The consequences of women’s use of donkeys for pastoral flexibility: Maasai ethnoarchaeology Fiona Marshall1,* and Lior Weissbrod2 1,2 Department of Anthropology, Washington University, St Louis MO 2 Zinman Institute of Archaeology, University of Haifa, Israel *Corresponding author. Department of Anthropology, Washington University, St Louis, MO 63130, USA. fmarshal@wustl.edu Abst r a c t / Zu s a m me n fa s s u ng Ethnoarchaeological research on donkey-use among Maasai pastoralists in Kajiado District, Kenya provides new insights into the intersection between donkey biology, behavior, and husbandry. Significant findings include de- tailed information on the gendered context of tasks for which donkeys are used, the amount of effort expended in donkey management, patterns of herd growth, and the influence of donkeys on pastoral household labor and rangeland management. Women’s use of donkeys to transport water and other crucial resources to people and reliance on donkeys for transport during long-distance residential movements reduces energy expended in domes- tic work, contribute to flexible mobility, and allows herders to reconcile often conflicting needs for water and pasture. Donkeys have relatively few costs and significantly enhance options for flexible mobility and survival in the study area. These results contribute to understanding donkey domestication processes, increasingly mobile African pastoral societies of the mid-Holocene, and strategies for sustaining ancient pastoral herds on long-dis- tance migrations and colonization events resulting from Saharan expansion. We hypothesize that reliance on donkeys for transport underlies the successful mobile responses of food producers to climate change in northeast- ern Africa and the spread of food production on the continent. Ethnoarchäologische Forschung zur Nutzung von Eseln durch die viehhaltenden Massai im Kajiado District, Kenia, liefert neue Einsichten in die einander überlappenden Bereiche der Biologie der Esel, deren Verhalten, und deren Haltung als Nutztiere. Signifikante Befunde schließen detaillierte Informationen betreffend des geschlechts- spezifischen Kontext ein, innerhalb dessen Esel genutzt werden, sowie Informationen über den Aufwand für das Eselmanagement, Art und Weise des Herdenwachstums, den Einfluss von Eseln auf die Arbeit in ländlichen Haushalten und bezüglich des Weidemanagements. Die Nutzung von Eseln durch Frauen zum Zwecke des Trans- portes von Wasser und anderen, für die Bevölkerung wesentlichen Ressourcen, sowie die Nutzung von Eseln für Transporte im Verlauf von Ortswechseln über große Distanzen reduziert den Energieaufwand für die häusliche Arbeit, trägt zur flexiblen Mobilität bei, und erlaubt den Hirten die notwendige Abwägung zwischen den häufig einander entgegenstehenden Bedürfnissen nach Wasser und Weideland. Esel sind wenig kostenintensiv und erhö- hen in signifikanter Weise die Optionen für eine flexible Mobilität und das Überleben in der untersuchten Region. Unsere Ergebnisse tragen zu dem Verständnis des Domestikationsprozesses von Eseln bei, zum Verständnis der steigenden Mobilität von ländlichen Sozietäten Afrikas während des mittleren Holozäns, und zum Verständnis der Strategien für das Aufrechterhalten der Viehherden im Verlaufe von Migrationen über lange Strecken und der Kolonisierungsereignisse in Folge der Saharaexpansion. Wir nehmen an, dass die Abhängigkeit von Eseln zu Transportzwecken die Grundlage für die erfolgreiche Reaktion nahrungsproduzierender Bevölkerungen auf den Klimawandel im nordöstlichen Afrika im Sinne von mobiler Lebensweise und der Verbreitung der Nahrungs- produktion über den Kontinent darstellte. Keywords: Africa, donkey, domestication, pastoralism Afrika, Esel, Domestikation, Pastoralismus I nt roduct ion the expansion of ancient Egyptian and Sumerian land- Domestication of the donkey influenced the course of based trade routes (Marshall 2007; Rossel et al. 2008). development of mobile pastoral societies in Africa and Today donkeys are depended upon for transport by 60 Fiona Marshall and Lior Weissbrod people living in arid and mountainous and resource- gions and time periods, information is needed on the poor areas of the world (Starkey 2000). Relatively little ways that contemporary mobile African herders use is known, however, regarding domestication of the and manage their donkeys. There has, however, been donkey and the influence of donkey’s on household little ethnographic, livestock management, or veteri- subsistence. Mitochondrial genetic data demonstrates nary research on donkeys. Most African pastoral eth- that the African wild ass is the ancestor of the donkey nographies indicate simply that donkeys are essential and suggests domestication in Africa. The presence of pack animals used by herders in arid environments two different mitochondrial clades suggests two sepa- where systematic mobility of people and herds is im- rate recruitments from the wild (Beja-Pereira et al. portant (Coppock 1994; Jacobs 1975; Klima 1970; 2004; Vilà et al. 2006). Nicolaisen & Nicolaisen 1997; Little and Leslie 1999; see Marshall 2007). Only two ethnographies, the Nico- Pastoralism was the earliest form of food production in laisens’ (1997) study of the Tuareg and Little and Leslie Africa and the three historically known African wild (1999) and colleagues’ research among the Turkana, ass, the Nubian (E. a. africanus), Somali or Dibokali provide substantial information on donkey-use and (E. a. somaliensis) and Atlas wild ass (so called E. a. there are no published studies of the size or demogra- atlanticus) (Groves 1986; Moehlman 2002) were dis- phy of donkey herds owned by pastoral families, or of tributed in areas occupied by early northeast African the details of their use or management. pastoralists. As a result geographic data, taken together with genetics, archaeology and linguistics suggest that It is clear, nonetheless, that herders in dry regions rely African pastoralists domesticated both clades of don- heavily on the superior physiological characteristics of key during the seventh millennium BP (Blench 2000; arid-adapted donkeys, which make them more efficient Beja-Pereira et al. 2004; Marshall 2007; Vilà et al. pack animals than cattle and tolerant of high levels of 2006). At this time, it has been hypothesized, climatic desiccation (Maloiy 1970). Donkeys also have more me- variability associated with development of the hyper- chanically efficient gaits than cattle, especially in rough arid Sahara led to increased pressure for pastoral mo- terrain (Dijkman 1991; Jones 1977; Yousef 1991). Mar- bility and intentional domestication of the desert- shall (2007) emphasized the significance of African pas- adapted African wild ass by northeast African herders toralist’s use of donkeys for transport during long-dis- (Marshall 2007; Marshall & Weissbrod n.d.). If this tance residential movements, for women’s daily tasks chronology is correct, domestication of the donkey such as collection of water or firewood and in the sus- would have followed at least 2000 years of cattle hus- tainable management of rangelands, but pointed to the bandry in the eastern Sahara and up to a thousand years need for more detailed field studies. In order to develop of a more complex pastoral system incorporating Asian archaeologically testable hypotheses regarding the mo- sheep and goats. We expect that during these early pe- tivations of ancient herders for capturing and managing riods pastoral mobility was more circumscribed, resi- African wild ass, the consequences of domestication for dential settlements more tethered to permanent water herding families, and to develop methods for identifying sources, and responses to climatic perturbations less early stages of domestication, we undertook systematic flexible than after domestication of the donkey. ethnoarchaeological research on the way that Maasai pastoral families use and manage their donkeys and on Archaeological evidence from Egypt suggests, how- the social and economic influence of donkey ownership ever, that the process of domestication was slow and in Kajiado District, southern Kenya. that ancient Egyptians also played a role in “morpho- logical domestication” of the donkey. Ten donkey skel- Social and geographic setting etons with pathologies indicative of transport-use bur- ied at Abydos in the funerary complex of one of the The Maasai are Nilotic speaking pastoralists with a earliest Egyptian kings dating to approximately 4950 relatively egalitarian social and political system char- BP (historic date 3000 BC) indicate the importance of acterized by decision-making by older men and clan, donkey-based transport to the early Egyptians (Rossel lineage and age-set sub-divisions (Galaty 1981, 1982; et al. 2008). Skeletally, however, the Abydos animals Merker 1910; Spear & Waller 1993; Spencer 1988). His- resemble African wild ass. This demonstrates that the torically, this framework facilitated social and eco- process of “morphological domestication” of the wild nomic co-operation across large areas as well as access ass continued in Egyptian contexts and that neither to distant water sources and pastures in times of pastoral husbandry in the range of the wild ass in Sa- drought, disease or warfare. Like most African pasto- hara, Sudan or Eritrea, nor early Egyptian donkey ralists, Maasai herders organize daily life and the loca- management resulted in rapid selection. tion of settlements around the needs of their herds. Families try to maintain access to large and varied ter- To better understand domestication processes and the rains, and rely on mobility to make efficient use of spa- economic and social roles of donkeys in different re- tial and temporal variability in rainfall, topography and Donkeys and Maasai ethnoarchaeology 61 pastures (Jacobs 1975; Grandin 1991; Grandin et al. tially and temporally patchy distribution of forage 1991; Western & Dunne 1979). Despite considerable (Bekure & Grandin 1991; BurnSilver 2003; Grandin et adherence to established social and economic institu- al. 1991). Aerial census data from 1977-1983 provide an tions, however, significant changes have taken place indication of the relative proportions of large wild during the course of integration of Maasai communi- mammals to livestock in Kajiado District over the last ties into regional and global economies. It is readily 30 years and rare quantitative data on the relative fre- apparent that the nature of these changes varies region- quency of donkeys, with estimates of 412,000 cattle, ally, locally, and from family to family (Homewood et 518,000 sheep and goat, 16,000 donkeys, 43,000 wilde- al. 2009; Worden 2007). In this section we briefly de- beest, 22,000 zebra, 7,000 eland and 8,000 giraffe scribe general and historic patterns and detail below (Bekure & Grandin 1991). transformations relevant to this study. Maasai families rely on cattle, sheep, goats and don- Strategies of livestock management are intricately con- keys and if labor is available graze herd segments, by nected to the social organization of Maasai households age, sex, or lactation status, in the locales best suited to and to men and women’s labor