1 CLIMATE CHANGE, LIVE LIHOOD AND HOUSEHOLD VULNER ABILITY IN EASTERN NIGER By SARAH LINDLEY MCKUNE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012
2 2012 Sarah Lindley McKune
3 To my colleagues, team, and friends in Tanout
4 ACKNOWLEDGMENTS I offer my sincere thanks to my doctoral committee co chairs, Drs. Brian Mayer and Sa ndra Russo, and members, Drs. Leonardo Villalon and Alyson Young. Without you, this research would not have happened in this form. I am additionally grateful to Dr. Sandra Russo who roped me into this PhD while I was floundering in an attempt to live ha ppily in Gainesville and work in US public health happily. Her support, guidance and friendship have served me immeasurably. With Dr. Mark Brown and the Adaptive Management of Water, Wetlands, and Watershed (AMW3) IGERT team, she helped me find a home, bot h professionally and personally. I thank the National Science Foundation and the entire AMW3 IGERT team for that opportunity and for their financial support of my doctoral program. The Center for African Studies and the Foreign Language Area Studies (FLAS ) Fellowship program, funded through the US Department of Education, provided financial support of my doctoral program and allowed me to study Arabic for two years, a lifelong goal fulfilled. I would like to thank the Livestock and Climate Change Collabo rative Research Support Program (LCC CRSP) for providing funding for my fieldwork in Niger, and for the encouragement and guidance of Drs. Julie Silva and Leo Villalon concerning how to conduct research in a volatile pocket of the African Sahel while raisi ng two small children. My heartfelt thanks must be expressed to my friends and research assistants, Drs. Sambo Bod and Malam Souley Bassirou, who enabled data collection when and where it might not otherwise have been possible. I learned immensely from th em both, and I hope this document is a reflection of that.
5 I will be ever grateful to my husband, J., and my kids, Matilda (2.5 years) and at the end were instrumental to its completion. And though only concepts in my mind at the time, Matilda and James were the reason I embarked on this journey; and they are certainly my motivation to complete it.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURE S ................................ ................................ ................................ ........ 11 LIST OF ABBREVIATIONS ................................ ................................ ........................... 12 ABSTRACT ................................ ................................ ................................ ................... 14 CHAPTER 1 OPENING REMARKS ................................ ................................ ............................. 16 National Context ................................ ................................ ................................ ..... 16 Introduction to Niger ................................ ................................ ......................... 16 Economics ................................ ................................ ................................ ........ 17 Environment ................................ ................................ ................................ ..... 20 Population and Health ................................ ................................ ...................... 23 People ................................ ................................ ................................ .............. 24 Ethnicity, Livelihood, and Water ................................ ................................ ....... 27 Research Questions ................................ ................................ ............................... 29 Research Design ................................ ................................ ................................ .... 31 Research Area and Sample ................................ ................................ ............. 31 Data Collection ................................ ................................ ................................ 33 Data Analysis ................................ ................................ ................................ ... 36 2 PASTORALISTS UNDER PRESSURE D OUBLE EXPOSURE TO ECONOMIC AND ENVIRONMENTAL CHANGE IN NIGER ................................ ... 37 Introduction ................................ ................................ ................................ ............. 37 Background: Pastoralists in the Contemporary Eco nomy ................................ ....... 39 Methodology ................................ ................................ ................................ ........... 43 The Case of Niger ................................ ................................ ................................ ... 45 Feedback Double Expo sure: Uranium Markets ................................ ................ 48 Context Double Exposure: Land Use Change, Markets, and Household Demographics ................................ ................................ ............................... 52 Concluding Remarks ................................ ................................ ............................... 59 3 PERCEIVED RISK OF CLIMATE CHANGE, ADAPTATION AND LIVELIHOOD VULNERABILITY IN EASTERN NIGER ................................ ................................ 62 Introduction ................................ ................................ ................................ ............. 62 Background ................................ ................................ ................................ ............. 65
7 Climate Change in Niger ................................ ................................ .................. 65 Pastoralism as a Sustainable Livelihood ................................ .......................... 66 Pastoralism and Vulnerability/Resilience ................................ .......................... 67 Research Methods ................................ ................................ ................................ .. 69 Key Variables ................................ ................................ ................................ .......... 71 Livelihood ................................ ................................ ................................ ......... 72 Perceived Risk of Climate Change ................................ ................................ ... 73 Vulnerability/Resilience ................................ ................................ .................... 74 Food Security ................................ ................................ ................................ ... 74 Adaptations/Coping Mechanisms ................................ ................................ ..... 75 Wealth ................................ ................................ ................................ .............. 76 Ethnicity ................................ ................................ ................................ ............ 77 Findings ................................ ................................ ................................ .................. 77 Percepti ons of Climate Change and Livelihood ................................ ................ 77 Perceptions of Climate Change, Coping Strategies, and Adaptations .............. 79 Coping Strategies, Adaptations, and Household Vulnerability/Resilience ........ 84 Discussion ................................ ................................ ................................ .............. 89 4 UNDERNUTRITION AND FOOD SECURITY IN NIGER A STUDY OF AG ROPASTORAL COMMUNITIES FOLLOWING THE 2005 AND 2010 FOOD CRISES ................................ ................................ ................................ .................. 95 Introduction ................................ ................................ ................................ ............. 95 Background ................................ ................................ ................................ ............. 96 Niger and Nutrition ................................ ................................ ........................... 96 Causes of Malnutrition ................................ ................................ ...................... 97 2005 Crisis ................................ ................................ ................................ ..... 102 2010 Crisis ................................ ................................ ................................ ..... 104 Methods ................................ ................................ ................................ ................ 105 Findings ................................ ................................ ................................ ................ 110 Distribution of Undernutrition 2005 and 2010 ................................ ................. 110 Case study: Dareram ................................ ................................ ............... 116 Case study: Kkeni ................................ ................................ .................. 119 Food Security and Undernutrition ................................ ................................ ... 122 Wealth ................................ ................................ ................................ ...... 127 Food security, HHH age, and ethnicity ................................ ..................... 127 Food security and livelihood ................................ ................................ ..... 128 Ethnicity ................................ ................................ ................................ ... 129 Discussion ................................ ................................ ................................ ............ 130 5 MOVING FORWARD: IMPLICATIONS OF RESEARCH FINDINGS .................... 136 Paper Review ................................ ................................ ................................ 136 Livelihood Index Development ................................ ................................ ....... 137 Factors Effecting Pastoral Vulnerability ................................ .......................... 139 The Need for Interdisciplinary Collabor ation ................................ ................... 141
8 APPENDIX A KEY INFORMANT INTERVIEW INSTRUMENT ................................ ................... 143 B FOCUS GROUP INSTRUMENT ................................ ................................ ........... 145 C HOUSEHOLD QUESTIONNAIRE INSTRUMENT ................................ ................ 153 D ANTHROPOMETRIC DATA COLLECTION INSTRUMENT ................................ 166 E METHODOLOGY FOR CREATION OF LIVELIHOOD INDEX ............................. 171 LIST OF REFERENCES ................................ ................................ ............................. 175 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 190
9 LIST OF T ABLES Table page 2 1 Livestock Loss and Coping Strategies Employed during 2005 and 2010 F ood Crises by Community ................................ ................................ ......................... 56 3 1 Nested Multivari ate Regression Models Examining the Effects of Demographic Characteristics, Livelihood, Wealth and Coping Strategies on Perceived Risk of Climate Change. ................................ ................................ .... 81 3 2 Nested Multivariate Regression Models Examining the Effects of Demographic Characteristics, Wealth and Coping Strategies on Perceived Risk of Climate Change Among Pastoral Households. ................................ ....... 83 3 3 Nested Multivariate Regression Models Examining the Effects of Demographic Characteristics, Wealth and Coping Strategies on Perceived Risk of Climate Change Among Agricultural Households. ................................ .. 85 3 4 Nested Multivariate Regression Models Examining the Effects of Demographic Cha racteristics, Coping Strategies and Livelihood on Self reported Resilience, All Households. ................................ ................................ .. 86 3 5 Nested Multivariate Regression Models Examining the Effects of Demographic Characteristics, Wealth and Coping Strategies on Self reported Resilience Among Pastoral Households. ................................ ............................ 88 3 6 Nested Multivariate Regression Models Examining the Effects of Demographic Characteristics and Coping Strategies on Self reported Resilience, among Agricultural Households ................................ ....................... 89 3 7 Percent of household heads with any education and mean household size by ethnicity. ................................ ................................ ................................ ............. 92 3 8 Comparison of coping strategies correlated with PRCC and vulnerability/resilience by livelihood ................................ ................................ .... 93 4 1 Sample Size for 2005 and 2010 Household Survey and Anthropometric Data 110 4 2 Descriptive statistics of sampled children, 2005 (N=200) and 2010 (N=290) ... 111 4 3 Percent of total* and severe** stunting (HAZ), wasting (WAZ), and underweight (WHZ) by community, 2005 ................................ ......................... 112 4 4 Weight for height z scores by community, 2005 ................................ ............... 112 4 5 Percent of total and severe stunting (HA Z), wasting (WAZ), and underweight (WHZ) by community, 2010 ................................ ................................ .............. 114
10 4 6 Weight for height z scores by community, 2010 ................................ ............... 114 4 7 Comparison of undernutrition (mean WHZ) between 2005 and 2010 by community ................................ ................................ ................................ ........ 115 4 8 Change in mean community undernutrition and self reported resilience .......... 116 4 9 Mean WHZ between food security groups ................................ ........................ 123 4 10 One way ANOVA tests to compare mean WHZ by food security and other predictor variables ................................ ................................ ............................ 1 24 4 11 Nested Multivariate Regress ion Models Examining the Effects of Demographic and Food Security Characteristics on Undernutrition in Children 6 60 months. ................................ ................................ ...................... 125 4 12 Nested Multivariate Regression Models Examining the Effects of Demographic Characteristics and Food Aid on Undernutrition in Children 6 60 months ................................ ................................ ................................ ......... 126 4 13 Linear regression model of undernutrition on wealth and demographic characteristics ................................ ................................ ................................ ... 127 4 14 Linear regression model of undernutrition on food security, head of household age, and interaction term ................................ ................................ 128 4 15 Linear regression model of undernutrition on food security, head of household age, and interaction term ................................ ................................ 128 4 16 Mean WHZ by ethnicity, 2010 data ................................ ................................ ... 129 4 17 Linear Regression model of WHZ on ethnicity ................................ .................. 129
11 LIST OF FIGURE S Figure page 1 1 Livelihood/food zones of Niger ................................ ................................ ........... 17 1 2 Five year moving averages of August rainfall at selected locations in Niger ...... 22 1 3 Evolution of number of days with annual mean minimum temperature greater ................................ ........ 23 2 1 Research Sites in Tanout District Niger ................................ .............................. 44 3 1 Map of Niger ................................ ................................ ................................ ....... 65 3 2 Map of Research Area in Eastern Niger ................................ ............................. 71 3 3 Agricultural pastoral Livelihood Continuum ................................ ........................ 72 3 4 Univariate Distribution of Key Variables ................................ ............................. 76 3 5 Distribution of Research Communities Along the Agricultural pastoral Livelihood Continuum by Average Household Livelihood ................................ ... 79 5 1 Outbreaks of Locust, drought, political u prising, and famine against the backdrop of changing rainfall and population growth ................................ ........ 140
12 LIST OF ABBREVIATION S AE Adult equivalent AQIM Al Qaeda in the Maghreb BARA Bureau for Applied Research in Anthropology BRC British Red Cross CA CDC Centers for Disease Control and Prevention DE Double Exposure FAO Food and Agriculture Organization FCFA Franc Communaut Financire Africaine (currency used in Niger and throughout much of West Africa) FEWSNET Famine Early Warning System Network GDP Gross domestic product HAZ Height for age z scores HH Household HHH Head of household IPCC Intergovernmental Panel on Climate Change IRIN Integrated Regional Information Networks (humanitarian news agency covering sub Saharan Afr ica) MS Microsoft MT Metric ton ( 1000 kg or 2204 pounds) MUAC Middle upper arm circumference PCA Principal component analysis PRCC Perceived risk of climate change RA Resident adult SPSS Statistical Package for the Social Sciences
13 TLU Tropical livestock u nit UA University of Arizona UNDP United Nations Development Programme UNICEF USAID United States Agency for International Development WAZ Weight for age z scores WHO World Health Organization WHZ Weight for height z scores
14 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CLIMATE CHANGE, LIVELIHOOD, AND HOUSEHOLD VULNERABILITY IN EASTERN NIGER By Sarah Lindley McKune August 2012 Chair: Sandra Russo Cochair: Brian Mayer Major: Interdisciplinary Ecology Climate change is projected to disproportionately affect arid and semi arid regions of the world, including the African Sahel. Niger, a Sahelian country and one of the poorest countries in the world, is home to an estimated 1.5 million pastoralists, those whose livelihood is based on livestock herding. Niger is chronically food insecure, and increasing frequency and severity of environment al shocks are testing household and community resilience. Utilizing household data collected following the 2005 and 2010 food crises, this research examines the relationship between climate change, livelihood (pastoralism/agropastoralism), and vulnerabilit y in a collection of three papers. The first investigates the role of livelihoods, coping mechanisms, and adaptations on perceived risk of climate change (PRCC). The second explores if and how coping mechanisms including migration, sedentarization, and a cceptance of food aid are affecting food security, and the predictive power of food security on undernutrition. The final paper utilizes a Double Exposure (DE) framework to analyze the impact of economic globalization and global climate change on the vul nerability of Tuareg pastoralists in Niger. As a collection, the three papers aim to enrich the current understanding of
15 household vulnerability in Niger, and allow for improved design and targeting of development programs that work to improve livelihood a nd food security in the region.
16 CHAPTER 1 OPENING REMARKS National Context Introduction to Niger The Republic of Niger is among the least developed countries in the world (United Nations Development Program (UNDP), 2011) The dire health situation throughout the country, exacerbated by widespread, absolute povert development status: 81.8% of Nigeriens live in severe poverty and life expectan cy is 44 years (UNDP, 2011) Malnutrition rates in the country indicate high prevalence of both acute and chronic malnutr ition as 39.9% of children under five are under weight for age and 54.8 % of children under five are under height for age (UNDP, 2011) More than 50% of the population lack access to appropriate water sources and health services and food security is an ongoing threat to the country (UNDP, 2007) National rates such as these dire as they may are minimize the more severe situation that exists among the poorest, most remote communities of the country. A former French colony, Niger is located in the heart of the West African Sahel, a semi arid stretch of land that runs along the southern border of the Sahara Des ert. Niger shares borders with Mali, Algeria, Libya, Chad, Nigeria, Benin and Burkina Faso. Niger is mostly flat, rarely rising 200 meters above sea level except in the Air Mountains in the north and the Djado Plateau in the northeast. Natural resources in clude uranium, coal, iron ore, tin, phosphates, gold, and petroleu m The country can be divided latitudinally into three zones: agricultural, pastoral, and desert. In Figure 1 1 below, these zones are broken down further into food zones of Niger, where the agricultural zone includes an irrigated rice zone, distinct cash crop zones, a rainfed agricultural
17 zone, and the agropastoral zone; the pastoral zone and desert are labeled individually. These zones follow rainfall levels, which decrease progressively as one moves north. Figure 1 1. Livelihood/food zones of Niger (Source: Famine Early Warning System Network [FEWSNET], 2012 ) Economics As previously stated, Niger is on e of the least developed and poorest countries in the world. Its economy is largely based on subsistence crops and livestock, though natural resource reserves namely uranium have played an important part of its economic history at various stages. In th uranium was of the most significant globally and substantially contributed to the GDP. A decline in demand and thus revenue in the late 1980s lead to a decline in uranium earnings from 22% of the GDP in 1987 to less than 16% by 1998 Recent resurgences in foreign interest in uranium bode well for the Nigerien economy; after a 20 year period of relative stability, renewed interest in alternative energy sent the global price of uranium skyrocketing in 2007, with prices jumping from $10 per pound in 2003 to a record $136 per pound by June 2007. Although prices have not remained that high,
18 uranium prices have remained above $50 per pound throughout 2011, and new exploration for uranium in Niger continues (IndexMundi, 2012) Uranium represents the highest percent of foreign exchange earnings within Niger (55.4%). Earnings from livestock sales are a distant second at 13.6%, with agricultural earnings trailing at 9.9% (U.S. Department of State, 2008) The lat ter two numbers are very difficult to measure, however, as much transport and trade occur on an informal basis, particularly of livestock, thus official numbers are expected to substantially underestimate the real value of trade in livestock and agricultur e. Countries in December 2000 In January 2006 this debt forgiveness was increased to By alleviating the burden of massive international debt incurred largely during the 1970s and 1980s, Niger was free to reallocate funds that would have previously serviced that debt to basic health services, education, infrastructur e, and other programs that aim to alleviate the widespread poverty plaguin g the country. Despite improved investments since this time, however, the situation remains dire. Niger shares a common currency (CFA franc) with six other members of the West African Monetary Union (ECOWAS). The French government supplements the internat ional reserves of the Central Bank of West African States (BCEAO), and, thus, the rate of exchange has historically been tied to the French franc (100 CFA to the French franc). This fixed rate of exchange was transferred to the Euro on January 1, 2002 (U.S. Department of State, 2008) These economic ties to France play a onomy. In 1994, France devalued the CFA, and its value in
19 Niger was cut in half overnight. This devaluation proved devastating in many ways, Anglophone neighbor to the sout h. Subsequent to the devaluation and a newly established level of price competition internationally, revenue from trade in livestock, millet, cowpeas, cotton and onions to Nigeria all increased (US Department of State's Bureau of African Affairs, 2008) e support of ECOWAS, Sahelien countries including Niger made development of regional markets a priority and set out to limit government interventions in the market and, in general, to liberalize trade within the region. Research indicates, however, tha t outcomes of these efforts have been mixed with some areas (geographic, market, and temporal) experiencing increased liberalization and decreased barriers, and others experiencing both informal and formal obstacles to entry into a free market (Beekhuis, 2007) Research has also shown that neoliberal policy reform on food production and livelihood security, which aimed to improve food security, actually increased reliance on imported foods and dramatically increased the vulnerability of urban populations in three West African countries including Niger (Moseley et al. 2010) As will be discussed in the case of the 2004/5 food cri sis in the next chapter, this mixed outcome of trade liberalization and restriction has the potential for catastrophic effects on household economics and, importantly, food security. In recent years, Niger has made strides to increase its appeal to foreign investors. In conjunction with UNDP, Niger has worked to revise codes affecting private sector
20 investment, including its investment code (1997 and 2000), petroleum code (1992 and 2000), and mining code (1993) (US Department of State's Bureau of African Affairs, 2008) With such changes, Niger hoped to better market itself to an audience of the global investors. The positive benefits of economic globalization, howe ver, have arguably not reached most parts of Niger. To date nearly half of the Nigerien substantial portion of this is in food aid. For a subsistence based economy like Niger the role of foreign donor resources and involvement is widely debated, and, as evidenced by the case of the 2005/6 food crisis (see Chapter 2), is sometimes seen as harmful to overall economic growth. Environment Covering a vast 1,267,000 km 2 of land, t wo thirds of Niger is desert and sub desert and d rought is a persistent problem. Less is permanent pastures, and 2% is forests and woodlands; an additional 70 percent of it is deser t (Geesing, 2008) Ninety four percent of Nigeriens live on 35% of the land, a nd at least 85% of the population is rural and relies on rain fed subsistence farming (IRIN, 2008) Chronic food insecurity affects 80 % of the population, and yearly 10 30 % of the populati on suffers more than a 50 % deficit in their cereal needs (BARA, 2006b) The climate limited agricultural zone is typically Sahelian ; a verage annual rainfall varies from 350 mm in the nor thern part of the zone to 600 mm in the south (Sivakumar, 1992) The average temperature in the rainy season (June September) ranges from 2 7 to 30 C and in the dry season from 20 to 45 C Relative humidity is very low, rarely exceeding a monthly average of 40 % even during the rainy season.
21 Winds are generally mild (8 20 kilometers per hour), however, high winds do occur at the beginning of the rainy season and when cooler Harmattan winds sweep off the desert from December to March. During this period dust storms are not uncommon. Rainfall is the principal determinant of agricultural production in Niger, as the amount of rainfall and its di stribution are unpredictable and variance occurs between important production thresholds. The minimum amount of rain needed to grow millet, Although w ater has never been abundantly available in this semi arid stretch of land (Bradley, 1971) changing rainfall patterns, due a t least in part to global climate change, have led to greater water scarcity in many communities scattered throughout the stark landscape (Mohamed et al. 2002) Although there w as a progressive southward shift of the 350 mm isohyet rainfall line a movement of 50 to 100 kilometers from roughly 1970 through the 1990s (Daouda, 1996a; Sivakumar, 1992) there is recent documentation of a re greening of the country in some regions (Sendzimir et al. 2011) However, t he declining rainfall levels in some areas have turned previously productive areas of the agriculture zone into food deficit zones (e.g., Filingu and Tanout). With water scarcity a defining characteristic of its semi arid zones, changing rainfall patterns and increasing temperatures both included in current climate change projections are threatening water supplies in the already fragile landscape Although there has been a significant reversal of the trend toward desertification that appeared inevitable in Niger through the 1980s, and res earchers are hopeful that this may reduce the vulnerability of the 4.5 million people in the improved area, the impact of rapid population growth and global climate change create uncertainties for the
22 extremely vulnerable population of the area (Reij, 2006; Sendzimir et al. 2011) Gl obal climate change is predicted to have significant effects in arid and semi arid regions of the world, including increased frequency and severity of extreme events, such as droughts and floods, and continued overall drying in the Sahel (Held et al. 2005; Huntingford et al. 2005; IPCC, 2007) Years of food crises in Niger appear to be increasing and periodicity of such events decreasing, as the country has experienced exceedingly high rates of childhood mortality and undernutrition in 1973, 1984, 2005, 2010, and again this year, 2012. These crises have been catalyzed by environmental shocks such as drought, pocketed rainfall shortages, and locust infestations, though the situations often become emergencies due to t he compounding effects of structural determinants including unsustainable farming practices, high rates of population growth, widespread poverty, poor infrastructure, and political instability. Overall trends of decreasing rainfall and increasing heat exac erbate the problem (see Figure 1 2 and Figure 1 3). Since the 1980s, Niger has become progressively less able to feed itself; even in good years, most rural families do not produce enough food to feed themselves for more than five or six months (Baro & Deubel, 2006) Figure 1 2. Five year moving averages of August rainfall at selected locations in Niger (Source: Sivakumar, 1992).
23 Figure 1 3 Evolution of number of days with annual mean minimum temperature Mohamed et al., 2002) Population and Health 6.1 million in 20 11 and it is projected to r each 3 0 million by 20 30 C urrent population growth is high : 3. 6 percent annually. At s every 25 years. Extremely high fertility rates are coupled with extremely high mortality rates, which act as a mitigating factor against ev en greater population growth. It is notable that 56% of the population is under 1 8 years of age, and 19.8 percent are children under five (UNICEF, 2011) Life expectan cy is approximately 4 4 years, and nearly one of every six newborns do es not live to see its 5 th birthday (UNICEF, 2011) The crowding of such a large population on to diminished arable and pastoral lands results in lower per capita food production and subsequent lower per capita food pace (6% per year), and the question arises about h ow all the non producing urban dwellers will be fed. In 2008 22 % wa s urban and was projected to increase to as much as 50 % by 2030 (UN Population Division, 2008) In 1975, Niamey had a population of 80,000 ; t oday, its population is well over one million.
24 Rapid urbanization has produced a growing group of urban poor, with associated malnutrition and food security issues. Since the 1980s, Niger has become progressively less able to feed itself and today it is one of the most food insecure countries in the world. Even in a good rainfall year, rural families who can produce enough food to feed themselves year round are rare. As noted earlier, m ost families produce only enough food to cover their needs for five or six months Diets are generally deficient in protein, calories, and essential vitamins and minerals Chronic malnutrition levels of 40 percent are recorded in many ar eas of Niger, and it is estimated that 40 percent of the rural population cannot satisfy its minimal calori c intake requirements (Baro & Deubel, 2006) child mortality is mainly due to malnutrition, Vitamin A deficiency, diarrhea, improper breastfeeding practices and poor food variety and utilization Exacerbating this situation, o nly 46 percent of the population has access to potable water (UNDP, 2007) Access to potable water drops to 35.9% among households designated as multidimens ionally poor (UNDP, 2011) As food production and arable land per capita are declining, Niger is becoming increas ingly dependent on food aid and imports Over 75 % either millet or sorghum, and such heavy reliance on mono cropping by region has heavy dietary overall health, and economic implications in such a fragile environmen t. People Two kilometer wide band that kilometer long southern border with Nigeria (IRIN, 2008) A majority of the population lives along the 300 miles of the Niger River which runs through the southwestern corner of the country and is
25 water sources Niger is made up of four major ethnic groups classified by langua ge and livelihood ; in order of magnitude, these groups include the Hausa (53%), Zarma (21%), Tuareg (11%), and Fulani (7%) (US Department of State's Bureau of African Affairs, 2008) Although ethnicity in Niger is fluid and intermarriage between ethnic groups is not uncommon, opportunities and discrimination, as well as cultural norms and household practices, are a shared experience by most within an ethnic group. As this research will show, ethnicity used as a control variable to capture a variety of differences related to the political ecology of the household is a significant predictor of self reported vulnerability and child undernutrition. Thus, a brief descri ption of each of the four major ethnic groups in Niger is included here. Although F rench is the official language of Niger, Hausa is the most widely spoken The Hausa people are predominately located in south central Niger, as the Hausaland stronghold is l ocated just south of the Niger/Nigeria border and is home to some 20 million Hausa. The Hausa maintain a very strong network that crosses not just the border with Nigeria but national borders throughout Sub Saharan Africa. They are found most often workin g as farmers, traders, or merc hants throughout the country. Because they control much of the trade with Nigeria, the Hausa in Niger are known for their economic influence and entrepreneurial spirit (Neef, 2000) The Zarma live predominately in southwestern Niger, lar gely along the Niger River and its left bank, and are descendants of the Songhai Empire, whose base was in neighboring Mali (US Department of State's Bureau of Africa n Affairs, 2008) Although historically warriors and hunters, the Zarma lived predominately as farmers an d fishermen until colonialism. Under French rule, the Zarma were among the first to
26 benefit from the colonial educati on system and have maintained a d ominant presence in central government and in civil service since independence (Neef, 2000) Zarma are also the dominant ethnic group in the capital city of Niamey and continue to hold dominance in national politics A majority of both Hausa and Zarma live as sedenta ry farmers, largely in the sou thern portion of the country. In contrast to the agricultural tradition of Hausa and Zarma households the Tuareg and the Fulani population maintain ties to their pastoralist heritage, practicing either nomadic or semi nomadic livelihoods that rely heavily upon animal husbandry. The Tuareg are of Berber (North African) de s cent, and are notoriously known as warriors and raiders of the historic trade routes that crisscrossed th e Sahara. During colonialism, the French made clear d istinction between the productive black African land holders and the nomadic Arab Berber herders who they portrayed as unproductive and landless (Marty, 1996) ods of devastating drought and widespread animal losses have forced many of the Tuareg of Niger to sedentarized (University of Saskatchewan) Because of their historically nomadic livelihood formal education has been limited among the Tuareg, despite literacy in their own scrip t (Tifinar), thus limiting economic opportunities for those who have ventured into urban areas (Hammel, 2001) However, renowned Tuareg silver and leatherwork can be found throughout Niger, as well as most of sub Saharan Africa. Throughout recent history and again in 2012, the Tuareg have rebelled against the national governments of Mali and Niger in an attempt to create an independent Tuareg state. The Tuareg are disenfranchised from Nigerien society, and the current return of
27 Tuareg immigrants to Mali and Niger from Libya with arms has fueled a new surge in violence. The Fulani are located throughout West Africa and have very dive rse cultural and social norms. The Fu lani of Niger have a pastoral history and were dominant to the Hausa and Zarma populations of Western Nige r during pre colonial periods. This power dynamic shifted under French colonial rule, as the Zarma and Hausa gained political and economic clout, while the Fulani and the Tuareg were ostracized and marginalized based on misconceptions of their landless livelihood (Neef, 2000) Nonetheless, the Fulani h ave deep connections to the Zarma and Hausa, more so than the Tuareg likely be cause of geographic proximity. Fulani who are agro pastoral have tended to settle near to Zarma or Hausa communities where complementary p roduction systems have fostered importan t trad e and cooperation, for example, milk and meat in exchange for millet (Neef, 2000) Ethnicity, Livelihood, and Water The historical distinction and relationship between sedentary agricultural populations and nomadic pastoral populations have important implica tion s for understanding access to resources as livelihoods have historically been nearly analogous to ethnic lines in Niger report on Human Right to Water pastoralists struggle to access water in Niger : ple actors working in the water sector tend to focus on water supply for the villages of the south, as water points in the pastoral north have often been the source of conflict The history of conflict in Niger has long discouraged donors from investi ng in water for marginal populations; the example here highlights the effect on one of the most sidelined groups the pastoralists, among whom water consumption (CARE USA and CARE Denmark, 2007)
28 Thus, beyond climate change and its impact on surface water availability, cultural and social processes appear to have hindered pastoral access to water when compared to their sedentary counterparts. Poor regulation and legislati ve inconsistencies also appear to encumber pastoral ist access to water. In 1993, Niger developed a Rural Code that defined access to all resources and economic activity in rural areas, including water points. Although the attempt of the Rural Code was to c larify pastoral rights and put into law that which previously was only customary law, the reality has not been entirely advantageous for pastoralists, particularly with regard to water (Lavigne Delville, 1999) The Rural Code states that pastoralists have a right to graze animals in all commonly held rangeland as well as in their home grazing territory. Home grazing territory is commonly defined as an area to which the herder returns for months at a time during the rainy season (Hammel, 2001) Within home grazing territory, a clan has priority access to water and grass, while all others are required to neg otiate access to these resources; thus common access to wells exists only on commonly held lands, where wells are less likely to exist. In addition, creation of modern wells must be associated with priority rights, which exist o nly in home grazing territo ry. This is inherently problematic given the mobile nature of pastoralists, particularly because they move in search of water, thus are more likely to be out of their home grazing territory when water is in short supply. Further complicating the situation, Niger has a Water Code, which presents national guidelines that govern the installation of water points throughout the country, which is not in alignment with guidelines set out by the Rural Code. The Water Code indicates that access to water is open to a ll, including outsiders and nomadic groups, at all times.
