most of which is in the form of biomass). And also the production of different types of alcoholic drink requires the input of different natural resources that have different environmental impacts. Alcohol is a more general term, applied to any organic compound in which a hydroxyl group (-OH) is bound to a carbon atom, which in turn is bound to other hydrogen and/or carbon atoms. The alcohol found in beverages is known as ethanol or ethyl alcohol. It is the only type of alcohol that is

1 safe to consume and a psychoactive substance, which means that it has the ability to change consciousness and to alter perceptions and behavior.

In many countries there are alcoholic beverages which are traditionally produced at the local level. This kind of production seems especially common in many African countries, where a wide variety of different beverages can be found. Many of these are produced by fermentation of seeds, grains, fruit, and vegetables or from palm trees which is a rather simple procedure. Through fermentation the alcohol content does not rise very high and often the beverages have a very short shelf life before they are spoilt. Distillation is a more complex procedure requiring more equipment, energy and time, but then the result is both more potent, has a longer shelf life date (WHO, 2004).

In Ethiopia, being one of African countries, different types of traditional alcoholic beverages such as Tella, Korefe, Shamit, katikala/ Araki and Borde are produced and consumed . Katikala1 (araki) Ethiopian traditionally distilled alcoholic beverage, is described by Selinus (1971) as follows:

“Araki is a distilled beverage. Ground gesho2 leaves and water are kept for three to four days and after that a kita3 made up of teff4 or other cereals and germinated barley or wheat are added. The mixture is allowed to ferment for five to six days and then distilled. In the villages distillation is carried out with primitive equipment made of gourds and wood. The local beer tella can also be distilled to produce arake. The araki can be redistilled and will then have higher alcohol content. The average alcohol content of dagim araki is around 45%. The term dagim in Amharic refers to ''second time'' and indicates that it is distilled a second time …” (Selinus, 1971).

It is well known that Arsi-Negele distilled katikala is highly demanded in the middle and southern part of Ethiopia including the main capital in Addis Ababa. Making of katikala in Arsi-Negelle town has a tremendous impact on the deforestation; since the process involves series of tedious steps all of which are done by women have health risk

2 associated with exposure to smoke and other products of combustion. This is intended to investigate the ecological impact, especially on the surrounding vegetation as well as some pollution effect caused by the process on human health.

1 Katikala (arake): local alcoholic beverage produced by distillation process 2 Gesho: is common name of Rhamnus prinoides used in making local alcoholic beverage in Ethiopia. 3 Kita: refers to baked bread made of whole-grain flour. 4 teff: an annual grass cultivated for its seed, which is used as a grain and a common name of Eragrostis tef.

3 1. 2 Statement of the Problem

The Rift Valley Lakes are very important in terms of biological resources. Their ecosystems support both aquatic and terrestrial biodiversity, such as migratory birds, wildlife, fishery resources and aquatic and terrestrial vegetation. The whole Rift Valley ecosystem, including its wetland drainage system and the uplands, is regarded as a rich strategic site for a wide variety of resident and migratory avifauna populations (Hillman, 1993).

However, Zerihun Woldu and Mesfin Tadesse (1990) pointed out that degradation of woodland is most remarkable in the central rift valley where woodlands were reduced to 4 percent of the original extent within a time span of 50 years. Among many areas exposed for such problems Arsi-Negelle Woreda is one of the area of Central Rift Valley. The, Langanoo, Abijata, and Shalla Lakes area of the mid rift valley is entertaining the same environmental degradation as that of the national catastrophe due to human interaction in existing ecosystem. For example, 30 years ago the woodland in the rift valley was continuous (Zerihun Woldu and Mesfin Tadesse, 1994 cited in Zerihun Woldu et al., 1999).

The mountain ranges on both sides of the Rift Valley have serious environmental problems, the impact of which accelerates the loss of biodiversity in the lakes. These problems are the result of a combination of social, economic and climatic factors, which have increased pressure on the natural resources of the Rift Valley Lakes (RVLs) and wetlands. This has caused the degradation of watersheds, increased soil erosion, decreased water quality and caused immeasurable loss to biological diversity (Lemlem, 2003).

In general the vegetation of the area is under threat by the combined forces of resource exploitation and forest degradation and particularly in Arsi-Negele woreda. One of the factors of degradation of the forest in the last few decades is the demand for firewood in Arsi Negele town which is derived from a lucrative business of brewing katikala, an

4 alcoholic drink locally consumed and highly demanded in many parts of the country. In the town Arsi-Negele almost every household distills katikala. Fuel wood supply to Arsi Negelle town is from surrounding area through every direction of the town. In addition, while fuelwood is used as a source of energy in the production process it causes health problems through indoor air pollution. The health risk associated with exposure to smoke and other products of combustion during production process is also serious. As it is pointed out by the World Bank (1992), in poor countries indoor air pollution from burning wood, charcoal, and dung endangers the health of 400 million to 700 million people.

The daily amount of ‘atella’ a by product of katikala extraction process is very high, which in turn have environmental impact if disposed improperly, however it is totally reused for animal fed. There is other form of impact such business has been contributing to food insecurity problems through in-taking large volume of grains for the production of alcohol in towns. The volume of input grains in which such activity has been demanding justifies the contribution of katkala production to food insecurity problems around Arsi-Negele. According to ADAA (2004) (unpublished report), katikala production in the town of Arsi-Negele alone consumes about 38.6 percent of wheat and 67.9 percent of maize produced on peasant farms in the woreda. Katikala consumption, on the other hand, has social, economic and health impacts. Katikala consumption can have adverse economic effects on the individual drinker and society as a whole .As it is stated in the WHO (2004), alcohol consumption can affect work performance through absences, work accidents, reducing productivity and unemployment, and increases medical and legal expenses and decreases eligibility for loans.

This katikala production system seems to work well for most families but is coursing problems because of environmental and socio-economic impact. This study aims to explore environmental problem associated with katikala production through out the whole process in Arsi-Negele. Generally in the rift valley particularly in the study area, it is very difficult to sustain the life’s of human, animal and wild life in the surrounding ecosystem (land and aquatic) with out taking precautionary and environmental life

5 supporting measures. Provision of alternative source of energy to decrease the rate of deforestation caused by the high fuel requirement of the distillation process and health risk associated with exposure to smoke and other products of combustion need to be put in place to sustain the environment.

Arsi-Negele was selected for this particular study because the area is well known by distillation of katikala and it is the main means of stay for the majority of residences in the town.

Previously, the impact of katikala production process on the vegetation and indoor air quality (IAQ) has not been studied. Thus, it is imperative to assess and document the direct and indirect ecological as well as health impacts of katikala distillation in the study area.

6 1.3 Objective of the Study

1.3.1 General objectives

The main objective of the study is: To assess impact of katikala production process on vegetation and indoor air pollution during distillation process and to investigate alternative solutions to the problem in the Arsi-Negele area.

1.3.2 Specific objectives

¾To identify patterns of fuelwood consumption by katikala distillers and non-ditillers households.

¾To identify hot spot areas of deforestation as a result of katikala distillation

¾ To study the degree of indoor air pollution as a result of katikala production and to recommend controlling mechanisms for indoor pollution.

¾ To identify the relation between pollutant concentration and different environmental factors.

7 2. REVIEW OF RELATED LITERATURE

General In many areas of the world, products derived from woody plants provide essential goods and services such as fuel for cooking and heating, construction material for housing, forage for cattle, and non woody products. Moreover, plants are important as they serve human populations through maintenance of cultural values and symbolic use and by being sources of raw materials. They also provide aesthetic values and maintain the environment. Hence, they are important for providing support to the earth’s life. For this reason human life has great impact on plant species diversity and vice versa (World Resources Institute (WRI), 1997). In Africa, 80% of all people use plants for the major part of their daily needs (Cunningham, 1988). Kaale (1985) stated that nearly 100 percent of the rural population, which accounts for over 80 percent of the total population in East Africa, relies entirely on wood energy as their main source of domestic fuel. According to the same author over 50 percent of the urban population in this region of Africa also depend on wood energy. Due to the increasing population pressures and lack of appropriate proper management strategies, the benefits are no longer readily available.

The interest in substituting of other forms of energy with energy from biomass is not satisfactory. Total energy consumption increases linearly with gross national product and fuelwood decrease with increased use of other energy resources (Earl, 1975). According to Bridgewater (1991) there is no indication that fuelwood can be replaced with other sources of energy in the foreseeable future. Rather FAO (1983) estimated an annual increase in fuelwood consumption by a rate of 5%. However, most people who used fuelwood as an energy source result in to suffer in bad living condition in addition to its environmental impact by contributing to vegetation degradation.

8 2.1 Forest Resources Degradation

The World Resources Institute (1997) estimate that only one-fifth of the world's original forest cover remains, largely in blocks of undisturbed frontier forests in the Brazilian Amazon and boreal areas of Canada and Russia. The destruction of forest and woodland resources has a complex implication on the status of the environment because vegetation cover and dead plant biomass are known to reduce soil erosion by intercepting and dissipating raindrops and wind energy. Having intercepted this rainfall, they facilitate the infiltration rate of water to the ground. Forests also improve the quality of the soil underneath by the addition of organic matter through litter fall and nitrogen fixation. Moreover, it protects the soil from extreme erosion by interweaving through its root net. Under this situation, lowest erosion rates have been recorded from undisturbed forests, with ranges from 0.004 to 0.5 t/ha per year (WRI, 1997).

Even though forests have the above benefits, they have been treated as free resources or open access, taken largely from land to which every one has the right to access. According to (Sayer et al., 1992) the heavy dependence on forests for fuel and building materials combined with population growth, contribute to increasing rate of forest and woodland destruction. Once forest land is converted to agriculture, erosion rates increase because of vegetation removal, over-grazing, and continuous cultivation. Also there is a better understanding that forests burnt in certain parts of the world are important contributors to greenhouse gases and contributing to climate change. Overall these changes affect the livelihoods of societies.

In Ethiopia, population pressure is inducing, the clearing of forests for agriculture fuelwood, and other purposes, and the attendant accelerated soil erosion, is gradually destroying the soil resource (Hurni, 1990). This is because natural forests are the main sources of wood for fuel, construction and industry, even though plantation forestry is also increasingly becoming important. In Ethiopia forests may have existed long before history was recorded, but the present day forest cover does not correlate with human population in recorded history, even though environmental problems such as droughts

9 may have also contributed to this phenomenon. The annual loss of natural forest cover has been estimated to be 150,000 to 200,000 ha/yr-1, at a rate of 171,000 Km 2 y-1 and in 1989 forest cover was estimated at only 2.7% of the Ethiopian land mass (EFAP, 1993). Historical estimates suggest some 87% of the Ethiopian highlands had forests and woodland covers, but this was reduced to 40% by 1950, 5.6% by 1980 and 2.4% by 1990 (Sayer et al., 1992). Today, little of the natural vegetation of the highlands remains, except for the southern and southwestern parts of the country. Consequently, large forest areas of the country are now exposed to heavy soil erosion resulting in a massive environmental degradation and serious threat to sustainable agriculture and forestry. It has been projected that, if the present rate of deforestation continues, by 2010, the area covered by natural forests will be reduced to scattered minor stands of heavily disrupted forests in remote parts of the country (EFAP, 1993).

Generally, deforestation can result in the loss of biodiversity also; which in turn results in declines in ecosystem integrity, and also genetic losses that may impede future scientific advances in agriculture and pharmaceutics. WHO et al., (1993) reported that, as many as 80% of the world population depend on herbal medicine for primary health care needs which are mainly derived from forests. The consequences of deforestation will therefore be felt by the many poor because of lack of cash to buy modern medicine. In addition, deforestation can also impact hydrological processes, leading to localized declines in rainfall, and more rapid runoff of precipitation, causing flooding and soil erosion (Dagnachew Legesse et al., 2003).

2.2 Consequence of Woodfuel consumption on the Environment

Heavy reliance on woodfuels can result in a range of negative environmental impacts, with both local and global consequences, including forest loss and degradation, health problems for households and charcoal producers where biofuels are combusted (Ezzati and Kammen, 2002), and increased greenhouse gas emissions (Bailis et al., 2005).

10 It is estimated that biomass combustion contributes 20-50% of global greenhouse gas (GHG) emissions (Crutzen and Andreae, 1990; IPCC, 1990). Household stoves, although individually small, are numerous and thus have the potential to contribute significantly to inventories of greenhouse gases (GHG), particularly in those many developing countries where household use is a significant fraction of total fuel use. The emissions of greenhouse gases from small-scale combustion of biomass are not well characterized (IPCC, 1997). From this study point of view, reviews of related point for forest ecosystem degradation and Indoor Air Pollution due to fuelwood consumption are as follows.

2.2.1 Impact of Woodfuel Consumption on Forest Ecosystems

Fuelwood has been identified as one of the most significant causes of forest decline in many developing countries. According to Osei (1993) estimate, firewood accounts for over 54% of all global harvests per annum, suggesting a significant and direct role of woodfuel in forest loss.

It is estimated that the extraction of wood from tropical forest for timber, charcoal burning and fuelwood constitute 68 % of the proximate causes of deforestation in Africa, 89 % in Asia and 51 % in Latin America (Geist and Lambin, 2001).

