CASE STUDY No.2
BIODIVERSITY MANAGEMENT IN WEST AFRICAN PASTORAL AND AGRO-PASTORAL SYSTEMS

RESPONSIBLE TECHNICAL DIVISION

Animal Production and Health Division Animal Production Service

Irene Hoffmann

AUTHOR

Irene Hoffmann

Chief,

FAO Animal Production Service

[email protected]

ACKNOWLEDGEMENTS

I am grateful to the DG VIII and DG XII of the European Union for financing the research under the projects “Range development and camel studies“,

“Range development in the endangered Sudan Savanna in Sokoto State, Nigeria“, and Development of pastoral and agro-pastoral livelihood systems in West Africa”. I also thank my colleagues from Justus-Liebig-University Giessen, Usman Danfodiyo University and the people in the Zamfara Reserve for their kind co-operation.

DIVERSITY IN NORTHERN NIGERIAN RISK-PRONE SYSTEMS

Northern Nigeria is a risk-prone environment, with high inter-annual variations in rainfall. Farmers and herders have developed strategies, such as mobility or flexible access to natural resources, to cope with climatic risks (Scoones et at., 1995). Traditionally, livestock and cropland agriculture were segregated and practised by different ethnic groups. However, these groups always interacted and exchanged their products (Van Raay, 1975; Mc Intire et al., 1992; Mohammed, 2000; Hoffmann et al., 2001). Today, the systems tend to be more integrated and traditional pastoral groups such as the Fulani are engaged in cropping, whereas traditional farmers such as the Hausa are increasingly keeping livestock.

THE SITE

Studies were conducted in the Zamfara Forest Reserve in northwest Nigeria between 1990 and 2000. The reserve is located between 6°30’ and 7°15’E, and 12°10’ and 13°05’N in the north of Zamfara State, sharing a border with the Niger Republic to the north, Sokoto State to the west and Runka Reserve of Katsina State to the east. Annual rainfall ranges from 500 mm in the north to 850 mm in the south with considerable inter-annual variations. The vegetation is of a northern Sudan savanna type.

Documents from the pre- and post-colonial eras, as well as recent research reports document the historical development of the area. From historical sources, Krieger (1954) concluded that Zamfara existed already in the 14th century. The area was virgin bush, probably inhabited by hunters from Katsina. It can be assumed that the region, which is now the two Forest Reserves, or at least a large part of it, was abandoned during the political and social disruptions following the jihad (holy war in 1804). Subsequently, the depopulated area became covered by a dense secondary savanna vegetation (Keay, 1949; Kueppers, 1998). The fact that dye basins, ruins of old settlements and fortification walls can be found in the reserve underline this assumption. Probably, the transhumant Fulani also avoided the area during this time. Due to its low population density and the formerly dense vegetation, an area of 2 394 km² in Zamfara became a forest reserve in 1916. It was used by the British colonists to secure their supply of tropical timber and to establish a game hunting area. In 1932, the vegetation was described as an open Combretum savanna of varying quality. Near villages and on farmlands most economically valuable tree species were absent.

TABLE 1

Some characteristics of livestock and cropping systems in northern Nigeria

Map of the Zamfara Reserve

The villages which today encompass the four enclaves in the reserve were founded between the end of the last century and the 1940s, after the establishment of the Forest Reserve. The first settlers were hunters of Hausa origin who later cleared the bush for cropland. Both hunting and bush clearing resulted in a decline in the amount of wildlife and opened up the vegetation. As a result, an influx of pastoral herds took place and grazing became more frequent in the 1930s (Hoffmann, 1998). Initially, the villages had a low population of less than 100 persons each. Taking the figures given in the Forestry Ordinance 1916, there were 19 households in Aja, 40 in Tsabre and 63 in Dumburum in 1916. Population estimates indicate that 27250 people lived in the enclave villages in 1994 (ARCA, 1995). The upsurge was due to natural population growth and to immigration after the droughts and famines during the twentieth century.

The reserve, established in 1918, covers today a total area of 2300 km² including four enclave villages (Dumburum, Shamashalle, Tsabre, Aja). About 50 villages are lined up on its western fringe. About 130 000 people live within or around the reserve and utilize its natural resources (ARCA, 1995). The area is populated by Hausa farmers and by Fulani pastoralists. The clear distinction between the livestock and cropping systems is becoming blurred. About 80 percent of the Fulani, sedentary and transhumant herders, are engaged in crop production (Schaefer, 1998).

In this paper, the use and management of biodiversity, both in the livestock and the cropping systems, will each be analyzed.

DIVERSITY IN THE RANGELAND-BASED LIVESTOCK SYSTEM

Most large ruminant livestock is kept by Fulani pastoralists. However, sedentary Hausa farmers are increasing their livestock holdings with small ruminants and some cattle. The reserve is an important rainy season grazing area for transhumant pastoralists as well as for the herds of the sedentary farmers living in the enclaves or the bordering villages. After the grain harvest, most livestock are fed on stubble. Later in the dry season, transhumant pastoral herds leave the region in search of pasture and water. However, about one third of the Fulani are sedentary and stay in the region with their herds throughout the year (Schaefer, 1998). Therefore, the livestock system encompasses transhumant and sedentary animals.