29 Thus, where modern wells and boreholes are being drilled, whether on home grazing territories of herders or in communities of sedentary populations, outsiders are citing the Water Code and demanding their right to water without negotiation (CARE USA and CARE Denmark, 2007) D espite historical tension between nomads and sedentary populations over land and access to water, unclear regulation and legisl ative inconsistencies manifest new tension and violent conflict around water access at wells. Research Questions Although famine was traditionally understood as a discrete consequence of external causes, it has recently begun to be reframed as a long ter m process with social and cultural drivers whose consequences are not equally distributed and exacerbate the worsening situation of the most vulnerable (Sen, 1981; Walker, 1989) An estimated 10% of the 15.6 million people living in Niger are pastoralists and live predominantly in the southwestern corner, near the Malian border, and in the northern stretches of the country (see the Pastoral Zone in Figure 1 1). The Tuareg and the Fulani, the largest ethnic groups among Nigerien pastoral populations, claim discrimination by the Hausa and the Zarma, ethnic groups who comprise a majority of the population and control much of govern ment and commerce throughout the country (Levinson, 1998) As pastoralists, the Tuareg and the Fulani are inextricably linked to the land and subject to variable weather patterns as they steer their herd s towards water and grazing lands. This research aims to investigate if and how global climate change is differentially affecting the vulnerability of agropastoral and pastoral populations in Niger. This research examine s how households in Tanout District Niger experience climate change and the relationship between livelihood, coping mechanisms adaptation and vulnerability on this experience I had the opportunity to work in this
30 area of Niger from 2005 2007. I arrived on the heels of the 2005 food cris is as part of a monitoring and evaluation team and visited 19 communities repeatedly over the course of 14 months. During this time I was privilege to conversations particularly among women and community leaders that repeatedly referenced the changes i n the natural environment. Though my background and interest at the time focused on health, these conversations piqued my curiosity and drive to understand the experience of these communities and their natural environment. The research presented here build s on baseline data colle cted as part of that project in 2005, the utilizing the 2005 and 2010 food crises as shocks around which perceptions of climate change livelihood decisions and nutritional consequences are examined. The research seek s to qualitati vely and quantitatively describe the relationship between climate change, livelihood vulnerability/resilience, and adaptation, as experienced by communities along an agricultural pastoral continuum The hypotheses of this research were: Hypothesis 1: There is a positive association between perceived risk of climate change and an increased rate of sedentariza tion as a livelihood adaptation; Hypothesis 2: There is a positive association between pastoralism and livelihood resilience, as measured by land degrad ation, vegetation cover, soil and/or crop productivity, water supply, wealt h, access to food, and mobility; and Hypothesis 3: Perceived risk of climate change disproportionally increases the vulnerability of pastoral populations, as measured by land degrad ation, vegetation cover, soil and/or crop productivity, water supply, wealth, access to food, and mobility, compared to agricultural populations, thus weakening the relationship described in Hypothesis 2. The research findings are presented in the form of three distinct papers, each written to be submitted independently for publication in a different peer reviewed journal article. The hypotheses identified above are all addressed within the second paper (Chapter 3). The specific aims of each paper are as f ollows :
31 Paper 1 (Chapter 2): Examine the vulnera bility of pastoralists in Niger by explain ing the interactions of economic globalization and global climate change in eastern Niger, and examine the consequences of those stressors and their interactions on f ood security among pastoral populations ; Paper 2 (Chapter 3): Investigate how households perceive and respond to climate change determine variation by livelihood, and identify if and how certain coping mechanisms and adaptations affect their live lihood vu lnerability/resilience; and Paper 3 (Chapter 4): Examine food in security and humanitarian food aid as risk factors associated with undernutrition in children 6 60 months and identify determinants of undernutrition for the 2005 and 2010 food crises. Becaus e I have written each of the three papers to stand alone, some information contained within each may be redundant to information contained in this introductory chapter or to the other papers. Each paper contains its own methods section, however the follow ing section is included to provide an introduction to the overall research design. Research Design Research Area and Sample This research was conducted in eastern Niger in the agropastoral and pastoral livelihood zones within the administrative department of Tanout The area was targeted for intervention by the British Red Cross (BRC) during the 200 5 food crisis, based on vulnerabilities identified by two rapid assessments conduct ed by University of Arizona (Bureau for Applied Research in Anthropology, 2005) I oversaw four phases of data collection in 19 communi ties in 2005/2006 as part of a 12 month evaluation of this project Baseline data research presented here comes from 359 households that were part of the evaluation. The 201 0 food crisis created an opportunity to understand if and how perceived climate cha nge and perceptions of risk around climate change (as manifested during two food crises) are affecting livelihood vulnerability/resilience Thus,
32 five of the six research sites selected for the 2010 research were among the larger sample of communities targ eted by the UA/BRC evaluation in 2005/2006. This research includes participants from seven communities along an agricultural pastoral continuum that ranges from fully sedentarized agricultural to fully nomadic pastoral. These sites are: Takoukout, Kkeni, Guinia Alhazaye, Dar eram, Farak, and Eliki/Djiptoji. The six s ample sites along this agricultural pastoral continuum were identified through me e t ings with local experts and researchers to identify communities among the 19 UA/BRC communities that had variou s livelihood compositions that would thus represent various positions along the continuum Although an ideal sample would have come entirely from those communities who participated in 2005, pastoral communities who practice minimal to no agricultural cult ivation were not included at that time. Thus, one additional site (Eliki/Djiptoji), identified by local pastoral researchers as meeting point ( ) where two communities of Tuareg and Wodaabe pastoral ists could be located, was included with the five UA/BRC sites so as to get representation along the continuum. Collectively, the six sites (seven communities) include communities whose range of household livelihoods include fully and historically agricultural households, newly sedentarized agricult ural households, semi nomadic pastoral households, fully and historically pastoral households, as well as some that occupy the space between. The sample for household data collection (individual household interviews and measurements of children between six months and five years of age ) is made up of the 19 households at each of five sites previously researched in 2005 and 38 new households at a pastoral for a total of 133 households Nineteen
33 households were used in each community in the 20 05 study, and those 19 households were targeted again for this research. Where loss to follow up occurred (n=5) a new household was selected so that a total of 19 households are included from each community. At Eliki/Djiptoji, 19 households were included from both the Wodaabe (Fulani) and Tuareg communities, constituting the 38 new households added to the sample For measures of child health and growth, where participant households included less than 40 total children, additional children were selected fro m household census lists (all children in rand omly selected households were measured) until data for at least 40 children were collected. Researchers conduct ed interviews at each site with key informants, identified as leaders ( chef du village or chef du t ribu ) and/or elders within the community A minimum of three interviews was conducted at each site. Focus groups were conducted among men and women (separately) in each community. Participants were selected to maximize appropriate diversity of age, ethnici ty, socio economic status, and livelihood where possible. The researcher team work ed with community leaders and census lists to identify appropriate participants. Data Collection This research employ s a comp rehensive mixed method approach including docum ent review, analysis of secondary data, and primary data collection via key informant interviews, focus groups, household interviews, anthropometric measurements and, to a lesser extent, participant observation (Bernard, 1995) I worked with two research assistants, both doctoral candidates researching pastoralists of the region who were affiliated with LASDEL in Niamey. We met in July of 2 010 to review my research design and finalize a grant proposal. I traveled to Niger again in October of
34 2010 to train the research assistants, translators, and interviewers. We pilot tested the research instruments in a community of Fulani pastoralists who had fled into Tanout during the crisis. Once instruments were finalized, the two research assistants coordinated fieldwork for the duration of data collection. All of the data collection was conducted in the primary household language of either Hausa or T amachek. Research team members were able to conduct interviews and focus groups in other languages, but the sample population did not require this. Research include d semi structured interviews with male and female (where available) key informants at each s ite (see Appendix D for key informant interview instrument) Key informants were asked to describe the history of the community, including the ethnic composition, livelihood composition, and history of migration/sedentarism Researchers use d key informants experience of the 1973, 1984, 2005 and 2010 droughts/food crises, and some of the coping strategies and adaptations employed during each. Prompting was used to investigate the role of sedentarization within the community. These i nterviews were used to investigate local manifestations of climate change, and to refine generic indicators of livelihood resilience to locally appropriate indicators of resilience for use in focus groups and household interviews Research also include d f ocus groups, conducted with men and women, separately, at each site (see Appendix C for focus group instrument). At the advice of key informants, where sites include d distinct livelihood groups or ethnic groups between which tensions were present, separate focus groups were held for each ethnic group as was the case in Eliki/Djiptoji The research includes focus groups to compare and
35 and to understand each communi of climate change and how it has changed over time. Additional research include d structured interviews, conducted with heads of household (see Appendix B for full survey instrument) Among pastoral populations, the definition of a household can be complex, as traditional methods, such as the use of a shared cooking pot or sleeping domicile, counts multiple households under one polygamous male. Thus, the definition of household used in this study is based on relation of male head of household to females, her d association, and food stores. All co wives are associated with the same household. I ndividual household interviews provide data on water, sanitation, hygiene, health, livelihood coping strategi es, food intake migration, livestock and harvest stocks, and household demogr aphic data The instrument replicate s similar data collected in November 2005 so to allow for comparison between the two points in time (Bryman, 2008) Additional questions were added, including perceptions of change in indicators of vulnerability/resilience between 2005 and 2010 Data collection also include d anthropometric measurements of all children between six months an d five years of age for each household in the sample Anthropometric scores were then calculated using data about: 1) the age of each child, ( using a local event calendar following Cobey & Cunningham, 1968; Tukei, 1963) ; 2) mid upper arm circu mference (MUAC), measured using a color coded MUAC tape; 3) length/height of children, using recumbent measuring boards; 4) weight, using hanging scales with nylon pants; 5) presence of diarrhea and/or fever ; and 6) enrollment in a
36 feeding program as repo rted by mothers Growth performance was measured using INFO software (CDC, 2008) which employs globally standardized z scores for weight for age (WAZ) height for age (HAZ) and weight for height (WHZ) The anthropometric da ta collection instrument is included in Appendix E. Data Analysis All independent and dependent variables have be en analyzed using univariate statistics to determine variation based on age, sex, ethnicity, and livelihood, as well as to determine central te ndencies and dispersion Goodness of fit has been verified for all models (R square method) in order to test the total variance explained by each model. Correlation coefficients have been analyzed to identify spurious correlations in the research analyses that follow. P rincipal components analysis has been used to develop all indices, allowing for analysis of perceived risk of climate change, livelihood, and vulnerability/resilience along a continuous scale.
37 CHAPTER 2 PASTORALISTS UNDER P RESSURE DOUBLE EXPOSURE TO ECONOMIC AND ENVIRONMENTAL CHANGE IN NIGER Introduction In order to better understand the complex systems driving human vulnerability, academics and policy makers are increasingly turning to collaborative research, where scholars from differe nt disciplinary backgrounds join forces to investigate the various dimensions of human well being, human actions, and global change. The goal of integrated science necessitates the development of frameworks that can bring together varied approaches, theori es, and methods in order to address related questions. Within interdisciplinary environments, frameworks take on analytical significance, in that they help bridge the gap between theory and applied research. In addition, they help build synergies among res earchers from different disciplinary backgrounds and open windows of opportunity for collaboration. For example, the double exposure (DE) framework (Leichenko & O'Brien, 200 8) allows multiple processes of global change to be analyzed concurrently and sequentially in order to better understand human vulnerability and possible points of intervention in the cycle. The DE framework illustrates how different global processes, s uch as environmental change and economic globalization, constantly alter the context in which individuals and communities are responding to change change stimulated, often, by one or the other processes that are occurring simultaneously or sequentially (Leichenko & O'Brien, 2008; Leichenko et al. 2010; O'Brien & Leichenko, 2000; Silva et al. 2010) This framework has been used in various contexts to reveal insights otherwise not apparent in the analysis of global Limpopo River Basin, the DE framework demonstrates how economic stressors are
38 causing small scale agriculture to be less well adapted and more vulnerable to future climatic events (Si lva et al. 2010) In India, application of the framework reveals how climate change and increased globalization et al. 2004) Adger et al. have used the framework to show how the vulnerability of coffee farmers in Mexico and Vietnam, while not geographically bounded, is intimately connected to large scale processes of global change (Adger et al. 2009) Increased understand ing of how exposure to multiple global change processes global climate change and economic globalization affects a particular geographic region or human population may improve the prospects of appropriate humanitarian aid or development intervention. T his may occur through identification of synergies in the two processes that lead to improvements in overall human well being, through a better understanding of the local variability of winners and losers, or through the identification of intervention point s that maximize impact and efficiency. By applying the DE framework to examine the vulnerability of pastoralists in Niger, this article aims to 1) explain the complex interactions between economic and environmental stressors in eastern Niger, 2) examine th e consequences of those stressors and their interactions on food security among pastoral populations, and 3) offer reflections on the framework to improve its analytic use. This article examines how economic and environmental processes are interacting to a ffect food security of pastoral populations. Secondary data from the literature on Niger and primary data collected in 2005 and 2010 for three pastoral communities in eastern Niger are analyzed within the DE framework to broaden our understanding of
39 how so me pastoralists are dealing with the interactive processes of globalization and environmental change at a local level, and their consequent direct and indirect effects on food security. The remainder is organized as follows: the literature on pastoralists and vulnerability is reviewed, including how the contemporary economic and climate change has affected their livelihood. Then the methodology used to conduct the analysis is outlined. Next the economic and environmental situation of Niger is introduced. Fi ndings from the analysis of feedback and context double exposures are then presented. The paper concludes with discussion and directions for future research. Background: Pastoralists in the Contemporary Economy Global climate change is predicted to have s ignificant effects in arid and semi arid regions of the world, especially in the African Sahel (H eld et al. 2005; Huntingford et al. 2005) a region that is also described as having very limited adaptive capacity (IPCC,2007; Reid & Vogel, 2006; Tschakert, 2007) In addition, pastoralists face increasing risk s associated with economic restructuring as they are precariously linked to the global market. Though they may seem isolated, pastoralists are significant contributors to international livestock markets (Davies & Hatfield, 2007) and are directly affected by international markets for natural resources, including uranium and oil (Keenan, 2008) Pastoralists, those whose livelihood depends upon the raising of livestock for consumption (meat and/or milk), social exchange, sale and trade, have historically been socially and politically marginalized (Hogg, 1997; Sen, 1981) In the Sahel, French colonial policies favored the farming communities of the agr icultural south over pastoral populations of the north (Dayak, 1992; Decalo, 1997; Fugelstad, 1983) By the late 1970s, pastoralists were blamed for environmental degradation, desertification, and
40 mis management of resources, as the rich, complex land use system they employ was not well understood (Fratkin, 1997; Hesse & Thebaud, 2006; Warren, 1995) Policies which promoted sedentarization, limited mobility, and privatization of lands and livestock (Hardin, 1968) which asserted that common property resources such as pastoral lands led to overgrazing and degradation of the environment (Lamprey, 1983; Lamprey, 1976; United Nations Conference on Desertification (UNCOD), 1977) Despite subsequent extensive research to indicate otherwise, many national policies throughout Africa have promoted abandonment of pastoralism as a livelihood, encouraging the settlement of pastoral people, defended by the promise of increased socioeconom ic opportunities (Campbell, 1984; Smith, 1998) Historically, distinct social adaptations (including exchanges of livestock, rest ocking alliances, dowries, traditional loan mechanisms, and support of the poor through livestock loans) have fostered high levels of resilience among pastoral populations, enabling human existence in unpredictable and otherwise uninhabited environments (Adger, 2000; Davies & Bennett, 2007; Fratkin, 2004; Friedel, 1991; Kratli, 2001; McCabe, 2007; Niamir Fuller, 2000; Thebaud & Batterbury, 2001; Westoby et al. 1989) Because of the erratic nature of rains and water supply in dryland ecosystems, pastoral systems operate in a disequilibrium state (Behnke et al. 1993) It is this characteristic of the livelihood that necessitates mobility, an essential strategy that allows pastoralists to take advantage of natural resources (e.g., water, grazing areas) that are scattered throughout an otherwise stark landscape (Behnke, 1994; Birch & Grahn, 2007; Galvin et al. 2008) Despite this resilience, however, their
41 environmentally embedded livelihood operates within fragile ecosystems, and pastoralists are increasingly vulnerable to the effects of climate change (Adger, 2003; Bohle et al. 1994; Hesse & Cotula, 2006; IPCC, 2007; Morton, 2007; White, 1991) Current climate change projections indicate that most dry lands, including the Sahel, will experience rising temperature, decreasing rainfall, and increasing number and severity of extreme weather events, including flooding and drought (Hesse & Cotula, 2006; IPCC, 2007) Pastoralists are highly vul nerable to extreme environmental fluctuations such as drought and rainfall shortages, which leave animals hungry, without water, and ill from cycles of undernutrition and disease. Severe or consecutive droughts lead to herd decimation, as seen in Niger dur ing the 1970s (Bohle et al. 1994; White, 1991) As their natural resource base becomes increasingly threatened, so too does the pastoral livelihood. Pastoralists are also vulnerable to market fluctuations. Pastoral communities have historically relied on consumption of milk and other animal products, but most do not have sufficient numbers of animals to be independent of agricultural communities; instead, they increasingly depend upon the sale of animals or animal p roducts for the purchase of grain (Hogg, 1997) Fluctuating terms of trade mean that the cost of grain is very high at the end of the dry season, the time when pastorali sts are most in need of grain and when livestock are thinnest and fetch the lowest price (White, 1 991) As livestock flood into market for sale, prices drop precipitously and grain prices soar, further reducing the purchasing power of pastoralists. These terms of trade work against pastoral people and increase their overall vulnerability (White, 1991) During drought
42 these terms are worsened, as the price of scarcely available grain goes up and l ivestock prices plummet, meaning greater loss of animals in exchange for less grain. Research indicates that the global economic system places pastoralists at a distinct disadvantage with regards to accessing markets and acquiring essential goods (Sen, 1981) First, the very mobility that allows them to cope with unstable environmental factors is largely incompatible with capitalist labor markets. Second, pastoralist populations tend to live in areas poorly served by transportation and other infrastructure, which increases their isolation and inability to use markets effectively (Smith et al. 2000) In addition, internationally and domestically promoted export based growth strategies have been adopted across the Sahel, and these often lead to the state appropriation of land for economic pu rposes, such as mineral exploration. This hinders pastoralists from maintaining traditional livelihoods and increases their reliance on markets. In many ways, neoliberal development policies have increased the need for cash income in order to access basic needs, such as food, education, and healthcare. This growing need for goods and services not produced by the household or community pushes pastoralists into deeper poverty. The complexity of the rapidly changing environmental and economic conditions facin g pastoralists necessitates a multidisciplinary approach to understand how the resilience of the pastoral livelihood system has come under sustained pressure in contemporary economies. Moreover, integrative frameworks may help identify effective coping mec hanisms and adaptations for pastoralists in altered environments. Identification and facilitation of appropriate, effective adaptations for vulnerable
43 populations, such as pastoralists, are essential skills for development practitioners working within the context of global climate change and economic globalization. Methodology This article applies the DE framework to examine how economic and environmental stressors may interact to affect food security in pastoral economies. As Brien (2008), feedback double exposure emphasizes the ways that actions prompted by globalization or global environmental change drive additional global change. Context double exposure emphasizes how global change processes alter the environment in such a way as to increase its vulnerability to future internal and/or external shocks. These two pathways serve as focal points in the application of the DE framework to the pastoral experience in Niger. Secondary data from literature on Niger and pastoralism dem onstrate feedback double exposure, as epitomized in the case of uranium markets and pastoral food security. Primary data collected just following the food crises of 2005 and 2010 are then used to illustrate context double exposure. Collectively, data from both sources populate the framework, the analytical platform for our improved understanding of vulnerability and food security among pastoral populations. Primary data were collected during field research in 2005 and 2010 in the pastoral and agropastoral zones of the administrative department of Tanout in eastern Niger. Three Tuareg communities make up the research sites included in this paper: Farak, Dareram, and Abdounez (see Figure 4 1). The area was originally targeted for intervention by the British Red Cross (BRC) during the 2005 food crisis, based on vulnerabilities identified by two rapid assessments conducted by University of Arizona (Bureau for Applied Research in Anthropology, 2005) Communities with the highest
44 vulnerability were selected for inclusion in the BRC intervention and were monitored for 1 2 months after the intervention. Farak was a beneficiary community and Dareram was a control community. Control communities were identified as those with similar but slightly less vulnerability than beneficiary communities and were thus not targeted for in tervention but included in monitoring to establish impact of the intervention. During the 2010 food crisis, seven communities, including Farak and Dareram, were targeted for inclusion in the research on pastoral vulnerability to climate change. Insecurity in the region limited mobility and willingness of locals to travel into some areas, leading to the replacement of Farak with the community of Abdounez. Thus, 2010 data include information on the communities of Dareram and Abdounez. Figure 2 1 Researc h Sites in Tanout District Niger ( Source: FEWSNET, 2012 ) This study draws on data collected during key informant interviews (at least three in each community), community focus groups (male and female in each community), randomly selected household surveys (19 households per community per year,
45 conducted with the HHH and, in cases of male HHH, the senior wife), anthropometric measurements of children under five (40 children per community per year, including children of targeted households), and on field obse rvations of the research teams from both periods. All data were collected in household preferred languages of either Tamachek or Hausa and were translated into French by research assistants, Nigerien doctoral students, and field assistants. Field assistant s were originally recruited and trained by the BRC in 2005 before conducting focus groups and household surveys throughout 2005 2007, the duration of the BRC project; the same individuals were available and able to assist by conducting household surveys in 2010. Data were collected by hand, using survey instruments and written notes, and were later analyzed using SPSS and through content analysis for identification of common themes relating to food insecurity. The Case of Niger Niger provides an illustrati ve case to examine interactive processes of environmental and economic change. Landlocked in the West African Sahel, Niger is divided into three zones: agricultural, pastoral, and desert. Rainfall is the principal determinant of agricultural production in Niger, as its quantity and distribution are unpredictable and variance occurs between important thresholds for crop production. Although water has never been abundantly available in this semi arid stretch of land (Bradley, 1971) changing rainfall patterns, arguably due to globa l climate change, have lead to greater water scarcity in communities scattered throughout the stark landscape (Ben Mohamed et al. 2002) Historically, there was a progressive southward shift of the 350 mm isohyet rainfall line a movement of 50 to 100 kilometers from roughly 1970 through the 1990s whi ch triggered the conversion of previously productive areas of
46 the agriculture zone, including Filingu and Tanout, into food deficit zones (Daouda, 1996b; Sivakumar, 1992) Changing rainfall patterns and increasing temperatures both included in current climate change projections are threatening water supplies in the already fragile l andscape. According to the Nigerien government, chronic food insecurity affects 80% of the population, and yearly 10 30% of the population suffers more than a 50% deficit in their cereal needs (Government of Niger, 2005) Niger is one of the most food insecure countries of the world, and severe events including drought, floods, and increasing d esertification threaten its fragile natural resource base (Held et al. 2005; Hesse & Cotula, 2006) In a country that is economically dependent upon crop production and livestock (nearly 80% of the population relies on subsistence agriculture), the outcome of global climate change has large implications. The negative consequences of increased temperatures and decreasing rainfall are tangible at household levels, as crop production falls and herd numbers decline precipitously in years of excessive heat or poor rains. Over the past 30 years, Niger has become increasingly less able to feed itself, and most agricultural families produce only enough food to cover their needs f or five or six months (Baro & Deubel, 2006) Diets are generally deficient in protein, calories, and essential vitamins and minerals, contributing to chronic undernut rition (Wuehler & Biga Hassoumi, 2011) ; it is estimated that 40% of the rural population cannot satisfy its m inimal caloric intake requirements (Baro & Deubel, 2006) As food production and arable land per capita decline, Niger becomes increasingly dependent on food aid and imports. Over 75% of monocropping has serious dietary, overall health, and economic implications.
47 In terms of economic change and approaches to development, Niger has tak en a number of steps within the last two decades to increase its appeal to direct foreign investors. In conjunction with UNDP, Niger has worked to revise codes affecting private sector investment, including its investment code (1997 and 2000), petroleum co de (1992 and 2000), and mining code (1993). With such changes, Niger hopes to increase export revenue by selling minerals to multiple countries. The positive benefits of economic globalization, however, have arguably not reached most parts of Niger. To dat resources, a substantial portion of which is almost always in food aid. globalized are not significa ntly engaged in the global economy (Kelley, 2002, p. 635) However, very few remain entirely unaffected by its progress or decline. For example, cell phone coverage and subsequent cell phone use in Niger has increased dramatically since 2005. While conducting research in Niger in the spring of 2005, the long awaited cell phone tower in Tanout, our research base, was completed and coverage turned on. This single act connected an othe rwise isolated town in the Sahel to the globe. Our Nigerien research team members were reunited with family by phone calls to France and Libya, using numbers that had been sent overland years, sometimes decades prior to this moment. Improved communication meant, for many families, an increased possibility of receiving remittances. Immediately, ad hoc systems for monitoring regional market prices of livestock mainly camel were created. And information on security issues including car jackings and other a cts of violence began traveling quickly, often restricting movement within certain areas of eastern Niger. Beyond ubiquitous cell
48 phone use and the now common sight of a road side solar charging station for cell phones, evidence of engagement in the global market is seen other places in pastoral Niger, including the regular use of large plastic sheets, made and sold by Chinese merchants, to collect large quantities of water from deep wells in a single draw. The Tuareg are Muslim, historically nomadic peopl e that speak a Berber language, Tamachek, and practice small scale agriculture, caravan trading, and livestock herding throughout the Sahel. Marginalization of Tuareg pastoralists in Niger was ignited under French colonial rule. The French favored the agri cultural Hausa and Zarma communities of the south over the pastoral populations of the north; policy decisions and legal codes that favored the agricultural sector extremely limited Tuareg rights and access to land, and upon independence, the Tuareg were e ssentially unrepresented within the new government (Abdalla, 2009; Hammel, 2001; Simanowtiz, 2009) This disenfranchisement has been doc umented in both policy and practice and has overlapping consequences to processes of global change (Emerson, 2011) During the 1970s, the region experienced severe drought and a ccompanying food insecurity. In that time period, the Tuareg lost an estimated 75% of their herds and thus thousands fled their homeland for cities or neighboring states (Libya or Algeria). When some 20,000 of them returned in the mid 1980s under the promi se of a new political regime, they were greeted by a national economic crisis and promises of humanitarian aid that never materialized leaving them feeling, again, marginalized and discriminated against (Bednik, 2008) Feedback Double Exposure: Uranium Markets globalization and global climate change (Leichenko & O'Brien, 2002) ; projections
49 ind icate that they will endure the greatest negative impact associated with these processes, and yet they are among the least responsible for the changes. However, their responses to these processes, as well as those of others within the region, are in some c ases driving additional change, further exacerbating their own vulnerability. The examples of uranium markets and land use change are explored here as both consequences and drivers of global change processes in Niger, both of which have important implicati ons on Tuareg vulnerability and food security. The process of economic globalization is currently engaging individuals, communities, and nations from around the world in a highly connected, interactive marketplace. Niger, too, is engaged in globalization. Niger is one of the leading exporters of uranium in the world. And though uranium mining may deliver economic benefit to an elite few within Niger, the negative impact of the uranium industry on the Tuareg pastoralists of the region is gaining attention (Abdalla, 2009; Action Against Hunger, 2006; Bednik, 2008; Keenan, 2008, 2009; Meyer, 2010) estimated 3,300 tons annually, Niger remains e xtremely poor, perhaps the poorest country in the world (Bednik, 2008; Keenan, 2008; United Nations Development Program (UNDP), 2011) When the French granted Niger its independence, they maintained control over the uranium mines in and around Arlit, and, until recently, (Bednik, 2008; Niger Uranium Limited, 2008) Arevo, the largest uranium mining company in the world, maintained the only two uranium mines that endured since the end of colonialism (Staff, 2008) But, since the early 1990s the Nigerien government has actively engaged
50 investors from China, Canada, India, Australia, South Africa, and numerous other countries in the exploration for and mining of uranium. In 2006, the French lost their were awarded licenses to explore for uranium, followed shortly thereafter by numerous other countries; that year Niger sold roughly 300 tons directly on the market (Bednik, 2008; Keenan, 2008) After a 20 year period of relative stability, renewed interest in alternative energy sent the global price of uranium skyrocketing in 2007, with prices jumping from $10 per pound in 200 3 to a record $136 per pound by June 2007. Although prices have not remained that high, uranium prices have remained above $50 per pound throughout 2011, and new exploration for uranium in Niger continues (IndexMundi, 2012) By October 2007, the Government of Niger had issued around 90 exploration permits, with 90 more under consideration (Keenan, 2008) To date, nearly two dozen countries have been granted exploration permits throughout the northern reaches of Niger, yet the Tuareg have not been consulted. The Tuareg, in whose ancestral lands the uranium deposits exist, claim continued marginalization, a lack of benefit from the presence of the mining companies, increased burden of disease due to the mines, and dire ecological consequences to their fragile ecosystem (Abdalla, 2009; Emerson, 2011; Keenan, 2008; Meyer, 2010) Grazing has been disrupted by exploration and mining, and the increased demand for water (required in the uranium exploitation process) is threatening traditional alternative livelihoo d activities of Tuareg, essential coping mechanisms during times of stress (Bednik, 2008; Emerson, 2011) Mining in the area is also compromising hu man health: dust and water show high levels of radioactivity (Keenan, 2008; Meyer, 2010) Greenpeace visited Arlit in March 2010 and documented
51 contamination of air, soils, and water, as well as the sale of radioactive scrap metal at markets throughout the pastoral region (France 24, 2010) s export income in 2009, and depending upon global trends in energy, projections indicate steady to significant increases in the near future (Mira, 2011; US Department of State, 2012) The Tuareg believe they have a right to benefit from the uranium mining, which until now they have not received. They are adamant that reckless mining efforts are disrupting the essential, yet delicate balance with the environment on whi ch their livelihood depends (Emerson, 2011) These grievances among others, including marginalization, discrimination, and a failure to meet the terms of the 1995 Peace Accord, have led to increased violence in the region since 2007. The Mouvement des Nigeriens pour la Justice (MNJ), a Tuareg rebel group organized in and around the region, has claimed a series of violent attacks, kidnappings, and other crimes, many of which have targeted the uranium industry and other westerners in the area. The group claims racist neglect and seeks increased representation in the military, government, and mining industry (McConnel, 200 7) The battle over uranium in the pastoral regions of Niger illustrates a double exposure of globalization and environmental global change with important feedback mechanisms: globalization, through colonialism and foreign direct investment, has sideline d the pastoral Tuareg from economic benefit while simultaneously creating increased global demand for energy. To date, that energy has required either oil, which has negative climatic impacts on the region through global warming, or uranium, which further exacerbates the presence of the mining companies and the marginalization of the Tuareg. The violence now gripping the region is arguably a consequence of the
52 aforementioned processes. This insecurity then acts as a barrier to regional and local economic de velopment by restricting pastoralists from utilizing locally adapted strategies, such as migration, and encouraging strategies that take them southward, outside the area of conflict. In addition, violence and insecurity deter international investors, human itarian aid, and development organizations that might otherwise bring economic or livelihood relief to the region. Context Double Exposure: Land Use Change, Markets, and Household Demographics This section utilizes data from three Tuareg communities locate d within the agropastoral and pastoral zones of Niger to examine how context double exposure is affecting food security. These historically pastoral communities are located geographically south of the areas most affected by uranium mining and, today, have varying degrees of mobility. Their experiences are used to illustrate the Tuareg experience of context double exposure. Throughout Niger, changes in the contextual environment are resulting from the local response of agricultural and pastoral communities to the ongoing drying of the region. As previously established, decreased rainfall has led to a southward shift of the 350 mm isohyet line, the amount required for cultivation of millet. This means that households that choose to stay in their historical ho melands suffer from decreasing yields over time, due to reduced rainfall. Similarly, the drying pattern has threatened the viability of historic livelihood systems in some pastoral communities. Herds must travel farther and farther afield in search of graz ing lands, and, yet, increased violence is limiting migration options. In addition, traditional coping strategies for maintenance of herds are no longer as effective as they once were. For example, in 2005 and in 2010
53 both communities reported that they we re no longer using the traditional restocking system of haba nai, where female livestock are loaned for a period of time long enough to produce offspring for the borrower, then returned to the original owner. Due to malnourishment of animals and the conseq uent length of time required for maturation and reproduction, in addition to the widespread loss of livestock among pastoral families due to illness, starvation, and sale, this strategy has been abandoned. Pastoralists are moving further afield in search of grazing lands and water, which often means traveling farther south, where average rainfall is slightly higher, earlier in the dry season and for longer stays. In doing so, pastoral communities encroach upon agricultural land and thus must negotiate acce ss to grazing land and water with agricultural communities (CARE USA & CARE Denmark, 2007) At the same time, decreasing yields have resulted in farmers claiming and trying to cultivate land in areas that have historically been used for pastoral grazing (Hammel, 2001) Land use changes among pastoralists and agricultural communities in response to climate change and globalization are p ushing the geographic boundaries of historic livelihood territories (Bassett & Zuli, 2000; Mortimore, 1998) increasing land degradation and conflict between agricultural and pastoral communities (Baro & Deubel, 2006; Thebaud & Batterbury, 2001) In addition, some pastoralists are choosing to sedentarize. Though some research argues its merits, sedentarization, the loss or decrease in pastoral mobility, has been linked to increased fertility infectious disease, and other indicators of vulnerability, thus this adaptation is not without additional risk (Blench, 2001; Hogg, 1986; Keenan, 2000a; Little, 1985; Niamir Fuller, 2000; Talle, 1999) Each of the Tuareg communities included in this research is more settled than their ancestors were, having
54 employed a range of coping strategies, including sedentarization, to adapt to changes in the natural and political environment. A brief introduction to each com munity and its current composition is presented below. Farak is a predominately Tuareg community of roughly 800 people, situated 38 km north of Tanout. A majority of the population grows no crops, though some produce small scale gardens for their own consu mption, as water is available from an open well. They continue to live in traditional, temporary shelters that are easily moved, though at least a portion of the population moves only very short distances around the open well. Others continue transhumant m igration. There is a health center, a veterinary input center, and a primary school within the environs of the camp and the well, which provides water year round. Focus groups reported that staff and supplies at the health center and veterinary input cente r are inconsistent; no one was available to speak with during data collection. Dareram is a predominately Tuareg community of roughly 200 people located 38 kilometers southeast of Tanout. Made up of three small settlements, most community members have sede ntarized, while others are semi nomadic and use the community as a base. In years of poor agricultural production, families reported they return to transhumance as a coping mechanism. In addition to livestock herding and agricultural cultivation, household s engage in small scale commerce, including jewelry making. There is no school, medical or nutritional service offered in town. The nearest school and health center is 10 km away (Adjiri) and the nearest nutritional center is 45 km away (Kokoram). There is no community well, but a seasonal lake provides water in the wet season.
