Human Ecology, VoL 19, No. 1, 1991

Indigenous Kikuyu Agroforestry: A Case Study of Kirinyaga, Kenya Alfonso Peter Castro I

This" article analyzes agroforestry practices among the Ndia and Gichugu Kikuyu of Kirinyaga, Kenya, at the turn of the century, before the onset of" colonial rule. It describes ways" in which people adapted to competing pressures for retaining and removing tree cover. It shows" how religious beliefs, tenure relations based on a communal property-rights regime, and farm forestry practices contributed to the conservation of trees. Such strategies were not aimed at reversing deforestation, but mitigating its" impact by incorporating valued trees" into local sociocultural and household production systems. The article points" out that indigenous agroforestry practices need to be viewed in the context of" local socioeconomic" and ecological differences. It also considers the impact of the caravan trade on land use during the late 1800s. Tree scarcity in the late precolonial era is" briefly contrasted with the area's "woodfuel crisis" of the 1980s. KEY WORDS: agroforestry; forests; firewood; indigenous technical knowledge; common property; conservation; IZdkuyu;Kenya

INTRODUCTION

Agroforestry is a form of land use that simultaneously or sequentially combines trees with crop or animal production (Lundgren, 1982). It has become increasingly accepted among professional foresters as a socially and ecologically sustainable land use strategy (see FAO, 1985; Falconer and Arnold, 1989; Gregersen et al., 1989; Cook and Grut, 1989; Kerkhof,

tDepartment of Anthropology, Syracuse University, 308 Bowne Hall, Syracuse, New York 13244. 1 0300-7839/91/0300qJ00150~,00/09 1991 Plenum Publi~hin,~Corpora,ion

2

Castro

1990; Guggenheim and Spears, 1991). 2 Yet, many agroforestry techniques have long been practiced by cultivators and herders throughout the world (Olofson, 1983; Niamir, 1990; Castro, 1991a). This article examines agroforestry practices among the Ndia and Gichugu Kikuyu of Kirinyaga, Kenya, at the turn of the century, prior to the onset of colonial rule. It describes ways in which people adapted to competing pressures for retaining and removing tree cover. In particular, the article shows how religious beliefs, tenure relations based on a communal property-rights regime, and farm forestry techniques contributed to the conservation of trees. It is argued that such practices were not aimed at reversing the process of deforestation. Rather, their goal was to mitigate its impact by incorporating valued trees into local sociocultural and production systems. The article points out that indigenous agroforestry practices need to be viewed in the context of local socioeconomic and ecological differences. It also considers the impact of the coastal caravan trade on land use during the late 1800s. Finally, localized tree scarcity in the late 1800s is contrasted with the area's "firewood crisis" of the 1980s. The purpose of this essay is to underscore the historical dynamism of local land use systems. Too often, present-day practices are viewed as if they are static forms of behaviors, without realization of the social and historical processes from which they emerged. Thus, contemporary agroforestry systems in central Kenya are not simply age-old traditions, but the outcome of processes spanning the precolonial, colonial, and independence eras (Brokensha et al., 1983; Castro, 1983, 1988, 1990, 1991b,c; Riley and Brokensha, 1988a).

SETTING AND SOURCES Kirinyaga is located on the southern slopes of Mount Kenya (5,199 m), the second highest mountain in Africa. It is Kenya's smallest rural district, covering 1,437 km 2 in Central Province. In the colonial era the area was included in the administrative boundaries of Nyeri and Embu Districts. Today, Kirinyaga contains some of the most productive farmland in the country because of its generally fertile soils and abundant rainfall. The district is the traditional homeland of the Ndia and Gichugu Kikuyu. The current population exceeds 300,000 (Kirinyaga District, 1984). 2Other useful sources on agroforestry include publications by the International Council for Research on Agrotbrestry, the ODI Social Forestry Network, and the Forests, Trees, and People Programme sponsored by the Food and Agriculture Organization of the United Nations and SIDA.

Indigenous Kikuyu Agroforestry

3

In pre-colonial times, the people of Kirinyaga were similar to other acephalous agriculturalists of central Kenya (see Routledge and Routledge, 1910; Kenyatta, 1938; Fisher, 1952; Lambert, 1956; Middleton and Kershaw, 1965; Saberwal, 1970; Muriuki, 1974; Mwaniki, 1974; Munro, 1975; Rogers, 1979; Fadiman, 1982; Glazier, 1985; Ambler, 1988; Riley and Brokensha, 1988a; Davison, 1989). Social organization centered around patrilineal kinship (clan, lineage, and household) and territorial groups (age- and generationsets). They resided in dispersed homesteads, usually situated along ridges, in the highlands with abundant rainfall and fertile soils. The area was brought under colonial control in 1904 (Mungeam, 1966). Four types of sources have been used to reconstruct indigenous agroforestry practices: European accounts of Kirinyaga at the time of initial contact, including colonial records, testimony about pre-colonial land use by Kikuyus and colonial officials at the Kenya Land Commission in 1932, descriptions from other parts of Kikuyuland and central Kenya, and data collected by the author in Kirinyaga, Kenya, during 1982-83. All of the sources required careful sifting and analysis. Some writers were misinformed or biased in their depiction or interpretation of native land use. These are offset by the descriptive excellence contained in other accounts by other explorers, officials, and ethnographers. None of the elders interviewed in contemporary Kirinyaga had been adult participants in pre-colonial life, but they were familiar with many aspects of local custom. In particular, they had first-hand knowledge of the common-property resource regime that existed before land privatization and the expansion of cash-cropping in the 1950s. The present study focuses on analyzing socioeconomic and historical dimensions of indigenous agroforestry, rather than on detailing every aspect of traditional tree use. For in-depth botanical identities and uses, see Leakey (1977c) and Riley and Brokensha (1988b). It is also beyond the scope of this article to analyze change and continuity in local agroforestry during the present century. As might be expected, however, indigenous agroforestry systems sometimes provided a cultural foundation for the numerous state-organized tree planting and woodland protection interventions in Kirinyaga since the earliest days of colonial rule (see Castro, 1983, 1988, 1990, 1991b).