Livestock management on rangeland and cropland

Cattle are the basis of livelihood for pastoralists and agro-pastoralists in the northern parts of West Africa. They are used for milk, manure, meat and draught power, and they serve as savings and insurance. Composition of pastoral herds, particularly cattle, has been analyzed repeatedly. Usually, the sex and age groups of individual herds, directly counted or resulting from interviews by the owner or herder, are recorded (Sutter, 1987; FDLPCS, 1992; Vabi, 1993; Amanor, 1995). Data on livestock numbers are generally hard to obtain in northern Nigeria. Collecting data on herd size by counting or interviews poses some problems. Firstly, pastoralists are reluctant to give information about their livestock numbers due to cultural reasons and fear of taxation (Gefu, 1992; Mohammed and Bello, 1994). Secondly, risk management practices in an ecologically and economically highly variable environment augment the problems of getting accurate data on livestock ownership. Such management practices lead to an overlay of ownership and management patterns, hence, the term 'herd' applies to the unit of management rather than to the ownership. This is specifically due to: the widespread exchange of animals in a social network, the division of herds into management units which might be herded far away from the owner's homestead, the herding of animals of different (absentee) owners in one herd, and the waged herder's reluctance to admit being a "poor" employee who has to herd for others.

The stocking density information for this paper was obtained from transect counts. Livestock density in the Zamfara Reserve was estimated by monthly repeated animal counts (June 1993 to May 1994) along five lines transects from North to South (Schaefer, 1998). Each transect crossed the reserve in east-west direction, thereby crossing transhumance routes at a right angle. Animal counts were then converted into tropical livestock units (TLU) of 250 kg liveweight.

An average of 1.5, 0.6 and 0.3 TLU per hectare (ha^-1) was found on rangeland during the rainy, early and late dry seasons. Over the whole year, an average of 0.84 head of cattle, 0.55 sheep and 0.38 goats was found per ha of range, resulting in a stocking rate of 0.81 TLU ha^-1. This stocking rate exceeds the recommended rate (Boudet, 1991). Cattle represent about 80 percent of livestock biomass. The highest density of cattle was observed in August with 2.3 head ha^-1. This coincides with the peak of rainfall, growth rate of the vegetation, and the feed supply of pastures in quantitative and qualitative terms. The continually decreasing animal density on rangeland during the dry season reflects the decrease in feed and water availability and the subsequent migration of pastoral herders out of the Zamfara Reserve (Figure 1).

FIGURE 1
Monthly stocking density on rangeland and cropland in the Zamfara Reserve (data: Schaefer)

No livestock was found in the cropland areas during the rainy season, when access is forbidden by the village authorities. Cattle numbers are kept at a rate of 1.6 head per ha of cropland from December to March, rapidly declining thereafter (Figure 1). In contrast, small ruminant stocking rate declined gradually from 0.3 to 0.1 animals ha^-1 during the dry season. The high livestock density on cropland leads to considerable nutrient input through manure (Hoffmann et al., 2001). After planting in May/June, no livestock is allowed in the cropland areas.

Stocking density of cattle during the rainy season varies with the distance to the village, although not significantly (Figure 2). At a distance of four km 2.3 heads of cattle were counted per ha. This value decreased by about 30 percent at six and eight km. The small ruminant counts did not show an effect of settlement distance. Possibly, six km represents the limit of village herds' grazing orbit, as was found in farmer interviews by Malami et al. (1998) and Hassan (2000). The slight but insignificant increase in to more than eight km distance from the village might be due to transhumant Fulani herds, who settle at a certain distance from the village and have a five km grazing orbit during the rainy season (Mohammed and Bello, 1994).

Grazing orbits are enlarged during the dry season due to search for feed and water. Fifty-four percent of interviewed pastoralists cover more than five km grazing distance from their settlements, including those transhuming southwards (Mohammed and Bello, 1994). In contrast to the rainy season, when livestock are concentrated on the rangeland, they are more evenly distributed in the landscape during the dry season. After the harvest, village authorities declare the opening of the fields. Therefore, livestock use rangeland and cropland. An average of 1.2 and 0.3 TLU ha^-1 was found on cropland and rangeland during the six months of dry season.

FIGURE 2
Mean stocking density of livestock in two seasons by distance to the villages (data: Schaefer)

Composition of herding units

When asked about their livestock holding in different surveys, the Fulani in the reserve gave figures ranging from 69 to 75 head of cattle, 33 to 43 sheep and 34 to 36 goats (Kyiogwom et al., 1994). Similar cattle, but lower small ruminant holdings were observed in settled Fulani in northern Cameroon (Vabi, 1993). These figures should be considered cautiously, because the Fulani are reluctant to give information about their livestock numbers. However, another approach of indirect estimates of pastoral herd size reveals similar results. The pastoralists told in interviews that three to four young men or children per household are engaged in livestock herding (Kyiogwom et al., 1994). The average herding unit in the Zamfara Reserve consisted of 20.2 cattle. Hence, 70 head of cattle per pastoral household seems to be a realistic estimate. These cattle holdings are above the figures given for agro-pastoral Fulani which vary between 32 in the Udubo Grazing Reserve (Gefu, 1992) and 50 in central Nigeria (Powell and Taylor-Powell, 1984; Powell, 1986). Evidently, sedentary pastoralists keep smaller herds than transhumant or nomadic ones (Gefu, 1992). Ownership of larger herds in the Zamfara Reserve might be due to the presence of nomadic and transhumant pastoralists (70 percent of the Fulani) (Mohammed and Bello, 1994; Schaefer, 1998) and the relative abundance of rangeland resources. Herd size of the Fulani is larger than that of Hausa. Seventy-seven percent of the Hausa farmers keep an average of 13 sheep, and 75 percent of the farmers keep 11 goats. They also own a few cattle, mainly adult males for draught purposes, but only seven percent of the farmers keep more than 10 cattle.