Even if forest is not entirely cleared, selective harvesting for fuelwood may change forest composition and ecosystem function (Struhsaker, 1997). For example, the harvesting of trees from old-growth forest may result in slow growing species being replaced with faster growing secondary species. Secondary growth is more susceptible to fire. If the burn frequency is sufficiently high, these disturbed forested areas, which were originally mature forest, can be converted to grassland that can be maintained indefinitely by fire. Unlike other tropical regions, many African mid-elevation secondary growth species do not provide food for frugivorous birds or primates (Struhsaker, 1997). Thus the replacement of old-growth trees with secondary growth reduces populations of many frugivores (Struhsaker, 1997). This may lead to the disruption of ecosystem, community or population structure and changes in resource and substrate availability or

11 the physical environment. Changes in ecosystem characteristics such as species diversity, nutrient output, and biomass as well as changes that reset succession in one or more sites may be significantly disturbed (Mooney and Godron, 1983; White and Pickett, 1985; Turner et al., 1993; Burrows, 1993).

Fuelwood gathered from the forested commons is the most important source of domestic energy in the rural areas of many developing countries (Cecelski, et al., 1979). About half of the world’s populations cook with biomass fuels, which provide around 35 % of energy supplies in the developing countries (World Bank, 1992). Fuelwood collection and consumption are intricately linked to natural resource management. There is a two- way relationship between fuelwood collection and deforestation. On the one hand, demand for fuelwood from commons and forests causes’ resource degradation to the extent that collection exceeds sustainable yield. Forest degradation, on the other hand, leads to a situation of fuelwood scarcity. Indeed, mounting global “fuelwood crisis” has been envisaged (Dewees, 1989). In addition, there are a number of other adverse consequences of forest degradation, including loss of biodiversity, deterioration of watershed management functions, release of carbon dioxide into the atmosphere, and soil erosion.

2.2.2 Fuelwood consumption pattern and Forest Degradation in Ethiopia

Approximately half of the wood cut annually worldwide is used as fuel, and of this amount, nearly 90 % is produced and consumed in developing countries, where firewood and charcoal constitute the primary source of energy for the poor (Parikka, 2004). This reliance is even more pronounced in sub-Saharan Africa, where wood fuels are the dominant energy source, both in terms of primary energy supply and the number of people relying on them (Bailis et al., 2005).

Bownder (1987) reported that many developing countries have a substantial reliance up on wood fuels. Akinbami et al., (2003), quantitatively expressed the reliance that biomass accounts for over 80 % of annual energy consumption for developing countries while it

12 was 2 % for developed nations. The case in Ethiopia is still greater than the developing countries’ average and it is estimated to be 89 % (Bownder, 1987). Alemneh (2003) state that in Ethiopia fuelwood burning contributes as one of the main factors for deforestation with biomass fuels supplying nearly 95 % of the country’s energy consumption.

In Ethiopia the demand for fuel wood is high. Biomass fuel contributes to 95 % of the total energy consumption in Ethiopia. Furthermore, 96.2% of the total consumption of the biomass fuel is used in the household sector; i.e. as firewood (Bekele., 2003). This has led to an increasing deficit of fuel (Mariame, 1996). Some studies suggest that 90 % of the annual forest harvest is used as firewood. Most households in Ethiopia use traditional means to prepare their food. A study states that 91 % of the households use BLT (branches, twigs and leaves); which makes BLT the most important category of fuel wood (Bekele, 2003).

According a study conducted by Anderson (1986), annual rate of consumption has reached the point where it is estimated to exceed the mean annual incremental growth of local tree stocks and forest reserves by 150 percent in Ethiopia. According to the World Bank (1984), traditional fuels (firewood, charcoal, twinges, leaves, straw, stalks, crop residues and animal dung) contributed about 94 percent of the gross energy supply of Ethiopia by 1989, which is one of the highest, in the world. Of the traditional fuels, fuelwood is the dominant one. There is a strong cultural preference for fuelwood, making it the fuel of first choice in Ethiopia (World Bank, 1984).

Currently fuelwood supply is getting scarce in Ethiopia as the supply source dwindles, being substituted by other form of biomass fuels such as dung, branches/leaves and agricultural residues which otherwise would have been used for soil improvement. On the other hand, use of dung as fuel is at the opportunity cost of using it as organic fertilizer (Newcomb, 1984). Even earlier, fuelwood demand and supply was not balanced from 1976 - 87, where demand exceeded supply by twofold with 42 million cubic meters while supply remained at 24 million cubic meters (Alemneh, 2003).

13 Firewood, dung, charcoal, crop residue and kerosene are sources of domestic energy in both urban and rural parts of the study area (ORS, 2004). The demand for firewood in Negele town is derived from a lucrative business of brewing katikala, an alcoholic drink locally consumed and highly demanded in many parts of the country. It was found that, most of Arsi-Negele residents are engaged in katikala production by using wood-fuel.

The World Bank (1984), states that "The most important issue in the future of Ethiopia is the supply of fuels, the related massive deforestation, and the resultant and insidious depletion of agricultural resources on which so much economic activity depends." This problem is one of the major problems affecting the future of Ethiopia and hence should stand as the nations’ top priorities to be considered by policy makers.

2.3 Impact of Indoor Air Pollution

It is conventionally believed that air pollution is associated with the contamination of urban air from automobile exhausts and industrial effluents. However, in developing countries, the problem of indoor air pollution far outweighs the ambient air pollution. There are four principal sources of pollutants of indoor air (Behera, 1995): (i) combustion, (ii) building material, (iii) the ground under the building, and (iv) bioaerosols. In developed countries the most important indoor air pollutants are radon, asbestos, volatile organic compounds, pesticides, heavy metals, animal dander, mites, moulds and environmental tobacco smoke. However, in developing countries the most important indoor air pollutants are the combustion products of unprocessed solid biomass fuels used by the poor urban and rural population for cooking and heating. Indoor air pollution can be traced to prehistoric times when humans first moved to temperate climates and it became necessary to construct shelters and use fire inside them for cooking, warmth and light. Fire led to exposure to high levels of pollution, as evidenced by the soot found in prehistoric caves (Albalak, 1997). A report of the World Health Organization (WHO, 1997) asserts the rule of 1000 which states that a pollutant

14 released indoors is one thousand times more likely to reach people’s lung than a pollutant released outdoors.

The use of solid fuels for cooking and heating is likely to be the largest source of indoor air pollution on a global scale. In households with limited ventilation (as is common in many developing countries), exposures experienced by household members, particularly women and young children who spend a large proportion of their time indoors, have been measured to be many times higher than World Health Organization (WHO) guidelines and national standards (Bruce et al., 2000; Smith, 1987).The United States Environmental Protection Agency’s 8-hour average carbon monoxide standard is 9 ppm or 1 mg/m3 (USEPA, 1997). In some areas exposure is longer because wood fires are used for heat during the night and the winter months (Achmadi, 1992). Particles with

diameters below 10 microns (PM10), and particularly those less than 2.5 microns in

diameter (PM2.5) can penetrate deeply into the lungs and appear to have the greatest potential for damaging health (USEPA, 1997).

2.3.1 Indoor Air Pollution Due to Combustion and Its Health Effect in Developing Countries

Household energy impacts on the health of the poor through a variety of physical, social and economic routes, but the most important direct health impact results from indoor air pollution (IAP) produced by burning biomass fuels (wood, dung, crop wastes) and coal in simple stoves with inadequate ventilation(WHO, 2000).

Household energy and indoor air pollution pose a substantial threat to the health of the world’s poor. In developed countries, modernization has been accompanied by a shift from biomass fuels such as wood to petroleum products and electricity. In developing countries, however, even where cleaner and more sophisticated fuels are available, households often continue to use simple biomass fuels (Smith, 1987). Although the proportion of global energy derived from biomass fuels fell from 50% in 1900 to around 13% in 2000, there is evidence that their use is now increasing among the poor (Albalak

15 ,1997). Poverty is one of the main barriers to the adoption of cleaner fuels. The slow pace of development in many countries suggests that biomass fuels will continue to be used by the poor for many decades.

The majority of households in developing countries burn biomass fuels in open fireplaces, consisting of such simple arrangements as three rocks, a U-shaped hole in a block of clay, or a pit in the ground, or in poorly functioning earth or metal stoves (Smith, 1987). Combustion is very incomplete in most of these stoves, resulting in substantial emissions which, in the presence of poor ventilation, produce very high levels of indoor pollution (WHO, 2000).

The greatest burden of mortality arises from indoor exposures in rural areas of developing countries. Estimates of the global burden of disease suggest that indoor air pollution is responsible for around 4 % of the disability-adjusted life years lost, meaning that its consequences are comparable with those of tobacco use and that they are only exceeded by those of malnutrition (16%), unsafe water and sanitation (9 %) and unsafe sex (4 %) (Murray, 1997).

By far the largest contribution to the disability adjusted life years lost arises from acute respiratory infections because of their high incidence and the mortality for which they are responsible among young children (Smith, 1999). In developing countries respiratory disease is the chief cause of death (Krugmann, 1989; WHO, 1992). The most important appear to be childhood acute lower respiratory infections, which remain the single most important cause of death for children aged under 5 years in developing countries (Bruce et al., 2000).

People in developing countries are commonly exposed to very high levels of pollution for 3–7 hours daily over many years (Engel et al., 1998). During winter in cold and mountainous areas, exposure may occur over a substantial portion of each 24-hour period (Norboo, 1991). Because of their customary involvement in cooking, women’s exposure is much higher than men’s (Behera et al., 1988). Young children are often carried on their

16 mothers’ backs while cooking is in progress and therefore spend many hours breathing smoke (Albalak, 1997).

Exposure can be reduced by means of improved stoves, better housing, cleaner fuels and behavioral changes. Cleaner fuels, especially liquefied petroleum gas, probably offer the best long-term option in terms of reducing pollution and protecting the environment, but most poor communities using biomass are unlikely to be able to make the transition to such fuels for many years (Nigel et al., 2000). Indeed, until recently, the main emphasis of stove programs has been to reduce the use of wood, and consequently there has been relatively little evaluation of reductions in exposure (Bruce, 1999).

2.3.2 Major Air Pollutants Released from Biomass Combustion

It has been estimated that more than half world’s households cook their food on the unprocessed solid fuels that typically release at least 50 times more noxious pollutants than gas (Smith, 1990).

The stoves used for cooking are not energy efficient. The fuels are not burned completely. The incomplete combustion of biomass releases complex mixture of organic compounds, which include suspended particulate matter, carbon monoxide, poly organic material (POM), poly aromatic hydrocarbons (PAH), formaldehyde, etc and the biomass may also contain intrinsic contaminants such as sulfur, trace metals, etc (De Koning, 1985).

Carbon Monoxide

Incomplete combustion of fuels produces carbon monoxide (CO). The CO and particle emission pose a serious problem when biomass fuels are used. Smith (1991) has estimated that about 5 g/meal carbon monoxide is released during the household cooking by using woodfuel. A study (Patel and Raiyani, 1995) by the National Institute of Occupational Health (NIOH), reported indoor air CO levels of 156 mg/m3 air during cooking by wood. The short-term health effects of CO exposure are dizziness, headache,

17 nausea, feeling of weakness, etc. The association between long-term exposure to carbon monoxide from cigarette smoke and heart disease has been described by several authors (Wynder, 1979).

Particulates

In recent years a large number of studies of health impact of suspended particulate air pollution have been undertaken in developing countries (WHO, 1997). These studies show remarkable consistency in the relationship observed between changes in daily ambient suspended particulate levels and changes in mortality. Smith (1996) estimated the health risk from exposure to particulate air pollution by applying the mean risk per unit ambient concentrations based on the results of some urban epidemiological studies (Hong, 1997). According to this study the range of risk was found to be 1.2 - 4.4 % increased mortality per 10 mg/m3 incremental increase in concentration of respirable

suspended particles (PM10). For the calculations of estimates, it was assumed that the health risk has linear relationship to exposure, the risk factors determined for urban centers of developed nations were used as standards; where the PM10 data were not available, and 50 % of suspended particulate matter (SPM) levels were considered as equivalent. The above assumptions may add to inaccuracy already inherent in such estimates. The United States Environmental Protection Agency’s standards for 24-hour 3 3 average PM10 and PM2.5 concentrations are 0.15 mg/m and 0.65 mg/m respectively.

2.3.3 The Gap in Pollutant Testing Methods

Development organizations attempting to promote improved stoves in developing countries often recognize the need for reduced smoke and other emissions from these stoves. Historically research organizations have used efficiency and cost as the primary criteria for rating rural stove suitability. This has provided little incentive to study the emissions from biofuel combustion. Technical difficulties in measuring rural stove emissions have also contributed to this information scarcity. However, the need for the development of emission testing methods is frequently stressed in the literature (WHO, 1992). In a study of the priorities of improved cook stove programme from around the

18 world, decreasing smokiness scored almost as high as increasing household fuel efficiency (Ramakrishna, 1992).

It has sometimes been assumed that efficiency is a sufficient measure for gauging emissions - that improving efficiency will naturally lead to reduced emissions. A number of studies have shown that this clearly is not the case (Smith, 1992; Ahuja et al., 1987). In fact it has been recognised that a conflict exists between high efficiency and low emissions (Shelton, 1982). The reason for this is simple to understand; efficiency is influenced by two largely independent factors, combustion efficiency and heat transfer efficiency. Combustion efficiency, a measure of how well the fuel is burnt, relates directly to emissions. Poor combustion efficiency means that the fuel is not completely burnt and therefore the products of incomplete combustion are emitted from the stove. However, heat transfer efficiency (how well the energy released from the wood is transferred to the pot) can be improved while, at the same time, decreasing combustion efficiency. This often yields an improved overall efficiency but increased emissions (Smith 1992; and Ahuja et al 1987). It is therefore vital to measure stove emissions and not to assume that a high efficiency stove will have low emissions.