PRESSURES TO RETAIN AND TO REMOVE TREE COVER Trees were a vital part of the pre-colonial Kikuyu resource base. They furnished fundamental items of material culture such as timber, fuel, fencing, foodstuffs, fodder, medicine, utensils, fibers, and weapons

4

Castro

(Middleton and Kershaw, 1965; Leakey, 1977a). Trees also served a range of other purposes. They provided environmental amenities, shade and windbreak, and were used as boundary markers. Selected groves and trees were sites of communal worship and local ceremonies, including initiation rites (Castro, 1990). Traditional dance grounds were sometimes situated in forest clearings. Thick woodland acted as a buffer against outside marauders, particularly the Maasai (Kenya Land Commission, 1934, p. 412). Not only were frontier homesteads protected by dense foliage (Leakey, 1977a), but Kikuyu warriors were adept at carrying out ambushes while hidden in trees (Meinertzhagen, 1957). The multiple uses of trees reflected both the area's diverse flora and the impressive botanical knowledge accumulated by the Kikuyu (Leakey, 1977a). In the 1930s, Leakey (1977a) recorded from Kiambu Kikuyu elders the names of over 400 different trees and plants used during pre-colonial times. Leakey (1977c, p. 1286) noted that his list was "not in any way exhaustive." Tree use in pre-colonial Kirinyaga was not as thoroughly documented, but early colonial accounts suggested that the situation was similar (Crawshay, 1902; Embu District, n.d.; Orde-Brown, 1925). This utilitarian knowledge of trees and forest resources derived from close interaction with, and dependence on, a range of local ecozones. In Kirinyaga, people utilized to varying extents the gradient of ecozones running down the southern slopes of Mount Kenya: afro-alpine moorland, bamboo and moist mountain forest, permanently settled farmland, and sub-humid woodland and grassland (see Kenya Land Commission, 1934, pp. 95-99, 249-258, 551-558; Stigand, 1913; Castro, 1983, 1988). An elderly informant from Nduine sublocation, in Ndia Division, readily illustrated this point during an interview by bringing out his bow, quiver, and arrows. 3 He had made the bow from a hardwood that grew in the warm Mwea Plains, the quiver was bamboo collected in the Mount Kenya Forest, and the arrows were prepared from sorghum stocks and "local trees." Regional trade relations allowed people access to forest products from distant areas (Routledge and Routledge, 1910; Ambler, 1988). Learning about trees was embedded in the enculturation process, with males and females becoming knowledgeable about different aspects based on the social division of labor (see Leakey, 1977a; Riley and Brokensha, 1988a). For example, boys grew familiar with the property of many trees while helping herd livestock or clear fields, while girls acquired knowledge of flora as they assisted in collecting forest products or preparing food. Informants in contemporary Kirinyaga emphasized that elders were 3Bows and arrows are rarely seen in contemporaryKirinyaga(exceptamong small children)--a

reflection of culture change and ordinances banning their display and use.

Indigenous KikuyuAgroforestry

5

significant repositories of information about vegetation, which they passed on to the young. Beekeepers and woodworkers also had specialized knowledge of trees (Leakey, 19774). The usefulness of trees conflicted with the need for agricultural land. The long process of settlement in central Kenya by Bantu-speaking peoples was accompanied by extensive clearing of primeval forest (Routledge and Routledge, 1910; Muriuki, 1974). 4 Over the course of many centuries, iron axes, cultivating knives, and fire gradually converged woodland into fields and pastures. Europeans who ventured into thickly inhabited Kikuyu country, including Kirinyaga, during the late 1800s and early 1900s commented on the lack of tree cover (see von Hohnel, 1968; Crawshay, 1902; Dickson, 1903; Hutchins, 1907, Cranworth, 1912; Orde-Brown, 1925). Gedge (1892, p. 526), who traveled through southern Mount Kenya in 1891, noted that "extensive clearances" caused the forest to be "some distance" from settlements and fields. The Routledges (1910, p. 7) observed that Nyeri appeared to be "one huge garden." Stigand (1913, p. 235) described Kikuyuland as a "roiling, almost treeless, cultivated country." Not only was land cleared for farms, but fires set to improve grazing grounds sometimes spread into nearby woodland (Maher, 1938). Deforestation has significant cultural dimensions. Oral traditions in Kirinyaga closely connected land clearing with the emergence of a distinct sociocultural identity. For example, another Ndia elder from Nduine stated that, "Ndemi ["the cutters"] was the first known generation [moiety] in Kirinyaga." He added, "They were the first Kikuyus to clear the bush, and they gave birth to Mathathi, the second generation. ''5 Lambert (1950) recorded that Ndia and Gichugu were the names of brothers who first cleared land in the district. Folk histories about the primordial colonization of southern Mount Kenya emphasized the role of forest clearance in establishing land tenure rights (Routledge and Routledge, 1910; Maher, 1938; Dundas, 1915; Mwaniki, 1974). Thus, the Ndia and Gichugu Kikuyus' sense of themselves was closely related to their forest environment and their transformation of it. It has been argued that certain indigenous cultures have a world view or ideology that causes them to mismanage natural resources, including forests. The Tzotzil of highland Chiapas, Mexico, for example, associated forests with dangerous deities, demons, and supernatural forces (Gossen, 1974). According to Collier (1975), this ideology strongly contributes to a cycle of Tzotzil land use that resulted in severe deforestation and soil 4Bantu-speaking populations were probably present in central Kenya by the eleventh century (Muriuki, 1974; Phillipson, 1985). ~Nderni and Mathathi are interchanged in other accounts (Muriuki, 1974).