The number of cattle or small ruminants observed per herding unit did not differ between cropland and rangeland. The average cattle number in 1 264 herds was 20.2 (median 13; range 1 to 183). Two-thirds of the herds were comprised of less than 20 head of cattle. Although the observation of herds does not allow direct conclusions on the ownership pattern, this figure can be explained by a high percentage of smaller herds owned by farmers, who keep only few cattle with a high proportion of work bulls (Hassan, 2000; Hoffmann et al., 2001). The pastoralists stated in the interviews that large herds are split and the cattle grouped by category. Bulls, cows, young stock and calves are then herded separately to better suit their feed requirements and walking ability (Schaefer, 1998).

The condition of cattle differed significantly by season. The cattle observed in the rainy and early dry seasons were in good condition, but in the late dry season, the majority was only in medium condition.

The major ecotypes or breeds in the area are the Bunaji (White Fulani), the Rahaji (Red Bororo), and the Sokoto Gudali. Rahaji is a dual-purpose breed producing milk and beef. Bunaji and Sokoto Gudali supply additional draught power, Bunaji is utilized principally for milk and the well-muscled Sokoto Gudali for meat and labour (FDLPCS, 1992).

Seven hundred thirty-four herds (58 percent) consisted of one breed. If herds are fairly pure, reproductive animals of a certain type are selected. This implies that the livestock owners have a good idea about adaptation of cattle for particular environments and purposes and select their bulls accordingly. Season and region within the reserve significantly influenced the distribution of these breeds.

The Rahaji, better adapted to the harsher arid environment, are more frequently found in the northern parts of the reserve. Rahaji is the most prestigious pastoral breed and best adapted to arid environments. They are sensitive to humidity-related diseases (Blench, 1999).

The Bunaji, which is the most important breed in 42 percent of the herds in the north, is more frequently found in the central and southern parts of the reserve, where it clearly dominates (62 and 90 percent, respectively). Docility of the breeds increases from Rahaji over Bunaji to Sokoto Gudali (FDLPCS, 1992). Therefore, Rahaji are solely kept by pastoralists - regardless of whether they farm or not. Bunaji are kept by pastoralists and agro-pastoralists, and Sokoto Gudali are mainly kept by farmers (Hausa and Fulani).

Replacement of Rahaji and Bunaji by Gudali cattle in the herds of settled Fulani was also observed by Vabi (1993) and Blench (1994). Hence, the distribution of cattle breeds reveals a deliberate choice of particular breeds for particular purposes, both for use and in cognizance of their adaptation to ecological conditions.

In 374 herds (30 percent) a second breed or more were observed: These herds consisted mainly of a combination of other breeds together with Bunaji (168 herds) or Rahaji (92 herds). Rahaji were counted together with Bunaji in 74 herds. A possible reason for finding herds with different breeds might be contract herding, i.e. herding of animals belonging to other families. A family herd is more likely to consist of a single breed because animals are inherited from family members and related to each other. Herds of consigned animals are more likely be composed of different breeds.

FIGURE 3
Cattle breeds on rangeland in the Zamfara Reserve, by season (data: Schaefer)

Model calculation of mobility

Figure 4 shows an estimate of the cumulated number of cattle in the reserve during the year, derived from extrapolation of transect counts to the whole area. From the lowest figure in December it can be assumed that sedentary Fulani and Hausa farmers keep

about 50000 estimated 16250 households in the enclaves and the bordering villages (ARCA, 1995; Hassan, 2000). Sixty-nine percent of these households own about four cattle, resulting in 45000 head of cattle. Assuming an average herd size of pastoral herds of 70 cattle, a maximum of 6 400 migratory pastoral households with 450000 heads of cattle can be found in the reserve at the peak of the rainy season in August. The real figures might be lower because the arid northern part of the reserve was not covered by the transects, and because the extrapolation of line transects to the whole area is necessarily very rough.

FIGURE 4

Cattle breeds on rangeland and cropland in the Zamfara Reserve, by region (data: Schaefer)

However, they clearly show that the predominant mobile grazing management involves mobility of a large number of livestock and humans (Figure 5). The absolute number of humans and animals involved, and their spatial and temporal movements, have social and ecological implications. On the social side, this implies functioning social and information networks, logistics, communication and the solution of conflicts over resources between the different user groups. Mohammed and Bello (1994), when asking the Fulani about their perception of problems related to transhumance, found that the incidence of livestock disease (32 percent), theft and conflicts with settled farmers (12 percent each), family split and social discrimination (two percent each) were mentioned as problems.

FIGURE 5
Total number of cattle, and their in- and out-migration in the Zamfara Reserve, by month

Rangeland biodiversity

Vegetation types and species’ number

Animals of pastoralists and farmers feed on the rangelands, particularly during the rainy season. Therefore, pasture quantity and quality have to be looked at. Ground cover of the herbaceous layer increased from the arid northern (29 percent) to the central and southern parts of the reserve (54 and 49 percent). On the same gradient, ground cover of the woody vegetation increased from about three percent tree and 37 percent shrub cover to about 40 percent tree and 46 percent shrub cover (Hoffmann et al., 1998), (Table 2).