55 Abdounez is a pastoral community of mostly Tuareg located 45 km north of Tanout, situated near a seasonal lake. The area was a traditional destination for herders to water animals a s they came south during the dryer months. However, due to widespread loss of livestock during the 1984 crisis, a group of Tuareg began adopting agricultural practices and settled in the area in 1990. There is a livestock market that occurs every Friday in Abdounez. There is also a primary school, a health center, and a cereal bank within the community. The community of Abdounez reports rapid increase in the number of family members to feed since settlement, but also cites the advantage that they are now reached by aid that flows in during food crises. They report new health problems associated with settlement, and though there are some that prefer being settled, most agree that the lack of milk in their diet is causing a new host of health problems. Food security of pastoral households is undermined by the longer term consequences of herd loss as compared to crop loss. When agricultural communities suffer from drought and have a loss of crops, they are able to start over with minimal to moderate impact on communities who lose livestock have to rebuild herds sometimes entire herds which takes an average of five to seven years. When herds in Dareram and Abdounez were decimated during the 1984 cri sis, the communities dispersed and families fled to Nigeria. Upon returning, neither community was able to rebuild herds to a sufficient size, and by 1990, the community of Abdounez decided to settle and begin growing crops in addition to herding. Having employed this coping mechanism of sedentarization, the context in these two Tuareg communities was quite different for the recent crises. Due
56 in part to their agricultural ventures, neither community dispersed during the 2005 or 2010 crisis (as some other pastoral communities did), and in 2010, both reported reduced mobility and sedentarization among common coping mechanisms (see Table 1). Additional coping mechanisms employed in research sites during the 2005 and 2010 crises included migration of young peo ple to Nigeria and Libya, consumption of famine foods (namely Boscia ), collection and sale of wood (as reported in focus groups), days of fasting, and reduction in number of meals all of which negatively affect nutrition and health, increasing overall vu lnerability (see Table I). Table 2 1. Livestock Loss and Coping Strategies Employed during 2005 and 2010 Food Crises by Community Dareram 2005 Farak 2005 Dareram 2010 Abdounez 2010 Average HH livestock loss 46.0 % 26.0 % 50.3% 39.4% Collect/eat fa mine foods 63.2 % 57.9% 44% 15.8% Migration en exode 94.7 % 84.2 % 38.9% 10.5% Day without food 26.3% 52.6% 5.6% 52.6% Decreased number of meals 47.4% 68.4% 84.2% 100% Sedentarization Not asked Not asked 55.6% 63.2% Reduced Mobility Not as ked Not asked 55.6% 52.6% No te: Household sample size = 19 per community except where missing data decreased sample size to 18 In addition to land use changes, including sedentarization, global changes are affecting these communities through marke ts and changes in household demographics. Tuareg pastoralists in Niger have varying access to regional markets for the exchange of goods; those who have decreased their mobility are generally situated further south than their nomadic counterparts and gener ally have increased access to markets, due to their proximity to crop producing communities. Their household food security can be greatly affected by their geographic access to markets, as well as their purchasing
57 power. In November 2005, pastoral communit ies of Dareram and Farak reported livestock losses (based on average decrease in total livestock units per household) during the crisis year of 46% and 26% respectively, and in November 2010, Dareram and Abdounez reported losses of 50% and 39% (Jahnke, 1982) While some of these livestock died due to starvation and disease, others were sold at market at significantly reduced prices in exchange for grain. A 2005 foc us group in Dareram indicated that grain prices at local markets reached nearly four times (1200 FCFA /tia) their normal price (300 350 FCFA /tia) by July 2004, the peak of the hungry season during the crisis year. During the 2010 crisis, grain prices rose b ut, according to 2010 focus groups, not as dramatically as they did during the 2005 crisis never surpassing 500 FCFA /tia. As a final example of context double exposure, globalization is affecting household composition, an important determinant of househ old vulnerability. Nigerien households regularly rely on the seasonal or short term migration of a family member, often an adult male, either to another region of the country or abroad. Sending someone en exode as it is called, is an important coping mech anism for families in both agricultural and pastoral communities, and globalization is changing this pattern. In 2005, 89% of households reported having a family member who migrated in search of work as a coping strategy for the crisis. Two of the most com mon destinations for this out migration of the labor force were Libya and Nigeria. In 2010, household data indicated a drop in out migration of adult men to 24% of households. Focus groups in Dareram and Abdounez indicate that this change is due to increa sed humanitarian aid (neither group received any aid in 2005, while both received aid from at least one NGO in 2010), history of past migration (some have lost family members to out migration after past
58 crises and no longer see it as an effective coping me chanism or have no additional members to send), and increased agricultural production within the communities. Globalization has also changed the demographics of households by limiting in country migration through increased violence and political upheaval, and increasing migration overseas through the increased economic disparity between Nigeria and Niger, largely driven by the oil economy of southern Nigeria. The activity of Al Qaeda in the Maghreb (AQIM), a radical Islamic militia who have been active in Niger since 2007, and increasing violence among the Tuareg in the region around and north of Tanout are discouraging movement of both local people and NGOs throughout the agropastoral and pastoral zones (Keenan, 2009) During data collection in 2010, field assistants blatantly refused to go into certain Tuare g held areas, and local entrepreneurs owning the only 4x4 vehicles available for rent in the region forbade travel into certain pastoral areas around Tanout (the reason Farak was not visited in 2010). As households have fewer economic options that rely on short term migration of family members to nearer destination they must rely more heavily upon long distance migration, if they are to engage in this historic coping mechanism of out migration. But this type of migration is contentious, as family members ar e gone for long periods of time, and the remaining family may or may not receive remittances in this time. Research findings suggests that families whose members migrate abroad experience increased vulnerability, despite the ased on loss of human resources within the household while awaiting remittances. Conversely, they may opt out of migration and risk increasing household vulnerability through a loss of potential income. The changing demographics of households as a result o f globalization and the ways it stimulates or
59 stymies migration has important implications on household vulnerability when food crises are stimulated. Concluding Remarks As the case of the Tuareg in Niger shows, issues of food security result from dynami c, interacting processes that operate at different scales. Using an interdisciplinary framework allows us to identify two multiscalar dynamics, and the causal mechanisms behind them, which alter the food security of pastoral populations. Perhaps the most i mportant finding of this study is the effects that climate change and globalization have on political stability in the region. First, climate change at the global scale has decreased the mobility of pastoralists. With their herds decimated by excessive hea t and droughts, people increasingly settle near water sources and attempt to position themselves in areas serviced by international aid organizations. Their inability to re grow herds after severe and/or consecutive droughts increases incentives for crop f arming, a choice that increases local level conflicts over land with agricultural communities already living in the area. At the same time, growing demand for uranium at the global scale has also contributed to increasing violence and insecurity at the reg distribution of benefits from uranium sales and the negative ecological and human health consequences of mining in the region. Resistance to state appropriation of land fo r mineral exploration, conducted by foreign multinational firms, has knock off effects that serve to further marginalize an already oppressed group. Conflict prevents international food aid from reaching those in need, and limits the use of migration to ne arby areas as a coping strategy for pastoralists. These political issues compound the negative effects of climate change and reduce the coping strategies available to
60 pastoral households. All the processes examined in this analysis interact in a way that w orks to decrease food security among the Tuareg, which suggests that their situation will worsen over time. While food insecurity is an obvious manifestation of poverty, global economic processes appear to be a thin driver of change to livelihoods. Althou gh better access to cash income would most likely improve their health and wellbeing, the processes that reduce their capabilities to earn income appear more directly related to weather and conflict. As interconnected as processes may be, some exert more f orce than others when looking at how change affects food security at the local level. And in this case, economic globalization plays a smaller role relative to other processes of change. While global economic processes may contribute to climate change or t o higher demand for uranium, it is the conflict and weather that, in this case, have the most direct influence in altering local livelihoods. Moreover, our findings suggest limited potential for economic globalization to bring benefits to the Tuareg. For e xample, the feedback double traditional lands decreases the ability of globalization to bring economic opportunity to the area and increases the potential for climate chang e to do harm. In terms of policy, our analysis underscores dynamics between structural and circumstantial determinants of vulnerability. There is no simple intervention to improve food security; it must be addressed structurally (e.g., development implic ations) as well as circumstantially (e.g., humanitarian/relief implications). Feedback pathways in this case illustrate how interconnected global change processes cannot be separated from context. The Tuareg have been structurally and institutionally margi nalized:
61 development programs with the best of intentions further their vulnerability by providing incentives for settlements; climate change threatens their relationship to natural resources; and they are being disproportionately (negatively) impacted by the main export opportunity (uranium) existing in country. Ultimately the use of the DE framework in this case provides a cautionary tale. The framework is designed to examine multiple processes of global change; however, the processes selected for exa mination by researchers may not necessarily be the most significant drivers of change. With an interdisciplinary framework, there remains the need for open mindedness when analyzing multiple processes of change. In this case, an additional driver, conflict had more influence on food security dynamics than the original processes that the framework introduces. When approached in this spirit of discovery, interdisciplinary frameworks may identify overlooked or underemphasized drivers of change essential inf ormation for effective aid and development intervention. However, no matter how interdisciplinary, a theoretical framework cannot substitute for theories themselves. In order to truly understand the complex drivers of vulnerability, the advancement of inte rdisciplinary theories is a critical next step for scholars and policy makers.
62 CHAPTER 3 PERCEIVED RISK OF CL IMATE CHANGE, ADAPTA TION AND LIVELIHOOD VULNERABILITY IN EAS TERN NIGER Introduction Pastoralism, a livelihood that depends upon the raising of d omestic animals for consumption (meat and/or milk), social exchange, sale and trade, is practiced throughout the African Sahel Among pastoralists, reliance on milk and animal products for subsistence a nd distinct social adaptations allow for a high level of resilience, essential for communities whose natural environment is unpredictable and requires constant coping and adaptation managers of risk dealing routinely with cons equences of climate variabil ity including drought induced livestock loss and increased competition for grazing land (pp. 492, Davies & Bennett, 2007) However, arid and semi arid regions of the world are projected to be among those most affected by global climate change, which poses new challenges increasingly erratic weather patterns (timing, du ration, and amount of rainfall), rising temperatures, and the deterioration of social resources to the historic adaptability and resilience of populations that inhabit these lands (Nia mir Fuller, 2000) In an attempt to understand how communities have coped and adapted thus far, indigenous knowledge, perceptions of risk, and adaptive capacity of farmers have become critical areas of research (Ben Mohamed et al. 2002; Mortimore & Adams, 2001; Nyong et al. 2007; Roncoli, 2006; Roncoli et al. 2002) Though some of this research has included pastoral experiences of climate change (Hesse & Cotula, 2006; Le Ho uerou, 1996; Morton, 2007) much of the research has focused on farming communities.
63 The term livelihood has been defined as the capabilities, assets (stores, resources, claims and access) and activities required for a means of living (Chambers & Conway, 1992, pp. 7 8) which prioritizes access over availability as a determinant of poverty and vulnerability (Sen, 1981) P astoral and agricultural livelihoods are neither permanent nor distinct in Niger; instead they are better conceived as theoretical opposite poles on a spectrum of possible liv elihoods Significant assumptions are made when pastoralists are subsumed into larger agricultural livelihood groups, but dichotomizing livelihoods into pastoral or agricultural groups based on ethnicity, community, or primary income source is an oversimpl ification of a complex set of experiences. Household livelihood is not static, particularly within the pastoral livelihood system. Households constantly shift strategies, coping and adapting to economic, environmental and personal shocks in an attempt to e ndure the ever changing environment in which they live a classic demonstration of what has come to be referred to as livelihood sustainability (Chambers, 1992; D avies, 1996; Scoones, 1998) This dynamism of livelihood, typical of pastoralists, has long been a seminal and defining characteristic of pastoral resilience. For example, a settled household may abandon farming practice and return to no madic pastoralism in times of drought (Galvin, 2009; Le Houerou, 1996) or a nomadic household may respond to consecutive droughts by reducing mobility, settling, and initiating crop production (McKune & Silva, 2012) The same household may look like farmers one year and pastoralists another, and coping strategies and adaptations may vary in their effectiveness due to the historical livelihood of the household and accompanying knowledge and skill sets. Thus, it is essential that global efforts to improve livelihoods determine if/how climate change is differentially
64 affecting pastoral, agropastoral, and agricultural communities, rather than assuming that communities with different means of subsistence will experience, interpret, and be affected equally by climate change and the strategies and adaptations used to manage climate related r isk Furthermore, any discussion of climate change that aims to improve the livelihood security of pastoralists needs to use a well defined, robust indicator of livelihood that allows for appropriate interpretation of the findings, given the dynamic nature of pastoral existence. Th is research investigates the development and use of a livelihood index to examine how agropastoral populations in eastern Niger perceive and respond to climate change, the role of livelihood in perceptions of climate related risk and if/how coping mechanisms and adaptations affect their livelihood vulnerability/resilience Data for this study was collected among seven communities in eastern Niger from November 2010 to February 2011. The study aims to augment existing literature b y 1) describing the relationship between perceived risk of climate change and livelihood within communities across an agricultural pastoral spectrum, 2) investigating whether perceived risk of climate change is correlated with the use of certain coping mec hanisms and adaptations, and 3) examining the consequences of certain livelihood associated coping mechanisms on pastoral or agricultural vulnerability. The remainder of this article is divided into four sections. The first reviews the literature on climat e change in Niger, the sustainability of pastoralism as a livelihood, and pastoralism and vulnerability/resilience. The second section describes research methods and defines key variables used in analysis. In the third section, findings from the research a re presented, organized by specific aim, as outlined above. The fourth and final section discusses implications and
65 constraints of using a livelihood spectrum to assess vulnerability among pastoral populations. Background Climate C hange in Niger Niger is an extremely poor, landlocked country in the West African Sahel, two thirds of which is desert or sub desert (see Figure 1). Tanout, the research base for this stud y, is located in eastern Niger within the Sahel Sahara zone, a water scarce region character ized by 200 350 mm annual rainfall (Ben Mohamed et al. 2002) Global climate change is predicted to have significant effects in arid and semi arid regions of the world, including substantial drying in the Sahel (Held et al. 2005; IPCC, 2000 (IPCC); IPCC 2007) Ra infall and temperature data suggest that rainfall has decreased in recent years by as much as 42% in Tanout (pre 1968 compared to 1968 1990) and that the average number of days with temperatures exceeding 30 C have increased from 0 to over 15 days in the second half of the 20 th century in nearby Maradi (Barb & Lebel, 1997; Battisti & Naylor, 2009; Ben Mohamed et al. 2002) Figure 3 1 Map of Niger (Source: FEWSNET, 2012)
66 Semi arid regions of the world such as th e Sahel typically experience a high level of variabil ity in rainfall and temperature. However, d espite regular high variability, climate data from Zinder, the largest town within reasonable distance from Tanout, for three periods (1951 1968, 1969 1984, 198 5 1998) show clear trends of climate change: a decrease in average annual rainfall (542 to 367 mm), a reduction in average length of rainy season (92 to 74 days), and later onset of rains ( from late June to mid July) (Lebel et al. 1992; van Duivenbooden et al. 2002) Although climate variation and extreme events such as droughts and floods are typical for the Sahel, climate change models predict increased variability, including increases in the number of very dry and very wet years in the next 90 years, and extreme events such as droughts and floods (Huntingford et al. 2005) These changes are predicted to have a number of dire consequences, and v ulnerability to these consequences is not equally distri buted ; vulnerability is expressed along social, power, poverty, and gender lines, leaving marginal groups women, ethnic minorities, and the poorest among the most vulnerable (Denton, 2000, 2002; London, 2001; Nel son et al. 2002) Pastoralism as a Sustainable L ivelihood The people who inhabit the Sahel are acc ustomed to both weather and climate variability, and the flexibility of the pastoral livelihood there reflects this ability to survive in g eographic regions where weather patterns and rainfall are unpredictable is a defining characteristic of their livelihood. As elsewhere, pastoralists in Niger rely on a range of economic activities, not solely the production of livestock through herding (Davies & Bennett, 2007; Hogg, 1997; McPeak & Barrett, 2001) These activities include livestock herding, jewelry making, salt trade, and petty commerce. This range of economic activities in which pastoral households a nd communities engage
67 often renders distinction between pastoral and non pastoral populations quite difficult (Davi es & Hatfield, 2007; Hogg, 1997) Livelihood diversification, where households spread their risk by investing across multiple economic sectors, has been widely researched as an instrumental means of decreasing the vulnerability of the household to externa l shocks and unpredictable income and production (Ellis, 1998; Little et al. 2001; Scoones, 1998) Increasingly, pastoral populations are sedentarizing and employing crop cultivation and small scale gardening for t he sake of livelihood diversification Unfortunately, research has shown that, for nomadic pastoralists of the Sahel, sedentarization for the sake of agricultural production is not improving livelihood security, due to the large costs associated with maint enance of different livelihoods (Pedersen & Benjaminsen, 2008) Several researchers have concluded that sedentarization of pastoral populations leads to increased rates of impoverishment and destitution (Blench, 2001; Hogg, 1986; Little, 1985; Niamir Fuller, 2000; Talle, 1999) ; others adhere to the benefits of pastoral sedentarization, particularly surrounding the benefit of increased access to markets (Campbell, 1984; Ensminger, 1992; Sato, 1997; Smith, 1998) Households within the stud y area have a range of economic activities that may be more or less significant from one year to the next and, increasingly, may involve some agricultural production. Pastoralism and Vulnerability/R esilience Vulnerability is a function of the sensitivity and exposure of a system to an external shock and the adaptive capacity of that system to absorb or recover from consequences of that shock (Adger, 2006; Kelly & Adger, 2000; Turner et al. 2003) Within the climat e change literature, adaptive capacity the ability of a system to adjust to a shock, minimize negative impact, and cope with consequences has gained important traction
68 in understanding, assessing, and intervening to reduce vulnerability (Adger, 2006; Gallopin, 2006; Smit & Wandel, 2006) Because global climate change models predict increased climate variability and extreme events in the Sahel, the adaptive capacity of population s to current climate variability se rves as a good indicator of their future vulnerability to global climate change (Elasha et al. 2005) Resilience is a concept and return to a stable state (Holling, 1995) In human systems, the word is used to describe populations that are able to cope and adapt in the face of external shocks, including climate extremes (Davies, 1996) Pastoralists are in herently resilient, as the livelihood is responsive to significant climatic variations (Adger, 2000; Ellis & Swift, 1988; Friedel, 1991; Niamir Fuller, 2000; Westoby et al. 1989) Among the most useful contributio ns to understanding vulnerability/resilienc e of thinking, which challenged the previously held assumption that rangelands are equilibrium based ecosystems (McCabe, 2007; Scoones, 1999) The theory proposes that arid and semi arid lands are persistent, non equilibrium systems characterized by ecological variability, unpredictability, and high resilience; it posits that ecological areas of great unpredictability and variability lend themselves to communal management (Niamir Fuller, 2000) for managing natural resources in arid and semi arid zone (pp. 65, Grell & Kirk, 1999) N omadic pastoralism is seen as an efficient, effective adaptation to instability inherent within the ecosystem (Ellis & Swift, 1988; Friedel, 1991; Westoby et al. 1989)
69 A key aspect of this resilience is mobility. Pastoral populations emplo y varying degrees of mobility, and they move in response to ecological, political, and social changes in the environment. A number of researchers have investigated the importance of mobility in pastoral populations and the consequences of decreased mobilit y (Ellis & Swift, 1988; Fratkin, 1992; Nathan et al. 2005; Roth, 1996; Schwartz, 1995) as more and more pastoral communities have initiated semi nomadic and even sedentary livelihoods (Keenan, 2000b) In semi nomadic communities women, children, and the elderly will often stay in a camp around a w ell for long periods of time, while men travel with herds shorter distances around the well or well system When grazing land is exhausted, the community may or may not move with the men to another well system, depending upon the time of year (Clarke, 1959) This change in livelihoods allows communities to practice small scale agriculture and gardening, as well as trade in non animal goods, such as jewelry Other important determinants of pastoral vulnerability include s ocial bonds and social capital, created through exchanges of livestock, re stocking alliances, dowries, traditional loan mechanisms, and s uppor t of the poor through gifts or loans of livestock (Davies & Bennett, 2007; Fratkin, 2004; McCabe, 2007; Niamir Fuller, 2000; Thebaud & Batterbury, 2001) Just as the functional diversity of an arid ecosystem co ntributes to its health and resilience, the range of options, adaptations and coping mechanisms available to a pastoral community contribute s significantly to its health and resilience. Research Methods This study includes analysis of information from doc ument review secondary data (data previously collected by others and available either publicly or used with permission) and new, primary data collected via key informant interviews, focus
70 groups household interviews and anthropometric measurements from 2010 2011. Fieldwork was based out of Tanout, Niger and was initiated during the period just post harvest (late October) in 2010. Seven communities representing livelihoods along an agricultural pastoral continuum were identified for inclusion in the stud y (see Figure 3 2). Originally, seven communities were selected from a pool of 19, previously identifie d for research through a vulnerability assessment during a 2005 British Red Cross food security intervention (Bureau for Applied Research in Anthropology, 2005) However, nomadic pastoralists were not well repr esented in the 2005 pool of communities, thus, two new pastoral communities with high levels of mobility and ethnic composition that contributed to sample diversity were added to the sample (Eliki and Djiptoji). A third new community (Abdounez), comparabl e in ethnic make up, livelihood, and location, replaced one of the 2005 communities (Farak) when political insecurity obviated the possibility of researchers accessing it in November 2010. In repeat communities, the same households that were interviewed in 2005 were targeted in 2010 (BARA 2006a) In new communities, 19 households were randomly selected from a census list for inclusion in the household survey sample, replicating household targeting techniques employed in 2005. In each communit y, researchers conducted m ale and female focus groups with randomly selected individuals and a minimum of three key informant interviews ( elders, teachers, health workers, and spiritual leaders ). Research and field assistants collected all data in the pri mary household languages of Tamachek and Hausa. Research assistants translated responses into French, and I analyzed all data in French.
71 Hand written data from surveys were entered into spreadsheets in Niger. Data were sorted, cleaned, and rechecked for ac curacy in order to reduce data entry error. Descriptive statistics were run for all independent and dependent variables. Data were transferred to SPSS software and quantitative analysis was conducted, applying Figure 3 2 Map of Research Area in Eastern Niger (Source: FEWSNET, 2012) tests of bivariate association estimating multiple linear regression models to examine the effects of independent variables on PRCC and vulnerability, while holding demographic, economic, and livelihood data constant (SPSS, 2 011) Qualitative data from focus groups and key informant interviews was analyzed using thematic coding, without the use of qualitative analysis software. Findings from the study are organized by research aim and presented here. Key Variables Key variabl es for this research include livelihood, perceived risk of climate change (PRCC), vulnerability/resilience, food security, health, wealth and various adaptations
72 and coping mechanisms. Brief definitions, methodologies, and univariate descriptions are given below for each. Livelihood A livel ihood index was created using information from 14 questions on the household survey and from more general information collected during focus groups interviews. This data included revenue sources, history of agriculture, h istory of pastoralism, self identification as pastoral, and use of migration as herding strategy. Ultimately, I established a seven item index to represent the agricultural pastoral continuum on which research household s continually move (see Figure 3 3 fo r index and see Appendix E for full explanation of index creation). Since most households fall somewhere bet ween exclusive reliance on either agriculture or pastoral production, each household is assigned a position (1 7) along the continuum representing t heir livelihood portfolio at that moment in time, rather than being identified as either agricultural or pastoral (n=128). For each po sition on the continuum in Figure 1 typical households are described in the text circle above Univariate distribution of the livelihood of households in the sample is presented in Figure 3 4 Figure 3 3. Agricultural pastoral Livelihood Continuum
73 P erceived Risk of Climate Change An index of perceived risk of climate change was developed based on a) climate change hazards relevant to the community, b) potential harm of each hazard and c) an assessment of the current situation (Slovic et al. 1982) Perceptions rather than measures of climate change are examined in this study, as perceptions may not mirror ground truth measures, such as land use or land cover change, and household and community decisi on maki ng is not always based on fact. Seven areas of potential climate change (hazards) were identified through literature review and focus group discussion s at the research site (Elasha et al. 2005; IPCC 2000; IPC C 2007) The hazards identified were: land degradation, condition of vegetation cover, soil and crop productivity, rainfall, desertification, heat/temperature, and l oss of indi genous species (plant or animal). Heads of household then assessed each hazard for 1) thei r capacity to do harm or damage and 2) the change that each household has observed locally. A perceived risk of climate change associated with each hazard was then calculated as the product of capacity to do har m and locally observed situation. The sum of the seven individual products was then used as a summary statistic or index for overall perceived risk of climate change PCA factor analysis indicated three components within the index Due to poor inter item correlation and internal reliabilit y ( = .425), an alternate index was created to include only condition of vegetation cover, crop and soil = .557) Both indices were binned into tertiles, representing high, medium, and low perceived risk of c limate change, and both were used in initial analysis (n=126 see Figure II ).