6

Castro

erosion. 6 The Kikuyu customarily viewed forests and trees as inhabited by spirits and other supernatural forces (Hobley, 1967; Leakey, 1977a). 7 They also regarded woodland as a habitat for dangerous animals and a hiding place for strange or threatening people. 8 It has sometimes been suggested that traditional Kikuyu culture predisposed or caused them to remove tree cover. An interesting example is from a World Bank (1980, p. 23) report on "Sociological Aspects of Forestry Project Design," which states, "Attitudes toward forests sometimes take on cultural shades . . . . In Kenya, the Masai practice range management, while the Kikuyu cut down trees to grow maize." There is substantial evidence to suggest that economic and demographic forces, rather than ideology or "cultural shades," propelled deforestation in precolonial central Kenya (Muriuki, 1974; Leakey, 1977a). The Kikuyu attempted to reconcile these competing pressures to retain and to remove tree cover. Restraints on cutting trees were included in customary tenure rights and land use practices, and these were reinforced by cultural beliefs regarding the nature of trees. It needs to be emphasized that their goal was not to halt deforestation, but to mitigate its impact. The application and effectiveness of such practices varied over time.

TENURE AND THE MANAGEMENT OF TREES The traditional landholding system exercised much influence over the management of trees. Customary land tenure in Kirinyaga can be classified as a communal-property rights regime (Feeney et al., 1990). 9 Primordial land rights were vested in descent groups--the patrilineal clan (muhiriga) and the more closely-defined lineage (mbari). The extended family (nyumba) and household (rnucii) had usufruct rights to land they occupied. Access and use rights were also defined within the context of local territorial or "neighborhood" groups. Thus, rights to trees were enmeshed in a web of social obligations involving kinsfolk and neighbors. Rights of first clearance were an important means of establishing claim to land in the southern Mount Kenya area. But trees did not have to be cut down to be involved in the assertion of land ownership. On the contrary, the planting or protection of trees were means of establishing 6See Wasserstrom (1978) for a critique of Collier's view. 7Similar accounts have been recorded in Meru (Fadiman, 1982). 8A few informants recalled that Dorobo (Okiek) sometimes hunted in the southern Mount Kenya Forest. 9Key works on the definition and nature of common-property resources include Hardin (1968), National Research Council (1986), McCay and Acheson (1987), Blaikic and Brookfield (1987), Berkes (1989), and Bromley and Cernea (1989).

Indigenous Kikuyu Agroforestry

7

rights to land. Folk histories of early Kirinyaga often included accounts of trees being used as markers for dividing land between and within clans. Mwaniki (1974, p. 297), for example, recorded that Ndemi generation "demarcated" boundaries using trees such as Kigelia africana and Markhamia hildebrandtii. During fieldwork, several informants recalled that descent groups sometimes used particular species to divide the landholdings of individual homesteads. If an already standing tree was not conveniently located, people used transplants or cuttings to set the boundary. An elder from Nduine sublocation, Inoi, stated: Each mbari planted its own type of tree [to mark boundaries]. For example, they would use mukoigo [Bridelia micanthra] and muu [Markhamia hildebrandtii], and others would use muringa [Cordia abyssiniea], muumbo [Ficus wakefieldii] and magumo [Ficus natalensis]. My mbari used muu, rnukoigo, and rnuringa.

All of the aforementioned species were commonly grown in Kirinyaga from cuttings or transplants (see Table I). By the late 1800s, communal controls had been placed on the clearing of forest by descent and neighborhood groups. Permission was supposed to be obtained from local elders and rightholders before harvesting trees or farming along the forest frontier on the slopes of Mount Kenya (Castro, 1988). In a memorandum prepared for the Kenya Land Commission in 1932, Lambert, the district commissioner of Embu, reported about traditional access to the forest: Natives from the lower areas who wished to cultivate in the higher land had to get the permission of the individual owners of pieces of land already demarcated and recognized as the individual's ngamba [plots]; they could not merely choose any convenient sites higher up; the upper boundaries of definitely owned ngambas were the upper limits of cultivation (Embu District Archives, 1939; Embu District Commissioner to the Kikuyu Province Commissioner, 9 December 1932, p. 1). t~

He added that cutting trees above the line of cultivation was prohibited by the local community. Small patches of hilltop and riverine woodland such as Njukiine Forest were also set aside by kin and neighborhood groups as "timber reserves" or woodlots (Castro, 1991b). However, people generally regarded such woodland as land banks, to be opened to cultivation as needed. Descent groups exercised rights of encumbrance over certain trees located on farmland. Two highly valued multi-purpose species, Cordia abyssinica (muringa) and Pygeurn africanurn (mweria), were always considered "clan property." Lineage elders also usually exercised control over any large, locally-valued timber tree on its territory. People, even the household l~ account contradicted the testimony of Orde-Brown, who served in Embu during the early years of colonial rule. Research in the late 1920s by Hopkins (n.d.) on landholding supported Lambert's findings.

Castro Table I. Common Indigenous Agroforestry Trees in Ndia and Gichugu a

Bridelia micanthra, mukoigo. Poles, timber, and firewood (propagation by cuttings, pollarding, drought resistant). Commiphora zimmermanni, mukungugu. Vine props, poles, utensils (cuttings, intercropping). Cordia abyssinica, muringa. Beehives, stools, mortars, well covers, and building timber (cuttings, volunteers, pruning, intercropping). Croton macrostachyus, mutundu. Poles, medicinal uses, and boundary markers (coppicing). Croton megalocarpus, rnukinduri. Poles, boundary marker, cattle shade (profusely seeding, fast growing). Erythrina abyssinica, mubuti or muhuti. Living fence, medicinal uses, weather indicator (cuttings, pruning, and some intercropping). Ficus natalensis, mugumo. Ceremonial and medicinal uses (cuttings). Kigelia africana, rnuratina. Fruit used as a fermenting agent (intercropping). Markhamia hildebrandtii and Markhamia platycalyx, muu. Poles, building timber, and firewood (cloning, volunteers, pollarding, intercropping). Pygeum afi'icanum, rnweria. Mortars, pestles, poles, building timber, and cattle enclosures (coppicing, pruning). Ricinus cornmunis, mubariki. Castor seed and oil (profuse seeds, fast growing). ~ table is meant to be illustrative rather than comprehensive. It lists in alphabetical order indigenous agroforestry trees traditionally utilized in Ndia and Gichugu Divisions, Kirinyaga District. Botanical and vernacular terms for the species are given, followed by major traditional uses and, in parentheses, properties of the tree which fostered its incorporation into local production systems. Source: Castro, 1983; Leakey, 1977c; Riley and Brokensha, 1988b.