TABLE 2

Species’ number and ground cover on natural rangeland in the Zamfara Reserve (Elsholz, 1996; Kueppers, 1998)

Although the number of species identified is only slightly lower than that found in the 1940s (Keay, 1949), there seems to be a change in the vegetation composition with an increase of unpalatable species. Herders and farmers are conscious observers of their environments. Answers from both interviewed Fulani herders and Hausa farmers, suggest evidence of a change in the plant community as a result of increased grazing pressure in the long run. According to them, vegetation cover was much denser and palatable forage species more frequent in earlier times when Zamfara Reserve was only lightly used for grazing livestock. Perennial grasses like Andropogon gayanus, which were abundant in earlier times according to the Fulani, have more or less disappeared (Eckert and Hoffmann, 1998). Their assessment is supported by Keay (1949). In the 1940s he still found perennial grasses in the herb stratum, but noticed that the woody vegetation was influenced by cattle grazing and fire. Today, rapidly maturing annual plants such as Zornia glochidiata or Dactyloctenium aegypticum are increasing (Kueppers, 1998; Schaefer, 1998). These species have good feed quality but little leaf biomass and are less susceptible to defoliation by grazing. Their ground cover reaches 25 percent. Cassia tora, the most important increaser which locally may reach 100 percent ground cover, is mostly unpalatable for grazing animals (Kueppers, 1998; Kreimer and Steinbach, 1998).

Characteristics and use of woody vegetation

Size and crown diameter tend to increase from north to south, following the rainfall gradient. Trees and shrubs found in the rangeland are lower and have a smaller crown than those found on cropland, although trees on cropland are more intensively used (Figure 6). This might indicate a higher age of trees in cropland.

FIGURE 6
Characteristics of woody species on rangeland and cropland in the Zamfara Reserve (CBH = circumference at breast height; data: Hoffmann and Malami, unpublished)

The tree composition in the fields reflects the former open savanna woodland which covered most of the area before settlement took place (Keay, 1949). Trees found on the fields were in most cases indigenous species left at random while the land was cleared and undesirable vegetation was removed. Other species were explicitly left over and used in a multi-purpose way except Prosopis africana, the characteristic species of the former dry deciduous forests (Aubreville, 1950), which is today mostly encountered as dead stumps since they are hardly removable by farmers. The majority of dead tree stumps is found on cropland.

The most common tree species found on farmland are Adansonia digitata, Anogeissus leiocarpus, Diospyros mespiliformis, Lannea microcarpa, Parkia biglobosa, Piliostigma reticulatum, Sclerocarya birrea, Tamarindus indica and several species of Ficus. They were left on the farmland because of their usefulness, but no effort was made to replace old trees or to increase their number. Only Adansonia digitata, Azadirachta indica and several fruit trees were planted (Moringa oleifera, Mangifera indica).

Traditionally, lopping of trees provides important feed resources during the dry season, especially for pastoralists with limited access to crop residues or stubble. Important browse species include Butyrospermum parkii, Vitex doniana, Ziziphus mauritania, Pterocarpus erinaceus, Khaya senegalensis, Adansonia digitata, Acacia seyal and Celtis integrifolia.

Due to intensive utilization (fodder, medicinal purposes, human nutrition and firewood), nearly all tree species are heavily lopped near the villages (Figure 7). However, at greater distances from the villages the number of lopped species and the degree of utilization were generally found to decrease. RIM (1991) reported that the proportion of trees utilized for browse alone was very small and most trees were lopped for multiple use. In Zamfara, no rejuvenation was found in 13 tree species (including Pterocarpus erinaceus, Khaya senegalensis and Celtis integrifolia). Adult specimens are utilized very intensively so that recruiting for these species is seriously threatened. The intensity of use, estimated by the use score, is closely linked to the usefulness of the trees as animal fodder. Only the unpalatable species were mainly untouched. Species which are valued because of their fruits in other West African countries like Lannea microcarpa, Sclerocarya birrea, Vitex doniana or Butyrospermum parkii were only lopped for browse (Kueppers, 1998). Since the Hausa know about the usefulness of those trees, they preserve the tree while using it to feed their animals to a certain extent (Eckert at al., 1998).

FIGURE 7
Use score of woody vegetation, by palatability and settlement density (Use score low = 1, high = 5, all branches removed from the tree) (data: Hoffmann and Malami, unpublished)

THE FARMING SYSTEM

There is a widespread belief that increasing population pressure will lead to further natural resources degradation (Drechsel et al., 2001). However, the “Boserupian“ development paradigm that population growth induces changes in factor scarcities which in turn alter the relative factor prices, stresses the role of people as producers and innovators. The improvement of yield-enhancing and land-conserving methods serves to increase the productivity of the available resources (Boserup, 1965). Moreover, not only technological but also institutional changes are induced in response to changing resource endowments (Hayami and Ruttan, 1985).

In closely settled zones of northern Nigeria, increased population density has been accompanied by more productive and diversified farming systems (Mortimore et al., 1990; Harris, 1998; Mortimore and Adams, 1998), based on close tree-crop-livestock integration. Densely populated zones in semi-arid northern Nigeria, such as the Kano Close Settled Zone (KCSZ), have supported population densities of more than 300 people per km² for centuries. There, livestock and crop systems become more integrated and cropping intensity as well as labour input increase with population density. Small ruminants increase whereas cattle decline. Livestock feeding becomes more labour-intensive because of the scarcity or complete loss of natural range. Crop residue, cut grass and browse are gathered to feed livestock kept in confinement. Manure production is an important component of livestock production (Harris, 1998).