74 Vulnerability/R esilience S ince vulnerability/resilience in this context are seen as representing opposite ends of th e same spectrum, they were measured using a vulnerability ind ex adapted from Elasha et al (Elasha et al. 2005) Generic indicators of vulnerability, taken from the previous study and appropriate for the Sahel, were presented to key informants at the research site who helped establish locally appropriate indicators. These indicators were then validated and specified throug h focus group discussions, and additional indicators were suggested. Ultimately eight items were used in the final survey. Locally specific indicators of vulnerability identified through this process and utilized in cons truction of the index were: land degradation, condition of vegetation cover, soil and crop productivity, health and sanitation, wealth, access to food, exode and m obility 1 Heads of household were asked to compare their current situ ation for each indicato r to the situatio n immediately following the 2005 food crisis (the same time period after two different crises). Respondents indicated whether their situation was worse, about the same, or better than in 2005. PCA factor analysis indicated three component s and a of .42 8 Thus, the index was edited to include only five items (removing land degradation, exode and mobility) resulting in an index with a slightly higher internal reliability (CA = .547). This index was then binned into tertiles, re presenting low stable and high resilience (n=126, see Figure 3 4) Food S ecurity Food security was defined at the World Food Summit of 1996 as the state in which all people at all times have access to sufficient, safe, nutritious food to meet their 1 Exode i s the local term used to describe the outmigration, predominately of young men, to other regions of Niger, neighboring countries (Nigeria and Libya), and Europe in search of paid work.
75 die tary needs and food preferences for a n active and healthy (Food and Agricultur e Organization, 1996) Accordingly, three dimensions of food security have been identified, which are important for both measurement and protection; these dimensions are food availability, food access, and food utilization (Webb et al 2006) The food security indicator for this study is calculated using a method that assesses the frequency of use and severity of coping mechanisms employed by households, and thus is a measure of food access (Ma xwell et al. 1999; Maxwell, 1996) The measure food. Changed eating habits, decreased portions at meals, decreased nutritional quality at meals, acceptance of gifts/lo ans from family and friends, decreased number of meals, collection of wild famine foods, day without eating, and sale of personal belongings are the self reported coping strategies included in the food security variable. Calculated based on reported use of the strategy and local severity rankings (which mirrored 15. For this study, households were assigned a status of mild, moderate or severe food insecurity, based on this scale (n=133, see Figure 3 4 ). Adaptations/Coping Mechanisms During household interviews, data were collected on various coping mechanisms and adaptations, referred to here as strategies. Households were asked to identify all strategies employed during the past year (year of crisis) from a list of possible responses generated during the 2005 crisis and verified in 2010 pilot tests. Households were then asked to identify how often each strategy is employed (commonly, only during exceptional years, or not at all). These strategies incl uded those coping mechanisms utilized to measure food security, as well as others including
76 sedentarization, reduced mobility, migration, removal of children from school, consumption of own harvest, consumption of own milk/meat, and acceptance of commercia l credit or loans. The frequency distribution of certain adaptations and coping str ategies is presented in Figure 3 4 (n=130 133, varying by strategy). Figure 3 4. Univariate Distribution of Key Variables Wealth The measure of wealth of households in th is study is estimated by assigning average regional market value (in F CFA ) to grain and livestock holdings at the time of
77 the survey for each household and dividing the sum of those values by the number of adult equivalent s (AE) per household Households w ere then categorized into groups representing poor, average, and wealthy tertiles of the sample population. Ethnicity Each of the regression models estimated in this research includes key demographic characteristics as control variables, including ethnicit y. Despite that ethnicity is fluid and intermarriage between ethnic groups is not uncommon in Niger, ethnic groups maintain a shared set of opportunities and disadvantages, as well as cultural norms and household practices, that are more similar within tha n across groups. The differences that exist across groups may have significant effects on the outcome of the model, thus ethnicity is included as a control variable. Findings Perceptions of Climate C hange and Livelihood A correlation test was run to assess bivariate association, generating a icient of .045. Although t his indicates a positive association between perceived risk of climate change (seven item index) and livelihood (the greater the perceived risk of climate change, th e more pastoral the livelihood ), t he correlation is not statistically significant (p=. 625 ), indicating significant chance that the observed association has occurred due to sampling error. The test was rerun using the modified version of the PRCC variable, and it was still not significant (p=.427). L inear regression was used to estimate a model that regresses perceived risk of climate change on livelihood The results are not significant (p=.528); however they are presented here in the interest of statistic s: the slope of the line is .033 indicating that for every unit increase in the independent variable, livelihood (moving toward pastoralism),
78 ther e is an increase of .033 in the dependent variable, perceived risk of climate change. The y intercept, the va lue of PRCC when l ivelihood is equal to 0, is 1.765 (where 1 equa ls low perceived risk and 3 equals high perceived risk of climate change ) Thus, the regression equation is: PRCC = 1.765 +.033 (Livelihood) None of the findings indicate a relationship betwe en perceived risk of climate change and livelihood Interestingly no significant correlation was found between the perceived risk of any individual hazard within the PRCC index (land degradation, land cover, soil productivity, rainfall, desertification, h eat, and loss of species) and livelihood. Also of interest, there was no significant correlation between PRCC and wealth (p=.129), ethnicity (p=.163), or dichotomized livelihood (p=.875); and a marginally significant positive correlation between PRCC and f ood insecurity (p=.075). In focus groups, all seven communities identified rainfall as the most significant and most critical environmental change. Specifically, the quantity and distribution of rainfall is identified as the environmental change with the g reatest impact on overall community wellbeing. All communities cite the loss of plant and/or animal species as problematic; two highly agricultural communities (Takoukout and Kkeni, average livelihood index of 2.76 and 2.47, respectively; see Figure 4 5) identified the invasion of non traditional plant species, those typical of desert landscape, as an indictor of climatic problems within the community. All communities discussed a loss of land cover (particularly trees) within the context of either rainfall deficit or harvesting of wood. Perceptions of climate change, based on focus group data, did not vary by livelihood i.e., across the livelihood continuum
79 Figure 3 5 Distribution of Research Communities Along the Agricultural pastoral Livelihood Conti nuum by Average Household Livelihood P erceptions of Climate C hange Coping S trategies and Adaptations In order to better understand climate related risk, households were asked about a number of strategies and whether they employed these strategies during the crisis year. Correlation tests indicate that PRCC (seven item index) is positively correlated with reducing amount of food (p=.095, n=125), reducing quality of food (p=.095, n=125), eating own meat or milk (p=.087, n=126), collecting famine foods to e at (p=.003, n=126) and bartering (p=.043, n=126) as coping mechanisms. A nested multivariate regression model was estimated to further investigate the relationship between coping strategies and PRCC, taking into consideration other potential confounders. The model includes the following demographic characteristics: household size; sex, age and education of the head of household (HHH); and ethnicity. In addition, wealth and livelihood are included in the model. The results indicate that, among coping strate gies identified as significantly correlated with PRCC through bivariate analysis, only the collection of famine foods maintains significance when confounders are included in the model Results also show that the sale of personal belongings and regular migr ation are coping strategies associated with a lower perceived risk of climate change (see Table 3 1) Interestingly, education is marginally
80 positively correlated with risk of climate change in the final version of the model. Also being Fulani, one of the major pastoral ethnic groups in Niger, is highly positively correlated with perceived risk of climate change, indicating a significantly higher reported risk of climate change compared to Hausa (the reference population). Finally, living in a pastoral geog raphic livelihood zone is marginally positively correlated with risk of climate change as well. Interaction terms were created to investigate a possible interaction between ethnicity and education. The interaction term was tested in a regression model con taining demographic and wealth variables, and no significant effect on PRCC was generated (p=.204, n=121). Similarly, interaction terms for PRCC and individual coping mechanism (collection of wild foods and regular migration) were created and tested; no si gnificance found (p=.128, n=121; and p=.472, n=121, respectively). Though certain coping strategies may be associated with agricultural communities and others with pastoralists, the previous findings stand independent of variations that may exist by liveli hood, as the livelihood index is included in the model. However, in order to better understand how climate change might differentially affect the use of coping strategies, the livelihood index was dichotomized into 1) agricultural and 2) pastoral groups: a gricultural households consisting of those whose livelihood index was less than four, and pastoral households consisting of those whose livelihood index was greater than or equal to four. While not perfect, this methodology allows for a comparison of strat egies employed by those who are more agricultural and those who are more pastoral. Bivariate tests for correlation indicate that households with a livelihood score of four or more (hereafter, pastoral) have a PRCC that negatively
81 correlates with the strate correlates with collecting famine foods to eat (p=.014, n=51) and reducing mobility (p=.083, n=51). In other words, pastoral populations who consume their own harvest Table 3 1 Nested Multivari ate Regression Models Examining the Effects of Demographic Characteristics, Livelihood, Wealth and Coping Strategies on Perceived Risk of Climate Change. Independent Variables Model 1 Model 2 Model 3 Model 4 Household Size .003 .003 .004 .014 Sex of HHH .058 .007 .018 .051 Age of HHH .001 .000 .001 .004 Education of HHH .044 .046 .047 .101^ Ethnicity Hausa --------Tuareg .188 .064 .034 .093 Fulani .495^ .734* .763* .713* Beri Beri .071 .065 .050 .231 Ge ographic Zone Agricultural --------Agropastoral .504 .547^ .518 .440 Pastoral .416^ .466^ .445^ .321 Livelihood index .093 .086 .096 Wealth .040 .001 Reduce amount of food .684 Reduce quality of food at meals .069 Eat own milk or meat .099 Collection of famine foods to eat .604** Barter .094 Reduce number of meals .342 Accept gifts 1.174 Sell personal belongings .314* Regular m igration .333* Constant 1.514 1.605 1.683 .745 Model Fit Subset F 3.133** 2.980** 2.744** 3.121** Adjusted R square .137 .141 .137 .260 N 122 122 122 122 Notes: **p<.01; *p<.05; .05<^p< .10 Values shown in each cell are unstandardized coefficients
82 have a lower pe rceived risk of climate change than pastoralists who do not; and pastoral populations who collect famine foods and those who have reduced their mobility both have a higher perceived risk of climate change than those who have not. Among households whose liv elihood index is less than four (a gricultural ), there is a meat (p=.065, n=72), eating famine foods (p=.038, n=72), and taking a commercial loan (p=.084, n=72). A regression model was estimated to better understand the effect of coping mechanisms on perceptions of climate change (seven item index) among pastoral populations (livelihood index greater than or equal to four). Household size is significantly negatively associated with PRCC, and education of the HHH is significantly positively correlated with PRCC. When wealth is included in the second model, the effect of household size on PRCC g oes away. However the effect of education remains throughout the second and third model 2 In addition, as previously noted, ethnicity has a significant effect on PRCC. Tuareg households have significantly lower PRCC than Fulani households, and this effect remains throughout all models. As presented in Table 3 2, once demographic and wealth characteristics are included in the model, the coping strategies that are significantly positively correlated with PRCC among pastoralists include accepting gif ts, collec ting famine foods, exode and reduced mobility. The use of these strategies is associated with a higher perceived risk of climate change Strategies that are significantly negatively associated with PRCC include regular migration, 2 (Among pastoralists only two of 56 sample households reported members with any education: one had some primary school, and one had completed primary school).
83 reducing quality of foods sending family members elsewhere to eat, reduced food spending, and removing a child from school Again, these represent coping strategies whose use is associated with a lower p erceived risk of climate change, net of demographic and wealth characteristic s. Table 3 2. Nested Multivariate Regression Models Examining the Effects of Demographic Characteristics, Wealth and Coping Strategies on Perceived Risk of Climate Change Among Pastoral Households Independent Variables Model 1 Model 2 Model 3 Household S ize .044* .038 .040 Sex of HH .297 .295 .539* Age of HH .007 .007 .001 Any Education HH 1.348* 1.362* 1.368* Ethnicity Fulani ------Tuareg .943* .938* 1.166* Geographic Zone Agricultural ------Pastora l .523 .504 .303 Wealth .050 .004 Reduce amount of food 1.223 Accept gifts 1.490** Collection of wild foods to eat .737** Regular Migration .958** Reduce quality of foods 1.951** Eat own milk or meat .197 Barter .080 Send family elsewhere .561** Reduce food spending 2.273** Exode .640** Remove child from school 1.664** Reduce mobility .385* Constant 1.585 2.045 7.365 Model Fit Subset F 5.920** 5.084** 4.228** Adjusted R square .37 1** .364 .716 N 51 51 49 Notes: **p<.00; *p < .05; .05 < ^p < .10 Values shown in each cell are unstandardized coefficients
84 Comparable analysis was conducted for the agricultural population (where livelihood index < 4). Bivariate analysis generated sign ificant correlations between perceived risk of climate change and reduced quality of food (p=.06, n=72), collecting wild foods for consumption (p=.030, n=72), regular migration (p=.09, n=72), and sending a child to work (p=.05, n=71) among agricultural hou seholds. Regression models were then estimated to analyze the relationship between these coping below in Table 3 3, indicates that collection of wild foods and accepting commer cial loans are significantly positively correlated with perceived risk of climate change, while the sale of personal belongings is negatively correlated with PRCC among agricultural households. This model explains 19% of the variance in PRCC among agricult ural households, compared to 72% explained by comparable pastoral models. Coping Strategies, A daptations and Household Vulnerability/R esilience The final aim of this research is to understand how certain strategies employed during crisis affect the vulne rability and resilience of households, and whether there is variation in the effect of certain coping strategies by livelihood. While strategies are employed to reduce vulnerability, some meet short term needs and may render households more vulnerable in t he long run, while others may reduce the vulnerability of agricultural communities, but increase the vulnerability of pastoralists, or vice versa. Based on self reported measures of vulnerability across all livelihoods, bivariate tests of correlation show that the following strategies positively correlate with high self re ported household vulnerability: reducing number of meals, reduc ing amount of food at meals fasting on certain days, and con
85 Table 3 3 Nested Multivariate Regre ssion Models Examining the Effects of Demographic Characteristics, Wealth and Coping Strategies on Perceived Risk of Climate Change Among Agricultural Households Independent Variables Model 1 Model 2 Model 3 Household Size .038 .039 .033 Age of HH .000 .001 .004 Sex of HH .153 .153 .238 Education of HH .514 .518 .685* Ethnicity Hausa ------Tuareg .596 .585 .618 Fulani 1.313* 1.330** 1.456^ Beri Beri .039 .045 .323 Geographic Zone Agricultural ------Agropastoral 1.145* 1.154* 1.114* Pastoral .892* .901* .812* Wealth .013 .008 Reduce quality of food .809 Collection of wild foods to eat .651** Regular migration .172 Send child to work .171 Reduced mobility .311 Accept commercial loan .506* Sell personal belongings .485* Constant .497 .480 .722 Model Fit Subset F 1.799^ 1.595 1.945* Adjusted R square .094^ .079 .191* N 70 70 69 Notes: **p<.01; *p < .05; .05 < ^p < .10 Values shown i n each cell are unstandardized coefficients Conversely, a ccessing grain banks and sending children to work are both correlated with low self reported household vulnerability. In other words, bivariate analysis indicates that when food supplies are protec ted through a grain bank and child labor opportunities exist, people report being less vulnerable. People report feeling more vulnerable when consumption patterns have had to change including reductions
86 Table 3 4 Nested Multivariate Regression Models E xamining the Effects of Demographic Characteristics, Coping Strategies and Livelihood on Self reported Resilience, All Households Independent Variables Model 1 Model 2 Model 3 Model 4 Household Size .019 .019 .021 .032 Age of HH .001 .001 .206 .28 2 Sex of HH .238 .227 .001 .003 Education of HH .019 .018 .021 .038 Ethnicity Hausa --------Tuareg .114 .148 .076 .247 Fulani .501* .437 .513 .317 Beri Beri .107 .108 .086 .050 Geographic Zone Ag ricultural --------Agropastoral .046 .032 .005 .170 Pastoral .090 .074 .056 .130 Livelihood index .023 .031 .053 Wealth tertiles .071 .057 Reduce number of meals .362 Reduce amount of food .370 Fasting day .093 Eat own milk or meat .167 Cereal bank .332^ Send child to work .255^ Constant 2.451 2.472 2.321 3.034 Model Fit Subset F 2.142* 1.920* 1.917* 2.267** Adjusted R square .070 .071 .078 .152 N 121 121 121 121 Notes: **p<.01; *p < .05; .05 < ^p < .10 Values shown in each cell are unstandardized coefficients in the number of meals or amount of food at meals, use of fasting days or having had to eat into their own stock. These findings were used to esti mate a regression model examining the effects of coping strategies on self reported vulnerability, within the context of potential
87 confounders. When demographic, livelihood, and wealth variables are included in the model, none of the strategies previously associated with high vulnerability are statistically significant; only those associated with low vulnerability (accessing grain banks and sending children to work) are marginally significant see (Table 3 4 ) Despite having included livelihood in the previ ous model, livelihood was again dichotomized to investigate the use of coping strategies and consequent vulnerability in pastoral communities. Bivariate analysis showed that, among all coping strategies, only accepting commercial loans was significantly co rrelated with vulnerability (decreased vulnerability, in this case). A regression model was estimated to see how coping strategies affect self reported resilience when demographic and economic characteristics of the household were included in the model. Re sults are presented in Table 3 5, showing a significant negative correlation between reducing the amount of food consumed and household resilience (indicating an increase in vulnerability with use of the strategy) and a significant positive correlation bet ween household resilience and accepting a commercial loan (decrease in vulnerability with use of the strategy), net of household demographic characteristics and wealth Additionally, the model shows that household size has a positive significant correlatio n with resilience (the larger the household, the less vulnerable the household) and age of HHH has a negative significant correlation with resilience (the older the HHH the more vulnerable the household). Once again, ethnicity has a significant predictive power on self reported resilience among pastoralists, with Tuareg households reporting much higher resilience than their Fulani counterparts.
88 Table 3 5 Nested Multivariate Regression Models Examining the Effects of Demographic Characteristics, Wealth and Coping Strategies on Self reported Resilience Among Pastoral Households Independent Variables Model 1 Model 2 Model 3 Household Size .066^ .075* .102* Age of HHH .016^ .016* .022* Sex of HHH .204 .202 .035 Education of HHH .096 .116 .109 Ethn icity Fulani ------Tuareg .994** 1.003** .887** Geographic Zone Agricultural ------Pastoral .069 .039 .224 Wealth .078 .116 Reduce amount of food 1.845^ Reduce quality of food 1.318 Accept c ommercial loan .456^ Constant 1.443 1.697 2.549 Model Fit Subset F 2.865* 2.642* 2.631* Adjusted R square .183* .301 .409^ N 51 51 49 Notes: **p<.01; *p < .05; .05 < ^p < .10 Values shown in each cell are unstandardized coefficien ts Comparable data was then analyzed for agricultural communities (where livelihood index < 4) in order to compare the impact of coping mechanisms of self reported vulnerability by livelihood. Bivariate analysis indicates significant correlation between and the collection of wild foods to eat (p=. 38, n=72). However, none of these was signi ficantly significant when included in regression models that control for demographic and wealth characteristics of the household. As seen in Table 3 6, among agricultural
89 households, having access to a cereal bank and regular migration were the only coping mechanisms significantly positively correlated with self reported resilience. Table 3 6 Nested Multivariate Regression Models Examining the Effects of Demographic Characteristics and Coping Strategies on Self reported Resilience, among Agricultural House holds Independent Variables Model 1 Model 2 Model 3 Household Size .059* .055* .063* Sex of HHH .018 .011 .007 Age of HHH .016* .024 .122 Any Education HHH .369 .325 .279 Ethnicity Hausa ------Tuareg .578 .718^ .517 Fulani 1.462** 1.255* .511 Beri Beri .132 .058 .149 Geographic Zone Agricultural ------Agropastoral .849^ .737^ .386 Pastoral .797* .686^ .258 Wealth .165* .184* Eat own harvest .314 Reduce amo unt of food .249 Collection of wild foods to eat .070 Cereal bank .489* Regular migration .462* Constant 2.612 2.403 1.994 Model Fit Subset F 2.346* 2.727* 2.545* Adjusted R square .149* .200* .267^ N 70 70 69 Notes: ** p<.00; *p < .05; .05 < ^p < .10 Values shown in each cell are unstandardized coefficients Discussion Analysis of data from seven communities in eastern Niger shows no significant difference in the perception of risk from climate change across the agricu ltural pastoral livelihood continuum. This finding is important, as future researchers and policy makers
90 will work with various groups to appropriately adapt to and cope with the negative consequences of climate change and will be required to identify subg roups within which the experience of climate change differs (e.g., sex, wealth, ethnicity, nearness to road, access to grain bank, etc.). It is important to note that the lack of significant findings may be in part due to the narrow range of livelihoods in cluded in the sample population. Otherwise stated, a sample population that included households with a broader range of agricultural and pastoral livelihood practices might find different results when measuring the perceived risk of climate change. Though this finding must therefore be validated in other locations and with other research tools, it suggests that livelihood is not a determining factor in the perception of household risk to climate change. However, perception of climate change is correlated w ith specific coping strategies and the nature of this relationship varies significantly by livelihood. Only the collection of famine foods is significantly positively correlated with perceived risk of climate change for all livelihoods. More often, differe nt coping strategies are associated with PRCC for harvest significantly positively correlates with PRCC among agricultural households and significantly negatively correla tes with PRCC among pastoral households. Though temporal sequence cannot be determined through this cross sectional survey, this s own harvest. Similarly, coping strategies concerning mobility differentially affect the pastoral populations of this study; bivariate analysis indicates that reduced mobility is positively associated with the perceived risk of climate change, and linear regression illustrates that regular migration is negatively
91 associated with the perceived risk of climate change. These findings reinforce the significance of mobility in pastoral resilience, as those who are more mobile are those reporting lower risk of c limate change. This research also indicates that certain strategies are associated with vulnerability. For pastoralists, reducing the quality of food and accepting commercial loans are both strategies associated with decreased vulnerability. However, this association varies between livelihood groups. Thus, because perception of climate change is associated with the use of certain coping strategies which themselves are associated with vulnerability, but vary based on livelihood it is too early to relinq uish the idea that the perception of climate change might be acting as a driver for certain coping mechanisms that have consequences on household vulnerability. In the example outlined above, if mobile pastoralists are more likely to report lower risk of c limate change, one possible explanation is that these populations reduce mobility as risk increases; equally possible, however, is that risk related to climate change increases as pastoralists decrease their mobility. Further research is needed to clarify this association. In analyzing the relationship between climate change, coping strategies, and household vulnerability, a number of confounding factors, included in the regression models, surfaced as significant determinants of PRCC. Not surprisingly educa tion, household size, ethnicity, and geographic livelihood zone all had a significant effect on PRCC within the regression models. Education and northern (pastoral) proximity of livelihood zone were positively correlated with PRCC, as was being Fulani (as compared to Tuareg, Hausa or Beri Beri). Among pastoral populations, Fulani have
92 significantly higher self reported vulnerability than Tuareg populations and have a significantly higher reported PRCC. These findings support qualitative information indicati ng that Fulani populations were severely affected by the 2010 crisis due to the widespread loss of cattle and donkey populations (the primary large livestock of the Fulani); camel populations (historically herded by Tuareg) were less affected. Variation in household size and education are presented in Table 3 7 below. This data is important when interpreting results. Table 3 7. Percent of household heads with any education and mean household size by ethnicity. with any education Mean household size Hausa (32) 9.4% 8.91 Tuareg (44) 2.3% 6.38 Fulani (31) 3.2% 9.97 Beri Beri (23) 17.4% 8.21 Because of the dynamic nature of pastoral livelihoods in Niger, this research developed and utilized a livelihood index, in lieu of ethnic ity or primary income source, in its attempt to identify and understand how pastoral households are experiencing and responding to climate change. Though the seven point livelihood index did not have a significant effect on either the PRCC or vulnerability models in this research, creating a dichotomous variable from the index allowed for a more robust definition of pastoralism than is typically used. By creating the index before dichotomizing the population, the complexity of pastoral livelihoods is better captured, which is meaningful in an environment like Niger, where identification as an agriculturalist or a pastoralist is nuanced and complex. Table 8 provides a comparison of findings among agricultural and pastoral communities identifying coping stra tegies that are correlated with
93 Table 3 8 Comparison of coping strategies correlated with PRCC and vulnerability/resilience by livelihood Agricultural (1 3 3 ) Pastoral (4 7) Descriptive characteristics of livelihood (livelihood index 1 7): Highly sedent arized; more permanent dwelling structure; largely reliant on crop cultivation for household economy; some small stock; minimal if any migration with livestock; some history of migration with livestock, but moving toward agricultural production Historicall y nomadic herders; some practice transhumance; highly mobile population; focus on management of livestock herds (camel and cattle); non permanent dwelling structures; reliance on livestock, milk and meat for household economy Coping strategies associa ted with increased risk of climate change Wild foods Accepting commercial loans Accepting gifts Collecting famine foods Exode Reduced mobility Coping strategies associated with decreased risk of climate change Sale of personal belongings Regular mi gration Reducing quality of foods Sending family members elsewhere to eat Reduced food spending Removing a child from school Coping strategies associated with increasing vulnerability None Reducing the amount of food consumed Coping strategies as sociated with increasing resilience Grain bank Regular migration Accepting a commercial loan 3 These numbers represent positions along the livelihood index.
94 climate related risk and coping strategies that are correlated with vulnerability/resilience for each group. Use of the l ivelihood index in this way may be meani ngful in the development of appropriate livelihood interventions. Although more research is required to determine the full utility of a livelihood index, it is reasonable to assume that use of a continuum, as a tool for generating rich distinction between livelihood groups, will allow for more appropriate interventions to improve pastoral livelihoods. As pastoral populations face the imminent threat of climate change in the Sahel, researchers and policy makers must be mindful to define exactly who the popu lation is that they are targeting, and what it is about that population that renders them vulnerable mobility, geographic location, herd health, or otherwise. As evidenced by this research, the coping strategies employed by households during food crisis and their effects on household vulnerability vary by livelihood, thus the definition of what makes someone a pastoralist must be carefully examined and articulated.