t h a t p o s s e s s e d usufruct rights to the p a r t i c u l a r parcel, h a d to s e e k t h e a p proval o f mbari e l d e r s b e f o r e felling such trees. I n f o r m a n t s a g r e e d t h a t a p a y m e n t Of a g o a t to the e l d e r s was e x p e c t e d . Newlyweds who n e e d e d building t i m b e r w e r e n o t subject to this r e q u i r e m e n t . E x t e n d e d families a n d h o u s e h o l d s w e r e r e c o g n i z e d as having exclusive rights to trees, especially o n e s within e n c l o s e d c o m p o u n d s a n d cultiv a t e d fields. Exclusivity was especially a c k n o w l e d g e d if the t r e e s h a d b e e n pollarded, coppiced, o r deliberately propagated. Rights o f e n c u m b r a n c e a n d

Indigenous Kikuyu Agroforestry

9

commonality obliged homesteads to share certain forest products with their kinfolk and neighbors. Indeed, a Kikuyu term for neighborhood, mwaki, denoted a group of households who assisted each other with embers to relight fires (Middleton and Kershaw, 1965). In many cases, kinfolk or friends asked permission to cut timber as a formality, an act of etiquette; people were not required "to beg" or to pay. Customary tenure rights also allowed neighbors to collect deadwood, shrubs, herbs, most wild fruits, and other forest products as long as the trees, crops, and other property were not damaged or disturbed. A rightholder (mwene githaka) could limit felling on his land. Such restrictions were placed on timber or other valued trees. Curses were sometimes placed on such trees to protect them from encroachment (Nyeri District, 1914). This was common with Kigelia africana (muratina), whose sausage-shaped fruits were valued as a fermenting agent for brewing. Even in the early 1980s many people in Kirinyaga believed that evil would befall anyone who took fruit from another's muratina. Rightholders sometimes placed curses on entire stands of trees (Leakey, 1977a; Nyeri District, 1914). The Eithaga clan, whose members in Ndia were reported renowned for their potent curses, "were believed to have the power of protecting forests" (Hobley, 1967, p. 183; Orde-Brown, 1925). Placing a curse on woodland was sometimes considered antisocial, however, and neighbors and kinsmen were known to pressure a rightholder into removing it (Dundas, 1915).

FARM-LEVEL STRATEGIES TO RETAIN TREE COVER Individual households used several strategies to retain tree cover on their particular holdings. Families sometimes set aside small plots of heavily wooded land to supply building material, fuel, and other items (see Routledge and Routledge, 1910; Dundas, 1915; Kenyatta, 1938; Hobley, 1967; Leakey, 1977a). Present-day practices suggested that these small stands of trees were situated in a variety of settings, including along steep ridges and at the head of narrow valleys and ravines. Such copses were protected during clearing, with trees cut as need arose. Probably more commonplace was the protection of individual or small groups of trees during clearing. When preparing land or a fallow plot for farming, certain trees would be spared the axe. Before setting fire to the plot, brush and trash would be removed from around the trees to protect them from the flames. This practice was motivated by spiritual and utilitarian concerns.

10

Castro

Selective cleating was influenced by traditional religious beliefs. The Kikuyus believed that trees possessed spirits capable of intervening in human affairs (Leakey, 1977c). A tree spirit was not upset when its abode was cut, as long as it had another tree nearby to go to. When clearing land, people were supposed to leave "a large and conspicuous tree" at intervals to absorb the spirits from the ones cut down (Hobley, 1967, pp. 31-32; Leakey, 1977c, pp. 1117-1118). Such a tree was called a murema kiriti ("one which resists the cutting of the forest"), and it was not supposed to be cut or allowed to fall without a ceremony transferring the spirits to another tree. The angry spirit would kill a person who reportedly failed to perform the ceremony within a short time span (Leakey, 1977c). Young people who used the wood from such a tree for fuel would "become ill or die," though senior male elders and very old women could burn it without danger (Hobley, 1967, p. 32). Sacred groves were also protected by supernatural sanctions (Castro, 1990). Trees were usually left standing for a more mundane reason: to maintain a convenient supply of forest products for family consumption. One of the most visible reasons why Kikuyus spared large trees was to hang beehives on them (Routledge and Routledge, 1910). Crawshay (1902, p. 33), an early European visitor to Kirinyaga, wrote, "Many a fine tree owes its existence to the bees." Ficus natalensis and Cordia abyssinica were favorite "bee trees," but any large tree with sprawling limbs was generally used to hang hives. Honey was especially valued for brewing purposes. Besides being a popular drink, honey beer was required for a number of ceremonies (see Leakey, 1977a). Certain tree species were preferred by the Kikuyu because they directly supplied multiple or special products. Such trees often possessed botanical properties--would seed profusely, were easily transplanted from volunteer seedlings or propagated from cuttings, had the capability of being pollarded or coppiced, or the ability to be intercropped--all of which facilitated their retention. Table I lists commonplace indigenous agroforestry trees in Ndia and Gichugu, including major uses and relevant botanical properties. The list is not exhaustive; other species were incorporated to varying extents into local production systems. Because of local differences in rainfall, soils, slope, and other ecological factors, there was no single pattern of agroforestry, but rather there were numerous variations on a similar theme of trying to fit trees into household production systems. Decision-making about trees was influenced by a range of factors, including the availability of labor and land, the crop mix, and individual ethnobotanical knowledge. What is important to emphasize is that Kikuyu households were utilitarian in their management and use of trees.