roles. The primary set- each. Historically, families obtained some grains from tlements – Enkang in Maa or boma in Swahili – are neighboring agriculturalists and maize flour, sugar and occupied by up to a dozen polygamous extended fami- tea are now procured from small trading centers or lies (10-50 people) who own between 20 and 350 head traveling salesmen. Cultivation has not been a Maasai of cattle per family and share labor for herding and practice until recently in much of Kajiado District, and animal protection (Arhem 1985; Bekure & Grandin even now it is not common in the relatively arid study 1991; Grandin et al. 1991; Homewood & Rodgers 1991). area. Herders move animals to pasture daily as well as The household forms the basic social unit of herd man- seasonally, when entire households may spend weeks agement and is composed of the herd owner, his wives, at a time with livestock in areas of productive tempo- children and often his mother or married sons and rary grazing. The main wet season residence areas brothers and their nuclear families (Grandin et al. (Emparnat sing.) are typically made up of up to 10 1991). Each household is economically independent, settlements and are located close to permanent water however, and each married woman within a household sources. Smaller more temporary settlements (Enkaron is the head of her own house. She is in charge of the sing.), are located further from water. The annual cycle domestic space, constructing the dwelling and caring of migration in the study area is small-scale and follows for her children (Talle 1987). Adult women and older a regular pattern of movement between wet and dry girls also undertake all cleaning, cooking, childcare, season settlement areas. Long-distance migrations on construction and maintenance of houses, milking of a larger-scale are driven by severe droughts, which livestock and collection of water and firewood (Arhem have recurred once or twice every decade at least since 1985; Homewood & Rodgers, 1991, Talle 1987, 1988). the 1950’s. Worden (2007) notes that under normal cir- Women also care for the family’s livestock when ani- cumstances livestock in the study area is moved less mals are corralled in the enclosures. Adult men are in now than in the past, but that animals are moved ear- charge of management of herds, organization of family lier and further during droughts. labor, care of livestock and construction of thorn fenc- es. Group decisions are based on discussions and con- Recent changes in Maasai economy and mobility pat- sensus, but are primarily in the hands of older men terns in southern Kenya have important implications (Arhem 1985; Homewood & Rodgers 1991; Spencer for understanding donkey use and for its relevance 1993). Historically, warrior age sets were responsible (sensu Wylie 1985, 2002) to archaeological cases. for defense of livestock and settlements. Boys and These include sedentarization, greater investments in young men were responsible for herding. In more re- the productivity of livestock, addition of local water cent times, once boys start going to school, girls spend sources such as bore holes, pipeline taps, and mechan- more time supervising livestock. Obtaining enough la- ically dug wells, and introduction of alternative means bor for flexible herd management in variable conditions of transport such as bicycles and motorized vehicles. has often, however, been a challenge for African pasto- Over the past several decades land tenure reform in ralists (Dahl & Hjort 1976). Maasai territories has also been associated with sub- stantial increase in the use-life of settlements in con- Although some of the territories currently occupied by junction with decrease in the number of households Maasai families receive more rain than other pastoral cohabiting settlements and cooperating in herd man- regions of Africa, the study area is relatively dry re- agement (Grandin 1991; Worden 2007). At the same ceiving average annual precipitation of 475-625 mm. time, there is evidence for increased input in livestock Rainfall is bimodal with significant intra- and inter- productivity based on contraction of daily grazing or- annual variability. Similar variability in soil and topog- bits, greater use of improved breeds, salt, and veteri- raphy combines with rainfall patterns to create spa- nary medicine and more frequent watering (Bekure &  62 Fiona Marshall and Lior Weissbrod Fig. 1: The location of the study area showing the distribution of Emparnat settlements Grandin 1991). Settlement in the study area is also 2010). As a result, our work in this area benefited from more tied to modern resources such as schools, health close familiarity with neighborhoods and the environ- centers and churches (Worden 2007). Nonetheless, be- ment and people felt comfortable discussing donkeys as cause livestock management continues to be oriented they related to many aspects of their lives. We inter- mainly towards household rather than market produc- viewed married women and men living in a settlement- tion and seasonal mobility remains essential in the dry cluster known as the Olobelibel neighborhood (Fig. 1). environment, donkeys still play a significant role in Women were regarded as the owners of and acknowl- facing the challenges of maintaining local herding sys- edged experts on donkeys. Twenty six women provided tems. Specific similarities between the present and the quantitative information on their herds, a total of 65 past, that we think are relevant (sensu Wylie 1985; donkeys. These women (each with their own house) be- 2002) to modeling ancient pastoral systems, include longed to eight of the ten polygamous households living unpredictable rainfall and patchy distribution of re- in the six enkangs or major settlements that made up the sources in semi-arid grasslands and reliance on cattle, Olobelibel settlement cluster (Fig. 1; Table 1). Each local sheep and goat herds, donkeys and mobility, as a strat- neighborhood such as Olobelibel maintains a single for- egy for survival in fluctuating climatic conditions. mal enkaron settlement which they use in typical dry seasons. During more severe droughts they may move to other areas and construct more ephemeral enclosures Met ho d s (Worden 2007: 32). Information on the enkaron used by the Olobelibel residents is documented in order to illus- Information on the role of donkeys in the Maasai study trate the role of donkeys in dry-season dispersal. Our area was collected at intervals during 2006 when one of decision to conduct the study within the defined social us (LW) was living there undertaking a broader ethno- environment of a single Maasai settlement neighbor- archaeological research project focusing on the ecology hood means that, although we do provide some quantita- of Maasai settlements and seasonal mobility (Weissbrod tive data, this study is largely qualitative. Donkeys and Maasai ethnoarchaeology 63 No. households Adult population size and composition Total Settlement in study Men Womena Young adultsb Donkeys S1 3 3 4 3 10 S2 1 2 5 0 7 S3 1 4 6 3 13 S4 1 3 5 - 8 S5 1 4 11 - 15 S6 1 1 2 0 3 a Some of the women of these households did not posses donkeys of their own during the time of the study. b Young adults refers to teenagers. 0 denotes zero, - denotes no data. Table 1: The study sample: number of settlements, households, people and donkeys Topics that we addressed in depth in interviews included chose not to try to define donkey age categories during the demography of donkey holdings, social and eco- interviews, however, and employed emic perceptions nomic roles of donkeys. We also focused on donkey of the categories, juvenile, adult and old. Based on ob- management practices including control and selection servations and discussions, most women interpreted in breeding, training, and general approaches to care- juvenile to mean unweaned donkeys, that is yearlings taking. We conducted four types of interviews: 1) or younger. Animals sufficiently aged to be no longer lengthy semi-structured interviews with seven women fit for work, were considered old. in participating households 2) informal questioning of women about their donkeys during water-collection Interviews were conducted with the aid of Joseph Par- trips in which we participated 3) opportunistic ex- teri Lekanaiya (JPL), who lived in the area and acted as change with women on the subject of donkeys on vari- research assistant and interpreter to L.W. To maximize ous occasions and 4) unstructured interviews with men accuracy of simultaneous translation during the inter- on their perspectives on the significance of donkeys for views, prior to the beginning of the study the interview issues of settlement use, residential mobility and herd- guide was translated in writing by L.W. and J.P.W. into ing strategies. It is important to note that we conducted Maa and the topics were discussed in detail. During the the study in 2006 after a severe 2-year drought that interviews questions were presented by L.W. in English resulted in some families loosing a number of their and translated to Maa by J.P.W. who was fluent in both donkeys. English and Maa. (Both L.W. and J.P. also spoke Swa- hili, but found the English to Maa translation worked The quantitative data on donkey holdings was gath- better in this setting). In order to preserve confidential- ered, as mentioned, from eight of the ten households ity, in the following results section we do not to specify occupying the six settlements of Olobelibel settlement the identities of participating individuals, households cluster. Additional, more detailed data on life-histories or settlements. of donkey holdings was contributed by five households. During lengthy interviews, seven women from these Information was cross-checked during the course of households discussed past holdings and forms of social multiple interviews and return visits to individual set- exchange of donkeys. Based on their life experiences, tlements and neighborhoods. Additional information these women also provided extensive contextualized was collected and information from interviews con- information on local knowledge of donkeys. The long firmed through participant observation. Data on the interviews were each roughly three hours long and spatial pattern of donkey use during daily and seasonal loosely based on the interview guide presented below movements of households participating in the study and devised by F.M. and L.W. (Table 2). We presented was gathered with the use of a GPS (see Fig. 2). It questions as open-ended in order to encourage elabo- should be emphasized that we use the quantitative data rate answers, which often raised further questions on on donkey demography and movements mainly to pro- topics of interest such as control and selection practices vide a general indication of relations between common in donkey management and breeding. Most topics un- donkey management practices and landscape use at der discussion, such as gender, were straightforward. Olobelibel, rather than as a statistical sample of the re- Age categories, though, could be more complex. We gional population. 