Although the Zamfara Reserve is a large and remote area where population pressure is not expected to play a big role, this does not hold true due to its particular forest reserve status. Except for the land allocated to the enclave villages at the establishment of the reserve in 1919 and under the Forestry Ordinance of 1957, bush clearing to enlarge the cropping area is not permitted under the Forestry Laws, but has always been practised to a limited extent. As pressure on cropland is high, shifting cultivation and fallow are no longer possible. An interpretation of air photographs and satellite images shows that about 2200 ha around the enclaves and about 8000 ha on the western fringe of the reserve had been illegally cleared (Hof, 2000). About two-thirds of the total encroachment took place between 1990 and 1994 (ARCA, 1995). Farm size estimates range between 1.5 and 1.8 ha for the enclaves Dumburum, Aja and Shamashalle, and 5.4 ha for Tsabre, with an average of 3.2 ha. Farms in the bordering villages, without legal restriction, are larger (4.3 ha) (Hassan, 2000). The average field size within the reserve is 1.1 ha (median 0.7 ha).

Soil diversity management

In the villages in the Zamfara Reserve, intensive application of farmyard manure was found to be the most important source of nutrients, followed by dung voided by livestock directly on the field (Hoffmann et al., 2001). No manuring gradient depending on the proximity to the village could be found, indicating intensive soil fertility management in all fields. Nutrient transfer from the rangeland to the cropland through farmyard manure application or corralled animals is an important strategy for maintaining soil fertility. This is usually based on exchange relations between pastoralists and farmers, where the pastoralists provide herd manure through corralling on fields and farmers remunerate them in grains. This manure contract fulfils ecological functions in terms of cropland soil fertility maintenance, and institutional functions as a response to pressure on natural resources. Livestock was corralled on 49 percent of the fields in the Zamfara Reserve later in the dry season. The majority of fields was corralled by Fulani livestock, that provided 1 306 kg of faeces ha^-1.

Any manuring practice resulted in local concentrations of the manure due to the rotation of manure application between different fields and within a field. Farmyard manure is rotated in a one to two year cycle, but it is widely spread on the field. A more pronounced local manure concentration was observed for the corrals, which covers only about 10 percent of the area of a field. On instruction of the field owner, the night-corral is regularly shifted to achieve an adequate spread of the manure, depending on the livestock species and the nutrient status and former yield of the soil. They result in a two to four year rotation, where one part of the field has been intensively manured during the preceding dry season, and the other parts in previous years (Hoffmann et al., 2001; Figure 8).

FIGURE 8
Model of the rotation of corrals, campsites and the spread of farmyard manure on a field in north-western Nigeria, resulting in spatial heterogeneity of soil fertility (Hoffmann et al., 2001)

Although even spreading of farmyard manure and faeces is the farmers’ goal, the rotation of both types of manure results in micro-variability on the field, and residual effects of nutrients, which have to be taken into account for the calculation of nutrient balances. The variable pattern of manuring and planting might also influence N-fixation of legumes. The farmers explained that manure scarcity is the main reason for the rotation. Since they aim at maintaining and equilibrating the soil fertility of their different fields, soils considered as fertile receive less manure than those considered to be poor. To better deal with the scarce manure, farmers take into account the slope of the field to minimize run-off and the risk of ”crop burning” on recent and intensively manured areas in case of low rainfall in the early growing period. Crop burning has also been mentioned by Sandford (1989). In years of good rainfall, spots with high soil nutrient content will result in high yields, whereas in years of little or poorly distributed rainfall, spots with low nutrient content will still produce minimal yields. According to the farmers, the second year of intensively manured spots results in good crop yields without risk of burning. Hence, such differences in soil fertility on one single field would help to reduce risk of crop losses and equalize yield variations. Brouwer et al. (1993) reported similar use of micro-variability for millet planting as a risk aversion strategy in Niger and argued that farmers aim for reliable but not maximum yields even in years of bad rainfall.

Biodiversity management on cropland: species, varieties and planting patterns

The farmers try to increase species and spatial heterogeneity through their choice of crop species, varieties and planting patterns. Most common features are a multi-species mix, a large number of local varieties for different targets and various crop species combinations and planting patterns. They will be explained in the following paragraphs.

The land in the reserve has been continuously cropped for the last 40 years. Fifty different crop species combinations were found, mostly in the relay intercropping system (Mané-Bielfeldt et al., 1998). In the vast majority of fields, subsistence gramineous staple crops were intercropped with more cash generating and nitrogen-fixing legume crops.
Vegetables, spices and some economic trees were grown to a lesser extent.

The most common combinations were millet (Pennisetum ssp.) and cowpea (Vigna unguiculata) with 35 percent of investigated fields, millet, cowpea and groundnut (Arachis hypogaea) (11 percent) and millet, sorghum (Sorghum bicolor) and cowpea in nine percent of the fields (Table 3). Crop rotations are uncommon. Pure millet cropping was rare (3.4 percent), while various two-crop combinations, millet plus another crop, occurred most frequently (47.5 percent). One-quarter of the fields was planted with three crops. Cotton (Gossypium hirsutum) is sometimes part of the system. In fadama land, low-lying areas that are seasonally flooded or have a high water table and can be cultivated during the dry season as well, rice (Oryza sativa) is grown, sometimes together with maize (Zea mays). However, these fields make up only a small percentage of the arable land in the enclaves.