Indigenous Kikuyu Agroforestry

11

The present-day incidence of the species in Table I should not be used as guides to their past distribution. Afforestation campaigns, agrarian intensification, and other changes in land use management during the twentieth century have influenced contemporary patterns (Castro, 1983, 1990, 1991b). For example, Comrniphora zimmermanni (mukungugu) was said to be "the commonest" farmland tree in Kikuyuland at the turn of the century (Stigand, 1913, p. 239). It was used as a boundary marker and a living prop for yam and sweet potato vines, plus its leaves furnished fodder for goats. The tree was easily propagated from cuttings, a significant factor in its distribution. Although still widely grown, this tree has been surpassed in local popularity and spatial distribution by exotics, including fast-growing and multipurpose Grevillea robusta (rnukema or mubariti) and various eucalyptus, plus indigenous species such as Markhamia hildebrandtii (see Castro, 1983; Brokensha et al., 1983). Several informants stated that Croton megalocarpm" (mukinduri) was grown more widely today than in the past. Trees that could be intercropped with food crops were especially valued. The intercropping of yams and sweet potatoes with Cornmiphora zimmermanni has already been mentioned. Markhamia hildebrandtii was often heavily pruned to obtain poles and firewood; then maize, beans, and other crops were planted around it. Maize was sometimes grown near the deciduous Cordia abyssinica and rnuhuti. Grain and pulses were also sown under rnuratina trees. Croton megatocarpus and Albizzia sp. (mukurue) commonly furnished shade for cattle. Several early colonial writers claimed that the Kikuyu practiced a particularly destructive form of shifting cultivation (Eliot, 1905; Hutchins, 1909; Cranworth, 1912; Baker, 1931). l~ This misconception greatly underestimated the diversity, stability, sustainability, and resilience of their indigenous agriculture (Allan, 1965). Kirinyaga households practiced a number of farming techniques, including bush fallowing of plots for varying periods. People took land out of farm production because of soil exhaustion and the growth of weeds (Embu District Archives, 1939). Such fields were brought back into cultivation long before they could revert to secondary forest growth. In some areas garden plots were under almost continuous production (Kenya Land Commission, 1934, pp. 551-558). Several informants stated that the only shifting cultivation they knew about was the taungya ("shamba" or "squatter") system of reforestation used by the Forest Department in the Mount Kenya Reserve. In general, fallowing allowed trees and bush to regenerate to some extent, but there was no ~lThe claim that cultivators along the southern edge of the Mount Kenya Forest practiced a destructive form of shifting cultivation was repeated by Orde-Brown and Lane at the Kenya Land Commission in 1932 (see Kenya Land Commission, 1932, pp. 6-14, 50-52).

Castro

12

long-term rotation involving farmland and trees. Trees and shrubs that grew in fallowing fields were a source of forest products such as firewood and fodder. Economic inequalities of the precolonial era were sometimes manifested or reflected in the use and the physical distribution of trees. The Routledges (1910, p. 433) recorded that people who were too "poor" to obtain mutton fat rubbed castor oil from the seed of Ricinus communis (mubariki) on their garments. An informant from Baricho, Mutira, Ndia, stated that having Cordia abyssinica near a homestead was an indicator of a prosperous family. "People liked muringa [Cordia abyssinica]. Where you had one, the family was well to do, having many cows and so on." Another informant suggested that Kigelia africana was also connected with prosperous livestock keepers, since the tree often grew on the sites of old cattle enclosures. The goal of indigenous agroforestry practices was to maintain tree cover, not to extend it. People did not want to halt deforestation, but mitigate its impact. They might protect a grove or tree by restricting felling or allow farmland to fallow for an extended period. The primary aim, however, was to keep a supply of useful trees near homesteads. Indigenous land use strategies--whether at the communal or household level--attempted to integrate trees into local farming and herding systems. Even the practice of fallowing was carried out with the objective of eventually returning the plot to cultivation. Although certain trees were sometimes deliberately propagated from cuttings or transplants, the goal was always limited and immediate, such as marking boundaries, providing fencing, or furnishing props. Indigenous agroforestry practices constituted a form of practical conservation that was highly selective in terms of species. While Cordia abyssinica, Bridelia micanthra, and similar multi-purpose trees might be protected, other species decreased in number. Some historical evidence and oral traditions suggested that species such as Heywood lucens (mutaigoko),

Drypetes gerrardii (munyenye), Ekebergia capensis (mununga), Blighia unijugata (muikuni), and Milletia oblata (mwanga) became increasingly scarce as settlements and fields expanded in the late nineteenth and twentieth centuries (see Hutchins, 1907; Embu District, n.d.; Maher, 1938).

DEFORESTATION IN THE LATE PRE-COLONIAL ERA Regional and the long-distance caravan trade expanded throughout central Kenya in the late 1800s (Saberwal, 1967; Marris and Somerset, 1971; Moris, 1973; Muriuki, 1974; Mwaniki, 1974; Beachey, 1967, 1976;