64 Fiona Marshall and Lior Weissbrod 1. How many donkeys do you have: male, female, juvenile, and old? 2. What are the kin relations among the donkeys? 3. Are the males castrated and if so why and by whom? 4. Were the donkeys bought, born to a female owned by you, or loaned? 5. Have you ever selected a specific male and/or female for mating? 6. How do you choose which donkey to buy? 7. Do females from your herd mate with a male from your herd that might be their offspring? 8. Can that be prevented? Have you ever received or given a donkey as a loan/gift and is there a benefit to the owner in lending any of 9. his donkeys? 10. Do you have any donkeys which are better than the others at carrying heavy loads? 11. Do you have donkeys that you know will not become sick? 12. Do you have any donkeys that you know will survive during drought and long-distance migration? 13. Are there any advantages or disadvantages to the white, black, or striped-legged donkeys? 14. How are juvenile donkeys trained and at what age does training begin? 15. What do you use your donkeys for? 16. How often do you take the donkeys to fetch water and how many donkeys do you take every time? 17. Who is in charge of the donkeys for other tasks? 18. What were donkeys used for long ago but are no longer used for nowadays? 19. How heavy are the loads that donkeys carry? 20. Is there a special enclosure for the donkeys in your settlement? 21. Do you bring the donkeys into the settlement at night? 22. Do you take the donkeys to graze or do they graze on their own? 23. What sicknesses do donkeys have? 24. Do you use modern or traditional medicine to treat sick donkeys? 25. Have you lost any of your donkeys to drought, disease, or predators? 26. What equipment do you use with the donkeys and how are loads arranged or donkeys mounted? 27. Do you know any stories or songs about donkeys? Table 2: Interview guide Fig. 2: The location of permanent Emparnat settlement clusters, the dry season Enkaron of the study settlements and donkey routes to water. Donkeys and Maasai ethnoarchaeology 65 Settlement Woman Male Donkeys Female Donkeys Total Donkeys N=6 N=26 juvenilea adult old juvenile adult old N=65 A 2 1 1 4 S1 B 1 1 2 C 2 1 1 1 5 A 1 1 2 4 B 1 1 1 3 S2 C 2 2 D 1 1 1 3 E 1 1 A 1 3 4 B 1 2 3 S3 C 2 1 3 D 1 1 2 E 1 1 2 A 1 1 B 1 1 S4 C 1 1 D 1 1 E 1 1 A 1 1 B 1 1 1 3 C 1 1 3 5 S5 D 1 1 E 1 1 2 F 1 1 G 3 4 7 S6 A 1 1 2 Table 3: Age and sex of donkeys belonging to 26 women in six settlements Re su lt s during the time of the study and were sharing the use of donkeys with co-wives or other women of the ex- We begin this section by providing quantitative data on tended family. The 2006 donkey population totaled 65 the demography of donkey holdings among the eight individuals. The number of donkeys in the care of each Maasai households participating in the study. We go on of the 26 women ranged between 1 and 7. The average to present a brief description of the geography of don- number of donkeys per woman in the sample was 2.5 key use within the settlement catchment area of the (N=26; SD=1.61). The overall male to female ratio was study settlements. The bulk of this section is devoted to approximately 1:2. The herds were dominated by adults presenting input from the various interviews organized that on average make up c. 79% (N=26; SD=30.88) of according to key topics: donkey use and management, the holdings of the individual households in the study. donkeys and residential mobility, women as caretakers The remainder of the herds consisted of juvenile and of donkeys, donkeys in social exchange and archaeo- old donkeys. Settlement S1 at the top of Table 3 con- logical residues. sisted of three separate households and had an excep- tionally high average number of donkeys per woman The demography of don key holdings (3.7; SD=1.5), with a male to female ratio of c. 1:3. The overall proportion of non-adult individuals was 5 of 11 Table 3 details the donkey holdings of 26 women from donkeys (45%). The average number of donkeys of Olobelibel and is broken down by settlement (N=6) and settlement S1 was lower, however, when we accounted household (N=8). Six additional married women from for an additional wife without donkeys in household A these households did not have donkeys of their own and the fact that at least one of the old donkeys was no 66 Fiona Marshall and Lior Weissbrod Fig. 3: Factors affecting the population structure of donkey holdings in the study area over the last decade. longer used for carrying water (Avg.=2.5). Conversely, size and composition of donkey herds belonging to six in settlement S4 where each of the women with don- of the Olobelibel households. The figure shows that keys had only a single individual the average was ex- mortality among donkeys born in Olobelibel was slight- ceptionally low. This can be explained by considerable ly higher than the numbers of local births over the same losses incurred during the drought of 2005-4. The period. The numbers of purchased and borrowed don- highest rate of drought related mortality was reported keys seem to compensate for this discrepancy in births from this settlement where people lost six of their don- versus deaths and reveal the significance of social ex- keys. In comparison, all of the other settlements to- change in household maintenance of donkey holdings. gether lost three animals. Information on the number of Although reproduction is central to demographic donkeys per family prior to the drought was obtained growth, herders make no attempt to control breeding in five of the interviews and for 20 women. The women (Marshall and Weissbrod n.d.). Female donkeys (jen- of S4 used to own three donkeys each on average nies) in oestrus will leave settlement areas in search of (N=20; SD=1.89) prior to the drought. These averages mates, and the males (jacks) in herds are difficult to may not reflect the number of donkeys available to in- control at such times. Castration of males is, therefore, dividual women for moves or for fetching water, the only form of intentional management that affects though, given that there is often considerable sharing mate choice and reproductive selection in Maasai don- among women of the same household or settlement. key herds (see below) (Marshall & Weissbrod n.d.). Information from the interviews identifies a number of The relative proportion of males and females owned by factors that affect the composition and size of donkey different families is variable, but males made up ap- holdings of individual Maasai households. Herd growth proximately one third of the donkeys in the neighbor- results from natural reproduction and a range of other fac- hood that we studied. The strength of male donkeys tors including the purchase and borrowing of donkeys. makes them attractive for transport-use and as a result Donkey herds consist of females and multiple males. In there is likely to be a higher ratio of male:female don- terms of reproductive rates, we were not able to collect keys in family holdings than there would be in cattle or detailed information on the age at which females first sheep and goat herds managed for milk and meat. Peo- come into oestrus, but according to Wilson (1991), this ple’s preferences in this matter are situationally vari- occurs at about 4 years of age. In the study area, females able, however. We found that the male to female ratio typically have one foal every other year. However, herders of donkeys acquired through trade and loans (c. 1:3) by do not try to control reproduction (Marshall & Weissbrod Olobelibel residents exceeds the overall male to female n.d.). Declines in herd size result from miscarriages, in- ratio we recorded for the settlement neighborhood (1:2). fant mortality and drought, disease and predation events. When examined separately, however, we found that all Donkeys are also lost when they stray. These strays may donkeys acquired through loans were female, whereas be returned or are de facto converted to stock-loans to traded donkeys show a male:female ratio approaching families in other neighborhoods. 1:2. People said that they desired both males and fe- males in order to maintain domestic breeding. But they Figure 3 presents information on the way in which ac- also noted that males are especially useful when fe- quisitions and losses over the last decade influenced the males are pregnant or lactating and typically not used Donkeys and Maasai ethnoarchaeology 67 Fig. 4: Household women with firewood and loaded donkeys returning from water for daily water collection. Selection of males or females meat of cattle slaughtered far from the residence or fire- in trade or social exchanges may also be constrained by wood and water needed during ceremonies were men- a number of other factors. Juvenile males have rela- tioned as exceptions, each by different women. A no- tively lower purchase rates than females do and as a tion expressed in one of the interviews is that unlike result gender selection in trade settings is influenced by regular use for water-collection, use of donkeys for how much money a person is willing to spend on a new daily transport of firewood is a ‘non-Maasai’ practice donkey. Nevertheless, more females than males were (Fig. 4). purchased. When it comes to stock loans or social ex- changes, the family lending the donkey will define the Donkeys and water: spatial patterns gender of the loaned animal depending on what they can spare and what they prefer to lend. Given the social Figure 2 is based on GPS data and illustrates the spatial value of this transaction, though, it is likely that every pattern of daily and seasonal cycles of donkey move- attempt will be made to accommodate the wishes of the ments. The location of the Olobelibel settlement cluster person who is borrowing the animal. is shown in relation to neighboring settlement clusters to the east and west, seasonal and permanent water Donkey use and management sources used by the households occupying the Olobeli- bel settlements to the north and east, and the location According to the women “donkey experts” that we of the southern temporary dry-season enkaron settle- talked to, seasonal household moves and transport of ment cluster used by people from Olobelibel. It is im- household water every few days during the dry season portant to note that vehicle traffic in the area of Olobe- are the main tasks of donkeys today. Donkeys were also libel is possible through an unofficial narrow dirt road important for carrying provisions such as maize meal, that connects to adjacent settlement clusters but not be- sugar and tea from regional shopping centers to Olobe- tween the settlements and the permanent water sources libel settlements until relatively recently (c. 1996) when (see also Fig. 1) (i.e. affects contact between settle- a number of shops were built locally allowing women ments more than women’s work of obtaining water). on foot or young men with bicycles to fetch basic sup- Hatched lines represent routes that women take with plies. Other tasks of donkeys were mentioned only an- donkeys on a regular basis from the emparnat settle- ecdotally by a number of the participants. Some re- ments to water. The three marked water points includ- membered rare incidents