The majority of households in Arsi-Negele burn biomass fuels in open fireplaces, consisting of such simple arrangements as three rocks for preparing meals and a U- shaped hole in a block of clay for distillation of katikala. Combustion is very incomplete in most of these stoves, in the presence of poor ventilation, produce very high levels of indoor pollution. Indoor air pollution is a major public health hazard for most of the resident. In addition a huge amount of fuelwood is consumed in the town; this is evident in the woreda, where most of the acacia trees were removed in the last 30 years (Fig 2.1).

19 50000

45000

40000 Water

Grazing land 35000 Open Forest

30000 High Forest

Disturbed Wood land; Scattered farm plots and 25000 settlement

Area (ha)Area Acacia Wood Land 20000 Intensively Cultivated Land

15000 Degraded Grass Land, retreated water body

Homestead 10000

5000

0 1973 1986 2000 2006 Years

Figure 2.1 Land Use / Land cover categories of different years of Arsi-Negele woreda (Source: Zelalem Amde, 2007)

The study is the first of its kind in the woreda in providing impact of katkala production process. In the light of the need to conserve the dwindling natural biomass of the country in general and in Arsi-Negele woreda in particular and for successful improvement of improved stove for such kinds of local industry the importance of this kind of research should not be under-valued.

20 3. MATERIALS AND METHODS

3. 1.The Study Area: Arsi-Negele Woreda

3.1. 1 Location and Description

Arsi-Negele is one of the weredas of the east shao zone of Oromia region, found in the Ethiopian rift valley system located 225 km south of Addis Ababa, the capital of Ethiopia. It is bordered on the south by Shashamene, on the southwest by Lake Shala which separates it from Seraro, on the west from the Southern Nations, Nationalities and Peoples Region, on the north by Adami Tullu and Jido Kombolcha with which it shares the shores of Lakes Abijatta and Langano, and on the east by the Arsi Zone. The major rift valley lakes of Abijata, Langano and Shalla are partly in Arsi Negele accounting for about 32% of the total area of the district. Most part of Shalla-Abijata national park is found inside Arsi Negele woreda. Arsi Negele has the highest number of rivers in the zone. They include Gedamso, Lephis, Huluka, Awede Jitu, Awede Gudo & Dadaba Gudo.

3.1.2 Climate and Topography

Conventionally, the Arsi-Negele woreda is divided into 3 major climatic zones based on altitude (low, mid and high altitude). The high altitude climatic zone occupies the largest area followed by mid and low altitude climatic zones. Average annual temperature varies from 10-25 °C while rainfall varies between 500-1000 mm (ORS, 2004). About 80% of the district is sub-tropical, while 20% belongs to the temperate agro-climatic zone.

The topography is slightly undulating in the highlands and almost flat in the lowlands. Except the southeastern part, most of the district’s elevation is between 1500 and 2300 metres. Gara Duro (3095m) is the highest peak in the district.

Figure 3.1 Location Map of Arsi-Negele Woreda

3.1.3 Geology and Soil

One of the defining geologic features of Ethiopia is the Great Rift Valley, slicing through the middle of the country from the Red Sea to Kenya. It is a geological relic of the critical weakening in the earth's crust along two roughly parallel faults which opened some 20 million years ago, the world's largest geological divide. Pumice, sand, soda ash, diatomite, salt, etc have been available. These construction materials and industrial minerals are being mined for different purposes. From construction materials, sand it the most exploited while soda ash is the widely exploited industrial mineral in the district.

The residual products of rock decay are rich in iron and Aluminum but poor in biologically important elements like potassium, calcium and sodium and 83 % of the soil

22 is classified as sandy loam, 9% as sand (EARO, 2002). Andosol soil type covers about 52.2% of Arsi-Negele, while Nitosols cover the remaining 47.8% (ORS, 2004).

3.1.4 Vegetation

Coniferous forests of podocarpus variety, woodland, and broadleaf forests prevail in the district. At the Rift Valley plain, open Acacia woodland dominates, and this gradually turns into dry open deciduous woodland of a transitional vegetation type (Eriksson et al., 2003). At mid-altitude i.e. between 2100-2600 m above sea level tropical dry evergreen montane forest dominates. Different plant communities comprise this section. At the lower sub-humid part a Podocarpus falcatus - Croton macrostachyus mixed forest exists, which gradually converts into the humid zone dominated by Podocarpus falcatus forest. These vegetation communities are all referred to as ‘Montane forests’ in many classification systems (Brown and Cocheme, 1969). Based on the description given by Lundgren (1971) the dominant tree both in size and frequency is Podocarpus falcatus.

The vegetation of the woodland at lowland part can be classified mainly as Acacia- Balanites with some thorny shrublands occurring around the lakes. The characteristics species of woody plants include various types of trees, shrub, sub-shrub and climbers with different distribution and abundance. Small area of woodland surrounding the head quarter of the National Park (which is protected from human interference) exhibits the initial complex of plant species diversity in the study area.

According to Pichi-Sermolli (1975), the present vegetation of the Rift Valley developed as a result of climatic episodes that took place in the last 3 millions years. During this period three ice ages affected the climate and vegetation formation. The woodlands surroundings the lakes and the four lakes Zeway, Langano, Abjiata and Shalla were separately formed as a result of uninterrupted lowering of the water level about 9,400 to 8,400 years B.C (Grove et al., 19875). Besides, there are also plantation species, which are exotic. The main species include Cupressus lustanica, Pinus patula, Eucalyptus globules (E.globules), E. grandis, and E. viminalis.

23 3.1.5 Land Use and Agriculture

2 The total area of the woreda is about 1396 km of which 52% is arable, 30% water bodies, 5% forest and 13% grazing and others. Arsi-Negele is characterized by crop- livestock based farming systems. It is rich in both crop production and livestock rearing. The woreda consists of 35 peasant associations (PAs) and 3 urban kebeles. Maize and wheat are the most important cereal crops grown in the district. Annual crops accounted for 95% of all crop lands in Arsi Negele. Grazing and forest lands covered 4.3% & 5.2% respectively. Degraded and others accounted for 60.6% of the district. The average farm size per household was 2.19 hectares, while the average farm oxen per household were 4(OSR, 2004).

3.1.6 Population

The district had a total population of 147,114 in 1997. The urban population was 17.8% in the same year. The ages groups 0-14, 15-64 and above 64 years were 50%, 48% and 2% respectively. About 51.3% of the urban and 50.7% of the rural populations was females.

The average family size for the district was 5.2 (5.3 for urban & 5.1 for rural). The population density of the district was 105.4 persons per km2 (ORS, 2004). The total population of the woreda is estimated to be 161,000 in 2000 out of whom 81% is rural (CSA, 2005).

24 3.2 Study Design

3.2.1 Household Survey

Structured interviews were mainly used to obtain both qualitative and quantitative data form the household. Data were collected through the structured questionnaires using random sampling. In total, 165 interviews giving an average household size of 6.56 persons per household were completed at the house hold level, representing 3% of the total population of Arsi-negele town.

During the household interviews information was obtained on the preferred fuelwood species, along with quantities; where they had obtained each piece of fuelwood. Six enumerates were selected from the study area for administering the questionnaire. The enumerates were carefully trained and pre-testing of the questionnaire was done prior to the start of the actual work. The questionnaire generated quantified data on species preference, amount of fuel wood use and marketing (Annex-1).

3.2.2 Market Survey

Structured market questionnaire was also included (Annex-3). One hundred five vendors were interviewed to obtain information on daily supply, location of fuelwood gathering, preference of fuelwood species to sell etc…The amount of fuelwood supplied to four market place i.e. Atena-Tera, Tureta, Mobile and Board counted for a week. In 3 conjunction to the counting study it is also differentiated the fuel carriers in terms of m , donkey drawn cart, horse drawn cart, donkey load, woman load, man load and also car (Annex-3). The counting study was conducted from 06.00 am-6.00 pm during the week.

3.2. 3 Woody Vegetation Survey

Vegetation sample was taken at the acacia woodland to crosscheck the information gathered through the questionnaire with the impact on the vegetation around the town. This was done at kebele Galle-akello, Shalla-billa, Keraru specifically the locality were Galle, Shalla and park headquarter, and Korre-Mendia village respectively where the

25 most preferred species were expected to exist. Additionally these areas were source of fuelwood during the sampling period of the study (Fig. 3.2).

Figure 3.2 Field Vegetation Survey Sample Point

Furthermore, on each four village three parallel transects 350m apart were established in each sample taking village making a total of 12 transects. The park area surrounding the headquarters has been protected from human activities relatively since the establishment of the park. Along each transect, plots were established at 300 m intervals using quadrants of 20 × 20 m. Within each sampling plot all individual woody plant species were recorded and their height and diameter at breast height (DBH) were measured.

26 Height was measured using a Clinometer while DBH was taken at 1.3m heights by using a Caliper. Young trees with a DBH of less than 2.5cm were considered as seedlings and those with DBH 2.5 – 10cm were recorded as saplings. Tree stumps appearing in every 20 x 20m plot were counted and recorded. A Garmin GPS (global positioning system) device was used to record co-ordinates of centre of sample plots and altitude.

3.2.4 Indoor Air Pollution Measuring

To test the impact of the katikala distillation process on indoor air quality (IAQ), monitoring of carbon monoxide (CO) and particulate matter (PM) was done during the whole distillation process for 31 households. These kitchens were used only for katikala distillation process.

Carbon monoxide was measured using HOBO Carbon Monoxide Data Logger (Onset Corporation, Bourne, MA). The HOBO CO monitors used in the study are commercial data-logging devices available at relatively low cost. It is easy and portable. The HOBO carbon monoxide logger has three measurement ranges. Data were collected using three channels, Channels-1, Cannel-2 and Channel-3 of the monitor, which cover the ranges 0.2–124.3 parts per million (ppm), 1–497.1 ppm, and 4 – 1988 ppm, respectively. As reported by the manufacturer, the typical accuracies of reading according to decrease as of channels -1 toward channel-3 though the range of reading to increase. Normally only one range should be selected. The concentration CO was recorded in 1-minute intervals between the hours from 6:00 for the tests starting in the morning and 19:15 for the tests ending in the afternoon. C-1 was used for CO concentration less than 124.3 ppm and C-2 was used for CO concentration greater than 124.3 ppm for every 1-minute intervals. No value greater than 497.1 ppm was encountered throughout the study period.

During the same period, the suspended particulate matter was measured in the same intervals. Measurement of suspended particulate matter was carried out using the UCB Particle Monitor (University of California, Berkeley). The UCB particle monitor is a

27 programmable continuous particle monitor. In addition to measurements of PM, the UCB also logs temperature and relative humidity.

Data were recorded about at a distance of 1m from the stove and at a height of approximately 1.5m. This area is assumed to be the breathing zone of the distillation process. The concentration of suspended particulate matter was averaged over and recorded in 1-minute intervals between the hours from 6:00 for the tests starting in the morning and 19:15 for the tests ending in the afternoon.

In addition to indoor pollutant concentrations, information on other relevant factors was collected, which included daily woodfuel use rate, volume of kitchen, mean indoor temperature, katikala produced per day and number of stove in a kitchen used for distillation process during the monitoring period (Annex-7).

Data were downloaded from the UCB and HOBO monitors using protocols of the instrument. These data were then exported to an Excel spreadsheet. Data from HOBO CO logger was downloaded and processed using the software (Boxcarpro) specifically designed for the instrument itself. Likewise, the UCB particle monitor has its own data downloading and processing software. The latest version of the software was used for this study. The software has two parts. The first part (UCB PM Manager) is used to launch the instrument, download the raw data, and process it. The other part of the software (UCB PM Browser) is used to read the downloaded data and export it to the excel spreadsheet.

The parameters used for making analysis on the indoor air conditions were average during distillation process, the maximum observed 15-minute concentration in the whole distillation process and an 8-hour average from 7:00 am-3:00 pm (7:00-15:00). Inhalation exposure is a function of the air pollutant concentration at the breathing zone and the duration that a person experiences it (Smith, 1993; Zartarian et al., 1997; Lioy, 1990).Mathematically, potential exposure (E) is calculated as

E = C average T, where C average = average concentration, T is the duration of time a person stays in a kitchen throughout an entire task (t from 0 to T).

3.3 Data analysis

SPSS statistical package and Microsoft excel spreadsheet was used for statistical analysis. Descriptive statistics (applied by the help of SPSS program and Microsoft Excel) was used to analyze and summarize (into tables and graphs) the data obtained through households and market interviews, market count, field observation and indoor air pollution monitoring.

The investigation focused on the impact of katikala brewing on indoor air quality, which was evaluated according to three time-based parameters for both CO and PM: average during distillation process, the maximum observed 15-minute concentration and an 8- hour average from 7:00 am-3:00 pm. For each parameter of interest, multiple regression testing was conducted using regression models. In the multiple regression models, independent but potentially influential variables were included appropriate. These variables included kitchen volume, daily woodfuel use rate, volume of kitchen, mean indoor temperature, katikala produced per day and number of stove in a kitchen. However, to take a conservative approach, all variables were retained in the multiple regression models even if non significant. The impact of reported fuelwood use on indoor air quality was examined by constructing bivariate regressions was conducted using

29 analysis of variance (ANOVA) of each indoor air pollutant (CO and PM) against fuel use rate for each of the time periods of interest; during the whole time of distillation process, 15-minute maximum, and 8-hour indoor pollutant morning period. To verify the relationship between PM and CO a simple linear model was developed....………………..