Indigenous Kikuyu Agroforestry

13

Ambler, 1988). Caravans from the coast, often with several hundred porters, offered a substantial, if erratic and sometimes troublesome, market for Kikuyu foodstuffs (Crawshay, 1902; Dickson, 1903; Leakey, 1977a). By the 1860s, people in Kiambu and other areas deliberately extended their cultivated area to meet the rising commercial demand (Kershaw, 1972; also see Miracle, 1974; Ambler, 1988). An estimated "one-third" of the farmland in southern Kikuyuland eventually became directed toward the caravan trade (Kershaw, cited in Molnos, 1972, p. 216). The situation was similar in Kirinyaga, where Gedge (1892, p, 526) observed in 1891 that, "The plantations appeared like market-gardens, so well were they kept." In fact, they often were "market-gardens," providing an "unlimited food supply" to Arab and Swahili caravans (Arkell-Hardwick, 1903, pp. 51-60). Dickson (1903, p. 39) reported in 1903 that Ndia and environs were "a great centre for the native caravans" (also see Saberwal, 1967, 1970; Moris, 1973; Mwaniki, 1974; Ambler, 1988). Indeed, Swahili traders established a permanent outpost at Kagio in southern Ndia (Stigand, 1913). 12 Farm surpluses from Kirinyaga were also exchanged in local and regional trade circuits, particularly with the Kamba (Moris, 1973; Ambler, 1988). Expanding market and population pressures overwhelmed tree maintenance strategies in some areas during the late 1800s. Acute tree scarcity reportedly emerged in places heavily trafficked by caravans, such as near Ngongo Bagas on the southern edge of Kikuyuland and Maruka in Murang'a (von Hohnel, 1968; Lugard, 1893; Dickson, 1903). From Nyeri, the Routledges (1910, pp. 6-7) observed, "In the heart of Kikuyu, except for a sacred grove here and there, scarcely a tree remains. As far as the eye can reach, in all directions, spreads one huge garden." Of course such gardens still included heavily pollarded and coppiced trees. The lack of tree cover in densely settled sections of southern Mount Kenya caused the 1891 Dundas expedition to have difficulties procuring firewood (Gedge, 1892). In general, though, acute tree scarcity was not widespread in Ndia and Gichugu, but was a highly localized phenomenon. Many places still possessed a smattering of tree cover. While quality timber trees might become locally scarce, fuel and other products could be obtained from fallowing fields and "bush" (Arkell-Hardwick, 1903, p. 59; Crawshay, 1902). The decrease in tree cover was not necessarily synonymous with increased land degradation. Instead, it often simply meant a shift in 12As late in 1911, there were 62 Swahili trading plots at Kagio (Fort Hall District, 1911). Officials expelled them in 1913, accusing the Swahili of "encouraging women to run away" from their husbands (Nyeri District, 1913).

14

Castro

cropping patterns, with grains, pulses, tubers, and other foodstuffs, or grassland for livestock, replacing trees. Techniques such as intercropping, nutrient recycling by burning or burying crop residues and other plants, and no-till farming using digging sticks and cultivating knives, made it possible to expand farmland without undermining soil management (Leakey, 1977a). 13 Still, tree scarcity often caused hardship, especially for women, who were responsible for collecting wood. The Routledges (1910, p. 122) recorded the considerable time and effort expended by women in densely populated, tree-scarce parts of Nyeri to collect firewood for their domestic needs. Women often left at 5:00 a.m. for a 4-hour walk to the forest, where they collected loads weighing up to 30 kg, then returned home in the evening. A robust trade in foodstuffs and forest products developed in Nyeri between the grain-abundant but wood-scarce region and the woodabundant but often food-short villages on the Nyandarua Forest frontier (Routledge and Routledge, 1910). The sale of firewood to caravans by women and children was also common in tree-scarce areas (von Hohnel, 1968). Several elders in contemporary Kirinyaga recalled a trade between villages near the Mount Kenya forest and the more densely settled areas. A man from Kanyei sublocation, Mutira, Ndia, stated, Forest region people used to come here to trade firewood. We exchanged maize and beans for wood . . . . They [the firewood sellers] came to the home areas. It was done by women who brought head loads. 14

Studies of the caravan trade often conclude that men, especially wealthy and powerful brokers, were its major local beneficiaries (Cavicchi, 1977; Ambler, 1988). In contrast, the foregoing analysis suggests that women in tree-scarce areas bore some of its costs in terms of increased difficulty in procuring firewood. The scarcity of trees in the late 1800s shared several characteristics with the "firewood crisis" in contemporary central Kenya. x5 In both cases, long-term demographic and market pressures had stimulated land clearing, and in a number of places acute tree scarcity had emerged. Yet, there are significant differences between the two eras. The present-day woodfuel crisis experienced by many households often derived from socioeconomic factors unknown in the precolonial era. In particular, land privatization 13Comprehensive overviews of indigenous sustainable farming practices in Africa are provided in Okigbo (1990). 14Colonial Paramount Chief Githai wa Giti of Ndia reportedly took advantage of the scarcity of firewood near Kagio market by setting up "a wood-pile, and ]exacting] compensation for it, thus making a good living out of passersby" (Madeira, 1909, p. 62). 15Key works on the fuelwood crisis in central Kenya include Haugerud (1984); Brokensha et al. (1983).

Indigenous Kikuyu Agroforestry

15

initiated in the 1950s altered local socioeconomic relations governing access to trees. Land "reform" ended the communal property regime, including rights of commonality to forest products. It also created an outright landless and generally impoverished population who often lacked the income to purchase alternative supplies (Castro, 1983; Brokensha et al., 1983; Riley and Brokensha, 1988a). 16

CONCLUSION This article has emphasized the importance of understanding indigenous agroforestry practices in a historical perspective. People in precolonial Kirinyaga developed a range of communal and farm-level strategies in order to incorporate trees into local production and sociocultural systems. These strategies included the communal protection of sacred groves and locally-valued timber species, the setting aside by households of small copses as woodlots, selective clearing, fallowing, and pollarding. As in perhaps most traditional resource management systems (see Berkes, 1987), this practical conservation was not based on "scientific reasoning" or some indigenous environmentalism regarding trees. Rather, a complex combination of factorswcultural beliefs and values, local social relations, economic concerns, population pressures--governed the use and management of trees. Commercial and demographic expansion in the late 1800s led to accelerated land clearing and localized tree scarcity. However, the decrease in tree cover was not necessarily synonymous with increased land degradation.