of donkeys carrying an ing two clusters of river-bed wells and a pipeline tap injured person to a local health facility or the very old (Nolturesh pipeline originating on Mt. Kilimanjaro; during residential shifts. Use of donkeys to haul the see Fig. 1) are communal sources where the individual 68 Fiona Marshall and Lior Weissbrod Fig. 5: Woman loading donkey at dry season wells households of Olobelibel possess rights of water utili- pasture located to the south of the settlements. The his- zation. The wells are dug into the sandy bed of the sea- toric strategy documented for Maasai herders has been sonal Eselenkei River and provide water year-round. to move away from permanent water during the wet The water must be hauled or carried to the surface and season and to reserve grazing close to permanent water packed onto donkeys or poured into troughs for live- for dry season use (Jacobs 1975; Lamprey & Waller stock (Fig. 5). The average length of the three routes to 1990). In recent times, however, the establishment of water from the emparnat settlements of Olobelibel is game reserves/parks and group ranches and increased 4.4 km (SD=0.7) (Fig. 2). Temporary water holes adja- population densities in the study area have resulted in cent to the settlements fill up during the rainy season less mobility and shifts in patterns of mobility (Worden when, according to our observations at Olobelibel, 2007). Wet-season residences or emparnat are now women collect and carry water from the holes without situated closer to permanent water and grazing further using donkeys. Enkaron occupation during the peak of from permanent water is reserved for the dry season. the dry-season entails the use of donkeys for transport- During the dry season households or parts of house- ing household goods from the emparnat, when either holds move to enkaron near dry season grazing. Unlike the whole or a portion of the household may shift, as pastoral societies in the deserts of northern Kenya, the well as longer trips (>14 Km) with donkeys to the per- Horn or the Sahara, Maasai mobility in this better wa- manent sources of water also utilized during occupa- tered and more developed region is relatively limited in tion of the emparnat. both spatial and temporal extent. Typically enkaron areas are located approx. 2 km away from emparnat Donkeys and residential mobility residences in the study area. In this section we describe different strategies of dry- In addition to the availability of water and grazing, a season residential movement of three Olobelibel house- complex series of factors, listed below, are taken into holds in relation to their donkey holdings. These ex- account in decisions to move. Newer considerations, amples illustrate the role of donkeys in facilitating or such as the number of school age children and avail- constraining seasonal mobility and are consistent with ability of schools also play into the choices that herders patterns of variation in L.W.’s broader study of mobil- make. Just as in the past, however, some level of sea- ity and settlement use in the area (Weissbrod in press). sonal mobility is essential in order to ensure the sur- It is important to emphasize that the location of Olobe- vival of households and their livestock herds. Proximate libel settlements as of other Maasai neighborhoods is decisions that householders make regarding seasonal in part a compromise between the need for water lo- movements are affected by neighboring households, cated in Eselenkei River wells to the north of the settle- the timing and intensity of rains and the onset of sea- ments, or the Kilimanjaro pipeline tap, and the need for sonal droughts, the size and composition of herds to be Donkeys and Maasai ethnoarchaeology 69 moved, and whether everyone or just portions of the emparnat, which is situated in another settlement clus- household will participate in the move. Risks associ- ter closer to the permanent water sources of the river. ated with disease, security and other situational factors At the height of the dry season in 2006, as in previous are also weighed. We focus here on the role that don- seasons, most of the household moved away from Olo- key-holdings play in such decisions. belibel back to their emparnat near the river and only the young men (moran pl.) shifted with cattle to the The first example that we examine is settlement S3 (see official enkaron area of Olobelibel within the dry sea- Table 3), where five women possessed 14 donkeys be- son grazing grounds and farther away from permanent tween them and a sixth woman as yet had no donkeys of water. Because relatively few donkeys were available her own. In addition to the use of donkeys, water was to them, members of household A (S1) needed to main- hauled to the settlement on occasion by bicycle and from tain most of the household and sheep/goat herds closer more distant sources by a small truck. During the time (<1.5 km) to water during the dry season and to rely, in of the study at the height of the dry season, members of part, on the women to carry water. settlement S3 decided to split between the emparnat and enkaron settlements. This was desirable because they In a third case, that of household A of settlement S6 at owned a number of pregnant and lactating cattle and Olbelibel, two women relied on two donkeys for access sheep/goats with young offspring and because two of the to water and one of these animals had a spinal defor- women had children who were attending school, which mity that limited transportation of heavy loads. Instead was closer to their wet season emparnat at Olbelibel. of moving nearer to dry season grazing in the south, This split was possible in part because there were enough this household evacuated their Olobelibel settlement donkeys for water to be collected for the dry season en- prior to the height of the dry season of 2006 in order to karon, which is a c. 14 km round-trip from the perma- move north, even closer to permanent water. They were nent wells dug in the sand of the seasonal Eselenkei accommodated in a settlement cluster adjacent to the River or from the Pipeline tap (compared to 8-9 km from river and permanent water of the riverbed wells. In pre- the emparnat; Fig. 2), and also to ensure regular supply vious years household S6 was able to remain at Olobe- to those that remained at the emparnat. libel throughout most years, including the dry season when the cattle (only) were moved to the dry-season Information from “elders” (the age class of older men grazing grounds and cared for by a stock associate involved in decision making) of Olobelibel regarding from settlement S5. The difference in the study year, residential movements indicates that this settlement had women said, was that they had lost donkeys in the re- a particularly high level of flexibility in decision-mak- cent drought. Additional factors affecting the mobility ing regarding residential movements. Discussions with of this household include the shortage of young adult women and men of the households in the settlement children for labor needs and small size of livestock demonstrated that this was strongly related to the avail- holdings during the time of the study. ability of donkeys for transport both for moving house- hold possessions (four women moved their households) Wo m e n a s c a r e t a k e r s of d o n k e y s i n and for regular collection of water over the long round- Maasai society trip distance to water from the enkaron (c. 14 km) and the shorter distance (c. 8 km) from the emparnat. Maasai women are the caretakers of donkeys and don- keys are generally considered women’s animals. Their A different scenario played out among members of use lightens the burden of work for women and without household A in settlement S1. The two women of this a donkey, women literally have greater weight to bear. household possessed four donkeys as their principal Nevertheless, donkeys are not overloaded, as in the ex- means of securing water. Given that one of these don- ample of household A (S6)—old donkeys are not used keys was old and no longer used for carrying water even when they are needed—and the way that women these women had a relatively low number of donkeys treat donkeys reflects Maasai attitudes to animals in gen- (Avg.=1.5). This household normally occupies settle- eral. Examining approaches to donkey caretaking among ment S1 at Olobelibel for a relatively short period an- women participants in the study provides important in- nually at the beginning of the dry season. In other sights into the level and organization of donkey manage- words in contrast to members of settlement S3, this ment in Maasai society. We learned that a load of ap- household uses Olobelibel as a strategic point of access proximately 50 L or 50 kg is the commonly acknowledged to the dry season grazing grounds prior to the onset of maximum load for donkeys during trips to collect water. the seasonal drought, rather like the temporary occupa- We were easily able to discuss and estimate volumes of tion of an enkaron, and divides the seasonal movement water carried because today women in the area no longer into two stages rather than one. In this way, members use gourds or leather water bags and rely on standard of household A in settlement S1 bring their herds clos- commercial plastic containers, often old vegetable oil or er to dry season grazing than they would be at their industrial fluid containers (some L stamped) (Fig. 5). 70 Fiona Marshall and Lior Weissbrod Fig. 6: Donkeys in settlement enclosure We know of only one exception to this weight rule among specific requirements of cattle, sheep and goat, for graz- Olobelibel residents, a case where the owner consid- ing, watering, land use, protection, and medicine, rather ered one particular donkey to be especially strong and than presenting additional challenges for household labor. occasionally loaded the animal with 60 L of water. Donkeys are only herded during periods when they are Women generally rested their pregnant or lactating fe- used regularly, typically the drier months. During these male donkeys (jennies) or loaded them with a lighter times children or young adults of the households usually weight. No-one loads juvenile donkeys (foals) until they herd the donkeys with the calves. During seasons with are one year old. As mentioned, very old donkeys may sufficient rain, when temporary water holes adjacent to be put out to pasture. settlements become filled, aggregated herds of donkeys from settlement neighborhoods can often be seen grazing The typical schedule that women follow for water-col- without supervision. Donkeys can also be seen at times to lection is a trip every other day, usually loading one return independently to settlements at sunset and other- donkey per woman each trip and utilizing 20 L daily wise may be corralled by the women. Figure 6 shows don- per woman. During the dry season when donkeys are keys in the settlement enclosure. Donkey diseases and worked most intensively, owners allow a number of maladies mentioned in the interviews are those character- days of rest between successive trips depending on the istic of other livestock and