95 CHAPTER 4 UNDERNUTRITION AND F OOD SECURITY IN NIGE R A STUDY OF AGROPASTORAL COMMUNI TIES FOLLOWING THE 2005 A ND 2010 FOOD CRISES Introduction Food insecurity is a chronic problem in Niger. The country is characterized by widespread poverty, a lack of infrastructure, low education, and erratic low rainfall. In addition, extreme climatic events associated with climate change, including floods, droughts, and extreme heat, are further disrupting the fragile systems of rainfed agriculture and livestock production that the population depends upon for food. Consequently, rates of malnutrition speci fically, undernutrition remain high, even in years of average or above average crop and livestock fodder production. Although malnutrition includes micronutrient deficiencies as well as overnutrition (obesity), the ongoing nutritional crisis in Niger is driven by undernutrition, or a lack of protein and/or calories. In years of crisis, such as the recent events of 2005, 2010, and the unfolding situation of 2012, undernutrition rates soar. The Nigerien government, along with national and international non governmental organizations, monitors rates of undernutrition, and early, heightened undernutrition rates, generated by ongoing surveillance of child growth, serve as an indicator of impending crisis. Nationally, the multiplicity of causes of undernutrition is complex and likely varies across crisis years, as does the national/international response triggered by crisis. In 2005, the national government refused to acknowledge and appropriately respond to the severity of the food crisis, and international resp onse was late and uncoordinated. Research indicates that the lessons learned by humanitarian aid groups in 2005 and changes implemented in policy and government response since then facilitated improved response during the 2010 crisis (Baker & Ngendakuriyo, 2011; Tsai, 2010) when once again, crop and
96 fodder production were low, and rates of undernutrition and child mortality were extraordinarily high. In 2010, response was timely, widespread and coordinated, compared to 2005 (Baker & Ngendakuriyo, 2011; Georgieva, 2011) This study examines risk factors associated with undernutrition in children 6 60 months in communities in eastern Niger. Utilizing data collected in the post harvest period just after the food crises in 2005 and 201 0, this study seeks to investigate the role of food security and humanitarian food aid as determinants of nutrition. The remainder of this article is arranged as follows. First, a review of the literature describes the nutritional situation of Niger, inclu ding a discussion of food insecurity as one of multiple causes of malnutrition, and provides information on the 2005 and 2010 crises. Next, the article outlines research methods that were utilized during data collection and data analysis. A presentation of research findings follows this chapter and includes 1) an analysis of undernutrition in the 2005 and 2010 crises, including qualitative explanations of the contrasting experiences of two communities during the 2005 and 2010 crises, and 2) a quantitative i nvestigation of the role of food security and humanitarian food aid in the 2010 crisis. Finally, the article closes with a discussion of key findings and their implications on future research and humanitarian relief and development activities. Background Niger and Nutrition Niger, a land locked country in the African Sahel, is consistently ranked among the least developed countries in the world, based on the United Nations Human Development Index (United Nations Development Program (UNDP), 2011) Its population is among the most rapidly growing in the world and relies almost entirely on
9 7 subsistence, rain comes from foreign assistance, and livestock and uranium are among the few exported goods. The population 16.1 million is projected to reach 30.8 million by 2010, due to the alarmingly high 3.6% growth rate. The child mortality rate is 167 per 1000 live b irths, one of the highest rates in the world (United Nations Development Program (UNDP), 2011) In this context, nutrition is in, arguably, a chronic state of emergency in Niger (B lackwell et al. 2010) The prevalence of total undernutrition (defined as weight for height Z scores [WHZ] of < 2 and/or edema ) in 2009, a non crisis year, was 12.3 percent in children under five, and prevalence of severe undernutrition (defined as WHZ o f < 3) was 2.1 percent (UNICEF, 2009) In Zinder region, the research area for this study, the prevalence of total undernutrition was 15.4% in 2009 slightly above the 15 percent threshold used by the UN and other aid organizations as an in dicator of emergency situation (UNICEF, 2009) Thus, even in normal years, nutrition is an urgent problem. Causes of Malnutrition It is widely understood that inadequate food intake particularly in the first two years of life leads to stunti ng and wasting in children. However, research has made clear that food availability, which is often restricted during food crisis, is not the only determinant of food intake or malnutrition (Smith & H addad, 2001) Breast feeding practices, food quality, and disease state within the household are among the many other factors that affect nutritional status in the household. In 1990, UNICEF produced a conceptual framework illustrating the causes of malnu trition at the household level that has been modified only slightly over the past 20 years (UNICEF, 1 990) The framework identifies
98 the immediate causes of malnutrition as inadequate food intake and poor health or disease. Importantly, the conceptual framework also identifies three underlying causes: poor health services and unhealthy environment (insuff icient clean water), inadequate care of women and children, and inadequate household food security. This section reviews each of these underlying causes of malnutrition and examines the contribution of each cause to the overall nutritional situation in Nig er. A poor health system and poor environmental conditions contribute to malnutrition by fostering disease, one of the immediate causes of malnutrition. In the context of a food crisis, poor health infrastructure is problematic because the population will be more vulnerable to malnutrition and other negative health outcomes because of the reduced state of health prior to the crisis. Poor health infrastructure and systems may exacerbate the problem of malnutrition in a crisis if the system lacks the ability to effectively accommodate increased numbers of malnourished or otherwise ill children caused by the crisis (Grobler Tanner, 2006) Although overall funding for health appears to be improving, as is evidenced by the near doubling of the percent of GDP spent on health from 3.6% in 2003 to 6.1% by 2009, Niger is still below the average for sub Saharan Africa (6.9% in 2009), and historic patterns of underfunding have lead to an extremely weak health system (UNDP, 2011) Recent research has shown that while private financial contributions in aid to Niger increased between the 1999 2002 and 2003 2006 time periods, government funding of healthcare decreased dramatically, a trend that is replicated throughout much of Sub Saharan Africa (Lu et al. 2010) In 2004, there were 377 physicians in Niger, or 3 per 100,000 people (World Health Organization 2006) A skilled birth attendant is present at only 33% of births and the total fertility rate is 6.9
99 (UNDP, 20 11) In general, health services are either missing entirely or largely dysfunctional. Basic health indicators demonstrate this point; under five mortality is 160 per 1,000 live births and life expectancy at birth is 54.7. Basic hygiene and sanitation is a problem, as 64% of the population lacks access to clean water and 89.3% lacks access to improved sanitation (UND P, 2011) Care of children particularly infants and young children, plays a significant role in determining their dietary intake and can be detrimental to nutrition, even in a context of food security (Brown et al 1988; Dettwyler, 1989; Shankar et al. 1998) Practices essential for infant and young child health include early exclusive breastfeeding, which has been shown to reduce mortality in developing countries (Edmond et al. 2006; Mullany et al. 2008) ; appropriate complementary feeding (Brown et al. 1988; Engle, 2002) ; and birth spacing (birth to birth) of at least three and no more than five years (Rutstein, 2005; Wuehler & Biga Hassoumi, 2011) A 2007 natio nal nutritional survey in Niger indicated that only 9% of children are exclusively breastfed for six months, as recommended by WHO, and less than 25% of children are fed within the recommended first hour of birth (Rpublique du Niger, 2007) Analysis of 2006 demographic and health data indicates that 50 75 percent of Nigerien children are not consuming age appropriate quantity or quality o f complementary foods (Wuehler & Biga Hassoumi, 2011) Food allocation within the household is also a key det erminant in dietary intake in Niger, affecting both the quality and quantity of food consumed, particularly among women and young girls who are often the last to eat. The low status of women in Niger problematizes nutrition, as the marginalized status of w omen has been linked to poor nutrition in children (Smith et al. 2003; Smith
100 & Haddad, 2001) Niger has one of the lowest female literacy rates in Africa (11.6% of women age15 24 are literate), one of the highest fertility rates (6.9) and one of the highest maternal mortality ratios (820 deaths per 100,000 live births; UNDP, 2011) All of these factors serve as an indicator of the low overall status of women (Shen & Williamson, 1999) The Gender Inequality Index is a tool developed and utilized by UNDP to illustrate the disadvantage of women through reproductive health, 0.724, ranking it 144 out of 146 countries (UNDP, 2011) The 2006 Demographic and Health Survey indicates that t he average age at first marriage another important indicator of is 15 in Niger, one of the lowest in the world (Macro International Inc., 2007) The poor status of women negatively impacts the health of the household, particula rly children. Where women have better access to and control over food resources, both women and children have better health outcomes, including lower rates of undernutrition. Where women lack this control, due in part to their status and the care society takes of women, children are much more likely to suffer from undernutrition. Food security is defined as a state where all people at all times have access to sufficient, safe, nutritious food to meet their dietary needs and food preferences for a n active and healthy (Food and A griculture Organization, 1996) There are three dimensions of food security: food availability, food access, and food intake (Webb et al. 2006) These dimensions are hierarchical each one a requirement for those dimensions which f ollow. The relationship between household food access (which requires food availability) and
101 household nutrition (a metric for food intake) is complex. Securing access to food does not necessarily eliminate malnutrition; it is a necessary but not sufficien t condition (Pinstrup Andersen, 2009) Even when fami lies are able to bring adequate amounts of nutritious food into the household, this does not ensure that everyone in the house will be well nourished feeding practices of infants and preferential feeding of boys illustrate how age and sex of a child may create increased risk for undernutrition. Household food insecurity is a chronic problem in Niger, as both availability and access are insecure. There are multiple ways that food security is routinely threatened: 82% of Nigeriens live in severe poverty, so financial access to enough quality food is a regular obstacle for many households (UNDP, 2011) ; seasonal variati ons in food availability for humans and livestock (in good years) mean fluctuations in household income and expenditure on food and, thus, regular annual periods of limited food supply or hungry seasons (Grobler Tanner, 2006) ; and political instability restricts movement, which may limit physical access for buyers and sellers to markets, and/or hamper humanitarian aid activities that aim to strengthen liveliho od and food security through various interventions (Tsai, 2010) Additionally, those who have money or other assets control the foo d that is available. Climatic events are also a major threat to food security in Niger, as c limate change is predicted to have a number of dire consequences on the agricultural sector, largely due to changing rain patterns, with increased incidences of bot h flooding and drought (Nelson et al. 2002) Increased frequency and severity of extreme climatic events, characteristic of climate change proj ections for the Sahel, have manifested recently in years of poor agricultural production and livestock fodder growth, limiting both
102 availability of grains and access through market exchange (Battisti & Naylor, 2009; Ben Mohamed et al. 2002; Held et al. 2005; IPCC, 2007) This situation, overlaid on a web of political and economic realities distinct to each year and underlain with continued inequities (e.g., gender, ethnicity), lead to situations of food crisis in 2 005 and 2010. The existing high burden of disease, low rates of early exclusive breastfeeding, weak health infrastructure, and poor health and sanitation render Nigeriens nutritionally depleted independent of crisis years and extremely vulnerable to ma lnutrition in the face of environmental shocks. 2005 Crisis The 2005 food crisis was the result of both structural determinants as well as distinct events that occurred that year. As previously outlined, a number of structural causes of food insecurity an d undernutrition persist in Niger; these include widespread poverty, rapid population growth, climate change, and unsustainable agricultural practices. In addition to these factors, the Nigerien government had been drawing down its National Food Reserve St ock for a number of years, so that by 2005 it was well below the 50,000 metric tons (MT) of reserved grains and the financial reserves to purchase an additional 60,000 MT that were supposed to be in place, based on historic food security policies (Wilding et al. 2005) This was the situation in 2004, when a regional cricket infestation decimated pastures and early cessation of rains caused scattered crop failure and deficits throughout the countr y. While the cricket infestation and poor rains certainly affected crop production, harvest deficit was 223,500 MT, only 11% below the 1999 2003 average (Government of Niger et al. 2004; USAID, 2005) There was, however, a 4.6 million MT fodder deficit and the impact of livestock famine on food s ecurity and human suffering was not appropriately included in the assessment
103 and response of the 2005 situation (Wilding et al. 2005) Overall, there was very poor analysis of the regional situa tion as well. The cricket infestation was foreseeable, as the damage was reported months before in neighboring countries of Mauritania and Senegal. Effective regional analysis of grain markets could have predicted a shortfall of grain availability, but ins tead Nigerien officials designed a plan that relied heavily on importation of grains from Nigeria and other neighboring states. Regional markets hampered this plan when borders were closed, prohibiting the importation of grains into Niger, and countries wi thin the region reduced exportation to Niger, both to meet their purchasing power is greater (Mousseau & Mittal, 2006; Wilding et al. 2005) Nigerien President Mamadou Tandja downplayed the severity of the situation, delaying subsid ization of foods and refusing distribution of free food until it was much too late to achieve positive impact (Tectonidis, 2006) Consequently, gra in prices soared to 2 3 times their normal prices during the lean season of 2005; undernutrition rates rose above 20 percent; 3.3 million people were affected by food shortage; and thousands of children died (McKune & Silva, 2012; Mousseau & Mittal, 2006; Tsai, 2010) The humanitarian response was delayed and unorganized, culminating in the bulk of aid arriving too late (Cornia & Deotti, 2008; Mousseau & Mittal, 2006; Wilding e t al. 2005) Nearly all of the aid that came into Niger during the 2005 crisis went through international relief organizations, rather than through the Nigerien government, where it might have had the added benefit of reinforcing existing systems and imp roving capacity of the state to respond (Wilding et al. 2005)
104 2010 Crisis Due to a shorter than normal rainy season, like 2004, the 2009 growing season resulted in reduced crop and livestock f odder production. After harvest, there was a national cereal deficit of 410,000 MT; an estimated 40% of agricultural villages lost half or more of their crop production and total grain yields were on average 30% below the (Tsai, 2010) As well, pasture production of fodder was estimated to meet only 67% of livestock requirements, which lead to massive and pre mature transhumance to the southern part of the country (FEWS NET, 2010) A national survey conducted in April 2010 indicated that 7.1 (47.7%) mill ion people, nearly half of the population were at risk of food insecurity and that 22.2% were extremely food insecure and 25.5% were moderately food insecure (Baker & Ngendakuriyo, 2011) A nutritional survey from late May/early June 2010 showed the rate of total undernutrition among children 6 60 months to be 16.7 % and of children 6 24 months to be 26.1% (National Institute of Statistics & Ministry of Public Health, 2010) As in 2005, President Tandja downplayed the situation; ho wever this position was quickly reversed militia government quickly put together an emergency appeal for international assistance. Recognition of the looming crisis and timel y appeal to international donors by the Government of Niger was followed by an improved response by the international community. Where previously there was a lack of coordination, NGOs clustered into four areas active during the response to ensure effectiv eness, efficiency and impact (Wilding et al. 2005) Aid arrived during the sourdure the hungry season, rather than weeks before the next harvest, and it went through an existing public system d esigned and structured to deal with this type of crisis: the National Framework for the Prevention
105 and Response to Food Crises ( Dispositif National de Prvention et de Gestion des Crises Alimentaires ) Rather than aid going almost entirely through internat ional aid organizations, as it did in 2005, it served to reinforce capacity of existing infrastructure within the Nigerien government. In addition to an improved response by a government that no longer approached food crise s as taboo and by an engaged and orchestrated international community, the regional markets were better functioning in 2010 which allowed for improved food security in Niger through international safety nets (Tsai, 2010) Despite these improvements, the 2010 crisis saw some setbacks compared to the 2005 crisis, due in large part to political security. In 2007, after a 20 year peace, unrest in the northern pastoral zones of Niger was reignited, catalyzing sporadic outbreaks of violence. The situation remained tenuous through the 2010 crisis. Political insecurity in the region restricted the movement of humanitarian and development assistance personnel into a priority zone, as well as the movement of Nigeriens who might otherwise have been mobile in these areas as part of a coping mechanism during the crisis. In addition, the insecurity in this area leading up to 2010 led to the withdrawal of a number of development an d economic programs that were investing in the area. The presence and activity of an al Qaeda cell, al Qaeda in the Islamic Maghreb (AQIM), in Niger further complicated aid efforts throughout the country; as the safety and security of expatriate staff work ing with international NGOs was jeopardized and aid workers were increasingly targeted, more and more NGOs pulled out of the region. Methods Research assistants collected data from November 2005 through February 2006 and from October 2010 through January 2011. Tanout, Niger served as the research
106 base and the University of Arizona and British Red Cross (BRC) Cash Distribution Monitoring and Evaluation project identified the original sample (Bureau for Applied Research in Anthropology, 2005, 2006a) In the BRC project, a research team visited 361 households from 19 communities four times over the course of a year, and col lected household and community level data through surveys, key informant interviews, focus groups, and anthropometric measurements. Data utilized in this research comes from one of four phases of this project. I served as a project manager on the BRC proje ct and was involved in data collection and analysis from November 2005 through the end of the project in February 2007. I designed the 2010 research to utilize and build upon the 2005 data. Thus, the 2010 research targets four of the 2005 communities, alon g with three new ones, for participation during the same post harvest time period following the crisis, and collects comparable nutritional data, as well as new data on mobility and climate change. The seven communities included in the sample represent li velihoods along an agricultural pastoral continuum. Some of these livelihoods are more reliant on livestock herds and high mobility, while others are more sedentarized and reliant on agriculture. Originally, the seven communities were selected from the 19 BRC communities, based on livelihood composition. However, because nomadic pastoralists were not well represented in the 2005 pool of communities, and because representation across the livelihood spectrum was important to the study, two nomadic pastoral co mmunities with high levels of mobility and an ethnic composition that contributed to sample diversity (Eliki and Djiptoji) replaced two of the communities originally targeted. A third new community (Abdounez), comparable in livelihood, ethnic make up, and location,
107 replaced one of the 2005 communities (Farak) when insecurity obviated the possibility of researchers accessing it in November 2010. In repeat communities, researchers targeted the same households interviewed in 2005 for inclusion in the 2010 sa mple (BARA 2006a) Where original households were not available (this occurred less than twice per site), research assistants randomly selected new households for inclusion. In new communities, research assistants randomly selected 19 househ olds from a census list for inclusion in the household survey sample, replicating household targeting techniques employed in 2005 in order to maximize consistency. In each community, researcher assistants conducted m ale and female focus groups with randoml y selected individuals; conducted a minimum of three key informant interviews ( elders, teachers, cereal bank managers, health workers, and spiritual leaders ); and collected anthropometric measurements of children 6 60 months. Focus groups of six to eight i ndividuals were conducted within the community, usually on mats in open air or in tent dwellings. Although the research team attempted to secure privacy, inevitably community members gathered to listen and sometimes contributed to the focus group discussio n. Researchers discouraged participation by non members, but community norms of inclusion often trumped researcher efforts and groups became as large as 12 15 people. Field assistants collected anthropometric measurements, including height, weight, age, a nd middle upper arm circumference (MUAC), on all children 6 60 months of age from targeted households. Where this number was less than 40, the sample included additional children from the community until 40 individuals were measured. In some communities, s ample households contributed more than 40 children, thus there is some
108 variation in the number of children per community. As well, in some communities the number of children coming from targeted households was quite low, thus the sample size of children fo r whom household level data is available was reduced (see Table 4 1 below). The food security variable for this research uses the perception of vulnerability and use of coping mechanisms as an indicator of household food access. I developed this variable using a method that assesses the frequency of use and severity of coping mechanisms employed by households during the crisis (Maxwell et al. 1999; Maxwell, 1996) Though researchers have recently developed a univer sally accepted and standardized measurement tool for food security (Household Food Insecurity Access Scale), this study uses a predecessor to that tool to ensure that data from 2005 and 2010 are comparable (Swindale & Bilinsky, 2006) The self reported coping strategies included in the food security variable include: changed e ating habits, decreased portions at meals, decreased nutritional quality at meals, acceptance of gifts/loans from family and friends, decreased number of meals, collection of wild famine foods, day without eating, and sale of personal belongings. Calculate d based on reported use of variable is a scale from 1 15. For this study, I dichotomized the variable and assigned households a status of either food insecure or secure. Each of the regression models estimated in this research includes key demographic characteristics as control variables, including ethnicity. Despite that ethnicity is fluid and intermarriage between ethnic groups is not uncommon in Niger, ethnic groups m aintain a shared set of opportunities and disadvantages, as well as
109 cultural norms and household practices, that are more similar within than across groups. The differences that exist across groups may have significant effects on the outcome of the model, thus ethnicity is included as a control variable. Two Nigerien doctoral students (both male) studying pastoralism in Niger served as research assistants. Four former Red Cross team leaders (one female, two male) served as field assistants. For two years th ey worked and received extensive training when we worked together on the monitoring and evaluation project that generated the 2005 data. Those who conducted anthropometric data were trained and certified by Mdecins Sans Frontires and have substantial exp erience in collecting child growth measures. Research and field assistants collected all data in the primary household languages of Tamachek and Hausa. Research assistants translated responses into French, and I analyzed all data in French. Research and fi eld assistants originally entered all data on site (in Niger) into Microsoft Excel then imported into either SPSS (SPSS, 2011) (CDC, 2008) I used EpiInfo to generate nutritional z scores (weight for age [WAZ], height for age [HAZ], an WHZ) based on CDC 2 000 growth curves. I then imported these data into SPSS where all other analyses were completed. 2005 and 2010 data sets included hierarchical data made up of household level data (n = 95 and 133, respectively) and child growth data (n = 200 and 290). I me rged these data into a single data set allowing for a comparison of 2005 and 2010 data. I then cleaned the data and screened key variables for distribution and normality. The dependent variable, undernutrition, defined here as WHZ of less than 2, is norma lly distributed in both the 2005 and 2010 data set.
110 Table 4 1 Sample Size for 2005 and 2010 Household Survey and Anthropometric Data Household data Child anthropometric data Linked child anthropometric and household data 2005 95 200 144 2010 133 290 1 69 One way ANOVA tests were used to compare mean undernutrition (WHZ) between groups. Bivariate tests were run to determine correlation between variables non parametric tests were necessary due to non normal distribution of data or use of categorical variables. Other analysis included chi square tests to establish correlations between categorical and dichotomous variables, except where sample size was too s linear regression to estimate multiple models that regressed undernutrition on food security and humanitarian food aid, as well as undernutrition on other potential determinants of undernutrition, in cluding various coping strategies, consumption patterns, health, and environmental determinants. I initially controlled for age and sex of the child in all models. Findings Distribution of Undernutrition 2005 and 2010 The 2005 sample population was compris ed of 200 children from five communities. The average child in this sample was almost 32 months of age and mildly malnourished (z score between 2 and 1, see Table 4 2). At the time of data collection, approximately 14% of children presented with fever, 1 5% with diarrhea, and over a quarter (25.5%) were enrolled in a feeding program. The 2010 sample population consists of 290 children from seven communities. Household data are available for 155
111 of these children. In certain cases, multiple children were sa mpled from the same household. The average child in the sample was approximately 30 months old, mildly malnourished, and from a household with moderate to high food insecurity (see Table 2). Nearly one third (31.7%) of all children presented with fever, ne arly one quarter (24.8%) presented with diarrhea, and nearly one fifth (19.0%) of all children were actively enrolled in a feeding program at the time of data collection. Table 4 2. Descriptive statist ics of sampled children, 2005 (N=200) and 2010 (N =290) Mean SD 2005 2010 Age (months) 31.7 15.9 29.5 15.2 Weight (kg) 10.8 3.1 11.3 5.8 Height (cm) 84.4 12.6 84.1 16.1 MUAC (mm) 139 13 137 12 Underweight z score 1.2 1.2 1.1 1.2 Food security (1 15, secure to insecure) NA 10.3 3.9* Percent 2005 2010 Male 46.5 46.2 Presenting with fever 13.6 31.7 Presenting with diarrhea 14.6 24.8 Currently enrolled in a feeding program 25.5 19.0 Received food aid NA 84.6* Note: Frequencies may vary due to missing values *N =155, relies on household data The rates of stunting, wasting, and underweight varied across communities in both 2005 and 2010. In 2005, total stunting rates (HAZ less than 2) ranged from a low 21.1 in Farak to a hig h 51.4 % in Dareram (see Table 4 3 for nutritional rates by community in 2005). Total wasting rates (WA Z less than 2) ranged from 16.2 % in Farak to 51.5 % in Dareram. Total underweight rates (WHZ less than 2), ranged from 13.3 % in Guinea Alhazaye 29.4% in Takoukout. All of the 2005 communities (Guinea Al hazaye, Kkeni, Takoukout, and Farak) had total underweight rates above the emergency threshold of
112 15%, except Dareram, where total underweight rate was 13.3%. Also of note, rates of severe underweight ranged from alarmingly high in communities of Takoukou t (11.8%) and Guinea Alhazaye (11.4%) to zero in Kkeni. Kkeni was the only community to have severe rates of underweight below the emergency threshold of 2%. Although rates of stunting and wasting are important measure of chronic and acute undernutritio n, the difficulty in obtaining reliable age data on children renders these measures less reliable than weight for height measures. Because WHZ are widely accepted as the most reliable summary measure of maln utrition, h ereafter, malnutrition will refer to t otal malnutrition, defined as the population with a WHZ of less than 2, unless otherwise noted. A comparison of means using one way ANOVA indicated no significant d ifference in undernutrition by community, though means ranged from 0.82 in Kkeni t o 1.57 in Guinea Alhazaye (p=0.16, n = 172, see Table 4 4 ) Table 4 3 Percent of total* and severe** stunting (HAZ), wasting (WAZ), and underweight (WHZ) by community, 2005 Community Stunting Wasting Underweight Total Severe Total Severe Total Severe Darera m 51.4 29.7 51.5 18.2 13.3 3.3 Guinea Alhazaye 41.0 23.1 47.1 20.6 28.6 11.4 Kkeni 32.3 16.1 48.6 28.6 15.2 0.0 Takoukout 28.1 6.2 40.6 12.5 29.4 11.8 Farak 21.1 2.6 16.2 2.7 21.6 2.7 Total (N =172) 35.0 15.8 40.4 16.4 21.9 5.9 Note: Frequencies m ay vary due to missing values Table 4 4. Weight for height z scores by community, 2005 Community Mean N Std. Deviation Dareram 1.11 31 1.33 Guinea Alhazaye 1.57 36 1.15 Kkeni 0.82 33 1.19 Takoukout 1.29 34 1.38 Farak 1.18 38 1.18 Total 1.2 0 172 1.26 Note: One way ANOVA comparing mean WHZ by community: F = 1.66, p=.16
113 Similar to 2005, the prevalence of stunting, wasting and underweight in 2010 varied across sample communities (see Table 4 5). Total stunting rates ranged from a low 14.3% in Abdounez to a high 44.2% in Djiptoji. Total wasting rates ranged from 17.1% in Abdounez to 56.1% in Kkeni. Total underweight rates ranged from 2.6% in Abdounez to a high 34.2% in Eliki. Abdounez and Dareram had rates of total underweight below emerge ncy threshold, while all other total underweight rates (Djiptoji, Eliki, Guinea Alhazaye, Kkeni, and Takoukout) were above the 15% emergency threshold. For severe underweight, only Abdounez had rates below the 2% emergency threshold. Though this variatio n appears similar to 2005, variation by community is greater in 2010; one way ANOVA comparison of means indicated a significant difference in undernutrition by community in 2010, ranging from .33 in Abdounez to 1.61 in Eliki (p=.000, n = 263, see Table 4 6). The communities of Eliki and Djiptoji, both predominately Fulani communities, have the lowest mean WHZ, while Dareram and Abdounez, both predominately Tuareg communities, have the highest mean WHZ Though Dareram is an agropastoral community today, all four communities have pastoral roots, and all households within Eliki, Djiptoji and Abdounez practice nomadic pastoral ism today. The contrasting experiences of these pastoral communities may be attributed to a number of practices that vary by ethn ici ty, many of which are beyond the scope of this research. However, of significance to this research may be the historical reliance of Fulani pastoral populations on donkeys In multiple focus groups and household interviews, participants shared stories of s tarving animals and repeatedly told researchers that they had no memory of a crisis where donkeys died
114 Table 4 5. Percent of total and severe stunting (HAZ), wasting (WAZ), and underweight (WHZ) by community, 2010 Community Stunting Wasting Underweight Total Severe Total Severe Total Severe Abdounez 14.3 4.8 17.1 2.4 2.6 0 Dareram 32.5 10.0 32.5 10 8.6 2.9 Djiptoji 44.2 11.6 58.1 25.6 29.3 4.9 Eliki 26.8 2.4 51.2 24.4 34.2 10.5 Guinea Alhazaye 26.8 12.2 36.6 14.6 32.5 5.0 Kkeni 39.0 22.0 56 .1 26.8 27.8 5.6 Takoukout 21.1 10.5 34.1 14.6 26.5 14.7 Total (n=263) 29.4 10.5 41.0 17.0 23.2 6.1 Note: Frequencies may vary due to missing values Table 4 6. Weight for height z scores by community, 2010 Community Mean N Std. Deviation Abdounez .33 39 .98 Dareram .84 35 .99 Djiptoji 1.37 41 1.4 Eliki 1.61 38 .99 Guinea Alhazaye 1.34 40 1.06 Kkeni 1.12 36 1.13 Takoukout 1.31 34 1.38 Total 1.13 263 1.20 Note: One way ANOVA comparing mean WHZ by community: F = 5.312, p=.000 Asked wh y they believed donkeys had died at such high rates in this crisis, community members indicated that rains were better further north than in the southern pastoral zone, so communities had traveled farther away than normal and arrived to find crowded pastur es. Many of the donkeys did not survive either the trip home or travel to the next area for grazing, whereas more camels survived these moves. The question of ethnicity will be further addressed in the discussion section. Communities were targeted to allow for repeat measures of undernutrition, thus mean undernutrition was compared for 2005 and 2010 in repeat communities. The community of Farak, initially included in the 2010 target sample, was not accessible during data collection in 2010 due to civil inse curity in the area. Research assistants
115 identified a replacement community, Abdounez, based on similarities in livelihood (pastoral), ethnic composition (largely Tuareg), geographic location of meeting point and historical migration patterns. Farak/Abdounez (see Table 4 7). Given that these are tw o distinct communities, the dyad was not considered a repeat measure from this point forward. Table 4 7. Comparison of undernutrition (mean WHZ) between 2005 and 2010 by community Community Mean WHZ N SD F P Dareram 2005 1.11 31 1.33 0.918 0.342 2010 0.84 35 0.99 Guinea Alhazaye 2005 1.57 36 1.15 0.834 0.364 2010 1.34 40 1.06 Kkeni 2005 0.81 33 1.19 1.168 0.284 2010 1.12 36 1.13 Takoukout 2005 1.29 34 1.38 0.005 0.946 2010 1.31 34 1.38 Farak/Abdounez 2005 1.18 38 0.98 0.933 0.001 2010 0.33 39 1.18 Total 2005 1.20 172 1.26 0.332 0.565 2010 1.13 263 1.20 In 2010, households were asked a series of questions to determine their overall self reported resilience (see Appendix B for full Household Questionnaire). For ea ch of seven items (land degradation, vegetation cover, soil productivity, health and sanitation, wealth, food security, and mobility), they indicated whether they thought their current situation was declining, about the same, or improving, compared to 2005 Based on this data, significant difference existed when comparing mean community self reported resilience (scale of 1 3, where 1 is more vulnerable and 3 is more resilient; p=.000, n=165). This measure of self reported resilience was compared to the aver age change
116 in community WHZ for the four repeat communities (see Table 4 8). Chi square test indicates a highly significant correlation between self reported resilience and change in mean community undernutrition (p=.000, n = 121) Table 4 8 Change in mean community undernutrition and self reported resilience Community Mean WHZ Mean self reported resilience Dareram .27 2.75 Guinea Alhazaye .23 1.62 Kkeni .31 1.85 Takoukout .02 1.56 Although analysis shows no significant difference in rates of un dernutrition between crisis years of 2005 and 2010, Dareram and Kkeni represent the greatest improvement and decline in nutritional status between 2005 and 2010; thus, an analysis of the two communities, their coping strategies, aid activities in the comm unity and other determinants of food and livelihood security are presented here. These data come from key informant interviews and focus groups at each location. Interviews and focus groups included discussion of the 2005 and 2010 crises, as well as histor ical years of crisis. Most often the events of 1973 and 1984 were referenced without prompting, but where past years of hardship were not volunteered, researchers asked specifically about these years, as both were events associated with a significant loss of human and livestock life in the region. Quantitative data from child growth surveys and household surveys are included, where relevant. Case study: Dareram Dareram is a community of about 200 people situated in three small settlements. Predominantly Tu areg, community members are agropastoralists of the Icherifan clan those that claim decent from the prophet Mohammed. Slaves of the noble Tuareg, their
117 ancestors escaped and formed the community of Dareram. Initially families continued to practice transh umance, and at least some part of the family would travel with small ruminants, camel and cattle toward Agadez. But this practice has diminished over time were pastoralists focus group, 2010). With unreliable rainfall and years of crisis repeating, mobility has declined and reliance on crop production has increased. Not everyone cultivated crops before, but now co mmunity members feel it is clear that animals will be lost with each crisis, so everyone has started cultivating crops. This community has historically relied on natural resource base of the area (such as famine food) during crisis years. Women would coll ect fodder to sell at market; household heads and families would collect wild foods to survive on through the crisis. Heads of household and young men would also travel en exode in search of work or food. In good years, youth travel en exode to earn money with which they buy their own animals to raise, but the strategy is employed by heads of household as well during times of crisis. In 1973, heads of household traveled south with camel and effectively secured food supply for those at home. When these effor ts fail, however, entire families flee the area, as happened in 1984 the worst crisis in living memory for this community. The entire community fled to Nigeria, Tanout, Zinder and Agadez. Most went to Nigeria where they stayed 2 3 years, and some have ne ver returned. Many social ties within the community were broken at this time, and community members see this as a turning point in their vulnerability to crisis.
118 The 2010 crisis was not as bad as 2005 for this community. In 2005 there was a rain deficit, a nd some of the households did not produce any grain. Regionally, the price of millet passed 1100 FCFA and was scarcely available at market. Those who were able to cultivate a harvest shared what they had with those whose harvest failed entirely; however, g eneralized production deficit lead to crisis within the community, and no outside aid was received. During the 2005 crisis, households reportedly collected Boscia (wild food) more than 50 km north of Dareram to bring home to their families to eat. In 2010, there was always grain at market and the price of millet did not surpass 500 FCFA. The entire community stayed in place, with the exception of young men who left en exode largely to Nigeria. The financial contributions of those exodants are cited as inst included small scale commerce (sale of tea and jewelry) and quick sale of livestock early in the crisis. By June, however, reserves were empty and the community was in crisis NGOs including Catholic Relief Services began distribution of substantial food aid to the community. This is in distinct contrast to 2005, when Dareram received no aid. 2010 aid came in four phases: 1) distribution of corn and sorghum (50 kg per 7 people ), 2) distribution of corn and beans (12.5 kg per person), 3) blanket distribution of 50 kg corn to all families with a child under five, and liter of oil per child under five, and 4) blanket distribution of 50 kg of corn per child under five. The commun ity cites the loss of men en exode and declining livestock holdings, due to death and sale, as the consequences of the 2010 crisis, and they attribute financial contributions of exodants and humanitarian aid as the most important determinants of their abil ity to endure the 2010 crisis.