30 4. RESULTS AND DISCUSSION

4.1 Fuelwood Consumption by Households

In Arsi-Negele town 87.3 % of the populations directly support their life by distillation of katikala and woodfuel has been continue to provide all the energy required for distillation of it. Fuelwood was purchased in many different measuring units. Analyzing was a little complicated to evaluate because most of the people gave their answers in other measures than kilograms. The most common measure concerning firewood at the study area was donkey load, donkey drawn cart. One donkey load weighs an average of 49.82 kg according to ministry of agriculture (1996) and one donkey drawn cart corresponds to four donkey load.

The per household woodfuel consumption rate in the town was equivalent to 69.38 kg /day and 387.52 kg /week with standard deviation 25.52 and 153.29 respectively. The rate was different for household (distillers and non-distillers) ranged between 12.46 and 124.55 kg per day. For non-distillers, the per household woodfuel consumption rate (17.82 kg per day) was lower than the average. However, for katikala distillers, the rates were higher than the average (76.98 kg per day). The difference is due to the need o f energy in the form of heat by using fuelwood for distillation process throughout the day. The large volumes of fuelwood consumed for katikala distillation coupled with other factor is driving rapid deforestation around the town.

Table 4.1 Daily Consumption rate (Kg/day)

Consumption Rate Minimum Maximum Mean Std. Deviation

12.46 124.55 69.3786 25.52070 Household

Distillers 37.37 124.55 76.9778 16.82637

Non-Distillers 12.46 37.37 17.8205 7.16343

31 4.1.1 Species Preferences of Households

Household’s fuelwood species preference was also included in the questionnaire. Twelve fuelwood species were identified as the most preferred and used by the respondents. There are different factor that acts as the determinant of preference to the distillers and non-distillers of the category.

Most of the households reported some plant species preferences for fuelwood use. A. etbaica was found to be the most frequently preferred (by 27.3 % of the household) species followed by dry wood materials and branches, which was chosen by 13.9 % of the household most of them are non- distillers which was preferred due to its low cost. Katikala distillers have the extensive preferential use of plant species for use as fuelwood as an energy source. Table 4.2 Fuelwood species preference of the consumers (household= Distiller+ Non-Distiller).

Species % Household % Distiller % Non-Distiller Main Determining sampled sampled sampled Factors

A. etbaica 27.3 29.2 14.3 Good Quality dry wood materials and Low Cost and branches 13.9 7.6 57.1 Availability

R. natalenis 10.9 12.5 - Good Quality Availability and Podocarpus falcatus. 9.1 10.4 - Good Quality Good Quality A. seyal 9.1 8.3 14.3 Good Quality A. tortilis 7.3 7.6 4.8 Availability and Junipers procera and 6.1 6.9 - Good Quality Cupressus lustanica Availability and Eucalyptus species 4.2 4.2 4.8 Good Quality Good Quality A. senegal 3.6 3.5 4.8 Good Quality B. aegyptiaca 3.6 4.2 - Low cost and Other types of species 4.8 5.6 - good quality

32 The respondents described a number of attributes or factors that they considered to be important for good fuelwood. The most important attributes were cost, availability and quality of the species. If the fuelwood species have good quality, the fuewood should have a hot flame, a flame that is long lasting, produce long lasting embers, and be easy to split and ignite. In addition the preferred species would have to burn without producing much smoke, and have a flame that does not produce sparks. Conversely, the most frequently mentioned negative attributes of fuewood are related to wood being difficult to

ignite, that it burns out easily, or produces too much smoke or ash.

The preference of woodfuel species specified by distillers in this study is not in agreement with what has been found out for non-distillers. Considering only the first three species, the arrangement of species in order of the magnitude of preferences given by the distillers in this study were: A. etbaica, was the most preferred tree species followed by R. natalenis, and then Podocarpus falcatus. However the non-brewers reveal somewhat different ordering: dry wood materials and branches, and then A. etbaica and A. seyal.

The variation between the characteristics and qualities of fuelwood that from the distiller point of view are: A. etbaica preferred first by distillers because only small quantity of the wood can burn like charcoal for a longer time. Thus the fire provides a lot of heat and leaves hot persistent glowing embers for distillation process. It can also have the property of good flammability. However, the basic reason R. natalnsis was preferred by fewer distillers than A. etbaica because it burn noisily and emits dangerous sparks that may harm their cloths during burning. Podocarpus falcatus was preferred by fewer number of distillers than A. etbaica and R. natalensis because of its resistance to cutting and splitting, slow drying rate and difficult to ignition, and it leaves much ash after combustion. On the other hand, the species has got good qualities such as slow continuous burning with release of much heat, and very good embers formation.

The variation between the characteristics and qualities of fuelwood that from the non- distiller point of views are: dry plant materials and branches preferred by most of the

33 non-distiller due to its relatively low cost, readily available A. etbaica preferred second by non-distiller because only small quantity of the wood is enough for different kinds of tasks. It can also be spitted to any required size, for fast ignition and flammability. The basic reason A. seyal, which is preferred by the same number of non-distillers as A. etbaica, has the characteristics of very fast ignition and flammability. Even though it is said to produce little embers after flame combustion, it produces enough heat for preparing different kinds of meals. It also releases little smoke, fast drying rate and ease of splitability. On the other hand, for katikala distillers using this species need more fuelwood to distil katikala although it has the above advantages.

Collectively acacia species were the most frequently mentioned and preferred fuelwood plants by consumers. Most plants exploited for fuelwood by distillers possess a woody growth habit and are trees that mostly grow wild and are indigenous. The impact of plant harvesting on species viability will depend on whether the harvesting is destructive or not (Kinyanjui, 1987). The harvesting of whole plants, as is the case with demand from katikala brewer, is considered destructive because it creates a larger negative impact on the survival of the individual and population than the harvesting of renewable parts like leaves or branches. Meanwhile, non-distillers use of fuelwood around the town appears to be sustainable due to their reliance on dry parts and branches of tree in addition to smaller amount of consumption for their day to day activity relative to distillers.

The influence of harvesting on species depends also on the growth habit of the species in question. Most of the plants vulnerable to high preference and harvesting rates in the study area are hardwood and indigenous one. Many woody plants take a long time to reach maturity and according to Solberg (1988) indigenous trees usually grow more slowly than exotics. Using old growth, hardwood species as fuel is not sustainable, particularly for such kinds of local industry, given the large amount of wood consumed. The main effects of unsustainable plant consumption practices are assumed to be a change in the vegetation structure, and in species composition.

34 4.1.2 Most frequently used Species

Species preference do not necessarily mean the most commonly used and/or available species. Most of distillers are often forced to use whatever is available at local market, including those from most preferred to less preferred in fuelwood quality. Since most preferred species are already in short supply, particularly for distillers, this rules out free choice or preference. The types of wood species used most frequently in the households are, as shown in the following table 4.3, Eucalyptus species for distillers and dry parts and branches of tree for non-distillers. The reasons are availability and relatively low cost. Eucalyptus species is supplied to all local market except Mobile by an association of micro and small enterprise; they are organized and licensed by the municipality of Arsi-Negele town to solve fuelwood scarcity especially for katikala production.

Table 4.3 Most frequently used fuelwood species by households.

Species % Households % Distillers % Non-Distillers sampled sampled sampled

Eucalyptus species 20 20.8 14.3 dry parts and branches of tree 18.2 11.8 61.9

A. tortillis 14.5 16 4.8

Podocarpus falcatus 12.1 13.9 -

Balanites aegyptica 6.1 5.6 9.5

A. senegal 6.1 6.9 -

R. natalanis 5.5 6.3 -

A. seyal 3.6 3.5 4.8 Junipers procera and Cupressus lustanica 5.5 4.2 -

Other types of species 10.3 11.1 4.8

35 This study has demonstrated that a preference for a species of distillers does not necessarily coincide with what fuelwood for katikala distillation used most frequently which are less preferred one. The shift of household woodfuel use from the more preferred to the less preferred one suggests that the most preferred species are already not readily available.

4.2 Fuelwood Supply and Source

Almost all of the households purchased woodfuel from the four local fuelwood markets (Board, Tureta, Atena-Tera, and Mobile) either directly from farmer or via retailer. It is known that 37% of the consumer purchase from Atena-Tera, while 28.5 % from Mobile and 20% from Tureta. The remaining 13.9 % was purchased from Board. No consumers reported ever obtaining woodfuel from surrounding forest, private farmland or communal wood lots.

Firewood selling in Arsi-Negele town ranges from a seasonal to a part time or full time year round occupation. Most of fuelwood-seller had multiple occupations. Farming was done by 72.4 % of the interviewed vendor. The quantity of fuelwood supplied by for sale varied from 30.63 - 797.12 kg per vendor with an average supplied of 162.15 kg.

Vendors have different sources of fuelwood these are Garra-Durro and around it, Around Abjaita, Shalla and Langano lakes, and out of Arsi-Negele woreda (Table 4.4). In addition, 9.5 % of vendors purchase from Kello-Durro market which is a local market found in the woreda and 3.8% of vendors are mediator (purchase the fuelwood from farmer and then resell it to consumer in the same local market).

The type of the fuelwood species varied significantly by location. From the market interview data12 species were listed within the top three most preferred species for fuelwood. From the interview there are clear local preferences for specific fuelwood species for sale. The indigenous trees are popular choices due to characteristics of fuelwood especially distillers criteria, such as heavy wood with slow burning strong fire, and long-lasting embers. Moss (1980) says, there is little preferences when the wood is

36 required for straight burning. However, when a commercial element enters the system, then selectivity and price differences begin to operate. The demands of this species lead to discriminate cutting of the individual species later on this lead to the non-existence of of specific species.

Table 4.4 Sources of fuelwood to Arsi-Negele town from different location. Garra-Durro and Around Abjaita- Around Out of Arsi-Negele around it Shalla lake Langano lake woreda Lepis, Ashoka, Adaba-tita, Shalla-Billa, Galle- Keraro, Hadhe- Adamitulu, Siraroo and Kebele Kersa – Gerra, Gambo-Ito Kello, Bulbula & Bosso, Bukuu- Alaba woreda Mudi-Arjo Walda, Goljota Podocarpus falcatus, , A.seyal, A.totillis, A. seyal, A. totillis, A. etbaica, R.natalanis,Cupressus Types of Juniperous procera, B.aegyptica, A. senegal B. aegyptica, A. lustania, A. seyal, A. Species Pygeum africanum, senegal totillis, B. aegyptica, A. supplied Croton macrostachys senegal

,Cupressus lustania, Eucalyptus species

Dodonea viscose, etc

Interviewed vendor (%) 54.3 11.5 5.7 15.3

Of those vendors who acquired their supplies by not purchasing nearly 73.6 % from surrounding vegetation, only 8.8% from their wood lot & private land. The most important sources of fuelwood were the trees found on the surrounding vegetation.

Among the various patterns (katikala brewing, domestic consumption, restaurant, retailer or mediator) of woodfuel use, the vendors who sell most of the time for katikala distillers were 56.2 %( the largest), who sells for any consumer including the brewer were 35.2 % and the rest for the mediator.

Availability of the vegetation around them and demand especially from katikala distiller are the main factor that acts as the determinant of preference to the fuelwood species by vender. And the fuelwood should also be easy to gather—for example some wood with spines is difficult to gather (like Acacia senegal) and be able to lose moisture fast.

37 Woodfuel vendors are seeking to cut a sufficient amount of wood of adequate quality by selection of the preferred species. However, selective harvesting may change forest composition and ecosystem function (Struhsaker, 1997).

The overwhelming impression among the interviewed vendors was that the woody vegetation had diminished during the last few years. Almost 88.7% of the community considered katikala producer are one of the main cause for declination of vegetation as well timber producer at Gara-Duro and around it and charcoal producer around Abjaiata, Shalla and Langano lake, but 98.9 % agreed that vegetation are declining due to the overexploitation of tree species for fuelwood demand at Arsi-Negele town for katikala brewing, timber and charcoal production in the woreda.

Despite the degradation of forest resource found in the woreda, the supply of fuelwood has remained. This is due mainly to the higher demand of fuelwood in the town for katikala production. The average daily supply of fuelwood and means of transportation of each category of all the different local market have been shown in table 4. 5 which was done by counting at four local markets throughout a week.

Table 4.5 Daily Supply of Fuelwood to Arsi-Negele town (Mean (+ SD)) Market Place Category /Unit Board Tureta Mobile Atena-Tera Total Donkey load 356.86 (+171.5) 769.29 (+243.25) 992.86 (+122.8) 1335.86 (+406.6) 3475.87 Donkey drawn cart 114.86 (+12.24) 67.57 (+14.41) 65 (+5.66) 460.43 (+549.70) 707.86 Horse load 0.00 4.00 (+2.31) 0.00 66.14 (+19.26) 70.14

Woman load 23.00 (+5.39) 11.29 (+3.20) 1.14 (+.90) 24.86 (+10.79) 60.29 Man load 0.57 (+0.79) 0.00 .43 (+.54) 0.00 1

Car (3.6 tons, ISZUZU) 1.00 (+1.53) .71 (+.76) 1.14 (+1.68) .71 (+.95) 3.56 Association(m3)* 47.43 (+14.50) 53.86 (+11.10) 0.00 71.71 (+10.80) 173 *Association of micro &small enterprise (M3) are organized to supply fuelwood mostly eucalyptus species .