variably treated with remedies number of donkeys of each household. In the dry sea- in common usage such as liquid salt for coughs, boscia son of August 2006, this was between 1 and 5 days of charcoal (Boscia sp.) or hot compress for wounds, regular rest among households that we studied. At this time spraying for ticks and fleas, and generalized veterinary women in households with particularly low numbers of medicine for other illnesses. donkeys—as a result of drought related losses—were sharing the quantity of water that could be transported One aspect of caretaking specific to donkeys among with available donkeys rather than increasing the fre- Maasai livestock is their habituation to carrying loads quency of trips beyond every other day. In addition, and associating with calves during daily grazing, which most of the women interviewed reported that they re- begin at the time of weaning (one year old). Tasks re- quire three donkeys for transporting the contents of lated to surgery are undertaken by men, who are re- their houses during residential movements. sponsible for castrating male donkeys in cases where this is considered desirable, as for example when a male Our information from Olobelibel indicates that donkey is especially aggressive or sometimes when young caretaking is well integrated into the general pattern of males fail to thrive (8 of 31 [38%] of the males in our Maasai economic activities and dovetails with the more sample were castrated). Particularly aggressive males Donkeys and Maasai ethnoarchaeology 71 may also have a stick threaded through their nostrils. the frequency and social context of immediate benefit Donkey ownership is sometimes marked by cutting off lending of donkeys, which may be less frequent than the tip of the ear or ears of males or females and this is lending of other livestock. Just as for cattle delayed advantageous in cases when animals stray. benefit from lending of donkeys, though, is related to strengthening social relations and relations of mutual Donkeys in social exchange aid. These are essential to long term survival in an un- predictable semi-arid environment in which cycles of The two main forms of social exchange involving don- drought and disease are common, and stock losses due keys in the study area are commercial trade in donkeys to theft, raiding or predation may also occur. through regional markets, traveling salesmen, or local partners in social trades and stock exchange or the Social exchange of donkeys was also examined in lending of donkeys usually among members of the same light of Maasai conceptions of ownership as applied clan and close relatives. Donkeys acquired through to donkeys. Our interviews with women and men commercial trade can originate from long distance from Olobelibel households verified the assertion that sources such as the regional livestock trading center in in Maasai society donkeys belong to married women. the town of Emali roughly 50 km distance from Olo- Nonetheless, in the examples of donkey exchange that belibel on the Nairobi-Mombasa highway. Conversely, came up during interviews, it was the married men, lending of donkeys occurs mostly locally. Among par- the heads of Maasai households and herd owners, who ticipants of the study we recorded cases of borrowing directly conducted or oversaw transactions. All the only (N=5), and no examples of lending out of don- recorded cases of donkey trade (N=13) were conduct- keys. It is important to consider the role of stockloans ed by men. In two cases of donkey borrowing that were or social exchange of donkeys in maintaining the rela- recorded in detail, women reported that the transaction tively even distribution of one to 7 donkeys per mar- was based on the clan/family ties of their husbands ried woman that we observed in Maasai households even though the exchange was initiated and executed (see Table 3). by the women themselves. When asked about the in- volvement of women in exchange of donkeys one of the A theme that recurred during questioning on specific other woman interviewed commented, in fact, that is a cases of borrowing of donkeys by Olobelibel residents ‘matter for men’. was that, as might be expected, it is typically owners with ‘plenty’ of donkeys who lend to relatives or clan In general, it seems that the issue of ownership in the mates with impoverished donkey holdings. In addition, borrowing of donkeys can remain vague and fluctuate it is generally expected that the party in need of don- with the ongoing nature of relations between the lender keys will initiate the transaction and that the loaning and the borrower. In discussing two of the cases of bor- party will initiate the return of the donkeys. Moreover, rowing that we recorded, women considered the don- loan donkeys may be considered in some cases to be key “loans” to be permanent. When asked about the permanent loans. An important aspect of the dynamic status of the progeny that their loan donkeys produce of social exchange based on lending of donkeys in- during the loaning period, however, the women stated volves perceived benefits to lenders with relatively high that these donkeys belonged to the lenders who can po- numbers of donkeys. From the perspective of borrow- tentially claim back the original individuals and their ers whom we interviewed, there may be two types of progeny in the future. Furthermore some disparities benefits to lending donkeys, these we designate “im- that we detected in the accounts of different women’s mediate” and “delayed” benefits. One example of an donkey holdings within a household indicated that loan immediate benefit is in a situation where the loaned donkeys may or may not be counted as part of a spe- donkeys are relocated to areas with seasonally abun- cific set of holdings. Taken together these data suggest dant grass, thus reducing the risk of donkeys going a flexible approach towards the issue of ownership of astray and loss from predation. It is significant that the donkeys in Maasai society. woman who provided this example also emphasized the ways in which this differs from better known prac- D o n k e y s a s w o m e n’s a n i m a l s , tices of cattle lending among stock partners. Cattle taboos and status loans are often seen as a strategy for spreading risk and strengthening social bonds (Ryan et al. 2000). As one In keeping with their role as women’s animals, most of woman explained, however, whereas households can the customs and taboos involving donkeys that we re- survive in the short term without some of their cattle, corded are associated with women. As noted above, in each individual donkey lost from the household even Maasai society donkeys belong to married women who for a short time can considerably affect the ability to rely heavily on donkeys for load-carrying. As one of secure sufficient water for people and livestock on a the women that we talked to in the area commented daily basis. Such considerations may, therefore, affect “without a donkey you have a hole in your head”, refer- 72 Fiona Marshall and Lior Weissbrod ring to indentation of the skull caused by the strap used belief, but do not attain the symbolic value of cattle. In to support heavy loads carried on women’s backs. Each fact, it is common for people to laugh when one speaks of the women of the household has specific rather than seriously of donkeys. The reasons for this disjunction shared hold of a set of individual donkeys. Women are unclear, but because of their slow growth rate don- from Olobelibel reported that donkeys may be passed key herds are not ideal for accumulation of wealth, and on within families as matrimonies from elder women donkeys are not considered to be food. Donkeys in Af- to daughters-in-law and in some cases to younger co- rica are also well known for the length of the penises, wives. Two different elder women whom we inter- their sexual activity and, when they are not herded, for viewed recalled turning over donkeys in their posses- copulating in public places. Furthermore donkeys are sion to a series of new daughters-in-law and younger noisy and their bray is intrusive. Some aggressive male co-wives usually, though, only to those women with donkeys, we were told in Olbelibel, “become a nuisance whom they shared a settlement. Within these two in the neighborhood”. It seems likely, however, that their households we recorded seven cases of donkeys given domestic use and association with work of the household to daughters-in-law and co-wives. and with women, plays a role in the apparent and perva- sive disjunction between the importance of Maasai don- Although social exchange of donkeys is typically con- keys for survival and their relatively low status. ducted by men or based on the social ties of men, wom- en provide important input regarding the preferred sex Archaeological residues and physical condition of donkeys to be acquired. Don- keys seem to play an important role within the concep- Donkeys are not eaten and their carcasses are removed tual world of Maasai women. One woman commented from settlements. The only donkey bones that we saw that “we do not eat donkeys because donkeys are like in the study area, drought related deaths, lay at least people; they do a woman’s work”. Certainly donkeys 200 m from the nearest settlement. As a result there is are heavily associated with Maasai women and inti- low likelihood of preservation of donkey skeletal re- mately involved in tasks that are essential to the run- mains in Maasai settlements. Preservation of deposits ning of the household. The taboo against eating don- of donkey dung in Maasai settlements is, however, a keys, though, applies to all genders, and seems to be possibility because settlements often contain distinct common to African pastoralists regardless of animist, donkey enclosures. Figure 7 presents diagrams of two Moslem, Christian or other beliefs. In some places, Olobelibel settlements with separate enclosures for however, this taboo is broken in secret especially in donkeys. In such enclosures we observed accumula- times of drought (Little & Leslie 1999). We saw no tions of donkey dung, suggesting that the enclosures evidence for this in the study area. specifically constructed to corral donkeys at night are also mainly used for that purpose. As a result, donkey Another important taboo concerning donkeys in Maa- dung deposits accumulate throughout the period of oc- sai society is the absolute prohibition against men cupation of the settlement. touching donkeys that have died. Dead donkeys should also not be touched by younger women or pregnant women, at the risk of infertility or abortion. According Discu ssion to custom, the bodies of donkeys that have died within settlements are ceremonially transported away from The results of this ethnoarchaeological study of don- the settlements. Elder women only, may carry the body key-use among Maasai pastoralists in Kajiado District, of the dead donkey. At this time they should cover their Kenya provide new insights into the intersection be- eyes and sing a song “nobody died, it’s only a donkey tween donkey biology and behavior and husbandry. who died” as they move the animal well away from the Important