TABLE 3

Main combinations of species on the crop fields in the three villages Dumburum, Shamashalle and Tsabre in the Zamfara Reserve in 1996 (figures are percentages of all investigated fields)

The farmers were asked about the sources of their crop seeds and the factors influencing their choice of varieties. Results showed that farmers stored millet, sorghum and cowpea seeds from their previous harvest or bought seeds on the local markets. Groundnut and cotton seeds are more commonly bought on the market than those of the other crops. Seven varieties each were identified for millet and cowpea, eight for sorghum and three for groundnut. The varieties of sorghum and cowpea are chosen for their yield of grain as well as for by-product depending on the specific needs of the farmer. Livestock keeping farmers will chose a variety yielding high amounts of by-product, thereby integrating fodder production with grain production. Other criteria used in the choice are availability of seeds or the length of the growing period. Sometimes a mixture of several millet or sorghum varieties was planted in the same field to extend the harvesting period and minimize risk of failure. The species, varieties and mixtures of crops grown on particular soils depended on how farmers rate the crop production potentials of the soils (Mané-Bielfeldt and Schaefer, 1996; Kyiogwom et al., 1998).

The main crop combinations with millet and cowpea were found to be planted in a great variety of patterns, with a wide range of different row and plant distances (Tables 4 and 5). Sorghum rows were more widely spaced than millet (2.2-2.3 m vs. 1.0-1.2 m). In both combinations with groundnut, the average row distance for millet was wider.

TABLE 4

Most frequent planting patterns of the main crop combinations in the Zamfara Reserve enclave villages

TABLE 5

Average inter- and intra-row distances for the main crop species in the Zamfara Reserve enclave villages

The high spatial heterogeneity of planting patterns impedes the calculation of yield averages per single crop. The following model calculations of grain and crop residue yields are derived from the interviews and relate to a multitude of crop mixtures (Figure 9).

FIGURE 9
Grain and crop residue yield estimates in villages in the Zamfara Reserve and bordering villages

The estimated grain harvest of 1 856 kg ha^-1 is sufficient to support of about two persons per ha on a cereal-based diet throughout the year (Oltersdorf and Weingärtner, 1996) and even leaves a surplus for sale. This finding is in agreement with the interview results that the people never experienced hunger, and that about 10 percent of the farmers indicated cereal sales (Mane-Bielfeldt et al., 1998).

CONCLUSIONS AND RECOMMENDATIONS

The case study of the Zamfara Reserve shows that farmers and pastoralists have developed similar strategies to use and manage biodiversity. In general, they are based on local knowledge of soils, wild and domestic plants, and livestock, on the temporal and spatial variations of access to natural resources, including mobility and flexible property rights, on the exchange of goods and services within and between systems, and a mix of income generating activities. This is typical for West African drylands, where both livestock and farming production systems have been maintained facing variable rainfall, demographic expansion and changing market conditions. All the strategies are based on high diversity, flexibility and adaptability in order to better deal with incommensurables, as was defined by Mortimore and Adams (1999):

• flexibility – the day-to-day decisions allocating resources of land, labour, capital, knowledge or other scarce factors in response to opportunities and constraints;
• adaptability – the year-by-year modifications or transformation of their production systems in response to economic or environmental realities; and
• diversity – of available choices, for example, of animals, farm technologies, fodder sources, farm crops, markets, or off-farm incomes.

The biodiversity management of the livestock system can be summarized as Use of rangeland biodiversity through the management of livestock mobility and deliberate choice of livestock breeds. The experiences of the rangeland-based livestock system in the Zamfara Reserve show that the predominant mobile grazing management involves mobility of a large number of livestock and humans. This has social implications: The mobile system implies functioning social and information networks, logistics, communication and the solution of conflicts over resources between the different user groups. It also has ecological implications: In the present set-up, rangeland is an open access resource. Stocking density in the Zamfara Reserve exceeds the recommended rate. The related high grazing pressure seems to have ecological implications, affecting the herbaceous and woody layers of the vegetation. As a result, rangeland biodiversity is being reduced. Long-term ecological studies are needed to show whether this trend is reversible. Most long-term studies done so far in West Africa indicate that a certain vegetation resting or patchiness of grazing favour vegetation recovery and species’ richness (Miehe, 1990; Hiernaux, 1998). The needed management practices, such as rotational grazing, however, cannot be achieved in an open-access situation. Ongoing studies investigate whether the rangeland in the reserve is common property resource with somehow regulated access or an open-access resource without any binding rules.

In order to generalize the experience of the cropping system of the Zamfara Reserve, it was found that soil fertility maintenance is a key issue for the farmers. Farming is based on relatively secure property rights to land. The farmers’ strategy is the active Management of soil and crop biodiversity, through maintaining diversity on various spatial and temporal scales, within and between species of crops and animals, and they do so to reduce risk. Farmers rely not only on close crop-livestock integration on the farms (crop-residue feeding, nutrient cycling), but also on exchange and interaction with the mobile livestock system. Animal genetic resources are exchanged between the two systems.