38 Woodfuel are widely used in the town. These fuels are supplied from the surrounding rural residences, retailers and small enterprises and are used for a variety of household sector such as katikla production process, preparation of meals etc. It is found out that 96 % of the supplied fuelwood is consumed by katikala distillers.

Commercial harvesting for fuelwood in the study area overrides ecological impacts from all other harvesting purposes because of powerful economic incentives, the preferences of species and the need of large amount of fuelwood used to produce katikala.

4.3 Impact on Vegetation

In general the areas in which vegetation sample were taken are low in tree species diversity, and only four woody species are encountered at the sampling plot. These were Acacia tortillis (A. tortillis), Acacia senegal (A. senegal), Acacia seyal (A. seyal) and Balanite aegypticus (B. aegypticus), they are listed in the following Table 4.6 which shows density and frequency of seedlings/saplings and matures trees, respectively. The tables exhibit numbers as density per hectares. Generally density was higher for the seedlings and saplings together than for mature trees in each village. In all the villages A. totillis had the highest percentage and significant number of mature tree (Table 4.6). Seedlings and saplings density A. totillis of also had higher percentage except for the park headquarter (Table 4.8).

Table 4.6 Density of trees (SS= seedlings+sapling, M= mature trees) per hectare with in villages and around park headquarter. Galle-Kello Shalla-Billa Keraro Park Headquarter Species SS M SS M SS M SS M A. tortillis 178 28 278 92 48 78 44 156 A. senegal 181 31 134 22 0 0 210 12 A. seyal 9 6 6 7 0 0 0 0 B. aegyptica 11 6 3 20 45 11 28 14 Total mean 379 71 421 141 93 89 282 182

There were two main forms of wood harvesting chopping and root digging in the study area observed. The residences around the vegetation dig out the chops for production charcoal after cutting the plant for different purpose. The main influenced factor for this activity is financial need at Negele town.

At the same time there are trends/evidences of discriminate cutting during household survey due to preference of specific species in the previous few years which have resulted in more damage of some species than others. No A. etbaica and R. natalensis encountered in the sampling plots. However, Beals (1968) studied the species composition of the woodland and stocking density of individual species occurring in the study area. The study revealed 124.5 stems/ha stocking density for A. etbaica and 84 stems/ha for that of R. natalensis. It was found that A. etbaica, was the most preferred tree species followed by R. natalensis for production of katikala. Distillation of katikala in this area might be contributing for the extinction of these species. If the present extensive fuelwood utilization of the woodland continues it is likely that A. seyal the next species that will be highly affected which is the third preferred species in this area.

Table 4.7 Size class distribution of trees comparison of species between the plots in the villages and the park headquarter.

Villages Park Headquarter

Seedlings Saplings Tree Seedlings Saplings Tree Species No SD % No SD % No SD % No % No % No % A. tortillis 143 8 55.4 25 2 62.5 66 3 67.4 6 4.2 38 27.3 156 85.7 A.senegal 93 6 36.1 12 1 30 18 1 18.4 115 80.4 95 68.4 12 6.6 A. seyal 5 1 1.9 0 0 0 2 .3 2.0 0 0 0 0 0 0.0 B. aegyptica 17 1 6.6 3 1 7.5 12 1 12.2 22 15.4 6 4.3 14 7.7 Sum 258 100 40 100 98 100 143 100 139 182 100

The distribution of saplings and seedlings did not correspond to the distribution of tree species, and varied between sites. Tree regeneration (seedling + sapling) was greatest at Galle-kello and Shalla-Billa, decreased at the park headquarter and very low at Keraro (Table 4.6). The number of regenerating tree species was greatest at Galle-Kello and Shalla-Billa. During the study period both areas were under greatest biotic pressure of fuelwood demand at Negele town, therefore, new juveniles were regenerating. The choice of the residence Langano lake water use for their livestock, seedling and sapling at Keraro is affected by the effect of cattle trampling and grazing the way to the lake. As indicated by Mohammed Abdi (1993) livestock impact limits the natural regeneration of the existing woodland in this area. The abundance of seedling and sapling in the woodland reflect that there were less pressure on these than on the trees. The abundance of older trees headquarter the park headquarter reflect that there were less pressure, which indicates that humans are responsible for the thinning of the woodland, because there are no large wildlife species dependent on these mature trees. This suggests that past harvesting activities have reduced mature tree density in the freely accessible areas.

Density differences within size class groups of the same species are shown in the following table 4.8. Numbers per hectare in descending order were seedlings, mature tree and sapling; but A. tortillis of mature tree had the highest number around the parks headquarter. As shown in the following table 78 % of A. tortillis was in mature stage and only 3 % was in seedling stage around the park headquarter plots. The fact that it is found that number of seedling is more than sapling show that the capability of regeneration and replacement is low in the study area.

41 Table 4.8 Size class distribution of trees within species comparison between the villages and Park Headquarter.

The plots Control

Seedlings Saplings Mature tee Seedlings Saplings Mature Species tee No % No % No % No % No % No % A. tortillis 143 61.1 25 10.7 66 28.2 6 3 38 19 156 78 A.senegal 93 75.6 12 9.8 18 14.6 115 51.8 95 42.8 12 5.4 A. seyal 5 71.4 0 0 2 28.6 0 0 0 0 0 0 B. aegyptica 17 53.1 3 9.4 12 37.5 22 52.4 6 14.3 14 33.3

Comparison of the intensity of tree cutting among the different village (Table 4.9) revealed that Galle-Kello was the most affected by cutting. Since the residence around the vegetation dig out the chops for production of charcoal it is difficult to interpret intensity of cutting by counting chops. However, it was understood that the recent dramatic increase in wood cutting in Galle-Kello. This is due to the weakening of keeping by militia of the Abijiata-Shalla Lake National Park. People tended to use the wood even at the cost of violating the keepers, with the intention described by Brown (1973) ‘if I do not cut it down some one else will.’

Table 4.9 Number, number per hectare and frequency of trees cut in the villages.

Village Number Number/hectare Frequency (%) Galle-Kello 68 189 49 Shalla-Billa 35 97 25 Keraru 37 103 26

Most of the resident engaged their life by distilling this local beverage may be due to the availability and proximity of vegetation around the town as well accessible to the main roads to the capital city, Addis Ababa. Moreover, as fuelwood harvesting forms the most important forms of the woodlands in the Arsi- Negele particularly for the production of

42 katikala have significant contribution for degradation of vegetation around the town because the need of about 69.83 kg/day which is from the household study.

The low number of mature tree encountered in the study area. The main effects of unsustainable plant consumption practices are assumed to be a change in the vegetation structure from woodland to bushland, and in preference for fuelwood at the near by town Negele may contribute changes vegetation species composition of the study area. And specific species like A. etbaica and R. natalensis which preferred by katikala distillers were extinct from the study site. The major purpose for cutting tree in the study area was for fuelwood sale at Arsi-Negele town due to high demand especially for katikala production and charcoal production. The four big market of fuewood trade observed in the Negele town may be explained by the high demand and purchasing power of the people living there most of them are katikala distillers. The preferred species declines on a local scale, even the less favored one also decrease both in abundance and distribution. In the study site due to the scarcity and absence of trees leads indiscriminate cutting of the existing species.

4.3.1 Impact on Different Species

The overall structural pattern of the woodland revealed that all sites are mainly dominated with A. tortillis and A. Senegal. When densities for individual species are considered differences are evident (tables 4.6, 4.7 and 4.8).Considering differences among the individual species, A. tortilis was dominant although it is preferred by greater number of consumer than A. Senegal. A. tortillis is a shade intolerant species; the mature tree must die before regeneration and subsequent recruitment can occur at any given site (Vesy-Fitzgrald, 1973 and Mualyosi, 1990). This shows that low productivity of the closed canopy of A. tortillis woodland and the opening of the canopy may allow regeneration. The success of A. tortillis also lies in it being highly drought resistance, more so than all other species in the study area; it grows in area with annual precipitation of as low as 50 mm (Maydell, 1990), while the other need more precipitation.

43 A. senegal was the next notable species in the study area (Table 4.2 & 4.3). Frequency of mature tree is the next to A. tortillis in the study area. Density of seedling and saplings was the second in the villages; while it was the first in the park headquarter. A. senegal seedling and sapling flourished better in the park headquarter However, the frequency of mature trees was much lower compared to A. tortilis. The low number at maturity may be because it is less drought resistant, high demand for charcoal and for sale at Negele town compared to its existing strand, and may need shed at early stage.

B. aegyptica was one of the least common species in the sample area (Table 4.2 & 4.3). Even the sapling and seedling not occurred in significant number. Occurrence Balanite aegyptica were highest at Shalla-Billa. Despite B. aegyptica being ecologically a very flexible species, for example drought resistant, undamaged by grass fires, growing in a variety of soils and in most topographical features (Mohammed Abdi, 1993), it is one of the least in number only more than that of A. seyal which is going to extinct in the study area. They are better flourish in the park headquarter. Seedling of this species is lower may be due to it is being browsed and fed by livestock.

A. seyal was the least in number of all species. Only 2 mature trees, 5 seedlings were recorded. The low number of A. seyal seedling, sapling, and tree suggests may be the specie preferable by the local people around the vegetation especially for a lucrative business of fuelwood sale at Negele market. Because A. seyal was the least prefer species for charcoal production compared with A. tortillis, A. senegal and B. aegyptica (Nazereth Fikru, 1997) although it is the one of the most preferable species specified by distillers. The total absence of A. seyal at the control plots shows that there may be other factors operating than the human factor. A. seyals said to prevail more in clayish soils and requires more water than A. senegal (Rahaman, (undated) cited in Mohammed Abdi, 1993). However, according to Mohammed Abdi (1993) finding the number of A. seyal was more than B. aegyptica before fourteen years ago.

44 4.4 Impact on Indoor Air Quality

This study attempted to assess the impact of katikalla distillation process on indoor air condition. The values of daily fuelwood use rate and other environmental factors showed wide ranges. These factors were analyzed to determine if there was any systematic variation of indoor air pollution among households due to these factors. Table 4.10 shows the means and standard deviations of the variables over all households.

Table 4.10: Wood fuel Use Rate & Other Factors

Minimum Maximum Mean Std. Deviation

48.00 107.00 70.6774 17.88181 Woodfuel use rate (kg/day)

20.31 31.72 26.0055 2.64686 Indoor temperature (°C) during distillation process

27.61 132.66 68.5781 27.69896 Kitchen volume (m3)

Number of stove in a 6.00 12.00 8.0323 1.64284 kitchen

Katikala produced 26.00 64.00 44.2258 10.35281 (liter/day)

For each parameter of interest i.e. average concentration during the whole distillation process, an 8-hour average concentration, the maximum observed 15-minute concentration of CO and PM were analyzed against these variables. The variables included kitchen volume daily, woodfuel use rate, volume of kitchen, mean indoor temperature, katikala produced per day and number of stove in a kitchen.

4.4.1 Average Pollutant Concentration

Table 4.11 shows average CO (ppm) and PM (mg/m3) during distillation process and 8- hour concentration, and histograms of the means for all households are shown in figures 4.1 and 4.2.

45 Table 4.11: Mean Pollutant Concentrations during distillation process.

Minimum Maximum Mean Std. Deviation

Average CO concentration 6.50 149.29 68.8065 31.64718 (ppm)

8-hour CO concentration 6.31 168.60 75.2465 34.35 (ppm)

Average PM concentration .73 5.86 3.1142 1.54644 (mg/m3 )

8-hour PM concentration .54 52.66 5.5677 9.23124 (mg/m3)

The minimum concentration of CO encountered during distillation process was 6.5 ppm, the kitchen in which this value recorded had good ventilation, with volume of 117.65 m3 , the wall of it is made of wood, the eaves between the wall and the roof is greater than 30 cm, the two windows and the door of the kitchen is opened throughout the distillation process. In addition the concentration of PM for this kitchen was .80 mg/m3, which is comparable with the minimum concentration which is .73 mg/m3. On the other hand, the maximum concentration of CO recorded kitchen had volume of 41.02 m3 made of mud, no eaves and window, the door is partially opened during the process.

6 Frequency

Mean =68.81 Std. Dev. =31.647 N =31 0 25.00 50.00 75.00 100.00 125.00 150.00 Average CO concentration (ppm) Figure 4.1 Histogram of mean concentration of CO (ppm) during distillation process.

6 Frequency 4

Mean =3.11 Std. Dev. =1.546 N =31 0 1.00 2.00 3.00 4.00 5.00 6.00 7.00 Average PM concentration (mg/m3 )

Figure 4.2 Histogram of mean concentration of PM (mg/m3) during distillation process

Further tests to search for the relationship between pollutant concentration and different factors were done using multiple regressions, which accounted for possible correlation between measured household characteristics. The multiple regression covariates included daily woodfuel use rate, kitchen volume, mean indoor temperature during distillation process, number of stove in a kitchen and katikala produced.

The multiple regression for average CO concentration show that daily firewood use rate, mean indoor temperature during distillation process, number of stove in a kitchen and daily Katikala production were statistically significant predictors of increased CO concentration during distillation process. Kitchen volume was shown to be not significant predictors of CO concentration (Table 4.12).

The multiple regression for average PM concentration show that no covariates were shown to be significant predictors of PM concentration during the process (Table 4.13).

47 Environmental factors and fuelwood consumption rate have played a significant role in IAQ (indoor air quality) variation. Both CO and PM showed a tendency to increase with factors value increase except for kitchen volume for PM.