findings include detailed information on the settlement. There is an additional taboo regarding don- significance and gendered context of tasks for which keys that relates to castration. This is one aspect of donkeys are used, the amount of effort expended in caretaking performed by men who are knowledgeable donkey management, patterns of herd growth, and the regarding the procedure and not by women. As we influence of donkeys on pastoral household labor and learned in one of the interviews, however, men are re- rangeland management. These findings contribute to stricted by taboo to castrating the donkeys of their own understanding domestication processes and the devel- household. opment and spread of pastoralism in Africa. Given their importance to women and essential role in In the study area donkeys belong to women and are residential mobility, the status of donkeys raises some strongly associated with women’s work. Women rely interesting questions. Their monetary value in this re- on donkeys for collecting water on a regular basis and gion is relatively high (c. 4000 KSh or $60 USD), and for carrying household utensils during seasonal resi- they are the subjects of social lending, and of ritual and dential movements. Donkeys are also, as demonstrated Donkeys and Maasai ethnoarchaeology 73 Fig. 7: Map of settlements showing the size and location of donkey enclosures relative to huts and other livestock enclosures in song and ceremony, ritually integrated into the world tasks, and the level of motivation for expending energy of women. Donkey-ownership, however, affects the en- on the donkeys. Animals were herded and brought in at tire community by significantly enhancing the flexibil- night during the dry season when people relied on them ity and resilience of pastoral families. Donkeys provide heavily for water and residential moves. transport, which helps to resolve fundamental tensions between the need for water and the need for grazing. Our results also demonstrate, though, that labor is rare- Families with donkey herds in the study area were able ly dedicated solely to the donkeys, which were pastured to move their households to dry season grazing areas with the calves close to the household. Donkeys also far from water (residential mobility). Access to pack graze and browse, flourish on meager food supplies, donkeys also permitted families to relocate strategi- and were considered by their caretakers to be some- cally in different directions (logistical mobility), ac- what less vulnerable to drought, predation and disease cording to the needs of women with young children, than other livestock. Despite these factors donkey pop- lactating animals with young, or cattle versus sheep or ulation growth was flat or negative. Reproduction in the goat herds. Simply put, donkeys represent mobility. study area did not outpace stock-loss resulting from The maintenance of people and herds far from the wa- stray animals and natural mortality. Natural mortality ter sources in the study area and efficient utilization of in the study area over the last decade resulted from re- pasture resources in different seasons and for different production, old age, predation, and cycles of drought sections of the herds depended to a great extent on the and disease. Donkey herds were maintained during this number of donkeys available. As a result, herders with period through a combination of reproduction, donkey- larger donkey holdings had more flexibility in the com- loans and purchases. African wild ass herds reproduce plex calculations involving livestock and rangeland slowly (Asa 2002; Moehlman 2002) and even though management in an arid environment. domestication has resulted in somewhat earlier sexual and social maturity (first foal at 4 to 5 years), donkey It has been previously argued that donkeys are lightly herd growth is relatively slow (Wilson 1991). Where managed by African pastoralists because they retain predation is a factor, and drought and disease cyclical, physiological and behavioral characteristics of African as in the East African grasslands, it is clear that main- wild ass, including territorial male breeding behavior, taining the size of donkey herds can be a problem. This fewer permanent social associations than other live- is compounded by the transport use of donkeys, which stock (Klingel 1974), and greater success at deterring results in people keeping relatively high proportions of predators and digging for water (Marshall 2007; Mar- males in herds, because they are strong and consistent- shall & Weissbrod n.d.). Maasai patterns of manage- ly available. Cattle or sheep/goat herds managed for ment in the study area are, however, more complex than maximum growth are made up of high proportions of predicted. The amount of effort that women in the study females and reproduce much more rapidly (Dahl & area expended on donkey management varied season- Hjort 1976). Donkey herds are, therefore, not an ideal ally corresponding directly to the level of use of the vehicle for accumulation of wealth (for cattle see Gran- animals, the need for predictable availability for daily din 1988; Schneider 1979). 74 Fiona Marshall and Lior Weissbrod We conclude, however, that donkeys have relatively with which they can be herded with cattle is likely to few costs and play an important economic role, signifi- have eased trade-offs between management of cattle cantly enhancing options for flexible mobility and sur- and wild ass/donkey herds. We predict based on this vival in the study area. This finding accords with Ja- study, that levels of management would have varied ac- cobs’ (1975) observations, based on earlier patterns of cording to use and the need for scheduled availability Maasai mobility in the region. Our results also support of donkeys. Nevertheless, during initial phases of do- broader assertions regarding the significance of don- mestication investment must have been high, and cap- keys for pastoral mobility and resilience in African tured wild ass tied up (hobbled) or penned in thorn rangelands (Marshall 2007). fences and provisioned for months before they were tame enough to herd. Nicolaisen (Nicolaisen & Nico- Archaeological implications laisen 1997: 165-166) documents capture and taming of wild ass by Tuareg herders during the 1950’s, noting We think that these findings are relevant (sensu Wylie that captured animals joined the domestic herds after c. 1985, 2002) to modeling three different phases of de- one to two months. The levels of management em- velopment of African pastoralism, domestication of the ployed by early herders have a direct bearing both on donkey, increasingly mobile pastoral societies of the selection processes during domestication and on meth- mid-Holocene, and long-distance migrations and colo- ods that can be used to identify donkey domestication nization events resulting from Saharan expansion. and their presence in pastoral sites. Cattle were domesticated and pastoralism developed in the eastern Sahara between approximately 10,000 and Reliance on donkeys for transport during long-distance 8500 cal BP (by 7500 uncal bp) (Barich 2002; Bradley residential movements and for daily tasks such as col- & Magee 2006; Close & Wendorf 1992; Garcea 2004; lection of water would have significantly enhanced the Gautier 1987b). These events coincided with climatic ability of early cattle pastoralists to reconcile needs for fluctuations and unpredictable aridity that marked the water and pasture. In seasonally and annually variable end of the Holocene wet phase in the Sahara (Kuper & arid lands such as the mid-Holocene Sahara (di Lernia Kropelin 2006; Kropelin et al. 2008). There were many 2002; Kropelin et al. 2008), these resources are avail- subsequent oscillations between wetter and drier phas- able in patchily distributed locations and at varying es, but the long term trend in northeastern Africa has times (Behnke et al. 1993). Our results demonstrate been one of increasing aridity. Sheep and goats were that in a semi-arid environment, such as that of our adopted from Asia by the mid-eighth millennium cal study area, one donkey can be used to carry at least 50 BP (c. 7000 bp) and incorporated into pastoral systems L of water for 8 km every other day, equivalent to two (Close 2002). Domestication of the donkey seems to days supply for a nuclear family. This represents a sav- have occurred during a period of rapid expansion of the ing in human labor, an absolute increase in the distance Sahara Desert by the early seventh millennium cal BP from water that it is viable to live, and—depending on (c. >6000 uncal bp) (Beja Pereira et al. 2004; Marshall how many donkeys are available—opens a variety of 2007; Rossel et al. 2008). During this period pastoral options for residential and logistical mobility and societies became increasingly mobile. The develop- rangeland management. These findings add detail to ment of the hyper-arid Sahara during the sixth millen- previous syntheses (Marshall 2007), illustrating ways nium cal BP (by c. 5000 bp) resulted in long distance in which use of donkeys to transport water and other migration of pastoral peoples out of the Sahara and crucial resources to people and livestock and vice versa colonization of Sahelian zones and regions as far south would have reduced energy expended in domestic as the equator in eastern Africa (Gifford-Gonzalez work, contributed to flexible mobility, and as a result, 2005; Marshall 2000). could have improved the health and food security of ancient pastoralists and their herds. African wild ass were distributed in arid lands of north- eastern Africa, and would have been brought into in- During the mid-Holocene early herders of the central creasing proximity with early cattle pastoralists as Sahara spent shorter periods of time in one location and herders were forced to live in drier areas during arid relied more heavily on sheep and goats (Gautier 1987a; climatic episodes. We hypothesize that the subsequent also Pöllath & Peters 2007). Between the mid-sixth and domestication of African wild ass was intentional and fifth millennia BP (c. 5000-4000 bp) herders moved as that their size and desert adaptation provided the initial far south as Lake Turkana in eastern Africa. By the motivation for capture (see also Marshall & Weissbrod mid-fifth millennium BP (approximately 4000 bp) they n.d.). There is no evidence so far, though, that they were established settlements in the West African Sahel (Gif- eaten in any numbers. The resilience and utility of wild ford-Gonzalez 2005; Holl 1998; Linseele 2007; Mac- ass/early donkeys as transport animals may have pro- Donald and MacDonald 2000; Marshall and Hildeb- vided key incentives for management. But the relative- rand 2002; Smith 1992). Our findings show that owning ly low investment requirements of donkeys and ease at least three donkeys per family would have made it Donkeys and Maasai ethnoarchaeology 75 possible for people to move with their herds and house- are not depicted in scenes of ancient Saharan campsites hold possessions. We hypothesize that reliance on don- though they are the subject of isolated and uncommon keys for transport