ACKNOWLEDGMENTS Research in Kirinyaga, Kenya, was supported by the National Science Foundation, the Intercultural Studies Foundation, and the University of California, Santa Barbara. The author also wishes to thank the Kenya National Archives, Nairobi, Kenya, the Embu District Archives in Embu, Kenya, and Special Collections, Bird Research Library, Syracuse University, Syracuse, New York~ The views expressed in this paper are those of the author and do not reflect the views of the supporting agencies. 16For an insightful analysis of woodfuel scarcity, especially the difference between physical and socioeconomic scarcities, see Dewees (1989).

16

Castro

REFERENCES Allan, W. (1965). The African Husbandman. Oliver and Boyd, London. Ambler, C. H. (1988). Kenyan Communities in the Age oflmperialism. Yale University Press, New Haven. Arkell-Hardwick, A. (1903). An lvory Trader in North Kenia. Longman, Green, London. Baker, R. St. B. (1931). Men of the Trees. Dial Press, New York. Beachey, R. (1967). The East African ivory trade in the nineteenth century. Journal of African History 8: 269-290. Beachey, R. (1976). The Slave Trade of Eastern Africa. Barnes and Noble, New York. Berkes, F. (1987). Common-property resource management and Cree Indian fisheries in subarctic Canada. In McCay, B. J., and Acheson, J. M. (eds.), The Question of the Commons. University of Arizona Press, Tucson, pp. 66-91. Berkes, F. ted.) (1989). Common Property Resources. Belhaven, London. Blaikie, P., and Brookfield, H. (eds.) (1987). Land Degradation and Soc&ty. Methuen, London. Brokensha, D. W., Riley, B. W., and Castro, A. P. (1983). Fuelwood Use in Rural Kenya. Institute for Development Anthropology, Binghamton. Bromley, D. W., and Cernea, M. M. (1989). The Management of Common Property Natural Resources. Discussion Paper, World Bank, Washington, D.C. Castro, A. P. (1983). Household Energy Use and Tree Planting in Kirinyaga. Institute of Development Studies Working Paper, University of Nairobi, Nairobi. Castro, A. P. (1988). Southern Mount Kenya and colonial forest conflicts. In Richards, J. F., and Tucker, R. P. (eds.), World Deforestation in the Twentieth Century. Duke University Press, Durham, pp. 33-55, 266-271. Castro, A. P. (1990). Sacred groves and social change in Kirinyaga, Kenya. In Chaiken, M. S., and Fleuret, A. K. (eds.), Social Change and Applied Anthropology. Westview Press, Boulder, pp. 277-289. Castro, A. P. (1991a). Social forestry: A cross-cultural analysis. In Wali, M. (ed.), Ecosystem Rehabilitation (Vol. 1). SPB Academic Publishing, The Hague, pp. 63-78. Castro, A. P. (1991b). Njukiine Forest: Transformation of a common-property resource. Forest and Conservation History 35: 160-168. Castro, A. P. (1991c). The southern Mount Kenya Forest since independence: A social analysis of resource competition. World Development 19: 1695-1704. Cavicchi, E. (1977). Problems of Change in Kikuyu Society. EMI, Bologna. Collier, G. (1975). Fields of the Tzotzil. University of Texas Press, Austin. Cook, C. C., and Grut, M. (1989). Agroforestry in Sub-Saharan Africa. Technical Paper No. 112, World Bank, Washington, D.C. Cranworth, L. (1912). A Colony in the Making. Macmillan, London. Crawshay, R. (1902). Kikuyu: Notes on the country, people, fauna and flora. Geographical Journal, 20" 24-49. Davison, J. (1989). Voices from Mutira. Rienner, Boulder. Dewees, P. (1989). The woodfuel crisis reconsidered: Observations on the dynamics of abundance and scarcity. World Development 17:1159-1172. Dickson, B. (1903). The eastern borderlands of Kikuyu. Geographical Journal 22: 36-39. Dundas, C. (1915). The organization and laws of some Bantu tribes in East Africa. Journal of the Royal Anthropological Society 45: 234-306. Eliot, C. (1905). The East African Protectorate. Arnold, London. Embu District (n.d.). Kikuyu and Embu trees. In Embu District Political Record Book (PC/CP/I/5/1). Kenya National Archives, Nairobi. Embu District Archives (1939). File LND 16/17/1, Land Commission, Embu. Fadiman, J. A. (1982). An Oral History of Tribal Warfare. Ohio University Press, Athens. Falconer, J., and Arnold, J. E. M. (1989). Household Food Security and Forestry. Food and Agriculture Organization of the United Nations, Rome. FAO (Food and Agriculture Organization of the United Nations) (1985). Tree Growing by Rural People. FAO, Rome.