119 Participants identified repeated years of poor rainfall as the greatest threat to food security and nutritional status within the community. Youth and household heads leave to find work, but they are not always successful. T hose that are left at home are no longer able to rely on natural resources (wild food) while they are gone. One must travel farther and farther north to find those wild foods that remain. The community relied heavily on humanitarian aid during the last cri sis and recognizes that there is no guarantee that there will be aid during the next crisis. Case study: Kkeni Kkeni is an agricultural community that is heavily reliant upon rainfed crop cultivation. Some households keep a small number of livestock (av erage TLU/RA = .19), and some practice off season gardening, cultivating tomatoes, lettuce, cabbage, potatoes and or squash, depending upon the year and water availability, but millet cultivation is the primary livelihood of most houses. Keeping livestock (particularly cattle) and off season gardening are largely the responsibility of women. Historically, the community also hunted. Around 1970, the number of wild animals, particularly guinea hen and dear, began declining, which largely eliminated hunting by this community. The community is made up of over 100 households, and access to water is a major constraint. There are two open cement wells that function during the wet season. Only one well functions during the dry season; it used for both animal and hum an consumption at that time, which leads to a lot of tension within the community concerning water use and distribution. The community is over 40 km from a health center, requiring six hours by cart or walking, and women have died in labor while traveling to the health center.
120 Coping strategies during crisis years in this community historically included gathering of wild plants for human and animal consumption, hunting, and migration particularly among the poorest families. A majority of the population f led to either Matameye or Zinder in 1973 and to Tessaoua in 1984. There was no aid at that time unless families relocated to refugee camps in cities or towns nearby, so once hunting and collection of wild foods failed, families fled. In 2005 and 2010, some aid reached the community of Kkeni, so many people stayed. In hope of receiving aid within the community, many households (24%, n=33) sent young men en exode to Nigeria, Libya, Agadez, or even Europe, in search of work. Both male and female community me mbers fear that this strategy is undermining the strength of the community because it is breaking families apart. In terms of rainfall deficit and crop production, residents indicate that the 2010 crisis was worse than 2005 in Kkeni, but not as bad as 19 73 or 1984. However, the impact, they argue, was worse in 2010 than 2005, 1984, or 1973, because there is now only one species of wild food to collect and no animals to hunt. Other factors worsening the 2010 crisis include the short time between 2005 and 2 010 crises, and the difference in amount of aid. In 2005 households received 120,000 FCFA from the Red Cross and a food for work project was active in the area. In 2010 families have received an average of 2 sacs (50 kg) of grain, in large part through bla nket feeding programs for children under five, and protective rations for children enrolled in feeding programs (36% of children in feeding program, n=41). However, before aid arrived, an estimated half to two thirds of the household heads sold livestock, purchased what grain stock they could, and then traveled en exode to seek work in either Agadez or nearby
121 communities that had sufficient crop production. Those who remained received aid. The community identifies consecutive crises, population growth, lac k of health infrastructure, shifting temporal and geographic distribution of rainfall, decreasing quantity of rainfall, lack of aid, and the loss of natural resources as key determinants of difficulty in managing the 2010 crisis. Based on this analysis of Dareram and Kkeni, the loss of indigenous species of animals to hunt and plants to collect as food during crisis years is an important, recently developed constraint on livelihood and food security in years of poor rainfall. And, while variation in rainf all and years of rain deficit are normal for the region, the short amount of time between crisis years is making resilience to subsequent shocks evermore difficult. Coping strategies identified by both communities as seminal to successful management of cri sis are the financial support of family members en exode and food aid from humanitarian aid organizations. While the percent of families reporting the use of exode during the 2010 crisis is comparable between communities (25% in Dareram and 24.2% in Kkeni ), the reported receipt of humanitarian aid largely food aid was very different. Focus group and key informant interviews show that Dareram was the beneficiary of at least four phases of food aid in 2010 and reportedly received no aid in 2005; converse and received minimal aid in 2010. This information is validated by household surveys, which indicate that 94% of households in Dareram and 63% of household in Kkeni received some form of food aid during the 2010 crisis. There is a statistical significance in the difference in aid across all villages in 2010, and aid is statistically significantly lower in Kkeni that in Dareram (Fishers exact test, p=.021, n= 38). In addition,
122 explanat ions of the type of aid received by households in Kkeni are largely linked to children under five rather than the general population. Receipt of aid is not significantly correlated with food security (p=.38), nor is it significantly correlated with undern utrition (p=.60) Temporal sequence of when aid arrived would be interesting to measure, but is not captured in these data. As indicated in Dareram, coping strategies may have already been employed by the time food aid arrives, thus complicating interpreta tion of the findings. Food Security and Undernutrition As previously described, food security is a necessary condition for good nutrition, though a number of other underlying factors interact to determine if a child is malnourished or not. This section of the study examines the effect of food security on malnutrition in Niger, net of demographic and underlying health differences, and the role of food aid as a buffer against food insecurity and malnutrition. For the purpose of this research, food security i s calculated using a method that assesses the frequency and severity of coping mechanisms employed by households during the crisis (Maxwell et al. 1999; Maxwell, 1996) Changed eating habits, decreased portions at meals, decreased nutritional quality at meals, acceptance of gifts/loans from family and friends, decreased number of meals, collection of wild foods, day without eating, and sale of personal belongings are the coping strategies included in the food securi ty variable. Respondents were asked if they had utilized the strategy during the crisis year and with what regularity they employ that strategy (rarely, if ever; sometimes, often; almost every year). Calculated based on reported use of the strategy and loc al severity rankings, the food security variable is a scale from 1 15. For this study, households were assigned a status of food secure or insecure, by dichotomizing this scale into lower and upper 50
123 percentile. The sample is made up of 146 children for w hom anthropometric data and household food security were available. It is important to note that this food security indicator identifies household food insecurity relative to the sample population. This is extremely important to remember, as the prevalence of food security in Niger is among the lowest in the world. Thus, even external to the study. Due to the pervasive nature of food insecurity in Niger, the variable was dic hotomized to allow for only one distinction in a group that is largely food insecure. In order to assess the relationship between food security, food aid, and undernutrition in the 2010 sample population, a number of statistical tests were computed. First a one way ANOVA to compare mean WHZ by group was calculated for food security (see Table 4 9), as well as a number of other predictor and control variables. Results from the individual one way ANOVA calculations are included in Table 4 10 below. Table 4 9. Mean WHZ between food security groups Mean N Std. Deviation Food secure 1.11 107 1.16 Food insecure 1.48 39 1.33 Total 1.21 146 1.21 These results show significant differences in mean undernutrition when data are grouped by ethnicity; level of mobility; self reported resilience; consumption of beans, oil or bread in the 24 hours prior to survey; household water source; household reported presence of community vaccine campaign; child presentation with diarrhea, and child presentation with fever.
124 Table 4 10 One way ANOVA tests to compare mean WHZ by food security and other predictor variables N F Sig. Food security 146 2.643 .106 Received humanitarian aid 155 .276 .600 Wealth (1 5, grain and livestock values) 155 0.294 .882 TLU/RA 153 1. 180 .235 Household size 155 1.706 .066** Sex of household head (male reference) 155 1.961 .163 Age of household head 155 1.023 .446 Education of household head (none reference) 153 .509 .477 Ethnicity (Hausa reference) 155 5.377 .022* Livelihood 151 .606 .659 Mobility (migration this year) 155 8.832 .003* Self reported resilience (7 item vulnerability index) 151 3.618 .029* Perceived risk of climate change 152 .019 .981 Child consumption of beans (last 24 hours) 153 4.488 .036* Child co nsumption of oil (last 24 hours) 154 6.304 .039* Child consumption of bread (last 24 hours) 152 14.785 .001* Child consumption of fish (last 24 hours) 153 1.557 .214 Child consumption of meat (last 24 hours) 153 .570 .451 Child consumption of milk (las t 24 hours) 153 .797 .373 Household water source 154 3.598 .030* Community vaccine campaign 149 4.274 .040* Child presentation with diarrhea 263 5.616 .019* Child presentation with fever 263 3.00 0 .084* Calculations indicate a marginally signif icant difference in average undernutrition by household size. Negative findings of interest include no significant difference in undernutrition by wealth, calculated as either the value of grain and livestock holdings or by tropical livestock unit per adul t resident (cite); no significant difference in undernutrition by age, sex, or education of the HHH; and no significant difference in mean undernutrition by aid.
125 Table 4 1 1. Nested Multivariate Regression Models Examining the Effects of Demographic and Fo od Security Characteristics on Undernutrition in Children 6 60 months Independent Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Child age .011* .013* .010** .009 .008 .008 Child Sex Male ------------Femal e .081 .112 .128 .168 .207 .206 Head of Household Sex Male ----------Female .467 .404 .277 .193 .190 Age of HHH .003 .005 .008 .008 .012 Education of HHH .192 .350 .283 .269 .274 Ethnicity Hausa -------Tuareg .351 .189 .176 .175 Fulani .225 .129 .010 .011 Beri Beri .434** .457** .574** .575** Food security .327 .372 .373 Child presentation with diarrhea .115 .114 Ch ild presentation with fever .501** .501** Community vaccine campaign .415 .409 Household water source .080 .078 Received food aid .016 Constant .841 .800 .748 .551 .647 .623 Model Fit Subset F 2.9 8* 1.50 1.88** 1.32 1.38 1.27 Adjusted R square .022* .016 .044** .020 .032 .026 N 263 153 153 144 141 141 Notes: *p<.05; **p<.10 Values shown in each cell are unstandardized coefficients
126 Four multivariate regression models were estimated to exam ine if and to what extent food security affected undernutrition, net of child and household demographic characteristics, in the 2010 food crisis in Niger. A fifth model was estimated to examine how health and disease affect the predictive power of food sec urity and aid on undernutrition. Results are presented in Table 4 11. Because bivariate tests for correlation indicate no significance between undernutrition and child sex, HHH age, or HHH sex, these variables were removed and the models were estimated aga in. These findings are presented below in Table 4 12. Table 4 1 2. Nested Multivariate Regression Models Examining the Effects of Demographic Characteristics and Food Aid on Undernutrition in Children 6 60 months Independent Variables Model 1 Model 2 Model 3 Model 4 Child age .008 .007 .007 .007 Education of HHH .206 .111 .192 .210 Food security .271 .270 .273 Child presentation with diarrhea .192 .190 Child presentation with fever .361 .362 Community vaccine campaign .5 99** .578** Household water source .312 .317 Received food aid .028 Constant 1.040 .955 1.224 1.550 Model Fit Subset F .1.278 1.254 1.733 1.509 Adjusted R square .004 .005 .035* .028 N 141 141 141 141 Notes: p<.05; **p<.10; p=.106 Values shown in each cell are unstandardized coefficients Each of the predictor and control variables, identified in one way ANOVA tests as having significant differences in mean WHZ, were included in linear regression models
127 to e stimate their independent effect on undernutrition. Subsequently these models were repeated with other control variables, in order to better understand the relationship. Some key findings from these tests are included here. Wealth There is a marginally sig nificant decrease in undernutrition with increasing wealth, when wealth is defined as TLU/RA (p=.072); this relationship is not significant once sex and age of child are included in the model (see Table 4 13). Table 4 13. Linear regression model of undern utrition on wealth and demographic characteristics Model 1 Model 2 TLU/RA .391** .334 Child age .010 Child Sex Male --Female .082 Constant 1.295 1.022 Model Fit Subset F 3.287** 1.969 Adjusted R square .015 ** .019 N 198 198 Food s ecurity, HHH age, and ethnicity There is an increase in effect of food security on undernutrition when the age of the HHH is included as a control variable in the linear regression mode; when HHH age is included, the previous ins ignificant relationship between fo od security and undernutrition (p= .106) becomes highly significant (p=.05). An interaction term (food security x HHH age) was created and tested, but was not significant (p=.219 see Table 14 ).
128 Table 4 14. Linear regressio n model of undernutrition on food security, head of household age, and interaction term Model 1 Model 2 Model 3 Food security .366 .476* .694* HHH age .010 .005 HHH age X food security .030 Constant 1.110 .599 .844 Model Fi t Subset F 2.643 1.994 1.843 Adjusted R square .011 .014 .017 N 146 146 146 Notes: *p<.05; **p<.10 Values shown in each cell are unstandardized coefficients Food security and livelihood Regression analysis shows that although livelihood is not a significant predictor of undernutrition (F=.887, p=.348) inclusion of livelihood in a model estimating the effect of food security on undernutrition increases the negative effect of food insecurity on undernutrition (p=.089) Table 4 15. Linear r egression model of undernutrition on food security, head of household age, and interaction term Model 1 Model 2 Model 3 Food security .366 .415** Livelihood (index) .086 .070 Constant 1.110 1.458 1.331 Model Fit Subset F 2.64 3 .887 1.479 Adjusted R square .011 .001 .007 N 146 151 142 Notes: *p<.05; **p<.10 Values shown in each cell are unstandardized coefficients
129 Ethnicity Findings from both qualitative and quantitative analyses underscore the role that ethnicity pla ys as a potential determinant of undernutrition. Thus, a few additional analyses were run on ethnicity. ANOVA test to compare means shows a significant difference between mean WHZ by ethnicity (see Table 4 15). Table 4 16. Mean WHZ by ethnicity, 2010 data Ethnicity Mean N Std. Deviation Hausa 1.06 50 1.22 Tuareg 1.68 27 1.20 Fulani 1.46 29 1.27 Beri Beri 1.38 49 1.14 Total 1.17 155 1.22 A linear regression model estimating the effect of ethnicity on undernutrition indicates that ethnicity as a block (Hausa is the reference group) is significantly correlated with undernutrition, though directionality varies by ethnicity (see Table 4 16). Table 4 17. Linear Regression model of WHZ on ethnicity Ethnicity Hausa --Tuareg .376 Ful ani .399 Beri Beri .322 Constant Model Fit F 2.694 Adjusted R square .032* N 155 Notes: *p<.05 ; **p<.10 Values shown in each cell are unstandardized coefficients Other key indicators of undernutrition, including food security (p=.00), water source (p=.01), and vaccine coverage (p=.00), also vary significantly across ethnic groups.
130 Discussion This research describes the nutritional situation in Niger just following the 2005 and 2010 food crises. Undernutrition rates are extreme ly high, well above emergency thresholds in most areas, despite widespread receipt of humanitarian food aid. This is particularly alarming given the knowledge that undernutrition rates will often drop by as much as 10% after the harvest the period during which these data were collected, implying even higher rates at other points during the year (Loutan & Lamotte, 1984) There is variability across communities in both years, though nutritional status varied significantly only in 2010. This research aimed to describe the relationship of foo d security and food aid on undernutrition in these communities. Based on this analysis, neither food aid nor food security is a significant predictor of undernutrition in children in this sample, and the effect of food aid on household food security is not significant. None of the estimated models generated an r square of greater than .04, indicating that less than 4% of the nutritional variance observed in the sample is explained by the models. Because health and food security were included in the model, t his finding indicates a need to better understand the context of care and feeding practices of children in these communities, as this may be the underlying determinant of undernutrition that is most important. Another important consideration is the possibi lity that the extremely high rates of undernutrition make it difficult to tease out specific drivers of undernutrition. For example, food insecurity may be linked to undernutrition, but given the limited variation in food security among these, the most iso lated of communities in the least developed country in the world, the relationship may not be measurable.
131 Despite these negative findings, some interesting observations may be made from the analysis. First, the role of ethnicity has significant predictive power for undernutrition in the study population. This finding underscores the earlier hypothesis that the context of child feeding practices largely linked to cultural, community and household norms and embedded in any ethnic discrimination that exists may be an important area for further understanding of undernutrition in this area. Also, the role of disease and health is a statistically significant driver of undernutrition (fever and vaccine coverage were statistically significant in various models) This finding reinforces the need to improve the very poor health status and environmental conditions in which Nigeriens live if nutritional crisis during years of poor rainfall are to be averted. Wenhold had documented the important role of nutritional e ducation in strengthening the link between agriculture and nutrition (Wenhold et al. 2007) I n the Nigerien scenario, efforts to increase nutritional and health education may doubly serve to protect communities from undernutrition. Another objective of the research was to examine other predictors of undernutrition. Practice and use of exode long distance migration of individual family members, is a leading coping mechanism in communities in this region of Niger. It is preferred by communities over full family migration because it allows families to stay in place, and it has gained traction in comm unities where there is a belief that they may receive aid in their communities, not just in refugee camps in larger towns and cities. And while community members see the practice as seminal to coping during crisis because of the external financial resource s it provides, the use of exode as a coping strategy is not without repercussions. The practice is blamed for a loss of solidarity with
132 the community and increased rates of divorce. Women are particularly critical of the practice, citing the impact on fami ly and community structure when highly productive men leave the community. The uncertainty of financial remittances from family members en exode and the certainty of the loss of their productivity within the household makes the decision to send members awa y additionally difficult. The experience of Dareram in 2005 illustrates the need to tailor the targeting of humanitarian food aid appropriate to the specific crisis. Dareram, a Tuareg community that has pastoral roots but has increasingly practiced agropas toralism, relies to a greater extent than most agropastoral communities on livestock. However, in 2005, Dareram received no aid, even though the shortfall of fodder was significantly worse than the shortfall of grain. If food aid is going to be effective, the nature of the crisis and critical analysis of who is rendered most vulnerable must be part of the rapid assessment conducted for appropriate targeting of beneficiaries. Hindsight has illustrated the vulnerability of pastoral populations to the 2005 fod der deficit, and not simply because of the increased mortality and sale of livestock. Pastoral populations are also affected by poor rainfall because it means that families are split earlier in the season to allow animals to travel south toward better graz ing areas, extending the length of time that women and children are separated from milk producing livestock. One of the findings of this research is the potential of humanitarian aid to act as a catalyst for mobility or immobility, a phenomenon well docum ented by other researchers (Abebe et al. 2008; Fratkin & Roth, 2005; McPeak & Barrett, 2001) Historically, communities fled their home location for town and cities during crisis, to ensure that aid organizations f ound them. As aid becomes more and more common, it is reaching many
133 of these stationary communities in situ. However, others, including nomadic pastoralists, are taking a lesson from agropastoralists of the past and are coming into towns or choosing to par tially sedentarize in order to avail themselves of emergency services, including food aid, that arrive in towns during crises such as 2005 and 2010. This situation was observed in 2010 during pilot testing of the research instruments. In October 2010, the research team convened in Tanout town and aimed to target agropastoral households in and around Tanout as participants in the pilot phase of data collection. After arriving in Tanout, however, the team quickly became aware of a refugee camp of nomadic past accommodations. All instruments were pilot tested in this community, and the experience of the nomadic Fulani is illustrative of others in the area. They were very hard hit by the 2010 crisis losing over 9 0% of their livestock, including donkeys. The 2005 crisis had been difficult for them, and they did not receive any aid. Having heard that aid was distributed in Tanout, once the severity of the 2010 crisis was clear, families fled into Tanout and set up c amp on the outskirts of town. Indeed, by the time researchers met with the community in October 2010, they had been beneficiaries of food aid distributions. Similarly, food aid can be seen as the driver of migration in the case of exode, as previously dis cussed. In Kkeni, individuals not families migrated, with the hope that those who stayed in place would benefit from aid distribution, which is exactly what happened; food aid was distributed to households with young children. Unfortunately, the very poor are often the first forced to leave the community in search of help. Thus there is a risk that the most vulnerable populations will lose most by having to migrate
134 away from home in search of help and receiving less, because they are not in place if an d when their home community is targeted for aid. Vulnerability to livelihood and food insecurity in Niger is closely tied to the natural resource base. Although Niger and much of the Sahel have experienced repeated drought and famine since the 1960s, a dr amatic reversal of vegetation cover has occurred in the Maradi and southern Zinder regions; in the past 20 years, over 200 million trees have been planted and are supporting the livelihood of some 4.5 million people (Reij, 2006; World Resour ces Institute, 2008) This success has been attributed to institutional changes in governance, livelihood and the biophysical environment, and researchers argue that people there are consequently less vulnerable to economic and environmental shocks (S endzimir et al. 2011) This re greening of the Sahel does not overlap with the study area of this research, but is located to its immediate south. Nationally, Ni ger lost an average of 37,050 hectares ( 1.90% per year ) of its forest cover between 1990 and 2000, for a total loss of 38.1% (741,000 hectares) (Food and Agriculture Organization, 2010) Land c over and land use change were intended to be in included in this research, but the author was not able to secure the appropriate satellite imagery in a timely manner. This information may hold important implications on findings of this research and overall understanding of vulnerability in the area, thus is an area in which the author intends to follow up. Qualitative data indicate that the availability of wild foods and wild animals for consumption during times of crisis has decreased significantly in the past 20 30 years, eliminating historically important coping mechanisms among agropastoral and pastoral populations of the area. Information on land cover and land use change in Niger may also help explain some of the
135 observations of this research, includin g the significant loss of certain animals in specific years, such as donkeys during the 2010 crisis. Food insecurity is going to be a long term issue for Niger. Climate change and continued high population growth are likely to worsen an already difficult s ituation. Current political instability in Mali and in Libya is sending hundreds of thousands of additional mouths into rural Niger, increasing the demand for food in areas that consistently operate at a food deficit. This year, the inflow of 250,000 Niger iens who have returned from Libya and the 65,000 Malians who have fled violence in their own country will certainly add to livelihood insecurity, as households struggle to meet their nutritional requirements. The need for improved food aid and livelihood d evelopment efforts in this area of Niger is more important than ever. Researchers have documented the long term impact of short term, early childhood nutritional shocks (Alderman, 2010; Martorell, 1999) Niger is experiencing its third shock in seven years; without improved response and distri years, the nutritional situation of children in Niger is likely to continue a downward spiral.
136 CHAPTER 5 MOVING FORWARD: IMPL ICATIONS OF RESEARCH FINDINGS The overall aim of this research was to investigate if and how climate change is differentially affecting the vulnerability of communities with different livelihoods in Niger. This study has sought to qualitatively and quantitatively describe the relationship between climate change, liveliho od vulnerability/resilience, and adaptation, as experienced by communities along an agricultural pastoral continuum Although each paper is a distinct, stand alone unit, whose specific findings have previously been discussed, collectively t his research 1) augments existing literature by introducing and applying a livelihood index to the assessment of vulnerability of pastoral populations, 2) contributes to the growing body of literature describing the interactive effects of climate change, globalization, an d political instability on pastoral vulnerability, and 3) underscores the need for improved interdisciplinary collaboration to understand pastoral vulnerability and resilience. This final chapter will briefly review the findings of each distinct paper befo re discussing these major implications of the collective research. Paper Review To review, in the first paper, the Double Exposure framework is used to understand how multiple global processes climate change and economic globalization are interacting t o affect the vulnerability of Tuareg pastoralists in Niger. Examination of the uranium markets, land use change, markets, and household demographics is used to illustrate how feedback and content double exposure to these processes are affecting pastoral Tu areg. The analysis of feedback and content double exposure pathways indicate that conflict which is both a consequence and driver of these global changes is important determinant of pastoral vulnerability and food security.
137 The second paper focuses on understanding how people are experiencing climate change. Perceived risk of climate change is measure and compared across a livelihood index in order to understand if and how the differential experience of climate change might be driving livelihood adapta tions and consequent health indicators. Findings indicate no significant difference in the experience of climate change by livelihood when using the livelihood index. Discussion regarding the revision and future development of the livelihood index is discu ssed in more detail below. Findings from this research also show that perceived risk of climate change is significantly correlated with certain coping strategies, and that strategies vary by livelihood. As well, this research shows a significant correlatio n between certain coping strategies and vulnerability, which varied by livelihood. Last, the third paper investigates undernutrition by examining the experience of sample communities during the 2005 and 2010 crises. Findings indicate that neither food aid nor food security is a significant predictor of undernutrition in children under five years of age, and that the effect of food aid on household food security also is not significant. The research illustrates alarmingly high rates of undernutrition well above the internationally recognized emergency thresholds thus findings must be interpreted within this context. The research indicates that ethnicity is significantly correlated with undernutrition and that the use of exode is an important coping mecha nism in modern day Niger. Livelihood Index Development Having reviewed the findings of each paper, I now turn to those broader implications of the research. Households with a greater reliance on livestock have fundamentally different needs than communiti es with greater reliance on agriculture. For
138 example, programs that aim to improve livelihood security may focus on crop productivity for agricultural communities and livestock health for pastoral communities. However, important livelihood differences exis t among pastoral households as well; some differences may include mobility, economic activity, education, environment, or access to health. The use of a livelihood index affords the researcher the following options: to analyze the effect of livelihood as a continuous variable (1 7, agricultural to pastoral); to analyze the effect of livelihood as a dichotomous variable (pastoral and agricultural subgroups); to investigate the effect of being more or less agricultural or pastoral within just one subgroup (e. g., to assess the impact of decreased mobility among pastoral populations); and/or to analyze, through repeat measures, the effect of changing household livelihoods over time. As well, the livelihood index may be applied at various scales: the same househo ld at different time periods, different households within a community, or communities across a wider area. Use of the livelihood index perception and response to stres sors. In each instance, the index fosters improved, nuanced understanding of the role of livelihood over traditional research that dichotomizes the livelihood variable. As shown in the second paper (Chapter 3), livelihood (defined by the livelihood index) is not significantly correlated with perceived risk of climate change. However, this is likely explained by the very different experiences of the Tuareg and Fulani pastoral populations during the 2010 food crisis, and is useful in illustrating both constra ints and potential of the livelihood index. Despite no significant difference in wealth (p=.229, N=61) or TLU/RA (p=.249, N=61), Tuareg and Fulani populations represented the
139 ethnic groups with the lowest and highest rates, respectively, of undernutrition (p=.048, N=155); self reported vulnerability (p=.00, N=166); food insecurity (p=.002, N=159); and PRCC (p=.00, N=166). Because these two ethnic groups made up the pastoral population of the study, their disparate experiences during the 2010 food crisis mas ked any pattern that might otherwise have existed by livelihood; in other words, differences in ethnicity trumped the potential impact of livelihood, in this scenario. This will not always be the case. Communities commented during pilot testing for this re search that the experience of Fulani and Tuareg communities was not normally so different; communities were acutely aware of the abnormal contrast in the experience of Fulani and Tuareg communities in 2005. Thus, use of the livelihood index will clearly be more meaningful when assessing dependent variables that vary between agricultural and food crisis, this might also be the case with perceived risk of climate change. In such a context, use of the index would allow for identification of important thresholds or points of intervention that would be missed with use of a dichotomized livelihood variable. Factors Effecting Pastoral Vulnerability The vulnerability of pastoral p opulations is seemingly in contradiction to their historic adaptability and documented resilience. This research has underscored that a synergistic set of factors is to blame, including rapid population growth, global climate change (reduction of rainfall and increase in temperature), economic globalization, and political instability. Collectively, these factors have altered recent pastoral experiences and are threatening to undermine the livelihood system of pastoralists in Niger (see Figure 5 1). Their pa st resilience may not prove sufficient given the onslaught of external shocks and stresses. The past year, 2011, was another year of poor rainfall
140 and subsequent production shortage in Niger, indicating a reduction in the periodicity of climate related sho cks. In addition, recent conflicts in Libya and Mali have sent hundreds of thousands of people fleeing into Niger, increasing the food demand on individual households and eliminating remittances from family members en exode for many households. Uprising of the Tuareg in Niger may soon follow: Tuareg who are demanding an independent state are driving the conflict in Mali, and many of those returning to Niger from Libya are returning to pastoral homelands. Al Qaeda in the Maghreb continues to operate througho ut the Sahel, threatening both existing and future development efforts. And as demonstrated in the examination of the global uranium market and the Tuareg experience, globalization is affecting Niger, despite its appearance as existing at the fringe of eco nomic globalization. Thus, it appears likely that agropastoral and pastoral populations in Niger will face mounting obstacles to livelihood security in the months and years ahead. Figure 5 1. Outbreaks of Locust, drought, political uprising, and famine a gainst the backdrop of changing rainfall and population growth (Source: Sendzimir, 2011).
141 The Need for Interdisciplinary Collaboration Interdisciplinary collaboration is necessary to ensure full understanding of the socio ecological system within which p astoralists operate. There has recently been substantial discussion of adaptive capacity, resilience, vulnerability, and the various contributions and language use of different disciplines (Engle, 2011; Gallopin, 200 6; Smit & Wandel, 2006) Historically, measures of vulnerability have focused on social components and measures of resilience have focused on ecological components. I developed the research presented here within the constraints of data collected in 2005 a nd created a measure of vulnerability based on a metric established to measure community resilience to climate change (Elasha et al. 2005) Reframing this research in the context of adaptive capacity within the socio ecological system, per Engle, might more effectively combine the insights from both resilience and vulnerability frameworks. Collaboration between various disciplines investigating pastoral vulnerability is not only necessary to ensure effective cross disciplinary learning, but also to ensure the integration of so cial science research that investigates household vulnerability with ecological and geospatial modeling that predicts vulnerability at a broader scale. Land cover and land use change data information that is important to understanding the vulnerability o f pastoral populations are missing from this research and would contribute to an understanding of household level vulnerability of pastoralists. Likewise, much of the research that uses such modeling techniques for assessing pastoral vulnerability to cli mate change does not sufficiently include social science indicators of vulnerability, such as food insecurity, malnutrition, or gender inequity. If future research is going to inform effective development policy to improve livelihood and food security of p astoral populations, then efforts must be made to ensure research collaboration
142 between social, ecological, and geographic scientists to maximize understanding of vulnerability within pastoral populations.