Table 4.12: Average CO concentration (ppm): Multiple Regression Estimates (N = 31, R2 = .763)

Standard Covariates Estimate error t Sig.

-136.128 40.776 -3.338 .003 (Constant)

.594 .310 1.912 .067 Wood use rate (kg/day)

Mean indoor temperature 3.663 1.579 2.319 .029 (°C) during distillation process

.094 .123 .763 .453 Kitchen volume (m3)

Number of stove in a 4.896 3.555 1.377 .181 kitchen

Katikala produced .496 .511 .971 .341 (litre/day)

Table 4.13: Average PM Concentration (mg/m3): Multiple Regression Estimates (N = 31, R2 =.502)

Standard Covariates Estimate error t Sig.

-3.342 2.886 -1.158 .258 (Constant)

.019 .022 .886 .384 Wood use rate (kg/day)

Mean indoor temperature .075 .112 .673 .507 (°C) during distillation process

-.001 .009 -.145 .886 Kitchen volume (m3) Number of stove in a kitchen .209 .252 .833 .413

Katikala produced .035 .036 .956 .348 (litre/day)

48 Kitchen volume shows a tendency to increase CO and decrease PM level of the kitchen but neither showed statistically significant relations in either bivariate or multivariate analysis. However according to (Cowlin et al., 2005) for CO, increased kitchen volume could result in decreased concentrations because the pollutant mass would be distributed in a greater air volume and the same dilution effect would exist for PM. In this case changes in kitchen ventilation like the type of walls and roofs in which the kitchen made of, the eaves between the walls and roofs providing gaps between them, the numbers of doors and windows and whether those window and door are opened or closed during the monitoring process are important factors to drive (venting) the bulk of combustion byproducts to the outdoor environment.

In addition an increased volume would result in a decreased surface to volume ratio which leads to deposition on surfaces that is a major mechanism of particle removal from an indoor environment; also an increase in volume provide the opportunity increase the average distance a particle has to travel before it can be deposited, thus increasing a particle’s atmospheric lifetime. This leads the pollutant to remain in the atmosphere of larger volume kitchen for a long time.

4.4.2 Maximum Pollutant Concentrations

The maximum 15-minute concentrations experienced during the distillation time were calculated for both CO and PM. In 19 households, the 15-minute maximum PM concentration occurred between 07:00 and 12:00; due to high frequency refueling of wood at the morning. In 2 households, this maximum occurred between12:00 and 14:00; in 8 households, this maximum occurred between 14:00 and 16:00; and in the remaining 2 households, this maximum occurred between 16:00 and 17:00. Table 4:14 shows the Maximum 15-Minute Average Pollutant Concentrations and Histograms of 15-minute maximum CO and PM are shown in figures 4.3 and 4.4.

49 Table 4.14: The Maximum 15-Minute Average Pollutant Concentrations

Minimum Maximum Mean Std. Deviation

15-minute maximum CO 15.52 244.33 115.3116 48.90565 concentration (ppm)

15-minute maximum PM 2.54 27.04 12.3134 6.37113 concentration (mg/m3)

8 Frequency

2 Mean =115.31 Std. Dev. =48.906 N =31

0 42.00 84.00 126.00 168.00 210.00 252.00 15-minute maximum CO concentration (ppm) Figure 4.3 Histogram of Maximum 15-Minute Average CO (ppm) Concentrations

6 Frequency 4

Mean =12.31 Std. Dev. =6.371 N =31 0 0.00 5.00 10.00 15.00 20.00 25.00 30.00 15-minute maximum PM concentration (mg/m3) Figure 4.4 maximum 15-minute average pm (mg/m3) concentrations

Multiple regressions were run testing the effect on 15 minute maximum CO and PM concentration from daily woodfuel use rate, kitchen volume, mean indoor temperature, during distillation process, number of stove in a kitchen and amount of katikala produced. Estimates of covariates from the multiple regression models and their significance are shown in tables 4.15 and 4.16.

Wood use (kg/day), mean indoor temperature (°C) during distillation process and number of stove in a kitchen were statistically significant predictors of maximum 15-minute CO concentration. An increase in stove number by one corresponded to about 11.4 ppm CO increase and a mean indoor temperature increase by 1°C corresponded to a 6 ppm CO. Most of the covariates were not statistically significant predictors of maximum 15-minute PM concentration except number of stove in a kitchen.

Table 4.15: 15-minute maximum CO concentration (ppm): Multiple Regression Estimates (N = 31, R2 = .616)

Covariates Estimate Standard error t Sig.

-182.285 81.535 -2.236 .034 (Constant)

.595 .626 .951 .351 Wood use rate (kg/day)

Mean indoor temperature 6.006 3.059 1.964 .060 (°C) during distillation process

.105 .238 .441 .663 Kitchen volume (m3)

Number of stove in a 11.373 6.392 1.779 .087 kitchen

51 Table 4.16: 15-minute maximum PM concentration (mg/m3): Multiple Regression Estimates (N = 31, R2 = .439)

Covariates Estimate Standard error t Sig.

-11.540 12.840 -.899 .377 (Constant)

.109 .099 1.109 .277 Wood use rate (kg/day)

Mean indoor temperature .256 .482 .532 .599 (°C) during distillation process

.000 .037 -.007 .994 Kitchen volume (m3)

Number of stove in a 1.161 1.007 1.153 .259 kitchen

Based on available knowledge of fuel combustion and air pollution and particulate exposures would arguably provide the best indicator of pollutant risks from fuel combustion. The WHO health based standard guideline values for 90 (ppm) for 15- minutes, and 10 ppm for 8-hours which is less than the recorded value during the monitoring process by about 25.31 (ppm) and 65.25 (ppm). No guideline value for average concentrations PM is recommended according to WHO (1999) at Geneva, evidence from epidemiological studies even at low levels of PM exposure encountered adverse effects on human health. According to the study conducted by (Hong. 1997) the range of risk was found to be 1.2 - 4.4% increased mortality per 10 mg/m3 incremental increase in concentration of respirable suspended particles (PM10). But, for example acceptable short-term PM exposure range for residential indoor air quality of Canada is < = 1 mg/m3 and acceptable long-term PM exposure range is < = 0.4 mg/m3, which are an order of magnitude less than from the measured one. The concentration of CO and PM during distillation process are greater than standard guidelines set by different organization.

52 4.4.3 The Relationship of Carbon Monoxide and Suspended Particulate Matter

The study has tried to check whether it might be possible to take CO as a surrogate indicator of particulate matter. High correlation has been reported by (Smith et al., 1998, Naeher et al., 1996, Park and Lee, 2003) between carbon monoxide and suspended particulate concentrations, indicating the possibility of using the former as an inexpensive marker for the more costly to measure suspended particulate matter.

To see if this study is in line with such arguments attempt was made to analyze the correlation of CO concentration collected by HOBO CO data logger and PM collected by UCB Particle monitor. Positive correlation has been found between carbon monoxide and suspended particulate concentrations, even though there is variation in values of CO and PM with Pearson Correlation coefficient of 0.724. The results of study show strong correlations between daily average concentrations of PM and CO.

A simple linear model which can be used for the prediction of suspended particulate matter concentration from that of carbon monoxide is developed as follows,

[PM] = 0 .679 + 0.035 [CO] with correlation coefficient R2 = 0.525, N=31 (Figure 4.5) Where [PM] and [CO] are the concentrations of particulate matter and carbon monoxide, respectively.

53 6.00

Observed 5.00 Linear

4.00

0.03539351680299277 * x + 0.6788912470714967

3.00

2.00 Average PM concentration (mg/m3) Average PM concentration

1.00

0.00 0.00 25.00 50.00 75.00 100.00 125.00 150.00 Average CO concentration (ppm) Figure 4.5 The relationship between CO (ppm) and PM (mg/m3)

These findings are consistent with those of (Ezzati et al., 2000), who looked closely at the potential of using of CO as a proxy for suspended particulate concentrations.

So this study is in line with such arguments and also checked these scientific arguments were reflected in the data collected process of this specific study. Even though the results of study show strong correlations between daily average concentrations of PM and CO, there is another problem with using carbon monoxide as the indicator of suspended particulate matter an exposure assessment using time-integrated perspective of a single kitchen. As Figure 4.6 illustrates, the instantaneous concentrations of the two pollutants for one of the households show weak correlation with Pearson correlation 0.126. Sampling took place between the hours of 6:54 in the morning and 19:09 in the evening over 1-min intervals. There is no significant relation between temperature, CO and PM.

The lack of correlation (R less than or = 0.26 for all households) quantitatively confirms the suspicion raised in (Naeher et al., 1996) and observed in one of the trials of (Ballard-

54 Tremeer and Jawurek ,1996) that the temporal relationship is highly dependent on the burning status of the fire.

CO (ppm) Temprature PM (mg/m3)

120 20

100 16

14 80

60 10

6 PM (mg/m3) Concentration

CO (ppm) Concentration and Temprature (oC) 20

0 0 06:54:00 07:15:00 07:36:00 07:57:00 08:18:00 08:39:00 09:00:00 09:21:00 09:42:00 10:03:00 10:24:00 10:45:00 11:06:00 11:27:00 11:48:00 12:09:00 12:30:00 12:51:00 13:12:00 13:33:00 13:54:00 14:15:00 14:36:00 14:57:00 15:18:00 15:39:00 16:00:00 16:21:00 16:42:00 17:03:00 17:24:00 17:45:00 18:06:00 18:27:00 18:48:00 19:09:00

Time

Figure 4.6 Concentration of CO and suspended PM for a single day for a representative household

4.4.4 Pollutant Concentration and Fuelwood Use Rate

The impact of reported fuelwood use on indoor air quality was examined by constructing regressions of CO and PM concentrations against fuel use rate for each of the time periods of interest; during the whole time of distillation process, 15-minute maximum, and 8-hour indoor pollutant. The bivariate regression estimates of fuelwood use impact

55 on CO concentrations are shown in table 5.17. Each correlation was statistically significant for wood use rate.

The largest influence of woodfuel use on CO was seen in the 15-minute maximum concentration; an increase in woodfuel use of 1 kg/day corresponded to a 1.79 ppm 15- minute maximum CO increase. During the distillation process average CO concentration, an increase in woodfuel use of 1 kg/day corresponded to a 1.49 ppm CO increase, and an increase in wood use of 1 kg/day corresponded to a 1.55 ppm 8-hour average CO increase. Similar effect can be seen for the average concentration during distillation. For these periods as well, an increase in wood use of 1 kg/day corresponded to a 1.49 ppm CO increase

Table 4.17: Fuel Use Impact on Indoor CO Concentrations: Bivariate Regression Estimates

Dependent Estimate Standard t Sig. R2 variable error

Average CO (ppm) during 1.485 .200 7.411 .000 .654 distillation process 15-minute maximum 1.790 .408 4.382 .000 .398 CO (ppm)

8-hour CO (ppm) 1.550 .232 6.677 .000 .606

Average PM concentration during distillation process, a 15-minute maximum and an 8- hour concentrations were also done against wood use rate. The bivariate regression estimates of fuel use impact on PM concentrations are shown in table 5.18.

None of the correlations between fuel use rates and PM were statistically significant, although all correlations between fuel use and PM concentrations were positive. As with CO, the biggest influence of fuel use on PM concentrations was seen in the 15-minute maximum values. For this interval, an increase in 1 kg/day of wooodfuel use corresponded to an increase of 0.197 mg/m3 PM. Increased fuel use of 1 kg/day

56 corresponded to an increase of 0.055 mg/m3 average PM during distillation process, and increased wood use of 1 kg/day corresponded to an increase of 0.113 mg/m3 8-hour average PM.

Table 4.18: Fuel Use Impact on Indoor PM Concentrations: Bivariate Regression Estimates Standard Dependent variable Estimate error t Sig. R2

Average PM 0.055 0.013 4.222 0.000 0.381 (mg/m3) during distillation process 15-minute maximum 0.197 0.058 3.403 0.002 0.285 PM (mg/m3 ) . 8-hour morning 0.113 0.097 1.164 0.254 0.045 refuel PM (mg/m3 )

Further tests to search for statistically significant differences in fuel use rate among stoves and other factors were done using multiple regression, which accounted for possible correlation between measured household characteristics (Table 4.19).

57 Table 4.19: Wood Use Rate: Multiple Regression Estimates (N = 31, R2 = .771)

Standard Covariates Estimate error t Sig.

-51.351 22.495 -2.283 .031 (Constant)

Number of stove in a 6.184 1.746 3.542 .002 kitchen

Katikala produced .419 .295 1.418 .168 (litre/day)

.029 .072 .404 .690 Kitchen volume (m3)

Mean indoor temperature (°C) 2.049 .845 2.425 .023 during distillation process

In this study, woodfuel use rate (kg/day) was consistently and significantly associated with number of stove. Fuelwood consumption was higher by 6.18 kg/day in households with an increase of one stove. It is known that number of stove was associated with fuelwood consumption rate, and fuelwood consumption rate was associated with IAQ, one might expect a relationship between stove number and IAQ.

58 5. CONCLUSION AND RECOMMENDATION

5.1 Conclusion

The impact of katikala distilation with natural vegetation seems to be evident from the present study. At present, most of the natural vegetation has changed into farmlands and degraded lands which are claimed as the major threat. Collection of firewood for sale by cutting of live trees and household consumption contribute for degradation of vegetation in the study area.

The most preferred species according to distillers preference for fuelwood use were A. etbaica followed by R. natalenis and Podocarpus gracilior ,and then A. seyal,. Due to the scarcity of these preferred species, consumers use the species that are available at the local market.