Policy recommendations include:

• Support farmers and pastoralists to maintain and enhance their diversity management.
• Support livestock mobility through flexible property rights and secure access of mobile pastoralists to pasture and water.
• Enable choice of breeds and varieties for various purposes and agro-ecological conditions.
• Support adaptation to future climatic and economic challenges.

REFERENCES

Amanor, K.S. 1995. Dynamics of herd structures and herding strategies in West Africa: A study of market integration and ecological adaptation, Africa 65(3): 351-393.

ARCA, 1995. Forest Reserve study: Ruma Kukar Janjarai and Zamfara Forests. Final report Vol. 1. The environment: present conditions and degradation factors. Katsina Arid Zone Programme, Federal Republic of Nigeria, National Planning Commission.

Aubreville. A., 1950. La flore forestière soudano-guinéene. Paris.

Blench, R. 1994. The expansion and adaptation of Fulbe pastoralism to subhumid and humid conditions in Nigeria, Cahiers d'études Africaines 133-135: 197-212.

Blench, R. 1999. Traditional livestock breeds: Geographical distribution and dynamics in relation to the ecology of West Africa, ODI working paper 122, 67 pp.

Boserup, E. 1965. The conditions of agricultural growth. Allen and Unwin, London.

Boudet, G. 1991. Manuel sur les pàturages tropicaux et les cultures fourragères. Manuels et précis d'élevage No.4, Ministère de la Coopération, IEMVT.

Brouwer, J. Fussell, L.K., Herrmann, L., 1993. Soil and crop growth micro-variability in the West African semi-arid tropics: a possible risk-reducing factor for subsistence farmers. Agric. Ecosyst. Environ., 45, 229-238.

Drechsel, P. Gyiele, L., Kunze, D., Cofie, O., 2001. Population density, soil nutrient depletion, and economic growth in sub-Saharan Africa. Ecological Economics 38: 251-258.

Eckert, B. Hoffmann, I. 1998. Statutory versus customary land allocation: Implications on land tenure and resource management in the Zamfara Reserve. In: Hoffmann. I. (ed.). Prospects of pastoralism in West Africa. Giessener Beiträge zur Entwicklungsforschung. Reihe I. Vol. 25. 169-178.

Eckert, B., Kueppers, K., Hoffmann I., 1998. Indigenous trees as an important resource in rural life - the case of Zamfara Reserve. In: Hoffmann. I. (ed.). Prospects of pastoralism in West Africa. Giessener Beiträge zur Entwicklungsforschung. Reihe I. Vol. 25. 228-237.

Elsholz. C. 1996. Studies on structure and botanical composition of the vegetation in Zamfara Reserve. Sokoto State. Nigeria. UDU/JLU EEC Linkage Programme. unpublished research report. Gießen.

FDLPCS (Federal Department of Livestock and Pest Control Services) 1992a. Livestock in Sokoto Sate, in Nigerian livestock resources. Vol II: National synthesis. Vol III: State reports. RIM, UK, 26pp.

FDLPCS (Federal Department of Livestock and Pest Control Services) 1992b. Nigerian livestock resources. Vol I: Executive summary and atlas. RIM, UK, 53pp.

Gefu, J.O. 1992. Pastoralist perspectives in Nigeria. The Fulbe of Udubo Grazing Reserve. Research Report No. 89. The Scandinavian Institute of African Studies, 106 pp.

Harris, F.M.A. 1998. Farm-level assessment of the nutrient balance in northern Nigeria. Agric. Ecosyst. Environ. 71, 201-214.

Hassan, W.A. 2000. Biological productivity of sheep and goats under agro-sylvo-pastoral systems in the Zamfara Reserve in north-western Nigeria. Cuvillier, Goettingen, 260pp.

Hayami, Y., Ruttan, V.W. 1985. Agricultural Development: An International Perspective, Baltimore. Johns Hopkins University Press.

Hof, A. 2000. Developing a GIS methodology for the Analysis of agro-pastoral livelihood and ecosystems in the Zamfara Forest reserve, NW Nigeria. MSc dissertation, Dept. of Geography, University of Durham, September 2000.

Hoffmann, I. 1998. Zamfara Reserve - Past and present. In: Hoffmann. I. (ed.). Prospects of pastoralism in West Africa. Giessener Beiträge zur Entwicklungsforschung. Reihe I. Vol. 25, 1-9.

Miehe, S. (1990). Inventur und Überwachung der Vegetation in Weideparzellen. Ergebnisse der Untersuchungen 1988-1990 und vorläufige Gesamtauswertung des Standweideversuchs nach 10jähriger Laufzeit. GTZ Projekt Desertifikationsbekämpfung durch agro-pastorale Landnutzungsmodelle. Eschborn (unpublished).

Hiernaux, P. 1998. Effects of grazing on plant species composition and spatial distribution in rangelands of the Sahel, Plant Ecology 138, 191-202.

Hoffmann, I., Gerling, D., Kyiogwom U.B., Mané-Bielfeldt, A. 2001. Farmers' management strategies to maintain soil fertility in a remote area in Northwest Nigeria. Agric. Ecosyst. Environ. 86 (3): 263-275.

Hoffmann. I., Willeke-Wetstein. C., Schäfer. C. 1998. Desription of a grazing ecosystem in Northwestern Nigeria. using environmental indicators. Animal Research and Development 48: 69-83.

Keay, R.W.J. 1949. An example of Sudan Zone Vegetation in Nigeria. J. Ecol. 37. 335-364.