143 APPENDIX A KEY INFORMANT INTERV IEW INSTRUMENT V illage /tribus Enqute des Informants Cls Nom du chef du village Responsabilit de participant au village Homme ou Femme
144 Composition de la communaut Merci beaucoup d'avoir accept de parler avec nous aujourd'hui S'il vous plat dcrivez nous votre communaut ; combien tes vous et quelle est votre histoire ici ? Moyens de subsistance Je vois que votre communaut est (agricole, pastoral, etc.) ; d ites moi un peu de votre vie. Avez vous cultiv? Invites possibles: Avez vous toujours cultiv? Quelles sont les cultures? Quelle est la part de la communaut cultive les cultures? Parlez moi de la migration dans cette communaut. Possible invites: Jusqu 'o dplacez vous ? Qui voyage lors de la migration /transhumance se produit? Qu'est ce qui peut catalys er la migration pour votre communaut? Histoire A ce qui concerne la productions (agricole/pastoral), a t il toujours t ainsi pour votre c ommunaut? Avez vous toujours survivre sur la base de votre (btail, des terres, une combinaison des deux)? La crise alimentaire Votre communaut a t confronte aux crise s alimentaire s passes comme en 1973, 1984 et 2005. Comment jugez vous la crise alimentaire a ctuelle par rapport ceux d'autres? Comment s'est passe pour votre communaut et comment a t elle fait face au cours de chacun de ces vnements ? Quelle a t la raction de la communaut toujours le mme? Comment et pourquoi a t elle chang? Quelle a t la crise alimentaire provoque par des vnements semblables? Ce qui a chang pour chaque vnement? Les changements climatiques Quels sont les changements dans l'environnement qui se sont pass dans votre vie? Invites possibles: la productivit de s produits (rcolte), la qualit des sols, le rythme des pluies, la quantit des prcipitations Est ce que c es changements o nt jou un rle dans toutes les crise s alimentaire que vous avez connu? Est ce que les rponses de votre communaut des moment s difficiles ont changs ? Pourquoi et de quelle manire?
145 APPENDIX B FOCUS GROUP INSTRUME NT
146 Group Focal Nom du Village /tribu Commune Zone Agro Group: Hommes Femmes (choisit un) Quelle est la mode principale de production dans votre village ? Avant et maintenant ?
147 Vous venez de survivre une crise alimentaire nationale. de la crise de ? Comment avez vous rpondu ? Quelles taient les stratgies de survivre, des adaptations, ou des a fait ici ? Est ? Expliquez. Quelles taient les consquences pour les femmes contre les hommes ?
148 Comme nt compare la crise actuelle aux crises historiques de 1973, 1984, et 2005 ? Est ce que la rponse communautaire tait pareille ou diffrent aux crises historiques ? Dcrivez vous plait.
149 Comment est ce a chang pendant l es annes passes ? ( Qualit de sol productivit du sol, pturage, rythme, quantit, et dure de la pluie, etc.)
150 Parmi les changements de vous avez observ, lesquels ce dont concerne la communaut le plus ? Comment et pourquoi ?
151 Est ce que la capacit de votre famille a rpondre aux crises, c omme celle de cette anne, a chang depuis 1973 ? Vous tes plus ou moins capable de faire face de telles crises ? C omment ?
152 Aimeriez vous rajouter quelque chose particulier not re discussion ?
153 APPENDIX C HOUSEHOLD QUESTIONNA IRE INSTRUMENT Questionnaire Number : |__|__|__|__|__|__| Household Interviews research team, under supervision of Mamadou Baro, received IRB approval and utilize it October 2010 to replicate the survey. A G ENERAL I NFORMATION A.1 2005 Head of Household A.2 Longitude : |__|__|, |__|__| __|__| Latitude : |__|__|, |__|__|__|__| A.3. 2005 household located? 1 = Yes 2 = No /____/ A.4 Name of current head of HH A.5 Name of interviewer A.6 Date of interview : /___/___/ /___/___/ 2010 A.7 Village/Tribe : ______________________ _____ A.8 Commune: ___________________________ A.9 Canton/ Subgroup ___________________________ A.10 Department ___________________________ A.11 Language of interview 1 = Hausa 2 = Djerma 3 = Tamashek 4 = Fulani 5 = Other ___________________ /____/ A.12 Ethnic group 1 = Hausa 2 = Djerma 3 = Touareg 4 = Fulani /____/
154 5 = Beri Beri 6 = Other ____________________ B. W ATER AND H EALTH B.0 What is your principal source of water? (Choose the best response ) 1 = Open well 2 = Protected well 3 = Water c hanneled from the well 4 = Running water 5 = Lake, seasonal lake, or river 6 = Other (specify ): /________________________________/ /____/ B.1 How many months of the year is this water source available? ? 1 = 0 2 months 2 = 3 8 months 3 = 9 11 months 4 = E ntire year /____/ B.2. What is the distance to the water source that you regularly use? 1 = Source is in the homestead 2 = Less than one kilometer 3 = 1 3 km 4 = 3 5 km 5 = More than 5 km /____/ B.3 What is the distance to the water sources that you regu larly use for your 1 = Source is in the homestead 2 = Less than one kilometer 3 = 1 3 km 4 = 3 5 km 5 = More than 5 km /____/ B.4 How do you collect water? 1 = Manually collected 2 = With the use of animals 3 = With the use of motorized v ehicles 4 = Other (specify) ____________ /____/ B.5 How do you transport water (from the principal source)? Choose all that apply 1 = Human (manual) transport 2 = Use of animals 3 = Transport by motorized vehicles 4 = Other (specify): /_______________ ______________/ /____/ /____/ /____/ B.6 Does your household have access to a latrine ? 1 = Yes 2 = No End of Section /____/ B.7 If C.4.0 = 1 ( yes): How far is the latrine from your house? 1 = Within the homestead (household concession) 2 = Less tha n 100 meters 3 = Between 100 200 meters 4 = Plus de 200 meters /____/ B.8 Has there been a vaccination campaign in this community during the past 12 months? 1 = Yes 2 = No /____/ B.9 If yes,: how many of your children were immunized and against what illnesses? Fill in the number of children who are vaccinated /____/ B.10 If yes: Against what diseases? Choose all that apply 1 = Polio 2 = Meningitis 3 = Measles 4 = Tetanus 5 = Other ( specify ) ___________ /____/ /____/ /____/ /____/ /____/
155 C H OUSEHOLD I NCOME C.1 What is the principal source of income for the household ? 1 = Sale of animals 2 = Sale of animal products 3 = Salaried h erder (shepherd) 4 = Caravan 5 = Seasonal work outside the village 6 = Established salaried work 7 = Small/petty commerce 8 = Temporary hire (for a salary) 9 = Cultivation and sale of crops 10= Others (specify) _______________________ /____/ C.2 What is the secondary source of income for the household ? 1 = Sale of animals 2 = Sale of animal products 3 = Salaried herder (shepherd) 4 = Caravan 5 = Seasonal work outside the village 6 = Established salaried work 7 = Small/petty commerce 8 = Tem porary hire (for a salary) 9 = Cultivation and sale of crops 10= Others (specify) _______________________ /____/ B.10.1 If the children were vaccinated against measles, respond to B.10.1, B.10.2 B.10.3. If No skip to B.11 : When were the children immunized? When were the children immunized ( the most recent ) ? ________________________________ B.10.2 Where did the chil dren receive their vaccines? ( Location of vaccination campaign) _________________________ B.10.3 Who administered the vaccination campaign ? Name of the organization that administered the vaccination campaign ________________________ B.11 How many househo ld children 0 5 years old have died in the past 12 months? Number of deceased children /____/ B.12 By what cause ? Cause 1 ____________________________ Cause 2 ____________________________ Cause 3 ____________________________ B.13 Has the woman of t he house (any) accessed prenatal services during the past 12 months? 1 = Yes 2 = No 3 = She (none) has had the need /____/ B.14 How many pregnancies has the senior wife in this household had ( including a current pregnancy) If the response is = 0 section C |____| pregnancies B.15 How many living children does she (senior wife) have ? |____| children
156 D A GRICULTURAL G OODS AND A CTIVITIES D.1 What is the statue of your property ? 1 = Owner 2 = Renter/tenant 3 = Sharecropper 4 = Owner of rented land 5 = Other ______________________ 6666 = Do not cultivate /____/ D.2 How much land do you own ? 1 = 0 0.5 hectare 2 = 0.5 1 hectare 3 = 1 1.5 hectares 4 = 1.5 2 hectares 5 = More than 2 hectares /____/ D.3 What are the principal crops that you cultivate? Choose all that apply 1 = Millet 2 = Sorghum 3 = Corn 4 = Beans 5 = Legumes 6 = Watermelon 7 = Other fruits 8 = Peanuts 9 = Others _______________ 6666 = Do not cultivate /____/ /____/ /____/ /____/ /____/ D.4 Input of Grain Products a. P revious stock b. 2009 Harvest Other inputs c. Tithe d. Gifts e. Purchase f. Reimbursement 4.1. Mil let 4.2 Sorghum 4.3 Corn 4.4 Beans 4.5 Other D.5 Output of Grain Products a. Consumed b. Reimburs ement Other outputs f Current Stock g. Seeds kept c. Tithe d. Gifts e. Sale 5.1 Mil let 5.2 Sorghum 5.3 Corn 5.4 Beans 5.5 Other UNITS OF MEASUER MUST BE IN TIA !! (SPACE FOR CALCULATIONS) ____________________________ ________________________________________________________ ___
157 D.6 Your current stock (including 2010 harvest) will last how many months? Fill in the number of months /____/ D.7 Have you already harvested this fall ? 1 = Yes 2 = No 6666 = Do not harvest /_ ___/ D.8 What portion of your household income comes from agricultural production ? Use small rocks, peas, etc. to demonstrate for participants. 1 = None 2 = Very small portion, less than 5% 3 = Half 4 = Most 5 = All /____/ D.9 Has your family always p lanted crops for an income? 1 = Yes 2 = No /____/ D.10 Is your family more, less, or equally reliant on agriculture as in the past? 1 = More 2 = Less 3 = About the same /____/ E. L IVESTOCK What is livestock count ? Code Type of animal a : How many did you have six months ago? b : How many have you destocked in the past 6 months? c : Reason sold d : Current stock Sold Consumed Lost/Died ( Number ) ( Number ) ( Number ) ( Number ) 1 = Buy good 2 = Health 3 = Travel 4 = Education 5 = Ceremony 6 = Animal food 7 = Clothes 8 = Taxes 9 = Other ( Number ) E.1 Chicken |__|__|__| |__|__|__| |__|__|__| |__|__|__| |__|__|__| |__|__|__| E.2 Goat |__|__|__| |__|__|__| |__|__|__| |__|__|__| |__|__|__| |__|__|__| E.3 Sheep |__|__|__| |__|__|__| |__|__|__| |__|__|_ _| |__|__|__| |__|__|__| E.4 Cow |__|__|__| |__|__|__| |__|__|__| |__|__|__| |__|__|__| |__|__|__| E.5 Donkey |__|__|__| |__|__|__| |__|__|__| |__|__|__| |__|__|__| |__|__|__| E.6 Camel |__|__|__| |__|__|__| |__|__|__| |__|__|__| |__|__|__| |__|__| __| E.7 Horse |__|__|__| |__|__|__| |__|__|__| |__|__|__| |__|__|__| |__|__|__| E.8 Were pastures this year sufficient for feeding your animals? 1 = Yes Go to E.10 2 = No /____/
158 E.9 If pastures were NOT sufficient, what did you do to feed your animal s? 1 = Migration of animals 2 = Purchase of livestock feed 3 = Liquidation of some livestock for income 4 = Nothing 5 = Other Choose all that apply /____/ /____/ /____/ E.10 What portion of household income comes from livestock production ? Use smal l rocks, peas, etc. to demonstrate for participants. 1 = None 2 = Very small portion, less than 5% 3 = Half 4 = Most 5 = All /____/ E.11 Has your family always practiced livestock herding for an income? 1 = Yes 2 = No /____/ E.12 Is your family more, le ss, or equally reliant on livestock herding as in the past? 1 = More 2 = Less 3 = About the same /____/ E.13 On a scale of 1 7, where 1 represents agricultural communities and 7 represents transhumant pastoral populations, where would you place your famil y? Use small stones, peas, etc. to create a scale in the sand and demonstrate positioning. Complete with a number 1 7 /____/ F I AM GONG TO READ A LI ST OF STRATEGIES A FTER EACH ONE PLEASE INDICATE IF THIS IS A STRATEGY THAT IS U SED BY YOUR HOUS EHOLD AND WITH WHAT FREQUENCY IT IS USED Strategy Have you used this strategy in the past year? Do you use this strategy every year, in exception years, or never? Cod e Strategy 1 = Yes 2 = No 1 = Every year 2 = Exceptional years 3 = Never F. 1 Eat/sel l own harvested grain |___| |___| F. 2 Change in eating habits |___| |___| F. 3 Reduce number of meals |___| |___| F. 4 Reduce amount of food at meals |___| |___| F. 5 Reduce food quality at meals |___| |___| F. 6 Day without eating |___| |___| F. 7 Send family members to other households |___| |___| F. 8 Reduce food expenditures |___| |___|
159 F. 9 Accept gifts of family and friends |___| |___| F.1 0 Eat/sell milk or meat of own animals |___| |___| F.1 1 Buy/trade food goods |___| |___| F.1 2 Use of communi ty food stocks ( cereal banks ) |___| |___| F.1 3 Sale of personal goods |___| |___| F.1 4 Collection of wild foods for consumption |___| |___| F.1 5 Familial credit or loan |___| |___| F.1 6 Commercial credit or loan |___| |___| F.1 7 Migration |___| |___ | F.1 8 Exceptional migration/transhumance |___| |___| F.1 9 Reduction in non food expenditures |___| |___| F.2 0 Barter |___| |___| F.2 1 Pull children out of school |___| |___| F.2 2 Offer children as labor |___| |___| F.2 3 Sedentarize |___| |___| F.2 4 Reduce mobility |___| |___| F.2 5 Others ( specify ) |___| |___| G. C REDIT AND L OANS G.1 Does your community have loan or credit structures ? 1 = Yes 2 = No |____| G.2 If yes, please describe them (formal, informal) 1 = Formal 2 = Informal 3 = Both |__ __| G.3 Over the past six months, have you benefitted from credit or loan in order to purchase food ? 1 = No, not one time (Go to H.1) 2 = once 3 = twice 4 = three times 5 = more than three times Credit Loan |____| |____| G.4 If yes wha t type of credit/loan did you access? Choose all that apply 1 = Parents/family/friends 2 = NGO or other organization 3 = Local loan agent 4 = Commercial loan agent 5 = Bank, credit office 6 = Community bank 7 = Cooperative 8 = Other ( Specify ) a. |___| b. |___| c. |___| G.5 What is the total amount of all loans/credit you have received in the past 3 Total |___|___|___|___|___|___| F FCFA
160 months ? ( loans/credit = money ) G.6 What were the terms of the loan/credit? Interest rate (%) |____| H. I NTERV ENTIONS /A ID H.1 Has your household received food aid of any sort from any organization in the past six months? 1 = Yes 2 = No |____| H.2 What aid did you receive ? Name the types of intervention/aid ____________________________ ____________________________ ____________________________ J. A LIMENTATION I. F OOD In the past 24 hours, please indicate when different members of your household have eaten? Household member Before breakfast Breakfast B etween breakfast and lunch Lunch Between lunch and dinner Dinner I.1 Household children 1 = Yes 2 = No I.2 Household women I.3 Household men In the past 24 hours, which members of your household have eaten the following: Food pr oduct Household member Children of the household Women of the household Men of the household 1 = Yes 2 = No I.5 Mil let |___| |___| |___| I.6 Sorghum |___| |___| |___| I.7 Rice |___| |___| |___| I.8 Other grains |___| |___| |___| I.9 Bread crackers, pasta |___| |___| |___| I.10 Fist |___| |___| |___| I.11 Poultry |___| |___| |___| I.12 Meat |___| |___| |___|
161 I.13 Eggs |___| |___| |___| I.14 Beans |___| |___| |___| I.15 Legumes |___| |___| |___| I.16 Greens |___| |___| |___| I.17 O il or fats |___| |___| |___| I.18 Fruit |___| |___| |___| I.19 Sugar |___| |___| |___| I.20 Salt |___| |___| |___| I.21 Milk or dairy product (yogurt) |___| |___| |___| J. V ULNERABILITY /R ESILIENCE How would you compare the situation in 2005 to the c urrent situation for each of the following Possible change J. 1 Land degradation 1 = Increasing 2 = Stable 3 = Decreasing |___| J. 2 Condition of vegetation cover 1 = Worsening 2 = Stable 3 = Improving | ___| J. 3 Soil and crop productivity 1 = Decreasing 2 = Stable 3 = Increasing |___| J. 4 Health and sanitation 1 = Decreasing 2 = Stable 3 = Increasing |___| J. 5 Wealth 1 = Decreasing 2 = Stable 3 = Increasing |___| J. 6 Access to food 1 = Decreasing 2 = Stable 3 = Increasing |___| J. 7 Transhumance 1 = Increasing 2 = Stable 3 = Decreasing |___| J. 8 Mobility/Migration 1 = Decreasing 2 = Stable 3 = Increasing |___| K. C LIMATE C HANGE I am going to list a few aspects of climate change that may be occ urring here. Please tell me what you see as the potential for these aspects to affect your household is (little, somewhat, or a lot), as well as how you perceive the current situation (worsening, getting better, etc.). Possible change Potential to do h arm Current situation K.1 Land degradation 1 = Little 2 = Somewhat |___| 3 = A lot Dj fait en haut
162 K.2 Condition of vegetation cover 1 = Little 2 = Somewhat |___| 3 = A lot Dj fait en haut K.3 Soil and crop productivity 1 = Little 2 = Somewhat |___| 3 = A lot Dj fait en haut K.4 Rainfall 1 = Little 2 = Somewhat |___| 3 = A lot 1 = Decreasing 2 = Stable |___| 3 = Increasing K.5 Desertification 1 = Little 2 = Somewhat |___| 3 = A lot 1 = Decr easing 2 = Stable |___| 3 = Increasing K.6 Heat 1 = Little 2 = Somewhat |___| 3 = A lot 1 = Decreasing 2 = Stable |___| 3 = Increasing K.7 Loss of indigenous species 1 = Little 2 = Somewhat |___| 3 = A lot 1 = Decreasing 2 = Stable |___| 3 = Increasing L. O BSERVATIONS L.1 Based on the information you have received and your own visual observations, how would you categorize the vulnerability of this household? Very vulnerable So mewhat vulnerable Not vulnerable L.2. Please explain the situation of this household with regards to mobility, with particular attention to the years since 2005. Any other observations : (please include household dwe lling type, visibility of livestock, farming equipment, etc. any indicators of livelihood should be captured here).
164 L. M EMBERS OF HOUSHOLD First and Family names Sex Age Marital status Literate Education Present Seasonal Migration Principal Occ upation Secondary Occupation Please begin with head of household 1 = M 2 =F ( in years or partial years ) 1 = married ; 2 = widowed ; 3 = single ; 4 = young (<13 years) 1 = Yes 2 = No 0= none 1= primary incomplete ; 2= primary complete ; 3= secondary ; 4= K oranic school only 1 = Yes 2 = No 1 = Yes 2 = No 0= none ; 1= agriculture ; 2= herding/livestock care ; 3= commercial ; 4= artisanal ; 5= civil servant ; 6= housekeeping ; 7= small trade (carpentry, blacksmith, etc. ) ; 8= other 1. (head of household) |____| |___|___| |____| |____| |____| |____| |____| |____| |____| 2. |____| |___|___| |____| |____| |____| |____| |____| |____| |____| 3. |____| |___|___| |____| |____| |____| |____| |____| |____| |____| 4. |____| |___|___| |____| |____| |____| |____| | ____| |____| |____| 5. |____| |___|___| |____| |____| |____| |____| |____| |____| |____| 6. |____| |___|___| |____| |____| |____| |____| |____| |____| |____| 7. |____| |___|___| |____| |____| |____| |____| |____| |____| |____| 8. |____| |___|___| |____ | |____| |____| |____| |____| |____| |____| 9. |____| |___|___| |____| |____| |____| |____| |____| |____| |____| 10. |____| |___|___| |____| |____| |____| |____| |____| |____| |____| 11. |____| |___|___| |____| |____| |____| |____| |____| |____| |____| 12 |____| |___|___| |____| |____| |____| |____| |____| |____| |____| 13 |____| |___|___| |____| |____| |____| |____| |____| |____| |____| 14 |____| |___|___| |____| |____| |____| |____| |____| |____| |____|
165 15 |____| |___|___| |____| |____| |____| |___ _| |____| |____| |____| 16 |____| |___|___| |____| |____| |____| |____| |____| |____| |____| 17 |____| |___|___| |____| |____| |____| |____| |____| |____| |____| |____| |___|___| |____| |____| |____| |____| |____| |____| |____| |____| |___|___| |____ | |____| |____| |____| |____| |____| |____| |____| |___|___| |____| |____| |____| |____| |____| |____| |____| |____| |___|___| |____| |____| |____| |____| |____| |____| |____| |____| |___|___| |____| |____| |____| |____| |____| |____| |____| |____| |___|___| |____| |____| |____| |____| |____| |____| |____| |____| |___|___| |____| |____| |____| |____| |____| |____| |____| |____| |___|___| |____| |____| |____| |____| |____| |____| |____| |____| |___|___| |____| |____| |____| |____| |____| |____| |____| |____| |___|___| |____| |____| |____| |____| |____| |____| |____| |____| |___|___| |____| |____| |____| |____| |____| |____| |____| |____| |___|___| |____| |____| |____| |____| |____| |____| |____| |____| |___|___| |____| |____| |____| |___ _| |____| |____| |____|
166 APPENDIX D ANTHROPOMETRIC DATA COLLECTION INSTRUMEN T
167 Anthropometric Questionnaire A General Information 1.1 Name of interviewer: 1.2 Date: 1.3 Village/Tribe :
168 Name of Head of Household Name of Child Child Age ( months) Sex 1 = M 2 = F Anthropometrics Vaccine card present 1 = yes 2 = no Participating in a nutritional feeding program? 1 = yes 2 = no Currently : Weight (kg) Height (cm) MUAC (cm) Diarrhea 1 = yes 2 = no Fever 1 = yes 2 = no Nursing 1 = yes 2 = no 1 2 3 4 5 6 7 8 9 10 11 12 13
169 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
170 29 30 31 32 33 34 35 36 37 38 39 40
171 APPENDIX E METHODOLOGY FOR CREA TION OF LIVELIHOOD I NDEX This research sought to develop a livelihood in dex rather than using a dichotomous variable for livelihood (agricultural versus pastoral) The final seven item index was created using several items, collected during household interviews Initially, information gathered and analyzed in the construction of the variable included the following sub items (parentheses indicate reference survey number): Principal revenue source (C1) Property ownership (D1) Principal crop production (D3) Timing of harvest (D7) Revenue portion from crops (D8) Family history of c rop production (D9) Comparison of present and past crop production (D10) Revenue portion from livestock (E10) Family history of pastoralism (E11) Comparison of present and past pastoralism (E12) Self identification of livelihood (E13) Migration (F17B) Geog raphic livelihood zone (Focus Groups) Qualitative history of migration (Focus Groups) Th r ough factor analysis and tests for multi colinearity, these sub items were then organized into 9 items to create a preliminary nine item index They were arranged int o the following items: Principal revenue source (C1) Lack of land, crops, and harvest (D1, D3, and D7) Revenue portion from crops (D8) Agricultural heritage (D8, D9, and D10) Revenue portion from livestock (E10) Pastoral heritage (E10, E11, and E12) Self i dentification of livelihood (E13) Livelihood zone (FG) History of migration (FG)
172 Detailed descriptions of how each item was coded and generated referencing household survey question numbers, is included here: Item1: Principal revenue source: If C1=1 (s ale of animals) or 2 (sale of animal products) then Item1 = 7 If C1=3 then Item1= 6 If C1= 9 then Item1 = 1 Else Item1 = 4 Item2: Lack of land, crops, and harvest If SubItem2.1 or SubItem2.2 or SubItem2.3 = 6666 then Item2 = 7 Else Item2 = 1 SubItem2.1 : Property ownership If D1 = 6666 then SubItem2.1 = 7 If D1 = 1 or 2 then SubItem2.1 = 1 Else SubItem2.1 = 4 SubItem2.2: Principal crop production If D3 = 6666 then SubItem2.2 = 7 Else SubItem2.2 = 1 SubItem2.3: Harvest If D7 = 6666 then SubItem2.3 = 7 E lse SubItem2.3 = 1 Item3: Revenue portion from crops If D8 = 1 then Item3 = 5 If D8 = 2 then Item3 = 4 If D8 = 3 then Item3 = 3 If D8 = 4 then Item3 = 2 If D8 = 5 then Item3 = 1 Item4: Agricultural heritage Item4 = SubItem4.1 SubItem4.2 SubItem4.3 SubItem4.1: Revenue portion from crops See above (Item3) SubItem4.2: Family history of crop production If D9 = 1 then SubItem4.2 = 1 If D9 = 2 then SubItem4.2 = 2
173 SubItem4.3: Comparison of present and past crop production If D10 = 1 then SubItem4.3 = 0.5 If D10 = 2 then SubItem4.3 = 2 If D10 = 3 then SubItem4.3 = 1 Item5: Revenue portion from livestock If E10 = 5 then Item5 = 5 If E10 = 4 then Item5 = 4 If E10 = 3 then Item5 = 3 If E10 = 2 then Item5 = 2 If E10 = 1 then Item5 = 1 Item6: Pastoral he ritage Item6 = SubItem6.1 SubItem6.2 SubItem6.3 SubItem6.1: Revenue portion from livestock See above (Item5) SubItem6.2: Family history of pastoralism If E11 = 1 then SubItem6.2 = 1 If E11 = 2 then SubItem6.2 = 2 SubItem6.3: Comparison of presen t and past pastoralism If E12 = 1 then SubItem6.3 = 0.5 If E12 = 2 then SubItem6.3 = 2 If E12 = 3 then SubItem6.3 = 1 Item7: Self identification of livelihood Item7 = E13 Item8: Livelihood zone Pastoral zone = 7 Agropastoral zone = 4 Agricultural zone = 1 Borderline pastoral zone = 6 Item9: Qualitative history of migration (Focus Groups) Communities were place along the agricultural/pastoral continuum based on focus group discussions about livelihood, historical livelihood, transhumance and migration p atterns (1 7) After running tests for reliability and assessing inter item correlation, the final seven item index was constructed using the following items: Principal revenue source (C1) Lack of land, crops, and harvest (D1, D3, and D7) Revenue portion f rom crops (D8) Agricultural heritage (D8, D9, and D10)
174 Pastoral heritage (E10, E11, and E12) Self identification of livelihood (E13) Migration (F17B)
175 LIST OF REFERENCES Abdalla, M. A. (2009). Understanding of the Natural Resource Confl ict Dynamics: The Case of the Tuareg in North Africa and the Sahel, Institute for Security Studies (Vol. Paper 194). Pretoria, South Africa: ISS. Abebe, D., Cullis, A., Catley, A., Aklilu, Y., Mekonnen, G., & Ghebrechirstos, Y. (2008). Impact of a commerci al destocking relief intervention in Moyale district, southern Ethiopia. Disasters, 32 (2), 167 189. Action Against Hunger. (2006). Humanitarian Context in Niger. Retrieved 11/10/11, from www.actionagains thunger.org Adger, W. N. (2000). Social and ecological resilience: are they related? Progress in Human Geography, 24 (3), 347 364. Adger, W. N. (2003). Social Capital, Collective Action, and Adaptation to Climate Change. Economic Geography, 79 (4), 387 404. Adger, W. N. (2006). Vulnerability. Global Environmental Change, 16 (3), 268 281. Adger, W. N., Eakin, H., & Winkels, A. (2009). Nested and teleconnected vulnerabilities to environmental change. Frontiers in Ecology and the Environment, 7 (3), 150 157. Alde rman, H. (2010). Safety Nets Can Help Address the Risks to Nutrition from Increasing Climate Variability. The Journal of Nutrition, 140 (1), 148S 152S. Baker, J., & Ngendakuriyo, A. (2011). 5 Year Evaluation of the Cetnral Emergency Response Fund, Case Stud y: Niger Ohain, Belgium: Channel Research and OCHA. Barb, L. L., & Lebel, T. (1997). Rainfall climatology of the HAPEX Sahel region during the years 1950 1990. Journal of Hydrology, 188 189 43 73. Baro, M., & Deubel, T. (2006). Persistent Hunger: Perspe ctives on Vulnerability, Famine, and Food Security in Sub Saharan Africa. Annual Review of Anthropology, 35 (1), 521 538. Bassett, T. J., & Zuli, K. B. (2000). Environmental discourses and the Ivorian savanna. Annals of the Association of American Geograph ers, 90 (1), 67 95. Battisti, D. S., & Naylor, R. L. (2009). Historical Warnings of Future Food Insecurity with Unprecedented Seasonal Heat. Science, 323 (5911), 240 244. Bednik, A. (2008). Niger's mine war Radioactive land is real price of uranium. Le Mon de diplomatique (France)
176 Beekhuis, G. (2007). Cross border trade and food markets in Niger: why market analysis is important for humanitarian action. Humanitarian Exchange Magazine Behnke, R. H. (1994). Natural Resource Management in Pastoral Africa. Dev elopment Policy Review, 12 5 27. Behnke, R. H., Kerven, C., & Scoones, I. (Eds.). (1993). Range ecology at disequilibrium : new models of natural variability and pastoral adaptation in African savannas London: Overseas Development Institute. Ben Mohamed, A., van Duivenbooden, N., & Abdoussallam, S. (2002). Impact of Climate Change on Agricultural Production in the Sahel Part 1. Methodological Approach and Case Study for Millet in Niger. Climatic Change, 54 (3), 327 348. Bernard, H. R. (1995). Research Met hods in Anthropology: Qualitative and Quantitative Approaches (Second Edition ed.). Walnut Creek: AltaMira Press. Birch, I., & Grahn, R. (2007). Pastoralism Managing Multiple Stressors and the Threat of Climate Variability and Change, Human Development O ccasional Papers New York: Human Development Report Office, UNDP. Blackwell, N., Augier, A., & Sayadi, S. (2010). Food crisis in Niger: a chronic emergency. The Lancet, 376 (9739), 416 417. Blench, R. (2001). 'You Can't Go Home Again': Pastoralism in the N ew Millennium London: Overseas Development Institute. Bohle, H. G., Downing, T. E., & Watts, M. J. (1994). Climate change and social vulnerability: Toward a sociology and geography of food insecurity. Global Environmental Change, 4 (1), 37 48. Bradley, P. (1971). The Ecological Survey of Sub Desert and Sahel Zone. The Geographical Journal, 137 (4), 462 466. Brown, K., Dickin, K., Bentley, M., Oni, G., Obassanjju, V., Esrey, S., et al. (1988). Complementary Feeding of Young Children in Developming Countries: a Review of Current Scientific Evidence WHO/NUT/98.1. Geneva: WHO. Bryman, A. (2008). Structured interviewing In Social Research Methods : Oxford University Press. Bureau for Applied Research in Anthropology. (2005). Cash Distribution Initial Assessment U A, Cash Distribution Project in Tanout, Niger Tucson: University of Arizona.