Old-growth indigenous species tree are becoming increasingly rare because of harvest patterns to satisfy the demand of distiller fuelwood preference by vendor. The first two preferred species, A. etbaica and R. natalenis, are non-existence in the study area. The fourth preferred species by distillers, A. seyal, is low in density in the study area.

Katikala distiller are being exposed to levels of CO and PM are higher than WHO guideline, which are set to protect against detrimental health effects from chronic exposure. It follows that these residents are being regularly exposed to unsafe levels of both pollutants.

CO monitoring will remain a useful tool in qualitative estimation of households in terms of emission levels. However, since small size monitoring devices like the one used in this study are available and episodic nature of exposure to indoor smoke is of critical importance, the usefulness of CO as a means for detailed instantaneous assessment of exposure to suspended particulate matter is limited.

59 The indoor air pollution level in the kitchen increase with amount of fuelwood used, average indoor temperature, number of stove in a kitchen, amount of katikala produced and there is no detectable interaction with kitchen volume in this particular study.

5.2 Recommendation

Introduce new forms of environmentally benign alternative energy sources of energy to substitute for fuelwood.

Fuel needs of the town need to be supplemented with eucalyptus species. Using eucalyptus for fuelwood supply may be strengthened as a means of meeting distiller requirements and taking pressure off natural vegetation this will give the natural vegetation the chance to recover.

The links between exposure to indoor air pollutant and different adverse effects on human health should be extensively studied due to the production of Katikala production process.

Traditional wood burning stoves that are used for Katikala distillation should be improved upon significantly and inexpensively by using advanced combustion efficiency. Improvements in this stove can be either through venting by use of flues or hoods and/or improvements in stove combustion efficiency. As like other types of stove, katikala distillation stove should be designed and introduced to reduce the emissions of indoor air pollutants and the quantities of fuel-wood used, that truly benefit the quality of the distiller life.

60 REFERENCE

Achmadi, U.F. (1992). Health Aspects of Biomass Fuel Use in Households - the Indonesian Experience, in Indoor Air Pollution From Biomass Fuel - Working papers from a WHO Consultation, World Health Organisation, Geneva: 3-12. ADAA (Africa Development Aid Association), (2004). Deforestation around Duro Mountain. Ahuja, D.R, Joshi, V., Smith, K.R., Venkataraman, C. (1987). Thermal Performance and Emission Characteristics of Unvented Biomass-burning Cookstoves: A Proposed Standard Method for Evaluation, in Biomass, Vol. 12, Elsevier Applied Science Publishers, Great Britain: 247-270. Akinbami, J., Salami, A., and Siyanbola, W. (2003). An Integrated Strategy for Sustainable Forest - Energy - Environment Interactions in Nigeria. Journal of Environmental Management 29:115-118. Albalak, R.(1997). Cultural practices and exposure to particles pollution from indoor biomass cooking: effects on respiratory health and nutrtitional status among the Aymara Indians of the Bolivian Highlands. University of Michigan Alemneh Dejene .(2003). The Nexus of Natural Resource Degradation, Food Security and Poverty in the Ethiopian Highlands. In: Towards Sustainable Agriculture and Rural Development in the Ethiopian Highlands, pp 13-16 (FAO). FAO Corporate Document Repository. Rome. Anderson, D. (1986). Declining Tree Stocks in African Countries, World Development, Vol 14:7. Bailis, R., Ezzati, M., Kammen, D. (2005). Mortality and greenhouse gas impacts of biomass and petroleum energy futures in Africa. Science 308: 98–103. Ballard-Tremeer, G. and Jawurek, H. H. (1996). Biomass Bioenergy. pp 419-430. Beals, E.W. (1968). Birds of a Euphorbia-Acacia Woodland in Ethiopia: Habitat and Seasonal changes. Journals of Animal Ecology, Vol. 39: 277-297. Behera, D. (1995). Health effects of indoor air pollution due to domestic cooking fuels. Indian J. Chest Dis Allied Sci 37: 237. Behera, D., Dash S., and Malik, S. (1988). Blood carboxyhaemoglobin levels following acute exposure to smoke of biomass fuel. Indian Journal of Medical Research, 88: 522–542.

61 Bekele Melaku. (2003). Forest property rights, the Role of the State, and Institutional Exigency: The Ethiopian experience. Bownder, B. (1987). Environmental Problems in Developing Countries. Progress in Physical Geography 2: 246-259. Bridgewater, J. (1991). An over view of thermal chemical biomass conversion, pp 73-105, In: Fuel for Thought, Richards, G.E., Department of Energy, UK. Brown, L.H. (1973). Conservation for Survival Ethiopia’s Choice Haile Sellassie I United Printers, Ltd. Addis Ababa. Brown, L.H. and Cocheme, J. (1969). A study of the agroclimatology of the highlands of Eastern Africa. FAO/UNESCO/. Bruce, N., Perez-Padilla, R., and Albalak, R. (2000). Indoor air pollution in developing countries: a major environmental and public health challenge. Bulletin of the World Health Organization, 78:1078–1092. Bruce, N. (1999). Lowering exposure of children to indoor air pollution to prevent ARI: the need for information and action. Arlington, Environmental Health Project, (Capsule Report 3). Burrows, C.J. (1993). Processes of Vegetation Change. UNWIN HYMAN, London. Cecelsi, E., Ddunkerley, J. and Ramsey, W. (1979).”Household Energy Resource for the future and the Poor in the Third World.” Report R-15, Washington, D.C China Development Gateway, (2006). Ecological Modernization Cowlin, S. C., Rachel B. K., Rufus E. and Smith, K. R. (2005). “Impact of Improved Stoves on Indoor Air Quality in Ulaanbaatar, Mongolia”. Energy Sector Management Assistance program (ESMAP) of World Bank CSA, (2005). Annual Statistical Report . Addis Ababa, Ethiopia. Crutzen, P.J., Andreae, M.O. (1990). Biomass burning in the tropics: impact on atmospheric chemistry and biogeochemical cycles. Science 250: 1669_1678. Cunningham, A.B. (1988). Development of a conservation policy on commercially exploited medicinal plants: a case study from southern Africa. In:Akerele O., Heywood, V., and Synge H. (1991). Conservation of medicinal plants. Cambridge, UK. pp. 337–54. Dagnachew Legesse, Vallet-Coulomb, C. and Gasse, F. (2003). Hydrological response of a catchment and land use change in tropical Africa: Case study south central Ethiopia. Journal of Hydrology 275: 67-85.

62 De Koning, H.W., Smith K.R. and Last, J.M. (1985). Biomass fuel combustion and health. Bulletin of the World Health Organization, 63: 11–26. Deweees, P. A. (1989). The Woodfuel Crisis Reconsidered: Observations on the Dynamics of Abudance and Scarcity. World Development, 8: 1159-72. Earl, D.E. (1975). Forest Energy and Economic Development, Oxford, UK, Clarendon pres. EARO, (2002). Terms of Reference for the 2002 Earo-Icra Field Study at Arsi Negele East Shoa Zone, Oromiya Region, Ethiopia. EFAP (Ethiopia Forestry Action Programme), (1993). Draft Final Report, Volume II.The challenge for Development. Draft Final Report. Addis Ababa, Ethiopia. 78pp. Ehrlich, P.(1988). The Loss of Diversity: Causes and Consequences. Washington, D.C.: National Academy Press, pp 21-27. Engel, P., Hurtado, E. and Ruel, M.(1998). Smoke exposure of women and young children in highland Guatemala: predictions and recall accuracy. Human Organisation, 54: 408–417. Eriksson, I., Teketay, D. and Granström, A. (2003). Response of plant communities to fire in an acacia woodland and a dry Afromontane forest, southern Ethiopia. Forest Ecolology and Management 177: 39-50. Ezzati, M. and Kammen, D. (2002). Household energy, indoor air pollution and public health: research and policy needs in developing countries. Annual Review of Energy and the Environment, 27:1–38. Ezzati, M., Bernard, M. Mbinda and Kammen, D. M. (2000). Comparisons of Emissions and Residential Exposure from Traditional and Improved Cookstoves in Kenya. American Chemical Society, 34: 578 – 583. FAO, (1983). Woodfuel Survey: Forestry for local community development programme, GCP/INT 7365/SWE:Rome. Geist, H.J., and Lambin, E.F. (2001). What Drives Deforestation? A Meta-analysis and Underlying Causes of Deforestation based on Subnational Case Study Evidence. Ciaco Printshop, Louvaine-la Neuve. Giampietro, M. (1994). Models of energy analysis to assess the performances of food systems. Agricultural Systems , 45:25 Groves, A. T., Street, F.A. and Goudie, A. S. (1975). Former Lake levels and climae changes in the rift valley of southern Ethiopia. The Geographical Jornal.141 (2):177-202.

63 Hillman, J. C. (1993). Ethiopia: Compendium of Wildlife Conservation Information. NYZS - The Wildlife Conservation Society - International, New York Zoological Park, Bronx, NY and Ethiopian Wildlife Conservation Organisation, Addis Ababa, Vol. 2:332 pp. Hong, C.J., Corvalan, C. and Kjellstrom, T. (1997). Air Pollution. In: Hong, C.J.(1997). Quantifying Global Health Risks: The Burden of Diseases Attributed to Selected Factors. Hurni, H. (1990). Degradation and conservation of the soil resources in the Ethiopian highlands. In: Messerli, B. and Hurni, H. (1990). African Mountains and Highlands: Problems and Perspectives, pp. 51-64. Marceline, Missouri (USA). IPCC (Inter-governmental Panel on Climate Change), (1997). In: Houghton, J.T. (1996). Guidelines for National Greenhouse Gas Inventories: Reference Manual. Bracknell, UK. IPCC (Inter-governmental Panel on Climate Change), (1990). Climate Change:The IPCC Scientific Assessment. Cambridge University Press, UK. Kaale, B.k. (1985). Highlights of Energy in East Africa, pp 29-34 IN: Siren, G. and Mitchel, C.P. (1985).Forest Ecology and Management and fuelwood crisis. Proceeding of IUFRO projects group P1.09.00 Meeting in Uppsala. Kinyanjui, M. Fuelling N. (1987). The importance of small-scale charcoaling enterprises. Unasylva; 39(3, 4):17–28. Krugmann, H. (1989). Review of Issues and Research Relating to Improved Cookstoves, in Canadian Journal of Development Studies, Vol. 10 (1):121-133. Canada. Lemlem Sisay. (2003). Biodiversity potentials and threats to the southern Rift Valley lakes of Ethiopia, Wetlands of Ethiopia, IUCN - The World Conservation Union. Lioy, P.J. (1990). Assessing total human exposure to contaminants: a multidisciplinary approach. Environmental Science and Techonology, 7: 938–945 Lundgren, B. (1971). Soil studies in a montane forest in Ethiopia. Research Notes Nr11. Department of Forest ecology and Forest Soils, Royal College of Forestry, Stockholm. Mariame, A. (1996). Appraisal of wood fuel supply and demand, a case for Shashemane town, South Ethiopia. MSc. SLU, Sweden, pp. 31. Maydell, H.J. (1990). Trees and Shrub Species of the Sahel Their Charecterstics and Uses. Ministry of Agriculture (1996). Woody biomass inventory for strategic planning, Biomass conversion factors for some selected units of measurement.

64 Mohammed Abdi. (1993). Impact of human activities on Abijata-Shalla lake National Park Ecosystem, Msc Thesis, Agricultural University of Norway, NURAGRIC, AAS. Mooney, H.A., Godron, M. (1983). Disturbance and Ecosystems: Components of Response. Springer-Verlag, New York Moss, R.P. (1980). The ecological and agricultural aspect of rural energy resources. Mualyosi, R.B. (1990). The dynamic ecology of Acacia tortillis woodland Manayara Natinal Park, Tanzania. Africa Journal of Ecology,28:189-199 Murray, C.J., Lopez, A. (1997). Global mortality, disability, and the contribution of risk factors: Global Burden of Disease Study. Lancet, 349: 1436–1442. Naeher, L. P., Leaderer, B. P.,Smith, K. R., Grajeda, R., Neufeld, L.,Mage, D., Boleij, J. S. (1996).The proceeding of Indoor Air: The 7th International Conference on Indoor Quality and Climate; Nagoya, Japan, Nazereth Fikru (1997) Sudy of Farmers Species Preference and Evaluation of Some Fuelwood Qualities: A case Sudy in Abijaita-Shalla Lakes National Park in Arsi Negele District, Ethiopia. Newcomb, K. (1984). An economic justification for rural aforestation: the case of Ethiopia. Energy development Paper No. 16. World Bank Energy Department. The World Bank, Washington D.C. Norboo, T. (1991). Domestic pollution and respiratory illness in a Himalayan village. International Journal of Epidemiology, 20: 749–757. ORS (The Oromia Regional State government), (2004). Socio Economic Profile of East Shoa Zone. Osei, W.Y. (1993). Woodfuel and deforestation—answers for a sustainable environment. Journal of Environmental Management 37: 51–62. Parikka, M. (2004). Global biomass fuel resources. Biomass and Bioenergy 27: 613–620. Patel, T.S. and Raiyani, C.V.(1995). Indoor air quality: Problems and perspectives. In: Shukla. P.R. (1995).Energy Strategies and Green House Gas Mitigation. Allied Publishers, New Delhi, pp.72. Park, E. and Lee, K. (2003). Particulate Exposure and Size Distribution from Wood Burning Stoves in Costa Rica. Indoor Air, 13: 253 – 259. Blackwell Munksgaard. Picchi-Sermolli, R.C.(1975). Geo-botany of Africa: Ethiopia and Somalia 13:15-132