Indigenous Kikuyu Agroforestry

17

Feeney, D., Berkes, F., McCay, B. J., and Acheson, J. M. (1990). The tragedy of the commons: Twenty-two years later. Human Ecology 18: 1-19. Fisher, J. (1952). Reports on the Kikuyu. Unpublished manuscript. Fort Hall District (1911). Trade centres at Kagio. In Fort Hall Political Record Book (PC/CP. 1/7/1). Kenya National Archives, Nairobi. Gedge, E. (1892). A recent exploration of the River Tana to Mount Kenya. Geographical Society Proceedings 14: 513-533. Glazier, J. (1985). Land and the Uses of Tradition Among the Mbeere of Kenya. University Press of America, Lanham. Gossen, G. (1974). Chamulas in the World of the Sun. Harvard University Press, Cambridge. Gregersen, H., Draper, S., and Elz, D. (1989). People and Trees. The World Bank, Economic Development Institute, Washington, D.C. Guggenheim, S., and Spears, J. (1991). Sociological and environmental dimensions of social forestry projects. In Cernea, M. M. (ed.), Putting People First (2rid Ed.). Oxford University Press, New York. Hardin, G. (1968). The tragedy of the commons. Science 162: 1243-1248. Haugerud, A. (1984). Economy, ecology, and the unequal impact of woodfuel scarcity in Embu, Kenya. In Barnes, C., Ensminger, J., and O'Keefe, P. (eds.), Wood, Energy and Households. Beijer Institute, Stockholm. Hobley, C. (1967). Bantu Beliefs and Magic (2nd ed.). Cass, London (2nd Ed. originally published 1922, 1937). Hopkins, J. G. (n.d.). Embu land tenure system. In Embu Political Records (EBU/45A. V.), Kenya National Archives, Nairobi. Hutchins, D. (1907). Report on the Forest of Kenya. HMSO, London. Hutchins, D. (1909). Report on the Forests of British East Africa. HMSO, London. Kenya Land Commission (1932). Evidence (Vol. 1). England. Kenya Land Commission (1934). Kenya Land Commission Evidence (Vol. 1). HMSO, London. Kenyatta, J. (1938). Facing Mount Kenya. Secker and Warburg, London. Kerkhof, P. (1990). Agroforestry h~ Africa. Panos Institute, London. Kershaw, G. (1972). The Land is the People. Ph.D. dissertation, University of Chicago. Kirinyaga District (1984). Development Plan 1984-88. Kenya Ministry of Finance and Planning, Nairobi. Lambert, H. E. (1950). The Systems of Land Tenure in the Kikuyu Land Unit. Communications, No. 22, School of African Studies, Cape Town. Lambert, H. E. (1956). Kikuyu Social and Politkal Institutions. Oxford University Press, London. Leakey, L. S. B. (1977a). The Southern Kikuyu before 1903 (Vol. 1). Academic Press, New York. Leakey, L. S. B. (1977b). The Southern Kikuyu before 1903 (Vol. 2). Academic Press, New York. Leakey, L. S. B. (1977c). The Southern Kikuyu before 1903 (Vol. 3). Academic Press, New York. Lugard, F. (1893). The Rise of Our East Afi'icaEmpire (Vol. 1). Blackwood, Edinburgh. Lundgren, B. (1982). Introduction. Agroforeslry Systems 1: 3-6. Madeira, P. (1909). Hunting in British East Africa. Lippincott, Philadelphia. Maher, C. (1938). Soil Erosion and Land Utilization in the Embu Reserve, Part 1. Department of Agriculture, Kenya Colony, Nairobi. Marris, P., and Somerset, A. (1971). African Businessmen. East African Publishing House, Nairobi. McCay, B. J., and Acheson, J. M. (eds.) (1987). The Question of the Commons. University of Arizona Press, Tucson. Meinertzhagen, R. (1957). Kenya Dialy. Oliver and Boyd, London. Middleton, J., and Kershaw, G. (1965). The Kikuyu and Kamba. International African Institute, London. Miracle, M. (1974). Economic change among the Kikuyu, 1895-1905. University of Nairobi, Institute of Development Studies Working Paper, Nairobi.

18

Castro

Molnos, A. (1972). Cultural Source Materials for Population Planning in East Africa (Vol. 2). East African Publishing House, Nairobi. Moris, J. (1973). The Mwea environment. In Chambers, R., and Moris, J. (eds.), Mwea. Weltforum Verlag, Munich, pp. 16-63. Mungeam, G. (1966). British Rule in Kenya, 1895-1912. Clarendon, Oxford. Munro, J. (1975). Colonial Rule and the Kamba. Clarendon, Oxford. Muriuki, G. (1974). A History of the Kikuyu, 1500-1900. Oxford University Press, Nairobi. Mwaniki, H. S. K. (1974). Embu Historical Text. East African Literature Bureau, Nairobi. National Research Council (1986). Proceedings of the Conference on Common Property Resource Management. National Academy Press, Washington, D.C. Niamir, M: (1990). Herders' Decision-Making in Natural Resources Management in Arid and Semi-Arid Africa. Food and Agriculture Organization of the United Nations, Community Forestry Note No. 4, Rome. Nyeri District (1913). History of District. In Nyeri District Political Records (NYI/9), Kenya National Archives, Nairobi. Nyeri District (1914). Handing Over Report. Kenya National Archives, Nairobi. Okigbo, B. N. (1990). Sustainable agricultural systems in tropical Africa. In Edwards, C. A., Rattan, L., Madden, P., Miller, R. H., and House, G. (eds.), Sustainable Agricultural Systems. Soil and Water Conservation Society, Ankeny, pp. 323-352. Olofson, H. (1983). Indigenous agroforestry systems. Philippine Quarterly of Culture and Society 11: 149-174. Orde-Brown, G. (1925). The Vanishing Tribes of Kenya. Seeley and Service, London. Phillipson, D. (1985). African Archaeology. Cambridge University Press, Cambridge. Riley, B. W., and Brokensha, D. W. (1988a). The Mbeere in Kenya (Vol. 1). University Press of America, Lanham. Riley, B. W., and Brokensha, D. W. (1988b). The Mbeere in Kenya (Vol. 2). University Press of America, Lanham. Rogers, P. (1979). The British and the Kikuyu 1890-1905: A reassessment. Journal of African History 20: 255-269. Routledge, W., and Routledge, K. (1910). With a Prehistoric People. Arnold, London. Saberwal, S. (1967). Historical notes on the Embu of central Kenya. Journal of African History 8: 29-38. Saberwal, S. (1970). The Traditional Political System of the Embu of Central Kenya. East African Publishing House, Nairobi. Stigand, C. H. (1913). The Land of Zh~j. Constable, London. von Hohnel, L. (1968). Discovery of Lakes Rudolf and Stefanie (Vol. 1). Frank Cass, London (originally published 1894). Wasserstrom, R. (1978). Population growth and economic development in Chiapas, 1524-1975. Human Ecology 6: 127-143. World Bank (1980). Sociological Aspects of Forestry Project Design. World Bank, Washington, D.C.