underlies the successful mobile re- rock art panels throughout the Sahara (Muzzolini 1996, sponses of food producers to climate change in north- 2000). Depictions of wild ass or ancient donkeys in eastern Africa during the mid- to late-Holocene and the ritual exist, however, in rock art of the Libyan Messak spread of food production in much of the continent. (Lutz and Lutz 1995: 97; Muzzolini 1996, 2000). The hypothesis that households were the domain of women, Our results suggest, however, that low levels of growth and collecting water was women’s work and strongly of donkey herds are likely to have presented a problem associated with donkeys, could be tested in the future where pastoralists were living at low densities. Contin- through the discovery of burials of women with don- ued recruitment from the wild and reduced disease- and key-related material culture such as donkey panniers predation-related mortality may have alleviated the and depictions of donkeys, or of rock art panels or don- problem of slow reproductive rates during the early key burials where women are associated with donkeys. phases of domestication of the donkey. A crucial period during which African pastoralists were highly mobile, Detection of donkeys in archaeological sites is, however, dependent on their donkeys, and moving into very low difficult. Marshall (2007) argued that because of the way density contexts, occurred later during southward mi- that donkeys are used and managed, the rarity of don- grations and colonization of Sahelian and eastern Afri- keys in pastoral sites does not accurately reflect their can regions occupied by hunter-gatherers. Sahelian significance to prehistoric pastoralists. In East Africa hunter-gatherers did not at first keep any livestock, oth- <five donkey bones are known from >300,000 Pastoral er pastoralists were few and far between, and herders Neolithic specimens analyzed. The results of this study would have found donkeys difficult to obtain through also indicate that where donkeys are not eaten, as in the social exchanges. We predict, therefore, that both strat- study area, and the intensity of management is relatively egies for maintaining social ties and stock exchange, low, donkey bones are absent from settlements. Signifi- and increased investment in donkey-management, would cantly, though, our results also demonstrate that geoar- have been especially important for migrating herders. chaeological and isotopic research on ancient sediments The mid- to late-Holocene is also associated with evi- from donkey pens constitute an important line of evi- dence for increasing ritualization of cattle (di Lernia dence for identification of donkeys on pastoral sites. 2006) and construction of conspicuous ceremonial struc- Corrals constructed by African herders specifically for tures across the Sahara (Gifford-Gonzalez et al. 2008; the use of donkeys have not previously been document- Paris 2000; Wendorf & Krölik 2001) and in Eastern Af- ed. In the study area, however, donkeys are often brought rica (Gifford-Gonzalez 1998). Some of these ceremonial into their own donkey pens within the settlement perim- centers may have functioned as seasonal points of aggre- eter thorn-fences at night and thick lenses of distinctive gation and stock exchange (Marshall et al. in press). donkey dung accumulate. As a result micromorphology, phytolith concentrations, mineral assemblages and the The related issue of whether the strong association that isotopic composition of organic nitrogen could be used we note between women and donkeys in the study area to detect the presence of ancient donkeys dung following also existed in northeastern Africa in the past, is more approaches developed by Shahack-Gross and colleagues complex. Household work, specifically the hard labor (Shahack-Gross et al. 2003, 2008) to identify cattle, of acquiring cooking and drinking water in arid lands sheep and goat dung in degraded sediments in prehis- and responsibility for houses and their contents in a toric African pastoral sites. Because equids are not ru- mobile society, are the keys to this gendered associa- minants, larger fragments of vegetation make donkey tion. Because the link between donkeys and women’s dung readily distinguishable from that of cattle, sheep work is broadly cross-cultural among recent African and goat. Pellet size and shape, hoof conformation and pastoralists (Marshall 2007), we suggest that it may trampling morphology should also distinguish equid have a deep history. Archaeologists do not yet know, from bovid accumulations. though, much about gender and household-organiza- tion, food supply and cooking in prehistoric African pastoral societies (Gifford-Gonzalez 1998). Neverthe- Conclu sion less, mid-Holocene rock art of the Tassili in Algeria depicts women erecting houses in pastoral camps (Holl This research provides insights into domestication pro- 2004). Holl (2004: 22) notes that in the panel that he cesses and the development of mobile pastoral respons- studied in Dr. Khen Shelter at Iheren (Composition es to Saharan climate change, the spread of food pro- One, Scene Two), almost all the women, but none of the duction into eastern and western African, and the men are involved in hut construction. Pots possibly long-term success of mobile pastoral strategies in Af- containing food or water are also a subject of some of rica. It points to significant linkages between women’s these panels (Holl 2004). Donkeys, on the other hand, roles as caretakers of donkeys in African pastoral soci- 76 Fiona Marshall and Lior Weissbrod eties, relatively low intensity of donkey management, Beja Pereira A., England P.R., Ferrand N., Jordan S., Bakhiet A.O., and the context and pattern of domestication of the don- Abdalla M.A., Mashkour M., Jordana J., Taberlet P. & Luikart G., 2004. African origins of the domestic donkey. Science 304: 1781. key in Africa. The addition of the donkey to the Afri- can package of early animal domesticates in the social- Bekure S. & Grandin B.E., 1991. economic context of mobile pastoralism undoubtedly Introduction. In: Bekure S., de Leeuw P.N., Grandin B.E. & played a key role in the development, long-term suc- Neate P.J.H. (eds). Maasai Herding: an analysis of the livestock cess, and spread of African pastoralism. As efficient production system of Maasai pastoralists in eastern Kajiado pack animals, donkeys would have facilitated the flex- District, Kenya: 1-5. Addis Ababa, Ethiopia: ILCA Systems Study No. 4. ibility and stability of early herding systems thereby contributing to the unique African pattern of delayed Blench R.M., 2000. domestication of plants and beginning of settled farm- A History of Donkeys, Wild Asses and Mules in Africa. In: Mac- ing-based village life (see Gifford-Gonzalez 2005; Donald, K.C. & Blench R.M. (eds). The Origins and Develop- Marshall & Hildebrand 2002; Marshall 2007; Neu- ment of African Livestock: Archaeology, Genetics, Linguistics mann 2005). More widespread ethnoarchaeological and Ethnography: 339-354. London: University College London Press. studies of donkeys and pastoral societies are needed, however, to explore contemporary social and regional Bradley D.G. & Magee D., 2006. variability and to develop finer grained temporally and Genetics and Origins of Domestic Cattle. In: Zeder M., Bradley regionally specific hypotheses regarding the role of D.G., Emshwiller E. & Smith B.D. (eds). Documenting Domes- donkeys in the development of African pastoralism. tication: New Genetic and Archaeological Paradigms: 317-328. Berkeley: University of California Press. BurnSilver S.B., 2003. Ack nowle dge me nt s Linking pastoralists to a heterogenious landscape: The case of four Maasai group ranches in Kajiado District, Kenya. In: Fox J., We thank Dr. Joris Peters for the invitation to contrib- Rindfuss R.R., Walsh S.J. & Mishra V. (eds). People and the ute to this volume and are grateful to Veerle Linseele Environment: 173-199. Boston: Kluwer Academic. and anonymous reviewers for their helpful comments. We also thank Henriette Obermaier and George McG- Close A., 2002. Sinai, Sahara, Sahel: The introduction of domestic caprines to lynn. Without the expert knowledge and gracious sup- Africa. In: Jennerstrasse (ed). Tides of the Desert-Gezeiten der port of the people among whom we worked in Kajiado Wüste: 459-469. Cologne: Henrich-Barth-Institut. District this research could never have been conducted. We are most grateful to Joseph Parteri Lekanaiya for Close A. & Wendorf F., 1992. his research assistance and dedicate this paper to the The beginnings of food production in the eastern Sahara. In: women of the neighborhoods. We are indebted to Dr. Gebauer A.B. & Price T.D. (eds). Transitions to Agriculture in Prehistory: 63-72. Madison: Prehistory Press. Jeff Warden for introductions. We are also grateful to the Kenya National Museums and Dr. Purity Kiura for Coppock D.L., 1994. their support and the Government of Kenya for permis- The Borana Plateau of Southern Ethiopia: Synthesis of pastoral sion to undertake research. This research was support- research, development and change, 1980-1991. Addis Ababa: ed by NSF grants BCS-0447369 and BCS-0536507. International Livestock Centre for Africa, Systems Study 5. Dahl G. & Hjort A., 1976. Having Herds: Pastoral Herd Growth and Household Economy. Bibl iog r aphy Stockholm: Stockholm Studies in Social Anthropology 2. Arhem K., 1985. Dijkman J.T., 1991. Pastoral Man in the Garden of Eden. Uppsala: Uppsala Research A note on the influence of negative gradients on the energy ex- Reports in Cultural Anthropology. penditure of donkeys walking, carrying and pulling loads. In: Fielding D. & Pearson R.A. (eds). Donkeys, Mules and Horses in Asa C.S., 2002. Equid Reproductive Biology. In: Moehlman P.D. (ed). Equids: Tropical Agricultural Development: 221-222. Edinburgh: Uni- Zebras, Asses and Horses: 113-117. Gland, Switzerland: IUCN. versity of Edinburgh. Barich B., 2002. di Lernia S., 2002. Cultural responses to climatic changes in North Africa: Begin- Dry climatic events and cultural trajectories: adjusting Middle ning and spread of pastoralism in the Sahara. In: Hassan F. (ed). Holocene pastoral economy of the Libyan Sahara. In: Hassan F. Droughts, Food and Culture: 209-225. New York: Kluwer Aca- (ed). Droughts, Food and Culture: 209-225. New York: Kluwer demic. Academic. Behnke R.H., Scoones I. & Kerven C., 1993. di Lernia S., 2006. Range Ecology at Disequilibrium: New Models of Natural Vari- Building monuments, creating identity: cattle cult as a social ability and Pastoral Adaptation in African Savannas. London: response to rapid environmental changes in the Holocene Sa- Overseas Development Institute. hara. Quaternary International 151: 50-62. Donkeys and Maasai ethnoarchaeology 77 Galaty J.G., 1981. Holl A., 2004. Land and livestock among Kenyan Maasai. Journal of African Saharan Rock Art. Walnut Creek: Altamira Press. Studies 16: 68-88. Homewood K., Kristjanson P. & Chenevix Trench P., 2009. Galaty J.G., 1982. Staying Maasai? Livelihoods, Conservation and Development Being “Maasai”; being “people-of-cattle”: ethnic shifters in East in East African Rangelands. New York: Springer Press. Africa. American Ethnologist 9: 1-20. Homewood K.M. & Rodgers W.A., 1991. Garcea E.A.A., 2004. Maasailand Ecology: Pastoralist Development and Wildlife An alternative way towards food production: The perspective Conservation in Ngorongoro, Tanzania. Cambridge: Cambridge from the Libyan Sahara. Journal of World Prehistory 18: 107- University Press. 