177 Bureau for Applied Research in Anthropology. (2006a). Final Monitoring Report of the First Direct Cash Transfer Project in the Sahel: An Innovative Response by the British Red Cr oss to the 2005 Food Crisis in Tanout University of Arizona and the British Red Cross. Bureau for Applied Research in Anthropology. (2006b). Review of Available Secondary Data: Niger 2005 2006. University Of Arizona and the British Red Cross. Campbell, D. J. (1984). Response to drought among farmers and herders in southern Kajiado District, Kenya. Human Ecology, 12 (1), 35 64. CARE USA, & CARE Denmark. (2007). The Human Right to Water Beyond Access. Atlanta: CARE USA. CARE USA and CARE Denmark. (2007). The Human Right to Water Beyond Access: CARE USA and CARE Denmark. CDC. (2008). EpiInfo (Version 3.5.1). Atlanta: CDC. Chambers, R. (1992). Rural appraisal: rapid, relaxed and participatory Brighton: Institute of Development Studies at the University of Su ssex, Discussion paper 311. Chambers, R. (2007). Who Counts? The Quiet Revolution of Participation and Numbers Brighton: IDS Working Paper 296. Chambers, R., & Conway, G. (1992). Sustainable rural livelihoods: practical concepts for the twenty first centu ry In IDS Discussion Paper 296 Brighton: Institute of Development Studeis. Clarke, J. I. (1959). Studies of Semi Nomadism in North Africa. Economic Geography, 35 (2), 95 108. Cobey, J. C., & Cunningham, N. (1968). Anthropometry in action: an evaluation of a local calandar used in determining ages of children in a Nigerian village. The Journal of Tropical Pediatrics, 14 132 138. Cornia, G. A., & Deotti, L. (2008). Niger's 2005 Food Crisis: Extent, Causes and Nutritional Impact : European Development Research NetworkWP 2008 15. Daouda, M., Ozer, P., and Erpicum, M. (1996a). Consquences de la scheresse sur Proceeding from the International Conference on Tropical Climatology, Meteorology and Hydrolo gy, Brussels, 22 24 May 1996, Brussels, Belgium Royal Meteorological Institute of Belgium and Royal Academy of Overseas Sciences.
178 Daouda, M., Ozer, P., and Erpicum, M. (1996b). Consquences de la scheresse sur pluies au Niger. Proceeding from the International Conference on Tropical Climatology, Meteorology and Hydrology, Brussels, 22 24 May, Brussels, Belgium Royal Meteorological Institute of Belgium and Royal Academy of Overseas Sciences. Davies, J., & Bennet t, R. (2007). Livelihood adaptation to risk: Constraints and opportunities for pastoral development in Ethiopia's Afar region. Journal of Development Studies, 43 (3), 490 511. Davies, J., & Hatfield, R. (2007). The economics of mobile pastoralism: A globa l summary. Nomadic Peoples, 11 (1), 91 116. Davies, S. (1996). Adaptable Livelihoods: Coping with Food Insecurity in the Malian Sahel London: MacMillan. Dayak, M. (1992). Touareg, la tragdie Paris: Jean Claude Latts. Decalo, S. (1997). Historical Dictio nary of Niger Lanham, MD and London: Scarecrow Press. Denton, F. (2000). Gender impact of climate change: a human security dimension. Energia News, 3 (3). Denton, F. (2002). Climate change vulnerability, impacts, and adaptation: Why does gender matter? Gen der and Development, 10 (2), 10 20. Dettwyler, K. A. (1989). Styles of Infant Feeding: Parental/Caretaker Control of Food Consumption in Young Children. American Anthropologist, 91 (3), 696 703. Edmond, K. M., Zandoh, C., Quigley, M. A., Amenga Etego, S., Ow usu Agyei, S., & Kirkwood, B. R. (2006). Delayed Breastfeeding Initiation Increases Risk of Neonatal Mortality. Pediatrics, 117 (3), e380 e386. Elasha, B. O., Elhassan, N. G., Ahmed, H., & Zakieldin, S. (2005). Sustainable livelihood approach for assessing community resilience to climate change: case studies from Sudan AIACC Working Papers Number 17. Ellis, F. (1998). Household strategies and rural livelihood diversification. Journal of Development Studies, 35 (1), 1 38. Ellis, J. E., & Swift, D. M. (1988). Stability of African pastoral ecosystems: alternate paradigms and implications for development. Journal of Range Management, 41 450 459. Emerson, S. A. (2011). Desert insurgency: lessons from the third Tuareg rebellion. Small Wars & Insurgencies, 22 (4), 669 687.
179 Engle, N. L. (2011). Adaptive capacity and its assessment. Global Environmental Change, 21 (2), 647 656. Engle, P. L. (2002). Infant Feeding Styles: Barriers and Opportunities for Good Nutrition in India. Nutrition Reviews, 60 S109 S114. Ensminger J. (1992). Making a Market: The Institutional Transformation fo an African Society New York: Cambridge University Press. FEWS NET. (2010). Markets and Food Security in the Eastern Basin (Benin, Niger, Nigeria, Chad) Washington, D.C.: USAID FEWSNET, CIL SS, FAO, and WFP. FEWS NET. (2012). Make Your Own Maps Niger: An interactive Microsoft PowerPoint Presentation for Building Maps. Retrieved February 10, 2012 Food and Agriculture Organization. (1996). Declaration on world food security. Proceeding from t he World Food Summit, Rome Food and Agriculture Organization. Food and Agriculture Organization. (2010). Global Forrest Resource Assessment Report Country Report Niger Rome: Food and Agriculture Organization of the United Nations. France 24. (2010, Marc h 29). GreenPeace slams Areva over radioactive contamination. Retrieved 12/11/11, from http://www.france24.com/en/20100329 greenpeace slams areva over radioactive contamination Fratkin, E. (1992). Drought and Development in Marsabit District, Kenya. Disasters, 16 119 130. Fratkin, E. (1997). Pastoralism: Governance and Development Issues. Annual Review of Anthropology, 26 235 261. Fratkin, E. (2004). Ariaal Pastoralists of Kenya: Studying pastoralism, drought, and development in Africa's arid lands. (Second Edition ed.). Boston: Pearson. Fratkin, E., & Roth, E. A. (Eds.). (2005). As pastoralists settle: social, health, and economic consequences of the pastoral sedentarization in Marsabit District, Kenya New York: Kluwer Academin Publishers. Friedel, M. H. (1991). Range condition assessment and the concept of thresholds: A viewpoint. Journal of Range Management, 44 (5), 422 426. Fugelstad, F. (1983). A h istory of Niger, 1850 1960 Cambridge and New York: Cambridge University Press. Gallopin, G. C. (2006). Linkages between vulnerability, resilience, and adaptive capacity. Global Environmental Change, 16 (3), 293 303.
180 Galvin, K. A. (2009). Transitions: Pasto ralists Living with Change. Annual Review of Anthropology, 38 (1), 185 198. Galvin, K. A., Reid, R. S., Behnke, R. H., & Hobbs, N. T. (Eds.). (2008). Fragmentation in semi arid and arid landscapes: Consequences for human and natural systems Dordrecht, the Netherlands: Springer. Geesing, D. a. D., H. (2008). Food and Agriculture Organization. Country Pasture/Forage Resource Profile. Retrieved December 10, 2008, from www.fao.org Georgieva, K. (2011). International respons e to Niger's hunger crisis. The Lancet, 377 (9759), 28. Government of Niger. (2005). Rapport Gnral de la 1er Runion Annuelle du Systme de Prvention et de Gestion des Crises Alimentaires Rpublique du Niger: CNPGCA, SAP et Cabinet Premier Ministre. Gov ernment of Niger, World Food Programme, & FEWSNET. (2004). Joint Harvest Assessment : Government of Niger, WFP, and USAID FEWSNET. Grell, H., & Kirk, M. (1999). The Role of the Donors in Influencing Property Rights over Pastoral Resources in Sub Saharan Afr ica In N. McCarthy, B. Swallow, M. Kirk & P. Hazel (Eds.), Property Rights, Risk, and Livestock Development in Africa Washington, D.C.: International Food Policy Research Institute. Grobler Tanner, C. (2006). Understanding nutrition data and the causes of malnutrition in Niger Washington, D.C.: USAID FEWSNET. Hammel, R. (2001). Securing land for herders in Niger. Niamey, Niger: AREN, Drylands Programme. Hardin, G. (1968). The tragedy of the commons. Science, 162 1243 1248. Held, I. M., Delworth, T. L., L u, J., Findell, K. L., & Knutson, T. R. (2005). Simulation of Sahel Drought in the 20th and 21st Centuries. Proceedings of the National Academy of Sciences of the United States of America, 102 (50), 17891 17896. Hesse, C., & Cotula, L. (2006). Climate Chang e and Pastoralists: Investing in people to respond to adversity. In I. I. f. E. a. Development (Ed.), Sustainable Development Opinion London: International Institute for Environment and Development. Hesse, C., & Thebaud, B. (2006). Will Pastoral Legislati on Disempower Pastoralists in the Sahel? Indigenous Affairs, 1/06 14 23. Hogg, R. (1986). The New Pastoralism: Poverty and Dependency in Northern Kenya. Africa, 56 (3), 319 333.
181 Hogg, R. (1997). Pastoralists, ethnicity and the state in Ethiopia London: Ha an in association with the Institute for African Alternatives. Holling, C. S. (1995). What barriers? What bridges? In L. Gunderson, C. S. Holling & S. S. Light (Eds.), Barriers and bridges to the renewal of ecosystems and institutions New York: Columbia U niversity Press. Huntingford, C., Lambert, F. H., Gash, J. H. C., Taylor, C. M., & Challinor, A. J. (2005). Aspects of climate change prediction relevant to crop productivity. Philisophical Translations of the Royal Society 360 (1463), 1999 2009. IndexMundi (2012). Uranium Monthly Price US Dollars Per Pound. Retrieved 1/17/12, from http://www.indexmundi.com/commodities/?commodity=uranium&months=60 IPCC. (2000). IPCC Spec ial Report on The Regional Impacts of Climate Change: An Assessment of Vulnerability. Geneva: UNEP/WMO. IPCC. (2007). Climate Change 2007: Impacts, Adaptation, and Vulnerability. In M. L. Parry, O. F. Canziani, J. P. Palutikof, P. J. van der Linden & C. E. Hanson (Eds.), Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (pp. 1000). Cambridge: Cambridge University Press. IRIN. (2008). Population explosion threatens development gains. UN Office f or the Coordination of Humanitarian Affairs Jahnke, H. E. (1982). Livestock Production Systems in Livestock Development in Tropical Africa Kiel, FRG: Kieler Wissenschaftsverlag Vauk. Keenan, J. (2000a). The father's friend: Returning to the Tuareg as an 'elder'. Anthropology Today, 16 (4), 7 11. Keenan, J. (2000b). [Untitled]. The Journal of the Royal Anthropological Institute, 6 (2), 369 370. Keenan, J. (2008). Uranium goes critical in Niger: Tuareg rebellions threaten Sahelian conflagration. Review of Afr ican Political Economy, 35 (117), 449 466. Keenan, J. (2009). Al Qaeda terrorism in the Sahara? Edwin Dyer's murder and the role of intelligence agencies. Anthropology Today, 25 (4), 14 18. Kelley, T. A. (2002). Squeezing Parakeets into Pigeon Holes: The Eff ects of Globalization and State Legal Reform in Niger on Indigenous Zarma Law. NYU Journal of International Law & Politics, 34 (3), 635 711. Kelly, P. M., & Adger, W. N. (2000). Theory and Practice in Assessing Vulnerability to Climate Change and Facilitat ing Adaptation. Climatic Change, 47 (4), 325 352.
182 Kratli, S. (2001). Educating nomadic herders out of poverty? Culture, education and pastoral livelihood in Turkana and Karamoja Sussex for the World Bank: Institute of Development Studies. Lamprey, H. (1983 ). Pastoralism yesterday and today: The overgrazing problem In F. Bourliere (Ed.), Ecosystems of the World (Vol. 3). Amsterdam: Elsevier Scientific. Lamprey, H. F. (1976). The UNEP MAB Integrated Project in Arid Lands: Phase III, Regional Project Document FP/1101 77 Nairobi: UNEP. Lavigne Delville, P. (1999). Harmonising formal law and customary land rights in French speaking West Africa London: IIED: Issue Paper 86. Le Houerou, H. N. (1996). Climate change, drought and desertification. Journal of Arid En vironments, 34 (2), 133 185. Lebel, T., Sauvageot, H., Hoepffner, M., Desbois, M., Guillot, B., & Hubert, P. (1992). Rainfall estimation in the Sahel: the EPSAT NIGER experiment. Hydrological Sciences 37 (3), 201 217. Leichenko, R., & O'Brien, K. (2002). Th e dynamics of rural vulnerability to global change. Mitigation and Adaptation Strategies for Global Change, 7 (1), 1 18. Leichenko, R., & O'Brien, K. (2008). Environmental Change and Globalization New York: Oxford University Press. Leichenko, R. M., O'Brie n, K. L., & Solecki, W. D. (2010). Climate change and the global financial crisis: A case of double exposure. Annals of the Association of American Geographers, 100 (4), 963 972. Levinson, D. (1998). Ethinc Groups Worldwide A Ready Reference Handbook Pho enix: Oryx Press. Little, P. D. (1985). Social differentiation and pastoral sedentarization in northern Kenya. Africa, 55 (3), 243 261. Little, P. D., Smith, K., Cellarius, B. A., Coppock, D. L., & Barrett, C. (2001). Avoiding Disaster: Diversification and Risk Management among East African Herders. Development and Change, 32 (3), 401 433. London, L. (2001). Pesticides and Women Workers in South Africa: A Question of Social Justice. Women and Environments, Fall 2001 Loutan, L., & Lamotte, J. M. (1984). Seaso nal variations in nutrition among a group of nomadic pastoralists in Niger. The Lancet, 323 (8383), 945 947.
183 Lu, C., Schneider, M. T., Gubbins, P., Leach Kemon, K., Jamison, D., & Murray, C. J. L. (2010). Public financing of health in developing countries: a cross national systematic analysis. The Lancet, 375 (9723), 1375 1387. Macro International Inc. (2007). Enqute Dmographique et de Sant et Indicateurs Multiples au Niger 2006. Calverton, Maryland, USA: Institut National de la Statistique, Ministre de l'conomie et des Finances, and Macro International Inc. Martorell, R. (1999). The nature of child malnutrition and its long term implications. Food and Nutrition Bulletin, 20 (3), 288 292. Marty, A. (1996). La division sedentaires nomades. Le cas de bouc le du Niger au debut de la periode coloniale: IRAM. Maxwell, D., Ahiadeke, C., Levin, C., Armar Klemesu, M., Zakariah, S., & Lamptey, G. M. (1999). Alternative food security indicators: revisiting the frequency and severity of `coping strategies'. Food Pol icy, 24 (4), 411 429. Maxwell, D. G. (1996). Measuring food insecurity: The frequency and severity of ''coping strategies''. Food Policy, 21 (3), 291 303. McCabe, J. T. (2007). Cattle Bring Us to Our Enemies: Turkana Ecology, Politics, and Raiding in a Diseq uilibrium System Ann Arbor: University of Michigan Press. McConnel, T. (2007, October 30). Niger: the uranium curse. Open Democracy McKune, S., & Silva, J. (2012). Pastoralists Under Pressure: Double Exposure to Economic and Environmental Change in Niger Unpublished Manuscript. Unpublished manuscript. McPeak, J. G., & Barrett, C. B. (2001). Differential Risk Exposure and Stochastic Poverty Traps among East African Pastoralists. American Journal of Agricultural Economics, 83 (3), 674 679. Meyer, C. (2010, April 2). Uranium Mining in Niger: Tuareg Activist Takes on French Nuclear Company. Spiegel Online International Mira, C. (2011). Oil, Uranium Projects Brighten Medium term Prospects in Niger, IMF Survey Magazine, Countries & Regions Washington, D.C.: Af rica Department of the International Monetary Fund. Mohamed, A. B., Duivenbooden, N. V., & Abdoussallam, S. (2002). Impact of Climate Change on Agricultural Production in the Sahel Part 1. Methodological Approach and Case Study for Millet in Niger. Clima tic Change, 54 327 348. Mortimore, M. J. (1998). Roots in the African dust Cambridge: Cambridge University Press.
184 Mortimore, M. J., & Adams, W. M. (2001). Farmer adaptation, change and 'crisis' in the Sahel. Global Environmental Change, 11 49 57. Morton J. F. (2007). The impact of climate change on smallholder and subsistence agriculture. Proceedings of the National Academy of Sciences, 104 (50), 19680 19685. Moseley, W. G., Carney, J., & Becker, L. (2010). Neoliberal policy, rural livelihoods, and urban food security in West Africa: A comparative study of The Gambia d'Ivoire, and Mali. Proceedings of the National Academy of Sciences, 107 (13), 5774 5779. Mousseau, F., & Mittal, A. (2006). Sahel: A Prisoner of Starvation? A Case Study of the 2005 Food Crisis in Niger Oakland, CA: The Oakland Institute. Mullany, L. C., Katz, J., Li, Y. M., Khatry, S. K., LeClerq, S. C., Darmstadt, G. L., et al. (2008). Breast Feeding Patterns, Time to Initiation, and Mortality Risk among Newborns in Southern Nepal. The Journal of Nutrition, 138 (3), 599 603. Nathan, M. A., Roth, E A., Fratkin, E., Wiseman, D., & Harris, J. (2005). Health and Morbidity among Rendille Pastoralist Children: Effects of Sedentarization In E. Fratkin & E. A. Roth (Eds.), As Pastoralists Settle New York: Kluwer Academic Publishers. National Institute of Statistics, & Ministry of Public Health. (2010). National Nutrition Survey Niamey, Niger: Institute Nationale des Statistiques. Neef, A. (2000). Ethnic Groups in Niger In P. L. F. Graef, and M. von Oppen (Ed.), Adapted Farming in West Africa: Issues, Pot entials, and Perspectives Stuttgart, Germany: Verlag Ulrich E. Grauer. Nelson, V., Meadows, K., Cannon, T., Morton, J., & Martin, A. (2002). Uncertain predictions, invisible impacts, and the need to mainstream gender in climate change adaptations. Gender and Development, 10 (2). Niamir Fuller, M. (2000). The resilience of pastoral herding in Sahelian Africa In F. Berkes & C. Folke (Eds.), Linking Social and Ecological Systems: Management Practices and Social Mechanisms for Building Resilience Cambridge: Ca mbridge University Press. Niger Uranium Limited. (2008). Uranium in Niger. Retrieved 12/16/2009, from http://www.niger uranium.com/ Nyong, A., Adesina, F., & Osman Elasha, B. (2007). The value of indigenous knowledge in climate change mitigation and adaptation strategies in the African Sahel. Mitigation and Adaptation Strategies for Global Change, 12 (5), 787 797.
185 O'Brien, K. L., & Leichenko, R. M. (2000). Double exposure: assessing the impacts of climate chan ge within the context of economic globalization. Global Environmental Change, 10 (3), 221 232. (2004). Mapping vulnerability to multiple stressors: climate change and glob alization in India. Global Environmental Change, 14 303 313. Pedersen, J., & Benjaminsen, T. A. (2008). One Leg or Two? Food Security and Pastoralism in the Northern Sahel. Human Ecology, 36 43 57. Pinstrup Andersen, P. (2009). Food security: definition and measurement. Food Security, 1 (1), 5 7. Glimpses from KwaZulu Natal. Global Environmental Change, 16 (2), 195 206. Reij, C. P. (2006). More success stories in Afri ca's drylands than often assumed. Proceeding from the Forum sur la Souverainet Alimentaire, Niamey 7 10 Novembre, 2006. Rseau des Organisations paysannes et de Producteurs Agricoles de l'Afrique de l'Ouest, Ouagadougou, Burkina Faso. Rpublique du Niger. (2007). Plan stratgique pour la protection, la promotion et le 2012) : Ministre de Sant Publique Niger. Roncoli, C. (2006). Ethnographic and participatory approaches to research on farmers' responses to climate predictions. Climate Research, 33 81 99. Roncoli, C., Ingram, K., & Kirshen, P. (2002). Reading the Rains: Local Knowledge and Rainfall Forecasting in Burkina Faso. Society and Natural Resources, 15 409 427. Roth, E. A. (1996). Traditional pastor al strategies in a modern world: An example from Northern Kenya. Human Organization, 55 (2), 219. Rutstein, S. O. (2005). Effects of preceding birth interval on neonatal, infant and under five years mortality and nutritional status in developing countries: evidence from the demographic and health surveys. International Journal of Gynecology and Obstetrics 89 S7 S24. Sato, S. (1997). How the East African Pastoral Nomads, Especially the Rendille, Respond to the Encroaching Market Economy. African Studies Mono graphs, 18 (3 4), 121 135. Schwartz, H. J. (1995). Ecological and Economic Consequences of Reduced Mobility in Pastoral Livestock Production Systems In E. Fratkin & E. A. Roth (Eds.), As Pastoralists Settle New York: Kluwer Academic Publishers.
186 Scoones, I. (1998). Sustainable rural livelihoods: a framework for analysis. IDS Working Paper, No. 72 Sussex, UK: Institute of Development Studies. Scoones, I. (1999). New Ecology and the Social Sciences: What Prospects for a Fruitful Engagement? Annual Review of Anthropology, 28 479 507. Sen, A. (1981). Poverty and Famines: An Essay on Entitlements and Deprivation Oxford: Clarendon Press. Sendzimir, J., Reij, C. P., & Magnuszewski, P. (2011). Rebuilding Resilience in the Sahel: Regreening in the Maradi and Zinde r Regions of Niger. Ecology and Society, 16 (3). Shankar, A. V., Gittelsohn, J., West, K. ., Stallings, R., Gnywali, T., & Faruque, F. (1998). Eating from a Shared Plate Affects Food Consumption in Vitamin A Deficient Nepali Children. The Journal of Nutrit ion, 128 (7), 1127 1133. Shen, C., & Williamson, J. B. (1999). Maternal mortality, women's status, and economic dependency in less developed countries: a cross national analysis. Social Science & Medicine, 49 (2), 197 214. Silva, J. A., Eriksen, S., & Om be, Z. A. (2010). Double exposure in Mozambique's Limpopo River Basin. Geographical Journal, 176 (1), 6 24. Simanowtiz, S. (2009). Blue men and yellowcake: the struggle of the Tuaregs in Mali and Niger. Contemporary Review (291), 70 74. Sivakumar, M. V. K. ( 1992). Climate change and implications for agriculture in Niger. Climatic Change, 20 (4), 297 312. Slovic, P., Fischhoff, B., & Lichtenstein, S. (1982). Why Study Risk Perception. Risk Analysis, 2 83 94. Smit, B., & Wandel, J. (2006). Adaptation, adaptive capacity and vulnerability. Global Environmental Change, 16 (3), 282 292. Smith, K. (1998). Sedentarization and market integration: New opportunities for Rendille and Ariaal women of northern Kenya. Human Organization, 57 (7), 459 468. Smith, K., Barrett, C. B., & Box, P. W. (2000). Participatory Risk Mapping for Targeting Research and Assistance: With an Example from East African Pastoralists. World Development, 28 (11), 1945 1959. Smith, L., Ramakrishnan, U., Ndiaye, A., Haddad, L., & Martorell, R. (2003). T he Importance of Women's Status for Child Nutrition in Developing Countries Research Report 131. Washington, D.C.: International Food Policy Research Institute.
187 Smith, L. C., & Haddad, L. (2001). How important is improving food availability for reducing ch ild malnutrition in developing countries? Agricultural Economics, 26 (3), 191 204. SPSS. (2011). IBM SPSS Statistics, Version 20. Staff. (2008). French Niger Uranium Mines under Direct Threat from Tuareg Nomads Space Daily from http://www.energy daily.com/reports/French_Niger_Uranium_Mines_Under_Direct_Threat_From_Tua reg_Nomads_999.html Swindale, A., & Bilinsky, P. (2006). Develop ment of a Universally Applicable Household Food Insecurity Measurement Tool: Process, Current Status, and Outstanding Issues. The Journal of Nutrition, 136 (5), 1449S 1452S. Talle, A. (1999). Pastoralists at the border: Maasai poverty and the development di scourse in Tanzania In D. M. Anderson & V. Broch Due (Eds.), The Poor are Not Us: Poverty and Pastoralism in Eastern Africa (pp. 106 124). Oxford: James Currey. Tectonidis, M. (2006). Crisis in Niger -Outpatient Care for Severe Acute Malnutrition. N Engl J Med, 354 (3), 224 227. Thebaud, B., & Batterbury, S. (2001). Sahel pastoralists: opportunism, struggle, conflict, and negotiation: A cast study from eastern Niger. Global Environmental Change, 11 69 78. Tsai, T. C. (2010). Food crisis no longer taboo in Niger. The Lancet, 375 (9721), 1151 1152. Tschakert, P. (2007). Views from the vulnerable: Understanding climatic and other stressors in the Sahel. Global Environmental Change, 17 (3), 381 396. Tukei, P. M. (1963). A calendar for the assessment of the ages of young Baganda children. Journal of Tropical Medicine and Hygiene, 66 42 44. Turner, B. L., Kasperson, R. E., Matson, P. A., McCarthy, J. J., Corell, R. W., Christensen, L., et al. (2003). A framework for vulnerability analysis in sustainability science Proceedings of the National Academy of Sciences, 100 (14), 8074 8079. U.S. Department of State. (2008). Background Note: Niger. Retrieved December 11, 2008 UN Population Division. (2008). Niger. Retrieved December 11, 2008, from www.globalis.gvu.unu.edu
188 UNDP. (2007). Human Development Report 2007/2008. New York, New York: United Nations Development Programme. UNDP. (2011). Sustainability and Equity: A Better Future for All Human Development Report 2011. New York, New York: United Nations Development Programme. UNICEF. (1990). Conceptual Framework for the Causes of Malnutrition Strategy for improved Nutrition of Children and Women in developing countries. New York: UNICEF. UNICEF. (2009). June 2009 Child S urvival and Nutrition Survey : UNICEF, WFP, and HKI. UNICEF. (2011). Niger Humanitarian Action for Children West and Central Africa. Geneva: UNICEF. United Nations Conference on Desertification (UNCOD). (1977). Proceedings from the United Nations Conferen ce on Desertification: Its Causes and Consequences. Proceeding, Nairobi UNEP. United Nations Development Program (UNDP). (2011). Human Development Report 2011 New York: UNDP. United Nations Development Programme (UNDP). (2007). Human Development Report 20 07/2008. New York, New York: United Nations Development Programme. University of Saskatchewan. People of Niger. Retrieved October 26, 2008, from www.usask.ca/nursing/international /niger/people.htm US Department of State. (2012). Background Note: Niger. Retrieved 2/24/12, from http://www.state.gov/r/pa/ei/bgn/5474.htm US Department of State's Bureau of African Affairs. (2 008). Background Note: Niger. Retrieved October, 2008 USAID. (2005). Niger Drought and Locusts SItuation Report #1, Fiscal Year 2005. Washington, D.C.: USAID. van Duivenbooden, N., Abdoussalam, S., & Ben Mohamed, A. (2002). Impact of Climate Change on Agricultural Production in the Sahel Part 2. Case Study for Groundnut and Cowpea in Niger. Climatic Change, 54 (3), 349 368. Walker, P. (1989). Famine Early Warning Systems: Victims and Destitution London: Earthscan. Warren, A. (1995). Changing Understan dings of African Pastoralism and the Nature of Environmental Paradigms. Transactions of the Institute of British Geographers, New Series, 20 (2), 193 203.
189 Webb, P., Coates, J., Frongillo, E. A., Rogers, B. L., Swindale, A., & Bilinsky, P. (2006). Measuring Household Food Insecurity: Why It's So Important and Yet So Difficult to Do. The Journal of Nutrition, 136 (5), 1404S 1408S. Wenhold, F., Faber, M., Van Averbeke, W., Oelofse, A., Van Jaarsveld, P., Van Rensburg WS, et al. (2007). Linking smallholder agricu lture and water to household food security and nutrition. Water South Africa, 33 (3), 327 336. Westoby, M., Walker, B., & Noy Meir, I. (1989). Opportunistic Management for Rangelands Not at Equilibrium. Journal of Range Management, 42 (4), 266 274. White, C. (1991). Increased vulnerability to food shortages among Fulani nomads in Niger In R. E. Downs, D. O. Kerner & S. P. Reyna (Eds.), The Political economy of African famine (Vol. 9). Philadelphia: Gordon and Breach Science Publishers. Wilding, J., Mossi, I. I., Edwards, D. L., Weisbaum, A., Aw, B., & Mander, T. (2005). Joint Independent Evaluation of the Humanitarian Response of CARE, CRS, Save the Children and World Vision to the 2005 Food Crisis in the Republic of Niger : CARE, CRS, Save the Children, and Wo rld Vision. World Health Organization. (2006). Country Health System Fact Sheet 2006: Niger Geneva: World Health Organization. World Resources Institute. (2008). Word Resources 2009 roots of resilience: growing the wealth of the poor ownership, capaci ty, connection. World Resources Institute. Washington, D.C.: World Resrouces Institute, in collaboration with United Nations Development Programme and the World Bank. Wuehler, S. E., & Biga Hassoumi, A. (2011). Situational analysis of infant and young chil d nutrition policies and programmatic activities in Niger. Maternal & Child Nutrition, 7 133 156.
190 BIOGRAPHICAL SKETCH Sarah McKune received a Bachelor of Arts degree in French and sociology from Wofford College in 1999 and a Master of Public Health from Rollins School of Public Health at Emory University in 2002. She spent 6 years working in domestic and global public health before pursuing her doctoral degree at the University of Florida in the School of Natural Resources and the Environment.