Kreimer. D., Steinbach. J. 1998. Control of Cassia tora Linn. in three pasture ecotypes in the Sudan Savanna in north-western Nigeria. In: Hoffmann. I. (ed.) Prospects of pastoralism in West Africa. Giessener Beiträge zur Entwicklungsforschung. Reihe I. Bd. 25. 87-95.

Krieger, K. 1954. Die Zamfarawa - Ein Stamm der Hausa in Nord-Nigeria. Geogr. Rundschau 6, 387-393.

Kueppers, K. 1998. Evaluation of the ligneous strata of the vegetation of the Zamfara Reserve. In Hoffmann. I. (ed.) Prospects of pastoralism in West Africa. Giessener Beiträge zur Entwicklungsforschung. Reihe I. Vol. 25, 41-47.

Kyiogwom U.B., Umaru, B.F., Bello, H.M. 1998. The use of indigenous knowledge in land classification and management among farmers in the Zamfara Reserve. In: Hoffmann, I. (Ed.). Prospects of pastoralism in West Africa. Giessener Beiträge zur Entwicklungsforschung, Reihe I, Vol. 25, pp. 220-227.

Kyiogwom, U.B., Mohammed, I., Bello, H.M., Maigandi, S.A., Schaefer, C. 1994. The economic situation of the livestock farmer in Zamfara, in: Range development in the endangered Sudan savanna in Sokoto State. Unpublished report, Giessen, 63-70.

Malami. B.S.. Hassan. W.A.. Tukur. H.M.. Maigandi. S.A. 1998. Communal grazing of sheep and goats in the Zamfara Reserve. In: Hoffmann. I. (ed.). Prospects of pastoralism in West Africa. Giessener Beiträge zur Entwicklungsforschung. Reihe I. Vol. 25. 191-195.

Mané-Bielfeldt, A., Schaefer, C. 1996. Evaluation of crop production in the enclaves of Zamfara reserve, NW Nigeria, Part 2. Unpublished research report, Giessen.

Mané-Bielfeldt, A., Schaefer, C., Gefu, J., Hoffmann, I., Mohammed, I., Steinbach, J. 1998. Crop production and utilisation in the Zamfara Reserve. In: Hoffmann, I. (Ed.). Prospects of pastoralism in West Africa. Giessener Beiträge zur Entwicklungsforschung, Reihe I, Vol. 25, pp. 238-246.

McIntire, J., Bourzat, D., Pingali, P. 1992. Crop-Livestock Interaction in Sub-Saharan Africa. Washington, D.C. The World Bank.

Mohammed, I., Bello, H.M. 1994. Transhumance: a rangeland resource management strategy in an endangered ecosystem, in Range development in the endangered Sudan savanna in Sokoto State. Unpublished report, Giessen, 73-80.

Mortimore. M., Adams, W.M. 1998. Farming intensification and its implications for pastoralism in northern Nigeria. In: Hoffmann. I. (ed.). Prospects of pastoralism in West Africa. Giessener Beiträge zur Entwicklungsforschung. Reihe I. Vol. 25. 262-273.

Mortimore. M., Adams, W.M. 1999. Working the Sahel: Environment and society in Nigeria. Routledge, London, UK.

Mortimore. M., Essiet. E.U., Patrick. S. 1990. The nature. rate and effective limits of intensification in the small holder farming system of the Kano Close-Settled Zone. Federal Agricultural Coordinating Unit. Ibadan.

Oltersdorf, U., Weingaertner, L. 1996. Handbuch der Welternaehrung - Die zwei Gesichter der globalen Nahrungssituation Verlag J.H.W. Dietz Nachfolger GmbH, Bonn - Hrsg. Deutsche Welthungerhilfe.

Powell, J.M. and E. Taylor-Powell 1984. Cropping by Fulani agropastoralists in central Nigeria, ILCA Bull. 19: 21-27.

Powell, J.M. 1986. Manure for cropping: A case study from central Nigeria. Expl. Agric. 22, 15-24.

RIM (1991). Woody vegetation cover and wood volume assessment in northern Nigeria. Report for the Federal Government of Nigeria. St. Helier. UK.

Sandford, S.G. 1989. Crop Residue/Livestock Relationships. In: Soil, Crop, and Water Management in the Sudano-Sahelian Zone. ICRISAT Proceedings of International Workshop, 11-16 Jan. 1987 in Niamey/Niger, pp. 169-182.

Schaefer. C. (1998). Pastorale Wiederkäuerhaltung in der Sudansavanne: Eine Untersuchung im Zamfara Forstschutzgebiet im Nordwesten Nigerias. Cuveillir. Göttingen. 217pp.

Scoones, I. (Ed.), Living with uncertainty. New directions in pastoral development in Africa. London, pp. 1-36.

Sutter, J.W. 1987. Cattle and inequality: herd size differences and pastoral production among the Fulani of northeastern Senegal, Africa 57 (2): 196-217.

The Forestry Ordinance (Chapter 75) Sokoto Native Authority Forest reserve No. 9 (Zamfara Forests) (Amendment) Order 1957, in Northern Regional Public Note 38 of 1957.

Vabi, M.B. 1993. Fulani settlement and modes of adjustment in the Northwest Province of Cameroon, ODI Pastoral Development Network Paper 35d. 10 pp.

Van Raay, H.G.T. 1975. Rural planning in a savanna region. University Press, Rotterdam.