65 Ramakrishna, J. (1992). Preliminary results of the EWC/FWD global survey of improved cookstove programs, in Indoor Air Pollution From Biomass Fuel - Working papers from a WHO Consultation, World Health Organisation, Geneva: pp.137-148. Sayer, J. ,Harcourt, C., and Collins, N. (1992). The Conservation Atlas of Tropical Forests. McMillan Publishers Ltd. Uk. Selinus R., (1971). The traditional foods of the Central Ethiopian highlands; research report No. 7. Scandinavian Institute of African Studies. Shelton, J. (1982). Wood Heating System Design Conflicts and Possible Resolutions, in Proceedings: Cooper, J.A. and Malek, D. (1981). International Conference on Residential Solid Fuels, Environmental Impacts and Solutions, Oregon Graduate Center, Oregon, USA: 873-880. Smith, K.R. (1999). The national burden of disease from indoor air pollution in India. In: Raw, G., Aizlewood, C., Warren, P. (1996). The 8th International Conference on Indoor Air Quality and Climate, Edinburgh. London, Construction Research Ltd, pp13–18. Smith, K.R.(1996). Indoor air pollution in developing countries: Growing evidence of its role in the global disease burden. In: Ikeda, K. and Iwata, T. (1995). Proceedings of 7th International Conference on IndoorAir and Climate. Institute of Public Health, Tokyo, pp.33. Smith, K.R. (1993). Fuel combustion, air pollution exposure, and health: the situation in developing countries. Annual Revision , Energy and Environment, 18: 529±566. Smith, K.R (1992). Biomass cookstoves in global perspectives: Energy, Health, and Global Warming, in Indoor Air Pollution from Biomass Fuel - Working papers from a WHO Consultation, World Health Organisation, Geneva: 165-186. Smith, K.R.(1991). Managing the risk transmission. Toxicol and Health 7: 319, Smith, K.R. (1990). Indoor air quality and the population transition. In: Kasuga, H. (undated). Indoor Air Quality. Springer Verlag, Berlin, pp.448, Smith, K.R. (1987). Biomass fuels, air pollution and health. A global review. New York, Plenum Press. Smith, K. R . and McCracken, J. P. (1998) Environment , pp 739-747.

66 Solberg, K. H.(1988). Fuelwood in agroforestry systems. In:Vedeld, P.(1988). Bioenergy and land use in Africa. NORAGRIC occasional papers. Series C, development and environment, vol. 2:73-93. Norway. Struhsaker, T.T. (1997). Ecology of an African Rain Forest. University Press of Florida, Gainesville, Florida. Turner, G.M., Romme,W.H., Gardener, R.H., O’Neill, R.V., Krtz, T.K. (1993). A revised concept of landscape equilibrium: disturbance and stability of on scaled landscapes. Landscape Ecology 8: 213–227. UNEP (2000). Global Environment Outlook . Earthscan Publication; pp. 52-3, 98-9, 120-22. USEPA (1997). Revisions to the National Ambient Air Quality Standards for Particles Matter. 62: 38651–38701. Vesey-Fitzgerald, D.F. (1973). Animal impact on the vegetation and plant succession in lake Manyara National Park, Tanzania. Oikos 24:314-325 White, P.S., and Pickett, S.T.(1985). Natural disturbance and patch dynamics: an introduction. In: White, P.S and Pickett, S.T., (1983), The Ecology of Natural Disturbance and Patch Dynamics. Wiley-Interscience, New York, TJ, pp. 1–13. WHO (2004). Global Status Report on Alcohol. World Health Organization Department of Mental Health and Substance Abuse, Geneva. WHO (2000). Addressing the Links between Indoor Air Pollution, Household Energy and Human Health, World Health Organisation, Washington, DC. WHO (1999). Air Quality Guidelines for Air Quality management, Geneva.

WHO (1997). Air Quality Guidelines for Europe. Regional Office, Copenhagen, Series No.23, WHO (1992). Epidemiological, Social and Technical Aspects of Indoor Air Pollution from Biomass Fuel, World Health Organisation, Geneva. WHO, IUCN, and WWF (1993). Guidelines on the Conservation of Medicinal Plants. IUCN, Gland, Switzerland. World Bank .(1992). World Development Report t and the Environment .Newyork : Oxford University Press. World Bank .(1984). Human Development Report, Oxford University Press. WRI (World Resources Institute), (1997). The last frontier Forests: Ecosystems and Economies on the edge. Washington D.C., USA.

67 Wynder, E.L. (1979) Forum: Workshop on carbon monoxide and cardiovascular disease. Prev Med 8: 261. Zartarian, V.G., Wayne, O.R., Duan, N., (1997). A quantitative definition of exposure and related concepts. J. Exposure Anal. Environ. Epidemiol. 7:411±437. Zelalem Amde, (2006). Land Use/ Land Cover Dynamics and Vegetation Vulnerability Analysis: a Case Study of Arsi Negele. MSc. Thesis, Addis Ababa University, Department Earth Science. Zerihun Woldu, Mesresha Fetene and Asferachew Abate. (1999). Vegetation under different tree species in Acacia woodland in the rift valley of Ethiopia. SINET: Ethiop. J. Sci. 22:235- 252 Faculty of Science Addis Ababa University. Zerihun Woldu, and Mesfin Tadesse. (1990). The status of the vegetation in the lake regions of the Rift Valley of Ethiopia and the possibilities of its recovery. SINET: Ethiop. J. Sci. 13:97-120.

68 ANNEX

Annex 1: House Hold Survey Questionnaire

General Information Questionnaire number_____

1. Wereda: ______Locality ______Kebele ______House No______2. Name of household (HH) head: ______Age______; Sex______3. Name of respondent (if different from HH head) ______; Age______; Sex____ occupation______4. Family size ______5. Interviewer Name______Signiture______

The House Hold Fuelwood Consumption Questionnaire.

1. Have you been brewing katikala? ( Yes/ No). 2. If your answer is yes, for how many years? A) 5 B) 8 C) 10 D) 12 E) Other (specify)______3. Where do you get firewood? A) Market B) Forest (surrounding vegetation) C) Farmland (private farmland) D) Communal wood lots E) other (specify):______4. Answer of Q-3 is (A) where do you buy your firewood most frequently? A) Board, B) Tureta, C) Mobile, D) Atena-Tera, E) Other (specify) ______

5. What types of wood species do you use most frequently in your home? A) Dodoti, Dare (Acacia etbaica) B) Wachu, White Galed Acacia (Acacia seyal) wako dimo wasiya

69 C) Tedecha, Deweni Grar (Acacia tortillas) lotoba D) Sabas, Kontir, Gum Arabic, Etan Zaf (Acacia senegal), E) Bedenno Desert Date (Balanites aegytiaca), F) Debobosso, Chakema ( Rhus natalanis),tatesa G) Tid (Junipers procera and Cupressus lustanica) H) Zigba (Podocarpus gracilior) I) Bahrzaf (Eucalyptus species) J) Dry parts and branches of tree K) Others (specify) ______6. What is your preferred wood species? A) Dodoti, Dare (Acacia etbaica) B) Wachu, White Galed Acacia (Acacia seyal) wako dimo wasiya C) Tedecha, Deweni Grar (Acacia tortillas) lotoba D) Sabas, Kontir, Gum Arabic, Etan Zaf (Acacia senegal), E) Bedenno Desert Date (Balanites aegytiaca), F) Debobosso, Chakema ( Rhus natalanis),tatesa G) Tid (Junipers procera and Cupressus lustanica) H) Zigba (Podocarpus gracilior) I) Bahrzaf (Eucalyptus species) J) Dry parts and branches of tree K) Others (specify) ______

7. What is the most important reason for choosing your specific fuelwood? A) Availability B) Low cost, C) Quality (Mention), if quality______D) Others (specify) ______3 8. How many (m , donkey drawn cart, horse drawn cart, donkey load, woman load, man load….) of wood do you burn per day? ______

70 3 9. How many (m , donkey drawn cart, horse drawn cart, donkey load, woman load, man load….) of wood did you burn last week for all uses? (From Sunday to Sunday)

______10. If you brew katikala, how many liters of it do you distil per day? A) 32 B) 40 C) 48 D) 52 E) other (specify)______

11. Do you distil katikala throughout the week? ( Yes / No ).

12. If no, how many days of the week? A. 6 B. 5 C. 4 D. 3 E. other (specify) ______

THANK YOU FOR YOUR COOPERATION AND PATIENCE Date of interviewing______Other comments (if any):______

71 Annex 2: Market Questionnaire

Questionnaire number_____

1. Wereda: ______Locality______Kebele______House No______2. Name of respondent ______; Age______; Sex ____occupation ______3. Family size______4. Interviewer Name______Signiture______

1. How long have you been selling firewood? Give the number of years. A) 2 B) 3 C) 5 D) 8 E) Other (specify) ______2. To whom do you sell fuelwood? A) Katikala brewer B) Business owner (restaurant, hotel…..) C) Domestic consumption only D) Any body E) Other (specify) ______3 3. How much fuelwood totally do you supply (in m , donkey drawn cart, horse drawn cart, donkey load, woman load, man load….) per day?

3 4. Can you give an estimate (m , donkey drawn cart, horse drawn cart, donkey load, woman load, man load….) of how much fuelwood you sell to a single katikala brewer? ______3 5. Can you give an estimate (m , donkey drawn cart, horse drawn cart, donkey load, woman load, man load….) of how much fuelwood you sell to a single domestic consumer? ______

72 6. What locality does your fuelwood come from? ______

7. From where is your fuelwood collected? A) Forest (surrounding vegetation) B) Farmland (private farmland) C) Communal wood lots D) If other (specify) ______8. Do you know the common species of trees you normally sell as fuelwood? ( Yes / No ) 9. If yes, can you please mention them? I) To katikala brewer______

II) To domestic consumer______10. Are there tree species that you prefer to sell? ( Yes/ No ) If yes, mention the top three______11. What is the most important reason for choosing your specific fuelwood species to sell? ______12. What is the most important reason for choosing specific fuelwood species by the consumer? ______

13. Do you have any problems in obtaining adequate fuelwood to sell? ( yes , No ) 14. If yes, what are they? ______15. Have you noticed any changes in the vegetation pattern? ( Yes/ ,No ) If yes, what What do you think the cause or who are the agent?

THANK YOU FOR YOUR COOPERATION AND PATIENCE Date of interviewing______Other comments (if any):______

73 Annex 3: Market supply counting format (Tally Method)

Market place ______

Name of counter______

Donkey Horse Donkey Horse Car * Woman Man load load load drawn cart drawn cart load

If car, type ______

74 Annex 4: Vegetation Data Collection Format

Kebele ______Locality (Village) ______Transect No ______Plot No______GPS Position ______Altitude______Forward Bearing______Backward Bearing______Slope______Land Use Type______Number of Trees ______Number of Seedlings ______Number of Stumps ______

Remark No Species Name Local Name Height Diameter at (Scientific) Breast Height (DBH)

- - -

75 Annex 5: HOBO CO Sampling Data Form

76 Annex 6: UCB Particle Monitor Sampling Data Form

77 Annex 7: Indoor Air Pollution: Post-recording Questionnaire

General information Household ID on Master ______Kebele ______House Number ______Name of household (HH) head: ______Age______; Sex______Name of respondent (if different from HH head) ______; Family size ______Age______; Sex____ Interviewer Name______Signiture______1. (Observation) what types of walls does the kitchen have? A) Wood and mud B) porous wood C) Other (specify)______

2. (Observation) what types of roofs does the kitchen have? A) Sheet metal B) thatch C) plastic with wood tiles D) Other (specify) ______

3. (Observation) are there open eves between the walls and roof of the kitchen area? A) Yes, less than 30 cm B) Yes, greater than 30 cm C) No D) Other (specify) ______

4. (Measurement) what is the length (longer dimension), ______meter, width (shorter dimension) ______meter, height ______meter of the kitchen. 5. (Measurement) how much fuelwood was used for distillation during the measurement period. ______kg.

6. What was the type of wood species used? ______Please comment on the quality of fuelwood you used in. A. high moisture (low moisture) B) low moisture (less smoky) C) Other (specify) ______

7. During the distillation process, how many liters of katikala produced? ______liters.

8. Was there anything unusual about your stove use pattern? If , yes______

78 9. Was there any other combustion inside or near by the kitchen, influencing the pollution levels? If yes how long did this other combustion last______?

10. How many doors are in the kitchen?

11. Were the doors in the main kitchen open during the monitoring period? A) No, not open at all B) Yes, partially open about half the time C) Yes, partially open about half the time D) Yes, fully open about half of the time E) Yes, fully open almost all of the time

12. How many windows are in the kitchen? ______

13. Were the windows in the main kitchen open during the monitoring period? A) No, not open at all Bntila) yes, partially open about half the time C) Yes, partially open about half the time D) yes, fully open about half of the time E) Yes, fully open almost all of the time

14. How is the ventilation in the main kitchen? A) Very good B) good C) average D) fair E) poor

15. Did any of the following weather conditions occur during the measurement period? A) no rain B) drizzle (light rain) C) rain D) other (specify)______

16. How windy was it during the measurement? A) no wind (calm) B) mild wind C) high wind

17. Were there any disturbances to equipments during the monitoring equipment? (Yes, No) If yes, describe______