154. Jacobs A.H., 1975. Gautier A., 1987a. Maasai Pastoralism in Historical Perspective. In: Monod T. (ed). The archaeological sequence of the Acacus. In: Barich B.E. (ed). Pastoralism in Tropical Africa: 406-425. London: Oxford Uni- Archaeology and Environment in the Libyan Sahara, 1978-1983: versity Press. 283-308. Oxford: Cambridge Monographs in African Archaeol- ogy 23, (B.A.R. International Series 368). Jones P., 1977. Donkeys for Development. Animal Traction Network for East- Gautier A., 1987b. ern and Southern Africa (ATNESA). Prehistoric men and cattle in North Africa: A dearth of data and a surfeit of models. In: Close A.E. (ed). Prehistory of Arid North Klima G.J., 1970. Africa: 163-187. Dallas: Southern Methodist University Press. The Barabaig: East African Cattle-Herder. New York: Holt Rinehart and Winston. Gifford-Gonzalez D.P., 1998. Early Pastoralists in East Africa: Ecological and Social Dimen- Klingel H., 1974. sions. Journal of Anthropological Archaeology 17: 166-200. Social Organization and Behaviour of the Grevy’s Zebra. Zeitscrift für Tierpsychologie 35: 37-70. Gifford-Gonzalez D.P., 2005. Kropelin S., Verschuren D., Lézine A.M., Eggermont H., Cocquyt C., Pastoralism and its Consequences. In: Stahl A. (ed). African Francus P., Cazet J.-P., Fagot M., Rumes B., Russell J.M., Darius F., Archaeology: A Critical Introduction: 187-224. Cambridge: Conley D.J., Schuster M., von Suchodolet H. & Engstrom D.R., 2008. Blackwell. Climate-driven ecosystem succession in the Sahara: the past 6000 years. Science 320: 765-768. Gifford-Gonzalez D. (general ed), Clark J.D., Crader D.C., Garcea E.A.A., Gifford-Gonzalez D., Smith A.B. & Williamson M.A.J., Kuper R. & Kropelin S., 2006. 2008. Climate-Controlled Holocene Occupation in the Sahara: Motor Adrar Bous: The Archaeology of a Granitic Ring Complex in of Africa’s Evolution. Science 313: 803. Central Sahara, Niger. Tervuren: Annales in Archaeology, Roy- al Africa Museum. Lamprey R. & Waller R., 1990. The Loita-Mara region in historical times: Patterns of subsis- Grandin B.E., 1988. tence, settlement and ecological change. In: Robertshaw P. (ed). Wealth and pastoral dairy production. A case study from Maas- Early Pastoralists of South-Western Kenya: 16-35. Nairobi, Ke- ailand. Human Ecology 16: 1572-1915. nya: British Institute in Eastern Africa. Grandin B.E., 1991. Linseele V. 2007. The Maasai: socio-historical context and group ranches. In: Archaeofaunal remains from the past 4000 years in Sahelian Bekure S., de Leeuw P.N, Grandin B.E. & Neate P.J.H. (eds). West Africa. Domestic livestock, subsistence strategies and en- Maasai Herding: an analysis of the livestock production system vironmental changes. Cambridge: BAR International Series. of Maasai pastoralists in eastern Kajiado District, Kenya: 21-39. Addis Ababa, Ethiopia: ILCA Systems Study No. 4. Little M.A. & Leslie P.W. (eds). 1999. Turkana Herders of the Dry Savanna. Ecology and Behavioral Grandin B.E., de Leeuw P.N. & de Souza M., 1991. Response of Nomads to an Uncertain Environment. Oxford: Ox- Labour and livestock management. In: Bekure S., de Leeuw P.N, ford University Press. Grandin B.E. & Neate P.J.H. (eds). Maasai Herding: an analysis of the livestock production system of Maasai pastoralists in east- Lutz R. & Lutz G., 1995. ern Kajiado District, Kenya: 71-82. Addis Ababa, Ethiopia: The secret of the desert: the rock art of Messak Sattafet and Mes- ILCA Systems Study No. 4. sak Mellet, Libya. Innsbruck: University Press Golf Verlag. Groves, C., 1986. MacDonald K. & MacDonald R.H., 2000. The taxonomy, distribution, and adaptations of recent Equids. The origins and development of domesticated animals in arid In: Meadow, R.H. & Uerpmann H.P. (eds). Equids in the Ancient West Africa. In: Blench R.M. & MacDonald K.C. (eds). The Ori- World:11-65. Wiesbaden: Dr. Ludwig Reichert Verlag. gins and Development of African Livestock: Archaeology, Genet- ics, Linguistics, and Ethnography: 127-162. London: UCL Press. Holl A., 1998. Livestock husbandry, pastoralism, and territoriality: The West Maloiy G.M.O., 1970. African record. Journal of Anthropological Archaeology 17: Water economy of the Somali donkey. American Journal of 143-165. Physiology 219: 1522-27. 78 Fiona Marshall and Lior Weissbrod Marshall F., 2000. Rossel S., Marshall F., Peters J., Pilgram T., Adams M.D. & The Origins and Spread of Domestic Animals in East Africa. O’Connor D., 2008. In: MacDonald K.C & Blench R.M. (eds). The Origins and De- Domestication of the Donkey: New Data on Timing, Processes velopment of African Livestock: Archaeology, Genetics, Lin- and Indicators. Proceedings of the National Academy of Sci- guistics and Ethnography: 191-221. London: University College ences 105: 3715-3720. London Press. Ryan K., Munene K., Kahinju S.M. & Kunoni P.N., 2000. Marshall F., 2007. Ethnographic perspectives on cattle management in semi-arid African pastoral perspectives on domestication of the donkey: A environments; a case study from Maasailand. In: MacDonald fi rst synthesis. In Denham T.P., Iriarte J. & Vrydaghs L. (eds). K.C. & Blench R.M. (eds). The Origins and Development of Af- Rethinking Agriculture: Archaeological and Ethnoarchaeologi- rican Livestock: Archaeology, Genetics, Linguistics, and Eth- cal Perspectives: 537-594. California: Walnut Creek. nography: 462-477. London: UCL Press. Marshall F. & Hildebrand E., 2002. Schneider H.K., 1979. Cattle before Crops: the origins and spread of food production in Livestock and equality in East Africa: the economic basis for Africa. Journal of World Prehistory 16: 99-143. social structure. Bloomington: Indiana University Press. Marshall F. & Weissbrod L., nd. Shahack-Gross R., Marshall F. & Weiner S.T., 2003. Domestication Processes and Morphological Change: Through Geo-ethnoarchaeology of pastoral sites: The identification of the Lens of the Donkey and African Pastoralism. Current An- thropology. livestock enclosures in abandoned Maasai settlements. The Journal of Archaeological Science 30: 439-459 Marshall F., Grillo K. & Arco L., nd. Prehistoric East African pastoralists and changing opportunities Shahack-Gross R., Simons A. & Ambrose S.H., 2008. for coping with climatic risk. In: Miller N., Moore K.M. & Ryan Identification of pastoral sites using stable nitrogen and carbon K. (eds). Sustainable lifeways: Cultural Persistence in an Ever- isotopes from bulk sediment samples: a case study in modern changing Environment. University of Pennsylvania, Philadel- and archaeological pastoral settlements in Kenya. Journal of phia: MASCA, the University Museum. Archaeological Science 35: 983-990. Merker M., 1910. Smith A.B., 1992. Die Masai. Berlin: D. Reimer. Pastoralism in Africa: Origins and Development Ecology. Lon- don: Hurst and Company. Moehlman P.D., 2002. Status and action plan for the African wild ass (Equus africa- Spear T. & Waller R., 1993. nus). In: Moehlman P.D. (ed). Equids: Zebras, Asses and Horses: Being Maasai: Ethnicity & Identity in East Africa. London: 2-9. Gland, Switzerland: IUCN James Currey Ltd. Muzzolini A., 1996. Spencer P., 1988. Les équidés des figurations rupestres Sahariennes. Anthropolo- The Maasai of Matapato: A Study of Rituals of Rebellion. gie XXXIV: 185-202. Bloomington: Indiana University Press. Muzzolini A., 2000. Spencer P., 1993. Livestock in Saharan Rock Art. In: MacDonald K.C. & Blench Becoming Maasai, being in time. In: Spear T. & Waller R. (eds) R.M. (eds). The Origins and Development of African Livestock: Being Maasai: Ethnicity & Identity in East Africa: 140-156. Archaeology, genetics, linguistics and ethnography: 87-110. London: James Currey Ltd. London: University College London Press. Starkey P., 2000. Neumann K., 2005. The history of working animals in Africa. In: MacDonald K.C. The Romance of Farming: Plant Cultivation and Domestication & Blench R.M. (eds). The Origins and Development of African in Africa. In: Stahl A. (ed). African Archaeology: A Critical Livestock: Archaeology, Genetics, Linguistics and Ethnogra- Introduction: 249-275. Cambridge: Blackwell. phy: 478-502. London: University College London Press. Nicolaisen J. & Nicolaisen I., 1997. Talle A., 1987. The Patoral Tuareg, 2 volumes. New York: Thames and Hudson. Women as heads of houses: The organization of production and role of women among pastoral Maasai in Kenya. Ethnos 52: 50-80. Paris F., 2000. African livestock remains from Saharan mortuary contexts. In: Talle A., 1988. MacDonald, K.C. and Blench R.M. (eds). The Origins and De- Women at a Loss. Changes in Maasai Pastoralism and their Ef- velopment of African Livestock: Archaeology, Genetics, Lin- guistics and Ethnography: 111-126. London: University College fects on Gender Relations. Stockholm Studies in Social Anthro- London Press. pology, No.19. Stockholm: Scandinavian Institute of African Studies. Pöllath N. & Peters J., 2007. Holocene climate change, human adaptation and landscape deg- Vilà, C., Leonard J.A. & Beja-Pereira A., 2006. radation in arid Africa as evidenced by the faunal record. In: Genetic documentation of Horse and Donkey Domestication. In: Bubenzer O., Bolten A. & Darius F. (eds). Atlas of Cultural and Zeder M., Bradley D., Emshwiller E. & Smith B.D. (eds). Docu- Environmental Change in Arid Africa: 64-67. Cologne: Heinrich menting Domestication: 342-353. Berkeley: University of Cali- Barth Institut. fornia Press. Donkeys and Maasai ethnoarchaeology 79 Weissbrod L., in press. Worden J.S., 2007. Biological indicators of occupation intensity: An environmental Fragmentation and settlement pattern in Maasailand – Implica- ethnoarchaeology of Maasai settlements. In: Dean R.M. (ed). The tions for pastoral mobility, drought vulnerability, and wildlife Archaeology of Anthropogenic Environments: 295-320 (Occasional conservation in an East African Savanna. PhD dissertation Col- Paper No. 37). Carbondale: Southern Illinois University Press. orado State University, Fort Collins, Colorado. Wendorf F. & Królik H., 2001. Wylie A., 1985. Site E-96-1: The complex structures or shrines. In: Wendorf F., The Reaction against Analogy. Advances in Archaeological Schild R., & Associates (eds). Holocene Settlement of the Egyp- Method and Theory 8: 63-110. tian Sahara Vol. 1: The Archaeology of Nabta Playa: 503-520. New York: Plenum Publishers. Wylie A., 2002. Thinking from Things. University of California Press: Berkeley. Western D. & Dunne T., 1979. Environmental aspects of settlement site decisions among pas- Yousef M.K., 1991. toral Maasai. Human Ecology 7: 75-98. Energy cost of locomotion in the donkey. In: Fielding D. & Pear- son R.A. (eds). Donkeys, Mules and Horses in Tropical Agricul- Wilson R.T., 1991. tural Development: 220, Edinburgh: University of Edinburgh. Equines in Ethiopia. In: Fielding D. & Pearson R.A. (eds). Don- keys, Mule and Horses in Tropical Agricultural Development: 185-188. Edinburgh: University of Edinburgh.